Proceedings of SerbiaTrib '13

SerbianTribologySocietyFacultyofEngineeringUniversityofKragujevacSERBIATRIB'1313 th InternationalConferenceonTribology15–17May2013,Kragujevac,SerbiaPROCEEDINGSSERBIATRIB'13

tribology in industryISSN 0354-8996VOLUME 332011.3

Serbian Tribology SocietyUniversity of Kragujevac,Faculty of EngineeringSERBIATRIB ‘1313 th International Conference on Tribology15 – 17 May 2013, Kragujevac, SerbiaPROCEEDINGSEDITORS: Miroslav Babić, Slobodan Mitrović

13 th International Conference on Tribology – SERBIATRIB ‘13PROCEEDINGSISBN: 978-86-86663-98-6Editors:Publisher:For the Publisher:Technical editor:Printed by:Circulation:Miroslav Babić, Ph.D.University of Kragujevac, Faculty of EngineeringSlobodan Mitrović, Ph.D.University of Kragujevac, Faculty of EngineeringSerbian Tribology SocietySestre Janjić 6, 34000 Kragujevac, SerbiaUniversity of Kragujevac, Faculty of EngineeringSestre Janjić 6, 34000 Kragujevac, SerbiaMiroslav Babić, Ph.D., Faculty of EngineeringBranko Ivković, Ph.D., Serbian Tribology SocietySlobodan Mitrović, Ph.D.University of Kragujevac, Faculty of EngineeringDragan Džunić, research assistentUniversity of Kragujevac, Faculty of EngineeringMarko Pantić, research assistentUniversity of Kragujevac, Faculty of EngineeringKOPIRNICA MAŠINAC,34000 Kragujevac, Serbia100 copiesCopyright © 2013 by Serbian Tribology Society and Faculty of EngineeringThe publication of this Proceedings was financially supported by the Ministry of Education,Science and Technological Development of the Republic of Serbia.

Supported byMinistry of Education, Science and Technological Development of the Republic of Serbia

International Scientific CommitteeAleksandar MarinkovićAleksandar RacAleksandar VenclAndrei TudorBharat BhushanBogdan SoviljBranko IvkovićDamir KakašDušan StamenkovićEmilia AssenovaFriedrich FranekGeetha ManivasagamHakan KaleliKonstantinos-Dionysios BouzakisLorena DeleanuMara KandevaMarion MerkleinMehmet Baki KaramisMiroslav BabićNiculae Napoleon AntonescuNikolai MyshkinNyagol ManolovSlobodan MitrovićSlobodan TanasijevićUniversity of Belgrade (Serbia)Serbian Tribology Society (Serbia)University of Belgrade (Serbia)University Politehnica of Bucharest (Romania)The Ohio State University (USA)University of Novi Sad (Serbia)Serbian Tribology Society (Serbia)University of Novi Sad (Serbia)University of Niš (Serbia)Technical University of Sofia (Bulgaria)Vienna University of Technology (Austria)VIT University (India)Yildiz Technical University (Turkey)Aristotle University of Thessaloniki (Greece)University Dunarea de Jos (Romania)Technical University of Sofia (Bulgaria)Friedrich Alexander Universität Erlangen Nürnberg(Germany)Erciyes University (Turkey)University of Kragujevac (Serbia)Petroleum-Gas University of Ploiesti (Romania)V.A. Belyi Metal-Polymer Research Institute ofNational Academy of Sciences of Belarus (Belarus)Technical University of Sofia (Bulgaria)University of Kragujevac (Serbia)University of Kragujevac (Serbia)Organising CommitteeHonorary President:Branko IvkovićPresidents:Miroslav BabićSlobodan MitrovićMembers:Blaža StojanovićFatima ZivićMarko PantićDragan DzunićSerbian Tribology Society (Serbia)University of Kragujevac (Serbia)University of Kragujevac (Serbia)University of Kragujevac (Serbia)University of Kragujevac (Serbia)University of Kragujevac (Serbia)University of Kragujevac (Serbia)

PrefaceThe International Conference on Tribology – SERBIATRIB, is traditionally organized by theSerbian Tribology Society every two years, since 1989. The previous conferences were held inKragujevac (1989, 1991, 1993, 1999, 2005, 2007 and 2011), Herceg Novi (1995), Kopaonik(1997), Belgrade (2001, 2003 and 2009). This year the 13 th International Conference onTribology – SERBIATRIB '13 also takes place on May 15-17, 2013 in Kragujevac.This Conference is organized by the Serbian Tribology Society (STS) and the University ofKragujevac, Faculty of Engineering. Organizing Scientific Conferences, STS plays a significantrole in helping engineers and researchers to introduce in the fundamentals of tribology and topresent their experience, solutions and research results.The scope of the 13 th International Conference on Tribology – SERBIATRIB '13 embraces thestate of art and future trends in tribology research and application. The following two aspects oftribology practice require special attention. Firstly, the requirement for higher productivity ofmachinery means that machines must operate under higher loads and at higher speeds andtemperatures, and that is why finding the right solutions for tribological processes is extremelyimportant. Secondly, the good tribology knowledge can greatly contribute to the saving ofmaterial and energy.The Conference program generally includes the following topics: fundamentals of friction andwear; tribological properties of solid materials; surface engineering and coating tribology;lubricants and lubrication; tribotesting and tribosystem monitoring; tribology in machineelements; tribology in manufacturing processes; tribology in transportation engineering; designand calculation of tribocontacts; sealing tribology; biotribology; nano and microtribology andother topics related to tribology.All together 76 papers of authors from 18 countries (Taiwan, Russia, Belarus, Ukraine,Germany, Poland, India, Pakistan, Nigeria, Slovenia, Croatia, Bosnia and Herzegovina, Italy,Romania, Bulgaria, Greece, Turkey and Serbia) are published in the Proceedings.Approximately 37 papers were submitted by the foreign authors and app. 39 papers by theSerbian authors. All papers are classified into five chapters: Plenary lectures (4) Tribological properties of materials and coatings (29) Tribology in machine elements (23) Tribometry (13) Trenje, habanje i podmazivanje (7) – papers written in Serbian languageIt was a great pleasure for us to organize this Conference and we hope that the Conference,bringing together specialists, research scientists and industrial technologists, and Proceedingswill stimulate new ideas and concepts, promoting further advances in the field of tribology. The

Editors would like to thank the Scientific and the Organizing Committee and all those who havehelped in making the Conference better. We would like to thank especially prof. Miroslav Babićand prof. Branko Ivković for the helpful suggestions and support.The Conference is financially supported by the Ministry of Education, Science andTechnological Development, Republic of Serbia.We wish to all participants a pleasant stay in Kragujevac and we are looking forward to seeingyou all together at the 14 th International Conference on Tribology – SERBIATRIB '15.Kragujevac, May 2013Editors

ContentsPlenary Lectures1. THE GREEN AUTOMOBILE – DEFINITION AND REALIZATIONWilfried J. Bartz ........................................................................................................................................................... 32. THE ECO-LABEL AND THE CONFLICT BETWEEN BIODEGRADABILITY AND ENVIRONMENTALLYACCEPTABILITY OF LUBRICANTSWilfried J. Bartz ........................................................................................................................................................... 33. ROUGHNESS AND TEXTURE CONCEPTS IN TRIBOLOGYN.K. Myhkin, A.Ya. Grigoriev ....................................................................................................................................... 44. RECENT DEVELOPMENTS IN COATINGS’ CHARACTERIZATRION FOR FACILITATING THE COATED TOOLLIFE PREDICTIONK.-D. Bouzakis, G. Skordaris, E. Bouzakis, N. Michailidis ......................................................................................... 10Tribological Properties of Materials and Coating5. PREDICTION OF COATED TOOLS PERFORMANCE IN MILLING BASED ON THE FILM FATIGUE ATDIFFERENT STRAIN RATESK.D. Bouzakis, R. Paraskevopoulou, G. Katirtzoglou, S. Makrimallakis, E. Bouzakis, P. Charalampous .................. 136. SELECTIVE TRANSFER OF MATERIALS IN THE ASPECT OF GREEN TRIBOLOGYEmilia Assenova, Gottlieb Polzer, Dr. Tsermaa, Mara Kandeva ................................................................................ 217. ABRASIVE WEAR AND WEAR-RESISTANCE OF HIGH STRENGTH CAST IRONCONTAINING Sn MICROALLOYMara Kandeva, Boryana Ivanova ............................................................................................................................... 268. INFLUENCE OF NANO-DIAMOND PARTICLES ON THE TRIBOLOGICAL CHARACTERISTICS OFNICKEL CHEMICAL COATINGSMara Kandeva, Dimitar Karastoianov, Boryana Ivanova, Viara Pojidaeva ................................................................ 319. WEAR BEHAVIOR OF AUSTEMPERED DUCTILE IRONWITH NANOSIZED ADDITIVESJ. Kaleicheva ............................................................................................................................................................. 3710. NICKEL COMPOSITE COATINGS MODIFIED BY DIAMOND NANOPARTICLESM. Kandeva, N. Gidikova, R. Valov, V. Petkov .......................................................................................................... 4211. TRIBOLOGICAL BEHAVIOR OF THERMAL SPRAY COATINGS, DEPOSITED BY HVOF AND APSTECHNIQUES, AND COMPOSITE ELECTRODEPOSITS Ni/SiCAT BOTH ROOM TEMPERATURE AND 300°C.A.Lanzutti, M. Lekka, E. Marin, L.Fedrizzi ................................................................................................................. 4612. MECHANOCHEMICAL SYNTHESIS OF NANOSIZED MIXED OXIDESN.G. Kostova, M. Kandeva, M. Fabian, A. Eliyas, P. Balaz ....................................................................................... 55

13. WEAR OF POLISHED STEEL SURFACES IN DRY FRICTION LINEAR CONTACT ON POLIMER COMPOSITESWITH GLASS FIBRESDorin Rus, Lucian Capitanu ....................................................................................................................................... 5814. EXPERIMENTAL INVESTIGATION OF FRICTION COEFFICIENT AND WEAR RATE OF COMPOSITEMATERIALS SLIDING AGAINST SMOOTH AND ROUGH MILD STEEL COUNTERFACESMohammad Asaduzzaman Chowdhury, Dewan Muhammad Nuruzzaman,Biplov Kumar Roy, Sohel Samad, Rayhan Sarker, Abul Hasnat Mohammad Rezwan ............................................. 6515. ABRASIVE WEAR RESISTANCE OF THE IRON- AND WC-BASED HARDFACED COATINGS EVALUATEDWITH SCRATCH TEST METHODAleksandar Vencl, Bojan Gligorijević, Boris Katavić, Bogdan Nedić, Dragan Džunić ................................................. 7516. TOPOGRAPHIC AND ELECTROCHEMICAL TI6AL4V ALLOY SURFACE CHARACTERIZATION IN DRY ANDWET RECIPROCATING SLIDINGZinaida Doni, Mihaela Buciumeanu, Liviu Palaghian ................................................................................................ 8017. FRICTION COEFFICIENT OF UHMWPE DURING DRY RECIPROCATING SLIDINGFatima Živić, Miroslav Babić, Slobodan Mitrović, Dragan Adamović, Svetlana Pelemis ........................................... 8718. THE POTENTIAL OF MAGNESIUM ALLOYS AS BIOABSORBABLE / BIODEGRADABLE IMPLANTS FORBIOMEDICAL APPLICATIONSFatima Živić, Nenad Grujović, Geetha Manivasagam, Caroline Richard, Jessem Landoulsi, Vojislav Petrović ........ 9219. ANALYSIS OF THE SURFACE LAYER FORMATION OF SINGLE CYLINDER ENGINE COMBUSTIONCHAMBER WITH PHOSPHOROUS-FREE AND CONVENTIONAL ENGINE LUBRICANTSL.Yüksek, H.Kaleli, D.Özkan, H. Hacikadiroğlu ......................................................................................................... 9820. TRIBOLOGICAL STUDY OF BIOCOMPATIBLE HYBRID ORGANIC MOLECULES FILM WITH ANTIBACTERIALEFFECTJ.H. Horng, C.C.Wei, S. Y. Chern, W.H. Kao, K.W. Chern, Y.S. Chen ................................................................... 10221. THE INFLUENCE OF CORROSION ON THE MICROSTRUCTURE OF THERMALLY TREATED ZA27/SIC PCOMPOSITESBiljana Bobić, Aleksandar Vencl, Miroslav Babić, Slobodan Mitrović, Ilija Bobić ...................................................... 10622. TRIBOLOGICAL CHARACTERISATION OF PBT + GLASS BEAD COMPOSITES WITH THE HELP OF BLOCK-ON-RING TESTConstantin Georgescu, Mihai Botan, Lorena Deleanu ............................................................................................. 11323. NORMAL FORCE INFLUENCE ON 3D TEXTURE PARAMETERS CHARACTERIZINGTHE FRICTION COUPLE STEEL – PBT + 10% PTFEConstantin Georgescu, Lorena Deleanu, Catalin Pirvu ............................................................................................ 11924. WEAR BEHAVIOUR OF COMPOSITES BASED ON ZA27 ALLOY REINFORCEDWITH GRAPHITE PARTICLESSlobodan Mitrović, Miroslav Babić, Ilija Bobić, Fatima Zivić, Dragan Dzunić, Marko Pantić .................................... 12425. WEAR PROPERTIES OF A356/10SiC/1Gr HYBRID COMPOSITES IN LUBRICATED SLIDING CONDITIONSBabić Miroslav, Stojanović Blaža, Mitrović Slobodan, Bobić Ilija, Miloradović Nenad, Pantić Marko, Džunić Dragan .... 12926. A REVIEW OF THE TRIBOLOGICAL PROPERTIES OF PTFE COMPOSITES FILLED WITH GLASS, GRAPHITE,CARBON OR BRONZE REINFORCEMENTMiloš Stanković, Aleksandar Vencl, Aleksandar Marinković .................................................................................... 135

27. WEAR CHARACTERISTICS OF HYBRID COMPOSITES BASED ON ZA27 ALLOY REINFORCED WITHSILICON CARBIDE AND GRAPHITE PARTICLESSlobodan Mitrović, Miroslav Babić, Nenad Miloradović, Ilija Bobić, Blaža Stojanović, Dragan Džunić .................... 14128. INFLUENCE OF OXIDATION LAYER GENERATED ON PREHEATED CONTACT PAIRS ON STATICCOEFFICIENT OF FRICTIONMarija Jeremić, Dragan Adamović, Slobodan Mitrović, Bojan Bogdanović, Aleksandar SimićSaša Ranđelović, Petar Todorović .......................................................................................................................... 14729. DYNAMICS OF SAMS IN BOUNDARY LUBRICATIONJelena Manojlović .................................................................................................................................................... 15330. INFLUENCE OF RICE HUSK ASH – SILICON CARBIDE WEIGHT RATIOS ON THEMECHANICAL BEHAVIOUR OF Al-Mg-Si ALLOY MATRIX HYBRID COMPOSITESK. K. Alaneme, T. M. Adewale ................................................................................................................................. 16031. TRIBOLOGICAL PROPERTIES OF NANOMETRIC ATOMIC LAYER DEPOSITIONSAPPLIED ON AISI 420 STAINLESS STEELE. Marin, A.Lanzutti,L.Fedrizzi ................................................................................................................................. 16932. PREPARATION AND CHARACTERIZATION OF QUATERNARY AMMONIUMSURFACTANTS ON MUSCOVITE MICAJelena Manojlović ..................................................................................................................................................... 17733. MO-C MULTILAYERED CVD COATINGSA. Sagalovych, V. Sagalovych ................................................................................................................................. 184Tribology of Machine Elements34. EQUILIBRIUM STATE FORMATION FEATURES OF SURFACE LAYERS OF MACHINE PARTSVyacheslav F. Bezjazychnyj, Alexander N. Sutyagin .............................................................................................. 19535. THE INVESTIGATION OF COATED TOOLS TRIBOLOGICAL CHARACTERISTICS INFLUENCE ON THECUTTING PROCESS AND THE QUALITY PARAMETERS OF THE PARTS SURFACE LAYERFomenko Roman Nikolaevich................................................................................................................................... 19836. MODELING SURFACE ROUGHNESS EFFECTS ON PISTON SKIRT EHL IN INITIAL ENGINE START UP USINGHIGH AND LOW VISCOSITY GRADE OILSMubashir Gulzar, S. Adnan Qasim, Riaz A Mufti ..................................................................................................... 20437. STRESSES AND DEFORMATIONS ANALYSIS OFA DRY FRICTION CLUTCH SYSTEMOday I. Abdullah, Josef Schlattmann, Abdullah M. Al-Shabibi ................................................................................. 21038. THE WAVINESS OF AN ABRASIVE WATER JET GENERATED SURFACEJ.Baralić, P.Janković, B.Nedić .................................................................................................................................. 21739. EFFECT OF REFRACTORY ELEMENTS ON WEAR INTENSITY OF THE SURFACE LAYERSIN THE ABRASIVE SOIL MASSJ. Napiórkowski, P. Drożyner, P. SzczyglakP P ............................................................................................................ 22240. EXPERIMENTAL ANALYSIS OF TOOTH HEIGHT CHANGING AT TIMING BELTSBlaža Stojanović, Lozica Ivanović, Andreja Ilić, Ivan Miletić .................................................................................... 226

41. CYCLO DRIVE EFFICIENCYTihomir Mačkić, Živko Babić, Nenad Kostić, Mirko Blagojević ................................................................................. 23042. TRIBOLOGICAL ASPECTS OF THE PROCESS OF WINDING THE STEEL ROPEAROUND THE WINCH DRUMMiloš Matejić, Mirko Blagojević, Vesna Marjanović, Rodoljub Vujanac, Boban Simić ............................................. 23443. APPLICATIVE MONITORING OF VEHICLES ENGINE OILPerić Sreten, Nedić Bogdan, Grkić Aleksandar ....................................................................................................... 24044. ADVANTAGES AND APPLICATIONS OF SELF-LUBRICATING PLASTIC BEARINGSAleksandar Marinković, Miloš Stanković .................................................................................................................. 24745. EFFECT OF VISCOSITY ON ELASTOHYDRODYNAMIC LUBRICATION BETWEEN PARALLEL SURFACESSUBJECTED TO HIGH ACCELERATIONUsman Ali Zia, Aamer A. Baqai, Waseem Akram .................................................................................................... 25146. INCREASING OF TOOL LIFE FOR HOT FORGING USING SURFACE MODIFICATIONMilentije Stefanović, Dragan Džunić, Vesna Mandić, Srbislav Aleksandrović, Dragan Adamović,Slobodan Mitrović .................................................................................................................................................... 26147. ANALYSIS OF TRIBOLOGICAL PROCESS DURING IRONING OF SHEET METAL MADE OF AlMg3Dragan Adamović, Milentije Stefanović, Srbislav Aleksandrović,Miroslav Živković, Fatima Živić, Marko Topalović .................................................................................................... 26548. OPTIMAL DESIGN OF A CAMMECHANISM WITH TRANSLATING FLAT-FACE FOLLOWER USING GENETICALGORITHMI. Tsiafis, S. Mitsi, K.D. Bouzakis, A. Papadimitriou ................................................................................................. 27049. INFLUENCE OF VARIOUS TYPES OF ROCK AGGREGATES ON SELECTION OF THE WORKING PARTSMATERIAL IN CIVIL ENGINEERINGV. Lazić, M. Mutavdžić, R. Nikolić, S. Aleksandrović, D. Milosavljević, B. Krstić, R. Čukić ..................................... 27550. TECHNO-ECONOMIC JUSTIFICATION FOR REPARATORY HARD-FACING OF MACHINE SYSTEMS'WORKING PARTSV. Lazić, R. Čukić, S. Aleksandrović, D. Milosavljević, R. Nikolić, B. Krstić, B. Nedeljković .................................... 28151. TRIBOLOGY ASPECT OF RUBBER SHOCK ABSORBERS DEVELOPMENTMilan Banić, Dušan Stamenković, Miloš Milošević, Aleksandar Miltenović ............................................................. 28652. EFFECTS OF USING OF MQL TECHNIQUE IN METAL CUTTINGGordana Globočki Lakić, Branislav Sredanović, Davorin Kramar, Bogdan Nedić, Janez Kopač ............................. 29253. TRIBOLOGICAL ASPECT OF RUBBER BASED PARTS USED IN ENGINEERINGDušan Stamenković, Milan Nikolić, Miloš Milošević, Milan Banić,Aleksandar Miltenović, Miroslav Mijajlović ............................................................................................................... 30254. POSSIBILITY OF REPLACING THE CHORINATED PARAFFINS IN METALWORKING FLUIDSMarica Dugić, Branka Kojić, Pero Dugić, Goran Dugić ............................................................................................ 30855. QUALITY OF PLASMA CUTTINGBogdan Nedić, Marko Janković, Miroslav Radovanović, Gordana Globočki Lakić .................................................. 31456. TRIBOLOGICAL ASPECTS OF SINTERED STEEL GEAR IN APPLICATION WORM-AND-GEAR SETAleksandar Miltenović, Milan Banić, Miroslav Mijajlović, Đorđe Miltenović .............................................................. 320

Tribometry57. PRELIMINARY STUDY ON THE SEIZURE TREND OF A MOM-THP WITH SELF-DIRECTED BALLSLucian Capitanu, Liliana – Laura Badita, Virgil Florescu, Dumitru Catalin Bursuc .................................................. 33158. ANALYZING THE INFLUENCE OF THE CONSTRUCTION ELEMENT POSITION ON TORQUE TRANSMISSIONBY FRICTIONMarija Jeremić, Bojan Bogdanović, Aleksandar Simić, Dragomir Miljanić, Petar Todorović,Sasa Randjelovic, Branko Tadić .............................................................................................................................. 34159. USE ALGORITHM FOR CONSTRUCTION 3D VISIBILITY GRAPHS TO DESCRIBE PLASTIC AND ELASTICDEFORMATION OF ROBOT LASER HARDENED SPECIMENSM. Babič, P. Kokol, M. Milfelner, P. Panjan, Igor Belič ............................................................................................ 34860. USE FRACTAL GEOMETRY TO DESCRIBE FRICTION OF ROBOT LASER HARDENED SPECIMENSM. Babič, P. Kokol, M. Milfelner, P. Panjan, Igor Belič ............................................................................................ 35161. USE NEW PROCESS IN ROBOT LASER HARDENING TO DECREASE WEAR OF SPECIMENSM. Babič, P. Kokol, M. Milfelner, P. Panjan, Igor Belič ............................................................................................ 35562. DIFFERENT WAYS OF FRICTION COEFFICIENT DETERMINATION IN STRIPE IRONING TESTS Aleksandrovic, M. Stefanovic, V. Lazic, D. Adamovic, M. Djordjevic, D. Arsic ..................................................... 35963. A NANOMECHANICAL APPROACH ON THE MEASUREMENT OF THE ELASTIC PROPERTIES OF EPOXYREINFORCED CARBON NANOTUBE NANOCOMPOSITESG. Mansour, D. Tzetzis, K.D. Bouzakis .................................................................................................................... 36464. SOME TRIBOLOGY STATE TESTS OF “EPDM” RUBBER BASED ON LABORATORY EXPERIMENTATIONSAbhijit Mukhopadhyay .............................................................................................................................................. 37365. APPLICATION OF 3D SOFTWARE PACKAGES FOR DESIGNING TRIBOMETER OF MODULAR TYPEIvan Mačužić, Branislav Jeremić, Petar Todorović, Marko Đapan, Milan Radenković, Marko Milošević ................. 38066. USING OF KALMAN FILTER AS A PROGNOSTIC TOOL FOR TRIBOLOGY PROCESSESIvan Mačužić, Petar Todorović, Marko Đapan, Milan Radenković, Branislav Jeremić ............................................ 38467. FRICTION COEFFICIENT ESTIMATION DURING FRICTION STIR WELDINGWITH THE SINGLE SHOULDERED WELDING TOOLMiroslav Mijajlović, Dušan Stamenković, Milan Banić, Aleksandar Miletnović, Miloš Milošević ............................... 38868. MEASUREMENT INSTRUMENTATION FOR DETERMINATION OF STATIC COEFFICIENT OF ROLLINGFRICTIONPetar Todorović, Ivan Mačužić, Branislav Jeremić, Marko Đapan, Branko Tadić ................................................... 39669. IMPLEMENATION SQL REPORTING SERVICE IN THE TRIBOLOGYCAL DATA BASESMilan Erić, Marko Djukić ........................................................................................................................................... 401Trenje, habanje i podmazivanje70. VEŠTAČKO STARENJE TIKSOLIVENE ZA27 LEGURE I ČESTIČNIH ZA27/SIC KOMPOZITAI. Bobić, M. Babić, A.Vencl, S. Mitrović, B. Bobić .................................................................................................... 40971. UTICAJ POVRŠINE PODLOGE NA KARAKTERISTIKE PREVLAKA CINKADesimir Jovanović, Bogdan Nedić, Milomir Čupović , Vlatko Matrušić ..................................................................... 414

72. DEFEKTACIJA REDUKTORA BKSH-335 ZA POKRETANJE TRAKASTIH TRANSPORTERABAGERA Sch Rs 630Svetislav Lj. Marković, Ljubica Milović, Bratislav Stojiljković .................................................................................... 42073. ISPITIVANJE MEHANIČKIH I STRUKTURNIH OSOBINA PREVLAKAOTPORNIH NA EROZIJU I VISOKE TEMPERATUREMihailo R. Mrdak ..................................................................................................................................................... 42674. IZBOR MERNE GLAVE DIFERENCIJALNOG PNEUMATSKOG KOMPARATORA ZA KONTROLUUNUTRAŠNJIH MERA MAŠINSKIH DELOVADragiša Skoko, Cvetko Crnojević, Mileta Ristivojević .............................................................................................. 43375. POVEĆANJE POUZDANOSTI PODSISTEMA KOPANJA ROTORNOG BAGERA PODEŠAVANJEMTRIBOLOŠKIH KARAKTERISTIKA REZNIH ELEMENATAVojin Vukotić, Dragan Čabrilo .................................................................................................................................. 44076. PONAŠANJE NEHRĐAJUĆIH ČELIKA U KOMBINIRANIM UVJETIMA TROŠENJAGoran Rozing, Antun Pintarić, Desimir Jovanović, Vlatko Marušić ......................................................................... 446Authors Index ................................................................................................................................................................. 453

Plenary Lectures13 th International Conference on Tribology – SERBIATRIB ’1315 – 17 May 2013, Kragujevac, Serbia

Serbian TribologySocietySERBIATRIB ‘1313 th International Conference onTribologyKragujevac, Serbia, 15 – 17 May 2013Faculty of Engineeringin KragujevacTHE GREEN AUTOMOBILE– DEFINITION AND REALIZATION –Wilfried J. Bartz 11 T+S Tribology and Lubrication Engineering, Technische Akademie Esslingen e. V., Ostfildern, GermanyAbstract: The green automobile has to be green from the cradle to the grave. This means that the life cycle assessmentshould be the required approach which means the environmental impact of raw material generation, of production,during use and of disposal and recycling have to be taken into account.The overall aim is the reduction of energy as well as health and safety of the mankind. Nowadays the resources ofenergy carriers are consumed faster than expected.The impacts mentioned above can be characterized by emissions, primary energy demand, consumption of resourcesand waste generation. Often these environmental impacts are evaluated as CO2-emissions, contribution to thegreenhouse effect and summer smog and as primary energy demand during life cycle of all energy consumers.Mostly these impacts are simplified to the contribution of lubricants by higher efficiency in powertrains, by longerlubricants lifetimes, by energy reductions during the production of automobiles and during their use. The lastmentioned aspects mainly mean the reduced fuel consumption by reducing the friction between all moving parts.Some of the aspects mentioned above are explained and evaluated in this presentation. As an result an overall energyreduction is possible.THE ECO-LABEL AND THE CONFLICT BETWEENBIODEGRADABILITY AND ENVIRONMENTALLYACCEPTABILITY OF LUBRICANTSWilfried J. Bartz 11 T+S Tribology and Lubrication Engineering, Technische Akademie Esslingen e. V., Ostfildern, GermanyAbstract: Often the environmentally acceptability is equated with the fast biodegradability. But for thedegradation process oxygen is necessary. If large amounts of lubricants will be introduced into theenvironment, for instance as an accident, so much oxygen has to be taken from the surrounding, that otherorganisms will suffer.Nevertheless the regulations to define environmentally acceptable lubricants, which are listed in theframework to receive the European Eco-Label do not consider this aspect. The criteria for the Eco-Labelinclude environmentally and human health hazards, aquatic toxicity requirements, biodegradability ,exclusion of specific substances, the imperative use of renewable raw materials and of course the technicalperformance.Details of these criteria will be explained in the presentation.13 th International Conference on Tribology – Serbiatrib’13 3

Serbian TribologySocietySERBIATRIB ‘1313 th International Conference onTribologyKragujevac, Serbia, 15 – 17 May 2013Faculty of Engineeringin KragujevacROUGHNESS AND TEXTURE CONCEPTS IN TRIBOLOGYN.K. Myhkin 1 , A.Ya. Grigoriev 11 Metal-Polymer Research Institute, Belarus, nkmyshkin@mail.ruAbstract: Current knowledge on scale and spatial organization of engineering surfaces is presented. Mainmethods of surface roughness analysis are discussed. A review of theoretical and practical problems intribology involving concept of rough surface is presented. Critical analysis of rough surface description inthe evolution from 2D to 3D set of parameters is carried out. Advantages and disadvantages of traditionaland modern approaches of surface analysis based on concepts of roughness and texture are discussed.Keywords: surface roughness, roughness parameters, texture, wear surface, debris.1. INTRODUCTIONSurface asperities influence practically all theaspects of solids contact. There are a lot of theoreticaland experimental data on mutual dependence ofroughness and such phenomena as adhesion, contactstiffness, abrasion and many others which occurduring friction and wear [1-3]. It is commonlyaccepted that understanding of friction phenomenais directly connected to analysis of surface structureand its transformation due to wear. So, at presentany friction and wear model is involving surfaceroughness parameters.Traditional concept of rough surface basedmainly on profile parameters is not fully satisfiedmodern trends in tribology. Development of 3Dparameters did not change the situation significantly.Only recently, a new view of surface spatial organizationwas introduced. The concept is knownas texture and it reflects the appearance of distinctivesurface pattern. The current paper presents areview of the problems of rough surfaces analysisin their evolution from statistical height and stepparameters of profiles to dimensionless and scaleinvariant representation of surface texture.2. ROUGHNESS ORIGINThe deviation of a real surface from the ideallysmooth are associated with the action of variousfactors, which can be divided onto structural, technological,and operational ones [4-6]. Their featuresdefine scale and textural properties which form surfacegeometric irregularities and as a result differenttypes of roughness.Structural roughness is inherently connectedwith discrete nature of solids. Having small but finitesizes, the atoms and molecules of solids arelocated within a certain distance from each other. Ifassume that the boundary of solids corresponds tosome constant potential of atomic interaction withthe surrounding phase, it is evident that its reliefhas a periodic nature. Surface image with atomicresolution presented on Figure 1a confirms that.Figure 1. а – AFM image of pyrolytic graphite surface(amplitude of surface deviation 0.43 nm); b – deformationof gold surface structure under effect of surfaceforces [7] (TEM); c – surface of spiral dislocation; d –surface of steel fracture.4 13 th International Conference on Tribology – Serbiatrib’13

The relief characteristics at this level are closelyrelated to surface forces and here significant quantum-mechanicaleffects occurs such as thermal oscillationsof atoms with amplitudes up to 10% percentof interatomic distances and spontaneouschanges in their position. Moreover various atomicstacking faults in the surface layer produce compensatingdeformations of the layer material (Figure1 b). However, from a practical point of view,these properties are not significant, at least for thepresent level of tribological problems.Strict periodicity of atomic roughness is characteristiconly for ideal crystals. Real solids are imperfect.The imperfections of atomic structureknown as dislocations form next level of physicalroughness (Figure 1 c).Crystalline structure of solids can form last levelof structural roughness. Unlike to previous types ofirregularities which characterized by sub-nanometerheights, the size of corresponded asperities is comparableto the size of crystallites and can reach upto hundreds of microns. Usually this type of reliefis observed on the fracture surfaces of metals (Figure1 d).Technological roughness is a result of mechanical,thermal or any other types of material processing.These surface deviations consist of periodicaland random components. Formation of periodicalcomponents results from the processes of copyingof tool cutting edges and roughness is affectedby technology (Figure 2). The appearance of therandom components is associated with material destructionduring chip formation and its adhesion tocutting edges (build-up), work hardening and fatiguefailure of surface layers, etc.Figure 2. Machined surfaces (Ra 1.6): a – cylindricalgrinding ; b – turningErrors in mounting of the parts during machining,the presence of elastic deformations and vibrationin the machine-tool system, cutting tool wear,and so on, leads to waviness and deviations of form(hour-glassing, faceting, barrelling, etc.) of the realsurface or the profile from the corresponding parametersspecified on the basis of design. Thesedeviations can be periodic or stochastic.Operation roughness. Main reasons of surfacedegradation of machine parts during operation arewear and corrosion. Nowadays it is generally acceptedthat mechanisms of wear and corrosion correspondedto certain morphological types of formedsurfaces and fracture fragments (wear particles oroxides). It is a basis of modern methods of tribomonitoringand triboanalysis [8,9].The theoretical background of the methods isprovided by phenomenological models of frictioncontact damage. While using these models the actualwear mechanism is established basing on classificationof friction surface morphology (Figure 3).Figure 3. Surface damage at friction: a – abrasion wear;b – plastic deformation; c – ploughing and adhesive fracture;d – fatigue wearThus, the delamination of thin material layersand the formation of exfoliation and spalling regionsare related to fatigue wear at cyclic elasticcontact. It is followed by the separation of materialdebris which points to plastic deformation of thesurface layer at excessive loading and lubricant filmtearing. The appearance linear relief of asperitieswith sharp edges indicates abrasive wear. Defectsshaped as deep tear-outs, delaminated thin filmspoint to adhesive and cohesive interaction of thecontact surfaces. Thus, analysis of wear debris andfriction surfaces allows for evaluation of the operatingconditions of tribosystem, condition of lubricant,and provides the opportunity to predict failureof the tribosystem and take measures to prevent it[10].3. SCALE STRUCTURE OF ROUGH LAYERAs it can be seen the heights of the surface asperitieslie in a wide range. On lower side they arelimited by the dimensions of the atomic and supermolecularformations, on the upper one by maximalheights which are proportional to the length of theexamined profile [11].13 th International Conference on Tribology – Serbiatrib’13 5

Figure 4. Diagram of the height and spacing parametersof surface asperitiesIt is evident that in this case there are no limitationson the existence of asperities in various dimensionalranges (both spacing-wise and heightwise).However, in spite of the fact that there is nouniversally substantiated criterion for distinguishingasperities on the basis of scale, at the presenttime there exists the concept of the surface as anensemble of asperities of four dimensional levels:macrodeviations, waviness, roughness, and subroughness[12].4. SURFACE MEASUREMENTIn studying the topography the need arises forthe solution of three basic problems: description ofthe surface, development of representative surfaceevaluation systems and technical realization of themeasurement processes. In spite of the fact that theseproblems are interdependent, the last problem isof special importance, since our theoretical concepts,and therefore our understanding of how anyparticular phenomena may take place on the surface,are based on the quantitative estimates. Thereforeit is evident that roughness measurements areof primary importance in studying the topography.Nowadays a great number of experimental methodsof surface measurement are used. Stylus methodsremain the most widespread; they yield resultsforming the basis for current standards. Opticalmethods involving electromagnetic radiation suchas light section, shadow projection, interferencetechniques etc. have became widely applied. Atomic-forcemicroscopy has found a wide spread in surfacemetrology now. Figure 5 represents some capabilitiesof different methods of roughness measurementand their vertical and lateral resolution. Ascan be seen there is no method for measuring fullrange of asperities deviations.Figure 5. Resolution of various methods of roughnessmeasurementThe foregoing concepts form the basis of therepresentation of a surface in such disciplines asmechanical engineering, machine design, technology,tribology, heattransfer, and so on. On the wholein this representation the surface is examined as therealization of a random field, the characteristics ofwhich are evaluated on the basis of twodimensionalprofilogram samples [12]. In this casethe system of topography estimates is based onanalysis of the histogram characteristics of the asperitiesin some range of their values.A characteristic feature of this approach is thefact that the mutual influence and interrelationshipof the asperities are generally ignored (except forthe fact that the surfaces may be classified as isotropicor anisotropic), i.e., the spatial organizationof the asperities is not taken into account. We canillustrate the ambiguity arising in the surface representationsin this case. Figure 6 a, b shows photographs(obtained on a scanning electron microscope(SEM)) of surfaces having different spatial structure.Table 1 present the results of a comprehensivestudy of their microgeometry.Figure 6. Two types of surface texturesTable 1. Roughness parameters of surfaces on Figure 6.RoghnessparameterSample on Figure 5abRa 3.20.5 2.30.3Rz 19.34.2 15.92.0Rmax 15.03.7 12.21.4S 56.718.6 43.27.6Sm 165.9 4.4 117.78.76 13 th International Conference on Tribology – Serbiatrib’13

It can be seen that in spite of the significant differencebetween the studied objects practically allthe quantitative estimates coincide in the limits ofthe measurement errors. It is impossible to determinecriteria on the basis of which we can judge thedifference between these surfaces.The principal reasons why it is not possible toevaluate the topographic properties of surfacessolely with the aid of histogram estimates were discussedin [14, 15]. Specifically, it was shown thaton the basis of these characteristics we cannot constructa satisfactory prognostic model, since in thefinal analysis it is valid only in the limits of thosevalues that were used for its construction. The useof this model may lead to unexpected results. Forexample, according to the authors of [15], by superposingthe parameters we can achieve a gooddescription of practically any phenomenon, includingthose not relating to the examined object. Consideringthat the modern instruments yield aboutfifty different characteristics (including 3D) thatmay be subsequently used, the drawbacks of thisapproach become still more evident [6]For more correct representation of the surface itis necessary to have the possibility of characterizingit as an object, having a definite topology. In tribologythis approach is formalized by concept of texture[16, 17].5. CONCEPT OF TEXTURESurface texture is rather difficult to define. Mostauthors agree that this notion reflects the features ofthe surface relief caused by the two-level model ofspatial relations of irregularities heights [18, 19].The mode of these relations at the local level governsthe shape of irregularities and at the globallevel, the position of irregularities relatively to eachother. To some extent, the concept of texture unitesthe ideas of treatment direction and irregularity directionaccording to the USSR Standards GOST2789–73 and 9378–93. The texture is outlined qualitativelyby several adjectives characterizing theshape and mutual position of irregularities, such asstepwise linear, facet random, spherical, sphericalradial, etc.There are numerous approaches to the descriptionof texture; however, all of them are reduced toone of the following: comparative and parametricalapproach, usage of invariant presentations, and parameterizationof visual content.Comparative methods are based on expert visualevaluation of the similarity of the object under investigationand the reference. The features of man’svision allow him to notice and identify minute distinctionsin roughness, texture, color, and shape ofobjects. The simplicity of comparative methods andthe fact that they provide sufficient accuracy formost applications encouraged their widespread use.Thus, for qualitative evaluation of roughness, samplereference surfaces are used according to GOST9378–93 (Figure 2).Comparative methods are very simple and inmost cases a set of references and an optical microscopeare sufficient to realize them. Their disadvantagesare the subjective and qualitative nature ofthe estimates obtained. To overcome them, theopinions and agreement of multiple experts areused [20].Parametric methods use different statistics ofsurface asperity heights and spacing, brightness,and color characteristics of their images.In order to evaluate texture properties, roughnessparameters are most often used, e.g., accordingto GOST 25142–82. Since 2007, the ISO 25178standard has been used to describe 3D surfaceproperties. However, they are mainly similar tostandard profile estimates and hence inherit all theirshortcomings [6, 15].The advantages of the parametric approach arein the simple interpretation of the respective estimates,while their weak descriptive ability can beconsidered a shortcoming. Nevertheless, when agreat number of similar characteristics are used, theapproach is capable to solve the problems with accuracysufficient for most practical applications.The essence of invariant representation is theapplication of normalized description methods, i.e.,the representation of analysis objects in a form independentof their scale and coordinate origin.The simplest procedure of invariant representationis based on the Fourier transform of surfaceheights [21]. With respective normalization, thecoefficients of the amplitude spectrum (Fourier descriptors)do not depend on the scale and positionand can be considered as a complete (single-valuedand reversible) system of features. More complicatedmethods use fractal compression of images andwavelet transforms [22, 23].Features are not defined in the given approach atall. It is believed that all elements of normalizedrepresentations are features. It is of no significancethat they can be numerous and do not have visualinterpretation. It is assumed that they are analyzedand classified by computer methods; therefore, thesize of the feature vector is not important. The approachis unsuitable for research because of the absenceof any visual and geometric interpretation.Parameterization of visual content is based onthe assumption that representative description oftexture is the only possible by means of estimatesreflecting the visual content of the objects underinvestigation.13 th International Conference on Tribology – Serbiatrib’13 7

Realization of the approach is based on the introductionof a structural element, i.e., the minimalvisually perceived region of the object under analysis.It is believed that the features of dislocation ofstructural elements relative to each other governlocal morphological properties at small distances,and global ones at large distances. Differences inthe choice of structural element type and descriptionof their mutual position are responsible for thevariety of the methods for realizing the given approach.One of the methods is based on the use of cooccurrencematrices (COM) [16, 17]. The use ofCOM is motivated by the known assumption thatthe second order probabilities of features derivedfrom the images reflect their visual content [24].In order to describe the texture by the givenmethod, a surface region is chosen as a structuralelement whose position is characterized by the directionof gradient G i and distance P i from the coordinatecenter (Figure 7 a). The mutual position oftwo structural elements is specified by the distancebetween them ρ and the difference invariant descriptionbased on COM (Figure 7 b), the numberof pairs of structural elements is counted with certainρ and g being present at the surface area underanalysis. Both height-coded and half-tone imagescan be used taken by various microscopy methodsarranged as upside lightening.Figure 7. Parameterization of visual contents of texture:a – scheme of determination of structure element; b –matrix of co-occurrence of texture elementsWith respective normalization, COM does not dependon scale and object position in the field of view.As with invariant presentations, each of the COMelements can be considered as a feature. However,since COM elements define the areas of the surface,and then the possibility arises of non-parametric comparisonof objects with visualization of their similarityor dissimilarity [25, 26]. The technique involvesmarking the areas whose COM elements have eitherclose or essentially different values on the images ofthe compared objects. In the former case, this allowsfor visualizing the similarity of the objects, and in thelatter their distinction. Figure 8 shows the results ofthe solutions of this problem.Figure 8. The visual matching of surface texturedifferences: a, b – surface of two types of hard drivemagnetic media; d – difference of the surface a from b(presence of “kidney-like” structures)The advantage of the approach consists in itsgeneral nature, allowing us to unite the features oftexture of objects. This makes analysis of theirmorphology and classification much easier. However,its realization is rather complicated.Prospects in development of texture analysis andapplications look very promising. The analyticaland computational tools in texture analysis are progressingquickly [27]. The progress in technologyresults in a possibility to use a variety of methodsfor making regular textures and micro-textures e.g.by laser [28] or patterning with rigid asperities [29].These technological advances can bring a lot offruitful applications in many areas of tribology.6. CONCLUSIONSSurface 3D organization can be described bydefinition of texture. Experience of image recognitiontheory can provide methods for rough surfacetexture description and visualization of texture similarity/dissimilarity.The description of a surfacetexture by special type of COM is in a good agreementwith the texture distinctions obtained by expertvisual perception. Texture analysis can be efficientlyapplied for solving practical tribologicalproblems in micro/nanoscale.REFERENCES[1] N.K. Myshkin, C.K. Kim, M.I Petrokovets: Introductionto Tribology, Cheong Moon Gak, 1997.[2] Micro/Nano Tribology, Ed. By B. Bhushan, CRCPress, 1998.[3] G. W. Stachowiak: Engineering Tribology. Butterworth-Heinemann,2005.[4] I.V. Dunin-Barkovskii and A.N. Kartashova:Measurement and Analysis of Surface Roughness,Waviness, and Noncircularity [in Russian], Moscow,1978.[5] T.R. Thomas: Rough Surfaces, Imperial CollegePress, London, 1999.[6] D. Whitehouse: Surface and Their Meausrement,Kogan Page Science, 2004.[7] M. Pruton: Surface Physics, Clarendon: Oxford,1985.[8] B.J. Roylance, R. Dwyer-Joyce: Wear debris andassociated wear phenomena – fundamental re-8 13 th International Conference on Tribology – Serbiatrib’13

search and practice, Proc. Inst. Mech. Eng. Part JJEng. Tribology, No. 214, pp. 79-105, 2000.[9] N.K. Myshkin, A.Ya. Grigoriev: Morphology: Texture,Shape and Color of Friction Surfaces andWear Debris, Journal of Friction and Wear, Vol.29, No. 3, pp. 192-199, 2008.[10] D.W. Anderson: Wear particle atlas, ReportNAEC-92-163.[11] R.S. Sayles, T.R. Thomas: Surface Topography asa Non-Stationary Random Process, Nature 271(5644), pp. 431–434, 1978.[12] N.B. Demkin and E.V. Ryzhov: Surface Qualityand Contact of Machine Parts [in Russian], Moscow,1981.[13] A.P. Khusu, Yu.R. Vitenberg, and V.A. Pal'mov:Surface Roughness (Theoretical-Probabilistic Approach)[in Russian], Moscow, 1975.[14] A.Ya. Grigoriev, N.K. Myshkin, O.V. Kholodilov:Surface Microgeometry Analysis Methods, SovietJournal of Friction and Wear, Vol. 10, No. 1, pp.138-155, 1989.[15] M. Zecchino: Characterization surface quality:why average roughness is not enough, AP. Notes ofIveco Instrument, December, pp. 24–30, 2003.[16] A.Ya. Grigoriev, S.A. Chizhik, N.K. Myshkin: Textureclassification of engineering surfaces with nanoscaleroughnes, Int. J. of Machine Tools andManufacture, Vol. 38, No. 5–6, pp. 719–724, 1998.[17] N. K. Myshkin, A. Ya. Grigoriev: Spatial characterizationof engineering surface, in Proc. of 3rd InternationalConference on Surface Engineering,Chengdu, China, pp. 54–61, 2002[18] R.M. Haralick: Statistical and structural approachesto texture, Proceedings IEEE, Vol. 67, No. 5, pp.786–804, 1979.[19] M. Sonka, V. Hlavac, R. Boyle: Image processing,analysis and machine, Boston: PWS publishing,1999.[20] A.Ya. Grigoriev, R. Chang, E.S. Yoon, H. Kong:Classification of Wear Particles by Semantic Features,Journal of Friction and Wear, Vol. 20, No. 2pp. 42-48, 1999.[21] Z. Peng, T.B. Kirk: Two-Dimensional Fast FourierTransform and Power Spectrum for Wear ParticleAnalysis, Tribology Int, Vol. 30, No. 8, pp. 583-590, 1997.[22] T.B. Kirk, G.W. Stachowiak, A.W. Batchelor:Fractal Parameters and Computer Image AnalysisApplied to Wear Particles Isolated by Ferrography,Wear, Vol. 145, pp. 347-365, 1991.[23] S.-H. Lee, H. Zahouani, R. Caterini, T.G. Mathia:Morphological characterization of engineered surfacesby wavelet transform, in Proc. of 7 th Int. Conf.on Metrology and Properties of Eng. Surfaces,Goteborg, pp. 182-190, 1997.[24] H. Tamura, S. Mori, T. Yamawaki: Texture featurescorresponding to visual perception, IEEETrans. SMC-8, Vol. 8, pp. 460–473, 1978.[25] N.K. Myshkin, A.Ya. Grigoriev, S.A. Chizhik,K.Y. Choi, M.I.: Surface Roughness and TextureAnalysis in Microscale, Wear, Vol. 254, pp. 1001-1009, 2003.[26] P. Podsiadlo, G.W. Stachowiak: Development ofadvanced quantitative analysis methods for wearparticle characterization and classification to aidtribological diagnosis, Tribology International,Vol. 38, pp. 887-892, 2005.[27] P.Podsiadlo, G.W.Stachowiak: Directional MultiscaleAnalysis and Optimization for Surface Textures,Tribology Letters, Vol. 49, pp. 179-191, 2013.[28] I. Etsion, State of the Art in Laser Surface Texturing,ASME J. Tribology, Vol. 127, pp. 248-253, 2005.[29] D.T. Nguen et al.: Friction of Rubber with SrfacesPatterned with Rigid Spherical Asperities, TribologyLetters, Vol. 49, pp. 135-144, 2013.13 th International Conference on Tribology – Serbiatrib’13 9

Serbian TribologySocietySERBIATRIB ‘1313 th International Conference onTribologyKragujevac, Serbia, 15 – 17 May 2013Faculty of Engineeringin KragujevacRECENT DEVELOPMENTS IN COATINGS’CHARACTERIZATRION FOR FACILITATING THE COATEDTOOL LIFE PREDICTIONK.-D. Bouzakis 1,2 , G. Skordaris 1,2 , E. Bouzakis 1,2 , N. Michailidis 2,31 Laboratory for Machine Tools and Manufacturing Engineering, Mechanical Engineering Department, AristotelesUniversity of Thessaloniki, Greece2 Fraunhofer Project Center Coatings in Manufacturing, in Centre for Research and Technology Hellas in Thessalonikiand in Fraunhofer Institute for Production Technology in Aachen, Germany3 Physical Metallurgy Laboratory, Mechanical Engineering Department, Aristoteles University of Thessaloniki, GreeceAbstract: Coated tools constitute the majority of the tools applied in material removal processes. The paperintroduces analytical-experimental methodologies for predicting film properties and cutting performance ofcoated tools. In a first stage, procedures for calculating stress-strain curves and fatigue critical loads ofcoatings by nanoindentations and impact tests respectively, at various temperatures determined, arepresented. In a further stage, methodologies for the assessment of the film adhesion by inclined impact testsand of the film brittleness by nano-impacts are described. Moreover, the effect of the cutting edge impactduration in milling on the tool performance is demonstrated and explained via impact tests at various forcesignal times. Finally, the potential of micro-blasting on PVD coatings at appropriate conditions to improvethe coated tool life is exhibited. In this context, a tool life increase is associated with the appropriateselection of micro-blasting conditions. The relevant results are evaluated by Finite Elements Method (FEM)supported procedures. The described procedures allow the prediction of coated tool cutting performanceand the effective adaption of the cutting conditions to the film properties, thus restricting the relatedexperimental cost.10 13 th International Conference on Tribology – Serbiatrib’13

Tribological Properties ofMaterials and Coatings13 th International Conference on Tribology – SERBIATRIB ’1315 – 17 May 2013, Kragujevac, Serbia

Serbian TribologySocietySERBIATRIB ‘1313 th International Conference onTribologyKragujevac, Serbia, 15 – 17 May 2013Faculty of Engineeringin KragujevacPREDICTION OF COATED TOOLS PERFORMANCE INMILLING BASED ON THE FILM FATIGUE AT DIFFERENTSTRAIN RATESK.D. Bouzakis 1,2,a , R. Paraskevopoulou 1,b , G. Katirtzoglou 1,2,c , S. Makrimallakis 1,2,d , E. Bouzakis 1,2,e ,P. Charalampous 1,f1 Laboratory for Machine Tools and Manufacturing Engineering, Aristoteles University of Thessaloniki, Thessaloniki, Greece2 Fraunhofer Project Center Coatings in Manufacturing, in Centre for Research and Technology Hellasin Thessaloniki, Greece and in Fraunhofer Institute for Production Technology in Aachen, Germanya bouzakis@eng.auth.gr, b paraskeu@auth.gr, c gkatirz@auth.gr, d makrimallakis_s@eng.auth.gr, e E.Bouzakis@certh.gr,f paschalischaralampous@gmail.comAbstract: The knowledge of coated tool wear mechanisms in milling is pivotal for explaining the filmfailure and selecting the appropriate cutting strategy and conditions. In this paper, tool wear experimentswere carried out in milling of four different steels using coated cemented carbide inserts. The variablestress, strain and strain rate fields developed in the tool during cutting affect the film-substratedeformations and in this way the resulting coating’s loads and its fatigue failure. For investigating theinfluence of cyclic impact loads magnitude and duration on the films’ fatigue of coated speciments, animpact tester was employed which facilitates the modulation of the force signal. The attained tool life upto the films’ fatigue failure was assosiated to a critical force for the film fatigue endurance and to thecutting edge entry impact duration. These factors converge sufficiently to the tool life in all examinedmilling kinematics and workpiece material cases.Keywords: milling, tool wear, entry impact duration1. INTRODUCTIONMilling operations are often associated withcomplicated cutting edge-workpiece contact andintensive tool impact loads. These facts render theprediction of the tool wear development a difficultto be achieved task [1, 2]. Recent investigationswith coated cemented carbide inserts revealed thatthe milling up or down kinematic, as well as thecutting parameters, significantly affect the stressfield developed in the cutting edge during thematerial removal and consequently the cuttingperformance [3, 4].The present paper introduces a method forcalculating the coated tool wear evolution inmilling. In such cutting procedures, repetitiveimpact loads with variable duration and magnitudesare exerted on the coated cutting edge, caused bythe interrupted material removal. Hence, it wasnecessary to quantify the effect of the cutting edgeentry impact duration on the coated tool fatiguefailure at various cutting loads. This was enabled bya developed impact tester, facilitating the appliedimpact force modulation [5].2. EXPERIMENTAL DETAILSIn the conducted investigations, peripheral andface milling experiments were conducted by a 3-axis numerically controlled milling center applyingmilling cutters of 17, 35, 57 and 90 mm effectivediameters. The geometry of the cutters and theemployed cutting inserts is exhibited in Figure 1.The cemented carbide inserts are coated by a TiAlNPVD film of ca. 3 μm thickness.13 th International Conference on Tribology – Serbiatrib’13 13

Figure 1. The employed milling cuttersThe chamfer of ca. 280 μm and edge radius 20μm respectively (see Figure 1) contribute to cuttingedge stabilization especially at elevated dynamicloads. This may lead to an effective avoidance ofcutting edge micro breakages, especially when thechip formation is not stable, as for example at thecutting edge entry into the workpiece materialduring up milling [3].The specifications of the applied workpiecematerial are displayed in Figure 2. Four differentsteels were used; the hardened steel IMPAX, thestainless steel 304 L and the hardened steelsNIMAX and 42CrMo4.Figure 3. The employed coatings and substratepropertiesThe mechanical properties of the applied coatingand substrate materials were detected bynanoindentations and a FEM-based algorithm,facilitating the determination of related stress-straincurves [6]. The elastoplastic film material laws aredemonstrated in Figure 3.For rendering possible the modulation of theimpact force characteristics such as of frequency,impact duration and force signal pattern, an impacttester has been employed, in which a piezoelectricactuator is applied for the force generation [5]. Bythis device, the fatigue behaviour of thin hardcoatings at different impact force patternsamplitudes and durations can be investigated.3. IMPACT FORCE AMPLITUDE ANDDURATION EFFECT ON COATINGS’FATIGUE FAILUREFigure 2. The employed workpiece material propertiesFor detecting the effect of the cutting edge entryimpact duration on the film fatigue failure, impacttests at forces of various durations and amplitudeswere carried out on the used coated inserts (seeFigure 4a).14 13 th International Conference on Tribology – Serbiatrib’13

edge without chamfer and smaller radius, versus thecutting length corresponds to a trapezoidal forcepattern at significantly lower entry impact durationof 0.036 ms.Considering these facts and the results exhibitedin Figure 4b, the chamfered coated cutting edgescan withstand to fatigue failure approximately atwo and half times higher entry impact forceamplitude. In this way, at the same stress level, thefilm failure of a chamfered cutting edge may appearin up milling after a longer cutting time comparedto an insert without chamfer. The temperaturedeveloped close to the transient region of thecutting edge between flank and rake amounts toabout 200 o C at a cutting speed of 200 m/min andchip tool contact time up to roughly 15 ms [4].Thus, in this cutting edge region, the crystallinestructure of the investigated TiAlN film remainsstable, no diffusion or oxidation takes place and thefilm fatigue, which can be investigated by theimpact test, is the prevailing factor.Figure 4. a) Triangular and trapezoidal impact forcesignals b) Effect of impact signal and entry impactdurations on the critical force amplitudeAll applied triangular force signals withdurations (FSD) of 10 ms, 20 ms and 35 ms and thetrapezoidal ones of 20 ms and 40 ms, which arepresented at the upper Figure 4a part, had aconstant signal growth time t e of 5 ms (entry impactduration t e ). In contrast, the displayed force signalsat the bottom of Figure 4a possess different entryimpact durations t e from about 0.5 ms up to 15 ms.These force signals are created by the piezoelectricactuator and measured by the piezoelectric forcetransducer.The effect of the force pattern on the criticalforce amplitude, which induces coating fatiguefailure after one million impacts, is monitored inFigure 4b. According to these results, the criticalfatigue force amplitude remains practicallyinvariable versus the force signal duration atconstant t e . On the other hand, t e affectssignificantly the film fatigue behaviour, as it isexhibited in the same diagram. An increase of theimpact entry duration t e from 0.05 ms up to 15 msresults in a significant critical fatigue impact forceamplitude augmentation from about 60 daN up to220 daN respectively. The cutting load signal, i.e.the stress course versus the cutting length, when achamfered cutting edge is used, resembles to atriangular force signal at entry impact duration of3.6 ms [3]. Moreover, the stress course on a cutting4. FLANK WEAR DEVELOPMENT VERSUSTHE CUTTING EDGE ENTRY IMPACTDURATIONThe contact conditions at the tool entry into thematerial in milling are pivotal for the tool wear [1,2, 4, 7, 8]. The impact load on the cutting edge atthe tool entrance into the workpiece materialdepends on the milling kinematic (up or down,peripheral or face), since these factors affect thedeveloped chip geometry and thus the stress fieldsof the coating versus the tool rotation. The entryimpact duration corresponds to the cutting time, upto the development of the maximum equivalentstress in the coating.For describing the effect of the entry impactduration on the tool wear in milling with coatedtools, the accumulated tool life is introduced. Thelatter parameter refers to a flank wear land widthVB of 0.15 mm. This parameter can be calculatedconsidering the undeformed chip length l cu , thecutting speed v and the attained number of cutsNC 0.15 up to the same VB according to the equationshown in the upper part of Figure 5a. In Figure 5aand 5b characteristic examples concerning theeffect of the entry impact duration on the tool lifeare exhibited. These examples refer to peripheraland face milling of different undeformed chiplengths. Further examples in milling at variousconditions, kinematics and materials are presentedin [3, 9, 10, 11]. As it can be observed in Figure 5a,at an undeformed chip length of roughly 80 mm, asimilar tool wear evolution in up and down, face orperipheral milling develops, leading to almost thesame accumulative tool life.13 th International Conference on Tribology – Serbiatrib’13 15

Figure 5. Flank wear land width versus number of cutsin various cases of face and peripheral millingMoreover, as it is demonstrated in Figure 5b,when up milling is applied, the flank weardevelopment is less intense compared to downperipheral or face milling at a chip length of about40 mm. The attained accumulative tool life in upmilling is approximately three times highercompared to those ones in down milling. Thisbehaviour can be explained, based on the developedcutting edge entry impact duration in the previouslydescribed cases.To highlight this effect, in Figure 6, the obtainedaccumulative tool life in the investigated peripheraland face milling cases is displayed versus thecutting edge entry impact duration t e . The curve inthis chart describes the effect of the cutting entryimpact duration on the accumulated tool life. Therelevant results were obtained in milling, at varioustool geometries, cutting kinematics and conditions[3, 9, 10, 11].In down milling, face or peripheral, atundeformed chip lengths l cu of ca. 40 mm, thecutting edge entry impact durations t e amount toapproximately 0.1 ms leading to the accumulativetool life diminishing.Furthermore, in up milling at an undeformedchip length l cu of ca. 40 mm, due to the smootherchip thickness growth at chip formation start, thecutting edge entry impact duration t e isapproximately 2.2 ms and the accumulative tool lifeincreases significantly compared to thecorresponding one in down milling.In contrary, in down and up milling, face orperipheral, at undeformed chip lengths l cu of about80 mm, the entry impact duration varies from 3.1 to5.4 ms and the accumulative life remains almost onthe same level.Considering Figure 6, it can be concluded thatentry impact duration larger than 2 ms leadpractically to almost the same accumulative toollife. Furthermore, it is obvious, that short entryimpact durations correspond to comparably lowercoating fatigue critical forces (see Figure 4) anddiminishes the coated tool life. Longer entrydurations improve the film fatigue behaviour, thusenhancing the coated tool life.The accumulated tool life in milling of theemployed hardened steel IMPAX versus the entryimpact duration at various cutting speeds isdisplayed in Figure 7. The accumulated tool life inmilling of the employed hardened steel IMPAXversus the entry impact duration, displayed in Figure7, can be described by the equations, displayed inFigure 7b, for the cutting speeds of 100, 200 and 300m/min.Similar experiments were conducted for allemployed hardened steels. Figure 8 illustrates theaccumulated tool life in milling of NIMAX, AISI304 L and the 42CrMo4 versus the entry impactduration at various cutting speeds. The obtainedaccumulated tool life of NIMAX is substantiallylower than the corresponding of IMPAX at thesame cutting speed and almost equal to 1/3 of that.Figure 6. Accumulated tool life in milling versus theentry impact duration16 13 th International Conference on Tribology – Serbiatrib’13

Figure 7. Accumulated tool life in milling of theemployed hardened steel IMPAX versus the entryimpact duration at various cutting speedsfunction of the cutting speed and the entry impactduration is:T0.15C3( v,te) CC41teC2e(1)The parameters C 1 , C 2 , C 3 and C 4 depend on thecutting tool and workpiece material data. Moreover,these parameters are functions of the cutting speedand the entry impact duration.Considering the entry impact duration, usingequation (1), the cutting tool life T 0.15 up to a flankwear land width VB equal to 0.15 mm can beestimated. Moreover, the number of cuts NC 0.15corresponding to a flank wear land width VB equalto 0.15 mm can be calculated based on theundeformed chip length and the cutting speed usingthe relation (2).T0 .15 NC0.15 lcuv (2)Bearing in mind that a number of cuts equal tozero corresponds to a tool wear VB also equal tozero and the number of cuts NC 0.15 is associated toVB equal to 0.15 mm, the evolution of the toolwear during milling can be calculated as describedin [9].6. COMPUTATION OF THE TOOL WEARIN MILLING AT CHANGEABLECUTTING CONDITIONSDuring milling a workpiece, the values ofparameters influencing the tool wear developmentsuch as chip length, chip thickness, entry impactduration etc. may vary in the successive tool paths.Considering these circumstances, for computing thetool wear developed during milling, themethodology explained in Figure 9, is applied [15,16].Figure 8. Accumulated tool life in milling of theemployed hardened steels versus the entry impactduration at various cutting speedsThis is due to comparatively higher hardness ofNIMAX. Moreover, it is obvious that due toreasons described in [11, 12, 13, 14] stainless steelis difficult to cut.5. THE DEVELOPED MODEL FORDESCRIBING THE WEAR EVOLUTIONON COATED TOOLS IN MILLING BASEDON CUTTING EDGE ENTRY IMPACTDURATIONThe general form of the equations, shown inFigure 7, describing the accumulated tool life as aFigure 9. Determination of tool wear evolution inmilling at various cutting conditions13 th International Conference on Tribology – Serbiatrib’13 17

Based on the cutting data of every tool path, thenumber of cuts NC i and furthermore the tool wearVB i at the end of a tool path (i) can be calculated,as demonstrated in this figure. The flank wear VB i-1developed in the previous tool path (i-1), is relatedto a number of cuts NC i-1 considering the cuttingdata of the actual tool path. The number of cuts NC idata of the actual tool path is added to the NC i-1and thus the flank wear VB i at the tool path (i) canbe determined. By this method the flank weardevelopment can be effectively predicted in allsuccessive cutting tool paths.7. AN APPLICATION EXAMPLE OF THEDEVELOPED METHODOLOGYThe analytical method for estimating the toolwear is applied in the case of a test part presented inFigure 10. Considering the initial and finalworkpiece’s geometry, the tool paths required toremove the raw material volume were defined usingthe commercial “OPUS-CAM” system [17].Figure 10. The employed test part and the tool pathsrequired for the material removalFigure 11. Determination of chip data along the toolpaths by a CAD/ CAM systemThe determined tool paths are presented in thelower part of Figure 10 too. The machining tookplace in forty z-levels. The raw material removalwas accomplished using up milling and downmilling as well. Both operations lead to the samefinal workpiece shape, but the tool wear behaviourin each case may be different.After the tool paths have been determined, the“Schnitte.dat” file is generated by OPUS, as shownin Figure 11. This file contains geometrical datarelated to the chips formed in each tool path. Morespecifically, the parameters illustrated in Figure 10,determined at certain distances from every tool pathinitial point are stored into the “Schnitte.dat” file.In the first column of the file, the tool position isdefined as a percentage p of the actual tool pathlength l i , whereas i is the number of the tool path.At every tool position, the angle φ ref of the first toolrake – workpiece contact, the corresponding entryangle φ ent at the maximum cutting edge penetrationinto the part material and the exit angle φ ex arestored. Moreover, in the following columns, theundeformed chip length l cu , the axial depth of cut a zand the chip width b are accumulated. The data ofthe “Schnitte.dat” file are further processed by thedeveloped Data - PREparation (DAPRE) software.Thus, various data, as for instance the entryimpact time per chip, the undeformed chip lengths,the tool –workpiece contact angle etc. can beprovided. Considering these data the coated toolwear evolution versus the number of cuts isdescribed and in this way, the conduct ofalgorithms for an analytical optimization of millingprocess towards attaining set targets is facilitated.18 13 th International Conference on Tribology – Serbiatrib’13

Figure 13. Calculated and measured flank weardevelopment versus the number of cuts8. CONCLUSIONSFigure 12. Histograms of the entry impact durationalong the tool pathsCharacteristic results of this methodology aredisplayed in Figure 12, where histograms of theentry impact time of the removed chips in both upand down milling kinematics are illustrated.In up milling almost all chips were cut at impactduration of approximately 4,8 ms. In contrary,when down milling is applied almost half chipspossess entry impact durations of less than 4 ms,while some of them are associated with impactdurations less than 1 ms. In this way, it is expecteda more intense wear evolution in down millingcompared to up one.It is has to be pointed out, that the more intensetool wear evolution in down milling of thisparticular test part compared to the up one, cannotstand for every milling case and depends on theworkpiece and the tool edge geometry and materialdata.For calculating the tool wear developed duringmilling of the test part, the introduced method inprevious paragraph was used. The flank wear landwidth VB versus the number of cuts NC wascalculated and experimentally detected. Themeasured and the calculated values of the tool wearevolution in both milling kinematics are presentedin Figure 13. The experimental results convergesufficiently with the calculated ones.The results described in this paper show thesignificant effect of the cutting edge entry impactduration on the coated tools wear evolution inperipheral and face milling. The effect of cuttingedge entry impact duration on the coated toolfatigue failure was investigated via an impact testerwith force signal modulation facilities. Moreover,based on the cutting edge impact duration, acalculation of the expected tool wear developmentcan be carried out. In this way, the selection ofoptimum cutting conditions and strategies inmilling with coated tools can be achieved.REFERENCES[1] H.-J. Jakobs, P. Winkelmann: AktuelleStandzeitfunktion für die Arbeitsgestaltung beimFräsen, Fertigungstechnik und Betrieb, Vol. 31, pp.352–356, 1981.[2] L.J. Dammer: Ein Beitrag zur Prozessanalyse undSchnittwertvograbe beim Messerkopfstirnfräsen,Dissertation, RWTH Aachen, 1982.[3] K.-D. Bouzakis, S. Makrimallakis, G. Katirtzoglou,E. Bouzakis, G. Skordaris, G. Maliaris, S. Gerardis:Coated tools’ wear description in down and upmilling based on the cutting edge entry impactduration, CIRP Annals - Manufacturing Technology|Vol. 61, No. 1, pp.115-118, 2012.[4] Bouzakis K.-D, Gerardis S, Katirtzoglou G,Makrimallakis S, Michailidis N, Lili E., IncreasingTool Life by Adjusting the Milling CuttingConditions According to PVD Films’ Properties,CIRP Annals – Manufacturing Technology, Vol.57, No. 1, pp.105–108, 2008.13 th International Conference on Tribology – Serbiatrib’13 19

[5] K.-D. Bouzakis, G. Maliaris, S. Makrimallakis:Strain rate effect on the fatigue failure of thinPVD coatings: An investigation by a novel impacttester with adjustable repetitive force,International Journal of Fatigue, Vol. 44, pp. 87-97, 2012.[6] K.-D. Bouzakis, N. Michailidis, G. Erkens: Thinhard coatings stress-strain curves determinationthrough a FEM supported evaluation ofnanoindentation test results, Surface and CoatingsTechnology, 142-144, pp. 102-109, 2001.[7] M. Kronenberg: Analysis of Initial Contact ofMilling Cutter and Work in Relation to Tool Life,Transactions of the ASME, pp. 217–228, 1946.[8] K. Okushima, T. Hoshi: The Effect of the Diameterof Carbide Face Milling Cutters on Their Failures,Bulletin of JSME, pp. 308–316, 1963.[9] K.-D. Bouzakis, R. Paraskevopoulou, G.Katirtzoglou, E. Bouzakis, K. Efstathiou: CoatedTools Wear Description in Milling FacilitatingConsiderations towards Sustainable Manufacturing,The 10th Global Conference on SustainableManufacturing, pp. 20-25, 2012.[10] K.D. Bouzakis, R. Paraskevopoulou, G.Katirtzoglou, S. Makrimallakis, E. Bouzakis, K.Efstathiou: Predictive model of tool wear in millingwith coated tools integrated into a CAM system,accepted for publication, CIRP Annals -Manufacturing Technology 2013.[11] K.-D. Bouzakis, S. Makrimallakis, G. Skordaris, E.Bouzakis, S. Kombogiannis, G. Katirtzoglou, G.Malliaris: Coated tool performance in up and downmilling stainless steel, explained by film mechanicaland fatigue properties, accepted for publicationWear, 2013.[12] L. E. Murr : Metallurgical effects of shock and highstrain-rateloading, In: T.Z. Blazynski (Ed.)Materials at high strain rates, Elsevier, Essex,England, pp.1-46, 1987.[13] J. Harding: The effect of high strain rates onmaterial properties, In: T.Z. Blazynski (Ed.)Materials at high strain rates, Elsevier, Essex,England, pp. 133-186, 1987.[14] B.L. Boyce, M.F. Dilmore: The dynamic tensilebehavior of tough, ultrahigh-strength steels atstrain-rates from 0.0002 s -1 to 200 s -1 , InternationalJournal of Impact Engineering, Vol. 36, No. 2,pp.263-271, 2009.[15] K.D. Bouzakis: Konzept und technologischeGrundlagen zur automatisierten Erstellungoptimaler Bearbeitungsdaten beim Wälzfräzen,Habilitation, TH Aachen, 1980.[16] K. Efstathiou: Automatic generation of optimumtechnological data for numerically controlledmilling based and on measurements of theworkpiece solid geometry, Ph. D. Thesis: AristotlesUniversity, Thessaloniki, 1991.[17] Opus-CAM, 2012, User Manual.20 13 th International Conference on Tribology – Serbiatrib’13

Serbian TribologySocietySERBIATRIB ‘1313 th International Conference onTribologyKragujevac, Serbia, 15 – 17 May 2013Faculty of Engineeringin KragujevacSELECTIVE TRANSFER OF MATERIALS IN THE ASPECT OFGREEN TRIBOLOGYEmilia Assenova 1 , Gottlieb Polzer 2 , Dr. Tsermaa 3 , Mara Kandeva 41 Society of Bulgarian Tribologists, Sofia, Bulgaria, emiass@abv.bg2 Tribology Unit, 08115 Schoenfels, Germany, professor-polzer@gmx.de3 University of Ulan Bator, Mongolia, professor-polzer@gmx.de4 Technical University-Sofia, Tribology Center, Sofia, Bulgaria, kandeva@tu-sofia.bgAbstract: One of green tribology’s principles touches the environmental implication of coatings and lubricants,development, optimization and implementation of ecology friendly manufacturing and implementation ofcoatings.The paper deals with green tribology in the aspect of wear reducing frictional coatings andregeneration of worn surfaces without joint dismantling. Copper frictional coatings in the case of nonabrasivetreatment of steel or cast iron surfaces, their production with the assistance of selective transfer of materialsbetween the friction surfaces are considered. In the example of frictional coating deposition, this phenomenonis supported by the rubbing of brass stick on steel cylinder surface under particular conditions of selectivetransfer in the presence of a special lubricant. Important is the extremely low wear of components coated undercondition of selective material transfer mode with wide practical application. The interdisciplinary character ofthe study and application of technologies for coating formation, layer growth techniques, surface texturing, etc.involves studies by specialists of different sciences.Keywords: green tribology, frictional coatings, surface design, selective transfer.1. INTRODUCTION1.1 Tribology and Green TribologyNatural resources have been cruelly consumedin the last three centuries and the earth is seriouslydamaged and polluted. Humanity has to survivefighting with the pollution and the deficiency ofmaterial, energy and cleanness. Generally speaking,this problem is mainly the result of the misuse ofour contact with nature. So, it is contact deficiencyand a way out could be sought in the science ofcontacts, i.e. in tribology. Radical knowledge andtechnologies of sustainability are needed toestablish new human way of thinking.Tribology is supposed to assist the knowledgeand technologies in the purpose to meet theexpectations of quality, reliability and environmentsustainability. Tribology comprises the knowledgeof friction, lubrication, wear, hermeticity and otherprocess between contacting surfaces. The modernconcept observes tribology problems as essentiallyinterdisciplinary. Typical tribological studiesinvolve the efforts of mechanical engineers,material scientists, chemists, physicists, and soforth. New areas of tribological studies have beendeveloped at the interface of various scientificdisciplines, for example, nanotribology,biotribology, geotribology, ecotribology, etc.Recently, the new concept of ‘green tribology’has been defined as ‘the science and technology ofthe tribological aspects of ecological balance and ofenvironmental and biological impacts’ by H. P. Jost[1,2]. The former notion was eco-tribology andstressed the interaction of contact systems with theenvironment [3,4]. Green Tribology means savingmaterials, energy, improving the environment andthe quality of life. The area of Green Tribology willdirectly affect the economy by reducing waste andextending equipment life, improve thetechnological and environmental balance, andimprove the sustainability and safety in the humansociety. Green Tribology reflects in fact thetribological aspects of ecological balance and ofenvironmental impacts, and is expected to directlyaffect the reducing of waste and thermal pollution,and extending equipment quality, reliability and13 th International Conference on Tribology – Serbiatrib’13 21

life, which are some of the key challenges facingthe societies today [5], [6], [7].1.2. Wear preventionTribological knowledge helps to reveal and healwear related problems. So, it is possible to improvequality significantly by measures preventing thereasons for failures related to wear of contactingsurfaces. What is wear? Wear is a process oftribological interaction resulting in physicochemicalloss of material (weight, size or shape)from the surfaces in contact. Most important formsof wear are abrasion, corrosion, erosion, attrition,fretting, thermal destruction, scuffing, pitting,etching, etc. [8]The specific field of green or environmentfriendlytribology emphasizes the aspects ofinteracting bodies, which are of importance formaterial, and energy sustainability and safety, andwhich have huge impact upon today’s environment.This includes essentially the control of friction andwear, being of importance for energy, resources andcleanness conservation [5]. One of the mostimportant tasks of Green Tribology is Minimizationof wear. Wear limits the lifetime of components andcreates the problem of their recycling. Wear can leadto high consumption of the natural resources. Wearcreates debris and particles that contaminate theenvironment and can be even hazardous for humans.Moreover, the large amount of heat generated in thecontact joints, also leads to its thermal distortion andfailure, and to pollution of the environment withmaterial waste and heat.Measures for minimizing wear are related tosurface processing, namely optimal materialselection and surface texturing, and/or coating thesurfaces. It leads to good health and preservation ofperformance quality of machines, equipment andproduction systems, and hence, material, energyand environment saving as a whole.1.3. Surface coatingsThere are various methods for surface coatingsdeposition, a diversity of approaches to study thebehavior of the coatings, and numerous areas oftheir application.Wear prevention coatings are applied in manyareas: production industry and power industry,marine, automotive and transportation industry,aerospace techniques, agriculture, food processing,mining and metallurgy, sporting equipmentindustry, electronics, packaging, robotics,renewable energy sources, waste treatment andmore. A great variety of parameters influences thequality of the coating, depending also on theapplication. Important characteristics are: thickness,porosity, microstructure, inclusions, cracks,microhardnes and adhesion and cohesion bondstrength [9]. Control is realized using variousstandardized test methods by means of tensile testmachines, scratch testers, etc.The paper concerns the method of coatingdeposition during friction process. It aims andfocuses on the procedures of obtaining and thestudy of copper frictional coatings under selectivetransfer mode in the case of nonabrasive treatmentof steel and cast iron surfaces.2. FRICTIONAL COATINGS UNDERCONDITIONS OF SELECTIVETRANSFER2.1. BackgroundTribologists have the task to keep the destructionas small as possible or to stop it, in order that thesystem comes to the equilibrium process betweendestruction and regeneration. Exactly this happens inthe process of selective transfer of material betweenfriction surfaces. In the case of frictional coatingsproduction, this phenomenon is assisted by rubbingof brass against steel under the special conditions ofselective transfer.D. N. Garkunov and G. Polzer are of the firstresearchers in theory and practice of selectivetransfer of material during friction coatingdeposition [10], [11], [12], [13]. Common workswere carried out connecting the Tribology Center inSofia and the Tribology Group of Prof. Polzer inZwickau, and recently in Schoenfels, Germany.What is friction coating deposition? A steel element(e.g. a shaft) to be coated is both subjected torotation and to the pressure of а brass stick in thepresence of a special lubricant, forming a bronzesteeltribocouple (See the principle in Fig. 1).Figure 1. Principle assembly for brass deposition onshaft (1 - surface active liquid; 2 - brass rod; 3 - shaftto be coated)The film forms on the friction surfaces in thebronze-steel tribocouple with glycerin lubricationpassing firstly through dissolution of the bronzesurface, where the glycerin acts as a weak acid. Theatoms of the elements (tin, zinc, iron, aluminum)absorbed in bronze outgo into the lubricant, asresult the bronze surface is enriched with copper.22 13 th International Conference on Tribology – Serbiatrib’13

Friction deformation of the bronze surface causesnew passing of elements into the lubricant, so thebronze layer is purified and it nearly contains onlycopper. Its pores fill with glycerin. Glycerin isreducer for copper oxides, hence the copper film isfree form oxides; it is very active with free ions andis highly adhesive for the steel surface. The steelsurface is covered by thin copper layer. Selforganizationand selective transfer of copper tosteel take place. Before the stabilized selectivetransfer, the process goes on until steel and bronzeare coated by 2 μm copper layers [12]. Mechanicaland chemical transformations take place; e.g.formation of surface active substances on thefriction surfaces; they interact chemically with thesurfaces and form chemisorbed layers (see Fig. 2).the special conditions of selective transfer ofmaterial. Different processes result. In the contactzones emerges reactive coating deposition withspecial properties: Copper is rubbed on the steelfriction surfaces with totally different electrochemicalpotential, and secondly, not only thecontent but also the structure in the friction surfacesis being changed [13].The compress forces at the rotation of the brassstick involve great pressure in the contact zonebetween stick and basic material due to the relativesmall contact surface, hence a positive gradient ofthe shear strength in depth direction of the frictionsurface according to I. V. Kragelsky [8].Figure 2. Formation of micelles and interaction ofsurface-active substances with bronze (as per [12])Some results of the basic studies and applicationin the area of copper frictional coatings arepresented below. Based on equation of thetheoretical physics, G. Polzer [11] had formerlyderived equations of self-organization at friction.Always when destruction problems are available innature, there is either a simultaneous growthprocess which involves equilibrium betweendestruction and regeneration or destruction leads toexponential destroying of the whole system, in ourcase the tribological couple.A self-organization in the system brass-glycerolsteelis observed and the obtained film – a coat withsignificant change of wear-resistance. Major resultis the low wear of components coated undercondition of selective material transfer mode.Important is also the reduction of the concentrationof hydrogen at the frictional surface and,respectively, the lower hydrogen wear. It is highlyimportant for practical applications that theinclination for welding and seizure [13] betweenthe friction surfaces is significantly lowered underconditions of selective transfer. A considerablepractical result is the possibility for dismantlingfreerestoration of worn units/couples.2.2. Experimental workThe phenomenon of direct coating deposition isassisted by the rubbing/deposition of brass underFigure 3. MBZ 3A Brass-coating device for slidingbushes (application in lathes)Figure 4. Brass-coating device for cylinder-bushing byboring machineAs a result, a tribological system appears whichcan bear higher loads at the influence of variousprocesses. Different machines were designed andconstructed at the Department Tribotechnik inZwickau’ Higher Technical School, correspondingto the principles of the frictional deposition and theideas of the selective transfer. Many pieces of the13 th International Conference on Tribology – Serbiatrib’13 23

devices „MBZ 1" for shaft coatings and „MBZ 3A" for application in rotating machines weremanufactured (see Figs.3, 4, 5, 6), e.g. the „MBZ 3A" for engine cylinders was produced in 30 items.Unfortunately there is not sufficient use of theadvantages of the deposition of copper frictionalcoatings in the overall practice.coatings, however, improve significantly thewearresistance against hydrogen wear [13].Figure 5. View of the brass-coating device MBZ 3AFigure 7. Hardness in different depth after frictionalcoating deposition on steelFigure 6. The brass-coating device applied in anautomatic machineSome diagrams referring to a part of the basicnew results are presented below. In Fig. 7 is giventhe variation of hardness in depth; so thestrengthening can also be obtained at differentrotation speeds.Fig. 8 shows the reduction of hydrogenconcentration of the friction surface in depth. Thehydrogen wear results from synergetic interactionof various surface phenomena: exoemission,adsorption, frictional destruction, which providehydrogen extraction from the frictional surfaces.Thermal gradient is also formed, as well aselectrical and magnetic fields; this leads tohydrogen diffusion in the metal, hydrogenconcentration in the subsurface layer and rapidwear of this layer [12]. Metal defect formation inthe friction deformed layer also increases the H 2concentration and augments the wear. FrictionalFigure. 8. Reduction of H 2 concentration at thefrictional surface in depthFigure 9. Wear distribution in upper dead point of enginecylinder after different sliding paths: for cylinder withfrictional brass coating (left) and uncoated (right)By means of brass frictional coating in differentconstructions of steel and cast iron it can beobtained not only the 10 - 20 % lowering of frictionforce, but also a changed wear distribution, whichis to be seen, e.g., for the upper death point in24 13 th International Conference on Tribology – Serbiatrib’13

engine cylinders of 2-cylinder-twotact-Ottomotorsafter various completed paths (see Fig. 9). This wasthe reason that the brass frictional coatings weresuccessfully applied in the practice of the companyPeißig in Zwickau, especially in highly loaded racemotors for more than 20 years too.3. CONCLUSIONGreen tribology should be integrated into worldscience and make its impact on the solutions forworldwide problems. Being a new field, greentribology has a number of challenges. A basic oneof them, minimising the wear, is being discussed inabove investigation of wear reduction possibilitiesthrough frictional coatings.The study of frictional coatings and theirapplication can be summarized in the following: Self-organization in the system brass-glycerolsteelunder selective transfer is observed andthe obtained film – a designed or controllablecoat with significant change of wearresistance– can be intentionally manipulatedto influence its properties during friction. Important features of the coating depositedduring friction under selective material transfermode: Low wear of components at nonabrasivetreatment of steel/cast iron, and lower hydrogenwear of the coated surfaces; lower inclinationfor welding and seizure between the frictionsurfaces; possibility for dismantling-freerestoration of worn units/couples.The practical implementation of brass-copperfrictional coating is of extreme importance and wasrealized in Germany, Russia, Kazakhstan, Poland, etc.The interdisciplinary character of the study andapplication of technologies for frictional coatingformation, layer growth techniques, surfacetexturing, etc. involves intervention by specialistsof different sciences. The work and collaborationbetween scientists of Russia, Germany, Poland,Bulgaria, Mongolia and Vietnam in this field wascarried out by the International Council forSelective Transfer and Frictional coatings,established in 1990 in London.REFERENCES[1] H. P. Jost: 30 th Anniversary and Green Tribology.Report of a Chinese Mission to the UnitedKingdom, 7-14 June 2009, issued by the TribologyNetwork of the Institution of Engineering &Technology[2] H. P. Jost: The Presidential Address, 5 th WTCKyoto, 2009.[3] W. J. Bartz: Ecotribology: Environmentallyacceptable tribological practices. TribologyInternational, Volume 39, Issue 8, August 2006, pp.728–733.[4] M. Kandeva, E. Assenova, M. Daneva:Triboecology as a methodological center of modernscience. Proceedings of the 2 nd EuropeanConference on Tribology ECOTRIB 2009, Pisa,Italy, 2009.[5] M. Nosonovski, B. Bhushan (edtrs): Greentribology. Biomimetics, Energy Conservation andSustainability, Springer Verlag, 2012.[6] R. Wood: NCats Newsletters, Univ. ofSouthampton, October 2011.[7] E. Assenova, V. Majstorovic, A. Vencl, M.Kandeva: Green tribology and quality of life,International Convention on Quality 2012, Belgrade(Serbia), 05-07.06.2012, Proceedings, p.p. 32-38,Published in: Advanced Quality, 40, 2, p.p.26-32,2012.[8] I.V. Kragelsky: Friction and Wear, Moscow,Mashinostroienie, 1968 (in Russian; available alsoin English)[9] A. Vencl, S.Arostegui, G. Favaro, F. Zivic, M.Mrdak, S. Mitrović, V. Popovic: Evaluation ofadhesion/cohesion bond strength of the thick plasmaspray coatings by scratch testing on coatings crosssections,Tribology International, 44, 11, 2011, p.p.1281-1288.[10] D.N. Garkunov: Triboengineering (wear and nondeterioration),Moscow Agricultural AcademyPress, Moscow, 2000 (in Russian)[11] G. Polzer: Der Erfahrungsaustausch:Reibbeschichten und selective Uebertragung, Publ.Bezirks-Neuerer-Zentrum, Gera, 1988.[12] D.N. Garkunov: Scientific Discoveries inTribotechnologies. No-wear effect under friction:Hydrogen wear of metals. MAA Publishing House,Moscow, 2007.[13] G. Polzer, E. Assenova, Dr. Tsermaa: Copperfrictional coatings under conditions of selectivetransfer, Tribological Journal BULTRIB - Sofia,Vol. 3, 2012.13 th International Conference on Tribology – Serbiatrib’13 25

Serbian TribologySocietySERBIATRIB ‘1313 th International Conference onTribologyKragujevac, Serbia, 15 – 17 May 2013Faculty of Engineeringin KragujevacABRASIVE WEAR AND WEAR-RESISTANCE OF HIGHSTRENGTH CAST IRON CONTAINING Sn MICROALLOYMara Kandeva 1 , Boryana Ivanova 21 Tribology Center, Technical University-Sofia, Bulgaria, kandeva@tu-sofia.bg2 Technical University-Sofia, Bulgaria, bsaykova@tu-sofiaAbstract: A procedure for the study of wear of high strength (spheroid) cast iron under conditions of dryfriction on surfaces with fixed abrasive following the kinematics scheme „pin-on-cylinder” with spiralmovement has been developed. Five type specimens of high strength cast iron without and with micro alloyof various Sn contents – 0,08; 0,02; 0,06; and 0,12 mass percents were studied. The experimental resultslead to graphs and diagrams of the relationships for the parameters mass and linear wear, wear rate andintensity, and wear-resistance depending on process time, sliding way and normal load.This study is connected with the completion of a PhD dissertation and of the tasks under the Project ДУНК-01/3 “University R&D Complex for innovation and transfer of knowledge in micro/nano-technologies andmaterials, energy efficiency and virtual engineering” funded by the Bulgarian Ministry of Education andScience.Keywords: tribology, high strength cast iron, micro alloying, abrasive wear, wear-resistance1. INTRODUCTIONBeing a natural composite material with steelmetal matrix with embedded graphite phase, thehigh strength (spheroid) cast iron provides acomplex of properties which make it different fromthe conventional Fe-C alloys.The mechanical and tribological properties arestrongly dependent on the composition, structure,and on the size and distribution of the graphiteinclusion, as well as on the presence of microalloyingelements both in bulk and surface layer.Tin (Sn) is most often used as alloying element.The usual quantities of less than 0.15 % do notinfluence the leaning to graphite adoption in thecrystallization process.Alloying of spheroid cast iron by Sn causesperlitization of the metal base, along with strengthand hardness increase by decrease of the relativeincrement of collision resilience. This influencesthe parameters of friction and wear in the contactjoints of machines [1,2,3].The paper aims study of the parameters of wearof high strength cast iron micro-alloyed by variousmass percent contents of tin (Sn) under conditionsof dry friction on a surface with fixed abrasiveparticles.2. MATERIALS, PROCEDURE ANDPARAMETERS OF WEAR2.1. MaterialsSample specimens of high strength cast ironwith the following mass percent contents of tin(Sn): 0,018%, 0,020%, 0,032% and 0,051%. Thechemical composition and the designation of thesample specimens are given in Table 1.Wedge-shaped sample specimens were obtainedthrough gravitational casting in the factory ”Osam”in the city of Lovech.Hardness was measured by means of Brinellhardness meter of the type 2109TB, using a steelball of diameter 10 mm and normal load 30 kN, by15 s hold time. [4]Table 2 shows specimens’ hardness.26 13 th International Conference on Tribology – Serbiatrib’13

Table 1: Chemical composition of sample specimens№ Chemicalelement, %Specimen’s number0 1 2 3 41 C 3,87 3,87 3,87 3,87 3,872 Sn - 0,018 0,020 0,032 0,0513 Si 1,55 1,55 1,55 1,55 1,554 Mn 0,34 0,34 0,34 0,34 0,345 P 0,029 0,068 0,063 0,075 0,0776 S 0,012 0,051 0,059 0,047 0,0607 Cr 0,030 0,030 0,030 0,030 0,0308 Mo 0,018 0,019 0,020 0,017 0,0189 Ni 0,024 0,024 0,024 0,024 0,02410 Co 0,013 0,017 0,014 0,013 0,01311 Cu 0,051 0,058 0,077 0,059 0,07012 Ti 0,0013 0,0013 0,0018 0,0015 0,001313 W 0,126 0,126 0,135 0,123 0,12614 Pb 0,039 0,039 0,043 0,040 0,03915 As 0,036 0,036 0,037 0,038 0,04016 Zr 0,003 0,003 0,003 0,003 0,00317 B 0,0083 0,0083 0,0074 0,0091 0,0088Table 2: Specimens’ hardnessSpecimen’s No. 0 1 2 3 4Sn, % - 0,018 0,020 0,032 0,051Hardness, НВ 179 197 203 262 2772.2 Procedure and device for abrasive wearstudyThe experimental study was realized by aprocedure and device for quick tests according tothe kinematical scheme „pin-on-disk”. Figure 1shows the functional scheme of the device. Theprocedure was elaborated in the Laboratory ofTribology at the Faculty of Industrial Technologyof the Technical University – Sofia. The actuallyvalid standards were taken into consideration [5,6].The studied cylindrical specimen 3 (the body)was mounted fixed in an appropriate holder of theloading head 6. Its position allows that the frontalsurface К enters in contact with the abrasive surface2 of the horizontal disk 1 (the counter-body). Thehorizontal disk 1 is rotating with constant rotationalspeed ω = const around its vertical axis. Thenumber of revolutions of the disk 1 is read by therevolution-counter 5.The device allows variation of the relativesliding speed between the specimen 3 and the disk1 using two manners: by changing the rotationalspeed of the disk through a control unit or byvariation of the distance R between the revolutionaxis of the counter-body 1 and the axis of thespecimen 3.Figure 1: Functional scheme of the device “pin-on-disk”The abrasive surface 2 of the counter-body 1 isbeing modeled through surfaces of impregnatedcarbo-corundum with hardness minimum 60%higher than the hardness of the tested coatingsaccording to the requirements of the standard.The procedure of the investigation comprises thefollowing sequence:1. The surfaces of all specimens, which are of equalcylindrical shape and size, are subjected tomechanical treatment in three stages – rough,grinding and polishing, up to obtaining the equalroughness Ra = 0 ,4 ÷ 0, 6 µ m .13 th International Conference on Tribology – Serbiatrib’13 27

2. The mass of the specimen is measured beforeand after a given sliding path (number of cycles ofinteraction) by means of electronic balance of thetype WPS 180/C/2 with accuracy up to 0,1 mg .Specimens are cleaned with a solution neutralizingthe static electricity before each measurement.3. The specimen 3 is fixed in the loading head 6 ina given position, and by means of system ofleverages the normal central load P is being set.2.3 Parameters of wearParameters of the studied mass and linear wear aregiven in Table 3.Table 3. Parameters of wearMass wearmass, [mg] m o -mwear rate, [mg/min]m o -m/twear intensity, [mg/m]m o -m/Sspecific intensity, [mg/mm 2 m] m o -m/A a S,absolute wear-resistance, [m/mg] S/m o -mspecific wear-resistance, [mm 2 m/mg] S.A a /m o -mLinear wearwear, [µm] h o -hwear rate, [µm/min]h o -h/twear intensity, [µm/m]h o -h/Sspecific intensityh o -h/A a S[µm /mm 2 m]absolute wear-resistance, [m/ µm] S/h o -hspecific wear-resistance [mm 2 m/ µm] S.A a /h o -hThe designations in the table are as follows: А а– apparent contact area of sliding; S – sliding path.The factor “comparative wear-resistance” ε isintroduced, which is non-dimensional and gives theratio between the absolute wear-resistance of thetested specimen and the wear-resistance of a chosenreference sample. A sample of high strength castiron without Sn micro-alloy was accepted asreference sample by the authors.All specimens are studied under equalconditions given in Table 3.Table 3: Test parametersnormal loadР = 10,3 [N]apparent contact areaA a = 78,5.10 -6[m 2 ]apparent contact pressure Р а = 13,12[N/cm 2 ]average sliding speedV = 13,1 [cm/s]type of the specimencylindricalmaterial density of the7,8.10 3 [kg/m 3 ]specimeninitial roughness of theRa = 0,4÷0,6specimen[µm]abrasive surface Corundum Р 3203. EXPERIMENTAL RESULTSA part of the experimental results for theparameters of wear are given in this paper in theform of graphs, tables and diagrams.605040302010wear, h [µm]0wear resistance, [m/µm]908070605040302010050403020100Figure 2. Variation of linear wear [µm] withthe friction path [m]Figure 3. Diagram of mass wear [mg] ofall specimens for two friction cycles100 300 500 700 900№ 0 - 0%Sn№ 1 - 0,018%Sn№ 2 - 0,02%Sn№ 3 - 0,032%Sn№ 4 - 0,051%SnFigure 4. Variation of wear rate [µm/min] withthe number of friction cycles43,532,521,510,500 23,2 69,6 116 162,4 208,8Road friction S, [m]38,548,12836,90% Sn 0,018%Sn 0,02%Sn 0,032%Sn 0,051%Sn0 0,018 0,02 0,032 0,051% Sn36,5N = 500 cl46,8N = 900 cl0% Sn0,018% Sn0,02% Sn0,032% Sn0,051% Sn35,537,531,5 31,3N = 500 clN = 900 clFigure 5. Variation of wear-resistance [m/µm] withthe Sn % contents for two friction cycles28 13 th International Conference on Tribology – Serbiatrib’13

Table 4: Comparative wear-resistance by using as reference sample high strength cast iron without Sn micro-alloyNumber of cyclesComparative wear-resistance, ε i,0Nε 1,0 ε 2,0 ε 3,0 ε 4,0N = 500 cl 1,38 1,06 1,23 1,23N = 900 cl 1,3 1,04 1,4 1,343,532,521,510,50Figure 6 Wear-resistance of cast-iron [m/µm] atvarious Sn % contents for friction cycles numberN=900 cl and N=500 clε = Ii/I01,41,210,80,60,40,200 % Sn 0,018% Sn 0,02% Sn 0,032% Sn 0,051% Sn0,018/0 0,02/0 0,032/0 0,051/0N=900 clN=500 clN=500 clN=900 clFigure 7. Diagram of the comparative wear-resistanceε i,0 = I i /I 0 by reference sample high strength cast ironwithout Sn micro-alloy for two friction cycles4. RESULTS ANALYSIS, OUTCOME ANDCONCLUSIONSThe above investigations confirm the authors’outcome of earlier studies, namely that microalloyingof high strength cast iron with Sninfluences its mechanical and tribologicalproperties [2,7].Increasing the Sn % contents leads to increase ofthe hardness of the high strength cast iron.The highest values of wear are for the specimenswithout Sn micro-alloy. All specimens containingSn show higher wear-resistance compared with castiron without Sn contents. A direct dependenceexists between the % contents of Sn and hardnessand wear-resistance of cast iron in the studied limitsof Sn contents. Deviation of this dependence isobserved for the specimen with 0,02% Sn contents.The same statement is to be seen in the earlierstudies of the authors.Maximum wear-resistance is obtained for0,032% Sn contents. At higher contents - 0,051%,the wear-resistance decreases. The wear-resistanceis equal for sliding path 500 cycles at 0,032% and0,051% contents, however the comparative wearresistancefor the cast iron with lower Sn contents(0,032%) is higher – Table 4. Although the authorshave no photos of the microstructure at this stage ofthe study, the last observation could be interpretedas result related to the non-homogeneousdistribution of the graphite phase in the structure ofthe specimen.Wear and wear-resistance are the parameters,which are most sensitive to the structure of materialand the time of wear process (the friction path). It ispossible that in some stages of the wearing processa structure of higher contents of the graphite phasesis available in the contact zone. The relationshipbetween wear and friction path under conditions ofabrasive wear is not linear function (Figs. 2 and 4).A period of running-in is observed, which is ofvarious duration for specimens with differentcontents of tin. The period of running-in will besubject of individual study.The obtained results are sign for the authors thatfuture systematic complex investigations on tin areneeded, including also comparative study with highstrength cast iron alloyed with copper.AcknowledgementThis study is connected with the completion of aPhD dissertation and of the tasks under theProject ДУНК-01/3 “University R&D Complex forinnovation and transfer of knowledge inmicro/nano-technologies and materials, energyefficiency and virtual engineering” funded by theBulgarian Ministry of Education and Science.REFERENCES[1] Я.Е. Гольдштейн, В.Г. Мизин, Модифицированиеи микролегирование чугуна и стали, М.,Металлургия,1986. Ya.E. Goldstein, W.G.Mizgin,Modifying and micro-alloying of cast iron and steel,Moscow, Metallurgy,1986 (in Russian)[2] П. Добрев, Ж. Калейчева, Б. Съйкова,Микролегиране на високояк чугун с калай, XIIIНационална конференция «Металолеене2007»,Ловеч, Машини, технологии и материали;P.Dobrev, J. Kaleicheva, B. Saikova, Micro-alloyingof high-stenght cast iron with tin, XIII Nat. Conf.13 th International Conference on Tribology – Serbiatrib’13 29

«Metal casting 2007», Machines, technologies andmaterials, Lovech, Bulgaria (in Bulgarian)[3] С. Бондаренко, И. Гладкий, Повышениеэксплоатационных свойств чугунов, работающихв условиях гидроабразивного износа, Вiсникхарнiвськ. державн. технiч. Университету,Харкiв, 2003, вып. 14, с. 388-391.S. Bondarenko, I. Gladkij, Improvement of theoperational properties of cast iron working underhydro-abrasive wear conditions, Kharkov, 2003,issue 14, pp. 388-391(in Russian)[4] В. Анчев, В. Тошков, Л. Василева, Ж.Захаридова, Ж. Калейчева, Й. Николов, Р.Петров, В. Симеонов,Ръководство залабораторни упражнения по материалознание,ИК „КИНГ”, 2001, рр. 112-113.V. Anchev, et al., Manual for laboratory works inmaterial science, Ed. house „KING” рр. 112-113 (inBulgarian)[5] БДС 14289-77, Метод за изпитване на абразивноизносване при триене върху закрепениабразивни частици, Bulgarian State Standard14289-77, Method for testing of abrasive wear atfriction on fixed abrasive particles (in Bulgarian)[6] М. Kandeva, D. Karastoyanov, A. Andonova: Wearand tribothermal effects of nanostructured nickelchemical coatings, Applied Mechanics andMaterials. Vols. 157-158 (2012), pp. 960-963.[7] B. Ivanova, M. Kandeva, J. Kaleicheva, R.Rangelov: Influence of Tin on the Structure andProperties of Spherographitic Cast Iron, 12thInternational Conference on Tribology, Kragujevac,Serbia, 11-13 May 2011, pp 42-45.30 13 th International Conference on Tribology – Serbiatrib’13

Serbian TribologySocietySERBIATRIB ‘1313 th International Conference onTribologyKragujevac, Serbia, 15 – 17 May 2013Faculty of Engineeringin KragujevacINFLUENCE OF NANO-DIAMOND PARTICLESON THE TRIBOLOGICAL CHARACTERISTICS OFNICKEL CHEMICAL COATINGSMara Kandeva 1 , Dimitar Karastoianov 2 , Boryana Ivanova 3 , Viara Pojidaeva 41 Tribology Center, Technical University-Sofia, Bulgaria, kandeva@tu-sofia.bg2 Institute of Information and Communication Technologies – Bulgarian Academy of Sciences, dkarast@clmi.bas.bg3 Technical University-Sofia, Bulgaria, bsaykova@tu-sofia.bg4 University of Mining and Geology „S t Ivan Rilsky” – Sofia, Bulgaria, vpojidaeva@abv.bgAbstract: Friction and wear of 10 types Ni chemical coatings, with and without heat treatment, containingnano-diamond particles of various size – 0;5 nm; 100 nm; 200 nm and 250 nm, are studied in the paper.Procedure and laboratory device for friction investigation in starting regime were developed. Experimentalresults for the influence of the particle size on the static friction force and the change of friction coefficienthave been obtained. Abrasive wear has been studied by means of the procedure developed by the authors forthe study of above coatings under conditions of dry friction on surfaces with fixed abrasive. The obtainedresults are related to the parameters linear wear, wear rate and wear-resistance. A part of this study isconnected with the tasks on the 7 FP Project „Acom In (Advanced Computing Innovations)” coordinated bythe Institute of Information and Communication Technologies at the Bulgarian Academy of Sciences, and theother part is carried out under the Project ДУНК-01/3 “University R&D Complex for innovation andtransfer of knowledge in micro/nano-technologies and materials, energy efficiency and virtual engineering”funded by the Bulgarian Ministry of Education and Science.Keywords: tribology, nano-diamond particles, coatings, friction, wear1. INTRODUCTIONNi chemical coatings are obtained through themethod of electro-less chemical deposition knownin the literature as „Electroless Niсkel”.From chemical point of view, chemicaldeposition is a deoxidization process whichdevelops between positive charged metal ions М z+and negative electrons е:z Me ze Me(1)where z is the valence of the metal ion.Coatings obtained through chemical depositiondiffer in the methods for procurement of theelectrons necessary for the deoxidization.In the galvanic (electrolyte) methods, electriccurrent is passed through the solution of the metalsalt (electrolyte) and the metal ions are reduced tothe corresponding metal atom Ме on the cathode(the coated detail). The cathode renders, and theanode obtains electrons, which are provided byexternal source – the electric current.At chemical Ni deposition an external source isnot needed for providing electrons. The necessaryelectrons are obtained as a result of chemicalreactions going between the solution and thesurface of the detail to be coated. As a consequencethe Ni metal ions of the solution obtain a givennumber of electrons depending on their valencepassing thus in state of neutral atoms (Ме). Theatoms gradually build the crystal grid of thecoating. In this case, the role of „supplier ofelectrons” is realized by different substances(chemical agents) called reducers (deoxidizers)from the solution [1].Imbedding of micro- or nano-sized particles ofvarious natures in the Ni matrix changes thephysico-mechanical and the tribologicalcharacteristics of the coatings.In connection with the improvement of theresource of tribosystems, a special interest fornanotribology represent Ni chemical coatings13 th International Conference on Tribology – Serbiatrib’13 31

containing in their structure particles of thenanosize scale [2,3]. Imbedding of nano-sizedparticles in the solution for the production of the Nicoating brings changes in the character of thecontact interactions on three levels: interaction ofnanoparticles with Ni ions into the solution with theelectrons, interaction of the built atom with thesurface of the detail and formation of the crystalgrid of the coating [4].The purpose of the present work is to studysome characteristics of contact friction and wear forNi chemical coatings, without and withnanodiamond particles of different size: 4 nm,100 nm, 200 nm and 250 nm.2. NICKEL CHEMICAL COATINGSTen types of coatings are studied, gathered in 5series with number given in Table 1.Table 1: Description of the specimens with coatings of chemical Ni containing nanodiamond particlesThickness of the№ ofseriesDesignation of theseries№ Designation of thecoatingCompositionof the coatingcoating before wear,h 1 , µmI N 1 N- Ni 25,562 N+ Ni+T o C 11,283 nD4- Ni+Di 4 nm 23,22IInD44 nD4+ Ni+Di 4 nm +T o C 85 nD100- Ni+Di 100 nm 27,94IIInD100 6 nD100+ Ni+Di 100 nm +T o C 7,127 nD200- Ni+Di 200 nm 26,24IVnD200 8 nD200+ Ni+Di 200 nm +T o C 9,149 nD250- Ni+Di 250 nm 30,5VnD250 10 nD250+ Ni+Di 250 nm +T o C 8,7Each series has its designation in Latin letters,correspondingly: N - Nickel coating without nanoparticles; nD - Nickel coating with diamondnanoparticles;The number after the letter D indicates theaverage size of the nanoparticles – 4 nm, 100 nm,200 nm, 250 nm. Each series includes two groupsof coatings: first group - coatings without heattreatment designated by the sign (-) and secondgroup - with heat treatment at 360 о С during 6 hoursdesignated by the sign (+).3. ABRASIVE WEAR3.1 Device and procedureExperimental study of abrasive wear of Nicoatings is realized by means of the test rig TABERABRASER according to the kinematical scheme„disk-on-disk” (Fig.1).The specimen 1 (the body) with depositedcoating 2 is in the shape of disk and is fixedappropriately on carrying horizontal disk 3 drivedby electrical motor 4 with a constant rotationalspeed =1[s -1 ]=const. The counter-body 5 is anabrasive disk (roller) of special material CS 10mounted on horizontal axis 6 in the device 8, bymeans of which is set the desired normal load Р inthe contact zone К. Thus, the body 1 and thecounter-body 5 are located on two crossed axes.Because of the constant rotational speed of thebody 1 and the constant nominal contact pressurep a , the friction in the contact zone К supportsconstant speed of rotation of the counter-body 5.The procedure of the experimental study onabrasive wear is realized in the following sequenceof operations:- clean-up, cleaning of lubricants and drying of theequal specimens. The specimens represent disks ofdiameter 100 mm and thickness 3 mm with thedeposited coatings;- measuring of roughness of the contact surfaces ofthe specimens before and after wear;- measuring of specimens mass m o before and itsmass m i after a given friction path L by electronicbalance WPS 180/C/2 of accuracy 0,1 mg. At everymeasurement the specimens are cleaned withappropriate solution against static electricity;32 13 th International Conference on Tribology – Serbiatrib’13

- measuring of coating thickness h 1 before wearand h 2 after wear by means of Pocket LEPTOSKOP2021 Fe device in 10 points of the surface; theaverage value is taken for thickness of the sample;- the specimen 1 is fixed on the carrying horizontaldisk 3; then the normal load Р is set. The frictionpath L is determined by the number of cycles readby the revolution counter 8.Figure 1. TABER ABRASER – device for study of abrasive wearThe procedure of the experimental study onabrasive wear is realized in the following sequenceof operations:- clean-up, cleaning of lubricants and drying of theequal specimens. The specimens represent disks ofdiameter 100 mm and thickness 3 mm with thedeposited coatings;- measuring of roughness of the contact surfaces ofthe specimens before and after wear;- measuring of specimens mass m o before and itsmass m i after a given friction path L by electronicbalance WPS 180/C/2 of accuracy 0,1 mg. At everymeasurement the specimens are cleaned withappropriate solution against static electricity;- measuring of coating thickness h 1 before wearand h 2 after wear by means of Pocket LEPTOSKOP2021 Fe device in 10 points of the surface; theaverage value is taken for thickness of the sample;- the specimen 1 is fixed on the carrying horizontaldisk 3; then the normal load Р is set. The frictionpath L is determined by the number of cycles readby the revolution counter 8.Abrasive wear for all coatings is obtained byfixed equal operating conditions – nominal contactpressure given with the normal load Р, averagesliding speed V and parameters of the abrasivesurface.The characteristics of the experiment are givenin Table 2.Table 2. Working parameters in the experiment:Apparent contact area Aa 0, 26 cm 2Nominal contact pressure pa 9, 42 N/cm 2Average sliding speed V 22,3 cm/sAbrasive material CS 10wearresistance and their change in time,respectively the friction path.Wear intensity is determined as mass (or linear)wear for unit friction path, and absolutewearresistance - as the reciprocate value of wearintensity.The relative wearresistance is the ratio betweenthe absolute wearresistance of the tested coatingand the absolute wearresistance of reference samplefor equal friction path (number of cycles).Two reference samples are used in the presentwork – Nickel coating without nanodiamondparticles with heat treatment and without heattreatment.3.2 Experimental resultsFigure 2. Dependence of wearresistance I h onnanoparticles size for coatings without heat treatmentThe parameters of mass and linear wear arestudied: speed, wear intensity, absolute and relativeFigure 3. Dependence of wearresistance I h onnanoparticles size for coatings with heat treatment13 th International Conference on Tribology – Serbiatrib’13 33

Figure 4.Wearresistance of Ni coatings without and withnanoparticles without and with heat treatment3.3 Analysis of the experimental resultsThe presence of nanodiamond particles affectsthe value and the character of the abrasive wear.This influence becomes more complicated alongwith the heat treatment of the coating.For size of nanoparticles 4 nm and 100 nmcoatings with heat treatment show higher wear, andfor size 200 and 250 nm the opposite effect isobserved – wear is lower than that of the casewithout heat treatment.The dependence of wear on nanodiamondparticles size is of nonlinear character, and thevarious coatings show different duration of therunning-in process.The boundary number of cycles N*, where thewhole coating is worn, is always bigger forcoatings without heat treatment.The highest wearresistance show Ni coatingswith nanodiamond particles of the size δ = 100 nmwithout heat treatment.4. STARTING CONTACT FRICTION4.1 TheoryFrom the point of view of process history, or intime cross-section of the process, tribosystemundergo three friction stages: starting, kinetic andpathological friction. The starting friction, knownas static friction in the classical mechanics, is doneunder conditions of preliminary microdisplacementin the contact zone and the tribosystem performsthe transition between static state (at rest) andmovement. Kinetic is the friction when the body ismoving upon the counterbody.The kinetic friction matches the stationary andthe pathological regimes of contact joint operation.The pathological friction is characterized byabrupt increase of friction with wear and seizure incontact.Figure 5. Variation of friction force and frictioncoefficient with displacementThe difference between starting friction force Т оand sliding friction force ТT To T(1)gives the jump in the friction force during systemtransition from state at rest and state of movement,and corresponds to the jump in the frictioncoefficient, i.e. o (2)The work of the starting friction force is givenby:A T T S PS (3)s o o o o oAnd the work of the kinetic friction force is:A T TS PSpLet present the ratio AsTo ApToPSo PS s .100 .100A T PSsooo(4)Then: s 100,%(5)oThe parameter s is called relative change ofthe starting friction; it is the ratio between the jumpof friction and the starting friction coefficient.Similarly, for the relative change of the kineticfriction p is obtained the expression: AsTo ApToPSo PS p .100 .100 .100A TPSorp p 100,%(6)4.2 Procedure and experimental resultsThe parameters of starting friction have beenstudied using a test rig with functional scheme asshown in Figure 5.34 13 th International Conference on Tribology – Serbiatrib’13

Figure 5. Functional scheme of the test rig for study ofstarting frictionThe experimental arrangement consists of body1 and counterbody 2, which form a contact. Thebody 1 is fixed in the holder 3 and is connectedthrough the nonelastic thread with thedynamometer 6 and micrometric screw 5.Tangential force is loaded on the body 1 near thecontact surface through slow turning of themicrometric screw. The normal load Р is set bymeans of the loading bodies 4.The body 1 is a prismatic sample of size30 х 50 х 8 mm made of duraluminium (Al), andthe counterbody 2 represents a round disk ofdiameter ф 100 mm and thickness 3 mm with thedeposited coating.The procedure of measurement the friction forceis of following sequence:- The specimen 2 with coating is fixed in the bedof the base, and the body 1 is mounted in the holder3, then they are put on the specimen 2.- The normal load is set by the loading bodies 4.- The elastic dynamometer 6 is put in the initialreset to zero.- The micrometric screw 5 is turned very slowlyand the pointer of the dynamometer 6 shifts withease. In the moment of shivering of the pointerbackwards, the indication of the dynamometer isread. The maximum value of the indicationcorresponds to the value of the starting frictionforce То.- The screw keeps on turning and the indicationsof the dynamometer are observed; they match thekinetic friction force Т after the jump of friction.- The dial of the dynamometer is calibrated inforce [N].- During the tests the body 1 is made of the samematerial but for each test with different coatings adifferent specimen of this material is used. Allspecimens of the body have equal size androughness Ra 0, 418 m.Table 3 shows the results of starting and kineticfriction, and the figures give some diagrams.353025201510500‐ 0+ 5‐ 5+ 100‐ 100+ 200‐ 200+ 250‐ 250+Figure 6. Diagram of starting friction force То0,10,090,080,070,060,050,040,030,020,01Figure 7. Diagram of the jump of friction force Δµ30252015105040353025201510500‐ 0+ 5‐ 5+ 100‐ 100+ 200‐ 200+ 250‐ 250+Figure 8. Diagram of the relative change ofstarting friction Ψs0‐ 0+ 5‐ 5+ 100‐ 100+ 200‐ 200+ 250‐ 250+Figure 9. Diagram of the relative change ofkinetic friction Ψр13 th International Conference on Tribology – Serbiatrib’13 35

Table 3. Experimental data of friction parameters№ Series То, [N] Т, [N] µо µ Δµ Ψs Ψр1 0- ( Ni Al ) 19,62 16,35 0,34 0,28 0,06 17,6 21,42 0+ ( Ni Al ) 21,8 18,53 0,38 0,32 0,057 15 17,83 5- ( Ni nD5 Al ) 15,26 10,90 0,266 0,19 0,076 28,6 404 5+ ( Ni nD5 Al ) 30,52 27,25 0,53 0,475 0,055 10,4 11,65 100- ( Ni nD100 Al ) 27,25 21,80 0,47 0,38 0,09 19,1 23,66 100+ ( Ni nD100 Al ) 21,8 19,62 0,38 0,342 0,038 10 11,17 200- ( Ni nD200 Al ) 20,71 17,44 0,36 0,30 0,06 16,7 208 200+ ( Ni nD200 Al ) 33,79 28,34 0,59 0,494 0,096 16,3 19,49 250- ( Ni nD250 Al ) 11,99 9,81 0,20 0,17 0,03 15 17,610 250+ ( Ni nD250 Al ) 19,62 16,35 0,342 0,285 0,057 16,7 204.3 Analysis of the experimental resultsA jump in the friction force is observed for alltested tribosystems however at different values ofstarting friction force and kinetic friction force.The jump is of different duration for thedifferent coatings.The influence of the size of nanodiamondparticles upon friction parameters is notunambiguous.The relationship between the starting frictionforce and the size of nanodiamond particles isstrongly nonlinear. This is most clearly expressedfor coatings with heat treatment. At coatingswithout heat treatment this relationship has clearmaximum for particles size 100 nm, howeverthe value of the maximum is lower than the twomaximums in the curve of the coatings without heattreatment.Genesis and variations in the friction forcesdepend directly on the formation and evolution ofthe contact spots, the latest depending on manyvarious factors too.REFERENCES[1] Gavrilov G., C. Nikolov: Electroless Nickel andComposite Coatings, Sofia, Technika, 1985.[2] Karaguiozova Z., T. Babul, A. Ciski, S. Stavrev:Influence of Cubic Nanostructured Additions on theProperties of Electroless Coatings, – IJNM, Vol. 5,No 1-2, pp. 129-138, 2010.[3] Kralov I., P. Sinapov, K. Nedelchev, I. Ignatov,Friction Induced Rail Vibrations, AIP, Vol. 1497,pp. 19-25, 2012.[4] Kandeva M., D. Karastoyanov and A. Andonova,Wear and tribothermal effects of nanostructurednickel chemical coatings, Applied Mechanics andMaterials, Vols. 157-158, pp. 960-963, 201236 13 th International Conference on Tribology – Serbiatrib’13

Serbian TribologySocietySERBIATRIB ‘1313 th International Conference onTribologyKragujevac, Serbia, 15 – 17 May 2013Faculty of Engineeringin KragujevacWEAR BEHAVIOR OF AUSTEMPERED DUCTILE IRONWITH NANOSIZED ADDITIVESJ. KaleichevaTechnical University of Sofia, Bulgaria, e-mail: jkaleich@tu-sofia.bgAbstract: The microstructure and properties of austempered ductile iron (ADI) strengthened withnanosized addtives of titanium nitride + titanium carbonitride (TiN + TiCN), titanium nitride TiN and cubicboron nitride cBN are investigated. The TiN, TiCN and cBN, nanosized particles are coated by electrolessnickel coating EFTTOM-NICKEL prior to the edition to the melt. The spheroidal graphite iron samples areundergoing an austempering, including heating at 900 о С for an hour, after that isothermal retention at 280о С, 2 h and 380 о С, 2h. The metallographic analysis by optical metallographic microscope GX41 OLIMPUSand hardness measurements by Vickers Method are performed. The structure of the austempered ductileiron consists of lower bainite and upper bainite.Experimental investigation of the wear by fixed abrasiveare also carried out. The influence of the nanosized additives on the microstructure, mechanical andtribological properties of the austempered ductile irons (ADI) is studied.Keywords: nanosized particles, austempered ductile iron, hardness, wear resistance1. INTRODUCTIONThe austempering of the iron-carbon alloys is anisothermal heat treatment, which reduces theinternal stresses and deformations and increases thedetails’ impact strength. The bainitic structure isformed at this type of heat treatment, which iswidely applicable in constructional steels andductile iron processing due to its high strength andincreased toughness [1-2]. Incompleteaustempering is also applied in case of heattreatment of some hypereutectoid and ledeburitesteels [3]. The possibility of wider practicalapplication of this heat treatment type requires anadditional data for the bainitic transformation iniron-carbon alloys with different compositionincluding alloys with nanomodifiers. Nanosizedparticles added to the iron melt in a small quantitytransform the graphite morphology from laminar tovermicular one [5], increase the graphite quantity[6] and change the matrix structure, whichincreases the cast iron wear resistance [4-6].The aim of the performed investigation is tostudy the tribological properties, the microstructureand hardness of austempered ductile iron,containing additives of nanosized particles –titanium nitride+titanium carbonitride (TiN +TiCN), titanium nitride TiN and cubic boron nitridecBN.2. MATERIAL AND INVESTIGATIONMETHODSThe composition of the austempered cast ironsamples is: Fe-3,55C-2,67Si-0,31Mn-0,009S-0,027P-0,040Cu-0,025Cr-0,08Ni-0,06Mg wt%. TheTiN, TiCN and cBN, nanosized particles are coatedby electroless nickel coating EFTTOM-NICKEL[7] prior to the edition to the melt. The nickelcoating improves the particles wetting into the meltand their uniformity distribution into the castingvolume.The ductile cast iron samples are undergoingaustempering, including heat treatment at 900 о С foran hour, after that isothermal retention at 280 о С, 2h and 380 о С, 2h.The austempered ductile iron samples’microstructure is observed by means of an opticalmetallographic microscope GX41 OLIMPUS. Thesamples surface is treated with 2 % HNO3 -C2H5OH solution. The hardness testing isperformed by Vickers method (Table 1).The experimental wear examination of the castand austempered ductile iron (ADI) is performed in13 th International Conference on Tribology – Serbiatrib’13 37

friction conditions of a fixed abrasive by acinematic scheme „pin - disc” using an acceleratedtesting method and device [6].Table 1. Nanoadditives, hardness and wear resistance.№ofsample1MicrostructureNanosizedadditiveHardnessHV10WearresistanceI- 314 5,9.10 62 upper TiN + TiCN 319 7,75.10 63 bainite TiN 317 6,13.10 64 cBN 312 7.10 65- 388 7,8.10 66 lower TiN + TiCN 413 5,46.10 67 bainite TiN 405 7,34.10 68 cBN 422 6,79.10 63. EXPERIMENTAL RESULTSThe cast iron structure consists of upper bainiteafter austempering at 380 о С for 2 hours and oflower bainite after austempering at 280 о С for 2hours (Figure 1). The bainite is an orientedneedlelike grain structure of α-phase (bainiticferrite), carbides and untransformed austenite. Theα-phase is formed in the low carbon austenite areaby a martensitic mechanism [1,2]. Upon coolingfrom the temperature of the isotherm to ambientone, a part of the untransformed austeniteundergoes martensitic transformation and other itspart remains as a retained austenite A in thestructure.Figure 1. Lower bainitic (c) and upper bainitic (a,b,d)microstructure. (a- sample1; b- sample 2; c, d- sample 8)The nanosized additives change the bainiticferrite morphology and the austenitic conversiondegree during the austempering (Figure1). Thehardness of the austempered, with upper bainiticstructure samples changes from 312 to 319 HV10(Figure 2a) and this one of the samples with lowerbainitic structure changes from 388 to 422 HV10(Figure 3a). The austempered samples hardnesswith lower bainitic structure is higher than this oneof the samples with upper bainitic structure, whichis explained with the different carbon satiety of theα-phase (bainitic ferrite) and with the varyingdegree of austenitic transformation in the lower andupper part of the bainitic area [1,2].The experimental data for massive wear m , thespeed of wear dm / dt , absolute intensity of wear iand absolute wear resistance I of the samples andtheir alteration with the time of the contactinteraction (Table 2, 3) are received. The massivewear m dependence on cycle’s number N (frictionroad) and massive wear speed dm / dt dependenceon the friction time t are presented in Figures 4 and5. Figure 2b and 3b show the wear resistance I ofaustempered ductile cast iron samples with upperand lower bainitic structure for the same frictionroad L 700 [m].38 13 th International Conference on Tribology – Serbiatrib’13

Table 2. Test results for massive wear, wear speed, intensity of wear and wear resistance (samples 1÷4).Friction road, S [m]140 280 420 560 700Cycles number , N 500 1000 1500 2000 2500Massive wear,m [mg]Wear speed, dm/dt[mg / min]Intensity of wear, iWear resistance, ITime , t [min] 2,35 4,7 7,05 9,4 11,75sample 1 22,3 32,7 38,6 42,4 46,5sample 2 19 24,4 28 32,8 35,3sample 3 20 25 32,6 38,2 44,8sample 4 16,3 24 27,1 33,8 39,1sample 1 9,49 6,96 5,48 4,52 3,96sample 2 8,08 5,19 3,97 3,49 3,0sample 3 8,51 5,32 4,62 4,06 3,81sample 4 6,94 5,11 3,84 3,6 3,33sample 1 0,406.10 -6 0,298.10 -6 0,234.10 -6 0,194.10 -6 0,169.10 -6sample 2 0,346.10 -6 0,222.10 -6 0,17.10 -6 0,149.10 -6 0,129.10 -6sample 3 0,364.10 -6 0,228.10 -6 0,198.10 -6 0,174.10 -6 0,163.10 -6sample 4 0,297.10 -6 0,218.10 -6 0,164.10 -6 0,154.10 -6 0,142.10 -6sample 1 2,46.10 6 3,36.10 6 4,27.10 6 5,15.10 6 5,9.10 6sample 2 2,89.10 6 4,5.10 6 5,88.10 6 6,69.10 6 7,75.10 6sample 3 2,75.10 6 4,39.10 6 5,05.10 6 5,75.10 6 6,13.10 6sample 4 3,37.10 6 4,59.10 6 6,1.10 6 6,49.10 6 7.10 6Table 3.Test results for massive wear, wear speed, intensity of wear and wear resistance (samples 5÷8).Friction road, S [m]140 280 420 560 700Cycles number , N 500 1000 1500 2000 2500Massive wear,m [mg]Wear speed, dm/dt[mg / min]Intensity of wear, iWear resistance, ITime , t [min] 2,35 4,7 7,05 9,4 11,75sample 5 14,2 20,4 24,4 29,3 35,2sample 6 26,4 33,6 37,7 44,9 50,2sample 7 14,6 21,7 27,2 34,8 37,4sample 8 15,9 23,7 29,4 35 40,4sample 5 6,04 4,34 3,46 3,12 2,99sample 6 11,2 7,15 5,35 4,78 4,27sample 7 6,21 4,62 3,86 3,7 3,18sample 8 6,76 5,04 4,17 3,72 3,44sample 5 0,259.10 -6 0,186.10 -6 0,148.10 -6 0,133.10 -6 0,128.10 -6sample 6 0,48.10 -6 0,306.10 -6 0,229.10 -6 0,204.10 -6 0,183.10 -6sample 7 0,266.10 -6 0,198.10 -6 0,165.10 -6 0,158.10 -6 0,136.10 -6sample 8 0,29.10 -6 0,22.10 -6 0,178.10 -6 0,159.10 -6 0,147.10 -6sample 5 3,86.10 6 5,38.10 6 6,75.10 6 7,5.10 6 7,8.10 6sample 6 2,08.10 6 3,27.10 6 4,37.10 6 4,9.10 6 5,46.10 6sample 7 3,76.10 6 5,05.10 6 6,06.10 6 6,31.10 6 7,34.10 6sample 8 3,45.10 6 4,54.10 6 5,62.10 6 6,27.10 6 6,79.10 6The wear resistance is a multifactorial parameterand to make its prognosis using the standardmeasured properties (hardness etc.) could bewrong, since these features are not always reliablecriteria for the steels’and irons’’ wear resistanceevaluation. The metastable structures in the ironcarbonalloys as a martensite, bainite and retainedaustenite have higher resistance to abrasive wear incomparison to this one of the stable structures(ferrite, pearlite etc.). The intensive strengthening isgoing off during the wear process due to dynamicstrain ageing of martensite and partialtransformation of the metastable retained austenitein a strain-induced martensite [8].The samples hardness with upper bainiticstructure without and with nanoadditives are similar13 th International Conference on Tribology – Serbiatrib’13 39

in values (312÷319 HV10). The wear resistance I ofthese cast irons consisting nanosized additives is inthe range between 6,13.10 6 ÷ 7,75.10 6 and it is with4 to 32 % higher than this one of the cast ironsamples without nanoadditives (I = 5,9.10 6 ).Samples hardness with lower bainitic structure is388÷422 HV10. The wear resistance of the castiron samples with lower bainitic structure withoutnanoadditives (I = 7,8.10 6 ) is higher than this one ofthe cast iron samples with nanoadditives ( I =5,46.10 6 ÷7,34.10 6 ). The obtained results for thewear resistance values of the tested cast ironssamples are probably related to the characteristicsof structures upper bainite, lower bainite, retainedaustenite and martensite during abrasive wear. Thedifferent quantitative proportion between thestructural components in the samples with andwithout nanoadditives defines the degree ofstrengthening and the resistance during abrasivewear due to strain ageing of the carbon sated α-solid solution (martensite and bainitic ferrite) andalso to partially transformation of the retainedaustenite into deformation martensite.Figure 3. Hardness HV10 (a) and wear resistance I (b)of austempered ductile iron samples without (5) andwith (6,7,8) nanoadditives.Figure 4. Dependence of the massive wear m on thecycles number N (a) and of the wear speed dm/dt on thefriction time t (b) (samples 1÷4).Figure 2. Hardness HV10 (a) and wear resistance I (b)of austempered ductile iron samples without (1) andwith (2,3,4) nanoadditives.40 13 th International Conference on Tribology – Serbiatrib’13

knowledge transfer in the field ofmicro/nanotechnologies and materials, powereffectiveness and virtual engineering”, contractDUNK-01/3.Figure 5. Dependence of the massive wear m on thecycles number N (a) and of the wear speed dm/dt on thefriction time t (b) (samples 5÷8).4. CONCLUSIONThe microstructure, hardness HV10 andtribological properties of austempered ductile castiron samples without and with nanosized additivesof titanium nitride + titanium carbonitride (TiN +TiCN), titanium nitride TiN and cubic boron nitridecBN are investigated. The nanosized particleschange the bainitc ferrite morphology in theaustempered iron structure. In the cast iron with aupper bainitic structure the nanosized additivesincrease the wear resistance with 4÷32 % incomparison to this one of the irons withoutnanoadditives. The results for the wear resistance ofthe irons with lower bainitic structure show thehighest value (I = 7,8.10 6 ) for the cast iron withoutnanoadditives.ACKNOWLEDGEMENTThe presented investigations are carried out andfinanced under the Project “University Complex forresearch and development of innovations andREFERENCES[1] G.V. Kurdyumov, L.M. Utevskiy, R.I. Entin:Transformations in Iron and Steel, Nauka,Moscow,1977 (In Russian).[2] H.K.D.H.Bhadeshia: Bainiteinsteels, 2 nd edit.London, Inst.of Materials, Cambridge, 2001.[3] J.Kaleicheva: Structure and Properties of High-Speed Steels after Austempering. Int. J.Microstructure and Materials Properties 2, pp. 16-23, 2007.[4] J. Li, M. Chen, H. Gao, Y. Zhao, Structures andProperties of Cast Irons Reinforced by TraceAddition of Modified SiC Nanopowders. ChineseJournal of Chemical Physics 20, pp. 625 – 631,2007.[5] Y. Wang, Z. Pan, Z. Wang, X. Sun, L. Wang,Sliding wear behavior of Cr–Mo–Cu alloy castirons with and without nanoadditives, Wear 271,pp. 2953– 2962, 2011.[6] J.Kaleicheva, M. Kandeva, Z. Karaguiozova, V.Mishev: Wear behavior of ductile cast irons withnanoparticle additives. Proceedings of the 3nd Int.Conf. on Diagnosis and Prediction in MechanicalEngineering Systems DIPRE12, May 31-June 1 st2012, Galati, Romania, Paper 38.[7] G.Gavrilov, C. Nicolov: Electroless Nickel andComposite Coatings, Tehnika, Sofia, 1985 (InBulgarian).[8] А.Makarov: Wear resistance increase of Ironalloys at the expense of metastable and nanocrystalstructure formation, DSc thesis, Ural Departmentof the Mechanical Engineering Institute of theRussian Academy of Sciences, Chelyabinsk, 2009(In Russian).13 th International Conference on Tribology – Serbiatrib’13 41

Serbian TribologySocietySERBIATRIB ‘1313 th International Conference onTribologyKragujevac, Serbia, 15 – 17 May 2013Faculty of Engineeringin KragujevacNICKEL COMPOSITE COATINGS MODIFIED BY DIAMONDNANOPARTICLESM. Kandeva 1 , N. Gidikova 2 , R. Valov 2 , V. Petkov 21 Technical University-Sofia, Centre of Tribology, Sofia, Bulgaria, kandeva@tu-sofia.bg2 Institute of Metal Science, Equipment and Technologies with Hydroaerodynamics Centre, Sofia, Bulgaria,radoslav.valov@gmail.com, vladimir2pe@yahoo.comAbstract: The study deals with composite Ni electro-chemical coatings on steel containing diamondnanoparticles with grain size up to 100 nm. Coatings were obtained at various concentrations of thenanoparticles in the electrolyte and at various process time.A procedure is developed for the study of wear parameters of the coatings under conditions of dry frictionwith abrasive surface. Experimental results for linear wear, wear rate, wear intensity and wear-resistancehave been obtained. The study is related and financed under the Technical University - Sofia ContractDUNK-01/3 “University R&D Complex for innovation and transfer of knowledge in micro/nanotechnologiesand materials, energy efficiency and virtual engineering” funded by the Bulgarian Ministry of Educationand Science.Keywords: tribology, wear resistance, abrasive wear, Ni-coatings, diamond nanoparticles1. INTRODUCTIONThe deposition of nickel coating on steel isapplied in industry to increase the corrosionresistance and the wear resistance of the surfaces[1].There is a lot of work done in the recent yearson the application of nanoparticles of differentmaterials and with various concentrations. Thisleads to improvement of the mechanical andtribological characteristics [2] [3].The objective of the present work is to study theinfluence of the diamond nanoparticles concentration onthe wear parameters of electrochemical nickel coatings.The wear is performed in dry friction conditions onsurface with firmly embedded abrasive particles.2. MATERIALS AND METHODSElectrochemical nickel coatings deposited onsubstrate of carbon steel C45 are studied.The electrochemical nickel depositing isperformed on cylindrical samples with diameter 30mm and height 10 mm. Standard electrolyte withcomposition: NiSO 4 .7H 2 O - 240 g/l,Na 2 SO 4 .10H 2 O – 150 g/l, NaCl - 15 g/l, H 3 BO 3 -20 g/l, is used. The pH of the solution is 5.0 - 5.5.The anode is made of nickel. The temperature ofgalvanization is 25 - 30 0 C.The diamond nanoparticles (ND) are producedby detonation synthesis and the grain size is up to100 nm. The nanoparticles are added in theelectrolyte as water suspension. The nickeldeposition is carried out with concentrations of thediamond nanoparticles 1, 5, 10 and 20 g/l at 3А/dm 2 current density. The time of theelectrochemical nickel deposition is 10 and 15 min.The galvanization process is performed after theactivation of the nanoparticles in the electrolyte andat continuous vigorous stirring during the nickeldeposition. The studied parameters are the Ni yield,the thickness of the layer, the microhardness andespecially the wear resistance of the coating. Theirchanges related to the parameters of thegalvanization as current density, depositionduration and the concentration of the diamondnanoparticles (C NDDS ) are studied.The data concerning the studied samples ofelectrochemical nickel coatings are presented intable 1. The coatings are obtained at differentdiamond nanoparticles concentration.42 13 th International Conference on Tribology – Serbiatrib’13

Table 1. Parameters of the electrochemical nickel coatingsSample No Coating Nanoparticlesconcentration, g/lCurrent density,А/dm 2Processduration, minYield of nickel,mg/cm 28 Ni - 3 10 4,879 Ni - 3 15 8,0617 Ni+nDi-1% 1% 3 10 11,1818 Ni+nDi-1% 1% 3 15 8,8928 Ni+nDi-5% 5% 3 15 8,3138 Ni+nDi-10% 10% 3 15 5,6648 Ni+nDi-20% 20% 3 15 7,053. DEVICE AND METHOD OFINVESTIGATION3.1 Device for investigation of the abrasive wearat dry friction on surface with fixed particlesThe experimental investigation is carried outaccording to the method and with the device foraccelerated test by the kinematic scheme “thumb –disc”. The device is presented schematically onfigure 1. The method is in conformity with theexisting standards [4].Figure 1. Functional scheme ofthe device “thumb – disc”The studied cylindrical sample 3 (body) with thedeposited coating K is installed immovably insuitable holder of the loading head 6. It ispositioned in such a way that its front surfacecontacts the abrasive surface 2 of the horizontaldisc 1 (counter body). The horizontal disc 1 rotateswith constant angular velocity ω = const around itsvertical axis. The number of cycles of disc 1 ismeasured with a cyclometer 5.The device permits change in the relative slidingspeed between the sample 3 and the disc 1 in twoways: by changing the angular disc velocity bycontrol bloc and by changing the distance Rbetween the rotation axis of the counter body 1 andthe axis of the sample 3.The abrasive surface 2 of the counter body 1 ismolded by surfaces of impregnated carbocorundum with hardness at least 60 % more thanthat of the tested coatings which corresponds to therequirements of the standard [4].2.2 Method for investigation of the abrasivewearThe method for investigation is performed in thefollowing sequence of operations:• Cleaning, degreasing and drying ofcylindrical samples with equal dimensionsand roughness;• Measurement of the mass of the samplebefore m 0 and after m i covering definitefriction distance S (number of cycles). Themeasurement is done with electronicbalance WPS 180/C/2 with precision 0.1mg. The samples are dipped in specialsolution to prevent accumulation of staticelectricity. The mass wear m is determinedas the difference of the two measurements;• Measurement of the thickness of thecoating before h 0 and after h i the wear atdefinite friction distance S with the devicePocket LEPTOSKOP 2021 Fe at 10 pointsof the surface and calculating the averageof the measured values;• The normal loading P is applied along thesample axis by a lever system in theloading head and the cycle number N isread on the cyclometer which correspond tothe friction distance S.The abrasive wear of all coatings is fixed at oneand the same working conditions which arepresented on Table 2.Table 2. Experimental parametersParametersValuesNominal contact pressure, p a 1,46 [N/cm 2 ]Average sliding speed, V15,5 [cm/s]Nominal contact area, A a 7,065 [сm 2 ]Abrasive surface Corundum Р 32013 th International Conference on Tribology – Serbiatrib’13 43

3.3 Wear parametersThe following parameters of mass and linearwear are studied:- Absolute mass m (linear h) wear;- Average rate of mass dm/dt, [mg/min] (lineardh/dt, [µm/min]) wear;- Absolute intensity by mass wear im , [ mg / m]asper the formulae:im= m/S(1)- Absolute intensity by linear wear hcorrespondingly:ihi , [ m/m]µ ,= h/S(2)- The friction distance S is calculated by thecorresponding number of cycles N and the distanceR between the axis of rotation and the mass centerof the nominal contact site by the formulae:S = 2πRN(3)- Absolute wear resistance by mass mI , [ / ]m mg :Im= 1/ im= S / m(4)- Absolute wear resistance by linear wear I h ,[ m/m]µ :Ih= 1/ ih= S / h(5)- Comparative wear resistance εie, - dimensionlessvalue, representing the ratio between the absolutewear resistance of the tested sample I i and theabsolute wear resistance of a standard sample I o .ε = (6)ie , Ii/IoSample with electrochemical nickel coatingwithout diamond nanoparticles is accepted as astandard in the present study.The index i indicates the percentage of thediamond nanoparticles.3. EXPERIMENTAL RESULTSThe obtained experimental results of the massand linear wear, the rate of wear, the absolute andthe comparative wear resistance are presented in theform of graphical relations, tables and diagrams.wear, mg6050403020100100 200 300 500 600Cycles8-Ni9-Ni17-Ni+nDi,1%18-Ni+nDi,1%28-Ni+nDi,5%38-Ni+nDi,10%48-Ni+nDi,20%Figure 2. The relation of the mass wear m [mg] from thenumber N of the friction cycleswear rate mg/min454035302520151050100 200 300 500 600Cycles8-Ni9-Ni17-Ni+nDi,1%18-Ni+nDi,1%28-Ni+nDi,5%38-Ni+nDi,10%48-Ni+nDi,20%Figure 3. The relation of the rate of mass wear from thewear distance10864203,532,521,510,505.346.558.544.125.074.48 9 17 18 28 38 48N=600 cl2.67Figure 4. Linear wear in [µm] at N = 600 cl.for each sample1,92,323,031,468 9 17 18 28 38 48N=600 clFigure 5. Rate of linear wear in [µm/min] at N = 600 cl.for each sample1,81,60,9544 13 th International Conference on Tribology – Serbiatrib’13

Figure 6. Wear resistance by linear wear I h , [µm/m] atN = 600 cl. for each sampleFigure 7. Diagram of the comparative wear resistanceε i,0 = I i /I 0Table 3. Comparative wear resistance of the samples compared to the standard – electrochemical nickel coating withoutdiamond nanoparticles.Number of cyclesNComparative wear resistance, ε i,eε 1,0 ε 5,0 ε 10,0 ε 20,0N = 300 cl 1,22 0,76 1,60 3,68N = 600 cl 1,27 1,05 1,21 2,004. ANALYSIS OF THE RESULTSIt is found that the nickel yield decreases with theincrease of the diamond nanoparticles concentration inthe electrolyte. The Ni yield acquires its highest valuesof about 12.0 mg/cm 2 at diamond nanoparticlesconcentration C NDDS = 1g/l, current density I = 3 A/dm 2and process duration t = 10 min.It is found also that the microhardness of the coatingat concentration of the diamond nanoparticles C NDDS =1g/l is 4800 МPa or 2.5 times more than that of purenickel coating (1950 MPa) and 1.7 times more than thatof coating at concentration of the diamond nanoparticlesC NDDS = 5 g/l (2800МPа).The increase of the abrasive wear with the frictiondistance has nonlinear character and is different for thedifferent coatings. This relation is linear only for coatingderived from electrolyte with diamond nanoparticlesconcentration 5 %.The abrasive wear rate is not constant value in time.The only exception is coating obtained from electrolytewith diamond nanoparticles concentration 10 %.The presence of diamond nanoparticles in theelectrochemical nickel coatings leads to increase of thewear resistance. The wear resistance is increased withthe increase of the diamond nanoparticles content in theelectrolyte and the relation is of nonlinear character.Coating with 20 % content of diamond nanoparticlespossesses the highest wear resistance – 52,1.10 -6 at N =600 cycles. This wear resistance is 2 times higher thanthe wear resistance of coating without nanoparticles.Out of the studied coatings the coating containing 1% nanoparticles and deposited with current density 3A/dm 2 and process duration 10 minutes possesses thelowest wear resistance.Comparing the wear resistance of the coatings obtainedat equal diamond nanoparticles concentration 1 % andcurrent density 3 A/dm 2 , the coating obtained at processduration 15 minutes has 2 times higher wear resistance.REFERENCES[1] I. Petrov, P. Detkov, A. Drovosekov, M.V. Ivanov,T. Tyler, O. Shenderova, N.P.Voznekova, Y.P.Toporov, D. Schulz: Nickel galvanic coatings codepositedwith fractions of detonationnanodiamond, Diamond and related Materials 15,pp. 2035-2038, 2006.[2] M. Kandeva, Il. Peichev, N. Kostova, K. Stoichkov,Complex Study of Surface Layers and Coatings, Journalof the Balkan Tribological Association 17, 2011.[3] M. Kandeva, D. Karastoyanov and A. Andonova,Wear and tribotermal effects of nanostructurednickel chemical coatings, Applied Mechanics andMaterials Vols. 157-158, pp 960-963, 2012.[4] БДС 14289-77, Меthod of abrasion test by frictionagainst embedded abragant grain13 th International Conference on Tribology – Serbiatrib’13 45

Serbian TribologySocietySERBIATRIB ‘1313 th International Conference onTribologyKragujevac, Serbia, 15 – 17 May 2013Faculty of Engineeringin KragujevacTRIBOLOGICAL BEHAVIOR OF THERMAL SPRAY COATINGS,DEPOSITED BY HVOF AND APS TECHNIQUES, ANDCOMPOSITE ELECTRODEPOSITS NI/SIC AT BOTH ROOMTEMPERATURE AND 300°C.A.Lanzutti 1 , M. Lekka 1 , E. Marin 1 , L.Fedrizzi 11 Università di Udine, Dipartimento di Scienze e Tecnologie Chimiche, Via del Cotonificio 108, 33100 Udine Italy,alex.lanzutti@uniud.itAbstract: Both the thermal spray and the electroplating coatings are widely used because of their high wearresistance combined with good corrosion resistance. In particular the addition of both micro particles ornano-particles to the electrodeposited coatings could lead to an increase of the mechanical properties,caused by the change of the coating microsctructure.The thermal spray coatings were deposited following industrial standards procedures, while the Ni/SiCcomposite coatings were produced at laboratory scale using both micro- and nano-sized ceramic particles.All the produced coatings were characterized regarding their microstructure, mechanical properties and thewear resistance. The tribological properties were analyzed using a tribometer under ball on diskconfiguration at both room temperature and 300 o C.The results showed that the cermet thermal spray coatings have a high wear resistance, while the Ni nanocompositeshowed good anti wear properties compared to the harder ceramic/cermet coatings deposited bythermal spray technique.Keywords: thermal spray coatings, nano-composite electrodeposits, Ni/SiC, micro-compositeelectrodeposits, HVOF, APS, dry sliding, 300°C).1. INTRODUCTIONThe thermal spray coatings are widely used formany industrial applications [1-13] because of thepossibility to deposit different type of materials,ranging from different metal alloys to ceramics, andtheir technological properties, in particular the highwear resistance even if they are used also ascorrosion barriers at both high temperaturedegradation or wet corrosion.The thermal spray coatings are mainly used forhigh temperature applications (oxidation resistanceor fused salts resistance). Usually these types ofcoatings are deposited with the addition of rareearths in order to inhibit the oxidative degradationprocesses [1-3]. Some technological processes areused to reduce the porosity of the coating and thusincrease both mechanical properties and the barriereffect to oxidative environments. Sidhu et al [2]have found that the laser remelting process increaseboth the mechanical properties and the oxidationresistance and leaves only a small amount ofporosity to the coating (

coatings and observed their good wear resistance,related to their high hardness, but low corrosionresistance, in wet wear tests. This behaviour isrelated to the low toughness that promote the crackenucleations and thus the permeation of corrosionmedia that enhances the undermining corrosion ofthe coating. Toma et al [8] showed that the additionof Cr to metal matrix increased the abrasion andcorrosion resistance. Fedrizzi et al [9] showed thatthese type of coatings can be used as an alternativeto hard chromium and their performance are alsoincreased if are used nano-sized powders. Murthyet al[11] showed that the coatings grinded have anhigher wear resistance because of the production ofan hard layer on surface. Other researcher [12]found out that the decrease in powder size increasethe coating performance because of both reductionof the porosity and increase of the mechanicalproperties. This enhances both corrosion and wearsprotection.The ceramic coating Cr 2 O 3 was subjected tonumerous studies in the area of both wear andcorrosion protection. It was shown by Ahn et althat, in the reciprocating wear tests, the wearmechanism is a plastic deformation of wear debristhat influence both the friction and wear behavior.This is related to a formation of CrO 2 layer underhertzian loads [4]. Bolelli et al [5] performed weartests at room temperature on different plasma sprayceramic coatings. He found that the Cr 2 O 3 coatingis the hardest and most anisotropic coating withhigh abrasion resistance, as confirmed by Leivo etal [6], while in sliding condition the material formsa compact tribofilm. The high temperature data forthis type of coatings is scarce.The Ni based electrodeposits are widely used asboth corrosion/wear barrier coatings in manyapplications ranging from high temperatureapplications to room temperature applications inboth dry and wet conditions [14-21].In this work composite coatings are alsoanalyzed . The introduction of reinforcing particlesis aimed to enhances both the wear and corrosionproperties, in particular if the nano-particles areembedded to the metal matrix they produce a nanostructuredmicrostructure.Garcia et al [14] showed that the increase ofwear properties at room temperature for microcompositeNi/SiC coatings is a function of particle’size. In particular the decrease of particle’ sizeleads to an increase of anti-wear properties.The reaserch group of Zimmermann et al [15-16] observed that the addition of sub-microsizedparticles to the coating leads to an increase of bothmechanical strength and toughness, if thereinforcing content is below the 2 wt%. Above the2wt% they showed that some particles’ coalescenceis possible, during the deposition, leading to tcoating embrittlement. They tried to add nanoparticlesto the coating and observed a notableincrease of the mechanical properties.Benea et al [17-18] in many works demonstratethat the addition of SiC nano particles, in Ni matrix,leads to an increase of the wear properties and theycalculated the relation between the microstructureof the coating and the wear performance.The wear properties of both micro compositeand nano composite Ni/SiC coatings at hightemperature were investigated by Lekka et al [19].They found an increase of anti-wear properties ofthe composite coating at both room temperature testand 300°C compared to the pure Ni coating.This work aimed to compare the wearperformance of the most used thermal spraycoatings with the anti-wear properties of the Nicomposite coating, highlighting the importantproperties of the composite coatings produced witha simple and cheaper technique compared to thethermal spray process.2. EXPERIMENTAL2.1 Samples productionFor all types of the deposits ASTM 387 F22steel plates (7×10 cm) and discs (d=5 cm) havebeen used as substrates (chemical composition inTable 1).Table 1: Chemical composition of steel substrate ASTM387 F22C Si Mn P Cr Mo Fe0.11 0.31 0.5 0.025 2.2 0.9 Bal.The thermal spray coatings have been depositedusing industrial procedures. The deposited coatingswere: NiCr 80/20 and NiCr 80/20 + Cr 2 O 3deposited by APS (Air Plasma Spray) techniqueand WC CoCr 18/4 deposited by HVOF technique.Regarding the Ni matrix coatings, three types ofdeposits have been prepared: pure Ni (to be used asreference). Ni containing microparticles of SiC andNi containing nanoparticles of SiC. Theelectroplating bath used was a high speed nickelsulfammate plating bath having the followingcomposition: 500 g/l Ni(SO 3NH 2 ) 2 . 4H 2 O, 20 g/lNiCl 2 . 6H 2 O, 25 g/l H 3 BO 3 , 1 ml/l surfactant (CH 3(CH) 11 OSO 3 Na based industrial product. Thedeposition was carried out using a galvanic pilotplant (12 l plating tank) under galvanostatic controlat 4 A/dm 2 , 50 °C, under continuous mechanicalstirring. The deposition time was 2.5h in order toobtain 70–80 μm thick deposits. For the production13 th International Conference on Tribology – Serbiatrib’13 47

of the composite coatings 20g/l of micro- or nanopowderswere added into the electroplating bath,dispersed using ultrasounds (200W, 24kHz) for30min and then maintained in suspension undercontinuous mechanical stirring during theelectrodeposition. The micro-particles have a meandimension of 2μm and a very irregular and sharpshape, while the nano-particles have a meandiameter of 45 nm [19].2.2 Samples characterizationThe specimens characterization includesmicrostructure, chemical composition,microhardness, wear resistance at both roomtemperature and 300°C and corrosion resistance intwo different environments.The microstructure of the specimens have beenanalysed by SEM (Zeiss Evo-40) + EDXS (Oxfordinstruments INCA) in cross section. Both the SiCcontent and the coatings’ porosity were calculatedusing an image analysis software [13]. For nanocomposite coating The SiC content was measuredthrough the measurements of RF GDOES (HR-Profile, Horiba Jobin Yvon), calibrated using 28CRM (Certified Reference Material) samples. Thesystem was set up using an Ar pressure of 650Paand a applied power of 50W. The micro-compositecoating were not analysed by the GDOES becauseof some issues related to the plasma erosion of thereinforcing particles [21].Micro-hardness measurements (HV 0,3 ) havebeen performed on cross section of the specimens.Wear tests have been performed using a CETRUMT tribometer in a ball-on-disc configuration atboth room temperature and at 300 °C. The testingparameters are summarized in Table 2. The volumeloss has been evaluated using a stylus profilometer(DEKTAK 150 Veeco). The wear rate K [10 −6mm 3 /Nm] has been calculated using the equationdescribed in [22].Table 2: Wear test parameters.Counter materialApplied loadTest radiusRotation speedSliding speedTest durationWC sphere (d 9.5mm)70N18mm300rpm0.565 m/s60 minFig.1: Microstructure of gr. 22 steel.The Gr 22 steel presents a ferritic microstructurewith some carbides precipitated in the metal matrix,that leads to the high creep strength of material.The carbides are mainly produced by Cr and Mo.The hardness of the material is about 180±20 HV 0,3and the ferritic grain size is about 45±15 µm.In Fig.2 are shown the SEM micrographsobtained for Thermal spray coatings and therelative data acquired by mechanicalcharacterization and image analysis. In Tab. 3 thethermal spray coatings’ properties are listed.a)b)3. EXPERIMENTAL RESULTS3.1 Microstructural characterizationIn Fig. 1 is shown the microstructure of the steelsubstrate.48 13 th International Conference on Tribology – Serbiatrib’13

c)a)Figure 2: SEM images and microstructuralcharacterization of thermal spray coatings: a) NiCr80/20, b) WC CoCr 18/4 and c) NiCr 80/20+ Cr 2 O 3 .b)Table 3: Results of thermal spray coatings’characterization.Coating Thickness[µm]Porosityvol.%HardnessHV 0,3NiCr 98±16 6.5 359±18WCCoCr 105±15 3.45 1027±21NiCr+Cr 2 O 3 (38+187) ±255.5+10.1 (341+1118)±24As can be observed, the three types of thermalspray coatings present different thickness andporosity. The porosity, acquired by image analysis,is higher for the coatings deposited by APStechnique compared to the HVOF deposits. Thisdifference could be related to both powders sizeand impact velocity, that is lower in APS techniquewith respect to HVOF. Indeed, the difference inkinetic energy of the molten powders, that is higherin HVOF technique, leads to a different density ondeposited coating. The hardness acquired isassociated to the material deposited and the valuesacquired are similar to data available in scientificliterature for thermal spray coatings [1-13].The SEM micrographs obtained on cross sectionof Ni/SiC composite coatings previously etched(acetic acid: nitric acid 1:1) are shown in Fig.3. InTab. 4 are listed the electrodeposited coatings’properties.Table 4: Results of Ni/SiC electrodepositscharacterization.Coating Thickness[µm]SiC wt.%HardnessHV 0.3Ni 78±7 - 172±7Ni/µSiC 73±8 0.8 247±8Ni/nSiC 75 ± 5 0.15 270 ±9Figure 3: SEM images and microstructuralcharacterization of electrodeposited coatings:a) pure Ni, b) Ni/µSiC and c) Ni/nSiC.The microstructure of the electrodeposits iscolumnar. In the case of the pure Ni the metalcolumns are oriented along the direction ofelectrical fields. The addition of SiC micro-particlesleads to a slight modification of the Ni columnsorientation and size. On the other hand, thecodeposition of SiC nano-particles leads to theformation of a fine grained deposit in which the Nicolumns are not oriented. The addition of microparticlesleads to a microstructure columnar with aslight modification of orientation, caused probablyc)13 th International Conference on Tribology – Serbiatrib’13 49

y the deviation of electrical field in proximity ofceramic particles that are in non-conductivematerial. On the contrary, the addition ofnanoparticles gives a grain refinement and a multiorientationof columns. The SiC amount is higher inthe micro-composite coatings compared to thenano-composite one. The addition of particles leadsto a noticeable microhardness increase due to boththe presence of the particles and the grainrefinement.All the analysed samples showed a surfaceRoughness Ra of about 0.5µm, obtained after thesurface’ grinding.In particular the thermal spray coatings showedalso other degradation mechanisms which arerelated to their microstructure.Room temperature 300°CNiCr+Cr2O33.2 Tribological characterizationAll the wear tracks obtained for both bare steel,thermal spray and composite electrodeposits at bothroom temperature and 300°C are shown in Fig. 4-6.In Fig. 4 the top views of the wear tracksobtained for gr.22 steel are shown, tested at bothroom temperature and 300°C.WC CoCrRoom temperature 300°CNiCrFigure 4: SEM images of the wear tracks obtained forthe bare steel at both RT and 300°C.The steel is subjected to triboxidative wear atboth room temperature and 300°C. At roomtemperature the oxide produced is adherent to metalsubstrate and very homogeneus. At 300°C the oxideproduced is concentrated on the wear track’ sides.This phenomenon is related to the loss ofmechanical properties of the substrates that permitsthe countermaterial to destroy the oxide layer, thatis thus deposited on the sides of the wear track. At300°C are highlighted the debris of both oxide andsteel along the borders of wear track.In Fig. 5 the top views of the wear tracksobtained for thermal spray coatings are shown,tested at both room temperature and 300°C.For the metal matrix deposits the degradationmechanism is a triboxidation at both roomtemperature and 300°C, which intensity is varyingin function of the coating material.Figure 5: SEM images of the wear tracks obtained forthe thermal spray coatings at both RT and 300°C.The Ni/Cr showed a material detachmentoriginated by contact fatigue phenomenon,aggravated by its porosity. At high temperaturetests the detachment is decreased due to a possiblehardness decrease of the material which allowed thesealing of porosity, thus reducing the contactfatigue failure. The cermet coatings were allsubjected to triboxidation of metal matrix, moreintense at high temperature. For the ceramicmaterial (Cr 2 O 3 ) the degradation mechanism atroom temperature is similar to the NiCr coatingwhile at high temperature test the degradationbecomes more intense. This is caused by a phasechange of chromium oxide which enhances thewear ratio and reduces at the same time the frictioncoefficient [4].In Fig. 6 are shown the top views of the weartracks obtained for Ni/SiC electrodeposits tested atboth room temperature and 300°C.50 13 th International Conference on Tribology – Serbiatrib’13

Room temperature 300°CNi/µSiCNiNi/nSiCFigure 6: SEM images of the wear tracks obtained forthe Ni/SiC deposits at both RT and 300°C.The Ni electrodeposits showed, at roomtemperature, a triboxidation with a descaling ofoxide which forms a third body between thecounter material and the wear track, thus leading tothe formation of secondary tracks related toabrasive wear. At high temperature the coatingsshowed a strong triboxidation. By EDXS analysison pure Ni coatings, it was detected that during thewear test the steel substrate was reached. For thecomposite Ni/SiC coatings it seems that the oxideproduced is more adherent to steel substrate at hightemperature. Probably both the grain refinementand the presence of ceramic particles are linking theoxide to the metal matrix. It seems that somecounter-material was transferred to the coatingsurface, due to a slight adhesion phenomena.The wear rate of all tested coatings at both roomtemperature and 300°C are shown in Fig. 7.Figure 7: wear rates graph at both room temperature and300°C for all the coatings tested.All the coatings protect the steel substrate duringthe test, except for the pure Ni coating that showed,at 300°C, the highest wear rate. It is possible toobserve that the cermet coating has the lowest wearrate compared to the other coatings at both testtemperatures. This is related to the high amount ofWC which is bonded by a metal matrix that hashigh oxidation resistance at the test temperatures.For this coating the wear resistance is associated tothe carbide component and the particles binding isrelated to the metal matrix which has a hightoughness. On the other hand, the NiCr coatingshowed a lower wear rate at high temperature tests,compared to the room temperature one, and thiscould be related to both high oxidation resistance ofthe material, that contains a high amount of Cr, andto the reduction of material detachment thatconsequently reduces the abrasive phenomena. Theceramic coating (Cr 2 O 3 ) showed a high wear rate at300°C tests due to phase change of chromium oxideunder tribological contact.The electrodeposits showed good wearresistance at room temperature, higher for the nanocompositecoating. This reduction in wear ratecould be related to the grain refinement ofmicrostructure of the metal matrix, which increasesalso the mechanical properties of the coating. Thiseffect is not visible in the micro-composite coatingsbecause the reinforcement particles are usuallydetached from the metal matrix leading to intensiveabrasive wear caused by hard particle third bodycontact between counter material and surface of thespecimen. At high temperature the mechanicalbehaviour of the coating is reduced, probablybecause of the hardness decrease. In this case thepure Ni coating is completely removed while themicro composite coating showed a better wearresistance, compared to the pure Ni one, but thewear rate values were still higher than the thermalspray coatings wear rates. The higher wear13 th International Conference on Tribology – Serbiatrib’13 51

esistance of the nano-composite coating at 300°Cis probably related to the higher mechanicalproperties of the metal matrix compared to theother electrodeposits.In fig 8-9 the COF (Friction Coefficients of thetested materials) are shown.COF [-]COF [-] COF [-]RT test300°C testRT test300°C testRT test300°C testTime [s]Time [s]Time [s]Fig.8: COF graphs at both room temperature and 300°Cfor the thermal spray coatings: a) NiCr 80/20, b) Wc CoCr (18/4), c) NiCr 80/20+ Cr 2 O 3 .a)b)c)For all the test performed on thermal spraycoatings, it is possible to observe that the COFvalues, at the end of the test, are comparablebetween the tests performed at differenttemperature, apart the ceramic coating that showeda lower COF at high temperature due to the changephase of ceramic oxide under hertzian loads. TheNiCr coatings showed a noisy COF graph becauseof the coating material detachment that producedabrasive particles that dissipated more energy,required to move the particles in the hertziansystem. At high temperature there is a start at lowCOF and, at regime, it reached the same values ofthe test at room temperature. Probably during thestart of the test the surface of the sample wascovered by a oxide layer produced during the heatup of the system. The presence of oxide decreasedthe surface energy in proximity of the hertziancontact of the two materials, reducing the frictioncoefficient. When the oxide was broken the contactbetween the two materials was between the WCand the Ni/Cr slightly oxidized.For the WC-CoCr coating the COF are slightlydifferent and this is caused mainly by the number ofthird body particles produced during the test.Indeed is possible that at high temperature theamount of abrasive particles, that are taking part tothe hertzian system, are higher due to the intensetriboxidation that cause probably a high amount ofdescaled oxide. At the end of the test part of theparticles are evacuated from the wear trackreaching a COF value comparable with the roomtemperature test.The COF acquired from the test performed at300°C is lower compared to the value acquired atroom temperature test. This behaviour is related tothe change phase of ceramic oxide that decreasedthe contact energy and thus the COF. The frictioncoefficient values are higher at the start of the testbecause of possible partial fragmentation of thematerial caused by brittle contact between thecountermaterial and the coating. This leads to havean high amount of third body particles that increasethe COF value, at the beginning, that is decreasing,during the test, because of particle’ evacuation formthe wear track caused by the relative motion of thetwo materials.The COF acquired for the Ni/SiCelectrodeposits is lower compared to the pure Nielectrodeposit. This could be related to bothdifferent mechanical properties of the compositematerial respect to the pure Ni and possibleinteractions of SiC particles with countermaterialthat could lower the surface energy and interactionof the 2 surfaces.52 13 th International Conference on Tribology – Serbiatrib’13

COF [-] COF [-]Fig.9: COF graphs at both room temperature and 300°Cfor the Ni/SiC composite coatings: a) Room temperaturetest, b) 300°C.At high temperature the COF graphs are verysimilar and this behaviour could be related to thechange of contact, compared with roomtemperature test, that is between Ni oxide and theWC sphere, instead of Ni slightly oxidized andWC.4. CONCLUSIONSTime [s]Time [s]In this work different type of coatings have beenanalysed deposited either by thermal spraytechniques or by electrodeposition. The coatingsdeposited by thermal spray are: NiCr (APS), WCCoCr (HVOF) and NiCr+Cr 2 O 3 (APS) . Theelectrodeposits are Ni/SiC coatings with nano- ormicro-sized particles embedded in metal matrix.The analysed coatings showed differentmicrostructure that depends on both depositedmaterial and deposition technique.Regarding the wear properties, The steelsubstrate showed the worst wear resistance at bothroom temperature and 300°C. This behaviour isrelated to the low mechanical properties of thissteel, that are decreasing as the temperatureincreases.All the tested metal matrix coatings underwenttriboxidation, that was increased at hightemperature test. The triboxidation behaviourdepends on metal oxidation resistance. The ceramica)b)coating was subjected to an intensive materialdetachment, caused mainly by the highinterconnected porosity of the thermal sprayedcoating. The detachment increased in function oftemperature because the ceramic oxide changedphase under the hertzian loads. For all the metalmatrix coatings was present a third body abrasioncaused mainly by both oxide descaling and ceramicreinforcement detachment from the metal matrix.Observing the wear rates, the WC CoCr coatingshowed the highest wear resistance at both roomtemperature and 300°C. This behaviour is related tothe microstructure of the deposit: the reinforcingparticles (WC) give high hardness also at hightemperature and the metal matrix (CoCr) increasesthe toughness of the coating and acts as binder forthe reinforcing particles. The electrodepositsNi/nSiC showed a wear behaviour that iscomparable with the WC CoCr one. For the nanocompositeelectrodeposits the synergy of both grainrefinement and nano-particles embedding leads toan increase of hardness at both room temperatureand 300°C. This effect probably enhances the wearresistance of the Ni metal matrix that is subjected tohertzian loads.The COF values are strongly dependent on thematerial analysed but it was observed, for thermalspray coating, similar COF values between theroom temperature test and the 300°C tests. Theceramic coating showed the lowest COF values athigh temperature caused mainly by the productionof brittle CrO 2 phase. The electrodeposits showedsome differences in the COF values between thehigh temperature tests and the room temperaturetests caused mainly by the change of hertziancontact from Ni slightly oxidized, at roomtemperature, to Ni strongly oxidized, at 300°C. Atroom temperature is visible a different in COFvalue between pure metal and composite coatings.This effect is related to the different mechanicalproperties of the coating and the possibleinteraction of reinforcing particles with thecountermaterial in the hertzian contact/motion.REFERENCES[1] N.F. Ak, C. Tekmen, I. Ozdemir, H.S. Soykan, E.Celik, NiCr coatings on stainless steel by HVOFtechnique, Surface and coatings technology, Vol.173-174, pp 1070-1073, 2003.[2] B.S. Sidhu, D. Puri, S. Prakash, Mechanical andmetallurgical properties of plasma sprayed andlaser remelted Ni-20Cr and stellite-6 coatings,Journal of Materials Processing Technology, Vol.159, pp 347-355, 2005.[3] H. Singh, D. Puri, S. Prakash, Some studies on hotcorrosion performance of plasma sprayed coatings13 th International Conference on Tribology – Serbiatrib’13 53

on Fe-based superalloy, Surface & coatingstechnology, vol 192, pp 27-38, 2005.[4] H.S. Ahn, O.K. Kwon, Tribological behaviour ofplasma-sprayed chromium oxide coating, wear, vol.225-229, pp814-824, 1999.[5] G. Bolelli, V. Cannillo, L. Lusvaraghi, T.Manfredini, wear behaviour of thermally sprayedceramic oxide coatings, wear, vol. 261, pp 1298-1315, 2006.[6] E.M. Leivo, M.S. Vippola, P.P.A. Sorsa, P.M.J.Vuoristo, T.A. Mantyla, wear and corrosionproperties of plasma sprayed Al2O3 and Cr2O3coatings sealed by aluminum phosphates, Journal ofThermal spray technology, vol. 6, pp205-210, 1997.[7] L. Fedrizzi, L. Valentinelli, S. Rossi, S. Segna,Tribocorrosion behaviour of HVOF cermetcoatings, Corrosion Science, Vol. 49, pp 2781-2799,2007;[8] D. Toma, W. Brandl, G. Maringean, wear andcorrosion behavior of thermally sprayed cermetcoatings, Surface and coatings technology, vol.138,pp 149-158, 2001.[9] L.Fedrizzi, S. Rossi, R. Cristel, P.L. Bonora,Corrosion and wear behaviour of HVOF cermetcoatings used to replace hard chromium, vol. 49,Electrochimica Acta, pp 2803-2814, 2004;[10] P.L. Ko, M.F. Robertson, wear characteristics ofelectrolytic hard chrome and thermal sprayed WC-10Co-4Cr coatings sliding against Al-Ni bronze inair at 21°C and at -40°C, wear, vol 252, pp 880-893, 2002.[11] J.K.N. Murthy, D.S. Rao, B. Venkataraman, Effectof grinding on the erosion behavior of a WC-Co-Crcoating deposited by HVOF and detonation gunspray process, wear, vol 249, pp592-600, 2001.[12] C.W. Lee, J.H. Han, J. Yoon, MC Shin, S.I. Kwun,A study on powder mixing for high fracturetoughness and wear resistance of WC-CO-Crcoatings sprayed by HVOF, Surface and coatingstechnology, vol. 204, pp 2223-2229, 2010.[13] S. Deshpande, A. Kulkarni, S. Sampath, H. Herman,Application of image analysis for characterizationof porosity in thermal spray coatings andcorrelation with small angle neutron scattering.Surface and coatings technology, vol. 187, pp 6-16,2004.[14] I. Garsia, J. Fransaer, J.P. Celis, Role of contactfrequency on the wear rate of steel in discontinuoussliding contact conditions, Surface and Coatingstechnology, vol. 141, pp 171-178, 2001.[15] A.F. Zimmermann, G. Palumbo, K.T. Aust, U. Erb,Mechanical properties of Ni silicon carbidenanocomposites, Material Science and Engineering,Vol A328, pp 137-146, 2002.[16] A.F. Zimmermann, D.G. Clark, K.T. Aust, U. Erb,pulse electroideposition of Ni-SiC nanocomposite,Material Letters, Vol. 52, pp 85-90, 2002.[17] L. Benea, P.L. Bonora, A. Borello, S. Martelli, F.Wenger, P. Ponthiaux, J. Galland, Preprataion andinvestigation of nanostructured SiC-Ni layers byelectrodeposition, Solid State Ionics, Vol. 151 pp89-95, 2002;[18] L. Benea, P.L. Bonora, A. Borello, S. Martelli, wearcorrosion properties of nano-structured SiC-Nicomposite coatings obtained by electroplating,Wear, Vol. 249, pp 995-1003, 2002.[19] M.Lekka, A. Lanzutti, A. Casagrande, C. deLeitenburg, P.L. Bonora, L. Fedrizzi, Room andhigh temperature wear behaviour of Ni matrixmicro. Nano-SiC composite electrodeposits, surfaceand coatings technology, vol. 206, pp 3658-3665,2012;[20] M. Lekka, P.L. Bonora, A. Lanzutti, S. Benoni, P.Caoduro, L. Fedrizzi, Industrialization of NimicroSiCelectrodeposition on copper moulds forsteel continuos casting, metallurgia italiana, vol 6,pp 1-7, 2012;[21] A. Lanzutti, E. Marin, M. Lekka, P. Chapon, L.Fedrizzi, Rf-GDOES analysis of compositemetal/ceramic electroplated coatings with nano- tomicroceramic particles’ size: issues in plasmasputtering of Ni/micro-SiC coatings. SIA, vol. 44,pp 48-45, 2012.[22] K. Holmberg, A. Matthews, Coatings tribology:Properties, Mechanisms, Techniques andApplications in Surface engineering, Elsevier, UK,2009.54 13 th International Conference on Tribology – Serbiatrib’13

Serbian TribologySocietySERBIATRIB ‘1313 th International Conference onTribologyKragujevac, Serbia, 15 – 17 May 2013Faculty of Engineeringin KragujevacMECHANOCHEMICAL SYNTHESIS OF NANOSIZED MIXEDOXIDESN.G. Kostova 1 , M. Kandeva 2 , M. Fabian 3 , A. Eliyas 1 , P. Balaz 31 Institute of Catalysis, Bulgarian Academy of Sciences, 1113 Sofia, Bulgaria, nkostova@ic.bas.bg2 Technical University of Sofia, Sofia, Bulgaria, kandeva@tu-sofia.bg3 Institute of Geotechnics, Slovak Academy of Sciences, 043 53 Kosice, Slovakia, balaz@saske.skAbstract: Fe 2 O 3 -ZnO nanosized mixed oxides samples were successfully synthesized using a simplemechanochemical method. The composites were characterized by X-ray diffraction (XRD), Fouriertransform infrared, and UV-visible diffuse reflectance spectroscopies The pattern of XRD shows broadeningin the diffraction peaks, indicating a decrease in the particle size of the samples with milling time.Keywords: nanotribology, mechanochemistry, ball-milling, X-ray diffraction1. INTRODUCTIONThe mixed oxides find application catalysts andsupport for catalysts, batteries, magnetic materialsand gas sensors [1, 2]. The mixed metals oxides areused in the chemical and pharmaceutical industry.The catalytic activity of mixed oxide systems isusually higher than that of the individual oxidecomponents. The research work is aimed atachieving the maximal efficiency. In order toachieve this aim two strategies are applied ingeneral: modification of the method of preparationand addition of dopants. The mixed oxides areusually obtained in the form of powders and theserepresent a substantial part of the industrialcatalysts due to their low price, easiness ofregeneration and their selective action. They areprepared by the sol-gel method [3], solid statereaction method [4], co-precipitation [5], citric acidmethod [6], solution-combustion method [7-8],thermal decomposition method [9],mechanochemical processing [10-12], gas-phaseflame synthesis and aerogel method [13].Mechanochemical processing is a method forproduction of nanosized materials [11]. Its mainadvantage is the option to synthesize nanopowdersat low temperatures by a simple one-step procedureof milling. Mechanochemistry is generallyperformed in high-energy ball mills using powderreactant mixtures. In this work we report on Fe2O3-Zno mixed oxides formation during milling in aplanetary ball mill.2. EXPERIMENTALMechanochemical synthesis of Fe 2 O 3 -ZnOmixed oxides was performed in a laboratoryplanetary mill Pulverisette 6 (Fritsch, Germany) byhigh-energy milling of hematite and ZnO. Thefollowing experimental conditions were applied forthe mechanochemical synthesis: loading of the mill,50 balls of 10 mm in diameter; material of millingchamber and balls was tungsten carbide; volume ofmilling chamber, 250 ml; room temperature;rotational speed of the mill planet carrier 400 min –1 ;milling time, 20 min.X-ray powder diffraction patterns (XRD) of thesamples were registered at room temperature with aTUR M62 apparatus with PC management and dataaccumulation, using HZG-4 goniometer with CoK radiation. The XRD lines were identified bycomparing the measured patterns to the JCPDS datacards.Specific surface area was determined by the lowtemperature nitrogen adsorption method in aGemini 2360 sorption apparatus (Micromeritics,USA).The phase evolution during high energy ballmilling was followed by Nicolet 6700 FTIRspectrometer (thermo Electron Corporation, USA).13 th International Conference on Tribology – Serbiatrib’13 55

The method of dilution of the studied sample inKBr at concentration 0.5 % was used.The diffuse reflectance UV–vis spectra weretaken with a Thermo Evolution 300 UV-VisSpectrophotometer equipped with a Praying Mantisdevice with Spectralon as the reference. Spectralonis a fluoropolymer, which has the highest diffusereflectance of any known material or coating overthe ultraviolet, visible, and near-infrared regions ofthe spectrum.As milling proceeds broadening of diffraction peaksis observed due to grain size reduction.3. RESULTS AND DISCUSSIONTextural properties of initial hematite andrutile and mechanochemically synthesized mixedoxides are presented in Table 1.Table 1. Samples composition and textural propertiesSamplecodeChemicalcomposition [%]Fe 2 O 3 ZnOSpecificsurface area,m 2 g -1PorevolumeZnO 0.0 100.0 5.5 0.004ZnO-MA 0.0 100.0 3.0 0.0020.3 FZ 0.3 99.7 5.3 0.0044.3 FZ 4.3 95.7 5.4 0.00414.3 FZn 14.3 85.7 5.3 0.00466 FZ 66.0 33.0 6.6 0.005The XRD patterns of the Fe 2 O 3 -ZnO mixedoxides with different iron content are presented inFig. 1.66 FZ14.3 FZ4.3 FZFigure 2. FTIR spectra of the Fe 2 O 3 -ZnO mixed oxidesFigure 2 shows the IFTR spectra themechanically activated Fe 2 O 3 -ZnO mixed oxides.All the samples show prominent absorption bandand shoulder at about 440 cm - 1 and 550 cm -1 ,respectively. Slight variation with increasing of ironcontent in the samples is due to grain size influencethe band position. The band at lower frequency of440 cm-1 corresponds to M-O stretching vibrationin octahedral site. This shift can be result of highercontent of oxygen vacancies present in the structureand created during mechanical activation [14].The diffuse reflectance of the ZnO and Fe 2 O 3 -ZnO mixed oxides are shown in Fig. 3. The spectraof the samples in UV region exhibit an absorptionpeak at about 220-250 nm attributable to isolated,tetrahedral coordinated species.0.3 FZZnO MA****ooo * o*o * oZnO+Fe 2O 310 20 30 40 50, degreeFigure 1. XRD patterns of the Fe 2 O 3 -ZnO sample. Themark (*) indicated the peaks corresponding to Fe 2 O 3 , andthe marks (o) – peaks corresponding to ZnO.The XRD diagram marks as ZnO+Fe 2 O 3corresponds to the starting mixture with molar ratioZnO:Fe 2 O 3 = 1:1 before milling. All the diffractionpeaks of mechanical activated zinc oxide and 0.3FZ sample can be well indexed to the hexagonalphase ZnO (JCPDS card no. 36-1451. A decrease inthe reflection intensity was observed during milling.Figure 3. UV-vis diffuse reflectance spectra of ZnO andFe 2 O 3 -ZnO mixed oxidesThe samples with larger iron content showintense absorption in wide wavelength range fromUV to visible light with absorption tail extendinginto infrared region. This means that Fe componentwas physically connected to the external surface ofZnO structure. The peak with maximum centred at56 13 th International Conference on Tribology – Serbiatrib’13

about 260 nm is related to isolated Fe 3+ cations,while the second band at about 350 nm correspondsto small oligonuclear (FeO) n species. At higher ironcontent the peak around 520 nm is observedindicating the presence of iron oxide particles.This peak can be assigned to symmetrical andspin forbidden d-d transitions of Fe 3+. [15]. Whenthe sample contains small iron content (sample 0.3FZ), the band slightly shifts to the right without tailbroadening. This means that the Fe component waswell inserted inter the framework of the ZnOstructure.In a ball mill intense mixing takes place andreactants are brought into intimate contact with eachother. Grinding reduces the particle size andincreases the surface area available for reaction. Newreactive surfaces are exposed during particle fractureand the introduction of dislocations increases thesurface reactivity. At the point of contact betweentwo grinding balls during a collision event, a highlylocalised triboplasma is formed giving energy forchemical reactions to occur.ACKNOWLEDGEMENTThis investigation was financially supported bythe National Science Fund of Bulgaria (projectDNTS/Slovakia 01/3), by the Slovak Research andDevelopment Agency (project APVV-0189-10),and through a bilateral project between BulgarianAcademy of Sciences and Slovak Academy ofSciences.REFERENCES[1] I.E. Wachs, K. Routray: Catalysis science of bulkmixed oxides ACS Catalysis Vol. 2, No. 6, pp. 1235-1246, 2012.[2] M.V. Reddy, G.V. Subba, B.V.R. Chowdari: Metaloxides and oxisalds as anode materials for Li ionbatteries, Chem. Rev. DOI: 10.1021/cr 3001884,2013.[3] B. Palanisamy, C.M. Babu, B. Sundaravel, S.Anandan, V. Murugesan: Sol-gel synthesis of mixedFe 2 O 3 /TiO 2 photocatalyst: Application fordegradation of 4-chlorophenol, J. Hazard. Mater.,Vol. 252-253, pp. 233-242, 2013.[4] S. Martha, K.H. Reddy, N. Biswal, K. Parida: Facilesynthesis of InGaZn mixed oxide nanorods forenhanced hydrogen production under visible light,Dalton Trans. Vol. 41, pp. 14107-14116, 2012.[5] K-T. Li, L-D. Tsai, C-H. Wu, I. Wang: Lactic acidesterification on titania-silica binary oxides, Ind.Eng. Chem. Res. Vol. 52, No. 13, pp. 4734-4739,2013.[6] H. Zhou, P. Hu, Z. Huang, F. Qin, W. Shen, H. Xu:Preparation of NiCe mixed oxides for catalyticdecomposition of N 2 O, Ind. Eng. Chem. Res. Vol.52, No. 12, pp. 4504-4509, 2013.[7] G.K. Pradhan, S.Martha, K.M. Parida: Synthesis ofmultifunctional nanostructured zinc-iron mixedoxide photocatalysts by a simple solutioncombustiontechnique, Appl. Mater. Interface, Vol.4, No. 2, pp. 707-713, 2012.[8] G. Lauramoy, R. Carbonio, E.L. Moyano: Mixedoxides as highly selective catalysts for the flashpyrolysis of phenacyl benzotriazole: One-potsynthesis of dibenzazepin-7-one, ACS Catal DOI:10.1021/cs3008335, April 3, 2013.[9] S.M.A. Shibli, J.N. Sebeelamol: Development ofFe 2 O 3 -TiO 2 mixed oxide incorporated Ni-P coatingfor electrocatalytic hydrogen evolution reaction,Intern. J. Hydrogen Energy, Vol. 38, pp. 2271-2282,2012.[10] P. Balaz, M. Achimovicova, M. Balaz, P. Billik, Z.Cherkezova-Zheleva, J.M. Criado, F. Delogu, E.Dutkova, E. Gaffet, F.J. Gotor, R. Kumar, I. Mitov,T. Rojac, M. Senna, A. Streletskii, K. Wieczorek-Ciurowa: Hallmarks of mechanochemistry:fromnanoparticles to technology, Chem. Soc. Rev., DOI:10.1039/c3cs35468g, pp. 1-67, 2013.[11] P. Balaz: Mechanochemistry in Nanoscience andMineral Engineering, Springer-Verlag, BerlinHeidelberg, 2008.[12] A.M. Al-Saie, A. Al-Shater, S. Arekat, A. Jaffar, M.Bououdina: Effect of annealing of the structure andmagnetic properties of mechanically milled TiO 2 -Fe 2 O 3 mixture, Ceramics International, Vo. 39, pp.3803-3808, 2013.[13] K.M. Shrestha, C.M. Sorensen, K.J. Klabunde:MgO-TiO 2 mixed oxide nanoparticles: Comparisonof flame synthesis versus aerogel method;characterization and photocatalytic activities,Journal of Materials Research Vol. 28, No, 3, pp.431-439, 2013.[14] M. Fabian, A. Eliyas, N.G. Kostova, J. Brianchn, P.Balaz: Photocatalytic activity of nanocrystallineghanite (ZnAl 2 O 4 ) synthesized by ball milling, Microand Nano Technology, Proceedings of InternationalMultidisciplinary scientific geoconference SGEM2012, 17-23 June 2012, Albena, BulgariaBulgaria,Vol. III, pp. 391-497, ISSN: 1314-1314-2704.[15] Y. Wang, Q. Zhang, T. Shishido, T. Takehira:Characterization of iron-containing MCM-41 andits catalytic properties, J. Catal. Vol. 209, pp. 186-196, 2002.13 th International Conference on Tribology – Serbiatrib’13 57

Serbian TribologySocietySERBIATRIB ‘1313 th International Conference onTribologyKragujevac, Serbia, 15 – 17 May 2013Faculty of Engineeringin KragujevacWEAR OF POLISHED STEEL SURFACES IN DRY FRICTIONLINEAR CONTACT ON POLIMER COMPOSITES WITH GLASSFIBRESDorin Rus 1 Lucian Capitanu 21 Mechanical Departament, Institute of Civil Engineering, 59 Plevnei Way, 050141, Bucharest, Romania, e-dorinrus@gmail.com2* Institute of Solid Mechanics of the Romanian Academy, 15 Constantin Mille, 010141, Bucharest, Romanialuciancapitanu@yahoo.comAbstract: It is generally known that the friction and wear between polymers and polished steel surfaces has aspecial character, the behaviour to friction and wear of a certain polymer might not be valid for a differentpolymer, moreover in dry friction conditions. In this paper, we study the reaction to wear of certain polymerswith short glass fibres on different steel surfaces, considering the linear friction contact, observing thefriction influence over the metallic surfaces wear. The paper includes also its analysis over the steel’s wearfrom different points of view: the reinforcement content influence and tribological parameters (load, contactpressure, sliding speed, contact temperature, etc.). Thus, we present our findings related to the fact that theabrasive component of the friction force is more significant than the adhesive component, which generally isspecific to the polymers’ friction. Our detections also state that, in the case of the polyamide with 30% glassfibres, the steel surface linear wear rate order are of 10 -4 mm/h, respectively the order of volumetric wearrate is of 10 -6 cm 3 /h. The resulting volumetric wear coefficients are of the order (10 -11 – 10 -12 ) cm 3 /cm andrespectively linear wear coefficients of 10 -9 mm/cm.Keywords: wear, composite thermoplastics, comparative wearing coefficient.1. INTRODUCTIONThe tribological behaviour of polymers hasdistinctive characteristics, some of them beingdescribed by Bowden and Tabor [1]. The mainconcept related to the polymers’ tribology iscomposed of three basic elements involved infriction: (i) junctions adhesion, their type andresistance; (ii) materials’ shearing and fracturethrough friction during the contact; and (iii) the realcontact area.Friction’s straining component results from thepolymer’s resistance to “ploughing” made by theasperities existing on the harder counter-face. Thepolymer’s surface asperities bear elastic, plastic andviscous-elastic strains, according to the material’sproperties. Friction adhesion component comes outof the adhesion junctions formed on the real contactspots between the paired surfaces. Friction adhesioncomponent in what the polymers are concerned isconsidered to be much greater than the strainingcomponent. Special attention should be granted tothe transfer films, these transfer films being the keyfactors determining the tribological behaviour ofpolymers and polymeric composites. In what theglass fibers reinforced polymer is concerned, wealso encounter a strong abrasive component [2].Several models were developed to describe thecontact adhesion. The Johnson-Kendall-Roberts(JKR) model, mentioned sometimes as the contactmechanics model [3-4] and the Derjaguin-Muller-Toporov (DMT) model [5] are the best known. Themodels’ comparative analysis [6] shows that theJKR model is applied to bodies with micrometricdimensions and larger than that, with polymerproperties, whilst the DMT model is valid forbodies with nanometer dimensions, with metalproperties.Several authors [7-17] studied the polymers’friction on hard surfaces. By using the method of58 13 th International Conference on Tribology – Serbiatrib’13

contact’s conformity [18] they obtain the hardness,the deformability value (index) (which describesthe coarse surfaces’ deformation properties), aswell as the elasticity module for organic polymerspolymethylmethacrylate – PMMA; polystyrene –PS; polycarbonate – PC, ultra high molecularweight polyethylene – UHMWPE. We alsodescribe the dependence of the imposed penetrationdepth, the maximum load and the straining speed,the hardness and the elastic modulus [18-22]. Thetypical penetrating depths are included within theapproximate 10 nm to 10 μm range, whilst theapplied loads are smaller than 300 mN.We can observe the fact that almost withoutexception, the ploughing is accompanied byadhesion and in certain conditions it may lead tomicro-cutting, which represents a supplementaryadding to increase the friction force.There are other mechanisms to dissipate theenergy while straining. For instance, whenever apolymer with viscous-elastic reaction slides on ahard surface, the energy dissipation is caused by thehigh losses through hysteresis. This strainingcomponent is known under the name of friction dueto elastic hysteresis [1]. The energy can, as well, betransported further, for instance through elasticwaves generated at the interface and coming out atinfinit, as, a nucleation and micro-cracksdevelopment within the material, consequence [20].The mechanical component consists in theresistance of the softer material to harder asperities’ploughing. The adhesion component comes of theadhesion links formed between the surfaces duringthe friction contact. We believe that for polymersthe adhesion molecular component exceeds by farthe mechanical one [20], and we can explain itthrough the generated films’ transfer on the metalcounter-face. The following factors considerablyaffect the friction force: the contact load, slidingspeed and temperature. The effects are notindependent. For instance, according to the contactload and contact speed, the temperature mayconsiderably vary, changing the friction mode [21].2. MATERIALS AND METHODSIn order to study the metallic counter-part’swear in dry contact with glass fibres reinforcedplastic materials we use Timken type frictioncouples (with linear contact), cylinder on plan,which allows us to attain high contact pressures,hence high contact temperatures. In this manner wenotice, whether and in which conditions the plasticmaterial transfer on the metallic surface appears, aswell as the influence of the glass fibres fillingduring this phenomenon, and its effect on thesurface’s wear. As we do not follow the polymer’swear, but only the polymer’s friction influence,over the samples’ metallic surfaces wear, we usethe unidirectional sliding movement.We perform the tests using experimentalequipment containing a Timken type linear contactfriction couple, continuously controlling the normaland friction loads, and contact temperature. Theunidirectional movement and the linear contactallow us to attain very high contact pressures andtemperatures. We build the friction couple out of aplastic cylinder Nylonplast AVE polyamide + 30%glass fibres, which rotates at different speedsagainst the polished surface of a steel plan disk.The cylinder has an outer diameter of 22.5 mm and10 mm height.We choose as sample steel disks with 18.2 mmdiameter and 3 mm thickness. We polish the disks’surfaces successively using sandpaper of differentgranulations (200, 400, 600 and 800) and, finally,we polish them on the felt with diamond paste. Weobtain mirror polished surfaces, with roughness R aof 0.05 µm. This metal surface’s quality allows usto eliminate the influence of the metallic surface’sstate on the friction coefficient’s evolution andvisualization, to make measurements using opticalmicroscopy and to accurately record the wear tracesappeared on the metallic surfaces.Fig.1 shows the friction couple (a) and itsinstallation within the experimental equipment (b).(a)Figure 1. Friction couple (a) and its installation in theexperimental equipment (b), where 1 - cylindrical liner;2 – steel disk sample; 3 – nut; 4 – hole; 5 - knife-edge.The friction couple is build out of a cylindricalliner (1) and a plane disk type sample (2). The lineris fixed with the help of a nut (3) on the drivingshaft (4), and the disk sample is placed in a specialhole made within the elastic blade (5). We build thesample disk base in such a manner so that the baseallows the sample to make small rotations aroundthe edge of a knife fixed in the sample’s bezel,perpendicularly on the driving arbour. In this waywe ensure a uniform repartition of the load on theentire linear contact between the liner and steelsample, even if there are small building orassembling imperfections. An electric engine putsthe shaft into a rotation movement using trapezoidaltransmission belts.The experimental device allows us tosimultaneously measure the normal and tangential(b)13 th International Conference on Tribology – Serbiatrib’13 59

(friction) efforts through resistive converter straingauges,assembled on the elastic blade (5). The useof a pair of converters strain-gauges connectedwithin the circuits of two strain-gauges bridges,offers us the possibility to make simultaneousmeasurements, while separately, gives us thepossibility to measure the normal and frictionforces. We apply the normal load to the elasticblade, through a calibrated spring system. Theinstallation allows us to register the friction forceon an X-Y recorder. We control the tests’ durationthrough an alarm clock and we measure the contacttemperature with the help of a miniaturethermocouple, connected to a millivoltmetercalibrated in 0 C.I used the uni-directional testing because thepurpose of investigations was the study of metallicsurface wear. We perform the tests, based onHooke's law, at normal loadings of 10; 20; 30; 40and 50 N, loadings which are adequate to somecontact pressures all calculated considering theelastic contact hypothesis, that is: 16.3; 23.5; 28.2;32.6 and 36.4 MPa (for Nylonplast AVE polyamidewith 30% glass fibres) respectively, we use slidingspeeds, adequate to the diameter of the plasticcomposite sample, which are: 0.1856; 0.2785;0.3713; 0.4641; 0.5570; 1.114 and 1.5357 m/s, andwhich resulted as a consequence of electric motor’sspeed and the belt pulleys’ primitive diameters.As we know [21], we may characterize amaterial’s wearing coefficient (percentage) bywearing factor k. Archard’s relation defines thisfactor:V u kNvt(1)where: V u – the wear’s material volume; N - the testload; v - the sliding speed; t - the test period; k –volumetric wearing factor.By dividing both of this relation’s terms (4) bynominal contact area A, we obtain:Which means that:V u/ A kNvt / A(2)h u* k pvt(3)where: h u - wear’s material depth; p - the pressureon the nominal contact area and k * is the linearwearing factor. Relation (6) expresses a general lawof the wear as a function of the contact pressure pand the length of the wearing path, so that L f vt.We could then write:respectively:k V / Nvt V / NL(4)uuf*k hu/ pvt hu/ pL(5)fConsidering the large area of the load (N) orpressure (p) and the relative speed values usedduring tests in order to evaluate the wearingreaction of the metallic counter-pieces amid thefrictional couples, we use comparative wearcoefficients K and K * , defined by:and:K V / L kN (cm 3 / cm) (6)K*uf* h / L k p (cm / cm) (7)ufWe consider these wearing coefficients withrespect to the period in which the frictional couplefunctions at different sliding speeds, under certainloading conditions (contact pressure).The main objectives of these tests are thedetermination of the volume of material removedby wearing, the mean depth of the wearied layers,the frictional factors and coefficients, for differentloading conditions.Coefficients k and k * are coefficients of the wearprocess, while the comparative factors K and K * arecoefficients of this process’s consequences, that is,the amount of resulted wear and reported to thelength of the friction pathway. They can bequalitatively expressed in units of wear volume ona measure of the length of the friction pathway (cm 3/ cm), as wear’s depth on a measure of the length ofthe friction pathway (cm / cm) or as wear’s weighton a measure of the length of the sliding frictionpathway (mg / cm). Coefficients K and K * have nomathematical implication (can not simplify).Using the procedure described in [22], at the endwe obtain the mean depth (8) and the volume ofworn metallic material (9):and:Vu2l/ 8r10,527NE1 E2LE1E2h (8)n i1Siqi0.351E1 E2Nlm/ E1E2(9)where l m is the mean width of the wear imprint.Practically, we have to measure the width ofwear imprints in three points established before,computing then the mean value of this width. Withthis value we can obtain the volume of wornmetallic material V u and the removed layer’s meandepth h mu .We study the wearing of the friction couple’smetallic component on linear contact Timkenmachinery, see Fig. 1. Almost all tests are madewithout lubricating the frictional surfaces, but thereare also tests with micro-lubricating.60 13 th International Conference on Tribology – Serbiatrib’13

In order to calculate the metallic component’swear, we use the method described above. Theequations (8) and (9) take into consideration, for thestudied materials, particular forms obtained byintroducing the interfering parameters numericalvalues, thus obtaining for a mean depth h mu and aworn material volume V u the following relations: Nylonplast AVE polyamide + 30% glass fibres/ steel:25h l 8r 6.94 10N (mm) (10)mum14V u 4.5510Nl m(mm 3 ) (11)The studies concerning the metallic semi-couplewear are generally based on the elastic contacthypothesis. For these plane half-couple the valuesfor the equivalent elasticity module for NylonplastAVE polyamide + 30% glass fibres, E = 20.25MPa. Assuming that the plastic liner does notcrush, we impose the condition p max 0.5H,where H stands for the Brinell hardness. Therequired condition allows us to establish thefollowing values of the maximum loadings (contactpressure) of the couple:p 1 = 16.3 MPa; p 2 = 23.5 MPa; p 3 = 28.2 MPa;p 4 = 32.6 MPa; p 5 = 36.4 Mpa.We perform the experimental tests consideringbroader domains to vary the relative speed andnormal loadings, or contact pressures. We usecouples with liner made from thermoplasticmaterial with linear contact on a steel surface(C120, Rp3, a.s.o.).Table 1. The results of the experimental tests performed inorder to determine the wear rate of metallic component.Frictional couple: Polyamide Nylonplast AVE +30% glassfibres / C120; ν = 18.56 cm/s.N (N)) t (hour)10 110 120 120 130 130 140 140 150 150 1Average wear rateh mu (10 -4 mm/h) V mu (10 -6 cm 3 /h)0.9649 0.13872.4798 0.44044.0336 0.83815.4874 1.30867,1635 1.8667(a)3. RESULTSTable 1 is the representation of the experimentaltests results, testing two friction couples, for one ofthe 8 different relative sliding speeds used. Table 1represents the results of the tribologicalexperimental tests, e.g. the mean values of the wearimprint depth h u (10 −4 mm), and the average valuesof the worn material volume V u (10 −6 cm 3 ). Theaverage width l m represents the arithmetical averagecalculated based upon 3 measured values of thewear trace’s width. By dividing h u and V u to theduration of experimental test, we obtain the valuesof the wear rate in terms of depth h mu (10 −4 mm/h)and volume V mu (10 −6 cm 3 /h).Based upon the methodology described above,we process the results obtaining the variationcurves of the wear with normal loading and relativespeed, presented in Fig. 2 (a) and (b), for two of thetested couples, Nyloplast AVE Polyamide + 30%glass fibres / C120 steel, and respectively NyloplastAVE Polyamide + 30% glass fibres / Rp 3 steel.(b)Figure 2. The results of variation curves of the wearvolume with normal loading and relative speed, fortested couples (a) Nyloplast AVE Polyamide + 30%glass fibres/ C120 steel and (b) Nyloplast AVEPolyamide + 30% glass fibres/ Rp 3 steel. Measurementerrors were ±1.5 %.These curves characterize only the testedfrictional couples (one combination of materials).Furthermore, we can make the comparativeevaluation of different couples only qualitatively.Thus, using relations (8) and (9) we obtain thevariation curves of the "comparative wearcoefficients" (as volume and depth), K (cm 3 / cm)and K * (mm / cm). These master-curves are plottedin Fig. 3 and Fig. 4 representing the two tested and13 th International Conference on Tribology – Serbiatrib’13 61

taken into discussion couples, for different normalloading values.While measuring the wear traces widths with thehelp of optical microscopy, we also takemicrophotographs, in order to identify the plasticK(10 -11 cm 3 /cm)1.6y = 1.587e NYLONPLAST AVE+30% glass/C120material’s transfer and the metallic surfaces’ wear-0.009xNYLONPLAST AVE+30% glass/Rp3y = 1.395emechanisms. These microphotographs prove that-0.009x1.4y = 1.138ethe wear mechanisms vary from one couple to-0.009x- N = 10 N1.2another, due to surfaces’ nature: metallic andy = 1.020e -0.010x- N = 20 N- N = 30 Ncomposite plastic material, especially their hardness1y = 0.8739e -0.009x- N = 40 N(59 HRC for C120 hardened steel and 62 HRC fory = 0.803e -0.011x0.8Rp3 hardened steel), the glass fibres content, 30%and 20%, the composite plastic materials’ elastoplasticcharacteristics while in contact with metallic0.6y = 0.664e -0.013x0.4surfaces. he glass-fibres torn from the polymermatrix.0.2y = 0.424e -0.019xv(cm/s)00 10 20 30 40 50 60 70 80 90 100 110 120 130Figure 3. The variation curves of the volumetriccomparative wear coefficients K (cm 3 / cm).NYLONPLAST AVE+30% glass/C120K*(10 -9 mm/cm)12NYLONPLAST AVE+30% glass/Rp310y = 12.608e -0.0253x(a)- N = 20 N8y = 8.4032e -0.0249x- N = 30 N- N = 40 N6y = 5.2364e -0.0253x4y = 8.8046e -0.022x2y = 6.4915e -0.0173xy = 5.4312e -0.0153xv(cm/s)00 10 20 30 40 50 60(b)Figure 5. Wear and plastic material transfer on C120 steelsurface, following the friction with Nylonplast AVE polyamidereinforced with 30% fine glass fibres (a), in experimentalconditions: v = 27,85 cm/s; N =20 N; T = 150 0 C; t = 60 minand (b) in experimental conditions v = 27,85 cm/s; N =30 N; T= 175 0 C; t = 60 min.Considering the same loading conditions, thetwo couples to which we make reference have adifferent behaviour. On C120 steel sample (Fig. 5),at a normal load of 20 N and a contact temperatureof 150 0 C, there are plastic material transferbridges, broadways on the wear traces (Fig. 5a), aswell as the glass-fibres torn from the polymermatrix. At 1750 C contact temperature,corresponding to a normal load of 30 N and acontact pressure of 2879.5 MPa, the plastic materialtransfer on the wear trace’s edge is obvious (Fig.5b), leaving the impression that the plastic matrixmelts and drips off on the wear trace’s exit edge.Considering the same mechanical stress40 K = 1.3950 e - 0,0090 v K * = 12,6080 e – 0,0253 v conditions (load and relative speed), theFigure 4 The variation curves of the linear comparative wearcoefficients K * (mm / cm).In Table 2 are listed the equations for thecomparative wear coefficients (the volumetric andthe depth ones), for C120 and in Table 3. for Rp3steel.Table 2. The variation curve of compatative wear corfficientequations for Nylonplast AVE Polyamide + 30% glassfibres/C120Load (N) K K *10 K = 0.8030 e - 0,0110 v20 K = 0,8739 e – 0,0090 v K * = 5,4312 e – 0,0153 v30 K = 1.1380 e - 0,0090 v K * = 6,4915 e – 0,0173 v40 K = 1.5870 e - 0,0090 v K * = 8,8046 e – 0,0200 vTable 3. The variation curve of compatative wear corfficientequations for Nylonplast AVE Polyamide + 30% glass fibres /Rp3Load (N) K K *10 K = 0.4240 e - 0,0190v20 K = 0,6640 e – 0,0130 v K * = 5,2346 e – 0,0253 v30 K = 1.0200 e - 0,0100 v K * = 8,4032 e – 0,0249 v62 13 th International Conference on Tribology – Serbiatrib’13

microscopic inspection of the Rp3 steel samples,while in friction contact, with the same compositeplastic material, reveals a less pronounced plasticmaterial transfer through adherence onto themetallic surface, visible on the left side in Figs 6 (a)and 6 (b), and if the test duration is double (120min), practically there is no plastic material transferas one can see in Fig 6 (c).(a)v= 27,85 cm/s; N =40 N; t = 60 min;T=217 0 C(b)v = 27,85 cm/s N =30 N; t = 120 min;T= 175 0 C(c)v = 27,85 cm/s; N =40 N; t = 120 min;T =237 0 CFigure 6. Wear and plastic material transfer on Rp3 steelsurface, following the friction with Nylonplast AVE polyamidereinforced with 30% short glass fibres.We consider that due to high registered contacttemperature (237 0 C) the transfer takes place forsure, but the transferred material is subsequentlyremoved through friction from the contact area,under the form of wear particles following the glassfibres abrasive action. After this stage, the abrasivewear due to glass fibres becomes predominant.It is possible that the less pronounced plasticmaterial transfer emphasized on the Rp3 steelsurfaces to be due to this steel’s chemicalcomposition and structure.We detect the same findings in the case of Norylpolyamide +20% glass fibres in friction on thesame steels, but to a lesser scale. In the case ofLexan 3412 polycarbonate reinforced with 20%glass fibres friction onto the same metallic surfacesand considering the same stress conditions,generally speaking there is no plastic materialtransfer. The transfer appears only if the loadreaches 40 N, which corresponds to a contactpressure of 3449.7 MPa, and when the contacttemperature reaches 251 0 C. We do consider thatprobably the polycarbonate has a lesser transfercapacity than the polyamide.4. DISCUSSIONThe wear’s rate values, considering the usedexperimental conditions, cover a large range. Forgreater clarity, they are presented in Table 4.Comparing the metallic element’s wear ratesvalues at v = 46.41 cm/s and N = 40 N, it resultsthat the polyamide reinforced with 30% glass fibresinduces to the C120 steel a wear of approximately1.110 times more higher than to the Rp3 steel. Wedo estimate that this phenomenon is due to Rp3samples’ higher hardness (62 HRC), in comparisonto those from C120 (59 HRC).Table 4. The variation curve of compatative wear corfficientequations for Nylonplast AVE Polyamide + 30% glassfibres/Rp3Friction coupleVolumetric wearrate ( 10 -6 cm 3 /h)Linear wear rate(10 -4 mm/h)v = (18.56 - 46.41) cm/s; N = 10 – 50 NPolyamide + 30% 0.139 – 1.621 0.965 – 8.549glass fibres/ C120Polyamide + 30% 0.214 – 1.369 2.382 – 6.004glass fibres / Rp 3Polycarbonate +20% glass fibres /C1200.244 – 1.309 3.592 – 6.366v = (46.41 - 111.4) cm/sPolyamide + 20% 0.440 – 2.578 3,269 – 6,794glass fibres / C120Polyamide + 20%glass fibres / Rp 30,473 – 2,549 3.792 – 6.627Normal loads and corresponding contactpressures for the linear friction contact used duringthis research, lead to very high contact temperatures(180-240 0 C) according to the applied normal loadand relative sliding speed (see also Fig. 6a).In several cases they exceed the polymer’smelting temperature, thus being transferred on themetallic surface together with glass fibresfragments. Part of the glass fibres is smashed andstill produced a predominant abrasive wear of themetallic sample’s contact area, while another part ispushed out on the contact’s exit edge, together witha multitude of ejected glass fibres.We notice that only in the case of the frictioncouple Nyloplast AVE Polyamide + 30% glassfibres / C120 steel, there is a large plastic materialtransfer onto the metallic surface, which justifiesthe assertion that the transfer through adhesiondepends on the nature and characteristics of thecontact materials. From a qualitative point of view,obviously there is the fact that initially the wearprocess manifests itself as a wear throughadherence and polymer transfer onto the metallicsurface, which subsequently transforms itself into aprocess of abrasive wear, which leads to the plasticmaterial removal clung onto the contact area. Inwhat the friction couples are concerned – also seeFig. 6a.The process’ intensity depends on the fibres’content. The larger it is, the higher the intensity is.Metallic surface mechanical properties (especiallythe hardness), has a distinct influence over theplastic material transfer and metallic surface wear.13 th International Conference on Tribology – Serbiatrib’13 63

5. CONCLUSIONThe diagrams’ analysis plotted in Figs. 3 and 4allows us to establish the variation equations for thecomparative volumetric wear coefficient K and forthe comparative depth wear coefficient K * , for steelin linear contact, while in friction with glassreinforced thermoplastics.The equations listed in Table 2 and 3, for thecomparative wear coefficients (the volumetric andthe depth ones), show that the variation is not alinear one, these coefficients evolvingexponentially. We also notice that the decrease ofthe K * coefficient with the increase of relative speedis faster than the decrease of the K coefficient.We consider that this effect is due to the factthat the thermoplastic material deforms under loadwhich means that for Timken type couples theincrease of the wear imprint width is more effectivethan that of the depth of the wear imprint. From thediagrams plotted here, one can notice that thevalues of wear coefficients for the metalliccomponent of the couple glass reinforcedthermoplastic/steel are in the domain (10 −11 ÷ 10 −12 )cm 3 /cm and respectively 10 −9 mm/cm. Thecomparative wearing coefficients and their mastercurvesvs. relative speed have a special importancefrom the practical point of view. Based on thesefindings we can establish an optimal couple ofmaterials from the design phase.ACKNOWLEDGEMENTSThe authors would like to thank the RomanianAcademy for its material and technical supportoffered in order to achieve these researches.REFERENCES[1] F.P. Bowden, D. Tabor. The Friction andLubrication of Solids, part I-II, Clarendon Press,Oxford,1964.[2] B. J. Briscoe, in: Friction and Wear of PolymerComposites, Ed. F. Klaus, Elsevier, Amsterdam,p. 25, 1986.[3] K.L. Johnson, K. Kendall and A. D. Roberts,Surface Energy and the Contact of Elastic Solids.Proc. Roy. Soc. A324, 301 1971.[4] K.L. Johnson, Contacts Mechanics, (CambridgeUniversity Press, Cambridge, 1987.[5] B. V. Deryagin, V. M. Muller and Yu. P. Toporov.Adhesive Contact Deformation of a SingleMicroelastic Sphere. J. Colloid Interface Sci., 53,314 1975.[6] K. L. Jonson and J. A. Greenwood. An AdhesionMap for the Contact of Elastic Spheres J. ColloidInterf. Sci., 192, 326, 1997.[7] D. Maugis, Adhesion of spheres: The JKR – DMTtransition using a Dugdale model. J. Colloid Interf.Sci., 150, 243 (1992).[8] I.V. Kragelskii, Friction and Wear (PergamonPress, Elmsford, 1982).[9] N.K. Myshkin, A.V. Kovalev. Adhesion andFriction of Polymers, in PolymerTribology. ImperialCollege Press, London, 2009.[10] V. E. Starzhynsky, A. M. Farberov, S. S. Pesetskii,S. A. Osipenko and V. A. Braginsky, PrecisionPlastic Parts and Thei Production Technology,(Nauka I Tekhnika, Minsk, 1992) (in Russian).[11] B. J. Briscoe, Wear of polymers: an essay onfundamental aspects. Tribology Int., 14, 231, 1998.[12] G. Jintang, Tribochemical effects in formation ofpolymer transfer film. Wear, 245, 100, 2000.[13] H. Unal, U. Sen and A. Mimaroglu, Friction andwear behavior of unfilled engineeringthermoplastics. 183–187. Tribol. Int., 37, 727 2004.[14] H. Unal and A. Mimaroglu, Sliding friction andwear behaviour of polytetrafluoroethylene and itscomposites under dry conditions. Mater. Des., 24(2003), 239–245Mater. Des., 24, 183 (2003).[15] Y.K. Chen, O.P. Modi, A.S. Mhay, A. Chrysanthou,J.M. O’Sullivan, The effect of different metalliccounterface materials and different surfacetreatments on the wear and friction of polyamide 66and its composite in rolling–sliding contact. Wear255, 714 (2003)[16] C.J. Schwartz and S. Bahadur: The role of fillerdeformability, filler–polymer bonding, andcounterface material on the tribological behavior ofpolyphenylene sulfide (PPS). Wear, 251, 1532 2001.[17] Y.M. Xu, B.G. Mellor: The effect of fillers on thewear resistance of thermoplastic polymericcoatings. Wear, 251, 1522 2001.[18] B.J. Briscoe, L. Fiori, E. Pelillo, Nano-indentationof polymeric Surfaces. J. Phys. D: Appl. Phys., 31,2395, 1998.[19] H. Shulga, A. Kovalev, N. Myshkin, V.V. Tsukruk,Some aspects of AFM nanomechanical probing ofsurface polymer films. European Polymer Journal,40, 949, 2004.[20] A. Kovalev, H. Shulga, M. Lemieux, N. Myshkin,V.V. Tsukruk, Nano-mechanical probing of layerednanoscale polymer films with atomic forcemicroscopy. J. Mater. Res., 19, 716, 2004.[21] G.M. Bartenev, V.V. Lavrentev, Friction and Wearof Polymers, (Elsevier, Amsterdam, 1981).[22] L. Capitanu, A. Iarovici, J. Onisoru – On polyamideand polycarbonate materials behaviour under dryfriction, The Annals of University “Dunarea de jos”Galati, fascicle VIII, Tribology, 2003, ISSN 1221-459064 13 th International Conference on Tribology – Serbiatrib’13

Serbian TribologySocietySERBIATRIB ‘1313 th International Conference onTribologyKragujevac, Serbia, 15 – 17 May 2013Faculty of Engineeringin KragujevacEXPERIMENTAL INVESTIGATION OF FRICTIONCOEFFICIENT AND WEAR RATE OF COMPOSITE MATERIALSSLIDING AGAINST SMOOTH AND ROUGH MILD STEELCOUNTERFACESMohammad Asaduzzaman Chowdhury 1 *, Dewan Muhammad Nuruzzaman 2 , Biplov Kumar Roy 1 ,Sohel Samad 1 , Rayhan Sarker 1 , Abul Hasnat Mohammad Rezwan 11 Department of Mechanical Engineering, Dhaka University of Engineering and Technology, Gazipur, Bangladesh,*asadzmn2003@yahoo.com2 Faculty of Manufacturing Engineering, University Malaysia Pahang, MalaysiaAbstract: In the present study, friction coefficient and wear rate of gear fiber reinforced plastic (gear fiber)and glass fiber reinforced plastic (glass fiber) sliding against mild steel are investigated experimentally. Inorder to do so, a pin on disc apparatus is designed and fabricated. Experiments are carried out when smoothor rough mild steel pin slides on gear fiber and glass fiber disc. Experiments are conducted at normal load 10,15 and 20 N, sliding velocity 1, 1.5 and 2 m/s and relative humidity 70%. Variations of friction coefficient withthe duration of rubbing at different normal loads and sliding velocities are investigated. Results show thatfriction coefficient is influenced by duration of rubbing, normal load and sliding velocity. In general, frictioncoefficient increases for a certain duration of rubbing and after that it remains constant for the rest of theexperimental time. The obtained results reveal that friction coefficient decreases with the increase in normalload for gear fiber and glass fiber mating with smooth or rough mild steel counterface. On the other hand, it isalso found that friction coefficient increases with the increase in sliding velocity for both of the tested materials.Moreover, wear rate increases with the increase in normal load and sliding velocity. The magnitudes of frictioncoefficient and wear rate are different depending on sliding velocity and normal load for both smooth andrough counterface pin materials.Keywords: Friction coefficient, wear rate, gear fiber, glass fiber, mild steel, normal load, sliding velocity.1. INTRODUCTIONNumerous investigations have been carried outon friction and wear of different materials underdifferent operating conditions. Several researchers[1-6] observed that the friction force and wear ratedepend on roughness of the rubbing surfaces,relative motion, type of material, temperature,normal force, relative humidity, vibration, etc. Theparameters that dictate the tribological performanceof polymer and its composites include polymermolecular structure, processing and treatment,properties, viscoelastic behavior, surface texture,etc. [7-10]. There have been also a number ofinvestigations exploring the influence of testconditions, contact geometry and environment onthe friction and wear behavior of polymers andcomposites. [11-13] reported that the tribologicalbehavior of polyamide, high density polyethyleneand their composites is greatly affected by normalload, sliding speed and temperature. [14-15]showed that applied load and sliding speed playsignificant role on the wear behavior of polymerand composites. They also showed that applied loadhas more effect on the wear than the speed forcomposites. Experiments were carried out onfriction and wear behavior of poly-ether-imide andits composites under different operating conditions[16-19]. Polymers and its composites areextensively used in sliding/rolling components suchas gears, cams, bearings, rollers, transmission beltsand grinding mills where their self-lubricatingproperties are exploited to avoid the need for oil orgrease lubrication with its attendant problems of13 th International Conference on Tribology – Serbiatrib’13 65

contamination [20,21]. However, when the contactbetween sliding pairs is present, there is theproblem of friction and wear. [22-24] demonstratedthat the friction coefficient can, generally, bereduced and the wear resistance increased byselecting the right material combinations.It was reported [25-27] that the influence of slidingspeed on friction and wear of polymer and itscomposite is greater than that of applied load thoughsome other researchers have different views. Unal etal. [28,29]reported that the applied load exerts greaterinfluence on the sliding wear of polymer and itscomposites than the sliding speed. Transfer film hasimportant effects on the tribological behavior ofpolymer and its’ composite. If the transfer film is thin,uniform and continuous, the wear loss and the frictioncoefficient are low [30]. The results by [31, 32]showed that tribological performance of polymermaterial can be improved significantly by fibrereinforcement or fillers. The reason was that thetransfer films formed and adhered close on the surfaceof counterface material during friction which resultedin the increase in wear resistance of the composites[31,10]. It was showed [33] that reinforcement offibre or filler significantly improves the tribologicalbehavior of polymeric material but this is notnecessarily true for all cases. Franklin [34] reportedthat wear behavior of polymers under dryreciprocating sliding conditions does not alwaysfollow the generally accepted engineering rule of‘higher sliding speed, the higher wear rate’. Theinfluence of normal load on the friction coefficientand wear rate of different polymer and compositematerials was investigated [35]and it was found thatthe values of friction coefficient and wear rate aredifferent for different materials. Several researchers[36-39] reported that friction coefficient of polymersand its composites rubbing against metal increases ordecreases depending on the range of sliding speed andsliding pairs. Researchers [40-43] have also observedthat the friction coefficient of polymers and itscomposites rubbing against metals decreases with theincrease in load though some other researchers havedifferent views. It was showed [45-47] that value offriction coefficient increases with the increase in load.Friction coefficient and specific wear rate values fordifferent combinations of polymer and its compositewere obtained and compared [27]. For all materialcombinations, it was observed that the coefficient offriction decreases linearly with the increase in appliedpressure values. Unal et al. [37,29] reported that theapplied load exerts greater influence on the slidingwear of polymer and its composite than the slidingvelocity. Friction and wear behavior of glass fiberreinforcedpolyester composite were studied andresults showed that in general, friction and wear arestrongly influenced by all the test parameters such asapplied load, sliding speed, sliding distance and fiberorientations [48]. Moreover, it was found that appliednormal load, sliding speed and fiber orientations havemore pronounced effect on wear rate than slidingdistance. Wang and Li [26] observed that the slidingvelocity has more significant effect on the slidingwear as compared to the applied load and variations ofwear rate with operating time can be distinguished bythree distinct periods. These periods are running-inperiod, steady state period and severe wear period,respectively. The friction and the wear behavior of thepolymeric material depend on the nature, thicknessand stability of the transfer film that is formed and onthe properties of the metallic counter face material[49]. Yang [50] studied the transfer ofpolytetrafluoroethylene (PTFE) on to 316 stainlesssteel and silicon wafers using infraredspectrophotometry and founds that it was stronglytime and temperature dependent and reached a steadystate after a certain period of contact. Tsukizoe andOhmae [33] showed that reinforcement of fiber orfiller significantly improve the tribological behaviorof polymeric material but this is not necessarily truefor all cases. Suresha et al. [38] showed that there is astrong inter-dependence on the friction coefficient andwear loss with respect to the applied loads for steelcomposites contact. It was found that the coefficientof friction and wear loss increase with the increase inapplied normal load for all the samples evaluated.From the aforementioned research works, it can beconcluded that friction coefficient of compositematerials at different normal loads and slidingvelocities differs significantly. Even now a day, theeffect of normal load and sliding velocity on frictioncoefficient and wear rate of composite materials suchas gear fiber and glass fiber sliding against differentcounterface surface conditions is less understood. Thismeans that more research work is needed for a betterunderstanding of friction coefficient and wear rate ofthese materials under different normal loads andsliding velocities for smooth and rough mild steelcounterfaces. Therefore, in order to understand moreclearly, in this study experiments are carried out toinvestigate the influence of normal loads and slidingvelocities on friction coefficient and wear rate of gearfiber and glass fiber. The effects of duration ofrubbing on friction coefficient of these materials arealso examined in this study.2. EXPERIMENTALA schematic diagram of the experimental set-upis shown in Fig. 1 i.e. a pin which can slide on arotating horizontal surface (disc). In this set-up acircular test sample (disc) is to be fixed on arotating plate (table) having a long vertical shaftclamped with screw from the bottom surface of the66 13 th International Conference on Tribology – Serbiatrib’13

otating plate. The shaft passes through two closefitbush-bearings which are rigidly fixed withstainless steel plate and stainless steel base suchthat the shaft can move only axially and any radialmovement of the rotating shaft is restrained by thebush. These stainless steel plate and stainless steelbase are rigidly fixed with four vertical round barsto provide the rigidity to the main structure of thisset-up. The main base of the set-up is constructedby 10 mm thick mild steel plate consisting of 3 mmthick rubber sheet at the upper side and 20 mmthick rubber block at the lower side. A compoundV-pulley above the top stainless steel plate wasfixed with the shaft to transmit rotation to the shaftfrom a motor. An electronic speed control unit isused to vary the speed of the motor as required. A 6mm diameter cylindrical pin whose contacting footis flat, made of mild steel, fitted on a holder issubsequently fitted with an arm. The arm is pivotedwith a separate base in such a way that the arm withthe pin holder can rotate vertically and horizontallyabout the pivot point with very low friction. Slidingspeed can be varied by two ways (i) by changingthe frictional radius and (ii) by changing therotational speed of the shaft. In this research,sliding speed is varied by changing the rotationalspeed of the shaft while maintaining 25 mmconstant frictional radius. To measure the frictionalforce acting on the pin during sliding on therotating plate, a load cell (TML, Tokyo SokkiKenkyujo Co. Ltd, CLS-10NA) along with itsdigital indicator (TML, Tokyo Sokki Kenkyujo Co.Ltd, Model no. TD-93A) was used. The coefficientof friction was obtained by dividing the frictionalforce by the applied normal force (load). Wear wasmeasured by weighing the test sample with anelectronic balance before and after the test, and thenthe difference in mass was converted to wear rate.To measure the surface roughness of the testsamples, Taylor Hobson Precision RoughnessChecker (Surtronic 25) was used. Each test wasconducted for 30 minutes of rubbing time with newpin and test sample. Furthermore, to ensure thereliability of the test results, each test was repeatedfive times and the scatter in results was small,therefore the average values of these test resultswere taken into consideration. The detailexperimental conditions are shown in Table 1.1151617214356131941112789 101 Load arm holder2. Load arm3. Normal load (dead weight)4. Horizontal load (Friction force)5. Pin sample6. Test disc with rotating table7. Load cell indicator8. Belt and pulley9. Motor10. Speed control unit11. Vertical motor base12. 3 mm Rubber pad13. Main shaft14. Stainless steel base15. Stainless steel plate16. Vertical square bar17. Mild steel main base plate18. Rubber block (20 mm thick)19. Pin holder.18Fig. 1. Block diagram of the experimental set-up.Table 1. Experimental Conditions.Sl. No. Parameters Operating Conditions1. Normal Load 10, 15, 20 N2. Sliding Velocity 1, 1.5, 2 m/s3. Relative Humidity 70 ( 5)%4. Duration of Rubbing 30 minutes5. Surface Condition Dry6. Disc material(i) Gear fiber reinforced Plastic(ii) Glass fiber reinforced plastic7. Roughness of Gear and Glass fiber, R a 0.70-0.80 m8. Pin material Mild steel9. Roughness of mild steel, R a(a) Smooth counterface: about 0.30 m(b) Rough counterface: about 3.0 m13 th International Conference on Tribology – Serbiatrib’13 67

3. RESULTS AND DISCUSSIONFigure 2 shows the variation of friction coefficientwith the duration of rubbing at different normal loadsfor gear fiber sliding against smooth mild steelconterface. During experiment, the sliding velocityand relative humidity were 1 m/s and 70%respectively. Curves 1, 2 and 3 of this figure aredrawn for normal laod 10, 15 and 20 N respectively.Curve 1 of this figure shows that during the starting,the value of friction coefficient is 0.104 and thenincreases very steadily up to 0.147 over a duration of20 minutes of rubbing and after that it remainsconstant for the rest of the experimental time. Thesefindings are in agreement with the findings ofChowdhury and Helali [4]. At starting of experiment,the friction force is low due to contact betweensuperficial layer of pin and disc. As rubbingcontinues, the disc material becomes worn andreinforced material comes in contact with the pin,roughening of the disc surface causes the ploughingand hence friction coefficient increases with durationof rubbing. After certain duration of rubbing theincrease of roughness and other parameters may reachto a certain steady value hence the values of frictioncoefficient remain constant for the rest of the time.Curves 2 and 3 show that for the high normal load, thefriction coefficient is less and the trend in variation offriction coefficient is almost the same as for curve 1.From these curves, it is also observed that time toreach steady state values is different for differentnormal load. From the obtained results it is found thatat normal load 10, 15 and 20 N, gear fibre takes 20, 17and 15 minutes respectively to reach steady friction. Itindicates that the higher the normal load, time to reachconstant friction is less. This may be due to the factthat the higher the normal load, the surface roughnessand other parameters take less time to stabilize.Friction coefficient0.250.200.150.100.0510 N15 N20 N0.000 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32Duration of rubbing (min)Fig. 2. Friction coefficient as a function of duration ofrubbing at different normal loads (sliding velocity: 1m/s, relative humidity: 70%, test sample: gear fiber, pin:mild steel, smooth).Figure 3 shows the effect of the duration ofrubbing on the value of friction coefficient atdifferent normal loads for gear fiber sliding against123rough mild steel counterface at sliding velocity 1m/s and relative humidity 70%. Curve 1 of thisfigure drawn for normal load 10 N, shows thatduring starting of the experiment, the value offriction coefficient is 0.153 which rises for 22minutes to a value of 0.195 and then it becomessteady for the rest of the experimental time. Almostsimilar trends of variation are observed in curves 2and 3 which are drawn for load 15 and 20 Nrespectively. From these curves, it is found thattime to reach steady friction is different fordifferent normal loads. At normal load 10, 15 and20 N, gear fiber-mild steel rough pair takes 22, 19and 16 minutes respectively to reach steady frictionThat is, higher the normal load, gear fiber-mildsteel rough pair takes less time to stabilize.Friction coefficient0.250.200.150.100.0510 N15 N20 N0.000 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32Duration of rubbing (min)Fig. 3. Friction coefficient as a function of duration ofrubbing at different normal loads (sliding velocity: 1m/s, relative humidity: 70%, test sample: gear fiber, pin:mild steel, rough).Friction coefficient0.250.200.150.100.0510 N15 N20 N0.000 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32Duration of rubbing (min)Fig. 4. Friction coefficient as a function of duration ofrubbing at different normal loads (sliding velocity: 1m/s, relative humidity: 70%, test sample: glass fiber, pin:mild steel, smooth).Figure 4 shows the effect of the duration ofrubbing on the value of friction coefficient atdifferent normal load for glass fiber sliding againstsmooth mild steel counterface. Curve 1 of thisfigure drawn for normal load 10 N, shows thatduring starting of the experiment, the value offriction coefficient is 0.123 which rises for 21minutes to a value of 0.167 and then it becomessteady for the rest of the experimental time. Almost12312368 13 th International Conference on Tribology – Serbiatrib’13

similar trends of variation are observed in curves 2and 3 which are drawn for load 15 and 20 N,respectively. From the obtained results, it can alsobe seen that time to reach constant friction isdifferent for different normal load and higher thenormal load, glass fiber takes less time to stabilize.Several experiments are conducted to observethe effect of duration of rubbing on frictioncoefficient at different sliding speeds for glass fibresliding against rough mild steel counterface andthese results are presented in Figure 5. Curve 1 ofthis figure drawn for normal load 10 N, shows thatduring starting of the experiment, the value offriction coefficient is 0.175 which rises for 22minutes to a value of 0.225 and then it becomessteady for the rest of the experimental time. Almostsimilar trends of variation are observed in curves 2and 3 which are drawn for load 15 and 20 Nrespectively. From these curves, it is found thattime to reach steady friction is different fordifferent normal loads. At normal load 10, 15 and20 N, glass fiber-mild steel rough pair takes 20, 18and 15 minutes respectively to reach steady frictionThat is, higher the normal load, glass fiber-mildsteel rough pair takes less time to stabilize.Friction coefficient0.250.200.150.100.0510 N15 N20 N0.000 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32Duration of rubbing (min)Fig. 5. Friction coefficient as a function of duration ofrubbing at different normal loads (sliding velocity: 1m/s, relative humidity: 70%, test sample: glass fiber, pin:mild steel, rough).Figure 6 shows comparison of the variation offriction coefficient with normal load for gear fibermildsteel smooth, gear fiber-mild steel rough, glassfiber-mild steel smooth and glass fiber-mild steelrough sliding pairs. Curves of this figure are drawnfrom steady values of friction coefficient shown inFigures 2-5 for gear fiber-mild steel smooth, gearfiber-mild steel rough, glass fiber-mild steel smoothand glass fiber-mild steel rough sliding pairs,respectively (to ensure the reliability of test results,each test was repeated five times and curves 1-3 ofFigures 2-5 represent average value of fiveexperiments). It is shown that the friction coefficientvaries from 0.147 to 0.108, 0.195 to 0.127, 0.167 to0.123 and 0.225 to 0.135 with the variation of123normal load from 10 to 20 N for for gear fiber-mildsteel smooth, gear fiber-mild steel rough, glass fibermildsteel smooth and glass fiber-mild steel roughsliding pairs, respectively. From the obtained results,it can be seen that the coefficient of frictiondecreases with the increase in applied load. It isknown that tribological behavior of polymers andpolymer composites can be associated with theirviscoelastic and temperature-related properties.Sliding contact of two materials results in heatgeneration at the asperities and hence increases intemperature at the frictional surfaces of the twomaterials which influences the viscoelastic propertyin the response of materials stress, adhesion andtransferring behaviors [27]. From the obtainedresults, it can also be seen that the highest values ofthe friction coefficient are obtained for glass fibermildsteel rough pair and the lowest values offriction coefficient are obtained for gear fiber-mildsteel smooth pair. The values of friction coefficientof gear fiber-mild steel rough pair and glass fibermildsteel smooth pair are found in between thehighest and lowest values. It is noted that the frictioncoefficients of gear fiber-mild steel rough pair arehigher than that of glass fiber-mild steel smooth pair.From this figure, it is also found that at identicalconditions, the values of friction coefficient of gearfiber and glass fiber sliding against smooth mildsteel counterface is lower than that of gear fiber andglass fiber sliding against rough mild steelcounterface. It was found that after friction tests, theaverage roughness of gear fiber-mild steel smoothpair, glass fiber-mild steel smooth pair, gear fibermildsteel rough pair and glass fiber-mild steel roughpair varied from 0.95-1.35, 1.25-1.65 and 1.55-1.75and 1.67-1.91 μm respectively.Friction coefficient0.250.200.150.100.05gear fiber-mild steel, smooth pairgear fiber-mild steel, rough pairglass fiber-mild steel, smooth pairglass fiber-mild steel, rough pair0.008 10 12 14 16 18 20 22 24Normal load (N)Fig. 6. Friction coefficient as a function of Normal load forgear and glass fiber for different counterface surfaceconditions (Sliding velocity: 1 m/s, relative humidity: 70%).Figures 7, 8, 9 and 10 show the variation offriction coefficient with the duration of rubbingat different sliding velocities for gear fiber-mildsteel smooth, gear fiber-mild steel rough, glassfiber-mild steel smooth and glass fiber-mild steel13 th International Conference on Tribology – Serbiatrib’13 69

ough sliding pairs, respectively at normal load15 N and relative humidity 70%. Curves 1, 2 and3 of Fig. 7 are drawn for sliding velocity 1, 1.5and 2 m/s respectively. Curve 1 of this figureshows that at initial stage of rubbing, the value offriction coefficient is 0.087 which increasesalmost linearly up to 0.123 over a duration of 17minutes of rubbing and after that it remainsconstant for the rest of the experimental time. Atstarting of experiment, the friction force is lowdue to contact between superficial layer of pinand disc. As rubbing continues, the disc materialbecomes worn and reinforced material comes incontact with the pin, roughening of the discsurface causes the ploughing and hence frictioncoefficient increases with duration of rubbing.After certain duration of rubbing the increase ofroughness and other parameters may reach to acertain steady value hence the values of frictioncoefficient remain constant for the rest of thetime. Curves 2 and 3 show that for the highersliding velocity, the friction coefficient is moreand the trend in variation of friction coefficient isalmost the same as for curve 1. From thesecurves, it is also observed that time to reachsteady state value is different for different slidingvelocity. From the results it is found that gearfiber-mild steel smooth pair at sliding velocity 1,1.5 and 2 m/s takes to reach constant friction 17,14 and 11 minutes respectively. It indicates thatthe higher the sliding velocity, time to reachconstant friction is less. This may be due to thehigher the sliding velocity, the surface roughnessand other parameters take less time to stabilize.From Figs. 8, 9 and 10, it can be observed thatthe trends in variation of friction coefficient withthe duration of rubbing are very similar to that ofFig. 7 but the values of friction coefficient aredifferent for gear fiber-mild steel rough pair,glass fiber-mild steel smooth pair and glass fibermildsteel rough pair.Friction coefficient0.250.200.150.100.051 m/s1.5 m/s2 m/s0.000 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32Duration of rubbing (min)Fig. 7. Friction coefficient as a function of duration ofrubbing at different sliding velocities (normal load: 15N, relative humidity: 70%, test sample: gear fiber, pin:mild steel, smooth).321Friction coefficient0.250.200.150.100.051 m/s1.5 m/s2 m/s0.000 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32Duration of rubbing (min)Fig. 8. Friction coefficient as a function of duration ofrubbing at different sliding velocities (normal load: 15N, relative humidity: 70%, test sample: gear fiber, pin:mild steel, rough).Friction coefficient0.250.200.150.100.051 m/s1.5 m/s2 m/s0.000 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32Duration of rubbing (min)Fig. 9. Friction coefficient as a function of duration ofrubbing at different sliding velocities (normal load: 15N, relative humidity: 70%, test sample: glass fiber, pin:mild steel, smooth).Friction coefficient0.250.200.150.100.051 m/s1.5 m/s2 m/s0.000 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32Duration of rubbing (min)Fig. 10. Friction coefficient as a function of duration ofrubbing at different sliding velocities (normal load: 15N, relative humidity: 70%, test sample: glass fiber, pin:mild steel, rough).Figure 11 shows the comparison of the variationof friction coefficient with sliding speed for gearfiber-mild steel smooth, gear fiber-mild steel rough,glass fiber-mild steel smooth and glass fiber-mildsteel rough sliding pairs. Curves of this figure aredrawn from steady values of friction coefficientshown in Figures 7–10 for gear fiber-mild steel32132132170 13 th International Conference on Tribology – Serbiatrib’13

smooth, gear fiber-mild steel rough, glass fibermildsteel smooth and glass fiber-mild steel roughsliding pairs. It is shown that the friction coefficientvaries from 0.123 to 0.165, 0.143 to 0.189, 0.137 to0.176 and 0.156 to 0.213 with the variation ofsliding speed from 1 to 2 m/s for gear fiber-mildsteel smooth, gear fiber-mild steel rough, glassfiber-mild steel smooth and glass fiber-mild steelrough sliding pairs respectively. From these resultsit is seen that the values of friction coefficientincrease almost linearly with sliding speed. Thesefindings are in agreement with the findings ofMimaroglu et al. and Unal et al. [27,28]. With theincrease in sliding speed, the frictional heat maydecrease the strength of the materials and hightemperature results in stronger or increasedadhesion with pin [27,51]. The increase of frictioncoefficient with sliding speed can be explained bythe more adhesion of counterface pin material ondisc. From the obtained results, it can also be seenthat the highest values of the friction coefficient areobtained for glass fiber-mild steel rough pair andthe lowest values of friction coefficient are obtainedfor gear fiber-mild steel smooth pair. The values offriction coefficient of gear fiber-mild steel roughpair and glass fiber-mild steel smooth pair arefound in between the highest and lowest values. Itis noted that the friction coefficients of gear fibermildsteel rough pair are higher than that of glassfiber-mild steel smooth pair. From this figure, it isalso found that at identical conditions, the values offriction coefficient of gear fiber and glass fibersliding against smooth mild steel counterface islower than that of gear fiber and glass fiber slidingagainst rough mild steel counterface. It was foundthat after friction tests, the average roughness ofgear fiber-mild steel smooth pair, glass fiber-mildsteel smooth pair, gear fiber-mild steel rough pairand glass fiber-mild steel rough pair varied from1.05-1.45, 1.35-1.78 and 1.67-1.88 and 1.76-1.98μm respectively.Friction coefficient0.250.200.150.100.05gear fiber-mild steel, smooth pairgear fiber-mild steel, rough pairglass fiber-mild steel, smooth pairglass fiber-mild steel, rough pair0.000.5 1.0 1.5 2.0 2.5Sliding velocity (m/s)Fig. 11. Friction coefficient as a function of Normal loadfor gear and glass fiber for different counterface surfaceconditions (normal load: 15 N, relative humidity: 70%).Variations of wear rate with normal load forgear fiber and glass fiber sliding against smooth orrough mild steel counterfaces are shown in Fig. 12.The experimental results indicate that the curvesdrawn showing the variation of wear rate from0.815 to 1.453, 1.135 to 1.751, 0.929 to 1.553 and1.638 to 2.35 mg/min with the variation of normalload from 10 to 20 N for gear fiber-mild steelsmooth, gear fiber-mild steel rough, glass fibermildsteel smooth and glass fiber-mild steel roughsliding pairs respectively. From these curves, it isobserved that wear rate increases with the increaseof normal load for all types of sliding pairs. Whenthe load on the pin is increased, the actual area ofcontact would increase towards the nominal contactarea, resulting in increased frictional force betweentwo sliding surfaces. The increased frictional forceand real surface area in contact causes higher wear.This means that the shear force and frictional thrustare increased with increase of applied load andthese increased in values accelerate the wear rate.Similar trends of variation are also observed formild steel–mild steel couples [52], i.e wear rateincreases with the increase in normal load.Wear rate (mg/min)3.02.52.01.51.00.5gear fiber-mild steel, smooth pairgear fiber-mild steel, rough pairglass fiber-mild steel, smooth pairglass fiber-mild steel, rough pair0.08 10 12 14 16 18 20 22 24Normal load (N)Fig. 12. Wear rate as a function of Normal load for gear andglass fiber for different counterface surface conditions(Sliding velocity: 1 m/s, relative humidity: 70%).Figure 12 also shows the comparison of thevariation of wear rate with normal load for gearfiber and glass fiber under different pin surfaceconditions. From the obtained results, it can also beseen that the highest values of the wear rate areobtained for glass fiber-mild steel rough pair andthe lowest values of wear rate are obtained for gearfiber-mild steel smooth pair. The values of wearrate of gear fiber-mild steel rough pair and glassfiber-mild steel smooth pair are found in betweenthe highest and lowest values. It is noted that thewear rates of gear fiber-mild steel rough pair arehigher than that of glass fiber-mild steel smoothpair. From this figure, it is also found that atidentical conditions, the values of wear rate of gear13 th International Conference on Tribology – Serbiatrib’13 71

fiber and glass fiber sliding against smooth mildsteel counterface is lower than that of gear fiber andglass fiber sliding against rough mild steelcounterface.Variations of wear rate with sliding velocity forgear fiber and glass fibre mating with smooth orrough mild steel counterfaces are presented in Fig.13. Curves show the variation of wear rate from1.167 to 1.778, 1.433 to 2.25, 1.258 to 1.95 and1.987 to 2.78 mg/min with the variation in slidingspeed from 1 to 3 m/s for gear fiber-mild steelsmooth, gear fiber-mild steel rough, glass fibermildsteel smooth and glass fiber-mild steel roughsliding pairs respectively. From these curves, it isobserved that wear rate increases with the increasein sliding speed for all types of materialcombinations. These findings are in agreement withthe findings of Mimaroglu et al and Suresha et al.[27,38]. This is due to the fact that duration ofrubbing is same for all sliding velocities, while thelength of rubbing is more for higher slidingvelocity. The reduction of shear strength of thematerial and increased true area of contact betweencontacting surfaces may have some role on thehigher wear rate at higher sliding velocity [51].Figure 13 also shows the comparison of thevariation of wear rate with sliding velocity fordifferent sliding pairs. From the obtained results, itcan also be seen that the highest values of the wearrate are obtained for glass fiber-mild steel roughpair and the lowest values of wear rate are obtainedfor gear fiber-mild steel smooth pair. The values ofwear rate of gear fiber-mild steel rough pair andglass fiber-mild steel smooth pair are found inbetween the highest and lowest values. It is notedthat the wear rates of gear fiber-mild steel roughpair are higher than that of glass fiber-mild steelsmooth pair.Wear rate (mg/min)3.02.52.01.51.00.5gear fiber-mild steel, smooth pairgear fiber-mild steel, rough pairglass fiber-mild steel, smooth pairglass fiber-mild steel, rough pair0.00.5 1.0 1.5 2.0 2.5Sliding velocity (m/s)Fig. 13. Wear rate as a function of Normal load for gearand glass fiber for different counterface surfaceconditions (normal load: 15 N, relative humidity: 70%).From this figure, it is also found that at identicalconditions, the values of wear rate of gear fiber andglass fiber sliding against smooth mild steelcounterface is lower than that of gear fiber andglass fiber sliding against rough mild steelcounterface. It is due to the fact that rough surfacesgenerally wear more quickly and have higherfriction coefficients than smooth surfaces.4. CONCLUSIONThe presence of normal load and sliding velocityindeed affects the friction force considerably.Within the observed range, the values of frictioncoefficient decrease with the increase in normalload while friction coefficients increase with theincrease in sliding velocity for gear fiber and glassfiber sliding against smooth or rough mild steel pin.Friction coefficient varies with the duration ofrubbing and after certain duration of rubbing,friction coefficient becomes steady for the observedrange of normal load and sliding velocity. Wearrates of gear fiber and glass mating with smooth orrough mild steel counterface increase with theincrease in normal load and sliding velocity. Thehighest values of the friction coefficient areobtained for glass fiber-mild steel rough pair andthe lowest values of friction coefficient are obtainedfor gear fiber-mild steel smooth pair. The values offriction coefficient of gear fiber-mild steel roughpair and glass fiber-mild steel smooth pair arefound in between the highest and lowest values.The friction coefficients of gear fiber-mild steelrough pair are higher than that of glass fiber-mildsteel smooth pair. At identical conditions, thevalues of friction coefficient of gear fiber and glassfiber sliding against smooth mild steel counterfaceis lower than that of gear fiber and glass fibersliding against rough mild steel counterface.As (i) the friction coefficient decreases with theincrease in normal load (ii) the values of frictioncoefficient increase with the increase in slidingvelocity (iii) wear rate increases with the increasein normal load and sliding velocity and (iv) themagnitudes of friction coefficient and wear rate aredifferent for smooth and rough counterface pins andtype of materials, therefore maintaining anappropriate level of normal load, sliding velocity aswell as appropriate choice of counterface surfacecondition and tested materials, friction and wearmay be kept to some lower value to improvemechanical processes.REFERENCES[1] Archard, J.F.: “Wear theory and mechanisms”,Wear Control Handbook, ASME, New York, NY,1980.72 13 th International Conference on Tribology – Serbiatrib’13

[2] Tabor, D.: “Friction and wear – developments overthe last 50 years, keynote address”, Proceedings ofthe International Conference of Tribology –Friction, Lubrication and Wear, Queen Elizabeth IIConference Centre, London,Institute of MechanicalEngineering, London, pp. 157-72, 1987.[3] Kukureka, S.N., Chen, Y.K., Hooke, C.J. and Liao,P.: “The wear mechanisms of acetal inunlubricatedrolling-sliding contact”, Wear, Vol.185, pp. 1-8, 1995.[4] Chowdhury, M.A. and Helali, M.M.: “The effect ofamplitude of vibration on the coefficient of frictionfordifferent materials”, Tribology International, Vol.41, No. 4, pp. 307-14, 2008.[5] Chowdhury, M.A. and Helali, M.M.: “The frictionalbehavior of composite materials under horizontalvibration”, Industrial Lubrication and Tribology,Vol. 61, No. 5, pp. 246-53, 2009a.[6] Chowdhury, M.A. and Helali, M.M.: “The frictionalbehavior of materials under vertical vibration”,Industrial Lubrication and Tribology, Vol. 61, No.3, pp. 154-60, 2009b.[7] El-Tayeb, N.S.M. and Mostafa, I.M.: “The effect oflaminate orientations on friction and wearmechanisms of glass reinforced polyestercomposite”, Wear, Vol. 195, pp. 186-91, 1996.[8] El-Tayeb, N.S.M. and Gadelrab, R.M.: “Frictionand wear properties of e-glass fiber reinforcedepoxy compositesunder different sliding contactconditions”, Wear, Vol. 192, pp. 112-17, 1996.[9] Bahadur, S. and Zheng, Y.: “Mechanical andtribological behavior of polyester reinforced withshort glass fibers”, Wear, Vol. 137, pp. 251-66, 1990.[10] Bahadur, S. and Polineni, V.K.: “Tribologicalstudies of glass fabric-reinforced polyamidecomposites filled with CuO and PTFE”, Wear, Vol.200, pp. 95-104, 1996.[11] Watanabe, M.: “The friction and wear properties ofnylon”, Wear, Vol. 110, pp. 379-88, 1968.[12] Tanaka, K.: “Transfer of semi crystalline polymerssliding against smooth steel surface”, Wear, Vol.75, pp. 183-99, 1982.[13] Bahadur, S. and Tabor, D.: “Role of fillers in thefriction and wear behavior of high-densitypolyethylene”, in Lee, L.H. (Ed.), Polymer Wear andIts Control, ACS Symposium Series, Vol. 287, ACSPublications, Washington, DC, pp. 253-68, 1985.[14] Pihtili, H. and Tosun, N.: “Effect of load and speedon the wear behavior of woven glass fabrics andaramid fiber-reinforced composites”, Wear, Vol.252, pp. 979-84, 2002a.[15] Pihtili, H. and Tosun, N.: “Investigation of the wearbehavior of a glass fiber-reinforced composite andplain polyester resin”, Composites Science andTechnology, Vol. 62, pp. 367-70, 2002b.[16] Bijwe, J., Tewari, U.S., Vasudevan, P.: “Frictionand wear studies of polyetherimide composites”,Wear, Vol. 138, pp. 61-76, 1990.[17] Bijwe, J., Indumathi, J., John Rajesh, J. and Fahim,M.: “Friction and wear behavior of polyetherimidecomposites in various wear modes”, Wear, Vol. 249,pp. 715-26, 2001.[18] Bijwe, J. and Indumathi, J.: “Influence of fibers andsolid lubricants on low amplitude oscillating wearof polyetherimide composites”, Wear, Vol. 257, No.5/6, pp. 562-72, 2004.[19] Bijwe, J., Indumathi, J. and Ghosh, A.K.: “Role offabric reinforcement on the low amplitudeoscillating wear of polyetherimide composites”,Wear, Vol. 256, No. 1/2, pp. 27-37, 2004.[20] Zhang, S.W.: ‘State-of-the-art of polymer tribology’,Tribol. Int., Vol. 31, Nos. 1–3, pp.49-60, 1998.[21] Kowandy, C., Richard, C. and Chen, Y.M.:‘Characterization of wear particles forcomprehension of wear mechanisms case of PTFEagainst cast iron’, Wear, Vol. 265, No. 11–12, pp.1714–1719, 2008.[22] Yamaguchi, Y.: Tribology of Plastic Materials:Their Characteristics and Applications to SlidingComponents, Elsevier, Amsterdam, 1990.[23] Hooke, C.J., Kukureka, S.N., Liao, P., Rao, M. andChen, Y.K.: “The friction and wear of polymersinnon-conformal contacts”, Wear, Vol. 200, pp. 83-94, 1996.[24] Lawrence, C.C. and Stolarski, T.A.: “Rollingcontact wear of polymers: a preliminary study”,Wear, Vol. 132, pp. 83-91, 1989.[25] Feyzullahoglu, E. and Saffak, Z.: ‘The tribologicalbehavior of different engineering plastics under dryfriction conditions’, Mater. Design, Vol. 29, No. 1,pp. 205–211, 2008.[26] Wang, Y.Q., Li, J.: ‘Sliding wear behavior andmechanism of ultra-high molecularweightpolyethylene’, Mater. Sci. Eng., Vol. 266, No. 1–2,pp. 155–160, 1999.[27] Mimaroglu, A., Unal, H. and Arda, T.: ‘Friction andwear performance of pure and glass fiber reinforcedpoly-ether-imide on polymer and steel counterfacematerials’, Wear, Vol. 262, No. 11–12, pp.1407–1413, 2007.[28] Unal, H., Sen, U. and Mimaroglu, A.: ‘Dry slidingwear characteristics of some industrial polymersagainst steel counterface’, Tribol. Int., Vol. 37, No.9, pp. 727–732, 2004.[29] Unal, H., Sen, U. and Mimaroglu, A.: ‘An approachto friction and wear propertiesofpolytetrafluoroethylene composite’, Mater.Design, Vol. 27, No. 8, pp. 694–699, 2006.[30] Sirong, Y., Zhongzhen, Yu., Mai, Y-W.: ‘Effects ofSEBS-g-MA on tribological behavior of nylon66/organoclay nanocomposites’, Tribol. Int., Vol.40, No. 5, pp. 855–862, 2007.[31] Bahadur, S. and Kapoor, A.: ‘The effect of ZnF2,ZnS and PbS fillers on the tribological behavior ofnylon 11’, Wear, Vol. 155, No. 1, pp. 49–61, 1992.[32] Wang, J., Gu, M., Bai, S. and Ge, S.: ‘Investigationof the influence of MoS2 filler on the tribological13 th International Conference on Tribology – Serbiatrib’13 73

properties of carbon fiber reinforced nylon 1010composites’, Wear, Vol. 255, No. 1–6, pp. 774–779,2003.[33] Tsukizoe, T. and Ohmae, N.: ‘Friction and wear ofadvanced composite materials’, FibreSci. Technol.,Vol. 18, No. 4, pp. 265–286, 1983.[34] Franklin, S.E.: ‘Wear experiments with selectedengineering polymers and polymer compositesunder dry reciprocating sliding conditions’, Wear,Vol. 251, No. 1–12, pp. 1591–1598, 2001.[35] Nuruzzaman, D.M., Chowdhury, M.A. andRahaman, M.L.: ‘Effect of duration of rubbing andnormal load on friction coefficient for polymer andcomposite materials’, Ind. Lubr. Tribol., Vol. 63,No. 5, pp. 320–326, 2011.[36] Benabdallah, H.: ‘Friction and wear of blendedpolyoxymethylene sliding against coated steel plates’,Wear, Vol. 254, No. 12, pp. 1239–1246, 2003.[37] Unal, H., Mimaroglu, A., Kadioglu, U. and Ekiz, H.:‘Sliding friction and wear behavior ofpolytetrafluoroethylene and its composites underdry conditions’, Mater. Design, Vol. 25, No. 3, pp.239–245, 2004.[38] Suresha, B., Chandramohan, G., Samapthkumaran,P., Seetharamu, S. and Vynatheya, S.: Friction andwear characteristics of carbon-epoxy and glassepoxywoven roving fiber composites, J. Reinf. Plast.Comp., Vol. 25, No. 7, pp.771–782, 2006.[39] Cho, M.H., Bahadur, S. and Pogosian, A.K.:‘Friction and wear studies using Taguchi method onpolyphenylene sulfide filled with a complex mixtureof MoS2, Al2O3, and other compounds’, Wear, Vol.258, No. 11–12, pp. 1825–1835, 2005.[40] Santner, E. and Czichos, H.: “Tribology ofpolymers”, Tribology International, Vol. 22, No. 2,pp. 103-9, 1989.[41] Tevruz, T.: “Tribological behaviours of carbonfilledpolytetrafluoroethylene dry journal bearings”,Wear, Vol. 221, pp. 61-8, 1998.[42] Tevruz, T.: “Tribological behaviours of bronzefilledpolytetrafluoroethylene dry journal bearings”,Wear, Vol. 230, pp. 61-9, 1999.[43] Anderson, J.C.: “The wear and friction ofcommercialpolymers and composites”, in Friedrich,K. (Ed.), Frictionand Wear and PolymerComposites, Composite MaterialsSeries, Vol. 1,Elsevier, Amsterdam, pp. 329-62, 1986.[44] Stuart, B.H.: “Tribological studies of poly (etherether ketone) blends”, Tribology International, Vol.31, No. 11, pp. 647-51, 1998.[45] Unal, H., Mimaroglu, A.: “Friction and wearbehavior of unfilled engineering thermoplastics”,Material Design, Vol. 24, pp. 183-7, 2003.[46] Unal, H. and Mimaroglu, A.: “Influence of testconditions on the tribological properties ofpolymers”, Industrial Lubrication and Tribology,Vol. 55, No. 4, pp. 178-83, 2003.[47] Suresha, B., Chandramohan, G., Prakash, J.N.,Balusamy, V. and Sankaranarayanasamy, K.: “Therole of fillerson friction and slide wearcharacteristics in glass-epoxy composite systems”,Journal of Minerals & Materials Characterization &Engineering, Vol. 5, No. 1, pp. 87-101, 2006a.[48] El-Tayeb, N.S.M., Yousif, B.F. and Yap, T.C.:“Tribological studies of polyester reinforced withCSM450-R-glass fiber sliding against smoothstainless steel counterface”, Wear, Vol. 261, pp.443-52, 2006.[49] Clerico, M. and Patierno, V.: “Sliding wear ofpolymeric composites”, Wear, Vol. 53, No. 2, pp.279-97, 299-301, 1979.[50] Yang, E.-L.: “Effect of crystalline and amorphousphases on the transfer of polytetrafluoroethylene(PTFE)onto metallic substrates”, Journal of MaterialsResearch, Vol. 7, No. 11, pp. 3139-49, 1992.[51] B. Bhushan: Principle and Applications of Tribology,John Wiley & Sons, Inc., New York, 1999.[52] M. A. Chowdhury, M. M. Helali: The Effect ofFrequency of Vibration and Humidity on the Wearrate, Wear, Vol. 262, pp. 198-203, 2007.74 13 th International Conference on Tribology – Serbiatrib’13

Serbian TribologySocietySERBIATRIB ‘1313 th International Conference onTribologyKragujevac, Serbia, 15 – 17 May 2013Faculty of Engineeringin KragujevacABRASIVE WEAR RESISTANCE OF THE IRON- ANDWC-BASED HARDFACED COATINGS EVALUATED WITHSCRATCH TEST METHODAleksandar Vencl 1 , Bojan Gligorijević 2 , Boris Katavić 2 , Bogdan Nedić 3 , Dragan Džunić 31 University of Belgrade, Faculty of Mechanical Engineering, Belgrade, Serbia, avencl@mas.bg.ac.rs2 Institute Goša, Belgrade, Serbia, bojan.gligorijevic@institutgosa.rs, boris.katavic@institutgosa.rs3 Faculty of Engineering, University of Kragujevac, Kragujevac, Serbia, nedic@kg.ac.rs, dzuna@kg.ac.rsAbstract: Abrasive wear is one of the most common types of wear, which makes abrasive wear resistancevery important in many industries. The hardfacing is considered as useful and economical way to improvethe performance of components submitted to severe abrasive wear conditions, with wide range of applicablefiller materials. The abrasive wear resistance of the three different hardfaced coatings (two iron-based andone WC-based), which were intended to be used for reparation of the impact plates of the ventilation mill,was investigated and compared. Abrasive wear tests were carried-out by using the scratch tester under thedry conditions. Three normal loads of 10, 50 and 100 N and the constant sliding speed of 4 mm/s were used.Scratch test was chosen as a relatively easy and quick test method. Wear mechanism analysis showedsignificant influence of the hardfaced coatings structure, which, along with hardness, has determinedcoatings abrasive wear resistance.Keywords: abrasive wear, scratch test, hardfacing, iron-based and WC-based materials, SEM-EDS.1. INTRODUCTIONMore than 50 % of all wear-related failures ofindustrial equipment are caused by abrasive wear[1]. The estimated costs of abrasive wear arebetween 1 and 4 % of the gross national product ofan industrialized nation [2]. For these reasons, theabrasive wear resistance is a subject of greatimportance in many industries, such as agriculture,mining, mineral processing etc.Hardfacing could be defined as “coatingdeposition process in which a wear resistant,usually harder, material is deposited on the surfaceof a component by some of the weldingtechniques”. In most cases, hardfacing is used forcontrolling abrasive and erosive wear, like inmining, crushing and grinding, and agricultureindustries (buckets, bucket teeth, mill hammers,ball mills, digging tools, conveyer screws, etc.[3,4]). Hardfacing is also used to controlcombinations of wear and corrosion, as encounteredby mud seals, plows, knives in the food processingindustry, pumps handling corrosive liquids, orslurries [5]. The hardfacing is considered aseconomical way to improve the performance ofcomponents submitted to severe wear conditions,with wide range of applicable filler materials [6,7].The iron-based filler materials have drawn muchattention due to their low cost and good resistanceto abrasion in the hardfaced condition. However,their use is limited in applications where highimpact loading is present, i.e. high-stress orgouging abrasion [8]. For this reason, efforts arebeing made towards the improvement of theirimpact and other properties [9]. The progress isachieved mostly by modifying the hardfacedcoating’s structure. Taking into account their lowprice and improved properties, the resistance toabrasive wear of the iron-based hardfaced coatingsis normally tested against the resistance of proven,but more expensive materials, such as WC-basedhardfaced coatings.Abrasive wear has been defined as “wear bydisplacement of material from surfaces in relativemotion caused by the presence of hard particleseither between the surfaces or embedded in one of13 th International Conference on Tribology – Serbiatrib’13 75

them, or by the presence of hard protuberances onone or both of the surfaces” [10]. The second partof this definition corresponds to pure two-bodyabrasion, where tested material slides againstharder and rougher counter face material, while thefirst part corresponds to the three- and two-bodyabrasion, respectively. Another interesting exampleof two-body abrasion is the abrasive erosion,which is the special case of erosive wear. Abrasiveerosion has been defined as “erosive wear in whichthe loss of material from a solid surface is due torelative motion of solid particles which areentrained in a fluid, moving nearly parallel to asolid surface” [10]. Scratch test offers a possibilityfor comparison of different materials relativelyeasy and in short period of time, with goodreproducibility [11]. In single-pass scratch test astylus (which tip is made of hard material) slideover the test sample producing a single scratch,which seems to be appropriate simulation of thetwo-body abrasion.In this study, the abrasive wear resistance of thethree different hardfaced coatings (two iron-based andone WC-based) was investigated and compared.2. EXPERIMENTAL DETAILS2.1 MaterialsThe filler materials (coating materials) weremanufactured by Castolin Eutectic Co. Ltd, Vienna.Their nominal chemical composition is shown inTable 1. The iron-based filler materials (basiccovered electrodes) were deposited by using theshielded metal arc welding (SMAW) process. TheWC-based filler material was deposited by oxy-fuelgas welding (OFW) process. The substrate materialwas the hot-rolled S355J2G3 steel.Table 1. Coatings composition, process and hardnessCoatingNominal chemicalcompositionHardfacingprocessHardnessHV 54541 Fe-Cr-C-Si SMAW 7395006 Fe-Cr-C-Si SMAW 7817888 T WC-Ni-Cr-Si-B OFW 677All coatings were deposited by hardfacing in asingle pass (one layer). The substrate preparationand hardfacing procedures (deposition parameters)are described elsewhere [9,12]. The measurementsof near-surface hardness are performed on thecross-section of hardfaced samples by Vickersindenter (HV 5), and presented in (Table 1).The samples for structure characterization areobtained by cutting the hardfaced materialsperpendicular to coatings surface. The obtainedcross-sections are ground with SiC abrasive papersdown to P1200 and polished with aluminasuspensions down to 1 μm. The polished surfaces areanalyzed by using the scanning electron microscope(SEM) equipped with energy dispersive system(EDS). The SEM-EDS analysis was performed atUniversity of Belgrade, Faculty of Mining andGeology by using the JEOL JSM–6610LV SEMconnected with the INCA350 energy dispersion X-ray analysis unit. The electron acceleration voltageof 20 kV and the tungsten filament were used.Before SEM-EDS analysis was performed, polishedsurfaces were 20 nm gold coated in a vacuumchamber by use of a sputter coater device.The Figure 1a shows the near-surface structureof the 4541 iron-based hardfaced coating. Theprimary austenite phase occupies more than a halfof volume (50.7 vol. %) and the rest is the lamellareutectic mixture of austenite and Cr-carbides [9].The 5006 material during solidification achievesnear-eutectic structure (Fig. 1b). A small sphericalprimary Cr-carbides are observed (9.1 vol. %) inthe eutectic matrix. Based on Electron microprobeanalysis (EMPA), both coatings 4541 and 5006contain (Cr,Fe) 7 C 3 primary and eutectic carbides.The Figure 1c shows a larger WC grains (60 vol.%), which are embedded in the Ni-Cr based matrix.2.2 Scratch abrasion testingAbrasive wear tests are carried out on the scratchtester under the dry conditions, in ambient air atroom temperature (≈ 25 °C). A schematic diagramof scratch testing is presented in Figure 2. Stylus(indenter) was pressed with selected normal loadFigure 1. The structures (SEM) of: (a) 4541, (b) 5006 and (c) 7888 T hardfaced coating; back-scattered electron images76 13 th International Conference on Tribology – Serbiatrib’13

(10, 50 and 100 N) against surface of the test sampleand moved with constant speed (4 mm/s), producingthe scratch of certain width and length (10 mm) onthe test sample. Indenter had Rockwell shape andthe cone was diamond with radius of 0.2 mm.Figure 2. Schematic diagram of scratch testingOn surface of each material under investigationat least three scratches are made with a gap betweenscratches of at least 1 mm. Before and after testing,both the indenter and the test samples are degreasedand cleaned with benzene. The wear scar widths onthe surface of the test samples are measured fromSEM images at the end of testing. The wear scarwidths and the known indenter geometry are usedto calculate the volume loss. After testing, themorphology of the test samples worn surfaces isexamined with SEM.3. RESULTS AND DISCUSSIONThe results of the wear tests are presented inFigures 3, 4 and 5. Taking into account significantdifferences in structure homogeneity of thehardfaced coatings (Fig. 1), the repeatability of theresults, in terms of standard deviations, issatisfactory (within 16 %). Wear rate of the testedmaterials (volume loss divided by scratch length)increases with normal loading, as expected. Thehighest wear exhibits coating 7888 T. Nevertheless,wear rates for all coatings are high, even forabrasive wear. The reason for this is primarily dueto the experimental conditions.The test conditions were specific, i.e. the speedswere very low (4 mm/s) and the contact stressesvery high. At the end of test, the normal stresseswere between 2 and 5 GPa, which depends on thematerial, i.e. scratch width and applied normal load.With these conditions, a high-stress or evengouging abrasion can be expected. With high-stressabrasion, the worn surface may exhibit varyingdegrees of scratching with plastic flow ofsufficiently ductile phases or fracture of brittlephases. In gouging abrasion, the stresses are higherthan those in high-stress abrasion, and they areaccompanied by large particles removal from thesurface, leaving deep groves and/or pits [8].The relation between the wear rate and thehardness of tested hardfaced coatings is shown inFigure 6. The first feature is that the abrasive wearrate decreases as the hardness increases, i.e. thehardest material (coating 5006) showed the highestabrasive wear resistance.Wear rate, mm 3 /mWear rate, mm 3 /mWear rate, mm 3 /m3.22.82.42.01.61.20.80.40.0Coating 4541(v = 4 mm/s)0.100.751.8810 50 100Normal load, NFigure 3. Wear rates of coating 4541 for differentnormal loads3.22.82.42.01.61.20.80.40.0Coating 5006(v = 4 mm/s)0.090.681.6710 50 100Normal load, NFigure 4. Wear rates of coating 5006 for differentnormal loads3.22.82.42.01.61.20.80.40.0Coating 7888 T(v = 4 mm/s)0.211.242.8810 50 100Normal load, NFigure 5. Wear rates of coating 7888 T for differentnormal loads13 th International Conference on Tribology – Serbiatrib’13 77

Wear rate, mm 3 /m3.503.002.502.001.501.000.500.002.881.240.211.880.7510 N50 N100 N1.670.680.10 0.09670 690 710 730 750 770 790Hardness HV5Figure 6. Wear rate vs. hardness of tested materials fordifferent normal loadsFor all applied loads, the relation betweenhardness and wear rate is non-linear. It is morecurved for higher loads (Fig. 6). This is connectedwith the coatings structure and exhibited wearmechanism. Coatings 4541 and 5006 exhibit mainlyploughing abrasive wear (Fig. 7a), while coating7888 T dominant type of abrasive wear is fracture(cracking) abrasive wear (Fig. 7b).Figure 7. The wear scar appearance (SEM) of: (a) 4541and (b) 7888 T hardfaced coating; 50 N normal load;back-scattered electron images4. CONCLUSIONScratch test offers relatively easy and quickcomparison of different materials on abrasive wear.Structure of tested coatings showed influence onthe dominant type of abrasive wear, which togetherwith coatings hardness determined coatingsabrasive wear resistance.Coatings with lower hardness showed lowerabrasive wear resistance, but the dependence(hardness vs. wear rate) was non-linear.In the case of iron-based coatings, dominanttype of abrasive wear was ploughing and in the caseof WC-based coatings, it was fracture (cracking)abrasive wear.ACKNOWLEDGEMENTThis work has been performed as a part ofactivities within the projects TR 34028, TR 35021and TR 35034. These projects are supported by theRepublic of Serbia, Ministry of Education, Scienceand Technological Development, whose financialhelp is gratefully acknowledged.REFERENCES[1] A. Rac: Osnovi tribologije, Mašinski fakultetUniverziteta u Beogradu, Beograd, 1991.[2] R.G. Bayer: Fundamentals of wear failures, in: ASMHandbook Volume 11, Failure Analysis andPrevention, ASM International, Metals Park, pp.901-905, 2002.[3] V. Lazić, M. Jovanović, D. Milosavljević, M.Mutavdžić, R. Čukić: Choosing of the most suitabletechnology of hard facing of mixer blades used inasphalt bases, Tribology in Industry, Vol. 30, No. 1-2, pp. 3-10, 2008.[4] V. Lazić, M. Mutavdžić, D. Milosavljević, S.Aleksandrović, B. Nedeljković, P. Marinković, R.Čukić: Selection of the most appropriate technologyof reparatory hard facing of working parts onuniversal construction machinery, Tribology inIndustry, Vol. 33, No. 1, pp. 18-27, 2011.[5] J.R. Davis, Hardfacing, weld cladding, anddissimilar metal joining, in: ASM Handbook Volume6, Welding, Brazing, and Soldering, ASMInternational, Metals Park, pp. 789-829, 1993.[6] EN 14700: Welding consumables – Weldingconsumables for hard-facing, European Committeefor Standardization, Brussels, 2005.[7] M. Šolar, M. Bregant: Dodatni materijali i njihovaupotreba kod reparaturnog navarivanja,Zavarivanje i zavarene konstrukcije, Vol. 51, No. 2,pp. 71-76, 2006.[8] A. Vencl, N. Manić, V. Popovic, M. Mrdak:Possibility of the abrasive wear resistancedetermination with scratch tester, Tribology Letters,Vol. 37, No. 3, pp. 591-604, 2010.78 13 th International Conference on Tribology – Serbiatrib’13

[9] B.R. Gligorijevic, A. Vencl, B.T. Katavic:Characterization and comparison of the carbidesmorphologies in the near surface region of thesingle- and double layer iron-based hardfacedcoatings, Scientific Bulletin of the "Politehnica"University of Timișoara, Transactions onMechanics, Vol. 57 (71), Special Issue S1, pp. 15-20, 2012.[10] OECD, Research Group on Wear of EngineeringMaterials, Glossary of Terms and Definitions in theField of Friction, Wear and Lubrication: Ttribology,Organisation for Economic Co-operation andDevelopment, Paris, 1969.[11] A. Vencl, A. Rac, B. Ivković: Investigation ofabrasive wear resistance of ferrous-based coatingswith scratch tester, Tribology in Industry, Vol. 29,No. 3-4, pp. 13-16, 2007.[12] A. Alil, B. Katavić, M. Ristić, D. Jovanović, M.Prokolab, S. Budimir, M. Kočić: Structural andmechanical properties of different hard weldedcoatings for impact plate for ventilation mill, Welding& Material Testing, Vol. 20, No. 3, pp. 7-11, 2011.13 th International Conference on Tribology – Serbiatrib’13 79

Serbian TribologySocietySERBIATRIB ‘1313 th International Conference onTribologyKragujevac, Serbia, 15 – 17 May 2013Faculty of Engineeringin KragujevacTOPOGRAPHIC AND ELECTROCHEMICAL TI6AL4VALLOY SURFACE CHARACTERIZATION IN DRY AND WETRECIPROCATING SLIDINGZinaida Doni 1 , Mihaela Buciumeanu 2 , Liviu Palaghian 31 “Dunarea de Jos” University of Galati, Romania, zinaida.doni@yahoo.com2 “Dunarea de Jos” University of Galati, Romania, mihaela.buciumeanu@ugal.ro3 “Dunarea de Jos” University of Galati, Romania, liviu.palaghian@ugal.roAbstract: This present paper shows the behavior of functional integrity state of a TI6Al4V alloy underreciprocating wear sliding conditions in a comparative way for two different counter materials, steel andceramic balls in dry and corrosive environment (3.5% NaCl). The surface integrity analysis of the dryreciprocating wear tests was based on the evolution of the roughness parameters with the applied load. Inthe case of reciprocating wear tests in corrosive environment the surface integrity analysis was based onelectrochemical parameters. Comparative analysis of the evolution of the roughness parameters with theapplied load shows a higher stability of the Ti6Al4V/Al 2 O 3 contact pair, while from the point of view of theelectrochemical parameters the tribological properties are worst than Ti6Al4V/steel ball contact pair.Keywords: reciprocating wear, light alloy, roughness, surface integrity.1. INTRODUCTIONThe required functional properties of the contactsurfaces under relative motion are closely related tothe surface integrity state. The modifications of thesuperficial layer during the machining process andsubsequently during service life have an importantrole in defining the surface integrity state. ThusGriffiths [1] proposed to define the surface integritystate based the mechanical, chemical, metallurgicaland topographical characteristics of the superficiallayer and their relation to the functionalperformance.Bellows and Tishler [2] stated that there are fivetypes of modifications of the superficial layerduring the machining process of a surface:mechanical, metallurgical, chemical, thermal andelectrical. These characteristics are changing duringthe service life. Besides other functionalcharacteristics, such as fatigue resistance,correlation between surface roughness parameters(Ra, Rz, Rk, etc.), and the wear rate has animportant role during the friction processes of thecontact surfaces under relative motion [3,4].Thus the surface integrity state can be definednot only due to the machining processes of thesurface, but also due to the operating processes.This can be called fuctional integrity state.Generally, it is intended that during the service lifeto maintain the same characterisics obtained fromthe machinig process or to have acceptablemodifications.The Ti6Al4V alloy is the most common titaniumalloy due to its physical, chemical and mechanicalproperties, such as: high strength, low density,excellent machinability and excellent corrosiveresistance. Some of the most widespreadapplications of these alloys include aircraft turbineengines, structural components and joints inaeronautics, structural elements in automotive andmaritime constructions, medical devices (dental andorthopedic implants).The spontaneously formation of a continue andstrong adherent oxide layer in air and also water (e.g. the marine environment, body fluids) providesextensive use of these alloys.Conventional ceramics materials such asalumina (Al 2 O 3 ) have excellent properties, such as:high hardness and good wear resistance, excellent80 13 th International Conference on Tribology – Serbiatrib’13

dielectric properties, high resistance to chemicalattack in presence of acids and alkali, high thermalconductivity, high resistance and stiffness, excellentformability and high purity. Due to these propertiesalumina is widely used in technical applications (e.g. automotive industry and in medical implants).The general properties of the two types ofmaterials (Ti6Al4V and Al 2 O 3 ) have leaded to theuse of these in applications where both wear andcorrosion resistant qualities are critical. In theseapplications, especially under relative motion andunder the action of external loads and in activechemical environments, it is mandatory to maintainthe surface integrity state during the service life.In the present paper is presented the evolution ofthe functional integrity state in a comparative wayfor two contact pairs (Ti6Al4V/Al 2 O 3 andTi6Al4V/steel ball) in dry and corrosiveenvironment.2. DRY WEAR AND TRIBOCORROSIONFORMULATIONThe analysis of the functional integrity stateunder dry wear conditions is based on the wellknown Archard's wear law [5]. It says that theamount of the material loss depends on theproperties of the contact surfaces, topographicalcharacteristics of the surfaces and test conditions.The most common form of Archard's equation isVSFn K(1)Hwhere V - the volumetric material loss of the body,K- the wear coefficient (it is dimensionless andalways less than unity), H - the hardness of thesofter body in contact, F n - the applied normal loadand S - the sliding distance.It has been analysed the amonut of material thatwas removed (wear loss) in the wear process. Thesurface analysis of the wear tests was based on theevolution of the depth of the wear track (based onthe material loss) with the applied load. Anotherparameter that was analyzed is the final topographyof the wear track by using the 3D roughnessparameters (Sa – Average Roughness, Sy - Peak-Peak Height, Sq - Root Mean Square Height, Sp -Maximum Peak Height, Sv - Maximum Pit Height).In the case of reciprocating wear tests incorrosive enviroment occur the degradationprocesss of surfaces by tribocorrosion. This processincludes the interaction between mechanical,chemical and electromechanical processes of wearthat lead to loss of weight by adding all theseeffects [6]:Wear=mechanical wear process+electrochemical(and/or chemical response) (2)This process includes the interaction ofcorrosion with: solid corrosive particles (debris),particles resulted due to abrasive processes, frettingprocesses, processes under biological solutionconditions, and triboxidation related to the mutualinteraction process under relative motion conditionsof the surfaces.Generally, oxide layers are formed after thecorrosive attack which protects the material fromfurther corrosive attack. But these oxide films aresusceptible to the tribological processes that willaccelerate the corrosion in these areas. The galvanicactivity that results between the worn and unwornsurfaces under electrochemical conditions [7], leadsto an anodic current I a , [8]:Ia kb1 2 Fn l f H 1 f0i d (3)where: k b – proportionality factor; l – sliding length;f– frequency of the reciprocating motion; F n –normal load; H– surface hardness; i - corrosioncurrent density; τ- time.Equation (3) can be written as [9]:Ip kbVs Fn H 1 2 Qp(4)where: V s – sliding speed; Q p - passivation chargedensity [5].Qpi d(5)0corrOn the other hand, electrochemical wear can bedetermined based on passivation currentIV p t Mn Fmol(6)where: t – time; M – molecular weight; ρ – density;n – valence; I – Faraday's constant.The corrosion rate [10,11] can be determinedbased on linear polarization and on the Stern-Geary’s equation [12] as follows:icorr a c2.31a c R p(7)where: β a şi β c cathodic and anodic Tafel slopes(figure 1); R r – polarisation resistance.Thus the tribocorrosion processes can beanalysed based on the evolution of theelectrochemical parameters β a , β c , I corr , E corr..If in the dry wear conditions the amount of thematerial loss is determined based on the rationF n /H, in the case of wear tests in the corrosionenvironment conditions the wear process isinfluenced by the electrochemical parameters.13 th International Conference on Tribology – Serbiatrib’13 81

Figure 1. Typical plot derived by the Tafelextrapolation method3. EXPERIMENTAL PROCEDURE3.1. MaterialsThe material contact pairs comparatively studiedin this work are: Ti6Al4V/Al 2 O 3 and Ti6Al4V/steelball.The mechanical properties of the materials usedin this study are presented in Table 1 and theirchemical composition is given in Tables 2 and 3.Table 1. Mechanical properties of Ti6Al4V alloyMaterialMechanical propertiesE σ 0.2 σ r ε r (%) HV(GPa) (MPa) (MPa)Ti6Al4V 115 989 1055 16,1 360Al2O3 300 - 2200 1100(96%)100Cr6 210 1034 1158 15 750Table 2. Chemical composition (weight %) of Ti6Al4ValloyElements Al V Fe Sn NiTi6Al4V 6.1 4.21 0.2 0.003 0,01Table 3. Chemical composition (weight %) of 100Cr6Elements C Si Mn P S Cr Mo100Cr6 0.93 0,15 0.25 0.026 0.15 1.35 0.103.2. Experimental test set-upReciprocating dry wear tests were carried out ona tribometer type CETR PRO 5003D. Theexperiments were carried out at a frequency of 1 Hzand the total stroke length of 3 mm during 3h, usinga reciprocating ball-on-plate configuration. Bearingsteel and Al 2 O 3 balls of 8 mm diameter were usedas counterpart. The experiments were carried out atthree normal loads 100, 120 and 140 N. Figure 4presents the schematic test configuration.Figure 2. Schematic specimen-pad contact testconfiguration (F n - normal load)The surface roughness parameters weredetermined using a 3D profilometer type CETRPRO500. The roughness parameters were obtainedby scanning a surface of 500x500 µm, in 200 pointon each line. Multiple measurements in differentareas on the wear track were carried out to obtainstable roughness values that can be representativefor entire wear track. The 3D images were analyzedby using a processing image soft - Scanning ProbeImagine Processor (SPIP). The 3D roughnessparameters used to describe the surface features areSa – Average Surface Roughness, Sy - Peak-PeakHeight, Sq - Root Mean Square Height, Sp -Maximum Peak Height, Sv - Maximum Pit Height.The initial surface roughness parameters studiedwere Sa =0.14 µm; Sy =3.96 µm; Sq =0.18 µm; Sp=3.34 µm; Sv =0.63 µm.In corrosive conditions the reciprocating wear testswere carried out under the same conditions, with thecontact pairs immersed in the electrolyte - an aqueoussolution of 3.5% NaCl (figure 3). The electrochemicalcharacteristics were obtained with a potentiostaticassembly.Figure 3. Schematic of the corrosion wear method4. EXPERIMENTAL RESULTS ANDDISCUSSIONSFigure 4 shows the profiles of the wear track atthe end of the test for Ti6Al4V/steel ball contactpair for the three normal loads used in this study.Figure 5 shows the evolution of the weight losswith increasing applied normal load for bothcontact pairs.Regarding the functional integrity in terms ofvariation with the applied load it is remarked theequivalent level of the weight loss of the steel ballat higher loads (120-140 N) (figs. 4 and 5).82 13 th International Conference on Tribology – Serbiatrib’13

parameter, Sv, was slightly influenced by themodification of the applied normal load.In the case of Ti6Al4V/Al 2 O 3 contact pair theroughness parameters that changed with increasingapplied normal load are Sa (figure 6a), Sq (figure6b) şi Sy (figure 6c).Figure 4. Evolution of the profiles with the appliednormal loadaabbFigure 5. Evolution of the weight loss with increasingapplied normal load: a Ti6Al4V/Al2O3 and b.Ti6Al4V/steel ballOn the other hand, in the case of Ti6Al4V/Al 2 O 3contact pair can be observed that the weight loss ofthe alumina ball increases with increasing appliednormal load (figure 5b).In the case of Ti6Al4V/Al 2 O 3 contact pair canbe observed a constant variation of the weight lossof the alumina ball counterpart (figure 5a) while inthe case of Ti6Al4V/steel ball to the steel ballcounterpart (figure 5b). Also, the weight loss washigher is the case of the steel ball counterpartcompared to the alumina ball counterpart.The evolution of the roughness parameters withincreasing applied normal load over the functionalintegrity (figure 6) shows that in the case ofTi6Al4V/steel ball contact pair the roughnesscdFigure 6. Evolution of the roughness parameterswith increasing applied normal load13 th International Conference on Tribology – Serbiatrib’13 83

Although the average roughness parameter Sa iscommonly used in the analysis of the evolution ofsurface topography, it does not allow tocharacterize the influence of roughness over thedegradation process of a surface or the load levelover the evolution of surface topography [13].The roughness parameter, Sv, (Maximum ValleyDepth) is closely related to load level. In this case(figure 6d) the constant evolution of this parameterfor Ti6Al4V/steel ball contact pair indicates a lowaffinity of the titanium alloy to the bearing steel.Similarly can be remarked the affinity of thetitanium alloy to ceramic materials (Al 2 O 3 ).The evolution of the roughness parameter Sq(Root Mean Square) gives indications about thedegree of flattening of the profile. It was observed aconstant flattening level in the case ofTi6Al4V/Al 2 O 3 contact pair. It gives indicationsabout the dimensional stability of this materialpairs, and consequently the possibility to use thisover a long period.abFigure 7. Evolution of the electrochemical parameters: a. Ti6Al4V/Al 2 O 3 and b. Ti6Al4V/steel ballThe evolution of the roughness parameter Sy(Peak-Peak Height) (figure 6c) refers to theinterdependence between surface roughness and itsimage. This is based mainly on the functional84 13 th International Conference on Tribology – Serbiatrib’13

dependence of roughness height and grey levelimage of surface, which means that the higher partsof the asperities correspond to higher intensitypixels. Also this parameter indicates the stability ofthe surface conditions with increasing normal load.This leads to a longer stability of the initial surfaceconditions during the service life of Ti6Al4V/Al 2 O 3contact pair. The evolution of previous mentionedroughness parameters shows that in terms ofsurface quality and functional maintenance theTi6Al4V/Al 2 O 3 contact pairs present a higherfunctional integrity level.Figure 7a and b shows the evolution of theelectrochemical parameters for both contact pairs.The functional integrity of the tests in corrosiveenvironment for both material pairs indicateddifferences in the evolution of the electrochemicalparameters that characterize the electrochemicalstate of the contact surfaces.Thus, in the case of Ti6Al4V/Al 2 O 3 contact pair(figure 7a) at low load levels (100N) theelectrochemical parameters E corr and I corr do notchange much with time. Parameters β a and β c havesignificant variation, with an increasing tendencyfor β c and a decreasing tendency for β a . Theincrease of the applied load changes the evolutionof those parameters with time. These will have anoscillatory tendency. In the case of Ti6Al4V/steelball can be observed a more pronounced oscillatoryevolution of all electrochemical parameters withtime (figure 7b) at higher load levels than forTi6Al4V/Al 2 O 3 contact pair (except for theevolution of parameter β c ).The analysis from the functional integrity pointof view based on electrochemical criterionsindicates a higher integrity level of Ti6Al4V/Al 2 O 3contact pair.Figure 8 shows the evolution of coefficient offriction with the sliding distance in the case of dryreciprocating wear tests.abFigure 9. Evolution of the coefficient of friction with thesliding distance for reciprocating wear tests in corrosiveenvironment: a. /Ti6Al4V/Al2O3 and b. Ti6Al4V/steelballThe variation of COF for dry reciprocating wearconditions (figure 8) is similar for both contactpairs used in this study. The COF has a slighthigher value in the case of Ti6Al4V/Al 2 O 3 thanTi6Al4V/steel ball contact pairs.Also in the case of reciprocating wear incorrosive environment the variation of COF issimilar for both contact pairs. The COF was notinfluenced by the load level. The Ti6Al4V/steelball (figure 9b) contact pair showed a more stableevolution of the COF at a low level than in the caseof Ti6Al4V/Al 2 O 3 contact pair (figure 9a).5. CONCLUSIONFigure 8. Evolution of the coefficient of friction with thesliding distance for dry reciprocating wear testsFigure 9 shows the evolution of coefficient offriction with the sliding distance in the case ofreciprocating wear tests in corrosive environmentfor both contact pairs.This present paper showed the behavior of aTI6Al4V alloy under reciprocating wear slidingconditions in a comparative way for two differentcounter materials, bearing steel and ceramic balls(Al 2 O 3 - 99.6%) in dry and corrosive environment(an aqueous solution of 3.5% NaCl). It aimed tohighlight the tribological characteristics that showsinvariability during the test and provides a highlevel of functional integrity of the surface.The conclusions drawn from this work are asfollows:13 th International Conference on Tribology – Serbiatrib’13 85

- in dry condition the Ti6Al4V/Al 2 O 3 contactpair showed a high functional integrity degree ofthe surfaces in terms of surface quality,characterized by roughness parameters Sa, Sq andSy, while for the Ti6Al4V/steel ball based on theroughness parameter Sv;- in the case of Ti6Al4V/steel ball contact pair abetter functional integrity (evaluated based on theweight loss) occurred for higher applied loads thanin the case of lower loads;- from the point of view of the electrochemicalbehavior a higher functional integrity occurs in thecase of Ti6Al4V/Al 2 O 3 contact pair at lowerapplied loads (assessed through parameters E corrand I corr );- the electrochemical parameters forTi6Al4V/Al 2 O 3 contact pair are at a lower levelthan those of the Ti6Al4V/steel ball contact pair;- the evolution of the roughness parameters andthe structural affinity between TI6Al4V alloy andthe bearing ball conduced to a higher functionalintegrity level from the point of view of theevolution of COF by comparison to Ti6Al4V/Al 2 O 3contact pair.REFERENCES[1] B. Griffiths, Manufacturing Surface Technology-Surface Integrity and Functional Performance,Prenton Press, 2001.[2] G. Bellows, D. N. Tisher, Introduction to surfaceIntegrity Report TM70-974, Cincinnati: GeneralElectric Co., 1970.[3] N. Aris, K. Cheng, Characterization of the surfacefunctionality on precision machined engineeringsurfaces, International Journal of AdvancedManufacturing, Vol. 38, pp. 402-409, 2008.[4] I.S. Jawahir, E. Brinksmeier, R. M'Saoubi, D.K.Aspinwall, J.C. Outeiro, D. Meyer, D. Umbrello,A.D. Jayal, Surface integrity in material removalprocesses: Recent advances, CIRP Annals -Manufacturing Technology, Vol. 60, pp. 603–626,2011.[5] J.F. Archard, Wear theory and mechanisms. In:M.B. Peterson, W.O. Winer, Wear ControlHandbook. ASME, 1980.[6] R. J. K. Wood, Tribo-corrosion of coatings: areview, J. Phys. D: Appl. Phys., Vol. 40, pp. 5502–5521, 2007.[7] P. Ponthiaux, F. Wenger, D. Drees, J.P. Celis,Electrochemical techniques for studyingtribocorrosion processes, Wear, Vol. 256, pp. 459–468, 2004.[8] S. Mischler, S. Debaud, D. Landolt, Wearacceleratedcorrosion of passive metals intribocorrosion systems, Journal of theElectrochemical Society, Vol. 145, pp. 750-758,1998.[9] J. Perret, Modélisation de la tribocorrosion d'aciersinoxydables dans l'eau à haute pression et hautetempérature, Thèse No 4727 (2010) ÉcolePolytechnique Fédérale De Lausanne.[10] S. Feliu, J. A. Gonzalez, S. Miranda, Possibilitiesand Problems of in Situ Techniques for MeasuringSteel Corrosion Rates in Large Reinforced ConcreteStructures, Corrosion Science, Vol. 47, pp. 217-238,2005.[11] H. H. Uhlig, R. W. Revie, Corrosion and CorrosionControl, 3rd Edition, John Wiley and Sons, NewYork, 1985.[12] M. Stern, A. L. Geary, ElectrochemicalPolarization, A Theoretical Analysis of the Shape ofPolarization Curves, Journal of ElectrochemicalScience, Vol.104, pp. 56-63, 1957.[13] V. Mereuta, M. Buciumeanu, L. Palaghian, 3DRoughness Parameters as Factors in Determiningthe Evolution of Effective Stress ConcentrationFactors in Fatigue Processes, Applied Mechanicsand Materials Vol. 248, pp 504-510, 2013.86 13 th International Conference on Tribology – Serbiatrib’13

Serbian TribologySocietySERBIATRIB ‘1313 th International Conference onTribologyKragujevac, Serbia, 15 – 17 May 2013Faculty of Engineeringin KragujevacFRICTION COEFFICIENT OF UHMWPE DURING DRYRECIPROCATING SLIDINGFatima Živić 1 , Miroslav Babić 1 , Slobodan Mitrović 1 , Dragan Adamović 1 , Svetlana Pelemis 21 Faculty of Engineering, University of Kragujevac, Serbia, zivic@kg.ac.rs, babic@kg.ac.rs, boban@kg.ac.rs, adam@kg.ac.rs2 Faculty of Technology in Zvornik, University of East Sarajevo, Bosnia and Herzegovina, alannica@gmail.comAbstract: This paper deals with the friction coefficient behaviour during dry reciprocating sliding ofUHMWPE in contact with alumina (Al2O3), within a range of velocities typical for hip implants. Five valuesof normal force (100 - 1000 mN) and three values of sliding speed (4 - 12 mm/s) have been observed. Realtime diagrams of the friction coefficient as a function of the sliding cycles were recorded for each test.Dynamic friction coefficient curves exhibited rather uniform behavior for all test conditions. Somewhatlarger values of friction coefficient could be observed during the running-in period in case of low loads(100-250 mN) and the lowest velocity (4 mm/s). In case of high loads and speeds, friction coefficient reachedsteady state values shortly after the beginning of the test.Keywords: UHMWPE, Dynamic friction coefficient, Reciprocating sliding.1. INTRODUCTIONUltra-high molecular weight polyethylene(UHMWPE) is a unique polymer with outstandingphysical and mechanical properties. Most notableare its chemical inertness, lubricity, impactresistance, and abrasion resistance. The first clinicalapplication of UHMWPE biomaterials started in1962 and continued with astonishing speed throughthe three decades of the clinical history (1962–1997), with a few clinically relevant innovationsoccurred beyond the removal of calcium stearateand changes in sterilization practice. Radiationcrosslinked UHMWPE materials had been recentlyintroduced to clinical practice around a year 2000.Today, second generation of radiation crosslinkedmaterials are in clinical use, and vitamin Estabilized UHMWPE has emerged as a new,internationally standardized biomaterial.At the November 2012 ASTM Meeting inAtlanta, GA, the UHMWPE working groupconsidered revisions to four UHMWPE-relatedstandards, including F648 (unfilled UHMWPEhomopolymer), F2102 (FTIR analysis ofoxidation), F2565 (Guide for CrosslinkedUHMWPEs), and a new standard for small punchtesting of medical polymers, including UHMWPE,based on F2183. The next ASTM meeting of theUHMWPE working group will be in May, 2013.All existing data on UHMWPE indicate that forvery elderly patients, artificial joints incorporatingconventional UHMWPE will continue to be used.On the other hand, the more recently-introducedalternative bearing technologies, includingcrosslinked UHMWPE, should provide the greatestbenefit to young patients (less than 60 years in age)who lead an active lifestyle and who need a totalhip replacement. For patients in need of kneearthroplasty, shoulder arthroplasty, or total discreplacement, conventional UHMWPE continues toprevail as the polymeric bearing material of choice.Polymers are large molecules synthesized fromsmaller molecules, called monomers. Plastics arepolymers that are rigid solids at room temperatureand generally contain additional additives.UHMWPE has been used as a bearing surface, intotal joint prostheses, for more than 45 years.Each year, about 2 million joint replacementprocedures are performed around the world, and themajority of these joint replacements incorporateUHMWPE. Despite the success of these restorativeprocedures, orthopedic and spine implants haveonly a finite lifetime. Wear and damage of theUHMWPE components has historically been one of13 th International Conference on Tribology – Serbiatrib’13 87

the factors limiting implant longevity. In the past 10years, highly crosslinked UHMWPE biomaterialshave shown dramatic reductions in wear in clinicaluse around the world. The orthopedic communityawaits confirmation that these reductions in wearwill be associated with improved long-termsurvival, as expected.UHMWPE has been used for fabricating one ofthe bearing components in various arthroplasties,such as acetabular cups, acetabular cup liners, tibialinserts, etc. [1, 2]. Total joints with componentsmade of this material can function for more thantwenty years if they are well designed and wellimplanted. Components made of UHMWPE haveperformed admirably in vivo. The only majorconcern is wear and the effect of the wear particleson the in vivo longevity of the prosthesis. Someapplications of UHMWPE in biomedical area areshown in Fig. 1.Figure 1. Application of UHMWPE [1]: a) cup; b), c),d) liners; e) total knee components; f) shoulderprosthesis system components.Polyethylene contains the chemical elementscarbon and hydrogen. Polyethylene is createdthrough polymerization of ethene (Fig. 2), formingan extremely long, chained molecule calledgenerally polymer. At a molecular level, the carbonbackbone of polyethylene can twist, rotate, and foldinto ordered crystalline regions.Figure 2. Left: The repeating unit of polyethylene;Right: Granulated polyethylene.At a supermolecular level, the UHMWPEconsists of powder (also known as resin or flake)that must be consolidated at elevated temperaturesand pressures to form a bulk material. Furtherlayers of complexity are introduced by chemicalchanges that arise in UHMWPE due to radiationsterilization and processing.The mechanical properties of polyethyleneimprove slowly with rising molecular weight of theproduct. A dramatic change in mechanicalproperties, however, appears when molecularweight of the polyethylene molecule exceeds onemillion. This appears when more than 35000ethylene groups are added together. Such product iscalled Ultra High Molecular Weight PolyEthylene.The molecular weight of the UHMWPE currentlyused in total joint components varies between 4 to 6millions. Every such UHMWPE molecule iscomposed of 160 to 215 000 ethylene groups.The molecular chain of UHMWPE can bevisualised as a tangled string of spaghetti over akilometer long. Because the chain is not static, butimbued with internal (thermal) energy, themolecular chain can become mobile at elevatedtemperatures. When cooled below the melttemperature, the molecular chain of polyethylenehas the tendency to rotate about the C-C bonds andcreate chain folds. This chain folding, in turn,enables the molecule to form local ordered,sheetlike regions known as crystalline lamellae.These lamellae are embedded within amorphous(disordered) regions and may communicate withsurrounding lamellae by tie molecules. Thelamellae are on the order of 10–50 nm in thicknessand 10–50 µm in length. UHMWPE has a white,opaque appearance at room temperature. Attemperatures above the melt temperature of thelamellae, around 137°C, it becomes translucent.UHMWPE exhibits the composite nature due tointerconnected network of amorphous andcrystalline regions.Generally speaking, many polymers undergothree major thermal transitions: the glass transitiontemperature (Tg), the melt temperature (Tm), andthe flow temperature (Tf). The glass transition (Tg)is the temperature below which the polymer chainsbehave like a brittle glass. Below Tg, the polymerchains have insufficient thermal energy to slide pastone another, and the only way for the material torespond to mechanical stress is by stretching (orrupture) of the bonds constituting the molecularchain. In UHMWPE, the glass transition occursaround 120°C.UHMWPE shows two key features, the first oneis the peak melting temperature (Tm), which occursaround 137°C and corresponds to the point at whichthe majority of the crystalline regions have melted.The melt temperature reflects the thickness of thecrystals as well as their perfection. Thicker andmore perfect polyethylene crystals will tend to meltat a higher temperature than smaller crystals.As the temperature of a semicrystalline polymeris raised above the melt temperature, it mayundergo a flow transition and become liquid.88 13 th International Conference on Tribology – Serbiatrib’13

Polyethylenes with a molecular weight of less than500,000 g/mol can be observed to undergo such aflow transition (Tf). However, when the molecularweight of polyethylene increases above 500,000g/mol, the entanglement of the immense polymerchains prevents it from flowing. UHMWPE doesnot exhibit a flow transition for this reason.UHMWPE is a linear, low-pressure,polyethylene resin. It has both the highest abrasionresistance and highest impact strength of anyplastic. Combined with abrasion resistance andtoughness, the low coefficient of friction ofUHMWPE yields a self-lubricating, non-sticksurface. Static and dynamic coefficients aresignificantly lower than steel and most plasticmaterials. Elastic modulus of UHMWPE isapproximately 0.69 GPa. ASTM F648-00 definesstandard specification for Ultra-High-Molecular-Weight Polyethylene powder and fabricated formfor surgical implants. To date it has proven to bethe best polymer material for use in total joints.Along with the extensive application ofUHMWPE, the understanding of polymer tribologyis becoming increasingly important. Many authorshave investigated different aspects of tribologicalperformance of UHMPWE [2-7]. The structuralfactors associated with surface mechanicalproperties (crosslinking, oxidation state, localorientation of polymer, crystallinity, etc.) can behighly variable and localized and may vary onmicron spatial scales or smaller [8]. Therelationship between UHMWPE mechanicalproperties and the in-vivo performance of afabricated form has not been determined. Whiletrends are apparent, specific property-polymerstructure relationships are not well understood. Themechanical properties are subject to variation as thefabrication process variables (such as temperature,pressure, and time) are changed. [ASTM F-648(2000)].Reciprocating sliding at different test devicesand from different aspects has been a subject ofinvestigations [9-11]. Different approaches forimprovement of existing UHMWPE materials havebeen tried. [10] investigated effects of nitrogen ionirradiation on tribological properties. [9]investigated friction and wear behavior of ultrahighmolecular weight polyethylene as a function ofpolymer crystallinity. [2, 7, 8, 13, 14, 15]investigated crosslinking and material behaviorwith different approaches to crosslinking.This paper deals with friction coefficientbehaviour during dry reciprocating sliding ofUHMWPE in contact with alumina (Al2O3), withina range of velocities typical for hip implants. Fivevalues of normal force and three values of slidingspeed have been observed.2. MATERIALS AND TRIBOLOGICAL TESTPolished rectangular flat UHMWPE sampleswere used for tests, supplied by the companyNarcissus Ada, Serbia. Sliding tests were done atball-on-flat configuration of CSM Nanotribometerin dry conditions. Alumina was used as a ballmaterial (diameter, 1.5 mm), since it is extremelyhard and chemically inert. Alumina is frequentlyused in combination with UHMWPE in artificialhip joints. Duration of each test was 3000 cycles (1cycle, 1.6 mm), which is enough to reach stabilesteady state of friction coefficient. During the test,the dynamic friction coefficients were recorded inreal time, using the built-in TriboX 2.9.0 software.Five values of normal force (100 - 1000 mN) andthree values of sliding speed (4 - 12 mm/s) havebeen tested. Maximum elastic contact stress(according to applied normal loads) were calculatedby Hertz method and compared to test conditions.Characteristics of conducted tribological tests aregiven in Table 1.Table 1. Tribological parameters.Normal load values, F nMaximum elastic contact stress(according to applied normalloads)Maximum linear speed values, v3. RESULTS AND DISCUSSION100 mN, 250 mN, 500 mN,750 mN, 1000 mN28.5 MPa, 38.6 MPa, 48.7MPa, 55.7 MPa, 61.3 MPa4 mm/s; 8 mm/s; 12 mm/sReal time diagrams of the friction coefficient as afunction of the sliding cycles (sliding distance) wererecorded for each test. Friction coefficient curve is ofsinusoid shape, whereat the opposite directions aremarked with + and - sign, denoting coefficient offriction in two different directions of sample moving.Good agreement with reported values of frictioncoefficient was obtained [6, 16, 17, 18].Dynamic friction coefficient curves exhibitedrather uniform behavior for all test conditions(Fig.3). Somewhat larger values of frictioncoefficient could be observed during the running-inperiod in case of low loads (100-250 mN) and thelowest velocity (4 mm/s), as shown in Fig. 3a. Incase of high loads and speeds, friction coefficientreached steady state values shortly after thebeginning of the test. Maximum contact pressuresin these tests were approximately from 30 - 60MPa, representing high contact stresses exhibited inhip/knee implants. Especially extreme loadingconditions are present in the knee implant system.Similar behavior (short running-in phase for thefriction coefficient) was also reported by otherauthors [18].13 th International Conference on Tribology – Serbiatrib’13 89

nanotribometer. Comparative diagrams of variationof the average values of the dynamic frictioncoefficient, with load and sliding speed, are givenin Fig. 4. It can be seen from presented diagramsthat the sliding speed exhibited no significantinfluence on the friction coefficient. Load increaseproduced very slight decrease of the frictioncoefficient, for all tested conditions.4. CONCLUSIONThis research showed that UHMWPE exhibitslow dynamic friction coefficient (average valuearound 0.1 for all tests) under dry conditions, incontact with alumina. It also showed that low loadsleaded to a bit longer running-in time, but for alltests steady friction state was achieved andmaintained throughout the test with very shortrunning-in periods. Sliding speed change showedno influence on the dynamic friction coefficient.Figure 3. Dynamic friction coefficient curve during drysliding: a) v=4 mm/s, F N =100mN; b) v=12 mm/s,F N =1000mN.ACKNOWLEDGMENTSThis paper is supported by TR-35021 project,financed by the Ministry of Education, Science andTechnological Development of the Republic ofSerbia.REFERENCESFigure 4. a) Friction coefficient as a function of thenormal load; b) Friction coefficient as a function of thenormal load and sliding speed.An average value of the dynamic coefficient offriction (denoted by 'friction coefficient' further inthe text) was calculated, for all test conditions, for asteady state period of friction, as the root meansquare function using the raw data obtained by the[1] S.M. Kurtz: UHMWPE Biomaterials Handbook,Elsevier, London, 2009[2] G. Lewis: Properties of crosslinked ultra-highmolecular-weightpolyethylene, Biomaterials, Vol.22, pp. 371-401, 2001.[3] S.W. Zhang: State-of-the-art of polymer tribology,Tribology International, Vol. 31, pp. 49-60, 1998.[4] P.A. Williams, I.C. Clarke: Understandingpolyethylene wear mechanisms by modeling ofdebris size distributions, Wear, Vol. 267, pp. 646-652, 2009.[5] S. Ge, S. Wang, N. Gitis, M. Vinogradov, J. Xiao:Wear behavior and wear debris distribution ofUHMWPE against Si3N4 ball in bi-directionalsliding, Wear, Vol. 264, pp. 571-578, 2008.[6] V. Banchet, V. Fridrici, J.C. Abry, Ph. Kapsa: Wearand friction characterization of materials for hipprosthesis, Wear, Vol. 263, pp. 1066-1071, 2007.[7] A. Kilgour, A. Elfick: Influence of crosslinkedpolyethylene structure on wear of jointreplacements, Tribology International, Vol. 42,No.11-12, pp. 1582-1594, 2009.[8] J.L. Gilbert, I. Merkhan: Rate effects on themicroindentation-based mechanical properties ofoxidized, crosslinked, and highly crystallineultrahigh-molecular-weight polyethylene, Journal ofBiomedical Materials Research Part A, Vol. 71A,No.3, pp. 549–558, 2004.90 13 th International Conference on Tribology – Serbiatrib’13

[9] K.S. Kanaga Karuppiah, A.L. Bruck, S.Sundararajan, J. Wang, Z. Lin, Z.H. Xu, X. Li:Friction and wear behavior of ultra-high molecularweight polyethylene as a function of polymercrystallinity, Acta Biomaterialia, Vol. 4, No.5, pp.1401-1410, 2008.[10] L. Fasce, J. Cura, M. del Grosso, G.G. Bermúdez, P.Frontini: Effect of nitrogen ion irradiation on thenano-tribological and surface mechanicalproperties of ultra-high molecular weightpolyethylene, Surface and Coatings Technology,Vol. 204, No.23, pp. 3887-3894, 2010.[11] M. Flannery, T. McGloughlin, E. Jones, C.Birkinshaw: Analysis of wear and friction of totalknee replacements: Part I. Wear assessment on athree station wear simulator, Wear, Vol. 265, No.7-8, pp. 999-1008, 2008.[12] M. Flannery, E. Jones, C. Birkinshaw: Analysis ofwear and friction of total knee replacements part II:Friction and lubrication as a function of wear,Wear, Vol. 265, No.7-8, pp. 1009-1016, 2008.[13] J.L. Gilbert, J. Cumber, A. Butterfield: Surfacemicromechanics of ultrahigh molecular weightpolyethylene: Microindentation testing,crosslinking, and material behavior, Journal ofBiomedical Materials Research, Vol. 61, pp. 270–281, 2002.[14] L.A. Pruitt: Deformation, yielding, fracture andfatigue behavior of conventional and highly crosslinkedultra high molecular weight polyethylene,Biomaterials, Vol. 26, No.8, pp. 905-915, 2005.[15] C. Zhu, O. Jacobs, R. Jaskulka, W. Köller, W. Wu:Effect of counterpart material and water lubricationon the sliding wear performance of crosslinked andnon-crosslinked ultra high molecular weightpolyethylene, Polymer Testing, Vol. 23, No. 6, pp.665-673, 2004.[16] H.J. Cho, W.J. Wei, H.C. Kao, C.K. Cheng, Wearbehavior of UHMWPE sliding on artificial hiparthroplasty materials, Materials Chemistry andPhysics, Vol. 88, No. 1, pp.9-16, 2004.[17] M.P. Gispert, A.P. Serro, R. Colaco, B. Saramango:Friction and wear mechanisms in hip prosthesis:Comparison of joint materials behaviour in severallubricants, Wear, Vol. 260, pp. 149-158, 2006.[18] S.K. Young, M.A. Lotito, T.S. Keller: Frictionreduction in total joint arthroplasty, Wear, Vol. 222,No. 1, pp. 29-37, 1998.13 th International Conference on Tribology – Serbiatrib’13 91

Serbian TribologySocietySERBIATRIB ‘1313 th International Conference onTribologyKragujevac, Serbia, 15 – 17 May 2013Faculty of Engineeringin KragujevacTHE POTENTIAL OF MAGNESIUM ALLOYS ASBIOABSORBABLE / BIODEGRADABLE IMPLANTS FORBIOMEDICAL APPLICATIONSFatima Zivic 1 , Nenad Grujovic 1 , Geetha Manivasagam 2 , Caroline Richard 3 , Jessem Landoulsi 4 ,Vojislav Petrovic 51 Faculty of Engineering, University of Kragujevac, Serbia, zivic@kg.ac.rs, gruja@kg.ac.rs2 Vellore Institute of Technology, Centre for Materials Engineering, Tamil Nadu, India, geethamanivasagam@vit.ac.in3 François Rabelais University of Tours, Laboratory of Mechanics and Rheology, EA 2640-France, caroline.richard@univ-tours.fr4 Pierre & Marie Curie University of Paris, Laboratory of Surface Reactivity, UMR 7197-France, jessem.landoulsi@upmc.fr5 AIMME - Instituto Tecnológico Metalmecánico, Unidad de Ingeniería de Producto, Valencia, Spain, vpetrovic@aimme.esAbstract: The potential of magnesium alloys as bioabsorbable / biodegradable implants for biomedicalapplications has been extensively studied as emerging direction. This paper gives a review of current topicsin this field. Research activities related to biomedical magnesium alloys have been pursued in two maindirections, orthopedic and cardiovascular implants, by investigating different aspects of alloying systemdesign, novel structures, degradation rate control, and surface modification methods. Magnesium alloys arecurrently considered for applications as load-bearing implant devices such as plates, screws and pins forrepairing bone fracture. Highly important direction of research is degradable coronary stents. Degradablevessel stents promote stable vessel regeneration, unlike permanent stents. Different combinations of alloyingelements have been investigated in order to decrease corrosion rate.Tribological issues are also importantfor understanding of different phenomenon related to prolongation of Mg alloys corrosion degradationtime/rate, such as tribocorrosion, corrosion fatigue, and fatigue crack growth behavior.Keywords: Mg alloys, Bioabsorbable / Biodegradable implants.1. INTRODUCTIONThe beginning of the resorbable implantsconcepts is related to using polymers withcontrolled dissolution rates: polylactides andpolyglycolides, back in 1970s [1]. But the problemassociated with use of polymers is their mechanicalproperties, where metals have better characteristicsand represent the promising field for advancements.The history of biodegradable magnesium (Mg)implants started shortly after the discovery ofelemental magnesium by Sir Humphrey Davy in1808 [2]. It is supposed that the pure magnesiumwires were used as ligatures to stop bleedingvessels of three human patients in 1878 [2]. Theyelaborated corrosion induced degradation propertiesof pure magnesium and concluded that corrosionrate depended on the wire size. From those firstattempts, many other solutions and ideas were tried,because magnesium has been recognised as thepromising material for efficient degradableimplants. Today, in vitro and in vivo study dataexists, as well as some clinical trials data, but not soextensively present as for other biomedical metalmaterials, because degradable materials and Mgalloys are still having many unresolved issues ifcompared to the development of Ti biomedicalalloys. Even today, several important drawbacks ofthe technology and material need to be resolvedbefore its wide application in clinical practice.Magnesium is the seventh most abundantelement in earth’s crust (2% of the total mass) andalso essential and major constituent element ofhuman body, nontoxic and biocompatibleaccordingly [3]. It belongs to the group of alkalineearth metals and cannot be found in elemental formin nature, but only in chemical combinations, sincebeing highly reactive. From aspects of the Mgalloys production, important mineral forms are:magnesite MgCO3 (27% Mg), dolomite92 13 th International Conference on Tribology – Serbiatrib’13

MgCO3•CaCO3 (13% Mg), and carnalliteKCl•MgCl2•6H2O (8% Mg), as well as sea water,which contains 0.13% Mg or 1.1 kg Mg per m3(3rd most abundant among the dissolved mineralsin sea water) [3]. However, Mg alloys are stilllacking wide application due to several reasons, oneof which is rather high price of base material andregarding degradable implants, too rapid corrosionrate when implanted, especially in solutionscontaining Cl -1 . In addition, there are certainlimitations related to the processing temperaturesand production protocols.The usage of magnesium (Mg) has historicallybeen limited by relatively high cost of productionand associated energy costs. However, Mg marketwill steadily increase, mainly due to the low costproduction in China. Magnesium is recovered byelectrolysis of molten anhydrous MgCl2, bythermal reduction of dolomite, or by extraction ofmagnesium oxide from sea water. The globalproduction of roughly 436,000 t (1997) is coveredby melt electrolysis to 75% and by thermalreduction to 25%. Cast magnesium alloys dominate85-90% of all magnesium alloy products, with Mg-Al-Zn system being the most widely used. Rareearth alloying additions increase cost and are ofuncertain supply. US and Canada dominatedmagnesium production during the 1990s, however,since the late 90s, China become the mainproducer. Today, China dominates the worldproduction because of the relatively low operatingcosts. In general, electrolytic producers in the westhave been replaced by Chinese pyrometallurgicalproduction via the Pidgeon process (Pidgeon,1944). For example, the capital cost for AustraliaMagnesium (AM) Process using an electrolyticroute was estimated to be $10,000/tonne Mg, whilethe capital cost for the Pidgeon process wasestimated to be $3,000/tonne Mg in 2008.2. MAGNESIUM ALLOYSMagnesium alloys are standardized by ASTMnorm and they are marked with letters (A, B, C,etc.), indicating main alloying elements, followedby the rounded figures of each weight in percentageterms. The alloy AZ91D, for example, is an alloywith a rated content of 9% aluminium (A) and 1%zinc (Z) [3]. The corresponding DIN specificationwould be MgAl9Zn1. The most common alloyingelements are given in Table 1.Different alloying elements influence theproperties of the pure magnesium, depending on thewanted characteristics. The main mechanism forimproving the mechanical properties isprecipitation hardening and/or solid-solutionhardening. One of the most important alloyingelements is aluminium (Al) which increase tensilestrength, by forming the intermetallic phaseMg17Al12. The use of rare earth elements (e.g. Y,Nd, Ce) has become significant, especially fordesigning the medical grade Mg alloy fordegradable implants, since they impart a significantincrease in strength through precipitationhardening. But almost all elements used so far foralloying mostly increase susceptibility to corrosion.Review of commercially produced Mg alloys todayis given in Table 2 [3].Table 1. ASTM codes for magnesium’s alloying elements [3].AbbreviationletterAlloyingelementAbbreviationletterAlloyingelementA aluminium N nickelB bismuth P leadC copper Q silverD cadmium R chromiumE rare earths S siliconF iron T tinH thorium W yttriumK zirconium Y antimonyL lithium Z zincMmanganeseTable 2. Chemical composition (in weight %; Mg is thebalance; Cu, Ni in general

Energy absorption during impact or loading isinherent property of the foams. Fig. 1 shows stress–strain behavior of the foams in compression, ingeneral [1]. From Fig. 1, it can be seen that threezones exists: I. Elastic region: deformation of the porewalls; II. A plateau of nearly constant flow stress andlarge strain (10–50%) and III. Densification regionwith steep increase of flow stress where the plasticdamage occurs. Relatively wide region of constantflow stress during compressive loading explains thefact that while foams are in this interval, anyincreasing strain hardly entails increasing stress. Iftension is observed, the stress–strain behavior offoams is approximately similar to ductile metals(curve b in Fig. 1).Figure 1. Stress - strain behaviour of foams.3. MAGNESIUM ALLOYS ASDEGRADABLE IMPLANTSThe application of Mg and its alloys fordegradable implants started with ligatures for bloodvessels (Huse, 1878, pure magnesium) and plates,arrows, wire, sheets, rods (Payr, 1905, puremagnesium) [2]. Mg alloys have been tried indifferent medical areas, such as: pure Mg for band,suture from woven Mg wires, fusiform pins (in1940); Mg–Al2%-wt. pure magnesium wires forclotting aneurysms (dog studies in 1951); Mg–Al2%-wt. for intravascular wires (rat studies in1980) [2]. The potential of magnesium alloys asbioabsorbable / biodegradable implants forbiomedical applications has been extensively studiedas emerging direction. Research activities related tobiomedical magnesium alloys have been pursued intwo main directions, orthopedic and cardiovascularimplants, by investigating different aspects ofalloying system design, novel structures, degradationrate control, and surface modification methods.Magnesium alloys are currently considered forapplications as load-bearing implant devices suchas plates, screws and pins for repairing bonefracture. Other metals currently used for boneimplants, such as stainless steels and titaniumalloys, have elastic modulus that are much higherthan natural bone, leading to unwanted stressshielding. The elastic modulus of magnesium andmany magnesium alloys are much closer to bone.The advantage of Mg alloys is favorable elasticresponse during loading (such as shown in Fig. 1).Also, the second surgery is avoided due to thedegradation of the implant after its function in thebody is finished. For example, compared with poly-96L/4D-lactide, the magnesium alloys AZ31 andAZ91 enhanced the osteogenesis response andincrease newly formed bone [4]. Investigationsshowed that Mg–6Zn, Mg–Ca and Mg–Mn–Znalloys gradually degrade within a bone and hadgood biocompatibility both in vitro and in vivo [4].Highly important direction of research isdegradable coronary stents. Degradable vesselstents promote stable vessel regeneration, unlikepermanent stents [5, 6, 7]. However, as a vesseldefect gets larger, stronger and degradablematerials are paramount for stable vesselregeneration. Vessel scaffolding is necessary onlyfor a certain, limited time, than the permanent stenthas no known advantage. A stent is a miniaturemesh tube, made of a biocompatible metal,biodegradable metal or polymer, placed inside of ablood vessel (cardiovascular, neurovascular andperipheral blood vessels) or a natural conduit(gastrointestinal, urinary and biliary tracts). Thestent acts as a scaffold, pushing against the internalwalls of the conduit/vessel to open a blocked areaand thereby enables natural flow and prevents thevessel from collapsing, narrowing or closing. Stentsdiffer greatly in their design, dimensions andmaterial, depending on application. Coronary stentsare now the most commonly implanted medicaldevice for angioplasty, with more than 1 millionimplanted annually. Currently used metallic stentspermanently remain in the artery and are associatedwith limitations such as continued mechanicalstress, transfer to the tissue, and continuedbiological interaction with the surrounding tissue.Also, within 6 months, 30-35% patients suffer fromrestenosis. They are associated with late stentthrombosis and artifacts when non-invasivetechnologies such as MRI and MSCT are used. Thestent presence is required for a period of 6 - 12months during which arterial remodelling andhealing is achieved. After this period the stentpresence within the blood vessel cannot provideany beneficial effects. With the development ofbiodegradable implants, the concept of biomaterialshas shifted from purely mechanical replacementdevices towards true biological solutions. Bioabsorbablestents (Fig. 2) aim to mechanicallyprevent vessel recoil without the permanentpresence of an artificial implant. The advantages ofbioabsorbable stents are to leave no stent behind,94 13 th International Conference on Tribology – Serbiatrib’13

fully compatible with MRI/MSCT imaging, and arenot associated with late stent thrombosis.Figure 2. Bioabsorbable magnesium stent (Biotronik,Berlin, Germany) [6].Magnesium (Mg) alloys which has beendeveloped in the last ten years showed greatpotential in cardiovascular applications wheretemporary stent is required. Biotronik Mg AlloyBalloon expanding stent with a delivery catheterhas been clinically tested to some extent [7].Different combinations of alloying elementshave been investigated in order to decreasecorrosion rate. The addition of aluminium (Al) andrare earth (RE) elements has been reported toincrease its strength and improve corrosionresistance. But Al could cause nerve toxicity andrestraining growth to human body. Alloycontaining RE is very expensive. In some alloys,the low cost Ca has been used as alloying elementand Mg–Ca alloy exhibited increased corrosionresistance. The addition of Zn into Mg–Ca alloysincreases the tensile strength, ductility, hardnessand the kinetics of age hardening. Commercial Mgalloys such as WE43 (Mg–Y–RE–Zr), AZ91 (Mg–Al–Zn), AZ31 (Mg–Al–Zn–Mn) are underinvestigation. There are number of studies relatedto corrosion mechanisms and degradation kinetics.Published results indicate that the mechanismsaffecting the corrosion behaviour of Mg-basedstents in service conditions can be: internalgalvanic corrosion; localized corrosion (pittingand filiform); stress corrosion cracking (SCC) andfatigue corrosion. Simultaneous effect of differentcorrosion mechanisms influences a stent inservice, which makes the identification ofcorrelations between in vitro and in vivoexperimental results very complex and manyissues has not yet been resolved. ASTM G31(Standard Practice for Laboratory ImmersionCorrosion Testing of Metals) offers generalguidance for metal corrosion testing, but there areno standardized procedures for biocorrosion ofbiodegradable metal materials.Low corrosion resistance of magnesium andits alloys, within the very aggressive environmentin the physiological system, is the vitalcharacteristic enabling degradability of the metalimplant. On the other hand, too rapid loss ofmechanical properties and/or toxic degradationproducts are major problems associated with Mgalloys. Also, the high corrosion rate producesrapid hydrogen gas evolution within the body.The pursued development direction is to controlthe speed of corrosion, along with optimizationof the biological response to these implants tomaximize recovery. Biologically compatiblesurface modifications through treatments orcoating systems have been investigated asprotective strategy in corrosive environments inorder to optimize implant properties [8, 9].Corrosion resistant coatings are commonly usedfor metals in many applications, but in area ofdergadable implamnts, these coatings need todegrade along with the magnesium, or yield tothe environment, leaving no harmful traces. Someof the tried solutions are: anodization, puremagnesium coating on Mg alloy, PVD coating ofaluminium, but all of them have some downsides.One of the most biocompatible coating optionsfor orthopedics is calcium phosphate coatings.Zhang et al. [9] investigated ion plating of Ticoatingon pure Mg for biomedical applications,inspired by good properties of both Ti and Mgalloys. They reported good results inimprovement of the corrosion resistance of Mgand promising further potentials, butcomprehensive testing of this new coating is yetto come.Another method for increasing the corrosionresistance of a magnesium alloy is the surfacestructure modifications. Magnesium alloysundergo microgalvanic corrosion when multiplephases exist in an alloy, one more cathodic thananother and in order to slow the corrosion rate isto modify the surface to homogenize it [8].Amorphous surface would eliminate theformation of galvanic cells between grains andboundaries. One such a way is to make the matrixa completely amorphous bulk metallic glass tocompletely remove corrosion difference due tocrystal structure in the metal [8]. AmorphousMgZnCa alloys have been tested in vivo to showreduced hydrogen evolution [10]. Ionimplantation is another method of surfacemodification to increase corrosion resistance,also creating a gradual transition between themodified surface and the bulk of the material.This makes strong, adherent surfaces that do nothave the problems of adhesion, thermal stresses,and crackings that separate secondary coatingphases tend to have [8]. Plasma immersion ionimplantation of Al, Zr, and Ti has been used tocreate corrosion resistance on AZ91.13 th International Conference on Tribology – Serbiatrib’13 95

4. FRICTION AND WEAR ISSUES RELATEDTO MG ALLOYSTribocorrosion is defined as an irreversibletransformation of a material from concomitantphysicochemical and mechanical surfaceinteractions occurring at tribological contacts [11].In general, metallic implants need to be tested totribocorrosion and wear issues, regardless of theirarea of application. The realised contacts betweenimplant material and either other implant material,or organic body constituent (e.g. bone) andelements of the aggressive physiological bodyenvironment, provoke certain responses (weardebris) that need to be investigated in more depth,especially for newly developed biomaterials. It isproven that even micro contacts within smallregions sometimes influence significant responses,ranging from changes in contact environment up toincreasing deterioration of implant surfaces due towear related processes. Two- or three-body contactsare frequently associated with tribocorrosion [12].Entrapped wear debris acts as an abrasive and isdefined as the third body. The main concernsrelated to the simultaneous action of corrosion andwear in biomedical systems are the ability of thepassive layer to withstand the mechanical stressesarising from wear, the ability of the metal surface torepassivate when the passive film is removed andthe resistance of the new repassivated surface toboth wear and corrosion [12]. Fretting has a biginfluence on the corrosion behavior of orthopedicdevices. The tribocorrosion and fretting corrosionmechanisms have been mainly related to in vitrolaboratory investigations. However, the in vivobehavior of metallic implants under combined wearcorrosion or fretting corrosion has been hardlystudied [12]. Also, the corrosion fatigue ofstructural magnesium alloys has been studied byseveral authors in NaCl and borate-bufferedsolutions, but generally focused on applications inelectronic, automotive and aerospace industries.The fatigue strength of magnesium alloys issignificantly reduced in humid environments, andthe fatigue limit drops drastically in NaCl solution[12]. Even though the corrosion behavior ofbiomedical magnesium-based alloys has beenextensively studied, corrosion fatigue has receivedlittle attention and degradable material mustmaintain appropriate mechanical strength duringthe healing of the fractured bone, to provide safeorthopedic device. Also, fatigue crack propagationbehavior needs to be studied for a wide variety ofbiodegradable Mg alloys [12]. Corrosion and wearresistances are frequently studied in synergy,because of the direct relationship between theseproperties and the biocompatibility of thebiomedical device.Regarding coronary stenting by using metallicimplants, analysis of the fatigue crack growthbehavior is of the utmost importance for its properfunctioning. There are my mathematical models inthe literature, but, in particular, the fatigue crackpropagation approach simulating crevice corrosionconditions in physiological solutions has beenhardly considered for biomedical magnesium alloys[12]. Very important aspect is tribology relatedphenomena during the production of magnesiumbiodegradable vascular stents minitubes, sincemagnesium alloys possess highly limited roomtemperatureformabilities [13]. Highly complexphysical, chemical and mechanical characteristicsof magnesium must be taken into consideration, inorder to keep the magnesium alloy characteristics(microstructure, etc.) unchanged under theinfluence of e.g. severe abrasive friction and wearduring cold drawing [13].Highly important is the understanding of theMg alloy stent behaviour at its positioning duringthe movement through the vasculature, at theinitial interventional cardiovascular treatment.Lubricious coatings have been used for over 20years [14] and the benefits are well established:(1) lower frictional force between the device andthe vessel reduces tissue damage and preventsvasospasm; (2) improved maneuverability aidsnavigation of complex lesions and facilitatesaccess to tortuous vascular sites leading toexpansion of the patient population that canbenefit from these treatments; and (3) reducesthrombogenicity. In addition, reduced frictionbetween the therapy catheters and supportcatheters leads to improved outcomes, reducedprocedure time, and, ultimately, reduced cost [14].The stent is commonly placed at the one fixedposition within a blood vessel, meaning that thereare no further movement between the stentmaterial and surrounding tissue, leading to theconclusion that wear plays no significant role.However, blood flow around the placed stent hasmicro influence and might provoke nano-weardebris, which is not investigated so far. Such tinywear debris represents a form of particulate matterin the vasculature and it is well known that if largeenough and in sufficient quantities, can causeocclusion of blood vessels and lead to tissuehypoxia and, ultimately, necrosis [14]. If alloyingelements of Mg alloys, such as Al, Zn areconsidered as well, it is obvious that this areaneeds further studies. Since Mg alloys stents hasnot been widely tried in clinical practice, there aremany tribology related questions to be addresses.96 13 th International Conference on Tribology – Serbiatrib’13

5. CONCLUSIONThe magnesium alloys as bioabsorbable /biodegradable implants for biomedical applicationsare highly promising materials, but some issuesneed to be resolved and extensive researchactivities are pursued throughout the world.Different aspects of alloying system design, novelstructures, degradation rate control, and surfacemodification methods have been tested, mainly inorder to increase corrosion degradation time.Significant attention is still needed related toproduction processes, tribocorrosion, fatigue crackgrowth behaviour, wear and friction processes andother complex issues when observed withinaggressive human body environment.ACKNOWLEDGMENTSThis paper is supported by TR-35021 project,financed by the Ministry of Education, Science andTechnological Development of the Republic ofSerbia.REFERENCES[1] J.A. Helsen, Y. Missirlis: Biomaterials, Biologicaland Medical Physics, Biomedical Engineering,Springer-Verlag, Berlin Heidelberg, 2010.[2] F. Witte: The history of biodegradable magnesiumimplants: A review, Acta Biomaterialia, Vol. 6, No.5, pp. 1680-1692, 2010.[3] K.U. Kainer (Ed.): Magnesium – Alloys andTechnology, Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim, 2003.[4] S. Zhang, X. Zhang, C. Zhao, J. Li, Y. Song, C. Xie,H. Tao, Y. Zhang, Y. He, Y. Jiang, Y. Bian:Research on an Mg–Zn alloy as a degradablebiomaterial, Acta Biomaterialia, Vol. 6, No. 2, pp.626-640, 2010.[5] H. Hermawan, D. Dubé, D. Mantovani:Developments in metallic biodegradable stents,Acta Biomaterialia, Vol. 6, No. 5, pp. 1693-1697,2010.[6] R. Erbel et al.: Temporary scaffolding of coronaryarteries with bioabsorbable magnesium stents: aprospective, non-randomised multicentre trial,PROGRESS-AMS (Clinical Performance andAngiographic Results of Coronary Stenting withAbsorbable Metal Stents), The Lancet, Vol. 369,No.9576, pp. 1869 - 1875, 2007.[7] R. Waksman et al.: Early- and Long-TermIntravascular Ultrasound and AngiographicFindings After Bioabsorbable Magnesium StentImplantation in Human Coronary Arteries, JACC:Cardiovascular Interventions, Vol. 2, No. 4, pp.312-320, 2009.[8] J. Waterman, M.P. Staiger: Coating systems formagnesium-based biomaterials - state of the art, In:W.H. Sillekens, S.R. Agnew, N.R. Neelameggham,S.N. Mathaudhu (Eds.): Proceedings of MagnesiumTechnology 2011, TMS, The Minerals, Metals &Materials Society, San Diego, California, USA, pp.403-408, 2011.[9] E. Zhang, L. Xu, K. Yang: Formation by ion platingof Ti-coating on pure Mg for biomedicalapplications, Scripta Materialia, Vol. 53, No. 5, pp.523-527, 2005.[10] B. Zberg, P.J. Uggowitzer, J.F. Loffler: MgZnCaglasses without clinically observable hydrogenevolution for biodegradable implants, NatureMaterials, Vol. 8, No. 11, pp. 887-891, 2009.[11] D. Landolt, S. Mischler, M. Stemp, S. Barril: Thirdbody effects and material fluxes in tribocorrosionsystems involving a sliding contact, Wear, Vol. 256,pp. 517-524, 2004.[12] R.A. Antunes, M.C.L. de Oliveira: Corrosionfatigue of biomedical metallic alloys: Mechanismsand mitigation, Acta Biomaterialia, Vol. 8, No. 3,pp. 937-962, 2012.[13] G. Fang, W.J. Ai, S. Leeflang, J. Duszczyk, J. Zhou:Multipass cold drawing of magnesium alloyminitubes for biodegradable vascular stents,Materials Science & Engineering C, doi:10.1016/j.msec.2013.04.039, In Press, 2013.[14] D.E. Babcock, R.W. Hergenrother, D.A. Craig, F.D.Kolodgie, R. Virmani: In vivo distribution ofparticulate matter from coated angioplasty ballooncatheters, Biomaterials, Vol. 34, No. 13, pp. 3196-3205, 2013.13 th International Conference on Tribology – Serbiatrib’13 97

Serbian TribologySocietySERBIATRIB ‘1313 th International Conference onTribologyKragujevac, Serbia, 15 – 17 May 2013Faculty of Engineeringin KragujevacANALYSIS OF THE SURFACE LAYER FORMATION OF SINGLECYLINDER ENGINE COMBUSTION CHAMBER WITHPHOSPHOROUS-FREE AND CONVENTIONAL ENGINELUBRICANTSL.YÜKSEK 1* , H.KALELİ 2 , D.ÖZKAN 3 , H. HACIKADİROĞLU 31 Researcher Dr., Yıldız Technical University, Faculty of Mechanical Engineering, Automotive Division34349 Beşiktaş-İstanbul/TURKEY, lyuksek@yildiz.edu.tr2 Professor Dr., Yıldız Technical University, Faculty of Mechanical Engineering, Automotive Division34349 Beşiktaş-İstanbul/TURKEY, kaleli@yildiz.edu.tr3 PhD. Student, Yıldız Technical University, Faculty of Mechanical Engineering, Energy Division34349 Beşiktaş-İstanbul/TURKEY, dogus_ozkan@hotmail.com*Corresponding author.Abstract: Phosphorus-free engine lubricants are gaining importance for preventing catalyst poisoningwhich is the major deactivation mechanism that causes three-way catalyst malfunction. The main purpose ofthis paper is to evaluate the mechanism of surface layer formation in combustion chamber of spark ignitionengines with phosphorous-free and phosphorous containing mineral engine lubricants. An experimentalendurance test was conducted for 100 hours at equal load conditions for each lubricant. Endurance testswere run with a laboratory engine test bench. Subsequently, engines dismantled and cylinder liners were cutaccurately to obtain specimens for microscopic examination. Optical microscopy, scanning electronmicroscopy and energy dispersive X-ray spectroscopy methods were used for evaluation.Elemental measurements on the surface which were obtained by X-ray spectroscopy were examinedstatistically. Results of the experiments showed that phosphorous containing lubricant deposited morecarbon and oxygen although less manganese and silica than the phosphorous containing rival. After the X-ray spectroscopy of the combustion chamber surface at top dead centre, iron element composition of thephosphorus-free lubricant was notably higher than phosphorus containing oil.Keywords: Phosphorous-free Oil, Combustion chamber, X-ray spectroscopy, Additive layer formation,Cylinder liner1. INTRODUCTIONMechanical losses are responsible for theapproximately 10-15% fuel energy loss. Half of themechanical losses are generated by the frictionbetween piston rings and cylinder liner. Therefore,the performance of the piston ring and cylinderliner tribological pair is the indicator of theperformance and the lifespan of an internalcombustion engine [1]. Particularly, wear aroundtop dead centre is the main limiter of effectiveengine life. Piston rings are intended to maintainthe dynamic sealing between crankcase and thecombustion chamber which minimize the powerlosses caused by the blow-by mass transfer throughthe crankcase during expansion stroke [2]. Successof this sealing mainly depends on the wear rate ofring and liner pack which are primarily the functionof the formed tribofilm on ring and liner surfaces.Engine lubricants are formulated to generate andsustain the protection against wear whilelubrication, cooling and cleaning of the surfaces arealso expected from lubricant. Modern enginelubricant can satisfy these demands with chemicalcompounds like anti-wear additives, anti-oxidantadditives, dispersants, detergents etc. [3]. Additivesalso enhance the chemical composition of baselubricant [4].98 13 th International Conference on Tribology – Serbiatrib’13

Zinc dialkyldithiophosphate (ZDDP) lubricantadditives are the most effective anti-wear and antioxidantadditives in the cost and performanceperspectives [5]. Therefore, it has been in use fordecades. Although, ZDDP mainly containsphosphorus which is a well-known poison for threewaycatalysts [6]. Environmental concerns becomemore dominant among authorities and individualswhich result stricter emission regulations [7].Automotive manufacturers are pressurised toproduce cleaner vehicles in all terms from well towheel. Harmful exhaust emissions are reducedthrough after-treatment systems like dieseloxidation catalysts, three-way catalyst and dieselparticulate filters. Poisoning phenomenon relatedwith lubricant has to be prevented, and hence, newand emerging technologies have become important.The solution is complex, namely reducing theamount of phosphorus, sulphur, zinc andmagnesium without deteriorating the performanceof the oil [8].This study intended to investigate the interactionbetween phosphorus containing (PC) and Nonphosphorusand non-ash containing lubricant(NPNA) on the combustion chamber surface ofinternal combustion engines. An endurance test wasconducted with two of the identical spark ignitionengines. Both engines aged during 100 hours undercertain load conditions which were determinedaccording to standards. Liner surfaces are examinedwith electron, optical microscopy and energydispersive X-ray spectroscopy techniques.2. EXPERIMENTAL DETAILSExperimental study consists of two equalendurance tests with two identical engines,specifications of which are listed in Table 1.Endurance tests were performed after a run-inperiod and a consecutive oil drain.Table 1. Specifications of test engines.DesignationValue/typeManufacturer/Model Honda/GX 200Type4-stroke air-cooled, SIAspirationNaturally aspiratedLubrication methodSplashNumber of cylinders 1Bore x Stroke (mm) 68 x 54Cylinder volume (cm 3 ) 196Compression ratio 8.5:1Crankcase oil capacity (l) 0.6Speed max (rpm) 3600Rated torque (Nm)12.4@2500 rpmRated power (kW)4.1@3600 rpmISO 8178 standard was selected as a reference todetermine load conditions and a direct currentgenerator was used to generate the brake load [9].An overall scheme of the test bench is shown inFigure 1. A specially formulated NPNA lubricantand a conventional PC lubricant were used for thetests, specifications of which are listed in Table 2under the courtesy of IDEMITSU KOSAN CO.LTD. Japanese petrochemical company.Figure 1. Schematic, CAD and real view of the testbench.Table 2. Specifications of test oils.Specification PC NPNASAE grade 10W30 10W30TBN (mgKOH/g) 5.73 3.13Viscosity 100 °C (cSt) 10.4 10.3Viscosity index 139 142Flash point (Celcius) 224 240Specific gravity@15 °C 0.874 0.862Ca content (wt %) 0.2 0Zn content (wt %) 0.09 0S content (wt %) 0.19 0.18P content (wt %) 0.08 0Operating conditions are summarized in Table 3for both types of engine oils. PC and NPNAlubricant were aged during 100 hours under certainloading conditions, at the end of the test, cylinderliner and piston rings were machined to obtainspecimens for microscopic analysis.Table 3. Details of load conditions.Specification PC NPNAEngine speed (rpm) 2500 2500Engine load (%) 75 75Endurance test duration (h) 100 100Ambient temperature (Celcius) 22±3 22±3Typical composition of cylinder liner wasrequired to obtain better assessment of the surfacewhich had been provided by the enginemanufacturer and these are listed in Table 4.Specimens had been ultrasonically cleaned with n-hexane.13 th International Conference on Tribology – Serbiatrib’13 99

Table 4. Typical composition of cylinder liner.ElementComposition % massFe 93.97P 0.3V 0.15C 3.00Si 2.00Mn 0.60Figure 2. Optical, SEM and EDS results of PC lubricant.3. RESULTS AND DISCUSSIONTop dead centre is the most complex surface forthe tribologists where the conditions are extreme.Combustion induced pressure gradient acts upon topring and hence contact between liner and ring surfacereaches top levels and conditions become severe.Furthermore, combustion gases and sootcontroversially affect the lubrication in TDC area,acidic compounds, unburned hydrocarbons and sootaccumulation on liner surface occurs with constantreplenishment cycle. In addition to the factorsexplained above, high gas temperature reduces oilviscosity which also has detrimental effect on linerringlubrication. Base number retention capability andhigh temperature and high shear rate viscosity(HTHS) of engine lubricant become excessivelyimportant on TDC lubrication as well as theperformance of anti-wear additive. Total base numberindicates the ability of engine oil to neutralize acidiccompounds which primarily originate fromcombustion chamber and transfer through the liner tothe crankcase. The more the base number retention,the lower the oxidative wear on liner surfaceespecially around TDC.Downsized engines with turbochargers are ongoingtrend to fulfil the requirements of CO 2 emissionreduction and fuel economy [10]. Low load fueleconomy and torque flexibility make these types ofengines favourable although increased boost levelsresult significantly higher contact pressures.Therefore, HTHS viscosity gains attention with risingin-cylinder pressure which is the wellness of lubricantperformance under severe operation conditions.Layer formation of anti-wear additive on TDCsurface is the main factor for decreasing the linerwear. Anti-wear additive acts like buffer between ringand liner asperities and prevents adhesion.Accumulated additives on surfaces can effectively bedetected through electron microscopy technique.Besides, it is possible to detect elemental distributionwith X-ray spectroscopy method.Equal tear-down processes were applied to testengines, which include dismantling of cylinder linersand piston rings. Scanning electron microscopy(SEM) and energy dispersive X-ray spectroscopy(EDS) were applied on TDC region of cylinder lineras shown in Figure 2 and Figure 3.Figure 3. Optical, SEM and EDS results of NPNAlubricant.Labels on figures indicate the location of theinspection: G1 designation borders the combustionchamber surface where the sweep motion of topring has ended G2 designation indicates linersurface covered by piston crown. Representation ofinspected surfaces are depicted in Figure 4, analysepoints were determined by considering the surfacelayers of additive accumulation.Figure 4. Schematic representation inspected surfaces.Figure 5. EDS result of combustion chamber surface.EDS measurements were applied for both of thesurfaces lubricated with test oils. Figure 5 showsthe elemental composition of layer on the surface ofcombustion chamber. Higher amount ofcarbonaceous deposit was detected with the PClubricant on combustion chamber surface. Similartrend observed for G2 surface as shown in Figure 6while atomic concentration is different.100 13 th International Conference on Tribology – Serbiatrib’13

C- Lower deposits with NPNA on combustionsurface indicate the effect of lubricant oncarbonaceous deposit formation.ACKNOWLEDGEMENTThis study was supported by Idemitsu KosanCo. Ltd. and the authors would like to thank Dr.Hiroshi Fujita for his contributions throughout thisstudy.Figure 6. EDS result of TDC surface.Carbon and oxygen levels indicate the oil filmon both of the surfaces while the G1 surfaceenvironment is quite different than G2. Highercarbon concentration of PC surface of combustionchamber can be attributable to accumulation ofswept portion of surface oil film.4. CONCLUSSIONNewly developed catalyst friendly phosphorusfreeengine oil and a conventional ZDDPcontaining engine oil were tested by applying 100 hlong endurance study. Two of identical engineswere aged under equal loading conditions thendismantled and analysed. The main subject of theinvestigation is the additive layer formation oflubricant on combustion chamber surface.Considering the measurements and observationsmade with optical microscope, SEM and EDS,findings can be summarized as;A- Higher level of deposit formation wasdetected with PC lubricant on bothcombustion chamber and piston shadedsurfaces.B- PC conventional lubricant mainly differsfrom NPNA with oxygen content oncombustion chamber surface; thisdiscrepancy originates from the type ofdeposited compounds.REFERENCES[1] B.G. Rosen, R. Ohlsson, T.R. Thomas: Wear ofcylinder bore microtopography, Wear, Vol. 198,pp. 271-279, 1996.[2] M.F. Jensen, J. Bottiger, H.H. Reitz, M.E. Benzon:Simulation of wear characteristics of enginecylinders, Wear, Vol. 253, pp. 1044-1056, 2002.[3] D. Troyer, J. Fitch: Oil analysis basics, NoriaCorporation, Oklahoma,USA, 1999.[4] Z. Pawlak: Tribochemistry of Lubricating Oils,Elsevier Science, 2003.[5] M.A. Nicholls, T. Do, P.R. Norton, M. Kasrai,G.M. Bancroft: Review of the lubrication ofmetallic surfaces by zinc dialkyl-dithlophosphates,Tribol Int, Vol. 38, pp. 15-39, 2005.[6] A.J.J. Wilkins, N.A. Hannington: The Effect ofFuel and Oil Additives on Automobile CatalystPerformance, Platinum metals review, Vol. 34, pp.16-24, 1999.[7] EUP, Official Journal of the European Union,(EC) No 715/2007, 2007.[8] T. Katafuchi, N. Shimizu: Evaluation of theantiwear and friction reduction characteristics ofmercaptocarboxylate derivatives as novelphosphorous-free additives, Tribol Int, Vol. 40,pp. 1017-1024, 2007.[9] 8178-4:2007 Reciprocating internal combustionengines, Exhaust emission measurement in: Part4: Steady-state test cycles for different engineapplications, ISO, 2007.[10] T.V. Johnson: Vehicular Emissions in Review,SAE Int. J. Engines, Vol. 5, pp. 216-234, 2012.13 th International Conference on Tribology – Serbiatrib’13 101

Serbian TribologySocietySERBIATRIB ‘1313 th International Conference onTribologyKragujevac, Serbia, 15 – 17 May 2013Faculty of Engineeringin KragujevacTRIBOLOGICAL STUDY OF BIOCOMPATIBLE HYBRIDORGANIC MOLECULES FILM WITH ANTIBACTERIALEFFECTJ.H. Horng* 1 , C.C.Wei 1 , S. Y. Chern 1 , W.H. KAO 2 , K.W. Chern 1 , Y.S. Chen 11 Department of Power Mechanical Engineering, National Formosa University, Taiwan2 Institute of Mechatronoptic Systems, Chienkuo Technology University, Taiwan*jhhorng@gmail.comAbstract: Optical glass is widely used on bioengineering and various utilities such as public touchscreendisplay and mobile devices. This work evaluates the feature of anti-bacterial and anti-adhesion on OTSmaterial that mixed with biocompatibility antibacterial agent coated on the optical glass. Test samples wereallocated with different bath and drying temperatures as well as reaction time. The result shows that in thecontact of angle experiments, pure OTS film and mixed antibacterial films have almost the same contactangle about 105° at the condition of reaction time of 12 hours and reaction temperature of 80 ° C.Theantibacterial test find that the order: antibacterial agent> OTS+ antibacterial agent(50%) > OTS+antibacterial agent(10%) > OTS. At the same operation condition, OTS mixed with 50% antibacterial agentis able to increase adhesion force between OTS film and lens. It suggests that the surface treatment of opticallenses involving OTS with 50% antibacterial solution is the most to increase the antifouling andantibacterial functions and enhance the adhesion function between films and lens surfaces.Keywords: self-assembled monolayer, adhesion force, friction, terminal group bonding, contact angle, antibacterial.1. INTRODUCTIONThe uses of SAMs in biomedicine utilities areincreasing rapidly, such as in biosensors,nonfouling surfaces, bioactive surfaces, and drugdelivery [1, 2]. Octadecyltrichlorosilane (OTS)monolayer is one of the most extensively studiedself-assembled monolayer [3-5]. Therefore, how toimprove the adhesion and anti-bacterialperformance of SAMs film becomes an attractivetopic in order to enhance device application andreliability. Bierbum [6, 7] noted that the substratesurface water layers are an important factor in theformation of OTS films. Bierbum explained thatOTS molecules initially spread vertically onsubstrate surfaces and been clustered after locatingactivation positions. Afterwards, other OTSmolecules spread to the cluster edges and formislands. The molecules then spread outwards andcause adsorbed molecules to form connections,finally forming tightly connected monolayers. In1998, Vaillant et al. [8] used atomic forcemicroscopy (AFM) and a Fourier-transforminfrared spectrometer (FTIR) to observe the processby which OTS molecules form films on substratesurfaces. The results showed that a larger amount ofwater in the solutions cause the OTS molecules toundergo a hydrolysis reaction and producepolymerization within the solution. Cloud-shapedor island-shaped molecule films form throughoutthe solution. In the contrast, solutions withcomparatively low proportions of water exhibitpoint distribution and OTS molecules igrowchaotically into liquid-like form. While the surfacediffusion makes OTS molecules absorbingmolecules within the solution, the tightly knit,island-shaped structures are formed by messymolecule films. Resch [9] also used AFM andfound that OTS molecules initially grow messilyand irregularly. With the passage of time,molecules covering the surface spread horizontallyand ultimately form tightly arranged molecular102 13 th International Conference on Tribology – Serbiatrib’13

films. Carraro et al. [10] examined formation ofOTS SAMs under different ambient temperatures.They discovered when the ambient temperaturefalls below 16°C, OTS first form islands or cloudsand then films. When the ambient temperature risesabove 40°C, the films grow evenly instead offorming islands. In addition, films form morequickly at lower temperatures. The formation ofOTS monolayer on a material surface is highlysensitive to several factors, which include thedensity of surface hydroxyl groups, reactiontemperature, reaction environment, reaction time,solvent used to deposit OTS water content of thesolvent concentration of OTS, solution age,roughness of the underlying substrate andcleaning procedures after SAM deposition [11].Therefore, how to improve the adhesion and antibacterialperformance of SAMs film becomes anattractive task in order to enhance deviceapplication and reliability.2. EXPERIMENTALThe optical lenses were ultrasonicated in acetoneand then rinsed with solvent tetrahydrofuran anddeionized water and immediately dipped in the OTSsolution containing approximately 40 ml. For thepreparation of SAMs film, OTS was dissolved inalcohol and prepared to a molar concentration of10 mM, and then mix in different proportions ofantibacterial agent (10%, 50%). The test pieces wereplaced in the solution at different bath temperaturesand duration times and a drying time of 10 min. Thetest pieces were then removed and set aside for 12 hrbefore being ultrasonicated in acetone for 5 min toremove loosely bound material and rinsed indeionized water and blown dry with nitrogen gas.The molecular structure of OTS is shown in Table 1.It’s hydrophobic properties comes from terminalgroup (CH 3 ). Main composition of biocompatibilityantibacterial agent is bioflavonoid and citric acid.For the experimental investigation ofhydrophobic properties for the different surface filmson the lens, FTA contact angle equipment was usedto measure contact angle, as shown in Figure 1.Larger contact angle indicates better hydrophobicand anti-fouling properties of surfaces. Contactangles were measured on both sides of the waterdrop. Droplet profiles were captured using a videocomprising of digital frames over a period of 12seconds and transferred to a computer for anglemeasurement. The adhesion force between surfacefilms and substrates were measured using atomicforce microscopy (AFM) by scratch mode. AFMalso was used to examine topography of samplesbefore and after SAM deposition by non-contactmode.Table 1. The molecular structure of OTSSAMs Molecular formula Head group Terminal groupOTS CH 3 (CH 2 ) 17 SiCl 3 -SiCl 3 -CH 3Figure 1. Contact angle equipment3. RESULTS AND DISCUSSIONIn the contact angle analysis of variousoperation conditions, the measurement data of eachtest piece was obtained from the mean of fivemeasurements. Figure 2(a) is a photo of the contactangle for OTS material. Figure 2(b) show thecontact angle changes with various reaction timesand bathe temperatures.Contact Angle (deg)120906030(a)020 40 60 80Temperature ( o C)(b)Figure 2. Contact angles (a) experimental photo (b)comparison chart for different reaction times andtemperatures.It shows that the higher the bathe temperature,the higher the contact angle. The higher the reactiontime, the higher is the contact angle. However, thevariation of contact angles of OTS+50%antibacterial agent films under various operation13 th International Conference on Tribology – Serbiatrib’13 103

conditions are all quite low. Bathing OTS+50%agent films at a bath temperature of 80 graduallyincreases the contact angle to approximately 105degree. The various reaction time and bathetemperature have extremely little influence on thecontact angle. In summary, the bath temperature of80 ° C and duration time of 12 hours was chosen asoperation condition in order to investigate theantibacterial characteristics of surface film on lenses.Ra (nm)25020015010050LensOTSOTS + 10% agentOTS + 50% agentantibacterial agent absorbed and stored in thetopographic valley of OTS film.The reliability and beauty requirement of thedisplay elements made from company becomeimportant in their service life. The lighttransmittance and film adhesion properties are oneof key performances of lens. In order to explore therelation between surface film and lighttransmittance of lens, Figure 4 shows thattransmittance of OTS film and antibacterial agenton the lens. This result indicates that the OTSsurface film will decrease the light transmittance oflens. However, the antibacterial agent hasextremely little influence on the transmittance oflens. The minimum value of transmittance is 93.6%under the film of OTS + 50% agent. It concludesthat all transmittance of surface films is acceptablefor industrial applications in our work.1200(a)100LensOTSOTS + 10% agentOTS + 50% agentRa=102 nmTransmittance (%)80604020(b)Figure 3. (a)The roughness values of the differentsurface films (b) 3-D topography image of the OTS +50% agent film.The various roughness values of differentsurface materials are shown in Figure. 3.Roughness test were conducted in air at a relativehumidity of about 50% using AFM by non-contactmode. The scanned detection range was 40 µm × 40µm. The various surface roughness value of differentsurface materials are shown in Figure 3(a). Thecomparison chart shows that antibacterial agent candecrease the surface roughness value of pure OTSfilms. The roughness value of OTS film surfaceadding 10% antibacterial agent is approximately 175nm, whereas the OTS film roughness value adding50% antibacterial agent was decreased toapproximately 100 nm. The 3-D topography imagefor the hybrid organic molecules film (OTS + 50%agent) is shown as Figure 3(b). The island-shapedstructures were formed on the surface, as mentionedin Vaillant’s work [8]. It shows hybrid organic filmexhibit uniform coverage the surface with regularpattern of island formation. It indicates thatFigure 4 The light transmittance of differentsurface films on lensThe film adhesion is another one of keyperformances of lens for reliability. Figure 5 showsthe effect of antibacterial agent on the critical loadof surface films on the lens. It shows thatantibacterial agent increases adhesion forcebetween OTS film and lens. Mixing antibacterialagent (50%) in OTS material increases the criticalload to approximately 104N. In summary, thesurface treatment of optical lenses involving OTS+Agent (50%) is the most capable of effectivelyincreasing anti-adhesion functions.In the antibacterial tests, staphylococcus aureuswere inoculated with different self-assembled film,and then after 24 hours to measure bacteria values(JISZ 2801:2010). Figure 6 is the comparison chartof the number of the bacteria for the differentsurface films. For the general lens surface, thebacteria number is about 135000 after 24 hours.The pure OTS film also has little antibacterialfunction. It shows that the bacteria number on OTSwith 50% antibacterial agents and pure antibacterialagent surface is less than 10. It is far lower than thebacteria value, 5.3 × 10 4 , on the OTS film.104 13 th International Conference on Tribology – Serbiatrib’13

PlateCount No.Critical Load ( N)1201101009080706050OTSOTS + 10% agentOTS + 50% agentFigure 5. Critical loads between surface films andsubstrate16000012000080000400000LenseOTSOTS + 10% agentOTS + 50% agentagentFigure 6. Effect of surface film material onantibacterial4. CONCLUSIONThis work studied the feature of anti-bacterialand anti-adhesion on OTS self-assembledmonolayers which mixed with biocompatibilityantibacterial agent that coated on optical lens. Theresults can be concluded as follows:1. Both OTS and mixed OTS film caneffectively increase the contact angle of alens surface at various bath temperatures aswell as duration time and reduces deviceadhesion force effectively.2. The adding of antibacterial agent has littleeffect on the contact angle and lighttransmittance of pure OTS film.3. The antibacterial agent can effectively reducethe surface roughness while increase theadhesion force and antibacterial abilities ofpure OTS film on lens surfaces (reactiontime = 12hr., reaction temperature = 80 ° C).ACKNOWLEDGMENTSThe author gratefully acknowledges common Lab.for Micro-Nano Science and Technology at NFU.

Serbian TribologySocietySERBIATRIB ‘1313 th International Conference onTribologyKragujevac, Serbia, 15 – 17 May 2013Faculty of Engineeringin KragujevacTHE INFLUENCE OF CORROSION ON THEMICROSTRUCTURE OF THERMALLY TREATED ZA27/SIC PCOMPOSITESBiljana Bobić 1 , Aleksandar Vencl 2 , Miroslav Babić 3 , Slobodan Mitrović 3 , Ilija Bobić 41 Institute ″Goša″, Belgrade, Serbia, biljanabobic@gmail.com2 Faculty of Mechanical Engineering, University of Belgrade, Belgrade, Serbia, avencl@mas.bg.ac.rs3 Faculty of Engineering, University of Kragujevac, Kragujevac, Serbia,babic@kg.ac.rs, boban@kg.ac.rs4 ″Vinča″ Institute of Nuclear Sciences, University of Belgrade, Belgrade, Serbia, ilijab@vinca.rsAbstract: The influence of corrosion on the microstructure of ZA27/SiC p composites was examined.The composites were produced by compo casting technique and then subjected to the thermaltreatment (T4 regime). Microstructural examinations were performed after 30-day exposure ofthermally treated composites in the sodium-chloride solution. Corrosion processes have occurredin the composite matrix. Corrosion did not affect SiC particles in the composites. The localprogress of corrosion in depth of the composite matrix was noticed in micro-cracks. Corrosionresistance of ZA27/SiC p composites was evaluated based on the mass loss of composite samplesduring the immersion test. It was found that corrosion resistance of the composites decreased withincrease in content of SiC particles. The applied thermal treatment had a negative influence on thecorrosion resistance of ZA27/SiC p composites.Keywords: Metal-matrix composites, ZA27 alloy, Compocasting, Thermal treatment, Corrosion,Microstructure1. INTRODUCTIONDomestic composites with base ZA27 alloy[1, 2] have been developed with an aim toobtain composites which maintain goodmechanical characteristics at elevatedtemperatures as well as to make compositeswith improved tribological propertiescompared to those of the matrix alloy.Particulate ZA27/SiC p composites were shownto possess significant tribological potentialbecause of high hardness and high wearresistance [3, 4]. Within this work, domesticZA27/SiC p composites were obtained bycompo casting technique. The matrix alloy,with 27 wt. % aluminum, belongs to zinc–aluminum foundry alloys with relatively highcontent of aluminum (ZA alloys). The alloy ischaracterized by good physical, mechanicaland technological properties (low density, highstrength and hardness, easy machinability) [5,6], by a substantial resistance to corrosion andhigh wear resistance [7–9]. This enabledcommercial application of ZA27 alloy as asignificant tribomaterial, especially for makingbearings and bushings.ZA27 alloy solidifies in the widetemperature range and is suitable forprocessing in the semi-solid state [10]. This ledto the application of compo casting techniquefor producing domestic particulate compositeswith base of ZA27 alloy. Micro-particles ofSiC [1], Al 2 O 3 [1, 2, 11, 12], graphite [9, 13]or ZrO 2 [14] were incorporated in the semi-106 13 th International Conference on Tribology – Serbiatrib’13

solid melt of ZA27 alloy using mechanicalmixing. Obtained composites were subjected tomicrostructural examinations [11, 12] andtribological tests [9, 13]. In addition,mechanical characteristics of the compositeshave been studied, at room temperature [11,12] and moderately elevated temperatures [2].However, corrosion behavior of domesticcomposites with base ZA27 alloy has not beentested so far.Physical, mechanical and corrosioncharacteristics of metal-matrix composites aredeeply influenced by the microstructure ofmetal matrices [15, 16]. It was shown thatthermal treatment affected the microstructureand properties of ZA27 alloy [17–19]. Abeneficial effect of T4 regime on ductility [17]and tribological characteristics of ZA27 alloy[18, 19] was noticed, although this thermaltreatment resulted with minor reduction inhardness and tensile strength [18]. In addition,it was shown that T4 regime affected themicrostructure and corrosion resistance of ascastZA27 alloy [20] and thixocast ZA27 alloy[21].ZA27 alloy is highly corrosion resistant inatmospheric conditions and natural waters [22].The most common form of corrosion in theseenvironments is general corrosion, whichenables calculations of the alloy corrosion rate,based on the weight loss of samples duringexposure in corrosive media. Immersion testsin chloride solutions have been usedfrequently, because chloride ions are present innumerous corrosive environments.Тhixocast ZA27 alloy is the base ofZA27/SiC p composites obtained by compocasting technique [23]. Accordingly, themicrostructure of thixocast alloy is actually themicrostructure of the composite matrix. It wasshown recently that thermal treatment (T4regime) negatively affected the corrosionbehavior of thixocast ZA27 alloy [21].However, there have been no published resultsuntil now, concerning the effect of T4 regimeon the microstructure and corrosion resistanceof ZA27/SiC p composites obtained by compocasting technique.Considering the importance of corrosionresistance for selection and application ofmetal-matrix composites, it was the aim of thiswork to study the influence of corrosion on thesurface appearance and microstructure of thethermally treated ZA27/SiC p composites.Corrosion resistance of the composites wasevaluated based on the weight loss of samplesduring immersion in the sodium-chloridesolution.2. EXPERIMENTAL2.1 MaterialsA domestic producer of zinc-aluminumalloys (RAR Foundry ® Ltd., Batajnica) hasprovided the master alloy for the experimentalwork. SiC particles (with average diameter of40 µm) were obtained from the domesticmanufacturer of abrasive products (GinićTocila ® Ltd., Barajevo).ZA27 alloy was conventionally melted andcasted in the Department of Materials Science″Vinča″ Institute. Chemical composition of thealloy is given in Table 1.Table 1. Chemical composition of ZA27 alloyElement* Al Cu Mg Znwt. % 26.3 1.54 0.018 balance*Concentration of other elements (Fe, Sn, Cd,Pb) is within acceptable limits.Compo casting technique was used formaking composites with 1, 3 and 5 wt. % SiCparticles. The particles were incorporated intothe semi-solid melt of ZA27 alloy with use ofmechanical mixing. Parameters of the appliedcompo casting aprocess nd description of theapparatus can be found in [23].Composite castings were subjected to a hotpressing, in order to reduce porosity andimprove the bond strength between the matrixand particulate reinforcements. Samples formicrostructural examinations and corrosiontesting were machine cut from the compositecastings. The samples were thermally treatedaccording to T4 regime: solutionizing at 370ºCfor 3 hours, with subsequent water quenchingand natural aging.13 th International Conference on Tribology – Serbiatrib’13 107

2.2 MethodsMicrostructural examinationsSurface morphology and microstructure ofthermally treated ZA27/SiC p composites wereexamined by optical microscopy (OM) andscanning electron microscopy (SEM). CarlZeiss optical microscope and JEOL JSM–5800scanning electron microscope were used.Cylindrical samples (5 mm in diameter and 8mm in height) were embedded in thepolymethacrylate and then they were groundand polished. Wet grinding was performedwith successively finer abrasive papers (240,360, 600 and 800 grit SiC), while polishingwas done using polishing cloth and diamondpaste (particles size up to 2 μm). Aqueoussolution of nitric acid (9 v/v % HNO 3 ) wasused for etching of the samples.The samples were rinsed with acetone anddried in the air before exposure in the testsolution (3.5 wt. % NaCl). After finishing ofthe exposure, the samples were prepared formetallographic examination in the usual way.Corrosion rate testingCorrosion rates of thermally treatedZA27/SiC p composites were calculated basedon the samples mass loss during exposure inthe test solution (immersion test). Preparationof the samples and testing procedure wereperformed in accordance with ASTM G31 [24].The samples (18 x 28 x 3 mm), in triplicate,were suspended vertically in the stagnantsodium-chloride solution (3.5 wt. % NaCl,pH=6.7) open to the atmosphere. The test wasperformed at room temperature (23 ± 2°C).After 30 days of exposure, the samples werewithdrawn from the test solution and rinsedwith distilled water. Corrosion products wereremoved from the surface of the samples bychemical procedure [25]. The samples werethen reweighed to determine the mass lossduring exposure in the test solution.Calculation of the average corrosion rateCR [mm/year] is based on the mass loss of thesamples Δm [g] during the immersion test:8.76 ⋅ ∆mCR =d ⋅ A ⋅τ(1)τ is the exposure time (720 hours), A is thesample surface [cm 2 ] and d [g/cm 3 ] is thecomposite density. Values of the compositedensity (for the composite with 1, 3 and 5 wt.% SiC particles, respectively) were calculated[23] and shown in Table 2.Table 2. Density of ZA27/SiC p compositesMaterial ZA27/1%SiC p ZA27/3%SiC p ZA27/5%SiC pd[g/cm 3 ]4.97 4.92 4.87The values in Table 2 were used to calculatecorrosion rates of the thermally treatedZA27/SiC p composites.3. RESULTS AND DISCUSSION3.1 Microstructure of thermally treatedZA27/SiC p compositesSurface appearance and microstructure ofthermally treated ZA27/SiC p composites wereexamined before exposure and after 30-dayexposure in the sodium-chloride solution.Surface appearance of the composites with 3wt.% SiC particles, is shown in Figure 1a, b.Figure 1. Surface appearance of the thermally treatedcomposite ZA27/3wt.%SiC p (OM, polished):a) before exposure, b) after 30-day exposure in 3.5wt.% NaCl.108 13 th International Conference on Tribology – Serbiatrib’13

SiC particles are uniformly distributed in themetal matrix (Fig. 1a). A few inclusions can benoticed on the surface of the composite sampleand mechanical damages on the edge of thesample. Applied thermal treatment had noeffect on the particles of reinforcement andtheir distribution in the composite matrix.The microstructure of composites wasrevealed by etching (Figure 2a, b). It can beseen in Fig. 2a that SiC particles are distributedin the regions of η phase and regions of phasemixture α+η. There are no voids, due to thefallout of SiC particles from the compositebase (e.g. during machining or metallographicpreparation of samples). This indicates goodbonding between SiC particles and the matrixalloy.Figure 2. Microstructure of the thermally treatedZA27/3wt.%SiC p composite (OM, etched):a) before exposure, b) after 30-day exposure in 3.5 wt.%NaCl.Main micro-constituents in the compositematrix are also visible in Fig. 2a. Themicrostructure of the composite base is nondendriticand is characterized by the presenceof large primary particles [21, 23]. The primaryparticles are complex; they consist of a core(rich in aluminum) and a periphery (composedof the phase mixture α+η). Interdendritic ηphase, rich in zinc, is located between theprimary particles. The microstructure of thecomposite matrix and the microstructure ofthixocast ZA27 alloy are morphologically verysimilar [23]. The applied thermal treatment (T4regime) has caused changes in the structure ofthe composite matrix. The region of phasemixture α + η was expanded, while the regionsof individual phases (α and η) were reduced. Inaddition, the size of primary particles of αphase was decreased for about 30 vol. % [21,23]. All this resulted with finer microstructureof the composite matrix. However, the increasein number of micro-cracks on the phaseboundaries η/α+η was noticed in the compositematrix after the thermal treatment [23].Thermal stress at boundary surfacesmatrix/particle, due to quenching withinthermal treatment, was preceded by thermalstress during solidification of the compositemixture. The stress can cause a localdeformation of the metal matrix aroundparticles of reinforcement [26], appearance ofmicro-cracks or fracture of the particles.Electrolytes can be retained in the microcracks,causing local progress of corrosionprocesses into depth of the composite base.The number of boundary surfacesmatrix/particle was significantly increased inthe ZA27/SiC p composites due to the presenceof SiC particles.. It can be assumed thatdislocation density was also increased duringcooling of the composite mixtures, due todifferent coefficients of linear expansion of thematrix ZA27 alloy and SiC particles.Corrosion processes have influenced thesurface appearance and microstructure ofthermally treated ZA27/SiC p composites. Thesurface appearance of the composites with 3wt. % SiC p , after 30-day exposure in thesodium-chloride solution, is shown in Fig. 1b.Large primary particles of α phase are visiblein the central area of the composite sample. Itcan be seen that corrosion has started at placesof mechanical damages, voids, inclusions.Corrosion processes occurred in the compositebase, in the regions of phase mixture α+η andregions of η phase. The local progress ofcorrosion was noticed in the micro-cracks andpores. The micro-cracks were probably formedduring solidification of the composite mixtures,in hot pressing as well as during quenching13 th International Conference on Tribology – Serbiatrib’13 109

within thermal treatment of the composites.The presence of micro-cracks negativelyaffected corrosion resistance of thermallytreated ZA27/SiC p composites.SiC particles were not involved in corrosionprocesses because of their inherent chemicalstability. However, these particles haveinfluenced corrosion behavior of ZA27/SiC pcomposites. The continuity of boundarysurfaces matrix/particle is disturbed in theclusters of SiC particles. On these placesmicro-pores and micro-cracks can be formed.Due to the retention of sodium-chloridesolution in these places, local progress ofcorrosion in depth of the composite matrix wasnoticed, as it was mentioned before.Figure 3. Corrosion products of the of thermally treatedZA27/3wt.%SiC p composite after 30-day exposure in3.5 wt.% NaCl (SEM): a) surface appearance, b) detail.During exposure in the sodium-chloridesolution, corrosion products were formed onthe surface of the composite samples. Spongy,white deposits of the corrosion products,mostly in the form of rosettes, are shown inFigure 3a, b.Microstructural examinations of thermallytreated ZA27/SiC p composites, after exposurein the sodium-chloride solution, made itpossible to gain some insight into the influenceof corrosion processes on the structure of thesecomposite materials. It was found thatcorrosion started in places of mechanicaldamage, voids, inclusions. Corrosion processeshave occurred mainly in the composite base,although in pores and micro-cracks, the localprogress of corrosion in depth of thecomposites was noticed. Corrosion processesdid not influence SiC particles.Results of the microstructural examinations,after exposure of thermally treated ZA27/SiC pcomposites in the sodium-chloride solution, arein accordance with results obtained during theimmersion test.3.2 Corrosion rate of thermally treatedZA27/SiC p compositesAfter finishing of exposure in the NaClsolution, corrosion products were removedfrom the surface of ZA27/SiC p compositesamples by chemical procedure [25]. It wasfound that corrosion attack was mostlyuniform, while corrosion processes took placepredominantly on the composite surface. Theaverage value of corrosion rate CR [mm/year]was calculated based on the mass loss ofcomposite samples during the immersion test.The results are presented in Figure 4.For the purpose of comparison, results of theimmersion test for thermally treated ZA27alloys (as-cast and thixocast) [21, 23] are alsopresented in Figure 4. It can be seen thatcorrosion rates of the composites are higherthan those of both ZA27 alloys (as-cast andthixocast).Corrosion resistance of the compositematrix (thermally treated thixocast ZA27 alloy)is higher than that of thermally treatedcomposites.Corrosion rate (mm/year)0.300.250.200.150.100.050.00ZA27 cast ZA27 thixo K1 K2 K3MaterialFigure 4. Corrosion rate of thermally treated ZA27alloys and ZA27/SiC p composites after 30-day exposurein 3.5 wt.% NaCl. K1 - ZA27/1wt.%SiC p , K2 -ZA27/3wt.%SiC p , K3 - ZA27/5wt.%SiC p .Corrosion rate of thermally treatedZA27/SiC p composites increases with increasein content of SiC particles, because of increase110 13 th International Conference on Tribology – Serbiatrib’13

in number of micro-cracks and clusters of SiCparticles. All above presented indicates lowercorrosion resistance of thermally treatedZA27/SiC p composites with higher content ofparticulate reinforcements.4. CONCLUSIONSParticulate ZA27/SiC p composites (with 1, 3and 5 wt.% SiC particles) were obtained bycompo casting technique and subsequentlysubjected to the T4 thermal treatment.Thermally treated composites were exposed inthe sodium-chloride solution for 30 days. Theinfluence of corrosion on the surfaceappearance and microstructure of thecomposites was examined. Corrosionresistance of the composites was evaluatedbased on the mass loss of composite samplesduring the immersion test. According to theresults presented, the following conclusionscan be proposed:1. SiC particles are uniformly distributed in themetal matrix of the particulateZA27/3wt.%SiC p composite that was madeby compo casting technique.2. Morphological changes and appearance ofmicro-cracks in the microstructure of thecomposite matrix were noticed after T4thermal treatment. However, the thermaltreatment had no effect on SiC particles andtheir distribution in the composite matrix.3. Corrosion process has influenced themicrostructure and surface appearance ofthermally treated ZA27/SiC p composites,after 30-day exposure in the sodiumchloridesolution. However, corrosion didnot affect SiC particles in the composites.4. Corrosion started at places of mechanicaldamages, voids, inclusions. Corrosionprocesses ocurred mainly in the compositematrix although the local progress ofcorrosion was noticed in the micro-cracks.5. Corrosion resistance of the compositematrix is higher than that of ZA27/SiC pcomposites.6. Corrosion rate of thermally treatedZA27/SiC p composites increases withincrease in content of SiC particles.7. Applied thermal treatment (T4 regime) hasshown a negative effect on the corrosionresistance of ZA27/SiC p composites.ACKNOWLEDGEMENTSThe Ministry of Education, Science andTechnological Development of the Republic ofSerbia has supported financially this workthrough the projects TR 35021 and OI 172005.The authors are gratefully acknowledged to theRAR ® Foundry Ltd., Batajnica (Belgrade,Serbia) and Ginić Tocila ® Abrasive ProductsFactory Ltd., Barajevo (Belgrade, Serbia), forproviding the master alloy and SiC particles, toperform this research.REFERENCES[1] I. Bobić: Development if Metal Forming Process inSemi-Solid State (Rheocasting and CompocastingProcess) and Process Controling Effects on Qualityof ZnAl25Cu3 Alloy Products, PhD thesis, Facultyof Technology and Metallurgy, University ofBelgrade, Belgrade, 2003.[2] B. Bobić, M. Babić, S. Mitrović, N. Ilić, I. Bobić,M.T. Jovanović: Microstructure and mechanicalproperties of Zn25Al3Cu based composites withlarge Al 2 O 3 particles at room and elevatedtemperatures, Int. J. Mat. Res. (formerly Z.Metallkd.), 101, pp. 1524–1531, 2010.[3] I.A. Cornie, R. Guerriero, L. Meregalli, I. Tangerini:Microstructures and Properties of Zinc-Alloy MatrixComposite Materials, in Proceedings of theInternational Symposium on Advances in CastReinforced Metal Composites, 24–30.09.1988.,Chicago, pp. 155–165[4] N. Karni, G.B. Barkay, M. Bamberger: Structureand properties of metal-matrix composite, J. Mater.Sci. Lett., 13, pp. 541–544, 1994.[5] E. Gervais, R.J. Barnhurst, C.A. Loong: An analysisof selected properties of ZA alloys, JOM, 11, pp.43–47, 1985.[6] E.J. Kubel Jr.: Expanding horizons for ZA alloys,Adv. Mater. Process., 132, pp. 51–57, 1987.[7] R.J. Barnhurst, S. Beliste: Corrosion Properties ofZamak and ZA Alloys, Tech. Rep., NorandaTechnology Centre, Quebec, Canada, 1992.[8] M. Babić, R. Ninković: Zn-Al alloys astribomaterials, Trib. Ind., Vol. 26, No. 1–2, pp. 3–7,2004.[9] S. Mitrović: Tribological Properties of ZnAl AlloyComposites, PhD thesis, Faculty of MechanicalEngineering, University of Kragujevac, Kragujevac,2007.[10] M. Flemings: Behavior of metal alloys in the semisolidstate, Metall. Trans. A, Vol. 22A, pp. 957–981, 1991.[11] I. Bobić, M.T. Jovanović, N. Ilić: Microstructureand strength of ZA-27 based composites reinforced13 th International Conference on Tribology – Serbiatrib’13 111

with Al 2 O 3 particles, Mater. Lett., 57, pp. 1683-1688, 2003.[12] I. Bobić, R. Ninković, M. Babić: Structural andmechanical characteristics of composites with basematrix of Rar 27 alloy reinforced with Al 2 O 3 andSiC particles, Trib. Ind., Vol. 26, No. 3–4, pp. 27–29, 2004.[13] M. Babić, S. Mitrović, D. Džunić, B. Jeremić, I.Bobić: Tribological behavior of composites basedon ZA-27 alloy reinforced with graphite particles,Tribol. Lett., 37, pp. 401–410, 2010.[14] Z. Aćimović–Pavlović, K.T. Raić, I.Bobić, B.Bobić: Synthesis of ZrO 2 Particles Reinforced ZA25Alloy Composites by Compocasting Process, Adv.Compos. Mater., 20, pp. 375–384, 2011.[15] L.L. Shreir, R.A. Jarman, G.T. Burstein: Corrosion,Butterworth- Heinemann, Oxford, 2000.[16] S.L. Donaldson, D.B. Miracle: ASM HandbookVolume 21, Composites, ASM International, 2001.[17] I. Bobić, B. Đurić, M.T. Jovanović, S. Zec:Improvement of Ductility of a Cast Zn–25 Alloy,Mater. Charact., 29, pp. 277–283, 1992.[18] M. Babić, A. Vencl, S. Mitrović, I. Bobić: Influenceof T4 heat treatment on tribological behavior ofZA27 alloy under lubricated sliding condition,Tribol. Lett., 2, pp. 125–134, 2009.[19] M. Babić, S. Mitrović, R. Ninković: R: TribologicalPotential of Zinc-Aluminium Alloys Improvement,Trib. Ind., Vol. 31, No. 1–2, pp. 15–28, 2009.[20] B. Bobić, J. Bajat, Z. Aćimović–Pavlović, M. Rakin,I. Bobić I: The effect of T4 heat treatment on themicrostructure and corrosion behaviour ofZn27Al1.5Cu0.02Mg alloy, Corros. Sci., 53, pp.409–417, 2011.[21] B. Bobić, J. Bajat, Z. Aćimović–Pavlović, I. Bobić,B. Jegdić, Corrosion behaviour of thixoformed andheat-treated ZA27 alloys in the natrium chloridesolution, Trans. Nonferr. Met. Soc. of China, to bepublished[22] F.C Porter: Corrosion Resistance of Zinc and Zincalloys, Marcell Dekker, New York, 1994.[23] B. Bobić: Examination of the Corrosion ProcessImpact on the Microstructure and MechanicalProperties of Zn27Al1,5Cu0,02Mg alloy CastingsReinforced with Silicon Carbide Particles, PhDthesis, Faculty of Technology and Metallurgy,University of Belgrade, Belgrade, 2011.[24] ASTM G31–72 (2004): Standard Practice forLaboratory Immersion Corrosion Testing of Metals.[25] ASTM G1–03(2011): Standard Practice forPreparing, Cleaning and Evaluation of TestSpecimens.[26] R.J. Arsenault, L. Wang, C.R. Feng: Strengtheningof composites due to microstructural changes in thematrix, Acta Metall. Mater., Vol. 39, No. 1, pp. 47–57, 1991.112 13 th International Conference on Tribology – Serbiatrib’13

Serbian TribologySocietySERBIATRIB ‘1313 th International Conference onTribologyKragujevac, Serbia, 15 – 17 May 2013Faculty of Engineeringin KragujevacTRIBOLOGICAL CHARACTERISATION OF PBT + GLASS BEADCOMPOSITES WITH THE HELP OF BLOCK-ON-RING TESTConstantin Georgescu 1 , Mihai Botan 1 , Lorena Deleanu 11 ”Dunarea de Jos” University of Galati, Romania, constantin.georgescu@ugal.roAbstract: The materials involved in this research study were produced by die moulding in order to obtainbone samples type 1A (SR EN ISO 527-2:2003). These composites have a matrix of polybutyleneterephthalate (PBT) commercial grade Crastin 6130 NC010, DuPont. The values for the glass beadsconcentrations were established at 10% and 20% (wt). Block-on-ring tests were run in order to characterizethe tribological behaviour of this friction couple (PBT and PBT composites with glass beads on steel). Theblock was manufactured by cutting parts from the bone samples, having the dimensions of 16.5 mm × 10 mm× 4 mm. The other triboelement was the external ring of the tapered rolling bearing KBS 30202, havingdimensions of Ø35 mm × 10 mm and was made of steel grade DIN 100Cr6. There were analysed thefollowing characteristics: friction coefficient (mean value over a test and scattering range), wear (wearrate). There are also presented particular aspects of the worn surfaces, as investigated from SEM images.Keywords: PBT composite, tribological behaviour, block-on-ring test, dry sliding.1. INTRODUCTIONMaterials based on PBT are obtained both byadding very different materials (nano and microfibre reinforcements [1], metallic or/and ceramicpowders, minerals [2], [3]), the result could beincluded in the class of composites, and byblending with other polymers polytetrafluoroethylene(PTFE) [4], polycarbonate (PC) [5],polyethylene (PE), SAN, epoxy resin, with fireresistant additives [6], both solutions directioningone or a set of the properties of PBT matrix.The adding materials in PBT are very diverse,almost all types known for the polymericcomposites (long and short fibres, particles andtheir mixtures), both at micro scale and nano scale.For tribological applications, the fibre nature is alsodiverse: glass, carbon, aramidic, titanates.Even if the specialized literature emphasis theinfluence of the adding materials in PBT, uponsome mechanical characteristics (traction limit andelasticity modulus) [2], [3], [7], these properties donot also reflect the tribological behaviour of thesematerials. This is why the testing of the polymericcomposites is of high importance and, even if theresults could not be extrapolated from thelaboratory tests on tribotesters, to the actual frictioncouple, these studies are useful in materials'ranking, when the designer is interest in a particularparameter or a set of characteristics [8], [9], [10].2. MATERIALS AND TESTINGMETHODOLOGYThe tested materials were produced by diemoulding in order to obtain bone samples type 1A(as required by the tensile test ISO 527-2) at theResearch Institute for Synthetic Fibres Savinesti,Romania, taking into account the producerspecification for moulding and heat treatment [11].These composites have a matrix of polybutyleneterephthalate (PBT), commercial grade Crastin6130 NC010, DuPont.The recipes for the composite materials based onPBT, included in this study, were elaborated by theauthors based on up-to-date documentation [1], [4],[11] and were designed in order to point out theinfluence of matrix and adding materials on thetribological behaviour in dry regime. Table 1presents their compositions and the abbreviationsused in this paper. The polyamide (PA) was addedin low concentration in order to have a better13 th International Conference on Tribology – Serbiatrib’13 113

dispersion of the micro glass beads. The blackcarbon was added for both technological andtribological reasons.Table 1. The tested materialsConcentration [%, wt]Material symbol Micro glassBlackPBTPAbeadscarbonPBT 100 - - -GB10 88 10 1.5 0.5GB20 77.5 20 2 0.5The tests were done using a block-on-ringtribotester, functioning on a CETR tribometerUMT-2 Multi-Specimen Test System.The ring was the external ring of the taperedrolling bearing KBS 30202 (DIN ISO 355/720),having the dimensions of Ø35 mm × 10 mm andwas made of steel grade DIN 100Cr6, having 60-62HRC and Ra = 0.8 μm on the exterior surface.The block was manufactured by cutting partsfrom the bone samples, having the dimensions of16.5 mm × 10 mm × 4 mm.The tests were run in dry condition, forcombination (F, v), F being the normally appliedload (F = 1.0 N, F = 2.5 N and F = 5.0 N) and vbeing the sliding speed (v = 0.25 m/s, v = 0.50 m/sand v = 0.75 m/s). The sliding distance was thesame for all tests, L = 7500 m.For evaluating the mass loss of the blocks, ananalytical balance METTLER TOLEDO was used,having the measuring accuracy of 0.1 mg.The SEM images were done with the help of thescanning electron microscope Quanta 200 3D,having a resolution of 4 nm, a magnification×1.000.000.3. EXPERIMENTAL RESULTS3.1 Friction coefficientIn order to compare the three tested materials, theextreme values and the average value of the frictioncoefficient were graphically presented in Figure 1 asa function of the sliding speed and the normal load.These values (the lowest value, the highest value andthe average one) were calculated based on therecorded values during each test (sampling rate being10 values per second). Thus, it could be appreciatedthe stability of the friction coefficient by the size ofthe scattering interval and an average energyconsumption by the average value of the frictioncoefficient.For actual applications working under similarconditions of speed and load, the author wouldrecommend the materials with a smaller scatteringinterval and lower values of the average frictioncoefficient.The low loads and speeds produce a largerscattering interval for the friction coefficient, but theload and speed increase makes the frictioncoefficient diminish the average value and to narrowthe scattering interval. A research report from NASA[12] had evidenced high average values of thefriction coefficient of over 0.6, for three polymerssliding against steel (the tribotester: polymeric ballon steel disk).From these research reports and theexperimentally obtained data during this study, theauthors point out the importance of the laboratorytests for evaluating the friction coefficient and othertribological characteristics.0.80.6µ 0.40.200.80.6µ 0.40.200.80.6µ 0.40.20TBP10BG20BGF = 1 NTBP10BG20BGTBP10BGv = 0.25 m/s v = 0.50 m/s v = 0.75 m/sMaterialTBP01BG02BGF = 2.5 NTBP01BG02BGTBP01BGv = 0.25 m/s v = 0.50 m/s v = 0.75 m/sMaterialTBP10BG20BGF = 5 NTBP10BG20BGTBP10BGv = 0.25 m/s v = 0.50 m/s v = 0.75 m/sMaterialFigure 1. Variation of friction coefficient of PBT andcomposites with different micro glass beads content, forthe sliding distance L = 7500 m20BG02BG20BG114 13 th International Conference on Tribology – Serbiatrib’13

PBT has the average values of the frictioncoefficient, , in the narrowest range, around thevalue 0.2. The increase of this average could beexplained by the elimination of the relatively bigwear particles that are characteristic for this polymer(see Figure 4). The values obtained for F = 5 N aregrouped under 0.2 for all the tested sliding speeds.The composites GB10 (PBT + 10% micro glassbeads) and GB20 (PBT + 20% micro glass beads)have the average value of the friction coefficientscattered on larger intervals, especially for thesmaller normal loads (F = 1 N and F = 2.5 N). For F= 1 N, it is hard to establish a dependency relation ofthe friction coefficient on the adding materialconcentration and the sliding speed. It could benoticed that for blocks made of GB20, there arelarger intervals.At the sliding speed of v = 0.25 m/s, the abrasivewear is predominant, the polymer being hung (torn)and drawn from the superficial layers as microvolumes,their size being greater at higher speeds(Figure 2). At the sliding speed of v = 0.75 m/s, theinfluence of the normal load on the average value ofthe friction coefficient is similar: increases from0.12 for F = 1 N, to ~0.2 for F = 5 N.the composites with same type of micro glass beadsadded in a polyamide matrix [13].The values of the friction coefficient have thetendency of being less dependent on the slidingspeed for the normal load F = 5 N; this recommendsthese materials for an exploitation regime withdifferent working speeds (differentiated speedsimposed by the technological process), withouthaving very different energy consumption levelswhen the speed is changing.The extreme values of the friction coefficient arecaused by the generation and the detaching of thewear debris, the ring passing over a bigger microglass beads, an agglomeration of micro glass beadsor fragments of some broken ones on the surface asremained after a preferential elimination of thepolymer from the superficial layer. In other studieson the polymeric composites with micro glass beads,there were no reports on fracturing the hard particles.For the composites with PBT matrix, the authorsnoticed breakings of the micro glass beads, generallythose of bigger diameters (20...40 m) being broken.Figure 3 presents four broken micro glass spheres(A, B, C and D) on an area of ~ 600 m × 600 m inthe central zone of the contact; the resultedfragments are embedded into the polymeric matrix.Such events taken place in the contact create highoscillations of the friction coefficient.Figure 2. SEM image of the block made of GB10, forv = 0.25 m/s, F = 5 N, L = 7500 mFor the blocks made of GB20, under F = 2.5 N,the scattering of the values for the friction coefficientis the largest. The probable cause would be themicro-cutting processes that will have a morereduced intensity when the sliding speed increases.There were not noticed processes of dragging themicro glass beads on the block surfaces, meaningthat the interface between the micro glass beads andthe polymeric matrix is harder to damage, ascompared to, for instance, the mobility of the microglass beads in the sliding direction, but also in thedepth of the superficial layer, as noticed in testingFigure 3. SEM image of a block made of GB10 – fourbroken micro glass beads (A, B, C and D). Testconditions: v = 0.25 m/s, F = 5 N, L = 7500 mFrom SEM images (Figure 4), the wear debriswere characterized as size and shape, many are madeespecially of polymer with only small glass debris(from fragmented micro glass beads) or small microglass beads (but rare). During the test, the weardebris adhere one to each other and are generally bigand rare (as compared to the wear debris resultedfrom other polymer in dry sliding against steel) and13 th International Conference on Tribology – Serbiatrib’13 115

they are volumic (Figure 4), not laminated and thin,as it is happening in the case of PTFE [14].Generally, small micro glass beads are evacuatedfrom the superficial layers and the polymer aroundthe bigger ones is detached. In this scenario, one ormore micro glass beads will support an individualload great enough to be broken.a) At the edge of the wear track from the ring(v = 0.5 m/s) determines diminishing the averagevalue of the friction coefficient characterizing thesecomposites, from 0.15...0.28, to 0.12...0.22. At v =0.5 m/s, both composites behave well, the frictioncoefficient becoming stable around the average valueof 0.2. The polymer is warming and, thus, it isreducing its mechanical properties and allows forgenerating a very thin viscous film that is notexpelled from the contact (as it happens with otherpolymer under high speed) and becomes afavourable factor in reducing friction also byembedding the glass beads in the soften matrix.At F = 2.5 N, the average value of the frictioncoefficient has a slightly tendency of increasingwhen the micro glass beads concentration areincreased.At F = 5 N, the values of the analysed parametersof the friction coefficient have been reduced (figure1), confirming the results obtained in other research[12] that the small loads generate a more intensefriction for the friction couple element(s) made ofpolymer or polymeric composites and hardcounterpart (steel). The normal force, for which thefriction coefficient begins to decrease, is dependingon the shape and size of the triboelements and on theworking conditions [15], [16].3.2 Wearb) Wear particles made of polymer and very smallfragments from the broken glass beadsFigure 4. Aspect of the wear particles generated duringthe test involving the sliding of the block made of GB20on the metallic ring. Test conditions: F = 5 N, v = 0.75m/s, L = 7500 mAt F = 1 N and v = 0.25 m/s, a larger scatteringinterval of the friction coefficient had resulted; thereare prevailing the micro-cutting process and eventsimplying the glass beads (overrunning of the hardasperities of the metallic ring, the breakage of themicro glass beads and rare shear of the hardasperities, the micro glass beads embedding into thepolymeric matrix). A doubling of the sliding speedTaking into account the commanding parametersinvolved in this study (the material, by theconcentration of the adding materials, the slidingspeed and the load) and the recent documentation onwear parameterization [15], [16], [17], [18], [19],[20], the authors selected the wear rate (k) foranalysing the experimental wear results obtainedduring this research.V mk F L F L3[mm /(N m)] (1)where F [N] – the normal force and L [m] – thesliding distance, V [mm 3 ] is the material volume lostby wear, Δm [g] is the mass loss of a block,calculated as the difference of the initial mass of theblock and its mass after being tested, ρ [g/mm 3 ] isthe density of the tested block material.The wear maps (see Figure 5) were plotted usingMATLAB R2009b, the wear parameter beingrepresented for each material as a function of thesliding speed and the normal force, with the help of acubic interpolation.For PBT (see Figure 5), one may notice asignificant increase of the wear parameter when thenormal force is decreasing - the cause could be theincrease of the heightening factor for the abrasivewear under low loads and the absence of a transferfilm on the hard surface due to the absence of the116 13 th International Conference on Tribology – Serbiatrib’13

mechanical pressure and thermal loading greatenough for initiating and maintaining an adherenceprocess.For all tested sliding speeds, the tendencycharacterizing the wear variation as a function ofload has a minimum zone around the value of 4 N.- an accentuated increase of the wear rate forloads smaller than 2.5 N, with higher values for thecomposite GB10;- for the composite GB10, the wear rate isdecreasing almost linearly when the load isincreasing and it is insignificantly decreasing whenthe sliding speed is increasing; k is smaller for thetwo composites with micro glass beads as comparedto the basic material (PBT), the lowest values beingrecorded for the composites, under the load F = 5 N;- at F = 5 N, for all the tested materials, the wearrate has a very low sensitivity to the variation of thesliding speed, the smaller values being obtained forthe composites.Thus, the wear rate diminishes when introducingglass beads in PBT. The wear is diminishing due tothe increase of the material resistance (see theresults for the composite GB10), but when themicro glass beads concentration becomes 20%, theabrasive component of the wear process increases,too.4. CONCLUSIONSAdding micro glass beads in PBT makes thefriction coefficient increase almost linearly with themicro glass beads massic concentration, with ~15%for each 10% of micro glass beads.An addition of 10% micro glass beads decreasesthe wear rate with ~20%. When the concentration ofmicro glass beads is increased, the decrease of thiswear parameter is smaller as compared to PBT, with~18%.REFERENCESFigure 5. The wear rate for PBT and the composites PBT+ micro glass beadsFor the composites PBT + micro glass beads,analysing Figure 5, the following conclusions couldbe drawn:- a zone with minimum values, for F = 5 N;[1] C.P. Fung, P.C. Kang: Multi-response optimizationin friction properties of PBT composites usingTaguchi method and principle component analysis,Journal of Materials Processing Technology, Vol.170, pp. 602-610, 2005.[2] G.S. Deshmukh, D.R. Peshwe, S.U. Pathak, J.D.Ekhe: A study on effect of mineral additions on themechanical, thermal, and structural properties ofpoly(butylene terephthalate) (PBT) composites, JPolym Res, Vol. 18, pp.1081-1090, 2011.[3] G.S. Deshmukh, D.R. Peshwe, S.U. Pathak, J.D.Ekhe: Evaluation of mechanical and thermalproperties of Poly (butylene terephthalate) (PBT)composites reinforced with wollastonite,Transactions of The Indian Institute of Metals, Vol.64, No. 1 & 2, pp. 127-132, 2011.[4] C. Georgescu, L. Deleanu: Influence of PTFEconcentration on the tribological characteristics ofPBT, in: Proceedings of the 7 th InternationalSymposium KOD 2012 Machine and IndustrialDesign in Mechanical Engineering, 24-26.05.2012,Balatonfured, Hungary, pp. 405-410.13 th International Conference on Tribology – Serbiatrib’13 117

[5] J. Wu, K. Wang, D. Yu: Fracture toughness andfracture mechanisms of PBT/PC/IM blends. Part VEffect of PBT-PC interfacial strength on the fractureand tensile properties of the PBT/PC blends, Journalof Material Science, Vol. 38, pp. 183-191, 2003.[6] R.J. Crawford: Plastics Engineering, 3 rd Edition,Butterworth-Heinemann, Oxford, 2002.[7] J.A. Brydson, Plastics Materials, 7 th Edition,Butterworth-Heinemann, Oxford, 1999.[8] H. Czichos, T. Saito, L. Smith: Springer Handbookof Materials Measurement Methods, SpringerScience-Business Media, 2006.[9] A. Dasari, Z.Z. Yu, Y.W. Mai: Fundamental aspectsand recent progress on wear/scratch damage inpolymer nanocomposites, Materials Science andEngineering, Vol. 63, pp. 31-80, 2009.[10] K. Friedrich: Advances in Composite Tribology,Vol. 8, Composites Materials Series, UniversitätKaiseslautern, Kaiserslautern, Germany, 1993.[11] ***: Crastin PBT. Molding Guide, available on-line:http://www2.dupont.com/Plastics/en_US/assets/downloads/processing/cramge.pdf[12] W.R. Jr. Jones, W.F. Hady, R.L. Johnson: Frictionand Wear of Poly(Amide-Imide), Polyimide andPyrone Polymers at 260°C (500°F) in Dry Air,NASA TN D-6353, Lewis Research Center, 1971.[13] L. Deleanu, G. Andrei, L. Maftei, C. Georgescu, A.Cantaragiu: Wear maps for a class of compositeswith polyamide matrix and micro glass spheres,Journal of the Balkan Tribological Association, Vol.17, No 3, pp. 371-379, 2011.[14] P. Samyn, J. Quintelier, W. Ost, P. De Baets, G.Schoukens: Sliding behaviour of pure polyester andpolyester-PTFE filled bulk composites in overloadconditions, Polymer Testing, Vol. 24, pp. 588-603,2005.[15] G.M. Bartenev, V.V. Lavrentev: Friction and Wearof Polymers, Elsevier, 1981.[16] B.J. Briscoe, S.K. Sinha: Wear of polymers, Proc.Inst. Mech. Eng. Part J. Engineering Tribology, Vol.216, pp. 401-413, 2002.[17] H. Czikos: Tribology – A System Approach to theScience and Technology of Friction, Lubricationand Wear, Elsevier Scientific Publishing Company,New-York, 1978.[18] K. Friedrich, A.K. Schlarb: Tribology of PolymericNanocomposites – Friction and Wear of BulkMaterials and Coatings, Tribology and InterfaceEngineering Series, 55, Elsevier, 2008.[19] N.K. Myshkin, M.I. Petrokovets, A.V. Kovalev:Tribology of polymers: Adhesion, friction, wear, andmass-transfer, Tribology International, Vol. 38, pp.910-921, 2005.[20] K. Holmberg: Reporting Experimental Results inTribology – Summary of the discussion on NewWear Rate Unit, in: OECD IRG 26th Meeting,5-6.10.2006, Lyon, France.118 13 th International Conference on Tribology – Serbiatrib’13

Serbian TribologySocietySERBIATRIB ‘1313 th International Conference onTribologyKragujevac, Serbia, 15 – 17 May 2013Faculty of Engineeringin KragujevacNORMAL FORCE INFLUENCE ON3D TEXTURE PARAMETERS CHARACTERIZINGTHE FRICTION COUPLE STEEL – PBT + 10% PTFEConstantin Georgescu 1 , Lorena Deleanu 1 , Catalin Pirvu 11 ”Dunarea de Jos” University of Galati, Romania, constantin.georgescu@ugal.roAbstract: This study presents the influence of the normal force on the surface quality of the friction couplesteel – polybutylene terephthalate (PBT) + 10% polytetrafluoroethylene (PTFE). There were calculated theaverage values of the amplitude and functional parameters, as obtained from investigating square areas on thewear tracks, with the help of a proposed methodology, for initial and tested surfaces generated on the blocksand on counterpart ring made of rolling bearing steel, for the following test conditions: three normal forces (F= 1 N, F = 2.5 N and F =5 N), three sliding speeds (v = 0.25 m/s, v = 0.50 m/s and v = 0.75 m/s) and a slidingdistance of L = 7500 m. The conclusion of the research study was that the tested normal force range has aninsignificant influence on the surface quality for the tested materials. This friction couple could berecommended for variable dry regimes.Keywords: PBT + PTFE material, surface texture parameters, block-on-ring test, dry sliding.1. INTRODUCTIONIn actual application for bearings and seals in dryregime, a self-lubricating polymeric material slideson a hard surface, proved to be tribologicallyefficient as compared to the sliding of a polymericmaterial on itself [1], [2]. The adhesion and abrasioncomponents of the friction and wear processessinergically influence themselves. For instance,extend of the junctions depends on the elasto-plasticdeformation of the asperities [3], [4], [5] and theycould not be separated. For the polymer-metalcontact, the deterioration of the polymer by elastoplasticdeformation is more intense and the adhesioncomponent increases for the harder surfaces [6]. Thegenerated transfer film characteristic for thepolymer-metal friction couple, also changes thesurface texture, depending of the polymer nature andthe working conditions [7], [1], [2].There are many published studies on thetribological behavior of polymeric materials [8],[7], [9], but few of them deal with the influence ofthe surface texture on the tribologicalcharacteristics of the polymeric materials and evenfewer reported how the working conditions affectthe surface texture. In 1970, Pooley and Tabor(quoted in [1]) pointed out that for PTFE, the valueof the friction coefficient is only slightly affectedby the surface quality when involving relativelysmooth ones, but with rough surfaces the wear andthe friction are intensified. Till now, the terms"smooth" and "rough" were used only in aqualitative way and there are no recommendedvalues of the texture parameters for particularapplications.Experimental studies proved that a change of thetexture parameters could significantly affect thefriction and the wear. There is why the authors ofthis research consider the texture evaluation, beforeand after testing, necessary for understanding anddirecting the tribological processes. Many polymericfriction couples are working with frequent startsand stops and the evolution of the surface texture isof great importance for improving the reliabilityand the durability of these tribosystems.For polymeric friction couples and especially forpolymer – metal contacts, the wear could be relatedboth to the amplitude and functional parameters.As resulted from the studied documentation[12], [13], the surface quality is frequentlydescribed by parameters as Sa (arithmetic averageof absolute values) and Sq (root mean squared).13 th International Conference on Tribology – Serbiatrib’13 119

The authors’ estimates that for studying the wornsurfaces and for obtaining correlations among thesurface parameters and the testing conditions, thefollowing parameters are more suitable: theparameters related to the maximum values of thetopography (Sz – the height difference between thehighest and lowest heights in the investigated area,Sv – the largest pit height, Sp – the largest peakheight) and the functional parameters (Svk –reduced valley depth, Sk – core roughness depth,Spk – reduced summit height).2. MATERIAL AND TESTINGMETHODOLOGYThere were selected the following testparameters: three sliding speeds (v = 0.25 m/s, v =0.50 m/s, v = 0.75 m/s), three applied loads (F = 1.0N, F = 2.5 N, F = 5.0 N), the sliding distance beingL = 7500 m for each test done at room temperatureand in a laboratory environment.In order to do this study, the profilometer LaserNANOFOCUS μSCAN [14] was used.For parameters' calculation it was used thesoftware SPIP 5.1.11 [15]. Figure 2 presents avirtual (rebuilt) image of the investigated zone withthe help of this software.The friction and wear behavior of PBT slidingagainst steel was evaluated with the help of aUniversal Micro-Tribometer UMT-2 and a blockon-ringtribotester. The geometry of the frictionalcouple is given in figure 1.Loadsamplea) The initial surfacerotating ringFigure 1. The shapes and dimensions of the frictioncouple block-on-ringThe polymeric blocks are prisms of 16.5 mm ×10 mm × 4 mm and they were obtained by injectionat ICEFS Savinesti, Romania, according to thespecifications of the producer from tractionsamples, cutting the blocks from the middle parallelzone of them.The polymeric blend has 90% (wt) PBT, thecommercial name being Crastin 6130 NC010 (assupplied in grains by DuPont) and 10% (wt) PTFE,commercial grade NFF FT-1-1T® Flontech, havingthe average size of the particles ~20 μm.The other element of the friction couple was theexternal ring of the tapered rolling bearing KBS30202 (DIN ISO 355/720), having the dimensionsof Ø35 mm × 10 mm and they were made of steelgrade DIN 100Cr6, having 60 - 62 HRC and Ra =0.8 μm on the exterior surface.b) The worn surface (F = 5 N, v = 0.25 m/s, L = 7500 m)Figure 2. Virtual images of the polymeric blocks madeof PBT + 10% PTFEMeasurements were done for blocks made of thepolymeric blend PBT + 10% PTFE and for theexternal rings of tapered rolling bearings, bothelements being involved in block-on-ring tests, forboth non-worn and worn surfaces.For evaluating the 3D parameters involved inthis study, there were selected three zones, each of500 m 500 m for the polymeric blocks and of100 m 100 m for the metallic rings, thesebeing reduced for reason of the surface curvature.All 3D measurements were done with a step of 5m. The distance between lines for 3Dmeasurements was also 5 m. The 3D parametersare calculated for all the values z(x, y), measuredon one area of 500 m 500 m on the block andone area of 100 m 100 m on the steel ring.120 13 th International Conference on Tribology – Serbiatrib’13

3. EXPERIMENTAL RESULTSTaking into account that PTFE has lowermechanical properties as compared to PBT [10], itwas considered necessary to study the influence ofthe normal force on the surface quality of thispolymeric blend, before and after testing.Figures 3, 4 and 5 present the average values ofthe amplitude and functional parameters, obtainedwith the help of the proposed methodology, for theinitial and tested surfaces generated on the blocksmade of PF10 (material symbol for the polymericblend PBT + 10% PTFE), for the tested conditions:three forces and three sliding speeds and a slidingdistance of L = 7500 m.μm141210m 8µ6420141210m 8µ6420141210m 8µ642014121086420PF10L = 7500 mInitial surfaceSa Sq Sp Sv SzF=1.0NF=2.5NF=5.0NPF10L = 7500 mv = 0.25 m/sSa Sq Sp Sv SzF=1.0NF=2.5NF=5.0NPF10L = 7500 mv = 0.5 m/sSa Sq Sp Sv SzF=1.0NF=2.5NF=5.0NPF10L = 7500mv = 0.75 m/sSa Sq Sp Sv SzFigure 3. The influence of the normal force on theaverage values of the dimensional amplitude parametersfor the blocks made of PF10 (PBT + 10% PTFE)The wear track surfaces are characterized byparametric values 2...3 times lower than those ofthe initial surfaces as they were obtained by themoulding technology.6420-26420-26420-26420-2PF10L = 7500 mInitial surfaceSskPF10L = 7500 mv = 0.25 m/sSskPF10L = 7500 mv = 0.5 m/sSskPF10L = 7500 mv = 0.75 m/sSskSkuSkuSkuSkuFigure 4. The influence of the normal force on theaverage values of the a dimensional amplitudeparameters for the blocks made of PF10The surface quality of this material is onlyslightly dependent on the normal force, at least forthe tested values (F = 1 N, F = 2.5 N and F = 5 N).Sa and Sq have very close values, regardlessthe force values, but Sp and Sz present a slightdecrease when the force increases, for the testdone with the sliding speeds of v = 0.25 m/s and v= 0.5 m/s.Sku (surface Kurtosis) values greater than 3.0indicate narrower height distribution due to the13 th International Conference on Tribology – Serbiatrib’13 121

particular ductile fracture of the polymer duringadhesion – abrasion wear. Ssk (surface Skewness)has values oscillating around zero, indicatingsymmetric height distributions. If Ssk < 0, thebearing surface has holes and if Ssk > 0 it is a flatsurface with peaks.μm43mµ 210431043210µm 243mµ 210PF10Initial surfaceSpk Sk SvkPF10L = 7500 mv = 0.25 m/sF=1.0NF=2.5NF=5.0NSpk Sk SvkPF10L = 7500 mv = 0.5 m/sF=1.0NF=2.5NF=5.0NSpk Sk SvkPF10L = 7500 mv = 0.75 m/sF=1.0NF=2.5NF=5.0NSpk Sk SvkFigure 4. The influence of the normal force on theaverage values of the 3D functional parameters for theblocks made of PF10It was noticed a slight decrease of thefunctional parameters when the load increases,for tests done with the sliding speed of v = 0.25m/s. For the other two tested speeds (v = 0.50 m/sand v = 0.75 m/s), this poor dependence on thenormal force was noticed only for Svk. Thegreater forces make this parameter to decreaseand this tendency could be justified by the elastoviscousbehaviour of the polymeric blend; it ispossible that the passing of the hard asperitieslaterally moves the softer material of thecounterpart accompanied by an elevation of thevalley bottoms between asperities, process alsoreported in [11], [1].4. CONCLUSIONFor the polymeric blend PBT + 10% PTFE,there was found no significant influence of thenormal force on the surface quality, for the testingconditions: range of force 1 N ... 5 N, range ofspeed 0.25 m/s ... 0.75 m/s and the sliding distance7500 m.The obtained results recommend the testedmaterial for friction couples functioning undervariable conditions (speed and load) in dry sliding.REFERENCES[1] G.W. Stachowiak: Wear – Materials, Mechanismsand Practice, John Wiley & Sons, England, 2005.[2] G.W. Stachowiak, A.W. Batchelor: EngineeringTribology, Butterworth-Heinemann, 2002.[3] D. Boazu, I. Gavrilescu: Contactul mecanic.Analiză cu elemente finite, EUROPLUS, Galaţi,2006.[4] S. Creţu: Contactul concentrat elastic-plastic,Politehnium, Iaşi, 2009.[5] N.B. Demkin, V.V. Izmailov: The Relationbetween the Friction Contact Performance and theMicrogeometry of Contacting Surfaces, Journal ofFriction and Wear, Vol. 31, No. 1, pp. 48-55, 2010.[6] P. Samyn, J. Quintelier, W. Ost, P. De Baets, G.Schoukens: Sliding behaviour of pure polyesterand polyester-PTFE filled bulk composites inoverload conditions, Polymer Testing, Vol. 24, pp.588-603, 2005.[7] K. Friedrich, A.K. Schlarb: Tribology of PolymericNanocomposites – Friction and Wear of BulkMaterials and Coatings, Elsevier, pp. 17-148,2008.[8] B.J. Briscoe, S.K. Sinha: Wear of polymers, Proc.Inst. Mech. Eng. Part J. Engineering Tribology,Vol. 216, pp. 401-413, 2002.[9] K. Friedrich, Z. Zhang, A.K. Schlarb: Effects ofvarious fillers on the sliding wear of polymercomposites, Composites Science and Technology,Vol. 65, pp. 2329-2343, 2005.[10] J.A. Brydson: Plastics Materials, 7 th Edition,Butterworth-Heinemann, 1999.[11] K.L. Johnson: Contacts Mechanics, CambridgeUniversity Press, Cambridge, 1987.122 13 th International Conference on Tribology – Serbiatrib’13

[12] L. Blunt De, X. Jiang: Advanced Techniques forAssessment Surface Topography, Elsevier, London,2003.[13] F. Blateyron: États de surface: la norme, Mesures,Vol. 787, pp. 44-47, 2006.[14] ***: NanoFocus AG μScan® – Instruction Manual.[15] SPIP The Scanning Probe Image ProcessorSPIP TM , Version 5.1.11, 2012, available on-line:http://www.imagemet.com/WebHelp/spip.html.13 th International Conference on Tribology – Serbiatrib’13 123

Serbian TribologySocietySERBIATRIB‘1313th International Conference onTribologyKragujevac,Serbia, 15 – 17 May 2013Faculty of Engineeringin KragujevaccWEARBEHAVIOUROF COMPOSITES BASED ONN ZA27 ALLOYREINFORCED WITH GRAPHITEPARTICLESSlobodan Mitrović 1 , Miroslav Babić 1 , Ilija Bobić 2 , Fatima Zivić 1 , Dragan D Dzunić 1 , MarkoPantić 11 Faculty of Engineering, University of Kragujevac, Kragujevac, Serbia,boban@ @kg.ac.rs, babic@kg.ac.rs, zivic@kg.ac.rs, dzuna@kg. ac.rs, pantic@ @kg.ac.rs2 INN "Vinca", Univerzitet u Beogradu, Beograd, Srbija, ilijab@vinca.rsAbstract: Itis well known that reinforcing the matrix with graphite particles effects on frictionproperties,while reinforcing the matrix with hard particles (Al 2 O 3 , SiC, Garnet. ..) increasess wear resistance of thematrix material, in thiscase ZA27alloy.Reinforcing the matrix by adding a the graphite also effects onmechanicalproperties of material. In this studyy authors made an attempt to investigate the effect of smalllamount of graphite reinforcement on wear behaviour of composites andd in order to preserve themechanicalproperties of the material. The composites with 1 and 2wt% of graphite particles were produced bycompocasting procedure. Wear behaviour of unreinforcedZA27 alloy and composites were studied using,computer aided block-on-disc tribometer, underr dry slidingconditions at different sliding speeds (0.25, 0.5and 1m/s) and normal loads (10N, 30N and 50N). The obtained resultss revealed that composites exhibitedbetter wearresistance in comparison to unreinforced ZA27alloy. Better wear properties of composites incomparisonto the unreinforced matrix alloy are a result of creation of o graphite rich film onthe contacttsurfaces.Keywords: Composite, ZA27, Graphite, Wear.1. INTRODUCTIONDue to wide potential applications, compositematerials have been investigated very intensivelyover the recent decades [1]. Metal matrixcomposites have emerged as an important class ofengineeringgmaterials because they provideopportunityto manage the material mechanical andtribological properties [2].Zinc-aluminium (ZA) alloys are importantbearing materials, especially suitable for high-loadand low-speed applications [3, 4]. ZA alloys arecharacterized by good tribological and mechanicalproperties, low weight excellent foundry castabilityand fluidity, good machining properties, low initialcost, and environmentally friendly technology.However, major limitations of ZAalloys aree itsinferior mechanical andwear properties on elevatedproperties. Also, this alloy exhibits dimensionalinstability at temperatures above 120°C [5].Various authors have reportedthat theincorporation of hard particles (SiC, Al 2 O 3 , zircon,garnet and glass) [6-17] improves wear resistance124of base alloy. Also, A many researchers have reportedthat MMCs reinforced r with graphite particlesexhibits low friction andd low wear rate, andsuggested that such behaviour is the result of self-lubricating graphite-rich film formation on thecontact surface [4, 18-20]. Mechanical properties ofZA27 alloy graphite reinforced are significantlychanged by varying the amount of graphite [21].Theincrease of o the graphite contentt within theZA27 matrix results in increase of ductility,compressive strength, corrosion resistance, but in adecrease of hardness. In spite of thesignificantdecrease in hardness, tribological tests showed thattaddition of graphite particles to ZA27 alloy matriximproved wear resistance off compositess [22].Based on presented literature review smalllamount of graphite will not degrade mechanicalproperties so An A attempt has been madeto evaluateethe dryslidingg wear behaviour off the ZA-27/ /graphite composites over a range of appliedloads and slidingspeeds. The unreinforced ZA-27alloy was tested as a referencematerial.The role ofgraphite in dry was discussed.13 th International Conference C onn Tribology – Serbiatrib’13

2. EXPERIMENTALTESTING2.1 MaterialThe ZA-27 alloy (27.5% Al, 2.5% Cu, 0.012%Mg, and balance Zn) was used as the base matrixalloy. The graphite particles of mean size 30 μmwere used as the reinforcement. The percentagee ofgraphite was 1 and 2 by weight. The compositespecimens were obtained by thecompocastingprocedure, which was executed bymixing in theisothermal regime.More detailed process of thecompocating proceduree could be found f elsewhere[4].a)After obtainingthe compositematerialssamples, it was necessary to perform the hotpressing to reduce porosity. The samples (blocks)for the tribological investigations weree then madefromthe ZA-277 as-cast alloy and pressed pieces.Microstructural characterization off the alloyswascarriedoutt using the optical microscopy onsamples, similar tothose used for wear testing. Thetypical OM micrographs m ofthe matrixalloy andcomposite are shown s in Fig. 1.Bulk hardness of all thee samples was measuredusing a Brinell hardness tester witha 2.5-mmmdiameter steel ballindenter b and at an applied load of625N. The load application time was 60 s. Themeanvalues of at leastfivemeasurements,conducted in different areas of each sample, showthat the composite attained lower hardness (115HB) than that of o the matrixx ZA-27 alloy (124 HB).Hardness of the matrix alloy was 124HB, whilehardness for composites reinforced with1 wt% and2 wt% of graphite particless were 119 HB and 115HB, respectively.In his investigationofmechanical properties of the cast ZA-27/ graphiteparticulate composites Seah found that withgraphite contenincrease hardness monotonicallydecreases significantly [23]. In fact, as the graphitecontent is increased from 0 to 5% the t hardnesssdecreases for about 27%.2.2Wear testsb)c)Figure 1. Optical microscopy of tested specimens: a)unreinforced matrix alloy ZA27; b) composite withh 1wt% of graphite particles; c) composite with 2 wt% % ofgraphite particles13 th International Conference on Tribology – Serbiatrib’13Samples forr tribological testing were made bycutting. Cuttingg was realised by machine saw withintensive cooling in orderto avoid changes ofsurface layers, due d to high temperature.Wear test were w carried out in a computer aideddblock-on-disk sliding s wear testing machine with thecontact pair geometry in accordance with ASTM G77– –05. More detailed description of the tribometerris available elsewhere [4].The test blocks (6.35x15.75x10.16mm) wereeprepared from ZA27 unreinforced alloy and fromcomposite withh 1% and 2% % of graphite particles.Alll samples prior to wear testare polished. Thecounter face (disc of 35 mmm diameter and 6.35 mmmthickness) was made of EN: HS 18-1-1-5 tool steelof 62HRC hardness. Thee tests weree performedunder dry sliding conditions at different slidingspeeds (0.25 m/s, 0.5 m/s, 1 m/s) and applied loads(10N, 30 N, 500 N). The duration of sliding was 10min. Each experiment was repeated fivetimes.The tests were performed at room temperature.tThewear behavior of the block was monitored interms of the wear scar width (Figure 2). Using thewear scar width and geometry of the contact pairthe wear volume (expressed in mm 3 ) wascalculated.125

Figure 2. The scheme of contact pair geometryload is presented on Fig.3. Presented plots suggestthat wear volume of all tested samples increasesswith normal load increase, , at all values of slidingspeed. Increasee in wear volume with ncreasing ofnormal load is more pronounced at higher slidingspeed (1 m/s) , as could be clearly seen if wecompare plots on o Fig. 3a with plots on Fig. 3c. Thisphenomenon iss more pronounced for unreinforcedmatrix alloy. According to Seah et al. [24] wear rateincreases monotonically with normal load increase.3. RESULTS AND DISCUSSIONWear volumeof tested ZA27/graphitecomposites,as well as unreinforcedZA27 alloy, asa function of sliding speed and normal load in drysliding conditions is illustrated on figures presenteddown below.a)a)b)b)c)Figure 4.Wear volume v of tested samples versus slidingspeed under different applied normal loads in dry slidingconditionsc)Figure 3.Wear volume oftested samples versus normalload for different sliding speeds in dry sliding s conditionsThe effect of normal load on wear volumee oftested composites, as well as the matrix alloyspecimens at different values of applied normal126The influence of slidingg speed on wear volumeof tested composites, as well as matrix alloyspecimens at constant values of applied normal loadis presented on Fig. 4. Fromm presented plots it couldbe clearly seenn that with increase of sliding speedwear volume of all tested specimens increases.Wear volume increase iss more pronounced atapplied load off 50N, and for unreinforced matrixalloy in comparison to the composites.Also, on the13 th International Conference C onn Tribology – Serbiatrib’13

figure 3a it could be seen that the wear of thecomposite with 2 wt% of graphite is almostnegligible under appliedload of 10N.Seah et al. [24] have confirmedthat the wearrate of as-cast ZA-27/graphite particulate decreasesmonotonically with an increase in sliding speed, , butin our caseit is inversely because of differentcontact geometry.portion of the contact surface. Presence of thegraphite film inn contact zone reduces the t metal-to--during drysliding,themetal/graphiteecompositestribo-influenced the graphitefilmm forming onthe contacttsurface of elements [18–20], whichacts as solidlubricantthatreducesmetal-to-metallcontactbetweenn the sliding surfaces. The formationof graphiterichlubricant film between the slidingsurfaces has beenexplainedd as a resultt of the softsecond phase (graphite)squeezing-out from thesubsurface toward the mating surfacedue toextensive plastic deformation metal contact between the sliding pairs.Many researchers have reported that[18].4.CONCLUSIONa)b)Based on the results presented in this t paper itcould be concluded: Generally wear w volumee of the composites arelesser in comparison to o the unreinforced matrixalloy. Higher content of graphite particles within thematrix alloyy results in higher wear resistance r ofmaterial, composite with 2 wt% of graphiteparticles has lesser values of wearr volumes incomparisonn to the composite with 1wt% ofgraphite particles, p under the same contacttconditions. Wear volume of all tested specimens increasesswith slidingg speed and normal load increase. Higher wear resistance of ZA27/graphitecomposites in comparison to the unreinforcedmatrix alloyy is a result of graphite film f formingon the surface of contact elements.ACKNOWLEDGEMENTTThestudy was financed byy Ministry of Education,Science and TechnologicaTal Development, Serbia,project No.35021.c)Figure5. Wear scars of tested specimens: a)unreinforced matrix alloy; b) and c) composites withh 1and 2 wt% of graphite particles, respectively.Figure 5 presents wear scarss of all testedspecimens Based on the wear scars it couldd beconcluded that the dominant wear mechanism wasabrasive wear, because of this parallel tracks withinthe wear scars of all tested materials. One canclearly notice that on the worn surface the blackgraphite film is smeared and it covers the large13 th International Conference on Tribology – Serbiatrib’13REFERENCES[1] Vasiliev, V.V., V Morozov, E.V.: Mechanics andAnalysis of Composite Materials. Elsevier, Oxford,2001.[2] Pruthviraj, R.D., R Krupakara, P.V.: Influence of SiCadditions onn mechanical l properties of the Zn–Alalloy (ZA-27). Int J Mater Sci 2(1) ), pp. 53–57,2007.[3] Babic, M., Ninkovic, R., Rac, A.: Sliding WearBehavior of o Zn-Al Alloys in Conditions ofBoundary Lubrication. L The Annals of University‘‘Dunarea De Jos’’ of Galati Fascicle VIII.Tribology, pp. p 60–64, 2005.[4] Babic, M., Mitrovic, S., Dzunic, D., Jeremic, B.,Bobic, I.: Tribological behavior of composites based127

on ZA-27 alloy reinforced with graphite particles.Tribology Lett., pp. 401-410, 2010.[5] Prasad, B.K.: Abrasive wear characteristics of azinc-based alloy and zinc-alloy/SiC composite.Wear 252(3–4), pp. 250–263, 2002.[6] Tjong, S.C., Chen, F.: Wear behavior of as-castZnAl27/SiC particulate metal-matrix compositesunder lubricated sliding condition. Metall MaterTrans A 28A, 1951–1955, 1997.[7] Prasad, B.K., Das, S., Jha, A.K., Modi, O.P.,Dasgupta, R., Yegneswaran, A.H.: Factorscontrolling the abrasive wear response of a zincbasedalloy silicon carbide particle composite.Composites A 28(4), 301–308, 1997.[8] Prasad, B.K., Modi, O.P., Khaira, H.K.: High-stressabrasive wear behavior of a zinc-based alloy and itscomposite compared with a cast iron under varyingtrack radius and load conditions. Mater SciEng A381, 343–354, 2004.[9] Prasad, B.K.: Abrasive wear characteristics of azinc-based alloy and zinc-alloy/SiC composite.Wear 252(3–4), 250–263, 2002.[10] Sharma, S.C., Girish, B.M., Kamath, R., Satish,B.M.: Effect of SiC particle reinforcement on theunlubricated sliding wear behavior of ZA-27 alloycomposites. Wear 213, 33–40, 1997.[11] Sastry, S., Krishna, M., Uchil, J.: A study ondamping behavior of aluminate particulatereinforced ZA-27 alloy metal matrix composites. JAlloys Compd 346, 268–274, 2001.[12] Sharma, S.C., Sastry, S., Krishna, M.: Effect ofaging parameters on the micro structure andproperties of ZA-27/aluminate metal matrixcomposites. J Alloys Compd 346, 292–301, 2002.[13] Bobic, I., Jovanovic, M.T., Ilic, N.: Microstructureand strength of ZA-27 based composites reinforcedwith Al2O3 particles. Mater Lett 57, 1683–1688,2003.[14] Modi, O.P., Rathod, S., Prasad, B.K., Jha, A.K.,Dixit, G.: The influence of alumina particledispersion and test parameters on dry sliding wearbehavior of zinc-based alloy. TribolInt 40, 1137–1146, 2007.[15] Sharma, S.C., Girish, B.M., Kamath, R., Satish,B.M.: Sliding wear behavior of zircon particlesreinforced ZA-27 alloy composite materials. Wear224, 89–94. 1999.[16] Ranganath, G., Sharma, S.C., Krishna, M.: Drysliding wear of garnet reinforced zinc/aluminummetal matrix composites. Wear 251, 1408–1413,2001.[17] BabicMiroslav, Slobodan Mitrovic, Fatima Zivic,IlijaBobic: Wear Behavior of Composites Based onZA-27 Alloy Reinforced by Al2O3 Particles UnderDry Sliding Condition, TribolLett 38, pp. 337–346,2010.[18] Riahi, A.R., Alpas, A.T.: The role of tribo-layers onthe sliding wear behavior of graphitic aluminummatrix composites. Wear 251, 1396–1407, 2001.[19] Yang, J.B., Lin, C.B., Wang, T.C., Chu, H.Y.: Thetribological characteristics of A356.2Al alloy/Gr(p)composites. Wear 257, 941–952, 2004.[20] Akhlaghi, F., Zare-Bidaki, A.: Influence of graphitecontent on the dry sliding and oil impregnatedsliding wear behavior of Al 2024-graphitecomposites produced by in situ powder metallurgymethod. Wear 266, 37–45, 2009.[21] Seah, K.H.W., Sharma, S.C., Girish, B.M.:Mechanical properties of cast ZA-27/graphiteparticulate composites. Mater Des 16(5), 271–275,1995.[22] Sharma, S.C., Girish, B.M., Kramath, R., Satish,B.M.: Graphite particles reinforced ZA-27 alloycomposite materials for journal bearing applications.Wear 219, 162–168, 1998.[23] Seah, K.H.W., Sharma, S.C., Girish, B.M.: Effect ofartificial ageing on the hardness of cast ZA-27/graphite particulate composites. Mater Des 16(6),337–341, 1995.[24] Seah, K.H.W., Sharma, S.C., Girish, B.M., Lim,S.C.: Wear characteristics of as-cast ZA-27/graphiteparticulate composites. Mater Des 17(2), 63–67,1996.128 13 th International Conference on Tribology – Serbiatrib’13

Serbian TribologySocietySERBIATRIB ‘1313 th International Conference onTribologyKragujevac, Serbia, 15 – 17 May 2013Faculty of Engineeringin KragujevacWEAR PROPERTIES OF A356/10SiC/1Gr HYBRID COMPOSITESIN LUBRICATED SLIDING CONDITIONSBabić Miroslav 1 , Stojanović Blaža 1 , Mitrović Slobodan 1 , Bobić Ilija 2 , Miloradović Nenad 1 , Pantić Marko 1 ,Džunić Dragan 11 Faculty of Engineering, University of Kragujevac, Kragujevac, Serbia, babic@kg.ac.rs, blaza@kg.ac.rs2 Institution Institute of Nuclear Sciences “Vinca”, University of Belgrade, Belgrade, SerbiaAbstract: This paper presents basic tribological properties of A356/10SiC/1Gr hybrid composites inconditions with lubrication. Hybrid composite specimen is obtained by compocasting procedure. A356aluminium alloy is used as a base matrix alloy, reinforced with 10wt% of SiC and 1wt% of graphite.Tribological tests are done on advanced and computer supported tribometer with block-on-disc contact pair.By the experimental plan, test is conducted under three different values of sliding speed, three differentvalues of normal load, different sliding distances, and also different lubricants. SEM and EDS are used forwear analysis. The analysis has shown the presence of MML, which means that there was transfer ofmaterial from steel disc to composite block.Keywords: Hybrid composites, aluminium, SiC, graphite, wear, lubrication, MML.1. INTRODUCTIONAluminium is the most attractive material inautomotive, airplane, space and precise devicesindustry. Improvement of mechanical andtribological properties of aluminium can beachieved through aluminium reinforcement with theproper material and through creating compositematerial. The most effective improvement of theseproperties is achieved through creating hybridcomposites with two or more types ofreinforcements. By adding the ceramicreinforcement, mechanical properties of the matrixare changed, but in that case problem ofmachinability occurs. To improve machinability,the graphite is added to composite materials that arealready reinforced with ceramic material. Presenceof graphite reduces mechanical properties (hardnessdecreases), but tribological properties are improved[1-5].Basavarajappa et al [6-8] have studied thetribological behaviour of hybrid composites withaluminum base Al2219 reinforced by SiC andgraphite. They studied the tribological properties ofhybrid composites with 5, 10 and 15% SiC and 3%Gr obtained with process of liquid metallurgy. Thetribological tests show that wear decreases withincreasing SiC content in the hybrid composite.With increasing sliding speed and normal load,wear rate of composites is growing. Mahdavi andAkhlaghi [9,10] have studied the tribologicalproperties of Al / SiC / Gr hybrid compositesobtained by In situ Powder Metallurgy process.Aluminum alloy Al 6061 is used as a base,reinforced with graphite 9% and 0 ÷ 40% SiC. Thetribological tests are done on tribometer with pin ondisc contact, and the composite with 20% SiC hasthe best properties. Further increase of SiC leads toincreased wear of hybrid composites.Suresh and Sridhara [11-14] have studied theeffect of SiC content and graphite on thetribological behaviour of hybrid Al / SiC / Grcomposites with aluminum base LM25 (Al-Si7Mg0.5) obtained by stircasting process.Ames and Alpas have [15] studied thetribological testing of hybrid composites with a baseof aluminum alloy A356 reinforced with 20% SiCand 3 ÷ 10% Gr. The tribological tests are done ontribometer with block on ring contact. The wear rateof hybrid composites is significantly lower than thewear rate of the base material withoutreinforcements, especially at low normal loads.Vencl et al [16,17] have studied the tribologicalbehaviour of hybrid composites with the A35613 th International Conference on Tribology – Serbiatrib’13 129

matrix reinforced with SiC, Al2O3 and graphite. Thetribological tests are done on tribometer with pin ondisc contact and show that the wear and frictioncoefficient decreases with addition of graphite.This paper presents tribological behaviour ofhybrid composites with aluminum base of A356alloy reinforced with SiC and Gr obtained withcompocasting procedure. The tests are done oncomputer aided block-on-disc tribometer underlubricated sliding conditions by varying the contactpairs (sliding speed and normal load).2. EXPERIMENT2.1 The procedure for obtaining compositesHybrid Al / SiC / Gr composites are obtained by themodified compo-casting procedure (infiltration ofparticles in the semi-solidified melt A356 alloy). subeutecticAl-Si alloys En AlSiMg0,3 (A356 alloy) isused as a basis. Using compocasting procedure, particlereinforcements are easily infiltrated / trapped. Thissolves the problem of wettability on the border baseand reinforcements. The cost of composite producingwith that process is much lower.Figure 1 shows the structure of the base materialA356 and the hybrid composite with 10wt%SiCand 1wt%Gr. When mixing composites, particles ofgraphite have become fragmented with regard tooriginal size of 35 µm. The picture shows thedistribution of SiC particulate reinforcements, thesize of 39 um.2.2 Plan of experiment and description ofequipmentTribological tests are done on advanced andcomputer supported tribometer with block-on-disccontact pair in accordance with ASTM G77standard. Contact pair consists of rotating disc ofdiameter Dd = 35 mm and broadness bd = 6.35 mm,and a stationary block of size 6.35x15.75x10.16mm 3 . The discs are made of steel 90MnCrV8hardness of 62-64 HRC with grinded surfaces.The tests were performed in lubricated slidingconditions on the samples with the best structural,mechanical and anti-corrosive properties.a)Figure 2. Tribometer.The values of sliding speed (0.25 m / s, 0.5 m / sand 1 m / s) and the normal loads (40N, 80N and120 N) are in accordance with the plan ofexperiment. The tests are performed for slidingdistance of 2400 m.b)Figure 1. The structure of: a) base material A356, andb) hybrid composite Al/10SiC/1Gr.Figure 3. Lubrication of the contact pair.All tests used the same hydraulic lubricant withimproved anti-wear properties, viscosity VG46(ISO 3848). Lubricant is housed in a small tank,130 13 th International Conference on Tribology – Serbiatrib’13

and lubrication is done so that the bottom of thedisc is immersed to up to depth of 3 mm into thesmall tank with lubricant, whose volume is 30 ml.During rotation of the disc, oil is continuouslyintroduced into the zone of the contact and makesboundary lubrication of contact pair (Figure 3).All experiments were repeated 5 times, and themean values of obtained values are taken asauthoritative.3. RESULTS OF TRIBOLOGICAL TESTSResults of tribological tests of hybrid compositeAl/10SiC/1Gr and basic material A356 are shownin the following diagrams.Wear, mm 3 x 10 -2Wear, mm 3 x 10 -28765432987654321A356, F 1 =40NA356, F 2=80NA356, F 3 =120NA356/10SiC/1Gr, F 1 =40NA356/10SiC/1Gr, F 2=80NA35610SiC/1Gr, F 3=120NV=0.25 m/s, lubrication00 500 1000 1500 2000 2500Sliding distance, mA356, F 1=40NA356, F 2=80NA356, F 3=120NA356/10SiC/1Gr, F 1=40NA356/10SiC/1Gr, F 2=80NA356/10SiC/1Gr, F 3=120NV=0.5 m/s, lubricationDiagrams of wear volume are formed on thebasis of wear scar which is obtained by measuringafter 150 m, 300 m, 1200 m, 2400 m, and they aregiven for all three values of sliding speed (Figure4).It is obvious that the wear rate of the hybridcomposites A356/10SiC/1Gr is several times lessthan the wear rate of the base material A356. Withincrease of sliding speed, wear rate of the hybridcomposite A356/10SiC/1Gr and the base materialare decreases. Wear rate dependence has almostlinear dependence for all values of the normal loads(Figure 5).Wear rate, mm 3 x10 -5 /m4.03.53.02.52.01.51.00.5s=2400 m, lubricationA356, F 1=40 NA356, F 2=80 NA356, F 3=120 NA356/10SiC/1Gr, F 1=40 NA356/10SiC/1Gr, F 2=80 NA356/10SiC/1Gr, F 3=120 N0.00.00 0.25 0.50 0.75 1.00 1.25Sliding speed, m/sFigure 5. Wear rate dependence on sliding speed.With increase of normal load, wear rateincreases. This increase is particularly pronouncedat the base material A356 (Figure 6).4.0s=2400 m, lubricationWear, mm 3 x 10 -2100 500 1000 1500 2000 2500Sliding distance, m87654321A356, F 1=40NA356, F 2=80NA356, F 3=120NA356/10SiC/1Gr, F 1=40NA356/10SiC/1Gr, F 2=80NA356/10SiC/1Gr, F 3=120NV=1 m/s, lubricationWear rate, mm 3 x10 -5 /m3.53.02.52.01.51.00.5A356, V 1=0.25 m/sA356, V 2=0.5 m/sA356, V 3=1.0 m/sA356/10SiC/1Gr, V 1=0.25 m/sA356/10SiC/1Gr, V 2=0.5 m/sA356/10SiC/1Gr, V 3=1.0 m/s0.00 40 80 120 160Load, NFigure 6. Wear rate dependence on normal load.00 500 1000 1500 2000 2500Sliding distance, mFigure 4. Wear volume for all three values of slidingspeed.Wear rate dependence on normal load andsliding speed for sliding distance of 2400 m, isshown in Figure 7.13 th International Conference on Tribology – Serbiatrib’13 131

compared to the base material. Decrease of wearrate occurs due to the effects of SiC from hybridcomposite in contact with a steel disc.Wear rate dependence on normal load and slidingspeed are shown in Figures 9 and 10 as the 3D plots.Wear rate is approximated by exponential functionwith a high correlation coefficient.Both tested materials show that at least wearoccurs at the maximum sliding speed of 1 m / s andthe minimum normal load of 40N.Figure 7. Wear rate dependence on normal load andsliding speed.After the tribological tests, SEM analysis isperformed for wear scar of base material A356 andhybrid A356/10SiC/1Gr composite, whose microphotosare shown in Figure 8.Figure 9. Wear rate dependence on the base material.a)Figure 10. Wear rate dependence on the hybridcomposites А356+10SiC+1Gr.b)Figure 8. SEM micro-photos of wear scar:a) base material A356,b) hybrid composite A356/10SiC/1Gr.4. ANALYSES OF OBTAINED RESULTSThe analyses of the obtained tribological resultsshow that the wear rate or wear volume is muchlower in the hybrid composites Al/10SiC/1GrSEM microscopy shows that due to the contactof the SiC composites and Si phases from the basicA356, wear of steel disc occurs. Fe particles enterthe surface layer of the composites and lead to thecreation of mechanically mixed layer (MML). Theformation of MML layer is characteristic ofaluminum alloys reinforced with SiC [18-23]. Ironaccumulates around the SiC particles taking aposition of small particle of graphite. At some partswhite lines appear enriched with iron oxides, whichare consistent with the sliding direction (Figure 11).132 13 th International Conference on Tribology – Serbiatrib’13

5. CONCLUSIONFigure 11. The accumulation of iron in the compositeA356/10SiC/1Gr, 120 N, 0.25 m/s.Figure 12 shows the SEM photograph of part ofthe hybrid composite A356/10SiC/1Gr. Wear scar isobtained by sliding speed of 0.25m/s and normal loadof 120N in conditions of lubrication. At higher loads,the dominant wear mechanism is abrasive wear. SiCparticles (darker) and iron particles (bright colours)are clearly visible on the scar. Confirmation of theseassumptions is obtained by EDS analysis, as shownin the two spectrums. The first spectrum shows thepresence of SiC particles, and the particles of ironand its oxides can be seen on second spectrum .Figure 12. EDS analysis A356/10SiC/1Gr, 120N,0.25m/s, SEM.Wear tests of hybrid compositesA356/10SiC/1Gr show their superior performancein relation to the base material A356. Appliedcompocasting modified procedure, in addition tolow prices, confirms the good distribution ofreinforcements in the composite.Wear rate on A356/10SiC/1Gr hybridcomposites is 3 ÷ 8 times lesser than the wear rateon the base material A356. It is especially bigdifference of wear rate at the lowest sliding speedof 0.25 m / s and maximum normal load of 120N.Wear rate decreases with decrease of normal loadand increase of sliding speed.SEM microscopy and EDS analysis confirm agood distribution of SiC reinforcements in thehybrid composite. Also, advent mechanically mixedlayer (MML) is obvious, respectively, theappearance of iron and its oxides in the hybridcomposite.ACKNOWLEDGMENTSThis paper presents the research resultsobtained within the framework of the project TR–35021, financially supported by the Ministry ofEducation and Science of the Republic of Serbia.REFERENCES[1] A. Vencl: Tribology of the Al-Si alloy based MMCsand their application in automotive industry, u:Magagnin L. (ed.), Engineered Metal MatrixComposites: Forming Methods, Material Propertiesand Industrial Applications, Nova Science Publishers,Inc., New York (SAD), pp. 127-166, 2012.[2] B. Stojanovic, M. Babic, S. Mitrovic, A. Vencl, N.Miloradovic, M. Pantic: Tribological characteristicsof aluminium hybrid composites reinforced withsilicon carbide and graphite. A review, Journal ofthe Balkan Tribological Association, Vol.19, No.1,pp. 83-96, 2013.[3] N. Miloradovic, B. Stojanovic: Tribologicalbehaviour of ZA27/10SiC/1Gr hybrid composite,Journal of the Balkan Tribological Association,Vol.19, No.1, pp. 97-105, 2013.[4] S. Mitrovic, M. Babic, B. Stojanovic, N.Miloradovic, M. Pantic, D. Dzunic: TribologicalPotential of Hybrid Composites Based on Zinc andAluminium Alloys Reinforced with SiC and GraphiteParticles, Tribology in industry, Vol. 34, No. 4,pp.177-185, 2012.[5] P. Ravindran, K. Manisekar, R. Narayanasamy, P.Narayanasamy: Tribological behaviour of powdermetallurgy-processed aluminium hybrid compositeswith the addition of graphite solid lubricant,Ceramics International, Vol. 39, No. 2, pp. 1169-1182, 2013.13 th International Conference on Tribology – Serbiatrib’13 133

[6] S. Basavarajappa, G. Chandramohan, K. Mukund,M. Ashwin, M. Prabu: Dry sliding wear behavior ofAl 2219/SiCp-Gr hybrid metal matrix composites,Journal of Materials Engineering and Performance,Vol. 15, No. 6, pp.668-674, 2006.[7] S. Basavarajappa, G. Chandramohan: Dry slidingwear behavior of hybrid metal matrix composites,Materials Science, Vol.11, No.3, pp. 253-257, 2005.[8] S. Basavarajappa, G. Chandramohan, A.Mahadevan: Influence of sliding speed on the drysliding wear behaviour and the subsurfacedeformation on hybrid metal matrix composite,Wear, Vol. 262, pp. 1007–1012, 2007.[9] S. Mahdavi, F. Akhlaghi: Effect of SiC content onthe processing, compaction behavior, and propertiesof Al6061/SiC/Gr hybrid composites, Journal ofMaterials Science, Vol. 46, No. 5, pp. 1502-1511, 2011.[10] F. Akhlaghi, S. Mahdavi: Effect of the SiC Contenton the Tribological Properties of Hybrid Al/Gr/SiCComposites Processed by In Situ PowderMetallurgy (IPM) Method, Advanced MaterialsResearch, pp. 1878-1886, 2011.[11] S. Suresha, B.K. Sridhara: Wear characteristics ofhybrid aluminium matrix composites reinforced withgraphite and silicon carbide particulates, OriginalResearch Article, Composites Science andTechnology, Vol. 70, No. 11, pp. 1652-1659, 2010.[12] S. Suresha, B.K. Sridhara: Effect of addition ofgraphite particulates on the wear behavior inaluminium–silicon carbide–graphite composites,Materials & Design, Vol. 31, pp. 1804–1812, 2010.[13] S. Suresha, B.K. Sridhara: Effect of silicon carbideparticulates on wear resistance of graphiticaluminium matrix composites, Materials & Design,Vol. 31, No. 9, pp. 4470-4477, 2010.[14] S. Suresha, B.K. Sridhara: Friction characteristicsof aluminium silicon carbide graphite hybridcomposites, Materials & Design, Vol.34, pp. 576-583, 2012.[15] W. Ames, A.T. Alpas: Wear mechanisms in hybridcomposites of graphite-20% SiC in A356 aluminumalloy, Metall.Mater.Trans. A; Vol. 26, pp. 85-98, 1995.[16] A. Vencl, I. Bobic, S. Arostegui, B. Bobic, A.Marinkovic, M. Babic: Structural, mechanical andtribological properties of A356 aluminium alloyreinforced with Al 2 O 3 , SiC and SiC + graphiteparticles, Journal of Alloys and Compounds, Vol.506, pp. 631-639, 2010.[17] A.Vencl, I.Bobic, B.Stojanovic, Tribologicalproperties of A356 Al-Si alloy composites under drysliding conditions, Industrial Lubrication andTribology, Vol. 66, No.3, 2014 (in print).[18] J. Rodriguez, P. Poza, M.A. Garrido, A. Rico: Drysliding wear behaviour of aluminium–lithium alloysreinforced with SiC particles, Wear, Vol. 262, pp.292–300, 2007.[19] L. Jung-moo, K. Suk-bong, H. Jianmin: Dry slidingwear of MAO-coated A356/ 20 vol.% SiCpcomposites in the temperature range 25–180°C,Wear, Vol. 264, pp. 75–85, 2008.[20] R.N. Rao, S. Das, D.P. Mondal, G. Dixit: Effect ofheat treatment on the sliding wear behaviour ofaluminium alloy (Al– Zn–Mg) hard particlecomposite, Tribology International, Vol. 43, pp.330–339, 2010.[21] M. Gui, S.B. Kang: Dry sliding wear behavior ofplasma-sprayed aluminum hybrid compositecoating, Metall. Mater. Trans., A Vol. 32A, pp.2383–2392, 2001.[22] A.K. Mondal, S. Kumar: Dry sliding wearbehaviour of magnesium alloy based hybridcomposites in the longitudinal direction, Wear, Vol.267, pp. 458–466, 2009.[23] R.N. Rao, S. Das, D.P. Mondal, G. Dixit: Drysliding wear behaviour of cast high strengthaluminium alloy (Al–Zn–Mg) and hard particlecomposites, Wear, Vol. 267, pp. 1688–1695, 2009.134 13 th International Conference on Tribology – Serbiatrib’13

Serbian TribologySocietySERBIATRIB ‘1313 th International Conference onTribologyKragujevac, Serbia, 15 – 17 May 2013Faculty of Engineeringin KragujevacA REVIEW OF THE TRIBOLOGICAL PROPERTIES OF PTFECOMPOSITES FILLED WITH GLASS, GRAPHITE, CARBON ORBRONZE REINFORCEMENTMiloš Stanković 1 , Aleksandar Vencl 2 , Aleksandar Marinković 21 University of Belgrade, Inovation Center of Faculty of Mechanical Engineering, Belgrade, Serbia,mstankovic@mas.bg.ac.rs2 University of Belgrade, Faculty of Mechanical Engineering, Belgrade, Serbia,avencl@mas.bg.ac.rs , amarinkovic@mas.bg.ac.rsAbstract: Polytetrafluroethylene (PTFE) is currently finding increasing utility due to its unique propertieslike high chemical resistivity, low coefficient of friction and high temperature stability. However, PTFEexhibits poor wear resistance, especially abrasion. The wear resistance of PTFE can be significantlyimproved by addition of suitable reinforcement (filler) materials. Among the most common filler materialsare glass fibers, graphite, carbon and bronze. In this paper, it is presented a review of tribologicalproperties of composite materials with PTFE matrix and above mentioned filler materials.Keywords: PTFE, composites, glass fibres, graphite, carbon fibres, bronze, friction, wear.1. INTRODUCTIONNowadays, there is very intensive growth in thelarge scale production of the fibre reinforcedpolymer composites since they posses certainadvantages over the metals. The advantages includelower density, less need for maintenance and lowercost [1]. Polymers and polymers reinforced withfibres are used for producing of various mechanicalcomponents, such as gears, cams, wheels, brakes,clutches, bush bearing and seals [2]. Considerableefforts are being made to extend the range ofapplications. Such use would provide the economicaland functional benefits to both manufacturers andconsumers. Many researchers have studied thetribological behaviour of polytetrafluroethylene(PTFE). Studies have been conducted with variousshapes, sizes, types and compositions of fibres. Ingeneral these composites exhibit lower wear andfriction when compared to pure PTFE.The most commonly used reinforcements(fillers) for tribological applications are carbon,graphite, bronze and glass. Generally the fillersimprove the wear resistance from 10 to 500 times,depending on the filler type and shape. On the otherhand, coefficient of friction of various PTFEcomposites is strongly dependent on filler crystalstructure and improvements are not so significant.Subject of this paper are composites with PTFEmatrix. Filler materials investigated in this paperare glass, graphite, carbon and bronze. Analysedtribological characteristics are coefficient of frictionand wear. In most cases, friction and wear testswere carried out on pin-on-disc tribometer atambient conditions (temperature and humidity).Counterpart material used in the experiments wasalways harder than composite. In most cases,counterpart material was steel. All tests werecarried out at dry sliding conditions. Overall resultsfrom all analysed papers are summarized andpresented in Table 1.2. FACTORS AFFECTING POLYMERCOMPOSITE WEAR AND FRICTIONThere are many factors that affect materialtribological properties. For polymer composites themost influential are [6]:Normal load: In order for a polymer compositeto function as a solid lubricant it must be able tosupport the load, as well as the tangential stressesinduced by sliding. At high loads severe wear13 th International Conference on Tribology – Serbiatrib’13 135

Table 1. Overview of the results from the analysed papers (for pure PTFE and its composites at dry sliding conditions)*Ref.[7]Testrig**Ballon-disc[8] Pin-ondisc[9] Pin-ondisc[10] Pin-ondisc[11] Pin-ondiscType (amount and size) of thefiller***1. C particles (18 %; 10 – 25 μm)+ Gr flakes (7 %; 25 – 50 μm);2. E-glass fibres(15 %; 10 × 50 – 75 μm);3. E-glass fibres(25 %; 10 × 50 – 75 μm)1. Glass fibres (17 %);2. Bronze (25 %);3. C (35 %)1. Glass particles(25 %; 40 μm);2. Bronze particles(40 %; 48 μm)1. Gr flake (2 %; 10 μm);2. Gr flake (5 %; 10 μm);3. Gr flake (10 %; 10 μm)1. Bronze (25 %);2. Bronze (40 %);3. Bronze (60 %)CounterpartmaterialAISI 440Csteel ball(d = 9 mm)AISI 440Csteel discEN 32hardenedsteel discStainlesssteel discAISI 400Csteel discLoad,N (MPa)5(pointcontact)5 – 30(0.2 – 1.1)60(0.6)25(0.2)Slidingspeed,m/sSlidingdistance,mCoefficient offrictionSpecific wear rate,mm 3 /Nm × 10 –6PTFE Composites PTFE Composites0.1 1000 0.11 0.13 – 0.16 950 90 – 7000.32 –1.281152 –46080.13 –0.790.11 – 0.711.5 2500 – –476 –943app.5676 – 290app. 4.4 –3351 8000 0.24 0.20 – 0.26 1650 10 – 190app.5 – 200(0.2 – 7.1) 0.32 – 2 2000 0.12 –0.22app.0.12 – 0.181000 1 – 100* The friction and wear values in the table are approximate and can be used only as a guidance, since the authors in most cases didnot presented the results in appropriate way; ** Pin – cylindrically shaped specimen (flat contact); *** C – carbon, Gr – graphiteoccur, characterized by brittle fracture or severeplastic deformation. On the other hand, at low loadsusually mild wear occur, characterized by the localplastic flow of the thin transfer film and surfacelayers (decreasing friction), together withdelamination wear.Contact area: The contact area will determine theprojected contact stresses. If the load cannot bereduced, one way of reducing stress is to increase theprojected contact area. However, if the area ofcontact becomes too large instead of the materialflowing across the counterpart surface, it will have atendency to build up, forming ridges, which cancause high localized stresses and higher adhesion,thus higher friction and wear. It is important to designa part with correct match up of load and contact area.Sliding speed: The high sliding speeds canproduce high temperatures due to friction heating.This may cause the polymer or the polymercomposite additives to degrade. However in somecases higher temperature might be beneficial to thelubricating process. In order to develop a surfaceshear film and/or a transfer film, the molecular chainmust have time to reorient. If one slides too fast overthese un-oriented chains, instead of reorienting, theywill fracture, leading to the production of large wearparticles and high wear. Thus it is important tochoose sliding speed for each particular polymer toensure the optimum performance.Counterpart topography: If the counterpartmaterial is too rough it can abrade the compositeand not allow a shear film or transfer film to form.Therefore, it could be generally accepted that, thesmoother the counterpart the lower the wear. Thishas certain limits, since it is also found that overpolishingtend to remove the counterpart softermatrix material, leaving the harder phases and/orparticles protruded above the surface.Temperature: At lower temperature the frictionand wear properties of most polymers are not asexceptional as they are at or above the ambienttemperature. At lower temperatures polymers losetheir relaxations ability, i.e. the movement of theirmain molecule chain do not obtains adequatedegree of freedom, and thus the polymer does notobtain a great deal of plasticity. High temperaturescan affect bonding between the filler material andpolymer matrix. They can also affect the lubricatingproperties of some additives in polymer composite,since these additives might desorbs gases at certaintemperature or even decompose.3. STATE-OF-THE-ART OF PTFECOMPOSITES TRIBOLOGICALRESEARCHESTribological behaviour of PTFE and itscomposites with filler materials such as carbonparticles, graphite flakes and E glass fibres (Table 2)was investigated by Khedkar et al. [7]. Experimentswere performed under the normal load of 5 N andsliding speed of 0.1 m/s. They found that the usedfiller additions increase wear resistance in allcomposites that were studied. The highest wear136 13 th International Conference on Tribology – Serbiatrib’13

esistance was found for composite containing 18vol. % of carbon and 7 vol. % of graphite (Figure1). The coefficient of friction values were from 0.11to 0.16 (Figure 2). This behaviour can be attributedto the presence of hard carbon particles, which areembedded within the matrix and impart additionalstrength to the composite. Wear testing and SEManalysis showed that three-body abrasion wasprobably the dominant mode of failure for PTFE +18 vol. % carbon + 7 vol. % graphite composite.Table 2. Composition (vol. %) of materialsDesignationSpecific wear rate, mm 3 /Nm × 10 –4AB109876543210MaterialPure PTFE75 % PTFE + 18 %amorphous carbon + 7 %hexagonal graphiteCharacteristic offiller material99 % pure PTFEpowderCarbon particles(diameter: 10 – 25μm); graphite flakes(size: 25 – 50 μm)C 85 % PTFE + 15 % E-glass E-glass fibres(diameter: 10 μm;D 75 % PTFE + 25 % E-glass length: 50 – 75 μm)5 N; 0.1 m/ssA B C DMaterialFigure 1. Average specific wear rate of PTFE and PTFEcomposites (adopted from [7])Coefficient of friction0.300.250.200.150.100.050.00ACBD0 200 400 600 800 1000Sliding distance, mFigure 2. Frictional behaviour of PTFE and PTFEcomposites (adopted from [7])Unal et al. [8] studied PTFE composites filledwith glass fibres (17 %), bronze (25 %) or carbon(35 %). Experiments were performed under loadrange from 5 to 30 N (0.18 – 1.06 MPa) and speedrange from 0.32 to 1.28 m/s. The results showedthat, for pure PTFE and its composites, thecoefficient of friction decrease with the increase inload. For the ranges of load and speed used in thisinvestigation, the coefficient of friction showedvery little sensitivity to the sliding speed and largesensitivity to the applied load, particularly at highload values. Figure 3 shows that sensitivity for thepure PTFE, but it is quiet similar for composites, aswell. Adding glass fibres, bronze and carbon fillersto PTFE were found effective in reducing the wearrate. The maximum reductions in wear rate andcoefficient of friction were obtained by PTFEreinforced with 17 % of glass fibres. The specificwear rate for PTFE + 17 % glass fibres was almosttwo orders of magnitude lower than for pure PTFE.By means of microscopy, it is noticed that thePTFE with glass fibre filler form a good thin anduniform transfer film which have positive influenceto the wear rate.Coefficient of friction0.300.250.200.150.100.050.000.32 m/s0.64 m/s0.96 m/s1.28 m/sPTFECoefficient of friction0.800.700.600.500.400.300.200.100.00PTFE5 N10 N20 N30 NFigure 3. Sensitivity of PTFE coefficient of friction tothe sliding speed (for 20 N load) and applied load (for0.32 m/s speed) (adapted from [8])A single influence of glass particles (25 vol. %;40 μm) and bronze particles (40 vol. %; 48 μm) onwear behaviour of PTFE based composites wasstudied by Mudasar Pasha et al. [9]. The tests weredone on a pin-on-disc tribometer with differentnormal loads (20 – 100 N, i.e. 0.2 – 1 MPa), slidingspeeds (1.5 – 5.5 m/s) and distances (500 – 2500m). The experimental results indicate that theweight loss increases with increasing load andsliding speed (Figure 4). The PTFE + 40 % bronzecomposite exhibits better wear resistance compareto the others (Figure 5). The transfer film formedon the counterpart surface, sliding against PTFE +40 % bronze, is smooth, thin and uniform, which13 th International Conference on Tribology – Serbiatrib’13 137

indicates that the formation of adhesive strengthbetween the transfer films and the counterpartsurface is strong. In the case of the counterpartsliding against the pure PTFE, the transfer film isvery thick.Figure 4. Weight loss vs. sliding speed at constantapplied load of 60 N and sliding distance of 1500 mFigure 6. Specific wear rate of PTFE with variouscontent of graphite flakesCompared to pure PTFE, composites showedstable coefficient of friction (Figure 7). The lowestcoefficient of friction was 0.20 for composites with2 and 5 wt. % of graphite. For composite with 10wt. % of graphite, the coefficient of friction wasslightly higher and close to the value obtained withpure PTFE. Anyway, variations of the coefficient offriction are very small.Figure 5. Wear curves of tested materials for 60 N loadand speed of 1.5 m/sEvaluation of the wear rate and coefficient offriction for graphite flake (2, 5 and 10 wt. %; 10μm) filled PTFE composites were studied byGoyal and Yadav [10]. It was performed on apin-on-disk tribometer under dry slidingconditions, at sliding speed of 1 m/s and 25 Nload (0.19 MPa), during 8000 m. A significantdecrease in wear of composites in compare topure PTFE is noticed. The wear rates ofcomposites with 5 and 10 wt. % of graphite weredecreased 22 and 245 times, respectively (Figure6). This decrease in wear rate is also attributed tothe formation of a thin and tenacious transfer filmon the counterpart surface.Figure 7. Coefficient of friction with various content ofgraphite flakesUnal et al. [11] also studied the friction andwear behaviour of pure PTFE and bronze (25, 40and 60 %) filled PTFE polymer composites underapplied load range from 5 to 200 N (0.18 – 7.07MPa) and sliding speed range from 0.32 to 2.0 m/s,using a pin-on-disc tribometer. The results showedthat bronze filled PTFE composite exhibited lowercoefficient of friction (Figure 8 and Figure 9) andhigher wear resistance (Figure 10) compared topure PTFE. The coefficient of friction of both –pure PTFE and bronze filled PTFE compositesdecreases when the applied load is increased from 5to 30 N (light condition). Above 30 N thecoefficient of friction remains stable.The PTFE filled with 60 % of bronze showedhigher wear resistance than pure PTFE. Thisbehaviour can be attributed to the presence of138 13 th International Conference on Tribology – Serbiatrib’13

onze, which is embedded within the matrixmaterial and impart additional strength to thecomposite. The applied load has shown moreinfluence on the wear behaviour of PTFE and PTFEcomposite than the sliding speed.In addition, it could be interested to present adiagram that shows dependence of specific wearrate on wt. % of filler materials (Figure 11) [12].Although there are filler materials which are notsubject of this paper, it might be interesting tocompare specific wear rate for different polymercomposites.Figure 8. Variation of coefficient of friction with slidingdistance (normal load: 50 N; sliding speed: 1.5 m/s)Figure 11. Specific wear rate vs. wt. % of filler materialfor various polymer composites(FEP – fluorinated ethylene propylene; PA6 – polyamide6; HDPE – high-density polyethylene; PA11 –polyamide 11; GF – glass fibre; PEEK – polyether etherketone; CNT – carbon nanotube)4. CONCLUSIONFigure 9. Variation of coefficient of friction withapplied load and sliding speed for PTFE andPTFE + 60 % bronze compositeIn this review paper, the tribological behaviourof PTFE composites, filled with glass fibres,graphite flakes, carbon, bronze or combination ofmentioned filler materials has been analysed.It is noticed that coefficient of friction usuallyremains in the range from 0.1 to 0.3. When purePTFE is compared with PTFE composites, there isonly slight decrease in coefficient of friction withalmost all analysed composites.On the other hand, regarding to wear of purePTFE and its composites, influence of fillermaterials is quiet significant. Presence of fillermaterial can increase wear resistance (decreasewear) up to 2 to 3 orders of magnitude. However, itcan be concluded that the best results, regarding towear resistance, were obtained by PTFE + bronzecomposites. Nevertheless, if we compare it to theother composites, that difference is not thatsignificant, i.e. similar wear resistance can beobtained with appropriate amount of othermentioned filler materials.ACKNOWLEDGEMENTFigure 10. Variation of specific wear rate with appliedload and sliding speed for PTFE and PTFE compositesThis work has been performed within theprojects TR 35021, TR 34028 and TR 35011. Theseprojects are supported by the Republic of Serbia,Ministry of Education, Science and TechnologicalDevelopment, whose financial help is gratefullyacknowledged.13 th International Conference on Tribology – Serbiatrib’13 139

REFERENCES[1] W. Brostow, V. Kovačević, D. Vrsaljko, J.Whitworth: Tribology of polymers and polymerbasedcomposites, Journal of Materials Education32, pp. 273-290, 2010.[2] K. Friedrich, P. Reinicke: Friction and wear ofpolymer-based composites, Mechanics of CompositeMaterials 34, pp. 503-514, 1998.[3] W.A. Glaeser: Materials for Tribology, Elsevier,Amsterdam, 1992.[4] N.K. Myshkin, A.Yа. Grigoriev, V.L. Basiniuk, A.I.Mardasevich, V.G. Kudritsky, I.N. Кavaliova:Equipments and materials for tribotesting in openspace on International Space Lab, Tribology inIndustry 33, pp. 43-46, 2011.[5] I. Zidaru, R.G. Ripeanu, I. Tudor, A.C. Drumeanu:Research regarding the improvements oftribological behavior in three cone bits bearings,FME Transactions 37, pp. 99-102, 2009.[6] P.V. Kulkarni, N.K. Chapkhhane: Development andtesting of PTFE based composite bearing materialfor turbine pump, International Journal ofEngineering and Advanced Technology (IJEAT) 1,pp. 15-20, 2012.[7] J. Khedkar, I. Negulescu, E.I. Meletis: Sliding wearbehavior of PTFE composites, Wear 252,pp. 361-369, 2002.[8] H. Unal, A. Mimaroglu, U. Kadıoglu, H. Ekiz:Sliding friction and wear behaviour ofpolytetrafluoroethylene and its composites underdry conditions, Materials & Design 25, pp. 239-245,2004.[9] B.A. Mudasar Pasha, D. Abdul Budan, S.Basavarajappa, S. Manjunath Yadav, B.A.Nizamuddin: Dry sliding wear behaviour of PTFEfilled with glass and bronze particles, Tribology –Materials, Surfaces & Interfaces 5, pp. 59-64, 2011.[10] R.K. Goyal, M. Yadav: Study on wear and frictionbehavior of graphite flake-filled PTFE composites,Journal of Applied Polymer Science 127, pp. 3186-3191, 2013.[11] H. Unal, E. Kurtulus, A. Mimaroglu, M. Aydin:Tribological performance of PTFE bronze filledcomposites under wide range of applicationconditions, Journal of Reinforced Plastics andComposites 29, pp. 2184-2191, 2010.[12] D.L. Burris, W.G. Sawyer: A low friction and ultralow wear rate PEEK/PTFE composite, Wear 261,pp. 410-418, 2006.140 13 th International Conference on Tribology – Serbiatrib’13

Serbian TribologySocietySERBIATRIB‘1313th International Conference onTribologyKragujevac,Serbia, 15 – 17 May 2013Faculty of Engineeringin KragujevaccWEARCHARACTERISTICSS OF HYBRID COMPOOSITES BASEDON ZA27 ALLOY REINFORCED WITH SILICON CARBIDE ANDGRAPHIITE PARTICLESSlobodanMitrović 1 , Miroslav Babić 1 , Nenad Miloradović 1 , Ilija Bobić 2 , Blažaa Stojanovićć 1 , DraganDžunić 11 Faculty of Engineering University of Kragujevac, Kragujevac, Serbia, boban@kg.acb.rs, babic@kg.ac.rs,mnenad@kg.ac.rs,blaza@kg.ac. .rs, dzuna@kg.ac.rs,2Institute of Nuclear Sciences ”Vinča”, University of Belgrade, Belgrade, Serbia, ilijab@vinca.rsAbstract: The paper presents the wear characteristics of a hybrid composite based on zinc-aluminium ZA27alloy, reinforced with silicon-carbide and graphite particles. The tested sample contains 5 vol.%of SiC and3 vol.% Gr particles. Compocastingtechnique has been used to preparee the samples. The experiments wereeperformed on a “block-on-disc” weree determinedd by varying the normal loads and sliding speeds. . The paper contains theprocedure for preparation of sample compositess and microstructure of the composite material and the baseZA27 alloy. The wear surface of the composite materialwas examined using the scanningelectroniccmicroscope (SEM) andenergy dispersive spectrometry (EDS). Conclusions were e obtained based on theobserved impact of thesliding speed, normal load and sliding distance on tribological behaviour of theobserved composite.Keywords: ZA27 alloy, hybrid composites, wear r characteristicstribometer under conditions of dry sliding. The wear volumes of the alloyand the composite 1. INTRODUCTIONThe composite material represents the solidconnection of the two or more constituents, whichare joined into the unbreakableconnection, forobtaining the better mechanical, tribological t andother characteristics.Metal matrix composites (MMCs) have attractedconsiderable attention recently because of theirpotential advantages over monolithic alloys [1 - 3].Zinc–aluminium (ZA) alloys have emergedd asimportant material fortribological applications,especially suitable forhigh-load and low-speedapplications. Commercially available ZA alloyss arecharacterized by good tribomechanical properties,low weight, excellentfoundry castability andfluidity, good machiningproperties, low initial cost,and environmentally friendly technology The zincalloy with increased content of aluminium is one ofthe alloys, which can beused for manufacturingg themetal matrix composites. The ZA27 alloyy isconsidered as the most prospective for f obtainingg thecomposites, since it is convenient as a substrate forseveral methods of composites manufacturing.Besides, it is also convenient for the heat treatmenttandplastic forming, thus it is possible to aposteriori influence the mechanical properties ofthe final products [4 - 7].Various methods and techniques were used tosubstantially improve wear behaviour of Zn–Al(zinc–aluminium) alloysswithout loweringsignificantly the mechanical properties of materials[6 - 8].Thanks to mentioned properties and goodcharacteristics,zinc-aluminium alloys have inspiredresearchers too reinforce them with differentdispersed reinforcement materials (SiC, Al 2 O 3 ,graphite and garnet) in order to obtain much moreenhanced mechanical and d tribological properties[9 - 16].Consideringg the above,ZA27/ SiC/graphitecomposites may be a good alternative to zinc-aluminium alloys in many industrial applications.13 th International Conference on Tribology – Serbiatrib’13141

2. EXPERIMENTALPROCEDURE2.1 Preparation of the compositesHybrid ZA27/SiC/Graphite composites havebeen successfully prepared using the compocastingprocedure.Hybrid ZA27/SiC composite reinforced r withgraphite was produced at the Laboratory formaterials of the Institute for nuclear sciences“Vinca”. The obtained chemical structure of ZA27alloy thatt was usedduring experimentalinvestigations coincides with thecorrespondingchemical structure defined by standard. Averagesize of SiC particles was 26 μm, while average sizeof graphite particles was15 μm.The applied compocasting procedure consistedof two phases. During the first phase, infiltrationn ofthe particles from the secondary phases into thesemisolid melt of the basic alloy was conductedwith constant mechanical blending. Obtainedcomposite casts were then subjectedto hot pressingduring the second phase. This was done in order todecrease porosity and get better connectionbetween thematrix andthe reinforcement particles.At the same time, better mechanical characteristicof the composite material were obtained.2.2 Structure of the sample materialsMicrostructure of ZA27 alloyand obtainedcomposite were observed by metallurgy microscopeand presented in Figure 1 and 2. Stucture of f thesample of ZA27 alloy ismainly dendrite (Figuree 1).Distinct uniformity of the structure was present,which indicates a favourable ratioof mechanicalproperties of the tested materials.Figure 2. Microstructure of ZА27/3% SiC/3% Grcomposite (magnification х 400)A uniform distributionn of SiC and graphiteparticles was present in tested composite materials.2.3 Testing methodsExperimental tests were performed at the Centreefor tribology of the Faculty of Engineering,University of Kragujevac. KThe tests of the ZA27/SiC/Grcomposite’sstribological characteristicswere performed on thecomputer supported tribometer with “block-on--disc” contact geometry (Figure 3). Tribometerrprovides variation of contact conditionss in terms ofshape, dimension and material of contact elements,normal contact load and sliding speed.Figure 1. Microstructure of ZA27 alloyFigure 2 shows the microstructure ofobtained composite.theFigure 3. The “block-on-disc” tribometerBased on the measured wear scar width on thecontact surfacee obtained by variationof normalloads and sliding speeds, the material wear volumewascalculated. The tests were performed in drysliding conditions, with variation of sliding speedlevels (0.25 m/ /s, 0.5 m/s and 1 m/s) and contacttload levels (100 N, 20 N and 30 N). The observedsliding distances during tests were: 30 m, 60 m,90 m, 150 m and 300 m.14213 th International Conference C onn Tribology – Serbiatrib’13

The test contact pair meets the requirements of theASTM G77-05 standard. It consists of the rotationaldisc with the diameter of D d =35 mm and the width ofb d =6.35 mm and of the stationary block of the width ofb b =6.35 mm, the length of l b =15.75 mm and the heightof h b =10.16 mm. The discs were made of 90MnV8steel with hardness of 62 HRC and the surfacesroughness of R a =0.40 m. The blocks were made of thetested ZA27/5%SiC/3%Gr.3. THE RESULTS AND DISCUSSIONThe variations of dry sliding wear volume loss arepresented in corresponding diagrams in the paper,depending on the sliding distance and for differentvalues of sliding speeds and contact loads. The resultsof wear for given hybrid composite and for ZA27alloy were presented in the same diagrams in order tounderstand the wear process evolution during testsand to make corresponding comparisons. Solid lineson the diagrams refer to the wear scar widths of thecomposite, while the wear scar widths of the ZA27alloy are denoted by dashed lines.The variation of dry sliding wear volume losswith the sliding distance for different applied loadsand for a sliding speed of 0.25 m/s is presented inFigure 4. All diagrams are given for the slidingdistance of 300 m.Wear, mm 3 x10 -3Figure 4. Variation of wear volume of ZA27 alloy andZA27/5%SiC/3%Gr composite against sliding distance fordifferent contact loads and for sliding speed of v=0.25 m/s.Wear, mm 3 x10 -3140012001000800600400200140012001000ZA27, F1=10 NZA27, F2=20 NZA27, F3=30 NComposite, F1=10 NComposite, F2=20 NComposite, F3=30 Nv 1 = 0.25 m/s00 50 100 150 200 250 300800600400200ZA27, F1=10 NZA27, F2=20 NZA27, F3=30 NComposite, F1=10 NComposite, F2=20 NComposite, F3=30 Nv 2 = 0.5 m/sSliding distance, m00 50 100 150 200 250 300Sliding distance, mFigure 5. Variation of wear volume of ZA27 alloy andZA27/5%SiC/3%Gr composite against sliding distance fordifferent contact loads and for sliding speed of v=0.5 m/s.The variation of wear volume loss of ZA27alloy and ZA27/5%SiC/3%Gr composite dependingon the sliding distance and for different appliedcontact loads and the sliding speed of 0.5 m/s maybe seen in Figure 5.Diagram in the Figure 6 presents the variation ofwear volume loss of ZA27 alloy andZA27/5%SiC/3%Gr composite depending on thesliding distance and for different applied contactloads and the sliding speed of 1 m/s.Wear, mm 3 x10 -314001200100080060040020000 50 100 150 200 250 300Figure 6. Variation of wear volume of ZA27 alloy andZA27/5%SiC/3%Gr composite against sliding distance fordifferent contact loads and for sliding speed of v=1 m/s.Wear, mm 3 x10 -3140012001000800600400200v 3 = 1 m/sZA27, F1=10 NZA27, F2=20 NZA27, F3=30 NComposite, F1=10 NComposite, F2=20 NComposite, F3=30 NZA27, F1=10 NZA27, F2=20 NZA27, F3=30 NComposite, F1=10 NComposite, F2=20 NComposite, F3=30 NSliding distance, ms = 300 m00 0.2 0.4 0.6 0.8 1 1.2Sliding speed, m/sFigure 7. Wear volume of ZA27/5%SiC/3%Gr compositeand ZA27 alloy depending on sliding speeds, for differentcontact loads and for sliding distance of 300 m.The wear volume losses of the alloy and thecomposite increase with the increase of the slidingdistance. The wear volume loss curves are of thesame character, both for alloy and for the observedcomposite material. The only difference may beseen in level of wear. At the beginning, a largerslope of the curves is noticeable, so there is theintensive initial wear of the composite material. Arapid increase of wear volume loss is characteristicfor sliding distance of approximately 35 m. Afterreaching the zone of constant wear, the wearvolume loss has slight, almost linear increase.Generally, the wear behaviour of the testedmaterials is characterized by very intensive wearduring initial period, after which there is a period ofstabilization. Wear of the composites was alwayssignificantly lower when compared to wear of thematrix ZA27 alloy.13 th International Conference on Tribology – Serbiatrib’13 143

The influence of the sliding speed on wearvolume for both materials is shown in Figure 7, fordifferent values of normal loads.The effects of the normal load on wear volumeof both composite and alloy is presented inFigure 8, for different values of sliding speeds andfor sliding distance of 300 mWear, mm 3 x10 -314001200100080060040020000 5 10 15 20 25 30 35Figure 8. Wear volume of ZA27/5%SiC/3%Gr compositeand ZA27 alloy depending on contact loads, for differentsliding speeds and for sliding distance of 300 m.Analytical and graphical variations in wear ratedue to changes of sliding speeds and normal loadsin dry sliding conditions are presented in Figures 9and 10. Exponential regression functions wereadopted. Corresponding regression functionscoefficients and curvilinear correlation indices wereobtained showing the good correlation betweenexperimental data and used empirical distributions.Wear rate of ZA27 alloy in dry slidingconditions is presented in Figure 9 and wear rate ofZA27/5%SiC/3%Gr composite in dry slidingconditions is shown in Figure 10. Variations of thewear rate as a function of the sliding speed andnormal load are graphically presented for the ZA27alloy and composite materials.Both tested materials share basically the samenature of wear process development in all contactconditions. The observed composite material hasbetter wear resistance, under the same test conditions.I, mm 3 /m x 10 -3432100.4ZA27, v1=0.25 m/sZA27, v2=0.5 m/sZA27, v3=1 m/sComposite, v1=0.25 m/sComposite, v2=0.5 m/sComposite, v3=1 m/s0.6v, m/s0.81ZA2710s = 300 mLoad, NFigure 9. Wear rate of ZA27 alloy in dry slidingconditions.1520Fn, N25304.44.243.83.63.43.2I, mm 3 /m x 10 -3Figure 10. Wear rate of ZA27/5%SiC/3%Gr compositein dry sliding conditions.The comparative histograms of the wear volumeformed after 300 m of sliding distance, dependingon the contact conditions (the sliding speed and thenormal force) for the basic ZA27 alloy andZA27/5%SiC/3%Gr composite materials are shownin Figure 11.Wear, mm 3 x 10 ‐31400120010008006004002002.51.50.5032100.40.6v, m/sZA27ZA27/5%SiC/3%GrZA27+5%SiC+3%Gr0.8110Figure 11. Comparative histograms of wear volume ofZA27 alloy and ZA27/5%SiC/3%Gr composite.Analysis of histograms in Figure 11 shows that atrend of increase of wear with the increase of normalload may be observed. The increase of sliding speedinduces also the increase of wear. This observation isvalid for both tested materials. It may be noticed thatthe wear of the tested ZA27 alloy is alwayssignificuly higher compared to wear of thecomposite with addition of the SiC and graphiteparticles.From the Figure 11, the influence of the normalload and sliding speed on the wear magnitude maybe clearly noticed. The wear rate increases both withthe increase of the normal force the increase of thesliding speed. The largest value of wear correspondsto the highest sliding speed (v 3 = 1 m/s) and to thehighest value of the normal contact load (F 3 = 30 N).For the lowest sliding speed (v 1 = 0.25 m/s) and thelowest load (F 1 = 10 N), the smallest wear valueswere recorded.Characterization of the microstructure of wearsurface for metal matrix composites is morecomplex than that of the metals or alloys and anunderstanding of wear mechanisms is far from1520Fn, Nv1‐F1 v1‐F2 v1‐F3 v2‐F1 v2‐F2 v2‐F3 v3‐F1 v3‐F2 v3‐F325302.82.62.42.221.81.61.41.21144 13 th International Conference on Tribology – Serbiatrib’13

complete. The SEM analysis maycontributee tobetter understanding of this mechanism.The SEMmicrograph of the worn surfacee atload of 10 N and at speed of 0.25 m/s for a slidingdistance of 300 m is presented in Figure 12 forr thetested composite material.Figure 14. Appearance of the worn surface of ZA27alloy photographed by SEM M in dry slidingconditions(vv 1 = 0.25 m/s, , F 1 = 10 N)Figure 12.SEM micrographs of wornsurfaces of theZA27/5%SiC/3%Gr compositeQualitative and quantitative chemical analysesof micro-constituents were performed using energydispersive spectrometry(EDS), Figure 13.Analysess confirm the presencee of constituentelements like: Zn, Al, SiC, Gr (C), as well as thepresence ofFe as a consequence of material transferfrom the counterpart to the compositeblock.Figure 15. Appearance of the worn surface ofZА27/5%SiC/3%%Gr compositephotographed by SEM indry sliding conditions (v 1 = 0.25 m/s, F1 = 10 N)4.CONCLUSIONFigure 13. EDS analysis of worn surface on theZA27/10%SIC/1%Gr compositeMicrostructure of worn surface of ZA27 alloy isshown in Figure 14.Microstructureofworn surface ofZА27/5%SiC/3%Grcomposite is given inFigure 15.Generally, the parallel grooves and scratches canbe seen over all the surfaces in the t directionn ofsliding. These grooves and scratches resulted fromthe contact between theworn surface of the testedmaterial andthe counterdisc of significantly higherhardness.This research was conducted inorder tocomplete the tribological knowledge on developeddcomposite materials with ZA27 alloy reinforced bythe SiC and graphite g particles. The goal was toconfirm the further possibilities for f broaderrapplication of given g composites as advanced tribo-materials, in different d technical systems becauseethey have excellent wear resistance when comparedwith the base ZA27 alloy.By monitoring the wear process throughhobservation of wear volume in dry slidingconditions, the following conclusions can be made: Wear of o the testedd compositeis smallerthan wear w of ZA27 alloy forr all appliedsliding speeds and normal loads. Wear process evolution hass the samecharacter for both tested materials (basicZA27 alloy anddZA27/10%SiC/1%Grcomposite).13 th International Conference on Tribology – Serbiatrib’13145

Values of the wear volume of the observedcomposite material increase with theincrease of normal loads.Wear volume also increases with theincrease of the sliding speed.ACKNOWLEDGMENTSThis paper presents the research results obtainedwithin the framework of the project TR-35021,financially supported by the Ministry of Education,Science and Technological Development of theRepublic of SerbiaREFERENCES[1] M. Babić, S. Mitrović: Tribological characteristicsof composites based on ZA alloy, (in Serbian),Monograph, Faculty of Mechanical Engineering,Kragujevac, 2007.[2] M. Babić, S. Mitrović, I. Bobić, TribologicalProperties of Composites with Substrate Made of theZA-27 Alloy Reinforced by the Graphite Particles,Tribology in industry, Vol. 29 No. 3&4, 2007.[3] S. Mitrović, M. Babić, B. Stojanović, N.Miloradović, M. Pantić, D. Džunić: TribologicalPotential of Hybrid Composites Based on Zinc andAluminium Alloys Reinforced with SiC and GraphiteParticles, Tribology in Industry, Vol. 34, No. 4,177-185, 2012.[4] M. Babić, A. Vencl, A., S. Mitrović, I. Bobić,Influence of T4 heat treatment on tribologicalbehavior of ZA27 alloy under lubricated slidingcondition, Tribology Letters, Vol. 36, No. 2, pp.125-134, 2009.[5] M. Babic, S. Mitrovic, B. Jeremic: The Influence ofHeat Treatment on the Sliding Wear Behaviour of aZA-27 Alloy, Tribology International, Vol. 43, No.1-2, pp. 16-21, 2010.[6] M. Babic, S. Mitrovic, D. Džunic, B. Jeremic, I.Bobic: Tribological Behaviour of Composites Basedon Za-27 Alloy Reinforced With Graphite Particles,Tribology Letters, Vol. 37, No. 2, pp. 401-410, 2010.[7] K.H.W. Seah, S.C. Sharma, B.M. Girish:Mechanical properties of as-cast and heat-treatedZA-27/graphite particulate composites, Composites,Part A 28A, pp. 251-256, 1997.[8] S.C. Sharma, B.M. Girish, R. Kramath, B.M. Satish:Graphite particles reinforced ZA-27 alloy compositematerials for journal bearing applications, Wear,Vol. 219, pp. 162-168, 1998.[9] B.K. Prasad: Abrasive wear characteristics of azinc-based alloy and zinc-alloy/SiC composite,Wear, Vol. 252, pp. 250-263, 2002.[10] S.C. Sharma, B.M. Girish, R. Kramath, B.M. Satish:Effect of SiC particle reinforcement on theunlubricated sliding wear behavior of ZA-27 alloycomposites, Wear, Vol. 213, pp. 33-40, 1997.[11] B.K. Prasad: Investigation into sliding wearperformance of zinc based alloy reinforced with SiCparticles in dry and lubricated conditions, Wear,Vol. 262, pp. 262-273, 2007.[12] S. Mitrović, M. Babić, I. Bobić: ZA-27 alloycomposites reinforced with Al 2 O 3 particles, Tribologyin Industry, Vol. 29, No. 3-4, pp. 35-41, 2007.[13] M. Babić, S. Mitrović, I. Bobić, F. Živić: Wearbehavior of composites based on ZA-27 alloyreinforced by Al2O3 particles under dry slidingcondition, Tribology Letters, Vol. 38, No. 3, pp.337-346, 2010.[14] A Vencl: MMCs based on hypoeutectic Al-Si alloy:Tribological properties in dry sliding conditions,Tribological Journal BULTRIB, Vol. 2, No. 2, pp.17-22, 2012.[15] N. Miloradović, B. Stojanović: Tribologicalbehavior of ZA-27/10SiC/1Gr Hybrid Composite,Journal of the Balkan Tribological Association, Vol.19, No. 1, pp. 97-105, 2013.[16] B. Stojanović, M. Babić, S. Mitrović, A. Vencl, N.Miloradović, M. Pantić: Tribological characteristics ofaluminum hybrid composites reinforced with siliconcarbide and graphite, Journal of the Balkan TribologicalAssociation, Vol.19, No.1, pp. 83-96, 2013.146 13 th International Conference on Tribology – Serbiatrib’13

Serbian TribologySocietySERBIATRIB‘1313 thtInternational Conference onTribologyKragujevac, Serbia, 15 – 17May 2013Faculty of Engineeringgin KragujevacINFLUENCE OF OXIDATION LAYER GENERRATED ONPREHEATEDCONTACT PAIRS ONSTATIC COEFFICIENT OFFRICTIONMarija Jeremić 1 , Dragan Adamović 1 , Slobodan Mitrović 1Bojan Bogdanović 1 AleksandarSimić 1 , Sašaa Ranđelović 1 , Petar Todorović 1 ,1 Faculty of Engineering, SestreJanjić 6, Kragujevac, Serbia, mjeremic88@yahoo.com, adam@kg.ac.rs,boban@kg.ac.rs, bogdanovicboki@@gmail.com, aleksandarsimickg88@gmail.com, sasarandjelovic@yahoo.com,petar@kg.ac.rsAbstract: The subject of the work includes theoretical considerations, conducting c experimentall tests whichhare subjected to the analysis of the test t results related to the determination of the coefficient of static friction,of contact pairs which were previously heat treated. Contact pairs were, before thee procedure to t determineethe coefficient of friction, heated to t temperatures of 50⁰C-350⁰C andd cooled to room temperature. Testresults showthat, depending on thethermal treatment of contact pairs, there is significant increase in thecoefficient of friction. The authors believe thatt the reasonss for the increase of thee friction coefficient arerelated to the creation of oxide and changes in the surface layer of contact pairs.Keywords: coefficient ofstatic friction, oxidationn layer, pre heat treatment, normal load1. INTRODUCTIONPreciselyprediction of static friction is of f theessence significance due to its application in manymodern tribological systems, such as chucks,clamps, seals, micro-electromechanical systems andso on. Importance of the evaluation of thecoefficient of friction isfigured long time ago andthat area studied some of the first scientists suchh asLeonardo da Vinci, Guillaume Amontons, CharlesAugustin Coulomb, George Rennie and manyothers [1]. As the force of friction occurs when twobodies are in contacts, based on the relative speedof movement it can be divided intofriction at t theidle status and the friction in state of motion. Startthe movement of any kind is related to the existenceof static friction. The static coefficient of frictiondepends onmany parameters, primarily from thesurface of contact, normal load, atmosphereandtemperaturee at whichh contact occurs, surfaceabsorption, quality ofsurfaces in i contact andmaterials of contact surfaces [2-6]. Many authorshave been examining the influence of roughnessparameters on the of contact surfaces for the staticfriction coefficient and came to the conclusion thattthe static coefficient of friction increases withincreasing surface roughness parameters [4, 5].Some authors have concluded that some s of theroughness parameters, such as skewness andkurtosis, havee more influence on the staticcoefficient of friction compared to other parameters[7] [8]. The friction force at idle phase increasesswith increasingg tangential displacements up to thevalue that is needed n to begin movement of thebodies in contact [9]. The main parameters of staticfriction are thee maximum force of static friction,which is realized at the moment of macrodisplacement,and corresponding value of themicro-movement. The influence of temperature andnormal load on the friction characteristics ofmaterials is topic of numerous scientificc works [10-13] ], especially when are concerned materials usedin process of hot h metal forming. Etsion and Amit[10] experimentally investigated theeffect ofnormal load on the coefficient of static friction f for avery smooth metal m surface. Normal load is in therange of 10 -3 N - 0,3N andd applied tosamples ofsmall diameter made of three different aluminumm13 th International Conference on Tribology – Serbiatrib’13147

alloys in contact with nickel-platedfinish. The testswere conductedin conditions of controlledhumidity and air cleanliness. The dramatic increaseof coefficient of static friction was noticed whenthe normal load reduced to the lowest level. Thiskind of behavior is attributed to the adhesion forceswhich havea significant role and which are moreprominent at low normal loads and smoothsurfaces. Tothe knowledge of authors the studyy ofstatic coefficient of friction under conditionss ofhigh temperature wasnot the topic of moreextensive theoretical and experimental works [14-17]. Behavior of the static and kinematic frictionn ofmaterials coupling at differentvalues oftemperaturee and contact pressures investigatedChaikittiratana, Koetniyom and Lakkam [ 17].Device which is used for performing of theexperiment is specifically constructed for thedetermination of slidingfriction at highertemperatures. It was established that at 100°C thereisn't significant change in the coefficient of frictionwhen the contact pressure varies. However, at atemperaturee of 200°C observed friction coefficientchange with the change of the contact pressure.There was an increase in the coefficient of frictionwhen the contact pressure increasedfrom0.147MPa - 0.252MPa. It is concluded that theincrease in temperaturee significantly increases thecoefficient of friction. One of the most importantconclusionss of this workis that the static s coefficientof friction increases with increasing temperature,which is partly a consequence of increase theplasticity of most contact materials at highertemperatures.The aimof this paper is to determine theinfluence ofpreviously heat treated contact pairss onthe value ofthe static coefficient of f sliding friction,with contact elements made of steel. Experimentalmeasurements were performed on instrumentationdeveloped by the authors, in order to accuratelydetermine the value of the staticcoefficientt ofsliding friction at highertemperatures and relativelylow values of contactt pressures in the variableradius of curvature of contact elements. The resultscould havea significant aplication in industrialapplications.accurately measure very small values of physicalquantities(normaland friction force). f Formeasuring small values of the friction force isnecessary to exist e electronic components in themeasuring chain (sensorss of force and otherelectronic components). Since the forcesensors, forreasons of measurementmreliability, should beoutside the zone of high temperatures,, it becomesimportant ,to the signals associated with smallldisplacement and a force, that they have to betransmitted mechanically from the high temperatureezone (the chamber in which the pair is heated todesired temperature) to thee sensor to quantify thevalue of the friction and normal contact load ofcontact pair. Mechanicaltransmission chain offriction force pulls the corresponding measurementerrors, especially when it comes to measure verysmall values of f force.The authors started from the idea that theprinciple of measuring m the static coefficient offriction over the steep plane have to be upgradeddandenable measurementin conditions of highhtemperatures and low values of contact pressure.Theprinciple of measurement of coefficient offriction over the steep plane (Figure 1) ) is based ongravity. Static coefficient c of friction, ass it is knownnis the ratio of the friction force and the force f normalto the surfacee of contact, where condition forequilibrium on a steep plane is given byexpression ∙sin. In the limiting caseof slidingfriction valid equation is:Fmggsin tg (1)N mggcoswhere are: value of st tatic frictioncoefficient;F - friction force; m - mass of body; g -acceleration off gravity; - steep angle of theplane.2. THEORETICAL CONSIDERATIONSMeasurement of thefriction coefficient at hightemperaturee in contact pairs is associated with anumber ofproblems of physicall and technicalnature. Issuesrelated to the reliability ofmeasurements are especially pronounced in themeasurement of the friction coefficient under lowcontact pressure. Problems related tothe fact that athigh temperatures in the contact pairs must148Figure 1. Thee balance of the body on a steep s planeMeasurement error e of static coefficient of frictionon this principle arising from error of anglemeasurement in i relation to an ideal horizontalposition when the t body which is located on a steepplane crossed from sleep p in a stateof motion(Figure 2), respectively:13 th International Conference C onn Tribology – Serbiatrib’13

tg( ) tg 100, [%] (2)tgwhere are: relative error of measurement; -angular error of measurement . If I one takes intoaccount that the frictioncoefficient tg , thenthe diagramgiven in Figure 2 candetermine therelative measurement error ()which is, amongother, function of coefficient of friction.Friction coefficient µ [-]Relative percent difference ε [%]21.81.61.41.210.80.60.40.2004 8 12 16 20 24 28 32 36 40Rotation of steep plane [ ]°Figure 2. Graphical representationof the relativeemeasurement error of the coefficient off friction over r thesteep planeOn the basis of eqquation (2) and the diagram(Figure 2) can be noted that values of thecoefficient of sliding friction μ> 0.1 with Δα = 1'correspond to the measurement error which is lessof 0.3%. This means that theprinciple ofmeasurement can be efficiently used to determinethe static friction coefficient at higher temperatures,especially if one takes into accountt that the slidingcoefficient of staticfriction at elevatedtemperatures for most materials is greater than n theabove values.44tops of some asperities a and in that case betweenmost asperitiess there is noo real physical contact.Block and the rollers are made of same material(theaforementioned). Experiments weree carried outwith increase of temperatures by 50⁰ C, startingfromthe temperature T 1 =50 ⁰ C up to temperatureeT 2 = 350 ⁰ C, andd after heating samples were left tocool to room temperature.Each measurement wasrepeated 10 times, so that a total of 350independent experimentsis performed at roomtemperature after heating the contact pair.Figure 3. Schematic representation of the contact pairThe coefficient of friction was determined usingspecially designed devicess (Figure 4). Workingprinciple of designed d Tribometer iss based onrotation of steep plane, which is done on amechanical principle to the accuracy of the readingsof 1’. Namely, rotation steep plane is carried out bymanually turning of nonius. . Leveling off Tribometerris done using level whichh provides precision ofreading which is i less than 0,02/1000 ⁄ .3. EXPERIMENTAL RESEARCHExperimentalstudieswere performed onsamples made of steel EN X160 CrVMo12 1 whichis subjectedto heat treatment hardening, with theaim to obtain the highh hardness of 64 HRC andwear resistance. Tribological contact pair is realizedby setting the rollers of diameters 4 mm, 6 mm, 8mm, 10 mmm and 12 mm, length 20mm, on prismblock (Figure 3). This simulates tribological contactwhich is, theoreticaly,realized on line. If we takeinto account that the mass of the rollers is smallthen we have small values of contact pressure.When rollers are in contact on flat surface s block, inthe absencee of temperature, contactt is achievedd on13 th International Conference on Tribology – Serbiatrib’13Figure 4. Measuring equipmentContact pairs were heated to temperatures from50 ⁰ C to 350 ⁰ C with the step of 50 ⁰ C, after whichhthey are cooled to roomtemperature.So,determining the size of the static coefficient offriction was performed withh contact pairs that wereeheat treated at temperatures indicated. This meansthat contact pairs in the process of determining thestatic coefficient of friction weree at roomtemperature. The T results of measurement of thecoefficient of friction f depending on the weight of149

ollers, that is, the normal contact loadtemperaturee are shown in Table 1:150dTF [N][mm][ ⁰ C]121086412108641210864121086412108641210864121086412108640.174101 0.120903 0.077378 0.043525 0.019345 0.174101 0.120903 0.077378 0.043525 0.019345 0.174101 0.120903 0.077378 0.043525 0.019345 0.174101 0.120903 0.077378 0.043525 0.019345 0.174101 0.120903 0.077378 0.043525 0.019345 0.174101 0.120903 0.077378 0.043525 0.019345 0.174101 0.120903 0.077378 0.043525 0.019345 0.174101 0.120903 0.077378 0.043525 0.019345 202020202050505050501001001001001001501501501501502002002002002002502502502502503003003003003003503503503503500.1965810.1852610.2956190.2438110.2296940.2102290.2124840.2181930.2855140.2273320.1908950.2045410.2056290.2436610.2400490.3104430.2844180.2890480.3309710.3515920.2400520.2286210.234370.3408610.4591680.2702270.2867180.3778370.3225730.5229930.2917250.3666280.5479850.5717560.3891940.2796390.3417810.3540480.3152330.462363andAccording to the data from Table 1 was formeddiagram of static friction coefficientwhichdepending on the normal load and temperatures towhich contact pairs are pre-treated.Figure 5. The static coefficient of friction whichdepending on o the normal l load and temperature4. DISCUSSIOONBased on theoreticalconsiderationsandexperimental investigationss it can be concluded thattthe physical principle p off measuringthe staticcoefficient of friction f overr the steep plane can beeffectively applied in testing conditions at highhvalues of temperature andd low valuess of contacttpressures.It requires a certainprecisionmeasurement instrumentation (precisionmeasuringequipment - Tribometer),Twhich is realized withequipment used for conducting this tests. Theaccuracy of the measurement was the angle of Δα =1'. A large number of measurements and a numberof repetitions of experiments under identicalconditions alloww statistical analysis of the t results ofmeasurements to t minimize random error.Experimental results of f investigation presentedin the form of 3D 3 diagram in Figure 5 indicates thattfor given conditions the temperature treatment t ofcontact pairs have substantial effect on the value ofstaticcoefficient of friction.Theessentialdependenceof frictioncoefficientfromtemperature, which can be seen from the t diagram,expressed that for temperature change from 50⁰ Cto 300⁰ C value of the static frictioncoefficientincreases nearlyy two times. Based on the foregoing,it is obvious that the changes of static frictioncoefficient arise from temperatureeffect onmaterial in contact. It is likely that preheating ofcontact pairs above the temperature values close to200⁰C leads too physical changes in the surfacelayers of contact pairs.With diagram shown inn Figure 5 iss possible toanalyze the effect of normal load on the staticcoefficient of friction. f Normal load in the range of0.019 to 0, 17 N is used during experiment. Fromthe diagram it can c be seen that the minimum valuesof the normal load l in the entire temperature range,13 th International Conference C onn Tribology – Serbiatrib’13

corresponding to the maximum value of thecoefficient of friction, and reverse. This proves thatthe static coefficient of friction under conditions oflow values of contact pressures andpreheat treatedcontact pairs, just depends on the level of contactpressure, which is in line with tests performedd indifferent conditions [ 10]. The increase in thecontact pressure can lead to lower frictioncoefficient and reverse.When analyzing theresults, theauthors of thiswork started from the next. Because of preheatingof contact pairs it comes to decrease of hardness,which can be the reason for the increase of thefriction coefficient. It ispossible, and more likely,that oxidation layer iscreated at the surfacee ofcontactpairs withdifferenttribologicalcharacteristics comparing to base material, steel ENX160 CrVMo12 1. Theoretically,changes in thestructure changes the mechanicall propertiess ofhardened steel. With temperature increase hardnessand strengthdecrease, while ductility and toughnessincrease. During the heating of the samples wasthree modes layoffs and low firingup to 200⁰⁰ C,the transformation of retained austenite 200⁰ C -300⁰ C andremoving internal stresses of 300⁰ C -400⁰ C. This, withincreasingtemperingtemperatures the hardness and strength decrease,while ductility and toughness increase. Oxidationoccurs as a chemical reaction between metal andoxygen from the atmosphere. Theoccurrencee ofoxidation ispossible at room temperature, however,the metals that are exposed to elevated temperaturesabove 200⁰⁰ C are able to createe more intenseoxidation layer.The authors were not able to carry out structureanalysisof surface layer withelectronicmicroscope, so the question of largeincrease inn thefriction coefficient hastwo options listed. It isunlikely that changes inhardness off steel EN X160CrVMo12 due to release 1 (Figure 6) from thevalue of 64 HRC to value of 60 HRC can lead too anincreasein the coefficientoffriction ofapproximately 70%. For this reason the authorsbelieve thatt significant increase in the coefficient offriction is very probably due to the creationn ofoxides at the surface of contact pairs.Figure 6. 6 Releasing diagram of material5. CONCLUSIIONSBased on these tests and analyzingthe results,conclusions were made, preheated contact pairsmade of steel EN E X160 CrVMo12 1, at relativelylowtemperaturesresultingleads tosignificant change in the value of staticcoefficientof friction. Results R of measuring the staticcoefficient of sliding s friction of the test material,under conditions of high temperaturee and smalllvalues of load contact, c indicating a verysignificanteffect of temperature and contact pressure on value.Theimpact of the minimumm values of normal loadin the entire temperature range, corresponding tothe maximum value of thee coefficientt of friction,andvice versa. It is very unlikely that the changesof hardness of steel EN X160 CrVMo12 1 due tothe release fromm the value 64 HRC to the t value 60HRC may leadd to an increase in the coefficient offriction of approximately 70%. From this reason theauthors believe that a significantincrease ofcoefficient of friction is probably a consequence ofcreation of oxides at the surface of contact pairs.REFERENCE[1] Dowson, D. , History of Tribology, Longman, NewYork, 1979[2] K.-H. ZumGahr, K. Voelker, Friction and wear ofSiCfiberreinforced borosilicate glass mated to steel, ,Wear 225–229, pp. 888–895, 1999.[3] P. Blau, The significance and use off the frictioncoefficient, Tribol. T Int. 34, , pp. 585–591 2001.[4] B. Ivkovic, M. M Djurdjanovic, D. Stamenkovic: TheInfluence of f the Contact Surface Roughness on theStatic Friction Coefficient, Tribologyin Industry,22, pp. 41-44, 2000.13 th International Conference on Tribology – Serbiatrib’13151

[5] U. Muller, R. Hauert: Investigations of thecoefficient of static friction diamond-like carbonfilms, Surface and Coatings Technology, 174 –175,pp. 421–426, 2003.[6] B. Polyakov, S.Vlassov, L.M. Dorogin, P. Kulis, I.Kink, R. Lohmus: The effect of substrate roughnesson the static friction of CuO nanowires, SurfaceScience 606, pp. 1393–1399, 2012.[7] N. Tayebi, A.A. Polycarpou: Modeling the effect ofskewness and kurtosis on the static frictioncoefficient of rough surfaces, TribologyInternational 37, pp. 491–505, 2004.[8] B. Bhushan, S. Sundararajan, W.W. Scott, S.Chilamakuri: Stiction analysis of magnetic tapes,IEEE Magnetics Transactions 33, pp. 3211–3213,1997.[9] E.L. Deladi: Static Friction in Rubber-MetalContacts with Application to Rubber Pad FormingProcesses, PhD thesis, University of Twente,Twente, 2006.[10] I. Etsion, M. Amit: The effect of small normal loadson the static friction coefficient for very smoothsurfaces, Journal of tribology, Vol.115, July 1993.[11] O. Barrau, C. Boher, R. Gras, F. Rezai-Aria: Wearmecahnisms and wear rate in a high temperaturedry friction of AISI H11 tool steel: Influence ofdebris circulation, Wear 263, pp.160–168, 2007.[12] H. Kumar, V. Ramakrishnan, S.K. Albert, C.Meikandamurthy, B.V.R. Tata, A.K. Bhaduri: Hightemperature wear and friction behaviour of 15Cr–15Ni–2Mo titanium-modified austenitic stainlesssteel in liquid sodium, Wear 270, pp. 1–4, 2010.[13] A. Chaikittiratana, S. Koetniyom, S. Lakkam:Static/kinetic friction behaviour of a clutch facingmaterial: effects of temperature and pressure,World Academy of Science, Engineering andTechnology 66, 2012.[14] H. Kumar, V. Ramakrishnan, S.K. Albert, C.Meikandamurthy, B.V.R. Tata, A.K. Bhaduri: Hightemperature wear and friction behaviour of 15Cr–15Ni–2Mo titanium-modified austenitic stainlesssteel in liquid sodium, Wear 270, pp. 1–4, 2010.[15] P. Mosaddegh, J. Ziegert, W. Iqbal, Y. Tohme:Apparatus for high temperature frictionmeasurement, Precision Engineering 35, pp. 473–483, 2011.[16] M. Worgull, J.F. Hetu, K.K. Kabanemi, M. Heckele,Hot embossing of microstructures: characterizationof friction during demolding, MicrosystTechnol 14,pp. 767–773, 2008.152 13 th International Conference on Tribology – Serbiatrib’13

Serbian TribologySocietySERBIATRIB ‘1313 th International Conference onTribologyKragujevac, Serbia, 15 – 17 May 2013Faculty of Engineeringin KragujevacDYNAMICS OF SAMS IN BOUNDARY LUBRICATIONJelena Manojlović 11 The Faculty of Mechanical Engineering, University of Niš, Niš, Srbija, jmanojlo@gmail.comAbstract: Surfactant molecules have some properties responsible for a number of remarkable phenomena,such as oriented adsorption of surfactants at surfaces and interfaces. The capability to self-assemble intowell-defined structures is often seen as being more important than their surface activity. When a surfactantsolution is in contact with a solid surface, the surfactant molecules adsorb onto the surface, ideally formingan adsorbed layer of high order, termed as self-assembled monolayer (SAM). Many surface properties areinfluenced by such a film, and therefore, SAMs offer the capability to form ordered organic surface coatings,suitable for various applications, such as wetting or corrosion protection. Due to the flexibility in choosingthe molecular architecture, organic molecules have many interesting applications, such as biosensors, inphotoelectronics, in controlling water adsorption or boundary lubricant coating. This paper focuses oncationic surfactants (quaternary ammonium surfactants), with some unique properties that are not present inother surfactants.Keywords: surfactants, self-assemble, boundary lubrication, nanotechnology.1. INTRODUCTIONA contact between two surfaces is of greatimportance in technology. At the interface of twomaterials, when they are brought together,separated or moved with respect to one another,contact formation, friction, wear or lubrication arethe processes that occur [1]. Friction has long beenthe subject of research. All machined metalsurfaces, as viewed through a microscope, havetheir own roughness, called asperity. Therefore, twosurfaces touch at an extremely small number ofpoints, and their true area of contact is a part oftheir apparent contacting area. In contact of twosurfaces, the number of asperities increases due toplastic deformation of some of them. Theconsequence of this is the appearance of theremoval of material from a surface in bearing underdynamic conditions, defined as wear [2]. In order toreduce wear, lubricants are employed between thesurfaces. Friction, wear and lubrication are thecenter of consideration in many tribological andtechnological problems. Having in mind that acontact occurs in numerous asperities, the researchof two contact surfaces, especially at the molecularlevel and the friction phenomenon at the nanometerscale, is studied by nanotribology, a branch oftribology.In order to categorize the friction propertiesbetween two surfaces, the "Stribeck curve" wasdeveloped. Machine elements may experienceseveral lubrication regimes, including full-film,mixed, and boundary lubrication. These regimesdepend on the properties of a lubricant andoperating conditions. In the case where speeds aretoo low and loads are too high to permitestablishing a hydrodynamic film, or when thedistance between contact surfaces is a fewnanometers or a few molecular layers, we candefine boundary lubrication (fig.1).Figure 1. Boundary lubricationThe boundary films have been the subject ofstudy for decades, since friction and wearphenomena are affected by these ultrathin films.Under friction, the dynamics of lubricants onsurfaces is very important, especially the molecularbehavior of lubricants in boundary lubrication. Thebehavior and dynamics of the boundary films,13 th International Conference on Tribology – Serbiatrib’13 153

formed during sliding, becomes more complex dueto change of some experimental parameters, such astemperature [3]. The computer simulation ofprocesses during sliding contact, when severalhundreds of atoms are involved, indicates thatatomic processes cannot be neglected, when wedescribe nanotribology experiments [1, 4]. For thatpurpose, several available methods can be includedfor research at a molecular level [5].2. BOUNDARY LUBRICATION BY SAMSAttractive model systems for boundarylubrication are organic self-assembled monolayers(SAMs). Preparing self-assembled monolayers isone of the most elegant ways to make ultrathinorganic films of controlled thickness. The processof self-assembly is considered as a very importantexample of equilibrium structural organization onthe molecular scale. Organic thin films are anemerging area of materials chemistry and areutilized in many application areas, such aselectronic components, as well as in biomedicalapplication [6]. There is also special interest in thepossibility of manufacturing molecular layers withparticular properties. Molecular self-assembly isrecognized as a powerful strategy for thefabrication of nanoscale structures [7].The interest in these systems has been furtherintensified in order to understand and solve friction,lubrication and related problems [8]. More recently,lubrication in a small-size system, such as themicroelectromechanical system (MEMS) ornanoelectromechanical system (NEMS), is a bigchallenge in scientific work, especially in the studyof new kind of lubricants. Different type ofmonolayers attached to sliding surfaces appears as agood candidate in MEMS lubrication. Therefore,the understanding of behavior between monolayersfilms is of great importance in tribological andnanotribological experiments.Due to very small thickness of monolayers(range of few nanometers), new tools are requiredfor this nanotribological studies. Widely used arethe following: surface-force apparatus (SFA), thescanning tunneling miscorscope (STM), the atomicforce and friction-force microscopes (AFM andFFM). Developed more than 40 years ago, the SFAis usually applied to study properties of molecularlythin films, confined between two molecularlysmooth macroscopic surfaces, with surfaceseparations at the angstrom level and forcesbetween them. A scanning tunneling microscope(STM) is an instrument for imaging surfaces at theatomic level [1]. With the development of a numberof powerful techniques in surface analysis, asmentioned above, academic interest in SAMs hasregained, because of the possibilities to investigatethe growth and the structure of such layers on thenanometer scale [9, 10].3. SURFACTANTS SELF-ASSEMBLYThe word “surfactant”, does not always appearin dictionaries, because it is a contracted form ofthe phrase SURFace ACTive AgeNT. Surfactantsare molecules essential to the chemical industry andin many products such as soaps, detergents,shampoos, softeners, pharmaceutical products, etc.Surfactant molecules have amhiphilic propertiesbecause they consist of two distinct parts - one thathas an affinity for the solvent, and the another onethat does not. This dual structure is responsible fora number of remarkable phenomena, such asmicelle formation in solution at a certainconcentration, the so-called critical micelleconcentration (cmc), and oriented adsorption ofsurfactants at surfaces and interfaces. Micelleformation has attracted a notable part of thesurfactant research, in order to investigate theformation of micelles [11, 12], their shape [13] ortheir interactions [14]. Systems below the cmc havenot been widely studied [15].The self-assembled monolayers can be preparedusing different types of molecules and differentsubstrates. A very often studied SAM model systemcomprises thiol molecules, adsorbed onto gold,silanes on an oxide surfaces, or alkanephosphatemonolayers, which was in detail reviewed byUlman [16]. The choice of the substrates, used inthe self-assembling process, is dictated by themolecules and their interactions, as well as the finalapplication.Self-assembled monolayers form spontaneously,when certain classes of molecules adsorb onto asolid surface from solution. When a surfactantsolution is in contact with a solid surface, thesurfactant molecules adsorb onto the surface,ideally forming an adsorbed layer of high order,termed as self-assembled monolayer (fig.2).Figure 2. a) An organized monolayer on a substrate,b) CTAB chainMany surface properties are influenced by sucha film, e.g. the hydrophobicity or the wetting or154 13 th International Conference on Tribology – Serbiatrib’13

electrostatics [17]. Due to the flexibility inchoosing the molecular architecture, organicmolecules have many interesting applications, suchas biosensors, for lubrication or in controlling wateradsorption. Therefore, in recent years, muchattention has been directed to the study of SAMs.However, a discrepancy still exists between thetheoretical understanding and the practicalimportance involved in the formation of suchlayers.Cationic surfactants are a small subgroup withsome unique properties that are not present in othersurfactants. In order to help understanding ofadsorption of cationic surfactants, adsorption ofquaternary ammonium surfactants onto inorganicsubstrates, such as mica, has been widely studied[18, 19, 20]. The process of adsorption has beeninvestigated by different techniques [21], such as x-ray photoelectron spectroscopy (XPS), the surfaceforces apparatus, (SFA) [15] or contact angle (CA)measurements [22]. This paper focuses onquaternary ammonium surfactants with a cationichead group, single-tailedhexadecyltrimethylammonium bromide, CTAB,with the molecular structureCH 3 (CH 2 ) 15 N + (CH 3 ) 3 Br - (figure 2b).4. SAMS PREPARATION ANDCHARACTERIZATIONA standard protocol, which can produce a welldefinedand reproducible hydrophobic CTAB filmon mica, does not exist. Namely, previous studiesof CTAB adsorption on various substrates,suggested that the behavior of CTAB is morecomplex than the behavior of other cationicsurfactants [22, 23, 24], but the reason for thissingularity has not been clearly determined [17].The original goal of these experiments was toproduce self-assembled monolayers and use themas model systems to study boundary lubrication.But, there was a problem concerning the resultsbeing repeated, as well as the characterization ofthe adsorbed CTAB layers on muscovite mica indetail. The various SAM morphologies, found onmica by the use of different adsorption protocols,demonstrate the influence of a large number ofexperimental parameters on the adsorption process,such as concentration, pH, temperature andhumidity. They are rarely described in theliterature.Figure 3. CTAB layer on muscovite micaIn order to characterize and determine theproperties of SAMs, two techniques have beenextensively used in this work, contact anglemeasurements and the atomic force microscopy(AFM). For example, freshly cleaved mica hascontact angle less than 10° and contact angle on theSAM produced by CTAB adsorption on mica canbe 140° [22]. A contact angle greater than 90° isdetermined on hydrophobic surfaces.In our experiments the contact anglemeasurements have been used in order to define thedegree of hydrophobicity of the modified micasurface and molecular order after surfactantadsorption, performed by using ultra pure water.All contact angle measurements were averagedover several samples.5. EXPERIMENTAL PROCEDUREWe made self-assembled monolayers ofquaternary ammonium surfactants on mica. Singletailedhexadecyltrimethylammonium bromide,CTAB, with the molecular structureCH 3 (CH 2 ) 15 N + (CH 3 ) 3 Br - , was purchased fromFluka. For further purification, CTAB wasrecrystallized from an ethanol/acetone mixture. Asa solvent, ultra pure water of resistivity 18.3MΩcmwas prepared using a Barnstead EASYpurebatch-fed water purification system. The samewater quality was also used for the sample rinsing,before drying with a clean nitrogen stream.The glassware and bottles used in theexperiments were consistently cleaned by piranhasolution and then rinsed with purified water toavoid any organic contamination. All the employedtools were previously cleaned in order to minimizethe occurrence of molecular contamination,particularly on the high-energy mica surface.Muscovite mica purchased from Spruce PineMica Company Inc. (USA) was used for theadsorption experiments. Small mica samples of 1-1.5cm 2 size, were cut by scissors. Then, they werefreshly cleaved on both sides before immersion intothe surfactant solution. The adsorption was13 th International Conference on Tribology – Serbiatrib’13 155

performed from the surfactant solution in a volumeof 20ml.In our first adsorption series withouttemperature control, a 1000ml stock solution of10 -2 M (~10cmc) CTAB was prepared at roomtemperature. Since the solubility of CTAB in waterwas low at room temperature, the solution washeated to 30-35°C. By dilution of this solution,prepared by adding the appropriate volume of ultrapure water, surfactant solution concentrationsranging from 10 -3 M (~cmc) to 10 -6 M (~cmc/1000)have been prepared. One option, called “CTABin/CTAB out“(figure 4), involves immersion andextraction from the surfactant solution at thenominal concentration. In the water dipping step,the mica samples were dipped for 30sec into 20mlof ultra pure water to remove the excess solutionand excess surfactant molecules.After the post-rinsing step, the modified micasurface was gently blown dry with nitrogen, beforethe AFM imaging or contact angle measurements.This type of protocol (at different concentrations)was repeated several times to also assess thereproducibility.During the first experiments in March, the roomtemperature was in the range 21±2°C, and a fewmonths later, in June, the conditions in thelaboratory were clearly different, 32±2°C. Asdocumented by the air temperature measuredoutside of our building, shown in figure 5, thetemperature during the summer 2003 has beenmuch higher than in spring. In several experimentsthe temperature was more than 30°C.Using the above described preparation protocolsa significant number of samples have beenprepared. The AFM images of two representativesamples obtained by the ″CTAB in/CTAB out″protocol at a concentration below the cmc areshown in figure 6. The AFM results observed at allsolution concentrations below the cmc (from 10 -4 Mto 10 -6 M), are very similar with the resultspresented in figure 6.a) 85°/30°Figure 4. ″CTAB in CTAB out″ experimentb) 82°/22°6. RESULTS OBTAINED WITHOUTTEMPERATURE CONTROLThe first two sets of experiments were realizedwithout temperature control, in March and June2003.Figure 5. Local temperature recorded in June 2003(measured at 12:40 PM in Zürich-SwissMeteo data)Figure 6. AFM images of CTAB on mica obtained withthe “CTAB in/CTAB out” protocol at a concentration of10 -4 M. Advancing and receding water contact angles arealso shown:a) in March 2003 and b) in June 2003A clear seasonal influence on SAMs adsorptionhas been observed in all experiments realizedwithout temperature control, regardless of theexperimental protocols. On the sample prepared inJune, where the air temperature outside of thelaboratory building was higher than in March(figure 5), a significant number of clusters of heightbetween 0.5nm to 3.8nm and the size around156 13 th International Conference on Tribology – Serbiatrib’13

250nm size have been observed, according to thegrey scale of the image, which represents a heightrange of 5nm (cf. Figure 6.b). The sample preparedin March, shown in figure 6.a, is more promising.The advancing and receding water contact anglewere measured on both samples exhibit hysteresis.Without temperature control, the reproducibility ofthose surfactant films was difficult to accomplish.Therefore, it was difficult to identify the mostpromising protocol for us.We have performed a significant number ofexperiments at 10 -2 M CTAB (i.e.10xcmc) solutionsby the protocol “CTAB in/CTAB out”, whichnicely document the crucial role of the temperature.The results are summarized in figure 7.Figure 7.a is representative of the sampleprepared at 27°C and is qualitatively different fromthe samples prepared at the lower temperatures.The bright spots observed in figure 7.b representsmall islands on mica with a height between 0.5-1.3nm. In figure 7.c we detect clusters of a height inorder of 23nm.a) 27°Cb) 24.5°C7. DISCUSSIONOur adsorption results have shown that themorphology, the structure and the stability of theadsorbed films are sensitive to the experimentalconditions, primarily temperature.One very important concept of surfactantsolution is the Krafft temperature, whose effect isof great importance in SAMs formation. The Kraffttemperature is the minimum temperature at whichsurfactants form micelles. Below the Kraffttemperature micelles cannot form. The Kraffttemperature is a point of phase change below whichthe surfactant remains in crystalline form, even inaqueous solution (figure 8). Around the Kraffttemperature, Tk, many physical properties of thesurfactant solution reflect this transition. Thetransition in CTAB solution around Tk clearlyoccurs over a range of temperatures. Although theKrafft temperature is a well-established concept,reported values of Tk for CTAB in water varyconsiderably, from 20°C [17] to 25°C [14]. Kraffttemperatures, close to room temperature,significantly complicate the explanation ofexperimental results.According to the AFM images, the sampleobtained in March 2003 (figure 6.a) revealed ahomogeneous CTAB film, in contrast to the sampleperformed in June (figure 6.b). The reason for sucha difference must be related to changes in thesolution structure.c) 22°CFigure 7. Series of AFM images showing thesurface morphology of CTAB coated mica by theprotocol “CTAB in/CTAB out” at 10 -2 M solution attemperatures: a) 27°C, b) 24.5°C and c) 22°C.13 th International Conference on Tribology – Serbiatrib’13 157

day later, at a room temperature of 22°C, reveals asubstantially different SAM as shown in figure 7.c.Upon cooling of micellar CTAB solution from theinitial value ~30°C to 22°C, the solution thenconsists of crystals, monomers and micelles. As aconsequence of such structural changes in thesolution, the complex morphology (figure 7.c) isnot surprising. A distinction between monolayer orbilayer formation is not readily possible from AFMimages above (figure 7). The applied test, notdescribed here, on the samples in severalexperiments, suggested bilayer formation even atconcentration 10 -4 M.8. CONCLUSIONFigure 8. The structural changes in the CTAB solutionin the performed experiments, such as heating/coolingcycle of solution (marked as hysteresis) and the dilutionof the solution, both above the cmcNamely, solution used in the both experiments(10 -4 M) has been prepared by dilution of a 10 -2 Mstock solution. The room temperature in March(21±2°C) was slightly below the Krafft temperatureof CTAB and we could expect that this stocksolution mainly consisted of monomers. The sameargument applies, of course, also to the dilutedsolution (10 -4 M).In June, however, a significant daily temperaturevariation, with temperatures clearly above theKrafft temperature, has been recorded, particularlyin the days before the described adsorptionexperiment. It is to be expected that the solutionstructure of the stock solution is metastable andcomplex. The diluted solution may thus not be inthermodynamic equilibrium and consists ofmicelles and monomers. The clusters seen on thesample prepared under such conditions (figure 7.b)can be interpreted as micelle, adsorbed in differentshapes and sizes.The results shown in figures 6 and 7 clearlydemonstrate the temperature influence on thesurfactant films, morphology formed in differentseasons of 2003. The variety of adsorbed filmmorphologies in uncontrolled conditions above thecmc can also be explained by structural changes inthe stock solution. Namely, warming up the highlyconcentrated stock solution (10 -2 M) to some 30°C(above the Krafft temperature) will result in theformation of micelles. Since the film shown infigure 7.b. has been adsorbed at a temperature of24.5°C, which is near the Krafft temperature, wemight expect the presence of some aggregates inthe solution and also on the mica surface. Some ofthe micelles are expected to transform into surfacelayers.The repetition of the same experiment oneThe performed experiments describe theadsorption of quaternary ammonium surfactantsonto anionic, atomically smooth, muscovite mica.The surfactant films on mica, formed according todescribed experimental protocols, werecharacterized by contact angle measurements andby AFM. We have observed that SAMs can havecompletely different properties, depending on themeteorological conditions, influenced bytemperature. These results suggested thattemperature can influence all steps in adsorptionprocedure, from solution preparation to the rinsingstep. The fact that the Krafft temperature range ofCTAB (~25°) is around room temperature, makesthis system appear particularly complex.The dynamic CA experiments and AFMmeasurements have shown that the exact protocolof solution and self-assembled monolayer (SAM)preparation can substantially influence the stabilityof the hydrophobic layer, as well as thehydrophobicity. The results indicate that themorphology and the homogeneity of SAMs dependon many parameters, and the main reason for that isprobably the molecular structure of the solution,controlled by the temperature and concentration ofthe solution.In this model case of quaternary ammoniumsurfactants, the formation of homogeneous, wellorderedand reproducible monolayers is a verychallenging task. In order to assess such complexsystems, systematic variation of a great number ofparameters was a necessary procedure.ACKNOWLEDGEMENTSThe author would like to gratefully acknowledgethe support in obtaining and understanding thepresented results provided by her mentors, NicholasD. Spencer and Manfred Heuberger, members ofthe Laboratory for Surface Science and Technology(LSST), a part of the Department of Materials at the158 13 th International Conference on Tribology – Serbiatrib’13

ETH Zurich, Switzerland, where all theexperiments were performed.REFERENCES[1] B. Bhushan, J. N. Israelachvili, and U. Landman:Nanotribology:friction, wear and lubrication at theatomic scale, Nature 374,pp. 607-616, 1995.[2] NASA: Lubrication, friction and wear, 1971.[3] Choa Sung-H. et al.: A model of the dynamics ofboundary film formation, Wear 177, pp. 33-45,1994.[4] B. Bhushan: Nanotribology and Nanomechanica II,Springer, New Tork, 2011.[5] K. Miyake, et al.: Effects of surface chemicalproperties on the frictional properties of selfassembledmonolayers lubricated with oleic acid,Tribology Online 7, pp. 218-224, 2012.[6] J. e. a. Swollen: Molecular Monolayers and Films,Langmuir 3, pp. 932-950, 1987.[7] J. a. M. Mellott: Supercritical self-assembledmonolayer growth, Journal of the AmericanChemical Society 126, pp. 9369-9373, 2004.[8] M. K. Chaudhury: Adhesion and friction of selfassembledorganic monolayers, Current opinion incolloid & interface science 2, pp. 65-69, 1997.[9] R. o. M. S. Carpick: Fundamental investigations oftribology with atomic force microscopy, ChemicalReviews 97, pp. 1163-1194, 1997.[10] D. K. Schwartz: Mechanisms and kinetics of selfassembledmonolayer formation, Annual Review ofPhysical Chemistry, Vol.52, pp. 107-137, 2001.[11] J. Oremusova et al.: Thermodynamics ofmicellization of hexadecylpyridinium bromide inaqueous and alcoholic (C-1-C-4) solutions,Collection of Czechoslovak ChemicalCommunications 65, pp. 1419-1437, 2000.[12] K. K. Ghosh: Thermodynamics of micelle formationof some cationic surfactants as a function oftemperature and solvents, Indian Journal ofChemistry, Sec. B. 37B, pp. 875-880, 1998.[13] K.Boschkova et al.: Lubrication in aqueoussolutions using cationic surfactants - a study ofstatic and dynamic forces, Langmuir 18,pp. 1680-1687, 2002.[14] PS. Goyal, VK Aswal: Micellar structure and intermicelleinteractions in micellar solutions: Results ofsmall angle neutron scattering studies, CurrentScience 80, pp. 972-979, 2001.[15] Y. a. J. I. N. Chen: Effects of Ambient Conditions onAdsorbed Surfactant and Polymer Monolayers,Journal of Physical Chemistry 96, pp. 7752-7760,1992.[16] A. Ulman: An introduction to ultrathin organicfilms: from Langmuir-Blodgett to self assembly, ed.Academic Press, Boston, 1991.[17] I.Grosse, K.Estel: Thin surfactant layers at the solidinterface, Colloid and Polymer Science 278, pp.1000-1006, 2000.[18] F. Zhao et al.: Adsorption behavior ofhexadecyltrimethylammonium bromida (CTAB) tomica substrates as observed by atomic forcemicroscopy, Science in China Ser. B Chemistry 48,pp. 101-106, 2005.[19] S.Perkin et al.: Stability of self-assembledhydrophobic surfactant layers in water, Journal ofPhysical Chemistry B 109, pp. 3832-3837, 2005.[20] JM Mellott et al.: Kinetics ofoctadecyltrimethylammonium bromide selfassembledmonolayer growth at mica from anaqueous solution, Langmuir 20, pp. 2341-2348,2004.[21] B.Y. Li et al.: Time dependent anchoring ofadsorbed cationic surfactant molecules atmice/solution interface, Journal of Colloid andInterface Science 209, pp. 25-30, 1999.[22] Sharma,B.G. et al.: Characterization of adsorbedionic surfactants on a mica substrate, Langmuir 12,pp. 6506-6512, 1996.[23] W.A.Ducker and Erica J. Wanless: Adsorption ofhexadecyltrimethylammonium bromide to mica:Nanometer-scale study of binding-site competitioneffects, Langmuir 15, pp. 160-168, 1999.[24] S. Manne, H. Gaub: Molecular-Organization ofSurfactants at Solid-Liquid Interfaces, Science 270,pp. 1480-1482, 1995.13 th International Conference on Tribology – Serbiatrib’13 159

Serbian TribologySocietySERBIATRIB ‘1313 th International Conference onTribologyKragujevac, Serbia, 15 – 17 May 2013Faculty of Engineeringin KragujevacINFLUENCE OF RICE HUSK ASH – SILICON CARBIDEWEIGHT RATIOS ON THE MECHANICAL BEHAVIOUR OFAL-MG-SI ALLOY MATRIX HYBRID COMPOSITESK. K. Alaneme 1 , T. M. Adewale 1,21 Department of Metallurgical and Materials Engineering Federal University of Technology, Akure, Nigeria,kkalaneme@gmail.com2 Faculty of Engineering and Physical Sciences, School of Materials, University of Manchester, Manchester, UnitedKingdomAbstract: The influence of rice husk ash (RHA) and silicon carbide (SiC) weight ratio on the mechanicalbehaviour of Al-Mg-Si alloy matrix hybrid composites was investigated. RHA and SiC mixed in weight ratios0:1, 1:3, 1:1, 3:1, and 1:0 were utilized to prepare 5, 7.5 and 10 wt% of the reinforcing phase with Al-Mg-Sialloy as matrix using two-step stir casting method. Density measurement, estimated percent porosity, tensileproperties, fracture toughness, and SEM examination were used to characterize the composites produced. Theresults show that the composites were of good casting quality as the estimated porosity values were less than2.5 % in all grades produced. For the three weight percent worked on, the tensile-, yield-, and specific strengthdecreases with increase in the weight proportion of RHA in the RHA-SiC reinforcement. However, the resultsshow that the composites with composition of 1:3 weight ratio RHA:SiC (25% RHA: 75% SiC) offerscomparable specific strength values with the single SiC reinforced Al composite grades. The strain to fractureswas invariant to the weight ratio of RHA/SiC for all weight percent but the composite compositions containingRHA had improved fracture toughness compared with the single SiC reinforced Al composite grades.Keywords: Aluminium matrix composites, rice husk ash, mechanical properties, scan electron microscopy,stir casting, silicon carbide.1. INTRODUCTIONAlternative sources of reinforcements that offer thepotential of producing Aluminium matrix composites(AMCs) at reduced cost while maintaining highperformance levels is attracting interests fromresearchers [1-2]. Compared to other engineeringmaterials, AMCs are noted for the rare combination ofproperties they offer such as high specific strength andstiffness, good wear and corrosion resistance, lowthermal coefficient of expansion, good hightemperature mechanical properties, and excellentthermal management potentials among others [3-5].Aluminium based matrices also have the advantagethat they are the cheapest among other competingmatrix materials (Copper, Titanium, Magnesium) formetal matrix composites (MMCs) development; andalso are amenable to processing using techniquesconventionally suited for the production of metals andalloys [6,7].The unique properties of AMCs are derivedfrom the material characteristics of both the matrixand the reinforcing phases [8]. The reinforcementsare responsible for the improved mechanical, wear,and high temperature properties of the AMCs [9-10]. Thus the type of reinforcement andreinforcement parameters such as size, volumefraction, distribution, shape, and orientation oftenaffect significantly the properties of AMCs [11].The use of cheaper source of reinforcements suchas industrial wastes (fly ash, red mud) [12-13] andagro wastes (rice husk ash, bamboo leaf ash,coconut shell ash) [14-15] for AMCs developmentis gaining popularity considering its advantage insolid waste recycling which has been a cause formajor concern over the years. Additional to theadvantages of low cost, availability in largequantities, and contributions to creation of a moreeco-friendly environment; is lower densities whichmost of the agro and industrial wastes possess in160 13 th International Conference on Tribology – Serbiatrib’13

comparison with the synthetic reinforcements suchas silicon carbide (SiC) and alumina (Al 2 O 3 ) [16].The properties achieved with the sole utilization ofthese cheaper source reinforcements have beenreported to be lower than that of the syntheticreinforced but with promise for use in semistructuraland thermal management applications[17]. The use of hybrid reinforcements utilizingSiC/Al 2 O 3 and agro waste ashes as a means ofimproving the properties of AMCs has attractedinterest recently with very encouraging resultsobtained [18-19].The present work is aimed at investigating theinfluence of the weight ratios of rice husk ash andsilicon carbide on the mechanical behaviourAluminium matrix hybrid composites having variedweight percent of both reinforcements. Themotivation for this work is to establish optimumRHA/SiC weight ratios required to achieveoptimized performance of low cost AMCs developedwith the use of rice husk. Literatures on the use ofsynthetic/agrowaste hybrid reinforcements forAMCs development are still very limited and there iscurrently none that the authors are aware of thatdiscusses the use of RHA and SiC as hybridcomposites in Al-Mg-Si alloy matrix.2. MATERIALS AND METHOD2.1 MaterialsAl-Mg-Si alloy billets with chemicalcomposition determined using spark spectrometricanalysis (Table 1) was selected as Aluminiummatrix for this investigation.Table 1. Elemental composition of Al-Mg-Si alloy.Elementwt%Si 0.4002Fe 0.2201Cu 0.008Mn 0.0109Mg 0.3961Cr 0.0302Zn 0.0202Ti 0.0125Ni 0.0101Sn 0.0021Pb 0.0011Ca 0.0015Cd 0.0003Na 0.0009V 0.0027Al 98.88For the hybrid reinforcing phases, siliconcarbide (SiC) and rice husk ash (RHA) wereselected. The silicon carbide procured was of highchemical purity with average particle size of 28µmwhile rice husks utilized for the processing of ricehusk ash was obtained from Igbemo-Ekiti, EkitiState (a rice producing community in south westernNigeria). Magnesium for improving wettabilitybetween the Al-Mg-Si alloy and the reinforcementswas also procured.2.2 Preparation of Rice Husk AshThe procedure adopted is in accordance withAlaneme et al [16]. It involves the use of a simplemetallic drum with perforations as burner for therice husk. Dry rice husks placed inside the drumwas ignited with the use of charcoal. The husk wasallowed to burn completely and the ashes removed24 hours later. The ash was then heat-treated at atemperature of 650 o C for 180 minutes to reduce itscarbonaceous and volatile constituents. Sieving ofthe bamboo leaf ash was then performed using asieve shaker to obtain ashes with mesh size under50µm. The chemical composition of the rice huskash from this process is presented in Table 2.Table 2. Chemical Composition of the Rice Husk AshCompound/Element (constituent) weight PercentSilica (SiO 2 ) 91.59Carbon, C 4.8Calcium oxide CaO 1.58Magnesium oxide, MgO 0.53Potassium oxide, K 2 O 0.39Haematite, Fe 2 O 3 0.21Sodium, NatraceTitanium oxide, TiO 2 0.202.3 Composites ProductionTwo step stir casting process was utilized toproduce the composites [20]. The process startedwith the determination of the quantities of rice huskash (RHA) and silicon carbide (SiC) required toproduce 5, 7.5, and 10 wt% reinforcementconsisting of RHA and SiC in weight ratios 0:1,1:3, 1:1, 3:1, and 1:0 respectively (which amountsto 0, 25, 50, 75, and 100% RHA in thereinforcement phase). The rice husk ash and siliconcarbide particles were initially preheated separatelyat a temperature of 250 o C to eliminate dampnessand improve wettability with the molten Al-Mg-Sialloy. The Al-Mg-Si alloy billets were charged intoa gas-fired crucible furnace (fitted with atemperature probe), and heated to a temperature of750 o C ± 30 o C (above the liquidus temperature ofthe alloy) to ensure the alloy melts completely. Theliquid alloy was then cooled in the furnace to asemi solid state at a temperature of about 600 o C.The preheated rice husk ash and SiC particles alongwith 0.1 wt% magnesium were then charged into13 th International Conference on Tribology – Serbiatrib’13 161

the semi-solid melt at this temperature (600 o C) andstirring of the slurry was performed manually for 5-10 minutes. The composite slurry was thensuperheated to 800 o C± 50 o C and a second stirringperformed using a mechanical stirrer. The stirringoperation was performed at a speed of 400rpm for10minutes before casting into prepared sandmoulds inserted with chills. The designations usedto represent each grade of the composites producedare presented in Table 3.Table 3. Composite Density and Estimated Percent Porosity.Sample CompositionDesignation RHA: SiCTheoreticaldensity(g/cm 3 )Experimentaldensity(g/cm 3 )%PorosityA0 0 wt% 2.700 2.655 1.675wt%B1 A (0:1) 2.721 2.700 0.77B2 B (1:3) 2.691 2.650 1.52B3 C (1:1) 2.660 2.640 0.75B4 D (3:1) 2.630 2.590 1.52B5 E (1:0) 2.599 2.579 0.777.5 wt%C1 A (0:1) 2.733 2.670 2.31C2 B (1:3) 2.689 2.640 1.82C3 C (1:1) 2.640 2.590 1.89C4 D (3:1) 2.595 2.570 0.96C5 E (1:0) 2.550 2.510 1.5710 wt%D1 A (0:1) 2.743 2.690 1.9D2 B (1:3) 2.680 2.650 1.11D3 C (1:1) 2.620 2.610 0.3D4 D (3:1) 2.560 2.50 2.34D5 E (1:0) 2.500 2.497 0.122.4 Density MeasurementThe experimental density of each grade ofcomposite produced was determined by dividingthe measured weight of a test sample by itsmeasured volume; while the theoretical density wasevaluated by using the formula:ρ Al-Mg-Si / RHA-SiCp = wt. Al-Mg-Si × ρ Al-Mg-Si + wt. RHA ×ρ RHA + wt. SiC × ρ SiC (2.1)where, ρ Al-Mg-Si / RHA-SiCp = Density of Composite,wt. Al-Mg-Si = Weight fraction of Al-Mg-Si alloy, ρ Al-Mg-Si = Density of Al-Mg-Si alloy, wt. RHA = Weightfraction RHA, ρ RHA = Density of RHA, wt. SiC =Weight fraction SiC, and ρ SiC = Density of SiC.The experimental densities were compared withthe theoretical densities for each composition of theRHA-SiC reinforced composites produced; and itserved as basis for evaluation of the percentporosity of the composites using the relations [20]:% porosity = {(ρ T – ρ EX ) ÷ ρ T } × 100% (2.2)Where, ρ T = Theoretical Density (g/cm 3 ), ρ EX =Experimental Density (g/cm 3 ).2.5 Tensile PropertiesThe tensile properties of the composites wasevaluated with the aid of tensile tests performedfollowing the specifications of ASTM 8M-91standards [21]. The samples for the test weremachined to round specimen configuration with 6mm diameter and 30 mm gauge length. The testwas carried out at room temperature using anInstron universal testing machine operated at astrain rate of 10 -3 /s. Three repeat tests wereperformed for each grade of composite produced toguarantee repeatability and reliability of the datagenerated. The tensile properties evaluated from thestress-strain curves developed from the tension testare - the ultimate tensile strength (σ u ), the 0.2% offsetyield strength (σ y ), and the strain to fracture (ε f ).2.6 Fracture Toughness EvaluationThe fracture toughness of the composites wasevaluated using circumferential notch tensile (CNT)specimens [22]. Samples for the CNT testing weremachined having gauge length, specimen diameter(D), notch diameter (d), and notch angle of 30, 6,4.5mm, and 60 o respectively. The specimens werethen subjected to tensile loading to fracture usingan Instron universal testing machine. The fractureload (P f ) obtained from the load – extension plotsgenerated from the CNT testing were used toevaluate the fracture toughness using the empiricalrelations by Dieter [23]:K 1C =P f /(D) 3/2 [1.72(D/d)–1.27] (2.3)where, D and d are respectively the specimendiameter and the diameter of the notched section.The validity of the fracture toughness valuesobtained was determined using the relations inaccordance with Nath and Das [24]:D ≥ (K 1C /σ y ) 2 (2.4)Three repeat tests were performed for eachcomposite composition and the results obtainedwere taken to be highly consistent if the differencebetween measured values for a given compositecomposition is not more than 2%.2.7 Microstructural ExaminationA JSM 7600F Jeol ultra-high resolution fieldemission gun scanning electron microscope (FEG-SEM) equipped with an EDS was used for detailedmicrostructural study and for determination of theelemental compositions of the composites.162 13 th International Conference on Tribology – Serbiatrib’13

3. RESULTS AND DISCUSSION3.1 MicrostructureFigure 1 shows some representative SEMmicrographs of the RHA - SiC reinforced AMCsproduced. It is observed that there is a good dispersionof the RHA and SiC particulates in the Al alloy matrixand little particle clusters are observed. Thus there isno significant problem of segregation orsedimentation which often occurs duringsolidification of MMCs having components withdifferent densities and wettability characteristics [25].(d)(a)(e)Figure 1. showing (a) SE image of the Al-Mg-Si/5 wt%SiC composite showing the SiC particles dispersed in theAl-Mg-Si matrix; (b) SE image of the 5 wt% hybridreinforced Al-Mg-Si/RHA-SiC composite havingRHA:SiC weight ratio of 1:3; (c) SE image of the 7.5wt% hybrid reinforced Al-Mg-Si/RHA-SiC compositehaving RHA:SiC weight ratio of 1:3; (d) SE image of the10 wt% hybrid reinforced Al-Mg-Si/RHA-SiC compositehaving RHA:SiC weight ratio of 1:3; and(e) SE image ofthe Al-Mg-Si/10 wt% RHA composite showing the RHAparticles dispersed in the Al-Mg-Si matrix.(b)(c)This shows that the two step stir casting processadopted for the production of the composites isreliable judging from the microstructures examinedin Figure 1.The EDS profiles of the particulates in thecomposites produced, some of which are presentedin Figures 2 and 3; show peaks of aluminium (Al),oxygen (O), carbon (C), iron (Fe), silicon (Si),calcium (Ca), sodium (Na) and magnesium (Mg).The presence of these elements confirm the presenceof SiC; as well as silica (SiO 2 ), alumina (Al 2 O 3 ),Potassium oxide (K 2 O), ferric oxide (Fe 2 O 3 ), andMagnesium oxide (MgO) which are constituentsderived from the rice husk ash (Table 2).13 th International Conference on Tribology – Serbiatrib’13 163

(a)(b)Figure 2. Showing (a) representative SEPhotomicrograph showing the reinforcing particlesdispersed in the Al-Mg-Si matrix; and (b) EDS profile ofthe particle in 2(a) confirming the presence of Al 2 O 3 ,SiO 2 , Fe 2 O 3 , K 2 O, CaO, SiC and Na.(a)3.2 Composite Density and Estimated PercentPorosityThe results of the composite densities andestimated percent porosity are presented in Table 3.It is observed from the results that the estimatedporosity values are not dependent on the weightpercent of the reinforcement phase or the weightratio of RHA to SiC. It is however noted that theestimated porosity levels are less than 4 % whichhas been reported to be the maximum permissiblein cast AMCs [26]. The low porosity levels of thecomposites supports our submission that the twostep stir casting method adopted for producing thecomposites is reliable. As a result of the lowerdensity of RHA (0.31 g/cm 3 ) in comparison to SiC(3.6 g/cm 3 ), it is expected that the density of thecomposites will reduce with increase in the RHAcontent in the composite as observed from Table 3.3.3 Mechanical BehaviourThe variation of tensile strength and yieldstrength of the composites produced is presented inFigure 4. It is observed that there is a generalincrease in tensile strength (Figure 4a) and yieldstrength (Figure 4b) with increase in weight percentof the RHA-SiC hybrid reinforcement. However,for specific weight percents of the hybridcomposites (that is B, C, and D series), it is notedthat the tensile and yield strength decreases withincrease in the weight proportion of RHA in theRHA-SiC reinforcement. For the compositescontaining 5 wt% of the reinforcing phase, it isobserved that 4.9, 8.9, 12.5, and 15.8 % reductionin tensile strength was obtained from thecomposites with weight ratio RHA:SiC of 1:3, 1:1,3:1, and 1:0 (that is containing 25, 50, 75, and 100% RHA) in comparison to the 5 wt% SiC singlereinforced Al matrix composite. For the compositescontaining 7.5 wt% of the reinforcing phase,reductions of 5, 9, 13.4, and 19 % were observedfor the compositions of 1:3, 1:1, 3:1, 1:0 RHA:SiCweight ratios respectively (in comparison with the7.5 wt% SiC single reinforced composite). In thecase of the composites containing 10 wt%reinforcements, reductions of 4, 8.1, 13.2, and 18.3% was observed in comparison to the 10 wt % SiCsingle reinforced Al matrix composite.(b)Figure 3. Showing (a) representative SEPhotomicrograph of some clustered particles dispersed inthe Al-Mg-Si matrix; and (b): EDS profile the particlesidentified in 3(a) confirming the presence of Al 2 O 3 ,SiO 2 , Fe 2 O 3 , SiC, Cao, and Na which are constituentsfrom the RHA-SiC hybrid reinforcement.164 13 th International Conference on Tribology – Serbiatrib’13

(a)(b)Figure 4. Showing (a) variation of tensile strength forthe monolithic Al-Mg-Si alloy, single reinforced andhybrid reinforced Al-Mg-Si/RHA-SiC composites; and(b) variation of yield strength for the monolithic Al-Mg-Si alloy, single reinforced and hybrid reinforced Al-Mg-Si/RHA-SiC composites.It has been well reported that particle reinforcedAMCs achieve improved strength due to loadtransfer from the matrix to the particles (directstrengthening) and creation of more dislocationswhich serve as constraints to plastic deformation bythermal mismatch between the particles and theAluminium matrix arising from their differences incoefficient of thermal expansion (indirectstrengthening) [27-28]. Thus even in a scenariowhere the particles are not sufficiently strong toinduce strengthening via the ‘direct route’ of loadtransfer from matrix to particles, the indirectstrengthening it could offer is adequate to inducesome strength improvements well and above that ofthe monolithic alloy. In the present case underinvestigation, the reduction in strength observedwith increase in the RHA content of the compositesis as a result of the decrease of the directstrengthening capacity of RHA which containspredominantly silica. Silica is noted to be a softerceramic with elastic modulus of 60-70 GPa, whichis within the range of Aluminium unlike SiC whichhas an elastic modulus of 400GPa. Thus theefficiency of load transfer from the Al matrix to theparticles (load carrying capacity) of the hybridparticulates will be dependent on the amount of SiCthan RHA. However, it should be noted that samplesB5, C5, and D5 which contain only RHA, show aprogressive increase in tensile strength and yieldstrength with the increased weight percent of RHAsupporting our hypothesis that the indirectstrengthening mechanism (which entails dislocationgeneration results in higher dislocation densities withincreased weight percent of the particles) can resultin modest improvement in strength with increase inthe weight percent of the reinforcing particles.The variation of the specific strength of thecomposites produced with weight ratio of RHA/SiCis presented in Figure 5. It is observed that thespecific strengths of the composites generallyincreased with increase in the weight percent of thereinforcing phase (that is RHA-SiC weightpercent). Also the specific strength values decreaseswith increase in the RHA content in the hybridreinforcement. However, the % decrease in specificstrength of the composites is generally lower incomparison with that of the ultimate tensile strengthanalyzed earlier. For the 5 wt% compositions, it isobserved that 3.1, 6.8, 8.75, and 11.9 % reductionin specific strength is obtained. For the 7.5 wt %compositions (grades) 3.93, 6.2, 10 and 13.9 %reductions were obtained. In the case of the 10 wt%grade, 2.6, 5.3, 6.54, and 11.9 % reductions wereobtained. The results show that the composites withcomposition of 1:3 weight ratio RHA:SiC (25%RHA: 75% SiC) can offer comparable specificstrength values at reduced cost of production of thecomposite since its difference is less than 4 % for thethree weight percents of reinforcement worked on.Figure 5. Variation of specific strength for themonolithic Al-Mg-Si alloy, single reinforced and hybridreinforced Al-Mg-Si/RHA-SiC composites.The results of the variation of strain to fractureof the composites with weight percentreinforcement and weight ratio RHA/SiC ispresented in Figure 6. It is observed that there is ageneral decrease in ductility of the composites withincrease in the weight percent of reinforcing phase13 th International Conference on Tribology – Serbiatrib’13 165

in the composites. Closer observation show that foreach weight percent of hybrid compositesproduced, the strain to fracture was invariant to theweight ratio of RHA/SiC. It can be inferred fromthe results that the ductility levels of the hybridcomposites is not compromised by the addition ofRHA in the hybrid compositions. Thus its capacityto sustain plastic strain without fracture is notimpelled by the addition of RHA.particles like most hard and brittle ceramic particleshave a higher tendency to undergo rapid crackpropagation [32].Figure 7. Variation of Fracture Toughness for themonolithic Al-Mg-Si alloy, single reinforced and hybridreinforced Al-Mg-Si/RHA-SiC composites.4. CONCLUSIONSFigure 6. Variation of strain to fracture for themonolithic Al-Mg-Si alloy, single reinforced and hybridreinforced Al-Mg-Si/RHA-SiC composites.The fracture toughness values determined by theuse of circumferential notched tensile (CNT)specimens are presented in Figure 7. The valuesobtained were reported as plain strain fracturetoughness because the conditions for valid K 1C(plain strain condition) was met with the specimendiameter of 6mm when the relation D ≥ (K 1C /σ y ) 2[24] was utilised to validate the results obtainedfrom the CNT testing. It is observed that thefracture toughness decreases with increase in theweight percent of the composites. But for specificweight percents of the composites (that is B, C, andD series) it is observed that the compositecompositions containing RHA had improvedfracture toughness results compared with the singleSiC reinforced grades of the composites. Thus theaddition of RHA appears to be beneficial in termsof improving the resistance to crack propagation ofthe composites making them slightly lesssusceptible to sudden crack failure in comparisonwith the single reinforced SiC composite grades.The mechanism of fracture in particle reinforced Almatrix composites have been reported by severalauthors [29-30]. The primary mechanisms offracture have been reported to be facilitated by oneor a combination of particle cracking, interfacialcracking or particle debonding [31]. In the presentcase, the improved fracture toughness of thecomposites containing RHA, is most likely due tothe reduced amount of relatively harder and brittleSiC particles in the composites [19]. The SiCThe mechanical behaviour of Al-Mg-Si alloymatrix composites containing 5, 7.5, and 10 weightpercent of RHA and SiC reinforcements prepared inweight ratios 0:1, 1:3, 1:1, 3:1, and 1:0 respectivelywas investigated. The results show that:1. The estimated porosity values are notdependent on the weight percent of thereinforcement phase or the weight ratio ofRHA to SiC. They were however less than2.5 % in all grades produced.2. There is a general increase in tensile strength,and yield strength with increase in weightpercent of the RHA-SiC hybridreinforcement. However, the tensile andyield strength decreases with increase in theweight proportion of RHA in the RHA-SiCreinforcement.3. The specific strength followed the same trendas the tensile and yield strengths; however,the % decrease in specific strength of thecomposites is generally lower in comparisonwith that of the ultimate tensile strength. Thecomposites with composition of 1:3 weightratio RHA to SiC (25% RHA: 75% SiC)offers comparable specific strength valueswith the SiC single reinforced grades of thecomposite.4. There is a general decrease in ductility of thecomposites with increase in the weightpercent of reinforcing phase in thecomposites. However, the strain to fracturewas invariant to the weight ratio ofRHA/SiC.5. The fracture toughness decreases withincrease in the weight percent of the166 13 th International Conference on Tribology – Serbiatrib’13

composites. But the composite compositionscontaining RHA had improved fracturetoughness compared with the single SiCreinforced grades.ACKNOWLEDGEMENTThe electron microscopy assistance rendered byDr. P. A. Olubambi of the Department of Chemicaland Metallurgical Engineering, Tshwane Universityof Technology, South Africa is appreciated.REFERENCES[1] S.D. Prasad, R.A. Krishna: Production andMechanical Properties of A356.2 /RHAComposites, International Journal of AdvancedScience and Technology, Vol. 33, pp. 51-58, 2011.[2] H. Zuhailawati, P. Samayamutthirian, C.H. MohdHaizu: Fabrication of Low Cost Aluminium MatrixComposite Reinforced with Silica Sand, Journal ofPhysical Science, Vol. 18, No. 1, pp. 47-55, 2007.[3] T.V. Christy, N. Murugan, S. Kumar: Acomparative study on the microstructures andmechanical properties of Al 6061 alloy and theMMC Al 6061/TiB2/12p, Journal of Minerals andMaterials Characterization and Engineering, Vol.9, No. 1, pp. 57-65, 2010.[4] P.K. Rohatgi, B.F. Schultz, A. Daoud, W.W.Zhang: Tribological performance of A206aluminum alloy containing silica sand particles,Tribol Int, Vol. 43, pp. 455–66, 2010.[5] K.K. Alaneme, M. O. Bodunrin: Corrosionbehaviour of alumina reinforced Al (6063) metalmatrix composites, Journal of Minerals andMaterials Characterisation and Engineering, Vol.10, No. 2, pp. 1153-1165, 2011.[6] P. Rohatgi, B. Schultz: Light weight metal matrixcomposites – stretching the boundaries of metals,Materials Matters, Vol. 2, pp. 16-19, 2007.[7] D.B. Miracle: Metal matrix composites - fromscience to technological significance, CompositesScience and Technology, Vol. 65, No. 15/16, pp.2526-2540, 2005.[8] K.K. Alaneme: Influence of Thermo-mechanicalTreatment on the Tensile Behaviour and CNTevaluated Fracture Toughness of Borax premixedSiCp reinforced Aluminium (6063) Composites,International Journal of Mechanical and MaterialsEngineering, Vol. 7, No. 1, pp. 96-100, 2012.[9] T. Senthilvelan, S. Gopalakannan, S.Vishnuvarthan, K. Keerthivaran: Fabrication andCharacterization of SiC, Al 2 O 3 and B 4 CReinforced Al-Zn-Mg-Cu Alloy (AA 7075) MetalMatrix Composites: A Study, Advanced MaterialsResearch, Vol. 622-623, pp. 1295-1299, 2012.[10] S.A. Sajjadi, H.R. Ezatpour, H. Beygi:Microstructure and mechanical properties of Al–Al 2 O 3 micro and nano composites fabricated bystir casting. Materials Science and Engineering A,Vol. 528, pp. 8765-8771, 2011.[11] S. Tahamtan, A. Halvaee, M. Emamy, M.S Zabihi:Fabrication of Al/A206-Al 2 O 3 nano/microcomposite by combining ball milling and stircasting technology, Materials and Design 2013,DOI:http://dx.doi.org/10.1016/j.matdes.2013.01.032.[12] P. Naresh: Development and characterization ofmetal matrix composite using red Mud anindustrial waste for wear resistant applications,PhD Thesis, Department of MechanicalEngineering, National Institute of TechnologyRourkela, India, 23 – 34, 2006.[13] J. Bienia, M. Walczak, B. Surowska B, J.Sobczaka: Microstructure and corrosionbehaviour of aluminium fly ash composites,Journal of Optoelectronics and AdvancedMaterials, Vol. 5, No. 2, pp. 493-502, 2003.[14] P.B. Madakson, D.S. Yawas, A. Apasi:Characterization of Coconut Shell Ash forPotential Utilization in Metal Matrix Compositesfor Automotive Applications, International Journalof Engineering Science and Technology (IJEST),Vol. 4, No. 3, pp. 1190-1198, 2012.[15] S.D. Prasad, R.A. Krishna: TribologicalProperties of A356.2/RHA Composites, Journal ofMaterials Science and Technology, Vol. 28, No. 4,pp. 367-372, 2012.[16] K.K. Alaneme, I.B. Akintunde, P.A. Olubambi,T.M. Adewale: Mechanical Behaviour of RiceHusk Ash – Alumina Hybrid ReinforcedAluminium Based Matrix Composites, Journal ofMaterials Research and Technology, Vol. 2, No. 1,pp. 60-67, 2013.[17] R. Escalera-Lozano, C. Gutierrez, M.A. Pech-Canul, M.I. Pech-Canul: Degradation of Al/SiCpComposites produced with Rice-Hull Ash andAluminium Cans, Waste Management, Vol. 28, pp.389-395, 2008.[18] K.K. Alaneme, P.A. Olubambi: Corrosion andWear Behaviour of Rice Husk Ash – AluminaReinforced Aluminium Matrix Hybrid Composites,Journal of Materials Research and Technology,2013 (In Press).[19] K.K. Alaneme, E.O. Adewuyi: MechanicalBehaviour of Bamboo Leaf Ash – Alumina HybridReinforced Aluminium Based Matrix Composites,Metallurgical and Materials Engineering, Serbia,2013 (In Press).[20] K.K. Alaneme, A.O. Aluko: Production and agehardeningbehaviour of borax pre-mixed SiCreinforced Al-Mg-Si alloy composites developedby double stir casting technique, The West IndianJournal of Engineering, Vol. 34, No. 1/2, pp. 80-85, 2012.[21] ASTM E 8M: Standard Test Method for TensionTesting of Metallic Materials (Metric), AnnualBook of ASTM Standards, Philadelphia; 1991.13 th International Conference on Tribology – Serbiatrib’13 167

[22] K.K. Alaneme: Fracture toughness (K 1C ) evaluationfor dual phase low alloy steels using circumferentialnotched tensile (CNT) specimens, MaterialsResearch, Vol. 14, No. 2, pp. 155-160, 2011.[23] G.E. Dieter: Mechanical Metallurgy, McGraw-Hill, Singapore; 1988.[24] S.K. Nath, U.K. Das: Effect of microstructure andnotches on the fracture toughness of mediumcarbon steel, Journal of Naval Architecture andMarine Engineering, Vol. 3, pp. 15-22, 2006.[25] B.F. Schultz, J.B. Ferguson, P.K. Rohatgi:Microstructure and hardness of Al 2 O 3nanoparticle reinforced Al–Mg compositesfabricated by reactive wetting and stir mixing,Materials Science and Engineering A, Vol. 530,pp. 87-97, 2011.[26] M. Kok: Production and mechanical properties ofAl 2 O 3 particle reinforced 2024 aluminiumcomposites, Journal of Materials ProcessingTechnology, Vol. 16, pp. 381-387, 2005.[27] K.K. Alaneme KK, A.O. Aluko: FractureToughness (K1C) and Tensile Properties of As-Cast and Age-Hardened Aluminium (6063) –Silicon Carbide Particulate Composites, ScientiaIranica, Transactions A: Civil Engineering(Elsevier), Vol. 19, No. 4, pp. 992-996, 2012.[28] N. Chawla, Y. Shen: Mechanical behaviour ofparticle reinforced metal matrix composites,Advanced Engineering Materials, Vol. 3, No. 6,pp. 357-370, 2001.[29] M.T. Milan, P. Bowen: Tensile and fracturetoughness properties of SiCp reinforced Al alloys:Effects of particle size, particle volume fractionand matrix strength, Journal of MaterialsEngineering and Performance, Vol. 13, No. 6, pp.775-783, 2004.[30] M.M. Ranjbaran: Low fracture toughness in Al7191-20% SiCp aluminium matrix composite,European Journal of Scientific Research, Vol. 41,No. 2, pp. 261-272, 2010.[31] K.K. Alaneme: Mechanical Behaviour of ColdDeformed and Solution Heat-treated AluminaReinforced AA 6063 Composites, The West IndianJournal of Engineering, Vol. 35, No. 2, pp. 31-35,2013.[32] Fracture Mechanics of Ceramics, Vol. 13 – Crack-Microstructure Interaction, R – Curve Behaviour,Editors (Bradt RC, Munz D, Sakai M,Schevchenko V Ya, White KW, Springer, FirstEdition, pp. 538, 2002.168 13 th International Conference on Tribology – Serbiatrib’13

Serbian TribologySocietySERBIATRIB ‘1313 th International Conference onTribologyKragujevac, Serbia, 15 – 17 May 2013Faculty of Engineeringin KragujevacTRIBOLOGICAL PROPERTIES OF NANOMETRIC ATOMICLAYER DEPOSITIONS APPLIED ON AISI 420 STAINLESSSTEELE. Marin 1 , A.Lanzutti 1 ,L.Fedrizzi 11 University of Udine, Department of Chemistry, Physics and Environment, Via del Cotonificio 108, 33100 Udine Italy,elia.marin@uniud.itAbstract: Atomic Layer Deposition (ALD) is a modern technique that allows to deposit nanometric,conformal coatings on almost any kind of substrate, from plastics to ceramic, metals or even composites.ALD coatings are not dependent on the morphology of the substrate and are only regulated by thecomposition of the precursors, the chamber temperature and the number of cycles.In this work, mono- and bi-layer nanometric, protective low-temperature ALD coatings, based on Al2O3 andTiO2, were applied on AISI 420 Stainless Steel in order to enhance its relatively low corrosion resistance inchloride containing environments. Tribological testing were also performed on the ALD coated AISI 420 inorder to evaluate the wear and scratch resistance of these nanometric layers and thus evaluate theirdurability.Scratch tests were performed using a standard Rockwell C indenter, under a variable load condition, inorder to evaluate the critical loading condition for each coating. Wear testing were performed using astainless steel counterpart, in ball-on-disc configuration, in order to measure the friction coefficient and toconfront the wear resistance. All scratch tests scars and wear tracks were then observed by means ofScanning Electron Microscopy (SEM) in order to understand the wear mechanisms that occurred on thesample surfaces.Corrosion testing, performed under immersion in 0.2 M NaCl solutions, clearly showed that the ALDcoatings have a strong effect in protecting the Stainless Steel substrate against corrosion, reducing thecorrosion current density by two orders of magnitude.The preliminary tribological results showed that ALD depositions obtained at low temperatures have abrittle behavior caused by the amorphous nature of their structure, and thus undergo delaminationphenomena during Scratch Testing at relatively low applied loads. During ball-on-disc testing, the coatingswere removed from the substrate, in particular for monolayer ALD configurations, which seem to have alower toughness when compared to bi-layer configurations..Keywords: Atomic Layer Deposition, Al 2 O 3 , TiO 2 , wear, AISI 420, Stainless Steel, Scratch Test, Ball-on-disc1. INTRODUCTIONMartensitic stainless steels are widely used for awide range of applications, mainly due to theirbalanced properties, as they couple relativelymicrohardness, mechanical resistance and corrosionresistance in many aggressive environments [1].For these reasons, martensitic stainless steels arenowadays applied for: knife blades [2], oil and gas[3], offshore platforms [4], turbine blades [5],components subject to abrasive wear at relativelyhigh temperature or aggressive environments [6].Even so, stainless steels could show insufficientcorrosion resistance in strongly aggressive mediacontaining Cl - and S 2- ions, at high temperature orvery high / low pH values [7]. For these reasons, inparticular circumstances Stainless Steels may needa further improvement of corrosion protection.Conventional treatments, such as painting, are13 th International Conference on Tribology – Serbiatrib’13 169

hardly applicable to Stainless Steel due to adhesionproblems between paint and the metal substrate [8].A great number of innovative treatments arenowadays under intensive study to improveStainless Steel corrosion resistance, such as plasmadetonation techniques [9], arc-ion plating [10], solgeldeposition [11], chemical conversion layers ofcerium [12] chromium [13] or other elements,Chemical Vapor Deposition [14], High-VelocityOxy-fuel Spray (HVOF) [15], plasma-nitriding[16], and Atomic Layer Deposition [17]. All thesetechniques may also be applied in order to improvethe tribological resistance of the substrate, as theygrant higher hardness and wear abrasion whencompared to stainless steel or other commonmetallic alloys [18-20].The modern concept of ALD is an extension ofthe ALE (Atomic Layer Epitaxy), patented by Prof.Suntola [21]. Suntola’s studies were mainly focusedon switching effects in chalcogenide nanometricfilms for solid state electronic devices [22][23][24],and lately extended to a wide range of amorphoussemi-conductive thin films [25]. ALD (as ALE)process involves a sequence of self-limiting surfacereactions. As evidenced in 1980 by Ahonen et al.[26], the self-limiting characteristic of each reactionstep differentiates ALE and ALD from otherchemical vapor deposition technolnologies. In ALDeach deposition cycle is clearly divided in foursteps: in the first step a precursor is injected in thedeposition chamber. The precursor is chosen so thatits molecules will not react with each other at thedeposition temperature. In ideal situations, a singlemonolayer is thus formed as a result of the reactionwith the substrate. In the second step, the chamberis purged with nitrogen or argon gas in order toremove the excess of reactant and prevent“parasitic” CVD deposition on the substrate, whichwill eventually occur if two different precursors arepresent in the deposition chamber at the same time.In the third step, the second precursor is injected inthe chamber. In the case of metal oxide layers, thisis an oxidant agent, usually simple H 2 O. The laststep of the deposition cycle is a second purge toremove the excess of reactant with purging gas.Closed-loop repetitions of the four basic stepstheoretically allow obtaining perfectly conformaldeposits of any desired thickness. By avoiding thecontact between the precursors throughout thewhole coating process, a film growth at atomiclayer control, with a thickness control within ~ 10pm, can be obtained.Interest in ALD has increased stepwise in themid-1990’s and 2000’s, with the interest focused onsilicon-based microelectronics [27]. Up to thepresent time, ALD processes have been used todeposit several types of nanometric films, includingseveral chemical compounds (e.g. AsGa, CdSe,…),metal oxides (e.g. Al 2 O 3 , CaO, CuO, Er 2 O 3 , Ga 2 O 3 ,HfO 2 , La 2 O 3 , MgO, Nb 2 O 5 , Sc 2 O 3 , SiO2, Ta 2 O 5 ,TiO 2 , Y 2 O 3 , Yb 2 O 3 , ZnO, ZrO 2 ), nitrides (e.g. TiN,TaN, AlN, GaN, WN, NbN), sulfides (e.g. SrS,ZnS), carbides (e.g. TaC, TiC), fluorides (e.g.CaF2, LaF 3 , MgF 2 ), pure metals (e.g. Ru, Ir, Ta,Pt), biomaterials (e.g. hydroxyapatite(Ca 10 (PO 4 ) 6 (OH) 2 )) and even polymers (e.g.Polyimides) [28,29].Results on the corrosion protection on stainlesssteel by ALD TiO 2 and Al 2 O 3 layers were alreadyobtained by Matero et al. (1999)[29], whichsupposed that the conformal ALD coatings couldimprove the corrosion resistance of different metalalloys. In 2007, Shan et al. [30] used TiO2 ALDlayers to protect an undefined stainless steel,obtaining only a limited effect. In 2011, Marin et al.[31], Diaz et al. [17] and Potts et al. [32] clearlyshowed that the residual porosity of ALD layersdecreases increasing the thickness of the layer thusimproving the protection of the substrate. In mostcases [31,17,32, 33] the nanometric ALD layersclearly showed a corrosion protection similar, if notsuperior to conventional protective technolniquesand thicker coatings, even if common industrialtests (salt spray) performed on Plasma EnhancedALD by Potts et al. [32] clearly showed a timelimitedcorrosion protection.In this work, characterization of ALD coatedAISI 316 L has been carried out, in order todetermine the possible use of nanometric ceramicALD coatings for the corrosion and tribologicalprotection of stainless steel.2. EXPERIMENTAL2.1 Samples productionDiscs of standard AISI 420 martensitic stainlesssteel (chemical composition wt.%: C = 0.035; P

coated and uncoated region after adhesive taperemoval.The ALD coating was deposited using a TFS500 reactor (Beneq Oy, Finland): Al 2 O 3 layers wereobtained using trimethylaluminium (Al(CH 3 ) 3 ) andH2O precursors and TiO 2 layers were obtainedusing titanium tetrachloride (TiCl 4 ) and H 2 Oprecursors. Both depositions were performed at atemperature of 120 °C. The low temperatureprocesses were chosen in order to obtain anamorphous structure for both layers. The number ofprecursor cycles for each deposition was calculatedusing a growth rate per cycle (GPC) of ~0.1nm/cycle for TiO 2 and a GPC of ~0.15 nm/cycle forAl 2 O 3 .The samples were coated using different ALDconfigurations, with an overall coating thickness ofabout 200 nm.2.2 MorphologyMorphological characterization was carried outusing Veeco’s Digital Instrument’s Nanoscope IIIaatomic force microscope (AFM) in tapping modeconfiguration, using a Bruker SCM-PIT tip(Antimony (n) doped Si, frequency: 60-100 kHz,elastic constant: 1-5 N/m, PtIr coated) and CarlZeiss EVO-40 scanning electron microscope (SEM)with an operating voltage of 20 kV. Analyses wereperformed on the stainless steel substrate and oncoated samples. In particular, SEM was used inorder to scan the surface of the sample andinvestigate the presence of deposition defectsand/or surface anomalies. AFM was mainly used toinvestigate the presence of surface morphologicaldefects on the coating that were supposed to behardly visible using SEM due to the coatingtransparency. AFM was also used at the interfaceregions between coated and uncoated substrate afteradhesive tape removal in order to obtaininformation about the overall thickness of thedeposits and confront it with the theoreticaldeposition rates values and the results obtainedfrom GDOES analyses.2.3 CompositionIn-depth compositional analyses were carriedout using Horiba Yobin-Yvon’s RF – GD ProfilerGDOES. Due to the difficulties in calibration ofGDOES for the analysis of Titanium andAluminum oxides, only qualitative compositionalanalyses were performed, even if a keen calibrationwas performed in order to obtain reliable singlelayer thickness values. GDOES technolnique isstrongly influenced by surface roughness, whichwas relatively high.2.4 Mechanical propertiesIn order to evaluate the resistance todelamination of the different ALD configurations,Vickers indentations were applied to the samplesunder different load conditions (HV0.1-0.2-0.5-1.0-2.0) and the delaminated areas have been thenmeasured using a specific image post-processingsoftware (Wayne Rasband, ImageJ 1.44p). As allindentation hardness tests, Vickers indentation aredependant from the mechanical characteristics ofthe substrate and can only be used to comparedifferent coatings applied on the same substrate andnot results from different substrates. As adhesion isstrictly connected to the surface roughness [33-35],Vickers adhesion tests will give reliable results onlyfor substrates with similar surface finishing.2.5 Electrochemical propertiesElectrochemical characterization of the differentsamples was performed using PotendiodynamicPolarizations. An AUTOLAB PGSTAT-20potentiostat was used in a standard three electrodesconfiguration. The reference electrode wasAg/AgCl and the counter electrode was a 99.99%pure Platinum wire. All measurements have beenperformed in a pH 6.5, 0.2 M solution of NaCl. Allpolarization curves were obtained using a scanspeed of 0.2 mV/s after 10 minutes of immersion ofthe samples, in order to stabilize the OCP. Thepotential has been increased from -200 mV respectto the OCP to a measured current density of about10-3 A/cm 2 .2.6 Wear propertiesWear properties were investigated using anindustrial CETR UMT tribometer in ball-on-discconfiguration, using a WC counter-material. Weartesting was performed for 1, 10 and 100 cyclesunder dry conditions, at a relatively low rotatingspeed (1 rps) and at different diameters (15 mm, 18mm, 21 mm). After testing, the wear tracks wereobserved using SEM and the volume losses werethen estimated using a stylus profilometer.3. EXPERIMENTAL RESULTS3.1 MorphologySEM resulted to be an inadequate technique forthe analysis of ALD layers, since no morphologicaldifferences were found between images obtained oncoated and uncoated regions and no morphologicalproperties of the ALD layers could be correctly13 th International Conference on Tribology – Serbiatrib’13 171

esolved using this technique, even at relativelyhigh magnifications, such as 20k or 50k.Figure 2: SEM images of Al2O3 coated sampleobtained at 1000 (a) and 5000 (b) magnificationsFigure 1: SEM images of Al2O3 coated sampleobtained at 1000 (a) and 5000 (b) magnificationsFig. 1 shows the SEM image obtained on theAl 2 O 3 coated samples, at relatively “low” (fig. 1a)and “high” (fig. 1b) magnifications. It is onlypossible to discriminate the presence of scars andscratches caused by the cutting and grinding of thestainless steel substrates, while no informationabout the presence of the ALD layer could beobtained from these images.Similar results were obtained in the case of theAl 2 O 3 /TiO 2 coated samples, at relatively “low” (fig.2a) and “high” (fig. 2b) mangifications. It can beobserved that, even for this sample, only scars andscratches caused by the mechanical cutting andgrinding of the sample can be observed.AFM observations gave similar results, withonly scratches and scars clearly visible on themorphological maps obtained. In the case of themaps obtained at the interface between coating andsubstrate, they were successfully used in order toestimate the overall coating thickness of thedifferent ALD layers (fig. 3).Figure 3: AFM image obtained at the interface betweencoated and uncoated areas of the Al 2 O 3 coated sample.The results obtained are resumed in Table 1:Table 1: Coating thickness as obtained by AFMmeasurementsCoatingThicknessµmAl2O3 102± 3Al2O3/TiO2 107 ± 53.2 CompositionGDOES thickness measurement accuracy wasstrongly influenced by the sharpness of theinterface region between coating and substrate.Conventionally, in this work the coating thickness172 13 th International Conference on Tribology – Serbiatrib’13

has been measured between the top surface and theintersection point between oxygen and iron signals.Since no reference materials were present foramorphous ceramics, a specific GDOES calibrationwas required. Sputtered crater’s depth wasmeasured using a stylus profilometer and thesputtering rate has been evaluated accordingly.Following this calibration GDOES results shall beconsidered semi-quantitative and for this reason thecomposition of the coatings and in particular ofAl2O3 layers is not stoichiometric.Fig. 4 shows the results obtained by GDOES onthe Al 2 O 3 coated sample.be observed up to 100 nm, while the Aluminiumsignals can be observed up to 170 nm.Figure 5: GDOES graph obtained for the Al2O3 coatedsample. Signals of Iron, Chromium, Titanium, Oxygenand Aluminum3.3 Mechanical propertiesFigure 4: GDOES graph obtained for the Al2O3 coatedsample. Signals of Iron, Chromium, Titanium, Oxygenand AluminumIt can be observed that the interface betweencoating and substrate is not sharp. This is caused bythe relatively high surface roughness obtained afterthe sample preparation. The iron signal, which isconsidered representative for the substrate, reachesa plateau at about 200 nm, which is influenced byboth coating thickness and surface roughness,meaning that, after 200 nm, only substrate signalscan be seen by GDOES. At about 75 nm, the twosignals of Iron and Oxygen are even, while at thesurface, both Oxygen and Aluminium show a peak.This behaviour is caused by the substrateroughness, which causes a strong signal overlap.For this reason, coating thickness could not becorrectly estimated using GDOES, but only thethickness range could be determined. Surface peakin oxygen content can also be explainedconsidering the well known hydrogen effect [36].In the case of the bi-layer formed by a layer ofAl2O3 followed by a second layer of TiO2, bothsignals of Ti and Al can be observed in the first150-200 nm of analysis, while Oxygen can beobserved for more than 300 nm. The Iron signalsshows a plateau after about 170 nm. It can beobserved that Titanium and Aluminium have a clearpeak in the coating region. The titanium signals canFigure 6: 3 N Vickers indentation on AISI 420 StainlessSteel coated with Al 2 O 3Vickers indentations performed on the coatedsample surface were observed using SEM, as theOptical Microscope field depth resulted to be toosmall to sharply resolve all the delamination detailson the deformed region in just one image. Thedimension of the Vickers indentation is related tothe applied load. Even if localized defects arehardly visible on the ALD coatings due to theirtransparency, larger areas with a complete coatingremoval resulted to be clearly visible on the SEMimages. A clear example of how the delaminatedareas have been evaluated is presented in Fig. 6 andFig. 7. Fig. 6, in particular, shows a Vickersindentation as obtained using a 3 N load on the 100nm TiO 2 coated sample, while Fig. 7 shows detailsof cracks and delamination obtained using a 10 Nload. The results obtained for the different samples13 th International Conference on Tribology – Serbiatrib’13 173

are plotted in Fig. 8, as a function of the appliedload.a passive behavior in the 0.2 M NaCl solution, witha corrosion current density between 10 -6 and 10 -7A/cm2 and a passive region from -0.25 to -0.1 Vwith respect to Ag/AgCl. The Open CircuitPotential (OCP) for uncoated AISI 316 L steelresulted to be about 0.1 V with respect to Ag/AgCl.Figure 7: Cracks and delamination in proximity of aVickers indentation (5N on Al 2 O 3 coated AISI 420)All coatings showed a clear dependence on theapplied load for the delaminated Area/Load ratio,which remains almost constant only at the highestindentation loads (9.8 N = HV1 and 19.6 N =HV2).It can be observed that the best behaviour isshown by the coating formed by two differentlayers, with a lower delamination, in particular atlow loading conditions. At high loading conditions,and in particular at 19.6 N, the two coatings shownalmost the same delamination areas.Data dispersion was considerably high so astatistical approach was followed, with about 20Vickers indentations per load for each sample.Figure 3: polarization curves obtained on both coatedsamples and on naked substrateThe two coatings showed two different shifts ofthe corrosion potential, a positive shift to about -0.15 V respect to Ag/AgCl in the case of the singlelayer Al 2 O 3 coating and a negative shift to about -0.35 V in the case of the bi-layered structure ofAl 2 O 3 and TiO 2 .In the case of the single layer coating, acorrosion current density reduction of about oneorder of magnitude, to about 3*10 -8 A/cm 2 , wasobserved, while in the case of the bi-layeredstructure, the corrosion current density reductionresulted to be more intense, reaching 3*10 -9 A/cm 2 .In the case of the single layer coating, the barriereffect resulted to be similar in extension to thepassive range of the naked AISI 420 substrate, atlower currents, while an extended barrier effect wasobserved in the case of the bi-layered structure.3.5 Wear propertiesFigure 8: delamination as a function ofthe indentation load3.4 Electrochemical propertiesPolarization curves for the different sampleswith a total coating thickness of about 100 nm areshown in Fig. 9. Uncoated AISI 420 clearly showsFigure 9: wear track obtained after just 10 cycles forAl 2 O 3 single layer174 13 th International Conference on Tribology – Serbiatrib’13

Fig. 9 shows the wear track obtained after 10cycles of ball-on-disc wear testing on the samplecoated with a single layer of Al 2 O 3 . It can beobserved that, even if the applied load is relativelylow, a wear track is clearly visible on the surface ofthe sample and, in particular, third body wear scarscan be observed. The pristine surface roughness ofthe sample has been completely removed from thesurface due to the tribological contact with the WCcounter-part.EDXS localized analysis shown that the ALDlayer has been completely removed from thesurface due to the tribological contact.It is possible that the clearly visible third bodywear on the surface of the sample has also beencaused by Al 2 O 3 particles detached from the samplesurface during testing.Increasing the number of cycles, it was possibleto observe the formation of a thick oxide layerinside the wear track, and strong adhesionphenomena, leading to a coarse wear track full ofirregular oxide depositions.applied to AISI 420 martensitic stainless steel usinga thermal ALD process based on H 2 O, TMA andTiCl 4 .The applied ALD layers and the substrate wheresuccessfully characterized using SEM, AFM,GDOES, stylus profilometer, electrochemicalequipment and an industrial tribometer in ball-ondiscconfiguration.The morphological characterization evidencedthat ALD layers are conformal and almost defectfreeeven at relatively high magnifications.GDOES testing correctly discriminated thepresence of the different ceramic layers on the AISI420 substrate.Adhesion testing showed that adhesion betweenAISI 420 and 200 nm ALD layers is relatively poorand cracks propagate from microhardness Vickersindentations.Electrochemical testing clearly showed that evenan ALD coating with a limited thickness of about200 nm is sufficient to strongly improve thecorrosion resistance of the martensitic stainlesssteel substrate, strongly reducing the corrosioncurrent densities and widening the passive range ofthis material.Wear testing results showed that ALD layersdeposited at the temperature of 120 °C onmartensitic stainless steel, are unable to granttribological resistance to the substrate, due to theirintrinsic brittleness and the relatively low hardnessof the substrate on which they were applied.REFERENCESFigure 10: wear track obtained after just 10 cycles forAl 2 O 3 /TiO 2 bi-layerFig. 10 shows the wear track obtained after 10cycles of ball-on-disc wear testing on the samplecoated with a bi-layer formed by Al 2 O 3 and TiO 2 .It can be observed that, as seen before for thesingle layer sample, the wear track is clearly visibleat 1000 magnifications. Scars of third body wearare present inside the wear track, even if the trackitself is smaller when compared to the trackobtained from the single layer sample.Even in this case, EDXS analyses shown that theALD layer has been completely removed from thewear track on the sample surface, due to thetribological contact with the WC counterpart.4. CONCLUSIONSNanometric ALD mono- and bi- layers, with atotal thickness of about 200 nm, were successfully[1] H. Zhang, Y.L. Zhao, Z.D. Jiang, Materials Letters59, 3370-3374; 2005.[2] P.J. Hidalgo, Surcae Technology 10, 193-208; 1980.[3] A. Turnbull, B. Nimmo, Environment-InducedCracking materials, 2008, 2, 483-492;[4] J.W. Boh, L.A. Louca, Y.S. Choo, Shreir’sCorrosion 3, 1802-1878, 2010.[5] D. Ferreno, J.A. Alvarez, E. Ruiz, D. Mendez, L.Rodriguez, D. Hernandez, Engineering FAilureAnalysis, 18, 256-270, 2011.[6] B. Fournier, M. Sauzay, C. Caes, M. Noblecourt, M.Mottot, A. Bougault, V. Rabeau, J. Man, O. Gillia,P. Lemoine, A. Pineau, International Journal ofFatigue 30 (10-11), pp. 1797-1812, 2008.[7] C. Liu, G. Lin, D. Yang, M. Qi, Surface and CoatingTechnology 200, pp. 4011-4016, 2006.[8] V.H.V. Sarmento, M.G. Schiavetto, P. Hammer,A.V. Benedetti, C.S. Fugivara, P.H. Suegama,[9] S.H. Pulcinelli, C.V. Santilli, Surface and CoatingsTechnology 204, pp. 2689-2701, 2010.[10] R. Di Maggio, L. Fedrizzi, S. Rossi, P. Scardi, ThinSolid Films 286, pp. 127-135, 1996.13 th International Conference on Tribology – Serbiatrib’13 175

[11] C. Wang, F. Jiang, F. Wang, Corrosion Science 46,75-89, 2004.[12] C.R. Tomachuk, C.I. Elsner, A.R. Di Sarli, O.B.Ferraz, Materials Chemistry and Physics 119, pp.19-29, 2010.[13] D. Pech, P. Steyer, J. P. Millet, Corrosion Science50, 1492-1497, 2008.[14] J. Kawakita, T. Fukushima, S. Kuroda, T. Kodama,Corrosion Science 44, 2561-2581, 2002.[15] C.X. Li, T. Bell, Corrosion Science 48, pp. 2036-2049, 2006.[16] C.X. Shan, X. Hou, K. L. Choy, Surface & CoatingsTechnology 202, pp. 2399–2402, 2008.[17] E. Marin, L. Guzman, A. Lanzutti, W. Ensinger, L.Fedrizzi, Thin Solid Films 522, pp. 283-288; 2012.[18] P. Corengia, F. Walther, G. Ybarra, S. Sommadossi,R. Corbari, E. Broitman, Wear, pp. 479-485, 2006.[19] O. Choumad, X. Badiche, P. Montmitonnet, Y.Gachon, Journal of Manufacturing Processes 15, pp.77-86, 2013.[20] J.B. Cambon, F. Ansart, J.P. Bonino, V. Turq,Progress in Organic Coatings 75, pp. 486-493, 2012.[21] T. Suntola, J. Antson, J. (1977) U.S. Patent4,058,430;[22] T. Suntola, Solid State Electron. 14 933, 1971.[23] T. Stubb, T. Suntola, O.J.A. Tiainen, Solid StateElectron., 15, 611, 1972.[24] T. Suntola, O.J.A. Tiainen, M. Valkiainen, ThinSolid Films 227, 1972.[25] T. Suntola, Thin Solid Films 34 9, 1976.[26] M. Ahonen, M. Pessa, T. Suntola, Thin Solid Films65, 301, 1980.[27] J. Lua, J. Aarik, J. Sundqvist, K. Kukli,1, A. Harsta,J.-O. Carlsson, J. Cryst. Growth 273, 510, 2005.[28] Bedair, S. M., The Encyclopedia of AdvancedMaterials, vol 1. Edited by Bloor D, Brook RS,Flemings, M.C., Mahajan, S., Oxford: Pergamon141, 1994.[29] R. Matero, M. Ritala, M. Leskelä, T. Salo, J.Aromaa, O. Forsén, J. Phys. IV France 09, 493,1999.[30] C.X. Shan, X. Hou, K. L. Choy, Surf. Coat.Technol. 202, 2147, 2008.[31] E. Marin, A. Lanzutti, L. Guzman, L. Fedrizzi, J.Coat. Technol. Res. DOI: 10.1007/s11998-011-9327-0, 2011.[32] S. E. Potts, L. Schmalz, M. Fenker, B. Dìaz, J.Światowska, V. Maurice, A. Seyeux,P. Marcus, G.Radnóczi, L. Tóth, M. C. M. van de Sanden, and W.M. M. Kessels, J. Electrochem. Soc. 158, 132, 2011.[33] M. Ritala, Appl. Surf. Sci. 112, 223, 1997.[34] J. Takadoum , H.H. Bennani, Surf. Coat. Technol.96, 272, 1997.[35] H.C. Barshilia, A.Ananth, J. Khan, G. Srinivas, Vac.8, 1165, 2012.[36] R. Payling, D.G. Jones and A. Bengtson, “GlowDischarge Optical Emission Spectrometry”, JohnWiley, Chichester, 1997.176 13 th International Conference on Tribology – Serbiatrib’13

Serbian TribologySocietySERBIATRIB ‘1313 th International Conference onTribologyKragujevac, Serbia, 15 – 17 May 2013Faculty of Engineeringin KragujevacPREPARATION AND CHARACTERIZATION OF QUATERNARYAMMONIUM SURFACTANTS ON MUSCOVITE MICAJelena MANOJLOVIĆ 11 The Faculty of Mechanical Engineering Niš, Aleksandra Medvedeva 14, 18000 Niš, Srbijajmanojlo@gmail.comAbstract: In order to reduce friction, lubricants are usually employed. We describe the possibility of selfassembledmonolayers (SAMs) being used as lubricants. Our aim was to produce homogeneousmonolayers of surfactants on substrate. Because the surfactant behavior in the solution can be changed incontact with a substrate, we considered the adsorption mechanisms of surfactant molecules at the solidliquidinterface to be important. SAMs were prepared by using different methods, and their stability andstructure was studied, but a homogeneous hydrophobic monolayer was difficult to realize. It has beenshown that the factors, such as the temperature during solution preparation, frequently neglected, can bevery important in the process of surfactant adsorption from solution.Keywords: adsorption, SAMs, stability, muscovite mica, contact angle1. INTRODUCTIONThe tribological research involves the study offriction, wear and lubrication. These threecomponents between materials in contact are offundamental importance in many areas, such asaircraft engines, automobiles, gears or bearing.Practically, in all modern mechanical machines themajor problems are friction and wear. Theadvanced technical applications and the inventionof new characterization techniques, lead to theappearance of the new field of tribology, known asmicro-tribology/nano-tribology, which involvesthe study of friction and wear at a very smalllength scales. Nanotribology part of particularimportance is called nanolubrication, defined as thestudy of ultra-thin lubricant films. The applicationof nanolubrication is important in a device whichneeds a lubricant layer of few nanometers, toreduce friction and wear. Therefore, research andtesting of nano-lubricants, needed formicroelectromechanical systems (MEMS)components, is a big challenge.An attractive model system that can be used forstudies in wetting, adhesion and friction of surfacesis the system of self-assembled organic monolayers(SAMs). The concept of self-assembly was invokedby Zisman, more than 50 years ago. At that time,the potential of self-assembly was not recognized,but the field of self-assembled monolayers hasconsiderably developed in the past 30 years. Due topossibility of controlling the physical and chemicalproperties of these systems in a systematic way, andin order to solve many practical friction connectedproblems by SAMs, the interest in their studysignificantly increases 1,2. The interest in thegeneral area of self-assembly, specifically inSAMs, comes from their observed importance inscience and technology.One example of the general phenomena of selfassemblyis the formation of monolayers, by selfassemblyof surfactant molecules at surfaces. Selfassembledmonolayers can be prepared usingdifferent types of molecules and differentsubstrates, by very simple process, which makesSAMs manufacturable and thus technologicallyattractive in surface engineering 3. Surfactantadsorption from solution can produce various SAMmorphologies. It has been shown that surfactantscan form monolayers, bilayers 4, or aggregates ofvarious shapes and sizes 5. Commonly studiedfactors during SAMs formation are properties of thesurfactant concentration - above the cmc 6 andbelow the cmc 7 different chain length and headgroup structure 4 or types of solid surfaces (oxide13 th International Conference on Tribology – Serbiatrib’13 177

surfaces, graphite, gold, silica or mica 8. Asignificant part of work has been done at aconcentration above the cmc 6, in order to clarifythe structure of the adsorbed layers, especially theaggregates on the surface. The properties of theadsorbed films on the solid surface seem to bedependent on experimental parameters, such asconcentration, pH, temperature and humidity, rarelydescribed in the literature.Adsorbed surfactants have been characterizedusing various techniques, such as x-rayphotoelectron spectroscopy (XPS) 9, contactangle measurements 10, ellipsometry 11. TheAFM was first used to image surfactant aggregateson the surface by Manne et al. 12. This techniqueindeed offers a convenient method for the imagingof molecular assemblies, i.e. in situ 13 or rarely inex situ. In order to test the stability and measure thethickness of SAMs on mica, a scratch test can beused 14. Due to the small thickness of SAMs(~nm), the development of AFM gives a powerfultool to the visualization of the adsorbed surfactanton solid substrates and facilitates the study of theadsorption mechanism at the molecular level. It canidentify defects on the sample and detect thestructure of formed layers on the substrate. Over thelast 20 years experiments performed with theatomic force microscope or with the othermeasuring techniques, have provided new insightsinto the physics of contact between singleasperities, friction, wear and lubrication on amolecular level.Many authors have aimed at investigating theadsorption of cationic surfactants from aqueoussolution onto a variety of solid surfaces, includinggraphite 12, silica 15 and mica 16. Cationicsurfactants have been the most studied ones, usingthe technique of the direct surface forcesmeasurements, particularly with the surfaces forcesapparatus 17. Most of the surface force work wasdirected towards studying the interaction betweensurfactant coated surfaces, for better understandingof hydrophobic forces 18. The surface forcetechnique has been used to probe the lubricationproperties of aqueous surfactant solutions 19. Dueto the possibility to simultaneous measuring of thethickness and the intermolecular forces ofsurfactant films, this technique has been used as aneffective tool in the study of surfactant modelsystem.Adsorption of cationic surfactants, especiallyquaternary ammonium surfactantscetyltrimethylammonium bromide-CTAB, ontorelatively simple inorganic substrates, such as mica,has been studied very often, as a good modelsystem for boundary lubrication. Depending on theconditions, it has been reported that CTAB adsorbson mica as a compact monolayer 19 , as a stablehydrophobic surface 20, as a bilayer 6, or formsaggregates 21. For the preparation of CTAB selfassembledfilms on mica, numerous adsorptionprotocols have been proposed in the literature. Theapplied procedures include the variation of manyparameters, such as different temperatures in SAMpreparation, the immersion time or the postadsorptionsample treatment 10. Differentconclusions about adsorption theories and theexistence of numerous mechanisms, underline thisbroadness 22. On the other hand, the Kraffttemperature is a very important quantity for CTAB/water solutions as previously reported 6. Due tothe important structural changes in CTAB solution,above the cmc at and above the Krafft temperature,this surfactant transition is still the area of researchinterest.Our attention here will be directed towardmodification of the mica surface by adsorption ofquaternary ammonium surfactants, with the aim toproduce hydrophobic and well-orderedhomogeneous monolayers. Adsorbed layers wereprepared on mica using different concentrations ofsurfactants in aqueous solution, below and abovethe critical micelle concentration, in order to checkthe hypothesis that the adsorption depends on thesolution structure. The adsorbed layers werecharacterized by contact angle (CA) measurementsand atomic force microscopy (AFM) imaging.2. MATERIALS AND METHODSThe self-assembled monolayers of quaternaryammonium surfactants on mica, we made bysingle-tailed hexadecyltrimethylammoniumbromide (CTAB), with the molecular structureCH 3 (CH 2 ) 15 N + (CH 3 ) 3 Br - . CTAB was purchasedfrom Fluka and for further purification wasrecrystallized from an ethanol/acetone mixture. Asa solvent, ultra pure water of resistivity 18.3Mcmwas prepared, using a Barnstead EASYpurebatch-fed water purification system. In the processof sample rinsing we used water of the same qualitybefore drying with a clean nitrogen stream.To avoid any contamination, the glassware andbottles used in the experiments were consistentlycleaned by piranha solution and then rinsed withpurified water. All the employed tools werepreviously cleaned in order to minimize theoccurrence of a molecular contamination,particularly on the high-energy mica surface.178 13 th International Conference on Tribology – Serbiatrib’13

Mica samples preparationFor the adsorption experiments, we usedmuscovite mica purchased from Spruce Pine MicaCompany Inc. (USA). Small mica samples of 1-1.5cm 2 size were cut by scissors and after thatfreshly cleaved on both sides and immersed into thesurfactant solution. The adsorption was performedfrom the surfactant solution in a volume of 20ml.of 18°C or 30°C. It is important to emphasize thatall the used chemicals, tools and substrates wereequilibrated at the defined temperature before theadsorption experiments.3. EXPERIMENTAL PARTSeveral experiments, described in the literature,suggested that the transition through the threephaseboundary is an important step, with asignificant influence on the surfactant filmmorphology. To distinguish effects due toadsorption at the solid-liquid interface and thedeposition at the three-phase boundary (TPB), wehave systematically varied theimmersion/extraction protocol and defined fourdifferent experiments. All four adsorptionprotocols, separately described below, have beenused with/without temperature control. The resultsobserved under the temperature controlledconditions have been described.Every experiment consists of four steps:adsorption, rinsing, drying and analysis (AFM andCA). Varying the immersion and extraction of micasample into and out of the surfactant solution, wehave defined four different adsorption protocols(figure 1).The first option, called “CTAB in/CTAB out“,involves immersion and extraction from thesurfactant solution at the nominal concentration.The second adsorption type (“CTAB in/diluteout”) uses immersion into the nominal surfactantsolution, a rest time and subsequent rapid(10seconds) dilution with pure water prior toextraction, in order to eliminate surfactantdeposition during the transition through the TPB.The third option (“dilute in /dilute out”) preventssurfactant deposition at the TPB in both steps,immersion into solution and extraction from it. Inthis case, CTAB is added to the solution after thesample is submerged.The fourth option (“dilute in/CTAB out”) allowsdeposition at the TPB during extraction only.In the water dipping step, the mica samples weredipped into 20ml of ultra pure water to remove theexcess surfactant molecules.The same set of adsorption experiments shownin figure 1 has been repeated at controlledtemperature in the laboratory, at 18C (below theKrafft temperature) and 30°C (above the Kraffttemperature). A stock solution, 250ml of 10 -3 MCTAB, has been prepared at controlled temperatureFigure 1. Schematic representation of four differentadsorption protocols used to discriminate differentdeposition mechanisms at the three-phase boundary. Allsamples were dipped into ultra pure water to strip offpossible excess CTAB (e.g. incomplete second layer)After the post-rinsing step, the modified micasurface was gently blown dry with nitrogen beforethe AFM imaging or contact angle measurements.Each type of protocol (at different concentrations)was repeated several times to also assess thereproducibility.The samples were imaged with an Atomic ForceMicroscope (Digital Instruments, Nanoscope IIIa),which was operated under ambient conditions. Theimages were systematically collected for differentscan sizes (i.e. 10µmx10µm, 5µmx5µm and1µmx1µm, and again 10µmx10µm) and they wererepeated by scanning several different areas on agiven sample. At least two samples of eachpreparation protocol were analyzed.The contact angle is the angle conventionallymeasured through the liquid, where aliquid/vaporinterface meets a solid surface. It quantifies thewettability of a solid surface by a liquid. On everysample advancing (maximal) contact angle and thereceding (minimal) contact angle are measured. Awater contact angle greater than 90° is determinedon hydrophobic surfaces. For example, freshlycleaved mica has a contact angle less than 10, anda contact angle on the SAM produced by CTABadsorption on mica can be 140.13 th International Conference on Tribology – Serbiatrib’13 179

For water contact angle measurements, after theplacing the droplet on the surface, the advancingcontact angle was measured, a , after which thedroplet was retracted, and the receding contactangle, r , recorded. The times for both contact anglemeasurements, were similar, in the range of a fewseconds duration for each.Using hexadecane as the probing liquid, we candefine a well-ordered monolayer on a substrate andthe expected CA of ordered surfactant layers arearound or larger than 40°. A high contact anglesuggests a high degree of order and “tails-up”molecules orientation in the layer. With hexadecaneonly a static contact angle has been measured byplacing 1µL drop on the sample. Measurementswere performed at room temperature conditions andhumidity. The contact angle values were detectedwith accuracy 1-2.spots, both the results for water(advancing/receding) and hexadecane are givenwith errors.4. RESULTSThe influence of the temperature on thesurfactant solution properties and the resultingSAM morphology, spurred us to control thetemperature during the solution preparation, as wellas during adsorption. Two distinct temperatureshave been chosen-one below (18C) and one above(30°C) the Krafft temperature of CTAB.CTAB adsorption results on mica observed atlow concentrations are illustrated in figure 2.abelow the Krafft temperature (at 18°C), and 2.babove the Krafft temperature (at 30°C). Thetemperature dependence of the observed filmmorphology at 18C and 30°C can be confirmed bycontact angle measurements. Due to the variationsdetermined in the different samples spots, both theresults for water (advancing/receding) are givenwith errors.In both groups of experiments, hydrophobicsurfaces with advancing contact angles between 75°and 90° have been obtained. The significanthysteresis suggests a chemical heterogeneity orroughness in the surfactant film 24.Considering the highest contact angles in bothmeasurements with water and hexadecane, the mostpromising hydrophobic surfaces have been obtainedat 18°C, using the protocol “CTAB in /CTAB out”,suggesting a significant surfactant adsorption at thethree-phase boundary. Therefore, this adsorptionprotocol has been selected for all the followingmeasurements at 18°C.The temperature dependence of the observedfilm morphology at 18C and 30°C can beconfirmed by contact angle measurements. Contactangles for various preparation protocols aresummarized in the tables below. Due to thevariations determined in the different samplesFigure 2. The comparison of AFM images of CTABmodified mica by the “CTAB in/CTAB out” protocol atcontrolled temperature: a) from saturated 10 -3 M solution(at 18°C) b) from 10 -4 M solution (at 30°C)Table 1. Contact angle of water and hexadecane onCTAB coated mica from the saturated 10 -3 M solutionprepared at 18°C18°C hexadecane waterCTAB in/CTAB out 272 85/35CTAB in/dilute out 192° 79/25Table 2. Contact angle of water and hexadecane onCTAB coated mica at 30°C from 10 -4 M solution30°C hexadecane waterCTAB in/CTAB out 172 85/15dilute in/dilute out 14 90/10dilute in/CTAB out 12 79/12CTAB in/dilute out 7 76/125. DISCUSSIONThe structure and the stability of the adsorbedfilms are sensitive to the experimental conditions,primarily temperature.180 13 th International Conference on Tribology – Serbiatrib’13

According to the AFM results in figure 2.a, theadsorbed film can be the homogeneous SAM on themica. Hysteresis in water CA indicates ahydrophobic but not homogeneous film. Similarcontact angles have been reported in the literature.The decay in CA values with time shows instabilityof the formed layer, due to desorption of theadsorbed molecules into the water drop. A similarobservation has been also made 24 and interpretedas water penetration into the monolayer, aphenomenon already known to Langmuir in 1938.The degree of water penetration can also beinfluenced by the local environment, e.g. relativehumidity or temperature 24.In all experiments hydrophobic surfaces withadvancing contact angles between 75° and 90° havebeen obtained. The significant hysteresis suggests achemical heterogeneity or roughness in thesurfactant film. Measurements of hexadecanecontact angle represent one of the most sensitivetools to determine the conformational order ofhydrocarbon thin films. High hexadecane contactangles are observed on hydrocarbon SAMs onlywhen the monolayers are densely packed. Usinghexadecane we have observed lower values than theexpected 40° for ideally ordered films.In order to relate our SAM preparation protocolsto previous work, we have compared our resultswith those described in the literature. Much workhas been devoted to the study of the adsorption ofCTAB on mica using a variety of adsorptionprotocols and measurement techniques. For bothsolution concentrations, below and above the cmc,the picture drown in the literature demonstrates ahigh degree of variability.The protocol proposed by Zhao 25 wasreported to produce cylinder aggregates on the micasurface at a solution concentration of 2cmc (figure3.a). Due to the measured height of ~6nm of thefeatures, the layer has been described as consistingof two bilayers.At the CTAB concentration 10 -5 M monolayersand bilayers have been observed 22, using asimilar adsorption protocol at room temperature(around 25°C), as shown in figure 3.b. The AFMwas operated in the surfactant solution.We can conclude that the literature confirms thateven small differences in the SAM adsorptionprotocol can significantly affect the surfactant filmmorphology on mica, especially above the cmc andthe Krafft temperature. These observations areconfirmed by our experimental results and there isthe need to explain these variations by structuralchanges of the surfactant solution 26. There aresignificant structural changes around the Kraffttemperature. Below the bulk cmc and below theKrafft temperature, an equilibrated solution isexpected to be free of micelles. At the Kraffttemperature, the solubility becomes equal to thecmc and micelles will form in the solution and thistemperature is very often described in the literature.But, reported values of T k for CTAB in water varyconsiderably ( from 20°C to 25°C), which is veryclose to a room temperature and complicates theexplanation of experimental results.Figure 3. a) AFM topographic image of a mica surfaceprepared unspecified temperature conditions, using“CTAB in/CTAB out” protocol (immersion time-1min),at 2cmc, without rinsing after removal from the solution,dried with nitrogen before AFM imaging 25; b) AFMimage of a CTAB adsorbed layer on mica in a 10 -5 Msolution at 25°C for 25min immersion time, observed inthe surfactant solution 22.It is important to note that most of theexperiments described in the literature do notmention the problem of temperature control, and itis not possible to reconstruct this importantparameter from the information provided. We havea clear evidence that temperature is the key factordetermining SAM morphology. Therefore, we havesystematically studied temperature effects byrigorously controlling the temperature during allprocedures.6. CONCLUSIONThe concept of self-assembly on surfaces hasbeen treated in this research and the experimentshave revisited the adsorption of CTAB onto amuscovite mica. These results demonstrate theinfluence of a large number of parameters on theadsorption process, as well as the morphology and13 th International Conference on Tribology – Serbiatrib’13 181

the molecular order of SAMs. The molecularstructure of the solution seems to be a key element.It is chiefly controlled by the temperature andconcentration of the solution. The fact that theKrafft temperature range of CTAB is around roomtemperature (25C), makes this systemparticularly complex.We have systematically studied the transferacross the three-phase boundary, under differentexperimental conditions and different adsorptionprotocols. The surfactant films on mica, formedaccording to different experimental protocols, werecharacterized by contact angle measurements andby AFM. It has been observed that the transitionthrough the three-phase boundary during sampleextraction can be a step of significant influence onthe SAM morphology.A high stability of the adsorbed films is veryrarely detected. The problem of reproducibility inSAMs formation can be observed by controlling thetemperature during all steps, or, working below thecmc. A reproducible stability of the resulting films,however, remains an issue.Acknowledgements: The author would like togratefully acknowledge the support in obtainingand understanding the presented results provided byher mentors, Nicholas D. Spencer and ManfredHeuberger, members of the Laboratory for SurfaceScience and Technology (LSST), a part of theDepartment of Materials at the ETH Zurich,Switzerland, where all the experiments wereperformed.REFERENCES[1] M. K. Chaudhury: Adhesion and friction of selfassembledorganic monolayers, Current opinion incolloid & interface science, Vol.2, pp. 65-69, 1997.[2] S. Manne et al.: Direct Visualization of SurfactantHemimicelles by Force Microscopy of the ElectricalDouble-Layer, Langmuir, Vol.10(12), pp. 4409-4413, 1994.[3] A. Ulman: Formation and Structure of Self-Assembled Monolayers, Chem. Rev., Vol. 96, pp.1533-1554, 1996.[4] H. Patrick et al.: Surface micellization patterns ofquaternary ammonium surfactants on mica,Langmuir, Vol. 15(5), pp. 1685-1692, 1999.[5] S. Manne, H. Gaub: Molecular-Organization ofSurfactants at Solid-Liquid Interfaces, Science, Vol.270(5241), pp. 1480-1482, 1995.[6] W. Ducker, E.Wanless: Adsorption ofhexadecyltrimethylammonium bromide to mica:Nanometer-scale study of binding-site competitioneffects, Langmuir, Vol. 15(1), pp. 160-168, 1999.[7] M. Fujii et al.: Heterogeneous growth and selfrepairingprocesses of two-dimensional molecularaggregates of adsorbedoctadecyltrimethylammonium bromide at cleavedmica aqueous solution interface as observed by insitu atomic force microscopy, Langmuir, Vol.15(10), pp. 3689-3692, 1999.[8] R. Atkin et al.: Adsorption kinetics and structuralarrangements of cationic surfactants on silicasurfaces, Langmuir, Vol. 16(24), pp. 9374-9380,2000.[9] K. Boschkova et al.: Lubrication in aqueoussolutions using cationic surfactants - a study ofstatic and dynamic forces, Langmuir, Vol. 18(5),pp. 1680-1687, 2002.[10] B.Y.Li et al.: Time dependent anchoring ofadsorbed cationic surfactant molecules atmice/solution interface, Journal of Colloid andInterface Science, Vol. 209(1), pp. 25-30, 1999.[11] K. Eskilsson and V.V. Yaminsky: Deposition ofmonolayers by retraction from solution:Ellipsometric study of cetyltrimethylammoniumbromide adsorption at silica-air and silica-waterinterfaces, Langmuir, Vol. 14(9), pp. 2444-2450,1998.[12] S. Manne et al.: Direct Visualization of SurfactantHemimicelles by Force Microscopy of the ElectricalDouble-Layer, Langmuir, Vol. 10(12), pp. 4409-4413, 1994.[13] J. a. M. Mellott: Supercritical self-assembledmonolayer growth, Journal of the AmericanChemical Society, Vol. 126(30), pp. 9369-9373,2004.[14] HY Nie et al.: Robust self-assembledoctadecylphosphonic acid monolayers on a micasubstrate, Langmuir, Vol. 21(7), pp. 2773-2778,2005.[15] T. Horr et al.: XPS film thickness and adsorptionstudies of alkyltrimethylammonium bromides andorganosilanes on silica surfaces, Colloids andSurfaces A: Physicochemical and EngineeringAspects, Vol. 102, pp. 181-190, 1995.[16] S. Nishimura et al.: AFM Studies of AmineSurfactant Hemimicelle Structures at the Mica-Water Interface, Colloids and Surfaces A:Physicochemical and Engineering Aspects, Vol.103(3), pp. 289-298, 1995.[17] M. Rutland, J. Parker: Surface Forces betweenSilica Surfaces in Cationic Surfactant Solutions -Adsorption and Bilayer Formation at Normal andHigh pH, Langmuir, Vol.10(4), pp. 1110-1121,1994.[18] A. e. a. Rennie: Structure of a Cationic SurfactantLayer at the Silica Water Interface, Langmuir, Vol.6(5), pp. 1031-1034, 1990.[19] R. Pashley, J. Israelachvili: A Comparison ofSurface Forces and Interfacial Properties of Mica inPurified Surfactant Solutions, Colloids andSurfaces, Vol. 2(2), pp. 169-187, 1981.[20] M. Fujii et al.: Two-dimensional arrangements ofadsorbed alkylammonium halides on cleaved micasurface, Langmuir, Vol. 17(4), pp. 1138-1142, 2001.182 13 th International Conference on Tribology – Serbiatrib’13

[21] B. G. Sharma et al.: Characterization of adsorbedionic surfactants on a mica substrate, Langmuir,Vol. 12(26), pp. 6506-6512, 1996.[22] G. Ceotto et al.: Ionic surfactant films imaged byatomic force microscopy, Journal of MolecularCatalysis A: Chemical, Vol 167(1-2), pp. 225-233,2001.[23] T. Davey et al.: Krafft temperature depression inquaternary ammonium bromide surfactants,Langmuir, Vol. 14(12), pp. 3210-3213, 1998.[24] Y. L. Chen et al.: Molecular MechanismsAssociated with Adhesion and Contact-AngleHysteresis of Monolayer Surfaces, Journal ofPhysical Chemistry, Vol. 95(26), pp. 10736-10747,1991.[25] F.Zhao et al.: Adsorption behavior ofhexadecyltrimethylammonium bromida (CTAB) tomica substrates as observed by atomic forcemicroscopy, Science in China Ser. B Chemistry,Vol. 48(2), pp. 101-106, 2005.[26] Manojlovic J.: Structure, morphology andhistory effects in surfactant self-assembly,Ph.D. thesis, ETH Zurich, Switzerland , 2006.13 th International Conference on Tribology – Serbiatrib’13 183

Serbian TribologySocietySERBIATRIB ‘1313 th International Conference onTribologyKragujevac, Serbia, 15 – 17 May 2013Faculty of Engineeringin KragujevacMO-C MULTILAYERED CVD COATINGSA. Sagalovych, V. SagalovychScientific technological Centre “Nanotechnologie”, Kharkov, UkraineAbstract: Production processes of multi-layered Mo-C coatings by the method of chemical vapor deposition(CVD) with the use of organometallic compounds were developed. Coatings are applied on technicalpurpose steel DIN 1.2379 (Х12Ф1) and DIN 1.7709 (25Х2МФ (ЭИ10) heat-treated ball with the high classof surface roughness (> 10). The average deposition rate was 50 μm/h. The optimal conditions of coatingsdeposition for different technological schemas were defined.Metallographic investigations of the obtained coatings were carried out.Tribological studies of the friction and wear characteristics of sliding friction in conditions of boundarylubrication of Μo-С multilayered CVD coatings shows, that coatings have low friction coefficients (0.075 -0.095) at loads up to 2.0 kN, showed high resistance to wear and are effective in increasing the stability ofthe pair for precision friction pairs of hydraulical units.Key words: CVD processes, multi-layered coatings, tribology.1. INTRODUCTIONEver-increasing requirements to raise durabilityand efficiency of various newly created parts andmechanisms working in friction conditions lead tosharp increase in demands to search new materialsworking in the conditions of frictional contact andtheir tribotechnical characteristics. Complexity andintegrated nature of these demands boost constantsearch of new materials and techniques of theirproduction. Especially acute problem of makingnew highly effective wear resistant materials is foradvanced industries of machine building – aircraftengineering, aggregate building, shipbuilding,rocket and space engineering.Raise of the functional properties of parts by useof protective coatings is very widespreadtechnological tool for today. Hardsurfacing, wearresistant, corrosion resistant, etc. functionalcoatings are widely applied in the industry.Among coating deposition methods are chemicalvapor deposition methods (CVD) based on thepyrolysis of gaseous metal containing compounds [1].The relative simplicity of realization of processes,absence of high demands to vacuum (in many casesprocesses take place at atmospheric pressure), highcoating deposition rates (up to a few millimeters perhour), and possibility of deposition of the evenqualitative coatings on figurine-shaped (includinginternal) surfaces with great value of relationship L/d,make these methods rather perspective for depositionof the functional coatings [1-3].Wei and Lo [4] carried out examination ofdeposition of Mo and Cr coatings at temperatures170-450 °С and pressure ~ 1 torr using a mixture ofcarbonyl with hydrogen. Upon that, they usedspecially prepared carriers made of SiC and SS304.The main emphasis in the work is placed onstudying of structure of obtained coatings.In the work [5] molybdenum was deposited on aporous ceramic carrier at atmospheric pressure.Chromium was deposited from carbonyl atatmospheric pressure also, but already on thepolished carriers in the work [6].It should be pointed out that application of CVDmethods to receive hardsurfacing wear- andcorrosion resistant coatings was limited yet. Fromthe point of view of practical application,development of processes of obtaining of chemicalvapor deposited functional coatings ongeometrically complex precision surfaces of highsurface finish class (above 10 grade) is of interest.The purpose of this study is development of processof deposition of multilayered Мо-С coatings by a184 13 th International Conference on Tribology – Serbiatrib’13

chemical vapor deposition method with useorganometallic compounds, studying multilayeredand tribological properties of the obtained coatingsand an estimate of possibility of their application inthe capacity of candidate materials for precisionfriction pairs.2. EQUIPMENT, TECHNIQUES, AND USEDPROCEDURESMaking Mo-С coatings was carried out by athermal decomposition of metal-containingcompound – molybdenum hexacarbonyl Mo (CO) 6 .Process development was carried out using gasphaseunit of Avinit installation intended fordeposition of the multilayered functional coatingsby means of complex methods (plasma-chemicalCVD, vacuum-plasma PVD (vacuum-arc,magnetron), and processes of ion saturation, andtreatment of surfaces by ions) united in onetechnological cycle), presented in [7].Process of coating application was controlled bymeans of temperature of the sample, workingpressure in the chamber, and a method ofevaporation of molybdenum hexacarbonyl.For heating of the sample, high-frequency heaterwas applied with working frequency of 3 MHz andeffective power ~ 0.2 kW. Pressure was controlled bydynamical changing of velocity of a pumping-out ofvacuum system within 2.6 … 13 Pa (0.02 … 0.1 torr).Carbonyl evaporation was carried out under twoprocess flow charts:• The process flow chart of excessiveevaporation when a considerable quantity ofcarbonyl was evaporated within a containervolume. Then, along the heated up steampipeline through the adjustable valve, the gasmixture immediately supplied to the sample.It allowed obtaining major streams ofcarbonyl and, accordingly, highconcentration of molybdenum in a reactionlayer. In this case the carbonyl streamdepended on temperature of the carbonyl.• Under other process flow chart (the residualatmosphere) evaporation of a small amountof carbonyl was made immediately insidechamber volume. Thus, the uniformconcentration of the reaction gas consistingof carbonyl and carbonic oxide vapors wasobtained. At that, molybdenum volumeconcentration in a chamber atmospheredecreased with the course of process.3. PROCEDURES OF EXAMINATIONDuring metallographic analysis the multilayeredand microlayer coatings on the basis of system Mo-C were deposited on samples.Samples were made of structural steelDIN 1.2379 (Х12Ф1) and heat steel DIN 1.7709(25Х2МФ (ЭИ10) which are the materialscommonly used in the industry.Samples made of steel DIN 1.7709 (25Х2МФ(ЭИ10) with a size of 20х10х5 mm were polishedaccording to factory production method up to 8grade surface roughness (R a =0.32 μm).Microhardness was НВ~900.Samples made of steel DIN 1.2379 (Х12Ф1),56... 61HRC, with a size of 10х10х10 mm werepolished to surface roughness of 10 grade(Ra=0.063 μm) to demanded geometricalparameters (nonflatness ≤ 0.001мм, surfaceroughness – Ra 0.08 μm).Tribological tests of antifriction and wearproperties and seizure of samples with coatingswere carried out with friction and wear machine2070 SMT-1 under the "cube" - "roller" test patternat an incremental loading (with increments of200N) in 1-20 MPa loading range according to theprocedures presented in [8, 9]. The linear slipvelocities - 1.3m/s. Time of tests in each cycle –150 seconds. Operating fluid is fuel ТS-1, GOST10227-86.For reproducibility of results of wearing tests,mating of face surfaces by size of the contact areawas controlled: not less than 90 % of a workingarea of each sample.During tribological tests there were registered:– Values of frictional force F tr , normal loadingN, contact pressure P, by which value mechanicallosses in tribosystems were estimated;– Temperature of devices was continuouslyrecorded in real time during the tests in immediateproximity (1 mm) from a friction zone, withapplication of the sliding thermocouple. Frictioncoefficients were determined as f = F tr /N.4. RESULTS4.1 Deposition of coatingsTechnological information for process ofcoating deposition on steel DIN 1.7709 (25Х2МФ(ЭИ10) is presented in Tables 1.1, 1.2 and 2.13 th International Conference on Tribology – Serbiatrib’13 185

Table 1.1 Mo/Mo-C coatings obtained at puffing of carbonyl from the container.Temperature, °С Pressure, Pa Exposition τ, min. Thickness δ, μm Deposition rate V, μm/min. Adhesion5.20 10 8 0.80 +++10.00 10 7 0.70 +++3505.30 15 17 1.13 +++11.00 15 10 0.67 +++8.80 30 25 0.83 +++11.00 30 31 1.03 ++5.60 5 3 0.60 ++5.40 10 8 0.80 +4005.50 10 6 0.60 +6.10 10 8 0.80 +5.00 15 10 0.67 -5.40 15 12 0.80 -7.60 5 6 1.20 +++4505.30 15 17 1.13 +++7.20 15 20 1.33 +++5.10 30 12 0.40 ++++++ - the coating is deleted only by pickling++ - insignificant chipping+ - a lot of chippingTable 1.2 Coatings Mo/Mo-C obtained in an atmosphere of the residual gases.Temperature, °С Pressure, Pa Exposition τ, min. Thickness δ, μm Deposition rate V, μm/min. Adhesion3.90 15 4 0.27 +++35010.0 15 4 0.27 ++12.0 30 10 0.33 +11.0 60 16 0.27 -2.80 15 0.5 0.03 -4003.20 30 1 0.03 -3.80 60 5 0.08 -4504.00 15 2 0.13 ++30 3.70 4 0.13 +Table 2. Parameters of CVD process of coating application.Sample No. Sample temperature, ° C Exposition, min. Microhardness H v , kg / μm 2 Thickness δ, μm23 430 10 2500 1525 360 10 1800 1025 290 10 2500 10Properties of coatings sharply differ in the settemperature interval. At 350°С and 450°С stableuniform deposition of coating with high value ofmicrohardness is observed: НВ=2200 at 350°С,НВ=1700 at 450°С. The thickness of a coating dependslinearly on cure time. The lower temperature, the loweradhesion to the initial sample is observed, while at450°С the coating has the good adhesion with a carrier.Deposition rate of the coatings obtained in anatmosphere of the residual vapors at temperature 350°Сis even during long enough time intervals; however,quality of a coating deteriorates in process ofmagnification of coating thickness that is obviouslyrelated with accumulation of sufficient internal stressesin a film.Parameters of CVD-process of coating applicationon samples made of steel DIN 1.2379 (Х12Ф1) areshown in table 2. After deposition of CVD coatingsamples were polished with diamond paste АSМ7/3with removal of 0.004-0.006 mm stock up to recoveryof flat surface accuracy.Change of working pressure within 0.01 … 0.1 torraffects Mo/Mo-C composition and leads toinappreciable decrease in adhesion which becomesmore appreciable at magnification of a thickness of acoating.4.2 Metallographic examinationMetallophysics measurements of the obtainedsamples are carried out on raster-type electronicmicroscope JSM T-300. Appearance of Mo-Ccoating on samples made of steel DIN 1.2379(Х12Ф1) (a traversal metallographic sample) withmarked zones of analysis (a) and an approximatechemical composition of analyzed zones (b) isshown in Fig. 1.186 13 th International Conference on Tribology – Serbiatrib’13

analysis (a) and an approximate chemical compositionof analyzed zones (b) is shown in Fig. 3.a) DIN 1.2379 (Х12Ф1)Point No. Si Cr Fe Ni Mo Ρ003 97.0 3.0004 96.5 3.5005 2.21 94.79 3.0006 3.17 94.83 2.0007 10.13 22.67 9.90 55.31 2.0008 0.21 6.85 92.13 0.8009 0.34 7.13 91.73 0.8b) Chemical composition of analyzed zones.Fig. 1. Appearance of Mo-C coating on samples made ofsteel DIN 1.2379 (Х12Ф1) with marked zones ofanalysis (a), and the chemical composition of analyzedzones (b).Appearance of Mo/Mo-C coating on the samplemade of steel DIN 1.2379 (Х12Ф1) in a mappingmode is shown in Fig. 2.Appearance of Mo-C coating on samples madeof steel DIN 1.7709 (25Х2МФ (ЭИ10)) (a traversalmetallographic sample) with marked zones of thea) DIN 1.7709 (25X2 МФ (ЭИ 10)Point No. Si Cr Fe Ni Mo С021 3.40 94.12 2.48022 3.32 94.01 2.67023 3.05 95.61 1.34024 3.40 94.80 1.80025 3.18 94.64 2.18026 2.72 1.98 93.69 1.57027 0.17 1.93 97.69 0.22028 0.25 1.88 97.55 0.32029 0.27 1.67 97.86 0.19b) Chemical composition of analyzed zones.Fig. 3. Appearance of Mo-C coating on samples made ofsteel DIN 1.7709 (25Х2МФ (ЭИ10)) with marked zonesof the analysis (a) and the chemical composition ofanalyzed zones (b).Fig. 2. Appearance of Mo/Mo-C coating on the sample made of steel DIN 1.2379 (Х12Ф1) in a mapping mode. Morecontent of the element there matches to more saturated color.13 th International Conference on Tribology – Serbiatrib’13 187

Metallophysics measurements have shown highenough extent of coincidence of phase compositionof a carrier material - steels DIN 1.2379 (Х12Ф1)and DIN 1.7709 (25Х2МФ (ЭИ10)) (zones 009and 029, accordingly).Photos of a microrelief of coating surfaces areshown in Figs. 4 and 5.Fig. 6. Dependence of friction coefficient on loads for afriction pair Mo-C/steel DIN 1.4021 20Х3МВФ (ЭИ 415).Fig. 4. A microrelief of a surface of the sample (steelDIN 1.7709 (25Х2МФ (ЭИ10)).Fig. 7. Dependence of friction coefficient on a loadingfor a friction pair Mo-C / Avinit coating (on the basis ofMo-N).Fig. 5. A microrelief of a surface of the sample (steelDIN 1.2379 (Х12Ф1)) after its polishing.4.3 Results of tribological testsMultilayered and microlayer coatings on thebasis of Mo-C system are deposited on basicsamples - cubes made of steel DIN 1.2379 (Х12Ф1)with hardness 56 … 61HRC (НВ~900) with theworking planes polished by diamond paste to reachrequired geometrical parameters (nonflatness - ≤0,001mm, surface roughness - R a 0,08 μm) forcarrying out of tribological tests. Parameters ofCVD-process of coating application on examinedsamples and properties of the obtained CVDcoatingsare shown in tab. 2. After deposition ofCVD coatings samples were polished by diamondpaste АSМ7/3 with with removal of 0.004-0.006mm stock up to recovery of flat surface accuracy.Results of tribological tests are presented inFigs. 6-8 and in Tab. 3.Fig. 8. Dependence of friction coefficient on a loadingfor a pair of a friction Mo-C / Avinit coating (on thebasis of Ti-Al-N).When carrying out tribological tests, the specialpriority has been given to studying of behaviour of thedeveloped coatings in tribological matings with steels.As is clear from the obtained data, and also byresults of carried out earlier tribological studying[10, 11], the best tribological parameters in contactpair with steel are determined for Mo-C coatings.Good tribological properties are exhibited byMo-C coatings in case of friction in pairs with hardand extremely hard coatings which have beenalready used in friction pairs of precision units inaircraft aggregate building [10-12].188 13 th International Conference on Tribology – Serbiatrib’13

Table 3. Friction coefficients of samples during tribological tests.Roller (coating)Avinit coating (on the basis of Mo-N)δ =1.5 μm, 2200HVAvinit coating (on the basis of Ti-Al-N)δ =1.5 μm, 3500HVThe sameRun in separatelyThe sameRun in separatelyDIN 1.402120Х3МВФ (ЭИ 415),cementation, 88HRCDIN 1.402120Х3МВФ (ЭИ 415),run in separatelyCubeApplied loading, kN(coating) 0.2 0.4 0.6 0.8 1 1.2 1.4Mo-C (23) 0.14 0.13 0.127 0.13 0.124 0.12 0.12Mo-C (24) 0.15 0.16 0.16 0.157 0.152 0.127 0.107Mo-C (24) 0.13 0.11 0.133 0.122 0.108 0.105 0.101Mo-C (25) 0.14 0.15 0.153 0.148 0.146 0.145 0.134Mo-C (23)scoringMo-C (24) 0.18 0.17 0.193 0.175 0.16 0.15Mo-C (25)scoringMo-C (25) 0.16 0.18 0.167 0.157 0.152 0.157 0.153Mo-C (23) 0.12 0.12 0.12 0.125 0.126 0.125 0.12Mo-C (23) 0.14 0.14 0.143 0.135 0.128 0.123 0.12Mo-C (24) 0.14 0.14 0.137 0.137Mo-C (25) 0.11 0.125 0.127 0.122 0.132 0.127 0.12MoC (24) 0.14 0.14 0.133 0.13 0.126 0.123 0.123Table 4. Estimate of aging traces on samples after tribological tests.Мо-С cube (No. 23) δ ≈ 15 μm, 2500HV; After a lapped finishing with diamond paste δ ≈ 10 μm.Ή Roller (coating)Test resultsDIN 1.4021The cube has aging traces with parameters (determined using the profilogram): -1 20Х3МВФ (ЭИ 415),cementation, ≥88HRCdepth - ≈0.4 μm; - width - 0.6 mm.The roller has a normal aging trace; signs of wear are visually absent.233 aAvinit coating (based on Mo-N)δ =1.5 μm, 2200HVAvinit coating (based on Ti-Al-N)δ =1.5 μm, 3500HVAvinit coating (on the basis ofTi-Al-N)δ =1.5 μm, 3500HVRun in separatelyThe cube has aging traces with parameters (determined using the profilogram): -depth - ≈0.5 μm; - width - 0.8 mm.The roller has a normal aging trace, signs of wear are visually absentThe cube has 2 seizure sites which are placed near to ribs of a cube, the mainaging trace has following parameters (determined using the profilogram):- Depth - ≈3.4 μm; - width - 1 mm.The roller has two ring furrows which are reciprocal to the seizure sites on thecubeThe cube has aging traces with parameters (determined using the profilogram): -depth - ≈1.2 μm; - width - 0.8 mm.The roller has a normal aging trace; signs of wear are visually absent.Mo-C Cube (No. 24) δ ≈ 10мкм, 1800HV; after a lapped finishing diamond paste δ ≈ 5 μm.No. Roller (coating)Test resultsDIN 1.4021The cube has large fretting in width of ≈7 mm4 20Х3МВФ (ЭИ 415),The roller has ring traces of cube material transportcementation, ≥88HRC55а66аAvinit coating (based on Mo-N)δ =1.5 μm, 2200HVAvinit coating (based on Ti-Al-N)δ =1.5 μm, 3500HVDIN 1.402120Х3МВФ (ЭИ 415),cementation, ≥88HRC,run in separatelyThe cube has aging tracesThe roller has a normal aging trace; signs of wear are visually absent.The cube has aging traces with parameters (determined using the profilogram): -depth - ≈18 μm; - width - 1.8 mm.The roller has a normal aging trace; signs of wear are visually absent.The cube has aging traces with parameters (determined using the profilogram): -depth - ≈19μm; - width - 1.9mm.The roller has a normal aging trace; signs of wear are visually absent.The cube has aging traces with parameters (determined using the profilogram): -depth - 0.4 μm; - width - 0.5 mm.The roller has a normal aging trace; signs of wear are visually absent.Mo-C Cube (No. 25) δ ≈ 10μm, 2000 … 2500HV; after a lapped finishing diamond paste δ ≈ 5 μm.No. Roller (coating)Test resultsDIN 1.4021The cube has aging traces with parameters (determined using the profilogram): -7 20Х3МВФ (ЭИ 415),cementation, ≥88HRCdepth - ≈0.3 μm; - width - ≈0.5 mm.The roller has a normal aging trace; signs of wear are visually absent.13 th International Conference on Tribology – Serbiatrib’13 189

899аAvinit coating (based on Mo-N)δ =1.5 μm, 2200HVAvinit coating (based on Ti-Al-N)δ =1.5 μm, 3500HVAvinit coating (based on Ti-Al-N)δ =1.5 μm, 3500HV run inseparatelyThe coating on a cube inside the trace is visually worn to the carrier.The cube has aging traces with parameters (determined using the profilogram): -depth - ≈4.4 μm; - width - 1 mm.The roller has a normal aging trace; signs of wear are visually absent.The cube has 2 seizure sites which are placed near to ribs of a cube; flaws on acoating were formed on the same plane of a cube.The roller has two ring furrows which are reciprocal to the seizure sites on thecube.Exfoliation of the coating on the given plane of a cube, along aging traces fromboth its sides, is available, parameters of aging traces (determined using theprofilogram): - depth - 19 μm; - width - 2 mm.The roller has a normal aging trace; signs of wear are visually absent.When friction of cubes with Mo-C coatings overrollers with Avinit coating (on the basis of Mo-N),low enough antifrictional parameters are alsoobserved.Rollers with extremely hard Avinit coatings (onthe basis of Ti-Al-N) exhibit higher frictioncoefficients, and there are cases of a scoring of Mo-C coating.Results of an estimate of aging traces in testedfriction pairs after tribological tests are shown inTable 4.It is noted, that in case of use of already run-inrollers, minimum friction coefficients are obtained,thus cases of a scoring of Mo-C coating are notavailable.Carried out tribological tests of Мо-С coatingstestify to efficiency of the developed coatings forprecision friction pairs («steel/coating» and«coating/coating») with the raised wear hardnessand low friction coefficient.5. CONCLUSIONS1. Process of application of multilayered Мо-Сcoatings by a chemical vapor depositionmethod with use of organometalliccompounds is developed. The multilayeredcomposite coatings on the basis of Mo-Csystem are obtained. Optimization ofprocesses of deposition of qualitative tightlyinterconnected coatings is carried out.2. The kinetics of coating deposition process isstudied. Coating deposition rate up to 100μm/hour is obtained.3. Metallographic examination confirmpossibility of the low-temperature depositionof qualitative very hard Mo-C coatings indeveloped CVD process, good adhesion tocarrier materials (to steels DIN 1.2379(Х12Ф1), DIN 1.7709 (25Х2МФ (ЭИ10))without decrease in strength properties of a steeland without a deterioration of an initial surfacefinish class is thus provided.4. Multilayered and nanolayer coatings onsamples for tribological tests are obtainedand tribological tests of samples with coatingsare carried out.5. Possibility of postoperative machining ofcoatings by industrial methods without lossesof the functional properties of coatings isproven.6. Tribological tests exhibit high tribologicalproperties of Мо-С coatings and testify toperspectivity of the developed coatings forselection of optimum constructions ofcoatings for raise of stability of precisionfriction pairs.REFERENCES[1] Ivanov V. E.: Crystallization of refractory metalsfrom vapor phase / [V.E.Ivanov,E.P.Nechiporenko, V.M.Krivoruchko, V.V.Sagalovich]. //M.—Atomizdat.—1974.[2] Sagalovich А.V. Coating the surface ofthe figurine-precision gas-phase method (CVD) /[А.V. Sagalovich, V.V. Sagalovich ea.]. //Physical surface engineering.—2011.— v. 9.—№3.— P. 229-236.[3] Powell K.Chemical vapor deposition / [K.Powell,J. Oxley, J. Blocher Jr.]. //Translated from theEnglish//M.—Atomizdat.—1970.[4] Wen-Cheng J. Wei. Processing and Properties of(Mo, Cr) Oxycarbides from MOCVD / [Wen-Cheng J. Wei, Ming-Hung Lo]. //Appl.Organometallic .—1998.—v. 12.—P. 201–220.[5] Skudin V.V. Obtaining of membranes using amethod of chemical vapor deposition in a reactorwith "cold" walls / [Skudin V. V, S. G. Streltsov].//Crit. Technologies. Membrany. .—2007.—№ 2(34).— P. 22-33.[6] Douard A. Reactivity of Cr (CO) 6 in atmosphericpressure CVD processes for the growth of variousmetallurgical coating [Douard A, Maury F.].//Rev. Adv. Mater. Sci.—2007.—v. 15.[7] .Sagalovich А.V. Avinit installation fordeposition of multilayered functional coatings[O.V.Sagalovich, O.V.Kononykhin, V.V. Popovea]. // Physical surface engineering. .—2010.—v.8.—P. 336-347.[8] Lyubchenko A.P.. Examination of wear ofvacuum-plasma coatings made of TiN at afriction on metal materials / [A.P. Lyubchenko,190 13 th International Conference on Tribology – Serbiatrib’13

A.K.Oleynik, V.M.Matsevity, ea.]. //Friction andwear.—1981.—No. 6.—P. 29-31.[9] Dudnik S.F.. Examination of friction and wearproperties of the ionic-plasma coatings oproduced on an aluminium alloy / [S.F.Dudnik,A.V.Sagalovich, V.V. Sagalovich, ea.]. //Physical surface engineering.—2004.—v. 2.—P.110-114.[10] .Sagalovich A.V. Examination of tribologicalproperties of multicomponent multilayeredcoatings of Avinit type / [A.V.Sagalovich, V.V.Sagalovich, A.V.Kononykhin, etc.]. //Bulletin ofKhNARU.—2011.—v. 53.-P. 87-94.[11] Sagalovych A. The Tribological Investigation ofMulticomponent Multilayered Ion-plasmaCoatings Avinit / [A. Sagalovych V. Sagalovych,A. Kononyhin ea.]. //Tribology in industry.—2011.—v. 33.—No. 2.-P. 79-86.[12] Sagalovych A. Experimental research ofmulticomponent multilayer ion-plasma Avinitcoatings / [A. Sagalovych V. Sagalovych, A.Kononyhin ea.]. // Physical surface engineering. .—2012.—v. 10.—P. 136-14813 th International Conference on Tribology – Serbiatrib’13 191

Tribology of MachineElements13 th International Conference on Tribology – SERBIATRIB ’1315 – 17 May 2013, Kragujevac, Serbia

Serbian TribologySocietySERBIATRIB ‘1313 th International Conference onTribologyKragujevac, Serbia, 15 – 17 May 2013Faculty of Engineeringin KragujevacEQUILIBRIUM STATE FORMATION FEATURES OFSURFACE LAYERS OF MACHINE PARTSVyacheslav F. Bezjazychnyj, Alexander N. SutyaginFederal State-Financed Educational Institution of High Professional Education “P. A. Solovyov Rybinsk State AviationTechnical University”, Rybinsk, Yaroslavl region, Russian Federation,e-mail: technology@rgata.ruAbstract: The article describes the problem of forming the equilibrium state of surface layer of friction units.The possibility of taking into account the parameters of the lubricant in the mathematical model of therelationship of the wear rate with the equilibrium parameters of surface layer of parts is considered.Possibility of technological support equilibrium parameters of surface layer of machine parts is discussed.Keywords: friction, wear rate, surface layer, machining, quality parameters, equilibrium state.1. INTRODUCTIONPractice of exploitation of machines andmechanisms shows that one of the major causes offailures of various kinds of equipment is thedestruction of its parts or wear. Destruction andwear usually begins with working surfaces, so thestate of the last often becomes the dominant factorin determining the reliability and durability ofmanufactured equipment.Study the wear of machine parts friction unitsindicate that the transition to the normal process ofwear most of them are directed to change theirinitial geometry.During the break-in such quality parameters ofmachine parts undergo significant changes asprofile deviation, waviness, roughness of theirsurfaces, as well as the physical and mechanicalcharacteristics of the surface layer. If time and thewear are large enough, the stable steady state isreached tribosystem when the geometry of partssurfaces becomes conformal.2. THE STUDIED PHYSICAL MODELResearch the wear can provide a theoreticaljustification of the possibility of existence of stablereproducing geometric shapes of surfaces and theirphysical and mechanical properties in wear processin the given conditions of relative motion andloading.This rationale is based on the fundamentalprinciples of thermodynamics of dissipativesystems, which tribosystem are.Such systems can exchange with other systemsand the environment energy and mass (eg, massworn out particles), so stationary statescharacterized by a constant gradient of entropy canoccurΔS = S° – S = const, (1)where S°– entropy of the equilibrium state of thesystem; S – instantaneous entropy of the system.In the steady state, all the processes of heat andmass transfer are independent of time, asdetermined by the configuration of the system onlyin general terms and conditions on its borders.Tribosystem exhibit properties of selforganizationconsist of to consistently reproducethe macroscopic space-time structures that can existonly through the exchange of flows of energy(matter).Under self-limitation created by nature, astribosystem located on the border of artificialdevices and natural systems, and the selforganizationcan arise from a chaotic state, that is,the initial conditions do not matter, and therunning-in period, of course, increases anddecreases life of the friction units.The equilibrium state of the surface isdetermined by the minimum of its free energy,which may differ from the minimum of surface,13 th International Conference on Tribology – Serbiatrib’13 195

which is responsible for one of the reasons for thedeviation of the real surface of the crystal from theinitially smooth and the emergence of so-callednatural roughness [3]. Such modifications ofsurface geometry are sub roughness, whichsignificantly exceeds the atomic scale roughnessarising at vibrations of the atoms or molecules dueto thermal fluctuations.The most appropriate criterion for selectingmaterials for the parts of tribosystems can be takenminimum value of frictional work [4].Friction work calculated by the formula [5]Wfr fFS , (2)frwhere F – normal force of friction pair elementsinteraction; f – a friction coefficient; S fr – frictiontrack.Tribosystem are open thermodynamic systemsthat exchange energy and matter with theenvironment. Friction is the process of convertingthe external mechanical energy into internal energyin the form of vibration and wave motion ofparticles of tribosystem followed by thermal,thermionic, acoustic, and other phenomena. Most ofthis energy is converted into heat and is given to theenvironment, the other - is to change the physicaland chemical state of the surface layers of thematerial. Dissipation of energy corresponds to anincrease of entropy (dS > 0).Energy balance of tribosystem according to thefirst law of thermodynamics describes by theequation.Wfr q W. (3)where q – energy of heat exchange with theenvironment, ΔW – change of internal energy is thesum of the energy used to change the structure ofthe material and energy of heating.At the same time, the work of the friction forceis the sum of the work of plastic deformation,hysteresis loss and the elastic deformation of thedispersion, that is, the work expended in theformation of new surfaces and associated with thesurface energy of solids [6, 7].The basis of the thermodynamic approach tofracture and wear of solids is energy mechanicalanalogy (deformation) and thermal (melting andsublimation) of failure.The energy spent on the deformation andfracture of solid bodies, compared with one of thethermodynamic characteristics of the material (heatof sublimation enthalpy of the solid and liquid state,latent heat of fusion). In this case, it is assumed thatthermodynamic properties are independent of thestructure of the material. The body is treated as acontinuous, homogeneous, isotropic medium with astatistically uniformly distributed structuralelements.Plastic deformation is considered as acombination of a large number of acts ofmicroscopic atomic-molecular rearrangementsassociated with the generation of sources ofdeformation (dislocations).Plastic deformation of the surface temperaturebelow the recrystallization temperature leads towork hardening of the surface layer and itsstrengthening. At widely differing hardnessstructural components of the material and repeatedexposure of loading occurs initially high wear ofsoft base; specific pressures acting on the solidcomponent thereby increasing, solid componentsare pressed into a soft base, some of them arebroken up and moved further under the forces offriction. As a result of such selective wear surfaceenriched solid structural components and gets stitchstructure that during wear of babbit, for example,according to research of M. M. Khrushchov andA. L. Kuritsyna [8].As a result of the interaction of the interfacedparts new surfaces are formed, which is followedby the energy release, γ ef , consumed for itsforming [6]: f F,Rz HV , (4)ef ,where F – the normal force of friction pair elementsinteraction; Rz – ten point height ofirregularities [9].Rough surface can be considered as a set ofirregularities randomly located on a perfect surfaceand having a random size, in other words, as therealization of the random field. This approachmakes it possible to represent the surface as a scalarrandom function [10]z = z(t, ω), (5)where the parameter t runs through the set of valuesof T, defined by the spatial arrangement of therough surface; ω – elementary event of aprobability space Ω.To determine the surface characteristics requiredfor calculations in tribology, often enoughknowledge only first two derivatives of the functionz(t, ω). Thus there is a need to calculate the densityof the joint distribution of several random variables.Quantify the contact of rough surfaces is animportant step in the development of physicalmodels of frictional interaction. It requiresconsideration of both the characteristics of theroughness, and the specific properties of thecontacting bodies, depending on their internalstructure, loading time and environmentalconditions.196 13 th International Conference on Tribology – Serbiatrib’13

Most of the friction units of products used inengineering works in conditions of oiling Thisrequires a comprehensive study of processes in thearea of friction as lubricant in some way facilitatesextraction of heat from the friction contact zone, theremoval of the zone of wear and corrosionprotection, and the protection of the frictionsurfaces and other structures from the effects of theenvironment and also seal gaps, etc.Taking into consideration that expression,V W / S (V fr W – the volume of the worn material, S fr –the friction track) represents the value of the wearrate J V [11], and changing of an inner energy isdetermined by the formula of specific energy ofdeformation ΔW accumulated in the material as aresult of dislocation forming [12] W f HVHV , , G, (6), 0 0where G – a displacement module of an examinedmaterial; α 0 – a parameter of interdislocationinteraction; HV – a microhardness of a surface layerof an examined part at the specified depth; HV 0 – amicrohardness of a undeformed material; based onenergy-based approach to the problem of definingthe relationship of the wear rate of work surfaces ofmachine parts with quality parameters of thesurface layer the wear rate functional relationshipwith geometrical (roughness) and physicomechanical(degree of work hardening) parametersof surface layer of machine parts during normaloperation can be represented aJV0,55FG(0,9fSRz F T (7)Sfr bal 0 ,2 4 )exp(200/ )2 2fr RzbalNbalHV0 0,2where J V – the wear rate [m 3 /m]; F – the normalforce of friction pair elements interaction [N]; f – afriction coefficient; S fr – the friction track [m]; Rz bal– the balanced roughness of the interfaced surfacesof the components [m]; 0,2 – the yield strengthconditional with the tolerance of 0,2% for the valueof the plastic deformation at stressing [Pa]; T – thetemperature in friction area [K]; N bal – the balanceddegree of hardening; HV 0 – a microhardness of aundeformed material [Pa].3. CONCLUSIONOn the basis of [13] can be defined relationshipequilibrium parameters of surface layer parts withmachining process parameters.The analysis of the experimental research resultsof the wear rate of contacted surfaces aftermachining has shown that the receivedmathematical model of the correlation between thewear rate and the technological requirements ofmachining allows for calculating the wear rate ofthe interfaced machine components after therunning-in period.REFERENCES[1] A. G. Suslov, A. M. Dalsky: Nauchnye osnovytehnologii mashinostroenija (Scientific Basis ofMechanical Engineering), - Mashinostroenie,Moscow, 2002.[2] A. V. Chichinadze, E. D. Brown, N. A. Boucher:Osnovy tribologii (trenie, iznos, smazka)(Fundamentals of tribology (friction, wear andlubrication)), - Mashinostroenie, Moscow, 2001.[3] V. S Kombalov: Ocenka tribotehnicheskih svojstvkontaktirujushhih poverhnostej (Evaluation oftribological properties of contacting surfaces), -Nauka, Moscow, 1983.[4] V. I. Butenko: Struktura i svojstva poverhnostnogosloja detalej tribosistem (Structure and properties ofthe surface layer of parts tribosystems), - Taganrog:TTI UFU, 2012.[5] U. N. Drozdov, V. G. Pavlov, V. N. Puchkov:Trenie i iznos v jekstremalnyh uslovijah:Spravochnik (Friction and wear in extremeconditions: Reference), - Mashinostroenie, Moscow,1986.[6] V. N. Kashcheev: Processy v zone frikcionnogokontakta metallov (Processes in the area of frictionalcontact metals), - Mashinostroenie, Moscow, 1978.[7] V. Weibul: Ustalostnye ispytanija i analiz ihrezultatov (Fatigue testing and analysis of theresults), - Mashinostroenie, Moscow, 1964.[8] I. P. Sukharev: Prochnost sharnirnyh uzlov mashin(The strength of hinged machine units), -Mashinostroenie, Moscow, 1977.[9] Russian State Standard 25142-82, 1988,Sherohovatost poverhnosti. Terminy I opredelenia.(Surface roughness. Terms and definitions),publishing company of State Standards, Moscow.[10] A. I. Sviridyonok, S. A. Chijik, M. I. Petrokovets:Mehanika diskretnogo frikcionnogo kontakta(Mechanics of discrete frictional contact), - Minsk,1990.[11] Russian State Standard 27674-88, Trenie,iznashivanie i smazka. Terminy i opredelenija,Moscow, 1988.[12] V. F. Bezjazychnyj, A. N. Sutyagin:Tehnologicheskoe obespechenie iznosostoikostidetalei mashin na osnove izucheniya nakoplennoienergii v poverhnostnom sloe detali prideformacionnom uprochnenii pri obrabotke(Machine components wear resistance technologicalassurance based on analysis of accumulated energyof part surface layer while strain hardening duringthe machining), Uprochnyauschie tehnologii ipokrytiya, 7, pp. 3-6.[13] V. F. Bezjazychnyj: Metod podobija v tehnologiimashinostroenija (Method of Similarity inmechanical engineering), Mashinostroenie,Moscow, 2012.13 th International Conference on Tribology – Serbiatrib’13 197

Serbian TribologySocietySERBIATRIB ‘1313 th International Conference onTribologyKragujevac, Serbia, 15 – 17 May 2013Faculty of Engineeringin KragujevacTHE INVESTIGATION OF COATED TOOLS TRIBOLOGICALCHARACTERISTICS INFLUENCE ON THE CUTTING PROCESSAND THE QUALITY PARAMETERS OF THE PARTS SURFACELAYERFomenko Roman NikolaevichFSBEIHPE «Rybinsk State Aviation Technical University» named after P.A. Solovjev (RSATU), Russia,fomenko85@mail.ruAbstract: The influence of cutting tools nanostructured coatings on the parameters of machined partssurface layer has been researched. The interaction between friction characteristics of coated tools and shearplane angle during machining has been determined. The results of different materials cutting with coatedcarbide-tipped tools have been shown.Keywords: nanostructured coatings, quality parameters of the surface layer, tools friction coefficient1. INTRODUCTIONThe most important production properties arereliability and endurance. These properties provideproduct safety and competitiveness. The main causeleading to breakdown of parts is fatigue cracks.Such cracks appear and propagate in thin surfacelayers of parts. In order to hamper crack growing,the surface layer has to exhibit certain features.They are: roughness, residual stress and strainhardening, which depend on the characteristics ofcutting operation.The cutting force, temperature of cutting, depthof wear hardening and degree of deformation arereferring to the main characteristics of cuttingoperation. These characteristics influence on theparts quality, reliability and endurance.Technological conditions of cutting such as toolsgeometry, processing conditions, work materialproperties and tooling material properties, includingtribological feathers, determine the characteristicsof cutting process. Therefore there is a need toselect optimal cutting conditions to provide therequirement parts quality. In order to select optimalcutting conditions, there is necessity to have aspecial methodic, which takes into considerationthe relationship between parts quality andtechnological conditions.2. TASKS OF RESEARCHAt the Rybinsk state of aviation technicaluniversity named after P. A. Solovjev (Russia)there was developed the methodic, which permit toestimate the optimal cutting conditions. On the baseof this methodic underlay a functional connectionbetween cutting rate, tools geometry and theparameters of surface layer, accuracy of machiningand the rigidity of manufacturing system, includingwork material and tool material properties.But all advanced tools have wear-resistantcoatings that exhibit specific properties. Wearresistantcoatings have low friction coefficient inconsequence of weak adhesion interaction ofcovering material with work material. Theyinfluence on the cutting process and qualityparameters of the surface layer. Tools coveringsreduce chips contact length with tools surface,cutting force, temperature of cutting anddeformation of cut allowance. It causes due toincreasing of a chip flow angle.Thus the main purpose of research was thecreation of the methodology for calculation oftechnological conditions of turning, which providesrequired quality and accuracy levels at the stage ofmachining and takes into consideration thetribological properties of coated tools.198 13 th International Conference on Tribology – Serbiatrib’13

In order to provide both high parts quality andmaximum tools life one should calculate so called«optimal cutting speed» v О . Optimal cutting rates(v О , S О ) correspond to the optimal cuttingtemperature. It is constant magnitude for the definecombination work – tool material [1]. Whenmachining with this temperature, maximum toolslifetime, minimal roughness of machined surfaceRa, minimal amount of surfaces defects have beenoccurred. Therefore these cutting rates should beused, when finishing work was performed for parts,which work in corrosive medium and hightemperature, because the surface layer has to containminimal amount of defects. For estimating of theoptimal cutting speed the equation is obtained byprof. Silin S. S. [1]:vOCO⋅ a ⎛ a b c ⎞a⎜1 ⋅ 1 ⋅ ρ ⋅θ=Pz⎟ , (1)1 ⎝ min ⎠where a 1 , b 1 – is the thickness and the width of cutrespectively [m]; а – is the coefficient of the temperatureconductivity of the work material [m 2 /s]; сρ – is thespecific heat capacity per unit volume[J/(m 3 · s · degree)]; θ – is the temperature in the cuttingarea, °С; n, C o – are coefficients, which depend on theproperties of work material; Pz min – is a minimalstabilized cutting force [N].But very often there is a need to select a cuttingcondition, which differs from the optimal ones.Therefore the opportunity to estimate thetechnological conditions of turning with taking intoconsideration the tribological properties of coatedtools, will provide the required quality and serviceproperties of parts at the stage of machining.The analysis of the mathematical models forestimating of the parameters of cutting process andquality of the surface level has shown, that themore important variable quantities are the shearplane angle β 1 and the adhesive component of thenfriction coefficient f M . Thus the main tasks of thescientific research were:1. To investigate the influence of tribologicalcharacteristics of coated tools on cutting process andthe parameters of surface layer.2. To define optimal cutting speed for tools withdifferent coatings.3. EXPERIMENTAL CONDITIONSThe wide range of cutting rates, different workmaterials and coated tools were selected forperforming of experiments (Table 1).In the capacity of tools were used the replaceableinserts 120412, material – VK6R (chemicalcomposition: Co – 6%, basis – WC) and TT7K12(chemical composition: Co – 12%, TiC – 1 %, TaC –7%, basis – WC). The different compositenanolaminated ion-plasmous coatings weredeposited on the replaceable inserts: (Ti;Si)N,(Ti,Si,Zr)CN and (Ti;Si;Al)N. Other groupreplaceable inserts was modified by implanting ofnanoparticle TiB 2 , Al 2 O 3 , Ta 2 O 3 and ZrB 2 in worksurface of tools. All selected coatings have beencharacterized by the minimal adhesive of the toolssurfaces with work material, and also they have beenprovided maximum tools lifetime. The machiningwas performed by the regular engine lathe NH 22.The temperature was measured by means of adynamic thermocouple of work material – toolingmaterial. The normal component of a cutting forcePz was measured by using the tool dynamometerDyna-Z, which was connected with personalcomputer (Figure 1). The tool dynamometer Dyna-Zis a self-sufficient measurement system, which canuse without an additional power source, atensometric station and DAQ board. And a precisionmeasured signal can be shown and saved in a veryuseful for operator form [3].Table 1. Experimental conditionsChanging parametersToolsgeometryCuttingrateWork materialHeat-resistant alloy(CrNi77ТiAlW)EI437Stainless steel(05Cr12Ni2Co3Мo2WV)EK26Cutting angle, γ° 5 8 0Relief angle, α° 10 12 10Lead angle, φ, φ 1 ° 45Nose radius, r, [mm] 1,2Titanium alloyOT4Deph t [mm] 0,25; 0,5; 0,75; 1Feed S [mm/rev] 0,07; 0,14; 0,2; 0,32Speed v [m/min] 14-170 33-190 15-130Tool material –VK6R TT7K12 VK6R ТТ7К12 VK6Rcarbide materialNanostructured coating (Ti,Si)N Ta 2 O 3 (Ti,Si)N Ta 2 O 3 (Ti,Si)N(Ti,Si,Al) ZrB 2 (Ti,Si,Al)N ZrB 2 (Ti,Si,Zr)CNNTiB 2 TiB 2 ZrB 2Al 2 O 3 Al 2 O 3 Al 2 O 313 th International Conference on Tribology – Serbiatrib’13 199

Fig. 1. The dynamometer Dyna-Z4. RESULTS AND DISCUSSIONThe experimental data of machinabilityinvestigation has been shown, that a cover of toolcan reduce a temperature θ in a cutting area on 50-70 °С, and a cutting force Pz can be reduced on 10-30% (Figure 2).Thus on the base of obtained powerdependences, one can make an equation ofmachinability to estimate of optimal cutting speedv О for different combination work material – coatedtool. The equations of machinability for consideredexamples have been given on Table 2.The optimal cutting speed of coated tool exceedsthe optimal cutting speed of uncoated tool. Thenfewer the coatings friction coefficient, then biggeroptimal cutting speed.In order to estimate the influence of coated toolson the parameters of surface layer, one have todetermine the influence of coated tools on a shearplane angle β 1 or a criterion B. This criterion is oneof the major parameter, which used for estimatingof roughness, residual stress and strain hardening inthe parts surface layer.Fig. 2. The dependence of a cutting force andtemperature on cutting conditions and tools cover; workmaterial – Stainless steel EK26; Tool material – carbidematerial VK6R; tools geometry: φ = φ 1 = 45°, γ = 8°; α= 7°, r = 1,2 mm; cutting rate: t = 1 mm; S = 0,32mm/rev; nanostructured coatings of tool: VK6R(without cover); (Ti; Si)N; (Ti; Si; Al)N;TiB2; Al 2 O 3.B = tg β 1 – Is the quantity, which defines thedegree of allowance plastic deformation and thedeformation of parts surface layer.The quantity β 1 was estimated by means ofTim’s I. A. formula using a chip reductioncoefficient k а , which was determinedexperimentally [1].cos( β1− γ )k a =, (2)sin βwhere γ – is a cutting angle.1Figure 3 shows the dependence of criterion B onthe technological conditions of operation.Table 2. The equations of machinability.Materials VK6R–EK26 VK6R–EK26–(Ti,Si)N VK6R–EK26–Al 2 O 3Equation of2,472,48⎛ ⎞⎛ ⎞⎛⎞machinability 2,31⋅a ⎜ a1⋅b1⋅ cρ⎟ 2,76 ⋅ a ⎜ a1⋅ b1⋅ cρ⎟ 4,76 ⋅ a ⎜ a1⋅ b1⋅ cρ⎟vO=0,72a⎜ ⎟ vO=0,7661 0,77 0 , 083a⎜ ⎟ vO=0,695⎝ t ⋅ St1 0,68 0 , 106a⎜⎟t1 0,737 0 , 044⎠⎝ t ⋅ S ⎠⎝ t ⋅ St⎠Frictioncoefficientf M0,44 0,35 0,16θ = 800 °Сv О [m/min],cutting ratet = 1 [mm];56 64 102S = 0,32[mm/rev]2,53200 13 th International Conference on Tribology – Serbiatrib’13

Fig. 3. The comparison the criterion В andtechnological conditions of operation; Work material –the stainless steel EK26, tool – TT7K12, coating – ZrB 2.It is clearly shown, that in the time of increasingof cutting speed v the criterion В increases too. It isthe reason for increasing of an angle of shear planeβ 1 . The angle of shear plane β 1 increases, becausethe materials ultimate stress σ В reduces by reason ofincreasing of rate of deformation and temperaturein the cutting area.On the base of experimental research theinfluence of different technological conditions onthe criterion В has been obtained. The quantity ofshear plane β 1 of coated tool increasesapproximately on 5-10 %. But experimentalequations are limited by technological conditions ofexperiments and couldn’t be used for otherconditions or other covers of tool. Therefore themethodology for estimating of a criterion B forother covers of tool has been developed. Thismethodology is based on the taking intoconsideration adhesive component of the frictioncoefficient f M of coated tool.For determination of the friction coefficient twoapproaches were used. According to the firstapproach, friction coefficient μ F was determined asa ratio of a tangential force to a normal force ofcutting:F FtanPy + Pх Ру ⋅ cosϕ+ Px ⋅ cos(90 − ϕ)µ = = =,(3)N РzPzwhere μ F – friction coefficient; Py, Px, Pz – componentsof a cutting force, [Н]; F tan – tangential force to a cutterface, [Н]; N – normal force to the cutter face, [Н]; Py –radial component of a cutting force, [Н]; Px – axialcomponent of a cutting force, [Н].On the figure 4 the dependence of criterion Band friction coefficient μ F on a dimensionlessv ⋅ a1complex Pe = , which defines theatechnological conditions of operation, has beenshown.The comparison of curves on the figure 4permits to create the proportion:В21Fµ ⋅ В1= (4)µF2The magnitude of unknown criterion B 2 can beapproximately estimated if the magnitudes ofF Fcriterion B 1 and friction coefficients µ1, µ2whichcorrespond to the tools with different coatings, areknown. But the determination of the frictioncoefficient μ F according to the first approachdoesn’t take into consideration the temperature inthe cutting area.The second approach has ‘not this shortcoming.According to the second approach fordetermination of the friction coefficient theadhesiometer was used (figure 5). It is known, thatthe friction coefficient:f = f D + f M , (5)where f D – deformation component of the frictioncoefficient; f М – adhesion (molecular) component of thefriction coefficient:Fig. 4. The dependence of criterion B and friction coefficient μ F on a dimensionless complex Pe; work material –Stainless steel EK26; tool material – carbide material VK6R; nanostructured coatings of tool: VK6R (withoutcover); (Ti;Si)N; (Ti;Si;Al)N; TiB 2 ; Al 2 O 313 th International Conference on Tribology – Serbiatrib’13 201

f M3 F ⋅ R= ⋅ , (6)4 N ⋅ rwhere R – radius of the disc, [m]; r – radius of theimpress on the sample, [m]; N – normal force, [H]; F –peripheral force on the disc, [H].quality and accuracy levels at the stage ofmachining and takes into consideration thetribological properties of coated tools, has beendeveloped. The methodology can estimatetechnological conditions of turning and solve aninverse task – it can estimate roughness, residualstress and strain hardening. The algorithm forcalculation of required technological conditions ofturning was implemented in the software (figure 7).Opportunity to choose oredit a function of frictionFig. 5. The flow chart of one-ball adhesiometer; 1 –samples of the work material; 2 – indenter of the toolmaterial; N – normal force, which impress the indenter[H]; F – peripheral force, which roll the disc, [H].Figure 6 shows the friction coefficient, whichwas determined for different temperatures andcombinations of work materials – coated indenter(pin).Figure 6. The influence of the temperature on afriction coefficient; work material – heat-resistant alloyEI437; tool material – carbide material H10F;nanostructured coatings of indentor: H10F (withoutcover); (Ti;Si)N; (Ti;Si;Al)N;TiB 2 ; Al 2 O 3.Thus if the magnitude of the criterion B 1 ofcover 1 and the functions of friction f M = f(θ) ofcover 1 and 2 like the dependence of frictioncoefficient on the temperature θ are known, themagnitude of a unknown criterion B 2 of cover 2can be approximately estimated by means ofcorrecting coefficient:cov er1Mcov er2Mfk = (7)fcov er1 cov er2where f M , f M – adhesion componentof the friction coefficient of work material withcover 1 and 2.On the base of obtained results of experimentsthe methodology for calculation of technologicalconditions of turning, which provides requiredFig. 7. The software for calculation of requiredtechnological conditions of turning.In order to check the obtained mathematicalmodels, the comparison of the experimental andcalculated data was performed. The investigation ofthe parameters of surface layer has been performedon the machined parts “ring”. The conditions ofturning of the parts “ring”: work material –stainless steel EK26; tool material – carbidematerial VK6R; tools geometry: φ = φ 1 = 45°, γ =8°; α = 7°, r = 1,2 mm; cutting rate: t = 0,75 mm; S= 0,2 mm/rev; nanostructured coatings of tool:(Ti;Si)N; (Ti;Si;Al)N.The results of experiments (Table 3) have beenclearly shown, that coated tool reduces themagnitude of the roughness, residual stress andstrain hardening in according with the magnitude offriction coefficient. The calculation of theparameters of the surface layer was performed bymeans of mathematical models presented in [2] andsoftware. The parameters of the roughness Ra andRz reduce on the average 5 %, therefore the maincause leading to the formation of the roughness aretools geometry, feed rate, vibration and so on, butnot the cover of tool. The strain hardening reduceson 20 % as compared with uncoated tool.In order to check our obtain data we havecompared the experimental and calculateddistribution diagrams of tangential residual stress.The using of coated tool leads to the considerablereduces of adverse tensile residual stresses.The distribution diagrams of the tangentialresidual stress are shown on a Figure 8.202 13 th International Conference on Tribology – Serbiatrib’13

Table 3. Experimental and calculated value of strain hardening h C and parameters of the roughness Ra and Rz.Cover Calcul. Exper. Δ, % Calcul. Exper. Δ, % Calcul. Exper. Δ, % Criterion ВRa, mkm Rz, мкм h CVK6R 1,84 1,42 29 8,4 6,8 23 37 50 26 0,95(Ti;Si)N 1,53 1,35 13 7 6,3 11 34 40 15 1,02(Ti;Si;Al)N 1,64 1,34 22 7,5 5,8 29 35 40 13 1,01The experimental distribution diagrams of thetangential residual stress were performed by meansof methodology of layer-by-layer electrochemicaletching. The using of coated tool leads to theconsiderable redaction of adverse tensile residualstress and its depth, the calculated data correlatewith the experimental ones.5. CONCLUSION (TIMES NEW ROMAN 11PT) - ALIGN LEFT1. The optimal cutting speed of coated toolexceeds the optimal cutting speed of uncoated tool;then less the coatings friction coefficient, then moreoptimal cutting speed.2. The using of coated tool leads to theconsiderable redaction of adverse tensile residualstress and its depth.REFERENCES[1] S. S. Silin: Similitude method of cutting,Manufacturing, Moscow, 1979.[2] V. F. Bezjazichnij: The method of similarity inmanufacturing engineering, Manufacturing,Moscow, 2012.[3] http://ooo-technolog.ru/Fig. 8. The distribution diagrams of the tangentialresidual stress of machined part.13 th International Conference on Tribology – Serbiatrib’13 203

Serbian TribologySocietySERBIATRIB ‘1313 th International Conference onTribologyKragujevac, Serbia, 15 – 17 May 2013Faculty of Engineeringin KragujevacMODELING SURFACE ROUGHNESS EFFECTS ON PISTONSKIRT EHL IN INITIAL ENGINE START UP USING HIGH ANDLOW VISCOSITY GRADE OILSMubashir Gulzar 1 , S.Adnan Qasim 1 , Riaz A Mufti 11 College of Electrical and Mechanical Engineering, National University of Sciences and Technology (NUST), Pakistan,mubashir_nustian@hotmail.comAbstract: The absence of fully developed fluid film lubrication between Piston and Liner surfaces is responsible forhigh friction and wear at initial engine start-up. In this paper flow factor method is used in two dimensional Reynolds’equation to model the effects of surface roughness characteristics on Piston Skirt elastohydrodynamic lubrication. Thecontact of surface asperities between the two surfaces and its after effects on EHL of piston skirt is investigated. Forthis purpose, two different grade oils are used to show the changing effects of viscosity combined with surfaceroughness on different parameters including film thickness, eccentricities and hydrodynamic pressures. The results ofthe presented model shows considerable effects on film thickness of rough piston skirt, hydrodynamic pressures andeccentricities profiles for 720 degrees crank angle.Keywords: EHL, Piston Skirt, flow factor, asperity, hydrodynamic, Vogelpohl parameter1. INTRODUCTIONIn initial engine start-up the piston and linersurfaces are not separated by an oil film whichcauses maximum wear and friction between the twosliding surfaces. The effects of physical contactsbetween the asperities of surfaces which are inrelative motion must be included in lubricationmodel to get a better understanding of rheology.Hamilton, Wallowit and Allen [1] were thepioneer for taken into account the roughness effectson lubrication phenomenon and their work datesback to 1966.They developed a theory ofhydrodynamic lubrication between two parallelsurfaces with surface roughness on one or both ofthe surfaces. The classical theory of lubricationdoes not predict the existence of any pressure incase of sliding flat parallel surfaces. Surfaceroughness helps in the pressure buildup between thetwo interacting surfaces, so provide a load supportand avoid collapse of two bodies. Early researchintegrated the roughness amplitude with the filmthickness and developed the modified onedimensional Reynolds’s equation but the presentedmodels did not cover different regimes and asperitycontacts and limited to one dimensional changes. Inthis prospective an exception is given in 1978 and1979 by Patir and Cheng [2][3]. Since thecontacting surfaces have an inherent roughness, soLambda Ratio or Tallian Parameter will be used asthe defining parameter between differentlubrication regimes [4]. In recent research the filmthickness parameter (λ) range has been investigatedand redefined for different lubrication regimes [5].The P.C. model was suitable for values of filmthickness ratio λ > 3 1.e; full film lubricationregime where asperity contacts were neglected[6][7].To minimize the wear and friction losses theelastohydrodynamic lubrication (EHL) model ispresented where λ is much lesser than a value of 3[5]. Thus the flow factor model provided by J.H.Tripp [8] is numerically modelled forhydrodynamic lubrication at initial engine start-up.Greenwood-Tripp asperity contact model is used toincorporate the asperity contact forces and asperitycontact friction force in EHL between the slidingsurfaces [9]. To incorporate the directionalbehaviour the Peklenik number [10] is defined forthe rough surfaces which are generated by normaldistribution using Fast Fourier Transform(FFT)[11][12]. Here isotropic surface roughnessproperty is used.204 13 th International Conference on Tribology – Serbiatrib’13

For developing the numerical model followingassumptions are taken:1. Lubricant is incompressible and thermaleffects are neglected.2. Non-Newtonian lubricant behavior isneglected.3. Pressure at the inlet is zero and surfaces areoil-flooded.4. Lubricant flow is laminar and turbulenceeffects are neglected.5. Leakage at the sides and edges is neglected.2. NOMENCLATUREC = Radial clearance between piston and liner =10micronsC f = Specific heat of lubricantC g = Distance from piston center of mass topiston pin = 0.2cmC p = Distance of piston-pin from axis of piston= 1 cmF = Normal force acting on piston skirtsF f = Friction force acting on skirts surfaceF fh = Friction force due to hydrodynamiclubricant filmF G = Combustion Gas force acting on the top ofpistonF h =Normal force due to hydrodynamic pressurein filmF IC = Transverse Inertia force due to piston massF ~ = Reciprocating Inertia force due to pistonICmassF IP = Transverse Inertia force due to piston pinmassF ~ = Reciprocating Inertia force due to pistonIPpin massF c = Asperity Contact ForceF fc = Friction force due to asperity contactG = Shear modulus of elastic lubricantI pis = Piston inertia about its centre of massM = Moment acting on piston skirtsM f = Friction moment acting on skirt surfaceM fh = Moment about piston pin due tohydrodynamic frictionM h = Moment about piston pin due tohydrodynamic pressureM c = Asperity Contact MomentM fc = Moment due o friction force of asperitycontactR = Radius of pistonU = Piston Velocitya = Vertical distance from skirt top to pistonpin= 0.0125mb = Vertical distance from skirt top to pistoncenter of gravity = 0.0015me t = Piston eccentricities at skirts top surfacee b = Piston eccentricities at skirts bottom surfaceё b = Acceleration of piston skirts bottomeccentricitiesё t = Acceleration of piston skirts topeccentricitiesh = Film Thicknessl = Connecting rod lengthm pis = Mass of piston = 0.295 kgm pin = Mass of piston-pin = 0.09 kgp = Hydrodynamic pressurer = Crank radius = 0.0418 mω = Constant crankshaft speed (engine speed)τ = Shear stressη A = Oil A viscosity = 0.016 Pa.s,η B = Oil B viscosity = 0.1891 Pa.s. = Connecting rod angle = Crank angle X ,y = Pressure flow factor along x and y-axisrespectivelys = Shear flow factor = combined root mean square (rms)roughness1 = rms roughness of piston skirt= 1.4µm2 = rms roughness of cylinder liner = 1.5µm3. MATHEMATICAL MODEL3.1. Equations of Piston MotionThe forces and moments are in the form of theforce and moment balance equations similar to thatdefined by Zhu et al [13]: a a1121aa2222 e Fh Fc Fs ( Ffh Ft eb Mh Mc Ms Mffc) tan (1) a b a11 mpin 1 mpin 1 (2a) L L a a b a12 mpin mpin (2b) L L 21aF I Lpin mpinbLa b (1 )b I(2c)pin 22 mpina b ( ) (2d)L L s~ ~ tan F F F(3)MsGGpIP~ F C F C(4)Using the Greenwood-Tripp’s Asperity ContactModel, the values of F c , F fc , M c and M fc can befound for EHL regime [9].ICgIC13 th International Conference on Tribology – Serbiatrib’13 205

3.2. Film Thickness EquationThe film thickness between the skirts and the linergiven by Zhu [9]:h C ety ( t ) cos x eb( t ) et( t )] cos x (6)L 3.3. Reynolds’ Equation ModellingModified 2-D Reynolds equation is given as [2]:2 3 p R 3 p hTS h x h y 6U( ) (7)x x L y y xxwhere x and y are Poiseulle or pressure flowfactors and s is Cuotte or shear flow factor [2][8].The boundary conditions are defined as [4]:pxp 0x p =0 when x 1 ≤ x ≤ x 2p( ,0) p( , L) 0In dimensionless form the 2-D Reynoldsequation is given by [9][4]:20(8) 3 p* R 3 p* hT* Sh*x h*y * (9)x* x* L y* y* x*x*Where by J. H Tripp [8] 1 3( 2) /( 1)][ / h]X y = ( 1/ ) X2212 (h, 1) (hs, 2)2 s 2 s 3 (hs , 2) ( / h)( 1)and is the Peklenik number [10]In order to read the pressure profiles conveniently,the Vogelpohl parameter M v is introduced [4]:21.5M v p*h*The Reynolds equation in terms of the Vogelpohlparameter is given as:2 2 2Mv R Mv . x Mv2 2 (1 / ) ( ) . x* L . y* . x* . x*2 . y R.Mv (1 / ) ( ) FMvG. y* L. y*where(10)2 2 2 2 2 2h* R h* h* R h* 2 20.75 1.5x* Ly* x* Ly*F 2h* h*2h* Rh*1.5 x* x* L y* y* h * h* S * x* x*G 1.5( h * )Mvi ,, jRC. M M C.M ML 2. C 2.C F v i j v il j v i jlv i ji1 , 1, , , 2 , , , ,. * C. (1/ *)M M . x * 2. C 2.C F21 2 i,j3 vi , 1, j vi , l,j1 2 i,j. * C. (1/ *)M M G. y *2. C 2.C F4 vi , 1, j vi , l, j i.j1 2 i,j3.4. Film Thickness in EHL RegimeEHL regime the film thickness includes filmInthickness in the rigid hydrodynamic regime and theelastic surface displacements etc. By consideringthe bulk elastic deformation, the lubricant filmthicknessequation takes the following form [14] h f ( , y) vh eh ;where f(θ ,y ) is neglected. The differentialsurface displacement is [14]:1 p ( x , y ) dydydv E rr( xxy220) ( y0)221 1 (1 v 1) (1 v2) E 2 E1E2 At a specific point (x o , y o )deformation is [4]:1v( x0,y0)Eap(x,y)dxdyr4. RESULTS AND DISCUSSIONthe elasticThe hydrodynamic lubrication and EHL modelsof the piston skirts at 500 rpm are developed afterincorporating the pressure flow and the shear flowfactors. Two different oils having viscosity 0.016Pa.s and 0.1891 Pa.s are used for a comparison andinvestigating the viscosity effects on differentparameters which include film thickness206 13 th International Conference on Tribology – Serbiatrib’13

,eccentricities and hydrodynamic pressure profilesat 720 degree crank rotation cycle.4.1 Piston EccentricitiesThe dimensionless eccentricities of the top andthe bottom surface of the piston skirts (Et and Eb)are plotted against the 720 degree crank rotationcycle. Figure 1(a) and 1(b) show eccentricityprofiles for Oil A at 500 rpm. The results areplotted between a range of 1 and -1 where thephysical contact between the sliding surfaces canoccur. At central value ‘0’ the motion is concentric.Figure 1(a) shows the dimensionless eccentricityprofiles in the hydrodynamic lubrication regimewhereas figure 1 (b) shows the similar profiles inthe EHL regime. The behaviour is shown for all thefour strokes where it can be seen that at the start ofcycle the piston and liner axis are concentric thendue to the secondary motion the profiles are highlydisplaced from the centre towards thrust side andnon thrust side, but for Oil A the physical contact isavoided as shown in Figure 1. For Oil B , thedimensionless eccentricities profiles forhydrodynamic and EHL regime are shown inFigure 4. Figure 4 (a) shows that the contact isestablished at lower surface as line is meeting with-1 in rigid hydrodynamic regime. However inFigure 4 (b) the EHL regime shows the physicalcontact is clearly avoided. This shows that theelastic deformation of asperities help in avoidingthe contact between interacting surfaces, thus helpin avoiding friction related wear.Comparison of eccentricities for both oilsprovides an interesting finding that the lowviscosity oil can be more helpful at initial enginestart-up speed of 500 rpm for rigid hydrodynamicregime as well as equally good for EHL regime.4.2 Hydrodynamic PressuresThree dimensional pressure fields and relatedpressure distribution are plotted for 720 degreecrank angle. Figure 2 (a), 2(b), 2(c), 2(d) show 3- Dhydrodynamic pressure profiles at 900, 4500, 6300and 7200 crank angles at 500 rpm. The positivepressures are developed over the piston skirt andvary as shown in Figure 2. In figure 2 (a), for Oil A,at 90 degrees crank angle the pressures are biasedtowards bottom of piston skirt and extended to themiddle of piston skirt. The peak pressure occurs atthe bottom of piston skirt. In figure 2 (b), for Oil A,at 450 degrees crank angle, the pressure field showsthat the hydrodynamic pressures are developed attop of piston skirt though a small ridge can be seenat bottom of piston Skirt. The peak pressures arelarger than the 90 degrees angle. In figure 2 (c), at630 degrees crank angle, the pressures are shiftedtowards top of piston skirt. In figure 2(d), at 720degrees the pressure profile is more steep anddeveloped at bottom of piston skirt showing the endof cycle. For Oil B, in figure 5(a), 5(b), 5(c), 5(d)show 3- D hydrodynamic pressure profiles at 900,4500, 6300 and 7200 crank angles at 500 rpmspeed.For the pressure fields it can be clearlyinvestigated that the hydrodynamic pressures aretotally shifted towards top of piston skirt at 450degrees crank angle while the case was not same incase of Oil A for similar conditions. The majorchange in shape of pressure filed can be observedfor 630 degrees crank angle where thedimensionless pressure is biased towards bottom ofpiston skirt instead of top as discussed for Oil A.Thus changing the viscosity of oil is affecting thedistribution of hydrodynamic pressures over pistonskirt.4.3 Hydrodynamic and EHL Film ThicknessFigure 3(a) shows the maximum and theminimum hydrodynamic film thickness for Oil A at500 rpm and 10 micron radial clearance. Themaximum film thickness is calculated before theapplication of load and on the other side theminimum film thickness is found after theapplication of load. The magnitude of minimumfilm thickness shows whether the film thickness iscapable of avoiding the contact between slidingsurfaces or not. In figure 3(a), the minimumhydrodynamic film start getting establish from startof cycle and reaches at a peak at power stroke anddecrease to minimum at end of exhaust stroke andcycle continues. Similar case can be seen for Oil Bin figure 6(a) , but the difference is evident at endof exhaust stroke where a second peak of filmthickness can be seen. In figure 3(b) and 6(b) EHLfilm thickness profiles are shown. By comparingboth profiles, it can be seen that in case of Oil A theEHL film thickness is greater in magnitude fordifferent crank angles as compare to Oil B. ThusOil A , which is low viscosity oil, will be morehelpful in avoiding the contact and wear betweenrough piston and liner surfaces.(a)(b)Figure 1. For Oil A , Dimensionless Eccentricities at500 rpm in (a) Hydrodynamic regime (b) EHL Regime13 th International Conference on Tribology – Serbiatrib’13 207

(a)(b)(a)(b)(c)(d)Figure 2. : For Oil A, 3-D Hydrodynamic pressurefields at 500 rpm at crank angle (a) 90 degree (b) 450degree (c) 630 degree (d) 720 degree(c)(d)Figure 5. : For Oil B, 3-D Hydrodynamic pressurefields at 500 rpm at crank angle (a) 90 degree (b) 450degree (c) 630 degree (d) 720 degree(a)(b)Figure 3. : For Oil A, At 500 rpm (a) Film thicknessprofiles (b) EHL film(a)(b)Figure 6. : For Oil B, At 500 rpm (a) Film thicknessprofiles (b) EHL film5. CONCLUSION(a)(b)Figure 4. For Oil B , Dimensionless Eccentricities at500 rpm in (a) Hydrodynamic regime (b) EHL RegimeTwo dimensional numerical models forhydrodynamic and EHL regimes were developed atinitial engine start-up speed for isotropic roughpiston skirt and cylinder. Two different grade oilswere used to investigate the different parametersaffecting the rough piston skirt wear phenomenon.The different rough surfaces of the interacting skirtsand the liner were considered by introducing thepressure and the shear flow factors in thelubrication model. For Oil ‘B’ having a viscosity of0.1891 Pa.s, the simulation results verify that aphysical contact between the rough skirts and theliner surfaces cannot be avoided in the rigid208 13 th International Conference on Tribology – Serbiatrib’13

hydrodynamic regime. However, for both oils, therough interacting surfaces deform elastically togenerate a sufficiently thick film in the EHLregime. The hydrodynamic pressures shifting occurfrom top of piston skirt to bottom at 630 degreescrank angle by changing Oil A to Oil B at 500 rpmand 10 micron radial clearance. Comparing bothoils for given conditions, Oil A is more suitable toavoid the contact and wear between interactingrough surfaces.REFERENCES[1] Hamilton, D. B., Walowit, J.A., & Allen, C.M. ATheory of Lubrication by Microirregularities.,Journal of Basic Engineering, Trans ASME, Ser.D., March 1966, pp. 177-185.[2] N. Patir, & H. S. Cheng, An Average Flow Modelfor Determining Effects of Three-DimensionalRoughness on Partial Hydrodynamic Lubrication,ASME Journal of Lubrication Technology, 100(1),pp 12-17, 1978.[3] N. Patir, & H. S. Cheng, Application of AverageFlow Model to Lubrication Between Rough SlidingSurfaces, ASME Journal of LubricationTechnology, 101 (2), pp220-230,1979.[4] G. W. Stachowiak & A. W. Batchelor Engineeringtribology, 3rd ed., Elsevier, pp. 328 2005[5] Dong Zhu and Q Jane Wang ,On the λ ratio range ofmixed lubrication, Proc IMechE Part J: Journal ofEngineering Tribology pp1010–1022,226(12)[6] Tripp JH, Hamrock BJ. Surface roughness effects inelastohydrodynamic contacts. In: Proc 1984 LeedsLyon Symposium on Tribology 1985:30–9.[7] H. G. Elrod, A General Theory for LaminarLubrication with Reynolds Roughness, ASMEJournal of Lubrication Technology, 101(1), pp 8-14,1979.[8] J. H. Tripp, Surface Roughness Effects in HydrodynamicLubrication: The Flow Factor Method,ASME Journal of Lubrication Technology, 105, pp458-463,1983.[9] J. A. Greenwood, & J. H. Tripp, The Contact ofTwo Nominally Flat Rough Surface, Proc.Institution of Mechanical Engineers (IMechE), UK,(185), pp 625-633,1971.[10] Peklenik, J., New Developments in SurfaceCharacterization and Measurement by Means ofRandom Process Analysis, Proc. Instn. Mech.Engrs., Volume 182, pp 108-125,1967-68.[11] Garcia, N.; Stoll, E.: Monte Carlo Calculation ofElectromagnetic-Wave Scattering from RandomRough Surfaces, Physical Review Letters, Volume52, Issue 20, pp. 1798-1801, 1984.[12] FFTW library - free collection of fast C routines forcomputing discrete Fast Fourier Transforms.Developed at MIT by Matteo Frigo and Steven G.Johnson.[13] Zhu, D., Hu, Y., Cheng, H.S., Arai, T., and Hamai,K.:, A numerical Analysis for Piston Skirt in MixedLubrication, Part 2 : Deformation Consideration,ASME Journal of Tribology,pp115-125, 1992.[14] S. A. Qasim, M. A. Malik, M. A. Khan, & R.A.Mufti, Low Viscosity Shear Heating in PistonSkirts EHL in the Low Initial Engine Start UpSpeeds, Tribology International, 44(10), pp. 1134-1143, 2011.13 th International Conference on Tribology – Serbiatrib’13 209

Serbian TribologySocietySERBIATRIB ‘1313 th International Conference onTribologyKragujevac, Serbia, 15 – 17 May 2013Faculty of Engineeringin KragujevacSTRESSES AND DEFORMATIONS ANALYSIS OFA DRY FRICTION CLUTCH SYSTEMOday I. Abdullah 1 , Josef Schlattmann 1 , Abdullah M. Al-Shabibi 21 Department of System Technology and Mechanical Design Methodology, Hamburg University of Technology,oday.abdullah@tu-harburg.de2 Department of Mechanical and Industrial Eng. College of Eng. / Sultan Qaboos UniveristyAbstract: The friction clutch is considered the essential element in the torque transmission process. In thispaper, the finite element method is used to study the stresses and deformations for clutch system (pressureplate, clutch disc and flywheel) due to the contact pressure of diaphragm spring and the centrifugal forceduring the full engagement of clutch disc (assuming no slipping between contact surfaces). The investigationcovers the effect of the contact stiffness factor FKN on the pressure distribution between contact surfaces,stresses and deformations. The penalty and Augmented Lagrangian algorithms have been used to obtain thepressure distribution between contact surfaces. ANSYS13 software has been used to perform the numericalcalculation in this paper.Keywords: Dry friction clutch, Stresses and deformations, pressure distribution, full engagement,2D axisymmetric FEM.1. INTRODUCTIONA clutch is a very important machine elementwhich plays a main role in the transmission ofpower (and eventually motion) from onecomponent (the driving part of the machine) toanother (the driven part). A common and wellknown application for the clutch is in automotivevehicles where it is used to connect the engine andthe gearbox. Furthermore, the clutch is used alsoextensively in production machinery of all types.When the friction clutch begins to engage, slippingoccurs between the contact surfaces (pressure plate,clutch disc and flywheel) and due to this slipping,heat energy will be generated in the interfacesfriction surfaces. At high relative sliding velocity,high quantity of frictional heat is generated whichlead to high temperature rise on the clutch discsurfaces and hence thermo-mechanical problemssuch as thermal deformations and thermo-elasticinstability can occur. This in turn, can lead tothermal cracking and high rate of wear. Thepressure distribution is essential factor effect on theperformance of the friction clutch because of theheat generated between contact surfaces during theslipping period dependent on the pressuredistribution.Al-Shabibi and Barber [1] used the finiteelement method to find the transient solution of thetemperature field and contact pressure distributionbetween two sliding disks. Two dimensionalaxisymmetric FE model used to explore analternative method based on an eigenfunctionexpansion and a particular solution that can be usedto solve the thermoelastic contact problem withfrictional heating. Both constant and varying slidingspeed is considered in this analysis. Results of thedirect finite element simulation have been obtainedusing the commercial package ABAQUS. Theresults from the approximate solution show a goodagreement with the results from the direct finiteelement simulation.Lee et al. [2] used finite element method tostudy the effect of thermo-mechanical loads on thepressure plate and the hub plate of the frictionclutch system. Three types of loads are taking intoconsideration the thermal load due to the slippingoccurs at the beginning of engagement, the contactpressure of diaphragm spring and the centrifugalforce due to the rotation. Two and threedimensional finite element models were performed210 13 th International Conference on Tribology – Serbiatrib’13

to obtain the temperature distributions and thestresses. The results show the significant effect ofthe thermal load on the temperatures and stresses;therefore it is desirable to increase the thickness ofthe pressure plate as much as possible to increasethe thermal capacity of the pressure plate to reducethe thermal stresses. High stress intensity valueoccurs around the fillet region of the window in thehub plate.Shahzamanian et al. [3] used numericalsimulation to study the transient and contactanalysis of functionally graded (FG) brake disk.The material properties vary in the radial directionfrom full-metal at the inner radius to that of fullceramicat the outer radius. The coulomb contactfriction is considered between the pad and the brakedisk.Two-dimensional finite element model used inthe work to obtains the pressure distribution, totalstresses, pad penetration, friction stresses, heat fluxand temperature during the contact for differentvalues of the contact stiffness factor. It was found,that the contact pressure and contact total stressincrease when the contact stiffness factor increasesand the gradation of the metal–ceramic hassignificant effect on the thermo-mechanicalresponse of FG brake disks. Also, it can beconcluded when the thickness of the pad increasesthe contact status between pad and disc changesfrom sticking to contact and then to near contact.Abdullah and Schlattmann [4-8] investigated thetemperature field and the energy dissipated of dryfriction clutch during a single and repeatedengagement under uniform pressure and uniformwear conditions. They also studied the effect ofpressure between contact surface when varyingwith time on the temperature field and the internalenergy of clutch disc using two approaches heatpartition ratio approach to compute the heatgenerated for each part individually whereas thesecond applies the total heat generated for thewhole model using contact model. Furthermore,they studied the effect of engagement time andsliding velocity function, thermal load anddimensionless disc radius (inner disc radius/outerdisc radius) on the thermal behavior of the frictionclutch in the beginning of engagement.In this paper the finite element method used tostudy the contact pressure and stresses during thefull engagement period of the clutches usingdifferent contact algorithms. Moreover, sensitivitystudy for the contact pressure is presented toindicate the importance of the contact stiffnessbetween contact surfaces.2. FUNDAMENTAL PRINCIPLESThe main system of the friction clutch consistsof pressure plate, clutch disc and flywheel as shownin figure 1.TTorqueFlywheelFigure 1. The main parts of clutch systemThermal +Contact pressureSlipping period(Transient case)Thermal +Contact pressure+Centrifugal effectFull engagementperiod(Steady-statecase)TimeFigure 2. The load conditions during theengagement cycle of the clutchWhen the clutch starts to engage the slippingwill occur between contact surfaces due to thedifference in the velocities between them (slippingperiod), after this period all contacts parts arerotating at the same velocity without slipping (fullengagement period). A high amount of the kineticenergy converted into heat energy at interfacesaccording to the first law of thermodynamics duringthe slipping period and the heat generated betweencontact surfaces will be dissipated by conductiont sClutch discPressure plate13 th International Conference on Tribology – Serbiatrib’13 211

etween friction clutch components and byconvection to environment, in addition to thethermal effect due to the slipping there is other loadcondition which is the pressure contact betweencontact surfaces. In the second period, there arethree types of load conditions the temperaturedistribution from the last period (slipping period),the pressure between contact surfaces due to theaxial force of diaphragm spring and the centrifugalforce due to the rotation of the contacts parts.Figure 2 shows the load conditions during theengagement cycle of the clutch, where t s is theslipping time and T is the transmitted torque byclutch.3. FINITE ELEMENT FORMULATIONThis section presented the steps to simulate thecontact elements of friction clutch using ANSYSsoftware. Moreover it gives more details about thetypes of contacts and algorithms which are used inthis software.The first step in this analysis is the modelling;due to the symmetry in the geometry (frictionallining without grooves) and boundary conditions ofthe friction clutch (take into the consideration theeffect of the pressure and centrifugal force loads,and neglected the effect of thermal load due to theslipping), two-dimensional axisymmetric FEM canbe used to represent the contact between the clutchelements during the steady-state period as shown infigure 3.There are three basic types of contact used inAnsys software single contact, node-to-surfacecontact and surface-to-surface contact. Surface-tosurfacecontact is the most commonly type ofcontact used for bodies that have arbitrary shapeswith relative large contact areas. This type ofcontact is most efficient for bodies that experiencelarge values of relative sliding such as block slidingon plane or sphere sliding within groove [9].Surface-to-surface contact is the type of contactassumed in this analysis because of the large areasof clutch elements in contact.In this work, it has been assumed two types ofload conditions effects on the clutch system duringthe steady-state period (full engagement period) thecontact pressure between clutch elements due to theaxial force by diaphragm spring and the centrifugalforce due to the rotation.The elements used for contact model are: “Plan13” used for all elements of theclutch (flywheel, clutch disc and pressureplate). “Conta172” used for contact surfaces thatare the upper and lower surfaces of clutchdisc. “Targe169” used for the target surfacesthat are the lower surface of the flywheeland the upper surface of the pressure plate.Figure 4 shows the details about schematic for allelements that has been used in this analysis.zTarge169rThe frictional liningFlywheelPressure plateAxialFigure 3. The Contact model for clutch systemPlane13FlywheelClutch discConta172Pressure plateFigure 4. Schematic elements used for the friction clutchelementsThe stiffness relationship between contact andtarget surfaces will decide the amount of thepenetration. Higher values of contact stiffness willdecrease the amount of penetration, but can lead toill-conditioning of the global stiffness matrix andconvergence difficulties. Lower values of contactstiffness can lead to certain amount of penetrationand low enough to facilitate convergence of thesolution. The contact stiffness for an element ofarea A is calculated using the following formula[10]: f e f i iTFkn dA(1)The default value of the contact stiffness factorFKN is 1, and it is appropriate for bulkdeformation. If bending deformation dominates thesolution, a smaller value of KKN = 0.1 isrecommended.There are five algorithms used for surface-tosurfacecontact type are: Penalty method: this algorithm used constant“spring” to establish the relationship betweenthe two contact surfaces (figure 5). The contact212 13 th International Conference on Tribology – Serbiatrib’13

force (pressure) between two contact bodies canbe written as follows:F k x(2)nnWhere F n is the contact force, k n is the contactstiffness and x p is the distance between two existingnodes or separate contact bodies (penetration orgap).pF nshown in figure 6. A mesh sensitivity study wasdone to choose the optimum mesh fromcomputational accuracy point of view. The fullNewton-Raphson with unsymmetric matrices ofelements is used in this analysis assuming a largedeflectioneffect. In all computations for the frictionclutch model, it has been assumed a homogeneousand isotropic material and all parameters andmaterials properties are listed in Table. 1.In this analysis also assuming there are nocracks in the contact surfaces and the actual contactarea is equal to the nominal contact area.k nx pωFlywheelFigure 5. The contact stiffness between two contact bodies Augmented Lagrange (default): this algorithm isan iterative penalty method. The constanttraction (pressure and frictional stresses) areaugmented during equilibrium iterations so thatthe final penetration is small than the allowabletolerance. This method usually leads to betterconditioning and is less sensitive to themagnitude of the constant stiffness. The contactforce (pressure) between two contact bodies is:F k x (3)nnWhere λ is the Lagrange multiplier component. Lagrange multiplier on contact normal andpenalty on tangent: this method applied on theconstant normal and penalty method (tangentialcontact stiffness) on the frictional plane. Thismethod enforces zero penetration and allowssmall amount of slip for the sticking contactcondition. It requires chattering controlparameters, as well as the maximum allowableelastic slip parameter. Pure Lagrange multiplier on contact normal andtangent: This method enforces zero penetrationwhen contact is closed and “zero slip” whensticking contact occurs. This algorithm does notrequire contact stiffness. Instead it requireschattering control parameters. This method addscontact traction to the model as additionaldegrees of freedom and requires additionaliterations to the stabilize contact conditions. Itoften increase the computational cost comparedto the augmented lagrangian method. Internal multipoint constraint: this method usedin conjunction with bonded contact and noseparation contact to model several types ofcontact assemblies and kinematic constraints.The axisymmetric finite element model of thefriction clutch system with boundary conditions ispzrClutch discPressure plateU x =0Table 1. The properties of materials and operationsParametersPressureContactsurfacesFigure 6. FE models with the boundary conditions.ValuesInner radius of friction material & axial cushion, r i [m] 0.06298Outer radius of friction material & axial cushion, r o [m] 0.08721Thickness of friction material [m], t l 0.003Thickness of the axial cushion [m], t axi. 0.0015Inner radius of pressure plate [m], r ip 0.05814Outer radius of pressure plate [m], r op 0.09205Thickness of the pressure plate [m], t p 0.00969Inner radius of flywheel [m], r if 0.04845Outer radius of flywheel [m], r of 0.0969Thickness of the flywheel [m], t f 0.01938pressure, p [MPa] 1Coefficient of friction, μ 0.2Number of friction surfaces, n 2Torque [Nm], T 432Maximum angular slipping speed, ω o [rad/sec] 200Young’s modulus for friction material, E l [GPa] 0.30Young’s modulus for pressure plate, flywheel &axial cushion, (E p , E f , and E axi ), [Gpa]125Poisson’s ratio for friction material, 0.25Poisson’s ratio for pressure plate, flywheel & axialcushion0.25Density for friction material, (kg/m 3 ), ρ l 2000Density for pressure plate, flywheel & axialcushion, (kg/m 3 ), (ρ p , ρ f , and ρ axi )780013 th International Conference on Tribology – Serbiatrib’13 213

4. RESULTS AND DISCUSSIONSSeries of computations have been carried outusing ANSYS13 software to study the contactpressure and stresses between contact surfaces ofclutch (pressure plate, clutch disc and flywheel)during a full engagement period using differentalgorithms and contact stiffness factor values.The variation of the contact pressure with discradius for both sides of clutch disc (flywheel sideand pressure plate side) using penalty andaugmented algorithms (FKN = 1) is shown infigures 7 and 8. From these figures, it can be seenthat the identical results when using penalty andaugmented (default) methods and approximatelythe same behaviour of contact pressure for bothsides of clutch disc. The maximum contact pressurevalues in the flywheel side and pressure plate sideare found to be 1.491 MPa and 1.524 MPa,respectively. The maximum and minimum contactpressure values occur at outer disc radius r o andnear inner radius (1.01r i ) for both cases,respectively.The contact pressure [MPa]1.61.551.51.451.41.351.31.251.21.151.11.051Penalty methodAugmented method0.065 0.07 0.075 0.08 0.085r[m]Figure 7. The variation of contact pressure with discradius (flywheel / clutch disc)The contact pressure [MPa]1.61.551.51.451.41.351.31.251.21.151.11.051Penalty methodAugmented method0.065 0.07 0.075 0.08 0.085r[m]Figure 8. The variation of contact pressure with discradius (pressure plate / clutch disc)The contact total stress [MPa]1.61.551.51.451.41.351.31.251.21.151.11.051Penalty methodAugmented method0.065 0.07 0.075 0.08 0.085r[m]Figure 9. The variation of total contact stress with discradius (flywheel / clutch disc)The contact total stress [MPa]1.61.551.51.451.41.351.31.251.21.151.11.051Penalty methodAugmented method0.065 0.07 0.075 0.08 0.085r[m]Figure 10. The variation of total contact stress with discradius (pressure plate / clutch disc)Figures 9 and 10 show the variation of totalcontact stresses with disc radius for both sides ofclutch disc. It can be seen, that the total contactstresses have the same behaviour of the contactpressure.Figures 11 and 12 demonstrate the variation oftotal displacement of clutch surfaces with discradius. It’s clear; the values of total deformations ofclutch disc (pressure plate side) are higher than thedisplacements values at the flywheel side. Themaximum values of total deformation in the clutchdisc at flywheel and pressure plate sides are foundto be 4.6529E -6 m and 2.84E -5 m, respectively.The variation of the contact pressure for usingdifferent algorithms and different values of FKNalong the radial direction at contact area of clutchdisc with flywheel is shown in figures 13 and 14. Itcan be noted for both cases (when using penaltyand augmented method), that the values of contactpressure increases when FKN increases. Thepercentage increasing in contact pressure whenFKN change from 0.01 to 10 is found to be 19.5%and 17.9% corresponding to penalty and augmentedmethods, respectively.214 13 th International Conference on Tribology – Serbiatrib’13

5E-06The total displacement [m]4.5E-064E-063.5E-063E-062.5E-062E-06Penalty methodAugmented methodThe contact pressure [MPa]1.7 FKN = 0.01FKN = 0.1FKN = 1FKN = 001.61.51.41.31.5E-060.065 0.07 0.075 0.08 0.085r[m]Figure 11. The variation of total displacement with discradius (flywheel / clutch disc)The total displacement [m]3E-052.8E-052.6E-052.4E-052.2E-052E-05Penalty methodAugmented method0.065 0.07 0.075 0.08 0.085r[m]Figure 12. The variation of total displacement with discradius (pressure plate / clutch disc)The contact pressure [MPa]1.7 FKN = 0.01FKN = 0.1FKN = 1FKN = 101.61.51.41.31.20.065 0.07 0.075 0.08 0.085r[m]Figure 13. The variation of contact pressure with discradius using Penalty method (flywheel / clutch disc)1.20.065 0.07 0.075 0.08 0.085r[m]Figure 14. The variation of contact pressure with discradius using augmented Lagrange algorithm (flywheel /clutch disc)5. CONCLUSIONS AND REMARKSThe variations of the contact pressure, totalcontact stress and total displacements of the frictionclutch using different contact algorithms anddifferent values of FKN are investigated. Twodimensionalaxisymmetric finite element model forthe contact elements of clutch were conducted toobtain the numerical results.The present work presents a simplified model ofclutch to determine the contact pressure betweencontact surfaces during a full engagement period.The conclusions obtained from the presentanalysis are summarized as follows:1. The value of FKN is very important andeffective on the values of contact pressure, thecontact pressure is directly proportional to FKNfor both contact methods (penalty andaugmented).2. The penalty method has sensitivity for FKNmore than the augmented method.3. The maximum and minimum values of contactpressure and total contact stress occur at outerdisc radius and inner disc radius, respectively.The permanent deformations and thermal crackson the contact surfaces of clutch if taken intoconsideration will affect the contact pressuredistribution and the actual contact area will change.These disadvantages will focus the contact pressureon small region compared with the nominal contactarea.REFERENCES[1] Abdullah M. Al-Shabibi and James R. Barber:Transient Solution of The UnperturbedThermoelastic Contact Problem, J. thermal stresses,Vol. 32, pp. 226-243, 2009.13 th International Conference on Tribology – Serbiatrib’13 215

[2] Choon Yeol Lee, Il Sup Chung, Young Suck Chai:Finite Element Analysis of an Automobile ClutchSystem, J. Key Eng. Materials, Vol 353-358, pp.2707-2711, 2007.[3] M.M. Shahzamanian, B.B. Sahari , M. Bayat, Z.N.Ismarrubie and F. Mustapha: Transient and thermalcontact analysis for the elastic behavior offunctionally graded brake disks due to mechanicaland thermal loads, J. Materials & Design, Vol. 31,Issue 10, pp. 4655–4665, 2010.[4] Oday I. Abdullah and Josef Schlattmann: The Effectof Disc Radius on Heat Flux and TemperatureDistribution in Friction Clutches, J. AdvancedMaterials Research, Vol. 505, pp. 154-164, 2012.[5] Oday I. Abdullah and Josef Schlattmann: FiniteElement Analysis of Dry Friction Clutch with Radialand Circumferential Grooves, Proceeding of WorldAcademy of Science, Engineering and TechnologyConference, Paris, pp. 1279-1291, 2012.[6] Oday I. Abdullah and Josef Schlattmann: Effect ofBand Contact on the Temperature Distribution forDry Friction Clutch, Proceeding of World Academyof Science, Engineering and TechnologyConference, Berlin, pp. 167-177, 2012.[7] Oday I. Abdullah and Josef Schlattmann: TheCorrection Factor for Rate of Energy Generated inthe Friction Clutches under Uniform PressureCondition, J. Adv. Theor. Appl. Mech., Vol. 5, no.6, pp. 277 – 290, 2012.[8] Oday I. Abdullah and Josef Schlattmann: FiniteElement Analysis of Temperature Field inAutomotive Dry Friction Clutch, J. Tribology inIndustry, Vol. 34, No. 4, pp. 206-216, 2012.[9] ANSYS Contact Technology Guide, ANSYSRelease 11.0 Documentation, ANSYS, Inc.[10] Mohr GA: A contact stiffness matrix for finiteelement problems involving external elasticrestraint, Compos Struct., Vol. 12, pp. 189–91,1979.216 13 th International Conference on Tribology – Serbiatrib’13

Serbian TribologySocietySERBIATRIB‘1313 thtInternational Conference onTribologyKragujevac, Serbia, 15 – 17 May 20133Faculty of Engineeringinn KragujevacTHE WAVINESS OFAN ABRASIVEWATERJET GENERAATED SURFACEJ.Baralić 1 , P.Jankovićć 2 , B.Nedić 31 University of Kragujevac, Technical Faculty Čačak, Serbia, , jbaralic@tfc. .kg.ac.rs2 University of Niš, Faculty of Mechanical Engineering, Serbia, jape@masfak.ni.ac.rs3 University of Kragujevac Faculty of mechanical engineering, , Serbia, nedic@kg.ac.rsAbstract: Abrasive water jet generated surface has appearance that is characteristic for all machiningprocedures with the beam of high-energy density. On the surface machined with these procedures,characteristic lines, which are traces resultingg from the passage of beam throughh the workpiece, can beobserved. When machining with abrasive water jet, on the generated surface occurs trace lines andwaviness. The appearance of waviness is more pronouncedin the lower part of thee machined surface. s Theaim of this study is to investigate the effect off traverse speed on the waviness appearance on machinedsurface andthe regularity at whichh the waviness occurs. During investigation, it was observedthat in thesame periodof time, atall traversee speeds andd the thickness of material that has been processed, there isalmost the same numberof waves.Keywords: waviness, trverse speed, abrasive water jet1. INTRODUCTIONAbrasivewater jet machining is one of the latestnon-conventional methods, whichh have recentlybeen increasingly usedin industryfor cuttingg ofvarious materials, fromthose with pronouncedplastic properties, to very-brittle materials. Themain advantage of thismachining process is thatthere is no occurrence of heat affected zone.Abrasivewater jet machining is i based on theprocess of erosion as the primary mechanism forthe removalof materialfrom the workpiece. Bitter[1] has defined erosion of material as damage,which occurs as a result of single impactt ofabrasive particle (whichh is located in i a fluid and ismoving at high speed).Hutchings [2] has definederosion as abrasive wear in which abrasive particles(which are also located in a fluid and movingrapidly) hitting the surface several times and thusleads to erosion of the material. The traces on thesurface machined with abrasive water jet, resultingfrom erosion of the workpiece material, are visibleunder a microscope. Their direction is changingwith depth of cutting, and follows the directionn ofthe cutting beam through the workpiece. On theabrasive water jet generated surface can be13 th International Conference on Tribology – Serbiatrib’13observed the curved lines that are characteristic forthistype of machining. Also, machinedsurface hasa pronouncedwaviness inn the lower part, and itincreases withh increasing g depth of cut. Theseeirregularities onn machined surface, and the taper ofcut, significantly constrainthe opportunities ofapplication of abrasive a waterjet machining.2. ABRASIVEE WATER JET MACHININGModern installations for abrasivewater jetmachining workk with waterr pressure over 400MPa,while the water jet reaches speedss of up to1000m/s. They consist of a driver and the executiveeandsupportingg components. The driver is a unitthat creates high-pressure water, while an executiveecomponenet is a cutting head. The systemcomponents of f abrasive water jet cutting machineare shown in Figure 1.Theway of working is the following: Hydraulic oilunder pressure of 535 MPa enters the hydrauliccylinder and intensifier. Because of large differenceein the diameters of the intensifier, water pressureereaches the value of 400 MPa or more. Intensifyingsystem is the key k of equipment. Pressure value inintensifier depends on the e ratio of cross sectionss217

area of thecylinders. This ratio usually rangesbetween 1:10 and 1:25 and is a constant. To changethe value of water w pressure, it is necessary tochange the oil pressure inn the hydraulic system.Figure 1. The system components of abrasive water jet cutting machine [3]In order to get a jet of water, whose pressure isapproximately constant,the most commonly duplexreciprocating pump- intensifier (DRP) is used. Thisis actually a complex cylinder which is double-joined with the back towards each other. Whenn theacting pump, with two high pressure cylinders,first cylinder completee stroke, piston rod movesback and compresses the water in the secondcylinder.218Figure 2. Fluctuations of pressure during machiningwith abrasive water jet [4][These cycles are repeated successively. Becausethe water can be compressed 12% under thepressure of 400 MPa, the initial stage of the pistontravel is used to compress the water [4]. Water isnot delivered into the system until the waterpressure reaches the set value. Therefore, theactions of draining water and absorbing water arediscontinuous, and thepressure is fluctuant. Toneutralize the pressure p fluctuations, it is necessaryto install an accumulator behind the intensifier.Figure 2 is a diagram which shows the pressureefluctuations in a conventionall intensifier and the phase-shifted intensifierr cylinders.3. ABRASIVEE WATER JET GENERATEDSURFACEAbrasive water w jet generated surface has acharacteristic appearance and is shown in Figure 3.Curved lines, showing s thee movement of abrasivewater jet through the workpiece material, can beobserved in the Figure 3. The topography of themachined surface and thee appearancee of curvedlines are the most importantmacroscopic propertiesof the surface machined with abrasive water jet.Based on the analysis a of these two characteristicsscanresult in significant informationn about themachining process.Roughness of the surface machined withabrasive water jet increases with increasing depthof cut [5]. Surfaces machined with abrasive waterjet are divided into two areas, fine machining zone(theupper zone) and roughh zone (the lower l zone).Irregularities that occur inn the upper zone of themachined surface are considered as microscopiccirregularities and are in the domain off roughness.Irregularities that occur inn the lower zone of themachined surface have macroscopicdimensions.These irregularities are in the domain of wavinesssin conventionall machining.13 th International Conference C onn Tribology – Serbiatrib’13

[11]. Figure 5 shows s the influence of traverse speedandabrasive mass flow on the waviness.10Abrasive mass flow rate [g/min]40220110440Figure 3. Appearance of the surface machined withabrasive water jetWaviness of the machined surface is alsoo animportant phenomenonn in the machining withabrasive water jet, Figure 4. It was found that thereis a primaryand secondary waviness on the surfacemachined with abrasive water jet [6]. Primarywaviness isa waviness with higher step values,while the secondary waviness has less step value.For all cutting parameters,iff the primarywavelength of the profile at the rough cutting zoneis smaller, the surface is smoother. . Guo [7] foundthat there isa dependence of the waviness step andfocusing tube diameter,while Kovacevic [8] founda correlationbetween waviness stepand a diameterof abrasive particles and the focusing tubediameter.Amplitude [mm]WavinessRoughnessWavelenghtFigure 4. Waviness and surface roughnessThe quality of the surface machined withabrasive water jet isinfluencedwith systemoperational process parameters such as traversespeed, waterjet pressure, abrasive flow rate,standoff distance, depthof cut and angle of cutting[9].Level of influence of certainparameterss isdifferent. The largest number of authors agrees thatthe most influential are traverse speed, s operatingpressure andthe abrasive flow rate. Traverse speedof the jet has a stronginfluence on the surfacefinish of the workpiecee and material removal rateWaviness[ m m]4.86420 120 240 360 480 600Traverse speed [mm/min]Figure 5. Parameter influence on the waviness [10]EXPERIMENTAL INVESTIGATIONSThe aim of o this study was to investigateeinfluence of traverse speedd on the appearance ofprofile waviness on machined surface. Theexperiments were conducted using a Byjet 40222abrasive water jet cutting machine (Bystronic AG,Switzerland). Aluminium A alloy AA-ASTM 60600(EN: AW-6060; ISO: AlMgSi) was used as aworkpiece material. Aluminum and its alloys arecharacterized by b high reflectivity and thermalconductivity. This makes them relatively difficult tocutwith lasers. Therefore, the machining withabrasive water jet is much more acceptable foraluminium alloy.Abrasive water w jet cutting involves a largenumber of variables that affect the cutting resultss(kerf width, taper and surface roughness, waviness).In the present study, s the influence of the followingparameters was investigated: traversee speed (thespeed at which the cutting head moves alongworkpiece during cutting operation) and materialthickness. The other process parameters were keptconstant using the standardd machine configurationn(d 0 =0.3mm;d A =1.02mm;p=380MPa;Q= =350g/min).The samples of aluminum alloy 6 and 10mmmthick were cut with different traverse speedssV= =(200, 300, 400, 4 500, 800 i 1000 mm/min). Onsuch machinedd samples, length at which the tenwaves were observed, was measured, Figure 6.13 th International Conference on Tribology – Serbiatrib’13219

5.CONCLUSIONL w10 L wFigure6. Measuringthe length of ten wavesFor thismeasured values (10L w ), based onFormula 1, the time needed to maketen waves (t 10 )was calculated.0L wt 10 10 VBased onthese values, according to Formula 2,the number of waves that were made in one minute(N) was calculated.t 10 10∙60t 10∙60 s(1)(2)Table 1 shows the images of 6mm thicksamples, obtained withdifferent traverse t speeds,and values for 10L w , t 10 and N.Table 1. 6mmm thick samplesBy analyzing the results, it was observed that achange in traverse speed affects the wavelength andheight on surface machinedd with abrasive water jet.As the traversee speed increases, the higher is thewavelength of profile waviness. Also, machiningwith higher traverse speed results in theincrease ofheight of profile waviness. The most interesting isthe fact that, regardless off the traverse speed andthickness of thee workpiece, , the numberr of waves inoneminute is i approximately the same. Thefrequency of the waves, andd also the curved lines, isapproximately constant c andd in this case ranges from460to 476 waves w per minute. This fact canprobably be explained with fluctuationss in the valueof the operatingg pressure of the abrasive water jet.In order to better explain the relationship betweenpressureoscillationsandprofile wavinesssfrequency, more detailed examination are required.Table 2. 10mm thick t samplesV[mm/min]200300Images ofsamples10L w[mm]t 10[s]N[1/min]Waves are a notclearly markedWaves are a notclearly markedV[mm/min]Images ofsamples10L w[mm]t 10[s]N[1/min]4008.4 1.266 476.2200Waves are notclearly marked50010.6 1.272 471.7300Waves are notclearly marked80016.8 1.266 476.24008.41.26 476.2100021.1 1.266 473.950010.61.272 471.7800100016.821.11.26 476.21.266 473.9ACKNOWLEDGMENTSSThe authors would like to thank to the Ministryof Science andd Technological Development of theRepublic of Serbia. Paper is result of technologicallproject TR35034: "The research of modern non-conventionaltechnologiesapplicationinmanufacturingcompanies with the aimof increasee22013 th International Conference C onn Tribology – Serbiatrib’13

efficiency of use, product quality, reduce of costsand save energy and materials" which is supportedby Ministry of Education and Science of theRepublic of Serbia.REFERENCES[1] J. Bitter: A studu of erosion phenomena, part I,Wear, No. 6, pp. 5-21, 1963.[2] I.M.Hutchings, R.E.Winter, J.E.Field: Severe Solidparticle erosion of metals: the removal of surfacematerial by spherical projectils, in:Proc.R.Sec.Lond.A., 348, London, pp.379-302[3] A. Akkurt: Cut Front Geometry Characterization inCutting Applications of Brass with Abrasive WaterJet, Journal of Mterials Engineering andPerformance, pp. 599-606, 2010.[4] J. Xu, B. You, X. Kong, Design and ExperimentResearch on Abrasive Water-jet Cutting MachineBased on Phased Intensifier, in: Proceedings of the17th World Congress, Seoul, Korea, July 6-11,pp.14846-14851, 2008.[5] J. Baralić, B. Nedić, P. Janković: The traverse speedinfluence on surface roughness in abrasive waterjetcutting applications, in: : Proceedings of the 12thInt.Conf. on Tribology, Kragujevac, Serbia, pp. 349-354, May 11 – 13, 2011.[6] A.W.Momber, R. Kovacevic: Principles of AbrasiveWaterjet Machining, Springer, London, 1998.[7] N.S. Guo, H. Louis, G. Meier: Surface structure andkerf geometry in abrasive water jet cutting:formation and optimization, in: Proc. 7-th Amer.Water Jet Conf., Vol. 1, Water Jet Techn.Ass.,ST.Louis, pp.1-25,1993.[8] R. Kovacevic, R. Mohan, Y.M. Zhang: Cuttingforce dynamics as a tool for surface profilemonitoring in AWJ, ASME J. Engng. For Ind. 117,pp.340-350, 1995.[9] J.Valiček, M. Držík, M. Ohlídal, V. Mádr, L.M.Hlaváč: Optical method for surface analyses andtheir utilization for abrasive liquid jet automation,in: Proc. of the 2001 WJTA American WaterjetConference, Minneapolis, pp. 1–11, 2001.[10] M. Hashish: A modeling study of jet cutting surfacefinish, PED-Vol.58, pp.151-167, 1992.[11] E. Lemma: Maximum depth of cut and mechanicsof erosion in AWJ oscillation cutting of ductilematerials, Journal of Materials ProcessingTechnology, Vol. 160, No. 2, pp.188–197, 2005.13 th International Conference on Tribology – Serbiatrib’13 221

Serbian TribologySocietySERBIATRIB ‘1313 th International Conference onTribologyKragujevac, Serbia, 15 – 17 May 2013Faculty of Engineeringin KragujevacEFFECT OF REFRACTORY ELEMENTS ON WEAR INTENSITYOF THE SURFACE LAYERS IN THE ABRASIVE SOIL MASSJ. Napiórkowski 1 , P. Drożyner 1 , P. Szczyglak 11 University of Warmia and Mazury in Olsztyn, Polandnapj@uwm.edu.pl, przemyslaw.drozyner@uwm.edu.pl, szczypio@uwm.edu.plAbstract: The paper presents the results of researches on tthe influence of vanadium and niobium on wearintensity of the surface layers in soil mass. The investigations were conducted in the laboratory conditions ona wearing machine MZWM -1. The study adopted two types of surface layers obtained by welding withadditon of niobium and vanadium. The resulting layers were tested in three kinds of soil masses: loamysand, ligh clay and ordinary clay. Significant differences in the wear between the layers, depending on thetype of soil were observed.Keywords: weld overlay layer, vanadium, niobium, soil abrasive mass, the wear process.1. INTRODUCTIONWear in abrasive soil mass is a natural processof destruction which intensity depends on kind ofsoil [2,5]. Significant here is the occurrence ofvarious phenomena of wear, depending on therandom changes in the soil, the working processparameters and material properties used in theworking elements. The choice of material forelements work in soil with optimal properties forspecific environmental conditions should bepreceded by an analysis of the nature and type ofwear on the surface and the surface layer [1,4]. Thecriterion for the selection of the material is chosenprimarily on the basis of well-definedheterogeneous phase composition of the layerstructure. It has been found that the abrasive wearresistance is the sum of the individual resistances[3].In the case of iron alloys it is just a matter ofcarbide phase. Selection of carbide must take intoaccount the specific characteristics of theinteraction between Fe-Cr-C and carbide formerselements. Some of them are refractory elementssuch as vanadium and niobium. These elementsmelting at very high temperatures are quite difficultto obtain in the pure state. In chemical terms arerelatively unreactive, and their reactivity decreaseswith increasing atomic number. Most of theniobium is used in the form of ferrosilicon, andmay be used in the form of NbC carbides. Onlyabout 6% of the total production of niobium isintended for Nb alloys production. Niobium has arelatively low solubility in iron- a (alpha), and ironγ(gamma). Niobium is added to the surface layer inan amount up to 10%. The addition of niobium tothe steel and heat resistance alloys gives the effectof precipitation hardening by intermetalliccompounds or by NbC. Pure vanadium has goodplasticity, it is easily workable and have a goodwelding properties under argon atmosphere. It isresistant to corrosion and influence of alkali. Thevanadium in the steel forms a very hard VCcarbides in combination with the resistance totempering heat makes it is used in engineeringconstructions.Aim of this study is to analyze the wear processin the soil mass in the context of the construction ofthe abrasive surface layers containing refractoryelements V and Nb.2. METHODSThe laboratory researches have beenconducted on a wear machine "spinning mass" type.The sample was a rectangular prism withdimensions 30x25x10 mm, cut from the weldoverlay padding (with additional materials) on steel222 13 th International Conference on Tribology – Serbiatrib’13

38GSA. The chemical composition, determined bymethods of classical chemistry, is as follows:C - 0.38%, Mn - 1.07%, Si - 1.17%, P - 0.028%,S - 0.02%, Cr - 0 , 18% Cu - 0.16% Al - 0.022%.The microstructure of the steel: martensite withbainite and troostyt. At the same time two samplesof one of each type were placed in the machine.The chemical composition of the surface layers areshown in Table 1.Table 1. Chemical composition of layers testedRootChemical composition [%]NbVC Coal 5.2 5Si Silicon 2.2 1.5Cr Chrome 29 23Nb Niobium 6.8 -V Vanadium - 10Other - 3.5 -Fe Iron rest restEach sample underwent a total of 20 000 meterswith speed of about 1.7 m/s. Measurement of themass of the sample was performed at each2 000 meters with use of laboratory scale withaccuracy of 0.0001 g, after the cleaning in anultrasonic cleaner. At that time the mass of soilwere exchanged with a new one. Samples hadoscillating movement.The study was conducted in three types ofabrasive soil mass (according to USDA) in loamysand, light clay and ordinary clay. Characteristicsare shown in Table 2. Granulometric evaluationwas performed using a laser particle size meter+ Hydro Mastersizer 2000. Humidity of the soilwas determined by measuring the weight of thedried solid at a temperature of 105 C. The studywas conducted on humid mass.Table 2. Characteristics of soil pulpGroupgrain size2,0-0,05mmdiametersandFraction [%]0,05-0,002 mmdiameterdustClay below0.002 mmin diameterHumidityweight%Loamysand77, 48 20.83 1.69 9-11Light clay 56.48 30.83 12.69 12-13Ordinaryclay26.86 48.62 24.52 13-15Microscopic examination was performed bylight microscopy methods - Neophot 52 microscopecoupled with a digital camera Visitron Systems.Scanning microscope JEOL JSM - 5800 LVcoupled with X-ray microanalyser Oxford ISISLINK - 300 were used for scanning electronmicroscopy and chemical compositionmicroanalysis. Samples were digested with 3%HNO3 (Mi1Fe) and electro-chromic acid.Hardness measurements of the surface layerwere determined by Vickers method in accordancewith DIN EN ISO 6507-1:1999. Measurementswere carried out with load of 1 kg (9.807 N) actingduring the 15s. To quantify the wear it wasassumed unit weight wear related to 1cm2 abradedsurfaces and road friction. Figure 1 presents wearmachine "spinning mass" type which was usedduring experiment.Figure 1. Photo of wear machine "spinning mass" type3. RESULTSFigures 2 and 3 shows macroscopic images ofthe construction of the layer containing Nb and V.Figure 2. Macroscopic picture of the construction of thelayers with Nb content. Visible traces of grinding (1).In the right part of the layer - macro cracks (2) and (3).WN – weld overlay, MP - pad material.13 th International Conference on Tribology – Serbiatrib’13 223

Figures 6 and 7 show the characteristics of thesurface layer containing vanadium. Outside thefusion zone, layer has a microstructure consistingof ledeburite, primary carbides of chromium andvanadium carbides.Figure 3. Macroscopic picture of the construction ofthe layers with V content. Visible traces of grinding (1).No evidences of macro cracks .WN – weld overlay, MP - pad material.Figures 4 and 5 show microscopic structure ofNiobium-containing layer. In addition to the narrowstrip of ferritic alloy, weld overlay hashomogeneous structure. Large initial separation ofchromium carbides and niobium carbide are visibleon background of ledeburite (Fig. 11).Figure 6. Microscopic images of the surface layermaterial "V". Ferrite tungsten alloy (mixture ofconstruction ledeburitic) with unevenly spaced primarychromium carbide precipitates (1) and fine carbides ofvanadium (2).Figure 4. Microscopic image of material "Nb" fusionzone. At the junction of the weld metal pad materialvisible "bar" of ferrite alloy from which crystallizeddendritic deposit separating the phases (1). In thematerial "pad" the microstructure of ferrite grains withbright dark areas of perlite. WN - layer of the deposit,FS - ferrite alloy, MP - pad material.Figure 5. The microstructure of the weld layer material"Nb" of chromium carbide precipitates (1) and niobium(2).Figure 7. Microscopic image of the plastic layer "V".Ledeburitic mixture of ferrite alloy and chromiumcarbides M7C3 type (1) and vanadium carbides (2).Table 2. Distribution of layers hardnessType of layer Material "Nb" Material "V"Hardness HV 10 HV 100.5 694 7831.0 676 7771.5 643 7332.0 638 7222.5 612 6923.0 593 6543.5 5 of 85 6394.0 540 6414.5 541 2665.0 550 2565.5 286 -The distance from the surface[mm]As is clear from the measurements of hardness(table 2) test layers have diverse hardness on thecross section. The layer of vanadium content isharder for almost 90 units of the layer containing224 13 th International Conference on Tribology – Serbiatrib’13

niobium. It should be noted that for both layersgradual decrease in cross-sectional hardness wasobserved. Figure 8-11 presents the results of wearin different abrasive masses.Figure 8. The course of wear layers in loamy sand.Figure 9. Mileage wear light layers in clay.Figure 10. Light clay wear layers in the clay normal4. CONCLUSIONThe results of the research carried out in threedifferent soil masses showed the complexity ofprocess of the abrasive wear in tribological discretenodes. The highest wear was recorded in layers ofsandy soil and it was higher by more than 2.5 timeshigher than in clay soils. Wear of layer withvanadium content was about 30% smaller than thelayer containing niobium. Causes of dependenceshould be sought in the mechanism of wear. In theevent of wear in the abrasive mass containing largeamounts of silica, there was the greatest contactwith the abraded area. A different analysis isrequired for results obtained during friction in claysoil masses. The friction in these soils had adifferent course than the sandy soil. Clay soil tendsto form aggregates of soil, hence there are many airpores, and thus much more discontinuities frictionsurface than in the case of sandy soil. In addition,the impact of the clay and dust on the process ofwear is negligible, but in combination with otherfactions can be multiplied. Greater wear in clayhard was observed due to the fact that the weight ofthe formed soil aggregates was greater than a lightloam. The intensity of wear for layers of light andordinary clay remained at the same level valueswherein layer containing vanadium wear lessintense than the layer containing niobium.REFERENCES[1] M.F. Buchely, J.C. Gutierrez, L.M. Le´on, A. Toro:The effect of microstructure on abrasive wear ofhardfacing alloys, Wear, 259, pp. 52–61, 2005.[2] J. Coronado: Effect of Abrasive Size on Wear,Reserch Group of Fatigue and Surfaces, MechanicalEngineering School, Universidad del Valle, CaliColombia, 2011.[3] J. Napiórkowski, L. Pękalska, G. Pękalski: Structureof material and its wear resistance in the soil,Tribologia 6, pp. 871-879, 1998.[4] J. Napiórkowski, G. Pękalski, K. Kołakowski:Badanie struktur i zużywania powłok napawanych wglebowej masie ściernej, Tribologia 3/2012,pp. 111-118, 2012.[5] J. Napiórkowski, P. Drożyner, K. Kołakowski, P.Mikołajczak: Analysis of tribological processes ofconstruction materials in the soil mass abrasivewear process, International Virtual Journal forScience, Technics and Innovations for the Industryyear VI, 11/2012, ISSN 1313-0226, pp. 27-29, 2012.Figure 11. Comparison of intensity of layers dependingon the type of mass13 th International Conference on Tribology – Serbiatrib’13 225

Serbian TribologySocietySERBIATRIB ‘1313 th International Conference onTribologyKragujevac, Serbia, 15 – 17 May 2013Faculty of Engineeringin KragujevacEXPERIMENTAL ANALYSIS OF TOOTH HEIGHT CHANGINGAT TIMING BELTSBlaža Stojanović 1 , Lozica Ivanović 1 , Andreja Ilić, Ivan Miletić 11 Faculty of Engineering, University of Kragujevac, Serbia, blaza@kg.ac.rs, lozica@kg.ac.rs, imiletic@kg.ac.rsAbstract: Timing belt drives present relatively new power transmitters that transmit power by friction andform contact. Load at timing belts is under direct influence of active surface of timing belt tooth. Duringexploitation, the height of tooth at timing belt decrease and, by that, reduction of active surface is provoked,load increases and working life decreases. Besides of tooth height, changing of tooth width at timing belt,also presents very important factor. The tribomechanical system at pulley teeth - timing belt teeth withheight changing during exploitation is analyzed in this paper. Experimental testing of tribologicalcharacteristics was done at custom design and made testing device at Center for power transmission atFaculty of Engineering in Kragujevac.Keywords: timing belt, tribomechanical system, timing belt teeth, friction, testing.1. INTRODUCTIONWorking life and reliability of timing belttransmitters are highly influenced by timing beltgeometrical dimensions that means timing beltpitch, its tooth height and width. Timing belts aremade of polymer materials reinforced by metallicmaterials and their dimensions vary duringexploitation. Highest variation of geometricaldimensions occurred during running-in period ofnew belt. During this period, changing of timingbelt pitch is highest, primarily influenced by plasticdeformations of side surfaces of its tooth. Afterrunning-in period, changings of geometricalcharacteristics are linear with same trend forconsidered values [1-4].The contacts of timing belt and pulley causechanging of geometrical properties. On the basis ofkinematic analysis of pulley and belt contact indetails, following three tribomechanical systemscan be identificated [5, 6]:1. belt teeth – pulley teeth2. side surface of belt – pulley rim3. inter teeth belt space – head of teeth pulleyFriction force in tribomechanical system beltteeth – pulley teeth is dominant factor in case oftooth height changing analyses.2. TRIBOMECHANICAL SYSTEM BELTTEETH – PULLEY TEETHDuring mashing of belt teeth with pulley teethcontact between side surfaces is done. The contactis at line, for the beginning, when belt teeth come inmash with pulley teeth. The mashing starts withimpact of belt teeth and pulley teeth. The belt teethdeform, due to its elastic properties and, by that,enlargement of contact surface is done. Afterenlargement of contact surface, and rotations of beltand pulley, belt teeth slide on side surface of pulley,when rolling with sliding type of friction ishappened.The value of friction force decreases withlength of sliding path, so its maximal value is at thebase of belt teeth (Fig. 1). Simultaneously, actingpoint of resulting component of normal forcemoves from head to base of the teeth. Normal forcevaries with parabolic dependence:NN ( ) 2i=− ⋅ l− l + N (1)lmax2 t max,twhere is:Nmax- maximal value of normal force,l - length of tooth profile andl - length of sliding path.t226 13 th International Conference on Tribology – Serbiatrib’13

Figure 1. Friction force on side surface of belt teethFriction force on side surface of belt teeth can bedetermined by following relation:where is:Foi⋅ µF = Ni⋅ µti=(2)cos βN i - normal force on belt teeth,μ - friction coefficient,( 2)F oi - peripheral force that act on belt teeth andβ - belt profile angle.3. TESTING OF TIMING BELTExperimental testing of tribologicalcharacteristics was done at custom design and madetesting device with open power loop at Center forpower transmissions at Faculty of Engineering inKragujevac [7-9]. Basic elements of testing deviceare (Fig. 2):1. driving unit,2. Cardan transmitter,3. input shaft with measuring devices,4. sensor for input shaft number of rotation,5. torque sensor on input shaft,6. considered power transmitter (timing beltpulley),7. output shaft,8. mechanical brake,9. tension mechanism and10. signal amplifier.Figure 2. Device for timing belt testing.Driving unit, type KR-11/2C (37-180 rpm -1 ),consists of electromotor (1) type ZKT90S-4 (totallyenclosed single phase asynchronous motor withcage rotor with thermal protection, size 90L, 4-poletype), friction power transmitter, and gear reductor.Design solution provides automatic regulation ofpressure between friction discs and compensation ofaxial gap due to wear. Changing of number ofrotations per minute is done manually, by rotationof wheel that by coupling of gear and bar, radially(vertically) move electromotor with conical frictiondisc from friction wheel. Driving unit (1) and inputshaft (3) are connected by Cardan transmitter (2).Input shaft (3) is design in the way to beelastically deformed under maximal torque load.Inductive sensor of number of rotations per minute(4), type MA1 is placed on input shaft, so as torquetransducer (5) that is formed of strain gauges andsignal transmitter MT2555A that is mounted byspecial adapter with battery compartment BK2801A.Input and output shafts (7) are connected byconsidered power transmitter (6), means timing belt- pulley system. Tension of timing belt is done bythe tension mechanism (9) with external threadedspindle. By spindle rotations the movements ofplate with output shaft and mechanical brake aredone.Mechanical brake is specially designed for openpower loop (Fig. 3). Breaking is done by acting ofbreaking pads on both sides of the disc. Regulationof force and torque is done manually by the meansof spring and screw.Mechanical brake obtain certain braking torque,means load torque on output shaft of timing belt –pulley power transmitter. Value of torque ispresented on digital display of the signal amplifierthat gets signal from measuring device on shaft bysignal transmitter EV2510A. The number ofrotations per minute of input shaft is alsodisplayed on amplifier gain that gets signal frominductive sensor and impulse receiver DV2556.13 th International Conference on Tribology – Serbiatrib’13 227

Working regime of input shaft at powertransmitter is measured and regulated inpresented way.Figure 3. Mechanical BrakeBy adaptations of joining elements with drivingunit from one side and output shaft equipped withmeasuring devices testing of various types of powertransmitters can be done on presented equipmentwith limitations in dimensions and load.4. TESTING RESULTSMeasuring of geometrical dimensions is done atZastava tool factory, Department of quality. Inorder to provide relevant analysis measuring of thefollowing values are done at eight teeth of timingbelt (Fig. 4):• pitch ( h ),• belt width ( b ) ,• distance between belt teeth ( t 1)and• belt height ( t ) .Figure 5. Measuring device –DIGIMARResults of measuring of the belt height changingduring exploitation at eight considered tooth arepresented at Tab. 1 by values and at Fig. 6 bydiagrams.Table 1. Change of timing belt height ∆ t = t − t [ μm ]o∆ tBelt teeth1 2 3 4 5 6 7 85 52 75 34 23 17 3 32 1310 53 77 38 30 23 15 39 1520 62 83 46 41 36 17 42 3050 65 96 49 52 57 26 53 51100 65 98 50 53 57 33 53 53150 67 101 54 53 57 35 57 53200 68 105 58 61 57 55 74 55250 69 109 58 65 67 63 82 63300 76 125 63 67 67 66 84 63Exploitationperiod [h]Figure 4. Basic geometrical properties of timing beltChange of timing belt height is considered inthis paper. Belt height is distance between head ofbelt tooth and backing surface. Measuring was doneby DIGIMAR measuring device (Fig. 5). Changingof belt height ( ∆ t)during testing can be calculatedby following relation:∆ t = t − t,where is:t - measured value of belt height andt - starting height of the belt.ooFigure 6. Changing of tooth timing belt height inexploitationEvaluations of the obtained results implicate thatbelt height decreases monotonely duringexploitation. During running-in period, that lasts forapproximately 20 hours, this changing is verysignificant and it is happening on all of consideredeight belt tooth. The changing during running-inperiod is caused by deformations of the belt, itspitch and width decrease. During period ofexploitation due to normal wear belt heightdecreases. During period of 20 hours to 50 hours of228 13 th International Conference on Tribology – Serbiatrib’13

exploitation this changing is significant. After 50hours of exploitation till 200 hours of exploitation,the very fast changing occurred at most of thetooth. Plastic deformation occurred during runninginperiod. Due to the fact that those deformationsare small during period of normal wear, cylindricalwear of tooth head are not significant, so heightchangings are small. After 200 hours ofexploitation changing of tooth heights aresignificant.On the basis of the further analyses, conclusionthat changings of all values are subjected to samedecrease function is implicated. But, on the basis ofthe analysis in details it is implicated that changingof belt height is bigger than changing of inter toothspace width. This fact leads to decrease of activetooth heights that are in contact with tooth of thepulley. If the reduction of belt width after 150hours is taken into consideration, it is implicatedthat value of nominal surface side of timing belttooth also decrease. As timing belt - pulleytransmitters transmit power by form contact andfriction, increase of timing belt pitch and decreaseof nominal active surface of tooth all together causefailures in exploitation of those transmitters.5. CONCLUSIONThe basic tribomechanical systems at timing belt– pulley transmitter are: timing belt teeth – pulleyteeth, belt side – rim of the pulley, inter space oftiming belt tooth – head of pulley tooth. Frictionforces are highest at side surfaces of timing belttooth and pulley tooth. Directions and values ofthose forces are under direct influence of meshingkinematics of timing belt transmitters.As the consequence of friction at side and headsurface of tooth, the reduction of belt height iscaused. Decrease of belt width and reduction of itsheight cause decrease of active contact surface, thatfurther cause increase of loads at tooth andsimultaneous decrease of transmitter coefficient ofefficiency. Average changing of timing belt toothheight relatively to starting value is 3.14 %.Considered changing of geometrical propertiesof timing belt causes significant influence toreliability and working life of timing belttransmitters.REFERENCES[1] R. Perneder, I. Osborne: Handbook Timing Belts:Princilpes, Calculations, Applications, Springer-Verlag Berlin Heidelberg, 2012.[2] S. Tanasijević: Mechanical drives: chain drives,timing belt drives, cardanic drives (in Serbian),Yugoslav tribological society, Faculty ofengineering from Kragujevac, 1994.[3] B. Stojanović, S. Tanasijević, N. Marjanović, L.Ivanović, M. Blagojević: Wear as the Criterion ofMechanical Transmitters Working Life, J BalkTribol Assoc 17, pp. 215-222, 2011.[4] B. Stojanović, M. Babić, N. Marjanović, L.Ivanović, A. Ilić: Tribomechanical Systems inMechanical Power Transmitters, J Balk TribolAssoc 18, pp. 497-506, 2012.[5] B. Stojanović, S. Tanasijević, N. Miloradović:Tribomechanical Systems in Timing Belt Drives, JBalk Tribol Assoc 15, pp. 465-473, 2009.[6] B. Stojanović, N. Miloradović, N. Marjanović, M.Blagojević, A. Marinković: Wear of Timing BeltDrives, J Balk Tribol Assoc 17, pp. 206-214, 2011.[7] B. Stojanović, N. Miloradović, N. Marjanović, M.Blagojević, L. Ivanović: Length Variation ofToothed Belt During Exploitation, Strojniškivestnik, Vol. 57, No. 9, pp. 648-654, 2011.[8] B. Stojanović, L. Ivanović, M. Blagojević: Frictionand Wear in Timing Belt Drives, Tribology inIndustry 32, pp. 33-40, 2010.[9] B. Stojanović, L. Ivanović, N. Miloradović: Testingof Timing Belt Drives, IMK-14 37, pp. 77-80, 2010.13 th International Conference on Tribology – Serbiatrib’13 229

Serbian TribologySocietySERBIATRIB ‘1313 th International Conference onTribologyKragujevac, Serbia, 15 – 17 May 2013Faculty of Engineeringin KragujevacCYCLO DRIVE EFFICIENCYTihomir Mačkić 1 , Živko Babić 2 , Nenad Kostić 3 , Mirko Blagojević 41 Faculty of Mechanical Engineering Banja Luka, Bosnia and Herzegovina, tihomir.mackic@unibl.rs2 Faculty of Mechanical Engineering Banja Luka, Bosnia and Herzegovina, zivkobabic@unibl.rs3 Faculty of Engineering University of Kragujevac, Serbia, nkostic@kg.ac.rs4 Faculty of Engineering University of Kragujevac, Serbia, mirkob@kg.ac.rsAbstract: Cyclo drives have a many good characteristics: high gear ratio, compact design, two-thirds of itsreduction components in contact at all times, reliability and long life in the most severe applications,minimal vibration, low noise, low backlash and extended operational life, high power density, wide variety ofinputs available,… One of the most important its characteristics is high efficiency.Two methods for determining of cyclo drive efficiency are presented in this paper. Their complete analyticalmodels are defined. The influence of various parameters on the cyclo drive efficiency is also analyzed(power, rotational angle of input shaft, gear ratio, …). The calculation of the cyclo drive efficiency by bothmethods is done for the real one-stage cyclo speed reducer. Concluding remarks and directions for futurework are presented at the end of the paper.Keywords: cyclo drive, cycloid, gear, power losses, efficiency1. INTRODUCTIONCycloidal speed reducer belong the group ofplanetary drives (Figure 1). Because of very widearea of application, production of cyclo drives hasgrowing character and wide area of application:processing equipment, conveyors, presses, mixers,food industry, automotive plants, spinningmachines, cranes,…The most important working characteristics ofcyclo drives are: wide range of possible gear ratios,quiet and reliable work, low level of noise andvibrations, exceptionally compact design, highefficiency rate, ... Lehmann gave the basicinformation about cycloidal gearing, [1]. Thedynamic behavior of a cyclo drives is presented inRefs. [2,3]. Kosse investigated the hysteresisphenomenon in cyclo drives and dampingproperties derived from dickey curves undertorsional impact load, [4]. Liu and other generateda new type of double-enveloping cyclo drive andcalculated the torsional stiffness, [5]. The influenceof friction on contact forces distribution ispresented in papers [6,7]. Sensiger developed a newmethod for cycloidal gear profile, efficiency andstress optimization, [8]. Chmurawa and Lokiecpresented the inside meshing and force distributionsof cycloid disk with modified profile, [9]. A newconcept of a two-stage cyclo drive is presented inpaper [10].Friction and wear have a greatest impact on thecyclo drive efficiency, [11, 12]. The calculation ofthe cyclo drive efficiency by two method (Malhotraand Gorla) for the real one-stage cyclo speedreducer is presented in this paper.Figure 1. Cyclo drive230 13 th International Conference on Tribology – Serbiatrib’13

2. EFFICIENCY OF CYCLO DRIVEEfficiency of cyclo drive primarily depends onthe resistance due to friction between the elementsof cyclo drive. Two methods for determining ofcyclo drive efficiency are presented in this paper:Malhotra method [11] and Gorla method [12].2.1 Malhotra method for calculating of cyclodrive efficiency, [11]The various sources of power loss in a cyclodrive are:• Rolling friction in the mounting of thecycloid disc on the input shaft,• Rolling friction between output rollers andholes in the cycloid disc,• Rolling friction between housing rollersand the cycloid disc,• Sliding friction in the mounting of theoutput rollers,• Sliding friction in the mounting of thehousing rollers.Design parameters are presented on Figure 2,and loads on Figure 3.W =⎛+ n⎜f+r2⎝+2πn∫ dW =0fs1fr1⋅ d2π⋅ Dm⋅ n n∫ ( θ)dθ +DFr 0 E2π⎞ n q⎟ ∫ ∑ F ( θ)dθ +Kj⎠ 0 j = 12πVK2⎛ ⋅ d ⎞ n q( u + 1) ⎜f+f s2 O ⎟ ∫ ∑ F ( θ)dθ(1)CR ⎜ r3 ⎟⎝2Kj⎠ 0 j = 1where: f r1 , f r2 and f r3 are lever arms of rollingfriction, f s1 and f s2 are sliding friction coefficients,D m is mean diameter of input shaft bearing, D r isinput shaft bearing rollers diameter, F E is bearingreaction, d VK is diameter of outpit mechanism pins,F Kj is force between output roller j and cycloid disc,q is number of output rollers, u CR is cyclo driveratio and d 0 is diameter of housing pins.Figure 3. Loads on cycloid discThe overall efficiency of cyclo drive is then:M a 2π− Wη = (2)M 2πwhere M a is input torque.aFigure 2. Geometry of cyclo driveFor the elemental rotation dθ of the cycloid disc,the rotations of the input shaft, output rollers andhousing rollers are n⋅dθ, n⋅dθ and (n+1)⋅dθ,respectively. The frictional work per rotation of theinput shaft can be determined as:2.2 Gorla method for calculating of cyclo driveefficiency, [12]Power loss due to the bearing friction could becomputed by means of the following equation:WMa M a( ω − ω )= ⋅(3)innerouterwhere: ω inner is bearing inner race speed and ω outeris bearing outer race speed.Power loss due to the friction between the pinsof the output shaft and the holes of the cycloid discis:13 th International Conference on Tribology – Serbiatrib’13 231

Ws= ∑KFKj Kjj = 1f⋅⋅υKj(4)where: f Kj is friction coefficient between the pins ofthe output shaft and the holes of the cycloid discand υ Kj is sliding speed between the pins of theoutput shaft and the holes of the cycloid disc.Power loss due to friction between thecylindrical rollers, the surface of the ring gear andtheir housing in the planet wheel is:presented in the paper, too (Figure 4, Figure 5 andFigure 6).Dependence of cyclo drive efficiency on inputpower is presented on Figure 4. Input power wasvaried in range from 3 kW to 5 kW. Increasing theinput power, cyclo drive efficiency is increasing,too (from 93% to 96%). Values of efficiencycalculated by Malhotra [11] and Gorla [12] methodare very similar.WNn= ∑ ρ ⋅ sin( arctg(f )) ⋅Nii = 1FNi⋅ ωNi(5)where: ρ is radius of cylindrical roller, f Ni is frictioncoefficient between the cylindrical rollers and theirhouses and ω Ni is relative rotational speed betweenthe cylindrical rollers and the cycloid disc.Vertical component of force F Ni is calculatedbased on following expression:The efficiency of cyclo drive can be calculatedas:P − ( W + W + W )EM Ma Kη =N (6)PEMFigure 4. Dependence of cyclo drive efficiencyon input power3. CALCULATION OF CYCLO DRIVEEFFICIENCYThe efficiency of cyclo drive by two presentedmethod is has calculated for input parameters inTable 1.Table 1. Cyclo drive parametersMarkP EMValue4,0 kWn EM 1420 min -1u CR 13f r1 , f r2 , f r3 f r1 = f r2 = f r3 = 0,003f S1 , f S2 f S2 = f S2 = 0,03d 08 mmD 014 mmq 8d vk8 mm14 mmD vkCyclo drive efficiency is calculated in programcreated in MATLAB. Values of cyclo driveefficiency are:• η = 94,55% (Malhotra method),• η = 95,03% (Gorla method).Analysis of the influence of input power P EM ,input number of revolutions n EM and gear ratio u CRon cyclo drive efficiency by both method isFigure 5. Dependence of cyclo drive efficiencyon input number of revolutionsDependences of cyclo drive efficiency frominput number of revolutions is presented on Figure5. Number of revolutions was varied from 1180min -1 to 1660 min -1 . Increasing the input numberof revolutions, cyclo drive efficiency decreases (forboth method, Figure 5).Dependence of cyclo drive efficiency on gearratio is presented on Figure 6. Gear ratio was variedin range from 11 to 16. Increasing the gear ratio(Malhotra method), cyclo drive efficiencydecreases from 95% to 94%. For Gorla method,cyclo drive efficiency increases from 94,8% to95,3%.232 13 th International Conference on Tribology – Serbiatrib’13

Figure 6. Dependence of cyclo drive efficiencyon gear ratio4. CONCLUSIONTwo methods for calculating of cyclo driveefficiency are presented in this paper (Malhotra andGorla method). Their complete analytical modelsare defined. The calculation of the cyclo driveefficiency by both methods is done for the real onestagecyclo speed reducer.By analyzing the results, it can be concluded thenext:• One-stage cyclo drive has very highefficiency,• Both method (Malhotra and Gorla) havevery similar values for efficiency,• With increasing of input power, cyclo driveefficiency is increasing too,• With increasing of input number ofrevolutions, cyclo drive efficiencydecreases,• With increasing of gear ratio, cyclo driveefficiency decreases (Malhotra method), orincreases, (Gorla method).REFERENCES[1] M. Lehmann: Calculation and measurement offorces acting on cycloidal speed reducer (inGerman), PhD Thesis, Technical UniversityMunich, Germany, 1976.[2] S.V. Thube, T.R. Bobak: Dynamic analysis of acycloidal gearbox using finite element method,AGMA Technical Paper, 2012.[3] M. Blagojević, V. Nikolić - Stanojević, N.Marjanović, Lj. Veljović: Analysis of cycloid drivedynamic behavior, Scientific Technical Review,Vol. LIX, No. 1, pp. 52-56, 2009.[4] V. Kosse: Using hysteresis loop and torsional shockloading to asses damping and efficiency of cyclodrives, 14 th International Congress on Sound andVibration, 9-12. July, 2007, Cairns, Australia.[5] J. Liu, S. Matsumura, B. Chen, H. Houjoh:Torsional stiffness calculation of double-envelopingcycloid drive, Journal of Advenced MechanicalDesign, Systems and Manufacturing 6, pp. 2-14,2012.[6] M. Blagojević, N. Marjanović, B. Stojanović, Z.Đorđević, M. Kočić: Influence of friction on theforce distribution at cycloidal speed reducer, 12 thInternational Conference on Tribology, 11-13.5.2011, Kragujevac, pp. 226-229.[7] M. Blagojević, M. Kočić, N. Marjanović, B.Stojanović, Z. Đorđević, L. Ivanović, V.Marjanović: Influence of the friction on thecycloidal speed reducer efficiency, Journal of theBalkan Tribological Association, Vol. 18, No. 2, pp.217-227, 2012.[8] J. Sensiger: Unified Approach to Cycloid DriveProfile, Stress, and Efficiency Optimization, Journalof Mechanical Design (ASME), Vol. 132, 2010.[9] M. Chmurawa, A. Lokiec: Distribution of loads incycloidal planetary gear (CYCLO) includingmodification of equidistant, in: 16 th EuropianADAMS user conference, 2001, Berchtesgaden,Germany.[10] M. Blagojević, N. Marjanović, Z. Đorđević, B.Stojanović, A. Dišić: A new design of a two-stagecycloidal speed reducer, Journal of MechanicalDesign (ASME) 133, 2011.[11] S.K. Malhotra, M.A. Parameswaran: Analysis of acycloidal speed reducer, Mechanism and MachineTheory 18, pp. 491-499, 1983.[12] C. Gorla, P. Davoli, F. Rosa, C. Longoni, F.Chiozzi, A. Samarani: Theoretical and experimentalanalysis of a cycloidal speed reducer, Journal ofMechanical Design (ASME) 130, 2008.13 th International Conference on Tribology – Serbiatrib’13 233

Serbian TribologySocietySERBIATRIB ‘1313 th International Conference onTribologyKragujevac, Serbia, 15 – 17 May 2013Faculty of Engineeringin KragujevacTRIBOLOGICAL ASPECTS OF THE PROCESS OF WINDINGTHE STEEL ROPE AROUND THE WINCH DRUMMiloš Matejić 1 , Mirko Blagojević 1 , Vesna Marjanović 1 , Rodoljub Vujanac 1 , Boban Simić 21 Faculty of Engineering, University of Kragujevac, Serbia,mmatejic@kg.ac.rs, mirkob@kg.ac.rs, vmarjanovic@kg.ac.rs, vujanac@kg.ac.rs2 D.O.O RAPP Zastava, Serbia, bsi@rappzastava.comAbstract: Proper winding of the steel rope around the winch drum is great importance, mostly for:prolonging the service life of the rope, reduction of deformations of the body and the sides of the drum if thewinding of the rope is multilayered, increasing of the safety factors, easier unwinding of the rope whilelowering the load, even running of the drive unit, etc. The focus of this paper is on the analysis of the frictionwhich occurs in the process of winding and unwinding the rope around the winch drum. Friction force is inits highest intensity when the rope passes from one layer to another, if the winding of the rope ismultilayered. As the result of the research, certain mechanisms of winding of the rope from the aspects of thefriction force were obtained, and the affect of the forces on the sides of the drum were analyzedKeywords: winch, drum, rope, friction, friction force.1. INTRODUCTIONModern technological achievements haveenabled a great progress in the ship equipmentindustry. Ship winches, as one of the mostimportant parts of the equipment, are highlydeveloped. Their dimensions, compared to the forcethey use, have been reduced and the degree ofefficiency has increased with the use of variouskinds of compact mechanical gear and drive. Thecapacities for winding of the ropes have beenincreased for any kind of use on the vessels, fromthe oceanographic researches on the bottom of theocean, to catching of the special kinds of fish.A lot of research into vessel winches have beendone: increasing of the degree of efficiency,reduction of dimensions, reduction of the mass ofwinches, improvement of the drive unit regardingthe compactness and the power they use,improvement of the drum for rope-winding as thebasic element of a winch, of the rope-windingsystem, development and analysis of the varioustypes of ropes for various uses, etc.Great attention is paid to examination anddevelopment of the ropes used for winches. Steelropes are mostly used for vessel winches, while thesynthetic ones are rarely used. Different types ofcables are used for winches for oceanographicresearches rather than ropes because of thetransmission of information from the researchdevices from the bottom of the ocean to the vessel.When it comes to the development of the new typesof ropes, great attention is paid both to the outertensile and twisting forces affecting the rope, andthe inner friction forces that occur between thewires of the rope, [1], [2], [3]. When examining theexisting types of ropes, it is very important toexamine the inner friction in the ropes, as well asthe failure mechanism of the ropes, [4], [5]. In thecase where the steel rope cannot be installed on awinch because of its large mass, synthetic ropes areused. Synthetic ropes are still not widely usedbecause they are still in the development andimprovement stages, [6]. Synthetic ropes can bemade with the molten core for the improvement ofits mechanical characteristics.Reduction of the dimensions and the increase inthe degree of efficiency of a winch is mostlyachieved by installing the compact mechanicalgear, gearboxes. Single-stage gearboxes with two,three or more drives are often used for increasingthe compactness of the vessel winch constructions,[7]. In addition, the use of planetary gearboxes is234 13 th International Conference on Tribology – Serbiatrib’13

common for their ability to be installed within thewinch drum.Improvement of the winch drum, as the basicelement of the winch, is based on the examinationof the effect of the forces on the drum,experimentally and using finite element analysis,[8]. Improvement of the drum is done by reducingthe thickness of the material from which the drumis made in noncritical places, while increasing it inthe critical ones, and inserting the requiredstructural stiffeners, [9], [10]. Optimizationmethods are also a possible approach ofimprovement, [11].Improvement of the system for proper windingof the rope around the drum of the vessel winch isbased on its synchronization with the number ofturns of the drum. The synchronization can be donein many ways. Some of the ways are themechanical synchronization with the powertransmission or installing a special driving enginegeneratorfor the system, which is synchronized,with the engine-generator of the winch.In this paper, a mathematical model of windingof a steel rope around the drum is presented. Themathematical model shows the correlation betweenthe friction and pulling forces using geometricalcharacteristics of winding on the winch drum andthe friction coefficient. Following this, the resultsof the algorithm, which are also developed in thispaper, mathematical model as well as comparisoncharacteristics of friction forces for differentcoefficients of friction during winding, are given.Finally conclusions and guidelines for furtherresearch have been presented.2.1 General case of winding of the ropearound the drumIn determining the general case of winding, annth winding on the nth layer is observed. Thewinding lies upon the two windings from theprevious layer (Figure 1).The forces FWR occur as the reaction of therope to the pulling force FW. The friction force Fµalso occurs. Further observation is done on thecross-section where the force FWR acts (Figure 2).In this cross-section, the perpendicular force FNoccurs due to acting of friction force Fµ.Figure 1. General spooling case2. MATHEMATICAL MODEL OF STEELROPE WINDING AROUND THE WINCH-DRUMThe winding of the steel rope around the winchdrum can be single-layered or multilayered. Themathematical model concerns the multilayeredprocess of rope winding, from the first to the lastlayer. The model could be used for defining singlelayeredwinding too, by excluding the upper layersfrom the model. The bevel of the rope due to thewinding is excluded from the mathematical model,because the bevel has an insignificant effect in themajority of winding cases. Also, the assumptionthat during winding the rope acts as an absoluteelastic body is taken, while the wound rope acts asa solid body.Figure 2. Cross-section of ropes in general caseSince the rope is of constant cross-section, thecenters of circumferences, which make the crosssectionof the rope, generally, form an equilateraltriangle. Horizontal component reaction forces ofthe windings of the lower layer FR are annulledbecause of the previous claim. The connectionbetween the perpendicular force FN and thereaction force FR is derived by setting the planarsystem of opposed forces, and the resulting isrelation (3):FN3FR (1)3After determining the reaction, it is necessary todetermine the friction force, which is caused by thetractive force. The friction force was determined by13 th International Conference on Tribology – Serbiatrib’13 235

observing the simplified tensile system of the upperlayer of rope over the lower layer (Figure 3) on asmall angle ±dφ/2 from the cross-section shown onthe Figure 2.The friction force Fµ in the expression (6) wascalculated as the total friction force. Because of theadopted assumption about the symmetry ofwinding, the friction force Fµ is divided into twoequal parts (8) (Figure 1) in order to make arelation between the friction force appearing in thecontact of the upper and lower layers.F W1Fμ1 (8)LC2 e 2.2 Special case of rope winding on the drumFigure 3. Basic part of the ropeFor the elementary part of the rope on the angle±dφ/2, the following equation system is derived:ddx : dF F dF cos F cos 0 (2)WR WRWR2 2d dy : dF F dF sin F sin 0 (3)N WR WRWR2 2Accepting the assumptions for the basic angle±dφ/2 (φ→0): cosdφ/2≈1, sindφ/2≈dφ/2, dFµ =µ·dF N i dF WR dφ = 0, and by solving equations (2)and (3), the connection between the inputtedpulling force F W and the reaction force within therope F WR :μFW FWR e(4)With further solving, the connection between thetractive force F W and perpendicular force F N isobtained, as well as the connection between thefriction force Fµ and the tractive force F W :F W1F 1 (5)N e 1Fμ F 1 (6)W e From expressions (4), (5) and (6) it can be seenthat all resulting values directly depend on thetractive force F W , coefficient of friction µ, and theangle of winding φ.Returning the values from expression (5) toexpression (1) the function of the reaction of therope on the lower layer (7) is derived depending onthe tractive force, coefficient of friction, and theangle of rope winding:3 F F W11 (7)R3 μ e In the special case of rope winding on the drum(figure 4), the crossover of the rope from layer nthto the nth+1 layer is considered. The critical wind isobserved in this case, during the crossover. In thiscase, in the beginning, the contact on the last windof the previous lower layer and the side of the drumoccurs, while in the second wind that contact is lost,and the rope changes to the general winding case.Figure 4. Special rope winding caseEstablishment of the connection between theforces of reaction of the lower layer of the rope F R ,of friction between the layers of the rope Fµ R(Fµ R = F R·µ 2 ), forces of reaction on the side Fµ S(Fµ S = F S·µ 1 ) and perpendicular force F N , in thiscase must be done by introducing the angle α. Thischanges in the interval of 0 ≤ α ≤ π/2, while theangle of the winding changes in the interval of 0 ≤φ ≤ 2π. The connection between angle α and theangle of winding φ is established according to thefact that for the angle φ = 2π [rad] the rope iswinding for the length of the diameter of the roped W, while angle α changes from 0 to π/2.x F cos F sin F 0 (9):R2 RSy : FRsin 2 FRcos FN 1 FS 0 (10)By solving the equations above the followingexpressions are obtained:236 13 th International Conference on Tribology – Serbiatrib’13

FNF (11)R sin cos11 21 22FNF (12)R sin cos11 21 2FNcos 2sin 1 sin cosF (13)S121FNcos 2sin 1 sin cosF (14)S12In the expressions (11), (12), (13) and (14) thefriction coefficient µ 1 is the friction coefficientbetween the rope and the drum of the winch, whileµ 2 is the friction coefficient between the windingsof the rope.3. FRICTION FORCES IN THE WINDINGOF THE STEEL ROPE AROUND THEWINCH DRUMFor this paper a mathematical model has beendeveloped which for given initial parameters ofwinding of the rope gives the friction forcesdiagrams, perpendicular forces, as well as thecomparative friction forces diagram for differentcoefficients of friction. The initial parametersaccording to which the forces were calculated aregiven in Table 1.Table 1. Initial parameters for force examinationSize description Designation ValueDrum length L 100 [mm]Wire diameter d W 10 [mm]Number of windedlayers1122n L 10Load weight m 250 [kg]Friction coefficient µ 1 0,25Friction coefficient µ 2 0,35Mathematical model for calculating the forcevalues has been developed in MS Excel software.The division of the winding angle was made with10, but was converted to radians for easierclarification. All of the output diagrams have thedivision in radians [rad] on their x - axis, and inNewtons [N] on the y – axis.3.1 Spooling of the rope on the first layerIn the spooling of the first layer it ischaracteristic that the friction force occurs only onone rope inlet. Because of the relatively smallfriction coefficient (µ1 = 0,25) during the first layerspooling, the friction force Fµ does not have a rapidincrease in the first two layers (φ ≈ 18 [rad])regarding the greater friction coefficients (Figure5).Friction force [N]3000250020001500100050000 5 10 15 20 25Spooling angle [rad]Figure 5. Friction force Fµ on the first layer dependenceon the spooling angleAt the end of the third layer spool (φ ≈ 18 [rad]),the friction force is almost equal to the pullingforce. It can be said with certainty that after thefourth spool, with the friction coefficient being µ1= 0,25, the whole load of tractive force is carried bythe friction force.3.2 Rope spooling in the crossover from onelayer to anotherThe rope spooling is most critical in thecrossover from one layer to another (Figure 6). Inthat case, only one critical winding is observedduring the crossover from layer n to layer n +1,until the rope turns to the general spooling case. Inthis case, it is very hard to determine the frictioncoefficient between the winding spool and the lastspool on the previous layer because of the changingtrajectory of the rope inlet. When the rope gets tothis position, the friction force appears in thecontact between the side of the drum and the spoolthat is being wound by the rope inlet, but also thefriction force appears in the contact between thelast spool on the previous layer and the spool beingwound. In this case, the assumed coefficient offriction which occurs in the rope inlet is the same asthe friction coefficient which occurs between therope spools (µ2 = 0,35).From the diagram on the Figure 6 which shows:the reaction force of the lower rope layer FR ,friction force between the rope layers FµR ,reaction force on the side FS and friction force onthe side FµS , their changes during the critical spoolin the crossover of the rope from one layer toanother are visible. The lower rope layer reactionforce FR has the steady increasing character fromthe beginning to the end of the critical spooling,and transcends the nominal value of the pullingforce almost by two.Fµ13 th International Conference on Tribology – Serbiatrib’13 237

Forces: Fr, Fµr, Fs, Fµs [N]50004000300020001000FrFµrFsFµsFriction force in line contact [N]14001200100080060040020000 5 10 15 20 25Spooling angle [rad]Fµlc00 1 2 3 4 5 6 7Spooling angle [rad]Figure 6. The reaction force of the lower rope layer FR ,friction force between the rope layers FµR , reactionforce on the side FS and friction force on the side FµSdependence on the spooling angleThe friction force between the layers FµR alsohas the increasing character from the beginning ofthe critical spooling and all the way to the pointwhen the rope crosses to the general spooling case,for it is directly related to the force FR by thefriction coefficient µ2, (12). The reaction force onthe side of the drum FS rises up to the spoolingangle of φ ≈ 1,5 [rad], while after reaching theextreme value it decreases to zero. Its value equalszero at φ ≈ 4, 2 [rad]. The friction force on the sideFµS acts similarly, because it is related to thereaction force on the side through the frictioncoefficient µ1, (12). When the forces FµS and FSreach zero, the perpendicular force transfers to thelast spool of the lower spooling layer exclusivelyby the friction force between the rope windings.3.3 General case of rope spoolingGeneral case of rope spooling for multilayeredspooling has the greatest share in the spoolingprocess. Generally, friction force occurs on the tworope inlets on the contact line of the winding beingspooled and the two spools from the previous layer.In this case the total friction force is divided intotwo equal parts, (8). Friction force FµLC in linecontact on the rope inlet gets close asymptoticallyto the half of the nominal value of the pulling force,(Figure 7). In the case of rope spooling onto thehigher layers, the general case, in difference withfirst layer spooling, the friction force carries thetotal tractive force (φ≈12÷15[rad]) between thesecond and the third layer.Figure 7. Friction force Fµ in general case of ropespooling dependence on the spooling angleIn this case a grater friction coefficient value isadopted from the one used for rope spooling ontothe first layer (µ 2 =0,35).3.4 Comparative diagrams for different frictioncoefficient valuesLastly the comparative friction force diagrams(Figure 8) and perpendicular force (Figure 9)diagrams for friction coefficients µ = 0,2 ÷ 0,4(initial parameters are also taken from Table 1) aregiven. For these resulting diagrams the general ropespooling case was used.Friction force in line contact [N]14001200100080060040020000 5 10 15 20 25Spooling angle [rad]µ = 0,20µ = 0,25µ =0,30µ =0,35µ =0,40Figure 8. Friction force Fµ for different frictioncoefficients dependence on the spooling angleIt can be seen from Figure 8 that for all frictioncoefficient values the friction force has the sameinitial value during spooling. As expected, thehighest number of spools for total transfer of pulingforce to friction force is necessary for the highestfriction coefficient (φ > 12 [rad]), while for thelowest friction coefficient the transfer is possibleonly after the third spooling (φ > 19 [rad]).With perpendicular forces the case is slightlydifferent (Figure 9). Perpendicular forces do notexceed the nominal value of the pulling force forthe value of the friction coefficient of 0,3.Perpendicular forces with the friction coefficient of 0,3 do exceed the value of nominal pullingforce.238 13 th International Conference on Tribology – Serbiatrib’13

Normal force [N]400035003000250020001500100050000 5 10 15 20 25Spooling angle [rad]µ = 0,20µ = 0,25µ =0,30µ =0,35µ =0,40Figure 9. Perpendicular force FN for different frictioncoefficients dependence on spooling angleEven though the perpendicular and frictionforces are in linear correlation by the frictioncoefficient, the differences in their behavior occurin the general spooling case because the frictionforce is distributed to the two components on therope inlets.4. CONCLUSIONUsing the mathematical model and friction forcecalculation, it has been shown that the friction forcein the rope spooling onto the winch drum processdoes not depend only on the friction coefficient, butalso in the position of the rope during the process. Ithas been shown that the greatest friction forcesoccur during the crossing of the rope from one layerto another. With the increase of the frictioncoefficient, the time needed for the pulling tofriction force transfer shortens. For the transfer witha friction coefficient of 0.2 more than four spoolsare required (φ > 24 [rad]), while for the transfer ofthe same pulling force with a coefficient of 0.3 lessthan three spools are needed (φ < 18 [rad]). Thefriction coefficient value depends mostly on therope material as well as on its characteristics, butfor the first layer the material and thecharacteristics of the drum have an equal share. Forthe transfer of smaller masses, ropes with a highfriction coefficient can be used, but this is notadvisable for greater masses.During the crossing of the rope from one layerto another, effect of the rope on the side of thedrum in the interval when the spooling angle is 0

Serbian TribologySocietySERBIATRIB ‘1313 th International Conference onTribologyKragujevac, Serbia, 15 – 17 May 2013Faculty of Engineeringin KragujevacAPPLICATIVE MONITORING OF VEHICLES ENGINE OILPerić Sreten 1 , Nedić Bogdan 2 , Grkić Aleksandar 31 University of Defence in Belgrade, Military Academy, Serbia, sretenperic@yahoo.com2 Faculty of Engineering in Kragujevac, Serbia, nedic@kg.ac.rs3 University of Defence in Belgrade, Military Academy, Serbia, iralex@eunet.rsApstrakt: Confirming the basic causes of failures and their elimination, control of certain phenomena, isdefining proactive maintenance, as a new method that reduces maintenance costs and prolongs the life ofassets. Determination of tribomehanical systems condition has very important role in development of frictiontheory and practice, wear and lubrication. There are used today different physical and chemical methodsand tribology methods for tribomehanical system diagnosis. Experience in technical systems exploitationshoved that the most effective failure prognosis is according to parameters, particles created as result ofwear, which are reliable indicators of wear. Analysis of oil samples which contain particles, created asresults of wear, enable evaluation of system tribology condition in different phases of system exploitation.The paper presents the physical chemical tests in the analysis of oils that are used for the assessment of hiscondition. Furthermore the results of experimental research of physical chemical characteristics engines oilwas sampled from engines of vehicles, which were in use. The research results are originating from theresearch of the paper authors.Keywords: Monitoring, lubrication systems, analysis oils, proactive maintenance, diagnosis.1. INTRODUCTIONModern trends of diagnosis in recent years, go tothe affirmation of the monitoring of oil, which hasresulted in growth of interest of producers and usersof oil. The reasons lie primarily in increasing thereliability, effectiveness, economy, and recentlymore and more present protection of theenvironment.Using Oil Analysis programs for engine oils hasseveral benefits: reduction of unscheduled vehicledowntime, improvement of vehicle reliability, helpin organizing effectiveness of maintenanceschedules, extension of engine life, optimization ofoil change intervals and reduction of cost of vehiclemaintenance.In application, oils change their propertiesthrough [1]: contamination by combustion productsand metal wear particles, consumption of additiveswhich is chemical and bears impact on importantoil functions and base oil oxidation.The primary role of engine oil is the lubricationof moving engine parts and reducing friction andwear of metal surfaces which provides the goodengine performance and its long life. In order toprovide a defined quality of engine oils duringproduction and for final products to meet theproduct specifications we need to know thephysical chemical characteristics of engine oils.Certain physical-chemical characteristics whichare significant for the quality of engine oils areachieved by adding additives to base oils. The mostfrequent additives are for: improving of viscosityindex-improvers, reducing pour point-depressants,maintaining engine cleanness-detergents anddispersants, preventing oxidation-antioxidants,preventing corrosion-corrosion inhibitors.2. LUBRICANT SERVICE LIFE ANDANALYSISTo know analytical properties of lubricants isthe base to make a decision in development,production and application of lubricants. Thelubricant classifications and approved systemspecify many performance characteristics and240 13 th International Conference on Tribology – Serbiatrib’13

analytical tests. The analytical tests are classicaland instrumental. Instrumental technical have theadvantages in small quantity of the sample andrapid analyze. As a part of the common proactivestrategy of the hydraulic systems maintenance,concept of on-line monitoring is introduce inpractice, recently [2], [3], [4], [5], [6]. It is acombination of the measurement procedures, bywhich sample of fluid is to be analyzed is takendirectly from the system and the results of themeasurements are continuously. On-line monitoringconsidering, first of all, control of cleanlinessclasses (according to ISO, NAS, SAE), control ofhumidity, viscosity, permittivity (acid),temperature…The following tests are the most used incondition monitoring: Spectrometric analysis,Analytical Ferrography, Rotrode FilterSpectroscopy (RFS), Infrared Analysis (FT-IR),Viscosity, Total Acid Number (TAN), Total BaseNumber, Water and Particle Count.Spectrometric analysis is a technique fordetecting and quantifying metallic particulates inused oil arising from wear, contamination andadditive packages. The oil sample is energized tomake each element emit or absorb a quantifiableamount of energy, which indicates the element’sconcentration in the oil. The results represent theconcentration of all dissolved metals and particles.The equipment for spectrometric analysis is thestandard equipment for oil analysis laboratoriestoday. It provides information on technical system,contamination and wears condition relativelyquickly and accurately. Spectroscopy is more-orlessblind to the larger particles in an oil sample,more precisely, to particles greater than 10 µm indiameter, which are more indicative of an abnormalwear mode [7].Analytical ferrography is a technique whichseparates magnetic wear particles from oil. Thoseparticles settle on a glass slide known as aferrogram. Microscopic examination enables todetermine the wear mode and probable sources ofwear in the technical system. Analyticalferrography is an exceptional indicator of abnormalferrous wear and it is inadequate for nonferrouswear.Rotrode Filter Spectroscopy (RFS) was firstintroduced in 1992. This spectrometric techniquedetects coarse wear metals and contaminants in aused oil sample. Diameter of those particles is upto 25 µm, but it excludes all additives. The coarseparticles are especially important. They are the firstindicators of abnormal wear situations.Fourier-Transform Infra-Red Spectroscopy is aspectrometric technique for detecting organiccontaminants, water and oil degradation products ina used oil sample. It monitors lubricant degradation(oxidation, nitration, sulfation, additive depletion)and liquid contaminants (water, glycol, fueldilution).Viscosity is the resistance of a fluid to flow andthe most important lubricant physical property. Thefluid is placed in a "viscometer" (a calibratedcapillary tube for precise flow measurementbetween two pre-marked points on the tube) andpre-heated to a given temperature in a "viscositybath" (which is usually oil-filled). After the oilreaches the desired viscosity temperature, gravityinfluencedflow of the oil is initiated in theviscometer and timed between two calibratedpoints. This time becomes the determinant for theresult.Total Acid Number (TAN) is a neutralizationnumber intended for measuring all acidic and acidactingmaterials in the lubricant, including strongand weak acids. It is a titration method designed toindicate the relative acidity in a lubricant. The TANis calculated from the amount of KOH consumed.The acid number is used as a guide to follow theoxidative degeneration of oil in service.Total Base Number (TBN) is a neutralizationnumber intended for measuring all basic (alkaline)materials in the lube (acid-neutralizing componentsin the lubricant additive package). The converse ofthe TAN, this titration is used to determine thereserve alkalinity of a lubricant. The TBN is highestwhen oil is new and decreases with its use. LowTBN normally indicates that the oil has reached theend of its useful life.Water can be detected visually if grosscontamination is present. Excessive water in asystem destroys a lubricant's ability to separateopposing moving parts, allowing severe wear tooccur with resulting high frictional heat. There areseveral methods used for testing the moisturecontamination (crackle, FT-IR water, centrifuge,Karl Fischer) each with a different level ofdetection (1000 ppm or 0.1 % for first threemethods and 10 ppm or 0.001 % for Karl Fischermethod).Particle Count is a method used to count andclassify particulate in a fluid according to acceptedsize ranges, usually to ISO 4406 and NAS 1638 [8].There are several different types of instrumentationon the market, utilizing a variety of measurementmechanisms, from optical laser counters to poreblockage monitors.3. THE RESULTS OF OIL ANALYSIS ANDDISCUSSIONIn this part are presented the results of oilanalysis examination during application in four-13 th International Conference on Tribology – Serbiatrib’13 241

stroke engines by physic-chemical methods in orderto evaluate possibilities of engine conditionmonitoring by oil analysis. This part presents theresults of experimental research of physic-chemicalcharacteristics of engines oil which was sampledfrom engines of PUCH 300GD, Pinzgauer 710 andIKARBUS IK 104P vehicles [9], [10].The research was carried out in two vehiclesPUCH 300GD (PUCH-1, PUCH-2), two vehiclesPINZGAUER 710M (PINZ-1, PINZ-2) and twovehicles IKARBUS IK 104P (IK104P-1, IK104P-2).The research was conducted through periodicsampling oil from engine vehicles listed above.Apart from the fresh oil (“zero” sample), samplesare taken after 1.000 km, 2.000 km, 3.000 km,4.000 km and 5.000 km for vehicles.The physical-chemical characteristics of oil inaccordance with standard methods are examined,shown in table 1.Table 1. Implemented tests and methods for examiningthe physic-chemical characteristics of oilCharacteristicMethodKinematic viscosity, mm 2 /s SRPS B.H8.022Viscosity IndexSRPS B.H8.024Flash Point (C) ISO 2592, ASTM D 92Pour Point (C) ISO 3016Water Content, mas.% ASTM D 95Total Base Number (TBN),mgKOH/gASTM D 2896Insoluble substances inpentane, %ASTM D 893Insoluble substances inbenzene, %ASTM D 4055Fe Content, %ASSCu Content, %ASSThe analysis was done on the fresh (new) oilsand oils that are used in the engines of vehicles.During the sampling of oil choice of the samplingwere conducted carefully according to the actual oilusage, which enabled each sample as representativeone.The wear mechanism of a tribologicallubrication system consists in the wear of contactsurfaces, and lubricant consumption. If there iswear of the contact surfaces, there are wearparticles present.Regardless of the availability of numerousmethods for diagnosing the physic-chemicalchanges of lubricants, in order to create a truepicture of the condition of lubricants from the usersystem, it is of importance to satisfy theprecondition of the possibility to obtain arepresentative sample. That is why it is extremelyimportant to take the sample in a proper way.Allowable values of deviation limits ofindividual characteristics of the oil are conditionedby the type of oil, working conditions and internalrecommendations of the manufacturer of lubricantsand users. Limited value characteristics of oils thatcondition the change of oil charging from engineare given in table 2. They represent the criteria forthe change of oil charge. Deviation of only onesource changes characteristics of oil charge, nomatter of what a characteristic is about.Table 2. Allowed values deviation of physico-chemicalcharacteristics of new and used oilPhysical-chemicalcharacteristics oil andproducts wearViscosity at 40C and100C, mm 2 /sMaximum allowedvariationMotor oil20%Viscosity Index, % 5 %Total Base Number (TBN),mg KOH/gr The fall to 50%Flash Point, C 20 %Water Content, % 0,2 %Products wear – Content Fe,ppm(μg/gr)100 ppmProducts wear – Content Cu,ppm(μg/gr)50 ppmUsed engine oil in examined vehicles are shownin table 3. Characteristics of zero samples of motoroil are shown in table 4, and the results used oilsamples in table 5.Table 3. Used engine oil in examined vehicles [9]Engine oil from engine of PUCH 300 GD vehiclesSAEclassificationAPIclassificationManufacturerFAMSAE 15W-40 API SG/CEKrusevacEngine oil from engine of PINZGAUER 710 M vehiclesSAEclassificationAPIclassificationManufacturerGALAXSAE 30/S3 -BeogradEngine oil from engine of IKARBUS 104 P vehiclesSAEclassificationSAE 15W-40APIclassificationAPI SG/CEManufacturerFAMKrusevacThe viscosity index is an empirical numberwhich shows how the viscosity of some oilschanges by increasing or reducing the temperature.High viscosity index shows relatively smalltendency of viscosity to change upon influence ofcertain temperature, as oppose of low viscosityindex which shows greater viscosity change withtemperature.242 13 th International Conference on Tribology – Serbiatrib’13

Table 4. Results of zero samples of oil from the engine[9]Type of motor oilCharacteristic FAMSAE 15W-40GalaxSAE 30/S3Color 3,0 3,0Density, gr/cm 3 0,881 0,902Viscosity at 40C,mm 2 /s104,81 104,63Viscosity at100C, mm 2 /s14,12 11,67Viscosity Index ─ ─Flash Point, C 230 240TBN,mg KOH/g10,5 9,8Table 5. The results of testing samples of used oil fromengines examined vehicles [9]SamplePUCH PUCH IK104 IK104 PINZ PINZ–1 –2 –1 –2 –1 –20 14,1 14,1 14,1 14,1 11,6 11,61 14,6 14,2 13,7 13,6 10,9 10,52 15,4 15,0 12,8 13,5 10,3 10,43 16,0 15,6 12,4 13,2 9,96 10,14 16,6 16,1 12,3 12,9 9,3 9,65 17,5 17,0 12,2 12,6 8,7 9,00 104,8 104,8 104,8 104,8 104,6 104,61 111,0 110,4 96,9 104,4 100,4 100,92 113,5 111,8 96,2 101,9 94,4 96,13 119,4 113,8 92,3 97,1 86,3 88,64 126,4 115,9 90,8 94,8 79,1 82,25 132,7 127,5 90,2 93,1 75,9 76,9Viscosity at100C, mm 2 /sViscosity at40C, mm 2 /sViscosityIndexFlashPoint,CTBN,mgKOH/gFe Content(ppm)Cu Content(ppm)0 135 135 135 135 100 1001 129 131 132 133 96 972 122 126 130 131 93 953 119 123 125 127 89 914 116 120 122 124 84 875 112 115 119 121 82 840 230 230 230 230 240 2401 220 215 217 212 196 1932 208 210 214 210 186 1773 205 204 213 202 168 1594 197 202 210 193 154 1435 192 188 189 184 136 1280 10,5 10,5 10,5 10,5 9,8 9,81 9,1 9,4 8,8 8,1 9,6 9,42 7,2 8,9 8,7 7,7 9,1 8,43 6,5 8,7 8,4 7,2 8,3 7,84 6,1 8,1 7,9 6,8 7,6 6,65 5,2 7,6 7,3 6,4 7,1 6,21 98,4 27,4 30,1 20,5 19 17,92 123 59,8 32,5 46,3 19,8 40,93 137,1 71,2 35,6 57,6 38,3 86,74 149,4 71,4 37,5 62,8 54,3 132,85 165,3 86,8 38,5 69,6 105,4 2611 4,9 2 1,5 3,2 3,5 3,32 5,9 3,4 1,9 5,1 4,1 3,83 6,7 3,7 3,2 6,3 5,3 64 7,3 3,9 4,4 7,7 6,9 8,15 7,9 5,4 4,9 9,1 8,7 9,7During the exploitation it is desired that theviscosity changes as lesser as possible with thechange of temperature. If during work temperaturemodes are changeable and cause major changes ofviscosity that may cause disruptions in thefunctioning of the system, which is a manifestationof increased friction, wear and damage.Change of engine oil Viscosity Index is shownin the figure 1. The decrease in the Viscosity Indexoil is evident for all vehicles, exceeding the limit of5 % (table 2).Viscosity Index1401301201101009080Max allowed decrease 5 %= 128,25 (SAE 15W-40)Max allowed decrease 5 %= 95 (SAE 30)0 1000 2000 3000 4000 5000Crossed kilometers, kmIndex viscosity of "0" sample:135 (SAE 15W-40)PUCH−1PUCH−2PINZ−1PINZ−2IK104P−1IK104P−2Figure 1. The change of Viscosity Index [9]The most important engine oils characteristic isthe viscosity defined as a measure of inner frictionwhich works as a resistance to the change ofmolecule positions in fluid flows when they areunder the impact of shear force, or in other words,it is the resistance of fluid particles to shear.The viscosity is a changeable category and itdepends on the change of temperature and pressure.A higher temperature reduces the viscosity andmakes a fluid thinner.Multigrade engine oils among numerousadditives always contain also viscosity indeximprovers. These additives are special types ofpolymers, which in small concentrationsignificantly improve engine oils rheologicalproperties, especially viscosity and viscosity index.However, during engine oils utilization,degradation of viscosity index improvers i.e. Breakdown of polymeric molecules occurs. It results inreduction of their molecular weight what leads toviscosity loss and oil film thickness decrease,which causes undesirable phenomena of frictionand wear.Reasons for the increase of viscosity lubricantsare as follows: oxidation of lubricants, cavitationsdue to foaming lubricants, dissolution of lubricantswith water, pouring and charging system viscosityfat greater than recommended and contamination ofsolid particles and products wear lubricants.The reasons for the reduction of lubricantsviscosity are: lubricants contamination of fuel (formotor oil), shearing additive for reclamationviscosity, drop point of flash, grinding molecules,lubricants contamination without solubility withwater, pouring and charging system viscosity lessfat than recommended, and the impact of liquid13 th International Conference on Tribology – Serbiatrib’13 243

cooling. Also, the causes may be high temperature,load, uncontrolled long interval use, insufficientamount of oil in the oil system, inefficient coolingsystems and the like.As expected, kinematic viscosity usuallydecreases in time due to fuel penetration, or - inwell maintained engines, there occurs a slightincrease as a result of the increase of the oilinsoluble, without fuel penetration.Figure 2. shows the changes viscosity at 40Cengine oils during exploitation.Viscosity at 40 ° C, mm 2 /s135125115105958575Viscosity of "0" sample:104,81 (SAE 15W-40)104,63 (SAE 30)Max. allowed increase 20%=125,77 (SAE 15W-40)Max. allowed decrease20% = 83,84 (SAE 30)0 1000 2000 3000 4000 5000Crossed kilometers, kmPUCH−1PUCH−2PINZ−1PINZ−2IK104P−1IK104P−2Figure 2. The change of viscosity at 40C [9]The increase viscosity at 40ºC engine oil isevident for PUCH-1 and PUCH-2 vehicles,exceeding the limit of 20%. The decrease viscosityat 40ºC engine oil is evident for PINZ (exceedingthe limit of 20%) and IK104P vehicles.Viscosity at 100 ° C, mm 2 /s2018161412108Viscosity "0" sample:14,12 (SAE 15W-40)11,67 (SAE 30)Max. allowed increase 20%=17,04 (SAE 15W-40)PUCH−1PUCH−2PINZ−1PINZ−2IK104P−1IK104P−2Max. allowed decrease 20%= 11,29 (SAE 15W-40)Max. allowed decrease 20% = 9,33 (SAE 30)0 1000 2000 3000 4000 5000Crossed kilometers, kmFigure 3. The change of viscosity at 100C [9]TBN is a neutralization number intended formeasuring all basic (alkaline) materials in the lube(acid-neutralizing components in the lubricantadditive package). The TBN is generally acceptedas an indicator of the ability of the oil to neutralizeharmful acidic byproducts of engine combustion.The TBN is highest when oil is new and decreaseswith its use. Low TBN normally indicates that theoil has reached the end of its useful life. TBN is ameasure of the lubricant's alkaline reserve, andmostly applies to motor lubricants. If a lubecontains no alkaline additives, there is little use todetermine a TBN, as there will likely be none.Combustion acids attack TBN, e.g., sulfuric acid,decreasing as it consumes.Figure 4. shows the changes of total basenumber (TBN) engine oils.TBN, mg KOH/g111098765TBN of "0" sample:10,5 (SAE 15W-40)Max allowed decrease50% = 4,9 (SAE 30)Max allowed decrease 50% = 5,25 (SAE 15W-40)0 1000 2000 3000 4000 5000Crossed kilometers, kmFigure 4. The change of TBN [9]PUCH−1PUCH−2PINZ−1PINZ−2IK104P−1IK104P−2The decrease TBN engine oil is evident for allvehicles. Until 5.000 km TBN value does notexceed the allowed limit, except for PUCH-1vehicle.Flash point, ° C240220200180160140120Flash point "0" sample:230 (SAE 15W-40)240 (SAE 30)Max allowed decrease20% = 192 (SAE 30)Max allowed decrease20% =184 (SAE 15W-40)0 1000 2000 3000 4000 5000Crossed kilometers, kmFigure 5. The change of flash point [9]PUCH–1PUCH–2PINZ−1PINZ−2IK104P−1IK104P−2Flash point represents data that shows whattemperature leads to open fire ignition by the steamcreated by oil heating. In engine oil analysis theflash point determines the presence of fuel oil,which is a consequence of poor motor (bad workinjectors). The reduction of flash point is due to thepenetration of fuel.Figure 5 shows the change of flash point forengine oils. The decrease in the flash point is244 13 th International Conference on Tribology – Serbiatrib’13

noticeable, and by the end of exploitation testingexceeds the allowed limits (20%, table 2) for PINZvehicles.Analysis of the contents of different metals thatare in the lubricant is very important. Metalparticles are abrasive, and act as catalysts in theoxidation of oils. In motor oils, the origin of theelements may be from the additives, the wear, thefuel, air and liquid for cooling. Metals from theadditives can be Zn, Ca, Ba, or Mg and thatindicates the change of additives. Metalsoriginating from wear are: Fe, Pb, Cu, Cr, Al, Mn,Ag, Sn, and they point to the increased wear inthese systems. Elements originating from the liquidfor cooling are Na and B, and their increasedcontent indicates the penetration of cooling liquidin the lubricant. Increased content of Si or Ca,which originate from the air, points to amalfunction of the air filter.Iron and copper content (figure 6 and 7), as aproduct of wear, in the oil charge to the end ofexploitation testing has a growing trend.Content of Fe, ppmContent of Cu, ppm121030025020015010086420500Max content of Fe ─ 100 ppmPUCH−1PUCH−2PINZ−1PINZ−2IK 104P−1IK104P−20 1000 2000 3000 4000 5000Crossed kilometers, kmFigure 6. The change of content Fe [9]Max content of Cu ─ 50 ppm0 1000 2000 3000 4000 5000Crossed kilometers, kmFigure 7. The change of content Cu [9]PUCH−1PUCH−2PINZ−1PINZ−2IK104P−1IK104P−2Content of iron is significantly above theallowable limits (100 ppm, table 2) for PUCH-1and PINZ-2 vehicles.Content of cooper is significantly below theallowable limits (50 ppm, table 2) for all vehicles.4. CONCLUSIONThe interpretation of used oils analysis is verycomplex, because the individual analyses areinterdependent. That is the reason why it isnecessary to know the entire oil analysis, and notbring conclusions based on individual analysisresults. It is also necessary to establish both normaland critical quality levels for specific oils in givenengines and under specific application conditions.The lubricant, being an inevitable factor in thetribomechanical system of engine has – apart fromthe usual lubricating role, also an important role indetecting the engine operation efficiency andcondition. This is achieved through a systematicmonitoring of oil in application and a permanentcontact between the motor oil manufacturer anduser.Analyses from used oil sample should always becompared with previous samples and finalconclusions should be based on “trend analysis”and has two closely related objectives: to obtaininformation on the lubricant drain intervals andpreventive maintenance of the machine.Investigations it was realized that there is achange of physical-chemical characteristics of oilfor lubrication in the engines vehicle. Thesechanges are in direct dependence on the state of allelements tribomechanical engines system, anddepending on their functional characteristics.REFERENCES[1] J. Denis: Lubricant properties analyses and testing,Editions Tehniq, Paris, 1997.[2] D. Grgić: On-line monitoring of oil quality andconditioning in hydraulics and lubrications systems,in: Proceedings of 10th SERBIATRIB '07,Kragujevac, Serbia, pp. 305-309.[3] I. Mačužić, P. Todorović, A. Brković, U. Proso, M.Đapan, B. Jeremić: Development Of Mobile DeviceFor Oil Analysis, Tribology in Industry 32, pp. 26-32, 2010.[4] V. Macian, B. Tormos, P. Olmeda, L. Montoro:Analytical approach to wear rate determination forinternal combustion engine condition monitoringbased on oil analysis, Tribology International 36,pp. 771–776, 2003.[5] L. Guan, X. L. Feng, G. Xiong, J. A. Xie:Application of dielectric spectroscopy for enginelubricating oil degradation monitoring, Sensors andActuators A: Physical 168, pp. 22–29, 2011.[6] V. Macian, R. Payri, B. Tormos, L. Montoro:Applying analytical ferrography as a technique to13 th International Conference on Tribology – Serbiatrib’13 245

detect failures in diesel engine fuel injectionsystems, Wear 260, pp. 562–566, 2006.[7] R.I. Taylor, R.C. Coy: Improved fuel efficiency bylubricant design, 2001[8] M. Piest and C.M. Taylor: Automobile enginetribology, 2000.[9] S. Peric, The development of a method of diagnosisthe condition from the aspect of physical-chemicaland tribological characteristics of lubricating oils ofvehicles, PhD thesis, Military Academy, Belgrade,2009.[10] S. Perić, B. Nedić: Monitoring oil for lubrication oftribomechanical engine assemblies, Journal of theBalkan tribological association 16, pp. 242-257,2010.246 13 th International Conference on Tribology – Serbiatrib’13

Serbian TribologySocietySERBIATRIB ‘1313 th International Conference onTribologyKragujevac, Serbia, 15 – 17 May 2013Faculty of Engineeringin KragujevacADVANTAGES AND APPLICATIONS OFSELF-LUBRICATING PLASTIC BEARINGSAleksandar Marinković 1 , Miloš Stanković 21 Mechanical Engineering Faculty, University of Belgrade, Serbia, amarinkovic@mas.bg.ac.rs2 Inovation Center of Mechanical Engineering Faculty, Serbia, mstankovic@mas.bg.ac.rsAbstract: Self-lubricating sliding bearings are widely used in industrial applications and we could assortthem in several groups towards manufacturing or lubrication. Some of them, such are oil-retaining porousbearings have been long time studied, but plastic bearings are not enough investigated where we have justsome of experimental results. Self-lubricating plastic bearings are produced on polymer basis which isoptimized with fiber reinforcement and solid lubricants. They are an ideal solution for machinery thatrequire clean and oil-free operation. Plastic bearings also perform well in dirty environments since there isno oil to attract dust and dirt. Authors of this paper describe main performances and reasons why this kindof bearings are current widely used. A few typical applications of plastic bearings are presented in thepaper, taking into account advantages in lubrication, production and maintenance costs in comparison withclassical rolling and sliding bearings.Keywords: self-lubricating, plastic sliding bearing, polymers, dry operating.1. INTRODUCTIONMost of machine and equipment manufacturersare trying to eliminate or at least to reducelubrication systems in aim to settle production costsdown without sacrificing machine performances.According to significant Bearing Companiesinvestigations, more than 50% of bearing failuresare lubrication related (Figure 1). In a study byMIT, USA it was estimated approximately $240billion is lost annually due to downtime and repairsto equipment damaged by poor lubrication [1].By eliminating lubrication from machinery,equipment manufacturers can minimize the costsand risks associated with maintenance for the enduser. Because of lubrication problem and costs,which are dominant during the working life, apossible solution is to apply dry running plasticsliding bearing.Plastic bearings are produced on polymer basiswhich is optimized with fibre reinforcement andsolid lubricants. They are an ideal solution formachinery that requires clean and oil-freeoperation. Plastic bearings are doing well in dirtyenvironments since there is no oil to attract dust anddirt.2. PLASTIC BEARING MATERIALSThere are several typical groups of materials forplastic bearings, taking into account their physicalmechanicalperformances [2]:- Thermoplastic materials- Phenol and Epoxy plastic materials- Elastomers- Multilayer plastic materialsFigure 1. Types of lubricated related bearing failures13 th International Conference on Tribology – Serbiatrib’13 247

Thermoplastics and thermoplastic materials arepolymers that turn to liquid when heated and turnsolid when cooled. They can be repeatedly remeltedand remolded, allowing parts and scraps to bereprocessed. In most cases they are also veryrecyclable. Some thermoplastics contain fillermaterials such as powders or fibres to provideimproved strength and/or stiffness. Products inthermoplastics could also contain solid lubricantfillers such as graphite or molybdenum disulfide.Others contain metal powders or inorganic fillerswith ceramics and silicates aimed to improve theirmechanical and tribological performances.different coatings or solid lubricants. Because ofcurrent great plastic bearing expansion in widerange of different applications, many companies areexploring and try to on the market with theirproducts. Most of them are sited in Europe or USAand have relatively long tradition, but last years lotof Far East companies are trying to overrun themby low cost products. Some of best knownmanufacturers of plastic bearings are multinationalCompany Igus® [4] (Figure 2) with main factoriesin Germany; famous bearing Company SKF hasalso some investigations and products in plastics;CSB Bearings [5]; Federal Mogul Germany withGlyco products; AFT Fluorotec (SW Plastics) UK;ISB Italcuscinetti Group Italia [6], etc.3. ADVANTAGES OF PLASTIC BEARINGSFigure 2. igus® lines of plastic bearings made from highperformance polymersPolyethylene (PE), Fluoroplastics (such asPTFE), Polyamide (PA) and Polyoxymethylene ( asPOM) are common plastic materials from thisgroup and in general, those are using in slidingbearing manufacturing. Detailed performancesstudy of those materials is not aim of this paper [3],but here used to be mentioned just characteristicsimportant for typical applications. If somebodyneeds plastic bearing for extreme load, than aHomopolymers or Copolymers (POM) with higheststrength are recommended. In high environmentaltemperature conditions of the bearing exploitationPolytetrafluoroethylene (PTFE) is useful with max.working temperature around 200 o C. From tribologypoint of view, materials such PTFE is, has thelowest friction coefficient value (between 0,02 and0,06 in dry conditions). If we need good wearresistance of the bearing, materials as Polyamide(PA) and POM plastic materials are recommended.Other plastic materials except above explainedgroup of Thermoplastics are not so common in use,but we could apply them in some special cases. Forexample, multilayer plastic materials are useful incombination with some metal as a matrix, withIf we are taking into account proper lubricationdelivery as a critical for the operation of ballbearings and most require continued maintenancefor re-lubrication, this is a starting reason forthinking about their replacement with plasticbearings. There are also additional parts required toprotect ball bearing from contaminants. Accordingto several Institute research, the leading cause ofbearing failure is due to contamination of thelubrication by moisture and solid particles. If aslittle as 0.002 percent water gets mixed into thelubrication system, it increases the probability offailure by 48 percent. Just six percent water canreduce the bearing lifetime by 83 percent.Ball bearings require seals to keep oil in andunwanted water and liquids out, as well as wipers /scrapers to keep dust and debris out. Seals only lastso long and do not perform well in dirty and dustyenvironments and can also increase friction in theapplication. In some applications where dust anddebris are prevalent during operation, seals andwipers may require frequent replacement.Figure 3. Comparing ball bearings to plastic bearings248 13 th International Conference on Tribology – Serbiatrib’13

4. PLASTIC BEARING APPLICATIONSRegarding their advantages, plastic bearings aregood solution for many applications in machinerythat require clean and oil-free operation. They alsoperform well in dirty environments since there is nooil to attract dust and dirt, like the agriculturalindustry. Some manufactures creates individualplanting row units using walking gauge wheels todeliver a consistent planting depth (Figure 4).successfully used with shafts made of bronze andsteels of different hardness and structure.Figure 4. Plastic bearing application in agricultureOil impregnated bronze bearings with graphiteplugs were used to facilitate this movement untilthey began causing severe problems. They wereeven requiring replacement two to three times aseason. But the bronze bearings were experiencinghigh wear and premature failure due to the veryabrasive conditions caused by high levels ofvolcanic ash in the soil, or the high salt content inthe air caused corrosion and seizure. By replacingall 144 bronze bearings with iglide® selflubricatingplastic bearings from igus®, the pick arms lif wasincreased by 5 to 6 times. The actual bearings cost70 to 80 percent less than bronze bearings and weremore reliable.Shipbuilding and hydraulic turbine buildinghave accumulated much experience with the use ofsliding bearings made of UGET carbon plastic [7].These include friction units of a driving rudder setof ships of different types and design (supports forrudders and rudder machines) with regard tostabilizers, interceptors, drives for actuators ofKingston valve type, and scupper screens, as wellas mast elevating extending devices andmechanisms (Figure 5). Sliding bearings have beenpreviously made of bronze, and shafts have beenmade of a corrosion resistant material having ratherlow antifriction characteristics, corrosion resistantsteel or titanium alloys. Therefore, in the absence ofreliable oil lubrication system there is a danger ofseizure of metallic bearings, which may result inthe failure of the whole mechanism. UGET carbonplastic containing poly functional epoxy resin andtissue of low module carbon fibre was developed.Bearings made of UGET carbon plastic areFigure 5. Sliding friction unit of a hydraulic turbineOne manufacturer specializes in vertical, form,fill and seal packaging equipment for handling awide range of products: from green beans to candyto detergent. The machines are capable of reachingup to 160 cycles per minute and withstanding loadsup to 15 pounds, while operating at speeds of 750feet per minute (Figure 6).The manufacturer had been using metal linearball bearings. After the metal bearings scored theshafts and leaked grease on some of the machines,the company decided to replace them with selflubricating linear plain bearings. To date, the linearbushings have surpassed the 10 million cycle markon some of the company’s packaging machineswith little to no noticeable wear.Figure 6. Packaging machine with plastic bushingsIn the quest to improve the way prostate canceris detected and treated, a team of researchers fromthe Worcester Polytechnic Institute (WPI) inMassachusetts have developed a specializedmagnetic resonance imaging (MRI) compatiblepiezoelectric actuated robot [1]. To facilitate13 th International Conference on Tribology – Serbiatrib’13 249

different types of motion, the robot uses a DryLin®linear guide system and iglide® plastic selflubricating plain bearings. The linear guidesfacilitate translational motion of the positioningmodule, which provides gross positioning for therobot’s needle driver. The needle driver is a vitalpart of the system, as it enables the rotation andtranslational movement of the “needle cannula”: aflexible tube inserted into the patient’s body cavityfor MRI-guided diagnosis and therapy (Figure 7).loads in dry conditions, such also in the presence ofmany lubricants (water, acids, alkali, oils, hydraulicliquids).- This kind of bearing can be used on softershafting, even anodized aluminium, which hasexcellent corrosion resistance and is usually lessexpensive and easier to machine than case hardenedmaterial or stainless steel.This paper is just a part of preview andintroduction in further researches of plasticbearings subjected to make simpler machinemaintenance and better energy efficiency. Becauseof great expansion and clear explained advantagesof plastic bearings application in several branchesof industry, not only investigations of new polymermaterials, but also deformation behaviour analysisin dry and conditions under different lubricantsused to be done.AKNOWLEDGEMENTFigure 7. Plastic bearings in magnetic resonance robotTwo plastic plain bearings are used in the frontand rear of the driver to constrain the needle guide.The bearings enable the robot’s motor to rotate theneedle using the mechanism by way of a timingbelt. This rotating needle would reduce tissuedamage while enhance targeting accuracy. Another10 plain bearings were used to create a revolutejoint, also known as a “pin joint” or “hinge joint”,to provide single-axis rotation.CONCLUSIONAn actual scientific and practical problem couldbeen solved concerning the development andapplication of high strength antifriction polymermaterials in machine building. According to manyresearches following by experience in lot of typicalapplications [8], we could summarize the mainbenefits of plastic bearings:- No maintenance- Oil free, dry-running;- Corrosion resistant;- Cost less than ball and other bearings;- Handle contamination well and often do notrequire seals or scrapers;- High damping characteristics for vibrations,ability to reliably work under static or dynamicThis work has been performed within theprojects TR-35021 and TR-35011. Those projectsare supported by the Republic of Serbia, Ministryof Education, Science and TechnologicalDevelopment, which financial help is gratefullyacknowledged.REFERENCES[1] Mat Mowry: The Thrue Cost of BearingLubrication, igus White Paper, igus inc.USA, 2012.[2] Momčilo R Janković: Exploitation Performances ofPlastic Radial Sliding Bearings made in Yugoslavia,Master Thesis, Mechanical Engineering FacultyUniversity of Belgrade, Belgrade, 1976.[3] S. P. Zakharychev and D. V. Otmakhov: UsingPolymer Adhesive Compositions in Design of SliderBearings, Polymer Science, Series D. Glues andSealing Materials, Vol. 4, No. 4, pp. 314–316.Pleiades Publishing Ltd. 2011.[4] Igus Bearings Catalogue, Igus inc.USA.[5] CSB Bi-metallic Composite Bearings, CSB France.[6] ISB Technical Catalogue, Italcuscinetti Group,Italia.[7] A. V. Anisimov, V. E. Bakhareva, and G. I.Nikolaev: Antifriction Carbon Plastics in MachineBuilding, Journal of Friction and Wear, Vol. 28, No.6, pp. 541–545, Allerton Press, Inc., 2007.[8] T.A. Osman: Effect of lubricant non Newtonianbehaviour and elastic deformation on the dynamicperformance of finite journal plastic bearings,Tribology Letters, Vol. 17, No. 1, pp. 31–40,Springer, 2004.250 13 th International Conference on Tribology – Serbiatrib’13

Serbian TribologySocietySERBIATRIB ‘1313 th International Conference onTribologyKragujevac, Serbia, 15 – 17 May 2013Faculty of Engineeringin KragujevacEFFECT OF VISCOSITY ON ELASTOHYDRODYNAMICLUBRICATION BETWEEN PARALLEL SURFACES SUBJECTEDTO HIGH ACCELERATIONUsman Ali Zia 1 , Aamer A. Baqai 1 , Waseem Akram 11 National University of Science & Technology(NUST), Pakistanusman_ali442@hotmail.com, aamerbaqai1@gmail.com, wasim6@gmail.comAbstract:Material wear due to friction is one of the most commonly experienced causes of material failure inany mechanical industry. Various studies have been conducted and as a result of these studies variouslubrication solutions have been proposed. Present work is an attempt to propose an Elastohydrodynamiclubrication solutio20n for a frictional wear problem experienced in an industrial application involving thesliding contact between two parallel surfaces subjected to high acceleration. The “Numerical model formixed lubrication” developed by Dong Zhu in 1990s has been modified to accommodate the constraints ofproblem at hand. The solution proposed predict the lubricant film thickness that when maintained betweenthe contacting surfaces can avoid full metal contact, which in turn shall avoid the material wear. The presentresearch is an attempt to comprehend the effect of change of viscosity and change of clearance between thesurfaces on the value of film thickness. Different grades of nonflammable, anticorrosive Perfluoropolyetherbased grease (Krytox) are used. The results obtained are in the form of graphs which calculates the value offilm thickness for one complete slide of one surface over the other. The results obtained by the numericalmodel are compared and found well-in-accordance with the experimental data available and with theanalytical predictions made by scholars in the past.Keywords: Tribology, Friction, Elastohydrodynamic lubrication, Krytox, Shear stress factor, hydrodynamicpressure1. INTRODUCTIONWear is the major cause of material wastage andloss of mechanical performance and any reductionin wear can result in considerable savings. Frictionis a principal cause of wear and energy dissipation.Considerable savings can be made by improvedfriction control. It is estimated that one-third of theworld's energy resources in present use is needed toovercome friction in one form or another.Lubrication is an effective means of controllingwear and reducing friction. Principles studied underthe field of tribology helps to analysis suchproblems of frictional wear. Lubricationphenomenon is being used to avoid fricational wearsince the time of its discovery.For any tribological studies the lubricationregime in which that particular machine/application is working is very important as it steersthe later research in the field. The outcome of mostof these researches is a lubrication solution that ifmaintained between the contacting surfaces canavoid dry metal-metal contact and inturndiminishes the chances of any mechanical wear. Alubrication solution can be studied under two maintypes of regimes:Fluid Film Lubrication andBoundary Lubrication.Out of these regimes, the fluid film lubrication isconsidered in case the required factor of safety ishigh so it is desired that the two contacting surfaceshave minimum chances of coming in contact witheach other. This requires a thick lubrication layer(exceeding a thickness of more that 1um) [1]between the surfaces so that enough hydrodynamicpressure exists to keep the surfaces apart.13 th International Conference on Tribology – Serbiatrib’13 251

Table 1: Nomenclature of Symbols.Terms Definition Value Terms Definition ValueF h and M hF fh andM fhF MBHydrodynamic Force andmomentfrictional force andmoment due to lubricantfilmInertial force due toprimary bodyCalculated byintegratinghydrodynamic pressuresCalculated byintegrating the ShearstressCalculated by Equation(3)M MBF MB andF FınF G Gas Force or Thrust Force Pre-defined term aYF FinInertial force due to FinCalculated by Equation(4)I BMoment due to massof primary bodyReciprocatinginertial forcesAcceleration of thebodyCalculated byEquation (5)Calculated byEquation (7)Calculated byEquation (8)Calculated byEquation (9)Angular moment of Inertia about thebody’s center of massxPerpendicular distance from point of application ofF MB to the center of gravity of FinC Constantη p Viscosity at Pressure ‘p’ p Concerned pressureη 0Viscosity at atmosphericpressuren Constantαpressure viscosity coefficientThe slope of this graph is α[16]Obtained throughexperimental dataApproximately 16Calculated by plotting the natural algorithm of dynamic viscosity versus pressure.For the case of boundary lubrication nohydrodynamic film is sustained. The coefficient offriction is very high and friction is proportional tothe applied load for a certain range of slidingvelocity and temperature.Fluid film lubrication is further classified ashydrodynamic and Elastohydrodynamiclubrication, based on the fact that the surfacesunder consideration show remarkable elasticdeformation or not. As evident from abovediscussion the selection of lubrication regime isvery important and is completely dependent on thecircumstances under which the surfaces come incontact.The present research is focussed on thelubrication solution for problem under the regimeof elastohydrodynamic lubrication. The presentresearch is devided into 7 sections. Section 2focusses on the literature studied and benchmarkedfor the present research. The problem underconsideration is explained in the section 3.The constraints explained in the senction 3 laiddown the basics for developement of amethodology for the solution, the same isexplained in the section 4. Section 5 explains theresults obtained on the basis of methodology ofsection 4. Section 6 concludes the work bydrawing the conclusions based on discussions ofsection 5.2. LITERATURE SURVEY:The studies in the field of tribology initiated in19 th century with the experiment of BeauchampTower when he noticed that the oil film providedin between the surfaces of a journal bearing tries topump out of a hole provided on top of the bearing.He conculded that the pumping phenomenon canbe explained with the generation of hydrodynamicpressure between the bearing surfaces.Reynolds in 1886 [2] provided themathematical solution for the generation of abovementioned pressure. The theory presented byReynolds provided the long waited analytical proofof the hydrodynamic pressure generated betweenthe surfaces. This hydrodynamic pressure helpskeep the surfaces apart and thus avoiding themetal-metal contact which inturn avoid themechanical wear.After Reynolds, many scholars tried to predictthe lubricant film thickness based on his work, forproblem related to different types of bearingsincluding thrust bearing, journal bearing and sliderbearing etc. The most notable work in this reagrdwas done in the field of automobiles to predict thelubrication solution for piston-liner assembly.G.M. Hamilton in 1972 [3] predicted the filmthickness solution for piston-liner problem whileconsidering the regime of hydrodynamiclubrication. Further in 1977 Hamrok and Dowson[4] formulated the empirical formulae for filmthickness calculation.252 13 th International Conference on Tribology – Serbiatrib’13

For Elastohydrodynamic film thickness solution,the first realistic model was provided by Ertel andGrubin [5]. After Grubin other significantcontribution in the field was made by Dowson andHigginson [6]. They described an iterativeprocedure that not only yielded a wide range ofsolutions during the next decade, but also enabledthem to derive an empirical minimum-filmthickness formula for line contacts.One of the most considerable works in thisregard was presented by Dong Zhu in 1991 [7, 8].He presented a model namely “Numerical Modelfor Piston-Skirt assembly under mixed lubrication”.He developed a new relationship for film thicknessbased on the eccentricities and clearance betweenthe piston and cylinder. It was also claimed that theempirical relationship developed previously byHamrok is not applicable for cases of high pressureand acceleration. Further in 2002 he extended hiswork [9-11] by considering a wider range ofparameters like Load, Speed and Materialproperties. He also supported his theory withexperimental data in a series of research papers.The aim of the present research is to verify theapplicability of Dong Zhu’s model on our presentassembly. However our present problem, describedin detail in the next section, has differentgeometrical and environmental constraints,therefore it requires certain modification in theoriginal model.3. PROBLEM DEFINITION:The problem at hand involves a cylindrical bodysliding inside a hollow tube. The sliding action ofthe cylidrical body is under high accelearation andtherfore involves high pressure environment. It isdesirable that the body does not have any wear atits surface as it can alter the trajectory followed byit during its further operation. Based on high factorof safety required and the deforamtion of the bodydue to high pressure and high acceleration,elastohydrodynamic lubrication regime isconsidered. The assembly is shown below.The force which forces the cylindrical body toslide is a thrust force at the base of the body. Thisfurther guides the research to be done in the field ofthrust bearing. The thrust force is also responsiblefor the wobbling action of the cylinder inside thetube this wobbling is in itself responsible for thegeneration of hydrodynamic pressure in thelubricants’ layers provided between the surfacesthat keeps the surfaces apart and avoid mechanicalwear as will be explained in more detail in latersections.Outer Tube-Stationary SurfaceFins ofBodyCylindrical Body– Sliding SurfaceThe present research tries to find a lubricationsolution in term of film thickness for the assemblyexplained earlier. The film thickness predictedthrough the results of the research, if maintained inbetween the contacting surfaces shall be able towithstand the thrust force avoiding the surfaces tocome in contact.The model preseneted by Dong Zhu is modifiedbased on the above mentioned constarints as shallbe described in the following section.4. MODIFIED NUMERICAL MODEL:To model the phenomenon, followingassumptions are made:• The lubricant is an incompressibleNewtonian fluid and the flow is laminar.• Side leakage, oil starvation and surfaceroughness factors are neglected.• No relative motion between the bodiesunder sliding motion.• An Iso-viscous case, that is, viscosity issame in the circumferential and slidingdirections.• The fully flooded inlet and Reynolds exitconditions are applied.• The surfaces of the ring and the liner areperfectly smooth.• Thermal effects are neglected.Based on the problem constraints and theassumptions made following modification are madein the Piston skirt model.13 th International Conference on Tribology – Serbiatrib’13 253

4.1. Modifying mixed lubrication model to EHLModelAs previsouly explained the lubrication regimeto be considered for our present case isElastohydrodynamic lubrcation whereas the regimeof Dong Zhu’s work [7] was mixed lubrcaiton. Inorder to incorporate this change, the terms (forcesand moments) related to the contact between thetwo bodies are neglected.F = F h + F cF f = F fh + F fcM = M h + M cM f = M fh + M fc4.2. Modification in Basic Dynamic ModelDue to different geometrical constraints thedynamic model is modified thorugh following setof equations. These equation are based onsummation of forces and moments. The forces andmoments acting around the Center of Gravity of thecylindrical body and fin are considered andequilibrium equations are applied at center ofgravity of fin. Final equations formulated bysimplifying the equilibrium equation are:WhereF h + F S + F fh = −F MB − F Fin (1)M h + M fh + M MB + F MB ∗ (x) = 0 (2)F MB = −m MB e ẗ + b L (ë b − e ẗ)(3)F Fın = m Fin ∗ aY (8)aY =F G(m MB + m Fin )4.4. Modification in Pressure-ViscosityRelationship(9)The numerical model for mixed lubrication usesthe Barus eqution to cater for the change inviscosity due to pressure. Studies have shown thatin case of high pressure problems the use of Barusequation can result in serious error in film thicknessand hydrodynamic pressur calculations andtherefore a new relationship developed by Chu etal. [14] in 1962 shall be used.Barusη p = η 0 e αp (10)Equation[13]Equation byChu et. alη p = η 0 (1 + C × p) n (11)5. RESULTS AND CONCLUSION:Following data is used as input for the numericalmodel.TermsSymbolUsedValueMass of primary body m MB 2000 KgMass of fin m Fin 10 KgViscosity of fluid (Krytox215, 226, 227)η 00.03204Pa.s0.04550 Pa.s0.07476 Pa.sF Fin = −m Fin e ẗ + a L (ë b − e ẗ)(4)Diameter of primary body D 0.45 mElastic Modulus E 69 GPaM MB = −I B (e ẗ − e b )/L ̈ (5)F S = F G + F MB + F Fın(6)Where e b & e t are eccentricities at the top andbottom of the tube.4.3. Modification in Acceleration ProfileThe acceleration profile is modified based on theresearch by D.K. Kankane et. al. [12] whilecalculating the in-bore velocity of a projectile:F MB = m MB ∗ aY (7)Thrust Force F G 17-32 KNFollowing results have been plotted usingMATLAB. The results are graphical, showing theeffect of change in clearance between thecontacting surfaces, viscosity and the thrust force.The graph is drawn between the film thickness andthe length of the outer tube, this gives us the valueof the film thickness that if maintained between thesurfaces throughout the length of the tube shallavoid any metal contact between the surfaces andtherefore will help diminish the mechanical weardue to it.254 13 th International Conference on Tribology – Serbiatrib’13

The results concluded for different grades ofKrytox at different values of clearance have beenattached at the end of the paper for referencepurpose however a summarized graph for both caseof thrust forces are plotted in MS Excel and areshown above.Based on the graphs attached at the end of thepaper and the summary graph displayed above,following is concluded:• The film varies almost linearly throughoutthe length of the tube. This can beexplained by the fact that during the slidingof one surface over the other the thrustforce increases and this increase in thethrust force causes an increase in thehydrodynamic pressure which is directlyproportional to the velocity of the slidingsurface. This increase in the hydrodynamicpressure requires an increased amount oflubricant to avoid metal-metal contact.• The small peaks in the beginning of thegraph can be explained with the help ofwobbling phenomenon that takes place dueto the impulsive nature of the forceprovided.• It is clear from the above graphs that thevalue of film thickness required, increaseswith increase in the viscosity of the liquidkeeping the clearance between the surfacesconstant. This conclusion is also supportedby the experimental results provided in theresearch carried out by Crook, A. W in1961 [15].• It is also concluded that with increase in theclearance between the surface the value offilm thickness increases this can also be13 th International Conference on Tribology – Serbiatrib’13 255

• explained with the direct relationshipbetween the film thickness and theclearance as provided by Dong Zhu [7, 8].• The experimental results already availablefor the present problem are for the case ofKrytox 226 while keeping the clearancelevel of 0.00002 are 0.25 µm that is closein accordance with the numericallycalculated value of 0.265 µm with an errorof 6%.• The summarized graph also show that thegreater the value of the viscosity the greaterthe slope of the line which show that theeffect of clearance between the contactsurfaces increases with increase in theviscosity of the lubricant.6. CONCLUSION:Based on the above discussion it is concludedthat the effect of clearance between the slidingsurfaces under Elastohydrodynamic lubricationregime increases with increase in the viscosity ofthe lubricant and also that the increase in the thrustforce requires a more thick layer of lubricant to beprovided between the surfaces to avoid mechanicalwear. The comparison of results with theexperimental data and the analytical data ofscholars from past, the applicability of the modifiedDong Zhu’s model for present case of slidingsurfaces subjected to high acceleration is verified.The present work can be extended to incorporatethe roughness of the contacting surfaces as well asto draw a comparison of Elastohydrodynamic andhydrodynamic lubrication regime’s circumstances.REFERENCES[1] B. J. Hamrok, S. R. Schmid, B. Jacobson:Fundamental of Fluid Film Lubrication. MarcelDekker Inc., 2004.[2] O. Reynolds: On the Theory of Lubrication and itsApplication to Mr. Beauchamp Tower's ExperimentsIncluding an Experimental Determination of theViscosity of Olive Oil, Phil. Trans., Roy. Soc.London 177, pp: 157-234, 1886.[3] G.M. Hamilton, S.L. Moore: The Lubrication ofPiston Rings--First Paper--Measurement of the Oil-Film Thickness Between the Piston Rings and Linerof a Small Diesel Engine, IMechE 188, pp: 253-261,1974.[4] B.J. Hamrock, D. Dowson: IsothermalElastohydrodynamic Lubrication of Point Contacts.Part III -Fully Flooded Results, J. Lubr. Technol.99, pp: 264-276, 1977.[5] Morales-Espejel, E. Guillermo, A.W. Wemekamp:"Ertel–Grubin methods in Elastohydrodynamiclubrication a review." Proceedings of the Institutionof Mechanical Engineers, Part J: Journal ofEngineering Tribology 222.1, pp: 15-34, 2008.[6] D. Dowson, G.R. Higginson, ElastohydrodynamicLubrication, Pergamon Press, Oxford, 1977.[7] D. Zhu, H.S. Cheng, A. Takayuki, K. Hamai: ANumerical Analysis for Piston Skirts in MixedLubrication-Part I: Basic Modeling ASME Journalof Tribology, 114, pp: 553-562, 1992.[8] D.Zhu, H.S. Cheng, A. Takayuki, K. Hamai: ANumerical Analysis for Piston Skirts in MixedLubrication-Part II: Deformation ConsiderationASME Journal of Tribology, 115, pp: 125-133,1993.[9] D. Zhu: Elastohydrodynamic Lubrication inExtended Parameter Ranges—Part I: Speed Effect,Tribology transactions, 45.4, pp: 540-548, 2002.[10] D. Zhu: Elastohydrodynamic Lubrication inExtended Parameter Ranges—Part II: Load Effect,Tribology transactions, 45.4, p: 549-555, 2002.[11] D. Zhu: Elastohydrodynamic Lubrication inExtended Parameter Ranges—Part III: EllipticityEffect, Tribology transactions, 46.4, pp: 585-591,2003.[12] D.K. Kankane, S.N. Ranade: Computation of InboreVelocity-time and Travel-time profiles fromBreech Pressure Measurements, Defence ScienceJournal, 53.4, pp: 351-356, 2003.[13] C. Barus, Isotherms, Isopiestics and IsometricsRelative to Viscosity, American Journal of Science45, pp: 87-96, 1893.[14] P.S.Y. Chu, A. Cameron, Pressure ViscosityCharacteristics of Lubricating Oils, Journal of theInstitute of Petroleum 48, pp: 147-155, 1962.[15] A.W. Crook, The lubrication of rollers II. Filmthickness with relation to viscosity and speed. Phil.Transactions of the Royal Society of London. SeriesA, Mathematical and Physical Sciences 254.1040,pp: 223-236, 1961.[16] G.W. Stachowiak, A.W. Batchelor, EngineeringTribology. Butterworth-Heinemann, 2005.256 13 th International Conference on Tribology – Serbiatrib’13

For Thrust Force of 40KNKrytox 215 (µ = 0.03204 Pa.s)Figure 1. Clearance = 0.00001 Figure 2. Clearance = 0.00002Figure 3. Clearance = 0.00003 Figure 4. Clearance = 0.00004Krytox 226 (µ = 0.0445 Pa.s)Figure 1. Clearance = 0.00001 Figure 2. Clearance = 0.0000213 th International Conference on Tribology – Serbiatrib’13 257

Figure 3. Clearance = 0.00003 Figure 4. Clearance = 0.00004Krytox 227 (µ = 0.07476 Pa.s)Figure 1. Clearance = 0.00001 Figure 2. Clearance = 0.00002Figure 3. Clearance = 0.00003 Figure 4. Clearance = 0.00004For Thrust Force of 32KN258 13 th International Conference on Tribology – Serbiatrib’13

Krytox 215 (µ = 0.03204 Pa.s)Figure 1. Clearance = 0.00001 Figure 2. Clearance = 0.00002Figure 3. Clearance = 0.00003 Figure 4. Clearance = 0.00004Krytox 226 (µ = 0.0445 Pa.s)Figure 1. Clearance = 0.00001 Figure 2. Clearance = 0.0000213 th International Conference on Tribology – Serbiatrib’13 259

Figure 3. Clearance = 0.00003 Figure 4. Clearance = 0.00004Krytox 227 (µ = 0.07476 Pa.s)Figure 1. Clearance = 0.00001 Figure 2. Clearance = 0.00002Figure 3. Clearance = 0.00003 Figure 4. Clearance = 0.00004260 13 th International Conference on Tribology – Serbiatrib’13

Serbian TribologySocietySERBIATRIB ‘1313 th International Conference onTribologyKragujevac, Serbia, 15 – 17 May 2013Faculty of Engineeringin KragujevacINCREASING OF TOOL LIFE FOR HOT FORGING USINGSURFACE MODIFICATIONMilentije Stefanović 1 , Dragan Džunić 1 , Vesna Mandić 1Srbislav Aleksandrović 1 , Dragan Adamović 1 , Slobodan Mitrović 11 Faculty of Engineering University of Kragujevac, Serbia, stefan@kg.ac.rs, dzuna@kg.ac.rs, mandic@kg.ac.rs,srba@kg.ac.rs, adam@kg.ac.rs, boban@kg.ac.rsAbstract: Techno-economic indicators in the hot forging of steel and other materials are highly dependenton the total life of forging tools, that is, the number of forged parts requires accuracy of regeneration afteretching. Key influences on the life of tools, like in every tribo-system are: the pieces of material, geometryand material tools and machines for forging and environmental conditions. Characteristics of hot forginghigh temperatures are in contact materials and tools, and local high working pressures, the dynamiccharacter of the load tools etc. Tool life is usually limited to the complex mechanisms of wear and tear, as aconsequence of cyclic loading, such as abrasive and adhesive wear and, thermal and mechanical fatigue,and plastic deformation. This paper presents an overview of opportunities for increasing the life of forgingtools by modern techniques for modifying the working surfaces of tools, according to comparative results ofdifferent methods, and gives appropriate recommendations.Keywords: Hot forging, tool life, wear, coatings1. INTRODUCTIONIn general, forging entails the sequentialdeformation of the workpiece material through anumber of different processes. Furthermore, eachforging operation comprises all the input variablessuch as billet material, dies, the conditions at thedie-workpiece interface, the mechanics of shapechange in the workzone, and the characteristics ofthe processing equipment, as illustrated in Fig. 1 [1]Thus, in designing and developing bulk metalforming processes, key technical problem areas thatmust be addressed include:- workipiece material-shape and size, chemicalcomposition and microstructure, flow propertiesunder processing conditions (flow stress in functionof strain, strain rate and temperature), thermal andphysical properties- dies or tools-geometry, surface conditions,material and hardness, surface coating, temperature,stiffness and accuracy- interface conditions -surface finish, lubrication,friction, heat transfer- workzone - mechanics of plastic deformation,material flow, stresses, velocities, temperatures- equipment used -speed, production rate, force andenergy capabilities, rigidity and accuracy .The understanding of these variables allows theprediction of the characteristics of the formedproduct, i.e., geometry and tolerances, surfacefinish, microstructure and properties.Figure 1. Variables of a bulk forming process [1]13 th International Conference on Tribology – Serbiatrib’13 261

Tools for hot forging are used in extremelydifficult conditions: impact and very highmechanical load, variable partial and generalthermical load, friction on the surfaces of tools andmore. Each tribological system in the field of hotforging is characterized by the following elements[2]:- contact pair, which consists of tool andworkpiece-metal that is plastically deformed, withappropriate structural and mechanical properties,- lubricant with properties relevant for hotprocess,- a machine that implements processing,- micro and macro environmentThe final characteristics of the finished pieceforgings,such as size, shape, surface quality andstructure depend on the values of systemparameters. Figure 2. shows a global approach tothe modeling of machining processes, whereas thequality of lubricants is main output.Figure 2. Tribo-modelling of hot forging process [2]2. CAUSES OF DAMAGE TO THE TOOLSTools for hot forging function during cyclicmechanical and thermal loads. The complexprocess of tear occurs as a result of the two loads.Due to the high pressures and inadequatelubrication regime, the intensity of wear and teardue to friction is high. The main causes of damageto the tool, Figure 3:- Wear as a result of tribological processes- Plastic deformation,- Thermal cracks,- Mechanical fatigue crack growth,- Breakdowns (die failures).Figure 3. Failure and damaging of forging diesTool wear. This type of wear occurs as a resultof relating the material particles from the surface oftools. Consequence of the occurrence of abrasiveand adhesion processes and the formation ofwelded layers that are later destroyed.Plastic deformation occurs under the influenceof high pressure to the tool. As a result of resistmaterials, the deformation of the walls of tools andmeasuring tools changes.Thermal cracks. They are caused by cyclicthermal change and show it as a grid cracks on thesurface etching. Microcracks are connected andgrow into large thermal cracking. They can grow onthe surface and in depth.Mechanical cracks are caused by mechanicalloading tools. Consequence of fatigue, pretreatmentand initial thermal cracking. These arecalled large cracks.Failure of tools is the result of hidden defects,bad thermal processing, errors in design, irregularexploitation and so on.Generally, due to the forging temperature beingwell above 1000 ◦C, the temperature of the surfaceof the tool temporarily exceeds 500 ◦C and thus thetempering temperatures of conventional hot worktool steel. In such a case, the hardness of the tool isreduced and the mechanical impacts during forgingoperations can easily cause plastic deformation aswell as abrasion of tool material, Figure 4 [3].Life of forging tools is a complex function ofseveral parameters, the most important being:structural, method design and exploitationconditions. Figure 5. demonstrates an interactionbetween these parameters.Based on the analysis of tool wear, we can givethe following recommendations for the design oftools [4]:- The choice of tool material with high resistance toabrasion,- The application of appropriate methods forimproving surface properties of tools,262 13 th International Conference on Tribology – Serbiatrib’13

- Reliable exploitation tools, primarily because ofthe importance of lubrication,- Use of an active role of friction,- New design tool that allows the hydrostatic orhydrodynamic lubrication.Figure 4. Microstructure of the convex radius ofdifferent hot work tool steels after 1000 forging cycles.Tool temperature, 200 ◦C; forging material, C45; forgingtemperature, 1100–1150 ◦C; lubricated contact; cycletime, 13 s; hardness of tools, 47 HRC.[3]Figure 5 . The interaction of construction, technologyand the industrial conditions of working tools [4]3. METHODS OF INCREASING THE LIFEOF TOOLSThe predominant causes of damage are the resultof tool wear and thermal and thermal-mechanicaldamage. Forging dies are exposed to highmechanical loads, which are accompanied byextreme thermal and tribological load for a verynarrow surface layer. As a result of preheating steelparts at temperatures above 1000 ° C, the size of thehard oxide particles come in contact zone causingvery strong abrasive wear. At elevated temperaturestribological conditions are very favorable for theprocesses of adhesion and a transfer of material toand from the surface of the tool. During executionof cyclic forging process heat is transferred due tocontact with the heated work pieces followed bycooling lubricant tools in the form of spray, whichis performed at room temperature. In this way,thermal shocks occurring within the tool material,resulting in high internal stresses initiate cracking.Further development of the cracks formed inparallel to the contact surface may lead to itspeeling, when these cracks meet with cracks normalto the contact surface. In case of insufficientcooling of tools, it is possible to slightly releasetool steel [5].To increase the life of hot forging tools,different surface modification techniques are usedsuch as welding, thermal application,electrodeposition, diffusion method and othercombined methods. These special methods ofsurface modification increase the hardness, wearresistance and corrosion resistance of the tool athigh temperatures. One possible way to satisfy allthe conditions of hot forging process is acombination of thermo-chemical surfacemodification process (ie, carbonization andnitriding) with coating processes (ie, PVD, CVD, orPACVD), which is known in the literature asduplex modification process surface [6].Figure 6 shows the results of testing the wearalloy tool steel to functioning in a warm envirnment(WNL-55NiCrMoV6) using various surfacetreatments: nitriding, sulphurizing, diffusionchroming, Cr plating, plasma spraying withmetallic coatings of Cr, WFe, WC-types,burnishing [4].Using spray-metal coatings leads to triplereduction of wear compared to conventionaltreatment tools. It also shows that the wear of thetool depends directly on the oxide that is generatedon the surface of pieces and to a lesser extent on theoxide surface tools. Cr and WC plasma sprayedcoatings should be applies to hot working tools.These coatings are characterized by considerablewear ressistance and thermal and imact fatiqueresistance.According to the investigations [7] the bestresults for die service life were obtained for weldoverlay coated dies, Table 1. Compared to receiveddies, the results showed an increase of 892%. Theresults were 206% better than TOKTEK Coatings,which held the second place.13 th International Conference on Tribology – Serbiatrib’13 263

Figure 6. The influence of surface treatment andworking on wear [4]The dies can be ranged from the best to worst asweld overlay coated dies, multilayer dies TOKTEKcoated, single layer AlTiN coated dies, plasmanitrided dies and dies as received. This range is alsovalid for the die polishing life, with the exceptionthat it is equal for TOKTEK and AlTiN coatings.Table 1. Number of polishing and total number of partsobtained with experimented dies [8]1stpolishing2ndpolishing3rdpolishing4thpolishing5thpolishingTotalprod.Asreceiv.OperationNitrided4. CONCLUSIONNumber of Forged PartsAlTiNcoatedTOKTEKWeld overlaycoated1st 2ndforg. forg.1440 3920 4810 4660 8690 83301260 2140 1980 2030 8160 84101320 2110 1860 1750 8240 7830- - 1720 1830- - 17804290 9420 12280132102781025930Tools for hot forging are applied in extremelydifficult conditions: impact and very highmechanical load, variable partial and generalthermical load, friction on the surfaces of tools andmore.With these tools, a special kind of coating oncontact surfaces is regularly applied, in order toextend their working life. There are very differentmethods for modifying the tool surface. The mostcommon are different types of heat treatment,which significantly affects the hardness of thesurface layers.Shown results are related to single-layer andmultilayer coatings. It's hard to make a generalconclusion regarding the most influentialparameters, but it is certain that the surfaceroughness has a significant effect on the adhesioncharacteristics of the coating. Also the surfacehardness to which the coating is applied is veryimportant as is the ability to carry the load inquestion.Coatings based on Cr and WC, and weldingprocedures show the best results. From thestandpoint of economy of the process of forging,welding is the best. The temperature and thepercentage of nitrogen in the nitration are veryimportant for tribological characteristics of contactlayers.The ratio of alloying components of chromium(Cr), molybdenum (Mo), and vanadium to carbonratio (H/C) is highly significant influence on thechange of tribological behavior of tool materialregarding wear resistance.ACKNOWLEDGEMENTThe authors wish to acknowledge the financialsupport from the Ministry of Education and Scienceof the Republic Serbia through the project TR34002.REFERENCES[1] V. Vazquez, T. Altan: New concepts in die design -physical and computer modeling applications,Journal of MPT 98, pp. 212-223, 2000.[2] M. Stefanović, R. Stanković: Importance ofTribological Investigations in the Hot Forging,YUTRIB 1995 Herceg Novi, pp, 181-182, 1995.[3] Barrau, C. Boher, R. Gras, F. Rezai-Aria: Analysisof the friction and wear behaviour of hot work toolsteel for forging, Wear, Vol. 255, pp. 1444–1454,2003.[4] M.Gierzynska-Dolna: Effect of the Surface Layer inIncreasing the Life of Tools for Plastic Working:Journal of MWT 6, pp. 193-204, 1982.[5] J. H. Beynon: Tribology of hot metal forming,Tribology International 31, pp. 73–77, 1998.[6] B. Navinsek, P. Panjan, F. Gorenjak: Improvementof hot forging manufacturing with PVD andDUPLEX coatings, Surf. Coat. Technol. 137,pp. 255–264, 2001.[7] M. Bayramoglua, H. Polat, N. Geren: Cost andperformance evaluation of different surface treateddies for hot forging process, Journal of MPT 205,pp. 394–403, 2008.264 13 th International Conference on Tribology – Serbiatrib’13

Serbian TribologySocietySERBIATRIB ‘1313 th International Conference onTribologyKragujevac, Serbia, 15 – 17 May 2013Faculty of Engineeringin KragujevacANALYSIS OF TRIBOLOGICAL PROCESS DURING IRONINGOF SHEET METAL MADE OF AlMg3Dragan Adamović 1 , Milentije Stefanović 1 , Srbislav Aleksandrović 1 ,Miroslav Živković 1 , Fatima Živić 1 , Marko Topalović 11 Fakultet inženjerskih nauka u Kragujevcu, S. Janjić 6, 34000 Kragujevac, adam@kg.ac.rs, stefan@kg.ac.rs,srba@kg.ac.rs, zile@kg.ac.rs, zivic@kg.ac.rs,Abstract: The paper gives detailed analysis of tribological processes which occur in ironing, their influenceon tool and formed material by means of specially developed physical model of drawing process. The typicalproperty of such cold forming procedures is multiple repetitions of operations, i.e. several thinnings in oneoperation, which leads to gradual increase of thickness of glued material layer on contact surfaces of thetool. At sufficiently large thickness of glued layers, plastic forming of created glued particles occurs, i.e.their tear-off, disruption of forming process stability and increase of surface roughness of work piece. Theobtained results indicate the basic influence of tribo-conditions on ironing process, tool durability andquality of obtained parts.Key words: Ironing, tribo-modeling, tribological processes1. INTRODUCTIONFriction accompanies all physical processes thatinvolve movement. It is also very important inmetal forming, during which only external frictionis considered. External friction occurs betweenmaterial which is being plastically deformed andmaterial of tool.External friction shows significant influence onthe course of plastic deformation, and thus theuseful properties of the final product as well as thetool lifetime. The friction force, i.e. its contactcomponents, show a significant influence on thestress field in a deformed metal, especially in itsouter layers that came in contact with work surfaceof tool. Stress field has influence on the course ofmetal forming, and thus the movement of deformedmetal surface on the acting surface of tool, and thismovement have influence on the friction force.During this process, certain types of reverse elasticreleases occur as well.This type of resulted external friction betweenthe plastically deformed metals and tool(technically dry friction with distinct adhesioneffect between contact surfaces, coupled withboundary friction), significantly influences thequality of the product. If the adhesions (compoundsformed by friction due to the "cold welding") areformed on surface of the tool they can be causes ofscratch and allowances on the surface of theproduct that deteriorates its quality.As a result of the friction notable changes incharacteristic of outer layers occur, with differentcharacter of the changes that happens on theproducts surface layers in comparison to changeson tool surface.The most of the work piece surface is in contactwith the working surface of the tool and will haveshare in the friction process only once, while toolsurface takes part in this process multiple times.The characteristics of tool deformation and thework piece deformation are also different. Theouter layer of the work piece (as well as the entirevolume), has lower yield stress than tool, resultingin plastic deformations, while, at the same time,tool generally remains in the zone of elasticdeformation. Given the fact that surface layer hasthe highest stress gradient, properties of outerlayers, for both product and tools, will be differentfrom properties of other parts of product and tools.As a result of the friction, tool wear occurs.Mechanism and the intensity of tool wear arefunctions of friction force magnitude and type ofthe friction [1].13 th International Conference on Tribology – Serbiatrib’13 265

Model of the phenomena that occurs in microzonesof contact during friction is shown in Figure1 [2]. One can distinguish three main stages ofinteraction between contact surfaces during frictionStage I. At this stage contact occurs as well asmechanical action between contact surfaces whichare initially covered with oxide layers, whereby theamount of oxide depends largely on the type ofprocesses (metal forming of hot or cold), andsusceptibility to oxidation of the metals that formcontact pairs. The dominant phenomenon at thisstage of contact is the plastic deformation, not onlyof the surface roughness but also, and in theconsiderable volume of the material.Stage IStage IIStage IIIFigure 1. Model of the phenomena that occur in the micro-areas of contact during friction [1]Stage II. As a result of molecular interaction,adhesive joints (compounds) are formed. Thequantity of joints depends largely on the geometryof the contact and the specific pressure.Stage III. This stage of the contact surfaceinteraction includes the destruction of adhesionjoints which were formed during the relativedisplacement of contact pair metals. Failuremechanism of contact joints can be very complex.In the first stage of joint destruction micro-slip willsurely occur, and therefore the complex phenomenaof movements and mutually dependent movementof dislocations. As a result of these phenomenasurface defects such as micro-cracks and micronotchescan be created.As a result of repeated displacement ofdeformed metals in regards to the surface of thetool, the effect of friction and the associatedforming and destruction of adhesive joints, toolwear occurs.In the case of metal forming, process ischaracterized by the fact that multiple repeatedoperations (forging or ironing) causes a gradualincrease in thickness of glued layer, which meansthat the sum of the individual joints goes into acontinuous layer.As a result of predominance of adhesive forceover resistance to plastic flow in the glued layer, infurther stage glued particles suffer from plasticdeformation until they are torn off and smeared.When a sufficiently large thickness of gluedlayers is formed as a result of repeated process ofplastic deformation (in a series of passages whichleads to increasing and decreasing of mutualinteraction), process begins which leads to theseparation of adhesive joint from tool surface bypeeling (shear) or tearing. This leads to thesignificant damage of the tool surface layers, andtherefore to the increase in surface roughness. Theoccurrence of peeling or tearing depends on thetype of formed joint.In case of contact pairs with higher chemicalaffinity, the strength of diffusion produced joints(solid solutions) may be greater than strength ofmaterial of contact pairs, causing the destruction ofjoint to occur in depth of less strong and nofortified material, which means it occurs at such adepths at which there is no more squeezing.In the case of diffusion-less joints, as well as forthe occurrence of brittle inter-metallic phases, thedestruction of joint will be based mainly on layeredpeeling of metal with less strength. Tearing off ofglued particles also occurs. Further relativedisplacement of contact elements (the work piece isplastically deformed in relation to the tool), makesthese glued particles to reappear on the surface ofthe friction. Afterwards they are compressed on thetool surface, which results in creation of groovingin "partner" made of material with less hardness(usually plastically deformed work piece). The typeand intensity of this secondary effect depends onthe hardness of the plastically deformed andstrengthened adhesive joints.266 13 th International Conference on Tribology – Serbiatrib’13

In addition, as a result of cyclic loading of thetool, on its surface there might be occurrence ofsuch defects such as intrusion and extrusion, as wellas micro-cracks, which are characteristic of themetal fatigue process.2. EXPERIMENTAL TESTSTests were conducted on the original tribomodelof ironing, which simulate the two-sidedh=max 70 mm8TF ir/262vF ir"A"F frP49F ir/21 - Die support2 - "Die"3 - "Punch" body4 - "Punch" front5 - Gauge with measuring tapes6 - Plates7 - Sheet metal strip (Test piece)8 - Termocouple9 - Potentiometer travel gauge1F Dsymmetrical zone of contact with the die and thepunch [3]. This model enables the realization ofhigh contact pressures with the respect of physicaland geometric conditions of the real process (thematerial of the die and material of the punch, thetopography of the contact surface, the angle of thedie cone - , etc.). Diagram and image of theaforementioned tribo-model is shown in Figure 2.175632Figure 2. Diagram and image of the tribo-model used in this researchDevice for testing of ironing was installed on aspecial machine designed for sheet metal testingERICHSEN 142/12.For the experiments presented in this paper, weused sheet made of aluminum alloy, AlMg3(.43)(according to EN: AlMg3 F24, and in text belowonly AlMg3). Mechanical properties of testedmaterial are given in Table 1.Table 1. Mechanical properties of tested materialMaterijalRp,MPaRm,MPaA,%AlMg3 201.1 251.0 12.0 0.135 0.405Contact pairs ("die" and "punch") are made ofalloyed tool steel with high toughness and strength,designated as Č4750 (EN: X160CrMoV121).3. EXPERIMENTAL RESULTSDuring ironing friction coefficients of die andpunch can have a wave-like (unstable) form (Fig.3). Friction coefficients alternately rise and fallwith irregular and approximately the sameamplitude and frequency. At a particular time,friction coefficients can have slightly increasing,constant or slightly decreasing flow.Very interesting explanation of this type offriction is given in the papers [4, 5]. It is consideredthat the wavy type friction coefficients occur whenthere is a micro-welding of roughness peaks in the"form of the islands".n,-r,-Friction coefficient on die, -Friction coefficient on punch, -0.200.150.100.050.000 10 20 30 40 50Ironing travel, h, mm0.50.40.30.20.1abba0.00 10 20 30 40 50Ironing travel, h, mmFigure 3. Examples of unstable friction coefficients onthe side of the die and punch: a-constant, b-decreasing,c-growingaSubsurface layer of the contact surface on theexit part of the die, which surface is about 80% ofthe total contact area between the tool and thematerial, suffers considerable distortion due toshear stress which is result of friction forces. Thisstress is approximately equal to the shear stress inthe weld zone. This zone is therefore called the"zone of quasi-welding". At the entrance part of thebc13 th International Conference on Tribology – Serbiatrib’13 267

die, where there is a layer of the lubricant which isnot yet squeezed out, there is a formation of socalled"nipple slip". During this process quasi-weldzone is steadily increased and thus coefficient offriction is increased as well. When the surface ofquasi-weld zone become equal to the entire frictionsurface, friction coefficient reaches its maximumvalue. In addition, due to strong frictionconnections, micro cracks in the subsurface layerare formed. Due to the continuous material inflowin the zone of deformation noticeable nipple isformed, which at some point, because of crackscaused by the weakening of the frictionalconnection, detach itself from the base material. Inthis way the quasi-welded zone is reduced andlubricant starts to penetrate places of brokenconnections, which reduces the friction coefficient.Chipped of metal fragments are trapped betweenthe die and the surface of sheet metal and are beingcontinuously moved towards the exit part of the die.When they came out of the zone of deformationcoefficient of friction will have a minimum value.Then, the aforementioned process continuouslyrepeats itself.If the inadequate lubricant is used coupled withgreater gripping forces, stickers are formed andcontact conditions are greatly deteriorated whichlead to the significant increase of drawing force foreach subsequent passage (Fig. 6).Sila izvlačenja, kN2 µm14121086420.25 mmFigure 5. Grooved surfaceIVIIIIIIAlMg300 10 20 30 40 50Put klizanja, mmFigure 6. Change of drawing force4. CONCLUSIONFigure 4. Aluminum glued particles on die and punchersurfacesSome lubricants, no matter the fact that theyproduce satisfactory results in steel plates, in caseof plates made of AlMg3 have very poor results.Their usage leads to intense gluing of aluminumonto tool, which is shown in Figure 4. Gluedparticles that are formed on the die during theironing can cause severe damage to the sheet metalsurface (galling) (Figure 5).In the case of ironing one of the maincharacteristic of this processes is the fact thatmultiple repeated operations lead to a gradualincrease in thickness of glued layer, which meansthat the sum of the individual resulting layerscrosses over into the continuous layer.As a result of predominance of the adhesionforces over resistance to plastic flow within gluedlayer plastic deformation of glued particles occursfollowed by their tearing off and smearing.When a sufficiently large thickness of the gluedlayers is achieved, as a result of the multiplerepeated process of plastic deformation, process ofseparating glued particle from tool material begins.This separation is done by pealing (shear) or tearingoff, and creates significant damage to the surfacelayers of tool, and therefore increases the surfaceroughness of the work piece. Such processes arecharacteristic of ironing sheet metal made ofaluminum alloys, where no adequate lubricant isapplied.268 13 th International Conference on Tribology – Serbiatrib’13

AcknowledgementThe part of this research is supported byMinistry of Education and Science, Republic ofSerbia, Grant TR32036 and TR34002REFERENCES[1] Леванов А.Н, Колмогоров В.Л. и др.:Контактное трение в процессах обработкиметаллов давлением, Металлургия, 1976.[2] Gierzynska M.: Tarcie zužycie i smarowanie wobrobce plastycznej metali, WNT, Warszawa, 1983.[3] Adamović D.: Ponašanje materijala u kontaktu priprocesima hladnog plastičnog oblikovanja savisokim radnim pritiscima, Doktorska disertacija,Mašinski fakultet u Kragujevcu, Kragujevac, 2002.[4] Kawai N., Nakamura T., Seko S.: The weldingmechanism in drawing of aluminum sheet, Journalof Engineering for Industry, Vol. 102, No. 8, pp.229-238, 1980.[5] Kawai N.: Laboratory simulation for galling inmetal forming, metal transfer and galling in metallicsystems, Proc. Conf. Orlando, Florida, Metal. Soc.AIME, pp.63-86, 1987.13 th International Conference on Tribology – Serbiatrib’13 269

Serbian TribologySocietySERBIATRIB‘1313th International Conference onTribologyKragujevac,Serbia, 15 – 17 May 2013Faculty of Engineeringin KragujevaccOPTIMAL DESIGN OFA CAMMECHANISMWITHTRANSLATING FLAT-FACE FOLLOWERUSINGG GENETICALGORITHMI. Tsiafis 1 , S. Mitsi 1 , K.D. Bouzakis 1 , A. Papadimitriou 11 AristotleUniversity of Thessaloniki, Departmentt of Mechanical Engineering, Greece,tsiafis@eng.auth.gr, mitsi@eng.auth.gr,bouzakis@@eng.auth.gr, thanpap22@gmail.comAbstract: The optimumm design of f a cam mechanism is a time consuming task, due to the numerousalternativesconsiderations. In the present work, the problem of designn parameterss optimization of a cammechanism with translating flat-facefollower is investigatedfrom a multi-objective point of view. The designparameters,just like the cam basecircle radius, the follower face width and thee follower offset can bedetermined consideringas optimization criteriaa the minimization of the cam size, off the input torque and ofthe contact stress. During the optimization procedure, a number of constraints regarding the pressure angle,the contact stress, etcare taken into account. The optimization approach, based on genetic algorithm, isapplied to find the optimal solutionss with respect to the afore-mentionedd objective function and to t ensure thekinematic requirements. Finally, the dynamic behaviour of the designed cam mechanism is investigatedconsideringthe frictional forces.Keywords: Cam mechanism, geneticalgorithms, , contact stress, optimization.1. INTRODUCTIONThe optimal design of cam mechanismm ishandled in many publications [1-5], where variousconstraints and methods are considered. Αnon-known as SUMT algorithm is used in [3]foroptimum synthesis of a disk cam mechanism withswinging roller follower.In [4] the designparameters are determined by the minimizationn oflinear programming technique with constraints,the maximum compressive stress at the contact areaof a cam-disk mechanism with translating rollerfollower, where the camprofile is described withthe aid of cubic spline functions. Tsiafis ett al.present in [ 5] a multi-objective procedure basedd ongenetic algorithmsto optimize the designparameters of a disk-cam mechanism with a rollerfollower.In the present paperthe problemof the designparameters optimizationof a cam mechanism witha reciprocating flat-face follower is investigated,using multi-objectiveoptimizationwith geneticalgorithm. The design parameters for this typee ofmechanism are the radius of the cam base circle,the follower face width andd the follower offset. Theoptimization iss achieved by the development ofprograms usingg the high level computing languageMATLAB withh the GA (genetic algorithm) toolboxapplication. Furthermore, the dynamical analysis ofthe designed mechanism m consideringfriction isinvestigated.2. MATHEMATICAL FORMULATIONA cam mechanism withh a translating flat-facefollower is shown in figuree 1. The camis assumedto have constant angular velocity. The profile of thecamcan be determined considering thekinematicalanddynamical requirementss of the mechanism.The design parameters under optimization arethe cam base circle c R b , thee width follower face Landthe follower offset e as shown in figure 1.The optimization of the design parameters of thecammechanism can beachieved by theminimisation off the cam size, of the torque requiredto drive the camm and the contact stress between thecamand the follower.27013 th International Conference C onn Tribology – Serbiatrib’13

Thepressure angle can be calculated by[1]:Figure 1. Cam mechanism with translating flat-facefollower.Therefore, it could be formulated as anoptimizationproblem, where the objective function(F) takes into account the cam size(F 1 ), the inputtorque (F 2 ) and the maximum contact stress (F 3 )::WithF 2 FF 1σ max1F P'21μ11μ ρ E1E2where T is the input torque, P is the t total normalload on the cam, vis thefollower velocity, ω iss thecamshaft angular velocity, σ max iss the maximumcontact stress between the follower and the cam, P’is the normal load per unit width of the contactingmembers, ρ is the radii of curvature of the cam,μ 1 and μ 2 arePoisson’s ratio for the cam and thefollower respectively and E 1 , E 2 arethe modulee ofelasticity ofthe cam andthe follower respectively.The weighting factors α, β and γ are used in orderto scale the contributionn of the corresponding termsin the objective function value. The minimizationof the objective function determines the optimumvalues of the unknown parameters. During theoptimizationproceduree the following functionalconstraints are imposed:a) The maximum value of the pressure anglemust be smaller than the maximum permitted:δ max

An important issue in genetic algorithms is thetreatment of constraints. For each solution of f thepopulation, the objectivefitnessvalues arecalculated. Furthermore, every solution is checkedfor constraints violation.4. NUMERICAL APPLICATIONThe introduced methodology is applied to findthe design parameters of a cam mechanism withtranslating flat-face follower where the followeroffset is set equal to zero (e=0).Figure 2 shows the kinematic equirements pertransient region of the indicated in this figurefollower displacement diagram.The functional requirements and the t materialproperties usedd in this investigation areinserted inFigure 3.The parameters involved in all tests, mainly in GAprocedure, are the same and selected as optimumsthroughmanyyapplied tests: populationofindividuals=20,cross probability=80%, elite count=2andthe maximum number off generations is 100.Consideringgkinematicrequirementsthedisplacement,velocity and acceleration of thefollower are determined (Figure 4).Figure 2. Kinematic requirements.Figure3. Materials properties andfunctionalrequirements.Figure 4. The follower motion diagrams.In general the weighting factors α, β and γ of thefitness functionn (1) are selected considering theimportance of the objectives that must beachieved bythe mechanism. A high value of the weighting factorα increases the importance off first part of the objectiveefunction (F 1 ) that is to obtainn a small camsize. Afterseveral tests the followingg weighting factors arechosen: α=0.1, , β=0.1 and γ=0.8. Running theMATLAB codes with above mentionedparameters,the following design parameters are obtained:R b =32.67 mm and L=53.21 mm. Forr constructed27213 th International Conference C onn Tribology – Serbiatrib’13

mechanism these parameters are finally set: R b b=35mm and L=50 mm.The camprofile is shown in Figure 5. The 3Dmodel of the designed cam mechanism is illustratedin Figure 6.Figure 5. Cam profile.Figure6: 3D model of the cam mechanism.5. FORCEANALYSISS OF CAM MECHANISSMCONSIDERING FRICTION FORCESIn this section the dynamic forceanalysis off thedesigned mechanism considering the friction forcebetween follower and its guide and the frictionforce between cam and flat face followerr isinvestigated.The force transmission of a radial cam with areciprocating flat-faced follower is shown in figure7, where P is the external load on the follower, μ isthe coefficient of friction between the follower stemand its guide, μ o the coefficient of friction f betweenthe cam andthe flat facefollower and d is the guidediameter.13 th International Conference on Tribology – Serbiatrib’13Figure 7: Forcee transmissionn of cam mechanism withtranslating flat-face follower.From the equilibrium e equations of horizontalandvertical forces and moments about the point Adandassuming that t difference the between μ N12dandμ N2is negligible, the forces Fc, N 1 and N 22are determined [1]:withandΝ2N1FCbPΓ aμ0ξb PΓ 0 a μ b 1 ξ PΓΓ b2aμ μμ0b12ξP ms cskwhere m is thee follower mass, s, s an nd s are thedisplacement,velocity and acceleration of thefollower respectively, c is the dampingcoefficient,k isthe spring constant, c s 0 is the initial compressionof the spring and F b is the follower weight.Furthermore, the friction forces arewritten as:Qμ F(11)0 Q 1Q 2μ μNμ μNandthe cam shaft torque due to the frictionis given by:d dTf Q 0 Rb s Q1 Q2(14)2 20ss F012b(8)(9)(10)(12)(13)273

considering thee spring force greater than inertiaforce corresponding to maximum deceleration, inorder to avoid the jump phenomenon.The parameter ξ is determined with the relation:ξ= (15-s)/b.The diagramm of friction forces versus cam angleis illustrated in Figure 9.In Figure 100 is inserted the diagram of the inputtorque with andd without friction.6.CONCLUSIONFigure 8. Constructed cam mechanism.Figure 9. Friction forces of cam mechanism withtranslatingflat-face followerFigure 10. Input torque with and without friction.In the designed and constructed mechanism(Figure 8) the data used in dynamic analysis are:μ=0.78, μ 0 =0.15, m=1 kg, k=3004 N/m, s 0 =13 mm,d= 50 mm and b=50 mm.1000k mThe damping coefficient is c 2ζ1000with ζ=0.1. The spring constant k is chosen274In this paper the optimization of f the designparameters of a cam mechanism with a flat-facedfollower is approached. For this taskthe multi-objective optimization with genetic algorithm isapplied using the high levell programming languageof MATLAB. The T optimization satisfies constraintsswhich are made in order to operate a cammechanism properly. This procedure is automatic,gives results fast and it appears to be reliable. r Thefinal results provide usefull informationn for a cammechanism synthesis and can be used as a basis offinal preference dependingg on the objectives thatthave to be succeeded.Subsequently,after the cam mechanismmsynthesis, the applied a friction forces arecalculated.Themost important conclusion is the fact that thefriction forces are a analogous with the action of thefollower movement. This means that inthe areas ofdwell the friction forces aresteady, whereas in theareas of rise or return the friction forces alter in analmost similar way. wREFERENCES[1][2][3][4][5][6]Chen Y. F., Mechanics and Design of CamMechanisms,Pergamon Press, USA, 1982.Norton R. ., Cam design and manufacturinghandbook, Industrial I Press, Inc., New York, 2002Bouzakis K. K D., Mitsi S., Tsiafis I., Computeraided optimum design and NC milling of planarcam mechanisms, International Journal of MachineTools and Manufacture, ,Vol. 37, No 8, pp 1131-1142, 1997. .Mitsi, S., Bouzakis, B K. -D., Tsiafis, J., Mansour,G.,Optimal synthesis of cam mechanism usingcubic spline interpolation for cam NC milling.Journal of the Balkan Tribological Association,Vol. 7, No 4, 4 2001, pp. 225-233.Tsiafis, I., Paraskevopoulou, R., Bouzakis, K.-D.,Selectionn of optimal design parameters for acam mechanism usingg multi-objective geneticalgorithm, Annals of the “Constantin Brancusi”University of TarguJiu, Engineering series, nr. 2,Romania, 2009, pp. 57- 66.Coley D., An A introductionon to genetic algorithms forscientists and engineers, World Scientific Press, 1999.13 th International Conference C onn Tribology – Serbiatrib’13

Serbian TribologySocietySERBIATRIB ‘1313 th International Conference onTribologyKragujevac, Serbia, 15 – 17 May 2013Faculty of Engineeringin KragujevacINFLUENCE OF VARIOUS TYPES OF ROCK AGGREGATESON SELECTION OF THE WORKING PARTS MATERIALIN CIVIL ENGINEERINGV. Lazić 1 , M. Mutavdžić 2 , R. Nikolić 1, 3 , S. Aleksandrović 1 , D. Milosavljević 1 , B. Krstić 1 , R. Čukić 11 University of Kragujevac, Faculty of Engineering, Sestre Janjić 6, 34000 Kragujevac, Serbia2 Road Building Company Kragujevac, Tanaska Rajića 16, 34000 Kragujevac, Serbia3 University of Žilina, Faculty of Civil Engineering, Univerzitna 1, 010 26 Žilina, SlovakiaAbstract: In this paper are presented results of theoretical and experimental investigations and mutual comparison ofvarious types of rock aggregates from the aspect of working parts wear of different machines for preparation anddeposition of the rock materials on roads. Here are considered only the most important types of building stones: limestone,dolomite marble, calcite-dolomite marble and andesite, which are exploited at four sites in Republic of Serbia. Thoseaggregates are convenient for manufacturing of certain layers of the driveway constructions on roads, streets, airports, asthe base layer on railways and for preparation of various types of asphalt and concrete. In selection of rocks for depositingon roads, it is necessary to know both their general and specific properties. It is necessary to conduct the mineralogicalpetrographicand physical-mechanical investigations and, if needed, certain special ones, as well.The civil engineering machines for manufacturing and building-in materials in various structural objects are exposed todifferent types of high loads, what is especially true for some of their working parts, which come into direct contact withthe rock materials. The working life of those machines is directly dependent on the type of the building material as wellas on maintenance. In exploitation, the construction mechanization is subjected to various types of corrosion and wear,and some of its parts are occasionally subjected to impact loads, as well. Certain parts are also frequently in contactwith various types of stones, sands, soil, asphalt, concrete and occasionally with water.Key words: rock materials, aggregates, minerals, working parts, civil engineering mechanization.1. INTRODUCTIONCivil engineering machines for manufacturingand building in materials during construction ofvarious building objects are subjected to differenttypes of loads, especially some of their workingelements, which are in direct contact with rockmaterials. Working life of construction machines'parts is directly dependent on the kind of the rockmaterials, properties of construction mechanizationworking parts and exploitation and maintenanceconditions. Those machines are, during operation,exposed to different types of wear and corrosion,some working part are even exposed to occasionalimpact loads. Some parts of construction machinesare frequently in contact with various kinds ofrocks, sands, soil, asphalt, concrete, sometimes areeven exposed to influence of water.Knowing physical-mechanical properties ofthe rock minerals are of a special importance, bothfor their exploitation and for their processing andbuilding in. Since the matter of speaking are thecomplex tribo-mechanical processes, in whichtake part different elements of constructionmechanization, rocks and third objects, it isespecially important to properly select material ofthe construction machines' working parts, as wellas the technology for reparation of the damagedand worn parts of those machines.Based on investigations of the construction rocksfrom four available sites, the useful data wereobtained both for design and reparation of theworking parts of machines for minerals' exploitation,their processing and building into roads.2. THE MOST IMPORTANT TYPES ANDPROPERTIES OF ROCK MATERIALSRocks mainly consist of seven groups ofminerals: silicates, carbonates, oxides, sulphates,13 th International Conference on Tribology – Serbiatrib’13 275

sulphides, chlorides and hydroxides. To get amore complete picture about number of differentminerals that rocks are made of, it is necessary toemphasize that only the silicate group containsabout 800 minerals, categorized into varioussubgroups. Mineral masses in the Earth’s crustcan be found in forms of compounds – as solidrocks or in the unbound – dispersed form. Thusthe rocks are being divided, according to strength,into weak, solid and exceptionally solid rocks,since minerals can be in the crystal, crystallite oramorphous form. Rock properties can besignificantly changed due to action of water, frostor heat; thus it is highly important to know thelaws of those changes. The most importantproperties of rocks are petrographic, physical,mechanical and technological [1-8]. All the rocksthat are contained in the Earth's crust can beclassified in three major groups: magmatic,sediment and metamorphic rocks.In this paper are investigated properties ofstones from the four near-by sites: limestone,dolomite marble, calcite-dolomite marble andandesite. The most exploited (over 70 %) is thelimestone from the "Vučjak" site, thus this particularstone was taken as representative for rocks'properties experimental investigations.3. DETERMINATION OF PETROGRAPHICPROPERTIESThe most important petrographic properties ofrocks are: mineral composition, structure andtexture. The structure of rocks consists of mineralcrystallite grains, whose shape and strengthdepend on way of coalescence during the rockformation. The texture of rocks consists ofminerals spatial distribution and occupancy.Petrographic properties of rocks were testedaccording to standard SRPS.B.B8.002:1989.The limestone site "Vučjak" mainly consists oforganogenic-detritic 1 limestone. The rock's textureis massive. i.e., the mineral grains are notregularly distributed within the substrate. InFigure 1 are presented macroscopic andmicroscopic appearances of this rock [6].The dolomite marble site "Samar" mainly consistsof the dolomite marble. The structure of thistype of rock is granoblastic. Texture is massiveand noncompact. In Figure 2 are presentedmacroscopic and microscopic appearances of thisrock [6].The calcite-dolomite marble site "Gradac"mainly consists of this type of rocks. The rock's1 Detritus (latin): aggregate of small particles ofcrushed rockstexture is homogeneous and compact. In Figure 3are presented macroscopic and microscopicappearances of this rock [6].The andesite site "Šumnik" mainly consists ofthis type of rock. The rock's texture is massive. InFigure 4 are presented macroscopic and microscopicappearances of this rock [6].4. EXPERIMENTAL DETERMINATION OFSOME ROCKS PHYSICAL PROPERTIESFor construction of roads the most frequentlyexperimentally tested are the following rocks'physical properties: specific mass, bulk mass,porosity, water absorption and compactness, sincethose properties directly influence changes ofmechanical and technological properties of rocksand their aggregates.4.1. Determination of specific mass (density) ofrocksIn Table 1 are presented average values ofspecific mass of various rocks' samples [6].4.2. Determination of rocks' porosityWater absorption is defined by ratio of watermass and mass of solid mineral substance; for varioustypes of rocks it ranges from 1.5 to 4.4% [6].4.3. Determination of rock compactionpossibilitiesDetermination of the rock bulk mass can serveas criterion for estimates of possibility of theircompacting in the infilling state, what is of a greatimportance in construction of building objects.Complete investigation of compaction possibilityof rocks requires: determination of granulometriccomposition, compacting possibility by theProctor test and the Californian capacity index (cf.[1], [3-4], [7], [9 - SRPS B.B8.030: 1986]).4.4. Influence of water, low and elevatedtemperatures on rock content changesWater, low and high temperatures significantlyinfluence some of the rocks' properties. In contactwith water at low temperatures (< 0° C) and hightemperatures (>100° C), the minor changes inrocks are observed, while at temperatures lowerthan - 25° C and higher than 500° C, thesignificant changes occur.Low temperatures (frost) significantly affectrock properties, especially if the temperatures arevariable. Alternating heating and cooling of rocks276 13 th International Conference on Tribology – Serbiatrib’13

causes big changes of their properties. All the dryrocks endure well the low temperatures actions,while the wet rocks and rocks completelysaturated with water have significantly lowerresistance, since their destruction occurs due tofreezing of water in cavities.a) b)Figure 1. Appearance of limestone structure: a) macroscopic appearance, b) microscopic appearance.a) b)Figure 2. Appearance of dolomite marble structure: a) macroscopic appearance, b) microscopic appearance.a) b)Figure 3. Appearance of calcite-dolomite marble structure: a) macroscopic appearance, b) microscopic appearance.a) b)Figure 4. Appearance of andesite rock structure: a) macroscopic appearance, b) microscopic appearance.13 th International Conference on Tribology – Serbiatrib’13 277

Table 1. Bulk masses of tested rock samples.Tested rock propertyLimestone- VučjakDolomitemarble- SamarCalcitedolomitemarble- GradacAndesite -ŠumnikBulk mass with pores, V2690 2780 2820 2630v m sSpecific mass without pores and voids, m V 2730 2870 2850 2750sssBulk mass coefficient, i 0.985 0.969 0.989 0.956vsAll types of rocks well endure action of elevatedtemperatures up to 500° C. Further temperature increaseleads to visible changes: loss of characteristicringing sound at impact, significant decrease ofstrength and appearance of crumbling or totaldestruction if poured over with cold water. At hightemperatures (over 850° C) more durable are rocksmade of minerals whose heat conductivity is significantlydifferent. More durable are firm and finegrainedrocks than the porous and coarse ones. Atthat temperature granites crack irregularly, whilethe sand rocks usually crack parallel to stratification.Limestones and marbles possess good strengthup to the calcification temperature (~ 800° C),when they transform into quicklime (CaO). At hightemperatures quartzites, quartz sandstones withsilicon binder, clays, serpentinite, serpentinite andchromite are stable. Fireproof bricks are made ofmagnesite and chromite.5. EXPERIMENTAL DETERMINATION OFMECHANICAL ROCK PROPERTIESOf all the mechanical properties, the mostfrequently investigated are compressive strength,hardness, elasticity, toughness and wear resistance.The rocks' mechanical properties tests are definedby corresponding standards [9- SRPS B.B8.-012:1987 do SRPS B.B8.018:1957].5.1. Determination of rocks hardnessThe basic mineral that all the limestones aremade of is calcite, thus their hardness is usuallyabout 650 HB. Hardness of calcite-dolomitemineral depends on percentage shares of calcite anddolomite; it is usually within limits 650-850 HB,while the dolomite marble hardness is usually about850-1150 HB. Hardness of various types ofandesites is within the wide range and it depends ontheir type. The biotitic andesite has hardness similarto limestone, 550-700 HB, ensitite andesite 750HB, amphibolic andesite 1000-1500 HB, whilecontent of quartz SiO 2 (55-65%) in andesite canincrease hardness up to 1800 HB. Hardness ofrocks has strong influence on their processing andapplication, but also on damage of the workingparts of the construction machinery duringprocessing and building-in of the rock materials.This is why one must use metals that possess highhardness, with carbides in the metal substrate, orrelatively soft steels, which can, under pressure orimpact load, provide martensitic transformation ofaustenite – like the Hadfield steels [10].5.2. Determination of rocks impact toughnessIn Table 2 are presented values of impact hardnessof rock materials from the four studied sites.Testing was conducted in three mutually perpendiculardirections (I-I is parallel to rocks' stratificationdirection; II-II is perpendicular to I-I and lies instratification plane and III-III is perpendicular todirection of the rock's stratification). This isimportant to emphasize, since rocks are highlyanisotropic due to stratification and schistosityinhomogeneityof rocks.Based on these results one can conclude thatthose rocks have relatively low impact toughness;according to the fracture site appearance the similarconclusion can be drawn as well, since the fracturesurface is rough and with sharp edges.5.3. Determination of rocks elasticityThis property is related to solid bound rocks andit depends on type and hardness of rock minerals,structure and texture and minerals freshness,moisture, strength and direction applying of load,etc. Fine-grained rocks have higher values ofelasticity modulus, than the coarse rocks of thesame composition.For the stratified and inhomogeneous rocks theelasticity modulus is higher in the directionperpendicular to than parallel to stratification andschistosity. By testing the limestone samples thefollowing values were obtained: average Poisson'sratio = 0.36, elasticity modulus E = 50247 MPa,shear modulus G = 18608 MPa and bulk modulus K= 59714 MPa.278 13 th International Conference on Tribology – Serbiatrib’13

Table 2. Impact toughness of some types of rocks.Impact toughness, MPaLimestone - VučjakDolomite marble- SamarCalcite-dolomitemarble - GradacAndesite -ŠumnikDirection I-I 22.40 17.00 27.20 13.40Direction II-II 24.20 20.60 26.10 17.20Direction III-III 28.80 24.60 28.30 22.40Average value 25.13 20.73 27.20 17.675.4. Determination of rocks' strengthExperimental determination of rocks' strength isusually done with at least three samples, mostfrequently with 5 samples, cut out from the rock inthree mutually perpendicular directions; samplesare of prismatic form; the whole is defined byadequate standards [9- srps b.b8.012:1987 to srpsb.b8.018:195].Compression strength. Determination of thecompression strength was done on dry and watersaturated cubic samples, cube edge is 40 ± 1 mm,with ground and plan parallel surfaces. The averagevalue of the compression strength for the limestonefrom site Vučjak (15 samples, 5 for each direction)was R cm = 131 MPa. Tests of samples from othersites were done in the same way. For obtaining thecompression strength after 25 cycles of freezing 3cubic samples were used; cube edge was 1000 ± 1mm (Table 3).Tensile strength. Determination of the tensilestrength is conducted less frequently though it isalso an important property. The tests were done onsamples of the same shape and dimensions as forthe compression strength; but only dry samples weretested. The average value of the tensile strengthof the tested limestone samples was R m = 6.00MPa, while in large number of references reportedvalue is about 1.5 MPa for the porous limestonesand 6.4 MPa for the firm limestones. Tests ofsamples from other types of rocks were done in thesame way and results are shown in table 4.Bending strength. Bending tests were done onone dry sample per each type of rocks, for thedirection which is perpendicular to stratificationand schistosity propagation, in order to obtain thebest possible results, [6].Shear strength. Determination of shear strengthwas done only on limestone from the Vučjak site bythe direct shear method, by the Casagrande testingdevice, [6].5.5. Determination of rocks' wear resistanceExperimental investigation of the rocks' wearresistance is defined by corresponding methods:the Bome method, the Los Angeles and the Devalprocedure, [9]. By the Bome method the wearresistance is checked on the cubic samples, whilethe other two methods are more frequentlyapplied for the rock aggregates; results arepresented in [6].Table 3. Compression strength of some rock materials.Rocks' type and originDrysamplesCompression strength, R cm , MPaWater saturatedsamplesSamples after 25freezing cyclesSoftening coefficient,K softLimestone - Vučjak 131 123 117 0.94Dolomite marble - Samar 150 136 130 0.91Calcite dolomite marble - Gradac 161 140 138 0.87Andesite - Šumnik 195 186 184 0.95Table 4. Tensile strength of some rock materials.Rocks' type and originTensile strength, R m , MPaDirection I-I Direction II-II Direction III-IIIAverage valuesLimestone - Vučjak 5.97 6.14 5.89 6.00Dolomite marble - Samar 5.26 5.42 3.72 4.80Calcite dolomite marble - Gradac 5.64 5.36 4.03 5.00Andesite - Šumnik 9.48 9.82 8.90 9.4013 th International Conference on Tribology – Serbiatrib’13 279

6. SELECTION OF MATERIALS FOR THECONSTRUCTION MECHANIZATIONWORKING PARTSLarge number of parts of the constructionmechanization (hoop's teeth, rippers, mixers' blades,knives for lifting and removing asphalt, knivesbladesfor channels digging, crushers impact beamsetc.) are, during operation, in indirect or directcontact with rock materials, and, depending on therole and function of the part, they are exposed tovarious types of wear [10-12]. Mainly one dealswith abrasive wear or a combination of severaltypes of wear.The main factors in selection of materials exposedto wear are their chemical composition andstructure. Those materials are mainly steels and(white) cast irons. The most resistant to wear arematerials of high hardness. Though that characteristiccan not be the only criterion, it is mainly usedfor quality estimates of materials exposed to wear.Depending on the degree of wear, machines'working parts could be replaced by the new ones,or could be subjected to reparatory hard-facing.When selecting the reparation technology,mechanism of abrasive wear is being analyzed,both theoretically and experimentally, taking intoaccount hardness and microstructure of parts, andrelated wear resistance in laboratory and realoperating conditions. That represents the basis forselection of the optimal procedure, technology andfiller metal for working surfaces regeneration.Investigations, until now conducted by these authors,have shown that the working life of the properlyregenerated parts exceeds several times thenew parts working life. Besides that, the machinedown-time, assortment and quantity of the necessaryspare parts are reduced. All these point tocomplexity of materials selection for the workingparts of construction mechanization, as well as toimportance of knowing the properties of the rockmaterials.7. CONCLUSIONIn this paper are presented investigation resultsof the rock materials physical and mechanical propertiestests, for four types of rock materials mostfrequently applied in construction of driveways.The conclusion was reached that those materials arefirm and homogeneous rocks, of medium to highhardness. It was determined that the compressionstrength values are on average 10-40 times higherthan the shear, bending and tensile strengths, whilethe wear resistance is good in majority of studiedrocks, being especially high for andesite.Obtained results enable estimates of the individualrock material's quality and point to the complexityof the material selection for the workingparts pf construction mechanization. Experimentalresults of rock materials properties and the complexworking conditions of the tribo-system, can serveand must be taken into account in selection of thebase metal and reparation technology of thedamaged parts of the construction mechanization.AcknowledgementThis research was partially financially supportedthrough grants TR35024, TR35021, TR34002 andOI174004 of Ministry of Education and Science ofRepublic of Serbia.REFERENCES[1] M. Janjić: Fundamentals of geology and engineeringgeology, Part I, Faculty of Civil Engineering,Belgrade, 1964.[2] M. Janjić: Engineering geodynamics, Faculty ofMining and geology, Belgrade, 1979.[3] M. Janjić: Engineering geology with geologyfundamentals, Scientific Book, Belgrade, 1982.[4] M. Luković: Engineering geology, Scientific Book,Belgrade, 1950.[5] N. Panjukov: Engineering geology, Civil EngineeringBook, Belgrade, 1963.[6] M. Mutavdžić: Reparatory hard-facing of parts ofmachines and devices in constructionmechanization, MS thesis, Faculty of MechanicalEngineering, Kragujevac, 2007.[7] R. Stojadinović: Soil mechanics – I, Faculty ofCivil Engineering, Belgrade 1978.[8] M. Muravljov: Civil engineering materials, CivilEngineering Book, Belgrade, 2002.[9] Standards: JUS B.B8.001 to JUS B.B8.086.[10] V. Lazić et al.: The working life theoretical andexperimental estimates for machine parts hardfaced using austenite-manganese electrodes,Materials end Technology, Vol. 46, No 5, pp. 547-554, 2012.[11] M. Mutavdžić et al.: Model investigations of thefiller materials for regeneration of the damagedparts of the construction mechanization, Tribologyin Industry, Vol. 30, No. 3&4, pp. 3-9, 2008.[12] V. Lazić et al.: Selection of the most appropriatetechnology of reparatory hard facing of workingparts on universal construction machinery, Tribologyin Industry, Vol. 33, No. 1, pp. 18-27, 2011.280 13 th International Conference on Tribology – Serbiatrib’13

Serbian TribologySocietySERBIATRIB ‘1313 th International Conference onTribologyKragujevac, Serbia, 15 – 17 May 2013Faculty of Engineeringin KragujevacTECHNO-ECONOMIC JUSTIFICATION FOR REPARATORYHARD-FACING OF MACHINE SYSTEMS' WORKING PARTSV. Lazić 1 , R. Čukić 1 , S. Aleksandrović 1 , D. Milosavljević 1 , R. Nikolić 1, 2 , B. Krstić 1 , B. Nedeljković 31 University of Kragujevac, Faculty of Engineering, Sestre Janjić 6, 34000 Kragujevac, Serbia2 University of Žilina, Faculty of Civil Engineering, Univerzitna 1, 010 26 Žilina, Slovakia3 High technical school in Kragujevac, Kosovska 8, 34000 Kragujevac, SerbiaAbstract: Research in the field of hard-facing of various parts of mechanical systems is being done for technical andtechno-economic reasons. The reasons for introducing the new reparation technologies by hard-facing are numerous:three quarters of all the mechanical parts of engineering systems could be regenerated or manufactured by hardfacing;the working life of the repaired part reaches or even exceeds the working life of a new part, while the workinglife of the hard-faced manufactured part surpasses several times the working life of the new part manufactured by someother technology. Large number of damaged and, frequently even broken parts causes terminations of the workingprocess. Thus, due to difficulties in procurement of new, mainly imported parts, the alternative solution must be appliedand that is regeneration by hard-facing.It is shown that the a proper choice of the hard-facing technology is related to the complex procedure of checkingthe quality of the hard-faced layer, what indicates that the reparatory operations could be performed only in specializedregeneration workshops, which are furnished with adequate equipment and corresponding expert and skilled staff. Theestimated net benefit for the analysed parts is exceptionally high, regardless of the fact that the additional external andinternal effects have not been quantified. After the successful application of these new manufacturing hard-facingtechnologies it would be possible to create the knowledge base and apply it in maintaining the parts of civil engineeringmachinery, forging equipment and other similar mechanical parts.Key words: regeneration, wear, hard-facing, costs, techno-economic analysis.1. INTRODUCTIONThe reasons for the introduction of technologyfor manufacturing and reparatory hard-facing arenumerous: research indicates that three-quarters ofall the mechanical parts can be regenerated andmanufacturing hard-faced, service life of repairedpart reaches or exceeds the service life of the newpart, service life of new in production hard-facedpart exceeds several times that of the new part,which was not hard faced, repair costs are reducedas well as the downtime due to purchasing a newpart, which increases productivity, financing costsand cost of storage are also reduced [1- 4]. A largenumber of damaged, and often broken, parts causetermination of the process, and the difficulties inthe procurement of new, mostly imported parts,must use an alternative such as hard-facingregeneration.In addition, the maintenance of thetechnical system should take in considerationmanufacturing of new parts by hard-facing, what isexpected to extend their service life with respect tothe new working parts.To perform the modelling of hard-facing ofworking parts, i.e., to prescribe generalregeneration procedures, it is necessary to performprevious studies on a number of models and realworking parts made of various types of steel andcast iron. While the surfacing almost every time is aunique job, because it requires the technologycustomized to each working part, it is possible toestablish general procedure for groups of similarparts and then to apply it [2,5-6].2. SELECTION OF THE OPTIMAL HARD-FACING TECHNOLOGYIn examining the state of the damaged parts oneshould first determine: whether the wear occurred duringthe normal exploitation or it appeared due to somemechanical damage; what is the degree of the part's wearis crucial for the decision whether it is cost-effective and13 th International Conference on Tribology – Serbiatrib’13 281

safe to use it in furter exploitation (to apply regenerationor the part) or should it be rejected. The size of expecteddeformation and residual stresses are also importantfactors in making such a decision [7-8]. Afterdetermination of the chemical composition of the basemetal and working conditions, it is possible to create thebasic conditions for the design of technologicalprocesses. Based on those facts and previouslyconducted detailed techno-economic analysis, themethod of regeneration should be chosen, taking intoaccount the local possibilities of the company. The basicrequirement is to obtain the required properties of theregenerated part and, ofcourse, the reliability of the partduring the estimated working life.To achieve the above requirements it is necessary tomake the proper selection of filler material for hardfacing.In some cases of reparation of working parts itis necessary to apply two or more kinds of additionalmaterial to insert an intermediate layer, the so calledbuffer layer, between the layer and the substrate. Thisreduces the large differences in the chemicalcomposition, structure and, consequently, the thermophysicalproperties, of the substrate and the deposit.Next follows the selection of the regeneration processparameters, resulting from the properties of the baseand filler metal, and form demands concerning the sizeand shape of regenerated parts. The final stage ofplanning, before the experimental surfacing, is theassessment of the necessity for implementation ofspecial measures and the previous, current andsubsequent heat treatment.For verification of the proposed technologies, thecomparative tests in laboratory and in workingconditions have been performed, and, in some cases,comparative test of imported parts, which were nothard-faced and the new-hardfaced parts. Laboratorytests are related to the microstructure, hardnessdistribution and tribological tests and working tests ofcomparing the working life of the new and repairedparts installed in the same machine [2-3,9].From the point of view of techno-economicanalysis, reparation welding technology is a complexset of different types of mandatory procedures, whichtake into account: the conditions of work, damageidentification, estimation of weldability, weldingprocess, filler material, welding and hard-facingregimes, heat treatment applied, model and real tests.Having in mind the complexity of the process, it isnecessary to determine the optimal technicaltechnologicalsolutions to bring the reparation processto a stage when it is possible to make a final decision,wheter to buy a new part or to repaire it.3. EXAMPLES OF IMPLEMENTED REPAIRSHere is considered the justification for applicationof the production and reparation hard-facing and it ispointed to profitabilityof repairs on examples ofdamaged forging hammer, forging press frame andlarge gear of eccentric presses. The subject matter is thereparatory welding and surfacing of the damaged orcracked forging hammers, broken and cracked frames,forging presses and large gear eccentric presses [10-11]. To determine the optimal technology of hardfacing,it was necessary to carry out tests on model andreal working parts. Test hard-facing and testing ofmodels have served to establish the initial reparationtechnology, and to "transfer" thus approvedtechnologies to the working parts, which are thenfurther checked under actual working conditions.This paper mainly deals with the techno-economicadvantages of the hard-facing technology, while thecomplete procedures of determining the optimaltechnology for each particular part were presented inpapers [2-3,5,9-11].3.1. Regeneration equipment’s for forginghammer and press frameFor regeneration of responsible parts with complexgeometries and large masses, made of material sutiblefor tempering, a detailed analysis of the working partsis required as well as the precisely proposed reparationtechnology.Hammers mallets and presses frames are exposed,during the long operation, to thermal fatigue due tocyclic temperature changes and to impact loads. Due tothe high costs and often to impossibility of purchasingthe new working parts, it is necessary to evaluate thepossibility of their repairs. Harsh working conditionssometimes lead to a complete fracture of the part andendangering the workplace safety. Figure 1 showsfracture of a forging mallet, which has originated fromfatigue crack propagation.Mallet of forging's hammer, shown in Fig. 2, andframe forging press, shown in Fig. 3, are primarilysubjected to impact compression loads, and, in partiallyto temperature gradient, that is thermal stresses causedby uneven temperature field [2,7-8]. After a long workof these parts, i.e., large number of repeated cycles onhammer mallet and on the press frame, visible crackswere observed, and on one portion of the frame and ona single occurred the complete fracture (Fig. 1 and 3).Taking into account that these are parts of largedimensions and complex shapes, and that componentsare subjected to dynamic and thermal loads, they aredimensioned on the basis of the increased safetydegrees; thus the special measures are required for themanufacturing and reparatory technologies, as well.Forging press frame is madeby casting in sand, from the medium carbon cast steel.On the other hand, pneumatic forging hammer mallet,as one of the most loaded mechanical parts, is made oflow alloyed steel for tempering.282 13 th International Conference on Tribology – Serbiatrib’13

Figure 1. Appearance of fractured forging hammer mallet.900920250200748.2150170 802102600480a) b)Figure 2. Forging hammer malleta) Sketch of mallet with observed crack; b) Apperance of the regenerated mallet of mass of 6000 kgSection S-S75S10075160S1211902400a) b)Figure 3. Frame of vertical forging press.a) sketch of the frame (1 – fracture site; 2 - observed cracks); b) regenerated, heat treated and machined partThe complete technology for regeneration ofdamaged forging hammers' mallets and presses'frames is shown in [6, 10-11]; here are presented onlythe techno-economic indicators for the mallets'regeneration.The following data are relevant for comparison:A. The price of the new part is: 83987 Є(This price includes price of a new part - 67470 Є,tax -12144 Є, the customs 3373 Є and the cost ofshipping and transportation services - 1000 Є);B. The total real costs of reparation of 4912 Єinclude: Identification and damage detection:3 days × 8 (nh*/day) × 10 (Є/nh) = 240 Є;13 th International Conference on Tribology – Serbiatrib’13 283

Machining of damaged area:10 days × 8 (nh/day) × 12 (Є/nh) = 960 Є; Selection of the optimal hard-facing technology:8 days × 8 (nh/day) × 15 (Є/nh) = 768 Є; Model testing:4 days × 8 (nh/day) × 12 (Є/nh) = 384 Є; Surfacing of real working parts:20 days × 8 (nh/day) × 10 (Є/nh) = 1600 Є; The costs of machining operations of surfaced areas:10 days × 8 (nh/day) × 12 (Є/nh) = 960 Є.nh = norm-hourBased on these data one can conclude that thetotal reparation costs are far lower than the costs ofa new part (less than 6%). Therefore, the "buy" or"repaire" dilemma is apparently resolved withoutmore detailed analysis of the positive effects thatmallet regeneration allows.3.2. Reparation of large gears - toothed hub ofan eccentric pressThe techno-economic analysis of reparatorywelding and hard-facing of the damaged teeth of acoupling hub with mass of 500 kg, shown in Fig. 4,is performed after the repair has already beenperformed, because it is a unique part that could notbe easily obtained. The coupling is exposed toharsh environmental conditions and is made ofalloyed steel for tempering. Since it is aconditionally weldable steel, it was necessary toprescribe a particular reparation technology. It wasestablished through previous model tests [5,10].The analysis of obtained results leads to the optimalhard-facing technology which is then "transferred"to the real part.Figure 4. Appearance a coupling hubIn economic "buy" or "repaire" analysis, anestimate of the more complete effects was notconducted, which should be performed by thebenefit-cost (BC) analysis, or more precisely byusing the life-cycle-cost (LCC) analysis, whichwould point to more precise and more clearadvantages of application of this advancedtechnology [3]. A comparative analysis wasperformed after two and a half years of the hardfacedrack hub couplings operation.As relevant for comparison the following datawere taken:A. Purchasing price of the new part: 26500 Є(This price includes price of a new part, the cost oftaxes, customs duties, freight forwarding servicesand transport).B. The total real reparation costs of 3380 Єinclude: Identification and damage detection: 1day × 8 (nh/day) × 10 (Є/nh) = 80 Є; Machining of damaged area:2 days × 8 (nh/day) × 12 (Є/nh) = 192 Є; Selection of the optimal hard-facing technology:1 day × 8 (nh/day) × 15 (Є/nh) = 120 Є; Model testing:3 days × 8 (nh/day) × 12 (Є/nh) = 288 Є; Surfacing of real working parts:10 days × 8 (nh/day) × 10 (Є/nh) = 800 Є; Costs of production services (processing hardfacedteeth and transport)1900 Є.Based on these data one can conclude that thetotal cost of repairs is significantly lower than thecost of purchasing a new part (less than 13%).4. CONCLUSIONThrough the proper selection and application ofthe reparatory and manufacturing hard-facingtechnologies, it is possible to achieve a number ofadvantages compared to the installation of newparts. This is primarily related to the extension ofthe service life of the analyzed parts, increase ofproductivity, reduction of downtimes, reduction ofinventory costs and other benefits derived byapplying the welding technology.It is shown that a proper choice of hard-facingtechnologies is associated with the complexprocedure of checking the quality of deposits, whatindicates that the repair work can be performedonly in specialized workshops for regeneration,which have adequate equipment and appropriateskilled staff. The expected net benefit for theanalyzed parts is very high, regardless of the factthat additional external and internal effects have notbeen quantified. After the successful284 13 th International Conference on Tribology – Serbiatrib’13

implementation of these new manugacturingsurfacing technologies in presented areas, it ispossible, by applying the similar procedure, to forma knowledge base and to use it for maintenance ofequipment for forging, and other similarmechanical parts.ACKNOWLEDGEMENTThis research was partially financially supported byMinistry of Education and Science of Republic of Serbiathrough grants TR35024, TR35021, TR34002 andOI174004.REFERENCES[1] R. Wasserman: How to save millions by reducinginventories of spare parts, Eutectic-CastolinInstitute for the Advancement of Maintenance andRepair Welding Techniques, 1971.[2] V. Lazić: Optimization of the hard-facingprocedures from the aspect of tribologicalcharacteristics of the hard-faced layers and residualstresses, PhD thesis, The Faculty of MechanicalEngineering, Kragujevac, 2001. (In Serbian)[3] R. Čukić: Techno-economic analysis of productionand reparatory hard-facing of various parts ofmachine systems, PhD thesis, University ofBelgrade, Serbia, 2010. (In Serbian).[4] V. Lazić et al.: Selection of the most appropriatetechnology of reparatory hard-facing of working partson universal construction machinery, Tribology inIndustry, Vol. 33, No 1, pp. 18-27, 2011.[5] M. Mutavdžić et al.: Model investigations of thefiller materials for regeneration of the damagedparts of the construction mechanization, Tribologyin Industry, Vol. 30, No. 3&4, pp. 3-9, 2008.[6] V. Lazić et al.: Reparation of damaged mallet forhammer forging by hard-facing and weld cladding,Technical Gazette, Vol. 16, No. 4, pp. 107-113, 2009.[7] A. Wernoski: Zmeczenie cieplne metali,Wydawnictwa Naukowo - Techniczne, Warszawa,1983. (In Polish).[8] R. Pasierb: Spawanie žarowytrzymalych stalichromowo-molibdenowo-wanadowych, WNT,Warszawa, 1982. (In Polish).[9] V. Lazić et al.: Selection of the optimum technologyof the forging dies reparation from the aspect oftribological characteristics, TRIBOLOGIA -TEORIA I PRAKTIKA, ROK XXXVI, NR 2/2005,pp. 11-30, 2005.[10] V. Lazić et al.: Development and application ofwelding technologies for regeneration of damagedworking parts of forging presses and hammers,Welding and Joining Technologies for aSustainable Development and Environment, 24-26.05. 2006., Timisoara, Romania, pp. 341-346.13 th International Conference on Tribology – Serbiatrib’13 285

Serbian TribologySocietySERBIATRIB ‘1313 th International Conference onTribologyKragujevac, Serbia, 15 – 17 May 2013Faculty of Engineeringin KragujevacTRIBOLOGY ASPECT OFRUBBER SHOCK ABSORBERS DEVELOPMENTMilan Banić 1 , Dušan Stamenković 2 , Miloš Milošević 3 , Aleksandar Miltenović 41 Faculty of Mechanical Engineering, University of Niš, Serbia, banic@masfak.ni.ac.rs2 Faculty of Mechanical Engineering, University of Niš, Serbia, dusans@masfak.ni.ac.rs3 Faculty of Mechanical Engineering, University of Niš, Serbia, mmilos@masfak.ni.ac.rs4 Faculty of Mechanical Engineering, University of Niš, Serbia, amiltenovic@yahoo.comAbstract: Rubber is a very flexible material with many desirable properties which enable its broad use inengineering practice. Rubber or rubber-metal springs are widely used as anti-vibration or anti-shockcomponents in technical systems. Rubber-metal springs are usually realized as a bonded assembly, howeverespecially in shock absorbers, it is possible to realize free contacts between rubber and metal parts. Inprevious authors research it was observed that friction between rubber and metal in such case have asignificant influence on the damping characteristics of shock absorber. This paper analyzes the developmentprocess of rubber or rubber-metal shock absorbers realized with free contacts between the constitutive parts,starting from the design, construction, testing and operation, with special emphasis on the development ofrubber-metal springs for the buffing and draw gear of railway vehicles.Keywords: rubber friction, rubber-metal spring, shock absorber, product development.1. INTRODUCTIONRubber or rubber-metal springs are widely usedin industry as anti-vibration or anti-shockcomponents giving many years of service. Theyhave several advantages in respect to metal springs(lower price, easier installation, lower mass,reduced corrosion, no risk of fracture and no needfor lubrication) [1]. However, they have one majordisadvantage reflected in insufficiently reliableservice life caused by rubber fatigue.Those elements are well established to controlvertical and lateral movements. Nowadays, themore demanding operating environment has madethe design of such components more challengingthan ever before. In addition to the design of therubber part itself the interface between the part andthe structure is also important.The properties of the rubber-metal spring aremainly influenced by a rubber compound. Rubbercompounds are generally composed of a baserubber (e.g. natural rubber), filler (e.g. carbonblack) and a curing agent (e.g. sulphur). Additionalcomponents may include antioxidants, adhesionagents, flame retardant agents and special processenhancingchemical additives. Common physicalproperties of rubber compounds are affected byevery ingredient of a rubber recipe independently ofor dependently on each other. The mixing andcuring process is also critical in determining theseproperties. Improving one compound propertyalways results in changing other properties, forbetter or for worse. Noted fact makes developmentof elastomeric based products a very complicatedtask. Up to appearance of modern computer aidedtools, the development of those products relied onlyon previous experience of the designer and trial anderror procedure. Such approach was inefficient,expensive and time consuming because it requirediterative procedure combined with excessiveexperimental testing to achieve desired mechanicalproperties.Rubber-metal springs are usually realized as abonded assembly, however especially in shockabsorbers, it is possible to realize free contactsbetween rubber and metal parts. In that case,connections between rubber blocks and metal platesare realized by applying pressure and resultingstatic friction. During the load cycle of the shockabsorber, apart the energy dissipation in rubber,286 13 th International Conference on Tribology – Serbiatrib’13

additional energy is dissipated due to frictionbetween rubber and metal parts.In previous authors research [2] it was observedthat friction between rubber and metal in such casehave a significant influence on the dampingcharacteristics of shock absorber. This paperanalyzes the development process of rubber orrubber-metal shock absorbers realized with freecontacts between the constitutive parts, startingfrom the design, construction, testing and operation,with special emphasis on the development ofrubber-metal springs for the buffing and draw gearof railway vehicles.2. DEVELOPMENT OF THE RUBBER-METAL SHOCK ABSORBERWith appearance of modern computer tools andvirtual product development, the developmentprocess of shock absorbers became more efficientdue to simulated experimental testing of virtualprototype. With virtual product development toolsit is possible to predict the absorbing capacity andservice life before the manufacturing of the productprototype which was not possible in classicaldevelopment process.The assembly of shock absorber with rubbermetalspring usually consists of a few rubber-metalelements separated with metal plates andprestressed with a central screw. A rubber-metalelement represents a metal carrier in the shape of acircular plate with natural or synthetic rubbervulcanized on both sides. Therefore, the advantagesof both component elements are involved: highabilities of displacement and amortization of rubberand large loads which are sustained by metal parts.These ensure the decrease of noise and amortizationof impact loads. Figure 1 shows the design ofbuffing gear spring assembly, while Figure 2 showsthe design of draw gear spring assembly.Figure 1. Rubber-metal spring assembly of buffing gearFigure 2. Rubber-metal spring assembly of draw gearAs already noted, these springs are used as antishockcomponents, so the main properties designermust take into account are absorbing capacity andstiffness of the spring. The most importantabsorbing characteristic of rubber is evaluated byits hysteresis. Hysteresis is the mechanical energyloss that always occurs in an elastic materialbetween the application and the removal of a load.If the displacement of a system with hysteresis isplotted on a graph against the applied force, theresulting curve is in the form of a loop. It dependsnot only on the elastomer type, but also on fillersand other compound ingredients as othermechanical properties.The authors defined a virtual productdevelopment procedure (Figure 3) for developmentof rubber-metal springs used in shock absorbers.The development procedure is based on applicationof modern viscoplastic rubber constitutive model(Bergström-Boyce), which besides higher accuracyof prediction, enables the assessment rubbercompound hysteresis and strain rate dependencewhich is not possible by application of hyperelasticmodels usual for rubber FE analysis. Theparameters of rubber constitutive model(Bergström-Boyce) are determined by uniaxialcompression at different strain rates and stressrelaxation test on the samples of the rubbercompound (35.7 x 17.8 mm) [3]. The samples arecompressed between hardened steel plateslubricated with machine oil in order to prevent thebarrelling of samples. Based on the performedexperiments, the database of model parameters forthe rubber compounds can be defined. Databasealso contains data about other significant propertiesof rubber compound, such as composition, commonmechanical properties, etc.The first step in procedure shown at Figure 3 isto determine the rubber compound for rubber-metalspring. From the formed database severalcompounds are selected based on criteria definedby widely known selection and service guide forelastomers [4] regarding the product specificrequirements (creep, low-temp stiffening, heataging,…) and the operating environment conditions(resistance to ozone, radiation, …). The selected13 th International Conference on Tribology – Serbiatrib’13 287

compounds are used for simulation of static anddynamic hysteresis of standardised specimens.The simulation of static hysteresis test isconducted on cylindrical test samples (35.7 x17.8 mm), according to the internal standardSIMF.92.006 [5]. The simulation dynamichysteresis (Yerzley hysteresis) test is carried out oncylindrical test samples (19.5 x 12.5 mm)according to the standard ASTM D 945-06 [6].The obtained results of static and dynamichysteresis are used for final selection of rubbercompound. As the main feature and an indicator ofthe quality of rubber-metal springs is energyabsorption capacity, the compound is selectedbased on criterion of highest static and dynamichysteresis obtained during simulation.The next step following the adoption of therubber compound is the selection of the appropriatestructural design of the basic rubber metal elementand their combination into a spring package. Asalready noted, the main problem in developingrubber-metal springs is that a designer cannotestimate how many basic rubber metal springelements need be combined in a serial set toachieve the required absorption capacity. Theamortizing ability of a rubber-metal spring package,and therefore the constitutive number of basicrubber metal elements, should be determined bysimulation using the finite element method. Basedon the required operating stroke, built-in measuresand assumptions about preloading of rubber-metalspring the initial design of the basic rubber metalelements is adopted. For instance, the initialgeometry of the basic element of buffer rubbermetalspring is shown on Figure 4. It consisted of ametal disk with openings and vulcanized rubberparts on both sides of the plate connected throughthe plate’s openings (Figure 4).Figure 4. The basic rubber metal spring element ofbuffing gearUpon the simulation of static hysteresis of basicelement, the number of elements in rubber-metalassembly can be easily determined as the ratio ofrequired spring absorption capacity and theabsorption capacity of the single element. Theprocedure is sometimes iterative to obtain desiredresults with required value of normal reaction forceduring impact and limited number of basic elementsdue to installation requirements.Figure 3. Procedure for development of shock absorberswith rubber-metal springFigure 5. S-N curves of the rubber-metal element [7]288 13 th International Conference on Tribology – Serbiatrib’13

The adopted geometry parameters are furtherimproved by optimisation. The optimizationprocedure is performed in order to improve thedesign of the spring element regarding its servicelife. By lowering the element stress levels values,the service life is prolonged which is obvious fromFigure 5. As an example, the optimisation basicrubber metal spring element of buffing gear isperformed by defining the design of experiment asa central composite design in simulation. Byvariation of the plate opening diameter (D o ), thenumber of openings (B o ) and radius on top andbottom of the rubber part of the element (R), thefunctional dependence of maximum equivalentstress from input parameters was obtained.Minimisation of the obtained functionaldependence results in optimal geometricdimensions of the basic spring element. Thefunctional dependence of maximal equivalent stressfrom input parameters is shown in Figure 6.Upon the definition of final geometry of theelement, the actual production and testing ofprototype are performed in order to validate designand to determine the accordance with the designrequirements.3. TRIBOLOGY ASPECT OF THE SHOCKABSORBER DEVELOPMENTRubber has a very high coefficient of frictionwhich can even reach value of μ = 4. High frictioncoefficient, and thus high grip of rubber, found itsway in many engineering applications; for example,rubber is not always bonded in bushes, since itsfrictional grip is almost equal to a bond. During theload of unbounded rubber metal assembly incompression the friction force occurs (Figure 7).Due to high grip between the rubber and the metalthe rubber is barrelling thus increasing the contactsurface.Figure 6. Functional dependence of maximal equivalentstress from the optimisation parametersFigure 7. Compression of rubber between steel plates: a)unloaded; b) loadedIt has been noticed that the amount of theaccumulated/absorbed energies of rubber-metalsprings loaded in compression greatly depends onthe contact between the rubber and the metal [2].Noted findings were also confirmed by otherauthors. For instance, Figure 8 shows the effect oflubricating the contact between rubber and metal incompression. Provided that the steel ends are cleanthe grip is almost equal to that of a bonded sample.Although bonded contact provides a highernormal reaction force during impact, the freecontact such as in draw and buffer gear rubbermetal spring assembly dissipates more energy asthere are friction induced energy losses due tocontact sliding. If the friction coefficient issufficiently high to ensure that significant sliding13 th International Conference on Tribology – Serbiatrib’13 289

etween metal and rubber will not occur shockabsorber will have better absorbing properties.Figure 8. Effects of surface conditions on thestress/deflection curve for rubber under compression [8]Significant sliding compromises the assemblystability and has a great effect on lowering ofnormal resulting force. Furthermore, the increase ofthe friction coefficient on the contact surfaces ofthe rubber element increases the shear stress and itsshare of the total stress also. The increasing shearstress further increases the total stress in theelement and the force which resists the deformationof the element. The increase of the shear stressshare of the total stress leads to the enhancedamortization capacity of rubber elements. As thehigh values of normal resulting force are the designrequirement, it is necessary to find the balancebetween the sliding allowance and resulting normalforce. It can be achieved by influencing thetribological contact parameters (lubrication, surfaceroughness of the metal part, contact pressure, …)and thus the friction coefficient value.Based on above it can be concluded that it is notpossible to actually perform the virtualdevelopment process of the shock absorber withrubber metal spring without the knowledge of thefriction coefficient value in contact between therubber and metal parts.The coefficient of friction of rubber is highlydependent on contact pressure. As the contactpressure between the rubber and the free metalplates in shock absorbers is approximately 20 MPa,it is extremely difficult or even impossible toexperimentally determine the actual value offriction coefficient at noted operating contactpressure.The compound friction coefficient can bepredicted based on experiments with rubberspecimens or based on existing data on normalreaction force in similar operating conditions. Bysimulation of experiments on rubber specimens orpreviously performed experiments, the frictioncoefficient can be determined by goal drivenoptimisation procedure. The value of frictioncoefficient will be approximately determined whenthe normal reaction force obtained by simulation isequal to experimentally obtained one.As an example, it is necessary to determine thefriction coefficient in contact between rubber withtrade name TG-B-712 (manufactured by companyTIGAR, Pirot) and metal plate at contact pressureof 3 MPa. The rubber specimen (with dimensions(35.7 x 17.8 mm) was compressed between steelplates at specific tribological conditions for which itwas necessary to determine the value of frictioncoefficient. The force-displacement data wasrecorded during the experiment (Figure 9).Figure 9. Experimental force – displacement dataFigure 10. Functional dependence between frictioncoefficient and force determined by virtual experimentThe friction coefficient was determined byvirtual experiment from which the functionaldependence between friction coefficient and normalresulting force was obtained (Figure 10). Based onrealistic experimental data (Figure 9) it is clear thatthe maximal resulting normal force correspond tofriction coefficient value of μ = 1.5.290 13 th International Conference on Tribology – Serbiatrib’13

4. CONCLUSIONTools of virtual product development enablesignificant cost and time savings in the process ofdevelopment of shock absorbers filled with rubbermetalsprings.But to employ the tools of virtual productdevelopment it is essential to have a value offriction coefficient in free contact between rubberand metal parts. Without the correct value offriction coefficient the proposed procedure outlinedin the paper would provide incorrect data which isnot suitable for shock absorber developmentprocess.As experimental determination of frictioncoefficient in contact between rubber and metal athigh values of contact pressure can be veryproblematic, the friction coefficient can beestimated by goal driven optimisation duringnumerical simulation of realistic experiments withexisting resulting force data.REFERENCES[1] V. Miltenović: Mašinski elementi-oblici, proračun,primena, Mašinski fakultet Univerziteta u Nišu,2009.[2] D. Stamenković, M. Milošević: Friction at Rubber-Metal Springs, 11 th International Conference onTribology - SERBIATRIB ’09, 13-15.05.2009,Belgrade, pp. 215-219.[3] M. Banić, et al.: Prediction of Heat Generation inRubber or Rubber-Metal Springs, Thermal Science,16 (Suppl. 2), pp. 593-606, 2012.[4] …:Vibration & motion control catalogue, LORDCorporation, 2008.[5] SIMF.92.006.[6] ASTM D 945 Standard Test Methods for RubberProperties in Compression or Shear.[7] R. K. Luo, W. X. Wu, P. W. Cook, W.J. Mortel:Fatigue design of rubber springs used in rail vehiclesuspensions, Proceedings of the Institution ofMechanical Engineers, Part F: Journal of Rail andRapid Transit, 217/3, pp. 237-240, 2003.[8] A. B. Davey, A. R. Payne: Rubber in engineeringpractice, Great Britain, 1964.13 th International Conference on Tribology – Serbiatrib’13 291

Serbian TribologySocietySERBIATRIB ‘1313 th International Conference onTribologyKragujevac, Serbia, 15 – 17 May 2013Faculty of Engineeringin KragujevacEFFECTS OF USING OF MQL TECHNIQUE IN METALCUTTINGGordana Globočki Lakić 1 , Branislav Sredanović 1 , Davorin Kramar 2 , Bogdan Nedić 3 , Janez Kopač 21 University of Banjaluka, Faculty of Mechanical Engineering Banja Luka, Bosnia and Herzegovina2 University of Ljubljana, Faculty of Mechanical Engineering Ljubljana, Slovenia3 University of Kragujevac, Faculty of Engineering, Kragujevac, SerbiaAbstract: In this paper an effect of using of minimal quantity lubrication (MQL) technique in turningoperations is presented. Experimental research was performed on carbon steel C45E. Technologicalparameters: depth of cut, feed rate and cutting speed were adjusted to semi-machining and roughing. Highervalues of feed and cutting speed were used, than recommended from literature, and different types of coolingand lubrication conditions in turning were applied. As a conventional procedure and technology, lubricationwith flooding was applied. As special lubrication technique the MQL was used. During research, monitoringof the cutting force, chip shape, tool wear and surface roughness was performed. Relations betweenparameters, material machinability and economy of process were analyzed.Keywords: lubrication, cutting, MQL, effects1. INTRODUCTIONThe future of metal machining by 2020 is in thedevelopment of flexible machining systems,economical and productive processes, energyefficientprocesses, production without waste,ecological production with a reduced quantity ofthe cooling and lubrication fluids and etc. Thisconclusion is based on the study National ResearchCouncil of the USA and other teams of researchers[1, 2, 3, 4].Figure 1. Possibilities to increase the productivity ofmanufacturingIncreasing of productivity (Fig 1.) is impossiblewithout utilization of modern tools and machines,modern types of cooling and lubrication fluids(CLF), CLF dosing techniques, and modernequipment. Progress is not possible withoutknowledge of the materials machinability andexpert systems for the selection of suitablemachining regimes based on machinability andprocess modelling [3]. Phases of research,obtaining and application of knowledge in the fieldof cutting process are:• study of materials machinability based onexperimental research,• modelling of the cutting process and• integration of knowledge in expert systemsand specialized databases.The main progress in developing of highproductive machining processes is realized in thearea of special CLF dosing techniques. One of thesetechniques is minimum quantity lubrication (MQL).The analysis of previous researches have shownthat this CLF dosing technique was applied forlower cutting speed (v c = 100 -150 m/min) [5, 6, 7].From the structure of the cost of machined part,it can be concluded that the cost of CLF participate15%, costs of tools 10% and costs of energyconsumption 4% of total costs (Fig 2).292 13 th International Conference on Tribology – Serbiatrib’13

Figure 2. Structure of the machining costsThe focus of researches presented in this paperwas on effects of different CFL dosing techniquesin the field of higher cutting speed (v c = 200 - 400m/min), which contributed to the expansion oftechnological fields. In order to analyse themachinability when applying standard and MQLCFL dosing techniques, analysis of machiningenergy balance, effect of chip formation, tool lifeand quality of machined surface were also included[8].Figure 3. Influence technique of lubrication onproductivity and efficiencyThe studies that are presented in this paper arerelated to the analysis of use of modern techniquesCFL dosing, with the aim of defining the directionsof increasing productivity and efficiency ofmachining process (see Fig. 3).2. EXPERIMENTAL SETUPThe material that was used for the experimentalresearches is the carbon steel C45E. This steelbelongs to the group of construction steels whichare used for essential parts in the machines andconstructions. Workpiece is cold-rolled steel barwith a diameter of 120 mm and length of 300 mm.Research was performed on universal latheBOEHRINGER with the following properties:8 kW of power, maximum spindle speed of2240 rev/min, and feed of 1.6 mm/rev. Carbidecutting tool for semi machining SNMG 1204 08NMX was used. Tool clearance angle was 10°, rakeangle 0°, and a tool tip radius was 0.8 mm withoutchip breakers. Tool holder was PSDN 2525 M12with inclination angle 45° (Fig. 4).Figure 4. Experimental setup on machineIn this research two different CLF dosingtechniques were analysed:• conventional flooding and• special dosing technique - MQL.In conventional flooding, CLF is dosed at thetop of machining zone, from a distance ofapproximately 150 mm. CLF was directed on nonmachinedworkpiece surface and rake surface ofinsert.In the MQL technique, CLF was dosed using aspecial device which utilizes a compressed air toform an oil mist in the mixing chamber (Fig 5).During machining with MQL technique, the toolwas protected from sudden changes of heat loads.The effects of rapid expansion and contraction ofthe tool material and the cracks appearance andcoatings cracking were avoided. During themachining, due to the effect of the spray, the toolwas enveloped with a thin layer of emulsion. InMQL technique spray nozzle was installed at adistance of 30 mm (L MQL = 30 mm), normal to thecutting edge, with an angle of 30° (ψ MQL = 30°)regarding to the rake face of tool (Fig 6.). Withsuch recommended position of the nozzle qualitylubrication of machining zone were ensured. Table1 shows the values of the hydraulic conditions forboth CLF dosing techniques.Figure 5. MQL device13 th International Conference on Tribology – Serbiatrib’13 293

Table 1. The values of pressures and flowTechnique of CLFdosingPressure p[MPa]Flow Q[l/min]Standard flooding 0.3 2MQL 0.3 0.0005penetration cutting force (F p ). For processing of themeasured signals LabVIEW software package withspecially designed program framework was used.The developed program framework enables datatransmission to software package MATLAB.Monitoring and measurement of tool wear wasperformed using a tool microscope TMMITOTOYO 510 equipped with high-resolutioncamera (Fig 7). Surface roughness was measuredusing a mobile measuring device MITOTOYOSURFTEST SJ 301. During the experiments a chipformation process was monitored as well.Figure 7. Measuring devices: force data acquisition(left) and tool microscope (right)In the first phase of research cutting forces (F c ,F f and Fp) for different combinations of inputparameters were measured. At the same time theformed chips were collected, and the shape wasevaluated with the purpose of technological framesdefinition (Fig 8).Figure 6. Position of conventional flooding (left) andMQL nozzle (right) during the experimentsTechnological parameters varied in theexperiments were as follows: depth of cut (a), feed(f) and cutting speed (v c ). Technological parameterswere adjusted to the semi- turning, with the use ofhigher values (Table 2). The total numbers ofexperiments for both dosing techniques were 72.Table 2. The levels of technological parametersParameterCutting deptha [mm]The levels of variation1 2 3 41.5 2.0 2.5 -Feed s [mm/rev] 0.224 0.280 0.355 0.400Cuuting speed v c[m/min]210 310 400 -KISTLER dynamometer was used to measurethree cutting force components in turning: maincutting forces (F c ), feed cutting forces (F f ) andFigure 8. Data flow in measuring and monitoringIn the second phase of the experimental teststool wear was measured, as follows: the values ofconcentrated wear (VB), tool wear on tool clearanceface (VB') and the size of the crater on the rake face(b w ). Due to wear of tools values of surfaceroughness were measured, as follows: mean valuesof roughness (R a ) and maximum height ofroughness (R y ).294 13 th International Conference on Tribology – Serbiatrib’13

3. ANALYSIS OF RESULTSThe results of experimental investigations whichconcerning the values of machinability parameters:cutting resistances, shape of chip, tool wear andsurface roughness is shown. Modelling wasperformed using regression analysis and artificialneural networks (ANN).From the analysis of the cutting forcescomponents (Fig. 9, 10 and 11) in case ofconventional CLF dosing application, it can beconcluded that the value of the cutting forcescomponents increase with increasing feed anddepth of cut. When MQL technology is applied, thevalues of cutting forces increase with feed anddepth of cut as well. The values of forcecomponents F f and F p were smaller than the valuesof F c component for both dosing techniques, whichis consistent with the theoretical assumptions.Penetration cutting forces F p and feed cuttingforces in MQL technique are greater thanconventional technique of CLF dosing.Figure 11. Values of penetration cutting forceOne of the main indirect indicators of themachining process condition is chips shape. Basedon chip shapes the following features can bedetermined: tool wear, surface roughness, theamount of generated heat and related phenomena.Conclusions based on chip shape were adopted onthe basis of recommendations from the literature [1,2, 3].Table 3. Chip shape during machining with conventionalfloodingDepth a [mm]Cuuting speed v c[m/min]Feed f [mm/rev]0.224 0.280 0.4001.5Figure 9. Comparison of the mean value of componentsof cutting force for both dosing techniques2.02102.5Figure 10. Values of feed cutting forcesIn table 3 forms of chips obtained whenmachining steel C45E while using conventionaltechnique of CLF dosing are shown. Based onanalysis of chip shapes it can be concluded thatmachining with lower feed rates (f = 0.224 mm/rev)and greater depths (a = 2.5 mm), creates anunfavourable chip shape. However, for the sameparameters, in conditions with higher speeds, itforms more favourable chips shape.13 th International Conference on Tribology – Serbiatrib’13 295

Table 4. Chip shape during processing with MQLtechniqueDepth a [mm]1.52.02.5Cuuting speedv c [m/min]210Feed f [mm/rev]0.224 0.280 0.400Some chip shapes during machining of the samesteel C45E, with the same tool, using MQLtechnique are shown in Table 4. Chips are dark,which indicates that in the machining area a largeramount of heat is generated and dissipated throughthe chips. It can be concluded that MQL techniqueprovides good effects of lubrication, but the badeffects of cooling the machining zone. Based on achip shape, it can be concluded that the use of MQLtechnique provide favourable shapes of chip for allanalyzed machining conditions.The technological areas for both techniques oflubrication and different depths of cut are shown infigures 12 and 13. They are based on the assessedchip suitability. From the analysis of thetechnological areas, it can be concluded that MQLtechnique offers a wider field of machining.Previous studies have shown that differenttechniques of lubrication have a great impact on thewear of tools as well. In our study the measuredparameters of tool wear were as follows: concentricwear (VB) and wear on the secondary surface oftool (VB'), and crater wear - see table 5 and 6. Thisparameters has direct impact on the machiningprocess, and thus on the machinability. Criterion oftool wear was VB = 0.3 mm. Tool life in the case ofMQL technique was for about 33% longer (see Fig.14). The parameters of the surface roughness -mean height of roughness (Ra) and maximumroughness (Ry) were measured depending on themachining time. As regimes for machining themean values of the obtained technological areashave been adopted: a = 2.0 mm and f =0.280 mm/rev, cutting speed v c = 320 m/min(Figure 14 and 15).Table 5. Tool wear during machining with conventionalCLF dosingTime T [min]Machining lenghtL c [m]0.93 2723.75 1286Secondary rakeTool wear on tool facePrimaryrakeClerance7.23 2080Figure 12. Technological areas for depth a = 1.5 mmFigure 13. Technological areas for depth a = 2.5 mmFigure 14. Comparative diagram of tool wear fordifferent techniques of CLF dosing296 13 th International Conference on Tribology – Serbiatrib’13

Figure 15. Surface roughness regarding the tool wear inmachining with conventional CLF dosingFigure 15 shows that tool wear take affects onthe surface roughness parameters R a and R y .Roughness is influential parameter in assessing themachinability, and in the flooding conditions ofmachining, and this parameter was increased fromthe initial 14 μm to 18 μm at the time when toolwear achieved criterion 0.3 mm. Tool life for thegiven machining conditions was T = 7.23 min.Table 6. Tool wear during machining using MQLtechniqueFigure 16. Changes of parameters Ra and Ry, dependingon the tool wear in machining with MQL techniqueThe increase of R a the parameter value duringmachining with MQL technique is shown in Fig.16. It can be seen, that due to a higher percentage ofheat generated during processing the value ofparameter R y is higher when machining with MQLtechnique than in conventional lubrication.Time T [min]Machininglenght L c [m]SecondaryrakeTool wear on tool facePrimary rakeClerance1.10 3395.92 1952Figure 17. Values of tool life for different lubricationtechniques9.65 3163In Table 6 the values of the tool wear parametersduring machining with MQL technique oflubrication is shown. The regime a = 2.0 mm, f =0280 mm/rev, v c = 320 m/min was applied. In thiscase, it can be concluded that the tool insertenveloped with a thin layer of emulsion, which isnot the case with conventional CLF dosing. Toolachieved full damage at T = 9.65 min. The highervalue of tool life with MQL technique compared toconventional flooding is the result of a thin filmthat completely covered the tool insert surface(Figure 17).Figure 18. Values of surface roughness parameters fordifferent lubrication techniques13 th International Conference on Tribology – Serbiatrib’13 297

Comparative analysis of results 1n the figures 17and 18 indicates that the MQL technique providesbetter results in aspect of tool wear and tool life, butslightly worse results in aspect of the surfacequality as consequence of thermodynamicprocesses in the cutting zone.4. MODELING OF RESULTSExamination of the experimental results wasperformed by multiple regression analysis. (see Fig.19). The output values from the regression modelshowed a significant correlation with theexperimentally measured values. The averagerelative error of the regression models does notexceed 5%. The models presented in the form ofregression equations can be used with highaccuracy of prediction. The models of main cuttingforce (F c ) have errors less than 2%, while the mainsquare errors for models of forces (F f ) and (F p ) arehigher. This corresponds to the theoreticalassumptions of cutting parameters behaviour in thecase of machining of steel C45E.the main cutting force and resultant cutting force.Cutting speed and feed have a great influence onthe force (F f ) and (F p ), and the resultant forces(F f,p ), especially during machining with MQLtechnique. It can be concluded that the increase offeed and depth of cut increases the value of thecomponents of the cutting force. Increasing cuttingspeed reduces the values of cutting force becausethere is no negative phenomenon, such as the burrson the tool edge.Table 7. Models of cutting forces with coefficients ofmachiningMaterial: C45ETool: SNMG 1204 08 NXMCollerationcoefficientRelative error (%)F c = 2485 · a 0.878 · f 0.844 · v -0.047 · K 1 0.99 1.9F f = 1154 · a 0.838 · f 0.391 · v -0.157 · K 2 0.93 3.9F p = 822 · a 0.589 · f 0.644 · v -0.052 · K 3 0.93 3.7Techniques K 1 K 2 K ‚3Conventional 1 1 1MQL 0.99 0.96 1.02Figure 19. Comparison of outputs regression model withexperimental resultsThe biggest errors is expected in the predictivemodels of forces (F f ) and (F p ) in machining withMQL technique; 4.93% and 4.44% respectively. Inaddition to modelling the of cutting forcecomponents, also a resultant force F f,p , which is aresultant of (F p ) and (F f ) was modelled. Theseresultant forces have a higher value than the maincutting force. The resultant force (F f,p ) is a mainindicator of tool wear, and with its growth usuallyintensive tool wear occurs.Analyzing the models, and their correspondingexponents (Figure 19 and Table 7), it can beconcluded that the depth of cut has the highest,while the cutting speed has the least influence onModelling of cutting force for differenttechniques of CLF dosing using regression analysiswas done. The developed models are presented inTable 7, where the influential factors representedby the corresponding coefficient Ki. Developedregression models have error less than 4%, whichindicates the high accuracy of the model. Theapplying MQL technique reduces the energyconsumption compared to the conventionallubrication technique. MQL technique should befavoured in highly productive processes.Often, multiple regression analysis is notsuitable for the modelling of complex processes,which depends on a many number of factors. Formodelling of such processes a large number ofexperimental data are needed. In our study ANNtechnique was applied. For this technique a specialmodule Neural Toolbox in the software packageMATLAB, is used. Modelling with ANN wasconducted using a model of two-layer neuralnetwork with forward propagation.298 13 th International Conference on Tribology – Serbiatrib’13

From the analysis of the diagram it can beconcluded that the output values of both types ofmodel correspond to experimental values. Meanrelative error of predicted values for forces F c , F fand F p in the model based on ANN is 1.01%,2.24% and 1.71% respectively, while for regressionmodel error is 1.85%, 3.55% and 2.92%respectively.Figure 20. Output values of the main cutting force fromregression model and ANN modelFigure 21. Output values of the feed cutting force fromregression model and ANN modelFigures 20, 21 and 22 give the measured andpredicted values for all cutting forces respectively.The results of predicting with the model based onANN show that the developed models can be usedfor modelling the cutting force, although the set oflearning, validation and testing include a relativelysmall number of combinations of input and outputvalues. In order to analyse the accuracy of themultiple regression model the output values of themodel are also shown on the same chart (Table 7).Figure 23. Modelled curves of tool wear for differentlubrication technique in machining steel C45EModelling of tool wear was performed with athird order polynomial function. It can beconcluded from Fig. 23 that matching of outputvalues of model with the experimental values isexcellent.The parameters of the surface roughness R a andR y were modelled with linear function dependingon the machining times and depending on the toolwear using regression analysis.Table 8. Models of surface roughness in dependence oftool wear for different lubrication techniquesMaterial: C45E; a = 2.0 mm; vc = 320 m/minTool: SNMG 1204 08 NXMConv.MQLR y = 15.15 · VB + 1.45 + (f 2 · 10 3 / (8 · r))R y = 30.50 · VB + 1.18 + (f 2 · 10 3 / (8 · r))Figure 22. Output values of the penetration cutting forcefrom regression model and ANN model13 th International Conference on Tribology – Serbiatrib’13 299

where is I i machinability index of i-th material, p iparameter value accepted for machinabilityevaluation of i-th material, p r parameter valueaccepted for machinability evaluation of referentmaterial.Figure 24. Models of parameters roughness for theconventional technique of CLF dosingFigure 26. Values of machinability index based onenergy aspectFigure 25. Models of parameters roughness for theMQL techniqueIf the analysis is carried out from the aspect ofmaterials machinability, R y is a more relevantfactor than R a , because parameter R y indirectindicator of process condition and a direct indicatorof quality as well (Fig. 24 and 25). R y models basedon regression analysis are presented in the form ofthe function of tool wear VB, feed f and the radiusof the tool tip r for both CLF dosing techniques.The second part of this function is taken from theknown empirical expression for the theoretical levelof roughness.Comparison of material machinability was madefor both lubrication techniques under consideration.Machinability index of i-th material in regard toreferent material is defined as:±( p / p ) 1 ⋅100%I =(1)iirFigure 27. Values of machinability index based on thetool life aspectFigure 28. Values of machinability index based onsurface roughness aspectExponent of ratio p i /p r has value of +1 in casethat increase of chosen parameter has positiveeffect on machining process development;otherwise it is a -1 if effect is negative. Results of300 13 th International Conference on Tribology – Serbiatrib’13

comparison from economic, energy consumptionaspect and the aspect of quality of machining arepresented in figures 26, 27 and 28.5. CONCLUSIONMachinability is very important category in theindustry. Based on experimental research and usingthe novel model, machinability of different coolinglubrication techniques can be concluded. Cuttingforces, intensity of tool wear and surface roughnesswere used as the machinability criteria. Analysisshows that turning with MQL is a good alternativefor conventional lubrication. It is important for costof machining and for ecology as well.Future research will be performed in area of lowcost technologies, high productive and hybridmachining processes.REFERENCES[1] W. Grzesik: Advanced machining processes ofmetallic materials: theory, modelling andapplication, Elsevier B. V., Netherland, 2008.[2] L.G. Globočki: Metal cutting: theory, modelling andsimulation, (in Serbian), Faculty of MechanicalEngineering , Banja Luka, BiH, 2010.[3] J. Kopač: Cutting forces and their influence on theeconomics of machining. Journal of MechanicalEngineering, Vol. 48, No. 3, pp. 121-132, 2002[4] F. Klocke, G. Eisenblatter: Dry cutting – state ofresearch, VDI berichte, No. 1399, 1998.[5] V.N. Gaitondea, S.R. Karnik, P. Davim: Selection ofoptimal MQL and cutting conditions for enhancingmachinability in turning of brass, Journal ofMaterials Processing Technology, Vol. 204, pp 459–464, 2008.[6] N.R. Dhar, M.T. Ahmed, S. Islam: An experimentalinvestigation on effect of minimum quantitylubrication in machining AISI 1040 steel, Inter.Journal of Machine Tools & Manufacture, Vol. 47,pp. 748–753, 2007.[7] K. Weinert, I. Inasaki, J.W. Sutherland, T.Wakabayashi: Dry machining and minimumquantity lubrication, CIRP Annals - ManufacturingTechnology, Vol. 53, No. 2, pp. 511 - 537, 2004.[8] B. Sredanović: Developed of model for universalmaterial machinability defining based on cuttingprocess parameters, MSc thesis, Faculty ofMechanical Engineering, (in Serbian), Banja Luka,BiH, 2012.13 th International Conference on Tribology – Serbiatrib’13 301

Serbian TribologySocietySERBIATRIB ‘1313 th International Conference onTribologyKragujevac, Serbia, 15 – 17 May 2013Faculty of Engineeringin KragujevacTRIBOLOGICAL ASPECT OF RUBBER BASED PARTS USED INENGINEERINGDušan Stamenković 1 , Milan Nikolić 1 , Miloš Milošević 1Milan Banić 1 , Aleksandar Miltenović 1 , Miroslav Mijajlović 11 University of Niš, Faculty of Mechanical Engineering in Niš, Serbiadusans@masfak.ni.ac.rs, milan.nikolic.nis@gmail.com, mmilos@masfak.ni.ac.rs, milan.banic@outlook.com,amiltenovic@yahoo.com, mijajlom@masfak.ni.ac.rsAbstract: In most of the cases, the friction is considered as a negative side-effect concerning energyloss following every process of the power transmission. However, the friction has significantpositive side effects, because it is an indispensable prerequisite for the movement of people,machines, transportation means and others. Efficiency of these movements mostly depends on thefriction between rubber and different materials such as metals, concrete, earth, wood, plastic, etc.Certain standards relating to measurement and determination of the friction characteristics ofrubber were established. However considering that tribology of the rubber is very complexproblem, numerous studies around the world are conducted. This paper gives an overview of someof the existing standards and conducted researches in this area. The paper also provides anoverview of theoretical and experimental studies of friction the rubber and the other materials,which are done at Faculty of Mechanical Engineering in Niš.Keywords: coefficient of static friction, rubber, tribology, slip resistance, standards.1. INTRODUCTIONMovement can be realized only and merely bythe friction, but during a motion, frictionpermanently causes different kinds of losses(energy dissipation, mass loss, movement loss).Therefore, friction is the process where positive andnegative effects manifest both. In a certainsituation, a large friction force is required, but inother situation small friction force is required.Because of that, understanding the tribologicalinteractions between the shoe and the floormaterials is important in order to enhance shoe andfloor design and to prevent slip and fall accidentsduring walking.Since the coefficient of friction measurementswere commonly adopted to evaluate slip potentials,it has been found that there were controversies inthe interpretation of measurement results. The study[1] was principally focused on broadening theknowledge base and developing new ideas onwhich improvements in the validity and reliabilityof slip resistance measurements might be made. Toachieve this goal, crucial problems on the currentconcept of slip resistance measurement wereextensively analysed by a tribological point of viewwhere principle of understanding the shoe-floorfriction and wear phenomena could be made. Basedon this approach, new theoretical models weresuggested in paper [1].This study discussed the limitations of presentconcept on slip resistance measurements andanalysed the seriousness of misinterpretations onslip resistance properties that were mainly causedby over-simplified conceptions on frictionphenomena between the shoe heels and floorsurfaces. Based on those critical analyses, a newparadigm on friction and wear phenomena betweenthe shoes and floors was proposed for the futureresearches on the slip resistance measurements.On the basis of totality of the experimental andthe simulation results as well as concepts somerecommendations for dealing with the tribology ofpolymer-based composites – in instruction as wellas industrial and research setting – are made in thepaper [2].302 13 th International Conference on Tribology – Serbiatrib’13

Advantages and disadvantages of traditional andmodern approaches of surface analysis based onconcepts of roughness and texture are discussed inthe paper [3]. Authors considered that traditionalconcept of rough surface based mainly on profileparameters is not fully satisfied modern trends intribology. This paper presents a review of theproblems of rough surfaces analysis in theirevolution from statistical height and stepparameters of profiles to dimensionless and scaleinvariant representation of surface texture. Theyconcluded that texture analysis can be efficientlyapplied for solving practical tribological problemsin micro/nanoscale.Paper [4] presents a study on the surface qualitypointing out the influence of relative sliding on thetopography parameters. A comparative study of thesurface topography, obtained by changing a singleparameter during the tests, may reveal at least aqualitative influence of this parameter that could beuseful for practicians.The authors in paper [5] investigated theboundary friction model, which is built up by thesurface topography. The model contained the effectof boundary film, adhesion, plough and lubrication.Based on the model, a coefficient for weakeningplough for the lubrication was proposed. Themodified model could fit for the working conditionof wet friction elements.In the paper [6] authors indicate that staticfriction is necessary for vehicle starting and runningand show comparative information of static frictionexperiment of prismatic steel samples slip andtribology studies of the wheel-rail contact.The new friction coefficient calculationprocedure based on the Molecular-mechanicaltheory of friction is proposed in the paper [7]. Thisprocedure considers roughness parameters andhardness of contact surfaces, as well as therelationship between the deformation component ofthe static friction coefficient and the total staticfriction coefficient determined experimentally forspecific tribological conditions. Studied tribologicalconditions in the research are related to the press fitjoints of railway vehicles drive unit components.The proposed model considers experimentalresearch of tribomechanical pairs at which plasticdeformations exist in the real area of contact.A review of standards and methods of slipresistance measuring provided by flooring andfootwear suppliers in United Kingdom is presentedin paper [8]. It can be seen that a lot of suppliersdidn’t specify date about the slip resistance of theirproducts.The lack of international standards for the slipresistance of ceramic tiles is stated in the paper [9].The paper considers recent and current potentialdevelopments in the international standardization ofslip resistance. It identifies some limitations of wetbarefoot ramp test, and suggests that changesshould be made.The paper [10] researches the friction betweenrubber and metal which can significantly influencesdamping characteristics of the rubber-metal springs.In the framework of the experimental research thathas being conducted the coefficient of the staticfriction between the rubber and metal has beenestablished in different contact conditions.Moreover, compressions of rubber-metal springsare also performed and force-deflection diagramsare recorded. In this way, the mutual influence ofthe static friction between the rubber and the metalpad and the accumulated/absorbed energy within arubber-metal spring is analyzed.Tribological approach of the contact footwearflooris the subject of research that has started atFaculty of Mechanical Engineering in Niš.Experimental research of static friction of footwearrubber samples and different types of floormaterials is presented in this paper.2. STATIC FRICTIONIn order to achieve vehicle wheel turning on theroad, it is necessary to have the drive torque as wellas a force of resistance in the wheel-road contact.Similarly, in order to make walking on the floorpossible, a drive force delivered by the legs and aforce of resistance in the footwear-floor contact areneeded. This resistance is the static force of slidingfriction. So, wheel rolling is achieved through thestatic friction force of sliding. Likewise, pedestriancan walk with the help of static friction force.Friction represents a resisting force that opposesrelative motion of bodies’ surfaces that are incontact. According to the state of moving, i. e. tothe resultant tangential force that induces movingthere are two types of friction. The static friction orthe stationary state friction that exists when theresultant tangential force is lower than thesummation of all resistances that oppose movingand the kinetic friction or the moving state frictionwhen the force that induces moving is greater thanthe summation of resistant forcesThe diagram F(s) in Fig. 1 shows that the forceincreases from the point O to the point A, where themaximal value of the force is achieved. That is thestatic friction force (Fs). The static friction forcerepresents a maximal tangential resistant force thatacts during so called boundary relativedisplacement. Boundary displacement (preslidingmovement) can be defined as a micro moving offrictional surfaces that goes before visible or macromoving of surfaces in mutual contact (the part OA13 th International Conference on Tribology – Serbiatrib’13 303

of the graphic in Fig. 1). Futhermore, preslidingmovement represents a limit up to which the staticfriction lows between frictional surfaces are valid.After this limit the kinetic friction lows are inaction. Therefore, the presliding movement is aperiod of relative movement characterized by anextensive increase of the reactive force and a smallincrease of movement. Press fit joints, screw andrivet connections, all types of friction transmitters(variators, belt transmitters, couplers), parkingbrakes etc. work in the mode of preslidingmovement.It can be seen that the force retains the value ofthe static friction force (Fs) for a short time periodand then decreases to the value of the kineticfriction force (Fk). This process is followed by anintensive increase of movement.There are the following influencing factors: floor,contamination, footwear, pedestrian factors,cleaning and environment.Footwear suppliers use a variety of terms todescribe their products, as like as ‘slip-resistant’,‘anti-slip’, ‘improving grip performance’ etc. andthese can often mislead customers. Slip-resistantindustrial footwear will normally have been testedaccording to European standards, but manymanufacturers and suppliers do not give helpfuladditional information, such as the degree of slipresistance and the types of work environment forwhich their products are most suited.The aim of the HSE’s project [8] was to collect andassess the slips safety information/literature providedby flooring and footwear suppliers in 2008 in GreatBritain. A significant proportion of flooring products(55%) did not make any reference to slip resistance orprovide any test data. No indication of slip resistancewas given for 47% of footwear products.F k – Kinetic friction forceF s – Static friction forceΔs - Presliding movementFigure 1. Static and kinetic frictionUnder permanent conditions and even for thesame material the coefficient of static friction valueis not a constant and may vary in a certain range.The alteration of friction coefficient values ismostly stochastic, so one can only speak about themean values of the friction coefficient.Friction coefficient values depend on differentparameters such as: nature and properties of theused materials, contact pressure value, thicknessand type of surface film, contact surfacesroughness, duration of the contact, chemicalinteraction, presence of external bodies in thecontact area, cleanness of contact surfaces,temperature of the surrounding environment,relative humidity, elasticity etc.Figure 2. The pendulum friction coefficient testA review of flooring test data showed that 54%was generated using the pendulum test (Fig. 2),33% using the ramp test (Fig. 3), 0.2% usingroughness measurements and 12.8% was generatedusing sled-type test methods, which in the opinionof HSE, can provide misleading results incontaminated conditions. The type of test used fromfootwear suppliers are: RAMP test 46 %, SATRAtest 40% and HSL RAMP test 14%.3. MEASUREMENT OF SLIP RESISTANCENearly 11,000 workers suffered serious injury asa result of a slip in 2007 in Great Britain [8]. A keyelement of HSE’s (Health and Safety Executive)work to reduce slips and trips is to raise awarenessof how slip risks can be controlled through the useof suitable flooring and footwear. Research by theHealth and Safety Laboratory has shown that acombination of factors contribute to slip accidents.Figure 3. The ramp friction coefficient test304 13 th International Conference on Tribology – Serbiatrib’13

The information provided by footwear andflooring manufacturers was not satisfactory. Manyfootwear manufacturers made vague claimssuggesting slip resistance and did not providesupporting data. Many flooring manufacturersavoid making reference to slip resistance altogetherand information is hard to find.Recommendation of the HSE project [8] is thatit was apparent that many suppliers did not considerslip resistance to be a selling point and did notplace significant emphasis on it. Currently, it isvery difficult to make comparisons betweenproducts due to the number of tests used andspecifications quoted. Where test data is provided,very little explanation is given and the laypersoncould be easily confused or misled. Footwear andflooring suppliers should be influenced to placemore emphasis on the slip resistance of theirproducts, and to use more standardized ways ofassessing slip resistance; this would allowcustomers to make comparisons and help them toselect the most appropriate product for their needs.Slip resistance properties of flooring materialsand footwear are covered by various standards inEurope. Some of the most common are: BS7976 – British standard that describesthe specification, operation and calibrationof the Pendulum test, used for assessmentof floor surface slipperiness under both dryand contaminated conditions. DIN51130 - Laboratory based ramp test,using cleated safety boots and motor oilcontamination. Results are reported as an Rvalue, on a scale from R9 to R13, with R9being the least slip resistant. DIN51097 - Laboratory based ramp test,using barefoot operators with soapy wateras the contaminant. Results are reported asClass A, B or C, with A being the least slipresistant. EN13845 - Laboratory based ramp testspecifically for resilient floor coveringswith enhanced slip resistance. The test usesstandardized footwear and soapy watercontamination.EN13287 - Laboratory based mechanicalslip resistance test for safety / occupationalfootwear. The test uses several surfaces andcontaminants to assess footwear.Because of the nature of complexity and factorsinvolved, the measured coefficient of frictionquantities show inconsistencies even as the sameshoe-floor combinations are employed. This facthas been recognized as a great concern whendifferent friction testers, sensors and/or protocolsare used worldwide.However, variations of the coefficient of frictionresults under the same test environments have notreceived much attention in this research area.Despite of this fact, most slip safety researcheshave reported that a particular shoe or floor surfaceresists the movement of a particular floor surface orone’s shoe sole across its surface.4. EXPERIMENTAL RESEARCH OF THESTATIC FRICTIONSlip accidents can happen for a number ofreasons: footwear, flooring, contamination andobstacles, cleaning, human factors, environment,etc. But footwear and flooring are the mostimportant for tribological research.Because of the existence of many differentstandards and methods for assess the slip resistance,measuring of friction coefficient on tribometer inlaboratory condition is very useful.Footwear is produced most from rubber, becauseof its properties. The rubber is elastic, soundproofand it has low gravity density and good tribologicalproperties.Experimental determination of the static frictioncoefficient between samples of footwear soles andflooring were held on Mechanical Faculty in Niš.Static friction force can be measured only in themoment of sliding beginning for the reason that innext moment, after sliding start, this values falls onfriction kinetic force value.Experimental model for establishing staticfriction coefficient, projected for this investigationand which will be used for further investigation, isshown in Fig. 4.Figure 4. Schematic review of device formeasuring static friction forceMeasuring process was done so that by theturning the screw skater start sliding and forcesensor fixed on skater pushes sample A (footwearsole sample). Sample A starts to slide on the sampleB that is fixed in the base of device and pushingforce is measured. Static friction force isestablished in the moment of sliding start.13 th International Conference on Tribology – Serbiatrib’13 305

Measuring system with experimental samples isshown in Figure 5.Figure 5. Measuring systemSamples used in this experimental investigationare with following characteristics: Footwear sole samples (sample A) areprism shaped and formed of soles cutoutglued on a piece of chipboard. Nominalcontact area is 30mmx30mm=900mm 2 . Forthis investigation there are four solesamples: new rubber with relief, worn(used) rubber with texture, new flat rubberand leather. For floor samples (sample B) are usedplates of laminate, rough ceramic tile andsmooth ceramic tile. Dimensions of platesare 60mmx75mm according the measuringdevice.Before testing all contact surfaces are cleanedwith acetone.Floor samples surface roughness was measured byroughness measuring device Mitutoyo Surftest SJ-301. Roughness measuring gave the following results:1. Laminate plate: Ra=0,9µm, R max =4,98µm,R z =3,25µm,2. Rough ceramic tile: Ra=12,85µm,R max =59,04µm, R z =43,93µm,3. Smooth ceramic tile: Ra=0,53µm,R max =3,44µm, R z =2,24µm.Measurements are done with weight (normalforce) variations so that contact pressure was:45kPa, 79kPa and 142kPa.Force sensor is produced by HBM, maximumforce which can be measured is 500N and samplerate is 100Hz. For each contact combination fivemeasuring were done.Contact surfaces are prepared in three ways: drycondition, wet condition and soap lubricated.Tables 1, 2 and 3 show measuring results forstatic friction coefficient for different materialcombination and lubricating. Marks in the tablesare: U1-new rubber with relief, U2-worn rubberwith texture, U3-new flat rubber, U4-leather, P1-laminate plate, P2-rough ceramic tile and P3-smooth ceramic tile.Table 1. Static friction coefficient of footwear solesamples and laminate floor sample (P1)µ U1/P1 U2/P1 U3/P1 U4/P1dry 0,54 0,83 0,96 0,52wet 0,39 0,66 0,67 0,65soap 0,43 0,60 0,46 0,70Table 2. Static friction coefficient of footwear solesamples and rough ceramic tile sample (P2)µ U1/P2 U2/P2 U3/P2 U4/P2dry 0,52 0,47 0,54 0,63wet 0,46 0,40 0,58 0,79soap 0,38 0,54 0,39 0,77Table 3. Static friction coefficient of footwear samplessole and smooth ceramic tile sample (P3)µ U1/P3 U2/P3 U3/P3 U4/P3dry 0,25 0,69 0,44 0,47wet 0,19 0,53 0,42 0,52soap 0,11 0,22 0,13 0,50Performed experiment shows that values of staticfriction coefficient are very unpredictable and random.Static friction coefficient of leather sample (U4) withpresence of lubricants (water, soap) increases that isopposed of rubber samples with lubricants wherecoefficient of friction decreases. Very interesting resultswere in combination of rubber sample with relief (U4)and smooth ceramic tile, respectively measured staticfriction coefficient is very small (0,25 in dry conditionuntil 0,11 lubricated with soap). That can be explainedwith small real contact area. Also, it can be concludethat for smooth ceramic tile coefficient of static frictionis smallest for each sample.Force (N)65605550454035302520151050Experimental test U3P3M3s5 5,2 5,4 5,6 5,8 6 6,2 6,4 6,6 6,8 7Time (s)Figure 6. Friction force-time diagram for flat rubber andsmooth ceramic tile (normal load 131N, dry condition)Force (N)Experimental test U4P3M2v40353025201510501,5 1,7 1,9 2,1 2,3 2,5 2,7 2,9 3,1 3,3 3,5Time (s)Figure 7. Friction force-time diagram for leather andsmooth ceramic tile (normal load 72,28N, wet condition)306 13 th International Conference on Tribology – Serbiatrib’13

Force (N)5,554,543,532,521,510,50Experimental test U1P3M1sa3,4 3,5 3,6 3,7 3,8 3,9 4 4,1 4,2 4,3 4,4Time (s)Figure 8. Friction force-time diagram for new rubberwith relief and smooth ceramic tile(normal load 41,82N, soap condition)Figures 6, 7 and 8 give the representativeexamples of recorded friction force in performedexperimental investigation.On the presented diagrams can be seen staticand kinetic friction force when footwear solesamples slides over floor samples. Vertical axisrepresents force in Newtons and horizontal axistime in seconds. Diagrams show that friction forceincreases from zero value to the maximum valuethat is static friction force, and then falls to thekinetic friction force.5. CONCLUSIONSDue to the lack of static friction force in contactfootwear-floor is often the reason for falls andinjuries it is necessary to pay more attention infootwear and floor production in part of tribologicalproperties. Certain standards about the slipresistance assessing are established in EU. Up tonow in Serbia there isn’t enough professionalinterest for this area, and it is left to the producersof footwear and floor.Because of the existence of many differentstandards and methods for assess the slip resistance,measuring of friction coefficient on tribometer inlaboratory condition is very useful.According the importance of this problem andexperience in earlier studies in the field of staticfriction, at Faculty of Mechanical Engineering inNiš is initiated research with the aim to determinetribological properties of rubber produces asfootwear.In that sense measurement of static frictioncoefficient between footwear sole and floorsamples was performed. For that purpose it wasdesigned measuring device for static frictionestimation. Measuring results show that staticfriction coefficient is stochastic and unpredictable.In further investigation it is necessary toimprove measuring system and include moresamples. Some samples should be industrial shoesand floors, tiles on public walkways, white stripeson pedestrian crosses the street and material otherrisky points where falls and accidents can happen.REFERENCES[1] I.J. Kim, H. Nagata: Research on Slip ResistanceMeasurements — A New Challenge, IndustrialHealth 46, pp.66–76, 2008.[2] W. Brostow, V. Kovačević, D. Vrsaljko, J.Whitworth: Tribology of polymers and polymerbasedcomposites; Journal of Materials Education32, pp. 273 – 290, 2010.[3] N. K. Myshkin, A.Ya. Grigoriev: Roughness andTexture Concepts in Tribology; Tribology inIndustry Vol. 35, No. 2; p.p. 97‐103; 2013.[4] N. Diaconu, L. Deleanu, F. Potecasu, S. Ciortan:The Influence of the Relative Sliding on the SurfaceQuality; Tribology in industry, Volume 33, No. 3,p.p.110-115; 2011.[5] W. Yanzhong , W. Bin , W. Xiangyu: Wet Friction-Elements Boundary Friction Mechanism andFriction Coefficient Prediction; Tribology inIndustry Vol. 34, No. 4; p.p. 198-205; 2012.[6] M. Đurđanović, D.Stamenković: Static frictioncauses movement, SERBIATRIB ’07 – 10 thInternational Conference on Tribology, p.p.. 215-218, Kragujevac 2007.[7] D. Stamenković, M. Milošević, M. Mijajlović,M. Banić: Estimation of the static friction coefficientfor press fit joints; Journal of the BalkanTribological Association 17, pp. 341-355, 2011.[8] M. Liddle: Assessment of slips safety information/literature provided by flooring and footwearsuppliers, RR747 Research Report, Health andSafety Executive, 2009.[9] R. Bowman: Slip resistance testing - Zones ofuncertainty, Bol. Soc. Esp. Ceram. 49, pp.227-238,2010.[10] D. Stamenković, M. Milošević: Friction at rubbermetalspring, SERBIATRIB ’09 – 11 th InternationalConference on Tribology, pp. 215-219, Beograd2009.[11] HSE: Assessing the slip resistance of flooring, Atechnical information sheet published by the Healthand Safety Executive, 2012.[12] D. Stamenković, M. Milošević: Experimentalinvestigation of Static Friction, InternationalConference ''POWER TRANSMISSIONS '03''Section III "Experimental Investigations andApplications", pp. 67-69.Varna Bulgaria, 2003.13 th International Conference on Tribology – Serbiatrib’13 307

Serbian TribologySocietySERBIATRIB ‘1313 th International Conference onTribologyKragujevac, Serbia, 15 – 17 May 2013Faculty of Engineeringin KragujevacPOSSIBILITY OF REPLACING THE CHORINATED PARAFFINSIN METALWORKING FLUIDSMarica Dugić 1 , Branka Kojić 1 , Pero Dugić 1 , Goran Dugić 11 Oil Refinery Modriča, Bosnia and Herzegovinamajad@modricaoil.com, branka@modricaoil.com, pero@modricaoil.com, gdugic@modricaoil.comAbstract: All components that are used for neat oil formulations for metalworking operations belong togroup of chemicals that, because of its direct application, can have a big influence on the environment andespecially on human health.With time, metalworking operations become more complex and demanding. New materials with variouscompositions are processed, and they demand metalworking fluids with improved antiwear properties andthat provide extreme load-carrying capabilities.Exactly these additions, known as EP additives, that had been until recently based on chlorinated paraffins,are a group of chemicals that need to be replaced as soon as possible with a components that areenvironmentally acceptable.In this paper, in laboratory conditions, we investigated synthetic polymeric esters, as one of the possiblesubstitute for chlorinated paraffins, in several neat oil formulations, which are used for variousmetalworking operations. From the above comparative results of analysis results it can be concluded thatsynthetic polymeric esters can serve like an adequate substitute for metalworking fluids which are designedfor easy operations.Keywords: neat metalworking oil, environmentally acceptable lubricants, EP additives, chlorinatedparaffins, synthetic polymeric esters.1. INTRODUCTIONFormulation of industrial lubricants, dependingof performance that lubricants must have, requires alot of different types of additives. Their role in thelubricants is to improve certain characterisics oflubricant or to give the lubricant brand newperformance. Usual types of additives are:corrosion and oxidation inhibitor, antifoaming,viscosity index modifier, pour point depressant,antifriction and antiwear additive and additives thatallow carrying heavy loads, usually called EPadditives.That large number of additives in particularlubricants formulation lead to a questions not onlyabout the effect of additives in the lubricants, butalso the influence of one additive to the another.Complex testing of antagonism and synergismof all components on particular characteristics indifferent lubricant formulations are being made. Itcan sometimes induce a big problem, especially inrecent times, when many components, due tounfavorable effect on humans and environment, arebeing completely banned for usage in lubricants, ortheir allowed dosage is greatly reduced.Research of new environmentally acceptablecomponents and their influence in the lubricant hasstarted many laboratory tests and in the same timetests of their applications in service.Analyzing the results of one survey, which theeditorial of scientific magazine Tribology &Lubrication Technology made, on the question atwhich additive has the most influence onperformance characteristics of lubricants, showedthat the answer was EP additives. [1]The subject of investigation in this paper istesting the antiwear and EP characteristics ofdifferent formulations of neat oil for metalworkingoperations, which in its composition have EPadditives. Exactly this extras, known as EP308 13 th International Conference on Tribology – Serbiatrib’13

additives, that had been, until recently, based onchlorinated paraffins are a group of chemicals thatneeds to be replaced as soon as possible withcomponents that are environmentally acceptable.In this paper, in laboratory conditions areprepared samples of several different formulationof neat oil which are intended for differentmetalworking operations. Among the otherstandard characteristics, are also tested antiwearand EP characteristics of samples containingsynthetic polymeric esters, as one of the possiblereplacement for chlorinated paraffines. The resultsof analysis are compared with the results of thesamples that are formulated with chlorinatedparaffins.2. EXTREME PRESSURE (EP) ADDITIVESPrimary functions of metalworking fluids areheat control, cooling the surfaces of tools and partsprocessed in the cutting zone and lubrication andenduring the extreme pressures and heavy loadingduring processing. It is particularly important inorder to achieve desired size, shape and degree ofsurface processing of workpiece, and longer toollifetime. Removal of cuttings from operating zoneis considered as secondary role, but it is veryimportant and it needs to be done continually.Corrosion protection of tools, machines andworkpieces is also important. Freshly processedmetal have tendency to corrode faster duringmachine operations, than protected metal.For metalworking it isn’t enough to use only onekind of lubricant, but every applicable processrequires specially formulated oil. In order toemphasize differences in characteristics of certainlubricants, in this case, of neat oils, and point outthe complexity of choosing the basestocks forformulations, we start with metalworkingprocesses, which are:- Removal (formation of cuttings), mostlynamed cutting operations- Shaping: pressing, rolling, drawing...- Treating (surface strengthening): quenching- Protection (corrosion protection: temporarywith interprocess, during transport...)Cutting operations can be considered as the mostcommon in metalworking operations. Basicmethods of metalworking cutting are:- Scraping- Planning (during operation it is identical toscraping, but the process is discontinuous- Drilling (beveling, widely considered:drilling, widening, reaming)- Milling (oblique and orthogonal cutting)- Grinding (removal of thin layer of materialduring the finest final processing)Operations of metal forming, such as: pressing,rolling, drawing (wire, profiles) are also veryused and are considered as demanding.2.1. EP additives and their role in metalworkingfluids for heavy operationsTraditional additive packages that are added toneat metalworking oils in the boundary lubricationregime remain on the metal surfaces and theycannot prevent increased friction, wear and thedamage to the tool. EP additives are needed toenable the application in the more difficultconditions of elevated temperatures and pressures.The main groups of EP additives are:1. Chlorine compounds2. Phosphorus additives3. Sulphur additives4. Prebasic sulfonates (calcium and sodium)The first three EP additives are activated inreactions with the metal surfaces in a certaintemperature range. Chlorinated additives areactivated at temperatures between 180 i 420 ⁰C,phosphorus are activated at higher temperatures andsulphur at even higher temperature range whichends at 1000 ⁰C. In the reaction with the metalsurface these three types of additives produce Fechloride,Fe-phosphide and Fe-sulphide, whichserve as a barrier to reduce friction and wear andelimination of welding.The fourth EP additive, prebasic sulfonate, actby other mechanisms, whose process doesn’tdepend on temperature, and operate below 500 o C.In each category of EP additives there areseveral different types. The main used chlorinatedadditives are chlorinated paraffin, or even called achlorparaffine. In case of phosphorus compounds,the most common type is phosphate ester. Withsuphur additives those are: sulphurated fats,sulphurated esters, sulphurated hydrocarbons andpolysulphide, which differ in the concentration offree (active) sulphur.In order to select EP additive in formulations formetalworking fluids, one must be familiar with theapplications itself, machining operations, state ofthe tools and to have defined expectedperformance. One of the most importantinformation is the possibility of early cancellationof tools before the operation has reached the righttemperature for EP additive to become activated.Besides the choice of EP additive, it is necessary toknow the synergism with other additives includedin the formulation, and also to reduce undesirableinteractions which can bring to instability ofproduct using certain combination (foaming,sludge).13 th International Conference on Tribology – Serbiatrib’13 309

The greatest influence on the formulation ofneat oils for heavier operations was limiting thechlorine content. For manufacturers of thesesubstances it was hard task to find an adequatereplacement in a very short period of time.Many studies have been done on the impact ofchlorinated paraffins on the environment, and itreached its expansion in mid-nineties.Chlorinated paraffines are divided in threeclasses based on the chain lenght:- Short-chain, C 10-13- Medium-chain, C 14-17- Long-chain, C 18-30The most commonly used chlorinated paraffinswere with 35-70 % of chlorine and with thehydrocarbon base with C 10-20 .These studies have shown that short-chainchlorinated paraffins have the biggest potential riskof influence on the environment especially if theyare not handled properly when they become wastematerial.[2]In summary, therefore, the lists of undesirablechemicals in the formulations of manymetalworking fluid include one of the mostcommonly used EP additives, chlorinated paraffin.Chlorine in lubricants, at elevated temperaturesand pressures reacts with hydrogen, and formshydrogen chloride and dioxine, and dissolvinghydrogen chloride in water forms hydrochloricacid.A special problem brings the impossibility ofrerafination of used products that contain chlorine,since chlorine acts a catalityc poison. If thisproducts are burned, incomplete combustion alsoforms hydrochloric acid, which comes inenvironmentally round cycle and leads to increasedenvironmental pollution. [3]3. EXPERIMENTAL PARTBecause of all findings on harmfulness ofchlorparaffins, new components are beingresearched, which might be more ecologicallyacceptable, with function to serve as a replacementfor chlorparaffins in all places where severemetalworking opperations are, and where it was,until recent, only EP additive.A new generation of EP additives has beendeveloped, that are, by their chemistry, compundsof sulphur and phosphorus, and new generation ofsynthetic esters is currently being tested in severemetalworking fluids. A special focus is onsynergism of synthetic esters and sulphur basedadditives.For research of antiwear and EP characteristicsof formulations that have chlorparaffin replacedwith two synthetic esters, we chose formulationsfor vertical broaching, drawing and final finishingoperations, like fine stamping.Characteristics of synthetic ester that were usedin tested formulations are listed in Table 1. InTable 2. are listed characteristics and methods usedfor metalworking samples testing for all threemetalworking operations. [4]Table 1. Characteristics of syn ester 1 i 2.Characteristic Syn ester 1 Syn ester 2AppearanceHazy. Light Slightly hazy.amber Light amberSpecific gravity @ 25 0 C 1,00 1,01Acid number, mg KOH/g 20

Figure 1. Test results for samples of welding points forvertical broaching operations with differentconcentrations of chlorparaffine.Figure 3. Test results for samples of welding points forvertical broaching operations with differentconcentrations of Syn ester 1.Figure 2. Test results for samples of wear scar forvertical broaching operations with differentconcentrations of chlorparaffine.In Figure 1 are shown welding point values incomparison to increase of chlorparaffins content,were broaching operation require chlorparaffinscontent of up to 30% in order to achieve weldingpoint of 7000 N, which will provide goodprocessing.In Figure 2 are shown wear diameter valuescompared to chlorparaffin content. Diameter weartest values in broaching opereations are not soimportant, welding point values.Table 4. Test results of samples for vertical broachingoperations which formulation contains Syn ester 1.Neat oil for vertical broachingSample 1 2 3 4Syn ester 1 % m/m - 2 6 10Additive A,% m/m 4 4 4 4Base oil, % m/m 96 94 90 86Results1 2 3 4Wear, 40 kg, mm 0,96 0,76 0,61 0,67Welding point, N 3800 4000 4800 4800V at 40 0 C, mm 2 /s 34,47 29,85 31,58 37,83Cu corrosion 4c 3b 3a 3bAcid number 0,81 1,85 3,95 6,21Figure 4. Test results for samples of wear scar forvertical broaching operations with differentconcentrations of Syn ester 1.With Syn ester 1 are prepared samples for finestamping operations, whose test results arecompared with the classical formulations containingoptimal concentration of chlorinated paraffine tomeet the performance which this operationdemands. Test results of welding points (Table 5.and Figure. 5.) shows that nor with the maximalconcentration of Syn ester 1 do not achievesatisfactory results. Because of that, samples withSyn ester 2 are prepared where test results ofwelding point has shown that nor with it do notachieve the result which in practical conditionscould meet all expected performance. (Table 6. andFigure 7.)Figure 5. Test results for samples of welding points forfine stamping operations with different concentrations ofSyn ester 1 and classical formulation with chlorparaffine.13 th International Conference on Tribology – Serbiatrib’13 311

Table 5. Test results of samples for fine stampingoperation whose forrmulation contains chlorparaffine(Sample 1) and Syn ester 1 (2-6).Neat oil for stampingSample 1 2 3 4 5 6Syn ester 1,% m/m- 5 8 10 20 28Hlorparafin 28 - - - - -Additivepackage,% m/m4 2 2 2 2 2Base oil,% m/m68 93 90 88 78 70Results1 2 3 4 5 6Wear, 40kg, mm0,71 0,42 0,40 0,42 0,41 0,41Weldingpoint, N4400 2000 280 2700 2800 2800V at 40 0 C 89,2 74,8 85,2 89,38 97,7 106,5Cucorrosion1a 1a 1a 3a 1a 3aAcidnumber1,46 5,35 6,62 7,90 13,4 18,16completly satisfies all performance demands forthat operation. With Syn estrer 2 optimal quantityof additive package there is a formulation whosetest welding results are close to those withchlorparaffin. (Table 7. and Figure 9.)Table 6. Test results of samples for fine stampingoperation whose forrmulation contains chlorparaffine(Sample 1) and Syn ester 2 (2-5).Neat oil for stampingSample 1 2 3 4 5Syn ester 2,% m/m- 6 8 10 12Hlorparafin 28 - - - -Additivepackage, 4 2 2 2 2% m/mBase oil,% m/m68 92 90 88 86Results1 2 3 4 5Wear, 40 kg,mm0,71 0,45 0,47 0,5 0,45Weldingpoint, N4400 2500 2400 4200 2400V at 40 0 C,mm 2 /s89,1 89,91 95,36 96,33 98,99Cu corrosion(3 h, 100 0 C)1 a 1a 1a 1a 1aAcid number 1,46 4,45 5,06 5,69 6,35Figure 6. Test results of wear diametar with samples forstamping operations with different concentrations ofSyn ester 1.Figure 8. Test results of wear diametar with forstamping operations with different concentrations of Synester 2 and classical formulation with chlorparaffine.Figure 7. Test results of welding points with samples forstamping operations with different concentrations of Synester 2 and classical formulation with chlorparaffine.For steel profile drawing operations there isformulation with the optimal amount ofchlorparaffins and additive packages, whichSlika 9. Test results of welding points with samples fordrawing operations with classical formulation withchlorparaffine and Syn ester 2.312 13 th International Conference on Tribology – Serbiatrib’13

Table 7. Test results of oil samples for drawingoperations with classical formulation with chlorinatedparaffine and Syn ester 2.Neat oil for drawingSample 1 Sample 2ChlorinatedSyn ester 2,3paraffine, % m/m% m/m10Additive package, %Additive package,18m/m% m/m7Base oil, % m/m 56 Base oil, % m/m 73Vegetable oil,Vegetable oil,23% m/m% m/m10Results1 2Wear, 40 kg, mm 0,44 Wear, 40 kg,mm 0,35Welding point, N 4000 Welding point,N 4200V at 40 0 C, mm 2 /s 44,26Vat 40 0 C,mm 2 /s40,02Cu corrosionCu corrosion(3 h, 100 0 3aC)(3 h, 100 0 C)1bAcid number,Acid number, mg4,67mg KOH/grKOH/gr7,81which are prepared with Syn esters, as areplacement for chlorinated paraffine, shows:- test result of welding points has showed that atheavy operations such as a broaching with Synester 1 they do not achieve good results, although itis attempted to use synergistic effect of esters andadditives with active sulphur.- with stamping operations, which is alsoconsidered as heavy, welding points test showedthat with two different formulated Syn estes it isimpossible to achieve good results, which will alsoaffect application conditions- based on test results for steel profiles drawingoperations, one can conclude that Syn ester 2 canbe used for formulations intended for thatoperations, with reduced quantities of otheradditives.This is the start of one comprehensive research,in laboratory and application conditions, that arebeing done with additive manufacturers. Testresults shows that it is hard to achieve replacementfor chlorparaffins in formulations of heavymetalworking operations, like broaching, deepdrilling, stamping and others.Constant arguing about usage of chlorparaffinsand their allowed chain length, confuse both usersand manufacturers. But it initiated many studiesregarding that problematic and efforts to findadequate replacement for chlorparaffins, as soon aspossible.REFERENCESFigure 10. Test results of diameter wear with samplesfor drawing operations with classical formulationwith chlorparaffine and Syn ester 2.4. CONCLUSIONTest results of samples of neat oils intended foroperations: broaching, fine stamping and drawing[1] N. Canter: Special Report: Trends in extremepressure additives, Tribology & LubricationTechnology, September 2007., pp.10-17[2] T. Kelley, R. Fensterhein, A. Jaques : “Voices &Views”, Chlorinated Paraffins, Vol.59, No.4, April2009.[3] L. Rudnick, “Syntetics, Minerals oils, and Bio basedLubricants”, 2006[4] Internal documentation of additive manufacturer13 th International Conference on Tribology – Serbiatrib’13 313

Serbian TribologySocietySERBIATRIB ‘1313 th International Conference onTribologyKragujevac, Serbia, 15 – 17 May 2013Faculty of Engineeringin KragujevacQUALITY OF PLASMA CUTTINGBogdan Nedić 1 , Marko Janković 2 , Miroslav Radovanović 3 , Gordana Globočki Lakić 41University of Kragujevac, Faculty of engineering, Kragujevac, nedic@kg.ac.rs2Fiat automobili Srbija, Kragujevac3University of Niš, Faculty of mechanical engineering, Niš4 Univerzitet u Banja Luci, Mašinski fakultet, Banja Luka, RS, BiHAbstract: The plasma arc cutting process severs metal by using a constricted arc to melt a localized area ofa workpiece, removing the molten material with a high-velocity jet of ionized gas issuing from theconstricting orifice. The ionized gas is a plasma, hence the name of the process. This paper analyzes qualityof cut in plasma arc cutting. Quality of cut in plasma arc cutting is defined using standard EN ISO 9013. Thecorrelation between quality of plasma cut Conclusions of other authors who investigated quality of plasmacut are also presented.In the second part of the paper, experimental investigation of plasma cut waspresented. Samples of steel plate thickness of 15 mm were used for creating 17 cuts. Obtained experimentalresults are consistent with theoretical considerations, as well as previous experimental results.Keywords: Plasma arc cutting, experiment, quality of cut, process parameters1. INTRODUCTIONThe efficient manufacture of high-quality platecomponents is quite difficult task. One of theeasiest methods of contour cutting steel is oxy-fuelcutting. With respect to oxy-fuel cutting, lasercutting, abrasive water jet cutting, and plasmacutting are new attractive advanced processes forcontour cutting of plate. They have numerousadvantages, namely, a narrow cut, a proper cutprofile, smooth and flat edges, minimaldeformation of a workpiece, the possibility ofapplying high feed rates, intricate profilemanufacture and fast adaptation to changes inmanufacturing programs [1].Plasma cutting is an industrial process that isessentially controlled by the operator who usesrecommendations given by the manufacturers of thecutting equipment. Those recommendations,however, reflect the point of view of themanufacturers’ business, which includes not onlyselling the cutting torches but also the consumables.Yet, the manufacturers’ recommendations usuallylead to solutions that are technically sound in termsof cutting quality, but do not necessarily correspondto the most cost-effective solutions on the user’spoint of view [2].As a result, the user attempts to optimize thecutting operations by trial-and-error every time it isneeded to setup the existing equipment for a newdifferent task.This requires the development of full studies andapply theoretical and experimental researchesamong all the technological system’s links, in order toestablish (chose) the optimal processing variant.2. PROCESS PARAMETERSAs in the case of other machining methods, atthe plasma arc cutting (PAC), in order to obtaingood results, it is very important to well know theprocess, this meaning to exactly know what are theparameters involved in the process and theirinfluences (fig. 1).Input parameters are those parameters that canbe controlled, and their values are known and canbe set by the operator. Output parameters related tothe quality of the obtained surface. Factors whichcannot be controlled coming from machines andwork environment.314 13 th International Conference on Tribology – Serbiatrib’13

Input parameters- Current intensity (Ip, A)- Plasma gas voltage (Up, V)- Working speed (V, mm/min)- Workpieces thickness (a, mm)- Type of material workpieces- Nozzle distance (b, mm)- Pressure gas (p, MPa)- Type of plasma gas- Cutting directionDisturbing factors- Current intensity variation- Inaccuracy of the equipment- Purity of the plasma gas- Etc.PlasmacuttingOutput parameters- Kerf width- Roughness of the obtain surfaces, Ra- Angle of bevel cut ( )- Melting of the top edge (r)- Dross formation- Cutting precision- Thickness of heat affected zone - ZIT- Perpendicularity, angularity tolerance (u)This standard defines terms like: kerf width,angle of bevel cut… that can be used to define thequality of the work piece.Squareness and angularity tolerance is definedas distance between two parallel straight lines thatlimit the upper and lower boundaries of the cut faceprofile at the teoretically correct angle, 90 degreesfor square cut edges, fig. 3. Standard establishes azone of significance for the measurement of Ureduced at the the top and bottom edge by distance,∆a, related to material thickness. This measure alsoapplies to concave and convex surfaces.Since several works in literature highlight thatthe torch cutting direction and the swirling directionof the plasma gas determine a different squarenessand angularity tolerance in the two sides of the cut[6], both left and right (UL and UR, respectively)were measured in order to highlight the asymmetricbehaviour of the plasma beam.Figure 1. Parameters of plasma arc cutting [3]3. QUALITY OF PLASMA CUTTING PROCESSEuropean standard "EN ISO 9013" "ThermalCutting" defined classification of thermal cutting,contains geometrical product specification andquality. Standard applies to materials suitable forcutting with oxy-fuel (from 3 to 300 mm), plasmacutting (1 to 150 mm) and laser cutting (0.5 to 40mm) [4], fig. 2.Top SpatterDross Build UpCut SurfaceDrag linesKerf WidthCut SurfaceBevel AngleTop EdgeRoundingFigure 2. Quality parameters of a plasma cut [5, 14]Using standard "EN ISO 9013" quality of thesurface is defined by the following parameters:- Squareness and angularity tolerance (u)- Average peak-to-valley height (Rz5, Ra) -roughness- Drag (n)- Melting of the top edge (r)- Possible formation of burrs or drops ofmolten metal on the bottom of the cut-edge.Figure 3. Squareness and angularity tolerance [5]Surface roughness is defined cut appearance,and gives information on whether the need forfurther processing. Parameter surface roughness ismean height of the profile Rz5, unit is μm.Surface roughness is influenced by more inputparameters, but the most influential are: cuttingspeed, current and material thickness. Based on thedeveloped mathematical model [7, 8], shows thatthe thickness of the material has the greatestinfluence on surface roughness. This is logicalbecause current and cutting speed are functions ofmaterial thickness. Surface roughness is a functionof material thickness defined by ISO 9013, basedon which we can see that the thinner material havelower roughness (Fig. 4).Roughness of the cutting edge is connected withstability of process. When the torch is too highpositioning from work piece plasma arc is a longand curve. This phenomenon leads to the formationof surface waves, and lynx, and therefore to ahigher Rz. When the cutting speed increases thetorch moves fast and plasma arc loses stability withview to the cutting front. Therefore plasma arc cannot remain perpendicular to front surface of thework piece, which is on surface cutting formedlynx. On the other side too low cutting speeds leadto excessive melting in a work zone, resultingappearance of furrows.It is known that surface roughness of the cut isnot same by depth. Experimental studies [9]13 th International Conference on Tribology – Serbiatrib’13 315

showed that diameter nozzle has a larger effect onsurface roughness on upper reaches of cutting (1mm from the top edge) than in the lower zones (5mm from the top edge). These studies demonstratedthat higher values of pressure gives lower values ofRz.Mean height of the profile Rz5, m15010050110+1,8 . aRange 4Range 3Range 270+1,2 . aRange 140+0,8 . a10+0,6 . a0 5 10 15 20 25 30Cut thickness a, mmFigure 4. Influence of material thickness on Rz5by standard "EN ISO 9013"Surface roughness is different for the left andright sides of the cut. Surface on the right are about25% rougher than on the left side [10].Drag (n) is the projected distance between twoedges of drag lines (lag lines) in the direction ofcutting (Fig. 5.). At extremely high speeds arcbecomes unstable and oscillates so the sparks andmolten metal form a line in the form of a "tail"(drag lines) Fig. 5. At high speeds drag angle variesfrom 60 º - 80 º, while at maximum speed this anglehas a value of 90 º and cut is lost. In the lower thirdof the cut the arc sweeps back steeply. It is probablethat the hot gas, with no tendency to attach metalwalls, leads the arc slightly at the bottom. Such asmall amount of molten metal from the output portis not ejected.Drag angleDrag lineDrossFigure 5. Dross, drag lines, drag angle [4, 11]Melting of the top edge (r) occurs due to highcutting speeds or long distance from the nozzle towork piece. Top edge can be with overhang.rrrsharp edge molten edge overhang edgeFigure 6. Melting [4]On the top edge may appear slag spatter,accumulation of hardened metal sprayed along theedges of the cut. Basically it is easily removed. Thisphenomenon occurs at high speed, the largedistance between the nozzle and the work piece, ifthe nozzle is wears. Slag spatter may also arisefrom the whirling motion of plasma gas. Thisphenomenon occurs when there is a big positiveangle, because along the bevel is a difference inpressure that ejected molten metal on top.In plasma cutting one of the biggest problemsthat arises is the dross (burr formation on thebottom of the kerf and spatter on the top of the kerf)Concentrations of dross will be higher in the worseside of cut. The amount of dross depends of lotparameters, but the most influential: types ofmaterials, cutting speed and currents [7]. There aretwo types of molten metal on the bottom of cutedge[12]:- low speed dross- high speed drossIf the cutting speed is too low plasma starts tosearch more material to cut. Then cut expanding tothe point where more molten metal does not eject.The molten metal is accumulating along the loweredge of the large bulbous-shaped form. Thusformed molten metal are easily remove. Meltedmetal formed by low-speed cutting followed by theoccurrence concave surface (Figure 7). It may bethat the molten metal and the bottom edge forms a"bridge" over which the sectioned piecesconnecting again. At extremely low speeds arc isturns off because there is not enough metal to holdarc. Increasing power or decrease distance betweenthe nozzle and cutting objects have the same effecton appearance of molten metal. The practicalcounter measures for this phenomenon is removingpart of the heat from the cutting zone, which isachieved by: reducing amperage or increase thedistance between the nozzle and the work piece.Figure 7. Low speed drossHigh speed of molten metal gives a rounded tipof cut edges (Figure 8). Molten metal on the bottom316 13 th International Conference on Tribology – Serbiatrib’13

edge formed into a thin line. At these speeds arcoften does not penetrate into the metal and can beshut down. Long distance between the nozzle andthe work piece, and a small amperage current canlead to the same phenomenon as the speed too high.Increasing the current or reducing distance betweennozzle and work piece leads to more heat in thecutting zone, which the effect of drag line is reduce,effectively reduce negative consequences of highspeeds.Figure 8. High speed drossAngle of bevel cut is angle between the cutsurface and the top surface of the work piece. Theangle of inclination can be positive (the upper partis smaller size then the bottom) or negative (lowerdimension is smaller than the upper). Plasma arccutting will usually result in an angle on the cutsurface of approximately 1 to 3° on the "good" sideand 3 to 8° on the "bad" side, when using torchesthat swirl the plasma. With larninar-flow torches,the angle on both sides is usually about 4 to 8° [14].Best sideWorst sideFigure 9. Positive angle of bevel cut [13]Arc first establishes a connection with a betterside of and releases heat. Cutting speed, current anddistance from the work piece also have influence onangle of bevel cut.If the cutting speed and the current have acorrect value, and part have big positive angle ofbevel cut, then the distance from the nozzle to workpiece is too much. If the cutting speed and thecurrent have a correct value, and distance from thenozzle and work piece is small, then angle of bevelcut will be negative. Optimal distance from thenozzle to work piece is a distance before angle ofbevel cut start to appear [15].Kerf width, the rule is that the cutting width atthe plasma cutting is about 1.5-2 times bigger thanthe size of the nozzle exit. Cutting width isinfluenced by the cutting speed. If the cutting speeddecreases, the cut is expanding.4. EXPERIMENTAL EXAMINATIONIn the experiment were done 17 samples. Asinput parameters used cutting speed V, mm/minand current intensity I, A. Steel plate material S235JRG2 (Č0361) was 15 mm thick. Electric current of60 A, 80 A, 100 A and 120 A was used incombination with the tablet, reduced and increasedspeeds value (Table 1).Table 1. Used speeds and current intensitysamples I, A V , mm/min1 60 4252 60 530 (tablet value)3 60 6354 80 4905 80 610 (tablet value)6 80 7307 80 8708 80 10559 100 53010 100 69511 100 870 (tablet value)12 100 105513 120 73014 120 87015 120 105516 120 1320 (tablet value)17 120 1585Quality of samples is defined by:- surface roughness Ra (measured by deviceTalysurf 6)- kerf width (St and Sb, defined in Fig. 10.)- dross width (Sd) and dross higth (h), Fig. 10.- dimension of molten metal on the bottom edgeof the cut. (defined in Fig. 11.)Ra was measured in thre points: near to upper edge(g), in the middle (s), near to lower edge (d).5040A40A20rStSbSdhA-AFigure 10. Measuring kerf width and molten metal onthe bottom of cut2,52,5Figure 11. Measured in three pointsUsing a current of 60 A quality cut can bedescribed as very poor due to large deposits ofgsd13 th International Conference on Tribology – Serbiatrib’13 317

waste material to the bottom of cut. Thereforerequires additional treatment, and just cutting itvery slowly (Fig. 12.).Figure 12. Cutting with 530 mm/min and 60 AFigure 13 shows the sample which was cuttingwith current intensity of 80 A and speed of 610mm/min (tablet value). The quality of the cut canbe characterized as acceptable. There are wasteparticles on the bottom, but not on a large extent.An additional treatment section is necessary. Usinga current of 80 A and a change speed, quality isapproximate when used in tablet value.Figure 13. Cutting with 610 mm/min and 80 AFigure 14. Cutting with 870 mm/min and 100 AUsing a current of 100 A obtained quality of thecut can be described as solid. There are wasteparticles on the bottom, but not on a large extent.An additional treatment section is necessary.Using the current with intensity of 80 A the bestquality is obtained by using the highest speed.Sample of 17 is considered best combination ofparameters because obtained fragment with the bestquality (Fig. 15.). The best quality is obtainedincreasing the speed by 20% of tablet speed value.Table 2. Measured values of Ra, kerf width and moltenmetal on the bottom of cutSamplesMeasur.Sb h SdRa µm St mmpointsmm mm mm1 g 10.5s 17 2.14 1.2 4.78 7.12d 30.72 g 12.41.13s 21.3 2.18 5.11 7.18d 40.43 work piece is not penetrated4gsd17.529.426.62.55 1.61 1.05 5.63567gsdgsdgsd15.235.823.014.124.324.71441552.45 1.38 0.95 4.982.30 1.18 2.23 6.642,01 1,11 4,45 7,388 work piece is not penetrated91011121314151617gsdgsdgsdgsdgsdgsdgsdgsdgsd35193213.013.09.812.48.813.012.012.410.116141817142010.311.010.89.912.98.310.19.82,51 1,75 3,74 10,322.54 1.54 1.16 5.212.45 1.36 0.93 7.292.36 1.23 0.93 8.402,68 1,82 1,15 7,582,46 1,67 0,80 7,832.64 1.45 0.523.072.55 1.29 0.66 3.242.4 1.03 notexist-318 13 th International Conference on Tribology – Serbiatrib’13

Lower values of current are not resulted withenough heat in the cutting zone, and therefore thequality of these clips are worse. With lowercurrents and much higher speed than recommendedmay happend that clip does not penetrate, like insamples 3 and 8.Based on of experimental results can concludethat an increase in speed reduces kerf width.Figure 15. Cutting with 1585 mm/min and 120 A5. CONCLUSIONSPlasma cutting is nonconventional technologythat represents the best relation between cost andquality value for money for most of the standardports and small series production types. In addition,the processing speed is far greater than thetechnology of machining, and quality is comparableto the laser cutting technology.Plasma cutting process may be used to cut anyconductive material, including carbon steel,stainless steel, aluminum, copper, brass, cast metalsand exotic alloys.Obtained experimental results are consistentwith theoretical considerations, as well as previousexperimental results.The best quality is obtainedincreasing the speed by 20% of tablet speed value,which indicates that in this area have a place forfurther research and improvements.ACKNOWLEDGEMENTThis paper is part of project TR35034 - Theresearch of modern non-conventional technologiesapplication in manufacturing companies with theaim of increase efficiency of use, product quality,reduce of costs and save energy and materials,funded by the Ministry of Education and Science ofRepublic of Serbia.REFERENCES1. M. Radovanovic, M. Madic: Modeling the plasmaarc cutting process using ANN, NonconventionalTechnologies Review – No.4/20112. P. Ferreira, I. Melo, A.Gonçalves-Coelho,A. Mourão: Plasma cutting optimization by usingthe response surface methodology, The annals of“DUNĂREA DE JOS” university of Galati fascicleV, technologies in machine building, ISSN 1221-4566, 20093. S.M. Ilii, L. Apetrei, I. Carp: Considerations concerningplasma arc cutting machining, InternationalConference on Manufacturing Engineering(ICMEN), Chalkidiki, Greece, 1-3 October 20084. International standard ISO 9013:2002(E)5. Plasma working group at Linde AG, Linde GasDepartment, and specialists from the companyKjellberg Finsterwalde Elektroden und MaschinenGmbH, Facts about plasma technology andplasma cutting Riverside Corporate Park10 JuliusAvenueNorth Ryde, NSW 2113Australia, 20116. E. Gariboldi, B. Previtali: High tolerance plasmaarc cutting of commercially pure titanium, Journalof Materials Processing Technology 160, pp. 77–89,2005.7. S.M. Ilii, M. CoteaŃă, A. Munteanu: Experimentalresults concerning the variation of surfaceroughness parameter (Ra) at plasma arc cutting of astainless steel workpiece, International Journal ofModern Manufacturing Technologies II, ISSN2067–3604, 20108. R. Bhuvenesh, M.H. Norizaman, M.S. AbdulManan: Surface roughness and mrr effect on manualplasma arc cutting machining, World academy ofscience, engineering and technology 62, 20129. M. Hatala, R. Čep, Z. Pandilov: Analysis of surfaceroughness and surface heat affected zone of steelEN S355J0 after plasma arc cutting, MechanicalEngineering–Scientific Journal 29, pp. 1–6, 2010.10. A.P. Hoult, I.R. Pashby, K. Chan: Fine plasma cuttingof advanced aerospace materials, Journal ofMaterials Processing Technology 48, pp. 825-831,1995.11. J.A. Hogan, J.B. Lewis, A. Jordan: Plasmaprocesses of cutting and welding, Naval SurfaceWarfare Center CD Code 2230- Design IntegrationTools, Building 192, Room 138 9500 MacArthurBlvd Bethesda, MD 20817-5000, 1976.12. D. Cook: Plasma arc cutting-cut quality problems,Welding design and fabrication, November 199813. J. Berglund: Cut cost calculation, Master of scienceprogramme industrial economic, Lulea university oftechnology, 2006.14. B.R.Hendricks: Simulation of plasma arc cutting,Peninsula Technikon, faculty of Engineering, CapeTown, 1999.15. J.V. Warren: The Contribution of arc voltagecontrol to quality plasma cutting, ESAB-30213,South Carolina, USA, 2005.13 th International Conference on Tribology – Serbiatrib’13 319

Serbian TribologySocietySERBIATRIB ‘1313 th International Conference onTribologyKragujevac, Serbia, 15 – 17 May 2013Faculty of Engineeringin KragujevacTRIBOLOGICAL ASPECTS OF SINTERED STEEL GEAR INAPPLICATION WORM-AND-GEAR SETAleksandar Miltenović 1 , Milan Banić 1 , Miroslav Mijajlović 1 , Đorđe Miltenović 21 University of Niš, Faculty of Mechanical Engineering in Niš, Serbia,aleksandar.miltenovic@masfak.ni.ac.rs, milan.banic@outlook.com, mijajlom@masfak.ni.ac.rs2 College of Textile, Leskovac, Serbia, milten2004@yahoo.comAbstract: Due to the low manufacturing costs, worm-and-gear set with the combination of a steelworm and a gear are used almost exclusively in automotive auxiliary drive units such as windowlifters, seat adjustments and windscreen wipers. Worm-and-gear sets are a simple and compact wayto achieve a high speed gear ratio. Tribological aspect of worm-and-gear set is very complex whilethere can occur different damage forms such as: wear, pitting or scuffing. The conditions, underwhich a damage form occurs, are not fully elucidated. In this paper are shown experiments thathave been carried out with gear made of sintered steel Fe1.5Cr0.2Mo with different treatmentmethods.Keywords: sintered steel, gears, worm-and-gear set, wear, pitting, scuffing1. INTRODUCTIONCrossed helical gears are used, for example, inautomotive auxiliary drive units such as windowlifters, seat adjustments, windscreen wipers, andalso in home appliances. The trend towardsincreased comfort in motor vehicles has led to theutilization of more than a hundred servo-drives inluxury class automobiles. Important advantages ofthe crossed helical gears are their easy andinexpensive design, good noise performance andhigh ratio that can be realized in one step.The use of gear wheels made of sintered metalcan increase the load capacity of crossed helicalgears. As in the case of plastic gear wheels, thelarge scale production of sintered metal gear wheelsrequires a special tool and no additional postproductioncosts.Hochmann [1] determines the load capacity ofmaterial pair steel/steel with grease. The testedgears with grease have lower load capacity thantested gears with oil. Crossed helical gears havedifferent transmission conditions, therefore, theseresults cannot be used.The lubricating film thickness, pitting andscuffing load capacity correlate significantly withthe base oil viscosity of grease. The addition of aspecial synthetic graphite as solid lubricant showsincreased wear in this research. The base oilviscosity is a decisive factor for calculating thelubricating film thickness of grease, as well as forcalculating the pitting load capacity according toDIN 3990 [2]. Performance data of tested lubricantsis available for calculating scuffing and pitting loadcapacity according to DIN 3990. This data takesinto account the influence of grease.2. CHEMICAL COMPOSITIONThe material combination steel/sintered metalhas been investigated only in few research projects.Researchers from the company Höganäs AB,Sweden [3] investigated sintered metal Astaloy Mo(Fe0.85Mo) and Astaloy CrL (Fe1.5Cr0.2Mo).320 13 th International Conference on Tribology – Serbiatrib’13

Table 1. Chemical composition of sintered steel Fe1.5Cr0.2Mo (%)element measure point 1 measure point 2 measure point 3 measure point 4 measure point 5C 0,254 0,243 0,238 0,253 0,273Si 0,054 0,047 0,043 0,057 0,053Mn 0,163 0,161 0,161 0,161 0,161P 0,009 0,009 0,009 0,009 0,009S 0,001 0,001 0,001 0,001 0,002Cr 1,521 1,517 1,526 1,51 1,509Ni 0,026 0,026 0,026 0,026 0,026Mo 0,21 0,208 0,209 0,211 0,21Cu 0,066 0,066 0,067 0,067 0,067Al

Table 3: Data of the test gear pair [6]ParametersDataCentre distance30 mmModule1.252 mmTransmission ratio 40Pressure angle 20Wheel materialFe1.5Cr0.2MoWorm material16MnCr5Speed 1500 – 10000 min -1Torque12-36 NmSynthetic oil Klüber GH6 1500Mineral oil Optigear BM 1500Grease: Klübersynth G34-130Figure 2. Viscosity-temperature behaviour for usedlubricants4.2 Mineral oilCastrol Optigear BM 1500 was used in theexperiment as the mineral oil with the sameviscosity as Klübersynth GH6 1500. Highperformance gear oil Castrol Optigear BM containsthe mineral oil-based additive packageMICROFLUX Trans (MFT). This combination ofload-active agents and additives adjusts to varyingloads and actively prevents wear. Difficulties in therun-in phase can be reduced and problems withwear, material fatigue on surfaces (pitting) andmicropitting can be avoided.Table 4: Data of lubricants4. LUBRICATIONFigure 1. Test benchFor the sake of comparison, lubrications testswere done under same parameters and withdifferent lubricants: synthetic oil, mineral oil andgrease.4.1. Synthethic oilSynthetic oil made by Klüber company(Klübersynth GH6 15000) was used in tests. Thisoil can withstand high scuffing load capacity andhas good wear protection. Furthermore, the oilreduces friction and has a flat viscosity-temperaturebehaviour. This oil is based on polyglycol and it ismainly used in transmissions with the materialcombination of steel/steel.The additive GH6 reduces the frictioncoefficient and wear, especially for worm gearswith the material combination steel/bronze. Verylow wear intensities for worm gears can beachieved with this oil. Klübersynth GH6 oil with aviscosity of 40 = 1500 mm 2 /s was used for testing.Table 4 shows the characteristics of used lubricants.Parameter GH6 1500 BM 1500 G34-130Temperature Range [°C] -20 to 160 -10 to 90 -30 to 130Density (20 o C) [g/cm 3 ] 1.08 0.93 0.87Kin. base oil viscos. DIN51562 ν 40 [mm 2 /s]1500 1507 150Kin. base oil viscos. DIN51562 ν 100 [mm 2 /s]231 75.6 16Consistency class DIN 51818, NLGI (grease) 04.3 GreaseKlübersynth G34-130 grease for small gears wasused in the grease tests. Klübersynth G34-130 isgrease based on synthetic hydrocarbon oil. Otheringredients are mineral oil and lithium soap asspecial polyureas, which serve as consistencyfactors. The grease has good anti-wear propertiesand can be used with the combination of steel andplastic.Figure 2 presents the dependence of kinematicviscosity on temperature for the selected lubricants.The oils Klüber GH6 1500 and Castrol OptigearBM 1500 have the same viscosity at thetemperature of 40 o C. At higher temperatures, theviscosity of mineral oil Castrol Optigear BM 1500is smaller than the synthetic oil Klüber GH6 1500.Lubricants dependence on the viscosity is veryimportant at low temperatures. They should still be322 13 th International Conference on Tribology – Serbiatrib’13

flowable at high temperatures and very viscous atextremely high temperatures. The best oil in thissense is the Klüber GH6 1500 oil.5. HERTZIAN CONTACT STRESSThe Hertzian contact stress has a significantinfluence on the wear rate and the width of the wearsurface. The Hertzian contact stress in the crossedhelical gears can be determined as the Hertziancontact stress of globoid wheel. It should be takeninto consideration that the Hertzian contact stresson the crossed helical gear depends on the width ofthe wear surface. Therefore, a correlation betweenthe width of the wear surface b V and thedimensionless ratio p m * [5] is introduced. TheHertzian contact stress by the dimensionlessparameter of the average Hertzian contact stressp m,V * is taken into account. The new Hertziancontact stress Hm can be calculate for the averageHertzian contact stress p m,V * according to Equation1, depending on the output torque T 2 , the E-ModuleE red and centre distance a s .4 p EredHm (1)*m, VT210003asAccording to the tests for sintered steelFe1.5Cr0.2Mo with sintered-hardening (S5), thevalue of E-Module is E 2 = 203759 N/mm 2 .Therefore, for the material combination of wormmade of 16MnCr5 and wheel made of sintered steelFe1.5Cr0.2Mo, the value of the reduced E-Moduleis E red = 227288 N/mm 2 .The Hertzian contact stress that depends on thewidth of the teeth p m,V * can be calculated usingEquation 3. The ratio of the width of the globoidwheel b 2H to the width of wear surface b v takes intoconsideration the increase of the Hertzian contactstress with decreasing width [6].0,8614 b 2Hp m, V pm(2) bvFigure 3 shows the resulting Hertzian contactstress of all tests with different lubricants andn 1 = 1500; 5000 und 10000 min -1 . In the first loadstep output torque is 12 Nm, each next step outputtorque T 2 is increased by 4 Nm, and the timeduration of load level is set at 40 hours.Higher torques leads to the increase in pressure.There is no great difference in pressure values ofHertzian contact stress for lubrication with mineraland synthetic oil for the rotation speed n 1 = 1500min -1 or sliding velocity v gs = 0.76 m/s. For greaselubrication, the Hertzian contact stress is onaverage half the size in relation to mineral andsynthetic lubricating oil.There is great difference in pressure values ofHertzian contact stress for lubrication with mineraland synthetic oil for the rotation speedn 1 = 5000 min -1 or sliding velocity v gs = 2.53 m/s.The values of Hertzian contact stress with syntheticoil are for about 65% higher compared tolubrication with mineral oil. Hertzian contactstresses for lubrication with synthetic oil goes up to1400 N/mm 2 .Herzian contact stress Hm [N/mm 2 ]Herzian contact stress Hm [N/mm 2 ]Herzian contact stress Hm [N/mm 2 ]160014001200100080060040020001000900800700600500400300200100016001400120010008006004002000n 1 = 1500 min ‐18 12 16 20 24 28 32 36 40n 1 = 10000 min ‐1Output torque T 2[Nm]synthetic oilmineral oilgreasesynthetic oilmineral oil4 8 12 16 20 24Output torque T 2 [Nm]n 1 = 5000 min ‐18 12 16 20 24 28 32 36Output torque T 2 [N]synthetic oilmineral oilFigure 3. Hertzian contact stress Hm for duration of theexperiment with different lubrication and n 1 = 1500;5000 and 10000 min -1For the rotation speed n 1 = 10000 min -1 orsliding velocity v gs = 5.05 m/s up to the number ofload changes N L = 0,9×10 6 values of Hertziancontact stress are greater for mineral oil lubricationas compared to synthetic lubricating oil. Then it13 th International Conference on Tribology – Serbiatrib’13 323

comes to the sharp increase of wear rate and toreduction of pressure. Lubrication with synthetic oilhas a smaller increase of wear rate and smallerreduction of pressure.Based on the foregoing analysis, it can beconcluded that the Hertzian contact stress dependson the choice of lubricant. Synthetic lubricating oilis the most favourable from the aspect of wear, andin terms of achieving hydrodynamic lubrication.Optimal lubrication conditions were obtained forthe rotation speed of n 1 = 5000 min -1 or slidingvelocity v gs = 2.53 m/s.6. DAMAGE TYPES6.1. WearFigure 4. Wear on wheel tooth surface withoutadditional treatment for T 2 = 36 Nm; t = 260 h;n 1 = 5000 min -1tribological system are: the gear wheel (basicbody), the worm (opposed body) and the lubricant(intermediate component).Experiments with wheels with differentadditional treatments provide basic knowledge ofsintered gears load capacity. Worm and wheel arein contact in a point. During operation, a change inthe tooth flank of the wheel appears due to wear.The worm forms on the tooth flank of the wheel, awear surface that has a shape that is identical toworm gear flank. Wear progress widens the wearsurface, which leads to a lower Hertzian pressure inthe tooth contact. After a certain period of operationunder intensive wear progress, the steady stateoccurs, where a necessary oil layer exists, so thatthe wear progress is minimal. Figure 4 shows theform of wear damage on tooth surface of wheelmade from material without additional treatment.Figure 5 compares all experiments with wheelsof different material variants after a trial of 100 hand an output torque of 20 Nm. The maximumwear, δ wn = 115 m, occurred on material S2 –material with case hardening. The minimum wear,δ wn = 7.8 m, occurred on S5 – sinter-hardening.Figure 4 shows the wear width of the wear surfaceon wheel from material S4 - “pyrohydrolysis” andS5 sinter-hardening for different speeds. TheFigure 5. Wear δ wn for all trials with different material variants [6]The wear describes the continuous loss ofmaterial from the surface of the basic body whichhas a relative movement with respect to a solid,liquid or gaseous mating with which it is in contact[7]. Wear has exclusively mechanical causes.Different from hardness or tensile strength, wear isnot a specific material property but a systemproperty which depends on the particulartribological system. In our case, the elements of thesmallest wear width occurred at input speedn 1 = 5000 min -1 . The reason for this is that the bestexperimental conditions, with regard to lubricationand wear, are at this input speed.6.2. WearA large pressure on surface does not lead to asudden failure of drive, but over the time, small324 13 th International Conference on Tribology – Serbiatrib’13

holes (pits) emerge in the shape of shell on toothflank. Pit peak always points in the slidingdirection. This damage occurs through a cyclicfatigue due to repeated elastic and plasticdeformations of the surface. The holes occur onlyafter a sufficiently large number of overrollings(from ca. 5x10 4 load cycles). If only initial pitting ispresent, the situation is not dangerous. Destructivepitting destroys the flank and causes failure due tonoise and fatigue. The pitting occurred on wheelsmade from materials S4, S5 and S6 by an outputtorque of 16 and 32 Nm after the trial time periodof 120 h to 240 h.Figure 6 shows initial and destructive pitting ontooth flank of wheel. In trials with material S1 –without additional treatment, initial pitting ocurredunder input speed n 1 = 5000 min -1 and outputtorque of 20 Nm. In trials with material S5 – sinterhardening,destructive pitting occurred under inputspeed n 1 = 5000 min -1 and output torque of 20 Nmand mineral oil.6.3. ScuffingThere is a difference between cold and warmscuffing. Both damage types are caused by the lackof lubricant in contact between teeth. Cold scuffingis relatively rarely seen. It occurs mainly at lowspeed (< 4 m/s) and between teeth that are havingrelatively high hardness and rough quality ofcontact surfaces. Warm scuffing occurs due to greatpressure and high sliding velocity between toothflanks. Under such a load, combined effects occurwhich lead to the increase in temperature thatdisrupts the lubricant film between tooth flanks,making the contact between tooth flanks direct anddry. This can cause a short local welding of theflanks which damages both flanks. Warm scuffingis characterized by strip-shaped bands in thedirection of the tooth height, and with the strongestexpression in the tooth addendum and tooth root.Scuffing on high-speed gears increases thetemperature and tooth forces, eventually leading toshaft fracture due to high damage on tooth flanks.Initial pitting: without additional treatmentT 2 = 20 Nm; t = 120 h; n 1 = 5000 min -1Lubricant: synthetic oilVariant with case hardeningT 2 = 20 Nm; t = 160 h; n 1 = 5000 min -1Lubricant: synthetic oilDestructive pitting: sinter-hardening variantT 2 = 20 Nm; t = 120 h; n 1 = 5000 min -1Lubricant: mineral oilFigure 6. Pitting on wheel tooth surface [6]Material with 2% copper additionT 2 = 28 Nm; t = 160 h n 1 = 5000 min -1Lubricant: synthetic oilFigure 7. Scuffing on tooth flanks of wheel for differentmaterial variants [6]13 th International Conference on Tribology – Serbiatrib’13 325

Figure 7 shows the scuffing on wheel toothflanks for different material variants. Withexception of S1, scuffing occurred on all materialvariants. On wheels made from materials S2 andS3, under load of T 2 = 16 Nm and T 2 = 20 Nm, thephenomenon of scuffing and significant increase ofwear surface and gear forces were observed. Underoutput torque of T 2 = 28 Nm, and input speedn 1 = 5000 min -1 scuffing occurred for S4 and S6.Gear loads rose and, suddenly, the failure of wormshaft occurred. On material variant S6 “2% copperaddition”, scuffing occurred on tooth root in trialswith input speed n 1 = 5000 min -1 and output torqueof T 2 = 28 Nm, and a very short running time of ca.10 minutes. Scuffing also occurred on wheel toothflanks with speed n 1 = 5000 min -1 and output torqueof T 2 = 36 Nm for material variant S5 sinterhardening.7. MAXIMUM OUTPUT TORQUEFigure 8. Comparison of maximum transmissible torquewith critical damage type for different materials andinput speed n 1 = 5000 min -1 [8]Figure 8 shows maximum transmissible torque,as well as the type of critical damage, in a bar chart.With an output torque of T 2 = 20 Nm, materials S2(case hardening) and S3 (case hardening and shotpeening) were damaged due to scuffing. MaterialsS4 (pyrohydrolysis) and S6 (2 % copper addition)were damaged by pitting when output torque wasT 2 = 24 Nm, and by scuffing at the value ofT 2 = 28 Nm. Without an additional treatment, S1had the most critical wear and some pitting atoutput torque of T 2 = 28 Nm. The wheels S5(sinter-hardening) had the greatest load carryingcapacity, its maximum transmissible torque being32-36 Nm, and the scuffing being the mostdangerous damage form in this case.8. CONCLUSIONSThe research in this paper shows thattribological parameter of worm-and-gear has greatinfluence on working characteristics in exploitationconditions. The material characteristics likemicrostructure, wear load capacity and damagetypes of molded parts can be significantlyinfluenced by additional treatments for sinteredsteel.The Hertzian contact stress of teeth in contactdepends on the choice of lubricants. The smallestwear occurred in experiments with synthetic oilwhen Hertzian contact stresses were largest.Synthetic oil can resist the pressures of 1400N/mm 2 at sliding velocity v gs = 2.53 m/s.All experiments with different material variantsshow that additional treatments have significantinfluence on wear. Under identical experimentalconditions, the maximum wear δ wn occurred in thetrials with wheels of material variant with casehardening (115 µm) and the minimum with wheelswith sinter-hardening (7.8 µm).The pitting was observed in wheels of materialvariants S4, S5 and S6 by output torque from 16 to32 Nm. The initial pitting on the tooth flankoccurred on material variant without additionaltreatment (trials with T 2 = 20 Nm; t = 120 h; n 1 =5000 min -1 ; lubricant: synthetic oil). Destructivepitting on tooth flank occurred in material variantsinter-hardening (T 2 = 20Nm; t = 120h; n 1 = 5000min -1 ; lubricant: mineral oil).With exception of S1, scuffing occurred on allmaterial variants. The scuffing was the critical typeof damage for wheels from material variants S2 andS3 under the output torque of T 2 = 20 Nm withn 1 = 5000 min -1 and by S5 under the output torqueof T 2 = 36 Nm. For wheels of material variant S6,scuffing occurred for trials with n 1 = 5000 min -1and output torque of T 2 = 20 Nm. For wheels ofmaterial variants S4 and S6, the critical type ofdamage was the combination of pitting andscuffing.Lubrication with synthetic oil is the mostfavorable from the aspect of wear and in terms ofachieving hydrodynamic lubrication. Optimalworking lubrication was obtained for the rotationspeed n 1 = 5000 min -1 or sliding velocity v gs = 2.53m/s.REFERENCES[1] M. Hochmann: Tragfähigkeit vonZahnradpaarungen bei Schmierung mitGetriebefetten, Forschungsbericht, DGMK, 2007.[2] DIN 3990: Tragfähigkeitsberechnung vonStirnrädern, 12/2002.[3] S. Dizdar, P. Johansson, A.B. Höganäs, 263 83Höganäs, Sweden: P/M Materials for GearApplication; EURO PM2007, Tolouse October 16,2007.326 13 th International Conference on Tribology – Serbiatrib’13

[4] W. Schatt, K-P. Wieters, Pulvermetallurgie-Technologien und Werkstoffe, VDI-VerlagDüsseldorf, 1994.[5] W. Predki, A. Miltenović, Influence of Hardeningon the Microstructure and the Wear Capacity ofGears Made of Fe1.5Cr0.2Mo Sintered Steel.International Journal “Science of Sintering” 42,pp. 183-191, (2010).[6] Miltenović, A.: Verschleißtragfähigkeitsberechnungvon Schraubradgetrieben mit Schraubrädern ausSintermetall, Dissertation Ruhr-Universität Bochum,2011.[7] DIN 50320: Verschleiß: Begriffe – Systemanalysevon Verschleißvorgängen – Gliederung desVerschleißgebietes, 12/1979.[8] Miltenović,A., Predki,W.: Damage Types ofCrossed Helical Gears with Wheels from SinteredSteel. International Journal “Science of Sintering”43, pp. 205-214, 2011.[9] Wendt, T.: Tragfähigkeit von Schraubradgetriebenmit Schraubrädern aus Sintermetall, DissertationRuhr-Universität Bochum, 2008.13 th International Conference on Tribology – Serbiatrib’13 327

Tribometry13 th International Conference on Tribology – SERBIATRIB ’1315 – 17 May 2013, Kragujevac, Serbia

Serbian TribologySocietySERBIATRIB ‘1313 th International Conference onTribologyKragujevac, Serbia, 15 – 17 May 2013Faculty of Engineeringin KragujevacPRELIMINARY STUDY ON THE SEIZURE TREND OFA MOM-THP WITH SELF-DIRECTED BALLSLucian Capitanu 1 , Liliana – Laura Badita 2 , Virgil Florescu 2 , Dumitru Catalin Bursuc 31 Institute of Solid Mechanics of the Romanian Academy, 15 Constantin Mille, 010141, Bucarest, Romanialuciancapitanu@yahoo.com2* National Institute for Research and Development in Mechatronics and Measurement Technique, 6-8 Sos Pantelimon,district 2, Bucharest, Romania badita_l@yahoo.com3 Mechanical Departament, Institute of Civil Engineering, 59 Plevnei Way, 050141, Bucharest, Romania,florescuvirgil@yahoo.com4 ”Carol I” National Defance Academy, Bucharest, Romania, catalin258@yahoo.comAbstract: This work continues the approach of one of our topics relating to a MOM THP with self-directedmovement balls. Experiments revealed a certain seizure in some strain conditions. Laboratory trials forballs/plane Hertzian contacts have been restarted in order to determine seizure behaviour depending on theroughness of the flat area. The trials have been carried out in BSF (body simulated fluid) lubricationconditions, much closer to the real operating conditions up against the initial tests with distilled water.Seizure burdens to different loadings and contact surfaces roughness influence over the seizure burden havebeen determined. Even though the minimum value of the wear must be the same with the minimum value ofthe surfaces roughness, given the experimental conditions, it came out from the trials results on wear that thelowest level of wear is acquired at a certain value of roughness, not at the lowest level of roughness.Keywords: MOM –THP with balls, self-directed movement, seizure, optimal conditions, wear scar, frictioncoefficient.1. INTRODUCTIONNowadays, the design solutions for Total HipProstheses are diverse encompassing for improvingthe materials used for prostheses elements andreshaping geometrically and/or tribologically theload transfer path. In such context Total HipProstheses with rolling balls have been found as apossible viable alternative design to currentindustrial products, based on low friction of rollingcontact, against sliding one (now used in mostindustrial designs).Different designs of Total Hip Prostheses withDifferent designs of Total Hip Prostheses withrolling bodies have been developed in order toimprove the tribological performances of theartificial joint. We could mention here the designwith ball train, proposed by Katsutoshi and Kiyoshi[1], the French “Supertête” prosthesis [2], or thedesign with conical rolling elements proposed byImperial College of Science, Technology andMedicine of London [3].The French design, obtained by “Fondation del’Avenir” in collaboration with “Ministère de laDéfense, Mission Innovation”, propose theinsertion of a frictional contact inside a bearing.The design suggested by Imperial College ofScience, Technology and Medicine of Londonconsists in a major modification of a elementsbetween the femoral part stem neck and themodular hip prosthesis by introducing a rollingbearing with conical femoral artificial head.The bearing rotation axis corresponds with theaxis of femoral stem neck, the rolling elementsbeing guided by both the external surface of stemneck and the internal surface of the ball replacingthe femoral head. But changing the contactmechanism from sliding to rolling in a hipprosthesis is not an easy task due to difficultiesencountered in establishing the load transfer path, acritical characteristic of tribological behavior ofjoint with large influence in functionality anddurability of prosthesis active elements.13 th International Conference on Tribology – Serbiatrib’13 331

Basically, the sliding contact between largesurfaces of femoral head and acetabular cup wasreplaced by a multitude of rolling contacts with adifferent pattern of stress distribution influenced byrolling elements position at some instant duringrelative movement between femoral and acetabularparts.In the present paper, the authors focus on theoriginal design proposed by them in [4], i.e. aMOM Total Hip Prosthesis with self directedrolling balls (see Fig. 1).Figure 1. Total Hip Prosthesis with self directed rollingbodies.A characteristic of this design solution is the factthat the artificial joint will work similar to aspherical bearing, having what is called a“compensation space”, i.e. enough free spacebetween the femoral and acetabular parts of theprosthesis to allow the movement of the balls [5].Previous research studies performed by theauthors focused on the determination of the initialposition of rolling balls due to geometricalrestraints of the assembly and on the estimation ofthe overall friction coefficients in dry andlubricated motion [5].The geometrical studies (see [6] and [7]) haveshown that generally the balls are located nonsymmetricallyand that the configuration for a givenspace and a given number of balls is not unique.Tribological studies performed by he authors (see[4], [6] and [7]) have shown very low values ofoverall friction coefficients (0.12 to 0.2 for dry jointand 0.006 to 0.009 in the presence of lubricants),leading to an enhanced functionality of theprosthesis itself. The present study will use theresults of the previous studies in order to determinethe load distribution through the balls bed and thecompressions generated between the joint elements(femoral head, rolling balls, acetabular cup).As we previously stated, a general study of theproposed design will target the followingmechanical aspects:– characterization of load transfer mechanismthrough the joint elements (statics of envelopingloads and/or dynamics studies of natural,physiological movements);– evaluation of tribological behavior of all jointelements (including contact mechanics of all activeinterfaces – femoral head-ball, ball-ball, ballacetabularcup);– estimation of functional threats and damagingmechanisms for the proposed design (i.e. cleardefinition of criteria for joint locking, fatigue ofprosthetic parts, wear of active elements of thejoint) and determination of influencing factors forall these unwanted phenomena.Lessons learned from previous attempts(structural overall analysis performed in [8]) lead todecoupling the statics and dynamics of the joint (FEanalyses) from the tribological behavior (separateanalytical evaluation) in order to savecomputational effort and assuming simplifications.The characteristics of the prosthesis underevaluation are as follows:- type: Total Hip Prosthesis (THP) with selfdirected rolling balls;- geometrical features: outer radius of femoralhead - 14 mm; radius of each rolling ball 1.25 mm;internal radius of acetabular cup - 16.5 mm;spherical cap for balls bed (subtended angle) -160°;Material used for components:- femoral head - Stellite 21;- acetabular cup - Ti6Al4V;- rolling balls - CoCrMo alloy.The methodology used for computing thenumber of balls needed for assembling the joint andtheir positions inside the artificial joint is that usedin [4]. The resulted configuration of artificial jointwas used in order to build the numerical model forload transfer path through the balls bed.The 3D numerical model is a large one – 58,784elements and 73,100 nodes, with numerous surfacesin contact, requiring high computational resourcesand significant time for simulation. Instead of usinga big model with multiple non-linearities, asimplified model was built based on the followingassumed hypotheses1. Femoral head and balls have been consideredrigid (their stiffness is much higher than acetabularcup stiffness).2. The compressive force and flexion drivemoment have been maintained constant.3. Linear elastic behavior of acetabular cup wasassumed.4. The compressive forces acting at the ball-toballcontact surfaces are smaller than thecompressive forces between balls and femoral head,respectively between balls and acetabular cup. This332 13 th International Conference on Tribology – Serbiatrib’13

assumption allows us to use, instead of sphericalballs, unidimensional nonlinear elements(compression only) connecting the spots of contactsbetween the balls and cup, respectively between theball and femoral head with the center of each ball.The train of balls was not actually modeled as itis; instead of 3D representation of the balls (Fig. 5),unidimensional contact elements have beenconsidered between the center of each ball and theactive surfaces of femoral and acetabular prostheticelements.2. SIMULATION METHODOLOGY ANDRESULTSThe 3D FE model was loaded by a compressive1 kN force and a flexion of the joint was consideredfor ~37.6° (i.e. a relative maximum displacement ofcircumferential points located on femoral head andacetabular cup equal with 4 times the diameter ofone rolling ball).After the loads on each ball have beendetermined (being categorized based on ballsregions rather than each ball itself) a local analysiswas performed for establishing the extremeHertzian contact parameters based on the followingmethodology [9]:- maximum pressure, given by*26PEp 30 ,(1)3 2 R- radius of contact spot, given by*26PEp 30 ,(2)3 2 R- mutual approach between bodies in contact,given by29P 3 ,(3)*216REwhere P is the applied compressing load, and R therelative curvature given by:1 1 1 (4)R R1R2After performing the geometrical assessment,based on the methodology presented in [4], itresults that the maximum number of balls neededfor the spherical joint is 199, distributed on 12consecutive rows [9] as follows:n 0 = 37; n 1 = 19; n 2 = 19; n 3 = 19; n 4 =19 ; n 5 = 19;n 6 = 19; n 7 = 19; n 8 = 14; n 9 = 9; n 10 =5 ; n 11 = 1.Images of the rolling balls positions for ϕ = 0and β = 0° ±15° (where ϕ and β are the azimuthand zenith angular coordinates in the sphericalcoordinate system associated with the femoralhead). One could notice from the results of themathematical analysis that the arrangement of theballs in the rolling space is asymmetrical and willnot be uniquely determined.After applying 1 kN compressive load onto theartificial joint having the balls train configured asresulted from the geometrical analysis [8], theloadings on each rolling ball during the 37.6°flexion were determined by FE analysis of adynamic nonlinear model of the entire joint.Several three instances have been selected forpresenting the results in both vertical a nd normalviews to the flexion plane in Figs. 2.0 13.68618.8 13.52927.6 19.800Figure 2. The compressive loadings transferred fromfemoral to acetabular parts of the prosthesis for differentinstances of flexion (between 0° and 37.6°); flexionlisted in left, maximum load listed in right part of thepictures).Its correspond to 1 - diameter, 2 - diameter, 3 -diameter and 4 - diameter relative displacementsbetween the acetabular and femoral parts of theprosthesis.By analyzing the plots, the followingconclusions could be drawn:a. Even for the initial condition, due toasymmetrical arrangement of the balls resultedfrom the geometrical analysis, there is someasymmetry of transferring the load path from thefemoral head to the acetabular part [8].b. During the flexion (especially for large angles)a part of the balls will not be loaded anymore,leading to an increase of the maximum forcetransmitted by intermediate of a rolling ball (from~1.35% of the total joint compression force - as forflexion angles lower than 18.8°, to ~1.98% of thetotal joint compression force - as for a flexion of37.6°) - Fig.2.c. By analyzing the loading of each ball row, ithas been determined (for the initial position, 0°13 th International Conference on Tribology – Serbiatrib’13 333

flexion) that the most loaded rows are those locatedclose to 40°…60° from the equatorial plane (therows located lower have small loads on each balls,and for the rows located higher each ball carries abigger load but the number of balls is low). Thedistribution of rolling balls loading versus thezenith positioning angle of the rolling ball ispresented in Fig. 3.Analyzing the graph, the following conclusionscould be drawn:1. As reported before, there is a slight asymmetryof the distribution even for the initial position. Thisasymmetry evolves with flexion leading tounloading of some balls located peripherallyoutside the hemispherical area characterized bycompressive loading pole.2. The peripheral balls located closer to thecompressive loading pole are generally highlyloaded, but the highest loaded balls remain thosepositioned in intermediate rows (between 40° and60° from the equatorial plane).Seizure burdens to different loadings andcontact surfaces roughness influence over theseizure burden have been determined.Even though the minimum value of the wearmust be the same with the minimum value of thesurfaces roughness, given the experimentalconditions, it came out from the trials results onwear that the lowest level of wear is acquired at acertain value of roughness, not at the lowest levelof roughness.For the tests we used the test rig presented inFig. 4. In this test rig, the friction pair is formedfrom a bush with spherical profile and a flat diskshaped with a diameter of 18 mm and a thickness of5 mm (see Fig. 5). The sphere's radius is r = 11.5mm.Figure 4. Experimental test rigFigure 3. The rolling balls loadings versus zenithpositioning angle of ball.For the extreme maximum loadings of the ballsduring the analyzed flexion, a preliminaryevaluation of tribological parameters of contactbetween femoral head and rolling balls and betweenrolling balls and acetabular cup has been performedby using formulae (1) - (3).3. LABORATORY TRIALSExperiments revealed a certain seizure in somestrain conditions. Laboratory trials for balls/planeHertzian contacts have been restarted in order todetermine seizure behaviour depending on theroughness of the flat area. The trials have beencarried out in BSF (body simulated fluid)lubrication conditions, much closer to the realoperating conditions up against the initial tests withdistilled water.Figure 5. Used friction coupleIn the static contact, compression stresses incontact spot, p max si p med (maximum pressure andmean pressure) are:3pmax 1.5 PE 2 / π r 2 (1 – μ 2 ) 2 (5)Pp med 2a(6)and the radius of the contact surface, a, is:a32 1.5 1PrE(7)334 13 th International Conference on Tribology – Serbiatrib’13

where P is the load, a – radius of the contactsurface, and r – radius of the sphere.In the friction couple components (see Fig. 9)are made of steel, the quantities of equations (5),(6) and (7) become:p max ≈ 5800 3 P (8)p med ≈ 1700 3 P (9)A1A2a ≈ 0.09 3 P (10)Attention was paid to processing of workingsurfaces of couples. Surface state defined bytopography, microstructure of surface layer andoxidation state has a major influence on the wearprocess.Due to the complexity of processing by abrasionof the surface, the most reliable way to ensure areproductible surface is stringently observance ofall processing phases, which are turning of theform, finishing turning, thermal treatment, andcorrection of profile by polishing andsuperpolishing of working surface.All the operations until smooth processing of theprofile are made by current technology, noting thatthe intensity of the process is kept low to protectthe structure of the surface layer of material.Super-finishing operation using metallographicground slides technique includes:- wet polishing with sandpaper, grain size 32 μmand 17 μm;- polishing with diamont slury, grain size 6 μmand 1 μm;- wet polishing with slury of 2000 Ǻ.Finally, the surfaces are washed with distiledwater, alcohol and then are dried.The maintenance of processed couples is madein closed vessels, on silica gel. Surfaces roughnesswas measured with a roughness tester withparametric transducer and recording. Theinstrument allow recording of surface profile, andalso determination of R a and r.m.s., defined as:Rar.m.s 1ll01ll0ydxy2dx(11)(12)In Fig. 6 (A1-A8) profiles (cross-cut) of thesurfaces used and of microphotographs obtained innormal lighting are presented.A3A5A7Figure 6. Profile and microfotograph of the surface withroughness R a = 0.015 μm (A1-A2), R a = 0.045 μm (A3-A4), R a = 0.075 μm (A5-A6) si R a = 0.19 μm (A7-A8).4. RESULTS AND DISSCUSION4.1. Evolution of the surface state in the wearprocessExperimental determinations were made onthese test conditions:- Load: variabile between P = 20 ÷ 300 N fordetermining of seizure limit. A load of 50 N wasused for wear tests.- Sliding speed: Main speed for determining thewear rate was u = 174 cm/s. To determine theinfluence of speed on the wear, the device allowsachieving speeds:u = 60 cm/s; u = 18 cm/s si u = 3,2 cm/s.- Lubricant: BSF (Body Simulated Fluid) with thedensity 1183 kg/m 3 and vascosity 0.84 Pa s(HyClone, SH30212.03).Using the parameters above mentioned, thecpuple operates in elastohydrodynamic regime.For the minimum thickness of the lubricant film,Archard [8] proposed the relationship:hr00.741 20.0740 u0 Er.84 r P A4A6A8(13)where: h 0 -minimum thickness of lubricant film; r-radius of the sphere; α-pressure coefficient ofvascosity; u-sum velocity; μ 0 -dynamic viscosity at13 th International Conference on Tribology – Serbiatrib’13 335

atmospfericic pressure; E-reduced elasticitymodulus; P-load.Relation (13) reproduces satisfacatory thedependence of h 0 by the main quantities u, μ 0 , r andP. The exact determination of the minimumlubricant film thickness depends on the knowledgeof pressure coefficient α for lubricant used and theaccuracy of the numerical coefficient ofrelationship (13). In working conditions, using anestimated value α, resulted a minimum lubricantfilm thickness h ≈ 0.06 μm.Spatial form of lubricant film in the loaded arearesults from Fig. 7 and 8, which represents twosections of lubricant film, longitudinal and crosssections (in relation to the movement) by thesymetry axes.Figure 7. Variation of lubricant thickness in the x = 0plane, function of load, at speed u = 8 cm/s.▪ 0.7 N; ▼ 1.1 N; 1.5 N; ◊ 3 N; ▪ 4.6 N; ▲ 7.8 N; ○10.8 N.Curves were obtained experimentally, underclose conditions to those used by Dowson [9]. Arenoticed significantly higher values for minimumthickness h 0 , even at speed u = 23 cm/s.To watch in good conditions the wear of fixedsurface, function of couple roughness, thefollowing solution was used: roughness of thecouple was focused on one of surfaces, in particularon themobile one. Fixed surface had always theminimum roughness achievable, meaning aboutR a ≈ 0.015 μm.As it is known, the composed roughness of thecouple, expressed as standard deviations, σ, is:σ 2 = σ 1 2 + σ 22(14)where σ 1 , σ 2 represent the standard deviations of thetwo surfaces.If one of the surfaces has small roughness, i.e.σ 1

determination (total running in time t = 10 min.,(sample 704); R a = 0.045 μm, t = 5 min., with thebush from the previous determination (total runningtime t = 20 min. (sample 705).A27A28A9A10A29A30A11A12A13A14Figure 9. Central transversal profile and image of wearscar (magnification x 68). R a = 0.015 μm, t = 5 min.A17A18A31A32Figure 11. Transversal profile and image of wearimprint for samples with R a = 0.045 μm, t = 5 min.(sample 703); R a = 0.045 μm, t = 5 min., but the bushfrom the previous determination (total running in timet = 10 min. (sample 704); R a = 0.045 μm, t = 5 min., butthe bush from the previous determination ((total runningtime t = 20 min. (sample 705).Wear evolution depending on the time fordifferent roughness is shown in Fig. 12. Byincreasing the running time from 5 min., to 30 min.,the wear does not increase only about 10%.Is noted the rapid reduction of wear ratefunction of time, except the surface with roughnessR a = 0.045 μm. In the first 3 seconds it is producedbetween 25% and 50% of wear at 30 min. Thiswear evolution explained by complying surfaces,which results in changing the lubrication regime.A19A20A21A22A23A24Figure 10. Central transversal profile and image of wearscar (magnification x 68). R a = 0.045 μm, t = 5 minFigure 12. Wear evolution function of time, for differentroughness.13 th International Conference on Tribology – Serbiatrib’13 337

Note that the amount of wear on the plateau iscaused by the initial wear (during the first fewseconds of operaton). This observation allows theuse of wear value at t = 5 min., as representativequantity for the existing operating conditions. Atthis time of operation, scattering of values is lover.In the case of surfaces with R a = 0.045 μm, thewear is so low that during the entire period of timeused, lubrication conditions remain aproximatelyunchanged.Evolution of the wear function of time, forroughness R a = 0.075 μm (samples 826, 827, 829,830 and 832, is shown in Fig. 13.A51A53A52A54A43A44A55A56A45A46A57A56A47A48Figure 13. Evolution of the wear function of time, forroughness R a = 0.075 μm (samples 826, 827, 829, 830 and 832)Different roughness used have caused not only adiference between worn volume value, but also inthe aspect of wear imprint (wear type). In the caseof surfaces with R a = 0.015 μm, R a = 0.075 μm,(t =3 sec.) and R a = 0.19 μm, (t =3 sec.), the wear isof adhesive type (metallic shape with prononcedscratches). In the case of surfaces with R a = 0.045μm, oxidative wear type is prevailling. Whilereducing the wear rate as a result of surfacecompliance, surfaces with R a = 0.075 μm and R a =0.19 μm, also goin in oxidative wear regime.Evolution of the wear function of time, forroughness R a = 0.19 μm (samples 832, 835, 849,836, 837 si 850, is shown in Fig. 14.For roughness R a = 0.015 μm, R a = 0.075 μmand R a = 0.19 μm, the results obtained for thevolume of worn material, are consistent with thestrain of contact, determined by the lubricant filmparameter h min / σ.A59A60Figure 14. Wear evolutionear function of time, forroughness R a = 0.19 μm (samples 835, 849, 836, 837 and850)For srfaces with roughness R a = 0.045 μm, therewere observed a running-in influence. After thefirst 5 minutes of operations, the entire contactsupraface is covered with oxide. The imprintobtained after another 5 minutes, with the samebush, has a particular form, oxidative wear arearestricting the half at outcome ofthe loaded area(samples A52 si A54). It follows that after running,lubrication condition improve.Figure 15 shows the central profile and theimage of wear imprinr (magnification x 68). R a =0.15 μm, t =30 min. Pr 997, (new bush).A61A62Figure 15. Central transversal profile and image of wearimprint. R a = 0.15 μm, t = 30 min., sample 997( new bush).Next, Figure 16 ilustrates the effect of runningon the wear behaviour of the surface with R a = 0.15338 13 th International Conference on Tribology – Serbiatrib’13

μm, compared with Figure 17, wich shows theeffect of running on the wear behaviour of surfacewith R a = 0.045 μm.A63A64Figure 16. The effect of running-in on the wearbehaviour of the surface with R a = 0.015 μm, t = 5 min,Pr 998. The bush from previous determination was used.A65A67A69A66A68A70Figure 17. The effect of running-in on the wearbehaviour of the surface with R a = 0.045 μm, t = 5 min,samplez 999, (new bush) 1000 and 1001 (with the bushfrom previous determination).Different roughness used has caused not only adifference between worn volume values, but also inthe aspect of wear imprint (wear type). In the caseof surfaces with R a = 0.015 μm, R a = 0.075 μm, (t=3 sec.) and R a = 0.19 μm, (t =3 sec.), the wear is ofadhesive type (metallic shape with pronouncedscars). In case of surfaces with R a = 0.045 μm,oxidative wear type is prevailing. While reducingthe wear rate as a result of surface compliance,surfaces with R a = 0.075 μm and R a = 0.19 μm alsogo in oxidative wear regime.In the case of super-finished surfaces, with R a =0.015 μm and R a = 0.045 μm, a favourableinfluence of running in is not observed.For rouhness R a = 0.015 μm, R a = 0.075 μm andR a = 0.19 μm, the results obtained for the volume ofworn material, are consistent with the strain ofcontact, determined by the lubricant film parameterh min / σ.For surfaces with rouhness R a = 0.045 μm, therewas observed a running-in influence. After the first5 minutes of operation, the entire contact surface iscovered with oxide. The imprint obtained afteranother 5 minutes, with the same bush, has aparticular form, oxidative wear area restricting thehalf at outcome of the loaded area (sample 1001).It follows that after running-in, lubricationconditions improve.In the case of super-finished surfaces, afavourable influence of running-in is not observed.4.2. Influence of initial roughness on the wearand friction coefficientSurface wear, a quantity determined by thefraction of area in contact, should be a monotonefunction of film parameter h / σ. The minimumvalue of wear should coincide with the minimumvalue of surface roughness.A large number of determinatons of wear, withfour roughness have been made: R a = 0.015 μm;R a = 0.045 μm; R a = 0.075 μm si R a = 0.19 μm.Mean values of volume of worn material for thefour roughness are:R a = 0.015 μm→V u = 7.0 x10 -5 mm 3 → μ = 0.038;R a = 0.045 μm→V u = 7.0 x10 -6 mm 3 → μ = 0.050;R a = 0.075 μm→V u = 3.7 x10 -5 mm 3 → μ = 0.038;R a = 0.190 μm→V u = 1.0 x10 -3 mm 3 → μ = 0.078.Simultaneously with the surface wear, thefriction coefficient was also measured.In Fig. 18 reprezentative images for three of thefour roughnesses are presented.A75A79A81A76A80A82Figure 18. The effect of running-in on the wearbehaviour of the surface with R a = 0.015 μm, t = 5 min,Pr 998. S-a folosit bucsa de la determinarea precedenta.13 th International Conference on Tribology – Serbiatrib’13 339

Existence of an optimal roughness can beexplained either by an effect on lubricant film or bya change in mecahanical properties of surface. Inthis case, at the optimal roughness, reduction of theratio h / σ is compensed by increase of wearresistence of the surfaces.5. CONCLUSIONSFrom determinatios of the evolution over time ofthe wear, it resulted that, in the experimentalconditions used, minimal wear occurs at a certainvaloare of roughness and not at the minimalroughness.Surprisingly, minimum friction coefficient doesnot coincide with minimal wear.The existence of a minimum in the wear curveresults for roughness R a = 0.045 μm. At the sametime the friction coefficient is minimal at roughnessR a = 0.045 μm. A mathematical relationshipbetveen friction coefficient and wear cannot beestablished.REFERENCES[1] B. Katsutoshi, G.E., S. Kiyoshi: US Patent 5092898/ 03.03.1992.[2] J.P. Davant: Chirurgie de la hance. Mieux vivreavec une prosthesis. Fonder l’avenir, 23, pp. 53-56,1995.[3] M. Sadeghi-Mehr: Investigations of Rolling ElementBearing for Hip Joint Prosthesis, PhD Thesis,Imperial College of Science, Technology andMedicine University. London, 1997.[4] A. Iarovici, L. Capitanu, V. Florescu, M. Baubec:“Hip Joint Prosthesis with rolling bodies”,Proceedings of the Romanian Academy – Series A:Mathematics and Physics, Technical Sciences,Information Science, 1, 1-2, pp. 37-44, S. 2001.[5] L. Capitanu, V. Florescu. New Concepts forImproved Durability in MOM Total HipEndoprostheses. A review. American Journal ofMaterials Science. Vol.2, No. 6, December 2012.[6] A. Iarovici, L. Capitanu, V. Florescu, M. Baubec, F.Petrescu. Hip Joint Prosthesis with Rolling Bodies,Part I – The Balls Arrangement Analysis.Proceedings of the Annual Symposium of theInstitute of Solid Mechanics – SISOM, pp. 251-258,2001,[7] A. Iarovici, L. Capitanu, V. Florescu, M. Baubec, F.Petrescu. Hip Joint Prosthesis with Rolling Bodies,Part II – Numerical and Graphical Examples.Proceedings of the Annual Symposium of theInstitute of Solid Mechanics – SISOM, pp. 259-263,2001.[8] A. Iarovici, L. Capitanu, J. Onisoru, V. Florescu, M.Baubec. Stresses and deformations analysis for HipJoint Prostheses with Rolling Bodies. Proceeding ofthe Annual Symposium of the Institute of SolidMechanics SISOM 2002, pp. 263-268.[9] Dowson D. Elastohydrodynamic lubrication.Interdisciplinary Approach to the lubrication ofconcentrated contacts. (Ed. P.P.KU) NASA SP –237, 1970.340 13 th International Conference on Tribology – Serbiatrib’13

Serbian TribologySocietySERBIIATRIB ‘1313 th International Conference onTribologyKragujevac,Serbia, 15 – 17 May 2013Faculty of Engineeringin KragujevacANALYZING THEINFLUENCE OF THECONSTRUCTIONELEMENTPOSITION ONTORQUE TRANSMISSIONBYFRICTIONMarija Jeremić 1 , Bojan Bogdanović 1 , Aleksandar Simićć 1 , Dragomirr Miljanić 2Petar Todorović 1 , Sasa Randjelovic 1 , Branko Tadić 11 Faculty ofEngineering, Sestre Janjić 6, Kragujevac, Serbia, mjeremic88@yahoo.com, bogdanovicboki@@gmail.com,aleksandarsimickg88@gmail.com, petar@ @kg.ac.rs, sasarandjelovic@@yahoo.com, tadic@kg.ac. rs2 Metalik, Nikšić, MontenegroAbstract: This paper is analysing the impact of the construction element position of ship winch drum on the effects oftorque transmission by friction in the mechanization welding process. The driving and driven n wheels (constructionelements) were examined for the general case of the load distribution. Based on this examination, the construction ofthe device that should provide the reliable torque transmission andthe movement of the drumm in the process of itswelding is proposed. This construction is characterized by a high level of flexibility and ability to change the frictiontorque basedon changing drum positionin regard too the driving and driven wheels (construction elements) s). With thisnew construction, problems related to the movement synchronization are avoided, unlike the all previouslyy knownconstructionss of this type, which lead tothe positive impact on the wear intensityy of friction gears.Keywords: friction, wear, transfer off torque, special device, marine winch drum1. INTRODUCTIONResearchdescribed in the paper is related too theproblems of friction torque transmission. Researchis applicative and connected with design of bootdevice (reversal) drum ship winchess during weldingprocess. Torque transmission is performed byfriction tribological contact with rubber and metal.Movement can be reached by using only the effectsof friction, but frictionduring movement alwaysbrings different types of losses. Connected withthis, knowledge of the value off coefficientt offriction is very important for every engineer anddesignerwho is involvedindesign anddevelopment of mechanical structures, whichperform their function through interaction ofsurfaces which move relatively. It is well knownthat the process of friction followss every kindd ofbody movement. Friction is a necessary processbecause only with effects of friction f can beachieved starting, moving, changing speed orstopping. On the other hand, during movement, as aconsequence of frictional resistance, that resistancemust be overcome in order to continue movement,which is why the energylosses exist. In addition to13 th International Conference on Tribology – Serbiatrib’13energy loss, friction is always accompanied withwear of material on contact surfaces, whichhproducesan additional costs andloss offunctionality off elements in n contact.Friction is, therefore, such a process which, atthe same time, manifests positive and negativeeeffects. It is therefore natural that there is atendency to eliminate its negative effects, or at leastminimize, and to t increase the positive [ 1]. A reviewof the knowledge of the friction force, as well asways for measurement aree presented by Peter J.Blau [2], he presented some of the most commonstandards-defined measuring methods for f static anddynamic friction coefficient, and its potential uses.Beginning of movement m of any kind is related tothe existence of static friction. Static coefficient offriction depends of manyy parameters, especiallyfromthe surface of contact, normal load, andtemperature off the atmosphere in which contacttoccurs, surface absorption, quality of contacttsurface materials [3-7]. Analyzing the rollingprocess [8] inn the absence of anyload, theconclusion is that t the energy losses are result ofcollisions of two moving rolling mass, i.e. mass ofrolling body with w the body on whose surface is341

olling. Rolling resistance of radial motion of f thecylinder at flat surface was investigated in the paper[9]. Tested results showed that the coefficientofrolling friction depends on the speed of cylindermovement. At low speeds of cylinder, thecoefficient of rolling friction increases duee toincreasing of substraterate of deformation.Forhigher speeds, however, the coefficient of rollingfriction is reduced, thereby loweringthe area off thedeformed surface. The maximum force f of frictionat the initial moment of slip has been investigatedon rubber-metal frictionpairs under conditionss ofconstant compressivedeformation of the rubberduring transition from the high-elastic to the glassystate questioned A. I. El'kin et al [10] in the paper.Filled butadiene-nitrilee rubber compounds werestudied in the temperature rangefrom +200 to−50°C. The temperature dependence of themaximum force of friction has a sharply expressedmaximum near the glass transition temperature. . Asthe temperature falls, the force of friction at firstincreases, inaccordancewith the molecular-kinnetictheory. As the temperature continues to fall, inn thetransition region maximum force off friction beginsto rise more sharply owing to a sharp increasee inthe volume-mechanicalfriction component.Thefall in the maximum force of friction below theglass transition point associated with a decrease inthe deformed volume of rubber due to shrinkageand with the reducedd mechanical loss factor.Persson et al [11] study the sliding friction forviscoelasticc solids, e. .g., rubber, on hard flatsubstrate surfaces. Consider first the fluctuatingshear stress inside a viscoelastic solid which resultsfrom the thermal motionof the atoms or moleculesin the solid. At the nanoscale the thermalfluctuationsare very strong and give rise to stressfluctuationsin the MParange, which is similar tothe depinning stresses which typically occurr atsolid-rubberr interfaces, indicating the crucialimportance of thermal fluctuations for rubberfriction on smooth surfaces. Developed a detailmodel which takes into account the influencee ofthermal fluctuations on the depinning of smallcontact patches (stresss domains) at the rubber-thesubstrate interface. The experiment led toconclusion that the amplitude of the surfaceroughness has a very small effect off friction slidingrubber. Theeffects of carbon and cellulose fiberson the tribological characteristicsof rubber-basedfriction materials examined Akbar et al in his paper[12]. Friction tests realized with different slidingspeeds anddifferenttemperatures, with theexaminationn of the microstructureand mechanicalproperties of the surfaces in contact. Experimentalresults showed that carbon fibers had a minor effecton the coefficient friction, but that increase wearresistance. Shanahan et al [13] investigated ithemechanism of adhesion that occurs inthe contacttpair rubber andd hard metal rolling bodies. The highhdegree of adhesion can bee apparent even at roomtemperature if the contact time and pressure reachsufficient values. Based onn obtained results it wasdetermined thatt energy, which is dissipated duringrolling, refers not only to the influence thattaccompanies histeresis h adhesive separation, butalsoto the losses caused by loadingwith largecylinder. The nature n of friction between the rubberrandthe solid substrate is very important for manytechnical applications. Friction of rubber issignificantly different fromm the friction betweenhard substancess such as metals and ceramics. In thepaper [14] it was proved that the tire hassignificantly favorable friction characteristics.In order too optimize the construction wereeperformedlargenumberof theoreticalconsiderationsand preliminary ideas were done fora detailed review and analysis of the literature thattexamines this issue. Research is based ondetermining the transmission of torque from thespecial device on marine winch drum, which isdone by means of friction between rubber andmetal.2. CONCEPTUAL DESING OF ROTATINGEQUIPMENT FOR WELDINDAn important aspect of implementing of anyautomated system to the manufacturing process isits price. Because of that, design should beanalyzed in detail determing which parts of thetechnical systemm are not meaningful to t automate.To define the appropriate general design, projectstarts with the technicall requirements of theproduct, discusses designn solutions of existingproducts with similar functions (Figure 1). Drive ofrotary positioner shown is realized by two electriccmotors and gears, because speed and variable speeddrive is required. The principle of operation of theelectric motor which w drivess pieces of the t structurerequires ensuring synchronization of movement,which is seriouss theoretical and practical problem.Figure 1. Rotating positioner34213 th International Conference C onn Tribology – Serbiatrib’13

Especially when the chain transfer torquetransmissionn is based on the basis of friction. Therequirementfor the proper operation of structures,in the presence of synchronization, is that themotors are with the same characteristics. One off theproblems of providing synchronized movement isthe possible differencee of phasess of propulsionmotors. In addition, great equalization currents canoccur that can cause permanent deformation off themotor coils. In addition to these phenomenacanoccur torque on the wheel axle, which can causeeven shaft breakage. Ifthere is a significant phaseshift of motors, engines must be harmonized, whichmeans, must be speedup or slow down. Theproblem ofnot synchronized movement has aparticularly negative impact on thefrictional wearof the drive wheels.Figure 2. Conceptual design off constructionConceptual principally solution system Apogondrum marine winch, during the drafting process isshown in Figure 2. General function of rotationalstructure is achieved (Figure 2) executing a seriesof partial functions. Rotary positioner is composedof more machine parts,sub-assemblies, subgroupsand groups linked to a functional unit. A flexibleshaft coupling is interconnected with the shaft andwheels so the torque and power from the engine istransmitted from one shaft to another point. Drivenwheels rotating function performed under theinfluence of rotationof the drum. Framesconstruction, rely on a base that is used to keeping adistance between frames is realized, dependingg onthe needs ofthe fastener.3. ANALYSISOFTRANSMISIONFORCESONDimensioning of steel structures is performedbased on knowledge of layout external actions(forces andmoments). Determining the abilityy ofconstructionn to convey given loads of force is basedon the characteristics ofthe materials used andd theallowable stress. Allowable stressess are prescribedfor the steelstructure and functionss are determinedby the choice of the material and the t character of13 th International Conference on Tribology – Serbiatrib’13the external load. Calculation of the construction isproof load and usability off the structure under theinfluence of prescribed loadd during thelife time ofconstruction. The process of designing, calculatinganddimensioning of construction is actually choiceof a calculationn model to credibly describe the realbehavior of thee structure under the prescribed load,buttaking into account thee optimal complexity ofthe model, the t possibility to calculate theperformance and profitability of individual parts ofthe structure. This work studied the loads that occurin the system. . System shown in Figure 3 wasexaminated as static s system, i.e. systemat rest stateandthe conditions that enable this state are used.External cause (external effects) on thesolid bodyis the force thatt causes a change in sleep mode andthe motion of the system. For the analysis of theeffects of forcess in the specific examplee of the drumwith diameter D = 1000 mm and length l = 20000mmm load is thee weight of the drum, which is G =50000 N. Based on these data dimensions ofconstruction were w developed. It is necessary todetermine the load l distribution, especially load offriction wheel as a function of drum position(Figure 3). Rotary positioner is designed that atwork most of the weight rests on the drum drivewheel. The whole system iss viewed as a horizontalsystem which is attached to the substrate. In thiscase, the main task is reduced to the determinationnof the resistance of the supports and the force offriction.Figure 3. Schedule forces onn the maximumm size of thedrummIn the beginning problem is solved imagininggthat the supports are removed and replaced by theiraction forces called c resistance actions, as in thiscase the normall force wheels and . It analyzesthe effect of thee friction forces that appear betweenthe drum andd wheel. How there are differentmaterials in the contact zone, i.e. theinteractionnbetween the rubber and steel, the coefficients offriction μ is different. Connected with this, for the3433

sliding friction, apropos coefficient of frictionbetween the drive wheel and the drum takes thevalue 0.6, and the value of coefficient ofrolling friction between the driven wheels anddrums 0.05. As a starting basis for solvinggiven problem is the determination of friction forcefrom the equation: ∙ ∙ In order that observed system was inequilibrium, the resultant of all the forces must beequal to zero. Since in the observed case of actionof horizontal and vertical forces, this implies twoequilibrium conditions. 0 0From the previous two conditions can becalculated unknown resistance supports N 1 and N 2 .It's actually a system of two equations with twounknowns. The first condition of equilibrium is thesum of all components of the forces which acthorizontally in the direction of the x axis is equal tozero: 01where are: , , - projection force F ti N on the x axis.The second condition is the sum of all forceswhich acting vertically on the given system is equalto zero: 0 2where are: , , - projectionforce on the y axis.Further analysis of the static equationsdepending on the angles that are withinthe values of 65⁰ to 82⁰ for angle ie. from 38⁰ to47⁰ for angle φ to get possible values of the forcesthat are shown in Table 1.The final expressions for the calculation of reactionto friction wheels are: ∙ cos cos90 3 ∙ cos cos90 ∙ cos cos90 ∙ cos cos90 where k is a dimensionless coefficient.4Table 1. The dependence the force of anglesThe dependence the friction force and theresistance supports on the angle φ F N ∙ μ ∙cosφ F N ∙ μ ∙sinφ ∙cos90 ∙sin90 The dependence the friction force and theresistance supports on the angle φ ∙ ∙ ∙ ∙ ∙90 ∙90 Static equilibrium conditions for the y axis is: ∙ ∙sin ∙sin90 ∙sin sin90 ∙ 56 ∙ 7 ∙ 8Showing force intensity in depending on thechanges of angles, ie. depending on the position ofthe drum, and with the same diameter and weight ofthe drum, were calculated and are presented inTable 2.Table 2. Force intensities depending on the anglesφ φ N 1[kN]N 2[kN]F t1[kN]F t2[kN]1 82 38 6.03 41.41 0.30 24.852 80 39 6.88 41.16 0.34 24.703 78 40 7.75 40.88 0.39 24.534 76 41 8.61 40.55 0.43 24.335 74 42 9.48 40.19 0.47 24.116 72 43 10.35 39.79 0.52 23.877 70 44 11.22 39.35 0.56 23.618 68 45 12.11 38.87 0.61 23.329 66 46 12.99 38.35 0.65 23.0110 65 47 13.83 38.01 0.69 22.80Based on analysis of obtained values of thenormal forces and frictional force from Table 1, itcan be observed that for settings in any position, thehighest intensity have a force N 2 , i.e. maximumload is on the drive wheels. Based on comparativeanalysis calculation, as well as most of thecalculation that are not shown in this paper,dependence of driving force (friction force) in thefunction of the drum position is established, i.e.dependance of angles . This dependency ofthe complex form , is shown in thediagram, which is shown in Figure 4.It can be seen from the diagram that the functionreaches a very high values of the force intensityaround angles which are close to 90 0 . It344 13 th International Conference on Tribology – Serbiatrib’13

is obvious that this isa so-calledwedge effect,which occurs at extremely high intensity oftangential force.Figure 4. Diagrammatic representation of theedependence of the friction drive wheel of the anglesOstensibly, this "extreme function" can be goodsolution from the driving forces aspect. However,to achieve such high values of driving force, thesystem must be exposed to very high valuess ofnormal load, in the order of 20 [kN]. Theapplication of such highh load is veryproblematic interms of the allowablestress levels between thedrum and drive wheel. For this reason, the authorsfind that the angles of 2 should be chosenin a range with much lower values, with which willnot be intensive wear and will be ensured stabilityof device.In Figure 5, the different dimensionsof the drumcanbe set on the positioner. The distance betweenthe rollers cann be, depending on needs, set byscrew. Movement of rolls is possible in i horizontalandvertical direction of thee positioner frame. f Drivefor rollers aree accomplished with one electriccmotor, which reduces the e speed by worm gear.Operation of the t second part of the carriage isachieved by a flexible shaft. In this way using offluid power the t synchronization off engine isavoided. Usingg a positioner in the productionprocess, which gives high efficiency and cheaperprocess of welding, has thee characteristics of light,reliability,performanceand widefield ofapplications. Workpieces are therefore placed in thebest possible way and in the most favorableposition in relation to the welders, and on the basissof technical documentation..4. CONSTRUCTIVEDEVICESOLUTIONOFCalculation of the structure derived in theprevious section provesportability and usabilityy ofthe structure under the influence of f prescribed loadduring the life time of construction. Conceptt ofspecial device (device for popositioningandrotation ofdrum during manufacturing process)presented has as main function of f torque transfer(power) on the basis of friction between rubber andsteel elements that are in direct contact. To ensurethe power transfer, it isnecessary that t the forcee offriction is greater than torque on the wheels. Inn theprocess of developmentas well as design structureshown, estimate segment of profitability aspect is akey elementin the design. This typeof positioner isspecifically designed and manufactured for thepurpose ofincreasing productivity, reducing theintensity for the welder, because all the weldingprocess canbe performed automatically. Weldinghead is managed by welder, while workpiecerotates. Welding circular cylinder (drum), using acrane to set the adjustable rollers trolley. Thedistance of cart can adjust the size of the cylinder.13 th International Conference on Tribology – Serbiatrib’13Figure 5. Different sized drum set up in the positionerFigure 6. Rotaryy positioner assembly: 1- wheel drive, 2-driven wheel, 3 - base, 4- frame positioner, 5 - flexibleshaft, 6 - drum345

Rotary positioner (Figure 6) is composed of anumber of mechanical parts, sub-assemblies, andsub-groups linked to a functional unit. Drive wheel(position 1) is one of the main structural elementsand plays an important role in the performance offunctions. Casing point consists of sides and basewich are welded together. Beside the basic functionto combine elements and enable the properfunctioning of the wheel, housing need to protectwheel from external influence. During welding ofsides paralelism should be taken into account toensure that shaft, which is fitted in the openings odsides, can smootly perform their function. Shaftallows rotating of all parts which are there,assembling at functional unit and transferring loads.Given that the direction of external loads arevertically, for reliable operation of the wheel radialbearings are fitted. Both bearing are protectedagainst atmospheric agents’ with cups. Worm gearwith electric motor serves to transmit power fromthe engine to the working machine, and to adjustrotational speed and torque needed for shaft.Subassembly of driven wheel (position 2) is easilyderived as opposed to the drive wheel subassembly.For the purpose of stiffening the case, it isnecessary to construct the corresponding ribs. Inthis case the stationary shaft and pressed throughthe opening of sides, the wheel while under theinfluence of the load rotates around it. The wheelsare made of steel coated with rubber; where rubberhas a function to reduce wear process. At the endsof the shaft exist thread for nut and washner whichlocates bearing. The distance between the bearinghousing and sides is provided with spacers. As eachobject, in general, can be created in several ways,analysis of possible variants is done nad optimalsolution is selected. Standard steel profiles are usedfor base (position 3). Length of profile is chosenoptimally, based on the analysis of cases in terms ofparameters essential for the performance of thedrafting process. The main function of the base is toensure the proper conduct of frame structureswithout the possibility of drift during operation.The connection between the frame and the base isaccomplished by means of threaded joints.Parallelism of the guide is provided with two rods,which are connected with bolts to linear guides.Frame positioner (position 4) is an importantelement of the structure. It consists of two standardsteel profiles assembled by welding. The functionof the frame is to connect wheels into onefunctional unit. Safety and reliability depends onthe position of the wheels and safety and reliabilityof threaded fasteners. Nuts and bolts are used toconnect subassembly of wheels with frame.Movement of the wheels enables frame length andthe distance depends on the size of the drum beingprocessed.5. DISCUSSIONThe introduction of new design into theproduction process has many advantages which areshown as follows: Improving the quality of welded joints isachieved thereby enabling semiautomaticwelding. During rotation of the positioner,the worker can freely perform single-passwelding continuously, without interruptionof the welding. The process is repeated untilthe drum is fully welded at the site. Reduced the preparatory time duringwelding Increase in profit resulting from reducedpreparation time. Based on the review, withthe introduction of the positioner, it ispossible to produce two winches on theyear, which also means, that theintroduction of new design will paid in ashort period of time.Technical features of the structure are suchthat it is possible to perform the procedureon the positioner gas cutting, as well aswelding.Reengineering processes need to improveorganizational structure, to allow the replacementof long-term measures and measures to make quickand drastic changes to increase the quality, reducethe cost, reduce execution time of the process,improve internal and external relationships,eliminate unnecessary activities, provide a pleasantatmosphere for work and eliminate unnecessaryactivities.6. CONCLUSIONSFrom review of the literature sources related tothe considered problem, it can be concluded thatthis area of research is very complex. Here ispresented concept of development of specialproducts (devices for positioning and rotation of thedrum), based on knowledge of engineering sciencegroup. In order to optimize the construction,theoretical considerations made a number ofconceptual designs and prepared a detailed reviewand analysis of the literature that examines thisissue. The proposed solution does not requiresynchronization device for movement of two drivemotors, which is a particular problem, which isdiscussed in the paper. The inclusion of a specialdevice, designed for the production gets reducedvalue processing, minimum use of materials, betterquality and more reliable weld construction346 13 th International Conference on Tribology – Serbiatrib’13

winches. The authors are inclined to think that aproper choice of frame and accessory devices forpositioning, funds can be recovered at the level ofthe two-year. It should be noted that this positionerwhich is semiautomated has positive effects onwelding proces in terms of safety. In particular, ifone takes into account that safety in the workplaceis one of the most important categories of businessin modern conditions of production. On the otherhand, the structure is very efficient and suitable forwide application in industrial practice. A veryimportant advantage of this design is that themachine can be made with a lot of low costelements. If we look at design, it is more than clearthat the design does not require a particularly highaccuracy and precision manufacturing. It wastended to meet function durring design of device, tobe easy for manufacture, to be easy formanipulation, to have lowest cost possible and tosatisfy safety requirements.REFERENCE[1] M. Đurđanović, D. Stamenković: Trenje mirovanjauslovikretanja, SERBIATRIB`07[2] Peter J. Blau: The significance and use of thefriction coefficient, Tribology International,Vol. 34, pp 585–591, 2001.[3] K.-H. Zum Gahr, K. Voelker, Friction and wear ofSiC fiberreinforced borosilicate glass mated to steel,Wear 225–229, pp 88–895, 1999.[4] P. Blau, The significance and use of the frictioncoefficient, Tribol. Int. 34, pp 585–591, 2001.[5] B. Ivkovic, M. Djurdjanovic, D. Stamenkovic: TheInfluence of the Contact Surface Roughness on theStatic Friction Coefficient, Tribology in Industry,Vol. 22, No. 3&4, pp 41-44, 2000.[6] U. Muller, R. Hauert: Investigations of thecoefficient of static friction diamond-like carbonfilms, Surface and Coatings Technology 174–175,pp. 421–426, 2003.[7] B. Polyakov, S. Vlassov, L.M. Dorogin, P. Kulis, I.Kink, R.Lohmus: The effect of substrate roughnesson the static friction of CuO nanowires, SurfaceScience, Vol. 606, pp 1393–1399, 2012.[8] D.A. Desai: What is rolling friction, Resonance 9,pp 52-54, 2004.[9] T. Pöschel, T. Schwager, N.V. Brilliantov, Rollingfriction of a hard cylinder on a viscous plane, TheEuropean Physical Journal B - Condensed Matterand Complex Systems 10, pp 169-174, 1999.[10] A.I. El'kin, A. I. Moskovkin, E. I. Veksel'man,Maximum force of friction in rubber-metal frictionpairs under conditions of constant compressivedeformation of the rubber, Mekhanika Polimerov,Vol. 3, No. 3, pp. 533–538, 1967.[11] B.N.J. Persson, A.I. Volokitin, Rubber friction onsmooth surfaces, The European Physical Journal E,21, pp 69-80, 2006.[12] Akbar Shojaei, Mohammad Arjmand, Amir Saffar,Studies on the friction and wear characteristics ofrubber-based friction materials containing carbonand cellulose fibers, Journal of Materials ScienceMarch 2011, Volume 46, Issue 6, pp 1890-1901[13] M.E.R. Shanahan, N. Zaghzi, J. Schultz, A. Carré,Hard Rubber/Metal Adhesion Assessment Using aHeavy Cylinder Rolling Test, Adhesion 12, pp 223-238, 1988.[14] Valentin L. Popov, Rubber Friction and ContactMechanics of Rubber, Contact Mechanics andFriction, pp 255-270, 2010.13 th International Conference on Tribology – Serbiatrib’13 347

Serbian TribologySocietySERBIATRIB ‘1313 th International Conference onTribologyKragujevac, Serbia, 15 – 17 May 2013Faculty of Engineeringin KragujevacUSE ALGORITHM FOR CONSTRUCTION 3D VISIBILITYGRAPHS TO DESCRIBE PLASTIC AND ELASTICDEFORMATION OF ROBOT LASER HARDENED SPECIMENSM. Babič 1 , P. Kokol 2 , M. Milfelner 3 , P. Panjan 4 , Igor Belič 51 Emo-Orodjarna d.o.o., Slovenia, babicster@gmail.com2University of Maribor, Faculty of Health Sciences, Slovenia,3 Tic-Lens d.o.o., Slovenia,4 Institute Jozef Stefan, Slovenia,5 Institute of Materials and Technology, Slovenia,Abstract: Visibility graphs have many applications. One of all applications is analyze trend line of marketgraph. Here we use 2D visibility graph for analyze it. Construction for 2D visibility graph is known. But inthis paper, we will present algorithm for construction 3D visibility graphs. 3D Visibility computations arecentral in any computer graphics application. Drawing graphs as nodes connected by links in 3D space isvisually compelling but computationally difficult. Construction of 3D visibility graph have big timecomplexity, thus require high professional computer or supercomputer. Article describe new method,algorithm for analyze 3D visibility graphs. We develop algorithm, which draws 3D visibility graphs, foranalyze microstructure pictures of robot laser hardened specimens. Microstructure of robot laser hardenedspecimens is very complex, but we can present it with 3D visibility graphs. Algorithm for construction 3Dvisibility graph is very usefull in many cases, including illumination and rendering, motion planning, patternrecognition, computer graphics, computational geometry and sensor networks, military and automotiveindustry. This new algorithm we use to patterns recognition. In materials science, deformation is a change inthe shape or size of an object due to an applied force (the deformation energy in this case is transferredthrough work) or a change in temperature (the deformation energy in this case is transferred through heat).With 3D visibility graphs we described elastic and plastic deformation of robot laser hardened specimens.For analysis of results, we use an intelligent system method, namely, a neural network, linear regression andsupport vector machine. We compare all methods.Keywords: Elastic and plastic deformation, visibility graph, robot, laser, hardening,1. INTRODUCTIONIn materials science, deformation is a change inthe shape or size of an object due to an appliedforce (the deformation energy in this case istransferred through work) or a change intemperature (the deformation energy in this case istransferred through heat). The first case can be aresult of tensile (pulling) forces, compressive(pushing) forces, shear, bending or torsion(twisting). In the second case, the most significantfactor, which is determined by the temperature, isthe mobility of the structural defects such as grainboundaries, point vacancies, line and screwdislocations, stacking faults and twins in bothcrystalline and non-crystalline solids. Themovement or displacement of such mobile defectsis thermally activated, and thus limited by the rateof atomic diffusion. Deformation is often describedas strain.Laser hardening is a metal surface treatmentprocess complementary to conventional ame andinduction hardening processes. A high-power laserbeam is used to heat a metal surface rapidly andselectively to produce hardened case depths of upto 1,5 mm with the hardness of the martensiticmicrostructure providing improved properties suchas wear resistance and increased strength. We will348 13 th International Conference on Tribology – Serbiatrib’13

find parameters of robot laser hardened cell,because we will reduce deformations. We observe amicrostructure of robot laser hardened patterns. Wefind plastic and elastic deformations.Depending on the type of material, size andgeometry of the object, and the forces applied,various types of deformation may result. The imageto the right shows the engineering stress vs. straindiagram for a typical ductile material such as steel.Different deformation modes may occur underdifferent conditions, as can be depicted using adeformation mechanism map. Laser hardening is ametal surface treatment process complementary toconventional ame and induction hardeningprocesses. A high-power laser beam is used to heata metal surface rapidly and selectively to producehardened case depths of up to 1,5 mm with thehardness of the martensitic microstructureproviding improved properties such as wearresistance and increased strength. We will findparameters of robot laser hardened cell, because wewill reduce deformations. We observe amicristructure of robot laser hardened patterns. Wefind plastic and elastic deformations.Figure 2. Robot laser hardening with differenttemperature and speedFirstly, we analize profile graph ofmicrostructure picture with visibility graph.Graph 1. Profile graph of surface hardened specimen2. MATERIALS AND METHODOur study was limited to tool steel of DINstandard 1.7225 (Fig. 1). The chemical compositionof the material contained 0.38% to 0.45% C, 0.4%maximum Si, 0.6% to 0.9% Mn, 0.025% maximumP, 0.035% maximum S and 0.15% to 0.3% Mo.Figure 1. Transverse and longitudinal cross-section ofhardened specimenThe specimen test section had a cylindrical formof dimension 25 × 10 mm (diameter × height).Specimens with porosity of about 19% to 50%,were prepared by laser technique, followed byhardening at T ∈ [1000, 1400] °C and v ∈ [2, 5]mm/s. We changed two parameters of the robotlaser cell: speed v ∈ [2, 5] mm/s with steps of 1mm/s and temperature T ∈ [1000, 1400] °C in stepsof 100 °C (Fig. 2).We develop new algorithm for constructionvisibility graph in 3D space. This algorithm we useto analyze mechanical properties of robot laserhardened specimens.Graph 2: 2D Visibility graph for graph 1Reason for develop constructing the visibilitygraph in 3D space is analyze mechanical propertiesof robot laser hardening. Robot laser hardenedspecimens have better microstructure mechanicalproperties after hardening. With 3D visibility graphwe describe complexity of microstructure (Fig. 2and Fig. 3) of hardened specimens. We will knowwhich extremes on graph of microstructure areconnected.3. RESULTOn Graph 3 is presented deformation beforelaser hardening. Graph 4 present deformation afterlaser hardening. On Graph 5 is presentedrelationship between hardness (HV) and depthbefore laser hardening. On Graph 6 is presented13 th International Conference on Tribology – Serbiatrib’13 349

elationship between hardness (HV) and depth afterlaser hardening.2.51.50.5Depth [um]200 400 600 800 1000Load [mN]Graph 3. Deformation before laser hardening4. CONCLUSIONWe made experiments hardened specimen. Wepresent deformation on material before and afterrobot laser hardening. We present relationshipbetween hardness and depth befor and afterhardening. In the future we want to exploredeformatiuon as a function of more parameters of arobot laser hardening. Robot laser cell parametersare strength, energy density, focusing distance,energy density in the focus, focus position,temperature and speed of hardening. We willinterested to investigate deformation in the twobeamlaser robot hardening (a laser beam is splitinto two parts).Depth [um]2.51.50.5200 400 600 800 1000Load [mN]Graph 4. Deformation after laser hardeningREFERENCES[1] B.M. Yu, P. Cheng: A fractal permeability modelfor bi-dispersed porous media, International journalof Heat Mass Transfer 45, pp. 2983–2993, 2002.[2] Z. Q. Chen, P. Cheng, C.H. Hus: A theoretical andexperimental study on stagnant thermal conductivityof bi-dispersed porous media, Int. Commun. HeatMass Transfer 27, pp. 601–610, 2000.[3] R. Pitchumani, B. Ramakrishnan: A fractalgeometry model for evaluating permeabilities ofporous preforms used in liquid composite molding,Int. J. Heat Mass Transfer 42, pp. 2219–2232, 1999.[4] B.B. Mandelbrot: The fractal geometry of nature.New York: W. H. Freeman, 1982:93.Graph 5. Hardness in depth before laser hardeningGraph 6. Hardness in depth after laser hardening350 13 th International Conference on Tribology – Serbiatrib’13

Serbian TribologySocietySERBIATRIB ‘1313 th International Conference onTribologyKragujevac, Serbia, 15 – 17 May 2013Faculty of Engineeringin KragujevacUSE FRACTAL GEOMETRY TO DESCRIBE FRICTION OFROBOT LASER HARDENED SPECIMENSM. Babič 1 , P. Kokol 2 , M. Milfelner 3 , P. Panjan 4 , Igor Belič 51 Emo-Orodjarna d.o.o., Slovenia, babicster@gmail.com2University of Maribor, Faculty of Health Sciences, Slovenia,3 Tic-Lens d.o.o., Slovenia,4 Institute Jozef Stefan, Slovenia,5 Institute of Materials and Technology, Slovenia,Abstract: This paper describes some of our experience in laser surface remelting, consolidating, andhardening of steels. The process of laser hardening with remelting of the surface layer allows us to veryaccurately determine the depth of modified layers. In this procedure, we know the exact energy input into thematerial. Heating above the melting temperature and then rapidly cooling causes microstructural changes inmaterials, which affect the increase in hardness. Mathematics and Computer Science are very useful in manyother Science. We use mathematical method, fractal geometry in engineering, exactly in laser technics.Moreover, with fractal geometry we analize complexity of robot laser hardened specimens. We analizespecimens hardened with different parameters of robot laser cell. So we changed two parameters speed v ∈[2, 5] mm/s and temperature T ∈ [1000, 1400] °C. In this work, we have used a scanning electronicmicroscope (SEM) to search and analyse the fractal structure of the robot laser hardened specimens.Friction is the force resisting the relative motion of solid surfaces, fluid layers, and material elements slidingagainst each other. The present study is intended to use new method, fractal geometry to describe completelyfriction of robot laser hardened specimens. Finally, concept of fractal geometry is applied to characterizethe microstructure and derive the useful relationship between fractal dimension and microstructuralfeatures. The modeling of the relationship was obtained by the four layer neural network.Keywords: Friction, fractal geometry, robot, laser, hardening,1. INTRODUCTIONFriction is the force resisting the relative motionof solid surfaces, fluid layers, and material elementssliding against each other.In nature we have many geometrical objectswhich are irregular and cannot be described withclassical Euclidian geometry. Thus we need a newmethod for describing the complexity andirregularity of objects. A relatively new method isfractal geometry. Recently, a concept of fractalgeometry which was originally developed for theanalysis of irregular features in nature has beenfinding increased applications in the fields ofmaterials science for the characterization ofmicrostructures. The key to fractal geometry is thefractal dimension, which describes the complexityof a fractal and geometrically irregularmicrostructure. Measuring fractal dimensions hasbecome a common practice for describing thestructural properties of roughness and hardness ofheat-treated materials. We use fractal geometry inlaser techniques. Laser hardening is a metal surfacetreatment process that is complementary toconventional aim and induction hardeningprocesses. A high-power laser beam is used to heata metal surface rapidly and selectively so as toproduce hardened case depths of up to 1.5mm withthe hardness of a martensitic micro-structure,providing improved properties such as wearresistance and increased strength.First, we hardened tool steel standard label DINstandard 1.7225 with a robot laser cell. Thechemical composition of the material contained0.38 to 0.45% C, 0.4% maximum Si, 0.6–0.9% Mn,0.025% maximum P, 0.035% maximum S and13 th International Conference on Tribology – Serbiatrib’13 351

0.15–0.3% Mo [18]. The specimen test section wasin a cylindrical form with dimensions of 25×10mm.We changed two parameters, speed v ∈ [2, 5] mm/sin steps of 1 mm/s, and temperature T ∈ [1000,1400] °C in steps of 100 °C. After hardening, wepolished and etching all specimens. Detailedcharacterization of their microstructure before andafter surface modifications was conducted using afield emission scanning electron microscope, JEOLJSM-7600F. We used the ImageJ program(available from the National Institute of Health,USA) to analyse these pictures. On these specimenswe took measurements of roughness and hardnessbefore and after robot laser hardening.Figure 3. Transverse and longitudinal cross-section ofhardened specimenThe specimen test section had a cylindrical formof dimension 25 × 10 mm (diameter × height).Specimens with porosity of about 19% to 50%,were prepared by laser technique, followed byhardening at T ∈ [1000, 1400] °C and v ∈ [2, 5]mm/s. We changed two parameters of the robotlaser cell: speed v ∈ [2, 5] mm/s with steps of 1mm/s and temperature T ∈ [1000, 1400] °C in stepsof 100 °C (Fig. 2).Figure 1. Microstructure before robot laser hardeningFigure 2. Microstructure of robot laser hardenedspecimen with 1000° C and 2 mm/s2. MATERIALS AND METHODOur study was limited to tool steel of DINstandard 1.7225 (Fig. 1). The chemical compositionof the material contained 0.38% to 0.45% C, 0.4%maximum Si, 0.6% to 0.9% Mn, 0.025% maximumP, 0.035% maximum S and 0.15% to 0.3% Mo.Figure 4. Robot laser hardening with differenttemperature and speedA profilometer (available from the InstituteJozef Stefan, Slovenia) was used for themeasurement of the surface roughness parameterR a (arithmetic mean deviation of the roughnessprofile) and hardness of the robot laser hardenedspecimens. SEM images were converted into binarydigital images (using the public domain software,ImageJ). The fractal characterization of materialsproperties as an applicable and potential tool hasbeen well documented. The key parameter in fractalgeometry is the fractal dimension, D, which shouldbe determined first before we use the concept andknowledge of fractal geometry to characterize themicrostructure of the robot laser hardenedspecimens. We calculated the fractal dimensionusing image processing of the SEM pictures incombination with implementation of a boxcountingmethod (algorithm) using the ImageJsoftware. The measure of the fractal object, M(L),is related to the length scale, L, through a scaling inthe form of Eq. (1):M(L)=L D (1)where M(L) is the surface area of a pore and D isthe fractal dimension of the sample. A twodimensionalobject, such as the SEM picture, canbe divided into N(ε) self-similarity smaller squares,352 13 th International Conference on Tribology – Serbiatrib’13

each of which is measured by the length ε. Thefractal dimension can be calculated according toEq. (2):D=lnN(ε)/lnε. (2)Characterization of surface topography isimportant in applications involving friction,lubrication, and wear (Thomas, 1999). In general, ithas been found that friction increases with averageroughness. Roughness parameters are, therefore,important in applications such as automobile brakelinings and floor surfaces. The effect of roughnesson lubrication has also been studied to determine itsimpact on issues regarding lubrication of slidingsurfaces, compliant surfaces, and roller bearingfatigue. Finally, some researchers have found acorrelation between the initial roughness of slidingsurfaces and their wear rate. Such correlations havebeen used to predict the failure time of contactsurfaces. A section of standard length is sampledfrom the mean line on the roughness chart. Themean line is laid on a Cartesian coordinate systemwherein the mean line runs in the direction of the x-axis and magnification is the y-axis. The valueobtained with the formula on the right is expressedin micrometers (Om) when y=f(a).Fractal dimension1,9651,961,9551,951,9451,941,9351,935 mm/s4 mm/s3 mm/s1,9252 mm/s1,920 50 100 150 200 250RoughnessExperimental dataFitting curve with neural networkGraph 3. Relationship between fractal dimension androughness R a in specimens hardened at different speedsat 1000 °CFractal dimension21,991,981,971,961,951,941,934 mm/s5 mm/s3 mm/s2 mm/s1,9234 234 434 634 834 1034 1234 1434RoughnessExperimental dataFitting curve with neural networkGraph 4. Relationship between fractal dimension androughness R a in specimens hardened at different speedsat 1400 °C4. CONCLUSIONGraph 1. Arithmetical mean roughness (Ra)3. RESULTWe studied the relationship between the fractaldimension, parameters of the robot laser cell androughness (friction).The paper presents the use of fractal geometry todescribe the mechanical properties of robot laserhardened specimens. We use a relatively newmethod, fractal geometry, to describe thecomplexity of laser hardened specimens. The mainfindings can be summarized as follows:1. A fractal structure exists in the robot laserhardened specimens.2. We describe the complexity of the robot laserhardened specimens using fractal geometry.3. We have identified the optimal fractal dimensionof tool steel hardened with different robot laserparameters.4. We use the box-counting method to calculate thefractal dimension for robot laser hardenedspecimens with different parameters.REFERENCESGraph 2. Roughness od robot laser hardenendspecimens[1] B.B. Mandelbrot. The fractal geometry ofNature. New York: W.H. Freeman, p. 93, 1982.[2] P.V. Yasnii, P.O. Marushchak, I.V.Konovalenko, R.T. Bishchak, Computeranalysis of surface cracks in structuralelements, Materials Science 46, pp. 833-839,2008.13 th International Conference on Tribology – Serbiatrib’13 353

[3] P.V. Yasnii, P.O. Marushchak, I.V.Konovalenko, R.T. Bishchak, Structuraldegradation and damage caused by a system ofcracks to the steel of metallurgical equipment,Materials Science 47, pp. 798-803, 2009.[4] C.C. Barton, Fractal Analysis of Scaling andSpatial Clustering of Fractures. In: Barton, C.C.and La Pointe, P.R. (eds.), Fractals in the EarthSciences. Plenum Press. New York, pp. 141-178, 1995.[5] L.W. Fan, Y.C. Hu, T. Tian, Z.T. Yu, Thepredication of effective thermal conductivitiesperpendicular to the fibres of wood using afractal model and an improved transientmeasurement technique, International Journalof Heat Mass Transfer 49, pp. 4116–4123,2006.[6] T. Ficker, Fractal strength of cement gels anduniversal dimension of fracture surfaces,Theor. Appl. Fract. Mech. 50, pp. 167–171,2008.[7] M.Q. Jiang, J.X. Meng, J.B. Gao, X.L. Wang,T. Rouxel, V. Keryvin, Z. Ling, L.H. Dai,Fractal in fracture of bulk metallic glass,Intermetallics 18, pp. 2468–2471, 2010.[8] A. Carpinteri, S.A. Puzzi, A fractal approach toindentation size effect, Engineering Fract.Mech. 73, pp. 2110-2122, 2006.354 13 th International Conference on Tribology – Serbiatrib’13

Serbian TribologySocietySERBIATRIB ‘1313 th International Conference onTribologyKragujevac, Serbia, 15 – 17 May 2013Faculty of Engineeringin KragujevacUSE NEW PROCESS IN ROBOT LASER HARDENING TODECREASE WEAR OF SPECIMENSM. Babič 1 , P. Kokol 2 , M. Milfelner 3 , P. Panjan 4 , Igor Belič 51 Emo-Orodjarna d.o.o., Slovenia, babicster@gmail.com2University of Maribor, Faculty of Health Sciences, Slovenia,3 Tic-Lens d.o.o., Slovenia,4 Institute Jozef Stefan, Slovenia,5 Institute of Materials and Technology, Slovenia,Abstract: The mechanism of wear is very complex and the theoretical treatment without the use of rathersweeping simplifications is not possible. The material intrinsic surface properties such as hardness, strength,ductility, work hardening etc. are very important factors for wear resistance, but other factors like surfacefinish, lubrication, load, speed, corrosion, temperature and properties of the opposing surface etc. areequally important. Robot laser surface-hardening heat treatment is complementary to conventional flame orinductive hardening. A high-power laser beam is used to heat a metal surface rapidly and selectively toproduce hardened case depths of up to 1.5 mm with hardness values of up to 65 HRc. Laser hardeninginvolves features, such as non-controlled energy intake, high performance constancy and accuratepositioning processes. A hard martensitic microstructure provides improved surface properties such as wearresistance and high strength. We describe a new technological process of hardening, which can decrease thewear of hardened specimens. The new process uses robot laser hardening with an overlapping laser beam.First, we hardened specimens using different velocities and temperatures and then repeated the process. Inaddition, we present how the speed and temperature affect the wear in two different processes of robot laserhardening. Furthermore, we present the improved results after hardening with the overlap process. Toanalyse the results, we used one method of intelligent system, neural networks and a relationship wasobtained by using a four-layer neural network. We compare both processes.Keywords: Wear, robot, laser, hardening, process of overlapping,1. INTRODUCTIONIn materials science, wear is erosion or sidewaysdisplacement of material from its "derivative" andoriginal position on a solid surface performed bythe action of another surface. Wear is related tointeractions between surfaces and more specificallythe removal and deformation of material on asurface as a result of mechanical action of theopposite surface. The need for relative motionbetween two surfaces and initial mechanical contactbetween asperities is an important distinctionbetween mechanical wear compared to otherprocesses with similar outcomes.The definition of wear may include loss ofdimension from plastic deformation if it isoriginated at the interface between two slidingsurfaces.However, plastic deformation such as yieldstress is excluded from the wear definition if itdoesn't incorporates a relative sliding motion andcontact against another surface despite thepossibility for material removal, because it thenlacks the relative sliding action of another surface.2. MATERIALS AND METHODOur study was limited to tool steel of DINstandard 1.7225 (Fig. 1). The chemical compositionof the material contained 0.38% to 0.45% C, 0.4%maximum Si, 0.6% to 0.9% Mn, 0.025% maximumP, 0.035% maximum S and 0.15% to 0.3% Mo[10].13 th International Conference on Tribology – Serbiatrib’13 355

Figure 1. Transverse and longitudinal cross-section ofhardened specimenThe specimen test section had a cylindrical formof dimension 25 × 10 mm (diameter × height).Specimens with porosity of about 19% to 50%,were prepared by laser technique, followed byhardening at T ∈ [1000, 1400] °C and v ∈ [2, 5]mm/s. First, we changed two parameters of therobot laser cell: speed v ∈ [2, 5] mm/s with steps of1 mm/s and temperature T ∈ [1000, 1400] °C insteps of 100 °C (Fig. 2). Secondly, we repeated theprocess (Fig. 3). In addition, we hardened thespecimens again with equal parameters of the robotlaser cell. The microstructure of the specimens wasobserved with a field emission scanning electronmicroscope (JSM-7600F, JEOL Ltd.). An irregularsurface texture was observed with a few breaks,which are represented by black islands (Fig. 4). Fig.5 presents the boundary between the hardened andnon-hardened material.Figure 2. Robot laser hardening with differenttemperature and speedFigure 3. Repeated process of robot laser hardeningFigure 4. SEM picture of robot laser re-hardenedspecimenFigure 5. The boundary between work-hardened andnon-hardened materialWe used the method of determining the porosityfrom SEM images of the microstructure. It isknown that in a homogenously porous material thearea of pores is equal to the volume of pores inspecimens. The SEM pictures were converted tobinary images (Fig. 6), from which we calculatedthe area of pores of all pictures using the ImageJprogram (ImageJ is a public domain, Java-basedimage processing program developed at theNational Institutes of Health). The area of pores oneach picture of the material was calculated and thenthe arithmetic mean and standard deviation ofporosity were determined. To analyze he possibilityof the application of fractal analysis to the heattreatedsurface, we examined the relation betweenthe surface porosity and fractal dimensionsdepending on various parameters of the robot lasercell. In fractal geometry, the key parameter is thefractal dimension D. The relationship between thefractal dimension D, volume V and length L, can beindicated as follows:V~L D (1)356 13 th International Conference on Tribology – Serbiatrib’13

Fractal dimensions were determined using thebox-counting method which has been proven tohave higher calculation speed and more accuracyby Dougan and Shi.the structure of a neural network until it can modelthe problem in the most efficient way. Neuralnetworks are models of biological neural structures.The starting point for most neural networks is amodel neuron, as shown in Fig. 7. This neuronconsists of multiple inputs and a single output. Eachinput is modified by a weight, which multiplieswith the input value.Figure 6. Calculation of fractal dimensions with boxcountingmethodTo analyse the results we used one method ofintelligent system; the neural network. Artificialneural networks (ANN) are simulations ofcollections of model biological neurons. A neuronoperates by receiving signals from other neuronsthrough connections called synapses. Thecombination of these signals, in excess of a certainthreshold or activation level, will result in theneuron firing, i.e., sending a signal to anotherneuron to which it is connected. Some signals act asexcitations and others as inhibitions to a neuronfiring. What we call thinking is believed to be thecollective effect of the presence or absence offirings in the patterns of synaptic connectionsbetween neurons. In this context, neural networksare not simulations of real neurons, in that they donot model the biology, chemistry, or physics of areal neuron. However, they do model severalaspects of the information combination and patternrecognition behaviour of real neurons, in a simpleyet meaningful way. This neural modelling hasshown incredible capability for emulation, analysis,prediction and association. Neural networks can beused in a variety of powerful ways: to learn andreproduce rules or operations from given examples;to analyse and generalise sample facts and to makepredictions from these; or to memorisecharacteristics and features of given data and tomatch or make associations with new data. Neuralnetworks can be used to make strict yes-nodecisions or to produce more critical, finely valuedjudgments. Neural network technology is combinedwith genetic optimisation technology to facilitatethe development of optimal neural networks tosolve modelling problems. Genetic optimisationuses an evolution-like process to refine and enhance3. RESULTFigure 7. A neuron modelGraph [1-2] present relationship betweenroughness R a and hardness in specimens hardenedat different speeds at 1000 °C with both process.Roughness2502001501005004 mm/s3 mm/s5 mm/s2 mm/s55 56 57 58 59 60 61HardnessExperimental dataFitting curve with neural networkGraph 1. Relationship between roughness R a andhardness in specimens hardened at different speeds at1000 °CRoughness1400120010008006004002003 mm/s2 mm/s4 mm/s5 mm/s057,7 57,8 57,9 58 58,1 58,2 58,3HardnessExperimental dataFitting curve with neural networkGraph 2. Relationship between roughness R a andhardness in specimens hardened at different speeds at1000 °C with process of overlapping13 th International Conference on Tribology – Serbiatrib’13 357

4. CONCLUSIONThe paper presents using fractal geometry todescribe the wear of robot laser-hardenedspecimens with overlap. We use the relatively newmethod of fractal geometry to describe thecomplexity of laser-hardened specimens.REFERENCES[1] B.B. Mandelbrot. The fractal geometry ofNature. New York: W.H. Freeman, 1982, p. 93.[2] P.V. Yasnii, P.O. Marushchak, I.V.Konovalenko, R.T. Bishchak, Computeranalysis of surface cracks in structuralelements, Materials Science 46, pp. 833-839,2008.[3] P.V. Yasnii, P.O. Marushchak, I.V.Konovalenko, R.T. Bishchak, Structuraldegradation and damage caused by a system ofcracks to the steel of metallurgical equipment,Materials Science 47, pp. 798-803, 2009.[4] C.C. Barton, Fractal Analysis of Scaling andSpatial Clustering of Fractures. In: Barton, C.C.and La Pointe, P.R. (eds.), Fractals in the EarthSciences. Plenum Press. New York, pp. 141-178, 1995.[5] L.W. Fan, Y.C. Hu, T. Tian, Z.T. Yu, Thepredication of effective thermal conductivitiesperpendicular to the fibres of wood using afractal model and an improved transientmeasurement technique, International Journalof Heat Mass Transfer 49, pp. 4116–4123,2006.[6] T. Ficker, Fractal strength of cement gels anduniversal dimension of fracture surfaces,Theor. Appl. Fract. Mech. 50, pp. 167–171,2008.[7] [7] M.Q. Jiang, J.X. Meng, J.B. Gao, X.L.Wang, T. Rouxel, V. Keryvin, Z. Ling, L.H.Dai, Fractal in fracture of bulk metallic glass,Intermetallics 18, pp. 2468–2471, 2010.[8] A. Carpinteri, S.A. Puzzi, A fractal approach toindentation size effect, Engineering Fract.Mech. 73, pp. 2110-2122, 2006.358 13 th International Conference on Tribology – Serbiatrib’13

Serbian TribologySocietySERBIATRIB ‘1313 th International Conference onTribologyKragujevac, Serbia, 15 – 17 May 2013Faculty of Engineeringin KragujevacDIFFERENT WAYS OF FRICTION COEFFICIENTDETERMINATION IN STRIPE IRONING TESTS Aleksandrovic 1 , M. Stefanovic 1 , V. Lazic 1 , D. Adamovic 1 , M. Djordjevic 1 , D. Arsic 11 Faculty of Engineering, University of Kragujevac, Serbia, srba@kg.ac.rsAbstract: The sheet metal stripe ironing laboratory test has been developed to study tribologicalappearances and performance of lubricants in ironing process. Most common way for friction coefficientdetermination in the test is use of different formulas which gives relation between active forces and reactivefriction forces. In application of such formulas some difficulties occurs because of improper frictioncoefficient values, especially at small intensities of tensile or drawing forces. In this paper for literatureapproaches were analyzed and after that defining of new formula were proposed. New formula was testednumerically and experimentally. Obtained results indicated that the suggested improvements give much moreacceptable values of friction coefficient. That fact is particularly significant in lubricant evaluation process.Keywords: Thick sheet metal, stripe ironing test, friction coefficient1. INTRODUCTIONIroning is technological process which combinecharacteristics of sheet metal forming and bulkforming. Thinnig strain reach over 25%, andcontact pressure over 1000 MPa [1]. Most oftenapplies in manufacture of cylindrical geometrypieces whose depth is much bigger than diameter,and bottom thickness is bigger than wall thickness.Ironing is normally applied following deepdrawing (or extrusion) when forming high, thinwalled cans. Such cans are used for beverages,cartridge cases, high pressure cylinders, housingsfor pumps and shock absorbers etc. World annualproduction (especially for beverage cans) are morethan billion pieces [2].Of the sheet metal forming processes, ironing isone of the tribologically most severe, owing to thehigh surface expansion and normal pressure at thetool-workpiece interface. This is particularlysignificant in the case of forming of pour fomabilitymaterials such as stainless steel, high strength steel,etc. [3]. Because of that, use of proper performacelubricants is very significant. In order to quantifythe performance of the individual lubricants, adifferent simulative test methods has beendeveloped. All the tests are modelling the processconditions in ironing. It is a very convenient to usecoefficient of friction at contact surfaces change asa criterion for lubricants evaluation.For this study one of classic stripe ironing testswas chosen [4]. By analysis of acting of drawingforce, side forces and friction forces well knownformula was determined. This particular formulaestablished the connection between tool geometry,forces and coefficient of friction. The formula wasused in different researches, [4, 5, 6, 7, 8] in genuinor modified form.However, by more accurate measurements of thedrawing force was shown that formula givesnegative friction coefficient values in range of forcesmaller intensities. That fact was indicated yet inarticle [5]. That was motive for making analysis ofseveral approaches with goal to obtain moreconvenient formula appropriate for abovementioned strip reduction test.2. DEFINING OF FRICTION COEFFICIENTFigure 1 shows scheme of the stripe ironing testtooling which models the symmetrical contact ofthe sheet with the die during the ironing process.The metal strip is being placed into the holding jaw.The jaw with the sample is moving from the bottomtowards the top, by the mechanical part of thedevice. The sample is being acted upon by the side13 th International Conference on Tribology – Serbiatrib’13 359

elements with force F D , which simulate theindustrial tool die and perform the ironing. Duringthe ironing process the recording of the drawingforce is being done at over the total length of thepunch travel, by the corresponding measuringsystem.force in the sliding process starting phase. Thisnotice was given yet in article [5] where wasassumed that cause of such a disadvantage isnegligence of the forces in narrow vertical zonebetween side element inclined surfaces. Scheme offorces at fig. 2 was formed according topropositions from that study [9]. After forceanalysis friction coefficient is given by:F 2F0.25 2tgD (4)Ftg 4FDFigure 1. Stripe ironing test modelTerm (1) gives friction coefficient μ dependenceon drawing force (F), side force (F D ) and inclinationangle α and that is well-known classic formula [4].F tg2FDF 2FDtg Ftg12FD Ftg2FDFcos 2FDsin 2F cos Fsin D(1)Similar term (2) was proposed in article [6]. Ifinstead of force F is inserted F/2 term (1) wasgiven.Fcos FDsin F cos Fsin DTerm (3) is using in article [2].Fcos 2FDsin F cos Fsin D(2)(3)Previous three formulas give negative frictioncoefficient values for smaller intensities of drawingFigure 2. Force acting scheme [9]Within a framework of the same study [9]intuitively was proposed different scheme of sideforces F D acting. It assumes that at inclined surfaceacting force F D /2 and at narrow vertical surface alsothe same force F D /2. In such conditions anotherversion of previous formula was given.F2F0.25 2tgD (4a)Ftg2FDAfter analysis of the previous formulas schemeof forces in fig. 3 was formed. Based onequilibrium equation of all the forces (for contactsurfaces at both sides) in vertical direction, frictioncoefficient is given by:F (5)2 sin 22aFDcos F 21 aFD2360 13 th International Conference on Tribology – Serbiatrib’13

Figure 3. Modified force acting scheme0.80.70.60.5F D =10000 NParameter a is determining distribution of sideforce F D between inclined and small vertical contactsurface and his value is in the range 0 to 1. It wasadopted a=0.7 in this case. Parameter a influence onfriction coefficient value is very small (about 1%).Figures 4 and 5 gives comparative overview ofall the 6 formulas whereat was adopted F D =10 kN(fig. 4) and F D =0 kN (fig. 5). Inclination angle was10 o . Drawing force is linearly increasing from 0 to9500 N and lies on x axis. Clearly can be seen thatformulas 1, 2 and 3 gives unreal negative frictioncoefficient values for smaller force F intensities.Use of 4 and 4a formulas is solving thisdisadvantage, but at the sliding process beginningfriction coefficient have positive nonzero alsounreal values. Only formula 5 gives frictioncoefficient values which starts from 0. That is inaccordance with ironing process course. At smallerintensities of side force F D friction coefficientvalues are probably higher then real.=10 o Formula 2Proposed formula0.4Formula 4-a0.30.20.10.0-0.1Formula 4 Formula 1-0.2-0.3Formula 3-0.40 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000F, NFigure 4. Friction coefficient dependencies on drawing forceAs a example of formula (5) application inlubricants quality evaluation experiment giving arethe figures 6 and 7. Experimental equipment isbased on tribo model from fig. 1 and describedwith more details in [9]. Sliding process was onephase with side forces 5, 10, 15 and 20 kN. Slidinglength was approximately 60 mm at speed of 100mm/min. Stripe material is low carbon steel sheetwith 2.5 mm thickness. L2 is special dry ecologicallubricant based on wax and metallic soap.Lubricant layer was obtained by dipping into bathwith proper solution and than drying. L3 is lithiumgrease with MoS 2 .13 th International Conference on Tribology – Serbiatrib’13 361

0.80.7F D =20000 N0.6=10 o Formula 20.50.4Proposed formula0.30.20.10.0Formula 4-a-0.1-0.2Formula 4 Formula 1-0.3Formula 3-0.40 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000F, NFigure 5. Friction coefficient dependencies on drawing forceBy fig. 6 and fig. 7 comparison can be seen thatfor lubricant L3 contact pressure has no substantialinfluence on friction coefficient. In the case of0.25lubricant L3 application friction coefficient isdecreasing with side force decreasing.Lubricant: L2DC 04Friction coefficient , -0.200.150.10F D = 15 kNF D = 5 kNF D = 20 kNFD = 10 kN0.050.000 10 20 30 40 50 60 70Sliding length h, mmFigure 6. Friction coefficient dependencies on sliding length362 13 th International Conference on Tribology – Serbiatrib’13

0.25Lubricant: L3DC 04Friction coefficient , -0.200.150.10F D = 15 kNF D = 5 kNF D = 20 kNF D = 10 kN0.053. CONCLUSION0.000 10 20 30 40 50 60 70Comparative analysis of application of the fourliterature formulas for the friction coefficientdetermining in stripe ironing test wasaccomplished in the first part of this study. Threeformulas give negative unreal friction coefficientvalues for smaller intensities of drawing force inthe sliding process starting phase. For one formula(in two versions) friction coefficient have positivenonzero but also unreal values at the slidingprocess beginning. These notices are indicatingthat previously mentioned formulas are inaccurate.Different formula was suggested in the secondpart of this study. Proposed formula enables todetermine acceptable friction coefficient valuesand dependencies. After performing of trialexperiments the results are indicating that proposedformula can be successfully applied in thelubricant evaluation during chosen stripe ironingtest process.REFERENCES[1] N. Bay, A. Azushima, P. Groche, I. Ishibashi, M.Merklein, M. Morishita, T. Nakamura, S. Schmid,M. Yoshida: Environmentally benign tribo-systemsfor metal forming, CIRP Annals - ManufacturingTechnology 59, pp. 760 – 780, 2010.[2] D. Adamovic, M. Stefanovic, V. Lazic, M.Zivkovic: Estimation of Lubricants for Ironing ofSteel Pieces, Tribology in industry 26, pp. 12-20,2004Sliding length h, mmFigure 7. Friction coefficient dependencies on sliding length[3] J.L. Andreasen, N. Bay, M. Andersen, E.Christensen, N. Bjerrum: Screening theperformance of lubricants for the ironing ofstainless steel with a strip reduction test, Wear 207,pp. 1-5, 1997.[4] D. Schlosser: Beeinflussung der Reibung beimStreifenziehen von austenitischem Blech:verschiedene Schmierstoffe und Werkzeuge ausgesinterten Hartstoffen, Bander Bleche Rohre 7/8,pp. 302-306, 1975.[5] P. Deneuville, R. Lecot: The study of friction inironing process by physical and numericalmodelling, Journal of Materials ProcessingTechnology 45, pp. 625–630, 1994.[6] H.C.E. van der Aa, M.A.H. van der Aa, P.J.G.Schreurs, F.P.T. Baaijens a, W.J. van Veenen: Anexperimental and numerical study of the wallironing process of polymer coated sheet metal,Mechanics of Materials 32, pp. 423-443, 2000.[7] S. Đačić, M. Stefanović, S. Aleksandrovic, D.Adamović: Characteristics of Friction in SheetMetal Sliding with Thickness Reduction,International conference SERBIATRIB 2011,Kragujevac, Serbia, Proceedings ISBN: 978-86-86663-74-0, pp. 366-369.[8] M. Stefanovic, S. Djacic, S. Aleksandrovic,D. Adamovic: Importance of tribologicalconditions at multi-phase ironing, 34 th Internationalconference ICPE 2011, Niš, Serbia, ProceedingsISBN: 978-86-6055-019-6, pp. 503-506.[9] M. Djordjević, S. Aleksandrović, V. Lazić, M.Stefanović, R. Nikolić, D. Arsić: Influence oflubricants on the multiphase ironing process,SEMDOK 2013, Zilina-Terchova, ProceedingsISBN 978-80-554-0629-9, pp. 22-26, 2013.13 th International Conference on Tribology – Serbiatrib’13 363

Serbian TribologySocietySERBIATRIB‘1313th International Conference onTribologyKragujevac,Serbia, 15 – 17 May 2013Faculty of Engineeringin KragujevaccA NANOMECHANICAL APPROACHON THE MEASUREMENTOF THE ELASTICPROPERTIESOF EPOXY REINFORCEDCARBON NANOTUUBE NANOCOMPOSITESG. Mansour 1 , D. Tzetzis 2 , K.D. Bouzakis 31 Aristotle Universityof Thessaloniki, Department of Mechanical Engineering, Greece, mansour@eng.auth.gr2 International Hellenic University, Greece, d.tzetzis@ihu.edu.gr3 Aristotle University of Thessaloniki, Department of Mechanical Engineering, Greece, bouzakis@eng.auth.grAbstract:The mechanical behaviourr of nanocomposite materials with multiwall carbon nanotube (MWCNT)reinforcements is investigated in thepresent paper. Epoxy nanocomposites with different weight percentagesof carbon nanotubes have been characterized following tensile tests andd nanoindentations. The objective ofthis work was to investigate the efficiency of the reinforcement provided by nanotubes and to examine theagreement between the mechanical properties of the epoxy nanocomposites obtained via a macroscale andnanoscale experimental methods. Higher increase in modulus was accomplishedad at weightt fraction ofnanotube reinforcementof 1%. Themodulus as measured by the tensile tests differed an average of 18%with the results obtained from the nanoindentaations, however by utilising a properr calibrationmethod thedata were corrected resulting to only a 3% difference. Themodulus results obtained from the experimentswere compared with theHalpin-Tsai model andd with the Thostenson-CChou model accounting for f the outerlayer interactions of thenanotube with the hosting matrix. Arelatively good agreement was found betweenthe models and the experiments.Keywords:Properties,Nanoindentation Testing, EpoxyMicroscopyNanocomposites, Multiwall Carbon Nanotubes, Elastic1. INTRODUCTIONEpoxy nanocomposites using carbon nanotubes(CNTs)have been intensivelyinvestigated,following the successful synthesis of CNTss in1991[1].CNTs have attracted considerable attentiondue to their uniquemechanical,surface,multifunctional properties and strong interactionswith the hosting matrixmainly associated to theirnano-scale features. Recent experiment shaveshown remarkable enhancements in elastic modulusand strength of polymer composites with anaddition ofsmall amounts of CNTs [2,3]. Amongthe various studies incorporating CNTs, Loos et. al[4] have investigated the matrix stiffness rolee ontensile and thermal properties of carbon nanotubeepoxy reinforced nanocomposites. They haveshown thatt the addition of a small amountt ofSWCNTs (0.25 wt.%)in soft matrices, greatlyincreased Young’s modulus and tensile strengthh of364such nanocomposites. The results showed that thetensile properties of soft epoxy matrices are muchmore influenced by thee addition of carbonnanotubes thann stiffer ones. Also, Kimet. al. [5]studied the effects of surface modification onrheologicaland a mechanicalpropertiesofCNT/epoxy composites. The CNTs were modifieddby acid treatment, plasmaa oxidation, and aminetreatment. The surface modified CNTs were wellldispersed in the t epoxy matrix and had stronginterfacial bonding with the polymer matrix. Thenanocompositecontainingg the modified CNTsexhibited higher storage and loss moduli and shearviscosity than those withh the untreated CNTs,because the surface treatments provide morehomogeneousdispersion of CNTs and strongerrinteraction between the CNT and the t polymermatrix. Gojny et al. [6] have investigated theinfluence of different d types of CNTs on themechanicalproperties of epoxybased13 th International Conference C onn Tribology – Serbiatrib’13

nanocomposites. The influence of filler content, thevarying dispersibility, the aspect ratio, the specificsurface area and the functionalisation on thecomposite properties was correlated to theidentified micro-mechanical mechanisms. Theresults showed that the produced nanocompositeshave enhanced the strength and stiffness along withan increase in fracture toughness.Despite the huge amount of experimental dataavailable in the literature there are still debatableresults concerning the elastic property and strengthof such nanocomposites. This is due to thecharacteristic difficulties inprocessing the CNTnanofillers in polymer systems, and thereby areliable theoretical correlation of the experimentaldata is still lacking. This is because thereinforcement capability of the CNTs in apolymeric matrix will depend on their amount aswell as on their arrangement within the matrixwhich plays a fundamental role in the load transfermechanism.On the other hand, in context with the highprices of the CNTs, there is a requirement forprocedures using small samples of nanocomposites,compared to the standard tensile test samples, inorder to acquire mechanical property data on whichtheoretical predictions can be based. Therefore,alternative approaches have been utilised fordetermination of the mechanical properties ofnanocomposites [7]. Nanoindentation is a simplebut powerful testing technique, which can provideuseful information about the mechanical properties(such as elastic modulus and hardness) of materials.It has been proven that the nanoindentationtechnique is the most accurate method forevaluation of the effect of carbon nanotubes on thedeformation behaviour [8].The aim of this work was to investigate themechanical properties of MWCNTpolymercomposites by nanoindentation. Elastic modulusand hardness are the properties measured by thenanoindentation technique and these werecompared by results obtained by uniaxial tensiletests as well as with popular arithmetic predictions.The morphology of the nanocomposites wasinvestigated by using a stereomicroscope andscanning electron micrographs.2. MATERIALSThe epoxy matrix investigated was a lowstrength bisphenol A and epichlorohydrin epoxyresin (Epikote 816, Hexion Specialty Chemicals)containing an added proportion of Cardura E10P(glycidyl ester of neodecanoic acid) as a reactivediluent. The hardener was amine curing agent(Epikure F205, Hexion Specialty Chemicals). Thenanofiller used, was multiwall carbon nanotubes(MWCNT’s). The carbon nanotubes were used asreceivedwithout any further treatment.Epoxy-based nanocomposites were prepared bymixing the nanotubes with an appropriate amountof the neat epoxy resin using an ultrasonic stirrer(Bandelin Electronic GmbH, model HD2200) for5min followed by high mechanical mixing. Thiswas followed by the addition of the hardener andfurther mechanical mixing. The mixture wasdegassed and then cast into release-agent-coatedspecial formed moulds in order to form the platesfor specimen fabrication. The plates were left tocure for 48hours followed by 2 hours post curing at90ºC. As a result, a series of specimens withnanofiller contents of 0.5% and 1% by weight wereobtained. Small specimens of 10x10mm were cutfrom the plates and polished in order to make thenanoindentation specimens.3. EXPERIMENTAL PROCEDURES3.1 Tensile TestsTensile tests were performed at roomtemperature (23°C) on a Zwick Z010 (Zwick,Germany) universal testing machine at a constantcrosshead speed of 1 mm/min. The measurementsfollowed the EN ISO 527 testing standard usingdumbbell shaped specimens. The specimens havinga 4 mm thickness were machined from the mouldedplates using a Computer Numerical Control (CNC)milling machine. The overall length of dumbbellspecimens was 170 mm. The length and width ofnarrow section were 10 and 4 mm, respectively. E-moduli were calculated within the linear part of thestress-strain curves. All presented data correspondsto the average of at least five measurements.3.2 Nanoindentation TestingNanoindentation tests involve the contact of anindenter on a material surface and its penetration toa specified load or depth. Load is measured as afunction of penetration depth. Fig. 1 shows thetypical load and unloading process showingparameters characterizing the contact geometry.This schematic shows a generic viscoelastic-plasticmaterial with the loading OA, and unloading AB´segments. The plastic work done in theviscoelastic-plastic case is represented by the areaW1 (OAB´). The area W2 (ABB´) corresponds tothe elastic work recovered during the unloadingsegment. In the case of purely elastic material, theunloading curve is a straight line (AB) and h r =h max(W2=0). In this case, penetration depth is thedisplacement into the sample starting from itssurface. Numerous details on the nanoindentation13 th International Conference on Tribology – Serbiatrib’13 365

measurement process inrelation to polymers cann befound in references [9-11].In the current workthe nanoindentations wereconducted on a Fischerscope H1000 device, with aresolution of 0.1 mN. The indenter has a Berkovichdiamond tip(the tip shape is a three-sided pyramid,with a triangular projected geometry and anincluded angle of 65.3°; tip radius 20 nm). Thenanoindentations madeon the surface of thespecimens appeared asan equilateral trianglee asshown in Fig. 2. Prior to an indentation, itheindenter was driven, under computer control,toward the specimen surface.After contact, theindenter was driven into the surface, to a depthh ofaround 0.6μm, at a constant loading rate of0.15mN/s, until a peak load of 4.8mN was reachedand subsequently the indenter was unloaded usingthe same rate. This peakload was then t held forr 5 s(in order to minimizethe effect of viscoelasticdeformationof the specimen, notably creep, onproperty measurements)and then the indenter wasunloaded, toa load of zero.The calculation method to determine themodulus and hardness of the fumed silica epoxynanocomposites was based on the work of Oliverand Pharr [12]. According to this method, thenanoindentation hardness as a function of the finalpenetration depth of indent can be determined by:Where β is a constant that depends on the t geometryof the indenter. For the Berkovich indenter, β=1. .034. The specimen elastic modulus (E(s ) can thenbe calculated as:1 2 1 1 (5)(6)Where Ε i,s , and ν i,s are the elastic modulus andPoisson’s ratio, respectively, for the indenter i andthe specimen. For F a diamond indenter, E i is 1140GPa and ν i is 0. 07. (1)Where P max is the maximum appliedload measuredat the maximum depth of penetration (h max ), A isthe projected contact area between the t indenter andthe specimen. For a perfect Berkovich indenter, Acan be expressed as a function of the contactindentation depth h f as:Figure 1. Schematic of indentation load-depth data of aviscoelastic-plastic where h max x is the maximum depth, h eis the elastic depth rebound, h r is the residual impressiondepth, h a is the displacement of the surface at theperimeter and h f is the contact indentation depth. 3√3 65.3 24.5(2)The contactt indentation, h f , can be determined fromthe following expression: where ε isa geometric constantt ε=0.75 for apyramidal indenter, S isthe contactt stiffness whichcan be determined as the slope of the unloadingcurve at themaximum loading point, i.e. The reducedd elastic modulus E r is given by:(3)(4)Figure 2. Schematic of thee loading and unloadingsurfaces of an a indentation (half-section)) with thecorresponding indentation depths.Thespecimen’ss hardness H and elastic modulus E swere obtained from the set of equations given above.36613 th International Conference C onn Tribology – Serbiatrib’13

4. RESULTS AND DISCUSSION4.1 MorphologyMicroscope images from the of cured MWCNTepoxy nanocomposites are shown inFig. 3.surface of the nanotube n aggregatesthematrix itselfencapsulates the nanovoids inside the agglomeratednanotubes.4.2 Tensile TestsThe stress–strainbehaviour of thenanocompositess under tension is shown in Fig. 4.Thespecimenss revealed a characteristic plasticbehaviour.Figure 3. Stereoscope images of epoxynanocompositeswith nanotube concentrations of: a) 0. 5%wt, b) 1%wt.The nanotubes showsignificantt agglomerationwhich is more pronounced in thecase of 1% %wtnanotube loading due to strong van der Waalsinteractionsleading to relativelyinsufficientdispersion despite the ultrasonic application andd thesubsequent mechanical mixing. An aggregateformation could only be achievedin the epoxymatrix while these aggregates at certain areasattract eachother forming greaterr assembliess asseen from the images. The structure of nanotubeclusters observed in all specimens was very similarirrespectivee of the percentage loading, thoughslightly higher densities of particle clusters areevidently for the 1%wt nanocomposites.Theprocessing of the epoxy nanocompositesbyultrasonic mixing produced a frothy and viscousdark solution that made the degassing procedurerelatively difficult. Also, it is believed that thenanovoids could not beeliminated in total despitethe degassing procedure and as during the curingperiod the epoxy matrix can reactt only with the13 th International Conference on Tribology – Serbiatrib’13Figure 4.Typical uniaxial tensile stress-strain curves ofepoxyy reinforced nanocomposites.The additionn of the MWCNTs slightly increasedthe strength ass reported inn other studies [2]. Thefracture surfaces of the tensile specimens wereeexamined using a scanning electron microscope.Thepure epoxyy resin samples showed characteristiccriver lines and a smooth surface as shown in Fig.5(a). This typee of fracture behaviour is i typical ofbrittle epoxy surfaces indicating low resistance r tospontaneouscrack propagationwhichwasmonitored during tensile testing of specimens.Fig. 5(b) shows the fracture behaviour obtainedfromthe MWCNT nanocomposites.In certainplaces the fracture is a mirror-like which reflectsthat the nanotubes were not dispersed evenly. As aresult, when the external tensile force was applied,debonding mayy have occurred at these areas. Also,the formationss of clusters produced a severelytortuous surfacee with certainn yielding regions.During the applied macroscopic tensile stresssthe local stresses around the aggregates ofMWCNTs (Fig. 5c) triggered yielding of the epoxy.Additionally, before b the onset of a critical crack,numerous microcracks were formed onthe tensileetestt specimen as a visually monitored during testing.Theaggregates may have induced crack branchingwhich in turn may have triggered multiple locallyielding of the matrix. mThe nucleation of the cracks may havedeveloped either within network of theclusters ofMWCNTs that have not infiltrated withepoxy resinor at the aggregates’ interfaces.367

4.3 NanoindentationFig. 6 illustrates typical load–displacementcurves of indentationsmade at a peak indentationload of 4.8mN on the pure epoxyresins and theMWCNT nanocomposites. No cracks were formedduring indentation as no steps orr discontinuitieswere found on the loading curves.The indentation depths at the peak load rangefrom around 0.5 to 0.6 μm. Lower indentationdepths areobserved for the t MWCNTnanocomposites as compared with the pure epoxysamples. The hardness and elastic moduluss isincreased as the concentration is increased. It iswell documented in the literature that the elasticmodulus has an increasing trend ass the percentageloading of MWCNTs is increasing [ 13].There is a significant difference in elasticmodulus as obtained from the nanoindentationtesting compared to theone of thetensile testss asshown in Table 1.Clearly the elastic modulus obtained from thenanoindentation testingtechniquee was 14-18%higher than the one obtained from the tensile tests.intrinsic errorss may leadd to results which aredifficult to explain in the case of softer, viscoelasticcsurfaces like the solidified epoxy resin r in thecurrent case.Table 1. Elastic moduli values as a derived fromexperiments.Material0% CNT0,5%CNT1%CNTE tensile(GPa)3,3 ± 0,124,5 ± 0,154,64 ± 0,18E nanoindentati ionE(GPa) modified (GPa)3,9 ± 0,122 3,375,22 ±,0,18 4,575,31 ±,0,22 4,75The process of nanoindentation measurementts isa relativelycomplicateprocedure, , especially forpolymeric materials asit has been reportedd invarious studies [11, 14]. The system compliancemay be tooo low to measure the material responseproperty for ‘soft’ materials like the epoxy resin.Also, the nanoindentation techniquee is based onn theelastic behaviour of the test material; thereby theviscoelasticc behaviour may cause an error in thecalculation of the elastic modulus. Moreover, thereare uncertainties in tip shape calibrationthatdirectly relate to the area functionn (A) whichh ismaterial dependent in most cases. The tip defect,which is always presentdue to technical limitationsin the fabrication of the indenter, may greatly affectthe assessment of the mechanical properties of f thetested surface at the first material layers. This isexacerbatedby the calibration procedure whichrequires a series of indentations upon the referencematerial at various depths and produces an intrinsicblunting effect on the calibrated tipat the deepestpenetrations, which donot correspond with thetip/machinebehaviour at theshallowestindentationsand so the final area functionextrapolatedmay not be exact. Therefore, the368Figure 5. SEM micrographs m of typical fracture surfacesof a) pure epoxy resin, b) MWCNT, c) MWCNT athigher magnificationAlso, for an epoxy resinn material, pile-ups and adistorted surface are usually observedaround thecrater of the nanoindentation. It is evident thereforeethat the typical calibration procedure whichhinvolves calibration on a reference material of awell-defined elastic modulus such as fused silica isnotsuitable for polymer materials. This isdocumented byy the observed differences in elasticmodulus between the nanoindentationresults andthe uniaxial tensile test measurements.13 th International Conference C onn Tribology – Serbiatrib’13

Figure 6. Loading and unloadin versuss depth profiles ofpure epoxy resin and MWCNT nanocomposites.Figure 7.Hardness versus of pure epoxy resin anddMWCNT nanocomposites.Nevertheless the elastic modulus results asmeasured byboth techniques revealed similar trends.Subsequently as suggested by other researchers [10,11] a material depending calibrationprocedure hasbeen utilized for the current measurements. Usingequations (1-6) from the Oliver and Pharr [12, 15], , themodified area function related to indentation depthwas obtained using the elastic modulus from a tensiletest of the pure epoxy resin which was 3.3 GPa. Usingthe new calibrated area function the elastic moduli ofthe nanocomposites wascalculated. The result off theelastic modulus based onthe modifiedd area function ismarked as modified nanoindentation. Clearly, themodified elastic modulusvalues shown in Table 1 arein good agreement with the elastic modulus fromtheuniaxial tensile tests. For MWCNTs nanocompositesthe elastic modulus is increasing ass measured fromboth the tensile tests and from the nanoindentationexperimentswith the proposed calibration technique.Fig. 7 also shows the hardnessof thenanocompositesas a functionMWCNTconcentration. In agreement withthe previousoutcomes the hardness follows the elastic modulustrend and increases in the case of MWCNTs ass theconcentration increases from 0.5%wt to 1%wt. Itshould be noted that when measuredat small scales,13 th International Conference on Tribology – Serbiatrib’13the hardness is larger than at largerr scales. Anexample of this phenomenon is theso called‘indentation sizee effect’ which can be observed as anincrease in hardness with decreasing indentation depth[16]. This effectt complicatess the determination of thematerial hardness at low indentation depths, given thesmall remaining impression. However, , the resultssobtained in the current study lie within valuesobtained from other studiess investigating MWCNTepoxy nanocomposites [17, 18].The hardness of the carbon nanotubes themselvessis higher than the t one fromm the epoxy resin r therebythisexplains the t small increase noticed in thepresented results.5. ELASTIC MODULUS PREDICTIONSDespite the outstanding mechanical properties ofnanotubes, thee nanocomposites involving suchnanofillers exhibit a very limited improvement ofmechanical performances,if compared to conventionaladvanced composites.Thisopposingbehavior canbeexplained byy considering that the reinforcingcontribution of MWCNTs iss yieldednot only by theiramount within the material, , but also bythe state ofdispersion, orientation,shapee and number of contactswithin the matrix system. All these features play acriticalrole on the t final reinforcement enhancement,andthey shouldd be taken into account if i possible inorder todevelopp reliable models for prediction ofnanocomposite effective e properties.The classicall micromechanics approaches for shortfibre reinforcedd compositess were employed in thiswork in order todevelop predictive models for theMWCNT nanocomposites.AA popular and widelyadopted model to predict the stiffness of MWCNTsnanocompositess is the Halpin-Tsaimodel.TheHalpin–Tsai model m [19] iss widely used in manyliterature references.It is based on a force f balancemodel and empirical data and it isusedwidely formacroscopic composites.Forr the moduli of randomlyoriented MWCNTs in the epoxy matrix, theHalpin–Tsai model may predictthe elastic modulusof the nanocomposites, E N NC,which is governed bythe following set of equations: 3 1 8 1 5 12 8 1 1 (8) 369(7)

2 12(9)(10)Where E MW WCNT and E m are the Young’s moduluss forthe MWCNTs and matrix respectively while v MW WCNTand l/d are the volume fraction andaspect ratioo ofMWCNTs respectively. From Eq. 7 it can be seenthat E NC strongly depends on the geometry of f theMWCNTs such as theiraspect ratio. The lengthh ofthe fibres ranges from 1-25μm while variousdiameterswere measuredas seen in Figure 8. TakingΕ MWCNT = 1GPa whichh is muchgreater thanE m =3.3GPa the predicted values versus the volumefraction of the nanotubes of E NC based on Eq. 7 isshown in Fig. 9.It can be seen that the Halpin–Tsaiformula for d=5nm gives a slight different valuee forE NC compared to the ones measured from thenanoindentation using the calibrationprocedure.Figure 8.Measurementss of the outer diameter of theMWCNTs.Figure 9. Comparison of the experimental modifiednanoindentation results with the Halpin-Tsai anddThostenson-Chou models.Thostenson and Chou [20]modifiedd the HalpinTsai theory towards its applicability tonanotubereinforced composites. Thostensonand Chouconsidered that, in the case ofMWCNT, only theouter shell would w carryy the load as logicallassumption off the relativelylow bonding withinner layers. According A to this assumption, theeffective MWCNT elasticmodulus was evaluatedby considering the application of all loads only totheouter crosssection(outer diameter andgraphite layerr thicknesss which iss taken ast=0.34nm). Eq. 11 has been derived in orderrcalculate the maximumobtainableE NC for acomposite withh aperfect distribution of the CNTsandimpregnation within the epoxy. Predictionscomputed by using Thostenson-Chou modelshow a reduced level of efficiencyfor largediameters while for d=3nm the prediction iscompared welll with the experimentally deriveddmodulus for 1% %wt (0.56%vf) MWCNTS.This occurss despite thee fact that as shown bytheSEM investigations there are locally highernanotube concentrationswithin the composite.Accounting for any errors associated with theexperimentallyy derivedvaluesthe results have tobe interpretedd as a lower boundary of theobtainable moduli.Additionally, the presence ofvoids developed during mixing the hardener withtheMWCNT/ /epoxy-suspension via ultrasonicmixing and mechanicall stirring may haverestrainedtheecompositesfrom their fulllmechanical performancepotential. The highhviscosity disabled a fully adequate degassing ofthenanocomposite with voids remaining in thematrix.The initial failuree had been caused bythese voids and expressedd itself in the t reduceddfracture strain in the tensile tests.It is clear therefore t that despite the fact thattthemodels used in this work are valuable toolstowards the prediction of the elastic modulus ofthenanocomposites they do not totally correctlyrepresent thevarious issues associated with thecontent, morphology and type of f nanotubesincorporating a variety of f diameters and lengths.Also, and most importantly they consider thenanotubes agglomerated-free compared with experimentaldata.whichh may bemisleading when 37013 th International Conference C onn Tribology – Serbiatrib’13

3 8 12 1 4 2 4 2 (11)6. CONCLUSIONThe nanoindentation technique has beensuccessfully utilised in order to study themechanical properties (i.e. hardness and elasticmodulus) of MWCNT/epoxy nanocomposites. Theindentation results revealed that the hardness andmodulus of the nanocomposites increase withhigher MWCNT concentrations. The elasticmodulus data obtained by nanoindentation arecomparable with those obtained by tensile testingwhen a suitable material calibration is applied. Theresults verify the capability of the nanointendationinstrumented technique to characterize themechanical properties of polymer nanocompositesusing small sample amounts. Elastic moduluspredictions using the Halpin-Tsai model haveshown comparable results with the experimentaldata, while the Thorsten and Chou model providedgood predictions by taking into account the outerlayer of the nanotubes.REFERENCES[1] S. Iijima, Helical microtubules of graphitic carbonNature, Vol. 354, pp. 56-58, 1991.[2] K. Sharma, M. Shukla, Experimental study ofmechanical properties of multiscale carbon fiberepoxy-CNTcomposites, Advanced MaterialsResearch, Vol. 383-390, pp. 2723-2727, 2012.[3] H.U. Zaman,P.D. Hun, R.A. Khan, K.B. Yoon,Effect of Multi-walled Carbon Nanotubes onMorphology, Mechanical and Thermal Properties ofPoly(ethylene Terephthalate) Nanocomposites,Journal of Applied Polymer Science, Vol. 128, No.4, pp. 2433-2438, 2013.[4] M.R. Loos, S.H. Pezzin, S.C. Amico, C.P.Bergmann, L.A.F. Coelho, The matrix stiffness roleon tensile and thermal properties of carbonnanotubes/epoxy composites, J Mater Sci, Vol. 43,pp. 6064–6069, 2008.[5] J.A. Kim, D.G. Seong, T.J. Kang, J.R. Youn,Effectsof Surface Modification on Rheological andMechanical Properties of CNT/Epoxy Composites,Carbon, Vol. 44, pp. 1898–1905, 2006.[6] F.H. Gojny, M.H.G. Wichmann,B. Fiedler, K.Schulte, Influence of nano-reinforcement on themechanical and electrical properties ofconventional fibre reinforced composites,Composites Science and Technology, Vol. 65, pp.2300–2313, 2005.[7] P.M. Nagy, D.Aranyi, P.Horváth, P.Pötschke, S.Pegel,E.Kálmán, Nanoindentation Investigation ofCarbon Nanotube–Polymer Composites, InternetElectron. J. Mol. Des., Vol. 5, pp. 135–143, 2006.[8] A.K. Dutta, D. Penumadu, B. Files,Nanoindentation testing for evaluating modulusandhardness of single-walled carbonnanotubereinforced epoxy composites, J. Mater.Res., Vol. 19, pp. 158–64, 2004.[9] M.R. VanLandingham,Review of instrumentedindentation, J. Res. Natl. Inst. Stand. Technol., Vol.108, pp. 249-265, 2003.[10] B. J. Briscoey, L.Fiori, E.Pelillo, Nano-indentationof polymeric surfaces, J. Phys. D: Appl. Phys., Vol.31, pp. 2395–2405, 1998.[11] M.R. VanLandingham, J.S.Villarrubia, W.F.Guthrie,G.F. Meyers, Nanoindentation of polymers:An Overview, Macromol. Symp., Vol. 167, pp. 15-43, 2001.[12] W.C. Oliver, G.M. Pharr, An improved technique fordetermining hardness and elastic modulus using loadand displacement sensing indentation experiments, JMater Res, Vol. 7, pp. 1564–1583, 1992.[13] A. Montazeri,J. Javadpour, A.Khavandi, A.Tcharkhtchi,A.Mohajeri, Mechanical properties ofmulti-walled carbon nanotube/epoxy composites,Materials and Design, Vol. 31, pp. 4202–4208, 2010.[14] M. Sánchez, J. Rams, M. Campo, A. Jiménez-Suárez, A.Ureña, Characterization of carbonnanofiber/epoxy nanocomposites by thenanoindentation technique, Composites: Part B,Vol. 42, pp. 638–644, 2011.13 th International Conference on Tribology – Serbiatrib’13 371

[15] W.C. Oliver, G.M. Pharr, Measurement of Hardnessand Elastic Modulus by Instrumented Indentation:Advances in Understanding and Refinements toMethodology, J Mater Res, Vol. 19, No. 1, pp. 3–20,2004.[16] C.S. Han, Influence of the molecular structure onindentation size effect in polymers, MaterialsScience and Engineering A, Vol. 527, No. 3, pp.619-624, 2010.[17] Sajjad M., B. Feichtenschlager, S. Pabisch, J.Svehla, T. Koch, S. Seidler, H. Peterlik, G.Kickelbickd, Study of the effect of the concentration,size and surface chemistry of zirconia and silicananoparticle fillers within an epoxy resin on thebulk properties of the resulting nanocomposites,PolymInt, Vol. 61, pp. 274–285, 2012.[18] P.M. Nagy, D. Aranyi, P. Horváth, P. Pötschke, S.Pegel, E. Kálmán, Nanoindentation Investigation ofCarbon Nanotube - Polymer Composites,InternetElectronic Journal of Molecular Design, Vol. 5, pp.135–143, 2006.[19] J.C. Halpin, J.L.Kardos, The Halpin-Tsai equations:A review,Polym. Eng. Sci., Vol. 16, pp. 344–352,1976.[20] E.T. Thostenson, T.W. Chou, On the elasticproperties of carbon nanotube-based composites:modeling and characterization, J Phys D: ApplPhys,Vol. 36, pp. 573–82, 2003.372 13 th International Conference on Tribology – Serbiatrib’13

Serbian TribologySocietySERBIATRIB‘1313th International Conference onTribologyKragujevac,Serbia, 15 – 17 May 2013Faculty of Engineeringin KragujevaccSOMETRIBOLOGY STATE TESTS OF “EPDM” RUBBERBASEDONLABORATORY EXPERIMENTATIONSAbhijit Mukhopadhyay**Associate Professor, Mechanical Engineering Department, Jadavpur University, Kolkata-700032,India,m_obiji@yahoo.comAbstract: Rubber is very useful and suitablee material for a wide variety of engineeringg and otherapplications. Use of rubber as engineering material is not new. However, in the recent time, itss applicationis gaining importance due to several other reasons. Recent and renewed researches on rubber materialreveal its suitability forvaried engineering applications. Several researches are also going onto enhanceethe propertyrequirements of rubberr for differentt applications.Rubber possesses large elasticity compared to metals, has greater damping capability, high internal frictionand can accumulate energy greater than that of steel or other metals. During thee deformation of rubberrmaterial, e.g., by compressive force, , internal damping of thematerial leads to energydissipation. This is thecause of hysteretic friction of rubber. Friction of rubber is of great practical importance at the same time ithas many disadvantagess too.Amongst various rubber Ethylene Propylene Diene Monomer (EPDM) rubber emerges as a dominantelastomer for major engineering applications in automobiles, constructions,electric andelectroniccindustries etc. The major properties of EPDM are its outstanding heat, ozone o and weather resistance ability.The resistance to polar substances and steam are also good.In automobiles EPDM rubber has a common use as seals. This includes door seals, window seals, trunkseals and sometimes hood seals. Frequently these seals arethe source of noise duee to the movement of thedoor versuss the car body. This is due to frictionn between the EPDM rubber parts and the mating surfaces.Thus the contact iteration between the t rubber sealing and the indenting object o must be known to optimize theperformance of rubber sealing. It I, however, need less tomention that the behavior of any viscoelasticcmaterial is very difficultto be predicted.In the present work various tribo-characteristics of EPDM rubber of o different hardness are evaluatedutilizing thelaboratory test facilitiess available in the frame work of the mechanical engineering department,production engineeringg department and other specialization of the Jadavpur University, Kolkata.Compressiontests have been carriedout using ‘Instron’ to determine the flow behavior of EPDM rubber ofdifferent hardness values in dry as well as in different lubricated conditions. The flow behaviorr like load -elongation curves, true stress - elongation curves true stresss - true strain curves have been drawn from theexperimental data. Abrasive wear behavior has been evaluatedusing a two-body abrasion tester and the patternabrasion hasbeen appraised through SEM/EDAX X study. It hasbeen proposed to study further the wear loss usinga pin-on-plate (POP) type tribometer and conduct t fretting weartest on the same s material, that is, EPDM.Experimental results revealed that the hardnesss values of EPDM rubber had significant effectt on the flowbehavior and wear characteristics.The hardness, again, depends on the carbon black (CB) concentration.Thus it canbe stated that the flowbehavior can be governed by controlling the CB concentration in theEPDM rubber. The results of different tests followed by comparativeanalysis have been furnished in the‘result and discussion’ section of this paper.Conclusion has been drawn accordingly, highlighting someof the important tribo-characteristics of EPDMrubber as well as shedding light on various possible areas of further researches those should beundertakennin the futureto come.Keywords: EPDM, compression, flow behavior, abrasion, SEM/EDAX.13 th International Conference on Tribology – Serbiatrib’133733

1. INTRODUCTIONTribological studies of rubber like materialss arenot new. However, in this recent time, being fuelledby several new researches and development inn thefield of viscoelastic materials as well as duee tovarious property requirements of rubber for severalengineering, domestic,sports and otherapplications, the performance evaluation of rubberis becomingvery demanding and gaining renewedresearch interest in different parts off the globe.Propertyprediction is the other driving force[2,17]. It means that theproperty of f engineeringg andother materials shouldbe predictable and thereshould havesome useable model in that regard. It is,however, needless to bementioned that t the propertyof any viscoelastic material, like rubber, is veryhard to be predicted. The frictionand wear database of rubbers are alsonot very promising due tothe fact that the rubbers used in such tribotestss arenot characterized adequately[1].Test configurations,parameter selections, experimentall conditions areall important factors which should be standardizedbefore comparing the tribotest data generatedd byvarious agenciesorresearchers. All thesenecessitate further study and iteration of tribotestdata for rubber to be used as engineering or othermaterials.Ethylenee Propylene Diene Monomer (EPDM)rubber is widely used as seals in automobile door,window, hood and otherparts. Theyare subjected towear and tear due to pressure, vibration, frictionn andexposure to extreme conditions of atmosphere.Though EPDM has outstandingheat, ozone,weather resistance ability and resistance to anypolar substance as well as steam is also very good,still some realistic tribotest data are yet too bedeveloped.In the present workflow behaviors of EPDMrubber of different hardness have been evaluated.EPDM specimens have been compressed d inbetweenflat MS platens andstress-strainrelationship, specific energy and loss factors havebeen computed subsequently forr this purpose.Similarly wear characteristics have been studiedd in atwo-body abrasion testing machine. SEM/EDAXstudies havealso been made to appraise the patternabrasion, immediately after two-body abrasiontesting.Table 1. The recipe code of different hardness of EPDM.IngredientShore Hardnesss55Å 5 60Å 70Å 80Å85ÅEPDMZnOSt.AcidPEG 4000FEF 550P Oil(2500)100511.280130100511.2130110100511.2160100100511.217090100511.218080SulpherHBSZDBSTMT1 11 11 1.51 0.71.21.51.20.71.2 1.21.5 1.51.2 1.20.7 0.7However it is i needless too mention that the actualproportions of various ingredients are trade secretandstrictly a ‘not-disclosedd grade’. Theabove tableis thus only a closely indicative one.Basic ingredients were pre-mixed for 6 minutes at a rampressure of1000 psi. Curatives were thenn added to the pre-mixedmaterials on a two t roll laboratory mill (330 × 150)at room temperature. Curatives are required toenhance various properties. The mixing time wasapproximately 101minutes. A constant friction f ratioof 1:1.25 was maintained m during rolling. .in a K4/2A-MK3(Alfred Herbert)Processingcharacteristics including optimummcuretime (t 90 ) and torque difference (Δm = Mh –M l ) were determinedwithOscillatingDiscRheometer equipped with computer assisted dataaacquisition system and supported by‘Rheosoft’software. Standard procedure as is observed in [5]andothers were followed in that regard but themachine used for f the purpose is Indiann one andspecific procedural steps for the said machine wereefollowed accordingly. M h and M l are high and lowMooney (torque) respectively. The torque wasmonitored as a function off time and the t optimummcuretimes were recorded from the correspondingrheographs, onee such graph is shown in Figure 1.2. EXPERIMENTALThe EPDM rubber specimen for the tribo tests inthis work were prepared in thelaboratoryy ofNational Engineering Limited (Rubber Division),Kolkata, following the recipe code as mentioned inTable 1.Figure 1. Torque-vs.-Time rheograph of an aspecimen.EPDM37413 th International Conference C onn Tribology – Serbiatrib’13

The material after qualifying the rheometricanalysiswas ready for molding operation. Shortcylindrical specimen ofdiameter φ (16.5±0.5) andheight h (12.5±0.5) were then prepared in the steamheated hydraulic press at a pressure of 3000 psi i andtemperaturee 150 0 C. Thematerial iss compressed inthe press for approximately 10 minutes. Themolding operation has been carriedout as perr IS:3400 (part-X) - 1977 specification.The extra spewof materials have beentrimmed by scissor aftermolding to give the specimen proper shape. Thedimension, specific gravity and shore hardnessvalues of all the samples have been measuredaccordinglyusing appropriate measuring tools andinstrumentss for the respective parameters.2.1 Compression testEach test specimen was placed axysymetricallyin between two flat mild steel platens. Much carehad been taken in such placement toensure an evenforce distribution on both the faces of the specimen.The required compressive load wass provided byy anInstron (model 8801; serial no. K 2342 with‘Dynacell’ load cell, made in England. Maximumworking pressure: 207 bar; dynamic load capacity:± 100 KN).The machine, as shownin Figure 2, isequipped with ‘8800: Instron SAXV9.3’ softwarebased data acquisition system. Only one fatiguecycle had been utilized at a frequency of 0.0055 Hzfor the application ofcompressive load on thespecimen. The height of each cylindrical specimenwas reducedd by 65%. Each test hadbeen replicatedtwice to observe the repeatability of the process.The compressive load followed byload relaxationdata had been recorded and later utilized to plott theloading andunloading curves.friction, during compression, in between the dieandwork piece mating surfaces leads to thedevelopment off an undesirable phenomenon knownnas ‘barreling’ or o ‘pan caking’. The metal spreadssover the die surface to increase its diameter when itis compressed in between two die halves. Frictionalforce opposes the t outward flow of metal near themating zone of f work piece and die halves. But thematerial at thee mid height of the specimen s isabsolutely free to flow in an out ward direction.This is the basic explanation of barreling [3]. Onesuchbarreling during compression under fixedcontact, using sand paper, is shown in i Figure 3.This undesirable frictional force may bereduced byusing some suitable s lubricant. Theeffect ofcompressive load on EPDMM specimen of differenthardness in dryy working condition, that is, withouttanylubricant, was w appraised by the present author[4]. Similarly, the flow characteristicss of EPDMrubber of different hardnesss under compressive loadin the presence of some lubricants were studied bythe present author [6]. It is not out of place tomention here that to reduce frictionand wearbetween two mating m and interacting surfaces somefilmof solid, liquid or gas is applied which isconsidered as lubricant[7] ]. Selectionof properrlubricant(s) depends on several considerations andshould be judicious [8]. Inn this experimental worksome lubricants were selected based on theliterature surveyy and considering the practical workenvironment [9-12]. The characteristicss studied forthispurpose includes i stress-strain relationships,specific energy and loss factors.Figure 3. Compression of EPDM specimenusing fixedcontact.Figure 2. ‘Instron’ equipped with data acquisition software.Five different statess of test had been appliedduring compression. One in dry condition, oneunder fixed contact and three with differentlubricants like, talc, water and grease. Increased13 th International Conference on Tribology – Serbiatrib’132.2 Abrasion testSpecimen of sizes s 70mm x 30mm x 2mm were cutfromvulcanizedd rubber sheets of 150mmm x 150mm x2mmm for conducting the abrasion tests. EPDM rubberrof three different shore hardness values was selectedfor this purpose. These are 55 Å, 70 Å and 80 Å375

espectively. Abrasion tests were carried out using atwo body abrasion testerTR-605 (Ducom) as shownin Figure 4. The machineis designed to conduct test asper ASTM D 6037 (Testmethod B) and/or ISO 8251[18]. The machine is equipped with a stepper motordrive (makes –My Com,24 V DC, and 40 W, modelno. IMS-200-220 AL) and requires an electricity of230 V x 1 φ x 50 Hz and100 W power. The wheel ofthe abrasivewear tester is made of stainless steelhaving a diameter of 50mm and width of 12mm.SiliconCarbide (SiC)water proof paper(Carborundam Universal) of grades ER 240, ER 220and ER 180were pasted on the top surface off thewheel for the purpose of abrasion. Appropriate sizes ofthe abrasivepapers weree cut and pasted on the wheelusing an adhesive (Feviquick). Rubber specimenswere placedin the desired position on the machinetable and clamped properly. The specimens were alsosubjected tonormal load using a dead weight. Theleverage action obtainedin the machine in use is 1:2,that is, a counter weight of 2N will apply 1NN ofnormal loadon the job. Then the specimens wereabraded against the abrasive paper under simulatedabrasive wear condition.3. RESULTS AND A DISCUSSIONSFor thislaboratoryexperimentationthreedifferentlevels of four factors have beenconsidered. The factors are hardness of EPDMrubber, abrasive grade, load on joband cycles andthe levels are low, medium and highh respectively. Itis needlesss to mention that if a full factorialexperimentation had been conducted with the fourfactors eachat three levels, as mentioned, then atotal of 81 experiments would have to be carriedd outand the number would have been multipliedaccordinglyfor replication. In the present study, anL 9 - orthogonal array has been selected basedd onTaguchi’s experimentaldesign technique [13] andthus only 9 experiments have been conducted. Thecombinations of different factors and levels as wellas the specific wear rate corresponding to eachcombinationn are shown in Table 3 inAppendix-I.Table 2, inn Appendix-I, indicates the flowcharacteristics of o EPDM rubber under compressiveeloadand in different working conditions, that is,withor without lubricants. The lubricants, asindicated in thee table, are selected based primarilyon the literaturee survey as well as from the real lifeexperience. The compressive load imparted byInstron and correspondingg height reduction dataahave been recorded through the dataa acquisitionsystem of the machine m andd later on different flowcharacteristics like l load-vs.-deflection,true stress-vs.-deflection and a true stress-vs.-truestain havebeen calculatedd using the indigenouslydevelopeddMATLAB codee in that regard. Some flow curvesare shown in Figure 5 (a), (b) and (c) as samples.L o a d ( K N )141210865560708085WATER4200 5 1015 20 2530 35Percentage deflection(a)40 45 501412556070TALC80Figure 4. Laboratory set up of a two-body abrasion tester.t r u e s t r e s s ( N / s q m m )10864285376t r u e s t r e s s ( N / s q m m )121000 5 10864200 0.1556070808515 20 2530 35Percentage deflection(b)FI XED CONTACT0.2 0.30.4 0.5t r u e s t r a i n40 45 500.6 0.7(c)Figure 5. Flow characteristic c curves are shown in (a), (b)and(c) under different conditions as indicatedd in the curves.13 th International Conference C onn Tribology – Serbiatrib’13

It has already been mentioned that t four factors,each at three levels, have been selected to conductthe abrasiontest. Table 3 indicates the said factor –level combinations and corresponding specific wearrate data which is obtained fromthe followingformula [11]:mW sFL,where, W s = specific wear rate (mmm 3 /Nm)Δm = mass loss recorded gravimetricallyy(gm)ρ = specific gravity of EPDM (gm/cm 3 )F = the normalload on the job (N)and, L = overall sliding distance(m)The worn surface morphology had been studiedfor each sample immediately after the abrasionusing a scanning electron microscope (SEM: JEOL,JSM-6360, model 75 82) to see thesmallest detailof the pattern abrasion in the range of 4 – 5 nm (4 –5 millionths of a millimeter). Thetest had beenconducted immediately after the experiment, thoughit is reported by Pandey et.al. [14] that in theirexperimentsfracture mode did not change withinn 72hours of storage before conducting SEM studiess andcoating etc. The worn surfaces had been coated witha very thinlayer of palladium (Pd) using ionsputtering machine (Auto Fine Coater: JEOL, JFC-to1600) prior to SEM studies. It is not out of placemention here that ioncoating is done on non-etc.)conducting specimen (like biological specimento be analyzed in SEM for quick and highlyefficient results. This is done mainly to preventcharging ofelectrons atthe sample[15]. For somesamples energy dispersive X-ray spectroscopy(EDAX) had also beendone in conjunction withSEM to find out thepercentages of differentelements inthe sample. Elemental mapping withEDAX is helpful to get insight into the chemicalchanges onthe surface and sub-surface of thesample[16].As no chemical reactionis taking placein the present case hence EDAX hass not done for allthe specimens.Figure 6 shows some typical pattern abrasion ofEPDM as obtained fromSEM studies. In figure 6(b)a chunk of rubber agglomerate hass been separatedleaving behind a groove (chunkingand grooving).Figure 6(c) reveals the ridge formation whichsupports theconcept ofrubber wear by the processof plowing.(a)(b)(c)Figure 6. Worn W surface morphology of EPDMcorrespondingg to trial no.1 (×200), 2 (×1000) and 6(×2000) as obtained from SEM studies.The graph of o the average specific wear rate isshown in Figuree 7.13 th International Conference on Tribology – Serbiatrib’133777

0.20.1950.190.1850.180.1750.170.165Figure 7. The specific wear rate of EPDM 55Å (1), 70Å(2) and 80Å (3).The curves reveal that the specific wear rate issmallest forthe softer rubber, that is, EPDM 55Å.The specific gravity is also smaller in casee ofEPDM 55Å, which depends on the hardness andhardness again dependson the carbon black (CB)content of the rubber. However, in i case of flowcharacteristics of EPDMit is revealed that EPDM70Å is better than others. The selection of materialdepends onthe actual requirements in specificapplication area.CONCLUSION1 2This experimental work is devotedd for flow as wellas wears characterizationof EPDM rubber of differenthardness in different experimental conditions. Thee testconditions were very difficult to be harmonizedd butmuch care was taken toobtain results as accurate aspossible accepting the noise factorss included inn theexperimentations. The results obtained are tabulated,graphed andanalyzed accordingly in the previoussection. Future work is proposed with different otherlubricants as well as inclusion of complex operatingenvironment, like extreme temperature and pressureconditions etc. It is also proposed toconduct frettingwear test as well as measurement of abrasion lossusing pin-on-plate (POP)type of tribometer.ACKNOWLEDGEMENTThe present work has been carried out inn theframe workof Mechanical Engineering Department,Production Engineeringg Departmentt and Metallurgyand Material Science Engineering Department ofthe Jadavpur University, Kolkata, India. The authoris extremelythankful tothe facultyand laboratorystaff members, the students who at any point of timewere involved in this work at any ways. The authoris also thankful to NEL(Rubber division), Kolkata,for their very kind support in providing the workspecimen for the tests, purely for academic interest.3Series1REFERENCE[1] Felhos D., Karger-KocsisKs J.: “Tribological Testingof Peroxidee Cured EPDM Rubbers With DifferentCarbon BlackContentsUnder Dry SlidingConditions Against Steel”, Tribology InternationalVol. 41, pp. 404-415, 2008.[2] Kalpakjian S., Schmid Steven R.: “ManufacturingProcesses for Engineering Materials” ”, 4 th. Edition,Pearson Education, 2009. .[3] Dieter George E.: “MechanicalMetallurgy”,McGraw-Hill, SI Metric Edition, 1988.[4] Mukhopadhyay Abhijit: “Flow Behavior of EPDMRubber of f Different Hardness Values UnderAxysymetricc Compressive Load in Dry WorkingCondition”, Proceedingss of the 12 th. internationalconference on Tribology“SERBIATRIB’11”,Kragujevac, , Serbia, 11- 13.05.2011, pp. 56-64.[5] Sohail Khan M.,Franke R. et.al.: “Friction andWear Behavior of Electron Beam Modified PTFEFilled EPDM Compounds”, Wear, Vol. 266, pp.175-183, 2009.[6] Mukhopadhyay Abhijit: “Effect of Some Lubricantson the Flow Characteristics of ‘EPDM’ Rubber UnderCompressivee Load”, Acceptedpaper in theinternational conference on Tribology, ‘BALTTRIB-2011’, 17-199 Nov., 2011, Kaunas, Lithuania.[7] Mukhopadhyay Abhijit: “Maintenance Tribology:A New Paradigm in Maintenance”,SURVEY (amanagement journal of IISWBM, Kolkata), Vol. 50No. 3&4, pp.1-7, 2010.[8] Lenard J.G. : “Tribology in Metal Rolling”, Annalsof CIRP, Vol. 49, No. 2, pp. 567-590, 2000.[9] Sofuoglu H. ., Rasty J.: “Flow Behaviorr of PlasticineUsed in Physical P Modeling of Metal FormingProcesses”, Tribology International, Vol. 33, pp.523-529, 2000.[10] El. Tayeb N.S.M., N Nasirr Md.R.: “Effect of SoftCarbon Black on Tribology of Deprotenised andpolyisoprene Rubers”, Wear, Vol. 262, pp. 350-361, 2007.[11] Karger-Kocsis J., Mousaa A., et. al.: “Dry Frictionand Sliding Wear of EPDM Rubbers Against Steelas a Function of Carbon Black Content”, Wear,Vol. 264, pp. 359-367, 2008.[12] MukhopadhyayAbhijit:“Reviewon theTribological State Test in Some Metal FormingOperation inn Industry”, Mechanica Confab, Vol. 1,No. 3, pp. 29-39, 2012.[13] Logothetis N.: N “Managing for Total Quality: FromDeming to Taguchi andd SPC”, Prentice Hall ofIndia, 2000.[14] Pandey K.N., Setua D.K., Mathur G.N., “Materialbehavior: Fracture F topography of rubber surfaces:an SEM study”, Polymerr Testing, Vol. 22, No. 3,pp. 353-359, 2003.[15] SEM/EDAXXanalysis,http://www. ML06\1\ANALYSISWITHSEM_EDS.PMD060317.[16] Ginic-Markovic M., Roy Choudhury N.,et.al., ,“Weatherability of coatedd EPDM rubber compoundby controlledUV irradiation” ”, Polymer37813 th International Conference C onn Tribology – Serbiatrib’13

Degradation & Stability, Vol. 69, No. 2, pp. 157-168, 2000.[17] Ludema Kenneth C.: “Friction, Wear, Lubrication:A text book in tribology”, CRC press, U.S., 1996.[18] Handbook of ‘Two Body Abrasion Tester: TR-605’, Ducom Instruments, http://www.ducom.com.APPENDIX – ITable 2. Flow characteristics data of EPDM.EPDMLubricantofdifferentDry Fixed Contact Talc Water Greasehardness A B C A B C A B C A B C A B C55Å 2.71 2.08 2.16 1.77 1.34 9.90 2.71 2.04 2.05 1.99 1.52 1.52 1.66 1.34 1.2560Å 3.35 2.41 2.53 2.65 2.01 6.78 2.64 2.05 2.04 2.72 2.05 2.05 2.36 1.75 1.7970Å 4.05 3.15 3.47 3.18 2.40 2.40 3.95 3.17 3.03 3.09 2.34 2.34 2.52 2.62 1.9180Å 9.03 5.95 6.75 8.36 6.38 2.01 7.88 6.29 5.24 6.78 5.05 4.83 6.94 3.91 5.2485Å 16.57 10.00 12.53 13.03 9.90 1.34 16.32 12.29 12.29 11.76 9.33 8.87 - - -[A: Load (KN) at 50% deflection; B: True stress (N/mm 2 ) at 50% deflection; C: True stress (N/mm 2 at a true strain of 0.7]Table 3. Factor-level combinations of the experiments as per Taguchi’s L 9 - orthogonal array and the abrasion loss data.TrialNo.Hardness(Shore;Å)AbrasivegradeLoad onjob (N)CyclesSpecific wear rate(mm 3 /Nm)1 st replication 2 nd replication 3 rd replication1 55 Very Fine 5 2000.1149 0.1038 0.10942 55 Fine 10 400 0.2641 0.1692 0.29523 55 Medium 15 600 0.2279 0.1507 0.17414 70 Very Fine 10 600 0.1930 0.1932 0.19215 70 Fine 15 200 0.1991 0.2234 0.19306 70 Medium 5 400 0.1640 0.1963 0.13537 80 Very Fine 15 400 0.1775 0.2142 0.19448 80 Fine 5 600 0.1782 0.2685 0.18489 80 Medium 10 200 0.1551 0.2098 0.195113 th International Conference on Tribology – Serbiatrib’13 379

Serbian TribologySocietySERBIATRIB ‘1313 th International Conference onTribologyKragujevac, Serbia, 15 – 17 May 2013Faculty of Engineeringin KragujevacAPPLICATION OF 3D SOFTWARE PACKAGES FORDESIGNING TRIBOMETER OF MODULAR TYPEIvan Mačužić 1 Branislav Jeremić 1 Petar Todorović 1Marko Đapan 1 Milan Radenković 1 Marko Milošević 11 Faculty of Engineering Kragujevac, University of Kragujevac, Kragujevac, Serbia,ivanm@kg.ac.rs, bane@kg.ac.rs, petar@kg.ac.rs, djapan@kg.ac.rs, radenkovic@kg.ac.rs, m.milosevic@kg.ac.rsAbstract: In this paper is presented advantages of modular type of tribometers using various types ofsoftwares, especially for modelling. Designing this type of tribometer allows significant saving regardingtime, space, people, raw materials as well as financial resources. Design of tribometers followed with testingand verifying, which implies simulation using various types of softwares, significantly improve the way andstandards of tribometer contruction. Selection of the desired parameters which would be used in theexperiments is very important and the first step in designing of the tribometer. Testing and verifying is thelast step when scientists and researchers get the result of the final evaluation whether or not to proceed toconstruction of the tribometer.Keywords: tribometer design, rapid construction, CAD software, simulation1. INTRODUCTIONThe machines and equipment for varioustribology experiments need to be maximal versatile.The versatility is reflected through possibility ofproviding variety of experiments. These types ofequipment are specific and mostly designed,constructed and used by researchers and scientistsfor very complex measurements and experiments.These equipment need to be, not just versatile butvery precise, providing accurate results which areused as an input for further experiments orcalculations.Tribometers are widely used for experimentalresearch for measurement of friction characteristicsunder laboratory conditions. This laboratoryconditions have to be pre-defined, beforemeasurements are performed. Also, equipment andmeasurements are designed for conducting differenttypes of experiments, to obtain results for furtherexperiments. Laboratory conditions mean variousparameters are constant, for example normal force,velocity, temperature and humidity [4]. Because ofdifferent types of the contact [5] versatility ofequipment is the most important characteristicwhich is required.There are many different types of tribometersfor different use [1] [2] [3] [6]. Different use meansdifferent structure of tribometer, different processesfollowed with different parts of tribometer. So,rapid development of tribometers based on softwarefor 3D modelling is required. The modulartribometer will be able to speed up process ofdesign and construction of tribometers dependingon type of experiment. Today, there are commercialuniversal tribometers which are not the same asmodular we want to present. Universal tribometersoffer the various options and different types ofexperiments on the same equipment. The maindifference between universal and modulartribometer is that on the universal all necessaryparts are already mounted on the same equipmentand just small corrections in software enablesworking on test equipment. Also, the additionalparts can be added for desired experiments [7]. Onthe modular tribometers researchers can choosewhich of the tribology parameters wish to monitor.So, after calculations and simulations of propertiesof equipment structure, constructors can proceedwith equipment construction.Modular tribometers are suitable for rapidconstruction of equipment and what is the mostimportant, for performance evaluation and380 13 th International Conference on Tribology – Serbiatrib’13

conformance to researchers’ requirements.Software for 3D modelling allows that beforeconstruction of the equipment to check, test, verifyand validate successful experiment’s output. Thismeans there is a possibility of predictingsuccess/failure of the experiments. The results ofthis prediction provide valuable significant datawhich could save, on the first place, financialresource and then time, space, people, rawmaterials, etc.The aim of this paper is to propose methodologyof using software for modelling tribometer whichhave to be designed and tested performingsimulation for selected parameters. The wholeprocess of testing, quality checking and finaladjusting of modular tribometer need to beconducted before construction. Every part oftribometer is compact in size, having the bestpossible characteristics for selected parameters,overcome technical difficulties and the endchecking is the tribometer functional.The paper is structured as follows: Secton 2present importance of using software for 3Dmodelling. Section 3 describes one example ofmodular tribometer and what type of experimentcan be carried out. The conclusions are given inSection 4.2. BENEFITS OF 3D TRIBOMETERMODELLINGCAD technology is ubiquitous for diverse arrayof fields, particularly engineering andmanufacturing [9]. It is integral part of the everyprocess where is needed to meet some goals such asreducing design to production lead time, betterengineering analysis, additional flexibility andfaster response for design modification [8]. Allthese benefits are reflected to manipulation withdesigning parts necessary for final construction.The greatest attention is given to: dimensioning of critical parts, which arenecessary for conducting successfulexperiments resulting to relevant outputs;dimensioning of measurement parts;possibility to construct various equipmentfor different experiments mean thatmodules have been already designed forrapid design;simulations which are very important to seeif all parts of the tribometer are properlyassembled, if all system is functional or notto react in time before spending resources,before construction. simulations have another advantage,researches can see and conclude is thereany overlapping of the work areas, someerrors, mistakes and fault decisions whichare made during equipment and processdesign;predicting values of the parameters whichare selected for experimental research(normal load, viscosity, stress generated inthe contact zone due to the given force,etc.).All steps in design and construction depend onrequirements from researches and scientists whatparameters would be considered. Also, conceptualdevelopment of modules is dependant ofrequirements.In our case, the most important parameters werelinked with basic parameters typical for hydraulicsand its components which can be found in realindustrial systems. Also, we monitor processeswhich appear between selected pin and plate in oilenvironment. Based on this we could simulate andcalculate if the required processes are possible tomonitor and get relevant results. The mostimportant parameters for our research were: liquid resistance (in our case liquid is oil)inside the container and stress in the contact zone due to given load.In the next section will be presented onemodular tribometer designed, tested andconstructed for tribological phenomena in hydrauliccomponents (pumps, motors, cylinders and valves).3. EXAMPLE OF EXPERIMENTALEQUIPMENTThe first step is defining the type and concept ofthe tribometer. In our case, whole concept is basedon analysis of hydraulic components andcharacteristics of wear processes which are occur inthat kind of industrial equipment. From theliterature [10] [11] linear sliding movement andabrasive wear mechanism are the most commonand the most important in the hydrauliccomponents. Regarding this fact, pin-on-platetribometer was selected to be designed andconstructed because we had wear process betweenpin and plate in oil environment. Also, it can beable to control some of the basic tribologyparameters such as load in contact, slide length,sliding speed, liquid resistance, etc.On the figure 1 is shown model of tribometerwhich is divided in 3 bigger units: experimental unit (positions: 1, 2 3 4and 7); control unit (positions: 6 and 8) and pneumatic drive unit (position: 5)All units need to be well connected andfunctional. Potential problems with the first startingup of the constructed tribometer can be prevented13 th International Conference on Tribology – Serbiatrib’13 381

using number possibilities of 3D modellingsoftware packages.8623Figure 1. Sub-divided components of tribometer modelFirst of all, whole design of the tribometer isbased on container with a plate and oil (1), placedon a horizontal linear guides (2), movingalternately, while pin is stationary. Pin bracket (3)is set to vertical linear guides (4) and given loadson the pin passed through the bar, which is also adynamometer for measuring of friction force. Drivesystem for reciprocation motion is pneumatic (5)with pneumatic cycle counter (6).On container with a plate (1) displacementtransducer is fixed (7), with function to measure, inreal time, container position and thus enable theaccurate determination of velocity and momentswhen container change movement direction. On atribometer base plate surface there is a connectionpanel for this transducer (8).Now will be described the most important partof the tribometer, the contact zone where tests areperformed (figure 2).1475Value of normal load in the contact zone isdefined by calibrated weights (1) where the forcestransferred through the shaft (2) with a sphericalend are to the pin bracket (3). Compensation ofown mass elements which are located on the pinbracket is performed through a spring with athreaded spindle (4). On pin bracket (3) set thesingle-axis piezoelectric vibration sensor (6) whichmeasures vibrations in the tribological contact inthe vertical direction.The close-up view of the above describedcontact zone is shown in figure 3.Figure 3. Close up view of contact zoneThere is a plate attached at the bottom ofaluminium container. Pin has cone top that fits intoa spherical end of the dynamometer bar and thuscarries the normal force evenly over the entiresurface of contact. Container has a volume of 500ml and is filled oil up with oil to about 1/3 of itsheight. Container is covered with a transparentcover on which there are connections for oilcirculation - suction and return. Suction line takesoil from the bottom of container on one side andreturns oil back to the surface on the other side ofcontainer. This is to ensure adequate oil mixingduring circulation.And, at the end the third part of the tribometer ispneumatic system (figure 4).147332Figure 2. System for setting load and forcemeasurementFigure 4. Pneumatic drive unit382 13 th International Conference on Tribology – Serbiatrib’13

This part is assembled of pneumatic cylinder(1), air-operated distribution valve 5/2 (2), twopneumatic limit switches (3), pneumatic logic valve3/2 (4), pneumatic cycle counter and the resetbutton. Limit switches limit stroke of cylinderpiston which is fixed to container. Switches arefixed and the stroke length is determined byvarying the length of a cylindrical end part ofcylinder (7) which activated limit switches. At bothcommand lines, which bring compressed air to thecylinder, set of throttle valves that regulate thespeed of the cylinder in both directions.4. CONCLUSIONIn the times where financial savings are moreimportant than results, there need to be the waywhich will satisfy both conditions. Besides that,there is a constant need for various types ofexperiments which need to be precise and its resultsneed to be accurate. Development of modulartribometers could be good solution for issuesrelated for rapid construction of tribometer andobtaining accurate results for selected parameters.Tribometer which is designed from modules testedand verified through simulations it is very simple toproceed to the next step regarding construction.Desired and requested parameters are alsoimportant because software could obtain possiblemistakes and failures in design before construction.ACKNOWLEDGMENTResearch presented in this paper was supportedby Ministry of Science and TechnologicalDevelopment of Republic of Serbia, GrantTR-35021, Title: Razvoj triboloških mikro/nanodvokomponentnih i hibridnih samopodmazajućihkompozita.REFERENCES[1] E. Sajewicz, Z. Kulesza: A new tribometer forfriction and wear studies of dental materials andhard tooth tissues, Tribology International 40,pp. 885–895, 2007.[2] S.K. Sinha, S.L. Thia, L.C. Lim: A new tribometerfor friction drives, Wear, Vol 262, pp. 55–63, 2007.[3] J.L. Dion, G. Chevallier, O.Penas, F.Renaud: A newmulticontact tribometer for deterministic dynamicfriction identification, Wear 300, pp. 126–135,2013.[4] K.D. Moerlooze, F. Al-Bender, H.V. Brussel: Anexperimental study of ball-on-flat wear on a newlydeveloped rotational tribometer, Wear 271,pp. 1005– 1016, 2011.[5] P.J. Blau: Friction Science and Technology, MarcelDekker, New York, 1996.[6] M.R. Kashani, E. Behazin, A. Fakhar: Constructionand evaluation of a new tribometer for polymers,Polymer Testing 30, pp. 271–276, 2011.[7] URL: www.microtest-sa.com/index.php?lang=en(accessed: 04.03.2013.)[8] M.K. Malhotra, M.L. Heine, V. Grover: Anevaluation of the relationship between managementpractices and computer aided design technology,Journal of Operations Management 19,pp. 307–333, 2001.[9] W.C. Regli, V.A. Cicirello: Managing digitallibraries for computer-aided design, Computer-Aided Design 32, pp. 119–132, 2000.[10] C.B. Mohan, C. Divakar, K. Venkatesh, K.Gopalakrishna, K.S. Mahesh Lohith, T.N. Naveen:Design and development of an advanced linearreciprocating tribometer, Bengaluru 562112, India2009.[11] G.W. Stackhowiak, A.W. Batchelor, G.B.Stachowiak, Experimental methods in tribology,Tribology Series 44, Elsevier Science, 2004.13 th International Conference on Tribology – Serbiatrib’13 383

Serbian TribologySocietySERBIATRIB‘1313th International Conference onTribologyKragujevac,Serbia, 15 – 17 May 2013Faculty of Engineeringin KragujevaccUSING OFKALMAN FILTER ASA PROGNOSTIC TOOLFOR TRIBOLOGY PROCESSESIvanMačužić 1 , Petar Todorović 1 , Marko Đapan 1 , Milan Radenković 1 , Branislav Jeremić 11 Faculty of Engineering, University of Kragujevac, Kragujevac, Serbia,ivanm@kg.ac.rs, petar@ @kg.ac.rs, djapan@kg.ac.rs, radenkovic@ @kg.ac.rs, bane@kg.ac.rsAbstract: The paper consider possibilities for performing ofprognostic procedure pfor tribology processes inhydraulic equipment using advanced mathematical tool called Kalman filter. It is an algorithmthat uses aseries of measurementsobserved over time, containing noiseand other inaccuracies,, and produces estimatessof monitored parameter that tend to be more precise than those based on a single measurement alone.Kalman filter operates recursively on streams of noisy input data to produce a statistically optimal estimateeof the underlying system state.This type of procedure could be usedd for prognostic of state of chosenparameter of tribologysystem with significantly accuracy. Efficiencyy of Kalman filter were tested onexperimental results of hydraulic oill contamination monitoring performed in laboratory conditions.Keywords: Kalman filter, prognostic, tribology processes, hydraulic equipment1. INTRODUCTIONPrognostics is a set of activities aimedd atassessing the remaining time tofailure for aparticular technical system or risk of presencee oroccurrence of one or more failuremodes in thefuture. Prognosticsefficiency can be quitesatisfactory for the failure modes that haverepeating time characteristics,followed byprogressive degradation of keyexploitationcharacteristics [1].In casess of failure modes with random andunexpected events, prognostic is a very difficulttask with uncertain results.Prognostic process could be based on the model,or on the measurement results. Prognostic basedd onthe measurement resultss includes the use of variousmathematical tools for monitoring and predicting,such as for example Kalman filter and its simplifiedversion known as alpha-beta-gammaa filter [2].2. KALMAN FILTERSignal filtering, and extracting the t useful signalfrom noise is a traditional problemin science andtechnology.Significant number of models andalgorithms was proposed and developed for solvingof this problem. In case when signal and noise384spectra lies inn different frequency bands, theirseparation cann be made with appropriate bandfilters.Another problem arises when the spectra of thesignal and noise overlapp and thenstatisticalmethods for the assessment and evaluation of thesignal should be b used to extract the signal. In suchcircumstancesit is not possible toobtain anaccurate absolute value off the signal, and all themethods of filtration are made only to t minimizeinterference.The first such analog signals filter suggested s byNorbet Winer in 1940. using the method of leastsquares. New stage in the development of thetheory of filtration began Rudolf Emill Kalman in1960., with publication of his capital work, "A NewApproach to Linear Filtering and PredictionProblems" in which w it was first introduced methodwill become known in science as the Kalman filter[3].The Kalmann filter is a mathematical tool thattcanbe used too assess thee value of variables indifferent formss of real situation. Mathematicallyspeaking Kalman filter evaluates the condition oflinear systems. It is a statistical technique thattcombines the statistical nature of systemfaults withknowledge of the systemm dynamics, , and those13 th International Conference C onn Tribology – Serbiatrib’13

matrixes describe the state of the system and theirevaluation.The significance of this method is that, on theone hand it gives excellent results in practice,while, on the other side, is theoretically attractivesince it has been demonstrated that all of the filtersthat are applied precisely this variation achieved byminimizing the error estimates [4], which is oftenalso called the optimal filter.Application of Kalman filter was recently, inaddition to traditional applications related to signalprocessing, automatic control systems andprocesses, and the projection of the trajectory andballistic missiles, and a significant number of newareas such as medicine systems, global positioningsatellite (GPS) navigation, computer vision,economics, and so on.In order to use the Kalman filter to remove noisefrom the signal, system, or process that isconsidered to be such that it can be approximated aslinear and present. For nonlinear systems, whichcan’t be present or approximated as linear, the socalled.Extended Kalman filter is developed as anextension of the theory of linear Kalman filter tononlinear systems [4].Linear systems are those that can be representedusing the following two equations. The first is theequation of state: (1)the other is the output of the system of equations: (1) (2)These equations are:A - matrix that shows the relationship of current andprevious step, B - matrix connections inputs and thecurrent state, C - matrix state and do themeasurements, k - time index, x - variable thatindicates the state of the system, in - known inputthe system, y - the output of the system beingmeasured, and w and z are noises where w is calledprocess noise and z - measurement noise.Each of these values is generally a vector thatcontains more than one element. Vector x containsall the information about the current state of thesystem but we are not able to measure directly.Therefore, we measure the value of the vector ywhich is a function of the vector x with the additionof measurement noise z. This means that over themeasured values of the vector y can assess the statesystem described by vector x.Introduced the assumption that the averagevalue of the process noise w and the measurementnoise z is zero during a time interval and that thereis no correlation between them, and that the twoforests have approximately a normal distributionwith covariance Sw and Sz. Based on the above canbe derived equations for the Kalman filter [4]: (3) (4) (5)Each of the three defined equations of Kalmanfilter includes a series of operations with matriceswhere the index T represent matrix transpositionand index -1 represent matrix inversion. Matrix K iscalled the Kalman gain and the matrix R estimationof error covariance.Obviously, the Kalman filter works recursivelyand takes only the value of the system state at theprevious time point to generate assessmentfollowing conditions (not required the entire historyof the state).The Kalman filter can be used in different waysto handle real signals where the results are quitedifferent nature and use. When the filter is used toestimate the previous state of the known history ofthe system (measured values) is obtained byeliminating the possibility of measuring noise inorder to level the curve that defines the state of thesystem changes over time in the past. If we estimatethe current state of the system is the result offiltering the measured signal [4], [5].The filter can be used for the prediction of thesystem in the near future if the state of the systemin estimated time is moved to next time interval inadvance. This means that the estimated state andthe actual state were time-shifted by an interval soin accordance with the assessment of the situationfor the time moment k+1 conducted on the basis ofestimated values of x from time to time andmeasured values at the time point k + 1. This ispossible because the movement of the estimatedstate for a time interval of pre-practice leads totemporal overlap of the estimated state x from k+1timing and the real state of x from k time.Practically on the basis of previous estimates anderrors in the actual situation new value is evaluatedfor the next time point.One of the great advantages of Kalman filter isits feature that it is not necessary to carry outdetailed modelling of system which condition isestimated. The reason for this lies in the recursivityprinciple of Kalman filter and the periodicrepetition of the process of assessment andcorrection, and built-in tendency to corrects andminimize error from step to step. This feature opensthe door to a wide use in solving various technicalproblems since the generation of an accurate modelof the real system is a very demanding and complextask [5].13 th International Conference on Tribology – Serbiatrib’13 385

3. KALMAN FILTERAPPLICATION FORRPROGNOSTIC OFTRIBOLOGYPROCESSES IN HYDRAULICThe essence of the idea for theapplicationn ofKalman filter for forecasting andprognosticc oftribology processes liesin the fact that it is a toolthat is least dependent on the accuracy of the modelof tribology system that is considered.Theprocedure involves projections toprognosticc inmathematical modelling of the behaviour oftribology parameters in time, whichh severely limitsthe application of other, model-based, prognostictools, sincethey are directly dependent on thecharacteristics of the model for a specific system.On the other hand, Kalman filter will providevery useful results even for very approximatemodels andalso for some standard, general modelsof system behaviour in time (that do not even havedirect link with the observed system),.It is clear that the Kalman filter works onlyy atdiscrete points in time, and its use is relatedd todigital signal processing. Complexmath, matrixtransformations and calculations, represent an easytask for modern computers and processors to thestabile and to mobile devices, which also allowedthe installation of Kalman filters in numerousportable monitoring devices.As an example of application of Kalman filterfor prognostic of tribology processes, trendingg ofcontamination level in hydraulic equipment will bepresented.Method of hydraulic oil contamination valuesprognostic (based on ISO4406 contamination levelcode) usingKalman filter is shown on Figuree 1.From 200 measured points, that define the values ofcontamination level, forparticles off defined size, 9points was allocated (8 is shown from T1 to T8) .ISO 4406 class of oil contaminationT9PT5PT7T7PT6T4T7T8PT3PT6PT5T4PT2T3projectionerrorT1T2PT1,Kalman filter defines point T2P so that t its valueis the same as the t value of the point T1. This is theinitial assumption that nothing will change.At the time of obtainingg the measured values ofother points - T2, T the projection error is i calculatedas the difference in point values T2 and T2P. Onthe basis of the projection error values andmeasured values point T2, Kalman filter performsthe projection of o the value e of the thirdpoint T3P.Then a new measured valuee of contamination T3 isreceived and new projectionn error is calculated andthe cycle is repeated.Practically the value of each new projected pointis function of the previous measuredvalue andprojection errors in the previous point.At Figure 2, , diagram obtained by the projectionof contamination using a Kalman filter for thecurve related to the measured contamination ofhydraulic oil is shown, together with diagram oferror projections. Prognostic process is i conducteddfor 40 points, which define thevalue ofcontamination,in first attempt and 10 points insecond one.ISO 4406class of oil contaminationprojection errorISO 4406class of oil contaminationprojection errornumber of f cyclesnumber off cyclesnumber of cyclesmeasured valuesprojection valuesmeasured valuesprojection valuesnumber of cyclesFigure 1. Kalman filter prognostic processIn this case, those are equidistant points,although, ingeneral, donot have to t be. Basedd onthe value ofthe first point of T1 andset parametersof Kalmanfilter define the value of the firstprojected point in the future (T2P). Since there isno additional information other than the valuee of386number of cyclesFigure 2. Kalman filter prognostic and projection errorAt Figure 3. diagrams obtained by the projectionof contamination using the Kalman filter f for oilsample from the test with the externall addition ofcontamination in i the contact zone is shown.Total of 3 diagrams are shown in Figure 3.refers to variants of Kalman filters with different13 th International Conference C onn Tribology – Serbiatrib’13

values influence of themeasurement noise (fromlower to higher)ISO 4406 class of oil contamination ISO 4406 class of oil contaminationISO 4406 class of oil contaminationnumber of cyclesnumber of cyclesnumber of cyclesFigure 3. Kalman filter prognostic with step-changee incontamination for different values of measurement noise4. CONCLUSIONBased on shownresults some generalconclusionss about process of prognostic using theKalman filter could be defined: Theexamples of practicall applicationn ofprognostic using Kalman filter obtainedverygood results in tracking of realmeasured values with acceptable projectionerror in case ofmeasured diagrams withoutsuddenand significant changes ofcontamination value. The biggest mistakee of projection, as a rule,is rightt at the beginning at first projectedpoint. Variations in the values of the measuredsignal and a noise measurement have a directimpact on the accuracy of the prognostic. The influence of the measurement noise onthe result of the projection canbe adjusteddusing the t definition of the value of thecorresponding parameter in the t Kalmanfilter equations. Increasing the value of thisparameter indicatess the presence of moreintensive measuringg noise and vice versa. In the case c of a sharp and abrupt prognosticusing a Kalman filter have visible and theexpected delay in the response to change. Itis clear that there is no method offorecasting,which can predicttheoccurrence of sudden, unexpected andabrupt changes in the values followed bythe diagnostic parameter. In anycase, theseephenomenapoint to theserioussirregularities and problems in the systemand certainly represent an alarmsignal. The great advantage of using Kalman filterlies in its fullindependenceandinsensitivity to the shape and characteristicssof the measured m contaminationtrend charts.ACKNOWLEDGMENTResearch presented in this paper was supportedby the Ministry of Education, Science andTechnological DevelopmenDnt of Republic of Serbia,Grant 35021REFERENCES[1] A. Muller, M.C. M Suhner, B. Iung: Formalisation of anew prognosis model for supporting proactivemaintenancee implementation on industrial system, ,Reliability Engineering & System Safety 93,pp. 234-253, , 2008.[2] J. Lee, J. Ni, N D. Djurdjanovic, H. Qiu, H. Liao:Intelligent prognostic ptools and e-maintenance,Computers in Industry 57, , pp. 476-489, 2006.[3] R.E. Kalman: A New Approach to Linear Filteringand PredictionProblems,Journal of BasicEngineering 82 (Series D) ), pp. 35-45., 1960.[4] D. Simon: Kalman Filtering, Embedded SystemssProgramming, 2001.[5] G. Welch, G. Bishop: An Introduction to theKalman Filter, Department of Computer ScienceeUniversity of North Carolina13 th International Conference on Tribology – Serbiatrib’13387

Serbian TribologySocietySERBIATRIB ‘1313 th International Conference onTribologyKragujevac, Serbia, 15 – 17 May 2013Faculty of Engineeringin KragujevacFRICTION COEFFICIENT ESTIMATION DURING FRICTIONSTIR WELDING WITH THE SINGLE SHOULDERED WELDINGTOOLMiroslav Mijajlović 1 , Dušan Stamenković 1 , Milan Banić 1 , Aleksandar Miletnović 1 , Miloš Milošević 11 University of Niš, Faculty of Mechanical Engineering in Niš, Serbia, mijajlom@masfak.ni.ac.rs,dusans@masfak.ni.ac.rs, banic@masfak.ni.ac.rs, amiltenovic@yahoo.com, mmilos@masfak.ni.ac.rsAbstract: Friction stir welding utilizes friction forces on the contact of the welding tool and workpieces withthe goal of heating and softening workpiece material before stirring and mixing it into the weld. The processof stirring, mixing and welding is quite complex: material of workpieces in the welding zone is drasticallydeformed/reformed, heated, translated, rotated and softened, and finally, deposed behind the welding tool tocool, plasticize, and recrystallize as a weld. In such conditions, it is difficult to recognize friction conditions,contact surface(s), and loads on the contact. There are no fully operational analytical models for estimationof the friction coefficient during friction stir welding. This paper is giving an overview on a frictioncoefficient research and presents experimental results from performed friction stir welding of aluminiumalloy 2024 T351. Experimental results are used as input for the modified analytical model for estimation offriction coefficient in friction stir welding.Keywords: Friction Stir Welding, Friction Coefficient, Heat Generation.1. INTRODUCTIONFriction is one of the most important parametersfor successful friction stir welding (FSW) process –this is a sentence that no researcher of FSW will tryto disapprove. Such influence to the process itselfhas motivated the inventor of FSW to use “friction”in the name of the process.However, friction in FSW has been neverpresented and explained as a parameter that can bemanipulated or adjusted in some manner to improvethe FSW process itself. For example, when greaterfriction in FSW is needed, welding tool (figure 1, b)must have threads, facets, keys etc., when lessheating is needed, welding tool has to travel fasterwhat will result in shorter contact between weldingtool and some particles of workpieces what resultswith less friction on contact.Nowadays experiences in FSW usage recognizetechnological parameters of the process (travel rate,rotation speed, duration of welding etc.) andgeometry of the welding tool (shape, dimensionsetc.) as best parameters for successful managementof quality of FSW. Principle of “trial and error” andparameters management were successful for FSWand it has been significantly improved. It is knownthat better the mathematical model explaining thephysical process is, the more applicable the processbecomes. “Try and error” principle uses nomathematical model for improvement but alwaysgives results and improvements. Its maindisadvantage is high resource / time consumption.In a certain way, friction is very important foralmost any aspect of the FSW, but its ambiguityand complex dependencies with the otherparameters of FSW make it difficult to use formanagement of the process. That is the main reasonwhy friction is the least investigated physicalprocess of FSW.2. SINGLE SHOULDERED FSWThe first application of the FSW was with thewelding tool having one probe and one shoulder(figure 1, a). Such construction requires an anvil inorder to make weld creation possible. There arenewer constructions of the welding tool with twoshoulders and/or more than one probe.Application and technology of FSW with awelding tool with one shoulder is in detailexplained in the literature [1-3].388 13 th International Conference on Tribology – Serbiatrib’13

Figure 1. Friction stir weldinga – principle of FSW, b – welding tool and its active surfaces, c – heat generation and transport [1]3. ESTIMATION OF THE FRICTIONCOEFICIENT DURING SINGLESHOULDERED FSWThe newest improvement and development ofthe model for estimation of the friction coefficientin FSW is 4 years old Kumar’s model [4] and relieson the estimation of the momentum of frictionwhich is afterwards, with adequate mathematicalmodel, transformed into the friction coefficient.There are several difficulties in application of sucha model:1. measuring the momentum of friction requiresspecific and limitedly applicablemeasuring/working configuration [1],2. friction coefficient estimated during FSW byKumar is a median value for all contacts surfaces –active surfaces of welding tool and workpieces(figure 1, b).3.1 Specific time moments of the FSWMijajlovic et al [Ref. 5, Figure 4] gives ascheme of welding tools engagement during FSW.It is important to define specific moments of timeduring welding.Probe tip is active surface that is fully engagedin the FSW process from the beginning of theplunging phase (t 0 ) until the end of the seconddwelling phase (t 4 ). At the beginning of theplunging phase probe tip slides over the top surfaceof welding plates and there is no significantplunging into material of the welding plates.Material of the welding plates is still capable toresist influence of the contact pressure on contactbetween probe tip and welding plates. Plungingforce is rising as the plunging phase on goes andeventually plunging force will be intensive enoughto produce contact pressure that will overcomeresistance of the material and welding tool willpenetrate into the material (in the moment oftime t ps’ ). This intensive plunging will enablecontact between probe side and material of weldingplates and increase of engagement of the probe side– it will reach some value until the end of theplunging phase (t 1 ). It will be kept steady or slightlywill decrease during first dwelling phase (from t 1 tot 2 ) and it will increase again during welding phase(after t 2 ). When welding tool stabilizes (in weldingphase, when it reaches constant speed, at themoment of t ps" ) probe side will reach maximalengagement.It will be kept relatively steady until the end ofthe second dwelling phase (t 4 ) and after will slightlydecrease until the minimal value – when weldingtool gets pulled out, at the end of the pulling outphase (t 5 ). Shoulder tip will involve in FSW processwhen firstly touches (t st ) the material of weldingplates that was pushed upwards while plungingphase lasted. Engagement of the active surface willincrease to the maximum when plunging phaseends (t 1 ), it will keep steady value until the end ofthe second dwelling phase (t 4 ) when it will drop tominimum [5].3.2 Contact over the probe tipThe probe tip (pt) of the welding tool is rathercurved than flat due to the better distribution of thecontact pressure [1]. Without concern on thetopology of the welding tool’s probe tip, when theprobe tip is pressing the workpiece while loadedwith the axial force F z (t) and torque T pt (t),equilibrium of the force and the torque (no relativemovement of the welding tool and workpieces, norrotation of the welding tool) is reached if: pt () tFz()[ t dt () d0()]tTpt() t T1() t3 (1)where: (t)= pt (t) – total coefficient of friction -coefficient of friction at pt, d(t) – diameter of the13 th International Conference on Tribology – Serbiatrib’13 389

probe, d 0 (t) – diameter of the technological hole inthe workpieces, t - time.In such condition, the momentum of frictionM fr (t) is:Mfr() tFz()[ t dt () d0()]t() t (2)3and therefore, friction coefficient at pt is:3 M fr ( t)() t pt() t , t0 t tps'Fz()[ t d() t d0()]t3.3 Contact over the probe tip and the probeside(2)The probe side (ps) of the welding tool iscylindrical or coned surface with or without thread[1]. The thread is of great significance for thewelding process, however, it makes greatdifficulties for the analysis of friction and it will beneglected in analysis.If only the probe side is in contact with theworkpieces, equilibrium between the forces,represented as the contact pressure at the probe sidep ps (t), and the torque T ps (t) is:2 ps () tdt () ht () pps()π tTps() t T2() t (3)2where: (t)= ps (t) – total coefficient of friction –coefficient of friction at ps, h(t) – height of theprobe (side) plunged into the workpieces.In such condition, the momentum of frictionM fr (t) is:Mfr2() tdt () htp () ps ()π t() t (4)2and therefore, friction coefficient at pt is:2 M fr ( t)() t ps() t , t24 t t5(5)dt () htp () ps ()π tWhen the probe tip and the probe side aresimultaneously involved in the contact, equilibriumof loads and the torque at the probe tip and theprobe side T pt+ps (t) can be expressed as:2() tdt () ht () pps()π tpt ps () 1() (6)T t T tIn such condition, the momentum of frictionM fr (t) is:2M fr () t T1() t T2()t (7)Assuming that the friction coefficients at theprobe side and the probe tip are the same (only as avalue): () t ps() t pt(),t tps't tst(8)transforming the equation (7), friction coefficientbecomes:6 M fr ( t)() t ,22 Fz()[ t d() t d0()] t 3 d() t h() t pps()πttps' t tst3.4 Contact over the probe tip, the probe sideand the shoulder tip(9)The shoulder tip (st) of the welding tool iscylindrical or coned surface with the greatest area[1, 2]. Shoulder tip is the last active surface of thewelding tool involving into the welding process.If only the shoulder tip is in contact with theworkpieces, equilibrium between the loads and thetorque at the shoulder tip T st (t) is:st() tFz()[ t Dt () dmax]Tst() t T3() t (10)3where: (t)= st (t) – total coefficient of friction –coefficient of friction at st, D(t) – diameter of the st,d max – maximal diameter of the probe.However, shoulder tip is never involved in thewelding process as the only active surfaces –shoulder tip is always involved in weldingsimultaneously with the probe tip and the probeside and in such case, equilibrium of loads and thetotal torque T tot (t) is:Ttot() t T1() t T2() t T3()t (11)In such condition, the momentum of frictionM fr (t) is:M fr () t T1() t T2() t T3()t (12)Assuming that the friction coefficients at theprobe side, the probe tip and the shoulder tip are thesame (only as a value): () t ps () t pt () t st (), t tstt t4(13)transforming the equation (12), friction coefficientis:6 M fr ( t)() t ,22 Fz() t A3 d() t h() t pps()π+2 t Fz( t)BA dt () d0(), t B Dt () dmax , tps'ttst3.5 Contact pressure at the probe side(14)Contact pressure at the probe side is mostlydelivered by the welding force F x (t). Since weldingforce is active only during the welding phase390 13 th International Conference on Tribology – Serbiatrib’13

(t 2 t

Second set of experiments was performed withthe conical welding tool - CWT (Figure 2, b),changing the rotation speed from lower to higherand starting with the maximal dimension of thetechnological hole and decreasing it to the 0 – fromminimal plunging force to the maximal. [-]1.210.80.60.40.20t 0=5.1 st ps=9.05 st st=24.2 st 1=27.4 s0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30t [s]Figure 6. Friction coefficient: CWT, n=265 rpm, d 0 =5 mmM fr / T [-] [-]10.90.80.70.60.50.40.30.20.10t 0=5.1 st ps=9.05 st st=24.2 st 1=24.7 s0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30t [s]Figure 7. Ratio M fr /T: CWT, n=265 rpm, d 0 =5 mm1.210.80.60.40.20t 0 =t ps' =3.1st pt=10.1 st st=20.3 st 1=24.3 s0 2 4 6 8 10 12 14 16 18 20 22 24 26t [s]Figure 8. Friction coefficient: CWT, n=265 rpm, d 0 =3.2 mmM fr / T [-]10.90.80.70.60.50.40.30.20.10t 0=t ps' =3.1 st pt=10.1 st st=20.3 st 1=24.3 s0 2 4 6 8 10 12 14 16 18 20 22 24 26t [s]Figure 9. Ratio of the momentum of friction and thetorque: CWT, n=265 rpm, d 0 =3.2 mm [-]1.210.80.60.40.20t 0=9.6 st ps'=22.7 st st=25.1 st 1=29.7 s0 2 4 6 8 10121416182022242628303234t [s]Figure 10. Friction coefficient: CWT, n=265 rpm, d 0 =2 mmM fr / T [-] [-]10.90.80.70.60.50.40.30.20.10t 0=9.6 st ps'=22.7 st st=25.1 st 1=29.7 s0 2 4 6 8 10121416182022242628303234t [s]Figure 11. Ratio M fr /T: CWT, n=265 rpm, d 0 =2 mm1.21.110.90.80.70.60.50.40.30.20.10t 0=2.2 st ps'=6.1 st st=14.05 st 1=18.1 s0 2 4 6 8 10 12 14 16 18 20t [s]Figure 12. Friction coefficient: CWT, n=265 rpm, d 0 =0 mm392 13 th International Conference on Tribology – Serbiatrib’13

M fr / T [-] [-]10.90.80.70.60.50.40.30.20.10t 0=2.2 st ps'=6.1 st 1=18.1 st st=14.05 s0 2 4 6 8 10 12 14 16 18 20t [s]Figure 13. Ratio M fr /T: CWT, n=265 rpm, d 0 =0 mm1.210.80.60.40.20t 0=t ps ' =5.1 st st=19.9 st 1=21.5 s0 2 4 6 8 10 12 14 16 18 20 22t [s]Figure 14. Friction coefficient: CWT, n=600 rpm, d 0 =5 mmM fr / T [-] [-]10.90.80.70.60.50.40.30.20.10t 0=t ps' =5.1 st st=19.9 st 1=21.5 s0 2 4 6 8 10 12 14 16 18 20 22t [s]Figure 15. Ratio M fr /T: CWT, n=600 rpm, d 0 =5 mm43.532.521.510.50t 0=t ps' =5 .5 8 st pt=13.5 st st=17.5 st 1=21.2 s0 2 4 6 8 10 12 14 16 18 20 22 24t [s]Figure 16. Friction coefficient: CWT, n=600 rpm, d 0 =3.2 mmM fr / T [-]10.90.80.70.60.50.40.30.20.10t ps'=5.58 st pt=11.68 st st=21.15 st 1=24.3 s0 2 4 6 8 10 12 14 16 18 20 22 24t [s]Figure 17. Ratio M fr /T: CWT, n=600 rpm, d 0 =3.2 mm [-]2.221.81.61.41.210.80.60.40.20t 0=t ps' =6 .5 st pt=11.68 st st=19.3 st 1=24.3 s0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30t [s]Figure 18. Friction coefficient: CWT, n=600 rpm, d 0 =2 mmM fr /T [-] [-]10.90.80.70.60.50.40.30.20.10t 0=t ps ' =6.5 st st=19.3 st pt=11.68 st 1=24.3 s0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30t [s]Figure 19. Ratio M fr /T: CWT, n=600 rpm, d 0 =2 mm1.210.80.60.40.20t 0=5.3 st st=14.8 st ps'=8.2 st 1=18.95 s0 2 4 6 8 10 12 14 16 18 20 22 24 26t [s]Figure 20. Friction coefficient: CWT, n=600 rpm, d 0 =0 mm13 th International Conference on Tribology – Serbiatrib’13 393

M fr / T [-]10.90.80.70.60.50.40.30.20.10t 0=5.3 st st=14.8 st ps'=8.2 st 1=18.95 s0 2 4 6 8 10 12 14 16 18 20 22 24 26t [s]Figure 21. Ratio M fr /T: CWT, n=600 rpm, d 0 =0 mm [-]43.532.521.510.5t 0=t ps' =7.8 st st=12.25 st 1=15.8 s00 2 4 6 8 10 12 14 16 18 20 22 24t [s]Figure 22. Friction coefficient: CWT, n=910 rpm, d 0 =5 mmM fr / T [-] [-]10.90.80.70.60.50.40.30.20.10t 0 =t ps' =7.8st st=21.15 st 1=24.3 s0 2 4 6 8 10 12 14 16 18 20 22 24t [s]Figure 23. Ratio M fr /T: CWT, n=910 rpm, d 0 =5 mm1.210.80.60.40.20t 0=t pt=5.6 st ps'=9.4 st st=19.9 st 1=24.7 s0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30t [s]Figure 24. Friction coefficient: CWT, n=910 rpm, d 0 =3.2 mmM fr / T [-]10.90.80.70.60.50.40.30.20.10t 0=t pt=5.6 st ps'=9.4 st st=19.9 st 1=24.7 s0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30t [s]Figure 25. Ratio M fr /T: CWT, n=910 rpm, d 0 =3.2 mm [-]2.62.42.221.81.61.41.210.80.60.40.20t 0=t pt=6.2st st=31.4 st ps'=20.5 st 1=35.8 s0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 3032 34 36 38 40t [s]Figure 26. Friction coefficient: CWT, n=910 rpm, d 0 =2 mmM fr / T [-]10.90.80.70.60.50.40.30.20.10t 0=t pt=6 .2 st st=31.4 st ps'=20.5 st 1=35.8 s0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40t [s]Figure 27. Ratio M fr /T: CWT, n=910 rpm, d 0 =2 mm1.2t ps'=11.5 s1t 0=t pt=5.5t st=21.05 s [-]0.80.60.40.20st 1=26.05 s4 6 8 10 12 14 16 18 20 22 24 26 28 30 32t [s]Figure 28. Friction coefficient: CWT, n=910 rpm, d 0 =0 mm394 13 th International Conference on Tribology – Serbiatrib’13

M fr / T [-]10.90.80.70.60.50.40.30.20.10t 0=t pt=5.5 st ps'=11.5 st st=21.05 st 1=26.05 s0 2 4 6 8 10 12 14 16 1820 22 24 26 28 30 32t [s]Figure 29. Ratio M fr /T: CWT, n=910 rpm, d 0 =0 mmMeasured values of torque and forces were usedfor calculating values of the friction coefficient andratio of momentum of friction and the torque(Figure 6 to Figure 29).6. DISCUSSION AND CONCLUSIONSThe first set of experiments with the“theoretical” welding tool has shown that weldingtool without thread at the probe can not be used forwelding of 2024 T351 alloy. Plunging of thewelding tool into workpieces was possible onlywhen diameter of the welding tool was the same asthe diameter of the technological hole in theworkpieces. During such experiment, appeared thatfriction coefficient, after initial stabilization,reaches almost constant value between 0.3 to 0.4what is prescribed value for the FSW of AL 2024T351 and the “theoretical” welding tool [1]. Theratio of momentum of friction and the appliedtorque has a value of 1 – there is no (or hasminimal) deformation in the contact.The conclusion was that the plunging of the“theoretical” welding tool in welding plates wasimpossible when small or no technological holepresent what implies that welding couldn't even getstarted.The second set of experiments was conductedwith the coned, threaded welding tool with theprescribed technological parameters. Welding waspossible, however, only welding with n=910 rpmhas given the qualitative welds. During allweldings, trends and values of the frictioncoefficient were identical. Friction coefficient wasrising from the beginning of the plunging until themoment of time when the shoulder tip involves inthe welding. Common values of maximal frictioncoefficient reach about 1 but it is not uncommon toreach values of 2-5 (what is in correspondence withthe literature-present values [1-4] but it is possibleto have peak values as imperfections of theproposed method for estimation). From thatmoment, friction coefficient drops down and at thebeginning of the first dwelling phase reaches valueof about 0.2 to 0.7. However, at the end of thedwelling phase, friction coefficient in allexperiments reaches the values of 0.2 to 0.5.The ratio between the momentum of friction andthe applied torque has the same trend for allexperiments. It rises up to the maximal value of 1and varies from 0.8 to 1, during every conductedexperiment. The results are in agreement with theexisting results [2, 3, 4].LITERATURE[1] Mijajlović, M., Investigation and Development ofAnalytical Model for Estimation of Amount of HeatGenerated During FSW (in Serbian), Ph. D. thesis,Faculty of Mechanical Engineering Nis, Universityof Nis, Nis, Serbia, 2012.[2] V. Soundararajan, M. Valant, R. Kovačević: AnOverview of R&D Work in Friction Stir Welding atSMU, MJoM, Metalurgija – Journal of Metallurgy,Association of Metallurgical Engineers of Serbia,204, 12, pp. 277-295, 2006.[3] M. Djurdjanović, M. Mijajlović, D. Milčić, D.Stamenković: Heat Generation During Friction StirWelding Process, Tribology in Industry, 1-2, pp. 8-14, 2009.[4] K. Kumar, C. Kalyan, S.V. Kailasa, T.S. Srivatsan:An Investigation of Friction during Friction StirWelding of Metallic Materials, Materials andManufacturing Processes, 24, pp. 438-445, 2009.[5] M. Mijajlović, D. Stamenković, D. Milčić,M. Đurdanović: Study About Friction CoefficientEstimation in Friction Stir Welding, Balkantrib 11,The 7th International Conference on Tribology,Proceedings, Thessaloniki, Greece, pp. 323-330,2011.[6] M. Mijajlović, D. Milčić, B. Anđelković,M. Vukićević, M. Bjelić: Mathematical Model forAnalytical Estimation of Generated Heat DuringFriction Stir Welding. Part 1, Journal of BalkanTribological Association, 17, pp. 179-191, 2011.[7] P. Ulysse: Three-Dimensional Modeling of theFriction Stir-Welding Process, Int. J. Mach. Tool.Manu. 42, pp.1549-1557, 2002.[8] H. Schmidt, J. Hattel, J. Wert: An Analytical Modelfor the Heat Generation in Friction Stir Welding,Modeling Simul. Mater. Sci. Eng. 12,pp. 143-157, 2004.[9] M. Mijajlović, D. Milčić, B. Anđelković,M. Vukićević, M. Bjelić: Mathematical Model forAnalytical Estimation of Generated Heat DuringFriction Stir Welding. Part 2, Journal of BalkanTribological Association, 17, pp. 361-370, 2011.13 th International Conference on Tribology – Serbiatrib’13 395

Serbian TribologySocietySERBIATRIB ‘1313 th International Conference onTribologyKragujevac, Serbia, 15 – 17 May 2013Faculty of Engineeringin KragujevacMEASUREMENT INSTRUMENTATION FOR DETERMINATIONOF STATIC COEFFICIENT OF ROLLING FRICTIONPetar Todorović 1 , Ivan Mačužić 1 , Branislav Jeremić 1 ,Marko Đapan 1 , Branko Tadić 11 Faculty of Engineering Kragujevac, Serbia, petar@kg.ac.rs, ivanm@kg.ac.rs,bane@kg.ac.rs, djapan@kg.ac.rs, btadic@kg.ac.rsAbstract: This paper is considering the influence of temperature, at normal load and bend radius of contactelements on the coefficient of rolling friction. Contact pairs are made of steel DIN 17230 (100Cr6).Measurement results in a condition of high temperatures, variation of the normal load and bend radius ofcontact element indicate complex influence of temperature in this specific test condition. Authors’ futureresearch would be in direction of determination of static friction coefficient on the higher temperatures ofcontact pairs made of different materials.Keywords: Measurement instrumentation, coefficient of rolling friction, high temperatures, inclined plane1. INTRODUCTIONFor every engineer and constructor, who isengaged in design and development of mechanicalconstructions, knowledge of friction coefficient isvery important and crucial. However, there aremultiple issues, doubts and problems regardingusing of friction coefficient values duringexperiments. These problems occur primarilybecause of poor applying of standard tables andunder which conditions these values are measured.This is all because friction coefficient values aredifferent from laboratory to laboratory, and dependon equipment, measuring methods and a number ofother parameters that may influence on diversity ofmeasured values. Peter J. Blau [1] has representedreview of friction force and ways of itsmeasurement. He made a list of standardmeasurement methods for static and dynamicfriction coefficient as well as the way of itspotential use.As we know, friction occurs when two bodiesare in contact and based on velocity of relativemotion, friction can be static or kinetic. The staticfriction coefficient depends on many differentparameters, primarily from surface contact, normalload, atmosphere conditions and temperature,surface absorption, quality of processing andmaterial in contact [2-5]. There have been severalstudies regarding the influence of surface roughnessparameters with the static friction coefficient andconcluded that the coefficient of static friction willincrease if surface roughness coefficient increases[3, 4]. Also, some of them concluded that someroughness parameters, like skewness and kurtosis,have a greater influence on coefficient of staticfriction compared to other parameters [6, 7].Complete understanding of the coefficient of staticfriction is impossible without various analyses ofmechanisms under which this is occurring. Thisissue is a goal for numerous research efforts [8-10].As a start, some authors represented conditionsunder which the value of static friction is greaterthan the dynamic friction value, in terms oftemperature influence on creep motion. Generally,at temperature above zero, static friction coefficientis higher compared to kinetic friction coefficientdue to different heat activated processes. But, wecannot say that the static friction coefficient hasonly one value because it depends on contact andinitial velocity. In order to determine static frictioncoefficient, Chang et al. [8] analyzed adhesionforce and load in contact at rough metal surfaces.The study showed that the coefficient of staticfriction depends on characteristics of the materialand topography of the surface in contact as well asthat depends on external load versus generaldefined friction law. In this paper, researchers were396 13 th International Conference on Tribology – Serbiatrib’13

experimentally determined that for specific externalload, coefficient of static friction will decrease ifplastic characteristic of material increases andsurface energy decreases. D.-H. Hwang et al [11]concluded that the coefficient of static friction ishigher if contact pair is made of different material -steel/alumina, while the lower value is determinedfor similar materials (steel/steel). This result isconsequence of “stick-slip” effect. Also, one ofconclusions was that the influence of surfaceroughness has less influence for similar materials incontact pair, as well as the increasing value ofnormal load affects on increasing coefficient ofstatic friction in contact pair of different materials,while there is no significant influence for contactpairs of the identical materials.Etsion and Amit [12] experimentally researchedthe influence of normal load on coefficient of staticfriction with very smooth metal surfaces in acontrolled laboratory conditions. Dramaticallyincreasing coefficient of static friction was noticedwhen the normal load is on the lowest level.Behaviour like this is assigned to adhesion forceswhich have more important function regardingsmall normal loads and surface smoothness.A small number of papers deal with coefficientof static friction under influence of temperature[13-16]. The most important conclusion thatauthors made in this papers is that the coefficient ofstatic friction will increase if temperature isincreased, which is resulted of increasing plasticcharacteristics of the most contact material atincreased temperature. Reviewing the literature isnoticed that experimental tests of coefficient ofstatic friction were performed on experimentalequipment with different design, construction anddifferent contact geometry. Also, very interestingare measurement instruments for static coefficientof rolling friction [17-19]. Friction characteristicsof rolling bearing elements depend on contact pairmaterial, design, tolerance, topography of contactsurfaces and lubricants. Authors in this papernoticed, during literature review, that there are noany papers which based their research attempts onstatic coefficient of rolling friction, while inconditions at higher temperature referring to issueabove, there was no paper found (when this paper iswritten).The aim of this paper is to determine influenceof temperature on static coefficient of rollingfriction on contact elements made of steel.Experimental measurements were performed oninstrumentation that authors designed, developedand constructed regarding very precisedetermination of static coefficient of rolling frictionat higher temperatures and relatively small valuesof contact pressure with changing radius bends ofcontact elements.2. THEORETICAL CONSIDERATIONAccording to literature, the static coefficient offriction increases with increasing temperature. It isfound that temperatures above 200°C lead toincreasing of coefficient of friction which can beinterpreted as a result of increasing plasticcharacteristics of material at increased temperature.The static coefficient of rolling friction tested incondition of increased temperature has not beensubject of either theoretical or experimentalresearch. The authors will determine the influenceof temperature, normal load and radius bend ofcontact elements on coefficient of rolling frictionby experimental methods. According to that,measure instrumentation is designed andconstructed, based on inclined plane principle.In the case of rolling friction (contiguous case –figure 1), coefficient of rolling friction isdetermined from formula 1 and 2:where are:M N e(1)M PM moment of resistance andMP rolling moment.From equations 3 and 4: F R(2)eF N f N(3)Ref tanα(4)Rand from the body balance at inclined plate (figure2), we get the following equation:N G cosα(5)esinα R N e G cosα e tanα (6)Rwhere are:f static coefficient of rolling friction;N normal force;e coordinate that defines resultant reactionposition N;R radius of rolling body;G the force of gravity;α angle of inclined plane.13 th International Conference on Tribology – Serbiatrib’13 397

The tribometer operates in principle of inclinedplane. Contact pair together with system for heatingand probe for temperature measurement is rotatedfrom horizontal to the desired angle α. The rotatedangle of inclined plane is measured with readingprecision of one minute, which for a wide intervalof possible values of the coefficient of rollingfriction causes the measurement error less than 3%.Figure 1 – The balance of rolling body at inclined plateThe authors’ starting point was from theoreticalassumption that the contact between ball and flatsurface in laboratory conditions will be achieved onthe small number of unevenness in a regard anumber of unevenness at higher temperatures.Further, it is assumed that due to thermal expansionof material in the contact zone, will result asincreasing of value e (figure 1). This means that asa consequence we will have an increase of rollingmoment resistance and a parallel increase of thecoefficient of rolling friction. The authors believethat there is some correlation between staticcoefficient of rolling friction and the value ofthermal dilatation of contact pair. If we have inmind the stochastic nature of real contact area andnonlinear temperature field, it is hard totheoretically quantify the influence of temperatureon coefficient of friction. Hence, in order toquantify the influence of various parameters oncoefficient of friction, authors will provide relativeextensive experimental research.3. EXPERIMENTAL TESTSExperimental tests were performed on a specialdesigned and constructed tribometer. The completemeasuring instrument is showed on figure 2. Also,all positions are marked with numbers anddescribed in following text. The tribometer consistsof three bigger parts, as follows:1 – Thermoregulator. The main aim of this part is tovary a temperature (in our case is 200°C). There aretwo small screens; one is showing desiredtemperature and another current temperature.2 – Block with thermocouple. Inside of this block,beside thermocouple, there is a system for heatingand probe for temperature measurement. Also, inthis part of tribometer the contact between objectand block is made.3 – Counterweight. This part enables to makerotating of the block with thermocouple with verygood precision.Figure 2 – Measurement instrumentation(1-Thermoregulator, 2-Counterweight,3-Block with thermocouple)The tests were performed with rolling balls ofdifferent diameters over channels with differentradius bends. Balls weight and balls diameter werein a range from 0.04 to 0.08N and from 2.32 to13mm respectively. Bend radius of the blockcovered a range from 2.5 to 8 mm. Balls and blockwere heated on selected temperatures, 20, 100, 150and 200°C. Chosen material for balls and block wassteel DIN 17230 (100Cr6) with hardness 62-66HRC. Hardness is achieved by quenching andtempering process. Ball roughness is Ra=0.002µm.The roughness of the block channels surface was ina range between: Ra=0.8-1µm. The figure 3 is adiagrammatic representation coefficient of rollingfriction dependence regarding temperature andnormal load.Figure 3 – Coefficient of rolling friction dependenceregarding temperature and normal load398 13 th International Conference on Tribology – Serbiatrib’13

4. DISCUSSIONAccording to the theoretical consideration,physical principle and characteristics of inclinedplane for coefficient of static friction measurementcan be applicable in the conditions with highertemperatures. The measurement error is function ofthe angle α and value of friction coefficient, asfollows:tan(α Δα) tanα 100[%] (7)tan(αawhere are:ε - measurement error andΔα – measurement error of angle.Measured coefficient of friction is in the rangefrom 0.01 to 0.05 and reading precision is oneminute based on computation, the measurementerror is less than 3%. This result is totallyacceptable. The results of experimental tests enableglobal overview of how larger number ofparameters influence on coefficient of rollingfriction. Besides variations of temperature and levelof normal load, variations were made to blockchannel radius (the second contact element).From diagram (figure 3) we can conclude thattemperature, which is selected for contact pairheating and normal load (ball weight) has largeinfluence on changing trend of coefficient of rollingfriction. The bend radius has indirect influence onreal contact surface, meaning that larger radiuscorresponds to 10% of lower values of frictioncoefficient. If we have in mind that increasingradius of block bend increases contact pressure thenit can be concluded that results correspond withliterature. In conjunction with above stated, it canbe concluded that to lower coefficient of frictioncorresponds higher contact pressure.Based on analyses of experimental results,generally it can be stated that contact temperaturehas significant influence on coefficient of rollingfriction. However, level of temperature influenceon coefficient of rolling friction is highly dependentfrom normal load value, especially in an area oflower values of normal load.5. CONCLUSIONThe research in the field of static friction isspread in a number of directions. The topicexplored by authors aimed to draw attention thatresearch in a field of static coefficient of rollingfriction have not been carried out in order toquantify complex influence of normal load, contactsurface and temperature on coefficient of rollingfriction. Through theoretical considerationpresented in this paper, authors hypothesized thatthere is necessary thermal potential in a contactzone for redistribution of contact pressure andincrease of rolling moment resistance attemperatures around 200°C. The instrumentationused for static coefficient of rolling frictionmeasurement in a condition of high temperaturesfunctions as inclined plane and enables satisfactorydetermination results of static coefficient of rollingfriction. In this paper the measurement error is lessthan 3%, for performed program of experimentalresearch, and regarding problems of measurementof very small friction forces this is completelysatisfactory. The measurement results of staticcoefficient of rolling friction for selected materials,in a condition of high temperatures, normal loadand bend radius of contact elements variation,indicate a complex influence of temperature in thetesting conditions.Scientists’ future research in this field should bedirected to experimental tests of different materialsin contact and optimization in order to determineminimal values of static coefficient of friction athigh temperatures.ACKNOWLEDGMENTResearch presented in this paper was supportedby Ministry of Science and TechnologicalDevelopment of Republic of Serbia, GrantTR-35021, Title: Razvoj triboloških mikro/nanodvokomponentnih i hibridnih samopodmazajućihkompozita.REFERENCES[1] P. J. Blau, The significance and use of the frictioncoefficient, Tribology International, Vol. 34, No. 9,pp. 585–591, 2001.[2] B. Ivkovic, M. Djurdjanovic, D. Stamenkovic, TheInfluence of the Contact Surface Roughness on theStatic Friction Coefficient, Tribology in Industry,Vol. 22, No. 3&4, pp. 41-44, 2000.[3] U. Muller, R. Hauert, Investigations of thecoefficient of static friction diamond-like carbonfilms, Surface and Coatings Technology, Vol. 174-175, pp. 421–426, 2003.[4] B. Polyakov, S. Vlassov, L. M. Dorogin, P. Kulis,I. Kink, R. Lohmus, The effect of substrateroughness on the static friction of CuO nanowires,Surface Science, Vol. 606, No. 17-18. pp. 1393-1399, 2012.[5] N. Tayebi, A. A. Polycarpou, Modeling the effect ofskewness and kurtosis on the static frictioncoefficient of rough surfaces, TribologyInternational, Vol. 37, No. 6, pp. 491-505, 2004.[6] B. Bhushan, S. Sundararajan, W.W. Scott, S.Chilamakuri, Stiction analysis of magnetic tapes,IEEE Magnetics Transactions, Vol. 33, No. 5, pp.3211-3213, 1997.13 th International Conference on Tribology – Serbiatrib’13 399

[7] J. S. McFarlane, D. Tabor, Relation between frictionand adhesion, in: Proceedings of the Royal Societyof London. Series A, Mathematical and PhysicalScience, Vol. 202, No. 1069, pp. 244-253, 1950.[8] W.R. Chang, I. Etsion, D. B. Bogy, Static frictioncoefficient model for metallic rough surfaces,Journal of Tribology, Vol. 110, pp. 57-61, 1988.[9] B.N.J. Persson, O. Albohr, F. Mancosu, V. Peveri,V.N. Samoilov, I.M. Sivebaek, On the nature of thestatic friction, kinetic friction and creep, Wear, Vol.254, No. 9, pp. 835-851, 2003.[10] J. M. Galligan, P. McCullough, On the nature ofstatic friction, Wear, Vol. 105, No. 9, pp. 337-340,1985.[11] D.-H. Hwang, K.-H. Zum Gahr, Transition fromstatic to kinetic friction of unlubricated or oillubricated steel/steel, steel/ceramic andceramic/ceramic pairs, Wear, Vol. 255, No. 1-6,pp. 365-375, 2003.[12] I. Etsion, M. Amit, The effect of small normal loadson the static friction coefficient for very smoothsurfaces, Journal of tribology, Vol.115, No. 3, 1993.[13] H. Kumar, V. Ramakrishnan, S. K. Albert, C.Meikandamurthy, B.V.R. Tata, A.K. Bhaduri, Hightemperature wear and friction behaviour of 15Cr-15Ni-2Mo titanium-modified austenitic stainlesssteel in liquid sodium, Wear, Vol. 270, No. 1-2,pp. 1-4, 2010.[14] P. Mosaddegh, J. Ziegert, W. Iqbal, Y. Tohme,Apparatus for high temperature frictionmeasurement, Precision Engineering, Vol. 35, No.3, pp. 473-483, 2011.[15] A. Chaikittiratana, S. Koetniyom, S. Lakkam,Static/kinetic friction behaviour of a clutch facingmaterial: effects of temperature and pressure,World Academy of Science, Engineering andTechnology Vol. 66, 2012.[16] M. Worgull, J. F. Hetu, K. K. Kabanemi, M.Heckele, Hot embossing of microstructures:characterization of friction during demolding,Microsyst Technol, Vol. 14, pp. 767-773, 2008.[17] K. G. Budinski, An inclined plane test for thebreakaway coefficient of rolling friction of rollingelement bearings, Wear, Vol. 259, No. 7-12, pp.1443-1447, 2005.[18] Ta-Wei Lin, A. Modafe, B. Shapiro, R. Ghodssi,Characterization of Dynamic Friction in MEMS-Based Microball Bearings, IEEE Transactions oninstrumentation and measurement, Vol. 53, No. 3,pp. 839-846, 2004.[19] D. N. Olaru, C. Stamate, A. Dumitrascu, G.Prisacaru, New micro tribometer for rolling friction,Wear, Vol. 271, No. 5-6, pp. 842-852, 2011.400 13 th International Conference on Tribology – Serbiatrib’13

Serbian TribologySocietySERBIATRIB ‘1313 th International Conference onTribologyKragujevac, Serbia, 15 – 17 May 2013Faculty of Engineeringin KragujevacIMPLEMENATION SQL REPORTING SERVICE IN THETRIBOLOGYCAL DATA BASESMilan Erić 1 , Marko Djukić 11 University of Kragujevac, Faculty of Engineering, Serbia, ericm@kg.ac.rs, djuka84yu@gmail.comAbstract: In the work it will be presented one of the way how to publish the results of the scientific andresearch work quickly and efficiently and these results are saved in thedata bases,and this is calledtribologycal researches. The developed technology in the Reporting Services enables usto avoid the writingof the application programmes or using the data in the other section of the softver,type Statistica or Excel,and it enables us directly to form th presentation part over the bases.Scientists and researchers,as the bestway of scientific communication,have both the role of the creators and the users and they have started to bepublishers anddistributers,and this technology enables them all of this.Keywords: tribological data bases, SQL Reporting Services, processing of the reports1. INTRODUCTIONToday, in the world of inrormation technology,reports are the main key so that we can publish theresults in the scientific researches. It can be said thatthe reports are the final and main step of the long andcomplex process of collecting, keeping,transformating and manipulating of the data.Creating of the reports is presentation of workingwith the data bases. All reports are not the same. Thevalue of the report is information. Information arenot just the data,but they are the data transformedintosomething usefull,something that has value. Thistransformation is extremely important. People canread and publish the data in many different ways andthat is the reason why the data are all around us,butwhat we need in the 21st century to to complete ourjob is well valued, correct, quick and appropriateinformation.There are a lot of software tools for creating thereports that enable everyone to access to ana opennumber of the data which are all around us.However, all these who use these data are notfamiliar to the work technics-and that can be a hugeproblem. The data used to be saved and transferedorally, in the written form and today mostly in anelectronic form in our computers-in the data basesbut the data are not usually from the same base andthey do not come from he same base. Extremelysmall number of reports actually has the data in thebase.Today if you want a report to be valued as goodand to be in the terms of standards of using theinfotmation technologies, it must be reliable, quick,to have a good presentation, to have flexiblefom,connectivity and in the end that it can be usedby yhe correct tool. One of the tool or technologywhose develompment still is in a progress isReporting Service (RS) which is the part ofSQLServer.All tools ana applications in RS are made usingthe API (Application Programming Interface).Reporting Services contains all that is necessary thatresearchers and well trained business users to publisha report. Completed reports are guided by a serverwhere are they. The final users to whom thesereports are mare for, have an efficient and full report.The process of correction and analyses by usingthe information of Reporting Service leads togeneration of knowledge of the data and it is mostlyknown in the IT world as Distributed intelligence(knowledge technology). In this case, ReportingServices can be seen as server-based platform withthe developed tools for generating, manipulationand publishing of the reports.Reporting Services evoluated into a sofisticatedreporting platform which gives new abilities ofthe efficient analyses and an atractive presentation13 th International Conference on Tribology – Serbiatrib’13 401

of information which are saved on the hard discs ofthe server, opening complely new dimension ofworking with the the data and reports.Reporting Services must be understandable, wemust be able to read them and they must point tothe date that we need for the analyses andverification of the results. To achieve the goal wewant,we can design the report that specific datashows like a table/chart or any other form that canbe understood, Figure 1.trends. The data are not for the use unless there is away that they can be shown in a way that the userscan understand them. Presentation Layer platformsenable different ways of presentation like MicrosoftOffice, Microsoft SharePoint, MicrosoftPerformance Point Server or some othercomparatible applications.Basic components and logic architecture ofReporting Services are shown in the Figure 3 [1].Figure 1. The picture of the standard report.The reports can be also shown in an ad hoc form.2. PROCESSING OF THE REPORTS USINGTHE REPORTING SERVICESReporting Services is a part of MS SQLplatform which offers opportunities of processingand manipulating the data, Figure 2 [1].Figure 2. Architecture of the MS SQL Server platform.Database Engine is for packing, processing andsecuring of the data. Integration Services supportsdifferent typrs of the data, which have the samesource of the dataand technologies as well as theirintegration Integration Services is mainly used fortransfering,transformation and reading of the data.Packets of the data of Integration Services aremainly used as sources for the reports. AnalysisServices represents multi-dimension base for thequick reporting and generating of the questions andFigure 3. Reporting Services logical architectureIn the centre of Reporting Services architectureis a server , web-orientied middle part whichaccepts the requests,processes them and on the baseof that generates the reports. An illustration shows asimple sheme of Report Server. Repot Servercommunicate with the the users in two ways: by urlor through web service. The component Reportprocessor is responsible for processing of thereports in so called run time. This means that thereport sends the data to an user,combining the datafrom the base with the parametres making the finalreport sent in the requested form.An important characteristic of ReportingServices is that the archecture can be enlargedthrough special modules which are calledextensions. When the standard extensions are notenough, programmers can extend theopportunities of RS by puttingin their ownextensions. Like the sources of the data,the userscan export the report results in a several mostpopular forms like Microsoft Excel, MicrosoftWord, Adobe Acrobat PDF, HTML, SCV, it canbe shown in the pictures or the new extensionscan be written for the sending of the report or insome other forms.The definition of the report and its adjuctmentsare saved in the data base of Report Server. ReportServer is implemented as two SQL Server bases(Report Server and Report Server DB) which areinstalled during their configuration. When weupload the report, Reporting Services saves thedefinition of Report Server in the data base whilethe other data base – Report Server DB containsand saves temporary information on the report andits thruthfullness.402 13 th International Conference on Tribology – Serbiatrib’13

2.1 The life cycle of the reportThe life cycle of the report is the events or theactivities of the report, start from the moment whenwe start creating it. In the Figure 4 we can see thatthe life cycle is made of Authoring, Managementand Delivery phases [1].requests and sends them to an application thatshould answer to these requests. As a part of theconfiguration of the Reporting Services, it must besaid that URL address report server i ReportManager. Reporting Services Windows service hasthree server applications: Report Manager, ReportServer Web Service and Background Processor.Behind the scene, this service in fact creates threenet applications which will host them.Figure 4. Work with the reportIn the Authoring phase, the author of the reportuses one of the Microsoft designer reports (ReportBuilder). When the report is completed, the authorcan upload it so it can be seen by the final users. Inthe Management phase, the administratorconfigurates the generated reports and developingsurrounding where it is going to be shown. Theadministrator can use Report Manager to organisethe report in the folders as well as to set the securitymeasures so that the access can be authorised to theusers. When it is configurated, the reports can beseen to only those to whom this right is authorised.The report can be seen by the final users typingURL address in the web searcher or alternatevilyusing the option schedule -through some channallike an e-mail.The designers of the report are the tools whichthe authors use for the definition of the data looks atthe moment of creating the reports. Since thetechnological knowledge and the experiences of theauthors can vary, it is not easy to create a designerreport that can satisfy the need of the all users.In theFigure 5 allthe designer tools for the creating of thereport are shown with their basic characteristics [2].Figure 5. Tools for creating the reports and theircomparisonIt is important to say that all mentioned reportdesigners support RDL standard (Report DefinitionLanguage).2.2 Physical architecture of Reporting ServiceIn the Figure 6 we can see the physicalarchitecture of the Reporting Services, which ismade of three Report Server aplications: ReportManager, Report Server Web Service andBackground Processor. In the physical architecturewe can see an implemented network interfacewhich includes Service Network Interfaces (SNI)which checks new requests HTTP.SYS.HTTP.SYS je HTTP driver which accepts theFigure 6. Reporting Services 2008 architectureReport Manager is an ASP.NET web applicationwhich enables management and the look into theabilities of the Reporting Services instanceconfigurated in the natural code. We can see ReportManager as a client application configurated withthe report of the server. Thanks to the same hostingmodel, configuration adjustments of the ReportManager and Report Server Web service are kept inthe same configuration file in rs report server.confingDue to this, Report Manager can add somenew extensions. For example, if the user developsnew extension, using C# or VB.NET we canconfigurate in the Report Managers a web controland later use it as we adjust the details of the report.Report Server Web service processes the reportsby using the systems on-demand. When the userclicks on the link pages where the reports are,he\she sends the requests to the Report Web,service accepts this request, processes this requestand returns the report to the client. To make iteasier integrations with the different types of thereports, Report Server Web service enables the useof URL and SOAP protocol and their integrationoptions.Background Processor is an application whichjob is to accept all the tasks which are in anunmarked mode. For example, when the descriptionof the event is accepted, Background Processorinterprets the description of the report and sends itto the final destination. Basically, its job is to13 th International Conference on Tribology – Serbiatrib’13 403

process the reports, not to communicate withReport Server Web service. Instead of this, both ofthe applications communicate with the ReportProcessor in the same time.Report Processor does not save the whole reportin the memory, but it is processing the report ondemand, as it is shown in the Figure 7.After the connection is finished, it is possible toprepare the data by hand, or write SQL request,which will be use full for the report. Projection willdefine the columns and selection will define thelines of the data base which will be used in thereport, Figure 9 [2].Figure 7. Grafic picture of the processing and the reportAt the moment when Report Processor notesthe new request, (request), it will take the data andit will match them with the report template makingthe middle form of the report. That report,processor saves in the Report Server Database.Point is that Report Processor takes and saves onlythe parts of the report, for example, grouping,sorting and etc. In the phase of the investigating ofthe report and saving it, Report Processor usesRender Object Model (ROM) an object which isforming the form that we can show. Textbox valuesand data are processing every time on-demandwhen we want to see the report.2.3 The connection of the tribological bases ofthe data with the reportsTribological data base, whose logic structureand content are shown in the [2], are connected toalready formed template file for the creating of thereport. Working surrounding is Visual Studio,where we can connect the data base to the reportfrom the file Report.rdl created in SQL ServerBusiness Intelligence Developer Studio. Thestarting point of the model form of the connectionis shown in the Figure 8 [3,4].Figure 9. Defining of the requests for the data selection.In the next step, we will select the way of theshowing the data. Presentation of the data by thediagram versus the charts has more visual effects.The diagram can be added in one or two ways,moving control Chart from Toolbox or by pressinga click of the right mouse on the desktop Insert →Chart, Figure 10 [3,4].Figure 10. Selection of thr type of the diagram.After we had sellected the diagram, we need toconnect the Design Body in the dialogue windowwith the values which are chosen from the data basewith the pararmetres of the NET control whichdescribe the centres of the diagrams.When it isclicked on the diagram, it will show in theright,down part where we need to put the columns- results of the requestsfrom the data bases (parts inthe command SELECT), Figure 11 [3,4].Figure 8. Adding Data Source…404 13 th International Conference on Tribology – Serbiatrib’13

Obviously the report is checked and the values areeasily seen, can be seen and are clearly shown.3.CONCLUSIONReporting services is a complex and moderntechnology with the tendency for furtheradjustments and enlargement. Because of the needand implementation for these kinds of technologies,it grows the need that we must know and use them.Reporting services is a great technology whichmakes the job easier to the business world, and itcan also be of great need to the scientists while theyare publishing the results from the data base, and itfacilitates the work to the IT techinicians.Figure 11. Adjustments of the centers of the parametersof the diagramsAfter we turn off View mode, BI platform will givethe Render report by processing the request andmaking the graphic interpretation based parametersand adjustments. In the Figure 12 we can see thefinal report of the base of the data,TRIBOLOGYCAL_RESULTS.Figure 12. The final report after the adjustment of theparametersREFERENCES[1] T. Lachev: Applied Microsoft SQL Server 2008Reporting Services, Prologika Press, USA, 2008.[2] M. Erić, S. Mitrović, M. Babić, F. Živić, M. Pantić:Application of Contemporary InformationTechnologies in Nanotribometry, Tribology inIndustry, Vol. 33, No. 4, pp. 159-163, 2011.[3] B. Larson: Delivering Business Intelligence withMicrosoft SQL Server 2008, The McGraw-HillCompanies, San Francisco, 2009.[4] B. Larson: Microsoft SQL Server 2008 ReportingServices, The McGraw-Hill Professional, SanFrancisco, 2009.[5] P. Turley, T. Silva, B. Smith and K. Withee:Professional Microsoft® SQL Server® 2008Reporting Services, Wiley Publishing, Indianapolis,2009.[6] P. DeBetta, G. Low, M. Whiteborn: IntroducingMicrosoft SQL Server 2008, Microsoft Press, 2008.[7] L. Davidson, K. Kline, S. Klein, and K. Windisch:Pro SQL Server 2008 Relational Database Designand Implementation, Springer-Verlag, New York,2009.13 th International Conference on Tribology – Serbiatrib’13 405

Trenje, habanje ipodmazivanje13 th International Conference on Tribology – SERBIATRIB ’1315 – 17 May 2013, Kragujevac, Serbia

Serbian TribologySocietySERBIATRIB ‘1313 th International Conference onTribologyKragujevac, Serbia, 15 – 17 May 2013Faculty of Engineeringin KragujevacVEŠTAČKO STARENJE TIKSOLIVENE ZA27 LEGURE IČESTIČNIH ZA27/SIC KOMPOZITAI. Bobić 1 , M. Babić 2 , A.Vencl 3 , S. Mitrović 2 , B. Bobić 41) INN "Vinca", Univerzitet u Beogradu, Beograd, Srbija, ilijab@vinca.rs2) Fakultet inženjerskih nauka, Univerzitet u Kragujevcu, Kragujevac, Srbija, babic@kg.ac.rs, boban@kg.ac.rs3) Mašinski fakultet, Univerzitet u Beogradu, Beograd, Srbija, avencl@mas.bg.ac.rs4) Institut "Goša", Beograd, Srbija, biljanabobic@gmail.comAbstrakt: ZA27 legura sa nedendritnom strukturom dobijena je tiksokasting postupkom. Čestični kompozitisa osnovom od navedene legure dobijeni su kompokasting postupkom, odnosno infiltracijom 5 vol.% i 10vol.% SiC čestica u poluočvrsli rastop legure. Uzorci tiksolivene ZA27 legure i kompozita bili su podvrgnutiprocesu starenja na 80, 120 i 160°C (T5 režim). Strukturna ispitivanja matrične legure i kompozita i merenjapromena tvrdoća tokom procesa starenja izvršena su u zavisnosti od temperature. Pokazano je da prisustvočestica ojačivača utiče na ubrzanje procesa starenja, odnosno kompoziti dostižu maksimalne vrednostitvrdoće za kraće vreme nego matrična legura. Sa povećanjem temperature starenja opadaju tvrdoće svihispitivanih materijala.Ključne reči: ZA27 legura, čestični kompoziti, kompokasting, veštačko starenje1. UVODZbog dobre kombinacije fizičkih, mehaničkih itehnoloških karakteristika ZA27 legura ima velikikomercijalni značaj. Koristi se izradu niza odlivakarazličlitih dimenzija. Takođe, poznata je kao leguraza izradu kliznih ležajeva [1, 2]. Svojstva leguremogu se modifikovati termičkom obradom.Standardom [3] definisana je samo jedna varijantatermičke obrade ZA27 legure. Proučavanje uticajaraznih režima termičke obrade na svojstva ZA27legure bilo je predmet rada više istraživača [4–6].Budući da u livenom stanju legura sadrži presićene,metastabilne faze, osnovni cilj istraživanja bio je dase poboljša homogenost legure i postigne dobrakombinacija mehaničkih svojstava. Kod ZA27legure koja sadrži Cu [2], cilj je bio da se poboljšadimenziona stabilnost proizvoda.Zbog širokog temperaturnog intervalaočvršćavanja (oko 100 ºC [1]), ZA27 legura jepogodna za preradu u poluočvrslom stanju,različitim postupcima. U zavisnosti od primenjenihprocesnih parametara, moguće je dobiti leguru čijustrukturu karakteriše nedendritna morfologija [7].Pokazano je da čestični kompoziti ZA27/SiCposeduju bolje fizičke i mehaničke karakteristike uodnosu na matričnu leguru [8, 9], kao i većuotpornost prema habanju. Posebno je proučavanuticaj procesa starenja na svojstva navedenihkompozita [10–12].Procesi starenja imaju za cilj da omogućestabilizaciju strukture i da poboljšajudimenzionalnu stabilnost proizvoda pre njihoveeventualne upotrebe na povišenim temperaturama.Ovo je posebno značajno kada se ima u vidunepovoljan uticaj bakra na stabilnost dimenzijaproizvoda od ZA27 legure na povišenimtemperaturama [1].Kompokasting postupak primenjen je zadobijanje čestičnih kompozita ZA27/SiC, koji supredmet ovog rada, s obzirom da je relativnojednostavan i perspektivan za dobijanje jeftinijihkompozitnih materijala.Cilj ovog rada je da se ispita uticaj veštačkogstarenja (u oblasti temperatura od 80 do 160°C) namikrostrukturu i tvrdoću tiksolivene ZA27 legure iZA27/SiC kompozita, koji su dobijenikompokasting postupkom. Rezultati ispitivanjapredstavljaju doprinos boljem razumevanju procesa13 th International Conference on Tribology – Serbiatrib’13 409

starenja navedenih materijala, što je od značaja zanjihovu praktičnu primenu.2. EKSPERIMENTALNI RAD2.1 Tikso /kompokasting postupciTiksokasting i kompokasting izvršeni supomoću aparature koja je ranije opisana [13].Za potrebe eksperimentalnog rada korišćena jelegura čiji hemijski sastava odgovara standardu [3].Tiksokasting postupak sastojao se iz dve faze. Uprvoj fazi dobijeni su odlivci ZA27 legure, koji su udrugoj fazi podvrgnuti toplom presovanju. Tokomprve faze vršeno je mehaničko mešanjepoluočvrslog rastopa ZA27 legure na temperaturiod 460°C. Na početku je primenjena brzinamešanja od 500 o/min, u trajanju od 2,5 min, u ciljuhomogenizacije poluočvrslog rastopa legure. Zatimje u toku 10 min vršeno intenzivno mešanje rastopa(pri brzini mešanja od 1200 o/min), bez promenetemperature. Po završetku mešanja, poluočvrslirastop legure je izliven u čeličnu kokilu,predgrejanu na 350°C. Dobijeni su odlivcidimenzija 30x20x120 mm, od kojih su mehaničkiizrađeni manji uzorci (30x20x5 mm), koji supodvrgnuti toplom presovanju. Toplo presovanje jeizvršeno pomoću specijalnog alata odtoplootpornog čelika, na 350°C, primenom pritiskaod 250 MPa. Posle toplog presovanja dobijeni suuzorci dimenzija 30x20x6 mm.Kompokasting postupak je, takođe, izveden udve faze; u prvoj fazi izvršeno je dobijanje odlivakaZA27/SiC kompozita, koji su u drugoj fazipodvrgnuti toplom presovanju. U okviru prve fazeizvršena je priprema poluočvrslog rastopa matričnelegure, unošenje SiC čestica (prosečne veličinaprečnika 24 μm) u poluočvrsli rastop i mešanjepoluočvrsle kompozitne smeše. Pri dobijanjukompozita sa 5 vol.% SiC čestica (u daljem tekstukompozit K1) vreme unošenja SiC čestica bilo je3,5 min, pri brzini mešanja 500 o/min. Pri dobijanjukompozita sa 10 vol.% SiC čestica (u daljem tekstukompozit K2) primenjena je ista brzina mešanja,ali je unošenje SiC čestica u poluočvrsli rastopmatrične legure trajalo 7 minuta. Da bi se smanjioukupni viskozitet poluočvrslog rastopa, povećana jeradna temperatura tokom unošenja SiC čestica; od465 do 470°C, u slučaju kompozita K1 i od 465 do475°C u slučaju kompozita K2. Po završenomunošenju ojačavajućih čestica izvršeno je kratko,homogenizaciono mešanje (2,5 min, pri brzini od500 o/min), a zatim intenzivno mešanje (1000o/min) u narednih 10 min. Posle toga izvršeno jeizlivanje kompozitnih masa u čeličnu kokilu,predgrejanu na 350°C.U drugoj fazi izvršeno je toplo presovanje, priistim parametrima koji su primenjeni kodtiksolivene ZA27 legure. Dobijeni su uzorci istihdimenzija, kao uzorci tiksolivene ZA27 legure.2.2. Strukturna ispitivanja i merenje tvrdoćeIspitivanje strukture vršeno je optičkommikroskopijom (pomoću optičkog mikroskopa CarlZeiss) i skening elektronskom mikroskopijom(pomoću skening elektronskog mikroskopa JEOLJSM – 5800).Ispitivanja su vršena na uzorcima dimenzija15x15x6 mm, koji su mašinski izrađeni iz otpresakatiksolivene matrične legure i kompozita. Uzorci subrušeni pomoću brusnih papira (80, 360 i 600grita), dok je poliranje obavljeno primenom tkanineza poliranje i paste (sa Al 2 O 3 česticama). Ispitivanjapomoću SEM rađena su na poliranim uzorcima,dok su ispitivanja primenom OM vršena napoliranim i nagriženim uzorcima. Za nagrizanje jekorišćen vodeni rastvor HNO 3 (9 v/v).Merenje tvrdoće izvršeno je pre nego što suuzorci matrične legure i kompozita bili podvrgnutiprocesu starenja, a zatim tokom procesa starenja, nauzorcima dimenzija 15x15x6 mm. Nezavisno odtemperature starenja, najveći broj merenja tvrdoćeobavljen je tokom prvog sata starenja. Merenje jezatim vršeno u određenim vremenskim intervalima.Za merenje je korišćen uređaj za merenje tvrdoćeKarl Frank GMBH. Rezultati merenja izraženi su uBrinelovim jedinicama (HB). Na svakom uzorkuvršeno je po pet merenja, za svaki vremenskiinterval tokom procesa starenja.3 REZULTATI I DISKUSIJA3.1 Strukturna ispitivanjaNa slici 1a prikazan je izgled mikrostrukturetiksolivene ZA27 legure. Mikrostruktura se sastojiod kompleksnih eliptičnih i globularnih polifaznihzrna. Zrna se sastoje od jezgra (svetle zone) koječini faza i periferije (smeša +η faza, sive zone).Heksagonalna η faze, praktično cink, nalazi seizmeđu eliptičnih zrna (tamne zone).Mikrostruktura prikazana na slici 1 posledica jeuticaja mešanja poluočvrslog rastopa legure (tokomprve faze tiksokasting postupka) i očvršćavanjapoluočvrslog rastopa ZA27 legure po završetkumešanja. Pod uticajem sila smicanja (proizvedenihmešanjem), kao i usled interakcija primarnih česticameđusobno i interakcija mešač–primarne čestice,zid lonca–primarne čestice, došlo je do usitnjavanjai transformacije primarnih čestica premaFlemingsovoj šemi [7].410 13 th International Conference on Tribology – Serbiatrib’13

uticaja na osnovu kompozita, a time i na processtarenja. Mesto čestica u strukturi kompozita zavisiod procesnih parametara u prvoj fazi kompokastingpostupka, kao i od načina očvršćavanja kompozitnemase posle završenog mešanja.3.2 TvrdoćaSlika 1. Mikrostruktura tiksolivene ZA27 legure(OM, nagriženo)Rezultat toga je da su u strukturi tiksoliveneZA27 legure prisutne primarne čestice eliptičnog ikružnog oblika, što znači da je došlo promenemorfologije. Međutim, osnovni elementimikrostrukture, u pogledu faza koje nastaju tokomočvršćavanja, kvalitativno su isti kao elementi umikrostrukturi livene ZA27 legure [1].Na slici 2 prikazan je izgled mikrostrukturekompozita K1 i kompozita K2.Promene tvrdoće sa vremenom starenjaprikazane su pomoću dijagrama na slici 3 (a–c), zasve ispitivane materijala i temperature starenja: 80,120 i 160°C. Starenje na 80 i 120°C vršeno jetokom 25 časova, dok je na 160°C starenje trajalo 5časova.Slika 2. Mikrostrukture čestičnih kompozita ZA27/SiC(SEM, polirano).a) K1 (ZA27+5 vol.%SiC), b) K2 (ZA27+10 vol.%SiC).Na slici 2a prikazana je mikrostrukturakompozita K1. Postignut je relativno dobarraspored SiC čestica u metalnoj osnovi. Izmeđučestica ojačivača vidi se struktura osnovekompozita (tiksolivena ZA27 legura), koja je ranijeopisana.Na slici 2b prikazana je struktura kompozita K2.Može se zapaziti prisustvo nakupina SiC čestica,što znači da se njihova pojava nije mogla izbeći pridobijanju kompozita sa navedenim parametrimakompokasting postupka. Pored nakupina SiCčestica, na slici se mogu videti i SiC čestice kojenisu u međusobnom kontaktu. Čestice su smešteneu oblasti η faze i zalaze u oblast smeše faza +η.Položaj SiC čestica je značajan zbog njihovogSlika 3. Promene tvrdoća tiksolivene ZA27 legure ikompozita K1 i K2 tokom starenja. Temperaturastarenja: a) 80°C, b) 120°C, c) 160ºC. Krive: 1-tiksolivena ZA27 legura, 2 - K1, 3 - K2.Na slici 3a prikazana je vremenska zavisnosttvrdoće na temperaturi starenja od 80ºC, zatiksolivenu ZA27 leguru i kompozite K1 i K2.Početna tvrdoća tiksolivene ZA27 legure (120 HB)13 th International Conference on Tribology – Serbiatrib’13 411

lago raste tokom prvih 10 min starenja. Tvrdoćazatim neprekidno opada sledećih 50 min, a posle 60minuta starenja počinje da raste. Tiksolivena ZA27legura postigla je maksimalnu vrednost tvrdoćeposle 5 časova starenja. Od tada, pa do krajastarenja, tvrdoća neprekidno opada, tako da jenajniža vrednost tvrdoće za tiksolivenu ZA27leguru izmertena posle 25 časova starenja.Pri starenju na 80°C tvrdoće oba kompozita (K1i K2) u početku rastu veoma brzo, tako da sumaksimalne vrednosti tvrdoće dostignute već posle10 min starenja na ovoj temperaturi. Posle togadolazi do brzog pada vrednosti tvrdoće, pri čemu jeoko 30 min primećen zastoj, a zatim vrednostitvrdoće pokazuju diskontinuitet u pogledu trenda,do isteka 60 min starenja. Posle ovog vremena,tvrdoće kompozita K1 i K2 kontinuirano opadajudo kraja prdviđenog vremena starenja.Na osnovu objavljenih rezultata drugih autora[14], možemo tvrditi da se, u slučaju kompozita,dislokacije na granici čestica/matrica ponašaju kaocentri kristalizacije čestica koje difunduju tokomstarenja. Na taj način, povećao se broj centarakristalizacije taloga [11]), što utiče na ubrzanjeprocesa starenja. Usled toga, došlo je do bržegdostizanja maksimalnih vrednost tvrdoće kodkompozita, nego kod tiksolivene osnovne legure.Sa porastom temperature starenja, menjaju sevrednosti tvrdoće. Na slici 3b prikazana jevremenska zavisnost promene tvrdoće, pri starenjuna 120°C, za tiksolivenu ZA27 leguru i kompoziteK1 i K2. Tvrdoća legure blago raste tokom prvih 5min starenja. Sa dijagrama se vidi da se na daljetvrdoća menja diskontinuirano. Mogu se uočiti dvamanja pika, jedan posle 60 min, a drugi posle 10časova starenja. Posle toga, tvrdoća tiksolivenelegure opada i najniža vrednost izmerena je posle20 časova starenja.ZA27 legura ima relativno nisku tačku topljenja(380°C, odnosno 653K). Zatezna ispitivanja napovišenim temperaturama [1] pokazala su da se većiznad 80°C čvrstoća pogoršava, a izduženjepovećava. Takođe, rezultati ispitivanja pritisnihkarakteristika livene i tiksolivene ZA27 legure [15],pokazuju da dolazi do značajnog pada vrednostigranice popuštanja σ 0,2 , kada se dostignetemperatura od 80°C.Pri povećanju temperature starenja od 80 na120ºC, značajno se intenziviraju procesi starenja.Ovo se odrazilo na promene vrednost tvrdoćetokom starenja na 120°C, kako kod matrične legure,tako i kod kompozita K1 i K2. U slučaju kompozitaK1, maksimalna vrednost tvrdoće dostignuta jeposle 20 min starenja. Tvrdoća zatim relativno brzoopada do isteka prvih 60 min starenja. Tokomsledećih 15 časova vrednost tvrdoće se ne menjaznačajno. Posle 25 časova starenja izmerena jenajniža tvrdoća kompozita K1. Tok promenatvrdoće kompozita K2 sličan je opisanom tokupromena kod kompozita K1, ali postoji razlika uvrednostima tvrdoće. Kompozit K2 dostižemaksimalnu vrednost tvrdoće posle 20 min starenjana 120°C. Tvrdoća zatim opada do kraja trećegsata starenja, posle čega nema znatnijih promenasve do isteka 15 sati starenja. Posle toga, tvrdoćaponovo opada i dostiže najnižu vrednost posle 25časova starenja.Opšte je poznato da je u uslovima starenja nanižim temperaturama brzina nukleacije talogavelika, ali je brzina rasta mala. Obrnuto, na višimtemperaturama starenja brzina nukleacije taloga jemanja, ali je brzina rasta velika. Prema tome, nanižoj temperaturi starenja broj centara nukleacije jeveći, međučestični prostor se smanjuje, što dovodido povećanja tvrdoće. Na osnovu ovoga, mogla seočekivati manja tvrdoća kod svih materijala koji sustareni na 120°C, u odnosu na njihovu tvrdoću pristarenju na 80°C. Ovo očekivanje ispunjeno je uslučaju tiksolivene ZA27 legure i kompozita K1,što je u saglasnosti sa [11, 14].Analizom krivih promene tvrdoće tokomstarenja kompozita, može se videti da semaksimalne vrednosti tvrdoće, u principu,povećavaju sa sniženjem temperature starenja ipovećanjem količine čestica ojačivača. Međutim, ukonkretnom slučaju starenja kompozita K2 na120°C, zapažen je paradoks, koji je primećen ranije[15]. Naime, maksimalna tvrdoća kompozita K2 pristarenju na 120°C nešto je veća od maksimalnetvrdoće istog kompozita pri starenju na 80°C.Pri starenju na 160°C (slika 3c) tvrdoće svihispitivanih materijala opadaju u odnosu na tvrdoćepostignute tokom starenja na 80 i 120°C. Napočetku starenja na 160°C, tvrdoća tiksoliveneZA27 legure opada u odnosu na polaznu vrednost.Posle 1 časa od početka starenja tvrdoća malo raste.a zatim neprekidno opada do kraja starenja. Sličnapromena tvrdoće konstatovana je pri starenjukompozita K1. Ovo navodi na zaključak da sudifuzioni procesi na ovoj temperaturi značajnoubrzani. Kod kompozita K2 dolazi do povećanjatvrdoće u odnosu na polazno stanje, tako da seposle 20 min starenja na 160°C dostiže maksimalnatvrdoća. Posle toga, tvrdoća kompozita K2neprekidno opada do kraja ispitivanja.Rezultati koji su dobijeni u ovom radu, ukazujuna mogućnost postizanja maksimalne tvrdoćekompozita posle relativno kratkog vremenastarenja.412 13 th International Conference on Tribology – Serbiatrib’13

5. ZAKLJUČCIStruktura tiksolivene ZA27 legure grublja je uodnosu na strukturu livene legure, što se odražavana usporavanje procesa starenja tiksolivene legure.Prisustvo čestica ojačivača utiče na ubrzavanjeprocese starenja, na svim primenjenimtemperaturama starenja.Kompoziti veoma brzo dostižu maksimalnevrednost tvrdoće, što je od velikog značaja zanjihovu praktičnu primenu.ZAHVALNOSTAutori se zahvaljuju Ministarstvu za obrazovanje,nauku i tehnološki razvoj Republike Srbije koje jefinansijski podržalo ovaj rad kroz projekte TR35021 i OI 172005. Takođe se zahvaljuju LivniciRAR ® doo, Batajnica (Beograd, Srbija) i Fabriciabrazivnih proizvoda Ginić Tocila ® doo, Barajevo(Beograd, Srbija), koji su obezbedili master leguru iSiC čestice, za ovo istraživanje.LITERATURA[1] E. Gervais, R.J. Barnhurst, C.A. Loong, AnAnalysis of Selected Properties of ZA Alloys,JOM 11 (1985) 43–47.[2] V. G. S. Mani, P. Sriram, N. Raman, S. Seshan,AFS Transactions, 88-100, pp.525-532, 1988.[3] BS EN 12844:1999 Zinc and zinc alloys.Castings. Specifications.[4] S. Murphy, T. Savaskan, Metallography ofZinc-25 % Al Based Alloys in the As–Cast andAged Conditions, Prakt. Metallogr. 24 (1987)204–221.[5] I. Bobic, B. Djuric, M.T. Jovanovic, S. Zec,Improvement of Ductility of a Cast Zn–25Al–3Cu Alloy, Mater. Charact. 29 (1992) 277–283.[6] P. Choudhury, K. Das, S. Das, Evolution of ascastand heat-treated microstructure of acommercial bearing alloy, Mater. Sci. Eng., A398, pp. 332–343, 2005.[7] M. Flemings, Behavior of metal alloys in thesemi-solid state, Metall. Transactions A, 22A,pp. 957-981, 1991.[8] I. A. Cornie, R. Guerriero, L. Meregalli, I.Tangerini, Cast Reinforced Metal Composites,ASM International USA, pp. 155-165, 1988.[9] N. Karni, G. B. Barkay, M. J. Bamberger,Structure and Properties of metal-matrixcomposite, J. Mater. Sci. Lett. 13, pp. 541-544,1994.[10] K. Seah, S. Sharma, P. R. Rao, B. Girish,Mechanical Properties of As-Cast and HeatTreated ZA27/Silicon Carbide ParticulateComposites, Materials & Design, 16, pp. 277-281, 1995.[11] S. C. Sharma, S. Sastry, M. Krishna, Effect ofaging parameters on the micro structure andproperties of ZA27/aluminite metal matrixcomposites, Journal of Alloys and Compounds346, pp. 292-301, 2002.[12] LI ZI-qani, ZHOU Heng-zhi, LUO Xin-yi,WANG Tao, SHEN Kai, Aging microstructuralcharacteristics of ZA-27 alloy and SiC p /ZA-27composite, Transactions of Nomnferous MetalsSociety of China, 16, pp. 98-104, 2006.[13] Z. Acimovic-Pavlovic, K. T. Raic,, I. Bobicand B. Bobic, Synthesis of ZrO2 ParticlesReinforced ZA25 Alloy Composites byCompocasting Process, Advanced CompositeMaterials 20, pp. 375–384, 2011.[14] I. Dutta, D.I. Bourell, A Theoretical andExperimental Study of Aluminium Alloy 6061-SiC Metal Matrix Composite to Identify theOperative the Operative Mechanism forAccelerated Aging, Materials Sience andEngineering A11, pp. 267-77, 1989.[15] B. Bobic, M. Babic, S. Mitrovic, N. Ilic, I.Bobic, M.T. Jovanovic, Microstructure andmechanical properties of Zn25Al3Cu basedcomposites with large Al 2 O 3 particles at roomand elevated temperatures, Int. J. Mat. Res.101pp. 1524–1531, 2010.[16] T. Savaskan, S. Murphy, Decomposition of Zn-Al Alloys on qench-aging, Materials Scienceand Technology, August 6, pp. 695-703, 1990.13 th International Conference on Tribology – Serbiatrib’13 413

Serbian TribologySocietySERBIATRIB ‘1313 th International Conference onTribologyKragujevac, Serbia, 15 – 17 May 2013Faculty of Engineeringin KragujevacUTICAJ POVRŠINE PODLOGE NAKARAKTERISTIKE PREVLAKA CINKADesimir Jovanović 1 , Bogdan Nedić 2 , Milomir Čupović 3 , Vlatko Matrušić 41 Zastava oružje AD, Kragujevac, Srbija, zo.tehnologija@open.telekom.rs2 Univerzitet u Kragujevcu, Fakultet inženjerskih nauka, Kragujevac, Srbija, nedic@kg.ac.rs3 Državni univerzitet u Novom Pazaru, Srbija, mcupovic@np.ac.rs4 Strojarski fakultet, Slavonski Brod, Hrvatska, vmarusic@sfsb.hrAbstract: Galvanske prevlake cinka se nanose da bi površina osnovnog materijala dobila odgovarajućasvojstava, kao što su: otpornost prema koroziji, hemijska postojanost, potreban estetski utisak i dr.Ispitivanja galvanskih prevlaka cinka usmerena su najčešće vezu Zn sa osnovnim materijalom, dok je veomamalo podataka o uticaju podloge na karakteristike prevlaka. Prethodna završna obrada ima veliki uticaj naformiranje fizičko-mehaničkih svojstava i strukture prevlake. U radu su prikazani rezultati istraživanjakarakteristika prevlaka cinka istaloženih na podlozi dobijenoj različitom završnom obradom sa različitomtvrdoćom i topografijom.Keywords: galvanske prevlake cinka, tvrdoća, topografija1. UVODUticaj vrste postupka obrade i uslova prethodneobrade kao i pripreme površina na koje se nanoseprevlake, odnosno tehnološkog nasleđa, je veomamalo istraživan. Površinski slojevi obrađenihpovršina dobijenih različitim postupcima obrade irežimima mogu imati različitu strukturu, što se teku period eksploatacije može ispoljiti. Prema tome,može se reći da se karakteristike površinskihslojeva formiraju kao rezultat različitih uslovaobrade u tehnološkom lancu izrade gotovog dela.Osnovni parametri koji se nasleđuju kroztehnološki proces izrade mogu se podeliti na dvegrupe. S jedne strane to su parametri vezani zasvojstva materijala: njegov sastav, strukturu,naponsko stanje i dr., dok su sa druge straneparametri vezani za makro i mikrogeometrijupovršina (geometrijski parametri) [1,2]. To ukazujena kompleksnost problema i potrebu izučavanja.Prevlake cinka se najviše koriste za zaštitučeličnih površina od korozije. Elektrohemijskeprevlake cinka mogu imati različite morfologije iteksture. Kada su ove pojave u pitanju većinaistraživanja se odnosi na uticaj standardnihparametra pri taloženju prevlaka kao što sugustina struje, temperatura, sastav kupatila [6-10],dok je znatno manje pažnja posvećeno značajupripreme površine čelične podloge. Istraživanjauticaja stanja podloge za taloženje prevlaka cinkapolirane mehanički [11.12] ili elektrohemijski [10-14]. pokazuju da se morfologija i tekstura prevlakena ovim površinama znatno razlikuju.Završna obrada površina ima veliki uticaj naformiranje fizičko - mehaničkih svojstava istrukture površinskog sloja. U radu se vršiistraživanje uticaja prethodne obrade površina idebljine prevlake na mehaničke i hemijskekarakteristike prevlake cinka.2. EKSPERIMENTALNA ISPITIVANJAKao osnova za nanošenje prevlaka odabran ječelik Č5730 (prema GOST-u 30HN2FA 1).Odabrani čelik se koriste za izradu cevi streljačkogoružja. Hemijski sastav osnove je prikazan u tabeli1. Uzorci za ispitivanje su pločice dimenzija 15 x10 x 6.3 mm (prema ASTM G 77). Nakon izradeuzoraka glodanjem, izvršena je termička obradapoboljšanjem na različite tvrdoće. Završna obradauzoraka vršena je na više načina, brušenjem sa višerežima, poliranjem, peskarenjem. Na ovaj način su414 13 th International Conference on Tribology – Serbiatrib’13

dobijene različite karakteristike površinskog sloja irazličite topografije površina uzoraka.Mikrogeometrija podloge za nanošenje prevlakasnimana je na kompjuterizovanom mernomuređaju Talysurf-6, koji omogućava kompleksnopraćenje kontaktnih površina. Korišćenjem ovogmernog sistema dobijena je informacija o početnojmikrogeometriji kontaktnih površina uzoraka.Tabela 1. Hemijski sastav osnove Č5730element hemijski sastav %1 C 0,27 -0,342 Mn 0,30 - 0,603 Si 0,17 -0,374 Ni 2,0 -2.45 Cr 0,60 - 0,906 Mo 0,20 - 0,307 V 0,10 -0,188 S max 0,0259 P max 0,02510 Cu max 0,25Nanošenje galvanskih prevlaka je vršeno upogonu za galvanizaciju fabrike "Zastava oružje" uKragujevcu. U tabeli 2, date su karakteristikeosnovnog materijala na koji je nanošena prevlaka.Tabela 2. Karakteristike podloge uzorakaUzorakbroj35VrstaobradeRamTvrdoćapodloge,HRC0.818 3845 0.719 39brušenje118 0.844 1910 0.720 37120.600 35šmirglanje33 0.550 3832 peskarenje 0.870 35Prevlake cinka su taložene u programiranomrežimu rada jednosmernom strujom, po zadatomplanu eksperimenta. U toku procesa taloženja,parametri jednosmerne struje su kontrolisani iregulisani u zadatim granicama. Korišćene anodesu napravljene od olova sa 10 % kalaja.Nanošenje prevlaka cinka je vršeno na sledećinačin:‣ alkalno bezcijanidno odmašćivanje saindustrijskim deterxentom,‣ ispiranje u protočnoj vodi,‣ dekapiranje u razređenoj hlorovodoničnojkiselini u odnosu 1:1,‣ ispiranje vodom,‣ elektro-hemijsko nanošenje prevlakecinka, sobna temperatura nanošenja prevlake jačina struje I =3 A/dm2 prosvetljavanje u 2% rastvoru HNO 3 utrajanju od 50 sekundi,‣ ispiranje u protočnoj vodi,‣ sušenje toplim vazduhom.Nanošenje prevlaka cinka je vršena tako što suuzorci postavljani u vertikalni položaj, usmereni naisti način. Tačkom "A" na uzorku (slika 1) jeoznačena gornja strana uzorka. Merenje lokalnedebljine prevlake cinka vršeno je na 15 tačakaprema šemi površine uzorka prikazanoj na slici 1.1,558,510ASlika 1. Šema mesta merenja debljine prevlakeKarakteristike (srednja vrednost debljine ihrapavost) istaloženih prevlaka date su u tabeli 3.Na slici 2, data je topografija prevlaka za uzorke10, 12, i 32.Tabela 3. Karakteristike prevlakaUzorakbrojRaprevlakem1,54,57,510,513,515Debljinaprevlake µm1 35 1.070 8.852 45 1.120 28.903 118 1.130 23.874 10 1.640 20.165 12 2.360 36.576 33 2.61 32.267 32 1.180 10.5713 th International Conference on Tribology – Serbiatrib’13 415

10Visina neravnina , µm50-5-100.00 0.25 0.50 0.75 1.00 1.25Duzina profila , mma) uzorak 10Visina neravnina , µm1050-5-100.00 0.25 0.50 0.75 1.00 1.25Duzina profila , mm10b) uzorak 12pre nanošenja prevlakeVisina neravnina , µm50-5-100.00 0.25 0.50 0.75 1.00 1.25Duzina profila , mmc) uzorak 32Slika 2. Topografija nakon nanošenja prevlaka zauzorke 10, 12 i 32Na slici 3. prikazan je izgled uzoraka pre i poslenanošenja prevlaka.posle nanošenja prevlakeb) uzorak 12pre nanošenja prevlakepre nanošenja prevlakeposle nanošenja prevlakea) uzorak 10posle nanošenja prevlakec) uzorak 32Slika 3. Izgled uzoraka pre i posle nanošenja prevlake3. ANALIZA REZULTATANa svim uzorcima prvo je izvršen pregledspoljašnjeg izgleda. Izgled prevlake praćen je416 13 th International Conference on Tribology – Serbiatrib’13

vizuelno, na dnevnoj svetlosti, pod uglom od45º. Površina prevlake kod svih uzoraka je sjajna iglatka. Nema dendrita, pregorelih i nepokrivenihmesta.Debljina prevlake cinka je merena na 15 mestaprema datom planu na slici 1. Na graficima, naslici 4, na osnovu rezultata merenja, prikazana jeraspodela debljine prevlake po površini uzorka.Na osnovu ovih grafika se može zaključiti dadebljina po širini i dužini uzorka odstupa ineravnomerna je. Najveće vrednosti za debljinuprevlake su izmerene na središnjem delu uzorka nadužini 13,5 mm od početka. Početak uzorka jeoznačen tačkom "A". To je tačka koja označavagornju stranu uzorka pri nanošenju cinka.Debljina prevlake, µm30252015105Širinauzorka,mm1.558.5Ispitivani uzorci su zadovoljili zahteve standarda.Prianjanje nanetih prevlaka cinka je dobro, nisuuočene promene na prevlaci cinka koje bi ukazalena odvajanje prevlake od osnovnog materijala,podloge.Koroziona stabilnost prevlaka cinka određivanaje praćenjem uzoraka tokom dužeg vremenaizlaganja dejstvu rastvora 3% Na Cl, u skladu saASTM B117-64 metodom. Ispitivani su uzorcirazličitih karakteristika (hrapavost i tvrdoćapodloge, debljina prevlake), tabela 2. Rezultatipraćenja korozione stabilnosti prevlaka cinka supokazali da nije došlo do pojave korozije, tako dase ne može uspostaviti veza između parametaraprethodne obrade i korozione otpornosti.Tribološkim ispitivanjima na tribometru blokon-diskmerena širina traga habanja na bloku i nataj način određivana otpornost na habanje kaoparametar habanja površini sa prevlakom cinka naispitivanim blokovima (slika 5).01.5 4.5 7.5 10.5 13.5Dužina uzorka, mma) uzorak 45Debljina prevlake, µm252015105Širinauzorka,mm1.558.501.5 4.5 7.5 10.5 13.5Dužina uzorka, mmb) uzorak 118Slika 4. Raspored debljine prevlakeIspitivanje metodom termičkog šoka vršeno jeprema standardu ISO 2819-1980. Uslovi ispitivanja:‣ Temperatura zagrevanje uzoraka T=200 0 C(prema standardu T =180 0 - 200 0 C),‣ Vreme zagrevanja 1 sat.‣ Kvašenje mlazom hladne vode.Posle zagrevanja prema uslovima datimstandardom uzorci se izlažu mlazu hladne vode.Prevlaka mora da ostane nepromenjena, ne sme dadođe do pojave odslojavanja prevlake sa osnovnogmaterijala, podloge.Slika 5. Trag habanja na blokuIspitivanje habanja prevlaka su vršena natribometru TR-95 sa kontaktom blok-on-disk uCentru za obradu metala rezanjem i tribologijuFakulteta inženjerskih nauka u Kragujevcu.Tribometar TR-95 omogućava variranje uslovakontakta sa aspekta oblika, dimenzija i materijala13 th International Conference on Tribology – Serbiatrib’13 417

kontaktnih elemenata, normalnog kontaktnogopterećenja i brzine klizanja. Ispitivanja se moguvršiti u uslovima sa podmazivanjem i bezpodmazivanja. Razvoj procesa habanja na blokumanifestuje se formiranjem i širenjem izraženogtraga habanja. Normalno opterećenje je bilo 10 N,a brzine klizanja 0,25 m/s, 0,5 m/s i 1 m/s. Ukupanput klizanja je bio 150 m. Realizovana ispitivanjasu bila sa graničnim podmazivanjem sa mineralnimhidrauličnim uljem Hidrovisk HD46 zbog velikogkoeficijenta trenja u slučaju trenja bezpodmazivanja i velikih vibracija u sistemu merenjasile trenja kod tribometra TR-95.Početni nominalni linijski kontakt između diskai bloka usled razvoja procesa habanja postajekontakt po određenoj površini, što kao posledicuima razaranje materijala, najpre u površinskomsloju bloka (slika 6). Promena širine traga habanjaima isti karakter za sve ispitivane uzorke samo jerazlika u nivou njihove pohabanosti.Slika 6. Širina traga habanja na blokuProces habanja karakteriše postizanjeodređenog nivoa, stabilizaciju i usporeni porastširine traga habanja tokom vremena ispitivanja. Naosnovu rezultata merenja širine traga habanja nabloku, formirani su u zavisnosti od uslova kontakt(brzina klizanja, normalna sila) pojedinačni i zbirnihistogram promene širine traga habanja (slika 7).Najveće širine traga habanja odgovaraju najmanjojbrzini klizanja.Ispitivani su uzorci različitih karakteristika(hrapavost i tvrdoća podloge, debljina prevlake).Ako se posmatra histogram, slika 7, može sezaključiti da se između hrapavosti površinauzoraka pre i posle nanošenja prevlake cinka iširine traga habanja na bloku ne može uspostavitiveza. Takođe, ne može se uspostaviti veza izmeđutehnologija obrade uzoraka (brušenje, peskarenje idr.) sa širinom traga habanja.Najmanje habanje je bilo kod uzoraka 10 i 33,odnosno, kod uzoraka sa velikom tvrdoćom podloge ivelikom debljinom prevlake.Najveće habanje je bilo kod uzoraka 32, 35 i118. Uzorak 118 ima najmanju tvrdoću a uzorci 32i 35 najmanju debljinu prevlake cinka.Širina traga habanja, mm1.51.20.90.60.30uzorak10Brzina klizanja:uzorak12uzorak33uzorak35uzorak32uzorak45uzorak118v = 0,25 m/s v = 0,5 m/s v = 1 m/sSlika 7. Širina traga habanja na bloku4. ZAKLJUČAKIspitivane prevlake različitih debljina su nanetena uzorke sa različitom topografijom i tvrdoćom.Rezultati ispitivanja vizuelnim pregledom, ispitivanjakorozione otpornosti i ispitivanja prianjanjeprevlaka cinka za osnovni metal, podlogu, supokazali da prevlake zadovoljavaju zahteve standarda.To znači da je valjano izabrana tehnologijapriprema površina i nanošenja prevlaka.Rezultati ispitivanja uticaja promene topografijepovršine na kvalitet prevlake su pokazali dananošenjem prevlake dolazi do značajne promenetopografije, odnosno povećanja hrapavosti ali da tone utiče na ostale karakteristike prevlake cinka.Realizovana ispitivanja i dobijenih rezultataukazuju na postojanje zavisnosti između tvrdoćepodloge, debljine prevlake i širine traga habanja pritribološkim ispitivanjima, ali uspostavljanjekorelatiuvnih veza je moguće realizacijom znatnovećeg broja eksperimenata.AKNOWLEDGEMENTOvaj rad je deo projekta TR35034 "Istraživanjeprimene savremenih nekonvencionalnihtehnologija u proizvodnim preduzećima sa ciljempovećanja efikasnosti korišćenja, kvalitetaproizvoda, smanjenja troškova i uštede energije imaterijala", koga finansira Ministarstvo prosvete inauke Republike Srbije.418 13 th International Conference on Tribology – Serbiatrib’13

LITERATURA[1] Зинченко В. М.: Технологическая наследственностьпри изготовлении деталей,Технология металлов, № 5, 2007.[2] П. И. Ящерицын : Технологическое наследованиеэксплуатационных параметров деталей машин.Справочник, Инженерный журнал № 9, 2004,[3] S. Đorđević, M. Maksimović, M. G. Pavlović, K. I.Popov, Galvanotehnika, Tehnička knjiga, Beograd,1998.[4] Volker Kunz, ZINTEK–TECHSEAL-TRI-COATCatalouge, Trebur 2001.[5] Nedić, B., Jovanović, D., Lakić Globočki, G.,Influence Of Previous Machining On CharacteristicsOf Galvanic Coatings, Serbiantrib - 12 thInternational Conference on Tribology, Kragujevac,2011.[6] B. Nedić, D. Jovanović, G. Lakić-Globočki:Scratch Testing Of Zn Coating Surfaces,Proceedings: 34th International conference onproduction engineering, Niš, 2011.[7] Raeissi, Saatchi, Golozar, Szpunar: Effect ofsurface preparation on zinc electrodepositedtexture, Surface & Coating Tehnology 197, 2005,[8] K. Deblauwe, A. Deboeck, J. Bollen, W.Timmermans, Proceeding of the 12th InternationalConference on Textures of Materials (ICOTOM-12), Montreal, Quebec, Canada, 9–13 August,1999,.[9] S. H. Hong, J. B. Kim, S. K. Lee, Mat. Sci. Forum408–412, 2002.[10] Kim, S. Ch. Hong, Proceeding of the 12thInternational Conference on Textures of Materials(ICOTOM 12), Montreal, Quebec, Canada, 9–13August, 1999,[11] D. Vasilakopoulos, M. Bouroushian,N. Spyrellis, Trans. IMF 79, 2001.[12] M. Ye, J.L. Delplancke, G. Berton, L. Segers, R.Winand, Surf. Coat. Technol. 105, 1998.[13] H. Yan, J. Downes, P.J. Boden, S.J. Harris, J.Electrochem. Soc. 143, 1996.[14] H. Park, J.A. Szpunar, Proceeding of the 12thInternational Conference on Textures ofMaterials (ICOTOM 12), Montreal, Quebec, LCanada, 9–13 August, p. 1421, 1999.[15] Standard ISO 2819-1980[16] Standard ASTM B117[17] StandardISO 453913 th International Conference on Tribology – Serbiatrib’13 419

Serbian TribologySocietySERBIATRIB ‘1313 th International Conference onTribologyKragujevac, Serbia, 15 – 17 May 2013Faculty of Engineeringin KragujevacDEFEKTACIJA REDUKTORA BKSH-335 ZA POKRETANJETRAKASTIH TRANSPORTERA BAGERA Sch Rs 630Svetislav Lj. Marković 1 , Ljubica Milović 2 , Bratislav Stojiljković 31 Visoka škola tehničkih strukovnih studija, Čačak, svetom@open.telekom.rs,2 Univerzitet u Beogradu, Tehnološko-metalurški fakultet, Beograd, acibulj@tmf.bg.ac.rs3 Muzej Nikole Tesle, Beograd, bratislav.stojiljkovic@tesla-museum.orgApstrakt: Remont jednog moćnog mašinskog sistema kao što je bager-glodar Sch Rs 630 je uslovljen visokimnivoom pouzdanosti, ispravnosti i funkcionalnosti svih njegovih sklopova. Iz navedenih razloga remontrazmatranog reduktora mora biti praćen detaljnom defektacijom, odgovarajućom tehnologijom regeneracije,ugradnjom kvalitetnih delova (zupčanika, vratila, ležajeva) i pravilnom montažom kako elemenata ureduktor, tako i montažom reduktora na sam bager. Svi navedeni faktori od presudnog su uticaja na budućirad reduktora. Takođe, neophodno je voditi računa o ekonomskoj opravdanosti izvođenja ovakvih radova. Uvelikim sistemima svaki neplanirani zastoj ovakvih mašina veoma mnogo košta. Nedopustivo je da uzrokzastoja bagera bude kvar jednog ovakvog reduktora.Ključne reči: bager, reduktor, defektacija, remont.1. UVODRotorni bageri su mašine velikih gabarita,kapaciteta i težina. Jedan od takvih je i rotornibager Sch Rs 630, nemačkog proizvođača KRUPP.Bager glodar je moćna i visoko produktivnamašina, koja se koristi na površinskim kopovima zaotkopavanje ugljenih i jalovinskih masa. Spada ugrupu rotornih bagera i ima teorijski kapacitet 4100m 3 /h, visina kopanja je 25 m, a dubina 6 m.Rotorni bager se sastoji od donje i gornjegradnje sa konstrukcijom. U donju gradnju spadatransportni mehanizam (šest transportnih gusenica)od kojih su dve sa hidrauličnim upravljačkimmehanizmom. Iznad transportnog mehanizma jeaksijalni kuglični ležaj koji omogućuje zaokretanjegornje gradnje za 360º. Radno područje zaokretanjagornje gradnje u odnosu na donju je ±105º, a brzinaokretanja gornje gradnje je 10 m/min. U donjugradnju spada još i kablovski bubanj za napajanjecelog bagera visokim naponom od 6000 V. Prečnikkabla je Ø78 mm, maksimalna rezerva kabla nabubnju je 1500 m. Pogonski naponi na spravi su500 V, komandni 220 V kao i naponi za osvetljenje,dok su naponi za ručne lampe i magnetne ventile 24V.Na donju gradnju, preko kugličnog aksijalnogležaja, oslonjena je konstrukcija odložne trake, a sagornje strane ovešena za vertikalni stub bagera.Odložna traka se zaokreće u krug nezavisno odgornje gradnje. Širina trake je 1600 mm, a brzinakojom se kreće je 4,2 m/s. Ispod odložne trakepostoji i posebna traka za otpadni materijal sasopstvenim pogonom i brzinom od 1 m/s.U sastav gornje gradnje glodara spada utovranatraka sa radnim točkom, koja je vezana za čeličnukonstrukciju i može se podizati i spuštati prekouređaja za dizanje (čeličnih užadi). Za čeličnukonstrukciju gornje gradnje vezana je i konstrukcijaprotivtega, kao i montažni kran koji služi zademontažu i montažu delova pri popravci.Radni točak je prečnika 10 m na kome jepostavljeno 20 kašika zapremine 0,63 m 3 . Brzinarezanja je 2,827 m/s. Broj istresa je 108-162 min -1 .Dozvoljeni nagibi za rad su 5%, a za transport 10%.Ukupna masa celog bagera je 1.499.641 kg, aspecifični pritisak na tlo je 0,1 MPa.Bagerom se upravlja iz dva komandna pulta: izkabine bageriste, koja je vezana za konstrukcijuprijemne trake, i iz kabine trakiste, koja se nalazi nakonstrukciji odložne trake.420 13 th International Conference on Tribology – Serbiatrib’13

Slika 1. Rotorni bager - glodar VII (Sch Rs 630)snimljen noćuReduktori pogona traka su zajedno saelektromotorima vezani na jedno postolje. Najednoj strani su zakačeni za rukavce pogonskogbubnja preko izlaznog zupčanika, pomoćuelastičnih steznih prstenova, a sa druge strane prekopokretnog zgloba za čeličnu konstrukciju trake. Satako izvedenom vezom osa pogonskog bubnja i osereduktora se uvek seku jedna u odnosu na drugupod 90º.Puštanje u pogon trake može se izvoditi na dvanačina, sa komandnog pulta, odnosno iz kabinebageriste kada se uključuje kompletni lanac zakretanje celog sistema. Drugi način je pogon na licumesta preko uređaja za deblokiranje (ovaj pogon sekoristi prilikom popravki i proba samo tog dela uzveliku opreznost i stručni nadzor).2. REDUKTOR BKSH–335Reduktor BKSH-335 je dvostepeni i koristi se zapogon prijemne i odložne trake na rotornimbagerima za eksploataciju uglja u rudarskombasenu Kolubara. Njegov zadatak je da snagu saelektromotora prenese na pogonski bubanj kojipokreće prijemnu, odnosno odložnu traku nabageru. Traka koju pokreće je beskonačna, širine1800 mm, debljine 30 mm i ukupne dužine u obasmera ≈80 m.Uslovi eksploatacije u kojima se reduktor koristisu prilično loši. Reduktor je pozicioniran na mestukoje je izloženo spoljašnjem uticaju vremenskih iatmosferskih neprilika, tako da radi u po kiši,snegu, prašini, mrazu, vrućini, oksidaciji... Svepomenute nepogode utiču direktno ili indirektno nastanje reduktora i njegove prateće opreme. Usistemu EPS-a kada je u pitanju eksploatacija ugljaizbegavaju se bilo kakvi neplanirani zastoji rotornihbagera. Zbog toga promene stanja reduktora, nijemoguće otkloniti do sledećeg planiranog zastojamašine. Usled toga su i česte pojave havarijeovakvih reduktora, jer otkaz jednog njegovogelementa povlači defekt drugog, pa dolazi dootkaza kompletne pogonske grupe.Za pogon se koristi sinhroni elektromotor snage160 kW, koji ima izlazni broj obrtaja 960 o/min.Prateći deo opreme elektromotora je i upuštač kojiima ulogu menjača. Pogonska grupa se startuje namanjem broju obrtaja, a zatim se posle kraćegvremenskog perioda brzina povećava, da bi tek nakraju dostigla punu vrednost od 960 o/min.Spojnica se sastoji iz kočionog točka koji jevezan za ulazno vratilo reduktora i prirubnicepričvršćene za vratilo elektromotora. Ovo jenajlošiji deo pogonske grupe. Naime, ove dvanezavisna dela međusobno su povezana zavrtnjimakoji jednim delom prolaze kroz gumu tvrdoće50÷60Sh, koja ima ulogu da amortizuje udareprilikom pokretanja pogonske grupe. Guma je čestološeg kvaliteta i vremenom dolazi do njenogdeformisanja što direktno utiče na saosnost vratilana kojima se nalaze.Slika 2. Šema reduktora BKSH-335(I – ulazno vratilo, II – međuvratilo, III – šuplje vratilo,1 – tanjirasti zupčanik, 2 – gonjeni cilindrični zupčanik,A – par konusno-valjkastih ležajeva 31319,B – cilindrično-valjkasti ležaj NU 2322,C – par konusno-valjkastih ležajeva 31320,D – dvoredi bačvasti ležaj 23044)Kućište reduktora je dvodelno i izrađeno je odČL0500. Ojačano je rebrima jer je dolazilo doprslina na bočnim stranama. Tokom godina radakućište je ojačano sa po još jednim rebrom okoglavčina i sada ih ima po tri.Slika 3. Kućište sa kočionim točkom13 th International Conference on Tribology – Serbiatrib’13 421

3. POSTUPAK REMONTA REDUKTORABKSH–335Slika 4. Kućište reduktoraUlazno vratilo (označeno sa I, slika 2) je odčelika za cementaciju Č 4520 i izrađeno je zajednosa pogonskim, konusnim zupčanikom, koji ima 13zubaca (z 1 =13), modula m e =8 mm. Vratilo je naulazu oslonjeno pomoću para konusno-valjkastihležajeva 31319 (A, slika 3), a na izlazu jecilindično-valjkasti ležaj NU 2322 (B, slika 2).Gonjeni konusni (tanjirasti) zupčanik (1, slika 2)izrađen je od čelika za cementaciju Č 4321 i spojensa glavčinom pomoću podešenih zavrtnjeva. Ima 37zubaca (z 2 =37). Glavčina je navučena nameđuvratilo (II, slika 2) i osigurana klinom, aizrađena je od čeličnog liva ČL0400.Međuvratilo (II, slika 2) je izrađeno od čelika zacementaciju Č 4321. U jednom delu je ozubljeno,broj zuba je z 3 =20, a modul m=6 mm. Oslonjeno jena jednom kraju preko para konusno-valjkastihležajeva 31320 (C, slika 2), a na drugom pomoćucilindrično-valjkastog ležaja NU 2322 (B, slika 2).Izlazni zupčanik (2, slika 2) je izrađen od čelikaza cementaciju Č 4321. Ima 88 zubaca (z 4 =88), amodul mu je m=6 mm. Navučen je na šupljevratilo.Šuplje vratilo (3, slika 2) je izrađeno od čelikaza poboljšanje Č 4732. Oslonjeno je bačvastimdvoredim ležajevima 23044 (D, slika 2). Krozšuplje vratilo je provučeno vratilo bubnja.Neposrednom vezom šuplje vratilo prenosi obrtnimoment na vratilo bubnja, čime ga pokreće,dovodeći traku u pogon, tako da pri najvećem brojuobrtaja reduktora, traka širine 1600 mm dostižebrzinu od 4 m/s.Podmazivanje reduktora je rešeno tako što sezupčanici, vratila i ležajevi podmazuju prirodnimputem. U reduktor se pre početka rada sipa 60 l uljatipa EP koje zadržava svoja svojstva i pritemperaturi od 70÷80°C. Ulje se sipa u za topredviđene komore do određenog nivoa i prilikomrada dolazi do prirodnog kruženja ulja bez dodatnihpumpi.Reduktor se demontira sa pogonske grupebagera-glodara i transportuje u radionicu na remont.Doprema se sa kočionim točkom, koji ga vezuje saspojnicom, a bez postolja. Remontne aktivnosti susledeće: obezbeđivanje adekvatnog radnog prostoraza demontažu, defektaciju i montažureduktora, rasklapanje i demontaža reduktora,defektacija delova i sklopova reduktora,regeneracija starih delova ili izrada novihdelova (koji su defektažom utvrđeni kaoneupotrebljivi u daljoj eksploataciji), kontrola izvršenih radova, kompletiranje reduktora, montaža podsklopova i sklopova, završna montaža i priprema reduktora zaprobni rad, probni rad reduktora i izrada izveštaja oispitivanju, transport reduktora na bager glodar.Radni prostor za reduktor je površine od oko180 m 2 . Ograđen je limenom ogradom visine 2 m.Snabdeven je svim potrebnim energetskimkapacitetima, kao što su: komprimovani vazduh, sistem za gasno zavarivanje, električna energija napona 380 V, iznad radnog prostora funkcioniše kran od40 t nosivosti, koji poseduje dve kuke,jednu nosivosti do 5 t i drugu nosivosti do40 t,regali postavljeni u dva nivoa duž zidovaprostorije, namenjeni za odlaganje sitnihdelova, dobro osnovno osvetljenje sa još četirireflektora postavljena po dijagonali, pokretni hidraulički agregat od 2,8 l,maksimalnog pritiska 600 bar,radni stolovi sa stegama.4. PRANJE I ODMAŠĆIVANJEReduktor se unutrašnjim transportom(viljuškarom) prenosi na pranje, gde se uklanjajusve spolja dostupne nečistoće. Pranje se izvoditoplom vodom pod pritiskom da bi se uklonilenaslage blata, uglja, masnoće od ulja...Posle pranja reduktor se doprema u radionicu,gde se obavljaju dalje aktivnosti utvrđenetehnološkim postupkom. Defektatori pristupajusvom osnovnom zadatku, pregledu reduktora,422 13 th International Conference on Tribology – Serbiatrib’13

stvaraju sliku stanja i svoja zapažanja unose udefektacioni list.Prvo se demontira spojnica pomoćuhidrauličnog agregata, protoka 6 l/min imaksimalnog pritiska 600 bar, i hidrauločnoguređaja koji stvara silu od 5·10 5 N. Iz reduktora senajpre istače staro ulje, zatim se kućište reduktorarazdvaja i skidaju se svi bočni poklopci. Vade seelementi (podsklopovi) iz kućišta: ulazno vratilo,međuvratilo i izlazni zupčanik sa šupljim vratilom.Kućište reduktora se zatim ponovo sastavlja,polutke se postavljaju jedna na drugu tako da buducentrirane i ponovo stežu bez bočnih poklopaca.Otvori za ležajeve u kućištu se odmašćuju,pripremaju za vizuelnu kontrolu stanja otvora imerenje. Podsklopovi se spremaju za defektaciju.Ozubljenja se pripremaju za vizuelnu imagnetnofluksnu kontrolu pranjem trihloretilenom.Ležajevi se odmašćuju i pripremaju za vizuelnukontrolu i merenja zazora. Vratila se odmašćuju, aotvori u šupljem vratilu se pripremaju za defektažu.Kao sredstva za pranje i odmašćivanje koristese: trihloretilen, famin, za ozubljenjenitrorazređivač, naročito prilikom ispitivanjamagnetnim fluksom i ultrazvukom. prečnici i pohabanost rukavaca vratila, pojava prslina na delovima reduktora(kontrolom magnetnim fluksom iliultrazvukom), pohabanost kočionog točka.Na kočionom točku se vrši merenje prečnikaotvora u spojnici Ø75H7. Vizuelnom kontrolomradne površine kočionog točka je ustanovljendefektni sloj (risevi), koji su nastali verovatnopohabanošću obloga na paknovima (slika 6). Točakse šalje na strugarsku obradu i balansiranje.Slika 6. Prikaz oštećenog površinskog sloja kočionogtočkaNa kućištu reduktora vizuelnom kontrolomproveravaju se varovi oko glavčina i svi podužnivarovi, kao i promene (lom ili prsline) osnovnogmaterijala, zatim stanje revizionih poklopaca, kao iotvora za ležajeve (Ø275H7 – na mestu ulaznogvratila; Ø260H7 – na mestu međuvratila; Ø340H7 –na mestu izlaznog vratila), koji su ispitivani imerenjem u tri pravca (slika 7).Slika 5. Šuplje vratilo sa izlaznim zupčanikom iležajevima5. DEFEKTACIJAPri defektaciji kontrolišu se: otvori za ležajeve na kućištu, ukupno stanje kućišta, ležajevi i njihovi radijalni zazori, ozubljenja, elementi za vezu na zupčanicima (otvori,žljebovi za klinove), radijalna bacanja vratila,Slika 7. Prikaz merenja otvora za ležajeve na kućištuPrilikom vizuelne kontrole ulaznog vratilaustanovljena su oštećenja na rukavcu na koji nailaziležaj koji se grejao u toku rada. Skinut jecementirani sloj, tragovi su vidljivi (slika 8).13 th International Conference on Tribology – Serbiatrib’13 423

Izvršeno je magnetnofluksno ispitivanje zubakonusnog zupčanika, rukavaca i prelaznih radijusa inisu uočene nikakve indikacije tipa spoljašnjihprslina. Istovremeno je izvršena kontrola ostalihrukavaca na ulaznom vratilu i ustanovljeno da sumere u granicama preporučenih tolerancija (Ø75k6,Ø80h8, Ø95j6 i Ø110k6). Na kraju je vršenaprovera radijalnog bacanja vratila. Postavljeno je nastrug, pročišćena su mu središna gnezda ikomparaterom su očitane mere bacanja vratila.Vuzuelnom kontrolom međuvratila ustanovljenoda nema oštećenja zubaca na ozubljenom deluvratila, ali su vidljiva oštećenja rukavca na koji senavlači par konusno-valjkastih ležajeva (slika 10).Merenjem je ustanovljeno da su prečnici drugihrukavaca u granicama tolerancija.Slika 8. Prikaz oštećenja rukavca ulaznog vratilaVizuelnom kontrolom tanjirastog zupčanikauočena je labavost veze između tanjirastogzupčanika i glavčine za koju je pričvršćen. Tokomrada došlo je do oštećenja otvora Ø18 mm. Otvorimoraju biti razbušeni na prvu veću meru da bi seponovo ostvarila veza podešenim zavrtnjevima.Kontrolisani su i otvor u zupčaniku i žljeb za klin,čije mere se nalaze u granicama preporučenihtolerancija.Slika 10. Međuvratilo sa oštećenim rukavcemVizuelnom kontrolom gonjenog cilindričnogzupčanika ustanovljeno je da nedostaje skoropolovina jednog zupca (slika 11). Zbog ovakvogoštećenja neophodna je regeneracija zupčanika.Slika 9. Međuvratilo sa tanjirastim zupčanikom iležajevimaSlika 11. Izlazni zupčanikI šuplje vratilo je dijagnostikovano vizuelnimmetodom i tehničkim merenjima. Na njemu nemavidljivih oštećena, niti tragova udara i naprsnuća.Merenjem je ustanovljeno da se sve mere nalaze ugranicama tolerancija.Defektacija ležajeva podrazumeva vizuelnukontrolu, merenje radijalnih zazora i uparivanjepaketa ležajeva. Vizuelnom kontrolom jeustanovljeno da nema oštećenja ležajeva niti424 13 th International Conference on Tribology – Serbiatrib’13

njihovih kotrljanih elemenata. Merenjem zazora naležajevima utvrđeno je: paket ležajeva na ulaznom vatilu (31319) jevan preporučenih tolerancija, pa jepotrebna zamena, ležaj na ulaznom vatilu NU 2322 imanedopušteni zazor i on se mora zameniti, ostali ležajevi su u granicama preporučenihtolerancija od strane proizvođača.prijanjanja paknova kočnice i izvršidinamičko uravnoteženje(balansiranje), na ulaznom vratilu izvrši zamena svihležajeva sa obaveznim uparivanjem paraležajeva 31319, pohabani rukavci ulaznog vratila imeđuvratila regenerišu se metalizacijom sabrušenjem na nominalne mere, gonjeni cilindrični zupčanik regenerišenavarivanjem polomljenog zupca.6. ZAKLJUČAKReduktor ulazi u redovan remont jednomgodišnje, za vreme planirane investicione popravkeili remonta, ili ukoliko se desi neki neplanirani kvarili havarija. Najčešći razlozi remonta reduktora su:zagrevanje ležajeva, curenje ulja iz reduktora,povećana bučnost reduktora, pohabanost rukavacavratila, lom zubaca zupčanika.Slika 12. Merenje radijalnih zazora ležajaNa osnovu rezultata merenja i ispitivanjadoneta je odluka da se: kočioni točak na krutoj spojnici obradistruganjem na prvu veću meru radi boljegLITERATURA[1] J. Jevtić: Reduktori – proračuni i konstrukcije,Privredni pregled, Beograd, 1976.[2] S. Marković, D. Josifović: Regeneracija zupčanika,monografija, Jugoslovensko tribološko društvo,Kragujevac, 1998.[3] S. Marković: Održavanje mašina i opreme, Višatehnička škola, Čačak, 2005.[4] M. Trbojević, M. Janković, J. Vugdelija, N. Plavšić,V. Latinović: Reduktori, Naučna knjiga, Beograd,1991.[5] www.rbkolubara.co.yu.DEFECTATION OF THE REDUCER BKSH-335 FOR THE ACTIVATIONOF THE BAND TRANSPORTER OF THE DREDGER Sch Rs 630Abstract: The repair of a powerful mechanical system such as dredger Sch Rs 630 is based on the high levelof reliability, accuracy and functioning of all its circuits. Due to the reasons listed above the rapair of thementioned reducer must be done with thorough defectation, suitable regeneration technology, embedding ofquality parts (gears, shafts, bearings) and proper installation of the parts into the reducer, and installationof the reducer onto the dredger itself. All the mentioned factors are essential and they influence the properoperation of the reducer. Also, it is necessary to pay attention to the economical adequacy of this type ofwork. In big systems every non-planned work failure brings enormous costs. Neither it is recommended noracceptable that the failure in the dredger work has its cause in a malfunction of the similar reducer.Keywords: dredger, reducer, defectation, repair.ACKNOWLEDGMENTThis work has been performed within the project TR 35011. This project is supported by the Republic ofSerbia, Ministry of Science and Technological Development, which financial help is gratefullyacknowledged.13 th International Conference on Tribology – Serbiatrib’13 425

Serbian TribologySocietySERBIATRIB ‘1313 th International Conference onTribologyKragujevac, Serbia, 15 – 17 May 2013Faculty of Engineeringin KragujevacISPITIVANJE MEHANIČKIH I STRUKTURNIH OSOBINAPREVLAKA OTPORNIH NA EROZIJU I VISOKETEMPERATUREMihailo R. Mrdak 11 Istraživački razvojni centar, IMTEL Komunikacije a.d., Beograd, Srbija, miki@insimtel.comApstrakt: Cilj ovog rada bio je da se ispitaju mehaničke i strukturne karakteristike termo-barijernihprevlaka TBC otpornih na eroziju i visoke temperature. Deponovana su tri tipa TBC sistema dvojnihprevlaka sastavljenih od metalne vezne prevlake NiCrAlCoY 2 O 3 i keramičkih izolacionih prevlaka ZrO 2 MgO,ZrO 2 Y 2 O 3 i ZrO 2 CeO 2 Y 2 O 3 . TBC sistemi prevlaka su deponovani atmosferskim plazma sprej postupkom(APS) na ohrapavljenim čeličnim substratima sa temperaturom od 160 – 180 °C. Prevlake su deponovane saoptimalnim parametrima depozicije prahova. Vezni slojevi su deponovani sa jednim prolazom plazmapištolja, a keramički slojevi sa četrnest prolaza. Procena kvaliteta slojeva je urađena ispitivanjemmikrotvrdoće metodom HV 0.1 i zatezne čvrstoće spoja ispitivanjem na zatezanje. Metalografska procenaudela mikro pora (image analiza) u strukturi veznih i keramičkih slojeva je urađena tehnikom svetlosnemikroskopije. Morfologije čestica praha je urađena na SEM-u. Analize izvršenih ispitivanja su omogućile dase odaberu TBC sistemi prevlaka sa najboljim mehaničkim i strukturnim karakteristikama.Ključne reči: atmosferski plazma sprej (APS), termo barijerne prevlake, mikrostruktura, interfejs,mikrotvrdoća, čvrstoća spoja.1. UVODTermo barijerne prevlake (TBC) su grupeprevlaka namenjene za posebne uslove rada zbogsvojih specifičnih karakteristika. Prevlake su našleširoku primenu kao zaštita delovima izloženimvisokim temperaturama, visokotemperaturnojabraziji čestica, oksidaciji i toploj koroziji. Osnovnauloga TBC je da obezbedi mogućnost radaosnovnog materijala na temperaturama koje suiznad granice njegove izdržljivosti spuštanjemstvarne temperature na donjoj površini veznog slojaTBC sistema za ΔT = 200 – 400 °C [1,2]. Keramičkisloj mora da zadrži nisku toplotnu provodljivosttokom dužeg izlaganja u uslovima eksploatacije i daima dobru otpornost na abraziju čestica, koja možeda javi u različitim vidovima [3]. Keramičkaprevlaka se u poređenju termičkih koeficijenatarazlikuje od komponente na koju se primenjuje.Ovaj sloj treba da bude usklađen sa osnovnimmaterijalom preko veznog sloja. Donji vezni slojkoji je tanji u odnosu na keramički sloj mora da imadobru vezu sa substratom i keramikom, sprečidifuziju na interfejsu, poveća otpornost na koroziju ismanji uticaj zaostalih napona u keramici [4-7].Kompozitni prah NiCr-Al-Co-Y 2 O 3 koji sekoristi za proizvodnju veznog sloja je obloženoksidom Y 2 O 3 koji je stabilizator u keramičkimprevlakama. Oksid Y 2 O 3 koji je istovremenoprisutan u metalnom i keramičkom prahupoboljšava vezu između metalnih i keramičkihslojeva [8]. Za proizvodnju keramičkih slojevakoriste se prahovi kao što su Metco 210NS –ZrO 2 MgO, Metco 202NS – ZrO 2 Y 2 O 3 i Metco205NS – ZrO 2 CeO 2 Y 2 O 3 i dr. [9-11]. Atmosferskiplazma sprej postupak (APS) je standardni proces zadeponovanje termalnih barijera. Čestice praha seubrizgavaju u protok plazma gasa i ubrzavaju usledprenosa brzine i temperature jona na čestice praha.Pod uticajem substrata, čestice se plastičnodeformišu i vezuju za substrat da bi se formiralaprevlaka. Proces omogućuje deponovanje širokogspektra prahova: legura, keramike, karbida, metalokarbida,metalo-keramike, kompozitnih prahova idr. Na kvalitet deponovanih prevlaka utiče velikibroj parametara. Opis atmosferskog plazma sprej426 13 th International Conference on Tribology – Serbiatrib’13

postupka i uticaja parametara depozicije praha nakvalitet prevlaka dat je detaljnije u radovima [12-15]. Jedan od veoma bitnih parametara koji utiče nakvalitet i trajnost TBC u eksploataciji je temperaturasubstrata na kojoj se deponuju slojevi. Premaranijim istraživanjima, plazma deponovanikeramički depoziti pokazuju lamelarnu strukturu saograničenom intelamelalnim vezivanjem [16]. Zbogtoga su u depozitu prisutne mikro pore kaozapreminske greške sa velikim koncentracijamanapona koji uzrokuju pojavu mikro pukotina.Ograničeno vezivanje lamela u depozitu smanjujevrednosti tvrdoće, modul elastičnosti, žilavost lomai toplotnu provodljivost odgovarajućeg materijala[17-19].Predgrevanje substrata se koristi za povećanjeadhezije prevlake, poboljšanje mikrostrukture imehaničkih svojstava [2, 20-22]. Svojstava TBCdepozita su uglavnom pod kontrolom morfologijedeponovanih čestica i interakcija/prianjanje međunjima. Jedan od uzroka propadanja keramičkihprevlaka je termičko opterećenje. Nagle promenetemperature mogu uzrokovati pucanje i odvajanjecelog TBC sistema sa substrata zbog značajnerazlike koeficijenata linearnog širenja keramike imetala. Drugi način propadanja keramičkih prevlakanastaje od erozije čestica na visokimtemperaturama. Ispitivanja otpornosti keramičkihprevlaka na eroziju čestica Al 2 O 3 veličine od 15 μmdo 53 μm, za isti nivo poroznosti, na 980 °C supokazala da najveću otpornost imaju keramičkeprevlake ZrO 2 CeO 2 Y 2 O 3 sa zapreminskim gubitkom1,72 × 10 4 cm 3 /gm. Za prevlaku ZrO 2 20Y 2 O 3zapreminski gubitak je nešto viši 1,95 × 10 4cm 3 /gm, dok je za prevlaku ZrO 2 MgO zapreminskigubitak bio najviši 3,03 × 10 4 cm 3 /gm [11].U ovom radu su ispitana tri tipa TBC sistemaprevlaka NiCrAlCoY 2 O 3 / ZrO 2 MgO,NiCrAlCoY 2 O 3 / ZrO 2 Y 2 O 3 i NiCrAlCoY 2 O 3 /ZrO 2 CeO 2 Y 2 O 3 . Prahovi su deponovani napredgrejanim substratima sa temperaturom od 160 –180 °C sa ciljem poboljšanja interlamelarne veze,mikrostrukture i mehaničkih svojstava. Izvršena sumetalografska ispitivanja sadržaja mikro pora ioksida u veznim slojevima i mikro pora ukeramičkim prevlakama. Za svaki sistem TBCprevlaka su ispitane mikrotvrdoće i zatezne čvrstoćespoja. Analize rezultata izvršenih ispitivanja suomogućile da se ustanovi koji sistem prevlaka imanajbolje strukturno mehaničke karakteristike.2. EKSPERIMENTALNI DETALJI2.1 MaterijaliMaterijal na kome su deponovane termalnebarijere je bio od nerđajućeg čelika X15Cr13 (EN1.4024) u termički neobrađenom stanju. Za izradutermalnih barijera su upotrebljena četiri praha firmeSulcer Metko (Sulzer Metco) sa oznakama: Metco461 (NiCr/Al/Co/Y 2 O 3 ), Metco 210NS (ZrO 2 MgO),Metco 202NS (ZrO 2 Y 2 O 3 ) i Metco 205NS(ZrO 2 CeO 2 Y 2 O 3 ) [8-11].Kompozitni prah NiCr/Al/Co/Y 2 O 3 je NiCrlegura sa 17,5 tež.% Cr obložena sa 5,5 tež.% Al,2,5 tež. % Co i 0,5 tež. % Y 2 O 3 . Prah koji jekorišćen u eksperimentu je imao raspon granulaciječestica od 45 μm do 150 μm. Temperatura topljenjapraha je 1400 °C. Na slici 1 je prikazama SEMmikrofotografija praha NiCr/Al/Co/Y 2 O 3 na kojojse vide čestice praha nepravilnog oblika.Slika 1. (SEM) morfologija čestica prahaNiCr-Al-Co-Y 2 O 3Keramički prah ZrO 2 MgO je proizvedentopljenjem. Prednost ove metode je predlegiranost iprirodna homogenizacija čestica praha. KeramikaZrO 2 MgO je predlegirana sa 25 tež.% MgO. Zaeksperiment je korišćen prah uglastog oblika sarasponom granulacije čestica od 10 μm do 53 μm.Temperatura topljenja praha je 2140 °C.Prah ZrO 2 Y 2 O 3 je proizveden tehnološkimpostupkom aglomeracije finih čestica keramike sa80 tež.% ZrO 2 i 20 tež.% Y 2 O 3 i suvimraspršivanjem. Tako proizveden prah ima sfernumorfologiju čestica. Raspon granulacije čestica jebio od 45 μm do 106 μm sa temperaturom topljenjapraha 2480 °C.Čestice praha ZrO 2 CeO 2 Y 2 O 3 su proizvedenehomogenizacijom u peći i procesom sferoidizacijeHOSP. Keramika ZrO 2 CeO 2 Y 2 O 3 je potpunopredlegirana sa 24 – 26 tež.% CeO 2 i 2 – 3 tež.%Y 2 O 3 . Morfologije čestica praha su prikazane naslici 2.SEM mikrofotografija pokazuje da su česticepraha sa sfernim morfologijama koje omogućavajuodličan protok praha u mlaz plazme i uniformnotopljenje. Prah koji je korišćen u eksperimentu jeimao raspon čestica od 45 – 90 μm satemperaturom topljenja 2480 °C.13 th International Conference on Tribology – Serbiatrib’13 427

Slika 2. (SEM) morfologija čestica prahaZrO 2 CeO 2 Y 2 O 3Uzorci za merenje tvrdoće i metalografskaispitivanja su pravougaoni 70 × 20 × 1,5 mm, doksu za zateznu čvrstoću spoja korišćeni cilindričniuzorci Ø25 × 50 mm.Merenja mikrotvrdoća su izvršena korišćenjemVikers dijamant piramide indenter i 100 gramaopterećenje (HV 0.1 ). Merenje je obavljeno u pravcuduž lamela, u sredini i na krajevima uzorka. Na petmesta sprovedeno je pet očitavanja a prikazane suminimalne i maksimalne vrednosti.Testovi zatezne čvrstoće spoja su vršeni nasobnoj temperaturi na hidrauličnoj opremi sabrzinom od 10 mm/min, za sva ispitivanja.Čvrstoća je izračunata tako što se opterećenjekidanja deli sa površinom poprečnog presekauzorka. Geometrija uzoraka je u skladu sa ASTMC633 standardom, koji je detaljnije obrađen u radu[23]. Korišćeni su u paru dva uzoraka, od kojih jeprevlaka deponovana samo na jednom od njih.Uzorci su zalepljeni lepkom i čuvani pod pritiskomjedni prema drugom u peći na temperaturi od 180°C za 2 sata. Za svaki grupu uzoraka urađene su triepruvete a prikazane su minimalne i maksimalnevrednosti. Mehaničke i mikrostrukturnekarakterizacije dobijenih prevlaka su izvršeneprema standardu Pratt & Whitney [24].Mikrostrukturna analiza prevlaka i imageanaliza udela mikro pora sa oksidima u veznimprevlakama i udela mikro pora u keramičkimprevlakama urađena je na svetlosnom mikroskopu.Morfologija čestica praha i morfologija površinekeramičke prevlake urađena je na SEM-u (skeningelektronskom mikroskopu).2.2 Plazma sprej parametriDepozicija prahova je urađena sa atmosferskiplazma sprej sistemom firme Plasmadyne i plazmapištoljem SG-100, sa odgovarajućim robotizovanimkontrolnim sprej uslovima. Plazma pištolj SG-100 sesastojao od katode tipa K 1083-129, anode tipa A2084-145 i gas injektora tipa GI 2083-113. Za svedeponovane prevlake kao lučni gas korišćen je Ar ukombinaciji sa He i snaga napajanja od 40 KW.Plazma sprej parametri depozicije su prikazani utabeli 1. Pre procesa deponovanja površine čeličnihsubstrata su hrapavljene sa česticama korunda veličineod 0,7 – 1,5 mm i predgrejane na temperaturu od 160– 180 °C. Vezne prevlake VP – NiCrAlCoY 2 O 3 sudeponovane sa jednim prolazom plazma pištoljadebljine 0,068 – 0,093 mm. Sve keramičke prevlakesu deponovane sa 15 prolaza plazma pištolja. PrevlakaA – ZrO 2 MgO je deponovana sa debljinom 0,486 –0,50 mm, prevlaka B – ZrO 2 Y 2 O 3 sa debljinom 0,375– 0,41 mm i prevlaka C – ZrO 2 CeO 2 Y 2 O 3 sadebljinom 0,425 – 0,470 mm.Tabela 1. Parametri depozicije prevlakaVP – NiCrAlCoY 2 O 3 , A – ZrO 2 MgO, B – ZrO 2 Y 2 O 3 iC – ZrO 2 CeO 2 Y 2 O 3Parametri depozicije VP A B CPlazma struja, I (A) 900 900 900 900Plazma napon, U (V) 38 38 38 38Primarni plazma gasprotok Ar, l/min47 47 47 47Sekundarni plazma gasprotok He, l/min12 12 12 12Noseći gasprotok Ar, l/min5 7 6 6Protok praha, g/min 60 50 50 50Odstojanje substrata, mm 115 100 90 903. REZULTATI I DISKUSIJA3.1 MikrotvrdoćaVrednosti mikrotvrdoće za sve vezne prevlakeNiCrAlCoY 2 O 3 su izmerene u rasponu od min. 279HV 0.1 do max. 322 HV 0.1 . Raspodele mikrotvrdoće uveznim prevlakama su posledica različite raspodelemikro pora i oksida u deponovanim slojevima. Ovevrednosti su potvrđene image analizom priodređivanju ukupnog sadržaja mikro pora i oksida uveznim slojevima. Na slici 3 su prikazane min. imax. vrednosti mikrotvrdoće keramičkih prevlaka A– ZrO 2 MgO, B – ZrO 2 Y 2 O 3 i C – ZrO 2 CeO 2 Y 2 O 3 .Kao što se i očekivalo za sve tri keramičkeprevlake dobile su se različite vrednostimikrotvrdoće kao posledica uticaja različitih vrsta isadržaja stabilizatora MgO, Y 2 O 3 i kombinacijeCeO 2 Y 2 O 3 . Najveće vrednosti mikrotvrdoće suizmerene u keramičkim slojevima A – ZrO 2 MgO saraspodelom mikrotvrdoće od min. 611 do max. 662HV 0.1 , a najmanje vrednosti mikrotvrdoće suizmerene u slojevima C – ZrO 2 CeO 2 Y 2 O 3 saraspodelom mikrotvrdoće od min. 525 do max. 542HV 0.1 . Raspodele mikrotvrdoće u keramičkimprevlakama su, kao i u veznim prevlakama,428 13 th International Conference on Tribology – Serbiatrib’13

posledica različite raspodele mikro pora ukeramičkim slojevima. Najmanja raspodelamikrotvrdoće je izmerena u keramičkim slojevimaC – ZrO 2 CeO 2 Y 2 O 3 , a najveća u keramičkimslojevima A – ZrO 2 MgO. Raspodele vrednostimikrotvrdoće u keramičkim slojevima su bile uskladu sa sadržajem mikro pora, što su potvrdileimage analize keramičkih prevlaka.Zatezna čvrstoća spoja, MPa7006005004003002001000611 662 514 552 525 542A B CPrevlakeSlika 3. Mikrotvrdoća prevlaka3.2 Zatezna čvrstoća spojaNa slici 4 su prikazane min. i max. vrednostizatezne čvrstoće spoja sistema prevlakaNiCrAlCoY 2 O 3 / ZrO 2 MgO, NiCrAlCoY 2 O 3 /ZrO 2 Y 2 O 3 i NiCrAlCoY 2 O 3 / ZrO 2 CeO 2 Y 2 O 3 .Zatezna čvrstoća spoja, MPa605040302010035 3646 47A B CPrevlake52 53Slika 4. Zatezna čvrstoća spoja prevlakaZa sve sisteme prevlaka su izmerene dobrevrednosti zatezne čvrstoće spoja. Predgrevanjesubstrata je omogućilo da se deponuju prevlake sadobrim vezivanjem lamela vezne prevlake zasupstrate i lamela vezne prevlake sa keramičkimlamelama. Veća temperatura substrata je uticala napovećanje adhezije prevlaka, mehaničkih svojstavai poboljšanje mikrostrukture, što su potvrdilametalografska ispitivanja.Najveće vrednosti zatezne čvrstoće spoja jeimao sistem prevlaka NiCrAlCoY 2 O 3 /ZrO 2 CeO 2 Y 2 O 3 sa rasponom od 52 – 53 MPa, anajniže vrednosti sistem prevlaka NiCrAlCoY 2 O 3 /ZrO 2 MgO sa rasponom od 35 – 36 MPa. OksidY 2 O 3 koji je istovremeno bio prisutan u veznoNiCrAlCoY 2 O 3 i keramičkoj prevlaciZrO 2 CeO 2 Y 2 O 3 je uticao na bolju inter-lamelarnukohezivnu čvrstoću između dve prevlake. Sistemprevlaka NiCrAlCoY 2 O 3 / ZrO 2 Y 2 O 3 je iz istograzloga imao veću vrednost zatezne čvrstoće spojasa raspodelom od 46 – 47 MPa u odnosu na TBCsistem NiCrAlCoY 2 O 3 / ZrO 2 MgO. Izmerenevrednosti su potvrdile da temperatura substrata imabitan uticaj na zatezne čvrstoće spoja. Za sveuzorke lom se dešavao na interfejsu između veznihslojeva i substrata.3.3 MikrostrukturaImage analiza veznih prevlaka NiCrAlCoY 2 O 3je pokazala da je ukupan sadržaj mikro pora ioksida u slojevima bio od min. 18 % do max. 22 %.Raspon ukupnog sadržaja mikro pora i oksida uveznim prevlakama su posledica različite raspodelemikro pora i oksida u deponovanim slojevima. Naslici 5 je prikazana mikrostruktura vezne prevlakeNiCrAlCoY 2 O 3 u sistemu sa keramičkomprevlakom ZrO 2 CeO 2 Y 2 O 3 . Vezni slojevi imajuuniformnu lamelarnu strukturu. Granice nainterfejsu između podloge i prevlake se jasno moguvideti. Na interfejsu sa podlogom i keramičkimslojem nisu prisutne mikro pukotine i makropukotine i ne postoji odvajanje slojeva prevlake iljuštenje sa metalnih substrata.Slika 5. Mikrostruktura prevlake NiCrAlCoY 2 O 3Kroz slojeve NiCrAlCoY 2 O 3 prevlake se jasnouočavaju tamni lamelarni oksidi i mikro pore.Podužne oksidne lamele oksida su formirane utečnom stanju u plazmi [8].SEM analiza morfologije površine keramičkeprevlake ZrO 2 CeO 2 Y 2 O 3 pokazuje potpunotopljenje i razlivanje keramičkih čestica naprethodno deponovani keramički sloj. Na slici 6 jeprikazana SEM mikrofotografija površinekeramičke prevlake ZrO 2 CeO 2 Y 2 O 3 .13 th International Conference on Tribology – Serbiatrib’13 429

Slika 6. SEM mikrofotografija površine prevlakeZrO 2 CeO 2 Y 2 O 3Na SEM mikrofotografiji je crvenom linijomzaokružena površina jedne istopljene i razlivenečestice praha. Potpuno istopljena čestica praha jeformirala tanak disk u sudaru sa površinomprethodno deponovanog sloja. Morfologijadeponovane čestice potvrđuje da su istopljenečestice u sudaru sa podlogom formirale pravilanoblik i kao takve ostvarile dobru vezu sa prethodnodeponovanim česticama. U poprečnom presekuprevlake deponovane čestice imaju lamelarnustrukturu.Image analiza ukupnog sadržaja mikro pora ukeramičkim prevlakama ZrO 2 MgO, ZrO 2 Y 2 O 3 iZrO 2 CeO 2 Y 2 O 3 je pokazala da slojevi imajurazličiti udeo pora. Najmanji udeo mikro pora jeizmeren u prevlaci ZrO 2 CeO 2 Y 2 O 3 sa sadržajem od14 %. U keramičkoj prevlaci ZrO 2 Y 2 O 3 je izmerenudeo mikro pora od 16 %, a u prevlaci ZrO 2 MgO jeizmeren udeo mikro pora od 17 %. Morfologijačestica praha je bitan parametar za ravnomeran tokpraha u plazmi i na njegovo topljenje. Keramičkiprah ZrO 2 CeO 2 Y 2 O 3 sa sfernom morfologijomčestica koje imaju glatku površinu su omogućileizvrsno i uniformno topljenje čestica u odnosu naklasične postupke izrade prahova [11].Ravnomerno istopljene čestice praha se pravilnijeoblikuju u sudaru sa substratom i deponuju slojevesa manjim sadržajem pora, koji imaju većukohezionu čvrstoću i zateznu čvrstoću spoja. Naslici 7 su prikazana mikrostrukture keramičkihprevlaka ZrO 2 MgO, ZrO 2 Y 2 O 3 i ZrO 2 CeO 2 Y 2 O 3 .Na interfejsu između keramičkih slojeva iveznih prevlaka nisu prisutne pukotine. Nauzorcima nije uočeno ljuštenje – piling keramičkihslojeva sa substrata. U keramičkim slojevima suprisutne tamne površine koje predstavljaju mikropore sa različitim udelima u zavisnosti od vrstekeramičke prevlake. Izmerene vrednosti sadržajamikro pora u keramičkim slojevima su u skladu samikrotvrdoćama i mikrostrukturama.4. ZAKLJUČAKSlika 7. Mikrostrukture keramičkih prevlaka: A –ZrO 2 MgO, B – ZrO 2 Y 2 O 3 i C – ZrO 2 CeO 2 Y 2 O 3 ;uvećanje 200×Na osnovu rezultata prikazanih u ovom radumože se zaključiti, da sistemi prevlakaNiCrAlCoY 2 O 3 / ZrO 2 MgO, NiCrAlCoY 2 O 3 /ZrO 2 Y 2 O 3 i NiCrAlCoY 2 O 3 / ZrO 2 CeO 2 Y 2 O 3 imajudobra svojstva. Predgrevanje substrata predepozicije prevlaka je omogućilo da se dobijuslojevi sa dobrim mikrostrukturama i mehaničkimosobinama. Keramički prah ZrO 2 CeO 2 Y 2 O 3 zbogsferne morfologije čestica praha je imao najmanjiudeo mikro pora i najbolju zateznu čvrstoću spoja.Oksid Y 2 O 3 koji je istovremeno bio prisutan uveznoj i keramičkoj prevlaci je uticao na boljuinter-lamelarnu kohezionu čvrstoću između dveprevlake. Sistem prevlaka NiCrAlCoY 2 O 3 /430 13 th International Conference on Tribology – Serbiatrib’13

ZrO 2 Y 2 O 3 je takođe zbog oksida Y 2 O 3 imao većuvrednost zatezne čvrstoće spoja od sistema prevlakaNiCrAlCoY 2 O 3 / ZrO 2 MgO.Najbolje karakteristike od svih sistema prevlakaje imao TBC sistem NiCrAlCoY 2 O 3 /ZrO 2 CeO 2 Y 2 O 3 .ZAHVALNOSTIAutor je zahvalan za finansijsku podrškuMinistarstva prosvete i nauke Republike Srbije(nacionalni projekti OI 174004, TR 34016).LITERATURA[1] B. Basler, R. Buergel, W. Hoffelner: Proc. of 1 st Int.Conf. on Plasma Surface Engineering, Germisch-Partenkirchen 1, pp. 347, 1988.[2] M.R. Mrdak, A. Vencl, B.D. Nedeljkovic, M.Stanković: Influence of plasma spraying parameterson properties of the thermal barrier coatings,Materials Science and Technology, Article in Press,doi: 10.1179/1743284712Y.0000000193, 2013.[3] A. Vencl, N. Manić, V. Popovic, M. Mrdak:Possibility of the abrasive wear resistancedetermination with scratch tester, Tribology Letters37, pp. 591-604, 2010.[4] E. Celik, I.H. Mutlu, E. Avci, Y.S. Hascicek: Y 2 O 3 -ZrO 2 insulation coatings on AgMg sheathed Bi-2212superconducting tapes by the sol-gel process, IEEETransactions on Applied Superconductivityb10,pp.1329-1332, 2000.[5] E. Celik, I.H. Mutlu, Y.S. Hascicek: Electricalproperties of high temperature insulation coatingsby the sol-gel method for magnet technology, IEEETransactions on Applied Superconductivity 10, pp.1341-1344, 2000.[6] E. Celik, E. Avci, Y.S. Hascicek: High temperaturesol-gel insulation coatings for HTS magnets and theiradhesion properties, Physica C: Superconductivity,Vol. 340, No. 2-3, pp. 193-202, 2000.[7] A.S. Demirkiran, E. Çelik, M. Yargan, E. Avci:Oxidation behaviour of functionally gradientcoatings including different composition of cermets,Surface and Coatings Technology 142-144, pp. 551-556, 2001.[8] Metco 461 Nickel Chromium-Aluminum-Cobalt-Yttria Composite Powder, Technical Bulletin 10-315, Sulzer Metco, 2000.[9] Metco 210 Magnesium Zirconate Powder, TechnicalBulletin 10-108, Sulzer Metco, 2000.[10] Metco 202NS Zirconium Oxide Composite Powder,Technical Bulletin 10-107, Sulzer Metco, 2000.[11] Metco 205NS PreAlloyed Ceria-Yttria StabilizedZirconia Powder, Technical Bulletin 10-338, SulzerMetco, 2000.[12] A. Vencl, M. Mrdak, I. Cvijović: Microstructuresand tribological properties of ferrous coatingsdeposited by APS (Atmospheric Plasma Spraying)on Al-alloy substrate, FME Transactions, Vol. 34,No. 3, pp. 151-157, 2006.[13] A. Vencl, M. Mrdak, M. Banjac: Correlation ofmicrostructures and tribological properties offerrous coatings deposited by atmospheric plasmaspraying on Al-Si cast alloy substrate, Metallurgicaland Materials Transactions A, Vol. 40, No. 2, pp.398-405, 2009.[14] M. Mrdak, A. Vencl, M. Ćosić: Microstructure andmechanical properties of the Mo-NiCrBSi coatingdeposited by atmospheric plasma spraying, FMETransactions, Vol. 37, No. 1, pp. 27-32, 2009.[15] M. Mrdak, A. Vencl: Uticaj parametara nanošenjaNiCr prevlake plazma sprej postupkom uatmosferskim uslovima na njene mehaničkekarakteristike i strukturu, Tehnička dijagnostika 10,pp. 9-14, 2011.[16] C.-J. Li, A. Ohmori: Relationships between themicrostructure and properties of thermally sprayeddeposits, Journal of Thermal Spray Technology 11,pp 365-374, 2002.[17] R. McPherson, B.V. Shafer: Interlamellar contactwithin plasma-sprayed coatings, Thin Solid Films97, pp. 201-204, 1982.[18] S. Kuroda, T.W. Clyne: The quenching stress inthermally sprayed coatings, Thin Solid Films 200,pp. 49-66, 1991.[19] C. Li, A. Ohmori, R. McPherson: The relationshipbetween microstructure and Young’s modulus ofthermally sprayed ceramic coatings, Journal ofMaterials Science 32, pp. 997-1004, 1997.[20] L. Bianchi, A.C. Leger, M. Vardelle, A. Vardelle, P.Fauchais: Splat formation and cooling of plasmasprayedzirconia, Thin Solid Films 305, pp. 35-47,1997.[21] V. Pershin, M. Lufitha, S. Chandra, J. Mostaghimi:Effect of substrate temperature on adhesion strengthof plasma-sprayed nickel coatings, Journal ofThermal Spray Technology 12, pp. 370-376, 2003.[22] P. Fauchais, M. Fukumoto, A. Vardelle, M.Vardelle: Knowledge concerning splat formation:An invited review, Journal of Thermal SprayTechnology 13, pp. 337-360, 2004.[23] A. Vencl, S. Arostegui, G. Favaro, F. Zivic, M.Mrdak, S. Mitrović, V. Popovic: Evaluation ofadhesion/cohesion bond strength of the thick plasmaspray coatings by scratch testing on coatings crosssections,Tribology International 44, pp. 1281-1288,2011.[24] Turbojet Engine – Standard Practices Manual (PN582005), Pratt & Whitney, East Hartford, USA,2002.13 th International Conference on Tribology – Serbiatrib’13 431

TESTING THE MECHANICAL AND THE STRUCTURAL PROPERTIES OFTHE COATING RESISTANT TO THE EROSION AND THE HIGHTEMPERATUREAbstract: The aim of this study was to investigate the mechanical and the structural characteristics of thethermo-barrier coatings (TBC) resistant to the erosion and the high temperature. The deposited are the threetypes TBC dual coating systems which were consisting of a metal-bonded coatings NiCrAlCoY 2 O 3 and aceramical insulating coatings ZrO 2 MgO, ZrO 2 Y 2 O 3 and ZrO 2 CeO 2 Y 2 O 3 . TBC systems of the coatings weredeposited, with process of atmospheric plasma spraying (APS), on the roughened steel substrates with atemperature of 160 – 180 °C. The coatings were deposited with the optimal parameters of depositionpowders. A bonding layers were deposited with a single pass of plasma gun, and ceramical layers with afourteen passages. Assessment of a quality layers was done by the testing microhardness with method HV 0.1and bond strength by the testing on tensile. Metallographic assessment proportion of micro-pores (imageanalysis), in the structure of the bonding and the ceramical layers, was done with the technique of the lightmicroscopy. The Morphology of the particles powder was done on the SEM. Analysis of the performedtestings have been enabled to choose the TBC systems of the coatings with the best mechanical andstructural characteristics.Keywords: atmospheric plasma spraying (APS), thermo-barrier coatings (TBC), microstructure, interface,microhardness, strength bond.432 13 th International Conference on Tribology – Serbiatrib’13

Serbian TribologySocietySERBIATRIB ‘1313 th International Conference onTribologyKragujevac, Serbia, 15 – 17 May 2013Faculty of Engineeringin KragujevacIZBOR MERNE GLAVE DIFERENCIJALNOG PNEUMATSKOGKOMPARATORA ZA KONTROLU UNUTRAŠNJIH MERA MAŠINSKIHDELOVADragiša Skoko 1 , Cvetko Crnojević 1 , Mileta Ristivojević 11 Mašinski fakultet u Beogradu, Srbija, dskoko@mas.bg.ac.rs, ccrnojevic@mas.bg.ac.rs, mristivojevic@mas.bg.ac.rsApstrakt: Apstrakt: U ovom radu je analiziran uticaj promene uticajnih faktora na rad pneumatskogkommparatora.. Instalacija koja se koristi u ovu svrhu omogućava promenu pritiska napajanja, promenuizlazne mlaznice i prečnoka mlaznice merne komore komparatora. Posebna pažnja je posvećenapneumatskoj osetljivosti i dobijeno je da zavisi od pritiska napajanja i prečnika mlaznice u mernoj komori.Utvrđeno je da pneumatska osetljivost raste sa smanjenjem prečnika mlaznice u mernoj komori. Takođe jedat izbor prečnika merne glave pneumatsko komparatora za kontrolu munutrašnjih mera mašinskih delova.Ključne reči: kontrola, merna glava, diferencijalni pneumatski komparator, mašinski delovi, unutrašnjemere, pneumatska osetljivost1. UVODUporedo sa razvojem proizvodnih mašinarazvija se tehnika kontrole i merenja proizvedenihkomada. Prvi uređaji za kontrolu su bili mehanički.Ovi uređaji su mogli da zadovolje potrebu upojedinačnoj i maloserijskoj proizvodnjijednostavnih komada. Razvojem savremeneindustrije XX-tog veka, koju karakterišuvelikoserijska proizvodnja komplikovanih oblika ivelika tačnost izrade, mehanički uređaji nisu moglida zadovolje u svim segmentima proizvodnje. Tonaročito dolazi do izražaja u automobilskoj i avioindustriji. Za merenje i kontrolu razvijeni supotpuno novi uređaji: mehanički, optički, električni,pneumatski, hidraulički i laserki uređaji. Najvažnijaprednost u odnosu na mehaničke je njihova tačnost.Tačnost mehaničkih uređajaje je 1 μm, a pomoćusavremenih uređaja mogu se meriti i deseti i stotidelovi mikrometra.Pneumatski merni sistemi se intenzivnorazvijaju tridesetih godina XX-tog veka, počev odnajednostavnijih [1], pa sve do vrlo savremenihuređaja [2]. Zbog velike mogućnost primene,jednostavne konstrukcije, lakog održavanja,jednostavnog rukovanja a iznad svega veliketačnosti, ovi uređaji su dominantni u kontrolimašinskih delova. Prednost ovih uređaja, u odnosuna ostale, je što komprimovan vazduh koji izlazi izmlaznica merne glave velikom brzinom, oduvamehaničke nečistoće i tanak sloj tečnosti zahlađenje komada i ostalih nečistoća. Na ovaj načinse smanjuje mogućnost pojave greške pri merenju ikontroli. Druga prednost je što se može kontrolisativiše mera istovremeno, bilo spoljašnjih iliunutrašnjih. Vrlo značajnu primenu ima dinamičkapneumatska metoda koja se primenjuje kod obrtnihkomada, znači u toku rada, bez zaustavljanja radamašine, vrši se merenje i kontrola obratka.Teorijske osnove ovog postupka su obrađene uliterarturi [.3], [.4], [.5]. Stepen tačnosti kontrolemašinskih delova zavisi od: izbora izlaznemlaznice, pritiska napajanja p a , prečnika prigušniceD u mernoj komori i prečnika merne glave. Uraduje prikazan postupak određivanja prečnika merneglave. Analiziran je uticaj pneumatske osetljivosti iodstupanja kontrolisanog mašinskog dela na izbordimenzija merne glave.2. EKSPERIMENTALNA INSTALACIJAZa odredjivanje polja pritiska na ravnoj površinimernog komada koristi se instalacija prikazana naslici 1 i slici 2. Instacija se sastoji od pneumatskogkomparatora PC, koji sadrži mernu (B1) ireferentnu granu (B2), izvora vazduha pod13 th International Conference on Tribology – Serbiatrib’13 433

pritiskom CAS, koji se reguliše regulatorompritiska PR, izlazne mlaznice N i sistema D-r i D-δza fina pomeranja ravnog mernog komada. Umernoj grani se nalazi mlaznica A prečnika D učijem grlu se ostvaruje kritično strujanje. U ovomradu prečnik D je imao vrednost 0,7 i 1 mm, kojiodgovara realnim primenana, a tretitan je slučajkada je D=6 mm što odgovara kada u dovodnojgrani napajanja mlaznice N nema pneumatskogprigušenja. Pritisci napajanja p a menjani su uintervalu od 2 do 5 bar. Svi pritisci mereni sumanometrima koji imaju tačnost 0,001 bar. Izlaznamlaznica N je unutrašnjeg pečnika 2 mm ispoljašnjeg 4 mm. Ostvarivanje željenih položaja ri δ postiže se tako što se iznad fiksnog mernogkomada W pomera izlazna mlaznica N sistemimapomeranja D-r i D-δp()pTolerance rangeareaa)Linearity zoneNominal dimenssionSlika 3. Geometrija izlazne mlaznice [5]Pojava podpritiska na mernom komadu jenepoželjna zbog skupljanja nečistoća, koje remetiispravan rad komparatora. Takođe pogodnimodabirom geometrije izlazne mlaznice teži se dapolje podpritiska bude što dalje od ose mlaznice.Analiza polja pritiska na rad pneumatskogdiferencijalnog komparatora je data u radu [5].Karakteristični oblici mlaznica su dati na slici 3.p [bar]5432Pritisak na mlaznici M-1 za =100 mp a=2 barp a=3 barp a=4 barp a=5 bar1b)CASPRControledpieceSlika 1. Shematski prikazi pritiska p(δ)PCB2B1Nrp(r,p(0,00 500 1000 1500 2000 2500 3000r [m]Slika 4. Dijagram p=f(r) za M-1 i p ap [bar]Pritisak na mlaznici M-2 za =100 m5p =2 bar4ap =3 bara3p = 4 barp = 5 bar2p a3.1 MlazniceADD-D-rSlika 2. Shematski prikaz - merna instalacijaPolje pritiska na površini mašinskog dela koji sekontroliše zavisi od geometrije izlazne mlaznice.Analizom uticajnih faktora na polje pritiska, bira setakva mlaznica da se potpuno eliminiše podpritisakna komadu koji se kontroliše ili da taj pritisak bude,ako je moguće, što manji.W10r [m]0 500 1000 1500 2000 2500Slika 5. Dijagram p=f(r) za M-2 i p aAnaliza uticaja pritiska napajanja i geometrijeizlazne mlaznice na polje pritiska na površinimašinskog dela koji se kontroliše prikazana je nadijagramima na 4 i 5.Analizom dijagrama, slika 4, vidi se da je poljepritiska pozitivno za mlaznicu M-1 po celoj dužinii za sve pritiske napajanja. Polje pritiska zamlaznicu M-2 je pozitivno za za p a =2 i p a =3 bar aza veće pritiske p a =4 i p a =5 bar je negativno, štonije dobro, slika 5434 13 th International Conference on Tribology – Serbiatrib’13

3.2 Pneumatska osetljivostOblast primene pneumatske metrologije jeograničena pravolinijskim delom dijagrama p=p(δ).Pogodnim odabirom prigušnice dobija seodgovarajuća karakteristika pneumatskogdiferencijalnog uređaja - osetljivost uređaja.Pneumatska osetljivost S predstavlja odnos priraštajpritiska p i rastojanja iz pravolinijskog delakrive zavisnosti totalnog pritiska od rastojanjamlaznice i površine mernog komadap a[bar]p a[bar]S [bar/m]S=. Δp / Δδ – ref [6]54321Mlaznica M-2, prigusnica D=1,2 mmp a=2 barp a=3 barp a=4 barp a=5 bar [ m]00 200 400 600 800 1000Slika 6. Dijagram p=f(δ). za M-2 i različite p a54321Mlaznica M-2, prigusnica D=0,7 mmp a=2 barp a=3 barp a=4 barp a=5 bar00 200 400 600 800 1000 [ m]Slika 7. Dijagram p=f(δ) a M-2 i različite p a0.0400.0350.0300.0250.020D=1,7D=1,2D=1,0D=0,7D=0,5Osetljivost za M-1Iz definicije pneumatske osetljivosti uređaja,proizilazi da pneumatski uređaj koji ima većeprigušenje ima veću osetljivost a pravolinijski deokrive zavisnosti pritiska i rastojanja δ ima većiugao nagiba u odnosu na x osu tj. prava je strmija.Pneumatski diferencijalni uređaj koji ima manjeprigušenje ulaznog pritiska u mernu komoru imamanju osetljivost a pravolinijski deo krive imamanji nagib u odnosu na x osu. Osetljivost uređajaje, dakle definisana nagibom krive. Osetljivostuređaja predstavlja tačnost uređaja. Opseg primeneuređaja definisan je dužinom projekcijepravolinijskog dela krive na osu x. Uređaji kojiimaju veću dužinu projekcije imaju veći opsegprimene, tj. veću širinu tolerancijskog polja.Pneumatski diferencijalni uređaji koji su predviđenida rade sa većim pritiskom napajanja imaju većiopseg primene.Na dijagramima sa slike 5 vidi se da je malaosetljivost zbog malog prigušenja D=1,2 mm. Nadijagramima , slika 6 se vidi da krive imaju velikipad tj. veliko prigušenje D=0,7 mm i velikuosetljivost.Na slici 7, data je osetljivost mlaznica M-1 i M-2 uzavisnosti od stepena prigušenja i pritiska napajanjap a . Veliko prigušenje daje veliku pneumatskuosetljivost i obrnuto. Isto tako se vidi da je linearnopovećanje osetljivosti sa povećanjem pritiskanapajanja p a. Takođe se vidi uticaj oblika mlaznicena osetljivost. Najveća osetljivost je za mlaznicuM-5, prečnik prigušnice D=0.7 mm i p a =5 bar iiznosi s=0,039bar/mm.4. ODREĐIVANJE PREČNIKA MERNEGLAVEPneumatski uređaj za kontrolu dimenzija jespecijalan manometar sa skalom, koji umesto skalepritiska ima skalu za očitavanje zazora δ (rastojanjevrha mlaznice i površine komada koji sekontroliše). Pneumatski komparator jetransformisani manometar. Za zadati pritisaknapajanja p a , izabrani prečnik prigušnice D iizabranu mlaznicu M, na uređaju se očitava zazor δza odgovarajuće vrednosti pritiska p a . ,,Takođe ukonkretnim izvođenjima pneumatskih komparatorane meri se pritisak p(δ) na površini komada koji sekontroliše, već pritisak u komori merne grane p0.0150.0100.0050.0001 2 3 4 5Slika 8. Osetljivost mlaznice M-1p [bar]13 th International Conference on Tribology – Serbiatrib’13 435

Merna glava za kontroluspoljašnjih dimenzijaKomad koji se kontrolišed d – donja granična merad - nazivna merad g – gornja granična meraD mg –prečnik merne glave+dd gd dp (δ)2D mgp2Δp0pop1Δδ1DgδDMerna glava za kontroluunutrašnjih dimenzijad mgD dKomad koji se kontrolišeSlika 9. Šematski prikaz dijagrama p=p(δ) i merneglave za kontrolu spoljašnje i unutrašnje mere [5]436 13 th International Conference on Tribology – Serbiatrib’13

p mg (δ), koji je zbog malih strujnih gubitakapribližno jednak pritisku pZavisnost p=p (δ) prikazana dijagramom na slici9. Na ovom dijagramu uočava se pravolinijski deo.Tačku 2 predstavlja početak pravolinijskog dela 1. 1. Tolerancijsko polje unutrašnje mere je simetričnodijagrama i nju karakteriše povećan pritisak p 2 a u odnosu na nultu liniju, tj. kada su odstupanja istamale vrednosti δ. Na uređajima se ne koristi tačka 2ES = EI.kao reperna, već tačka ispod nje ,,g”., koja imakarakteristiku manji pritisak a veće δ. Ta tačkad mg = D -2 δ 0 - prečnik merne glavepredstavlja gornju granicu tolerancijskog poljakomada koji se kontriliše. Prevedeno u kontroli δ 0 – rastojanje merne glave od površine komadaspoljašnjih dimenzija to je komad sa maksimalno koji se (horizontalna koordinata tačke ,,0”).dozvoljenom merom i na uređaj za očitavanje tubismo stavili reper es, što predstavlja gornjegranično odstupanje. Uobičajeno je da je taj reperna merilu plave boje. Komad koji bi imao većeodstupanje od zadatog bi morao da ide na doradu.Pri kontroli unutrašnjih dimenzija ova tačka bipredstavljala minimalnu dozvoljenu meru i nauređaj za očitavanje tu bismo stavili reper kojioznačava minimalno odstupanje EI. U tom slučajuova tačka predstavlja donju graničnu meru.Povećanjem rastojanja δ smanjuje se pritisak.Naredna karakteristična tačka na pravolinijskomdelu dijagrama p=f(δ) je tačka ,,0”, koja uglavnompredstavlja sredinu dužine pravolinijskog deladijagrama. Ovo je značajna tačka ovog dijagramajer se koristi za određivanje dimenzija merne glave. Slika 10. Prikaz simetričnog tolerancijskog poljaTakođe njena horizontalna projekcija predstavlaunutrašnje merenultu liniju u tolerancijama.Treća tačka na ovom delu dijagrama je tačka 2. Tolerancijsko polje unutrašnje mere nalazi,,d”. Karakteriše je manji pritisak i veliko δ. Onaiznad nulte linije, radi se preslikavanjepredstavlja drugu granicu tolerancijskog poljatolerancijskog polja na tolerancijsko polje ,,J s ”.komada koji se kontroliše. Pri kontroli spoljašnjihOvo polje se naziva ekvivalentno tolerancijskodimenzija to je minimalno dozvoljena mera. Napolje, a postiže se povećanjem prečnika merneskali pneumatskog uređaja se stavi reper kojiglave za vrednost C = EI/2 +ES/2.označava ei, i predstavlja donje granično d mg = D -2δ 0 +(EI/2+ ES/2) - prečnik merne glaveodstupanje. Pri kontroli spoljašnje dimenzijekomada, ako je stvarno donje odstupanje veće odzadatog, komad se odbacuje jer mu je stvarna meramanja od najmanje dozvoljene. To važi i zakontrolu unutrašnjih dimenzija. Horizontalnaprojekcija tačke ,,d” na dijagramu koristi se kaoreper na skali uređaju i predstavlja ES, gornjegranično odstupanje. Svaka stvarna mera koja imaveće odstupanje od zadatog je loša mera i komad seodbacuje.4.1. Određivanje prečnika merne glave zakontrolu unutrašnjih dimenzijatolerancijsko polje. Pri određivanju prečnika merneglave mora se voditi računa o položajutolerancijskog polja u odnosu na nazivnu meru.Imamo tri različita slučaja:Merna glava za kontrolu unutrašnjih dimenzijaje valjak koji po obimu ima dve ili više identičnihmlaznica. Svaka tolerisana mera ima parametre: D -nazivnu meru, D g – gornju graničnu meru, D d –donju graničnu meru, ES – gornje graničnoodstupanje, EI – donje granično odstupanje i T -Slika 11. Prikaz nesimetričnog tolerancijskog poljaunutrašnje mere - tolerancijsko polje iznad nulte linije3. Ako se tolerancijsko polje unutrašnje merenalazi ispod nulte linije, odstupanja su negativna.13 th International Conference on Tribology – Serbiatrib’13 437

Dato tolerancijsko polje se preslikava utolerancijskog polja ,,J”. Ovo polje se nazivaekvivalentno tolerancijsko polje, a postiže sesmanjenjem prečnika merne glave za vrednostE = |EI/2 +ES/2I|.d mg = D -2 δ 0 -|ES/2+EI/2|- prečnik merne glavekonvergentno-divergentne mlaznice merne granepneumatska osetljivost se povećava porastompritiska napajanja komprimovanog vazduha.Pomoću diferencijalnog pneumatskog komparatorakontrolišu se spoljašnje i unutrašnje tolerisanemere. Za svaku tolerisanu meru mora se odreditiprečnik merne glave D mg i d mg . Prečnik merne glavezavisi od položaja tolerancijskog polja u odnosu nanazivnu meru, pritiska napajanja p a i δ 0 .ZAHVALNOSTRad je rezultat istraživanja realizovan u okviruprojekta TR 35029. Rad je finansiran odMinistarstva prosvete, nauke i tehnološkog razvojaRepublike Srbije na čemu sam im zahvalan.LITERATURASlika 12. Prikaz nesimetričnog tolerancijskog poljaunutrašnje mere - tolerancijsko polje ispod nulte linije2. ZAKLJUČAKNa osnovu eksperimentalnih rezultataprikazanih u ovom radu dolazi se do opštihzaključaka:Iako se različita geometrija izlazne mlazniceprvenstveno koristi radi eliminisanja vrtložne zoneizmedju mlaznice i kontrolisanog komada (ref. [2] i[5]), dati rezultati merenja pokazuju da segeometrijom izlazne mlaznice može uticati i napneumatsku osetljivost komparatora i na promenumaksimalnog tolerancijskog polja kontrolisanogkomada.Na pneumatsku osetljivost pneumatskogkomparotera se prvenstveno utiče promenomprečnika konvergentno-divergentne mlaznicemerne grane. Povećanje pneumatske osetljivosti sepovećava smanjenjem prečnika mlaznice mernekomore, ali se i na taj način i sužava maksimalnotolerancijsko polje kontrolisane mere Za istegeometrijske parametre izlazne mlaznice i[1] Fortier, M.,: Application industrielles desécoulements gazeux à la vitesse critique,Revenu chaleur et Industrie, N°299, p.145.2.(1950)[2] Crnojevic C., Roy G., Bettahar A. and FlorentP.: Influence of regulator diameter andinjection nozzle geometry on flow structure inpneumatic dimensional control systems.Transacions of ASME, Journal of FluidsEngineering, Vol. 119., pp. 609-615 (1997).[3] Crnojević C., Skoko D.,: O nekim pojavamakoje utiču na rad pneumatskog komparatora.XXXI JUPITER konferencija, Zlatibor 2005.Zbornik radova (na CD-u), str. 5.14-5.17.[4] Roy, G., Crnojevic C., Bettahar A., FlorentP.and Vo-Ngoc, D., "Influence of nozzlegeometry in radial flow applications,"International Conference on Fluid andThermal Energy Conversion, Proc. Vol. 1,pp.363-368 (1994), Bali, Indonesia[5] Skoko, D., Magistarski rad, Mašinski fakultetBeograd 2007.[6] Skoko D., Crnojević C., : Eksperimentalnoodređivanje osetljivosti pneumatskogkomparatora. XXXII JUPITER konferencija,Zlatibor 2007. Zbornik radova , strana -5.43.438 13 th International Conference on Tribology – Serbiatrib’13

PNEUMATIC PROBE HEAD SELECTION FOR MACHINE PARTSINTERNAL MEASURES CONTROL WITH DIFFERNTIAL PNEUMATICCONTROLERApstract: In this paper, we analyze the influence of various parameters on the work of pneumatic controller.For that purpose, experimental installation is used, which enables the variation of the pressure ofcompressed air source, output nozzle and the diameter of the measuring branch nozzle of pneumaticcontroller. Special attention is devoted to pneumatic sensitivity and it is determined that it depends on thediameter of measuring branch nozzle and supply pressure. In addition, it is determined that pneumaticsensitivity increases with the decrease of the diameter of measuring branch nozzle. In this paper, we treatedof the selection of pneumatic measuring head of diferential pneumatic controler for control internalmeasures of machine partsKeywords: control, pneumatic measuring head, pneumatic controller, machine parts, internal measure,pneumatic sensitivity13 th International Conference on Tribology – Serbiatrib’13 439

Serbian TribologySocietySERBIATRIB ‘1313 th International Conference onTribologyKragujevac, Serbia, 15 – 17 May 2013Faculty of Engineeringin KragujevacPOVEĆANJE POUZDANOSTI PODSISTEMA KOPANJAROTORNOG BAGERA PODEŠAVANJEM TRIBOLOŠKIHKARAKTERISTIKA REZNIH ELEMENATAVojin Vukotić 1 , Dragan Čabrilo 1Rudnik i Termoelektrana Gacko, Bosna i Hercegovina, Republika SrpskaAbstrakt: Poznato je da prilikom obrade tla, u ovom slučaju skidanja jalovone i kopanja korisne mineralnesirovine korištenjem rotornih bagera, rezni elementi mašina koji su u kontaktu sa tlom trpe opterećenje ihabaju se. Rezni elementi na rotornom bageru su kofice i zubi. U ovom radu učinjen je pokušaj da semodeliranjem geometrijskih parametara zuba na kofici isti prilagode režimu i kinematici rezanja tla i smanjise nivo habanja istih. Cilj je da se poveća pouzdanost rotornog bagera preko povećanja pouzdanostipodsistema kopanja, čiji je glavni dio rotorni točak sa reznim elementima.Vršeno je snimanje vremenskeslike stanja podsistema kopanja za nekoliko varijanti geometrijskog oblika zuba i proračunavanaodgovarajuća pouzdanost.Naravno, usvojena je geometrija zuba uz koju je postignut najveći nivo pouzdanosti.Ključne reči: rotorni bager, pouzdanost, zubi, habanje, geometrija1. UVODRotorni bageri su danas najrasprostranjenijeosnovne mašine na površinskim kopovima. Služeza uklanjanje jalovine ili za kopanje korisnemineralne sirovine.Po svom sastavu to je vrlo složen mašinskisistem. Sastavljen je od niza podsistema, koji svakiponaosob obezbjeđuje određenu funkcijuneophodnu za rad bagera.Podjela na podsisteme je uglavnom sledeća: podsistem za kretanje bagera, podsistem kopanja, podsistem prijemnog transportera, podsistem odlagajućeg transportera i podsistem za zakretanje gornjekonstrukcije.Ukoliko samo jedan od pomenutih podsistemanije u funkciji, tada ni rotorni bager kao cjelina nezadovoljava namjeni, odnosno ne obavlja funkcijukriterijuma u propisanim granicama.Zapaženo je i to da je različit uticaj naraspoloživost rotornog bagera, stanje ispravnostiodnosno pouzdanost pojedinih podsistema.Teorijskim razmatranjima i ispitivanjima krozeksploataciju ovih mašina [2], došlo se dozaključka da je podsistem kopanja najnepouzdanijii da njegova pouzdanost najviše snižava ukupnupouzdanost bagera.Na osnovu određenih istraživanja na rotornombageru ЭR – 1250 u sklopu gatačkog ugljenogbasena koje je provodio autor ovog saopštenja, [4]može se A B C dijagramom prikazati učešćeotkaza po podsistemima rotornog bagera kao na sl.1, gdje je:Slika 1. Analiza otkaza na bageru440 13 th International Conference on Tribology – Serbiatrib’13

1. Podsistem kopanja2. Podsistem prijemnog transportera3. Podsistem odlagajućeg transportera4. Podsistem za kretanje5. Podsistem obrtne plarformeOčigledno je, sa pomenutog dijagrama, dapodsistem kopanja u ukupnim otkazima participirasa otprilike 44%.Podsistem kopanja sastoji se od tri glavnasklopa: pogonskog elektromotora, reduktora iradnog točka sa vedricama. Naravno svaki od ovihsklopova učestvuje u ukupnim zastojimapodsistema kopanja sa svojim zastojima.U ovom radu, predmet interesovanja i analize jesklop radnog točka sa vedricama, a sami elementkoji se tretira jeste zub i uticaj radne sredine nahabanje istog.2. OPŠTI PRISTUP PROBLEMU HABANJAHabanje mašinskih dijelova je pojava koja seneminovno javlja na mašinama u toku njihova rada,odnosno eksploatacije. Posljedica je relativnogkretanja i pomjeranja dijelova u sklopu kojiostvaruju fizički kontakt.Spada u štetne pojavne oblike u toku životnogciklusa mašine, javlja se intenzivnije u posljednojetapi toga ciklusa a ovim problemom se bavekonstruktori mašine , rukovaoci i održavaoci iste,kako bi ovo štetno dejstvo sveli na što manju mjeru.Naučna oblast koja se bavi ovom pojavomnaziva se tribologija.Dva su osnovna cilja koja se stavljaju predtribologe, a to su: da mašina što duže traje, da ta mašina troši što manje energije,a u poslednje vrijeme, javlja se i treći; da mašina ima što veću pouzdanost.Zadovoljavanjem ovih ciljeva minimiziraju se iukupni troškovi proizvodnje sa konkretnommašinom.2.1 Abrazivno habanjeNajveći broj mašinskih elementata radi uuslovima abrazivnog habanja. Pojava jekarakteristična po tome što površina koja je ukontaktu dolazi u dodir sa tvrdim česticama –abrazivima. Abrazivi mogu biti sadržani u radnojsredini ( slučaj kod alata za obradu rezanjem,obradi tla, i sl. ) ili su na drugi način dospjeli dokontaktnih površina ( iz vazduha, u mazivu,prašina, pijesak, čestice blata, … ).Proces abrazivnog habanja je u uskoj vezi sakoličinom i vrstom abrazivnih čestica, njihovomtvrdoćom i oblikom, kvalitetom i stanjem tarnihpovršina, pritiskom i temperaturom u zoni dodira,te brzinom relativnog kretanja i karakteromkretanja.Abrazivno habanje je u dosta slučajeva bilopredmet interesovanja istraživača. Uspostavljena jezavisnost između intenziteta habanja u smislupromjene zapremine dijela na koji se habanjeodnosi i normalnog opterećenja spregnutih površinai tvrdoće abrazivnih čestica.Presudan uticaj na intenzitet abrazivnoghabanja, po mišljenju mnogih autora, [1] i [3] imaopterećenje odnosno aktuelna sila na kontaktnimpovršinama. Intenzitet habanja je upravoproporcionalan sa veličinom opterećenja, a zavisi iod vrste i oblika tarnih površina, uglova u kontaktu,vrste trenja, režima rada te naročito od vrsteobrađivanog materijala, odnosno tla koje se kopa, uslučaju rotornih bagera.I pored svih preventivnih mjera, habanje se nemože potpuno spriječiti ali se na tu pojavu možeznačajno uticati ako se prethodno spozna njenasuština, te mehanizam nastanka i razvoja.2.2. Habanje zuba rotornog bageraProblem habanja radnih elemenata na vedricamarotornih bagera prisutan je još od samogpočetka primjene ovih mašina, koje su se napovršinskim kopovima pokazale kao vrloproduktivne i ekeonomične mašine.U procesu kopanja dolazi do intenzivnoghabanja reznih površina izazvanih suvim trenjemklizanja minerala kombinovanog sa značajnimdinamičkim udarima. To dovodi do neželjenihpromjena na elementima mašine u vidu odvajanjačestica materijala sa posmatrane površine zuba.Na proces habanja reznih elemenata, osimmaterijala koji se kopa utiče i niz drugih faktora,kao što su: materijal od koga su izrađeni reznielementi, konstruktivne karakteristike zuba, režim rada rotornog bagera i specifični otpori kopanjuU procesu kopanja dolazi do klizanja materijalapo reznim elementima pri čemu nastaje neželjenohabanje zuba. Proces habanja se nesumnjivoodražava na geometrijske karakteristike reznihelemenata, na njihov vijek trajanja i opštuupotrebljivost, a indirektno na kapacitet rotornogbagera, potrošnju energije i opštu spremnost bageraza normalnu eksploataciju. Očigledno je dafenomen habanja ima veliki uticaj na pogonskuspremnost i pouzdanost rotornog bagera, te ga jepotrebno pažljivo proučiti kako sa aspekta uzroka ipreventivnog djelovanja tako isto i u oblastisanacije i uklanjanja eventualnih posljedica.13 th International Conference on Tribology – Serbiatrib’13 441

Kroz praksu i terenska ispitivanja odavno jepoznato da povećano habanje zuba na vedricamarotornih bagera, za posljedicu, ima sledećenegativnosti: povećane otpore kopanju smanjen specifičan učinak bagera [t/h] povećane vibracije mašine povećana specifična potrošnja energije povećane troškove održavanja smanjenu pouzdanost bagera.3. RADNA OPTEREĆENJA I VRSTEHABANJA ZUBA ROTORNOG BAGERAEksploatacioni uslovi za zube na rotornombageru koji su u neposrednom kontaktu sa radnomsredinom koja se kopa, su bez sumnje, vrlo teški isloženi. Habanje zavisi prvenstveno od osobinamaterijala koji se kopa, otpornosti materijala zubana habanje ili bolje reći od konstruktivneprilagođenosti zuba uslovima kopanja nakonkretnom lokalitetu, odnosno kopu, kao i odrežima rada bagera.U ovom slučaju, zubi su izloženikombinovanom dejstvu abrazije i udara, koji semanifestuju kroz nagle skokove opterećenja zuba utoku rada bagera.Abrazivno habanje u konačnom obliku rezultiraodnošenjem materijala sa radnih površina zuba i tona takav način da nakon što bude “ skinut “ jedanpovršinski sloj, dolazi pod udar procesa habanjasledeći površinski sloj, itd.Ako uporedimo dejstvo abrazije sa brušenjem ilirezanjem kao procesima mehaničke obradematerijala, prva konstatacija je da većina abrazivnihčestica ima negativan ugao rezanja. Zbog toga oneizazivaju na površinama, po kojima se taru,karakteristične ogrebotine-abrazivne brazde,praćene velikom plastičnom deformacijom itečenjem materijala.Istovremeno efekat rezanja, koji varira odčestice do čestice, zahvaljujući izvjesnomsmicajnom dejstvu, čiji je uzrok opet negativanrezni ugao, produkuje mikro strugotinu, a to jenačin odnošenja materijala – abrazije.U toku procesa habanja bagerskog zuba prisutnaje situacija, što se geometrije zuba tiče, kao naslici 2.Početna kontura vrha zuba prikazana je linijom“0”. Linije “1” do “6” prikazuju konture zuba ufunkciji vremena habanja u radu. Kada se zubpohaba za veličinu h b koja je za zube postojećekonstrukcije otprilike 30÷40 mm, na odrađenih200÷250 radnih časova geometrijski izgled vrhazuba predstavljen je linijom – “6”. Tada se vršiskidanje pohabanih zuba, stavljaju se novi ilireparirani, a pohabani šalju na reparaturu.Slika 2. Faze habanja zubaKapacitet rotornog bagera zavisi uglavnom odrada podsistema za kopanje. Tu je veliki doprinospravilnih i oštrih zuba kada je i kapacitet najveći. Ufunkciji vremena dolazi do zatupljenja zuba usljedabrazivnog habanja, odnosno do promjene njihovoggeometrijskog oblika. Usljed zatupljenja zubapovećava se i otpor kopanju, a s tim u vezi dolazido povećanog opterećenja prenosnika, pa i čitavekonstrukcije podsistema kopanja.Konačno, ovo ima za posljedicu promjenurežima rada bagera, dolazi do smanjenja njegovogkapaciteta, što se negativno odražava na ekonomskeefekte proizvodnje.4. PRILAGOĐAVANJE GEOMETRIJEOBLIKA ZUBAZub rotornog bagera, koji je u ovom radupredmet interesovanja, predstavlja kao pojam –mašinski element. Kada tretiramo konstrukciju bilokog mašinskog elementa onda se tu prvenstvenomisli na definisanje: geometrijskog oblika, materijala, kvaliteta obrade i termičke obradeOvom prilikom ćemo se ograničiti na aspektgeo-metrije zuba. Da bi na izvjestan načinkvantificirali nivo po-habanosti zuba, pristupilo semjerenju pojasa pohabanosti. Usvojen je dužinskiparametar h b prema sl.2, kao pokazatelj istrošenostizuba. Ova veličina je mjerena u određenimvremenskim razmacima pa su na osnovu tihmjerenja konstru-isani dijagrami zavisnosti h b =f(t),sl.3.442 13 th International Conference on Tribology – Serbiatrib’13

Slika 3. Kriva habanjaNa tom dijagramu uočljive su tri zone. U prvojdolazi do početnog intenzivnog habanja zuba.Kasnije zona II , zub se ravnomjerno i umjerenohaba, da bi se u zoni III počeo intenzivno habati itada se pristupa zamjeni zuba. Opredeljujućaveličina za zamjenu zuba je veličina t 0 – teorijskivijek trajanja. Poznato je iz prakse da se zub nemijenja striktno po isteku perioda t 0 nego posleperioda t z , koji se nalazi u tolerantnoj zoni t tol uodnosu na t 0 . Veličina t z se može iskustvenodefinisati relacijom:t z = ( 1±0,2) t 0Geometriju zuba smo modelirali tako da smo zaizbranu geometrijsku varijantu snimali krivuhabanja zuba i upoređivali izdržljivost zuba prematako konstruisanoj krivoj habanja uz praćenjepopratnih pojava uzrokovanih habanjem zuba.4.1. Varijanta sa originalnim zubimaZatečeno stanje zuba na rotornom bageru je, ugeometrijskom smislu, potpuno isto kao što jeproizvođač bagera isporučio u originalnom obliku,sl.4.Slika 5. Geometrijski izgled vedrice saoriginalnim zubimaMjerenja parametra habanja h b , sl.2 ikonstruisanje krive habanja, sl.3 vršena su zanekoliko zuba na najopterećenijem mjestu navedrici - čeoni zub (rezač) pa na dijagramu, sl.6dajemo izgled krive habanja za te pozicije zuba.Konstatovano je da zub izdrži 200 ÷ 250 radnihčasova dok se pohaba u nivou od h b 40 mm.Slika 6. Kriva habanja za originalni zub4.2. Varijanta IIUčinjena je izmjena samo u domenu geometrijezuba, a kvalitet materijala od koga je zubnapravljen nije mijenjan. Predložena jegeometrijska forma na osnovu savremenih stručnihsaznanja, te iskustava drugih rudnika koji koristerotorne bagere, sl.7 dok su na sl. 8 definisaniosnovni geometrijski parametri zuba učvršćenog navedrici prema radnoj sredini.Slika 4. Originalni oblik bagerskog zubaGeometrijski izgled zuba i vedrice sarelevantnim geometrijskim parametrima dat je nasl.5Slika 7. Oblik bagerskog zuba varijante II13 th International Conference on Tribology – Serbiatrib’13 443

Nakon tih prepravki dobili smo geometrijskiizgled zuba kao na sl. 9 sa geometrijskimparametrima u zahvatu prema radnoj sredini kao nasl.10.Slika 8. Geometrijski izgled vedrice sazubima varijante IIKonstatovan je mirniji rad bagera jer jeobezbjeđeno da zubi postepeno prodiru u radnusredinu, postoji klin i u vertikalnoj i u horizontalnojravni, a radni dio zuba je znatno duži u odnosu naoriginalnu varijantu zuba. Vedrica je manjeotvorena prema bloku koji se kopa, što ima zaposljedicu manje “struganje” vedrice od radni blok.Prema instrumentima u kabini rukovaoca bagerazaključeno je da se otpori kopanju savlađuju uzmanju angažovanu snagu mašine.Takođe, konstatovan je značajan porastčasovnog kapaciteta bagera, što se objašnjavapovećanjem radne dužine zuba i manjim leđnimuglom. Prilikom ispitivanja ove konstrukcije zubasu primijećena i izvjesne negativnosti. Naime, usledojačanog i previše isturenog leđnog dijela zuba,uočeno je da se taj leđni dio previše ukopava uradnu sredinu. Drugim rječima, taj leđni dio ne ideu otkopanu masu, koju je prethodno vrh zubarazorio, nego on svojim leđnim, isturenim dijelom“gnječi” radnu sredinu. Taj razlog, uz previšedugačak radni dio zuba (L=180 mm) dovodio je dotoga da su zubi savijani, plastično deformisani ilomljeni u dijelu drške i vrata zuba. [5].Zbog ove negativnosti nije se ni pristupalokonstrukciji krive habanja za ovu vrstu zuba.4.3. Varijanta IIIPoslije uočenih nedostataka po prethodnojvarijanti, ispitivanja su obustavljena i pristupilo seispravkama na geometrijskom izgledu zuba: dužina radnog dijela skraćena je naL=165 mm, smanjeno je ojačanje leđnog dijela, zub je prema radnoj sredini otvoren na13°14’7”, izvršeno je bolje pasovanje drške zuba u“džep” i eliminisano je aksijalno pomjeranje zuba u“džepu”.Slika 9. Geometrijski izgled zuba varijante IIISlika 10. Geometrijski izgled vedrice sazubima varijante IIIOsnovna zapažanja u vezi rada bagera saovakvim zubima su: Bager je imao miran rad bez udara ivibracija sa izraženim amplitudama. Ovo seobjašnjava time što je konstrukcija zubasamim vrhom prilagođena da postepenoulazi u radnu sredinu i izaziva manje otporekopanju i manje podrhtavanje bagera.Tome je doprinijela i prilagođenost leđnogugla zuba u sprezi sa vedricom premaradnoj sredini. Časovni kapacitet bagera je veći za 15 do20% u odnosu na rad sa originalnimzubima. Proces kopanja se odvija uz značajnomanju angažovanu snagu mašine nego sazubima originalne izvedbe. Manji su otpori kopanju. Leđni dio zuba je manje isturen pa nemaneželjenog opterećenja koje je zaposljedicu imalo lomljenje zuba.444 13 th International Conference on Tribology – Serbiatrib’13

Mjereni su parametri habanja i prilikomkonstrukcije krive habanja za čeone zube jekonstatovano da je znatno duži vijektrajanja ove konstrukcije i da iznosi 300 ÷350 radnih časova, sl.11.Rezultati ovih istraživanja mogu se primjeniti savelikom pouzdanošću na kopove sa sličnim radnimuslovima.LITERATURASlika 11. Kriva habanja čeonog zuba varijante IIIBudući da je radna dužina ove konstrukcije zubaza 20 mm veća od odgovarajućeg dijela naoriginalnom zubu to se reperna dužina parametrahabanja h b od 40 mm može povećati na 50 mm, paimajući to u vidu radni vijek zuba ove konstrukcijemože iznositi i do 500 radnih časova.Ova geometrijska varijanta usvojena je, urađenagarnitura zuba za potpunu zamjenu na rotornombageru i kroz duži period eksploatacije u dobrojmjeri potvrdila pretpostavke iz eksperimentalnefaze.[1] S. Tanasijević: Osnovi tribologije mašinskihelemenata, Naučna knjiga, Beograd 1989.[2] Z. Jugović: Uticaj abrazivnog habanja zubarotornog bagera na njegovu pouzdanost, Zbornikradova, Yutrib ’91 Kragujevac 1991.[3] B. Ivković: Osnovi tribologije u industriji prerademetala, Građevinska knjiga, Beograd 1990.[4] V. Vukotić: Prilog istraživanju efektivnosti složenihenergetskih sistema sa stanovišta integralnesistemske podrške (logistike) sa posebnim osvrtomna područje I BTO sistema rudnika “Gacko”,Magistarski rad, Mašinski fakultet Mostar 1983.[5] V. Vukotić: Povećanje pouzdanosti podsistemakopanja rotornih bagera poboljšanjem tribološkihkarakteristika reznih elemenata, Doktorska disertacija,Mašinski fakultet Kragujevac, 2002.ZAKLJUČAKU sklopu provedenih istraživanja na rotormbageru ER 1250 koji je u eksploataciji u sklopurudnika „Gacko“ učinjen je pokušaj da se istiprilagodi uslovima radne sredine preko optimizacijegeometrijskih parametara reznih elemenata – zuba.Na osnovu polaznih teoretskih postavki izoblasti pouzdanosti proizvodnih sistema, tekarakteristika triboloških procesa na reznimelementima vršena su ispitivanja pouzdanostipodsistema kopanja rotornog bagera koristeći upočetku zube originalne konstrukcije isporučiocabagera a posle u nekoliko varijanti zube predloženegeometrijske izvedbe. Modeliranjem tribološkihkarakteristika zuba na rotornom bageru podešavanjemgeometrijskih parametara istih, kao što jeprikazano u radu, rezultat poboljšanja kroz predloženugeometrijsku varijantu zuba imamo: Manje otpore kopanju, Manji iznos habanja zuba, Veću eksploatacionu pouzdanostpodsistema kopanja, a u vezi s tim i samogrotornog bagera, Veći radni vijek zuba, Manje troškove održavanja bagera, Mirniji rad bagera i Veći specifični kapacitet bagera.13 th International Conference on Tribology – Serbiatrib’13 445

Serbian TribologySocietySERBIATRIB ‘1313 th International Conference onTribologyKragujevac, Serbia, 15 – 17 May 2013Faculty of Engineeringin KragujevacPONAŠANJE NEHRĐAJUĆIH ČELIKA U KOMBINIRANIMUVJETIMA TROŠENJAGoran Rozing 1 , Antun Pintarić 2 , Desimir Jovanović 3 , Vlatko Marušić 41 Elektrotehnički fakultet Osijek, Hrvatska, goran.rozing@etfos.hr2 Elektrotehnički fakultet Osijek, Hrvatska, antun.pintaric@etfos.hr3 Zastava oružje, Srbija, j.desimir@gmail.com4 Strojarski fakultet Slavonski Brod, Hrvatska, vlatko.marusic@sfsb.hrApstrakt: Glavna karika u proizvodnji sirovog ulja svakako je pužna preša, koja služi za isprešavanje icijeđenje ulja iz samljevenog i zagrijanog uljnog sjemenja. U tijeku tog procesa dolazi do odnošenja česticametala s radnih površina preše, iz čega se može zaključiti da tribosustav čine, radni dijelovi preše isuncokretovo sjemenje. Osim navedenog neizbježnog trošenja, u ovom radu analiziran je i utjecajkorozijskog trošenja dijelova pužne preše nastalog kao posljedica agresivnog djelovanja medija. Navedenatrošenja ukazuju da kod pužne preše prevladavaju kombinirani uvjeti trošenja. Na ispitnim uzorcimaizrađenim od austenitnih korozijskih postojanih čelika AISI 316L i AISI 304 provedeno je nitrokarburiranjeu cilju povećanja tvrdoće u površinskom sloju, koja dovodi do povoljnijih triboloških svojstava. Za potrebeeksperimentalnog rada provedena su ispitivanja korozijskog ponašanja uzoraka, ispitivanja kemijskogsastava osnovnog materijala, mehaničkih svojstava i analiza mikrostrukture. Zaključeno je da se mogućipristup produljenju vijeka dijelova sastoji ne samo u materijalu nego i u izboru postupka toplinske obrade,kojom će se postići povećanje tvrdoće odnosno povoljnija tribološka svojstva uz zadovoljavajuću korozijskuotpornost.Ključne reči: trošenje, pužna preša, nehrđajući čelici, nitrokarburiranje, korozijska otpornost1. UVODProblemi trenja i trošenja u praksi su vrlokompleksni zbog odvijanja mnogostrukihtriboloških procesa. Stoga je nužna brižljivaneposredna i posredna analiza svih komponenata iutjecaja u tribosustavu 1. Jedan od takvihkompleksnih primjera trošenja je i pužna preša,koja služi za isprešavanje i cijeđenje ulja izsamljevenog i zagrijanog uljnog sjemenja. Procescijeđenja jestivog ulja ovisi o mnogo parametarakoji se mijenjaju u ovisnosti o vrsti uljnogsjemenja, načinu njegove pripreme i tipu preše.Nakon djelomičnog ljuštenja i kondicioniranjavodenom parom, pripremljeno sjemenjemehaničkim putem se cijedi u pužnoj preši. Utijeku tog procesa dolazi do odnošenja česticametala s radnih površina preše, iz čega se možezaključiti da tribosustav čine, radni dijelovi preše isuncokretovo sjemenje. Uzroci trošenja radnihdijelova preše (segmenata pužnice, jarmova inoževa cjedilne korpe) su djelovanje česticamikroabraziva SiO 2 x nH 2 O u suncokretovomsjemenu [2]. Mikroabraziv sadržan u ljuscisuncokreta troši radne dijelove preše a taj proces jenemoguće izbjeći. Trošenje se manifestiraoštećivanjem napadnih bridova radnih dijelovapreše tj. smanjenja njihovih dimenzija i promjenegeometrije profila. Kao rezultat toga dolazi dosmanjenja efikasnosti cijeđenja ulja [3]. Osimspomenutog neizbježnog trošenja nakon 4,5 godine(16630 sati) rada nastupilo je oštećenje izazvanodjelovanjem agresivnog medija (kisele supare), amanifestiralo se pojavom tribokorozije nadosjednim dijelovima reduktora koje su u kontaktus dosjednim površinama cjedila. Tadašnji pristupproduljenju vijeka istrošenih dosjednih površinasastojao se od zamjene istrošenih površinapoluprstenovima izrađenim od korozijskipostojanih čelika (EN X2CrNiMo18, AISI 316L,EN) i (EN X5CrNi18-10, AISI 304) u sirovomstanju. Opisana trošenja ukazuju da kod pužne446 13 th International Conference on Tribology – Serbiatrib’13

preše prevladavaju kombinirani uvjeti trošenja kojise mogu opisati kao proces koji vodi ka degradacijimetalnih materijala, koja je rezultat mehaničkogkontakta, kombiniranog sa korozijskim djelovanjemagresivne okoline.2. ISPITIVANJA UZORAKA/DIJELOVA UPUŽNOJ PREŠI ZA ZAVRŠNO PREŠANJEIz iskustva proizilazi da je trenje i trošenjematerijala svojstvo sustava, jer na procese osimmaterijala elemenata tribosustava utječekonstrukcijska izvedba tribosustava, vrste i načinopterećenja te naprezanja, način podmazivanja idrugi čimbenici. Stoga se svaki problem morarješavati individualno, ali uzimajuči u obzirtemeljne parametre i utjecajne veličine utribosustavu 1. Nakon što su poluprsteni dvijekampanje rada pužne preše za završno prešanjekapaciteta 100t/dan bili u uporabi, premaprethodnom dogovoru, preša je rastavljena u ciljuvađenja ispitnih poluprstena. Obavljena je vizualnakontrola površina na kojima su bili ugrađeni ispitnipoluprsteni. S obzirom na to da su svi poluprsteniradili u istim uvjetima, odnosno da se vizualno neuočavaju razlike u izgledu poluprstena dosjedne izaptivne površine, ocjenjeno je da su za potrebeistraživanja u ovome radu i donošenjeodgovarajućih zaključaka dovoljni poluprsteni sreduktora i s dosjedne površine cjedila. Na slici 1prikazan je vanjski izgled poluprstena na svimpovršinama prije skidanja za potrebe ispitivanja.pregled površina svih skinutih poluprstena.Kakateristično je uočiti da se golim okom neuočavaju pojave korozijskih oštećenja, ali da suprisutni tanki slojevi praškastih taloga,najvjerojatnije mješavine korozijskih produkataosnovnog materijala kućišta/cjedila (GS-42CrMo4)i sitnih čestica mliva. Izgled površine poluprstenana vratu kućišta reduktora i polutke cjedila prešeprikazan je na slici 2.Slika 2. Poluprsten prije demontaže s praškastimtalozima nakon ispitivanja uporabom u preši tip EP 16Bitno je istaknuti da se na površinama osnovnogmaterijala kako kućišta reduktora, tako i cjedilauočava prisustvo taloga ali i da su te površineintenzivno oštećene korozijom, pri čemu se dubinaoštećenja može procjeniti na 3 do 4 mm.Dimenzionalnom kontrolom pomoću pomičnogmjerila utvrđeno je da nije došlo do smanjenjadebljine niti jednog poluprstena, na svima jeizmjerna debljina 7 mm. Detaljnim pregledomvanjskih površina svih poluprstena, promatranjempod SEM TOPO (skening elektronskimmikroskopom), utvrđeno je da se mogu uočititragovi nastali kao posljedica abrazijskog trošenjasitnim česticama mliva, zatim adhezijom uslijedkontakta poluprstena vrata reduktora spoluprstenom dosjedne površine, ali i oštećenja uformi rupica kao posljedica korozije, slika 3. Naunutarnjim površinama, koje su bile dotegnute naosnovni materijal, promatranjem pod SEM TOPOuočeno je lokalno rupičasto oštećenje, slika 4.Slika 1. Vrat kućišta reduktora i površine polutke cjedilaprije skidanja poluprstenaOdgovarajućim strugačima pažljivo su skinutiuzorci korozijskih produkata za potrebe kemijskeanalize. Utvrđeno je da u sastavu dominirakorozijski produkat željezni oksid tipa FeO/Fe 2 O 3 ,ali i da su prisutni tragovi čestica organskogporijekla iz mliva. Puno je bitniji podatak da jeanalizom kisele supare utvrđeno da se radi okiselini karbonilnog tipa (zbog prisustva isparljivihmasnih kiselina) i još važnije da izmjereni pHsupare iznosi iznosi oko 5,2. Obavljen je vizualniSlika 3. Karakteristični izgled vanjske površinepoluprstena nakon uporabe, SEM TOPO13 th International Conference on Tribology – Serbiatrib’13 447

Slika 4. Karakteristični izgled unutarnje površinepoluprstena nakon uporabe, SEM TOPOTo je oštećenje najvjerojatnije posljedicadjelovanja kombinacije kiselina karbonilnog tipa ivodene pare koje su se nakon kondenziranja„slijevale“ preko dijelova preše i ipak prodrle uzonu kontakta poluprsten/osnovni materijal, bezobzira na to što su napravom stegnute, pri čemu jesilikonska brtva trebala onemogućiti prodor supare.Izabrani nadomjesni materijali AISI 316L i AISI304 po svom kemijskom sastavu spadaju u grupuaustenitnih nehrđajućih čelika, koji imaju primjenuu prehrambenoj i procesnoj industriji. Naime,ukupno su u dva ciklusa bili u proizvodnomprocesu oko 5000 sati (kroz dvije kalendarskegodine). U odnosu na materijal kućišta GS-42CrMo4 znatno su se pokazali postojanijima. Napoluprstenima izrađenim iz varijantnih materijalaTablica 1. Rezultati ispitivanja kemijskog sastava uzoraka 316L i 304utvrđeni su isti mehanizmi trošenja, pri čemu nisuuočene značajnije razlike u intenzitetu, kakoabrazijskog i adhezijskog tako i korozijskog, bezobzira na razlike u kemijskom sastavu. Tvrdoćaugrađenog nadomjesnog materijala je iznosila oko170180 HV, što se u pogonskim uvjetimapokazalo relativno dostatno, a to potvrđuju imikroskopski snimci (slika 3 i 4) koji pokazujutragove abrazijskog trošenja, adhezije ali i rupičastekorozije. Prva dva mehanizma trošenja su prisutnijana vanjskoj strani poluprstena cjedila, dok je trećimehanizam prisutniji na unutarnjoj strani. Razlogepojave abrazijskog trošenja treba tražiti u sastavumliva točnije u sadržaju SiO 2 x nH 2 O iz ljuskesuncokreta kao glavnog nositelja abrazivnihsvojstava ljuske suncokreta. Pojavu rupičastekorozije može se pojasniti zbog prisutnosti tzv.“kisele supare“.3. EKSPERIMENTALNI DIOIspitivanja kemijskog sastava provedena su nauzorcima oba varijantna čelika u dostavnom stanju.Kemijskom analizom materijala određen je sastavprisutnih elemenata. Za određivanje kemijskogsastava korištena je spektrometrijska metoda, aispitivanja su izvršena uređajem BELEC. Rezultatiispitivanja kemijskog sastava uzoraka (materijalEN X2CrNiMo18-14-3, AISI 316L i materijal ENX5CrNi18-10, AISI 304) prikazani su u tablici 1.MaterijalKemijski sastav [%]C Mn Si Cr Ni Mo V W Ti, Fe316L 0,048 1,224 0,438 16,71 10,08 2,124 0,124 0,185 0,112 68,40304 0,045 1,295 0,651 18,02 8,11 0,414 0,114 80,19 0,007 70,68Tvrdoća i otpornost trošenju austenitnihnehrđajućih čelika može se bitno povećati, a da pritome ne dolazi do značajnog gubitka otpornosti nakoroziju. Jedan od pristupa kako bi se povećalapovršinska tvrdoća i otpornost trošenju čelika jepostupak nitriranja koji nudi visokodimenzijskustabilnost obratka 4. Nitriranje je postupakotvrdnjavanja površine difuzijom dušika upovršinske slojeve i promjena kemijskog sastavačelika [5]. Obzirom na to da na varijantnimmaterijalima poluprstena nisu nakon uporabeuočena korozijska oštećenja koja bi svojimintenzitetom bila uzročnik prestanka funkcionalnograda pužne preše, zaključeno je da bi senitrokarburiranjem varijantnih materijala moglodoprinjeti bitnom povećanju tvrdoće u površinskomsloju koja bi se odrazila na povoljnija tribološkasvojstva i produljenje vijeka tribosustava pužnihpreša za završno prešanje. Postupaknitrokarburiranja bio je sljedeći: uzorci su prvopredgrijani na temperaturu υ p =380 o C, u trajanju od3 sata i potom uronjeni u solnu kupku (volumena1m 3 ) zagrijanu na 580 o C u trajanju od 5 sati. Nakontoga uzorci su hlađeni na zraku.3.1 Ispitivanje strukture varijantnih materijalanakon nitrokarburiranjaMetalografska ispitivanja uzoraka oba varijantnamaterijala nakon nitrokarburiranja daju cjelovitusliku o njihovom mikrostrukturnom stanju. Analizamikrostrukture uzoraka obrađenih postupkomnitriranja omogućava promatranje i ruba i jezgreispitnog uzorka. Mikrostruktura rubnog dijelanitrokarburiranih uzoraka čelika AISI 316L i AISI304 prikazana je na slici 5.448 13 th International Conference on Tribology – Serbiatrib’13

a) b)Slika 5. Karakteristična mikrostruktura nitriranog čelika, povećanje 240xa) uzorak 316L, b) uzorak 3043.2 Ispitivanje mikrotvrdoća varijantnihmaterijala nakon nitrokarburiranjaIspitivanje mikrotvrdoća provedena su uređajemDURIMET Leitz metodom Vickers HV 0,025(opterećenje 0,25 N) i HV 0,5 (opterećenje 5 N), nakojemu je obavljeno i mjerenje tvrdoće osnovnihvarijantnih materijala. Rezultati ispitivanjamikrotvrdoća uzoraka oba čelika metodom VickersHV0,025 dijagramski su prikazani na slici 6a) b)Slika 6. Dijagramski prikaz toka mikrotvrdoća iodređivanja dubine nitriranog sloja uzorka čelika, a) 316L i b) 304Izmjerene vrijednosti tvrdoća poboljšanog uzorkačeličnog lijeva (materijal GS-42CrMo4) kreću seod 230 do 280 HV 0,5.3.3 Elektrokemijska korozijska ispitivanjamodificiranih površina uzorakaU eksperimentalnom dijelu ispitana suelektrokemijska svojstva nekih čelika oznaka AISI316L, AISI 304 i GS-42CrMo4 u zasićenojvodenoj otopini s CO 2 , vrijednosti pH 4,8 do 5 pritemperaturi 50°C, kako bi se simulirali stvarniuvjeti agresivne okoline u kojima se odvija trošenjeradnih dijelova preše. Uzorci za ispitivanjepripremljeni su na dimenziju Ø16x8 mm.Elektrokemijska korozijska DC ispitivanjaprovedena su sukladno normi ASTM G5-94 [6] nauređaju Potentiostat/Galvanostat Model 273AEG&E uz primjenu programa SoftCorr III uLaboratoriju za zaštitu materijala, Fakultetastrojarstva i brodogradnje u Zagrebu. Mjerenja suprovedena u odnosu na referentnu zasićenu kalomelelektrodu (ZKE) poznatog potencijala + 0,242 Vprema standardnoj vodikovoj elektrodi. Određenisu parametri opće korozije: korozijski potencijal(E cor ), gustoća korozijske struje (j cor ), brzinakorozije (v kor ), polarizacijski otpor (R p ), pitingpotencijal (E pit ) i zaštitni piting potencijal (Ezpit).Korozijski potencijal Ecor određen je mjerenjempromjene potencijala u vremenu od 1000 s.Polarizacijski otpor materijala R p je određen izTafelovog dijagrama za podruĉje polarizacije ±20mV u odnosu na korozijski potencijal. Rezultatielektrokemijskih ispitivanja prikazani prikazani suu tablici 2.13 th International Conference on Tribology – Serbiatrib’13 449

Tablica 2. Rezultati elektrokemijskih korozijskih ispitivanja uzorakaMaterijal/stanjeβ AV/dekβ KV/dekJ corA/cm 2 V cormm/godR pcm 2 AISI304, nitrirano 0,078 0,103 6,62 0,067 3282AISI316L, nitrirano 0,988 0,039 6,71 0,069 5254GS-42CrMo4, poboljšano 0,088 0,598 86,62 1,003 250Ciklička potenciodinamička polarizacijska mjerenjaprovedena su na uzorcima AISI 316L, AISI 304 iGS-42CrMo4 u zasićenoj vodenoj otopini s CO 2 ,vrijednosti pH 4,8 do 5 pri temperaturi 50°C. Naslici 7 prikazan je dijagram cikličke polarizacije, ana slici 8 makrostrukturalne snimke uzoraka.Slika 7. Dijagram cikličke polarizacije nitriranog uzorka 316L, 304 i poboljšanog uzorka GS-42CrMo4Slika 8. Makro prikaz površine nitriranog uzorka 316L, 304 i poboljšanog uzorka GS-42CrMo4nakon cikličke polarizacije4. ANALIZA REZULTATA I ZAKLJUČAKNa poluprstenima izrađenim iz varijantnihmaterijala različitog kemijskog sastava utvrđeni suisti mehanizmi trošenja, pri čemu nisu uočeneznačajnije razlike u intenzitetu, kako abrazijskog iadhezijskog tako i korozijskog trošenja. Analizomuvjeta rada pužnih preša za završno prešanjezaključeno je da se abrazivno djelovanje vrlotvrdog SiO 2 x nH 2 O (oko 1100 HV) ne možeizbjeći, ali se može smanjiti povećanjem tvrdoćedijelova preše. U tome smislu zaključeno je da bi senitriranjem varijantnih materijala moglo doprinijetibitnom povećanju produljenja vijekareprezentantnog tribosustava pužnih preša zazavršno prešanje. Na uzorcima izrađenim iz450 13 th International Conference on Tribology – Serbiatrib’13

varijantnih materijala u dostavnom stanju i unitriranom stanju te na uzorcima osnovnogmaterijala provedena su ispitivanja otpornosti naelektrokemijsku koroziju, kako bi se simuliraliuvjeti agresivne okoline u kojima se odvija raddijelova preše. Rezultati elektrokemijskihispitivanja prikazanih u tablici 2, ukazuju damaterijali AISI 316L i AISI 304 imaju dvadesetputa veće vrijednosti polarizacijskog otpora R p uodnosu na materijal GS-42CrMo4. Vrijednostbrzine korozije za austenitne korozijski postojanečelike je podjednaka i četrnaest puta je manja negoza čelični lijev GS-42CrMo4. Iz dijagrama cikličkihpolarizacija (slika 7) vidljivo je da nitrirani uzorciaustenitnih čelika 316L i 304 ne pokazuju sklonostrupičastoj koroziji niti koroziji u procjepu, dokmaterijala GS-42CrMo4 pokazuje sklonostrupičastoj koroziji i koroziji u procjepu štopotvrđuje makrostrukturalna snimka površinenakon ispitivanja. Analizom navedenih podatakamože se potvrditi da se postupkomnitrokarburiranja u solnoj kupki znatno povećavatvrdoća površine što je bio jedan od traženihzahtjeva. Istovremeno dolazi do povećanja brzinekorozije (0,067 0,069 mm/god), ali ta brzina jeznatno manja od granične vrijednosti koja iznosi0,1 mm/god, prema kriteriju primjenjivosti metala sobzirom na prosječnu brzinu prodiranja općekorozije.LITERATURA[1] T. Filetin, K. Grilec: Postupci modificiranja iprevlačenja površina, Hrvatsko društvo zamaterijale i tribologiju, Zagreb, 2004.[2] V. Ivušić, V. Marušić, K. Grilec: Abrasionresistance of surface layers, VTT Symposium 180&COST516 Tribology Symposium, VTT TechnicalResearche Centre of Finland, pp. 201-210, 1998.[3] G. Rozing, M. Katinić, V. Marušić: Nekespecifičnosti utjecaja dominatnog mehanizmatrošenja na pristup održavanju pužnih preša, 16.Međunarodno savjetovanje ODRŽAVANJE 2010Zagreb: HDO Hrvatsko-društvo održavatelja, pp.120-126, 2010.[4] B. Vasiljević, B. Nedić: Modifikovanje površina,Univerzitet u Kragujevcu, Mašinski fakultet uKragujevcu, Jugoslovensko društvo za tribologiju,Kragujevac, 2003.[5] A. Triwiyanto, P. Hussain, A. Rahman, M.C. Ismail:The Influence of Nitriding Time of AISI 316LStainless Steel on Microstructure and TribologicalProperties, Asian Journal of Scientific Research,Vol.6, pp .323-330, 2013.[6] ASTM G5 – 94: Standard Reference Test Methodfor Making Potentiostatic and PotentiodynamicAnodic Polarization Measurements.STAINLESS STEEL BEHAVIOR UNDER COMBINED CONDITIONS OFWEARAbstract: The key element in the production of raw oil is definitely the worm press, which is used forpressing and extrusion of oil from ground and heated oil seeds. During this process, metal particles areworn from the working surfaces of the press, which indicates that the tribosystem consists of working partsof the press and sunflower seeds. Next to the aforementioned unavoidable wear, damage due to aggressivemedia was also observed. Wear described above indicate that at worm press overcome combined wearconditions. On the test samples made of austenitic staineless steel AISI 316L and AISI 304 nitrocarburisingwas conducted to increase hardness in surface layer, which leads to better tribological properties. In theexperimental part of paper, there were tested corrosion behaviour of samples, chemical composition of basematerial, mechanical properties and microstructure analysis. It was concluded that possible extension of lifetime consists not only in material but also in heat treatment selection, by which increase of hardness will beachieved with reference to better tribological properties with satisfactory corrosion resistance.Keywords: wear, worm press, stainless steel, nitrocarburising, corrosion resistance13 th International Conference on Tribology – Serbiatrib’13 451

Authors IndexAAbdullah O. I. 210Adamović D. 87, 147, 261, 265, 359Adewale T. M. 160Akram W. 251Al-Shabibi A. M. 210Alaneme K. K. 160Aleksandrović S. 261, 265, 275, 281, 359Arsic D. 359Assenova E. 21BBabić M. 87, 106, 124, 129, 141,409Babić Ž. 230Babič M. 348, 351, 355Badita L-L. 331Balaz P. 55Banić M. 286, 302, 320, 388Baralić J. 217Bartz J. Wilfried 3Baqai A. A. 251Belič I. 348, 351, 355Bezjazychnyj V. F. 195Blagojević M. 230, 234Bobić B. 106, 409Bobić I. 106, 124, 129, 141, 409Bogdanović B. 147, 341Botan M. 113Bouzakis E. 10, 13Bouzakis K.-D. 10, 13, 270, 364Buciumeanu M. 80Bursuc D. C. 331CCapitanu L. 58, 331Charalampous P. 13Chen Y.S. 102Chern K.W. 102Chern S. Y. 102Chowdhury M. A. 65Crnojević C. 433ČČabrilo D. 440Čukić R. 275, 281Čupović M. 414DDeleanu L. 113, 119Doni Z. 80Drożyner P. 222Dugić G. 308Dugić M. 308Dugić P. 308ĐĐapan M. 380, 384, 396Đordjevic M. 359Djukić M. 401DžDžunić D. 75, 124, 129, 141, 261FFabian M. 55Fedrizzi L. 46, 169Florescu V. 331EEliyas A. 55Erić M. 401GGeorgescu C. 113, 119Gidikova N. 42Gligorijević B. 75Globočki Lakić G. 292, 314Grigoriev A.Ya. 4Grkić A. 240Grujović N. 92Gulzar M. 204HHacikadiroğlu H. 98Horng J.H. 102IIlić A. 226Ivanova B. 26, 31Ivanović L. 226JJanković M. 314Janković P. 217Jeremić B. 380, 384, 396Jeremić M. 147, 34113 th International Conference on Tribology – Serbiatrib’13 453

Jovanović D. 414, 446KKaleicheva J. 37Kaleli H. 98Kandeva M. 21, 26, 31, 42, 55Kao W.H. 102Karastoianov D. 31Katavić B. 75Katirtzoglou G. 13Kojić B. 308Kokol P. 348, 351, 355Kopač J. 292Kostić N. 230Kostova N.G. 55Kramar D. 292Krstić B. 275, 281LLandoulsi J. 92Lanzutti A. 46, 169Lazić V. 275, 281, 359Lekka M. 46MMačkić T. 230Mačužić I. 380, 384, 396Makrimallakis S. 13Mandić V. 261Manivasagam G. 92Manojlović J. 153, 177Mansour G. 364Marin E., 46, 169Marinković A. 135, 247Marjanović V. 234Marković S. Lj. 420Matejić M. 234Matrušić V. 414, 446Michailidis N. 10Mijajlović M. 302, 320, 388Miletić I. 226Milfelner M. 348, 351, 355Miloradović N. 129, 141Milosavljević D. 275, 281Milošević Miloš 286, 302, 388Milošević Marko 380Milović Lj. 420Miltenović A. 286, 302, 320, 388Miltenović Đ. 320Miljanić D. 341Mitrović S. 87, 106, 124, 129, 141,147, 261, 409Mitsi S. 270Mrdak M. R. 426Mufti R. A. 204Mukhopadhyay A. 373Mutavdžić M. 275Myhkin N.K. 4NNapiórkowski J. 222Nedeljković B. 281Nedić B. 75, 217, 240, 292, 314,414Nikolaevich F. R. 198Nikolić M. 302Nikolić R. 275, 281Nuruzzaman D. M. 65OÖzkan D. 98PPalaghian L. 80Pantić M. 124, 129Panjan P. 348, 351, 355Papadimitriou A. 270Paraskevopoulou R. 13Pelemis S. 87Pintarić A. 446Sreten S. 240Petkov V. 42Petrović V. 92Pirvu C. 119Pojidaeva V. 31Polzer G. 21QQasim S. A. 204RRadenković M. 380, 384Radovanović M. 314Ranđelović S. 147, 341Rezwan A. H. M. 65Richard C. 92Ristivojević M. 433Roy B. K. 65Rozing G. 446Rus D., 58SSagalovych A., 184Sagalovych V. 184Samad S. 65Sarker R. 65Schlattmann J. 210Simić A. 147, 341454 13 th International Conference on Tribology – Serbiatrib’13

Simić B. 234Skoko D. 433Skordaris G. 10Sredanović B. 292Stamenković D. 286, 302, 388Stanković M. 135, 247Stefanović M. 261, 265, 359Stojanović B. 129, 141, 226Stojiljković B. 420Sutyagin A. N. 195Szczyglak P. 222TTadić B. 341, 396Todorović P. 147, 341, 380, 384, 396Topalović M. 265Tsermaa 21Tsiafis I. 270Tzetzis D. 364VValov R. 42Vencl A. 75, 106, 135, 409Vujanac R. 234Vukotić V. 440ZZia U. A. 251ŽŽivić F. 87, 92, 124, 265Živković M. 265YYüksek L. 98WWei C.C. 10213 th International Conference on Tribology – Serbiatrib’13 455

CIP - Каталогизација у публикацијиНародна библиотека Србије, Београд621.89(082)66.017:531.43(082)539.375.6(082)INTERNATIONAL Conference on Tribology (13 ;2013 ; Kragujevac)Proceedings / 13th InternationalConference on Tribology - SERBIATRIB '13,15-17 May 2013., Kragujevac, Serbia ;[organized by] Serbian Tribology Society[and] Faculty of Engineering, University ofKragujevac ; editors Miroslav Babic, SlobodanMitrovic. - Kragujevac : Serbian TribologySociety : Faculty of Engineering, 2013(Kragujevac : Kopirnica Masinac). - [12], 455str. : ilustr. ; 24 cmRadovi na srp. i engl. jeziku. - Tekststampan dvostubacno. - Tiraz 100. - Str.[5-6]: Preface / editors. - Bibliografija uzsvaki rad. - Abstracts. - Registar.ISBN 978-86-86663-98-61. Бабић, Мирослав [уредник] [аутор додатногтекста] 2. Serbian Tribology Society(Kragujevac)a) Трибологија - Зборници b) Машинскиматеријали - Триболошке особине - Зборнициc) Хабање - Зборници d) Мазива - ЗборнициCOBISS.SR-ID 198310412

tribology in industryISSN 0354-8996VOLUME 332011.3