annual 2010/2011 of the croatian academy of engineering - HATZ
annual 2010/2011 of the croatian academy of engineering - HATZ
annual 2010/2011 of the croatian academy of engineering - HATZ
Create successful ePaper yourself
Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.
ANNUAL <strong>2010</strong>/<strong>2011</strong><br />
OF THE CROATIAN ACADEMY<br />
OF ENGINEERING
CROATIAN ACADEMY OF ENGINEERING<br />
Annu. Croat. Acad. Eng. ISSN 1332-3482<br />
ANNUAL <strong>2010</strong>/<strong>2011</strong><br />
OF THE CROATIAN ACADEMY<br />
OF ENGINEERING<br />
Editor-in-Chief<br />
Vilko Žiljak<br />
Zagreb, 2012
Published by<br />
Croatian Academy <strong>of</strong> Engineering,<br />
28 Kai St., 10000 Zagreb, Croatia<br />
Editor-in-Chief<br />
Pr<strong>of</strong>. Vilko Žiljak, Ph.D.,<br />
Vice president <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
Editorial Board<br />
Pr<strong>of</strong>. Stanko Tonkovi, Ph.D.<br />
Pr<strong>of</strong>. Vilko Žiljak, Ph.D.<br />
Pr<strong>of</strong>. emer. Zlatko Kniewald, Ph.D.<br />
ISSN 1332-3482<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
Annu. Croat. Acad. Eng.<br />
Prepress and Press<br />
INTERGRAFIKA, Zagreb<br />
Circullation<br />
300
Table <strong>of</strong> contents<br />
B. Pribievi, D. Medak, A. apo<br />
GEODETIC CONTRIBUTION TO THE GEODYNAMIC RESEARCH<br />
OF THE AREA OF THE CITY OF ZAGREB .................................................... 11<br />
Drago Katovi, Andrea Katovi, Marija Krnevi<br />
SPANISH BROOM (SPARTIUM JUNCEUM L.) .............................................. 23<br />
Matanovi Davorin, Moslavac Bojan, Nediljka Gaurina-Meimurec<br />
THE WAY TO REDUCE PIPE WEARING WHILE DRILLING ...................... 38<br />
Dinko Mikuli, Vladimir Koroman<br />
FUNDAMENTALS OF MILITARY PRODUCTION DEVELOPMENT IN<br />
THE CROATIAN WAR FOR INDEPENDENCE FROM 1991 TO 1993 .......... 46<br />
Darko Stipaniev<br />
INTELLIGENT FOREST FIRE MONITORING SYSTEM – FROM<br />
IDEA TO REALIZATION ................................................................................... 58<br />
Igor Petrovi, Davorin Kovai<br />
LABORATORY TESTING AND NUMERICAL MODELLING OF MBT<br />
WASTE DEFORMABILITY ............................................................................... 74<br />
Darko Dujmovi, Boris Androi, Ivan Lukaevi<br />
BEAM-TO-COLUMN JOINT MODELLING TO EC3 ...................................... 89<br />
Darko Dujmovi, Boris Androi, Josip Piskovic<br />
MODELLING OF JOINT BEHAVIOUR IN STEEL FRAMES ...................... 109<br />
Juraj Božievi, Marijan Andrašec<br />
DEVELOPMENT STRATEGY OF ECOINDUSTRAL PARK<br />
RAŠA - EIPR ..................................................................................................... 121<br />
Marijan Bošnjak<br />
POSIBLE MODEL OF MATHEMATICAL EVALUATION OF<br />
BEHAVIOUR DISORDER THERAPY ............................................................ 136
Mario Žagar, Ivica Crnkovi, Darko Stipaniev, Maja Štula, Juraj Feljan, Luka<br />
Lednicki, Josip Maras, Ana Petrii<br />
DICES: DISTRIBUTED COMPONENT-BASED EMBEDDED<br />
SOFTWARE SYSTEMS ................................................................................... 154<br />
Martin Žagar, Hrvoje Mlinari, Josip Knezovi<br />
FRAMEWORK FOR 3D MOTION FIELD ESTIMATION AND<br />
RECONSTRUCTION........................................................................................ 168<br />
Gaurina-Meimurec Nediljka, Pašic Borivoje, Matanovi Davorin<br />
LABORATORY EVALUATION OF MUD DIFFERENTIAL STICKING<br />
TENDENCY AND SPOTTING FLUID EFFECTIVENESS ............................ 184<br />
Ružica unko, Sanja Ercegovi Raži<br />
USE OF PLASMA TECHNOLOGY FOR MODIFICATION<br />
OF TEXTILES ................................................................................................... 199<br />
Žaneta Ugari-Hardi<br />
IMPORTANCE OF SALT CONTENT REDUCTION IN<br />
BAKERY PRODUCTS ...................................................................................... 213
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 7<br />
Croatian Academy <strong>of</strong> Engineering is sponsoring scientic meetings, symposia and<br />
conferences organized by <strong>the</strong> Departments and Centers <strong>of</strong> <strong>the</strong> Academy. Important<br />
events in <strong>the</strong> Croatian Academy <strong>of</strong> Engineering in <strong>2010</strong> and <strong>2011</strong> are:<br />
<strong>2010</strong><br />
25th Annual Assembly <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering was held on<br />
March 27th, <strong>2010</strong> where Awards <strong>of</strong> <strong>the</strong> Academy for 2009 were granted and <strong>the</strong><br />
new Statute <strong>of</strong> <strong>the</strong> Academy adopted.<br />
In accordance with <strong>the</strong> provisions <strong>of</strong> <strong>the</strong> new Statute <strong>of</strong> <strong>the</strong> Academy, a new categorization<br />
<strong>of</strong> membership in <strong>the</strong> Academy was implemented. The most important<br />
innovation is <strong>the</strong> termination <strong>of</strong> Associate Members <strong>of</strong> <strong>the</strong> Academy and <strong>the</strong> transition<br />
<strong>of</strong> Associates and Members into <strong>the</strong> unique status <strong>of</strong> a Full Member <strong>of</strong> <strong>the</strong><br />
Academy.<br />
Annual 2009 <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering with <strong>the</strong> new Who is Who<br />
in <strong>the</strong> Croatian Academy <strong>of</strong> Engineering was published.<br />
President <strong>of</strong> <strong>the</strong> Republic <strong>of</strong> Croatia Pr<strong>of</strong>. Ivo Josipovi, Ph.D. visited <strong>the</strong> Academy<br />
on September 15th, <strong>2010</strong>.<br />
In November <strong>2010</strong> <strong>the</strong> Academy submitted its comments on <strong>the</strong> draft <strong>of</strong> <strong>the</strong> Law<br />
on Science to <strong>the</strong> Ministry <strong>of</strong> Science, Education and Sports <strong>of</strong> <strong>the</strong> Republic <strong>of</strong><br />
Croatia.
8<br />
<strong>2011</strong><br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
Croatian Academy <strong>of</strong> Engineering became a partner in EU SETA - South East<br />
Transport Axis. The holder <strong>of</strong> <strong>the</strong> project activities from <strong>the</strong> part <strong>of</strong> <strong>the</strong> Academy is<br />
its Center for Trafc Engineering and distinguished experts from <strong>the</strong> Department<br />
<strong>of</strong> Transport <strong>of</strong> <strong>the</strong> Academy have taken part in <strong>the</strong> project.<br />
On January 14th, <strong>2011</strong> <strong>the</strong> Academy organized a conference entitled "Engineering<br />
Ethics and <strong>the</strong> Croatian Economy" at <strong>the</strong> Faculty <strong>of</strong> Electrical Engineering and<br />
Computing in Zagreb.<br />
In <strong>the</strong> organization <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering and <strong>the</strong> Ministry <strong>of</strong><br />
Science, Education and Sports <strong>of</strong> <strong>the</strong> Republic <strong>of</strong> Croatia, under <strong>the</strong> auspices <strong>of</strong> <strong>the</strong><br />
President <strong>of</strong> <strong>the</strong> Republic <strong>of</strong> Croatia, Central Celebration <strong>of</strong> <strong>the</strong> 300th Anniversary<br />
<strong>of</strong> <strong>the</strong> Birth <strong>of</strong> Ruer Boškovi was held at <strong>the</strong> Vatroslav Lisinski Concert Hall on<br />
May 17, <strong>2011</strong>.<br />
On <strong>the</strong> same day, after <strong>the</strong> celebration, 26th Annual Assembly <strong>of</strong> <strong>the</strong> Croatian<br />
Academy <strong>of</strong> Engineering was held upon which Academy Awards for <strong>2010</strong> were<br />
presented.<br />
Co-organized by <strong>the</strong> Biotechnical Center <strong>of</strong> <strong>the</strong> Academy, Faculty <strong>of</strong> Food Technology<br />
and Biotechnology <strong>of</strong> <strong>the</strong> University <strong>of</strong> Zagreb and <strong>the</strong> Centre for Environment<br />
Protection and Development <strong>of</strong> Sustainable Technologies <strong>of</strong> <strong>the</strong> Academy,<br />
international symposium "The life and Achievements <strong>of</strong> Pr<strong>of</strong>. Emer. Vera Johanides<br />
" was held on September 28th, <strong>2011</strong>, and her memorial bust was unveiled<br />
in front <strong>of</strong> <strong>the</strong> House <strong>of</strong> <strong>the</strong> Academy in 28 Kai St., Zagreb.<br />
Bulletin <strong>of</strong> <strong>the</strong> Academy in Croatian and English "Tehnike znanosti / Engineering<br />
Power vol. 10 (1) <strong>2011</strong> was published.<br />
2012<br />
Bulletin <strong>of</strong> <strong>the</strong> Academy in Croatian and English "Tehnike znanosti / Engineering<br />
Power vol. 10 (special issue) <strong>2011</strong> entirely devoted to Ruer Boškovi and 300th<br />
anniversary <strong>of</strong> his birth was published.<br />
In March 2012 Croatian Academy <strong>of</strong> Engineering signed Agreement on Scientic<br />
and Technical Cooperation with <strong>the</strong> Academy <strong>of</strong> Medical Sciences <strong>of</strong> Croatia,<br />
Croatian Academy <strong>of</strong> Legal Sciences and <strong>the</strong> Academy <strong>of</strong> Forestry Sciences.<br />
Organized by <strong>the</strong> Scientic Committee for Agriculture and Forestry <strong>of</strong> <strong>the</strong> Croatian<br />
Academy <strong>of</strong> Sciences and Arts and Biotechnical Center <strong>of</strong> <strong>the</strong> Croatian Academy<br />
<strong>of</strong> Engineering, a scientic conference "Food as <strong>the</strong> Foundation <strong>of</strong> Health and<br />
Longevity" was held on June 18th 2012.
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 9<br />
In September 2012 representatives <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering met<br />
with <strong>the</strong> delegation <strong>of</strong> <strong>the</strong> Chinese Academy <strong>of</strong> Engineering (CAE) in Zagreb.<br />
CAE representatives were: Pr<strong>of</strong>. Pan Yunhe, Executive Vice President <strong>of</strong> CAE,<br />
Pr<strong>of</strong>. Zheng Xiaoguang and Wang Xiaowen, Ph.D., Deputy Directors <strong>of</strong> <strong>the</strong> Department<br />
for International Cooperation <strong>of</strong> CAE, and Ms. Zhang Song, Secretary<br />
to Pr<strong>of</strong>. Pan Yunhea. On behalf <strong>of</strong> <strong>the</strong> Academy Chinese delegation was received<br />
by Pr<strong>of</strong>. V. Žiljak, Ph.D.,Vice President, and members <strong>of</strong> <strong>the</strong> Academy N. Peri,<br />
Dean <strong>of</strong> <strong>the</strong> Faculty <strong>of</strong> Electrical Engineering and Computing, and Pr<strong>of</strong>. M. Cifrek,<br />
Ph.D. The meeting was also attended by a group <strong>of</strong> scientists from <strong>the</strong> Faculty <strong>of</strong><br />
Electrical Engineering and Computing.<br />
At <strong>the</strong> conference “The Future <strong>of</strong> Printing” held in September <strong>of</strong> this year Pr<strong>of</strong>.<br />
Anayath Rejendrakumar from India gave a comprehensive overview <strong>of</strong> <strong>the</strong> state<br />
and development <strong>of</strong> graphics technology in <strong>the</strong> world. The lecture was held at <strong>the</strong><br />
Croatian Chamber <strong>of</strong> Commerce in <strong>the</strong> presence <strong>of</strong> many engineers and members<br />
<strong>of</strong> <strong>the</strong> Academy.
10<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
In November 2012 <strong>the</strong> Department <strong>of</strong> Systems and Cybernetics and <strong>the</strong> Centre for<br />
Development Studies and Projects <strong>of</strong> <strong>the</strong> Academy started "Talks about <strong>the</strong> Present<br />
and Future <strong>of</strong> Engineering and Biotechnological Sciences," as an activity <strong>of</strong> <strong>the</strong><br />
Academy members who would at <strong>the</strong>ir monthly meetings reect upon important<br />
development issues.<br />
Information about <strong>the</strong> events associated with <strong>the</strong> Academy could be found on our<br />
web site: www.hatz.hr
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 11<br />
Geodetic Contribution to <strong>the</strong> Geodynamic Research<br />
<strong>of</strong> <strong>the</strong> Area <strong>of</strong> <strong>the</strong> City <strong>of</strong> Zagreb<br />
B. Pribičević, D. Medak, and A. Đapo<br />
University <strong>of</strong> Zagreb,<br />
Faculty <strong>of</strong> Geodesy<br />
email: {bpribic, dmedak, adapo}@ge<strong>of</strong>.hr<br />
ABSTRACT<br />
Paper presents 11 year long interdisciplinary research <strong>of</strong> geodynamic processes <strong>of</strong> <strong>the</strong> area<br />
<strong>of</strong> <strong>the</strong> City <strong>of</strong> Zagreb. The Geodynamic GPS-Network <strong>of</strong> <strong>the</strong> City <strong>of</strong> Zagreb represents <strong>the</strong><br />
longest and <strong>the</strong> most intensive research effort in <strong>the</strong> eld <strong>of</strong> geodynamics in Croatia. Since<br />
<strong>the</strong> establishment <strong>of</strong> <strong>the</strong> Network in 1997, several series <strong>of</strong> precise GPS measurements<br />
have been conducted on specially stabilized points <strong>of</strong> Geodynamical Network <strong>of</strong> City <strong>of</strong><br />
Zagreb with purpose <strong>of</strong> investigation <strong>of</strong> tectonic movements and related seismic activity <strong>of</strong><br />
<strong>the</strong> wider area <strong>of</strong> <strong>the</strong> City <strong>of</strong> Zagreb. The Network has been densied in 2005 in <strong>the</strong> most<br />
active region <strong>of</strong> nor<strong>the</strong>astern Mount Medvednica. Since <strong>the</strong>n, several GPS campaigns have<br />
been conducted. Processing <strong>of</strong> observation data was done with scientic s<strong>of</strong>tware GAMIT/<br />
GLOBK, developed by MIT. From this series <strong>of</strong> GPS measurements geodetic model <strong>of</strong><br />
tectonic movements has been created. In <strong>the</strong> area <strong>of</strong> interest, independent geological investigations<br />
have been conducted through even longer period <strong>of</strong> time which resulted in<br />
geological model <strong>of</strong> tectonic movements. The correlation coefcient between geodetic and<br />
geologic model has been calculated and shows high degree <strong>of</strong> correlation thus giving credibility<br />
to both methods <strong>of</strong> research. Systematic analysis has been conducted over geodetic<br />
and geologic results giving as a result unique interdisciplinary model <strong>of</strong> crust movements<br />
over wider Zagreb area. This interdisciplinary interpretation <strong>of</strong> obtained geodetic movements<br />
leads to a new scientic insight about geodynamics <strong>of</strong> <strong>the</strong> City <strong>of</strong> Zagreb area. The<br />
results <strong>of</strong> this scientic research will be used to delineate zones <strong>of</strong> potential earthquake<br />
hazard or tectonically caused landslides.<br />
Keywords: geodesy, GPS, geodynamics, GAMIT, tectonics<br />
Introduction<br />
It has been known for a long time that <strong>the</strong> wider area <strong>of</strong> Zagreb is geodynamically active,<br />
(Prelogovi, Cvijanovi 1981) (Kuk et al. 2000), (Kuk et al. 2000a), (Tomljenovi, 2002)<br />
however, <strong>the</strong>re were no data on <strong>the</strong> actual size <strong>of</strong> those movements, nor on <strong>the</strong>ir spatial orientation.<br />
The rst geodetic research in this direction was made through <strong>the</strong> implementation<br />
<strong>of</strong> <strong>the</strong> project <strong>the</strong> “Basic GPS network <strong>of</strong> <strong>the</strong> City <strong>of</strong> Zagreb” (oli et al. 1999) (Medak,<br />
Pribievi 2001).<br />
Through <strong>the</strong> realization <strong>of</strong> <strong>the</strong> project “Basic GPS-Network <strong>of</strong> <strong>the</strong> City <strong>of</strong> Zagreb” in 1997,<br />
Croatian capital got a modern, geodetic foundation <strong>of</strong> high accuracy. Network was planned<br />
as <strong>the</strong> basis for investigations <strong>of</strong> tectonic movements and related seismic activity <strong>of</strong> <strong>the</strong><br />
wider area <strong>of</strong> <strong>the</strong> City <strong>of</strong> Zagreb. Basic part <strong>of</strong> <strong>the</strong> network consists <strong>of</strong> 43 specially stabilized<br />
geodetic points to meet <strong>the</strong> specic criteria for geodynamic points. After <strong>the</strong> second<br />
series <strong>of</strong> GPS measurements in year 2001, <strong>the</strong> network has become “Geodynamic Network<br />
<strong>of</strong> <strong>the</strong> City <strong>of</strong> Zagreb” (Medak, Pribievi 2001), (Medak, Pribievi 2002), (Medak,
12<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
Pribievi 2003). The City <strong>of</strong> Zagreb has recognized <strong>the</strong> importance <strong>of</strong> this project and<br />
GPS campaigns have been performed in years 2004, 2006, 2007 and in 2008 (Pribievi et<br />
al 2004), (Medak et al 2007a).<br />
After six series <strong>of</strong> GPS-measurements in period from 1997 to 2008, <strong>the</strong> analysis <strong>of</strong> <strong>the</strong> results<br />
with scientic s<strong>of</strong>tware GAMIT/GLOBK show signicant movement on GPS points<br />
as a result <strong>of</strong> geodynamic activity in <strong>the</strong> research area (Medak, Pribievi 2004), (Medak,<br />
Pribievi 2006), (Pribievi et al. 2007), (apo 2005), (apo 2009). From <strong>the</strong> analysis results<br />
<strong>the</strong> geodetic model <strong>of</strong> tectonic movements has been created and scientic comparison<br />
with geologic model created on <strong>the</strong> basis <strong>of</strong> age-long research. The correlation coefcient<br />
between geodetic and geologic model has been calculated and shows high degree <strong>of</strong> correlation<br />
thus giving credibility to both methods <strong>of</strong> research. Systematic analysis has been<br />
conducted over geodetic and geologic results giving as a result unique interdisciplinary<br />
model <strong>of</strong> crust movements over wider Zagreb area (apo et al. 2009), (apo 2009).<br />
Geodynamic Network <strong>of</strong> <strong>the</strong> City <strong>of</strong> Zagreb<br />
Underlying network has already in its preliminary design devised a dual role: rst it was<br />
<strong>the</strong> basis for <strong>the</strong> establishment <strong>of</strong> homogeneous GPS network <strong>of</strong> <strong>the</strong> City <strong>of</strong> Zagreb, while<br />
since 2001 it has became <strong>the</strong> “Geodynamic Network <strong>of</strong> <strong>the</strong> City <strong>of</strong> Zagreb. Such role was<br />
possible primarily due to development <strong>of</strong> GPS technology and achieving high accuracy,<br />
which made possible measurement <strong>of</strong> geodynamic movements (Blewitt 1993), (Bock et al.<br />
1993), (Segall, Davis 1997). This fact now allows geodesy to actively participate toge<strong>the</strong>r<br />
with o<strong>the</strong>r pr<strong>of</strong>essions and disciplines in <strong>the</strong> important interdisciplinary research for <strong>the</strong><br />
City <strong>of</strong> Zagreb.<br />
The aim <strong>of</strong> such a special geodetic network is determination <strong>of</strong> actual geodynamic movements<br />
over a longer period <strong>of</strong> time with very high accuracy. This research represents a<br />
qualitative contribution to <strong>the</strong> geodetic pr<strong>of</strong>ession in a variety <strong>of</strong> interdisciplinary research,<br />
such as monitoring <strong>of</strong> earthquake prediction indicators in Zagreb and its surroundings.<br />
Thus, modern geodesy with its high accuracy enters in an entirely new area <strong>of</strong> research<br />
with purpose <strong>of</strong> determining new parameters microseismic zoning and landslide monitoring<br />
in <strong>the</strong> area <strong>of</strong> <strong>the</strong> City <strong>of</strong> Zagreb. Today, in this type <strong>of</strong> research, toge<strong>the</strong>r with geologists,<br />
seismologists, geotechnicists, geophysicists and o<strong>the</strong>r scientists, <strong>the</strong>ir signicant<br />
contribution may be given by <strong>the</strong> geodesists (Altiner 1999.), (Altiner et al. 2001), (Medak<br />
et al. 2002), (Grenercy, 2005), (Pinter et al. 2004).<br />
It goes without saying that this segment <strong>of</strong> <strong>the</strong> research is <strong>of</strong> great importance because <strong>the</strong><br />
City <strong>of</strong> Zagreb has <strong>the</strong> highest concentration <strong>of</strong> population and industry in <strong>the</strong> Republic <strong>of</strong><br />
Croatia. Therefore, it is certain that even now <strong>the</strong> city administration should think about<br />
measures to be taken in order to preserve human life and property in case <strong>of</strong> natural disasters<br />
such as earthquakes, landslides and similar. Figures 1-7 show cracks and damages on<br />
<strong>the</strong> objects in <strong>the</strong> Zagreb area due to tectonic movements.
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 13<br />
Figure 1: a, b) Cracks in <strong>the</strong> wall <strong>of</strong> <strong>the</strong> church <strong>of</strong> St. Marie in Granešina with marked direction<br />
<strong>of</strong> <strong>the</strong> sliding – size <strong>of</strong> <strong>the</strong> cracks is between 5-10 cm (c) Example o <strong>the</strong> cracks parallel to <strong>the</strong><br />
orientation <strong>of</strong> local compression stress above south church window; (d) same crack extended to<br />
<strong>the</strong> road between <strong>the</strong> church and <strong>the</strong> rectory (Medak et al 2007b)<br />
Figure 2. Gornja Planina, Podizeljska Street 12 – a) crack on <strong>the</strong> concrete oor parallel to <strong>the</strong><br />
fault from zone Stubica-Kašina b) In <strong>the</strong> house yard a concrete curb with a crack <strong>of</strong> 1.5 cm in<br />
size, created in just one year. (apo 2005)<br />
Figure 3. Example <strong>of</strong> cracks due to movements in <strong>the</strong> zone <strong>of</strong> Zagreb fault - Vidovec – cracks 4<br />
cm wide (apo 2009)
14<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
Figure 4. a) Donja Planina, Planinarska Street 21. Whole house has been cut with <strong>the</strong> crack <strong>of</strong><br />
NNW-SSE direction. On <strong>the</strong> facade a signicant horizontal right shift <strong>of</strong> 2 cm is notable. The<br />
facade was made two years prior; b). Donja Planina 13- south facade – cracks with horizontal<br />
right movement with size 3cm (apo 2009).<br />
Figure 5. During 2006 year in <strong>the</strong> period <strong>of</strong> a relatively larger displacement amplitudes measured<br />
at points Geodynamic network, in fault zone Stubica-Kašina landslides were activated. Particularly<br />
activated was <strong>the</strong> well known landslide in <strong>the</strong> Planina Gornja. Figure shows <strong>the</strong> cracking <strong>of</strong><br />
<strong>the</strong> road (apo 2009).<br />
Figure 6. a) In Gornji Mikulii on <strong>the</strong> route <strong>of</strong> main fault <strong>of</strong> <strong>the</strong> Zagreb zone <strong>the</strong> landslide has<br />
been reactivated in 2008. b) Particularly noticeable are shifts <strong>of</strong> <strong>the</strong> ground north <strong>of</strong> <strong>the</strong> road,<br />
where considerable damage to fences has been recorded (apo 2009).
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 15<br />
Figure 7: In <strong>the</strong> spring <strong>of</strong> 2008 a new landslide in Vidovec was created. Landslide is located<br />
within <strong>the</strong> route <strong>of</strong> <strong>the</strong> main faults <strong>of</strong> <strong>the</strong> Zagreb fault zone. In <strong>the</strong> overlying wing <strong>of</strong> <strong>the</strong> cliffs<br />
<strong>the</strong>re is a fault in <strong>the</strong> relief. Due to tectonic activity <strong>the</strong>re was an escarpment destabilization and<br />
<strong>the</strong> activation <strong>of</strong> landslides (apo 2009).<br />
Design and <strong>the</strong> stabilization <strong>of</strong> <strong>the</strong> Geodynamic Network<br />
Given <strong>the</strong> size <strong>of</strong> <strong>the</strong> City <strong>of</strong> Zagreb and <strong>the</strong> need to encompass <strong>the</strong> wider area, <strong>the</strong> Geodynamic<br />
network covers an area <strong>of</strong> app. 700 km 2 . The main characteristic <strong>of</strong> this mostly raster<br />
formatted network is that <strong>the</strong> distance between points is approximately seven kilometers in<br />
a sparsely populated area, and in urban areas <strong>of</strong> <strong>the</strong> City, density is slightly higher. Namely,<br />
given <strong>the</strong> fact that <strong>the</strong> City <strong>of</strong> Zagreb is seismically very active, <strong>the</strong> Network is designed so<br />
that <strong>the</strong> recent structural fabric <strong>of</strong> <strong>the</strong> Zagreb area is optimally covered. Network points are<br />
placed in relation to fault zones to positions that will show <strong>the</strong> truest geodynamic movements<br />
in <strong>the</strong> research area. Figure 8 shows <strong>the</strong> placement <strong>of</strong> <strong>the</strong> points in <strong>the</strong> Zagreb area<br />
and positions <strong>of</strong> main faults and epicenters <strong>of</strong> largest earthquakes in <strong>the</strong> area.<br />
Figure 8. Locations <strong>of</strong> <strong>the</strong> Geodynamic network points and positions <strong>of</strong> main faults and epicenters<br />
<strong>of</strong> largest earthquakes in <strong>the</strong> area (apo 2005)
16<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
Thanks to <strong>the</strong> signicant increase in accuracy, dependence <strong>of</strong> geodetic data on time (<strong>the</strong><br />
fourth dimension) is apparent. That is why special attention has been paid to <strong>the</strong> stabilization<br />
<strong>of</strong> <strong>the</strong> points <strong>of</strong> <strong>the</strong> geodynamic network. The aim was to ensure <strong>the</strong> stability <strong>of</strong> points<br />
over a longer period <strong>of</strong> time and provide accurate reoccupation <strong>of</strong> GPS antenna. In this way<br />
determination <strong>of</strong> actual geodynamic displacements was enabled with very high accuracy<br />
over a longer period (Solari, 1999), (Schmitt, 1985), (Gerasimenko et al. 2000).<br />
Works on <strong>the</strong> stabilization <strong>of</strong> points <strong>of</strong> <strong>the</strong> Geodynamic network <strong>of</strong> <strong>the</strong> City <strong>of</strong> Zagreb were<br />
performed according to <strong>the</strong> design created by <strong>the</strong> experts from <strong>the</strong> University <strong>of</strong> Zagreb,<br />
Faculty <strong>of</strong> Geodesy, with <strong>the</strong> interdisciplinary help <strong>of</strong> engineers, geologists and seismologists.<br />
The Geodynamic network consists <strong>of</strong> more than 40 specially stabilized points on wider<br />
area <strong>of</strong> <strong>the</strong> City <strong>of</strong> Zagreb and covers area <strong>of</strong> about 700km2 (oli et al 1999), (Medak,<br />
Pribievi 2002).<br />
In total <strong>the</strong>re were 33 pillars built, 32 <strong>of</strong> which come with <strong>the</strong> above ground stabilization,<br />
and only one with <strong>the</strong> sub terrain stabilization (1028 King Tomislav Square). The remaining<br />
7 points <strong>of</strong> <strong>the</strong> Geodynamic network has a different stabilization, which also meets all<br />
<strong>the</strong> stability requirements that are set before this special network (Medak, Pribievi 2002)<br />
(Pribievi et al 2007).<br />
As shown on <strong>the</strong> gure 1, most <strong>of</strong> <strong>the</strong> monuments have pilots that go up to 14 meters deep<br />
to consolidated ground. On <strong>the</strong> top <strong>of</strong> every monument is a steel mark with winding for <strong>the</strong><br />
special extension thus insuring stability <strong>of</strong> points through long periods <strong>of</strong> time and precise<br />
GPS antenna reoccupation.<br />
Figure 9. Specially stabilized monument on Geodynamic network <strong>of</strong> <strong>the</strong> City <strong>of</strong> Zagreb<br />
After seven years <strong>of</strong> intensive research, it became clear that <strong>the</strong> Geodynamic network must<br />
be densied in <strong>the</strong> seismotectonically most active area. Densifying is <strong>of</strong> particular importance<br />
for <strong>the</strong> preliminary assessment <strong>of</strong> <strong>the</strong> landslide susceptibility in those urban areas<br />
that lie on <strong>the</strong> clay layers in <strong>the</strong> nor<strong>the</strong>rn part <strong>of</strong> <strong>the</strong> zone: Šestine Grmošica, Granešina,<br />
Mikulii, Vugrovec and Kašina. In year 2005, ve new geodynamic points were established<br />
in that part <strong>of</strong> Zagreb (Medak et al. 2007a).
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 17<br />
GPS-observations<br />
Since 1997 GPS measurements on <strong>the</strong> Geodynamic network <strong>of</strong> <strong>the</strong> City <strong>of</strong> Zagreb were<br />
conducted through GPS measurement campaigns every few years. Each campaign consists<br />
<strong>of</strong> two to three 24 hour sessions. First two campaigns in 1997 and 2001 had tree 24 hour<br />
sessions with 15 second observation interval, thus giving 5670 epochs. All later campaigns<br />
had 24 hour sessions but with 30 second interval, giving 2880 epochs (Medak et al 2007a)<br />
(apo 2009).<br />
Figure 10. Distribution <strong>of</strong> points according to <strong>the</strong> observation sessions (2004, 2006, 2008)<br />
All conducted campaigns had only Trimble GPS receivers and antennas. There were 8<br />
conducted GPS campaigns: 1997, 2001, 2003, 2004, 2005, 2006, 2007 and 2008. Only<br />
ve <strong>of</strong> those were observation <strong>of</strong> complete Geodynamic network: 1997, 2001, 2004, 2006,<br />
and 2008 (41 points). Two campaigns were conducted for observing densication points<br />
2005, 2007 (11 and 21 points) (Pribievi et al. 2007) (apo et al. 2009). Table 1 shows all<br />
conducted campaigns including number <strong>of</strong> points and used instruments<br />
Table 1: GPS campaigns on Geodynamic network <strong>of</strong> <strong>the</strong> City <strong>of</strong> Zagreb<br />
Campaign Date sessions points receivers<br />
Zagreb 1997 27.-29.10.1997. 2 43 27<br />
Zagreb 2001 25.-28.06.2001. 3 40 16<br />
Zagreb 2003 22.-23.06.2003. 1 13 13<br />
Zagreb 2004 17.-20.06.2004. 3 39 13<br />
Zagreb 2005 10.-11.09.2005. 1 11 11<br />
Zagreb 2006 22.-25.06.2006. 3 41 13<br />
Zagreb 2007 13.-15.07.2007. 2 21 13<br />
Zagreb 2008 10.-13.06.2008. 3 41 13
18<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
Figure 11: Velocity rates and <strong>the</strong>ir error ellipses for period 2006-2007 (apo 2009), (Wessel 2004)<br />
Table 2: Statistical representation <strong>of</strong> absolute values <strong>of</strong> velocity rates for 2006–2007 in mm/yr.<br />
v<br />
v<br />
v hz<br />
mm/yr<br />
v H<br />
mm/yr<br />
min. 0,2 0,57 1,43 0,60<br />
max. 10,70 19,36 19,37 50,27<br />
avg. 3,14 4,86 6,43 16,25<br />
Velocity rates calculated for <strong>the</strong> complete period from 1997 to 2008 are much smaller due<br />
to different active zones <strong>of</strong> <strong>the</strong> Zagreb area from year to year giving thus sometimes opposite<br />
velocity directions for <strong>the</strong> same points. Velocity rates for <strong>the</strong> period 1997 to 2008<br />
are shown in Figure 12 and Table 3 shows statistical representation <strong>of</strong> absolute values <strong>of</strong><br />
velocity rates for period 1997–2008 in mm/yr.<br />
Table 3: Statistical representation <strong>of</strong> absolute values <strong>of</strong> velocity rates for 1997–2008 in mm/yr.<br />
v v<br />
v hz<br />
v H<br />
mm/yr mm/yr<br />
min. 0,03 0,04 0,12 0,02<br />
max. 3,93 3,42 4,34 17,48<br />
avg. 0,83 1,03 1,45 1,97
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 19<br />
Figure 13: Geodetic velocity model created using IDW interpolation<br />
with respect to faults model<br />
(1)<br />
Table 4: Correlation coefcient by Spearman formula<br />
Calculating <strong>of</strong> correlation coefcient by above given formula (1) for Spearman rank correlation<br />
we get rs=0.94 with 1% signicance level (or 99% in Gaussian model).<br />
The correlation coefcient between geodetic and geologic model shows high degree <strong>of</strong> correlation<br />
thus giving credibility to both methods <strong>of</strong> research (apo 2009).
20<br />
CONCLUSION<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
Geodynamic study <strong>of</strong> <strong>the</strong> City <strong>of</strong> Zagreb has started in 1997 and since 2008 <strong>the</strong>re has<br />
been seven GPS campaigns for <strong>the</strong> purpose <strong>of</strong> determining geodynamic movements on <strong>the</strong><br />
points <strong>of</strong> <strong>the</strong> Geodynamic network <strong>of</strong> <strong>the</strong> City <strong>of</strong> Zagreb.<br />
Through <strong>the</strong> project Geodynamic study <strong>of</strong> <strong>the</strong> City <strong>of</strong> Zagreb, based on independent multiple<br />
precise geodetic and geological measurements, original models <strong>of</strong> tectonic movements<br />
on Zagreb area have been created. Geodetic model is based on precise GPS satellite<br />
measurements, and geological model is based on <strong>the</strong> long-term geologic measurements<br />
and studies. For this purpose data collected through GPS measurements on seven GPS<br />
measurement campaigns in <strong>the</strong> period since 1997 until 2008. Year has been used. GPS<br />
measurements were processed in <strong>the</strong> scientic GAMIT/GLOBK s<strong>of</strong>tware, designed at MIT<br />
specically for processing GPS measurements on <strong>the</strong> geodynamic network. It calculated<br />
<strong>the</strong> geodynamic movements on <strong>the</strong> points <strong>of</strong> <strong>the</strong> Geodynamic network using modern methods<br />
<strong>of</strong> Kalman ltering.<br />
The result <strong>of</strong> <strong>the</strong> velocity models at <strong>the</strong> points Geodynamic network were obtained for <strong>the</strong><br />
period 1997- 2001, 2001-2004, 2004-2006, 2006-2007, 2007-2008, 2006-2007-2008 and<br />
cumulatively for <strong>the</strong> whole period from 1997-2008. The maximum absolute value <strong>of</strong> displacement<br />
for <strong>the</strong> total solution in <strong>the</strong> horizontal direction is 4.3mm/yr in vertical direction<br />
17.5mm/yr. The largest absolute values <strong>of</strong> displacements were obtained during <strong>the</strong> period<br />
2006-2007 and are amounted to 19.4mm/yr and 50.3mm/yr in <strong>the</strong> horizontal and vertical<br />
direction respectively. It is evident that it represents signicant geodynamic movements.<br />
However, it should be noted that <strong>the</strong> mean value <strong>of</strong> velocity is 3 mm/yr.<br />
Velocity model <strong>of</strong> geodynamic movements for <strong>the</strong> whole period <strong>of</strong> <strong>the</strong> study has been created<br />
using IDW interpolation method with <strong>the</strong> inclusion <strong>of</strong> fault models <strong>of</strong> <strong>the</strong> Zagreb area.<br />
In that way <strong>the</strong> original geodetic model <strong>of</strong> <strong>the</strong> velocity eld for <strong>the</strong> wider area <strong>of</strong> <strong>the</strong> City<br />
<strong>of</strong> Zagreb.<br />
Geostatistical analysis has been performed and <strong>the</strong> correlation coefcient has been calculated<br />
for <strong>the</strong> geodetic and geological model <strong>of</strong> <strong>the</strong> geodynamic movements, by using<br />
Spearman correlation coefcient formula. The resulting value <strong>of</strong> 0.94 with signicance<br />
level <strong>of</strong> 1% (or 99% in Gaussian model) indicates a high degree <strong>of</strong> correlation between<br />
<strong>the</strong> geodetic and geological model <strong>of</strong> <strong>the</strong> geodynamic movements in <strong>the</strong> subject area. This<br />
proves <strong>the</strong> credibility <strong>of</strong> eleven year long research independently conducted by geological<br />
and geodetic methods.<br />
In conclusion, it can be stated that as part <strong>of</strong> this research a unique interdisciplinary model<br />
<strong>of</strong> crust movements over wider Zagreb area has been created for <strong>the</strong> rst time. It can be<br />
applied to precisely dene <strong>the</strong> boundaries <strong>of</strong> seismic micro-zoning <strong>of</strong> <strong>the</strong> City <strong>of</strong> Zagreb.<br />
On <strong>the</strong> o<strong>the</strong>r hand, results <strong>of</strong> <strong>the</strong> study can help in decision making process about reconstruction<br />
and adaptation <strong>of</strong> important structures (reinforcement <strong>of</strong> foundations and structural<br />
elements) and also <strong>the</strong>y can be used to more accurately dene <strong>the</strong> zones <strong>of</strong> landslides<br />
caused by tectonic movements.<br />
REFERENCES<br />
1. Altiner, Y. (1999). Analytical Surface Deformation Theory for Detection <strong>of</strong> <strong>the</strong> Earth’s<br />
Crust Movements. Springer Verlag.<br />
2. Altiner, Y., Marjanovi-Kavanagh, R., Medak, D., Medi, Z., Prelogovi, E., Pribievi,<br />
B., et al. (2001). Is Adria a Promontory or does it exist as an Independent Microplate?<br />
In J. Sledzinski (Ed.), Proceedings <strong>of</strong> <strong>the</strong> EDS G9 Symposium “Geodetic and Geodynamic<br />
Programmes <strong>of</strong> <strong>the</strong> CEI (Central European Initiative), 25-30 March 2001.
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 21<br />
Reports on Geodesy (p. 225-229). Nice, France: Warsaw University <strong>of</strong> Technology.<br />
3. Blewitt G. (1993): Advances in Global Positioning System technology for geodynamics<br />
investigations: 1978–1992. In Contrib. Space Geodesy Geodyn.: Technol. Geodyn.,<br />
Geodyn. Ser., ed. DE Smith, DL Turcotte, 25: 195–213. Washington, DC: Am.<br />
Geophys. Union. 213 pp.<br />
4. Bock Y., D., Agnew, Fang, P., Genrich, J., Hager, B., Herring, T., Hudnut, K., et al.<br />
(1993). Detection <strong>of</strong> crustal deformation from <strong>the</strong> Landers earthquake sequence using<br />
continuous geodetic measurements. Nature, 361 , 337-340<br />
5. oli, K., Prelogovi, E., Pribievi, B., Švehla, D. (1999). Hrvatski geodinamiki<br />
projekt CRODYN i GPS mreža Grada Zagreba. In A. Baji (Ed.), Znanstveni skup<br />
Andrija Mohorovii - 140. obljetnica roenja: zbornik radova (p. 141-152). Zagreb:<br />
Državni hidrometeorološki zavod.<br />
6. Dong, D., Herring, T., & King, R. (1998). Estimating regional deformation from a<br />
combination <strong>of</strong> space and terrestrial geodetic data. Journal <strong>of</strong> Geodesy, 72 (4), 200-<br />
214.<br />
7. apo, A. (2005). Obrada i interpretacija geodetskih mjerenja na geodinamikoj mreži<br />
Grada Zagreba. magistarski rad. Zagreb: Geodetski fakultet, 12.05. 2005., 174 str.<br />
Voditelj: Medak, Damir.<br />
8. apo, A (2009): Korelacija geodetskog i geološkog modela tektonskih pomaka na<br />
primjeru šireg podruja Grada Zagreba. doktorska disertacija. Zagreb: Geodetski<br />
fakultet, 08.05. 2009, 199 str. Voditelj: Pribievi, Boško.<br />
9. apo, Almin; Pribievi, Boško; Medak, Damir; Prelogovi, Eduard (2009): Correlation<br />
between Geodetic and Geological Models in <strong>the</strong> Geodynamic Network <strong>of</strong> <strong>the</strong> City<br />
Of Zagreb. Reports on geodesy. 86 (2009), 1; p 115-122.<br />
10. Grenerczy, G., Sella, G., Stein, S., Kenyeres, A. (2005). Tectonic implications <strong>of</strong> <strong>the</strong><br />
GPS velocity eld in <strong>the</strong> nor<strong>the</strong>rn Adriatic region. Geophys. Res. Lett., 32 .<br />
11. Gerasimenko, M. D., Shestakov, N. V., Teruyuki, K. (2000). On optimal geodetic network<br />
design for fault-mechanics studies (Vol. 52). Earth Planets Space.<br />
12. Herring, T., King, R., McClusky, S. (2006a). Documentation for <strong>the</strong> MIT Global Kalman<br />
lter VLBI and GPS analysis program: GLOBK 10.3. Cambridge.<br />
13. Herring, T., King, R., McClusky, S. (2006b). Documentation for <strong>the</strong> MIT GPS analysis<br />
s<strong>of</strong>tware: GAMIT 10.3. Cambridge.<br />
14. Kuk, V., Prelogovi, E., Dragievi, I. (2000). Seismotectonically Active Zones in <strong>the</strong><br />
Dinarides. Geol. Croatica, 53 (2), 295-303.<br />
15. Kuk, V., Prelogovi, E., Sovi, L., Kuk, K., Mari, K. (2000a). Seizmološke i seizmotektonske<br />
znaajke šireg zagrebakog podruja. Graevinar, 52 (11), 647-653.<br />
16. Medak, D., Pribievi, B. (2001). Geodynamic GPS-Network <strong>of</strong> <strong>the</strong> City <strong>of</strong> Zagreb<br />
- First Results. In The Stephan Mueller topical conference <strong>of</strong> <strong>the</strong> European Geophysical<br />
Society: Quantitative neotectonic and seismic hazard assessment: new integrated<br />
approaches for environmental management. Balatonfüred, Hungary.<br />
17. Medak, D., Pribievi, B. (2002). Geodinamika mreža Grada Zagreba. In T. Baši<br />
(Ed.), Zbornik Geodetskog fakulteta Sveuilišta u Zagrebu povodom 40. obljetnice<br />
samostalnog djelovanja 1962-2002. pp. 145-156. Zagreb.<br />
18. Medak, D., Pribievi, B., Prelogovi, E. (2002). Determination <strong>of</strong> <strong>the</strong> recent structural<br />
fabric in <strong>the</strong> Alps-Dinarides area by combination <strong>of</strong> geodetic and geologic methods.<br />
In M. Brilly (Ed.), Raziskave s podroja geodezije in geozike. p. 57-64. Ljubljana:<br />
Slovensko združenje za geodezijo in geoziko.<br />
19. Medak, D., Pribievi, B. (2003). Processing <strong>of</strong> Geodynamic GPSNetworks with<br />
GAMIT S<strong>of</strong>tware. Reports on Geodesy, Warsaw University <strong>of</strong> Technology, 64 (1),<br />
75-84.
22<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
20. Medak, D., Pribievi, B. (2004). Processing <strong>of</strong> Geodynamic GPS networks in Croatia<br />
with GAMIT S<strong>of</strong>tware. In N. Pinter, G. Grenerczy, J. Webber, S. Stein, & D. Medak<br />
(Eds.), The Adria Microplate: GPS Geodesy, Tectonics and Hazards (Vol. 61, p. 247-<br />
256). Veszprem, Hungary: Springer.<br />
21. Medak, D., Pribievi, B. (2006): Processing <strong>of</strong> geodynamic GPS-networks in Croatia<br />
with GAMIT s<strong>of</strong>tware, The Adria Microplate, GPS Geodesy, Tectonics and Hazards /<br />
Pinter, Nicholas et al. (eds).<br />
22. Medak, D., Pribievi, B., apo, A. (2007a). Progušenje toaka Geodinamike mreže<br />
Grada Zagreba u podsljemenskoj zoni. Geodetski list , 61(84)(4), 247-258.<br />
23. Medak, D., Pribievi, B., Prelogovi, E., apo, A. (2007b). Primjene geodetskogeodinamikih<br />
GPS-mjerenja za monitoring tektonski uvjetovanih klizišta. In Simpozij<br />
o inženjerskoj geodeziji (p. 229-241). Beli Manastir.<br />
24. Pinter, N, Grenerczy, G., Webber, J., Stein, S., Medak, D. (Eds.) (2004), The Adria<br />
Microplate: GPS Geodesy, Tectonics and Hazards .Vol. 61. Veszprem, Hungary:<br />
Springer.<br />
25. Prelogovi, E., Cvijanovi, D. (1981). Potres u Medvednici 1880. godine. Geol. vjesnik<br />
(34), 137-146.<br />
26. Pribievi, B., Medak, D., Prelogovi, E. (2004). Geodinamika prostora Grada Zagreba.<br />
Geodetski list , 58(81)(1), 51-65.<br />
27. Pribievi, B., Medak, D., Prelogovi, E., apo, A. (2007). Geodinamika prostora<br />
Grada Zagreba. Zagreb: Geodetski fakultet Sveuilište u Zagrebu. Znanstvena monograja.<br />
28. Schmitt, G. (1985). Review <strong>of</strong> network design: Criteria, risk functions, design ordering.<br />
In Grafarend & F. Sanco (Eds.), Optimization and design <strong>of</strong> geodetic network (p.<br />
610). Berlin etc.: Springer.<br />
29. Segall, P., Davis, L. J. (1997): GPS Applications for Geodynamics and Earthquake<br />
Studies. Annual Review <strong>of</strong> Earth and Planetary Sciences. Vol. 25: 301-336. May 1997.<br />
30. Solari, M. (1999). Suradnja srednje europskih zemalja u geodeziji i geodinamici.<br />
In A. Baji (Ed.), Znanstveni skup Andrija Mohorovii - 140. obljetnica roenja:<br />
zbornik radova (p. 165-177). Zagreb: Državni hidrometeorološki zavod.<br />
31. Šoši, I. (2004). Primijenjena statistika. Zagreb: Školska knjiga, Udžbenici Sveuilišta<br />
u Zagrebu.<br />
32. Tomljenovi, B. (2002). Strukturne znaajke Medvednice i Samoborskog gorja. Doktorska<br />
disertacija. Rudarsko-geološko-naftni fakultet, Zagreb.<br />
33. Wessel, P., Smith, W. H. F. (2004). The Generic Mapping Tools – Technical Reference<br />
and Cookbook (4th ed.).
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 23<br />
Drago Katović, Ph.D.<br />
Spanish Broom (Spartium Junceum L.)<br />
University <strong>of</strong> Zagreb<br />
Faculty <strong>of</strong> Textile Technology<br />
Zagreb, Prilaz baruna Filipovića 28A<br />
e-mail: dkatovic@ttf.hr<br />
Andrea Katović PhD<br />
University <strong>of</strong> Calabria, Faculty <strong>of</strong> Engineering Department <strong>of</strong> Chemical Engineering<br />
and Materials Via P.Bucci, Cubo 44 A, Rende (CS), Italy<br />
(e-mail: katovic@unical.it)<br />
Marija Krnčević B.Sc. <strong>of</strong> Ethnology and Archeology,<br />
Museum <strong>of</strong> Šibenik, Gradska vrata 3<br />
marija.krncevic@muzej-sibenik.hr)<br />
LA GINESTRA<br />
O IL FIORE DEL DESERTO<br />
Qui<br />
su l’arida schiena<br />
Del formidabil monte<br />
Sterminator Vesevo,<br />
La qual null’altro allegra arbor né ore<br />
Tuoi cespi solitari intorno spargi<br />
Odorata ginestra,<br />
Contenta dei deserti …<br />
Giacomo Leopardi<br />
Abstract<br />
The Spanish Broom (Spartium Junceum L.) plant is almost a forgotten textile raw material.<br />
This paper presents a review <strong>of</strong> procedures practiced in <strong>the</strong> past and nowadays for obtaining<br />
bers from this plant. It discusses new discoveries about its use as a component <strong>of</strong><br />
special composite bers, as well as its possible benets in households <strong>of</strong> poorer limestone<br />
areas. It also reports on come chemical and physical properties <strong>of</strong> <strong>the</strong> bers extracted from<br />
local plants.<br />
Key words: Spanish Broom, maceration, composites, ash, Klason lignin, cellulose, hemcellulose,<br />
FTIR, SEM, microwave, composites<br />
Introduction<br />
Whilst ax and hemp have mostly been used as textile raw material <strong>of</strong> cellulosic origin
24<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
in plains, in coastal areas <strong>of</strong> <strong>the</strong> Mediterranean wild Spanish Broom has been used as<br />
raw textile material since ancient times. However, Spanish Broom (Spartium Junceum L.)<br />
(Croatian: žuka or brnistra; Italian: ginestra) is almost completely forgotten today, and it is<br />
only sporadically mentioned as raw textile material. The habitat <strong>of</strong> Spanish Broom is <strong>the</strong><br />
Mediterranean area <strong>of</strong> south Europe, south-west Asia and north-west Africa. In Italy, in <strong>the</strong><br />
Mediterranean area <strong>of</strong> olive groves, it climbs <strong>the</strong> altitude <strong>of</strong> 975 meters. In Turkey, Syria<br />
and Palestine, it reaches altitudes <strong>of</strong> 1,700 meters. It is regarded weed in <strong>the</strong> USA and New<br />
Zealand; where <strong>the</strong>re is a tendency to eradicate it in order to save indigenous plants.<br />
Figure 1 Spartium Junceum L.<br />
There are numerous anatomic adjustments to dry soil visible in microscopic structures <strong>of</strong><br />
Spanish Broom’s vegetative organs. One <strong>of</strong> <strong>the</strong>m is xerophtytic (xerophytes – types <strong>of</strong><br />
plants adjusted to dry climate) adjustment <strong>of</strong> <strong>the</strong> leaf during its short life span and transformation<br />
<strong>of</strong> its internal structure dominated by <strong>the</strong> chain <strong>of</strong> parenchyma. The upper part<br />
<strong>of</strong> <strong>the</strong> stem took over <strong>the</strong> function <strong>of</strong> leaf, whereas sclerenchyma’s bers and conductive<br />
elements take over <strong>the</strong> majority part <strong>of</strong> its secondary units. An unusual feature <strong>of</strong> <strong>the</strong> root’s<br />
primary structure is its underdeveloped endoderm, and <strong>the</strong> secondary structure points to<br />
its storage and mechanical roles. [1]. Morphological structures <strong>of</strong> Spanish Broom, more<br />
precisely its longitudinal and cross-section <strong>of</strong> shoots were recorded with a FE-SEM (Field<br />
Emission-Scanning Electron Microscope) <strong>of</strong> <strong>the</strong> company TESCAN, with a steamer and<br />
unit for EDX analysis (Energy Dispersive X-Ray Analysis). Before <strong>the</strong> microscopic recording,<br />
samples were processed for 180 seconds in a steamer with gold and palladium, using<br />
operative voltage <strong>of</strong> 5-15 kV. The results <strong>of</strong> images show that Spanish Broom’s shoots<br />
have two basic layers: rigid and woody inner layer, porous in <strong>the</strong> middle part, making <strong>the</strong><br />
plant light, and an outer layer, skin intertwined with sinewy bers. The cross-section <strong>of</strong>
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 25<br />
Spanish Broom also points to epidermis, parenchyma <strong>of</strong> <strong>the</strong> bark with chlorophyll, endoderm,<br />
bundles <strong>of</strong> root bers and bundles <strong>of</strong> pericyle bers. Head shaped units <strong>of</strong> root and<br />
pericycle bers, connected in an almost continuous ring, are also visible.<br />
Figure 2 SEM micrograph <strong>of</strong> cross-section <strong>of</strong> Spanish Broom vermenes (SEM) Hitachi S-3400 N<br />
Cross-section <strong>of</strong> <strong>the</strong> woody shoot <strong>of</strong> Spanish Broom, which give rigidness to <strong>the</strong> shoot, and<br />
enables passage <strong>of</strong> cell juices alongside <strong>the</strong> shoot<br />
Figure 3 SEM micrograph <strong>of</strong> Cross-section <strong>of</strong> <strong>the</strong> woody (lignin) part
26<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
Figure 4 SEM micrograph <strong>of</strong> Cross-section <strong>of</strong> bers <strong>of</strong> Spanish Broom<br />
The cross-section <strong>of</strong> technical bers shows <strong>the</strong>ir irregular shape, with visible regular elementary<br />
circular bers in some places.<br />
Figure 5 Longitudinal micrograph SEM <strong>of</strong> Spanish Broom technical bers with visible elementary<br />
bers (SEM FE/Mira Tescan);
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 27<br />
Figure 6 Elementary ber <strong>of</strong> Spanish Broom, recorded with scanning electron microscope (SEM<br />
FE/Mira Tescan);<br />
THE APPLICATION OF SPANISH BROOM IN HISTORY<br />
Spanish Broom is a plant that was known even to <strong>the</strong> ancient Romans. In <strong>the</strong> past, Greeks,<br />
Romans and Carthaginians used it as raw material for <strong>the</strong> manufacture <strong>of</strong> ropes, nets, bags,<br />
sails. They also used it for covering ro<strong>of</strong>s and even clothing. The ower <strong>of</strong> <strong>the</strong> Spanish<br />
Broom is important for <strong>the</strong> coastal apiculture. Aristotle and Pliny praised <strong>the</strong> honey produced<br />
from Spanish Broom. In general, out <strong>of</strong> all ancient writers about Spanish Broom, it<br />
was Pliny Senior who wrote <strong>the</strong> most. He called it genista and he writes: Ginestra quoque<br />
vinculi,… ores apibus gratissimi (Spanish Broom is used for tying and making crumples,…<br />
and bees <strong>of</strong>ten land on its owers).<br />
However, in several places he mentions a plant called Spartum (Latin form <strong>of</strong> <strong>the</strong> Dioskorides’<br />
description <strong>of</strong> <strong>the</strong> plant sparton). After a detailed analysis <strong>of</strong> <strong>the</strong> application and<br />
processing, it is regarded <strong>the</strong> same plant today, or ra<strong>the</strong>r (Spartium Junceum L.). Already in<br />
<strong>the</strong> period <strong>of</strong> <strong>the</strong> ancient Rome, Spanish Broom elds (genestium – or ager in later Latin)<br />
were cultivated as any o<strong>the</strong>r eld, and <strong>the</strong> plant or sprouts were sown in ploughed up furrows,<br />
as <strong>the</strong> Roman agronomist Columella describes. According to Vergil, hedges <strong>of</strong> Spanish<br />
Broom were planted in addition to willow, hazel, elder and o<strong>the</strong>r plants. Pliny wrote<br />
that <strong>the</strong> sowing and planting <strong>of</strong> Spanish Broom was crucial for peasants. Amongst o<strong>the</strong>r<br />
things, its branches provide excellent material for tying vine and young trees.<br />
The most intensive production <strong>of</strong> Spanish Broom as raw material for obtaining textile fabric<br />
was in Italy during <strong>the</strong> 1930’s, when <strong>annual</strong> production <strong>of</strong> Spanish Broom, due to economic<br />
sanctions in Italy (war in Abyssinia) and <strong>the</strong> proclamation policy <strong>of</strong> self-sufciency,<br />
reached 700,000 tones. It was planted on 300,000 hectares and <strong>the</strong>re were 61 processing<br />
factories [2].
28<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
However, <strong>the</strong> industrial processing <strong>of</strong> <strong>the</strong> textile raw material completely extinguished in<br />
<strong>the</strong> mid 20th century. There were only several associations <strong>of</strong> Spanish Broom producers<br />
left in Europe that obtained subsidized EU projects with <strong>the</strong> goal <strong>of</strong> maintaining and improving<br />
old crafts. However, during <strong>the</strong> past 10 years <strong>the</strong>re has been a growing interest in<br />
Spanish Broom as raw material in <strong>the</strong> production <strong>of</strong> composite materials.<br />
Findings <strong>of</strong> Liburnian ships, named Serilias, dating from <strong>the</strong> 2nd century, prove that Spanish<br />
Broom was used quite early in our parts <strong>of</strong> <strong>the</strong> Adriatic. The special feature <strong>of</strong> this type<br />
<strong>of</strong> ship is <strong>the</strong> way parts <strong>of</strong> its structure were connected by sewing individual elements. All<br />
dilemmas surrounding this type <strong>of</strong> ship building were resolved by ancient writers, who<br />
also transposed <strong>the</strong> name <strong>of</strong> <strong>the</strong> vessel constructed using this sewing technique. Ancient<br />
writers left us with data about <strong>the</strong> special type <strong>of</strong> vessel used by Liburnians and Histrians<br />
and <strong>the</strong>ir construction techniques. Marcus Verrius Flacus, grammarian from <strong>the</strong> Augustan<br />
period, ga<strong>the</strong>red grammatically interesting, but faintly familiar words in his work De verborum<br />
signicatu. He also mentions <strong>the</strong> name serilia for Istrian and Liburnian ships. This<br />
record came to us through Sextus Pompeious Festus, who lived in <strong>the</strong> 200 AD. He writes:<br />
According to Verrius, Serilia is <strong>the</strong> name for Istrian and Liburnian vessels, compressed<br />
with ax and Spanish Broom ropes. Its name thus derives from <strong>the</strong> Latin word conserere<br />
(string toge<strong>the</strong>r) and contexere (to weave). In his play Niptra, Pacuvius writes: No wooden<br />
pin hold <strong>the</strong> ship’s haul toge<strong>the</strong>r. It is ra<strong>the</strong>r sewn using ax and Spanish Broom ropes. He<br />
used a descriptive expression and invented word for ropes spun from Spanish Broom. [1].<br />
In this area, Spanish Broom was used throughout <strong>the</strong> Middle Ages, according to several<br />
historical records. The oldest record, even though it only mentions Spanish Broom, does<br />
point to its processing for textile purposes. This data is recorded in <strong>the</strong> collection <strong>of</strong> laws<br />
and regulations Statuta et leges civitatis et insulae Curzlae (1214 – 1558). According to<br />
chapter XCVIII, it was prohibited to ga<strong>the</strong>r on moles <strong>of</strong> Zavalatica, Prigradica and Blace,<br />
since <strong>the</strong>y were probably places where women came to soak ax and Spanish Broom. Provisions<br />
for <strong>the</strong> protection <strong>of</strong> useful plants were prescribed quite <strong>of</strong>ten. One <strong>of</strong> <strong>the</strong> examples<br />
is <strong>the</strong> Statute <strong>of</strong> Bra from 1656[3].<br />
The next preserved record on Spanish Broom was written in Glagolitic alphabet and dates<br />
from 1674. It describes <strong>the</strong> area <strong>of</strong> Šibenik. It is a testament and reads: Ostavlan Matii<br />
svekrivi jednu kanicu novu i brnistru, ka j(e) g(o)tova. It is not quite clear what does nished<br />
Spanish Broom actually refer to.<br />
In <strong>the</strong> book Viaggio in Dalmazia, written by <strong>the</strong> famous Italian travel writer Alberto Fortis<br />
in 1774, he describes his visit to <strong>the</strong> island <strong>of</strong> Murter including <strong>the</strong> processing and use <strong>of</strong><br />
Spanish Broom: “The favorite craft <strong>of</strong> Betinjans, population <strong>of</strong> <strong>the</strong> western part <strong>of</strong> <strong>the</strong> island,<br />
is picking, soaking and weaving <strong>of</strong> Spanish Broom which <strong>the</strong>y nd on <strong>the</strong> coast <strong>of</strong> Istria<br />
and islands <strong>of</strong> Kvarner. They make fabric from it, which <strong>the</strong>y use for making sacks and<br />
sometimes shirts and skirts for peasants. There is no doubt this plant could produce far better<br />
artifacts if processed in a more rened manner. They use <strong>the</strong> sea to soak <strong>the</strong> bundles.”<br />
The cultivation <strong>of</strong> Spanish Broom is unfamiliar today in our parts. However, it was widely<br />
spread in <strong>the</strong> past. The shortage <strong>of</strong> textile in 1916 motivated Friar Andrija Rajkovi to use<br />
<strong>the</strong> people’s experience in Spanish Broom processing and use it to a greater extent, even<br />
in industry. He explained to authorities at that time <strong>the</strong> need, benet and possibilities <strong>of</strong><br />
processing Spanish Broom, sending several samples <strong>of</strong> ready-made fabric. In <strong>the</strong> same<br />
memo he writes: For <strong>the</strong> past 40 or 50 years, Spanish Broom has been widely used, but now<br />
manufacturing <strong>of</strong> this type <strong>of</strong> fabric is completely neglected. He also attached a guide How
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 29<br />
to process Spanish broom fabric. [3]<br />
During <strong>the</strong> ‘30s <strong>of</strong> <strong>the</strong> past century, foresters widely recommended <strong>the</strong> cultivation <strong>of</strong> Spanish<br />
Broom (Spartium junceum) in <strong>the</strong> coastal karst areas, more precisely M. Ani Ph. D<br />
in Forest Journal in 1937 and engineer Stane Benko in his booklet on Spanish Broom,<br />
published by <strong>the</strong> Ministry <strong>of</strong> Industry and Mining, in Zagreb, 1946. Member <strong>of</strong> <strong>the</strong> Institute<br />
for Forestry Research Ante Premuži explains <strong>the</strong> possibility <strong>of</strong> using Spanish Broom<br />
for economic purposes in his report Systematic cultivation <strong>of</strong> Spanish broom in our karst<br />
areas, published in <strong>the</strong> Forest Journal. He wrote that diverse biological properties could<br />
be put to good use for forestation <strong>of</strong> poor soil <strong>of</strong> <strong>the</strong> coastal limestone area since Spanish<br />
Broom grows on clay and limestone soil <strong>of</strong> neutral, acid and alkali reaction. Positive sides<br />
<strong>of</strong> its cultivation were highly praised, giving <strong>the</strong>se studies a more propaganda character.<br />
Alongside <strong>the</strong> coast <strong>of</strong> east Adriatic, people used and processed Spanish Broom in various<br />
ways, but it was also harvested in Istria and <strong>the</strong> coastal area <strong>of</strong> Dubrovnik in <strong>the</strong> period <strong>of</strong><br />
Austrian-Hungarian Monarchy, and later between <strong>the</strong> two World Wars. It was <strong>the</strong>n harvested<br />
by <strong>the</strong> army, tied up in bundles and exported to Italy for manufacturing military suits.<br />
There were many attempts <strong>of</strong> industrial processing in our country. Let us remember <strong>the</strong><br />
previously mentioned initiative <strong>of</strong> Rajkovi, which spread across <strong>the</strong> entire Adriatic, gaining<br />
<strong>the</strong> status <strong>of</strong> industrial processing. Immediately after <strong>the</strong> World War I in 1919, <strong>the</strong>re<br />
was an attempt in Omiš to process Spanish Broom in bulk. A factory was also built for its<br />
processing. However, <strong>the</strong> rst attempt failed since Spanish Broom had been soaked in <strong>the</strong><br />
sea for six months, <strong>the</strong>reby rotting. There were no fur<strong>the</strong>r attempts from <strong>the</strong>n on. Only<br />
after <strong>the</strong> end <strong>of</strong> World War II, Spanish Broom was rediscovered, mostly due to a shortage<br />
<strong>of</strong> textile raw material during <strong>the</strong> post-war period.<br />
The processing actually started two years after <strong>the</strong> W. W. II (1947). Around two trucks <strong>of</strong><br />
Spanish Broom, harvested near Rabac and Labin (Istria), were turned over to <strong>the</strong> district <strong>of</strong><br />
Pula to send it for processing in <strong>the</strong> factory in Fažana. It was also decided to build factories<br />
in Vodice and Zakuac near Omiš. There was a second attempt to revive <strong>the</strong> production,<br />
this time using state-<strong>of</strong>-<strong>the</strong> art methods, quite near where <strong>the</strong> rst plant had been constructed<br />
in 1919. The process <strong>of</strong> maceration was chemical, and a device for hot-air drying was<br />
used. In <strong>the</strong> paper factory in Rijeka, a department was turned into a laboratory for testing.<br />
In order to save on transport <strong>of</strong> raw material, someone suggested turning an old ship into a<br />
oating factory, which would cruise in <strong>the</strong> Adriatic and process Spanish Broom, quite expensive<br />
to harvest. The results <strong>of</strong> this test production were taken over for fur<strong>the</strong>r processing<br />
by certain spinning mills in Croatia and Slovenia, and so rst test products were actually<br />
produced. However, <strong>the</strong> process stops <strong>the</strong>re. The whole initiative, which had started positively,<br />
stopped abruptly, allegedly due to high purchasing price, and so traditional processing<br />
remained in <strong>the</strong> hands <strong>of</strong> <strong>the</strong> people. [3].<br />
There were o<strong>the</strong>r attempts to process Spanish Broom industrially. In Vodice, between 1946<br />
and 1949, a factory was built for <strong>the</strong> processing <strong>of</strong> Spanish Broom. It worked until 1954,<br />
when it was shut down due to unprotability. The same factory is scheduled for demolition<br />
today.<br />
According to former workers <strong>of</strong> <strong>the</strong> factory (Tvornica Žuke) in Vodice <strong>the</strong> technological<br />
procedure for obtaining bbers was as follows: branches <strong>of</strong> Spanish Broom with <strong>the</strong>ir<br />
shoots were soaked in sodium alkali for several days. After resting in pools with <strong>the</strong> alkali,<br />
bers would separate from <strong>the</strong> o<strong>the</strong>r parts. Bundles <strong>of</strong> Spanish Broom would be subject<br />
to water jets under pressure <strong>of</strong> four to six atmospheres. The water jets would separate b-
30<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
ers from <strong>the</strong> connective tissue, transferring <strong>the</strong>m to pools for rinsing <strong>of</strong>f <strong>the</strong> alkali. After<br />
air drying, periodic exposure to rain would have a positive effect on <strong>the</strong> quality. Fibers in<br />
<strong>the</strong> shape <strong>of</strong> hemp were <strong>the</strong>n mechanically cleansed from any eventual remainders <strong>of</strong> <strong>the</strong><br />
connective tissue. They were pressed in bales and sent for fur<strong>the</strong>r processing in <strong>the</strong> factory<br />
Duga Resa. The resulting technical bers <strong>of</strong> Spanish Broom were used for producing ropes.<br />
The original use <strong>of</strong> Spanish Broom can be found in Petriani near Zadar. Freshly cut shoots<br />
were chopped and beaten into powder. This powder was sprinkled on a wet mesh, and <strong>the</strong><br />
ingredients which it would absorb, would protect it against rotting. Spanish Broom shoots<br />
were also used for manufacturing smaller-sized coops. [3]<br />
Several associations <strong>of</strong> Spanish Broom producers remain in Europe today. The European<br />
Union allocated subsidies to Italy with a goal <strong>of</strong> preserving and developing old trades.<br />
There is also a craft shop which produces sails made <strong>of</strong> Spanish Broom for <strong>the</strong> reconstruction<br />
<strong>of</strong> old boats. In Pirovac, <strong>the</strong> Porart Association organizes traditional harvesting and<br />
maceration <strong>of</strong> Spanish Broom, once practiced in our parts. (Figure 7-9).<br />
Figure 7 Harvesting <strong>of</strong> Spanish Broom<br />
Figure 8 Tying shoots in bundles - fašine
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 31<br />
Figure 9 Maceration <strong>of</strong> Spanish Broom in <strong>the</strong> sea<br />
There are several items made from Spanish Broom in our museums, or ra<strong>the</strong>r its combination<br />
<strong>of</strong> o<strong>the</strong>r bers. The sheet – lancun – from <strong>the</strong> Island <strong>of</strong> Pag (Jakišnica), 207x160 cm, was<br />
weaved in four threads in approximately 1945. It consists <strong>of</strong> two parts and it was washed and<br />
whitened by damping in dew and sun exposure. It is in good shape, with a few smears.<br />
Figure 10 Lancun (sheet) made <strong>of</strong> Spanish Broom from Pag (collection <strong>of</strong> <strong>the</strong> Museum <strong>of</strong><br />
Ethnography in Split)<br />
Figure 11 Sheet (collection <strong>of</strong> <strong>the</strong> Town Museum <strong>of</strong> Šibenik)
32<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
Cover from <strong>the</strong> Island <strong>of</strong> Capri (local name: klašnje), 130x80 cm. It is mostly made <strong>of</strong><br />
Spanish Broom, whilst old pieces <strong>of</strong> cloth were used for weft. They were <strong>of</strong> uniformed<br />
size, depending on <strong>the</strong> width <strong>of</strong> <strong>the</strong> weaving loom.<br />
Figure 12 Cover (basically Spanish Broom) from Capri (collection <strong>of</strong> <strong>the</strong> Town Museum <strong>of</strong><br />
Šibenik)<br />
The Faculty <strong>of</strong> Textile Technology, within <strong>the</strong> framework <strong>of</strong> <strong>the</strong> Textile Science Research<br />
Center, analyzed a sample <strong>of</strong> a fuse from a Roman ceramic oil lamp found in Split. The<br />
oil lamp dates from <strong>the</strong> 3rd or 4th century AD. Considering this item is so rare, we were<br />
interested from which material <strong>the</strong> fuse was made <strong>of</strong>.<br />
Figure 13 Roman ceramic oil lamp (lucerna)<br />
For <strong>the</strong> purpose <strong>of</strong> <strong>the</strong> fuse analysis, several images were taken with a scanning electron<br />
microscope (SEM) <strong>of</strong> <strong>the</strong> company Tescan Mira/FE .
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 33<br />
Figure 14 - 20 SEM images <strong>of</strong> <strong>the</strong> fuse (variations in zoom)
34<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
The images show bers covered with ashes, but not completely burned (grey part <strong>of</strong> <strong>the</strong><br />
sample). O<strong>the</strong>rwise, <strong>the</strong>y would not be identiable with precision. There are also black<br />
clumps <strong>of</strong> dirt (larger shapeless pieces)<br />
For <strong>the</strong> purpose <strong>of</strong> analyzing <strong>the</strong> fuse FTIR (Fourier Transform Infra Red) specters <strong>of</strong> bers<br />
were also recorded, which might have been used in <strong>the</strong> 3 rd or 4 th century AD in our parts:<br />
cotton, ax, hemp and Spanish Broom. The resulting curves <strong>of</strong> <strong>the</strong> mentioned bers were<br />
compared with <strong>the</strong> spectrum <strong>of</strong> <strong>the</strong> fuse sample.<br />
Figure 21 Fuse spectrum<br />
After attempts <strong>of</strong> separating parts <strong>of</strong> <strong>the</strong> fuse (grey part) from <strong>the</strong> black part, which mostly<br />
contains dirt, <strong>the</strong> difference in FTIR specters was evident, although insufcient for a denite<br />
conclusion. The differences, which relate to a total <strong>of</strong> spikes, double spike at wavelength<br />
ca. 2900 cm -1 , and at 1700 cm -1 (arc) and 1540 cm -1 (spike), represent <strong>the</strong> existence<br />
<strong>of</strong> non-cellulose natural bers (wax, pectin, lignin). At rst glance that would mean it is not<br />
cotton or Spanish Broom. However, this would be acceptable only if we take into consideration<br />
special methods for ber processing, practiced after <strong>the</strong> 19 th century.<br />
It was <strong>the</strong>n assumed, based on <strong>the</strong> color and SEM images, that <strong>the</strong> fuse was actually a<br />
remainder <strong>of</strong> incomplete burning process. FTIR ber specters were <strong>the</strong>n recorded after<br />
burning, and it was concluded it was probably not cotton.<br />
THE APPLICATION OF SPANISH BROOM IN FUTURE<br />
In <strong>the</strong> past, natural bers were not taken into account as reinforcements for polymeric<br />
materials because <strong>of</strong> some problems associated with <strong>the</strong>ir use. Lack <strong>of</strong> good interfacial<br />
adhesion, low degradation temperature, properties variability depending on <strong>the</strong> quality <strong>of</strong><br />
harvest, age and body <strong>of</strong> <strong>the</strong> plant, as well poor resistance against moisture make <strong>the</strong> use<br />
<strong>of</strong> natural ber reinforced composites less attractive than syn<strong>the</strong>tic. However, <strong>the</strong> production<br />
<strong>of</strong> composites reinforced with syn<strong>the</strong>tic bers and matrices (glass, carbon, aramid)<br />
requires a large amount <strong>of</strong> energy, and <strong>the</strong>y are difcult to recycle or <strong>the</strong>y are completely<br />
unrecyclable. Moreover, natural bers in cases <strong>of</strong> re decrease <strong>the</strong> content <strong>of</strong> toxic gases<br />
<strong>of</strong> combustion. [10].<br />
For <strong>the</strong> past several years <strong>the</strong>re has been an increase in <strong>the</strong> number <strong>of</strong> scientic studies
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 35<br />
relating to <strong>the</strong> production <strong>of</strong> Spanish Broom as raw material for composite materials [6-<br />
8]. The advantage <strong>of</strong> Spanish Broom over ax and hemp is that it can grow in <strong>the</strong> most<br />
unfavorable limestone soil; it is resilient to draught; and once planted it can be used during<br />
a period <strong>of</strong> up to twenty years, whilst ax and hemp demand high quality soil each<br />
year. Its cultivation on rocky areas would have several benets. For <strong>the</strong> purpose <strong>of</strong> analyzing<br />
<strong>the</strong> possibility <strong>of</strong> using Spanish Broom as composite material, inborn samples were<br />
harvested in <strong>the</strong> surroundings <strong>of</strong> Šibenik. Freshly picked shoots contained up to 35% <strong>of</strong><br />
moisture, compared with <strong>the</strong> mass <strong>of</strong> samples dried in standard conditions. Samples were<br />
tied in bundles, soaked in sea water for 21 days and <strong>the</strong>n bers were separated from <strong>the</strong><br />
remaining woody part, using a mechanical procedure. This traditional procedure <strong>of</strong> extracting<br />
Spanish Broom was compared with several contemporary processes, using NaOH,<br />
with subsequent <strong>the</strong>rmal processing at 130 °C and rapid decompression. [9]. Processing<br />
would last for approximately six hours, which is signicantly less compared with water<br />
processing between 20 and 80 days. According to <strong>the</strong> latest studies <strong>of</strong> <strong>the</strong> Faculty <strong>of</strong> Textile<br />
Technology – Institute for Textile Technology and Ecology – <strong>the</strong> possibility <strong>of</strong> microwave<br />
processing is currently undergoing research. According to present and promising results,<br />
total maceration period was reduced to 10 minutes, with a signicant decrease in <strong>the</strong> use<br />
<strong>of</strong> applied chemicals.<br />
Figure 22 Energy for production <strong>of</strong> some bers<br />
Energy for production <strong>of</strong> Spanish Broom depends on <strong>the</strong> type <strong>of</strong> applied maceration and<br />
ranges within <strong>the</strong> framework <strong>of</strong> o<strong>the</strong>r natural bers, according to our research.<br />
Cotton Flax Hemp Spanish broom<br />
Tensile strength (MPa) 300-700 500-900 310-750 750<br />
Figure 23 Some natural bers properties<br />
Spanish Broom bers were extracted from <strong>the</strong> plant’s branches harvested in Dalmatia<br />
(Šibenik area) which contained 35 % <strong>of</strong> moisture compared with branches dried under
36<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
standard conditions. The IR spectra <strong>of</strong> <strong>the</strong> ber sample obtained after alkaline extraction<br />
was recorded on <strong>the</strong> ATR-FTIR spectrometer (Spectrum 100, Perkin Elmer), while cotton,<br />
ax and hemp bers were taken as reference samples. From <strong>the</strong> recorded FT-IR spectra <strong>of</strong><br />
<strong>the</strong> ber samples it is visible that all bers have peak characteristic <strong>of</strong> cellulose although<br />
<strong>the</strong>y present some differences (Fig. 24.) The intensity <strong>of</strong> signal at 1734 cm-1 corresponding<br />
to <strong>the</strong> C=C es<strong>the</strong>r band related to pectin is higher in ax than in hemp bers while this<br />
band is not found in Spanish Broom ber sample. On <strong>the</strong> contrary, only a weak peak at<br />
ca. 1500 cm-1 corresponding to <strong>the</strong> C=C in-plane aromatic vibrations from lignin can be<br />
observed in <strong>the</strong> case <strong>of</strong> Spanish Broom bers, providing an ulterior evidence <strong>the</strong> applied<br />
chemical treatment was adequate for <strong>the</strong> almost complete removal <strong>of</strong> non-cellulosic compounds<br />
from <strong>the</strong> Broom bers. On <strong>the</strong> o<strong>the</strong>r hand, <strong>the</strong> two little sharp peaks observed over<br />
a broader peak in <strong>the</strong> area <strong>of</strong> 2850 to 2950 cm -1 , attributed to <strong>the</strong> CH 2<br />
and CH groups <strong>of</strong><br />
long alkyl chains <strong>of</strong> waxes, are present in <strong>the</strong> spectra <strong>of</strong> ax and hemp.<br />
Figure 24 FTIR (Fourier Transform Infra Red) spectra <strong>of</strong> some bast bers compared to <strong>the</strong> cotton<br />
ber<br />
In Fig. 25 <strong>the</strong> FTIR spectra <strong>of</strong> Spanish Broom bers obtained after four different extraction<br />
procedures are presented. According to <strong>the</strong> previously highlighted peaks that point <strong>the</strong><br />
presence <strong>of</strong> pectin, lignin and/or wax, it can be seen that <strong>the</strong> procedures B and D resulted<br />
in <strong>the</strong> best removal <strong>of</strong> <strong>the</strong> non-cellulosic compounds <strong>of</strong> <strong>the</strong> ber. Fiber treatment (B) with<br />
sodium hydroxide at 105 °C for 2 h removes completely wax and lignin while a weak peak<br />
related to pectin at 1740 cm -1 is still visible. On <strong>the</strong> o<strong>the</strong>r hand, ber treatment (D) with<br />
<strong>the</strong> same alkaline solution but heated in a microwave oven for only 10 minutes (900 W)<br />
resulted in greater removal <strong>of</strong> pectins and maybe not total elimination <strong>of</strong> waxes. The latter<br />
treatment seems very promising, especially from <strong>the</strong> economical point <strong>of</strong> view.
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 37<br />
Figure 25 FTIR spectrum <strong>of</strong> bers obtained from Spanish Broom applying four different procedures<br />
compared to <strong>the</strong> cotton ber.<br />
REFERENCES<br />
1. Kovaevi Z., Krnevi M., Katovi A. Katovi D.: Brnistra – zaboravljena tekstilna<br />
sirovina Tekstil <strong>2010</strong> –in press<br />
2. http://www.lammatest.rete.toscana.it/lammatest/documenti/ginestra_manuale.pdf<br />
Accessed November <strong>2010</strong><br />
3. Stojanovi A.: Brnestra (Žuka- Spartium Junceum) (1962) Etnološki Zavod Filoz<strong>of</strong>skog<br />
fakelteta Sveuilišta u Zagrebu, 4, 3-45<br />
4. Katovi D., Katovi A., Krnevi M.: Spanish Broom - History and Perspective (Spartium<br />
Junceum), Journal <strong>of</strong> Natural Fibers <strong>2011</strong> - in press<br />
5. Nekkaa S., Chebira F., Haddaoui N.(2006): Effect <strong>of</strong> Fiber Treatment on <strong>the</strong> Mechanical<br />
and Rheological Properties <strong>of</strong> Polypropylene/Broom Fiber Spartium Junceum<br />
Composites Journal <strong>of</strong> Engineering and Applied Sciences 1(3), 278-283<br />
6. Cerchiara,T., Chidichimo G., Ragusa, M.I., Belsito E.L., Liguorib, A. Arioli A. (<strong>2010</strong>):<br />
Characterization and utilization <strong>of</strong> Spanish Broom (Spartium junceum L.) seed oil Industrial<br />
Crops and Products 31, 423–426<br />
7. Avella M., Casale L., Dell’Erba R., Focher B., Martuscelli E., Marzetti A.M. (1998):<br />
Broom Fibers as Reinforcing Materials for Polypropylene-Based Composites J. Appl<br />
Polym Sci 68, 1077–1089<br />
8. Gabriele B., Teresa Cerchiara T., Salerno G., Chidichimo G., Vetere M.V. Alampi C,<br />
Gallucci M.C., Conidi C., Cassano A.( <strong>2010</strong>): A new physical-chemical process for<br />
<strong>the</strong> efcient production <strong>of</strong> cellulose bers from (Spartium junceum L.) Bioresource<br />
Technology 101, 724–729<br />
9. Cerchiara T., Chidichimo G., Gallucci M.C. Vuono D., (<strong>2010</strong>): Effects <strong>of</strong> Extraction<br />
(Spartium junceum L.) Fibres, Fibres & Textiles in Eastern Europe 2(79) 13-16<br />
10. Cristaldi G. Latteri A. Recca G., Cicala G. (<strong>2010</strong>) : Woven Fabric Engineering , Composites<br />
Based on Natursal Fibre Fabrics, Sciyo, Rijeka, pp.317-342, ISBN 978-953-<br />
307-194-7
38<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
THE WAY TO REDUCE PIPE WEARING WHILE DRILLING<br />
Matanović Davorin, Moslavac Bojan, Nediljka Gaurina-Međimurec<br />
University <strong>of</strong> Zagreb<br />
Faculty <strong>of</strong> Mining, Geology and Petroleum Engineering<br />
Pierottijeva 6, 10000 Zagreb, Republic <strong>of</strong> Croatia<br />
Tel. +385 1 5535 835; fax: +385 1 48 36 074<br />
E-mail: dmatan@rgn.hr<br />
Abstract<br />
While drilling <strong>the</strong>re is an additional load on <strong>the</strong> coiled tubing due to friction with <strong>the</strong> well<br />
walls/rocks regardless <strong>the</strong> drilling mode; rotating or sliding. According to recent investigations<br />
fatigue is <strong>the</strong> major factor controlling <strong>the</strong> service life <strong>of</strong> coiled tubing. The rate<br />
<strong>of</strong> fatigue damage accumulation is affected by a number <strong>of</strong> operating factors: mechanical<br />
damage, corrosion damage, damage incurred by exceeding <strong>the</strong> ultimate capacity <strong>of</strong> <strong>the</strong><br />
string, and damage due to cycling plastic strain or fatigue.<br />
Determination <strong>of</strong> <strong>the</strong> actual friction coefcient is essential in dening <strong>the</strong> project technical<br />
feasibility. As <strong>the</strong> part <strong>of</strong> drilling system, drilling uid can impact on <strong>the</strong> reduction <strong>of</strong> friction<br />
coefcient by implementation <strong>of</strong> lubricants. Selection <strong>of</strong> lubricants, <strong>the</strong>ir compatibility<br />
with drilling uid and rocks was determined through laboratory testing. The results are<br />
shown and elaborated in <strong>the</strong> paper.<br />
Key words: drilling, pipe wearing, lubricants<br />
Introduction<br />
Horizontal wells are more expensive to drill than vertical wells so increased production<br />
must <strong>of</strong>fset increased drilling costs. In drilling horizontal or highly deviated wells, more<br />
serious problems appear than in drilling vertical wells. These problems are: poor hole<br />
cleaning, excessive torque and drag, hole lling, pipe sticking, well bore instability, loss <strong>of</strong><br />
circulation, formation damage, poor cement job, and difculties at logging jobs. From that<br />
reason, <strong>the</strong> principle factors that require extra consideration for choosing a uid for drilling<br />
a horizontal well are: formation damages, hole cleaning, hole stability and lubricity. Good<br />
lubricating properties <strong>of</strong> <strong>the</strong> drilling uid can improve greatly drilling operation efciency.<br />
When drilling long sections <strong>of</strong> highly deviated wells, <strong>the</strong>re is considerable contact between<br />
<strong>the</strong> drill string and <strong>the</strong> borehole wall, generating frictional resistance to movement. The<br />
friction between <strong>the</strong> drill pipe and <strong>the</strong> wall will increase torque and power required to rotate<br />
<strong>the</strong> drill pipe and <strong>the</strong> stress on <strong>the</strong> drill pipe. It could also interfere with running <strong>the</strong> pipe<br />
in and out <strong>of</strong> <strong>the</strong> hole. The frictional resistance to movement can be large enough to be <strong>the</strong><br />
limiting factor in horizontal and extended-reach drilling. Since <strong>the</strong>re are many factors that<br />
affect torque and drag, it is sometimes difcult to detect what is causing increased down<br />
hole friction.<br />
Regardless <strong>of</strong> <strong>the</strong> drilling uid or lubricant used, down hole friction can be reduced by
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 39<br />
conditioning <strong>the</strong> uid to achieve a thin, slick, compressible lter cake, and by practicing<br />
good hole-cleaning techniques to minimize <strong>the</strong> cuttings bed. In order to design <strong>the</strong> best<br />
suitable drill-in uid for horizontal well, drilling problems generally encountered in each<br />
eld should be studied and laboratory tests planned and conducted.<br />
Systematic approach to drilling optimization<br />
To optimize a drilling process <strong>the</strong>re is a need to develop methods and processes that will<br />
improve drilling efciency. The method must be systematic, logical and must have quantiable<br />
approach. The applicable method is one that denes and quanties two sets <strong>of</strong><br />
parameters for <strong>the</strong> drilling operation: performance objectives and requirements, as shown<br />
in Fig. 1. The gure is taken from reference [1] but is slightly changed to apply for coiled<br />
tubing drilling with down hole positive displacement motor.<br />
Fig. 1 Systematic approach inuence diagram<br />
To achieve minimal interval cost we have to maximize on-bottom time and effective rate <strong>of</strong><br />
penetration. First branch determines parameters that are not direct drilling parameters, but<br />
depend on well path and used equipment. Second branch determines parameters that are<br />
<strong>of</strong> greatest inuence in drilling process. Bit design and used down hole motor combination<br />
efciency according to observed formation characteristics are essential. To obtain optimal<br />
performance <strong>of</strong> bit-motor combination optimal ow rate must be obtained and proper uid<br />
used.<br />
In deviated or horizontal wells proper determination <strong>of</strong> weight on bit in any moment or<br />
point <strong>of</strong> <strong>the</strong> well trajectory is essential. The determination <strong>of</strong> compressive load in vertical<br />
section, build section or horizontal section depends on well trajectory and coiled tubing<br />
dimensions. Buckling <strong>of</strong> coiled tubing and lock-up condition is <strong>the</strong> next limitation that<br />
must be considered.<br />
Because <strong>of</strong> different possible well paths and segment length and slope (Fig. 2) it is necessary<br />
to determine forces and stresses in coiled tubing.
40<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
Fig 2. Possible well paths and pipe buckling in <strong>the</strong>m<br />
When coiled tubing is run into horizontal section, it is subjected to compressive load due<br />
to frictional drag. The axial compressive load is highest at <strong>the</strong> end <strong>of</strong> <strong>the</strong> built section and<br />
lowest at <strong>the</strong> far end <strong>of</strong> <strong>the</strong> well. When compressive load exceeds <strong>the</strong> critical buckling load,<br />
<strong>the</strong> coiled tubing will initiate a sinusoidal buckling. Fur<strong>the</strong>r increase <strong>of</strong> <strong>the</strong> compressive<br />
load can result in a helical buckling <strong>of</strong> <strong>the</strong> coiled tubing. Helical buckling can also occur in<br />
<strong>the</strong> vertical section when “slacking-<strong>of</strong>f” weight at <strong>the</strong> surface to push coiled tubing through<br />
<strong>the</strong> well. In horizontal section coiled tubing will buckle when <strong>the</strong> axial compressive load<br />
exceeds critical (sinusoidal) buckling load. When <strong>the</strong> axial compressive load reaches helical<br />
buckling load, helical buckling occurs, and is fully formed.<br />
In <strong>the</strong> build-up section, coiled tubing under axial compression will be pushed against <strong>the</strong><br />
lower side <strong>of</strong> <strong>the</strong> well bore before <strong>the</strong> onset <strong>of</strong> sinusoidal or helical buckling. In this section<br />
buckling will not easily occur because <strong>of</strong> distributed lateral force, and <strong>the</strong> need <strong>of</strong> larger<br />
buckling load in build-up section than in straight well bore.<br />
When pushing coiled tubing into horizontal section with no helical buckling present, <strong>the</strong><br />
axial load simply increases linearly along <strong>the</strong> tubular. This is due to <strong>the</strong> friction drag from<br />
<strong>the</strong> tubular weight, without considering any additional frictional force from sinusoidal<br />
buckling, since usually it is very small. In vertical section without helical buckling <strong>the</strong>re<br />
is <strong>the</strong>oretically no frictional drag and <strong>the</strong> axial compressive load decreases linear. If helical<br />
buckling occurs in <strong>the</strong> vertical section due to “slack-<strong>of</strong>f”, <strong>the</strong> axial load distribution<br />
decreases nonlinear:
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 41<br />
Maximum axial compressive load that is transmitted to <strong>the</strong> bottom <strong>of</strong> vertical section to<br />
kick-<strong>of</strong>f point (KOP) by “slacking-<strong>of</strong>f” is [2]:<br />
(1)<br />
Where: (F b,max<br />
) is maximal axial compressive load at <strong>the</strong> bottom <strong>of</strong> vertical section, (E) is<br />
Young’s modulus <strong>of</strong> elasticity, (I) is moment <strong>of</strong> inertia <strong>of</strong> tubular, (we)- tubular weight in<br />
uid, (r) is radial clearance between well bore and tubular, () is coefcient <strong>of</strong> friction and<br />
(H) is vertical well bore depth.<br />
The compressive axial load at <strong>the</strong> end <strong>of</strong> <strong>the</strong> build section (EOC) is:<br />
(2)<br />
Where: (F eoc<br />
) is axial load at <strong>the</strong> end <strong>of</strong> build section, (F kf)<br />
is axial load at <strong>the</strong> kick<strong>of</strong>f point,<br />
(R) is radius <strong>of</strong> well bore curvature and (e) is natural logarithm base.<br />
“Lockup” refers to situation where <strong>the</strong> bit or packer load can not be increased by “slacking<strong>of</strong>f”<br />
weight at <strong>the</strong> surface, or <strong>the</strong> tubulars can not be pushed fur<strong>the</strong>r in <strong>the</strong> well bore. And<br />
<strong>the</strong> maximal horizontal reach without helical buckling in horizontal section will be:<br />
(3)<br />
Where: (L h<br />
) is horizontal section length.<br />
It is obvious that coefcient <strong>of</strong> friction is <strong>of</strong> importance when calculating loads on tubular.<br />
Measurements conducted by Johancsik at all [3] have shown that it can vary from 0,2 to 0,4<br />
in real well bore situations. The reduction <strong>of</strong> friction will allow greater horizontal lengths<br />
with smaller pipe wearing. The capability to do so has two main benets. First, deep,<br />
highly deviated wells can be planned to minimize torque and drag by selection <strong>of</strong> most<br />
appropriate well path. Second, more complete knowledge <strong>of</strong> drill string loading allows<br />
choose <strong>of</strong> drill string components according to systematic approach that considers <strong>the</strong> extra<br />
forces involved. They have showed that tension increment and torsion increment depend<br />
on coefcient <strong>of</strong> friction and normal force:<br />
Where: (F t<br />
) is axial tension acting at lower end <strong>of</strong> element, ( ) is average inclination<br />
angle <strong>of</strong> element, (F n<br />
) is normal force acting on element, (M) is increase in torsion over<br />
length <strong>of</strong> element, and (r p<br />
) is characteristic radius <strong>of</strong> drill string element. In Eq. 4, <strong>the</strong> plus<br />
sign is for upward motion and <strong>the</strong> minus sign is for downward motion.<br />
(4)<br />
(5)
42<br />
LUBRICANTS<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
Friction reduction can be achieved by using a drilling uid with good lubricating qualities<br />
such as oil-base uids, polymer uids, and glycol-derivative systems [4]. Oil-based uids<br />
<strong>of</strong>ten are chosen for extended reach wells because <strong>the</strong>y are highly inhibitive and exhibit<br />
good lubrication properties. Unfortunately, <strong>the</strong>ir use in environmentally sensitive areas is<br />
discouraged or even prohibited [5]. There are also a large number <strong>of</strong> lubricant additives<br />
available; <strong>the</strong>se should be tested for uid and formation compatibility as some <strong>of</strong> <strong>the</strong>m<br />
can have detrimental effects. In general, lubricants may be <strong>of</strong> a mechanical nature (beads,<br />
nut plug, etc.), general borehole or lm lubricants, or extreme-pressure (metal-to-metal)<br />
lubricants.<br />
Natural grease and oils abandoned because <strong>of</strong> cheaper oxidant stable mineral oils, are now,<br />
because <strong>of</strong> ecological aspect, used again. Rape oil and castor oil are used in production <strong>of</strong><br />
biodegradable lubricants. Rape oil has also o<strong>the</strong>r favorable properties: good lubricity, high<br />
viscosity, good adhesion on metal surfaces and very good load bearing. Problems on high<br />
and low temperatures and with foaming can be solved by adding additives. When added<br />
in water based mud (WBM) such problems are solved with <strong>the</strong> use <strong>of</strong> emulsiers, antifoaming<br />
agents and for warehouse purposes pour point depressants.<br />
Lubricants are designed to reduce torque to increase horsepower at <strong>the</strong> bit by reducing <strong>the</strong><br />
coefcient <strong>of</strong> friction. In water based drilling uids certain oils, graphite powder, glycols<br />
and soaps are used for this purpose. To be economic, a drilling uid lubricant must reduce<br />
torque and drag, and operating costs by performing <strong>the</strong> following functions:<br />
• reduce possibilities <strong>of</strong> twist-<strong>of</strong>fs,<br />
• shorten trip time,<br />
• lessen <strong>the</strong> conditions for differential sticking, and<br />
• lower <strong>the</strong> amount <strong>of</strong> energy necessary to run <strong>the</strong> rig.<br />
Depending on <strong>the</strong>ir chemical composition and state <strong>of</strong> dispersion or solubility in <strong>the</strong> base<br />
drilling uid, lubricants:<br />
• can coat metal surfaces, reducing <strong>the</strong> adhesion <strong>of</strong> steel to <strong>the</strong> lter cake;<br />
• can be incorporated into <strong>the</strong> lter cake and provide better uid-loss control (resulting in<br />
thinner cakes); and<br />
• can be incorporated into <strong>the</strong> lter cake to reduce <strong>the</strong> yield stress <strong>of</strong> <strong>the</strong> cake.<br />
In many uid systems, it was found that <strong>the</strong> most effective additives are those operating by<br />
more than one <strong>of</strong> <strong>the</strong> above mechanisms.<br />
LABORATORY TESTS AND RESULTS<br />
Laboratory tests <strong>of</strong> <strong>the</strong> effect <strong>of</strong> two commercially available lubricants L1 and L2, and new<br />
developed lubricant Horma-138 [7] on rheological, ltration and lubricating properties <strong>of</strong><br />
selected drill-in uids were carried out. They are water-dispersible lubricants designed to<br />
decrease <strong>the</strong> coefcient <strong>of</strong> friction in all water-base muds, reducing torque and drag. They<br />
also have a unique wettability characteristic which lowers <strong>the</strong> potential <strong>of</strong> BHA balling [7].<br />
Inuence <strong>of</strong> addition different concentrations (0%, 2%, 3%) <strong>of</strong> three lubricants on rheological<br />
and ltration properties in different drill-in uids were carried out according API<br />
standard using Fann viscometer and API lter press. The three unweighted, water based
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 43<br />
drill-in uid systems were tested. Fluids are designated as WBM 1, WBM 2 and WBM 3<br />
(Table 1).<br />
Lubricating properties were studied using <strong>the</strong> Lubricity Tester. Briey a steel test block that<br />
simulates <strong>the</strong> wall <strong>of</strong> <strong>the</strong> hole is pressed against <strong>the</strong> test ring by a torque arm. The torque is<br />
measured by <strong>the</strong> intensity <strong>of</strong> current that is required to turn <strong>the</strong> ring at a constant rotation<br />
when immersed in <strong>the</strong> uid that is tested. The API procedure (RP 13B) was followed for<br />
<strong>the</strong> tests, in laboratory tests <strong>the</strong> rotational velocity was 60 min-1 (60 rpm) and <strong>the</strong> force applied<br />
was 16,95 Nm (150 lb/inch). Results are given as a torque value that can vary from 0<br />
to 50. The lower <strong>the</strong> value, <strong>the</strong> better <strong>the</strong> lubricating properties. The coefcient <strong>of</strong> friction<br />
is calculated from torque reading using adequate formulas.<br />
Table 1. Effect <strong>of</strong> various lubricants on mud properties and lubricating ability<br />
It is obvious that adding <strong>of</strong> HORMA-138 lubricant has <strong>the</strong> greatest inuence on plastic<br />
viscosity and yield point in comparison to uid without lubricants. For all three lubricants<br />
<strong>the</strong>re is an signicant reduction <strong>of</strong> coefcient <strong>of</strong> friction (COF). Fig. 3 is an illustration <strong>of</strong><br />
<strong>the</strong> inuence <strong>of</strong> lubricant concentration on COF in selected water based muds.
44<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
Fig. 3. Efciency <strong>of</strong> selected lubricants in mud WBM 1<br />
The comparison <strong>of</strong> different lubricants inuence on lubricating properties <strong>of</strong> WBM is<br />
shown in Fig. 4. It can be seen that with 3% <strong>of</strong> HORMA-138 lubricant in WBM 1 and<br />
WBM 2 <strong>the</strong>re is a 60% reduction <strong>of</strong> COF. In WBM 3 because <strong>of</strong> original excellent lubricating<br />
properties <strong>the</strong>re is only 12% <strong>of</strong> COF reduction.<br />
Conclusions<br />
Fig. 4. Comparison <strong>of</strong> various lubricants impact on mud lubricating ability<br />
All used lubricants have certain but not important impact on rheological and ltration properties<br />
<strong>of</strong> tested uids, which is o<strong>the</strong>rwise desirable.<br />
By adding 2 to 3% <strong>of</strong> lubricant in WBM satisfactory reduction <strong>of</strong> COF is achieved.<br />
New developed lubricant HORMA-138 is in range <strong>of</strong> commercially available lubricants<br />
and is applicable for eld testing.<br />
Achieved COF reduction when implemented in force and horizontal reach calculations
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 45<br />
shows that for <strong>the</strong> same horizontal reach <strong>the</strong>re is <strong>the</strong> need for smaller axial loads on pipe,<br />
that means smaller amount <strong>of</strong> stresses imposed on pipe while drilling.<br />
Nomenclature<br />
E - Young’s modulus <strong>of</strong> elasticity, Pa<br />
e - natural logarithm base, 2,21828<br />
F b,max<br />
- maximal axial compressive load at <strong>the</strong> bottom <strong>of</strong> vertical section, N<br />
F eoc<br />
-axial load at <strong>the</strong> end <strong>of</strong> build section, N<br />
F n<br />
- axial load acting on element, N<br />
F kf<br />
- axial load at <strong>the</strong> kick<strong>of</strong>f point, N<br />
F t<br />
- axial tension acting at lower end <strong>of</strong> element, N<br />
H - vertical well bore depth, m<br />
I - moment <strong>of</strong> inertia <strong>of</strong> tubulars, m4<br />
L h<br />
- horizontal section length, m<br />
M - increase in torsion over length <strong>of</strong> element, N•m<br />
r - radial clearance between well bore and tubulars, m<br />
r p<br />
- characteristic radius <strong>of</strong> drill string element, m<br />
R - radius <strong>of</strong> well bore curvature, m<br />
w e<br />
- tubular weight in uid, N/m<br />
- friction factor<br />
- Ludolf’s number, 3,14<br />
References<br />
1. Wolfson L, Mensa-Wilmot G, Coolidge R. ( 1998) Systematic Approach Enhances<br />
Drilling Optimization and PDC Bit Performance in North Slope ERD Program, SPE<br />
50557, SPE Annual Technical Conference and Exhibition, New Orleans, 27-30 September,<br />
p. 1-12<br />
2. Wu J, Juvkam-Wold HC. (1993) Drilling and Completing Horizontal Wells With<br />
Coiled Tubing, SPE 26336, 68th Annual Conference and Exhibition, Houston, 3-6<br />
October, p. 221-233<br />
3. Johancsik CA, Friesen DB, Dawson R. (1984) Torque and Drag in Directional Wells-<br />
Prediction and Measurement, Journal <strong>of</strong> Petroleum Technology, June, p. 987-992<br />
4. Argillier J-F, Audibert A, Janssen M, Demoulin A. (1996) Development <strong>of</strong> a New<br />
Non-Polluting Ester Based Lubricant for Water Based Muds: Laboratory and Field<br />
Tests Results, SPE 36862, European Petroleum Conference, Milan, Italy, October 22-<br />
24, p. 401-410<br />
5. Bol GM. (1986) Effect <strong>of</strong> Mud Composition on Wear and Friction on Casing and Tool<br />
Joints, SPE Drilling Engineering, October, p. 369-376<br />
6. Evans N, Langlois B, Audibert-Hayet A, Dalmazzone C, Deballe E. High Performance<br />
Emulsiers for Syn<strong>the</strong>tic Based Muds, SPE 63101, SPE Annual Technical Conference<br />
and Exhibition, Dallas, TX, October 1-4, p. 1-11<br />
7. Gaurina-Meimurec N. (1998) Horizontal Well Drill-In Fluids, The Mining, Geological,<br />
Petroleum Engineering Bulletin, vol. 10, Zagreb, Croatia, p. 73-76
46<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
FUNDAMENTALS OF MILITARY PRODUCTION<br />
DEVELOPMENT IN THE CROATIAN WAR FOR<br />
INDEPENDENCE FROM 1991 TO 1993<br />
Dinko Mikulić<br />
University <strong>of</strong> Apllied Sciences Velika Gorica<br />
Zagrebačka cesta 5, 10410 Velika Gorica, Croatia<br />
e-mail: dinko.mikulic@vvg.hr<br />
Tel. 01/6230-761, Mob. 099 6222-503<br />
Vladimir Koroman<br />
Brodarski institut d.o.o.,<br />
Av. V. Holjevca 20, 10020 Zagreb, Croatia<br />
e-mail: vladimir.koroman@hrbi.hr<br />
Tel. 01/6504-401, Mob. 098 211-324<br />
Summary<br />
There were several phases in organizing <strong>the</strong> military production during <strong>the</strong> Homeland War<br />
from 1991 to 1993. In <strong>the</strong> rst phase, emphasis was placed on human creativity and top<br />
level improvisation in producing armoured vehicles and light arms. In <strong>the</strong> second stage, <strong>the</strong><br />
military armament and equipment production development fundamentals were established<br />
by forming The Sector for Development, Production and Scientic Research. In <strong>the</strong> third<br />
phase, a systematically organised massive arms and equipment military production and<br />
overhaul had begun by forming The Production Department. In this paper, based on an<br />
example <strong>of</strong> armoured vehicles development, <strong>the</strong> fundamentals <strong>of</strong> Croatian military production<br />
are explained, which fullled <strong>the</strong> military missions and reconciled <strong>the</strong> requests, thus<br />
serving <strong>the</strong> purpose <strong>of</strong> <strong>the</strong> country's liberation.<br />
Key words: armoured vehicles, development and production, Croatian War <strong>of</strong> Independence<br />
(Homeland war)<br />
Introduction<br />
The production <strong>of</strong> military equipment for Croatian National Guard (Zbor narodne garde,<br />
ZNG) soldiers began in <strong>the</strong> summer <strong>of</strong> 1991, a year <strong>of</strong> war. The equipment was produced<br />
manually in small and large plants as well as in workshops and garages. Firstly, trucks<br />
were armoured and given as present to <strong>the</strong> Croatian soldier who ghts for freedom. It was<br />
a help to <strong>the</strong> police in <strong>the</strong> rst place, <strong>the</strong>n to <strong>the</strong> Croatian National Guard. Such improvised<br />
vehicles gave protection from infantry weapons and enabled <strong>the</strong> completion <strong>of</strong> various<br />
logistic tasks, from food and ammunition transport to <strong>the</strong> transport <strong>of</strong> <strong>the</strong> wounded.<br />
Since Croatia didn't have any armoured vehicles at its disposal, excluding a part <strong>of</strong> <strong>the</strong><br />
police personnel carriers, <strong>the</strong>se handcrafted armoured vehicles in <strong>the</strong> hands <strong>of</strong> <strong>the</strong> brave<br />
Croatian Homeland War defenders as seen on TV screens, strongly raised people's morale<br />
and gave a huge impetus to faster production and o<strong>the</strong>r defence nances. Approximately<br />
fty rms participated in <strong>the</strong>ir production, mostly in Zagreb, <strong>the</strong>n in Split, Slavonski Brod,<br />
Rijeka etc.
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 47<br />
1 The beginning <strong>of</strong> organized production<br />
First it began in <strong>the</strong> Croatian Parliament and <strong>the</strong> Government <strong>of</strong> <strong>the</strong> Republic <strong>of</strong> Croatia.<br />
Direct contacts between <strong>the</strong> Headquarters <strong>of</strong> <strong>the</strong> military command and <strong>the</strong> economy resulted<br />
in <strong>the</strong> admission <strong>of</strong> company's leaders to <strong>the</strong> Croatian Parliament, as well as <strong>the</strong> Defence<br />
Minister, lieutenant general Martin Špegelj, and associates. Defence Minister Deputy<br />
V. Kikerec and Mr M. Hlad, a Parliament representative, accepted in <strong>the</strong>ir <strong>of</strong>ces persons<br />
and company requests for development and production <strong>of</strong> military equipment among<br />
which were also <strong>the</strong> <strong>of</strong>fers for <strong>the</strong> production <strong>of</strong> armoured vehicles.<br />
Having graduated from <strong>the</strong> Technical Military Academy from rnomerec in Zagreb, volunteers,<br />
military-technical experts <strong>of</strong>fered <strong>the</strong>ir pr<strong>of</strong>essional services to <strong>the</strong> Ministry <strong>of</strong><br />
Defence leaders. As pr<strong>of</strong>essional advisers, <strong>the</strong>y took over <strong>the</strong> coordination <strong>of</strong> development<br />
and production <strong>of</strong> various equipment. The development coordination at Zagreb-based<br />
companies TAZ, Hidroelektra-Mehanizacija, Autodubrava, Autoservis-Borongaj and o<strong>the</strong>rs<br />
soon followed as well as <strong>of</strong>fering pr<strong>of</strong>essional instructions for vehicle armouring, body<br />
armours, small individual weapons, etc.<br />
1.1 The sector for Development, Production and Scientific Research<br />
Systematic development <strong>of</strong> defence funding began by forming The Sector for Development,<br />
Production and Scientic Research. President <strong>of</strong> <strong>the</strong> Republic <strong>of</strong> Croatia Franjo<br />
Tuman, PhD, issued <strong>the</strong> Decision on <strong>the</strong> Appointment <strong>of</strong> Gojko Šušak as <strong>the</strong> Defence<br />
Minister on 16 September 1991, while Croatian Prime Minister Franjo Greguri, PhD,<br />
passed <strong>the</strong> Decision on <strong>the</strong> Appointment <strong>of</strong> Vladimir Volarevi, PhD, as <strong>the</strong> Defence Minister<br />
Deputy, who took over <strong>the</strong> organization <strong>of</strong> military equipment and armaments development<br />
and production and also formed The Sector for Development, Production and<br />
Scientic Research (temporary department at <strong>the</strong> Ministry <strong>of</strong> Defence).<br />
The rst staff <strong>of</strong> <strong>the</strong> just formed sector comprised military experts, mostly lecturers from<br />
<strong>the</strong> Technical Military Academy from rnomerec in Zagreb and experts from <strong>the</strong> Army<br />
Depot Mainentance Bregana, who as volunteers enlisted for Croatian military service, and<br />
who were competent for specic eld <strong>of</strong> military equipment and armaments development.<br />
These were <strong>the</strong> following Croatian army <strong>of</strong>cers:<br />
Dinko Mikuli, MSc, motor vehicles; Zdenko Matijaši, MSc, motor vehicles; Vjekoslav<br />
Stojkovi, PhD, motor vehicles; Mijo Vrhovski, PhD, motor vehicles; Simeon Kovaev,<br />
PhD, motor vehicles; Nikola Gambiroža, PhD, explosives; Mirjana Gambiroža Juki,<br />
PhD, explosives, Milan Prpi, MSc, explosives, Mladen Pleše, PhD, explosives; Mladen<br />
Barkovi, PhD, electronic devices; Artur Gorianin, MSc, electronic devices; Aljoša<br />
Božikovi, MSc, radars; Mirko Jukl, MSc, radars; Mihalj Strmeki, MSc, air defence;<br />
Vladimir Lebinac, MSc, electronic devices, Goran Prokopec, MSc, telecommunication;<br />
Ivan Pokaz, MSc; Mirko Kukolj, MSc, armaments; Milan Ivanuševi, MSc, NBC equipment;<br />
Josip Petrovi, electronic devices; Josip Vinter, overhaul; Vinko Turk, artillery<br />
weapons; Ivan Peica, air defence, Vitoš Bebi, <strong>engineering</strong>; Zdravko Paveli, armaments;<br />
Petar Radonji, navy.<br />
The cooperation was established between The Sector for Development, Production and<br />
Scientic Research <strong>of</strong> <strong>the</strong> Croatian Ministry <strong>of</strong> Defence and <strong>the</strong> Department <strong>of</strong> Energy <strong>of</strong><br />
<strong>the</strong> Croatian Ministry <strong>of</strong> Industry for <strong>the</strong> purpose <strong>of</strong> coordination and unobstructed work <strong>of</strong><br />
pr<strong>of</strong>essional advisers <strong>of</strong> <strong>the</strong> Ministry <strong>of</strong> Defence in companies.
48<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
1.2 The list <strong>of</strong> some companies working on <strong>the</strong> production <strong>of</strong> armoured vehicles<br />
Autodubrava, Zagreb, Brodosplit, Split, TAZ – Bus Factory Zagreb, Autoservis Borongaj,<br />
Zagreb, Hidroelektra-Niskogradnja-Mehanizacija, Zagreb, Industrogradnja, Zagreb,<br />
Tehnika, Zagreb, TPK, Zagreb, J.Gredelj, HŽP Zagreb, Jedinstvo, Zagreb, Tehnomehanika,<br />
Marija Bistrica, .akovi Specijalna vozila, Sl.Brod, elik/TVIL, Križevci, Torpedo,<br />
Rijeka, Rijekaprojekt, and o<strong>the</strong>rs.<br />
Production <strong>of</strong> some armoured vehicles, Semptember-December, 1991<br />
Key meetings at companies<br />
Key meetings <strong>of</strong> <strong>the</strong> representatives <strong>of</strong> The Sector for Development, Production and Scientic<br />
Research <strong>of</strong> <strong>the</strong> Croatian Ministry <strong>of</strong> Defence and <strong>the</strong> Department <strong>of</strong> Energy <strong>of</strong> <strong>the</strong><br />
Croatian Ministry <strong>of</strong> Industry with representatives <strong>of</strong> <strong>the</strong> rms which <strong>of</strong>fer <strong>the</strong> military<br />
production <strong>of</strong> armoured vehicles, were held as follows:<br />
Meeting in <strong>the</strong> factory Torpedo Rijeka, 1 October 1991<br />
Meeting in <strong>the</strong> factory . akovi, Slavonski Brod, 15 October 1991<br />
Meeting in <strong>the</strong> shipyard Brodosplit, Split, 4 December 1991<br />
The basic manufacturers' questions were about <strong>the</strong> necessary number <strong>of</strong> vehicles and <strong>the</strong><br />
way or <strong>the</strong>ir development as well as <strong>the</strong> possibility <strong>of</strong> project nancing. In <strong>the</strong> production<br />
<strong>of</strong> armoured personnel carrier based on <strong>the</strong> truck chassis, emphasis must be put on chas-
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 49<br />
sis standardization while for armouring <strong>the</strong> usage <strong>of</strong> steel plates (plates for ship) 8-10 mm<br />
thick. As a technical institute Brodarski institut <strong>of</strong>fered <strong>the</strong> help <strong>of</strong> its experts for <strong>the</strong> purpose<br />
<strong>of</strong> superstructure and certain systems development.<br />
Fig. 1. Armoured personnel carriers, Otoac, Gacka, Krešimir i Domagoj; Industrogradnja,<br />
Zagreb, 1991<br />
1.3 A letter to <strong>the</strong> producers <strong>of</strong> military equipment and armament, 10 November 1991<br />
In cooperation with <strong>the</strong> Ministry <strong>of</strong> Industry and after talks with potential armaments and<br />
military equipment manufacturers, Defence Minister Deputy V. Volarevi sent <strong>the</strong> manufacturers<br />
<strong>the</strong> following letter:<br />
The Republic <strong>of</strong> Croatia<br />
The Ministry <strong>of</strong> Defence<br />
Zagreb, 10 November 1991<br />
Class: 213-01/91-01/30<br />
Reg. no.: 512-08-91-1<br />
Armaments and military equipment production<br />
Please activate all available capacities for mass production <strong>of</strong> <strong>the</strong> assigned – arranged<br />
armaments and military equipment assortment for <strong>the</strong> needs <strong>of</strong> <strong>the</strong> armed battle and<br />
war operations. All produced quantities with <strong>the</strong> highest degree <strong>of</strong> achieved reliability –<br />
quality should be delivered to <strong>the</strong> commanders <strong>of</strong> <strong>the</strong> units in your territories, while <strong>the</strong><br />
necessary documentation should be made about <strong>the</strong> delivery.<br />
Defence Minister Deputy<br />
Vladimir Volarevi, PhD
50<br />
Typical example <strong>of</strong> Commands’ requests<br />
The Republic <strong>of</strong> Croatia<br />
Starigrad-Paklenica Command <strong>of</strong> <strong>the</strong> Croatian Army<br />
No.: 155-1-9/<br />
Date: 13 November 1991<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
The Ministry <strong>of</strong> Defence<br />
Fax 041 451-882<br />
Dinko Mikuli<br />
For <strong>the</strong> purposes <strong>of</strong> our command, please assign us one armoured vehicle with <strong>the</strong> following<br />
characteristics:<br />
- It must be protected against <strong>the</strong> shooting and cumulative effects<br />
- The possibility <strong>of</strong> mounting <strong>the</strong> machine-gun<br />
- Such vehicle is urgent to us<br />
Ivan Cerovac<br />
Commander<br />
1.4 Report on motor vehicles production and development, 20 November 1991, to <strong>the</strong><br />
Defence Minister Deputy<br />
1. The production <strong>of</strong> an armoured <strong>engineering</strong> vehicle Straško is running according to<br />
<strong>the</strong> delivery plan (15 delivered)<br />
2. The production <strong>of</strong> armoured personnel carriers in Brodosplit rms is nished. The<br />
works are nished at Hidroelektra, Tehnika-Bjelovar.<br />
3. Armour steel HPA-10 <strong>of</strong> various thicknesses is provided for vehicle armouring, placed<br />
in <strong>the</strong> vicinity <strong>of</strong> TAZ (Tvornica autobusa Zagreb)<br />
4. The production <strong>of</strong> safety car tires with lling mass, placed and examined in Hidroelektra-Mehanizacija,<br />
Žitnjak, Zagreb.<br />
5. The development <strong>of</strong> mine cleaner with disks, <strong>the</strong> examination <strong>of</strong> its effectiveness is<br />
under way<br />
6. The project <strong>of</strong> multi-purpose armoured vehicle is prepared in cooperation with Brodarski<br />
Institute Zagreb, on <strong>the</strong> TAM 150 T11 6x6 chassis<br />
7. Tactical-technical requirements are prepared for development <strong>of</strong> armoured light vehicle.<br />
8. Preparations for <strong>the</strong> possibility <strong>of</strong> cooperation on terrain vehicle are being made<br />
9. The project <strong>of</strong> shielded medical buses in TAZ is being prepared.<br />
10. The production <strong>of</strong> military truck on TAM 110 T7 4x 4 chassis is being prepared.<br />
After production <strong>of</strong> rst armoured vehicles, <strong>the</strong> manufacturers gained a valuable experience,<br />
so <strong>the</strong>y asked for new available armoured vehicles, trucks type TAM 260, TAM<br />
170, TAM 190 and similar. However, <strong>the</strong>re wasn't a sufcient number <strong>of</strong> such vehicles<br />
at disposal because a huge number <strong>of</strong> vehicles was mobilized, besides building companies<br />
which had <strong>the</strong> most trucks would be left without work resources. It is estimated that<br />
approximately 200 various armoured vehicles are handcrafted and <strong>the</strong>re were over 1000<br />
people working on <strong>the</strong>m, in more than 50 rms and workshops in <strong>the</strong> Republic <strong>of</strong> Croatia.
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 51<br />
2 Croatian Army requirements and organized production <strong>of</strong> armoured<br />
vehicles<br />
Considering <strong>the</strong> war situation, <strong>the</strong> Croatian Army required increasingly better military<br />
equipment, with better technical protection characteristics, with greater mobility and better<br />
possibility <strong>of</strong> combat activity as well as <strong>the</strong>ir supporting logistics, spare parts and intervention<br />
team. In this way <strong>the</strong> Croatian Army wanted to increase tactical and strategic mobility<br />
<strong>of</strong> <strong>the</strong> Croatian Army at <strong>the</strong> whole operation zone <strong>of</strong> <strong>the</strong> Republic <strong>of</strong> Croatia.<br />
Requirements for better equipment were based upon three important factors: armour protection,<br />
mobility and repower. Commanders demand <strong>the</strong> mobility <strong>of</strong> repower and this<br />
is possible only by means <strong>of</strong> armoured combat vehicles for it can lead to faster change in<br />
conducting combat operations. It was not enough anymore that <strong>the</strong> soldier could only be<br />
transported at <strong>the</strong> battleeld protected by armour, moreover repower from various arms<br />
was required as well as supporting technical logistics. Such vehicles from which infantry<br />
can lead <strong>the</strong> battle should be produced and leave <strong>the</strong> vehicle only in emergency case and<br />
reconnaissance. The basis <strong>of</strong> it can be multi-purpose light combat vehicles on wheels intended<br />
for various combat activities in <strong>the</strong> system <strong>of</strong> mechanized reconnaissance units,<br />
forces for fast activities and general support forces. Thus, armoured vehicles on wheels<br />
complement with armoured vehicles on tracks during tactical activity.<br />
Means <strong>of</strong> transport for troop transport, military trucks, ambulance vehicles, towing vehicles<br />
for MB-120 mm heavy mortars were required from <strong>the</strong> Croatian army as a priority,<br />
and <strong>the</strong>n requiring better armoured vehicles with <strong>the</strong> possibility to apply weapons. For <strong>the</strong><br />
purpose <strong>of</strong> vehicle standardization for <strong>the</strong> army multi-purpose vehicles are recommended<br />
in order to be used for various tasks.<br />
The Sector for Development, Production and Scientic Research intensies development<br />
and production for defence, which speeds up <strong>the</strong> implementation <strong>of</strong> projects for armoured<br />
vehicles and supporting logistics development.<br />
During <strong>the</strong> 1991 - 1993 period <strong>the</strong> following production <strong>of</strong> armoured vehicles, trucks and<br />
overhaul support was established and realized:<br />
- 120 mm Self-propelled mortar, year 1991<br />
- Military truck TK 130 T-7 4x 4 Torpedo, year 1992<br />
- Light Armoured Vehicle (LAV) 4x 4, year 1993<br />
- Army Maintenance Depot, Zagreb, year 1991<br />
2.1 Development and production <strong>of</strong> 120 mm self-propelled mortars<br />
It is estimated that <strong>the</strong> Croatian Army has at its disposal a sufcient number <strong>of</strong> terrain vehicles<br />
(TAM 150 T11 6x6) which can meet tactical-technical requirements <strong>of</strong> multi-purpose<br />
armoured vehicle production. The project was given to a scientic-research institution –<br />
Brodarski Institute Zagreb, which already has experts and experience in <strong>the</strong> development<br />
<strong>of</strong> complex military objects. It was <strong>the</strong> rst complex development project <strong>of</strong> <strong>the</strong> Ministry<br />
<strong>of</strong> Defence which was initiated with <strong>the</strong> scientic-research institution. The complete<br />
preparation <strong>of</strong> <strong>the</strong> project was led by pr<strong>of</strong>essional adviser in The Sector for Development,<br />
Production and Scientic Research, Dinko Mikuli, MSc, and <strong>the</strong> head <strong>of</strong> <strong>the</strong> Special Projects<br />
Department in Brodarski Institute, Vladimir Koroman, MSc. The implementing body<br />
– Brodarski Institute obliged to design, construct and deliver <strong>the</strong> prototype and establish<br />
<strong>the</strong> production <strong>of</strong> <strong>the</strong> multi-purpose armoured vehicle with a built-in MB-120 mm mortar<br />
and PAM 12.7 mm heavy machine-gun. There are three project phases:<br />
1. The construction <strong>of</strong> a functional model to full functionality
52<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
2. The construction <strong>of</strong> a prototype based on <strong>the</strong> approved functional model<br />
3. The organization <strong>of</strong> production based on <strong>the</strong> approved prototype<br />
Deadline for <strong>the</strong> completion <strong>of</strong> <strong>the</strong> 1st phase works was 10 December 1991. Within <strong>the</strong><br />
Institute, people specialising in various technical elds worked day and night in <strong>the</strong> preparation<br />
<strong>of</strong> <strong>the</strong> rst functional model and prototype. Technical documentation was prepared<br />
according to <strong>the</strong> simultaneous <strong>engineering</strong> principle (simultaneous phase overlap), regarding<br />
<strong>the</strong> armour design, as well as hydraulic suspension for <strong>the</strong> impact amortization, mortar<br />
tting and ammunition storage. The vehicle is designed for endurance according to possible<br />
impacts <strong>of</strong> <strong>the</strong> MB-120 mm heavy mortar, produced by “Prvomajska-Alatni strojevi<br />
/ INAS”- Žitnjak.<br />
Several manufacturers participated in <strong>the</strong> 1st and 2nd development phase: TAZ Zagreb(for<br />
armour), Strojogradnja Samobor (for hydraulics), Rade Konar – special devices and systems<br />
(laser cutting <strong>of</strong> sheet metal), Obrada Rovinj (mortar chassis at <strong>the</strong> vehicle), Elmech<br />
(inert mines – mines for testing) and o<strong>the</strong>r manufacturers. The testing <strong>of</strong> <strong>the</strong> self-propelled<br />
mortar functional model was carried out on 22 November 1991 on Velešec polygon in <strong>the</strong><br />
vicinity <strong>of</strong> Velika Gorica.<br />
SoMB-120 mm self-propelled mortar model was placed on hydraulic props and levelled.<br />
Recording equipment was connected as well. The usage <strong>of</strong> mortars was carried out by a<br />
group <strong>of</strong> military <strong>of</strong>cers <strong>of</strong> <strong>the</strong> General Staff <strong>of</strong> <strong>the</strong> Croatian Armed Forces. 9 inert mines<br />
were red. The vehicle was checked and <strong>the</strong>re weren’t any deformations on it. With <strong>the</strong><br />
change <strong>of</strong> functional model to prototype, a prototype was completely nished and prepared<br />
for nal testing on endurance and accuracy <strong>of</strong> targets shooting, with 99 red inert mines.<br />
Fur<strong>the</strong>r testings on mortar endurance and accuracy showed <strong>the</strong> correctness <strong>of</strong> suspension<br />
solution and <strong>the</strong> way <strong>of</strong> re control system. At <strong>the</strong> beginning <strong>of</strong> 1992, <strong>the</strong> prototype was<br />
given to <strong>the</strong> 108th Brigade ZNG, Slavonski Brod. A self-propelled mortar prototype was<br />
used during <strong>the</strong> Homeland War.<br />
According to <strong>the</strong> 3 rd phase <strong>of</strong> <strong>the</strong> contract - <strong>the</strong> organization <strong>of</strong> production based on <strong>the</strong> approved<br />
prototype – two more SoMB-120 mm self-propelled mortars were produced in TAZ<br />
factory according to <strong>the</strong> documentation and supervision <strong>of</strong> works by Brodarski Institut.<br />
Note:<br />
In time <strong>of</strong> completion <strong>of</strong> self-propelled mortar production, Ivan ermak was appointed<br />
as <strong>the</strong> Defence Minister Deputy for logistics and forms Logistics Sector which includes<br />
several departments: Department for Production, which takes over <strong>the</strong> organization <strong>of</strong> military<br />
production (from The Sector for Development, Production and Scientic Research),<br />
Procurement Department, Technical and Trade Department, Construction Department and<br />
Telecommunications Department. Colonel Nikola Gambiroža, PhD, was appointed as<br />
Head <strong>of</strong> The Department for Production on 9 January 1992.<br />
Both just produced SoMB-120 mm self-propelled mortars were delivered to <strong>the</strong> Croatian<br />
Army on 7 January 1991 before Operation ''Maslenica'' in which self-propelled mortars<br />
were successfully used in liberating operation <strong>of</strong> Zadar hinterland.<br />
The contribution <strong>of</strong> Brodarski Institute to <strong>the</strong> development <strong>of</strong> Croatian Army equipment<br />
and armaments, among which was also SoMB self-propelled mortar, was shown when<br />
Croatian President Franjo Tuman and Minister <strong>of</strong> Defence Gojko Šušak came to visit<br />
Brodarski Institute on 31 March 1993.
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 53<br />
Fig.2. Functional model <strong>of</strong> SoMB 120 mm self-propelled mortar with 12.7 mm,<br />
Kiseli, Brodarski Institut, Zagreb, 1991<br />
Fig.3. The testing <strong>of</strong> SoMB 120 mm self-propelled mortar prototype, Velešec, 1992<br />
2.2 Development and production <strong>of</strong> military trucks TK 130 T-7 4x4 Torpedo<br />
The idea <strong>of</strong> military truck production based on TAM 110 T7 4x 4 chassis fell on fertile<br />
ground in <strong>the</strong> tractor and construction machines factory Torpedo, Rijeka. After <strong>the</strong> production<br />
<strong>of</strong> armoured <strong>engineering</strong> vehicle Straško/HIAV on <strong>the</strong> universal excavator chassis, <strong>the</strong><br />
management board decided to make preparations for <strong>the</strong> domestic production <strong>of</strong> military<br />
trucks in cooperation with <strong>the</strong> factory TAM Maribor. For <strong>the</strong> purpose <strong>of</strong> domestic production<br />
share increase, Boženko Vukovi, MSc and Zdenko Novak, MSc, reconstructed<br />
<strong>the</strong> cabin and cargo crate, which was completely produced in Progres rm-Jastrebarsko.<br />
Nedjeljko Ški, MSc, proposed <strong>the</strong> tting <strong>of</strong> 130 HP. domestic diesel-engine Torpedo,<br />
which would be technically and logistically better than 110 HP engine. Feasibility study<br />
justied <strong>the</strong> truck domestic production. A new manufacturing department for vehicle production<br />
was prepared within <strong>the</strong> factory. In <strong>the</strong> production system Director Neven Dedi<br />
and Head <strong>of</strong> Procurement Department Marelo Jurman worked out a plan for mounting
54<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
parts procurement, thus <strong>the</strong> production in <strong>the</strong> new manufacturing department could start.<br />
A part <strong>of</strong> <strong>the</strong> chassis (frame, axle, gear box) must have been imported from <strong>the</strong> manufacturer<br />
TAM Maribor, while everything else was manufactured in Croatia (diesel engine<br />
Torpedo, <strong>the</strong> cabin, cargo crate, tarpaulin, o<strong>the</strong>r parts). Thus, vehicle price around 50% <strong>of</strong><br />
purchase price <strong>of</strong> <strong>the</strong> similar imported vehicle was acceptable. A domestic share <strong>of</strong> approximately<br />
60 % <strong>of</strong> product price was relevant. Product Quality Regulation was made, while<br />
all vehicles before <strong>the</strong> delivery passed all necessary polygon tests on Šljunara polygon,<br />
under <strong>the</strong> Grobnik automobile race-track. The vehicles had provided logistical support, a<br />
2-year guarantee and spare parts for <strong>the</strong> next ten-year-period. It was a successful decision<br />
in truck production, according to <strong>the</strong> guidelines <strong>of</strong> war industry, <strong>the</strong> reduction <strong>of</strong> costs and<br />
vehicle standardization in <strong>the</strong> Croatian Army.<br />
First ''0'' series vehicles comprising 20 trucks were delivered at <strong>the</strong> end <strong>of</strong> 1992, <strong>the</strong>n series<br />
comprising 200 trucks were delivered in 1993-1994 period. It provided <strong>the</strong> tactical and<br />
strategic mobility <strong>of</strong> <strong>the</strong> Croatian Army during <strong>the</strong> Homeland War. In this time Torpedo<br />
TK-130 4x 4 military vehicles could be mostly seen on Croatian roads. This means that <strong>the</strong><br />
production <strong>of</strong> Torpedo TK-130 4x 4 trucks with <strong>the</strong> capacity <strong>of</strong> 3 tons(it can accommodate<br />
12 soldiers with complete equipment) was a very important factor <strong>of</strong> <strong>the</strong> Croatian Army<br />
capability during <strong>the</strong> Homeland War. Besides, a series production <strong>of</strong> military trucks proves<br />
that <strong>the</strong> cooperation with foreign partners on <strong>the</strong> basis <strong>of</strong> economical protability is possible<br />
even during <strong>the</strong> laid embargo.<br />
2.3 Development and production <strong>of</strong> light armoured vehicles LOV 4x4<br />
On <strong>the</strong> basis <strong>of</strong> good experience <strong>of</strong> ‘’0’’ series military truck production, an initiative was<br />
accepted that <strong>the</strong> manufacturer Torpedo <strong>of</strong>fers <strong>the</strong> solution <strong>of</strong> <strong>the</strong> Light Armoured Vehicle<br />
on <strong>the</strong> same TK 130 T7 4x 4 vehicle chassis. A huge degree <strong>of</strong> unication <strong>of</strong> both products<br />
was achieved in this way, as well as <strong>the</strong> low price <strong>of</strong> development and production. After<br />
that, appropriate tactical-technical requirements for light armoured vehicles (LOV 4x4)<br />
were prepared. The weight <strong>of</strong> 6.5 tons light armoured vehicles will be propelled by an<br />
intensied 150 HP diesel engine, with <strong>the</strong> maximum vehicle speed 120 km/h. The vehicle<br />
crew is made <strong>of</strong> 2 + 8 soldiers.<br />
Two items <strong>of</strong> ''0'' series LOV OP are developed in 1992. Main engineer Zdenko Novak,<br />
MSc, designed a modern form <strong>of</strong> Armoured Combat Vehicle. The vehicle height was below<br />
2 metres (1860 mm). The front part <strong>of</strong> <strong>the</strong> vehicle was well designed, <strong>the</strong> door construction<br />
for fast and easy entering and exiting <strong>of</strong> <strong>the</strong> driver, <strong>the</strong>n for fast entering and exiting <strong>of</strong> <strong>the</strong><br />
crew. The power <strong>of</strong> diesel engine was increased to150 HP, so that <strong>the</strong> vehicle got more<br />
available power for terrain mobility. In cooperation with experts from Brodarski Institut,<br />
contemporary methods were used to check <strong>the</strong> rmness <strong>of</strong> <strong>the</strong> armour with regard to vibrations.<br />
The capacity <strong>of</strong> LOV OP include 2 tons artillery and 8 soldiers.<br />
The main protection <strong>of</strong> Light Armoured Vehicles is <strong>the</strong> armour made <strong>of</strong> special steel which<br />
protects <strong>the</strong> crew from all types 7.62 x 51 AP (NATO protection) ammunition which would<br />
be red from <strong>the</strong> distance <strong>of</strong> 30 metres as well as <strong>the</strong> 155 mm shell fragments. 6-8-9 mm<br />
thick Armox special steel plates are applied as well as <strong>the</strong> safety glass. Due to <strong>the</strong> greater<br />
re exposure <strong>of</strong> <strong>the</strong> armoured surfaces, <strong>the</strong> vehicle got <strong>the</strong> greater armour thickness. On <strong>the</strong><br />
bottom side <strong>of</strong> <strong>the</strong> vehicle an additional two-layered antitank mines protection was made.<br />
Trial factory testings <strong>of</strong> <strong>the</strong> Light Armoured Vehicles are carried out on Šljunara polygon,<br />
above Rijeka. The representatives <strong>of</strong> <strong>the</strong> factory, <strong>the</strong> Ministry <strong>of</strong> Defence, <strong>the</strong> Faculty <strong>of</strong><br />
Mechanical Engineering and Naval Architecture and Brodarski Institute were present. The<br />
vehicle demonstration was very good. It is dened that <strong>the</strong> braking regulator must be tted<br />
on <strong>the</strong> rear axle, Run Flat tires must be provided and making <strong>the</strong> opening <strong>of</strong> <strong>the</strong> ro<strong>of</strong> doors
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 55<br />
and rotation <strong>of</strong> turret easier.<br />
The nal testing <strong>of</strong> light armoured vehicles as a type LOV IZV 4x4 with <strong>the</strong> 12.7 mm<br />
heavy machine-gun was carried out on polygon <strong>of</strong> Jastrebarsko barracks. During <strong>the</strong> testing<br />
Chief <strong>of</strong> General Staff <strong>of</strong> Armed Forces Anton Tus and representatives <strong>of</strong> Armoured<br />
units <strong>of</strong> <strong>the</strong> Croatian Army were present and very satised with <strong>the</strong> possibilities <strong>of</strong> barriers<br />
overcoming, especially cross water-trench capability as well as <strong>the</strong> speed <strong>of</strong> entering and<br />
exiting <strong>of</strong> <strong>the</strong> soldiers from <strong>the</strong> armoured vehicles.<br />
Fig. 4. The testing <strong>of</strong> LOV 4x4 OP prototype, Jastrebarsko, 13 February 1993.<br />
Fig. 5. Testing <strong>of</strong> LOV RAK 24/128 mm prototype, Zeevo, Šibenik, 1993.
56<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
After <strong>the</strong> operational capacity <strong>of</strong> <strong>the</strong> armoured vehicle was shown, a serial quantity <strong>of</strong> different<br />
types <strong>of</strong> vehicles was arranged. The delivery <strong>of</strong> <strong>the</strong>se vehicles increased <strong>the</strong> tactical<br />
and strategic capability <strong>of</strong> <strong>the</strong> Croatian Army and its readiness to reject <strong>the</strong> enemy forces<br />
and end <strong>the</strong> war. Light Armoured Vehicle was developed in <strong>the</strong> following types: Armoured<br />
Personnel Carrier (LOV OP 4x4), Reconnaissance Vehicle (LOV IZV 4x4), Electronic<br />
Warfare Vehicle (LOV ED 4x4), Command Vehicle (LOV Z 4x4), Multiple rocket launcher<br />
(LOV RAK 24/128 4x4), Nuclear-Biological Vehicle (LOV ABK 4x4) and Reconnaissance<br />
patrol and artillery re control system (LOV UP).<br />
12.7 mm air-defence machine-gun is a part <strong>of</strong> <strong>the</strong> standard equipment <strong>of</strong> all Light Armoured<br />
Vehicles. The Reconnaissance Vehicle for electronic warfare uses an additional<br />
armaments – 20 mm RT-20 hand gun and 8-tube launcher system Obad (8 RL 60 M93),<br />
60 mm calibre, with range <strong>of</strong> 8000 m. In order to increase artillery repower or artillery<br />
rocket combat a type <strong>of</strong> RAK 24/128 Light Armoured Vehicle is developed. This vehicle is<br />
equipped with <strong>the</strong> multiple rocket launcher with 28 tubes <strong>of</strong> 128 mm diameter. The rocket<br />
range is 8.550 m, up to 13.500 m. A prototype LOV-2 as an improved type <strong>of</strong> previous series<br />
is developed. It comprised covered winch on <strong>the</strong> front part, easier vehicle conducting,<br />
better brakes, better electronic equipment, telecommunications system and working area.<br />
A complex development project <strong>of</strong> military LOV 4x4 Torpedo linked 4 important factors:<br />
construction, technical, conducting and economic. It gave its logistic identity. Logistic <strong>engineering</strong><br />
requirements were set already in <strong>the</strong> 1st phase <strong>of</strong> tactical-technical requirements<br />
setting so that <strong>the</strong> total logistic support for lime time is assessed through <strong>the</strong> ’’iceberg’’<br />
costs. Production and service price <strong>of</strong> such vehicles is more acceptable than <strong>the</strong> price <strong>of</strong><br />
such imported vehicles. The price <strong>of</strong> <strong>the</strong> basic LOV OP 4x4 is 6 times lower than <strong>the</strong> <strong>of</strong>fered<br />
price <strong>of</strong> basic Piranha 6x6 armoured vehicle, which is <strong>the</strong> pro<strong>of</strong> <strong>of</strong> <strong>the</strong> possible costeffectiveness<br />
<strong>of</strong> <strong>the</strong> fur<strong>the</strong>r production.<br />
This project <strong>of</strong> <strong>the</strong> original LOV design, low contour and domestic production proved <strong>the</strong><br />
sufcient domestic pr<strong>of</strong>essional knowledge and it showed that <strong>the</strong> army can rely on <strong>the</strong><br />
domestic science and production in dealing with <strong>the</strong> complex military systems in <strong>the</strong> hardest<br />
war conditions.<br />
2.4 Army Maintenance Depot<br />
Staff <strong>of</strong> The Sector for Development, Production and Scientic Research examined <strong>the</strong><br />
disposable halls used for manufacture in Jedinstvo company in Jankomir in Zagreb and<br />
prepared <strong>the</strong> necessary documentation on maintenance realization and <strong>the</strong> overhaul <strong>of</strong> <strong>the</strong><br />
Croatian Army military technology. Colonel Mijo Vrhovski, Col. Josip Vinter, Zdenko<br />
Matijaši, Col. Vjekoslav Stojkovi, Col. Josip Petrovi and o<strong>the</strong>r military <strong>of</strong>cers <strong>of</strong> <strong>the</strong><br />
Croatian Army were working on this project.<br />
A unit <strong>of</strong> army maintenance depot support was formed on <strong>the</strong> basis <strong>of</strong> this project on 14<br />
October 1991, rst for <strong>the</strong> overhaul <strong>of</strong> armoured means. The development <strong>of</strong> technical support<br />
system for <strong>the</strong> development <strong>of</strong> armoured vehicles began in this way, <strong>the</strong>n began <strong>the</strong><br />
formation <strong>of</strong> <strong>the</strong> 310th logistic support brigade.<br />
The 310th logistic support brigade signicantly contributed to <strong>the</strong> development and<br />
maintenance <strong>of</strong> military equipment and armaments. Among o<strong>the</strong>r things, it maintained<br />
<strong>the</strong> armoured vehicles <strong>of</strong> different manufacturers and it gave help at <strong>the</strong> lowlands elds<br />
throughout <strong>the</strong> Republic <strong>of</strong> Croatia. Army Maintenance Depot, as a part <strong>of</strong> <strong>the</strong> Logistics<br />
Command, became a unit <strong>of</strong> <strong>the</strong> Armed Forces responsible for <strong>the</strong> implementation <strong>of</strong> <strong>the</strong><br />
overhaul highest level (<strong>the</strong> 3rd level) <strong>of</strong> armaments means, Croatian Army military equipment<br />
and it is also a support to <strong>the</strong> units <strong>of</strong> <strong>the</strong> Croatian Armed Forces which take part in<br />
<strong>the</strong> international peacekeeping missions.
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 57<br />
3 Conclusion<br />
During <strong>the</strong> Serbian aggression against <strong>the</strong> independent Republic <strong>of</strong> Croatia, people’s motivation<br />
and organized directing <strong>of</strong> equipment and armaments projects enabled <strong>the</strong> production<br />
<strong>of</strong> improvised combat means. The units <strong>of</strong> <strong>the</strong> Croatian Army required <strong>the</strong>se vehicles<br />
in <strong>the</strong> rst 3 years <strong>of</strong> <strong>the</strong> Homeland war. The rst quantities <strong>of</strong> <strong>the</strong> armoured personnel<br />
carriers justied <strong>the</strong>ir purpose. Regarding <strong>the</strong> direct commanders' demand and soldiers' information<br />
from <strong>the</strong> eld, <strong>the</strong> Croatian Army was satised. Around 1000 people from many<br />
rms and workshops took part in <strong>the</strong> handcrafted production <strong>of</strong> 200 armoured vehicles.<br />
Requirements for military equipment caused <strong>the</strong> need for war industry directing in order to<br />
ensure <strong>the</strong> logistics <strong>of</strong> Croatian Army equipment in a faster and more rational way. Apart<br />
from <strong>the</strong> protection <strong>of</strong> people from infantry weapons using <strong>the</strong> sandwich armour, <strong>the</strong>se<br />
vehicles could not provide <strong>the</strong> necessary mobility <strong>of</strong> repower. Therefore, <strong>the</strong> development<br />
projects <strong>of</strong> contemporary armoured vehicles with armaments on chassis <strong>of</strong> <strong>the</strong> highly<br />
mobile military trucks, SoMB-120 (6x6) self-propelled mortars and Light Armoured Vehicles<br />
(LOV 4x4) were initiated. The developed domestic armoured vehicles were used<br />
in <strong>the</strong> three crucial country liberating operations ’’Maslenica ’’, ’’Bljesak ’’ and ’’Oluja ’’.<br />
Their usage in <strong>the</strong> hands <strong>of</strong> <strong>the</strong> Croatian Army soldiers, <strong>the</strong>ir operation and <strong>the</strong>ir production<br />
is carried out with regard to military tasks as well as in view <strong>of</strong> foreign contemporary<br />
solutions. A series production <strong>of</strong> military trucks proves that <strong>the</strong> cooperation with foreign<br />
partners is possible even during <strong>the</strong> embargo.<br />
In <strong>the</strong> end, <strong>the</strong>re were several phases in organizing <strong>the</strong> military production during <strong>the</strong><br />
Homeland War from 1991 to 1993. In <strong>the</strong> rst phase, emphasis was placed on human creativity<br />
and top level improvisation in producing armoured vehicles and light arms. In <strong>the</strong><br />
second stage, <strong>the</strong> military armament and equipment production development fundamentals<br />
were established by forming The Sector for Development, Production and Scientic Research.<br />
In <strong>the</strong> third phase, a systematically organised massive arms and equipment military<br />
production and overhaul had begun by forming The Production Department. In this paper,<br />
based on an example <strong>of</strong> armoured vehicles development, <strong>the</strong> fundamentals <strong>of</strong> Croatian<br />
military production are explained, which fullled <strong>the</strong> military missions and reconciled <strong>the</strong><br />
requests, thus serving <strong>the</strong> purpose <strong>of</strong> <strong>the</strong> country's liberation.<br />
4 References<br />
1. Ivkovi Z (2001) Hrvatski oklopnjaci 1991.-1993.: Ministarstvo obrane Republike<br />
Hrvatske, Zagreb, 2001.<br />
2. Poster (1996) Runo izraena oklopna vozila Hrvatske vojske 1991-1992: Galerija<br />
Zvonimir, Zagreb, 1996.<br />
3. Gambiroža N (1992) Proizvedeno u Hrvatskoj: Hrvatski vojnik, br. 17/92, Zagreb,1992.<br />
4. Bandula D (1997) Hrvatska vojna industrija: Velebit, 31.12.1997, Zagreb.<br />
5. Mallet J (1998) Croatian tanks and rocket launchers hits <strong>of</strong> Eurosatory 1998: DAILY,<br />
Janes Defence Weekly, Le Bourget, Pariz, 1998.<br />
6. Mikuli D (1991/1992) Kolegij blok: vlastite bilješke, Zagreb, 1992.<br />
7. Mikuli D (1991/1992) Oklopljena vozila: slikovni album, Zagreb, 1992.<br />
8. Mikuli D (1992/1993) Oklopna borbena vozila: slikovni album, Zagreb, 1993.
58<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
Intelligent Forest Fire Monitoring System – from idea<br />
to realization<br />
Darko Stipaničev<br />
Department for Modeling and Intelligent Systems & Center for Wildfire Research<br />
University <strong>of</strong> Split Faculty <strong>of</strong> Electrical Engineering,<br />
Mechanical Engineering and Naval Architecture<br />
R.Boškovića 32, 21000 Split, Croatia<br />
darko.stipnaicev@fesb.hr<br />
Abstract<br />
Wildres, including forest res, are natural phenomena that cause signicant economic<br />
damage and have quite devastating effect to <strong>the</strong> environment all over <strong>the</strong> world. Early re<br />
detection on one side and quick and appropriate intervention on <strong>the</strong> o<strong>the</strong>r one, are <strong>of</strong> vital<br />
importance for wildre damage minimization. The re season 2003 was quite severe one,<br />
particularly in Split and Dalmatia County. Provoked by great damages caused by 2003<br />
wildres in autumn 2003 a project was initiated at University <strong>of</strong> Split Faculty <strong>of</strong> Electrical<br />
Engineering, Mechanical Engineering and Naval Architecture Department for Modeling<br />
and Intelligent Systems, having <strong>the</strong> main goal to nd a way how advanced information –<br />
communication technologies (ICT) could be used to improve wildre prevention and protection.<br />
After three years <strong>of</strong> research and testing in 2006 <strong>the</strong> Intelligent Forest Fire Monitoring<br />
System called iForestFire was presented. iForestFire belongs to <strong>the</strong> last generation<br />
<strong>of</strong> wildre monitoring and surveillance systems, having a lot <strong>of</strong> innovative and advanced<br />
features. Since 2006 it has been applied in various Croatian national and nature parks, but<br />
also <strong>the</strong> whole Istria region is covered by iForestFire network having 29 monitoring stations<br />
and 7 operational centers. The paper describes main features <strong>of</strong> iForestFire system<br />
with emphasize on its development and implementation.<br />
Key words: wildre, wildre detection, smoke detection, intelligent system<br />
1 Introduction<br />
Wildre or wildland re is any uncontrolled burning <strong>of</strong> natural vegetation (grass, shrub,<br />
forest timber, litter and slash) in <strong>the</strong> wilderness area. When <strong>the</strong> vegetation layer is more<br />
precisely known, <strong>the</strong> more specic names could be used like forest re, grass re or bush-<br />
re. Wildres represent a constant threat to ecological systems, infrastructure and human<br />
lives. According to prognoses, wildres, including re clearing in tropical rain forest, will<br />
halve <strong>the</strong> world forest stand by <strong>the</strong> year 2030. In Europe, up to 10,000 km2 <strong>of</strong> vegetation is<br />
destroyed by res every year, and up to 100,000 km2 in North America and Russia. Wild-<br />
res are responsible for approximately 20% <strong>of</strong> CO2 emission into <strong>the</strong> atmosphere (Kührt<br />
et al, 2001).<br />
Croatia belongs to countries with high wildre risk. In summer season, seven coastal<br />
counties in Croatia, including in particular <strong>the</strong> Adriatic islands, are permanently exposed<br />
from high to very high re risks. This is due to meteorological conditions, densely spaced<br />
conifer forests and a lot <strong>of</strong> tourists. According to Croatian Forests data (Žaja, 2008) from
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 59<br />
1992 to 2007 <strong>the</strong>re were 4.851 wildres in Croatia and <strong>the</strong> burning area was 251.910 ha.<br />
Fire seasons 2000 and 2003 were particularly severe with 706 and 532 wildres and <strong>the</strong><br />
total damage caused by wildres was huge. For example in 2003 only in Split and Dalmatia<br />
County <strong>the</strong>re were 130 wildres, <strong>the</strong> total burned area was 9.700 ha, <strong>the</strong> direct damage<br />
caused by wildres (reghters interventions and post-re terrain recovery) was 16 mil.€,<br />
and <strong>the</strong> indirect damage, taking into account energetic equivalent <strong>of</strong> lost woody biomass,<br />
was assessed to 66 mil.€ (Stipaniev et al., 2004, Stipaniev et al. 2007). Wildres in<br />
2003 were particularly catastrophic on Split and Dalmatia County islands. For example,<br />
on islands Hvar and Bra <strong>the</strong> burning area was between 1/4 and 1/3 <strong>of</strong> total islands area<br />
and <strong>the</strong> small island Biševo near Vis has been totally burnt. Figure 1 shows photos <strong>of</strong> few<br />
2003 res on Split and Dalmatia County islands photographed by <strong>the</strong> author <strong>of</strong> this paper.<br />
Figure 1 – Wildres on Dalmatian islands in 2003<br />
Provoked by great damages caused by 2003 wildres in autumn 2003, a project was initiated<br />
at Department for Modeling and Intelligent Systems University <strong>of</strong> Split Faculty <strong>of</strong><br />
Electrical Engineering, Mechanical Engineering and Naval Architecture, having <strong>the</strong> main<br />
goal to nd a way how advanced information – communication technologies (ICT) could<br />
be used to improve wildre prevention and protection. After detailed survey <strong>of</strong> research<br />
topics and implementation <strong>of</strong> ICT systems in wildre prevention task in o<strong>the</strong>r countries<br />
subjected to wildres, we have focused ourselves primarily on <strong>the</strong>se three topics:<br />
- Automatic early wildre detection,<br />
- Calculation <strong>of</strong> micro location wildre risk index, and<br />
- Wildre behavior and propagation simulation.<br />
The rst topic is <strong>the</strong> most important one, because <strong>the</strong> only effective way to minimize damage<br />
caused by wildres is wildre early detection and fast and appropriate reaction, apart<br />
from preventive measures. Great efforts are <strong>the</strong>refore made to achieve early wildre detection,<br />
which is traditionally based on human wildre surveillance, realized by 24 hours observation<br />
by human observers located on selected monitoring spots. In Croatia <strong>the</strong> human<br />
wildres surveillance is mainly organized by Croatian Forests – <strong>the</strong> governmental organization<br />
responsible for protection and exploitation <strong>of</strong> forests in state ownership. Human observers<br />
are usually equipped only with standard binoculars and communication equipment<br />
and <strong>the</strong>ir observation area is only <strong>the</strong> area covered by <strong>the</strong>ir sight <strong>of</strong> view.<br />
A ra<strong>the</strong>r new, technically more advanced approach to human wildre surveillance is video<br />
cameras based human wildre surveillance and monitoring when remotely controlled<br />
video cameras are installed on monitoring spots, and <strong>the</strong> human observer is located in <strong>the</strong><br />
observation center. Such system could be used not only for early re detection, but also for
60<br />
distant video presence on re remote location.<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
Figure 2 – The difference between human wildre surveillance (left) and video camera based<br />
human wildre surveillance (right)<br />
The video cameras based human wildres surveillance has many advantages in comparison<br />
to direct human observation from monitoring spots. Few <strong>of</strong> <strong>the</strong>m are:<br />
- A wider area could be covered, because one human observer could monitor few video<br />
monitoring eld units.<br />
- Cameras are usually equipped with power zoom (optical zoom with 22 x magnication)<br />
so <strong>the</strong> observer could easily inspect suspected areas.<br />
- System usually has video storing capabilities, at least for <strong>the</strong> last couple <strong>of</strong> days, and that<br />
is quite useful for post-re analysis.<br />
The main limitation <strong>of</strong> video cameras based human surveillance is that re detection depends<br />
entirely on <strong>the</strong> human observation. The observer is located in more comfortable environment,<br />
<strong>the</strong> observation center, but he (or she) has to carefully watch multiple computer<br />
monitors at <strong>the</strong> same time, so problems like fatigue, boredom and loss <strong>of</strong> concentration<br />
could be encountered. That was <strong>the</strong> main reason for introduction <strong>of</strong> various forms <strong>of</strong> automatic<br />
and advanced automatic wildre surveillance and monitoring systems and networks.<br />
The system described in this paper, named iForestFire ® – Intelligent Forest Fire Monitoring<br />
System (Croatian name is IPNAS ® – Inteligentni Protupožarni NAdzorni Sustav) (iForest-<br />
Fire, <strong>2011</strong>), belongs to <strong>the</strong> last generation <strong>of</strong> advanced automatic wildre surveillance and<br />
monitoring systems, having a lot <strong>of</strong> innovative and advance features. It was entirely developed<br />
at University <strong>of</strong> Split Faculty <strong>of</strong> Electrical Engineering, Mechanical Engineering and<br />
Naval Architecture, so iForestFire is a good example <strong>of</strong> university – industry cooperation,<br />
because today it is a successful commercial product installed in various Croatian national<br />
and nature parks and Istria County, but also a product that have signicant export opportunities.<br />
In <strong>the</strong> rest <strong>of</strong> this paper, a short introduction to automatic wildre monitoring and surveillance<br />
systems will be given, main features <strong>of</strong> iForestFire system will be described with<br />
emphasis to its advanced and innovative aspects and nally a short discussion about iForestFire<br />
development, implementation and commercialization will be added.<br />
2 Automatic wildfire surveillance and monitoring system<br />
The research and system development in <strong>the</strong> area <strong>of</strong> automatic wildre surveillance system<br />
was extended in <strong>the</strong> last couple <strong>of</strong> years. There are two main types <strong>of</strong> automatic wildre
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 61<br />
surveillance systems: satellite systems based on satellite wildre monitoring and terrestrial<br />
systems based on wildre monitoring from ground monitoring stations. In this moment<br />
terrestrial systems are more useful because early re detection is not <strong>the</strong> only one task that<br />
contemporary automatic wildre surveillance and monitoring system has to fulll. Beside<br />
<strong>the</strong> automatic detection <strong>of</strong> wildres, <strong>the</strong> distance video presence on re remote location is<br />
almost <strong>of</strong> equal importance and today’s satellite systems could not fulll that task. Satellite<br />
systems could be used only for wildre detection with limited spatial and time resolution.<br />
Therefore <strong>the</strong> automatic wildre surveillance and monitoring ground-based or terrestrial<br />
system used as an enhancement <strong>of</strong> human based wildre surveillance are dominant today.<br />
In ground-based or terrestrial systems different kinds <strong>of</strong> re detection sensors could be<br />
used:<br />
- Video cameras sensitive in visible spectra where wildre detection is based on smoke<br />
recognition during <strong>the</strong> day and re ames recognition during <strong>the</strong> night.<br />
- Infrared (IR) <strong>the</strong>rmal imaging cameras where wildre detection is based on detection <strong>of</strong><br />
heat ux from <strong>the</strong> re (Arrue et al, 2000).<br />
- Optical spectrometry that identies <strong>the</strong> spectral characteristics <strong>of</strong> smoke (Forest Fire<br />
Finder, <strong>2010</strong>).<br />
- Light detection and ranging (LIDAR) systems that measure laser light backscattered by<br />
<strong>the</strong> smoke particles (Utkin et al, 2003)<br />
- Radio-Acoustic Sounding System (RASS) for remote temperature measurements and<br />
<strong>the</strong>rmal sensing <strong>of</strong> a particular forest region (Sahin et al, 2009).<br />
- Acoustic Volumetric Scanner (VAS) that recognizes <strong>the</strong> re acoustic emission spectrum<br />
as a results <strong>of</strong> acoustic re sensing (Viegas et al., 2008).<br />
- Sensor network based system, where a number <strong>of</strong> sensor nodes (in most cases wireless<br />
sensors) are deployed in forest, measuring different environmental variables used for re<br />
detection. There are lot <strong>of</strong> different approaches, from more or less standard wireless sensor<br />
network (Byungrak et al, 2006), application <strong>of</strong> so called Fiber Optic Sensor Network<br />
(FOSN) developed within <strong>the</strong> EU-FIRE project (Viegas et al, 2008) to exotic proposals<br />
where animals have to be used as mobile biological sensors equipped with sensor devices<br />
(Sahin, 2008).<br />
Each technology has its advantages and disadvantages. Most <strong>of</strong> <strong>the</strong>m are promising, but<br />
<strong>the</strong>y are still in experimental stage, particularly sensor networks, VAS, RASS and LIDAR<br />
systems. For example LIDAR - light detection and ranging system is used to carry out<br />
chemical detection from great distances and has <strong>the</strong> potential to be an efcient system for<br />
wildre detection. However, it requires <strong>the</strong> lighting <strong>of</strong> <strong>the</strong> horizon with a laser beam that<br />
causes public health risks, besides not being very feasible from <strong>the</strong> economic point <strong>of</strong> view.<br />
Because <strong>of</strong> that, today in commercial use are mostly video based system equipped with<br />
cameras sensitive in visible spectra and/or infrared spectra and systems based on optical<br />
spectrometry. Optical spectrometry is ra<strong>the</strong>r new technology. It is based on a chemical<br />
analysis <strong>of</strong> <strong>the</strong> atmosphere by an optical spectrometry system. A telescope is coupled with<br />
optical sensor connected to a spectrometer unit with optical cable. The system analyzes<br />
<strong>the</strong> way <strong>the</strong> sunlight is absorbed by <strong>the</strong> atmosphere. It is quite efcient having <strong>the</strong> smallest<br />
number <strong>of</strong> false alarms, but <strong>of</strong> course it has a lot <strong>of</strong> disadvantages too. The main one is that<br />
it scans <strong>the</strong> space above <strong>the</strong> tree crowns on horizon, so <strong>the</strong> smoke has to be higher than <strong>the</strong><br />
horizon. The second one is dubious night detection when standard video camera is used<br />
rst to detect light and <strong>the</strong>n optical spectrometry is used to detect re ames. Because <strong>of</strong><br />
that in commercial optical spectrometry based systems, <strong>the</strong> video cameras in visible spectra<br />
are also included. Infrared systems are good choice for wildre detection, but <strong>the</strong>ir price<br />
is still quite high in comparison to video cameras sensitive in visible spectra, <strong>the</strong>y have<br />
limited space resolution and <strong>the</strong>y could not be used for distant video presence. Therefore,
62<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
contemporary systems for wildre detection based on infrared cameras are usually also<br />
equipped with cameras sensitive in visible spectra.<br />
The conclusion is that almost in all commercially available systems, <strong>the</strong> camera sensitive<br />
in visible spectra is also present. As an old adage said: “The best hunting solution is<br />
to kill two rabbits with one shot.” we think that today <strong>the</strong> most suitable solution for terrestrial<br />
automatic wildre detection and monitoring systems is to use cameras sensitive<br />
in visible spectra, particularly from <strong>the</strong> price/quality point <strong>of</strong> view. If you want to build a<br />
network, you need a lot <strong>of</strong> monitoring devices and <strong>the</strong> price <strong>of</strong> various advanced re detection<br />
systems are much higher than today’s high quality video cameras. Additional feature<br />
<strong>of</strong> today’s video cameras is <strong>the</strong>ir dual sensitivity. They are usually color cameras sensitive<br />
in visible spectra during <strong>the</strong> day, and black and white cameras sensitive in near IR spectra<br />
during <strong>the</strong> night, so <strong>the</strong> detection capabilities, particularly in sunrise/sunset parts <strong>of</strong> <strong>the</strong><br />
day are greatly improved. The second reason why visible spectra video cameras based<br />
system are <strong>the</strong> best choice is because <strong>the</strong>ir way <strong>of</strong> detecting wildres is <strong>the</strong> most close to<br />
human based wildre detection. Human wildre observers primarily use his (or her) vision<br />
sensor (eyes) to detect wildre. Sometimes humans use additional visual enhancement<br />
devices like binocular to check suspicious areas, but humans never use o<strong>the</strong>r sensors for<br />
early wildre detection. During our research and development phase we have spoken with<br />
a lot various human wildre observers and no one <strong>of</strong> <strong>the</strong>m said that he (or she) is using<br />
for example hearing (ears) to detect wildres. That was <strong>the</strong> reason why our efforts from<br />
<strong>the</strong> beginning were to develop ground-based wildre monitoring system based on visible<br />
cameras sensitive in visible spectra.<br />
Off course we were not <strong>the</strong> only ones who have thought <strong>of</strong> that way. In various countries<br />
that encounter high risk <strong>of</strong> wildres, various terrestrial systems based on cameras sensitive<br />
in visible spectra were developed and proposed. In all <strong>of</strong> <strong>the</strong>m automatic wildre detection<br />
was based on smoke recognition during <strong>the</strong> day and ame recognition during <strong>the</strong> night.<br />
The main disadvantage <strong>of</strong> those systems is ra<strong>the</strong>r high false alarms rate, due to atmospheric<br />
conditions (clouds, shadows, dust particles), light reections and human activities. Therefore,<br />
systems are usually designed as semi-automatic systems, which means that a human<br />
operator supervises <strong>the</strong> automatic wildre detection and his (or her) decision is <strong>the</strong> nal<br />
one. After <strong>the</strong> re alarm is generated and suspicious part <strong>of</strong> <strong>the</strong> image is marked, <strong>the</strong> human<br />
operator conrms or discards <strong>the</strong> alarm. The task <strong>of</strong> a human operator is not to monitor<br />
camera displays all <strong>the</strong> time, like in video cameras based human wildre surveillance<br />
mentioned in previous section, but only to conrm or discard possible re alarms. If <strong>the</strong><br />
human operator is not sure about a re alarm, he (or she) could switch <strong>the</strong> system to manual<br />
operation and make additional inspections using camera pan, tilt and zoom features. Using<br />
such semi-automatic surveillance system, human operator efciency is highly improved.<br />
One operator can manage more video monitoring units but also his (or her) fatigue is<br />
greatly reduced.<br />
iForestFire - Intelligent Forest Fire Monitoring System belongs to this category. It is an<br />
innovative, cloud computing based, semi-automatic wildre detection system with quite<br />
advance distant video presence capabilities.<br />
3 Main features <strong>of</strong> iForestFire - Intelligent Forest Fire Monitoring Systems<br />
iForestFire is integrated and intelligent video based wildre surveillance and monitoring<br />
system. Wildres are detected in incipient stage using advanced image processing and<br />
image analyses methods. Intelligent re recognition algorithms analyze images automatically,<br />
trying to nd visual signs <strong>of</strong> wildres, particularly wildre smoke during <strong>the</strong> day and<br />
wildre ames during <strong>the</strong> night. If something suspicious is found, pre-alarm is generated
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 63<br />
and appropriate image parts are visibly marked. The operator inspects suspicious image<br />
parts and decides is it really <strong>the</strong> wildre or not. The system is capable to work with both<br />
types <strong>of</strong> cameras: video cameras sensitive in visible and near IR spectra and real IR <strong>the</strong>rmal<br />
imaging cameras, but video cameras sensitive in visible spectra are preferred.<br />
Theoretical background <strong>of</strong> iForestFire is innovative and newly introduced wildre observer<br />
network <strong>the</strong>ory based on three-layer sensor network architecture, formal <strong>the</strong>ory <strong>of</strong><br />
perception and notation <strong>of</strong> observer (Stipaniev et al, 2007a; Šeri et al, 2009). Wildre<br />
observer, illustrated in Figure 3, is <strong>the</strong> core element <strong>of</strong> iForestFire system. It has three horizontal<br />
layers: data or sensor layer, information or service layer and knowledge or application<br />
layer, vertically interconnected by low-level or data observer working as proprioception<br />
unit (syntactic and semantic validation <strong>of</strong> sensors and sensors data) and two high-level<br />
observers, <strong>the</strong> image re observer and <strong>the</strong> decision re observer, working as exteroception<br />
units (making conclusions based on sensory data).<br />
Figure 3 – Wildre observer is organized as three-layer observer network<br />
iForestFire is a cloud computing or Web Information System (WIS) which means that <strong>the</strong><br />
operator could be located on any location with broadband Internet connection and his (or<br />
her) user interface is standard Web browser. The system is based on eld units and a central<br />
processing unit. The eld unit includes <strong>the</strong> day & night, pan/tilt/zoom controlled IP based<br />
video camera and IP based mini meteorological stations connected by wired or wireless<br />
LAN to a central processing unit where all analysis, calculation, presentation, image and<br />
data archiving are done. The system is also an example <strong>of</strong> Future Generation Communication<br />
Environment (FGCA) where all applications and services are focused on users,<br />
and “user” in our case is <strong>the</strong> natural environment, having <strong>the</strong> main task its own wildre<br />
protection. For such environment behavior <strong>the</strong> term environmental intelligence (EI) was<br />
introduced (Stipaniev et al, 2007a; Šeri et al, 2009).
64<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
iForestFire is both integral and intelligent user friendly system. Its organizational structure<br />
is shown in Figure 4.<br />
Figure 4 – The structure <strong>of</strong> iForestFire system<br />
It has three data bases: data warehouse with input images and alarm images, SQL database<br />
with meteorological data and temporal information <strong>of</strong> alarm images and GIS database with<br />
all relevant GIS data. iForestFire has ve working modes:<br />
- Manual Mode – user-friendly camera manual control by pan, tilt and zoom.<br />
- Automatic Mode – automatic re detection based on images captured by video cameras<br />
in <strong>the</strong> visible and near infrared spectra.<br />
- Archive Mode – video and meteorological data archive retrieval using various userfriendly<br />
procedures.<br />
- Simulation Mode – re behavior modeling and re-spread simulation using meteorological<br />
data and various GIS layers.<br />
- Fire Risk Calculation Mode - micro location re risk index calculation using, not only<br />
meteorological data, but sociological parameters connected with forest res too.<br />
iForestFire is integral because it is based on three different types <strong>of</strong> data:<br />
- Real time video data. Digital video stream is used in both, automatic and manual system<br />
modes. In automatic mode <strong>the</strong> video stream is a source <strong>of</strong> images for automatic wildre<br />
detection and in manual mode <strong>the</strong> video stream is used for distant video presence and distant<br />
video inspection.<br />
- Real time meteorological data. The meteorological data is used in <strong>the</strong> post-processing<br />
unit for false alarm reduction, but it is important for wildre risk calculation during <strong>the</strong><br />
monitoring phase and wildre spread behavior modeling during <strong>the</strong> re-ghting phase.<br />
Main meteorological parameters are measured using high tech IP based ultra sound mini<br />
meteorological station (iMeteo, <strong>2011</strong>), also developed during iForestFire project.<br />
- GIS (Geographical Information System) database. GIS system stores not only informa-
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 65<br />
tion on pure geographical data (elevations, road locations, water resources etc.), but also<br />
all o<strong>the</strong>r relevant forest re information related to a geographic position, like re history,<br />
land cover – land use, roads and forest corridors and similar. These data are used for userfriendly<br />
camera pan/tilt control but also <strong>the</strong>y are quite useful for reghting management<br />
activities. GIS data is essential for Simulation Mode and Fire Risk Calculation Mode.<br />
iForestFire is intelligent because it is based on articial intelligence (AI), computational<br />
intelligence (CI) and distributed intelligence (DI) technologies like:<br />
- Multi - agent based architecture. The system s<strong>of</strong>tware organization is based on agent<br />
architecture. Intelligent s<strong>of</strong>tware agents are responsible for sensors integrity testing, image<br />
and meteorological data collecting, syntactic and semantic image and data validation, image<br />
and data storing, image pre-processing processing and post-processing and pre-alarms<br />
and alarms generation. All agents share <strong>the</strong> same ontology and speak <strong>the</strong> same agent communication<br />
language (ACL) (Šeri et al, 2009). To demonstrate <strong>the</strong> system complexity, let<br />
us mention that on one server having 5 monitoring locations with 16 preset positions on<br />
each video unit, more <strong>the</strong>n 300 agents are working in parallel.<br />
- Advanced image processing and analyses algorithms. In its automatic mode, <strong>the</strong> wildre<br />
detection is based on various advanced image processing, image analyzing and image understanding<br />
algorithms. Various algorithms work in parallel based on advanced motion detection,<br />
advanced image segmentation, res smoke dynamic pattern analysis, color-space<br />
analysis and texture analysis (Krstini et al,2009; Krstini et al, <strong>2011</strong>). Typical detection<br />
result for monitoring station located in Buzet region (Istria) is shown in Figure 5.<br />
- Advanced procedures for false alarms reduction. In post-processing analysis, various<br />
methods derived from intelligent technologies eld are used to reduce <strong>the</strong> number <strong>of</strong> false<br />
alarms, as for example advanced image processing techniques (Jakovevi et al, 2009),<br />
rule-based expert system, data fusion algorithms (Stipaniev et al, 2007b) and integration<br />
<strong>of</strong> re risk index calculation with automatic adjustment <strong>of</strong> detection sensitivity (Bugari<br />
et al, 2009). Algorithms have a number <strong>of</strong> tuning parameters, but our experience was that<br />
users adjust <strong>the</strong>m rarely. The poorly adjusted parameters sometimes cause overly false<br />
alarm generation. That was <strong>the</strong> reason why we have introduced <strong>the</strong> possibility <strong>of</strong> automatic<br />
parameter adjustment based on meteorological data fusion and augmented reality features.<br />
Results <strong>of</strong> re risk index calculation are used to automatically increase or decrease system<br />
detection sensitivity on various image regions. Also a powerful QoS (Quality <strong>of</strong> service)<br />
was developed, particularly related to wildre observer detection quality evaluation<br />
(Jakovevi et al, <strong>2010</strong>). QoS is used particularly as a tool for fur<strong>the</strong>r improvements <strong>of</strong><br />
detection quality.<br />
- Augmented reality. The system is geo-referenced, so for every image pixel <strong>the</strong> corresponding<br />
geo-coordinate could be known and vice versa. The augmented reality features,<br />
now in experimental phase, based on fusion and integration <strong>of</strong> GIS information and real<br />
time video images are used in both, automatic and manual mode. Two examples <strong>of</strong> augmented<br />
reality use in automatic mode are automatic adjustment <strong>of</strong> detection sensitivity and<br />
determination <strong>of</strong> smoke location geo-coordinates. In manual mode important GIS information<br />
could be shown on video screen like toponyms, coordinates, altitudes, but GIS data are<br />
also used in advanced cameras manual control.<br />
In system design phase particular attention was given to create a user-friendly system. All<br />
iForestFire modules and components could be reached and administrated through dynamic<br />
and interactive Web pages, where real time video and meteorological data are shown toge<strong>the</strong>r<br />
with GIS data and user friendly interface for camera pan/tilt/zoom camera control.<br />
From <strong>the</strong> beginning, <strong>the</strong> reghters were involved in experiments with iForestFire system<br />
prototype, so <strong>the</strong> nal user interface was designed taking into account <strong>the</strong>ir advices. Figure<br />
5 shows a typical camera control screen, and a typical re alarm screen.
66<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
Figure 5 - iForestFire camera control screen with various manual control modes and typical re<br />
alarm screen at location Buzet in Istria. The user can accept or discard <strong>the</strong> generated alarm.<br />
For right decision about reghting intervention, both <strong>the</strong> early re detection and appropriate<br />
judgment about <strong>the</strong> potential re danger are important. That is <strong>the</strong> reason why from <strong>the</strong><br />
reghters’ point <strong>of</strong> view, both automatic detection <strong>of</strong> wildres and manual camera control<br />
modes are <strong>of</strong> equal importance. Because <strong>of</strong> that, we have implemented in iForestFire various<br />
user-friendly procedures for cameras manual control like:<br />
- Geo-referenced camera map control. The user can control camera pan movement by<br />
simple clicking on geo-referenced map. The camera control system is integrated with GIS,<br />
so it automatically detects and informs <strong>the</strong> user is <strong>the</strong> chosen point visible from <strong>the</strong> camera<br />
location or not.<br />
- Geo-referenced one-click multiple cameras map control. In regions where <strong>the</strong> monitoring<br />
cameras network is established (like Istria County) <strong>the</strong> so-called one-click multiple cameras<br />
control was implemented. The user simply click on geo-referenced map and in background<br />
visibility <strong>of</strong> that location is calculated for all cameras in neighborhood, appropriate<br />
azimuth and elevation angles are calculated and all cameras are automatically pan and tilt<br />
moved to show chosen location (Stipaniev et al, 2009).<br />
- Camera control using panorama image. In left upper corner <strong>of</strong> Figure 5 <strong>the</strong> 360o panorama<br />
image is shown. By simply clicking on panoramic image camera moves to chosen<br />
position by both pan and tilt.<br />
- Camera control using preset positions. Camera could by simple click on preset thumbs,<br />
moved to preset positions, pre-dened by pan, tilt and zoom.<br />
- Virtual pan-tilt-zoom commands and joystick emulation. Virtual commands are shown<br />
in upper right part <strong>of</strong> Figure 5. Simple, self-explaining virtual commands for pan tilt and<br />
zoom camera control were implemented, toge<strong>the</strong>r with joystick s<strong>of</strong>tware emulation.<br />
iForestFire has a powerful archive retrieval methods for input images, generated alarm images<br />
and meteorological data. All <strong>of</strong> <strong>the</strong>m could be easily reached and analyzed using various<br />
image and image data retrieval procedures based on advanced Internet technologies.<br />
Each monitoring station is equipped with advanced IP based ultra sound mini meteorological<br />
station used for measuring <strong>the</strong> most important meteorological parameters: air<br />
temperature, relative humidity, air pressure, wind speed, wind direction and wind gust.<br />
Additionally it is possible to measure o<strong>the</strong>r meteorological parameters important for wild-
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 67<br />
re behavior and system performances, as for example isolation, precipitation, moisture,<br />
ground temperature or lighting activity. Meteorological data are used in Automatic Mode<br />
for false alarms reductions in post-processing procedures, but also in our experimental Fire<br />
Risk Calculation Mode and Simulation Mode. Figure 6 shows <strong>the</strong> system interfaces for<br />
micro-location wildre risk calculation and simulation <strong>of</strong> wildre propagation.<br />
Figure 6 – Experimental system for micro-location wildre risk calculation (left) and simulation<br />
<strong>of</strong> wildre propagation (right)<br />
In Fire Risk Calculation and Simulation Mode external meteorological services are also<br />
used, for example <strong>the</strong> results <strong>of</strong> meteorological simulations performed by simulation model<br />
ALADIN-HR. These data are automatically collected from <strong>the</strong> servers <strong>of</strong> Meteorological<br />
and Hydrological Service <strong>of</strong> Croatia twice a day. Figure 6 shows ALADIN-HR wind data<br />
superimposed on re risk and simulation <strong>of</strong> re propagation maps.<br />
iForestFire is also award wining ICT project. In 2008 iForestFire won rst price (Tesla<br />
Golden Egg) on VIDI e-novation award, competition founded by VIDI publication and<br />
Rudjer Boskovic institute. It was elected as <strong>the</strong> most prominent and innovative ICT project<br />
in Croatia in 2008.<br />
4 Development and commercialization <strong>of</strong> iForestFire system<br />
Development <strong>of</strong> Intelligent Forest Fire Monitoring System started in autumn 2003 as a<br />
seminar on <strong>the</strong> postgraduate (Mr.Sc.) study at Faculty <strong>of</strong> Electrical Engineering, Mechanical<br />
Engineering and Naval Architecture, provoked by great damages caused by 2003 wild-<br />
res. The student Damir Krstini (who was also <strong>the</strong> voluntary reghter at island Hvar) enrolled<br />
<strong>the</strong> course Digital Image Processing and Analyses. The results <strong>of</strong> his seminar entitled<br />
“Wildre smoke segmentation” were more than promising, so we have decided to apply for<br />
technology project, and in 2003 we have received a grant TP-03/0023-09 ''System for early<br />
forest re detection based on cameras in visible spectra" supported by <strong>the</strong> Ministry <strong>of</strong> Science,<br />
Education and Sport <strong>of</strong> Republic Croatia. The grant funds were enough to start more<br />
intensive research, but not enough to nished it, so a part <strong>of</strong> our research were also funded<br />
by ordinary project 023-0232005-2003 "AgISEco Agent-based intelligent environmental<br />
monitoring and protection systems", by support <strong>of</strong> Split and Dalmatia County authorities<br />
through a study "Holistic approach to forest re protection in Split and Dalmatia County"<br />
and by our own resources.<br />
In initial system development four researchers were involved - Darko Stipaniev as a team<br />
leader, Maja Štula as a designer <strong>of</strong> overall Web based information system, Damir Krstini
68<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
as a designer <strong>of</strong> wildre detection algorithms and Ljiljana Šeri as a designer <strong>of</strong> wildre<br />
observer agent architecture. From <strong>the</strong> beginning <strong>the</strong> main reghters adviser was Tomislav<br />
Vuko, <strong>the</strong> vice commander <strong>of</strong> Croatian reghters for Adriatic Coast and Islands.<br />
Research started in 2003 with wildres video materials collecting necessary for detection<br />
algorithm development. Controlled res were burned at island Hvar in cooperation with<br />
voluntary reghters. Lot <strong>of</strong> video material was recorded and soon <strong>the</strong> rst version <strong>of</strong> <strong>of</strong>fline<br />
detection algorithm was developed. Figure 7 shows examples <strong>of</strong> rst algorithm detections<br />
in <strong>the</strong> typical landscape <strong>of</strong> island Hvar.<br />
Recording and collecting wildre images and video sequences is quite important for detection<br />
algorithm development and testing so this activity was carried out all <strong>the</strong> time. Today<br />
our database has more than 2.500 images selected and segmented by reference (ground<br />
truth) human observer. Figure 8 shows one typical example. On manually segmented image<br />
all regions were distinguished, because in <strong>the</strong> latest version <strong>of</strong> our detection algorithm<br />
region context based wildre alarm reduction has been applied so it was important to have<br />
manually segmented, not only wildre smoke – no smoke regions, but also o<strong>the</strong>r regions<br />
like sky, water, vegetation, man-made objects etc.<br />
Figure 7 – Examples <strong>of</strong> detection images in <strong>the</strong> typical landscape for Croatian coasts and islands<br />
(island Hvar 2003)<br />
Figure 8 – Input image and corresponding manually segmented image regions represented by<br />
various gray tonalities
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 69<br />
After successful laboratory implementation, <strong>the</strong> real life eld-testing was performed during<br />
re seasons 2004 and 2005. Three experimental monitoring stations were installed on<br />
Marjan Hill near Split, Vidova gora on island Bra and FESB faculty building in Split.<br />
These eld tests were performed in cooperation with Fire Brigade Administration for <strong>the</strong><br />
Coast, Protection and Rescue Operations Administration, Ministry <strong>of</strong> Interiors in Divulje<br />
so <strong>the</strong> main monitoring centers were in Divulje and <strong>the</strong> second one in our laboratory. Figure<br />
9 shows <strong>the</strong> experimental system layout and Figure 10 <strong>the</strong> experimental station on<br />
Marjan hill.<br />
Figure 9 – The experimental system layout used for iForestFire prototype<br />
testing in 2004 and 2005<br />
After <strong>the</strong> testing period a lot <strong>of</strong> improvements in both, hardware and s<strong>of</strong>tware design were<br />
implemented and in 2006 <strong>the</strong> commercialization phase has started and <strong>the</strong> rst real life<br />
monitoring system was installed in National Park Paklenica supported by Ministry <strong>of</strong> Culture<br />
<strong>of</strong> Republic Croatia. The system was partially developed as a technological project<br />
supported by Ministry <strong>of</strong> Science, Education and Sport <strong>of</strong> Republic Croatia and in contract<br />
signed with <strong>the</strong>m our obligations were also <strong>the</strong> system commercialization. Two models<br />
were possible – to establish a new spin-<strong>of</strong>f company or to nd a strategic partner. In that<br />
moment, <strong>the</strong> second option was more convenient for us, so we found <strong>the</strong> strategic partner<br />
in company Lama d.o.o. from Split (iForesFire, <strong>2011</strong>). Lama was responsible for system<br />
promotion, selling, installation and maintenance, but fur<strong>the</strong>r system improvements, development<br />
and research were and remain until today our obligations.<br />
The development <strong>of</strong> system like iForetFire is a never-ending story. The system version<br />
is now 2.7 (October <strong>2010</strong>) and beside <strong>the</strong> initial researchers involved in system development<br />
from <strong>the</strong> begineeng, in today’s version signicant contributions were given also by<br />
Toni Jakovevi (advanced IP based meteorological station, detection algorithm), Marin<br />
Bugari (Web GIS based system features), Josip Maras (component based system architecture),<br />
Petar Jeri (s<strong>of</strong>tware developement) and Kaja Radi (electronic components design<br />
and realization).
70<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
Figure 10 – The experimental station on Marjan hill<br />
Since 2006 <strong>the</strong> system was successfully applied in various Croatian national and nature<br />
parks (7 monitoring stations and 5 operational centers), but <strong>the</strong> most advanced is network<br />
application in Istria County. Istria system called Istria iForestFire Net (Stipaniev et al,<br />
<strong>2010</strong>) has 29 monitoring stations and 7 operational centers, mutually interconnected using<br />
encrypted VPN and hardware rewalls developed by our strategic partner – <strong>the</strong> company<br />
Lama. Figure 11 shows system layout and few screen prints <strong>of</strong> that system. iForestFire also<br />
has a quite promising export potentials. In March <strong>2011</strong> two demo units are in installation<br />
phase, one in Greece and <strong>the</strong> o<strong>the</strong>r in Portugal.<br />
Last but not least it is important to emphasize that iForestFire has initiated a lot <strong>of</strong> scientic<br />
research, too. As a result <strong>of</strong> research connected with system development three PhD <strong>the</strong>ses<br />
were written and defended (D.Krstini, Lj.Šeri and T.Jakovcevic), two PhD <strong>the</strong>ses are in<br />
preparation phase and lot <strong>of</strong> scientic papers were published in journals, books and conference<br />
proceedings. A specialized Web portal dedicated to wildre observers and smoke<br />
recognition has been created and maintained (http://wildre.fesb.hr), but also our interest<br />
in more general scientic wildre research has resulted in establishment <strong>of</strong> Center for<br />
Wildre Research (http://cipop.fesb.hr).
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 71<br />
Figure 11 – Istria iForestFire Net – advanced wildre monitoring network in Istria County<br />
5 Conclusion<br />
The only effective way to minimize damages caused by wildres is wildre early detection<br />
and fast and appropriate reaction, apart from preventive measures. Great efforts are<br />
<strong>the</strong>refore made to achieve early forest re detection, which is traditionally based on human<br />
surveillance. This paper shows how modern ICT technologies could be used for automatic<br />
wildre detection and monitoring. It describes <strong>the</strong> advanced wildre surveillance and monitoring<br />
system named iForestFire and entirely developed at University <strong>of</strong> Split Faculty <strong>of</strong><br />
Electrical Engineering, Mechanical Engineering and Naval Architecture.<br />
iForestFire is a practical realization <strong>of</strong> <strong>the</strong> observer network <strong>the</strong>ory. Observer network<br />
was dened as an advanced sensor network described using formal <strong>the</strong>ory <strong>of</strong> perception<br />
and a notation <strong>of</strong> <strong>the</strong> observer. iForestFire is both integral and intelligent system. Integral<br />
because it is based on various data types (images, meteorological data, GIS data) and intelligent<br />
because it has a lot <strong>of</strong> features derived from various intelligent technologies (multiagent<br />
architecture, advanced image analysis and image recognition algorithms, advanced<br />
procedures for false alarm reduction, augmented reality).<br />
The system was developed as a technological and scientic research project, but today it is<br />
commercial system widely used for advanced wildre surveillance and monitoring in various<br />
Croatian national and nature parks and Istria region, so it is a good example <strong>of</strong> successful<br />
university – industry cooperation, particularly because its development has provoked a<br />
lot <strong>of</strong> scientic research activities resulting in a number <strong>of</strong> PhD <strong>the</strong>sis and published papers<br />
in journals, books and conference proceedings.
72<br />
References<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
1. Arrue B C, Ollero A, Martinez de Dios J R (2000) An intelligent system for false alarm<br />
reduction in infrared forest-re detection. IEEE Int. Systems, May/June 2000, 64-72.<br />
2. Bugari M, Jakovevi T, Stipaniev D (2009) Automatic adjustement <strong>of</strong> detection<br />
parameters in forest re video monitoring system, MIPRO, May 2009, Opatija, pp.<br />
270-275<br />
3. Byungrak S, Yong-Sork H, and Jung-Gyu K, (2006) A Design and Implementation<br />
<strong>of</strong> Forest-Fires Surveillance System based on Wireless Sensor Networks for South<br />
Korea Mountains, IJCSNS Int. Journal <strong>of</strong> Computer Science and Network Security,<br />
V.6 No.9B, pp.124-130<br />
4. Forest Fire Finder NGNS IS, http://www.ngns-is.com/html/orestas/fff_intro_eng.<br />
html (Accessed March, 1 <strong>2011</strong>)<br />
5. iForestFire (<strong>2011</strong>) Itelligent Forest Fire Monitorig System, http://iforestre.fesb.hr<br />
(Accessed March, 1 <strong>2011</strong>)<br />
6. iMeteo (<strong>2011</strong>) Advance Ultrasound IP based Mini Meteorological Station, http://imeteo.fesb.hr<br />
(Accessed March, 1 <strong>2011</strong>)<br />
7. Jakovevi T, Stipaniev D, Krstini D, (2009) False alarm reduction in forest re<br />
video monitoring system, MIPRO 2009, May 25-29, Opatija, 264-269<br />
8. Jakovevi T, Šeri Lj, Stipaniev D, Krstini D, (<strong>2010</strong>) Wildre smoke detection algorithms<br />
evaluation, VI Int. Conf. on Forest Fire Research, Coimbra, Nov 15-18 <strong>2010</strong>.<br />
9. Krstini D, Jakovevi T, Stipaniev D, (2009) Histogram-Based Smoke Segmentation<br />
in Forest Fire Detection System, Information Technology and Control, Vol.38,<br />
No.3, 2009. 237-244<br />
10. Krstini D, Kuzmani Skelin A, Slapniar I (<strong>2011</strong>) Fast Two-Step Histogram-Based<br />
Image Segmentation, IET image processing. 5 (<strong>2011</strong>), 1; 63-72<br />
11. Kührt E, Knollenberg J, Mertens V (2001) An automatic early warning system for forest<br />
res, Annals <strong>of</strong> Burns and Fire Disasters, vol. XIV, n. 3, Sept. 2001<br />
12. Sahin Y, Animals as Mobile Biological Sensors for Forest Fire Detection, Sensors<br />
2007, 7, 3084-3099<br />
13. Sahin Y G, Turker Ince S, Early Forest Fire Detection Using Radio-Acoustic Sounding<br />
System, Sensors 2009, 9, 1485-1498<br />
14. Stipaniev D, Hrasnik B (2004) Integral, holistic model <strong>of</strong> wildre prevention in Split<br />
and Dalmatia County, Expert Study for Split and Dalmatia County, FESB Split, 2004.<br />
(231 pages) (in Croatian)<br />
15. Stipaniev D, Hrastnik B, Vuji R (2007) Holistic Approach to Forest Fire Protection<br />
in Split and Dalmatia County <strong>of</strong> Croatia, Wildre 2007 Int.Conference, Sevilla, Spain,<br />
May 2007.<br />
16. Stipaniev D, Bodroži Lj, Štula M (2007a) Environmental Intelligence based on Advanced<br />
Sensor Networks, Proc.<strong>of</strong> 14th Int.Conference on Systems, Signals and Image<br />
Processing, Maribor, Slovenija, 27-30.6.2007<br />
17. Stipaniev D, Bodroži Lj, Štula M (2007b) Data Fusion in Observer Networks, Proc.<br />
<strong>of</strong> Second (IEEE) International Workshop on Information Fusion and Dissemination<br />
in Wireless Sensor Networks. Pisa, Italia, 08.10.2007, 1-6<br />
18. Stipaniev D, Bugari M, Bodroži Lj (2009) Integration <strong>of</strong> Forest Fire Video Monitoring<br />
System and Geographic Information System, Proc. <strong>of</strong> 51st Int.Symp ELMAR<br />
2009, Zadar, Sept 2009, pp.49-52<br />
19. Stipaniev D, Štula M, Krstini D, Šeri Lj, Jakovevi T, Bugari M (<strong>2010</strong>) Advanced<br />
automatic wildre surveillance and monitoring network, VI International Conference<br />
on Forest Fire Research, Coimbra, Porugal, Nov. 15 – 18, <strong>2010</strong>, 053, 15 pages
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 73<br />
20. Šeri Lj, Stipaniev D, Štula M (2009) Observer network and forest re detection,<br />
Information Fusion, DOI: 10.1016/j.inffus.2009.12.003, on-line from Dec 28, 2009.<br />
21. Utkin A, Vilar R, Feasibility <strong>of</strong> forest-re smoke detection using lidar, International<br />
Journal <strong>of</strong> Wildland Fire 2003, 12(2) 159 – 166<br />
22. Viegas X, Pita L P, Nielsen F, Haddad K, Calisti Tassini C, D’Altrui G, Quaranta<br />
V, Dimino I, Acoustic and <strong>the</strong>rmal characterization <strong>of</strong> a forest re event, Proceedings<br />
<strong>of</strong> SPIE, Remote sensing <strong>of</strong> re, San Diego CA , 2008, vol. 7089, pp. 708904.1-<br />
708904.12<br />
23. Žaja D (2008) The role <strong>of</strong> Croatian Forests d.o.o. in ghtings again wildres, International<br />
Workshop – new Methodes and Approaches to Wildres Prevention and Protection,<br />
Dec, 8-11 2008., Makarska (in Croatian)
74<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
Laboratory testing and numerical modelling <strong>of</strong> MBT<br />
waste deformability<br />
Igor Petrović, Ph.D.<br />
University <strong>of</strong> Zagreb, Faculty <strong>of</strong> Geotechnical Engineering<br />
Hallerova aleja 7, Varaždin, Croatia<br />
e-mail: igor.petrovic@gfv.hr<br />
Pr<strong>of</strong>. Davorin Kovačić, Ph.D.<br />
University <strong>of</strong> Zagreb, Faculty <strong>of</strong> Geotechnical Engineering<br />
Hallerova aleja 7, Varaždin, Croatia<br />
e-mail: davorin.kovacic@gfv.hr<br />
Abstract<br />
The paper deals with <strong>the</strong> experimental investigation and numerical simulation <strong>of</strong> <strong>the</strong> mechanical<br />
behaviour <strong>of</strong> samples from a “Mechanical Biological Treatment” waste deposit<br />
(MBT waste deposit). Laboratory tests with samples <strong>of</strong> different moisture contents were<br />
conducted to determine <strong>the</strong> basic geotechnical characteristics as well as <strong>the</strong> stress-strain<br />
relations under loading and reloading conditions. When it is required to evaluate <strong>the</strong> loaddeformation<br />
characteristics <strong>of</strong> specimens containing relatively large-diameter grain particles,<br />
like MBT waste for example, <strong>the</strong> standard oedometric apparatus with cell (sample)<br />
diameter <strong>of</strong> 56 mm is evidently inadequate. Therefore, at <strong>the</strong> University <strong>of</strong> Zagreb, an<br />
oedometer <strong>of</strong> 500 mm in internal diameter was designed, fabricated, assembled and evaluated.<br />
As only limited experimental results were available, for numerical simulations a simplied<br />
hypoplastic model was used. For <strong>the</strong> calibration <strong>of</strong> this model it was sufcient to conduct<br />
oedometer tests and to use a simple procedure to estimate <strong>the</strong> angle <strong>of</strong> friction. The model<br />
proved to be satisfactory for modelling <strong>the</strong> compaction behaviour <strong>of</strong> MBT waste observed<br />
in experiments.<br />
Key words: compressibility, hypoplasticity, MBT waste, oedometer, stiffness<br />
1 Introduction<br />
The waste management is actually one <strong>of</strong> <strong>the</strong> most important issues related to <strong>the</strong> protection<br />
<strong>of</strong> <strong>the</strong> environment. Remarkably high amount <strong>of</strong> money is invested in waste management,<br />
including <strong>the</strong> separation, recycling, transport and permanent landlling <strong>of</strong> municipal solid<br />
waste. According to <strong>the</strong> Waste Management Strategy <strong>of</strong> <strong>the</strong> Republic <strong>of</strong> Croatia (2005), a<br />
more intensive process <strong>of</strong> rehabilitation and improvement <strong>of</strong> sanitary conditions on a large<br />
number <strong>of</strong> uncontrolled waste dumps started in 1990s.<br />
The objective <strong>of</strong> <strong>the</strong> Strategy is to establish framework within which Croatia should reduce<br />
<strong>the</strong> amounts <strong>of</strong> waste it produces, and to be able to manage it in a sustainable manner. Taking<br />
<strong>the</strong> assessment <strong>of</strong> current situation and a vision <strong>of</strong> an appropriate waste management<br />
system as a starting point, <strong>the</strong> Strategy sets <strong>the</strong> objectives and <strong>of</strong>fers suggestions for <strong>the</strong>ir<br />
gradual achievement by 2025.
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 75<br />
An ideal project <strong>of</strong> waste management in <strong>the</strong> Republic <strong>of</strong> Croatia corresponds to <strong>the</strong> socalled<br />
non-landll concept. In order to achieve this project it would be necessary to close<br />
<strong>the</strong> circle starting from avoidance <strong>of</strong> waste production, reduction <strong>of</strong> quantities and harmfulness,<br />
recycling and recovery (mechanical, biological, energy recovery) to <strong>the</strong> use <strong>of</strong> inert<br />
residue.<br />
The Strategy has set quantitative targets that determine <strong>the</strong> dynamics <strong>of</strong> meeting strategic<br />
objectives (Table 1). Deadlines are determined with <strong>the</strong> expected time delays in relation to<br />
EU regulations.<br />
Therefore <strong>the</strong> following basic activities are to be performed:<br />
- rehabilitation and closure <strong>of</strong> <strong>the</strong> existing open dumps,<br />
- extension and improvement <strong>of</strong> those landlls which will remain in operation,<br />
- construction <strong>of</strong> new waste management centres in which <strong>the</strong> inevitable waste should be<br />
landlled.<br />
Table 1. Quantitative targets for landlls<br />
Targets<br />
Year<br />
2005. <strong>2010</strong>. 2015. 2020. 2025.<br />
Regional centres for waste<br />
management<br />
0 1-2 2-3 3 4<br />
County centres for waste management<br />
0 3-7 7-10 10-14 14-21<br />
“Ofcial landlls” * 187 100 50 30 14-21<br />
Percentage <strong>of</strong> remedied landlls<br />
(% from <strong>the</strong> number determined<br />
in 2000).<br />
5 65 75 85 100<br />
*Group “Ofcial landlls“ consists <strong>of</strong> <strong>the</strong> following categories dened above: legal land-<br />
lls, landlls in <strong>the</strong> process <strong>of</strong> legalization, <strong>of</strong>cial landlls, and landlls by agreement<br />
A waste landll has to be regarded as an engineered structure for which long-term safety<br />
has to be proved on <strong>the</strong> basis <strong>of</strong> an appropriate geotechnical design. This means that <strong>the</strong><br />
waste body has to be stable against slope failure. Fur<strong>the</strong>rmore, <strong>the</strong> deformations behaviour<br />
and stresses imposed on subsoil, liner systems and internal structures have to be assessed.<br />
Figure 1 gives schematic illustration <strong>of</strong> possible failure mechanisms and deformations<br />
problems (Jessberger et al., 1995)
76<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
Figure 1. Stability and deformation <strong>of</strong> a landll<br />
The study <strong>of</strong> <strong>the</strong> mechanical behaviour <strong>of</strong> municipal solid waste by monitoring <strong>the</strong> existing<br />
landlls and by laboratory testing <strong>of</strong> waste samples was named “waste mechanics” in<br />
<strong>the</strong> handbook issued by German Geotechnical Society in <strong>the</strong> chapter: Principles <strong>of</strong> Waste<br />
Mechanics in <strong>the</strong> Geotechnical Design <strong>of</strong> Landlls (1993).<br />
The results <strong>of</strong> a survey conducted in <strong>the</strong> period 1992 - 2000 (for both continental and coastal<br />
parts <strong>of</strong> <strong>the</strong> Croatia) show that <strong>the</strong> average <strong>annual</strong> composition <strong>of</strong> municipal solid waste<br />
has a biodegradable portion <strong>of</strong> 74,5 % (Domanovac et al., 2002). Waste Management Plan<br />
<strong>of</strong> <strong>the</strong> Republic <strong>of</strong> Croatia for 2007-2015 (2007) requires that until 2020, <strong>the</strong> biodegradable<br />
portion <strong>of</strong> municipal solid waste should be reduced to 35 % <strong>of</strong> <strong>the</strong> mass portion <strong>of</strong> biodegradable<br />
municipal solid waste produced in 1997. According to Ordinance on <strong>the</strong> methods<br />
and conditions for <strong>the</strong> landll <strong>of</strong> waste categories and operational requirements for waste<br />
landlls (2007), which is based on Council Directive 1999/31/EC <strong>of</strong> 26 April 1999 on <strong>the</strong><br />
landll waste, only pre-treated waste is allowed to be dumped in landlls. Most <strong>of</strong> county<br />
waste management plans anticipated mechanical biological treatment (MBT) as <strong>the</strong> optimum<br />
waste treatment method.<br />
The goal <strong>of</strong> mechanical biological waste treatment is <strong>the</strong> recycling <strong>of</strong> separated fractions<br />
(which leads to <strong>the</strong> reduction <strong>of</strong> landll volume) on one side, and <strong>the</strong> biological stabilisation<br />
<strong>of</strong> <strong>the</strong> waste on <strong>the</strong> o<strong>the</strong>r side. One <strong>of</strong> <strong>the</strong> consequences <strong>of</strong> <strong>the</strong> waste treatment is <strong>the</strong><br />
change <strong>of</strong> physical, biological and chemical properties <strong>of</strong> <strong>the</strong> material to be landlled,<br />
compared with <strong>the</strong> wastes which have been landlled before. It can be concluded that <strong>the</strong><br />
knowledge <strong>of</strong> mechanical parameters <strong>of</strong> MBT waste for <strong>the</strong> design <strong>of</strong> landlls, which will<br />
be built within waste management centres, is necessary. The extended operation period <strong>of</strong><br />
a landll due to <strong>the</strong> reduction <strong>of</strong> landll volume is directly dependent on <strong>the</strong> deformability<br />
properties <strong>of</strong> MBT waste which are in <strong>the</strong> focus <strong>of</strong> <strong>the</strong> present research.<br />
Most commercial geotechnical laboratories have been utilizing some kind <strong>of</strong> Casagrandetype<br />
oedometer (Casagrande, 1936) for routine measurements <strong>of</strong> deformability parameters<br />
<strong>of</strong> ne-grained soils for decades. O<strong>the</strong>r names for this test are: one-dimensional compression<br />
test, <strong>the</strong> conned compression test and <strong>the</strong> consolidation test. The apparatus contains<br />
an oedometer cell that can be installed in a loading frame so that a predetermined vertical<br />
consolidation stress can be applied to <strong>the</strong> specimen. The soil specimen is restrained laterally<br />
by a steel oedometer cell. The top and bottom surface <strong>of</strong> <strong>the</strong> specimen are in contact
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 77<br />
with porous stone discs, so that <strong>the</strong> excess pore water pressure induced in <strong>the</strong> specimen by<br />
<strong>the</strong> applied loading can be dissipated through simultaneous upward and downward drainage<br />
during consolidation. The specimen has to be fully submerged in water during <strong>the</strong> test<br />
to eliminate <strong>the</strong> surface tension <strong>of</strong> water at <strong>the</strong> drainage surfaces, so as to allow unrestricted<br />
drainage <strong>of</strong> water. The test is considered to be one dimensional as both specimen deformation<br />
and drainage occur in <strong>the</strong> vertical direction only. Oedometer samples usually have a<br />
diameter <strong>of</strong> 56 mm and <strong>the</strong> thickness-diameter ratio 1:3 to 1:4.<br />
Based on <strong>the</strong> above mentioned consideration, <strong>the</strong> manufacturing <strong>of</strong> large oedometer <strong>of</strong><br />
500 mm in internal diameter, suitable for testing MBT waste samples, was initiated within<br />
<strong>the</strong> scientic project “Characterization <strong>of</strong> municipal solid waste”. The basic outline <strong>of</strong> <strong>the</strong><br />
project was given by Petrovi (2008). The manufacturing <strong>of</strong> large oedometer started in<br />
September 2008 and lasted until March 2009 when it was assembled at <strong>the</strong> Faculty <strong>of</strong><br />
Geotechnical Engineering.<br />
2 Basic geotechnical characteristics <strong>of</strong> tested MBT waste material<br />
The MBT waste samples used for testing were imported from an MBT waste landll in<br />
Austria as in <strong>the</strong> Republic <strong>of</strong> Croatia <strong>the</strong>re is no MBT facility yet installed. Naturally <strong>the</strong>re<br />
was no experience in laboratory testing <strong>of</strong> MBT samples so far.<br />
Prior to laboratory testing <strong>the</strong> waste was dried in <strong>the</strong> open air for two weeks. After drying<br />
<strong>the</strong> equilibrium moisture content was determined according to ASTM D 2974 as in case <strong>of</strong><br />
organic soil samples. The samples were dried in <strong>the</strong> oven at <strong>the</strong> temperature 1050C. The<br />
equilibrium water content was 7 %.<br />
Figure 2 shows <strong>the</strong> particle size distribution curve for <strong>the</strong> air dried MBT waste material according<br />
to HRN.U.B1.018. It can be seen that <strong>the</strong> largest particle diameter does not exceed<br />
30 mm and <strong>the</strong>refore fulls <strong>the</strong> requirement that <strong>the</strong> largest particle diameter should not<br />
exceed <strong>the</strong> 1/5 <strong>of</strong> <strong>the</strong> sample height which, for <strong>the</strong> oedometer used, is 40 mm. Since <strong>the</strong><br />
ne-grained portion <strong>of</strong> <strong>the</strong> waste material is less than 10 %, a particle size analysis using<br />
<strong>the</strong> hydrometer test procedure was not conducted. According to <strong>the</strong> Unied Soil Classication<br />
System (USCS), <strong>the</strong> MBT waste can be classied as a coarse grained material, as<br />
<strong>the</strong> particle sizes lie in <strong>the</strong> range between 0.05 and 30 mm. The uniformity coefcient and<br />
coefcient <strong>of</strong> curvature are Cu = 26 and Cc = 1.5, respectively. Therefore <strong>the</strong> tested MBT<br />
waste is a well-graded material and it can be well compacted.<br />
Figure 2. Particle size distribution curve <strong>of</strong> MBT waste
78<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
The average density <strong>of</strong> <strong>the</strong> solid particles, s, was determined according to <strong>the</strong> ASTM D<br />
854 on <strong>the</strong> air free samples with maximum particle size <strong>of</strong> 4.75 mm using a pycnometer <strong>of</strong><br />
a known volume. The particles larger than 4.75 mm were removed by sieving. The value<br />
s = 2147 [kg/m 3 ] was determined.<br />
Detailed description <strong>of</strong> <strong>the</strong> laboratory testing <strong>of</strong> basic MBT waste parameters is given by<br />
Petrovi et al. (<strong>2011</strong>b).<br />
3 Design, fabrication and assembly <strong>of</strong> a large oedometer<br />
Since standards for testing compressibility behaviour <strong>of</strong> MBT waste do not exist it was<br />
concluded that a large oedometer must be designed, manufactured, assembled and evaluated<br />
in accordance with standards for testing compressibility behaviour <strong>of</strong> soils. This is quite<br />
in line with <strong>the</strong> fact that waste mechanics had been developed on <strong>the</strong> basis <strong>of</strong> soil mechanics<br />
principles. The most relevant standards used for dening <strong>the</strong> technical specications<br />
<strong>of</strong> large oedoemeter are British (BS 1377: Part 5: 1990) and American (ASTM D 2435)<br />
standards for compressibility testing as well as Croatian standard Eurocode 7: Geotechnical<br />
design – Part 2 - Ground investigation and testing (HRN EN 1997-2:2008). Technical<br />
specications <strong>of</strong> new oedometer were dened in collaboration with several mechanical<br />
<strong>engineering</strong>, industrial management and IT companies from Varaždin.<br />
The diameter <strong>of</strong> <strong>the</strong> new oedometer cell is 500 mm and <strong>the</strong> height is 200 mm. It means that<br />
minimum height-diameter ratio 1:2,5 required by <strong>the</strong> British standard in order to minimize<br />
<strong>the</strong> effect <strong>of</strong> <strong>the</strong> side friction forces, is satised. Maximum vertical pressure on sample is<br />
2000 kN/m 2 . The force (pressure) on specimen is transmitted by means <strong>of</strong> hydraulic loading<br />
actuator. Pressure transducer has a minimum accuracy ± 0,05 % <strong>of</strong> full range output<br />
(FRO). Displacement transducer has a minimum accuracy ± 0,05 % FRO and minimum<br />
stroke <strong>of</strong> 9 cm. The lower and upper lids, as well as <strong>the</strong> pressure plate, are made from<br />
structural steel (all protected against corrosion). Compression tests are fully automated<br />
(computer-controlled).<br />
Figure 3 shows a nal set-up <strong>of</strong> <strong>the</strong> large oedometer system while Figures 4, 5, 6 and 7<br />
present selected parts <strong>of</strong> a large oedometer during <strong>the</strong> production phase.<br />
Figure 3. Final set-up <strong>of</strong> <strong>the</strong> large oedometer system
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 79<br />
Figure 4 presents an upper and lower lid <strong>of</strong> oedometric cell. Maximum deformation <strong>of</strong> lids<br />
at highest load should not exceeds 0,7 mm. Figure 5 presents hydraulic loading actuator<br />
while Figure 6 presents oedometric cell ring polished to burnish so that friction between<br />
sample and cell ring is reduced to minimum. Figure 7 presents drainage porous plates<br />
which are required for dissipation <strong>of</strong> <strong>the</strong> excess pore pressure. Drainage porous plates<br />
are made from 1 mm thick stainless steel. The clogging <strong>of</strong> <strong>the</strong> drainage holes should be<br />
prevented with <strong>the</strong> use <strong>of</strong> <strong>the</strong> lter paper or geotextile placed between <strong>the</strong> specimen and<br />
drainage plate.<br />
Figure 4. An upper and lower lid <strong>of</strong> oedometric cell<br />
Figure 5. Hydraulic loading actuator<br />
Figure 6. Cell ring
80<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
Figure 7. Porous drainage plate<br />
As <strong>the</strong> innovative parts <strong>of</strong> <strong>the</strong> large oedometer are still subject <strong>of</strong> legal protection (patent<br />
registration), detailed technical description <strong>of</strong> <strong>the</strong> apparatus can not be given before <strong>the</strong><br />
completion <strong>of</strong> <strong>the</strong> legal procedure.<br />
4 Oedometric compression tests and friction angle<br />
4.1 Sample preparation<br />
In <strong>the</strong> present paper <strong>the</strong> properties <strong>of</strong> two MBT waste samples are presented, where Sample<br />
A is characterized by an initial moisture content <strong>of</strong> w = 65 % and Sample B is characterized<br />
by an initial moisture content <strong>of</strong> w = 41 %.<br />
In order to obtain a homogenous distribution <strong>of</strong> moisture content through <strong>the</strong> whole specimen<br />
after mixing with water, <strong>the</strong> wetted samples were left untouched for <strong>the</strong> next 24 hours.<br />
Then <strong>the</strong> samples were installed in <strong>the</strong> oedometer cell in ve layers. On average, each layer<br />
was 4 cm thick and pre-compacted to a dened initial void ratio.<br />
The basic geotechnical parameters for both samples at <strong>the</strong> beginning <strong>of</strong> <strong>the</strong> test are presented<br />
in Table 2 respectively. Herein denotes <strong>the</strong> density, w <strong>the</strong> moisture content, d <strong>the</strong><br />
dry density, s <strong>the</strong> average density <strong>of</strong> solid particles, e <strong>the</strong> void ratio and S <strong>the</strong> degree <strong>of</strong><br />
saturation.<br />
Table 2 Basic geotechnical parameters <strong>of</strong> tested samples<br />
w d s e0 S<br />
[kg/m 3 ] [%] [kg/m 3 ] [kg/m 3 ] / [%]<br />
Sample A 1366 65 828 2147 1,59 88<br />
Sample B 1093 31 834 2147 1,57 42<br />
4.2 Consolidation procedure<br />
The initial pressure obtained from <strong>the</strong> pressure plate was 4,7 kN/m 2 . A sequence <strong>of</strong> loadings<br />
was applied, where <strong>the</strong> next load step was applied after <strong>the</strong> completion <strong>of</strong> consolidation<br />
<strong>of</strong> <strong>the</strong> preceding loading step. The loading steps are 36, 82 and 180 kN/m 2 , followed<br />
by unloading to 36 kN/m 2 , <strong>the</strong>n reloading to 365 kN/m 2 and nally unloading to 4,7 kN/m 2 .
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 81<br />
4.3 Oedometric test results and <strong>the</strong>ir evaluation<br />
Figure 8 presents stress-strain relationship for samples A and B respectively.<br />
Figure 8. Stress-strain relationship for one-dimensional compression<br />
The graphs on Figure 8 can be used in order to determine secant compressibility moduli for<br />
usual load increments (Table 3)<br />
Table 3. Secant compressibility moduli<br />
Load 4,7-50 50-100 100-200 200-360<br />
Increment [kN/m 2 ] [kN/m 2 ] [kN/m 2 ] [kN/m 2 ]<br />
Sample A 566 1190 1515 3666<br />
Sample B 657 1316 1639 2860<br />
The values from Table 3 are compared with test results published by o<strong>the</strong>r researchers<br />
(Table 4).<br />
Table 4. Compressibility moduli published by o<strong>the</strong>r researchers<br />
Load (0) 25-50 50-100 100-200 200-400 400-600 280-420 Fraction<br />
increment/<br />
Ref.<br />
[kN/m 2 ] [kN/m 2 ] [kN/m 2 ] [kN/m 2 ] [kN/m 2 ] [kN/m 2 ] [mm]<br />
Duelmann<br />
(2002)<br />
730 1480 2460 4920 / / 0-30<br />
840 1990 1870 3290 / / 0-20<br />
800 1000 1800 / / / 0-20<br />
Kuehle- 1070 1590 1680 2880 / / 0-40<br />
Weidemeier<br />
(2003)<br />
500 1100 1600 2800 / / 0-40<br />
940 1490 2440 3030 / / 0-60<br />
600 1300 2000 2800 / / 0-60
82<br />
Bidlingmaier<br />
et al. (1999)<br />
Ziehmann<br />
(1999)<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
355 1670 1947 2904 4514 / < 60<br />
239 905 1570 2973 4977 / < 100<br />
/ / / / / 5100 /<br />
/ / / / / 7900 < 60<br />
By comparison <strong>of</strong> <strong>the</strong> values <strong>of</strong> compressibility moduli from Table 3 and Table 4 it can be<br />
concluded that <strong>the</strong> values obtained for <strong>the</strong> load increments 0 – 50 (25-50), 50 – 100 and<br />
100 – 200 kN/m2 t well with <strong>the</strong> values <strong>of</strong> compressibility moduli by o<strong>the</strong>r researchers.<br />
The values obtained for load increment 200 – 360 kN/m 2 are somewhat higher than <strong>the</strong><br />
values published by Bidlingmaier et al. (1999) and by Kuehle-Weidemeier (2003). Yet according<br />
to Duellmann (2002) i Ziehmann (1999) it is possible to obtain even higher values<br />
<strong>of</strong> compressibility moduli for approximately same load increment.<br />
In addition, Figure 9 presents comparison <strong>of</strong> measured stress-strain curves with stress-strain<br />
curves <strong>of</strong> MBT waste published by Duelmann (2002) and Kuehle-Weidemeier (2003).<br />
Figure 9 Comparison <strong>of</strong> measured stress-strain relationships with stress-strain relationships<br />
published by o<strong>the</strong>r researchers<br />
Figure 9 shows a good agreement, with respect to <strong>the</strong> relative deformation, <strong>of</strong> <strong>the</strong> selected<br />
stress-strain relationships with stress-strain relationships published by Duelmann (2002)<br />
and Kuehle-Weidemeier (2003). It should be also emphasized that <strong>the</strong> largest particle size<br />
<strong>of</strong> MBT samples used by Duelmann (2002) and Kuehle-Weidemeier (2003) corresponds<br />
well to <strong>the</strong> largest particle size <strong>of</strong> <strong>the</strong> tested MBT waste material. The elasto-plastic behaviour<br />
<strong>of</strong> MBT waste dened by Duelmann (2002) has been also conrmed.<br />
Detailed description <strong>of</strong> <strong>the</strong> oedometric compression tests <strong>of</strong> MBT waste samples is given<br />
by Petrovi et al. (<strong>2010</strong>) and Petrovi et al. (<strong>2011</strong>c).<br />
4.4 Friction angle<br />
In order to estimate <strong>the</strong> critical friction angle c, <strong>the</strong> angle <strong>of</strong> repose <strong>of</strong> <strong>the</strong> dry waste material<br />
was determined. The measurements were conducted on two separate samples at <strong>the</strong><br />
bottom and at <strong>the</strong> top <strong>of</strong> <strong>the</strong> repose in twelve different points. The values measured were<br />
as follows: 370, 310, 310, 410, 380, 380, 350, 400, 410, 380, 500, 430. If <strong>the</strong> highest and<br />
lowest values are considered to be unrealistic and are disregarded, <strong>the</strong>n <strong>the</strong> mean value is<br />
390 with a standard deviation <strong>of</strong> ± 2.450.
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 83<br />
5 Constitutive model<br />
In order to model <strong>the</strong> behaviour under oedometric conditions rst a relation between <strong>the</strong><br />
void ratio e and <strong>the</strong> mean effective pressure p for monotonic compression is considered. In<br />
particular, <strong>the</strong> decrease <strong>of</strong> e with an increase <strong>of</strong> p is described by <strong>the</strong> following exponential<br />
function:<br />
(1)<br />
where <strong>the</strong> constant e0 denotes <strong>the</strong> void ratio for p 0, hoe has <strong>the</strong> dimension <strong>of</strong> stress and n<br />
is a dimensionless constant. It can be noted that function (1) is similar to <strong>the</strong> approximation<br />
function for isotropic compression proposed by Bauer (1996), i.e.<br />
(2)<br />
While in <strong>the</strong> function by Bauer <strong>the</strong> quantities eio, , and hs are dened for <strong>the</strong> isotropic<br />
compression curve starting from <strong>the</strong> loosest possible state <strong>of</strong> <strong>the</strong> grain skeleton, <strong>the</strong> parameters<br />
eo, p, n and hoe in Eq. (1) are related to monotonic oedometric compression. Herein<br />
<strong>the</strong> mean pressure reads p = - (11+ 22+ 33)/3 = -(1+2K0) 11/3, where 11 is <strong>the</strong> vertical<br />
stress, 22 = 33 <strong>the</strong> lateral stress under zero lateral strain and K0 denotes <strong>the</strong> pressure<br />
coefcient at rest, i.e. K0= 22 / 11= 33 / 11. Usually K0 is obtained in experiments<br />
by measuring <strong>the</strong> vertical and horizontal stresses simultaneously. As this was not carried<br />
out in <strong>the</strong> experiments conducted, it is necessary to make estimation for <strong>the</strong> value <strong>of</strong> K0.<br />
The most established semi-empirical function in <strong>engineering</strong> practice is <strong>the</strong> relation proposed<br />
by Jaky (1944), where K0 is related to <strong>the</strong> critical friction angle c according to: K0<br />
= 1 - sin c.<br />
A comprehensive review <strong>of</strong> experimental data obtained from different soils, however,<br />
shows that K0 not only depends on <strong>the</strong> friction angle but also on <strong>the</strong> current density and<br />
<strong>the</strong> pressure level. Bauer (1997), for instance, reported a remarkable agreement between<br />
experiments carried out under different densities and <strong>the</strong> corresponding predictions with<br />
<strong>the</strong> hypoplastic model proposed by Bauer (1996) and Gudehus (1996). The framework<br />
<strong>of</strong> hypoplasticity is <strong>the</strong>refore suitable to get an approximation for K0 and to model <strong>the</strong><br />
inelastic and non-linear stress strain behaviour under axisymmetric loading and unloading.<br />
In particular for oedometric boundary conditions <strong>the</strong> constitutive equations (3-4) are<br />
proposed, which present a simplied version <strong>of</strong> <strong>the</strong> hypoplastic model by Bauer (1996) and<br />
Gudehus (1996).<br />
(3)<br />
(4)
84<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
Assuming a constant grain density, <strong>the</strong> balance equation <strong>of</strong> mass leads to a relation between<br />
<strong>the</strong> rate <strong>of</strong> <strong>the</strong> void rate, , and <strong>the</strong> volumetric strain rate, , i.e. equation (5) can<br />
be derived:<br />
(5)<br />
Herein, is <strong>the</strong> time derivative <strong>of</strong> <strong>the</strong> Cauchy stress are normalized<br />
quantities, is <strong>the</strong> strain rate, and is <strong>the</strong> rate <strong>of</strong> void ratio. Factor is related to <strong>the</strong> critical<br />
friction angle c and for general stress paths also depends on <strong>the</strong> Lode angle in <strong>the</strong><br />
deviator plane (Bauer, 2000). For axisymmetric loading paths factor reduces to (Bauer<br />
& Herle, 2000):<br />
(6)<br />
In Eq. (3) and Eq. (4) <strong>the</strong> density factor >0 is related to <strong>the</strong> limit void ratios as outlined<br />
in more detail by Bauer (1996) and Gudehus (1996). In particular 1 for a loose material. In <strong>the</strong> present paper is assumed to be constant.<br />
The compression law (1) is embedded in <strong>the</strong> stiffness factor using <strong>the</strong> consistency condition<br />
for oedometric compression, which leads to <strong>the</strong> following relation:<br />
(7)<br />
Under monotonic oedometric compression <strong>the</strong> following expression is obtained from Eq.<br />
(3) and Eq. (4) for <strong>the</strong> stress ratio<br />
(8)<br />
K 0<br />
in Eq. (8) is also called <strong>the</strong> pressure coefcient at rest. It depends on <strong>the</strong> critical friction<br />
angle c and <strong>the</strong> density factor . From Figure 10 it can be seen that Eq. (8) allows a ner<br />
adaptation <strong>of</strong> K 0<br />
. In particular higher friction angles c and lower values <strong>of</strong> <strong>the</strong> density<br />
factor give signicantly lower values for K 0<br />
while Jaky’s relation is independent <strong>of</strong> <strong>the</strong><br />
current density.<br />
The hypoplastic model proposed for MBT waste material includes ve constants. Taking<br />
into account Eq. (8) for K 0<br />
, a critical friction angle <strong>of</strong> c = 39º and =1.1, <strong>the</strong> values e o<br />
,<br />
h oe<br />
and n <strong>of</strong> <strong>the</strong> approximation function (1) can be calibrated. For Sample A, a comparison<br />
<strong>of</strong> <strong>the</strong> results obtained from Eq. (1) with <strong>the</strong> experiments is shown in Figure 11. For <strong>the</strong><br />
numerical simulations discussed in Section 6 <strong>the</strong> constants used are presented in Table 5.
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 85<br />
Figure 10. Comparison <strong>of</strong> K 0<br />
values obtained with <strong>the</strong> hypoplastic constitutive model (HM) for<br />
different density factor and friction angles c with <strong>the</strong> function by Jaky (1944)<br />
Table 5. Constants used for numerical simulations<br />
hs<br />
[kN/m 2 ]<br />
n e0 fd<br />
[0]<br />
Sample A 3519 0.389 1.75 1.1 39<br />
Figure 11. Comparison <strong>of</strong> <strong>the</strong> oedometric compression relation according to Eq. (1) with <strong>the</strong><br />
experimental results for Sample A<br />
6 Comparison <strong>of</strong> numerical simulations with experiments<br />
In <strong>the</strong> following <strong>the</strong> results obtained from numerical simulations <strong>of</strong> oedometric compression<br />
tests are compared with <strong>the</strong> corresponding experiments by Petrovi (<strong>2010</strong>). All simulations<br />
are conducted under drained conditions. Figure 12 show a good agreement between<br />
<strong>the</strong> results obtained from <strong>the</strong> numerical simulation (solid curve) and <strong>the</strong> experimental data<br />
(dots). Is should be noticed that with <strong>the</strong> presented hypoplastic model only loading and<br />
unloading can be modelled. For <strong>the</strong> simulation <strong>of</strong> loading cycles an extension <strong>of</strong> <strong>the</strong> constitutive<br />
model is needed as proposed for instance by Bauer & Wu (1992), Niemunis &<br />
Herle (1997).
86<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
Figure 12. Oedometric compression and extension <strong>of</strong> Sample A<br />
Detailed description <strong>of</strong> <strong>the</strong> numerical modelling <strong>of</strong> MBT waste deformability is given by<br />
Petrovi et al. (<strong>2011</strong>a).<br />
7 Conclusions<br />
A 500 mm internal diameter Casagrande-type oedometer has been designed, manufactured,<br />
and assembled at <strong>the</strong> University <strong>of</strong> Zagreb. The manufacturing costs were about 17.000,00<br />
EUR. It was a ra<strong>the</strong>r demanding task to produce a laboratory device capable to give reliable<br />
results with very limited nancial resources. In order to achieve this goal many innovative<br />
and unusual technical solutions were applied. Some <strong>of</strong> <strong>the</strong> innovative parts <strong>of</strong> <strong>the</strong><br />
large oedometer are currently in <strong>the</strong> process <strong>of</strong> legal protection (patent registration). The<br />
laboratory results obtained with <strong>the</strong> new device prove that even with limited resources it is<br />
still possible to achieve a respectable scientic result as long as scientists keep <strong>the</strong>ir minds<br />
wide open.<br />
The results obtained from <strong>the</strong> evaluation program indicate that <strong>the</strong>y are comparable to <strong>the</strong><br />
results published by o<strong>the</strong>r authors. The particle size distribution curve conrmed that <strong>the</strong><br />
MBT waste can be classied as a coarse grained material. A good agreement <strong>of</strong> stressstrain<br />
curves for samples A and B proves that new apparatus can provide repeatability <strong>of</strong><br />
<strong>the</strong> oedometer test. It has been veried that <strong>the</strong> magnitude <strong>of</strong> vertical relative deformation<br />
is in <strong>the</strong> range 15 - 25 % for given load range. The elasto-plastic behaviour <strong>of</strong> MBT waste<br />
has been also conrmed. Simplied hypoplastic model proved to be satisfactory for modelling<br />
<strong>the</strong> compaction behaviour <strong>of</strong> MBT waste observed in experiments<br />
A complete presentation <strong>of</strong> <strong>the</strong> laboratory testing and numerical modelling <strong>of</strong> MBT waste<br />
deformability toge<strong>the</strong>r with <strong>the</strong> recommendations for future research is given by Petrovi<br />
(<strong>2010</strong>).<br />
8 Acknowledgements<br />
The nancial support <strong>of</strong> <strong>the</strong> Ministry <strong>of</strong> Science, Education and Sports <strong>of</strong> <strong>the</strong> Republic <strong>of</strong><br />
Croatia for <strong>the</strong> project "Characterization <strong>of</strong> municipal solid waste" (160-0831529-3031) is<br />
gratefully acknowledged.
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 87<br />
References<br />
1. Bauer E (1996) Calibration <strong>of</strong> a comprehensive hypoplastic model for granular materials.<br />
Soils and Foundations., Vol. 36(1), pp 13-26<br />
2. Bauer E (1997) The critical state concept in hypoplasticity. 9th International Conference<br />
on Computer Methods and Advances in Geomechanics, IACMAG 97, Wuhan,<br />
China, Balkema, pp 691–696.<br />
3. Bauer E (2000) Conditions for embedding Casagrande’s critical states into hypoplasticity.<br />
Mechanics <strong>of</strong> Cohesive-Frictional Materials, 5. pp 125-148<br />
4. Bauer E, Wu W (1992) A hypoplastic model for granular soils under cyclic loading, In:<br />
Kolymbas (ed.) Proceedings <strong>of</strong> <strong>the</strong> International Workshop on Modern Approaches to<br />
Plasticity, Elsevier, pp 225-245<br />
5. Bidlingmaier W, Scheelhaase T, Maile A (1999) Langzeitverhalten von mechanischbiologisch<br />
vorbehandeltem Restmuell auf der Deponie, Abschlußbericht zum Teilvorhaben<br />
3.1 des BMBF-Verbundvorhabens „Mechanisch-biologische Behandlung<br />
von zu deponierenden Abfaellen“, Universitaet Gesamthochschule Essen, Fachbereich<br />
10 – Bauwesen, Fachgebiet Abfallwirtschaft, 1999<br />
6. Casagrande A (1936) The Determination <strong>of</strong> Preconsolidation Load and its Practical<br />
Signicance. Proceedings, 1st International Conference on Soil Mechanics Foundation<br />
Engineering, Cambridge, Vol. 3, pp. 60–64.<br />
7. Council Directive 1999/31/EC <strong>of</strong> 26 April 1999 on <strong>the</strong> landll waste, Ofcial Journal<br />
<strong>of</strong> <strong>the</strong> European Communities, L 182, 16.7.1999.<br />
8. Domanovac T, Orašanin R (2002) Municipal solid waste composition for continental<br />
and coastal part <strong>of</strong> Croatia. In: Milanovi Z. (ed.) Proceedings <strong>of</strong> VIIth International<br />
Symposium Waste Management Zagreb 2002, mtg-topgraf, Zagreb, pp 61-68 (in Croatian)<br />
9. Duellmann H (2002) Untersuchungen zum Einbau von MBA-Abfaellen auf der<br />
Zentraldeponie Hannover, Laboruntersuchungen zum Verdichtungs-, Durchlaessigkeits-,<br />
Last-Setzungs- und Scherverhalten. Februar 2002. Im Auftrag des Abfallwirtschaftsbetriebes<br />
Hannover<br />
10. Geotechnics <strong>of</strong> Landll Design and Remedial Works Technical Recommendations –<br />
GLR, Second Edition, Ernst & Sohn, Berlin, 1993.<br />
11. Gudehus G (1996) A comprehensive constitutive equation for granular materials. Soils<br />
and Foundations, Vol 36(1), pp 1-12<br />
12. Jaky J (1944) The Coefcient <strong>of</strong> Earth Pressure at Rest. Journal for Society <strong>of</strong> Hungarian<br />
Architects and Engineers, pp 355-358<br />
13. Jessberger HL, Kockel R (1995) Determination and assessment <strong>of</strong> <strong>the</strong> mechanical properties<br />
<strong>of</strong> waste materials. In: Sarsby (ed.) Waste Disposal by Landlls – GREEN'93,<br />
Balkema, Bolton, pp 313-322<br />
14. Kuehle-Weidemeier M. (2003) Landlling and properties <strong>of</strong> MBP waste, Proceedings<br />
Sardinia 2003, Ninth International Waste Management and Landll Symposium, S.<br />
Margherita di Pula, Cagliari, Italy<br />
15. Niemunis A, Herle I (1997) Hypoplastic model for cohesionless soils with elastic<br />
16. strain range. Mechanics <strong>of</strong> Cohesive-Frictional Materials, Vol. 2(4), pp 279-299<br />
17. Ordinance on <strong>the</strong> methods and conditions for <strong>the</strong> landll <strong>of</strong> waste, categories and<br />
operational requirements for waste landlls (2007) Ofcial Gazette <strong>of</strong> <strong>the</strong> Republic <strong>of</strong><br />
Croatia, No. 117/07 (in Croatian)<br />
18. Petrovi I (2008) Deformability <strong>of</strong> mechanically and biologically treated municipal<br />
solid waste, In: Milanovi Z. (ed.) Proceedings <strong>of</strong> Xth International Symposium Waste<br />
Management Zagreb 2008, Gospodarstvo i okoliš, Zagreb, pp 329-334 (in Croatian)
88<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
19. Petrovi I (<strong>2010</strong>) Modelling <strong>the</strong> behaviour <strong>of</strong> mechanically and biologically treated<br />
municipal solid waste. PhD Thesis, University <strong>of</strong> Zagreb, Faculty <strong>of</strong> Civil Engineering,<br />
Zagreb, Croatia (in Croatian)<br />
20. Petrovi I, Kovai D (<strong>2010</strong>) Measurement <strong>of</strong> stiffness modulus <strong>of</strong> mechanically and<br />
biologically treated waste, In: Milanovi Z. (ed.) Proceedings <strong>of</strong> XIth International<br />
Symposium Waste Management Zagreb <strong>2010</strong>, Udruga za gospodarenje otpadom,<br />
Zagreb,(in Croatian)<br />
21. Petrovi I, Bauer E (<strong>2011</strong>a) A simple hypoplastic model for simulating <strong>the</strong> mechanical<br />
behaviour <strong>of</strong> MBT waste, In: The Second International Symposium on Computational<br />
Geomechanics (ComGeo II), Cavtat, Croatia, 27-29 April, <strong>2011</strong><br />
22. Petrovi I, Szavits-Nossan V, Štuhec D (<strong>2011</strong>b) Laboratory testing <strong>of</strong> waste after biomechanical<br />
treatment. Graevinar 63(1), pp 43-53 (in Croatian)<br />
23. Petrovi I, Szavits-Nossan V, Kovai D (<strong>2011</strong>c) Deformability <strong>of</strong> mechanically and<br />
biologically treated municipal solid waste, Graevinar 63(3), pp 255-264 (in Croatian)<br />
24. Waste Management Strategy <strong>of</strong> <strong>the</strong> Republic <strong>of</strong> Croatia (2005) Ofcial Gazette <strong>of</strong> <strong>the</strong><br />
Republic <strong>of</strong> Croatia, No. 130/05 (in Croatian)<br />
25. Waste Management Plan <strong>of</strong> <strong>the</strong> Republic <strong>of</strong> Croatia for <strong>the</strong> period 2007-2015 (2007)<br />
Ofcial Gazette <strong>of</strong> <strong>the</strong> Republic <strong>of</strong> Croatia, No. 85/07 (in Croatian)<br />
26. Ziehmann G (1999) Veraenderung des mechanischen Verhaltens durch die mechanische<br />
und biologische Vorbehandlung, Deponierung von vorbehandelten Siedlungsabfaellen.<br />
Veroeffentlichungen des Zentrums fuer Abfallforschung der Technischen<br />
Universitaet Braunschweig, Heft 14, S. 1 – 9
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 89<br />
BEAM-TO-COLUMN JOINT MODELLING TO EC3<br />
Darko Dujmović<br />
University <strong>of</strong> Zagreb, Faculty <strong>of</strong> Civil Engineering; Croatian Academy <strong>of</strong> Engineering,<br />
Zagreb, Croatia<br />
Boris Androić<br />
IA Projektiranje Structural Engineering L.t.d.; Croatian Academy <strong>of</strong> Engineering, Zagreb,<br />
Croatia<br />
Ivan Lukačević<br />
University <strong>of</strong> Zagreb, Faculty <strong>of</strong> Civil Engineering, Zagreb, Croatia<br />
ABSTRACT<br />
The modelling <strong>of</strong> beam to column joints is analyzed taking into account real behaviour<br />
<strong>of</strong> such joints as described by <strong>the</strong> Mj- curve. General considerations relating to <strong>the</strong> new<br />
joint design philosophy, as based on Eurocode 3, are given in <strong>the</strong> initial part <strong>of</strong> <strong>the</strong> paper.<br />
The method <strong>of</strong> components, an analytical method for characterization <strong>of</strong> joints, is <strong>the</strong>n described.<br />
Components enabling bolted joint between beams and columns are identied and<br />
analyzed. The use <strong>of</strong> <strong>the</strong> described modelling procedure is shown by means <strong>of</strong> a specially<br />
devised computer program CoP (Connection Program).<br />
Key words: beam to column joint, modelling, component method, bolted joint, application<br />
<strong>of</strong> <strong>the</strong> new approach<br />
1 INTRODUCTION<br />
Laboratory research and <strong>the</strong> progress <strong>of</strong> numeric methods have stimulated <strong>the</strong> development<br />
<strong>of</strong> a more realistic approach to dividing joints for everyday <strong>engineering</strong> practice. The component<br />
method is an efcient way <strong>of</strong> solving complex behaviour <strong>of</strong> steel structure joint.<br />
The component method, as well as <strong>the</strong> mechanical models which describe individual joint<br />
types can be used to dene basic joint characteristics. These are: stiffness, bending resistance<br />
and rotation capacity.<br />
2 BEAM-TO-COLUMN JOINT MODEL<br />
Joint analysis requires <strong>the</strong> following to be taken into consideration:<br />
• material non-linearity (plasticity, strain-hardening),<br />
• non-linear contact and slip,<br />
• geometrical non-linearity (local instability),<br />
• residual stresses,<br />
• complex geometrical congurations.<br />
For practical purposes, <strong>the</strong> joint analysis is conducted on simpler models. This is <strong>the</strong> reason<br />
<strong>the</strong> design procedure known as <strong>the</strong> components method has been adopted in Eurocode 3<br />
(EC3). The method suits <strong>the</strong> simplied mechanical spring and rigid links model very well.<br />
The joint is represented by adequate rigid and exible components. They represent individual<br />
joint parts, Fig. 1.
90<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
Fig. 1. Component model <strong>of</strong> beam-to-column joint<br />
Typical bolted joint components are:<br />
• column web panel in shear,<br />
• end-plate in bending,<br />
• beam ange in bending,<br />
• beam web in tension,<br />
• beam ange and web in compression,<br />
• bolts in tension<br />
• welds.<br />
When using <strong>the</strong> component method, <strong>the</strong> rst characteristics to be considered are those <strong>of</strong><br />
resistance, stiffness and ductility in individual components. These characteristics are combined<br />
to reach an estimate <strong>of</strong> mechanical joint behaviour.<br />
A non-linear force-deformation (F-) curve can be used to show individual joint behaviour,<br />
Fig. 2.<br />
Fig. 2. Force-deformation (F-) curve <strong>of</strong> ductile component: a) actual behaviour and b) bi-linear<br />
approximation<br />
Simpler behaviour idealisations are possible. The simplied components model <strong>of</strong> EN<br />
1993-1-8, for instance, combines <strong>the</strong> joint bending response with <strong>the</strong> column web panel<br />
response to shear. This type <strong>of</strong> equivalent rotational spring has been shown in Fig. 3.
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 91<br />
Fig. 3. Equivalent rotational spring model<br />
Fig. 4. Application <strong>of</strong> <strong>the</strong> component method to a welded joint<br />
The application <strong>of</strong> <strong>the</strong> component method requires <strong>the</strong> following steps:<br />
• Identication <strong>of</strong> <strong>the</strong> active components from a global list <strong>of</strong> components (20 different<br />
components currently codied in EN 1993-1-8).<br />
• Evaluation <strong>of</strong> <strong>the</strong> force deformation response (F-) <strong>of</strong> each individual basic component<br />
(initial stiffness, design resistance etc.).<br />
• Assembly <strong>of</strong> <strong>the</strong> active components, using representative mechanical model shown in<br />
Fig. 2, for <strong>the</strong> evaluation <strong>of</strong> <strong>the</strong> mechanical characteristics <strong>of</strong> <strong>the</strong> whole joint (initial stiffness,<br />
design resistance etc. or <strong>the</strong> whole deformability M- curve).<br />
These steps are illustrated in Fig. 4. in simple case <strong>of</strong> a beam-to-column welded joint.
92<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
Understanding <strong>the</strong> mechanical behaviour <strong>of</strong> different joint components allows <strong>the</strong> analysis<br />
<strong>of</strong> a large number <strong>of</strong> different joint congurations. The purpose <strong>of</strong> <strong>the</strong> component method<br />
is to determine <strong>the</strong> F- curve for each individual spring. Only <strong>the</strong> linear stiffness <strong>of</strong> each<br />
component is required to estimate initial joint stiffness. The ductility estimate, however,<br />
requires knowledge <strong>of</strong> <strong>the</strong> non-linear F- response for each component.<br />
The following points encompass only <strong>the</strong> components that are relevant for <strong>the</strong> beam-tocolumn<br />
joint. For <strong>the</strong> joint ductility estimate, <strong>the</strong> joint components are divided into three<br />
groups, analogous with <strong>the</strong> cross-section classication:<br />
• components with high ductility,<br />
• components with limited ductility,<br />
• components with brittle failure.<br />
3 DESIGN RESISTANCE OF COMPONENTS<br />
3.1 Components with high ductility<br />
These components are represented by <strong>the</strong> F- curve which in <strong>the</strong> rst part shows linear<br />
elastic behaviour and in <strong>the</strong> second part allows an increase in deformation with an increase<br />
in force. The component deformation capacity is very large.<br />
• Column web panel in shear<br />
Extensive tests show <strong>the</strong> column web having a resistance reserve in shear. The following<br />
equation gives <strong>the</strong> design shear resistance <strong>of</strong> unstiffened column web panel, V wp,Rd<br />
, for<br />
single-sided framed joints:<br />
(1)<br />
where:<br />
f y,wc<br />
A vc<br />
M0<br />
<strong>the</strong> yield strength <strong>of</strong> <strong>the</strong> column web,<br />
<strong>the</strong> shear area <strong>of</strong> <strong>the</strong> column,<br />
<strong>the</strong> partial factor for resistance <strong>of</strong> cross-sections whatever <strong>the</strong> class is.<br />
In <strong>the</strong> case <strong>of</strong> welded cross-sections, <strong>the</strong> shear column area, Avc, coincides with <strong>the</strong> web<br />
area, whereas in <strong>the</strong> case <strong>of</strong> rolled sections it is given by:<br />
(2)<br />
where:<br />
A c<br />
b c<br />
t fc<br />
t wc<br />
r c<br />
<strong>the</strong> total area <strong>of</strong> <strong>the</strong> column,<br />
<strong>the</strong> ange width,<br />
<strong>the</strong> ange thickness,<br />
<strong>the</strong> web thickness,<br />
<strong>the</strong> root-radius <strong>of</strong> <strong>the</strong> web-ange junction.<br />
Eq. (1) disregards <strong>the</strong> axial force in <strong>the</strong> column. Using <strong>the</strong> Von Mises yield criterion it is<br />
possible to determine <strong>the</strong> reduced resistance value which takes into consideration <strong>the</strong> axial<br />
column force. The 0,9 reduction coefcient partially solves this problem, which has also
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 93<br />
been accepted into EN 1993-1-8. The resistance <strong>of</strong> <strong>the</strong> unstiffened column web in shear for<br />
single-sided joints, V wp,Rd<br />
, is:<br />
(3)<br />
The shear deformation <strong>of</strong> <strong>the</strong> column web, s<br />
, in <strong>the</strong> overall initial rotation <strong>of</strong> joint is given<br />
by:<br />
(4)<br />
where:<br />
V<br />
G<br />
<strong>the</strong> shear force on <strong>the</strong> column web taken as 2F i<br />
(F i<br />
denoting <strong>the</strong> force in each<br />
bolt row and i <strong>the</strong> bolt row),<br />
<strong>the</strong> shear modulus.<br />
Thus <strong>the</strong> corresponding axial stiffness, K wp<br />
, is:<br />
(5)<br />
where:<br />
z<br />
E<br />
<strong>the</strong> lever arm between <strong>the</strong> compressive and <strong>the</strong> tensile areas,<br />
<strong>the</strong> modulus <strong>of</strong> elasticity.<br />
Dividing <strong>the</strong> axial stiffness, K wp<br />
, with <strong>the</strong> modulus <strong>of</strong> elasticity E <strong>the</strong> Eq. (5) is:<br />
(5a)<br />
where:<br />
k wp<br />
<strong>the</strong> stiffness coefcient <strong>of</strong> column web in shear, <strong>the</strong> designation in EN 1993-1-8<br />
is k1.<br />
From Eq. (4) it can be observed that <strong>the</strong> stiffness <strong>of</strong> this component depends on <strong>the</strong> applied<br />
shear force on <strong>the</strong> column web. Given that, in general, internal forces transmitted by<br />
<strong>the</strong> lower and upper column and (for internal nodes with unbalanced moments) left beam<br />
may also be present, <strong>the</strong> applied shear force must also be modied by a transformation<br />
parameter , EN 1993-1-8, to deal with this effect. For different joint congurations and<br />
unbalanced bending moments, <strong>the</strong> stiffness coefcient, k<br />
WP<br />
, Eq. (5a), must be modied by<br />
<strong>the</strong> transformation parameter :
94<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
(6)<br />
For stiffened column web shear deformation can be disregarded (k wp<br />
= ). It ought to be<br />
noted that in slender webs instability becomes relevant, but has so far not been included<br />
in EC 3.<br />
• End-plate in bending<br />
The deformation <strong>of</strong> this component is evaluated using <strong>the</strong> simple substitute model, T-stub.<br />
It represents <strong>the</strong> tension joint zone behaviour, and is shown in Fig. 5.<br />
Fig. 5. Substitute model, <strong>the</strong> T-stub<br />
Three possible failure modes <strong>of</strong> <strong>the</strong> equivalent T-stub ange are assumed, as shown in Fig.<br />
6.<br />
The following failure modes are possible:<br />
Fig. 6. Failure modes <strong>of</strong> <strong>the</strong> T-stub ange<br />
Mode 1: End-plate yielding without bolt failure
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 95<br />
(7a)<br />
Mode 2: Simultaneous yielding <strong>of</strong> end-plate with bolt failure<br />
(7b)<br />
Mode 3: Bolt failure without end-plate yielding<br />
where:<br />
m <strong>the</strong> distance between <strong>the</strong> bolt centreline and <strong>the</strong> face <strong>of</strong> <strong>the</strong> weld connecting <strong>the</strong><br />
beam web to <strong>the</strong> end-plate, see Fig. 7,<br />
n <strong>the</strong> effective distance to <strong>the</strong> free edge, , but , see Fig. 7,<br />
F t,Rd<br />
<strong>the</strong> design tension resistance <strong>of</strong> a bolt,<br />
<strong>the</strong> total value <strong>of</strong> F t,Rd<br />
for all bolts in <strong>the</strong> T-stub.<br />
F t,Rd<br />
Design bending resistance <strong>of</strong> an end-plate depending on failure mode is:<br />
(7c)<br />
(8a)<br />
(8b)<br />
where:<br />
tf<br />
fy<br />
<strong>the</strong> effective width <strong>of</strong> <strong>the</strong> end-plate in bending,<br />
<strong>the</strong> thickness <strong>of</strong> <strong>the</strong> end-plate,<br />
<strong>the</strong> yield strength <strong>of</strong> <strong>the</strong> end-plate.<br />
Dimensions <strong>of</strong> an equivalent T-stub ange are shown in Fig. 7.<br />
For end-plate bending resistance, <strong>of</strong> <strong>the</strong> possible three failure modes, <strong>the</strong> minimum value<br />
is used.<br />
The T-stub stiffness coefcient analytical equations can be derived from classical beam<br />
<strong>the</strong>ory, when <strong>the</strong> effective width is correctly estimated, so that:<br />
(9)
96<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
where:<br />
k b,p<br />
<strong>the</strong> stiffness coefcient <strong>of</strong> <strong>the</strong> end-plate in bending, <strong>the</strong> designation in EN 1993-<br />
1-8 is k5,<br />
t p<br />
is <strong>the</strong> thickness <strong>of</strong> a end-plate.<br />
Fig. 7. Dimensions <strong>of</strong> an equivalent T-stub ange<br />
• Column ange in bending<br />
This component behaves similarly to <strong>the</strong> end-plate in bending, so <strong>the</strong> equivalent T-stub<br />
approach can also be utilized. The exception is <strong>the</strong> case when <strong>the</strong> ange is stiffened. The<br />
unstiffened ange is supposed to have <strong>the</strong> same levels <strong>of</strong> ductility and stiffness. The resistance<br />
equations dependent on failure modes are given in <strong>the</strong> following:<br />
Mode 1: Column ange yielding without bolt failure<br />
(10a)<br />
Mode 2: Simultaneous yielding <strong>of</strong> column ange with bolt failure<br />
(10b)<br />
Mode 3: Bolt failure without column ange yielding<br />
(10c)<br />
For column ange bending resistance <strong>the</strong> value which is used is <strong>the</strong> minimum out <strong>of</strong> <strong>the</strong><br />
three failure modes.
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 97<br />
The component stiffness coefcient is:<br />
(11)<br />
The symbols have <strong>the</strong> same meaning as in <strong>the</strong> case <strong>of</strong> <strong>the</strong> end-plate, with <strong>the</strong> exception that<br />
<strong>the</strong> column and not <strong>the</strong> end-plate ange characteristics are used. The designation <strong>of</strong> this<br />
stiffness coefcient in EN 1993-1-8 is k4.<br />
• Beam web in tension<br />
For bolted end-plate joints, <strong>the</strong> design tension resistance <strong>of</strong> <strong>the</strong> beam web, Ft,wb,Rd, is:<br />
(12)<br />
where:<br />
b eff,t,wb<br />
t wb<br />
f y,wb<br />
<strong>the</strong> effective width <strong>of</strong> <strong>the</strong> beam web in tension, equal to <strong>the</strong> effective length <strong>of</strong><br />
<strong>the</strong> equivalent T-stub representing <strong>the</strong> end-plate in bending,<br />
<strong>the</strong> thickness <strong>of</strong> beam web,<br />
<strong>the</strong> yield strength <strong>of</strong> <strong>the</strong> beam web.<br />
The initial stiffness for this component can be taken to be innite.<br />
3.2 Components with limited ductility<br />
These components are characterized by an F- curve with limit point after which <strong>the</strong> curve<br />
falls.<br />
• Column web in compression<br />
This component achieves limited ductile behaviour when <strong>the</strong> curve falls after <strong>the</strong> maximum<br />
resistance has been reached. The resistance for this component can be grouped according<br />
to two different types <strong>of</strong> criteria:<br />
- crushing resistance,<br />
- buckling resistance.<br />
The crushing resistance must take into account <strong>the</strong> interaction between:<br />
- <strong>the</strong> local stresses that arise from <strong>the</strong> shear stresses in <strong>the</strong> panel zone,<br />
- <strong>the</strong> vertical normal stresses due to axial load and bending moment in <strong>the</strong> column,<br />
- <strong>the</strong> horizontal normal stresses transmitted by <strong>the</strong> beam anges.<br />
Using <strong>the</strong> von Mises yield criterion, <strong>the</strong> design crushing resistance, Fc,wc,Rd, is:<br />
(13)<br />
The effective width <strong>of</strong> <strong>the</strong> column web in compression, b eff,c,wc<br />
, for bolted end-plate joints<br />
is:
98<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
(14)<br />
where:<br />
a p<br />
<strong>the</strong> effective thickness <strong>of</strong> <strong>the</strong> weld, Fig. 8,<br />
s = r for rolled column sections,<br />
<strong>the</strong> length obtained by dispersion at 45º through <strong>the</strong> end-plate.<br />
s p<br />
Fig. 8. Transverse compression on an unstiffened column<br />
Reduction factor, k wc<br />
, accounts for <strong>the</strong> inuence <strong>of</strong> vertical normal stress due to axial force<br />
and <strong>the</strong> bending moment, com,Ed<br />
, and is given in <strong>the</strong> following:<br />
(15)<br />
Reduction factor accounts for <strong>the</strong> shear interaction, and equals:
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 99<br />
(16)<br />
with<br />
(17a)<br />
(17b)<br />
where:<br />
is <strong>the</strong> transformation parameter from EN 1993-1-8.<br />
The buckling resistance, F c,wc,Rd<br />
, is taken approximately using <strong>the</strong> Winter equation:<br />
(18)<br />
The reduction factor for plate buckling (column web panel), , is:<br />
or<br />
(19a)<br />
(19b)<br />
The normalised plate slenderness (column web panel),<br />
, is:<br />
(20)<br />
The initial deformation <strong>of</strong> <strong>the</strong> c component is:
100<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
(21)<br />
where:<br />
N<br />
A c<br />
d c<br />
h c<br />
<strong>the</strong> resultant compressive force, taken as 2Fi (Fi denote <strong>the</strong> force in each bolt<br />
row and i <strong>the</strong> bolt row),<br />
<strong>the</strong> effective web area in compression zone, Ac = twc•beff,c,wc,<br />
<strong>the</strong> depth between column llets,<br />
<strong>the</strong> beam depth minus beam ange thickness.<br />
So that <strong>the</strong> initial (axial) stiffness becomes:<br />
(22)<br />
Dividing <strong>the</strong> axial stiffness, K c,wc<br />
, with <strong>the</strong> modulus <strong>of</strong> elasticity E <strong>the</strong> Eq. (22) is:<br />
(22a)<br />
It must be noted that for <strong>the</strong> calculation <strong>of</strong> <strong>the</strong> stiffness coefcient, kc,wc, a reduction <strong>of</strong> <strong>the</strong><br />
effective width used for <strong>the</strong> resistance calculation is adopted (0,7×b eff,c,wc<br />
). The designation<br />
<strong>of</strong> this stiffness coefcient in EN 1993-1-8 is k 2<br />
.<br />
• Column web in tension<br />
Excluding instability occurrences, <strong>the</strong> resistance <strong>of</strong> this component approaches <strong>the</strong> column<br />
web in compression. The design resistance <strong>of</strong> an unstiffened column web, F t,wc,Rd<br />
, equals:<br />
(23)<br />
where <strong>the</strong> symbols mean <strong>the</strong> same as earlier, with t (tension) replacing c (compression). It<br />
shall be noted that EN 1993-1-8 disregards <strong>the</strong> inuence <strong>of</strong> vertical stresses arising from<br />
<strong>the</strong> column.<br />
Analogous to <strong>the</strong> preceding case, <strong>the</strong> initial deformation <strong>of</strong> this component w<br />
is:<br />
(24)<br />
where:<br />
T <strong>the</strong> resultant tensile force, taken as 2Fi (Fi denoting <strong>the</strong> force in each bolt row<br />
and i <strong>the</strong> bolt row),<br />
A t<br />
<strong>the</strong> effective web area in <strong>the</strong> tensile zone, A t<br />
= t wc<br />
×b eff,t,wc<br />
,
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 101<br />
d c<br />
h t<br />
<strong>the</strong> depth between column llets,<br />
<strong>the</strong> distance from <strong>the</strong> tensile force to <strong>the</strong> centre <strong>of</strong> compression.<br />
Initial (axial) stiffness equals:<br />
(25)<br />
Dividing <strong>the</strong> axial stiffness, K t,wc<br />
, with <strong>the</strong> modulus <strong>of</strong> elasticity E <strong>the</strong> Eq. (25) is:<br />
(25a)<br />
This means that Eq. (25a) provides <strong>the</strong> stiffness coefcient, k t,wc<br />
, <strong>of</strong> <strong>the</strong> column web in tension.<br />
The designation <strong>of</strong> this stiffness coefcient in EN 1993-1-8 is k3.<br />
• Beam ange and beam web in compression<br />
This component, <strong>the</strong> beam ange and web in compression adjacent to <strong>the</strong> connection <strong>of</strong><br />
beam, provides a limitation to <strong>the</strong> resistance <strong>of</strong> <strong>the</strong> joint. It is <strong>the</strong>refore necessary to determine<br />
maximum component resistance, using:<br />
(26)<br />
Its initial stiffness is taken as innity.<br />
3.3 Components with brittle failure<br />
These components behave linearly until collapse. Before failure <strong>the</strong>y show very small deformation.<br />
• Bolts in tension<br />
The bolts show a linear force-deformation (F-) response up to failure. The resistance and<br />
initial stiffness <strong>of</strong> each bolt are:<br />
(27)<br />
(28)<br />
where:<br />
A s<br />
f ub<br />
L b<br />
<strong>the</strong> tensile area <strong>of</strong> <strong>the</strong> bolt,<br />
<strong>the</strong> ultimate tensile strength <strong>of</strong> bolts,<br />
<strong>the</strong> sum <strong>of</strong> <strong>the</strong> thickness <strong>of</strong> <strong>the</strong> connected plates (<strong>the</strong> column ange and <strong>the</strong> endplate),<br />
<strong>the</strong> thickness <strong>of</strong> <strong>the</strong> washers and <strong>the</strong> half thickness <strong>of</strong> <strong>the</strong> nut and <strong>the</strong> bolt<br />
head.
102<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
• Welds<br />
Welds are virtually undeformable (Kw = ). A rigid-plastic model is adopted as adequate,<br />
and <strong>the</strong> design resistance is:<br />
(29)<br />
where:<br />
a<br />
f u<br />
w<br />
<strong>the</strong> effective thickness <strong>of</strong> <strong>the</strong> weld,<br />
<strong>the</strong> ultimate tensile strength <strong>of</strong> <strong>the</strong> weld,<br />
<strong>the</strong> correlation factor.<br />
4 ASSEMBLY OF THE COMPONENTS<br />
Components assembly is <strong>the</strong> third and <strong>the</strong> last step <strong>of</strong> <strong>the</strong> components method. It consists<br />
<strong>of</strong> assembling components in such a way that <strong>the</strong> mechanical characteristics <strong>of</strong> <strong>the</strong> entire<br />
joint can be reached. The relationship between <strong>the</strong> components characteristics and <strong>the</strong> joint<br />
components is based on “<strong>the</strong> distribution <strong>of</strong> internal forces in <strong>the</strong> joint”. The manner in<br />
which <strong>the</strong> external forces bearing on <strong>the</strong> joint are distributed to joint components is determined<br />
for <strong>the</strong> given set (analogy with structural member cross-section).<br />
Fig. 9 shows <strong>the</strong> example <strong>of</strong> a spring model for bolted joint with end-plate. Each component<br />
possesses stiffness coefcient, k i<br />
. The springs are connected serially and/or parallely,<br />
and <strong>the</strong> initial rotational stiffness, S j,ini<br />
, is:<br />
(30)<br />
The bending resistance <strong>of</strong> <strong>the</strong> joint, M j,Rd<br />
, equals:<br />
(31)<br />
where:<br />
F tr,Rd<br />
<strong>the</strong> effective design tension resistance <strong>of</strong> bolt-row r,<br />
h r<br />
<strong>the</strong> distance from bolt-row r to <strong>the</strong> centre <strong>of</strong> compression,<br />
r <strong>the</strong> bolt-row number.
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 103<br />
Fig. 9. Components represented by springs on bolted beam to column joint<br />
5 PRACTICAL APPLICATION OF THE COMPONENTS METHOD<br />
The example shows <strong>the</strong> new approach to steel beam-to-column joint design adopted in<br />
EC3, in practical application. The beam to column joint in Fig. 10 is a real industrial portal<br />
frame joint with an inclined beam ( = 5°), <strong>the</strong> length <strong>of</strong> L = 19,5 m, and h = 6,32 high (in<br />
<strong>the</strong> ridge H = 7,20 m). A haunch 1,5 m long supports <strong>the</strong> beam. The design <strong>of</strong> this kind <strong>of</strong><br />
joint is based on <strong>the</strong> traditional approach and was intuitively held to be rigid.<br />
The second step is <strong>the</strong> design <strong>of</strong> mechanical joint characteristics (joint No. 1) conducted<br />
based on <strong>the</strong> EC3, Part 1.8, and <strong>the</strong> values are:<br />
• Initial rotational stiffness S j,ini = 260 863 kNm/rad,<br />
• Bending resistance M j,Rd = 554 kNm.<br />
In order to agree with <strong>the</strong> designer’s expectations, every value has to be higher than:<br />
• <strong>the</strong> inexibility limit, based on <strong>the</strong> rotational stiffness 129 769 kNm/rad,<br />
• <strong>the</strong> maximum bending moment, transferred to <strong>the</strong> joint in global frames analysis, equals<br />
M j,Ed<br />
= 416 kNm.<br />
Since both conditions were fullled, <strong>the</strong> joint design can be considered satisfactory.
104<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
Fig. 10. Detail <strong>of</strong> <strong>the</strong> joint<br />
The joint stiffness pro<strong>of</strong> based on <strong>the</strong> EN 1993-1-8 has not been conducted, and <strong>the</strong> joint<br />
was simply taken as inexible based on <strong>the</strong> designer’s experience.<br />
According to <strong>the</strong> EC3 additional efforts are needed to check whe<strong>the</strong>r <strong>the</strong> rigid joint could<br />
be considered unnecessary.<br />
Before any fur<strong>the</strong>r discussion, <strong>the</strong> following needs to be established:<br />
• A lack <strong>of</strong> knowledge necessary to design joints, when <strong>the</strong> designer’s joint geometry is<br />
dened, leads to a systematic application <strong>of</strong> transversal stiffeners <strong>of</strong> <strong>the</strong> column (in both<br />
<strong>the</strong> tension and compression joint zones), toge<strong>the</strong>r with thick end-plates,<br />
• As a result <strong>of</strong> excessive stiffening <strong>the</strong> internal force distribution is <strong>of</strong> a linear kind, which<br />
explains why <strong>the</strong> elastic design approach has traditionally been used for determining bolt<br />
forces,<br />
In such design procedures, <strong>the</strong> bolts are weak joint components and <strong>the</strong> plate approach to<br />
design can be considered valid, but it <strong>of</strong>ten results in an excessive dimensioning <strong>of</strong> joints<br />
and uneconomical joint framings, since <strong>the</strong> execution depends on <strong>the</strong> stiffness level to such<br />
an extent. Stiffness even prevents easy assembly at site. Finally, <strong>the</strong> lack <strong>of</strong> ductility related<br />
to bolt failure is far from satisfactory, when joint ductility is taken into consideration.<br />
Part 1.8 <strong>of</strong> <strong>the</strong> EC 3 suggests as an alternative a set <strong>of</strong> rules for design, where <strong>the</strong> type <strong>of</strong><br />
failure has not been chosen in advance. The stiffness and resistance characteristics for all<br />
participating components have been taken into account in <strong>the</strong> design. In o<strong>the</strong>r words, real<br />
joint characteristics are calculated, and <strong>the</strong> designer is left with <strong>the</strong> decisions to change <strong>the</strong><br />
joint geometry as needed, choosing a ductile type <strong>of</strong> failure, thus making <strong>the</strong> joint more<br />
resistant, stiffer and more ductile.<br />
In <strong>the</strong> case at hand, <strong>the</strong> joint characteristics exceed requirements (<strong>the</strong> resistance is higher<br />
than required, and <strong>the</strong> rotational stiffness <strong>of</strong> 129 769 kNm/rad is enough to secure rigidity<br />
in <strong>the</strong> joint, as was supposed in <strong>the</strong> frame analysis). Simplifying joint geometry thus<br />
decreases <strong>the</strong> joint execution costs, and those <strong>of</strong> <strong>the</strong> entire structure based on that. A more<br />
signicant fact here is that in this way we design joints that will achieve <strong>the</strong> required levels<br />
<strong>of</strong> reliability. Intuitive stiffness forecasts, though made to create a more reliable joint, may<br />
have an effect that is just <strong>the</strong> opposite.<br />
The following table shows different structural joint congurations. They are under consideration<br />
with <strong>the</strong> goal <strong>of</strong> simplifying joint geometry, taking into account <strong>the</strong> frame analysis<br />
requirements.
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 105<br />
• Joint No. 1: 6 bolt rows, leng<strong>the</strong>ned end-plate, 6 stiffeners <strong>of</strong> column web (three pairs).<br />
• Joint No. 2: 5 bolt rows, 6 stiffeners <strong>of</strong> column web.<br />
• Joint No. 3: 5 bolt rows, 4 stiffeners <strong>of</strong> column web (compression and tension zone).<br />
• Joint No. 4: 5 bolt rows, 2 stiffeners <strong>of</strong> column web (compression zone).<br />
• Joint No. 5: 5 bolt rows, no stiffeners <strong>of</strong> column web.<br />
• Joint No. 6: 4 bolt rows, 2 stiffeners <strong>of</strong> column web (compression zone).<br />
• Joint No. 7: 3 bolt rows, 2 stiffeners <strong>of</strong> column web (compression zone).<br />
• Joint No. 1, which <strong>the</strong> designer has suggested, fulls <strong>the</strong> frame structure analysis requirements,<br />
where <strong>the</strong> stiffness is such that <strong>the</strong> joint can be considered rigid, and <strong>the</strong> bending<br />
resistance is greater than <strong>the</strong> bending moment, which was reached through global frame<br />
analysis.<br />
• Since <strong>the</strong> joint resistance is signicantly greater than required, it is concluded that <strong>the</strong><br />
extended end-plate is unnecessary, and <strong>the</strong> upper bolt row is eliminated, joint No. 2. This<br />
decreases joint stiffness, but within rigidity limits.<br />
• The third step (joint No. 3) is to reduce <strong>the</strong> number <strong>of</strong> stiffeners, by nei<strong>the</strong>r eliminating <strong>the</strong><br />
middle stiffener, which does not decrease joint stiffness nor bending resistance. The middle<br />
stiffener is thus ineffective.<br />
• Next, <strong>the</strong> stiffener is eliminated from <strong>the</strong> tension zone (joint No. 4), and joint bending<br />
resistance and stiffness are reduced, although still completing <strong>the</strong> requirements <strong>of</strong> global<br />
analysis.<br />
• Following that, <strong>the</strong> stiffener would also be eliminated from <strong>the</strong> compression zone. That<br />
sort <strong>of</strong> joint, however (joint No. 5), does not satisfy requirements, since <strong>the</strong> resistance is<br />
lower than <strong>the</strong> action effect.<br />
• Because <strong>the</strong> relevant mode <strong>of</strong> failure is <strong>the</strong> column web in compression, now a step is<br />
taken back to joint No. 4 with stiffeners in <strong>the</strong> compression zone, with <strong>the</strong> elimination <strong>of</strong> 1<br />
row <strong>of</strong> bolts. That kind <strong>of</strong> joint (joint No. 6) is entirely satisfactory.<br />
• The number <strong>of</strong> bolts is <strong>the</strong>n reduced to 3, and joint no. 7 is no longer satisfactory as far as<br />
bending resistance is concerned (<strong>the</strong> spacing between <strong>the</strong> rst and <strong>the</strong> second row <strong>of</strong> bolts<br />
have been increased from 90 to 120 mm).
106<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
Table 1. Joint congurations with corresponding resistance and stiffness values, as<br />
well as frame analysis requirement fullment<br />
Joint No. 6 represents <strong>the</strong> solution. Bending resistance is reduced to an acceptable level,<br />
and as far as stiffness goes, <strong>the</strong> joint still belongs to <strong>the</strong> rigid joint.<br />
For each <strong>of</strong> <strong>the</strong>se structural solutions, Table 1 gives <strong>the</strong> resistance and stiffness values, as<br />
well as whe<strong>the</strong>r <strong>the</strong> frame analysis requirements have been fullled.<br />
This analysis shows very clearly that even for rigid joints, <strong>the</strong> new approach to joint design<br />
based on EC 3 can have serious benets. This is <strong>the</strong> result <strong>of</strong> <strong>the</strong> clearer denition <strong>of</strong> <strong>the</strong><br />
inexibility notion, and <strong>the</strong> usage <strong>of</strong> advanced design models for joint stiffness and resistance<br />
determination.<br />
6 CONCLUSION<br />
An entirely different approach that consists in a realistic representation <strong>of</strong> joints in structure<br />
modelling is shown. Joints in a broader sense connect all types <strong>of</strong> non-monolithic<br />
structural elements and are <strong>the</strong>refore <strong>the</strong> most important within structural elements and
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 107<br />
within <strong>the</strong> entire structure. In spite <strong>of</strong> that and owing to <strong>the</strong> lack <strong>of</strong> knowledge about <strong>the</strong>ir<br />
actual characteristics and <strong>the</strong> way <strong>of</strong> modelling, joints have been so far neglected and very<br />
simplied assumptions related to forming <strong>of</strong> models have been adopted. This has resulted<br />
in expensive structural solutions that <strong>of</strong>ten have very low reliability.<br />
This new approach can be successfully applied in every-day <strong>engineering</strong> practice and fur<strong>the</strong>r<br />
savings can be made due to <strong>the</strong> decrease in <strong>the</strong> self weight <strong>of</strong> structures or reduced<br />
costs <strong>of</strong> execution and assembly. The shown example represents <strong>the</strong> practical application<br />
<strong>of</strong> <strong>the</strong> new approach according to EC 3.<br />
Fur<strong>the</strong>r research <strong>of</strong> joint behaviour should be conducted by use <strong>of</strong> methods <strong>of</strong> structural<br />
reliability <strong>the</strong>ory. With <strong>the</strong> collection <strong>of</strong> statistical data on joint components that would be<br />
obtained in laboratory, <strong>the</strong>ir stochastic models could be established. By solving <strong>the</strong> limit<br />
state equations <strong>the</strong> comparison <strong>of</strong> <strong>the</strong> reliability level <strong>of</strong> different types <strong>of</strong> joints would be<br />
make possible. This way optimal solutions can be chosen in line with general reliability<br />
requirements according to EN 1990: 2001, Eurocode: Basis <strong>of</strong> structural design.<br />
REFERENCES<br />
1. Zoetemeijer, P. (1974) A design method for <strong>the</strong> tension side <strong>of</strong> statically-loaded bolted<br />
beam-to-column joints, Heron, pp 1-59<br />
2. Yee, Y.L.; Melchers, R.E. (1986) Moment-rotation curves for bolted connections,<br />
Journal <strong>of</strong> Structural Engineering 112 3, pp 615-635<br />
3. Janss, J.; Jaspart, J.P. (1987) Strength and behaviour <strong>of</strong> in plane weak axis joints and<br />
<strong>of</strong> 3-D joints, In: Bjorhovde, R.; Colson, A.; Zandonini, R. (Editors): Connections in<br />
steel structures, Proceedings <strong>of</strong> <strong>the</strong> International Workshop on Joints, Elsevier Applied<br />
Science, New York<br />
4. Zoetemeijer, P. (1990) Summary <strong>of</strong> <strong>the</strong> research on bolted beam-to-column connections,<br />
Report 25-6-90-2. Faculty <strong>of</strong> Civil Engineering, Stevin Laboratory - Steel Structures,<br />
Delft University <strong>of</strong> Technology, Delft<br />
5. European Convention for Constructional Steelwork (1992) Analysis Design <strong>of</strong> Steel<br />
Frames with Semi-rigid Joints, Publication 67, Brussels: ECCS<br />
6. Weinand, K. (1992) SERICON – Databank on Joints in Building Frames, Proceedings<br />
<strong>of</strong> <strong>the</strong> 1st COST C1 Workshop, Strasbourg, pp 28-30<br />
7. Shi, Y.J.; Chan, S.L.; Wong, Y.L. (1996) Modelling for moment rotations characteristic<br />
for end-plate joints, Journal <strong>of</strong> Structural Engineering 122 11, pp 1300-1306<br />
8. Bursi, O. S.; Jaspart, J. P. (1997) Benchmarks for Finite Element Modelling <strong>of</strong> Bolted<br />
Steel Connections; Journal <strong>of</strong> Constructional Steel Research, Vol. 43, No. 1-3, pp 17-<br />
42<br />
9. Cruz, P. J.S; da Silva, L. S.; Rodrigues, D. S; Simões, R. A. D. (1998) Database for <strong>the</strong><br />
semi-rigid behaviour <strong>of</strong> beam-to-column connections in seismic regions, Journal <strong>of</strong><br />
Constructional Steel Research 46 1-3, pp 233-234<br />
10. Huber, G.; Tschemmernegg, F. (1998) Modelling <strong>of</strong> steel connections, Journal <strong>of</strong> Constructional<br />
Steel Research 45 2, pp 199-216<br />
11. Kuhlmann, U.; Davison, J.B.; Kattner, M. (1998) Structural systems and rotation capacity,<br />
Proceeding <strong>of</strong> COST Conference on Control <strong>of</strong> <strong>the</strong> Semi-rigid Behaviour <strong>of</strong><br />
Civil Engineering Structural Connections, Liège, pp 167-176<br />
12. Weynand, K.; Jaspart, J.-P.; Steenhuis, M. (1998) Economy Studies <strong>of</strong> Steel Building<br />
Frames with Semi-Rigid Joints, Journal <strong>of</strong> Constructional Steel Research 46 1-3<br />
13. Kuhlmann U. (1999) Inuence <strong>of</strong> axial forces on <strong>the</strong> component: web under compression,<br />
Proceeding <strong>of</strong> COST-C1 Working Group Meeting, C1/WG2/99-01, Thessaloniki<br />
14. Faella, C.; Piluso, V.; Rizzano, G. (2000) Structural Steel Semirigid Connections, The-
108<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
ory, Design and S<strong>of</strong>tware, CRC Press LLC, 2000 N.W. Corporate Blvd., Boca Raton,<br />
Florida 33431<br />
15. CoP S<strong>of</strong>tware, (2002) Version 2002R03, RWTH Aachen, MSM Liège, ICCS Ho<strong>of</strong>ddorp<br />
16. da Silva, L. S.; Lima, L.; Vellasco, S.; Andrade, S. (2002) Experimental behaviour <strong>of</strong><br />
end-plate beam-to-column joints under bending and axial force, Database reporting<br />
and discussion <strong>of</strong> results. Report on ECCS-TC10, Meeting in Ljubljana<br />
17. European Committee for Standardization (CEN) (2002) “EN 1990: 2002, Eurocode 0:<br />
Basis <strong>of</strong> structural design“, Final draft, Brussels<br />
18. Dujmovi, D.; Androi, B.; Skeji, D. (2003) Modeliranje prikljuaka elinih okvirnih<br />
konstrukcija, Graevinar 55 6, pp 339-348<br />
19. Dujmovi, D.; Skeji, D.; Androi, B. (2003) Modeliranje prikljuka nosa-stup prema<br />
Eurokodu 3, Graevinar 55 7, pp 397-405<br />
20. Girão Coelho, A.M. (2004) Characterization <strong>of</strong> <strong>the</strong> ductility <strong>of</strong> bolted end plate beamto-column<br />
steel connections, PhD Thesis. University <strong>of</strong> Coimbra, Coimbra, Portugal<br />
21. Girão Coelho, A. M.; Bijlaard, F. S. K.; da Silva, L. S. (2004) Experimental assessment<br />
<strong>of</strong> <strong>the</strong> ductility <strong>of</strong> extended end plate connections, Engineering Structures 26, pp<br />
1185-1206<br />
22. European Committee for Standardization (CEN) (2005) EN 1993-1-8:2005, Eurocode<br />
3: Design <strong>of</strong> steel structures, Part 1.8: Design <strong>of</strong> joints, May 2005, Brussels<br />
23. Skeji, D.; Dujmovi, D.; Androi, B. (2008) Reliability <strong>of</strong> <strong>the</strong> bending resistance<br />
<strong>of</strong> welded beam-to-column joints, Journal <strong>of</strong> Constructional Steel Research, Vol.64,<br />
No.4, pp 388-399
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 109<br />
MODELLING OF JOINT BEHAVIOUR IN STEEL FRAMES<br />
Darko Dujmovic<br />
University <strong>of</strong> Zagreb, Faculty <strong>of</strong> Civil Engineering;<br />
Croatian Academy <strong>of</strong> Engineering<br />
Zagreb, Croatia.<br />
Boris Androic<br />
IA Projektiranje Structural Engineering L.t.d.,<br />
Croatian Academy <strong>of</strong> Engineering,<br />
Zagreb, Croatia<br />
Josip Piskovic<br />
University <strong>of</strong> Zagreb, Faculty <strong>of</strong> Civil Engineering,<br />
Department for Structural Engineering,<br />
Zagreb, Croatia.<br />
ABSTRACT<br />
The interaction between steel frames and <strong>the</strong>ir joints is explained, and <strong>the</strong> joint dimensioning<br />
method according to EC 3 is presented. Four basic terms related to novel approach to<br />
joint dimensioning are described. These terms are: characterization, classication, modelling<br />
and idealisation. An emphasis is placed on <strong>the</strong> advantages <strong>of</strong> component method,<br />
adopted in EC 3, when compared to <strong>the</strong> procedure currently used for joint modelling. Economic<br />
bases for <strong>the</strong> use <strong>of</strong> <strong>the</strong> new joint dimensioning method are also explained.<br />
KEY WORDS: Joints, frame structures, component method, characterization, execution<br />
costs<br />
1 INTRODUCTION<br />
Traditionally <strong>the</strong> beam-column joints have been considered as pinned without any resistance<br />
and stiffness (simple joints) or as completely rigid with full resistance (continuous<br />
joints). This approach is especially convenient for frame structures that are classied as<br />
braced and non-sway, where most beam-column joints do not demand for a moment transfer<br />
(pinned joints). Joints that must transfer moments are usually a constituent part <strong>of</strong> a<br />
bracing system. Therefore, <strong>the</strong>y must be completely rigid. In reality, both <strong>of</strong> <strong>the</strong>se assumptions<br />
for simple and continuous joints can be incorrect and uneconomical, and <strong>the</strong>y present<br />
only <strong>the</strong> limit cases <strong>of</strong> realistic behaviour <strong>of</strong> joints described by <strong>the</strong> relationship <strong>of</strong> <strong>the</strong><br />
bending moment M and <strong>the</strong> rotation angle .<br />
Because <strong>of</strong> insufcient knowledge about <strong>the</strong> behaviour <strong>of</strong> joints in <strong>the</strong> region between limit<br />
cases, procedures have been adopted within <strong>the</strong> traditional approach to designing which<br />
completely neglected this range, and in many cases <strong>the</strong>y provide uneconomical solutions<br />
or solutions that are not so reliable.<br />
The basic problem was <strong>the</strong> fact that had not been taken into consideration and that is that<br />
<strong>the</strong> structural properties <strong>of</strong> joints must be harmonised with structural properties <strong>of</strong> elements<br />
that are being considered. That means that this insufciency was a particular problem in<br />
terms <strong>of</strong> global structural analysis. By introducing a new <strong>engineering</strong> discipline called<br />
structural modelling, <strong>the</strong> insufciencies <strong>of</strong> <strong>the</strong> traditional approach had began to be consid-
110<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
ered. Numerous laboratory researches and <strong>the</strong> development <strong>of</strong> numerical methods encouraged<br />
a faster development <strong>of</strong> <strong>the</strong> idea regarding a more realistic division <strong>of</strong> joints for <strong>the</strong><br />
needs <strong>of</strong> every-day <strong>engineering</strong> practice.<br />
European standards for structures prescribed a system <strong>of</strong> joint division that classies joints<br />
regarding stiffness (nominally pinned, semi-rigid and rigid), and resistance (nominally<br />
pinned, partial-strength and full-strength). Also, joints can be classied regarding ductility,<br />
although <strong>the</strong> criteria for this classication have not clearly been prescribed within Eurocode<br />
3 yet.<br />
2 STRUCTURAL BEHAVIOUR OF JOINTS<br />
A connection is dened as a set <strong>of</strong> physical components that mechanically fastens <strong>the</strong><br />
elements that it connects. It is considered that <strong>the</strong> connection is concentrated at <strong>the</strong> place<br />
where fastening is realized, for example, where <strong>the</strong> end <strong>of</strong> <strong>the</strong> beam and <strong>the</strong> column for <strong>the</strong><br />
beam-column joint around <strong>the</strong> stronger axis meet. When <strong>the</strong> connection and <strong>the</strong> appropriate<br />
zone <strong>of</strong> interaction between <strong>the</strong> connected structural elements are considered toge<strong>the</strong>r,<br />
<strong>the</strong> expression joint is used, Figure 1. Depending on <strong>the</strong> number <strong>of</strong> elements in <strong>the</strong> plane<br />
that are connected, single-sided, Figure 1.a), and double-sided joints congurations, Figure<br />
1.b), are dened.<br />
It is a well-known fact that rotational behaviour <strong>of</strong> actual joints ranges between two limits:<br />
xed and pinned. Figure 2. illustrates such behaviour in an elastic area <strong>of</strong> a single-sided<br />
joint <strong>of</strong> a column and a beam.<br />
Fig. 1. Joints and connections<br />
Fig. 2. Classication <strong>of</strong> joints according to rotational stiffness<br />
Figure 2.a) presents a rigid joint. It is assumed that all elements <strong>of</strong> <strong>the</strong> joint are stiff enough.<br />
There is no difference between rotations at <strong>the</strong> ends <strong>of</strong> <strong>the</strong> structural elements connected
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 111<br />
in <strong>the</strong> joint. The joint rotation is a node rotation usually used in frame structure analysis<br />
methods. A joint that does not have stiffness, Figure 2.b), is nominally pinned. A connected<br />
beam behaves like a simple supported beam. For cases in between, <strong>the</strong> stiffness <strong>of</strong> a joint<br />
is nei<strong>the</strong>r zero nor innite. The moment that <strong>the</strong> joint transfers will have as a consequence<br />
<strong>the</strong> difference <strong>of</strong> between absolute rotations <strong>of</strong> two connected structural elements, Figure<br />
2.c). In this case this concerns a semi-rigid joint.<br />
The simplest way <strong>of</strong> introducing <strong>the</strong>se notions within global analysis is by means <strong>of</strong> a rotational<br />
spring placed between <strong>the</strong> ends <strong>of</strong> <strong>the</strong> structural elements that are being connected.<br />
Rotational stiffness <strong>of</strong> spring S is <strong>the</strong> parameter which links <strong>the</strong> carried-over moment M j<br />
with <strong>the</strong> appropriate rotation , which in turn is <strong>the</strong> difference between absolute rotations<br />
<strong>of</strong> two connected structural elements. Rotational stiffness is dened as <strong>the</strong> slope <strong>of</strong> curve<br />
M j<br />
- which depends on <strong>the</strong> properties <strong>of</strong> <strong>the</strong> joint.<br />
When this rotational stiffness equals zero, or when it is relatively low, <strong>the</strong> joint is in <strong>the</strong><br />
class nominally pinned. On <strong>the</strong> o<strong>the</strong>r hand, when rotational stiffness S is innite, or relatively<br />
large, <strong>the</strong> joint is in <strong>the</strong> class <strong>of</strong> a rigid joint. In all o<strong>the</strong>r in-between cases, <strong>the</strong> joint<br />
belongs to <strong>the</strong> class <strong>of</strong> semi-rigid joints.<br />
Fig. 3. Modelling <strong>of</strong> joints for <strong>the</strong> case <strong>of</strong> an elastic global analysis<br />
For semi-rigid joints, loads will cause <strong>the</strong> bending moment M j<br />
as well as <strong>the</strong> appropriate<br />
rotation between <strong>the</strong> connected structural elements. The moment and <strong>the</strong> appropriate<br />
rotation are linked by a relationship whic depends on <strong>the</strong> structural properties <strong>of</strong> <strong>the</strong> joint.<br />
This relationship is illustrated in Figure 3, where, for <strong>the</strong> purpose <strong>of</strong> simplication, it is<br />
presumed that <strong>the</strong> global analysis was performed with linear elastic assumptions.<br />
Within <strong>the</strong> global analysis, <strong>the</strong> effect that semi-rigid joints have in relation to <strong>the</strong> effect<br />
<strong>of</strong> rigid joints or nominally pinned ones, is not only <strong>the</strong> modication <strong>of</strong> displacement but<br />
also <strong>the</strong> distribution and size <strong>of</strong> internal forces and bending moments in <strong>the</strong> structure. The<br />
concept <strong>of</strong> semi-rigid joints has been introduced into Eurocode 3 for <strong>the</strong> design <strong>of</strong> steel<br />
structures regarding static loads.<br />
3 JOINT REPRESENTATION<br />
3.1 Introduction<br />
Research regarding joints was focused mostly on <strong>the</strong> following aspects:
112<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
• evaluation <strong>of</strong> mechanical properties <strong>of</strong> joints by means <strong>of</strong> rotational stiffness, moment<br />
resistance and rotational capacity,<br />
• procedures <strong>of</strong> analysis and dimensioning <strong>of</strong> frames including <strong>the</strong> behaviour <strong>of</strong> joints.<br />
From this approach it had been understood that <strong>the</strong>re is a step in between <strong>the</strong>se ones that is<br />
to be taken into consideration, with <strong>the</strong> goal <strong>of</strong> consistent integration <strong>of</strong> <strong>the</strong> actual response<br />
<strong>of</strong> <strong>the</strong> joint within <strong>the</strong> frame analysis, and it is called <strong>the</strong> joint representation.<br />
The joint representation includes four steps as follows:<br />
• joint characterisation: determination <strong>of</strong> stiffness, resistance and ductility <strong>of</strong> <strong>the</strong> joint with<br />
<strong>the</strong> help <strong>of</strong> <strong>the</strong> Mj - curve that describes its behaviour,<br />
• classication <strong>of</strong> joints: provides boundary conditions for <strong>the</strong> application <strong>of</strong> traditional<br />
joint modelling types, i.e. pinned or rigid,<br />
• joint idealisation: derivation <strong>of</strong> a simplied Mj - curve so that it is adjusted to <strong>the</strong> specic<br />
procedures <strong>of</strong> frame global analysis, e.g. <strong>the</strong> linear idealisation for <strong>the</strong> elastic analysis,<br />
• joint modelling: <strong>the</strong> manner in which <strong>the</strong> joint is physically presented in lieu <strong>of</strong> <strong>the</strong> frame<br />
analysis.<br />
3.2 Joint characterisation<br />
A more exact yet also more expensive way to characterisation <strong>of</strong> deformability and resistance<br />
<strong>of</strong> joints is laboratory research. This way is basically limited to research activities<br />
and is not recommended for every day practice. On <strong>the</strong> o<strong>the</strong>r hand, numerous ma<strong>the</strong>matical<br />
models have been developed that can be placed into four major sets as follows:<br />
• curve tting,<br />
• simplied analytical models,<br />
• mechanical models,<br />
• analysis by means <strong>of</strong> nite elements.<br />
The procedure for <strong>the</strong> characterisation <strong>of</strong> mechanical properties <strong>of</strong> joints adopted by Eurocode<br />
3, Part 1.8, is based on <strong>the</strong> component method. The identication <strong>of</strong> various components<br />
that constitute <strong>the</strong> joint (bolts, welds, stiffeners) provides a good picture <strong>of</strong> <strong>the</strong><br />
complexity <strong>of</strong> <strong>the</strong> joint analysis. The analysis <strong>of</strong> a joint requires exact consideration <strong>of</strong><br />
many occurrences: non-linearity <strong>of</strong> materials (plasticity, strain-hardening), non-linear contact<br />
and slip, geometrical non-linearity (local instability), residual stresses and complex<br />
geometrical congurations. Although numerical procedures that apply non-linear nite elements<br />
can take into consideration all <strong>the</strong>se complexities, <strong>the</strong>y require long-lasting procedures<br />
and are very sensitive to <strong>the</strong> possibility <strong>of</strong> modelling and analysis.<br />
Therefore, for practical reasons, <strong>the</strong> planned approach must be based on simple models that<br />
leave out many variables. The component method precisely ts <strong>the</strong> simplied mechanical<br />
model consisting <strong>of</strong> springs and rigid links. In this <strong>the</strong> joint is simulated by means <strong>of</strong> <strong>the</strong> appropriate<br />
selection <strong>of</strong> rigid and exible components. These components represent specic<br />
parts <strong>of</strong> <strong>the</strong> joint that, depending on <strong>the</strong> type <strong>of</strong> loading, make an identied contribution to<br />
one or more <strong>of</strong> its structural properties.<br />
Basic components that are included in Eurocode 3 are as follows:<br />
1. Column web panel in shear<br />
2. Column web in transverse compression<br />
3. Column web in transverse tension<br />
4. Column ange in bending<br />
5. End-plate in bending<br />
6. Flange cleat in bending<br />
7. Beam or column ange and web in compression
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 113<br />
8. Beam web in tension<br />
9. Plate in tension or compression<br />
10. Bolts in tension<br />
11. Bolts in shear<br />
12. Bolts in bearing (on beam ange, column ange, end-plate or cleat)<br />
13. Concrete in compression including grout<br />
14. Base plate in bending under compression<br />
15. Base plate in bending under tension<br />
16. Anchor bolts in tension<br />
17. Anchor bolts in shear<br />
18. Anchor bolts in bearing<br />
19. Welds<br />
20. Haunched beam.<br />
When using <strong>the</strong> component method, <strong>the</strong> characteristics <strong>of</strong> resistance, stiffness and ductility<br />
<strong>of</strong> basic components are evaluated rst. These characteristics are <strong>the</strong>n combined with <strong>the</strong><br />
goal <strong>of</strong> achieving <strong>the</strong> mechanical characteristics <strong>of</strong> <strong>the</strong> entire joint. As a result, <strong>the</strong> application<br />
<strong>of</strong> <strong>the</strong> component method is according to that fairly wide, since it can be applied to<br />
all steel joints for which <strong>the</strong> characteristics <strong>of</strong> constituent components can be identied. In<br />
general, each <strong>of</strong> <strong>the</strong>se components is characterized by a non-linear relationship between<br />
<strong>the</strong> force and deformation (F-), although more simple idealisations are also possible.<br />
Several models made <strong>of</strong> spring and rigid links are proposed, which consist <strong>of</strong> <strong>the</strong> same<br />
basic components.<br />
The application <strong>of</strong> <strong>the</strong> component method on steel joints requires <strong>the</strong> following steps:<br />
1. Selection <strong>of</strong> <strong>the</strong> relevant (active) components from <strong>the</strong> general list <strong>of</strong> components (20<br />
different components have thus been fur<strong>the</strong>r dened in Part 1.8, EC 3).<br />
2. Evaluation <strong>of</strong> <strong>the</strong> force-deformation response <strong>of</strong> each component.<br />
3. “Assembly” <strong>of</strong> <strong>the</strong> active components for <strong>the</strong> evaluation <strong>of</strong> <strong>the</strong> moment-rotation (Mj -)<br />
joint response, by means <strong>of</strong> a representative mechanical model.<br />
Its application can suit various levels <strong>of</strong> element “processing”, e.g. a simplied characterisation<br />
<strong>of</strong> components is possible whenever only <strong>the</strong> resistance or <strong>the</strong> initial stiffness <strong>of</strong> a<br />
joint is sought.<br />
Familiarization with <strong>the</strong> mechanical behaviour <strong>of</strong> various joint components enables <strong>the</strong><br />
analysis <strong>of</strong> a large number <strong>of</strong> various joint congurations with a relatively low number <strong>of</strong><br />
components. The key element in <strong>the</strong> component method <strong>the</strong>refore relates to <strong>the</strong> characterisation<br />
<strong>of</strong> <strong>the</strong> F- relationship for each particular spring (component). For <strong>the</strong> evaluation<br />
<strong>of</strong> <strong>the</strong> initial stiffness <strong>of</strong> a joint only <strong>the</strong> linear stiffness <strong>of</strong> each component is necessary,<br />
while for <strong>the</strong> evaluation <strong>of</strong> ductility it is necessary to know <strong>the</strong> non-linear F- relationship<br />
<strong>of</strong> each component. Classication <strong>of</strong> joints<br />
3.2.1 Classification according to stiffness<br />
Classication according to stiffness for rigid, semi-rigid and pinned joints is performed by<br />
simply comparing <strong>the</strong> design stiffness <strong>of</strong> <strong>the</strong> joint with two stiffness boundaries, Figure 4.<br />
With <strong>the</strong> goal <strong>of</strong> simplication, <strong>the</strong> stiffness boundaries are set in such a way that <strong>the</strong>y allow<br />
a direct comparison with <strong>the</strong> calculated initial stiffness Sj,ini <strong>of</strong> <strong>the</strong> joint, for any type<br />
<strong>of</strong> joint idealisation that is being used in <strong>the</strong> analysis and is presumed beforehand.<br />
The boundaries <strong>of</strong> classication according to stiffness are as follows:
114<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
• rigid joint<br />
(unbraced frames),<br />
(braced frames),<br />
• semi-rigid joint<br />
(unbraced frames),<br />
(braced frames),<br />
• pinned joint<br />
EI/L denotes <strong>the</strong> exural stiffness <strong>of</strong> <strong>the</strong> connected beam.<br />
Fig. 4. Boundaries <strong>of</strong> classication according to joint stiffness<br />
3.2.2 Classification according to resistance<br />
Classication according to resistance consists <strong>of</strong> comparing <strong>the</strong> design joint resistance moment,<br />
M j,Rd<br />
, with <strong>the</strong> boundaries <strong>of</strong> “full-strength” and “pinned”, Figure 5.<br />
Boundaries <strong>of</strong> classication according to resistance are as follows:<br />
• full strength joint<br />
• partial-strength joint<br />
• pinned joint<br />
M full strength<br />
denotes <strong>the</strong> design resistance <strong>of</strong> <strong>the</strong> weaker structural element in <strong>the</strong> connection.<br />
It shall be noted that <strong>the</strong> classication based on experimental Mj- joint characteristics was<br />
not taken into consideration because only <strong>the</strong> design properties were considered.
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 115<br />
3.2.3 Classification according to ductility<br />
Fig. 5. Boundaries <strong>of</strong> classication <strong>of</strong> joint stiffness<br />
Adequate experience and appropriate detailing result in <strong>the</strong> so-called pinned joints that<br />
exhibit a sufcient rotation capacity. That means <strong>the</strong>y can sustain <strong>the</strong> rotations imposed<br />
on <strong>the</strong>m.<br />
In <strong>the</strong> case <strong>of</strong> moment resistant joints a notion <strong>of</strong> ductility class that includes rotation capacity<br />
is introduced. This problem has not been dealt with in more detail within Eurocode<br />
3, although extensive research is being done regarding this problem. In order to evaluate<br />
<strong>the</strong> ductility <strong>of</strong> a joint one can apply an analogy with <strong>the</strong> classication <strong>of</strong> cross-sections<br />
according to rotation capacity. Thus, joints can be classied into three classes concerning<br />
ductility, see Figure 6.<br />
Fig. 6. Ductility classes <strong>of</strong> joints – classication according to rotation capacity
116<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
Class 1 - Able to reach MRd and with a sufciently good rotation capacity to allow plastic<br />
design <strong>of</strong> <strong>the</strong> frame.<br />
Class 2 - Able to reach MRd but with a reduced plastic rotation capacity. A plastic verication<br />
<strong>of</strong> <strong>the</strong> section is anyway allowed.<br />
Class 3 - Where brittle failure (or instability) limits <strong>the</strong> moment resistance and does not<br />
allow full redistribution <strong>of</strong> <strong>the</strong> internal forces in <strong>the</strong> joint.<br />
3.3 Joint modelling<br />
The following types <strong>of</strong> joint modelling have traditionally been considered:<br />
• for rotational stiffness: rigid or pinned;<br />
• for resistance: full-strength or pinned.<br />
However, as far as joint rotational stiffness and economical joint dimensioning are concerned,<br />
<strong>the</strong>y can be semi-rigid. This provides a new possibility <strong>of</strong> joint modelling: semirigid/full-strength<br />
and semi-rigid/partial-strength. Eurocode 3 takes into consideration<br />
<strong>the</strong>se possibilities with three joint modelling types, Table 1:<br />
• continuous:<br />
covers only <strong>the</strong> rigid/full-strength case,<br />
• semi-continuous: covers <strong>the</strong> rigid/partial-strength, semi-rigid/full strength and<br />
semi-rigid/partial-strength cases.<br />
• simple:<br />
covers only <strong>the</strong> pinned case.<br />
Table 1. Joint modelling types<br />
Stiffness<br />
Resistance<br />
Full-strength Partial-strength Pinned<br />
Rigid Continuous Semi-continuous -<br />
Semi-rigid Semi-continuous Semi-continuous -<br />
Pinned - - Simple<br />
The notions continuous, semi-continuous and simple have <strong>the</strong> following meaning:<br />
• continuous:<br />
<strong>the</strong> joint ensures a full rotational continuity between <strong>the</strong> ele<br />
ments it connects;<br />
• semi-continuous: <strong>the</strong> joint ensures only a partial rotational continuity between<br />
<strong>the</strong> elements it connects (connecting elements);<br />
• simple:<br />
<strong>the</strong> joint prevents any rotational continuity between <strong>the</strong><br />
elements it connects.<br />
These meanings are linked to <strong>the</strong> type <strong>of</strong> <strong>the</strong> global analysis <strong>of</strong> <strong>the</strong> frame. In <strong>the</strong> case <strong>of</strong> an<br />
elastic global analysis <strong>of</strong> <strong>the</strong> frame, only <strong>the</strong> stiffness properties <strong>of</strong> <strong>the</strong> joint are important<br />
for joint modelling<br />
In <strong>the</strong> case <strong>of</strong> an ideal plastic analysis, <strong>the</strong> main characteristic <strong>of</strong> <strong>the</strong> joint is resistance. In<br />
all o<strong>the</strong>r cases, <strong>the</strong> properties <strong>of</strong> stiffness as well as resistance are a guide for <strong>the</strong> manner in<br />
which joints are to be modelled.<br />
The above listed possibilities as well as <strong>the</strong>ir physical interpretation for analysis are provided<br />
in Table 2 and in Figure 7, respectively
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 117<br />
Table 2: Joint modelling and frame analysis<br />
Type <strong>of</strong> frame analysis<br />
Joint modelling<br />
Elastic-perfectly<br />
Elastic<br />
Rigid plastic plastic and elastoplastic<br />
Continuous Rigid Full-strength Rigid/full-strength<br />
Semi-continuous Semi-rigid Partial-strength<br />
Rigid/partialstrength<br />
Semi-rigid/fullstrength<br />
Semi-rigid/partial<br />
strength<br />
Simple Pinned Pinned Pinned<br />
Fig. 7. Simplied joint modelling for frame analysis<br />
3.4 Joint idealisation<br />
The Mj- curve that describes non-linear behaviour <strong>of</strong> joints is not convenient for everyday<br />
practice. However, it can be idealised without a more signicant loss <strong>of</strong> exactness.<br />
One <strong>of</strong> <strong>the</strong> simplest idealisations is <strong>the</strong> one done by means <strong>of</strong> <strong>the</strong> elastic-perfectly plastic<br />
relationship, Figure 8. The advantage <strong>of</strong> such modelling is in <strong>the</strong> fact that it is similar to <strong>the</strong><br />
one that is applied for modelling <strong>of</strong> member cross-sections that are subjected to bending.<br />
The M j,Rd<br />
moment that corresponds to <strong>the</strong> yield plateau is within EC 3 called design moment<br />
resistance. It can be considered as a pseudo-plastic moment resistance <strong>of</strong> <strong>the</strong> joint.<br />
The effects <strong>of</strong> strain-hardening and <strong>the</strong> possible membrane effects are neglected, which<br />
explains <strong>the</strong> difference in Figure 8 between <strong>the</strong> actual M- curve and <strong>the</strong> yield plateau in<br />
<strong>the</strong> idealisation.<br />
The value <strong>of</strong> constant stiffness <strong>of</strong> joint S j.ini/<br />
has been explained in a number <strong>of</strong> papers,<br />
and practical values are provided in EC 3, Part 1.8. The coefcient results from a highly
118<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
expressed non-linearity <strong>of</strong> <strong>the</strong> Mj- joint curves in comparison to <strong>the</strong> M- curves for structural<br />
elements.<br />
Fig. 8. Bi-linearisation <strong>of</strong> <strong>the</strong> M- curves<br />
Actually <strong>the</strong>re are different possibilities <strong>of</strong> Mj- joint curve idealisations. The choice <strong>of</strong><br />
one <strong>of</strong> <strong>the</strong>m depends on <strong>the</strong> type <strong>of</strong> global analysis <strong>of</strong> <strong>the</strong> frame used, and EC 3 lists <strong>the</strong>m<br />
according to <strong>the</strong> following:<br />
• elastic idealisation for <strong>the</strong> elastic analysis,<br />
• perfectly plastic idealisation for <strong>the</strong> perfectly plastic analysis,<br />
• non-linear idealisation for <strong>the</strong> elastic-plastic analysis.<br />
4 CHOICE OF A JOINT MODEL<br />
Joints presented as rigid or pinned in an analysis should be dimensioned in such a way that<br />
<strong>the</strong>y satisfy <strong>the</strong> conditions for a classication <strong>of</strong> a rigid or pinned joint. The model for a<br />
semi-rigid joint can be more or less complex. It can be modelled as a spring for which M-<br />
relationship is in <strong>the</strong> range between linear elastic and non-linear, see Figure 9. The elastic<br />
global analysis requires that <strong>the</strong> behaviour <strong>of</strong> a joint be modelled as linear elastic. The elastic<br />
– perfectly plastic analysis requires a bi-linear joint model. So, <strong>the</strong> type <strong>of</strong> global analysis<br />
directly inuences <strong>the</strong> complexity <strong>of</strong> <strong>the</strong> adopted joint model. It should be mentioned<br />
that formation <strong>of</strong> plastic hinges in <strong>the</strong> joints is allowed in <strong>the</strong> application <strong>of</strong> plastic analysis.<br />
Fig. 9. Presentation <strong>of</strong> different Mj- relationships<br />
5 ECONOMIC JUSTIFICATION OF THE NEW APPROACH<br />
A modern approach to <strong>the</strong> behaviour <strong>of</strong> joints is based on <strong>the</strong> realistic behaviour <strong>of</strong> joints<br />
as semi-rigid. It is founded on economic justication <strong>of</strong> a structure in which <strong>the</strong> joints have<br />
been considered as semi-rigid. To obtain <strong>the</strong> minimum price <strong>of</strong> a steel structure two design
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 119<br />
strategies are possible regarding <strong>the</strong> execution <strong>of</strong> <strong>the</strong> joint:<br />
• Simplication <strong>of</strong> joint details which results in reducing <strong>the</strong> costs <strong>of</strong> workshop execution.<br />
It is important for unbraced frames where joints transfer signicant moments and those are<br />
rigid joints.<br />
• By reducing cross-sections <strong>of</strong> elements which results in decreasing <strong>the</strong> expenses for material.<br />
It is signicant for braced frames with nominally pinned joints.<br />
Generally speaking both strategies will lead to <strong>the</strong> application <strong>of</strong> semi-rigid joints. In <strong>the</strong><br />
case <strong>of</strong> rigid joints an economic solution can be achieved if <strong>the</strong> stiffness <strong>of</strong> joint is close to<br />
<strong>the</strong> classication limit <strong>of</strong> semi-rigid joint.<br />
Different economic studies show that <strong>the</strong>re is a possibility <strong>of</strong> saving with <strong>the</strong> application<br />
<strong>of</strong> semi-rigid joints and <strong>the</strong> adequate design <strong>of</strong> 20-25% in <strong>the</strong> case <strong>of</strong> unbraced frames<br />
and 5-9% in <strong>the</strong> case <strong>of</strong> braced frames. With <strong>the</strong> assumption that <strong>the</strong> price <strong>of</strong> steel frames<br />
amounts to about 10% <strong>of</strong> <strong>the</strong> total price <strong>of</strong> a commercial building and about 20% for an<br />
industrial building a decrease in total costs can be estimated to 4-5% for unbraced frames.<br />
For braced frame systems saving <strong>of</strong> 1-2% is possible. The costs <strong>of</strong> material and workshop<br />
execution depend on <strong>the</strong> relative stiffness <strong>of</strong> <strong>the</strong> joint, Figure 10.<br />
Fig. 10. Qualitative relation <strong>of</strong> a steel structure depending on <strong>the</strong> relative stiffness <strong>of</strong> <strong>the</strong> joint<br />
With an increase <strong>of</strong> <strong>the</strong> relative stiffness <strong>of</strong> joints <strong>the</strong> costs <strong>of</strong> material is decreased, curve<br />
A, whereas <strong>the</strong> costs <strong>of</strong> labour grow, curve B. For <strong>the</strong> total costs that are a sum <strong>of</strong> <strong>the</strong> costs<br />
<strong>of</strong> curves A and B, <strong>the</strong> minimum can be found and <strong>the</strong>refore <strong>the</strong> optimum relative stiffness<br />
as well. In many cases <strong>the</strong> value that leads to <strong>the</strong> optimum design <strong>of</strong> a structural system by<br />
taking into consideration <strong>the</strong> minimum <strong>of</strong> <strong>the</strong> total costs does not lie in <strong>the</strong> area <strong>of</strong> nominally<br />
pinned nor in <strong>the</strong> area <strong>of</strong> semi-rigid joints.<br />
Looking at <strong>the</strong> trends in <strong>the</strong> past few decades it is obvious that <strong>the</strong> costs <strong>of</strong> labour grow<br />
compared to <strong>the</strong> costs <strong>of</strong> <strong>the</strong> used material (see curve B*). As a result, see Figure 10, it is<br />
clear that a fast development <strong>of</strong> <strong>the</strong> optimum stiffness <strong>of</strong> joints shifts more towards semirigid<br />
joints. Therefore, in an attempt to nd an economic solution for steel structures <strong>the</strong><br />
use <strong>of</strong> semi-rigid joints will become increasingly more interesting.<br />
6 CONCLUSION<br />
The design <strong>of</strong> structural joints that is proposed in Eurocode 3, is based on <strong>the</strong> research<br />
conducted over <strong>the</strong> past few decades. Eurocode 3 gives practical guidelines for <strong>the</strong> characterisation<br />
<strong>of</strong> most <strong>of</strong> steel joints (beam-to-column joints, beam and column splices, column<br />
base plates etc.) <strong>of</strong> frames made <strong>of</strong> I or H cross sections, loaded by static load. Fur<strong>the</strong>r<br />
research is focused on extending <strong>the</strong> guidelines for characterisation <strong>of</strong> joints as composite
120<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
joints (concrete and steel) and joints made <strong>of</strong> o<strong>the</strong>r types <strong>of</strong> cross sections and o<strong>the</strong>r types<br />
<strong>of</strong> load.<br />
The new approach, semi-rigid joints – semi-continuous structure, gives more possibilities<br />
than <strong>the</strong> traditional approach in type <strong>of</strong> framing. This is based on <strong>the</strong> fact that <strong>the</strong> joint<br />
characteristics in <strong>the</strong> design have been considered as variables that are chosen to comply<br />
with <strong>the</strong> particular requirements <strong>of</strong> a specic project.<br />
For braced frames, <strong>the</strong> new approach allows <strong>the</strong> use <strong>of</strong> smaller cross sections <strong>of</strong> beams<br />
and a reduction <strong>of</strong> <strong>the</strong> costs <strong>of</strong> <strong>the</strong> structural frame. For unbraced frames <strong>the</strong> new approach<br />
reduces <strong>the</strong> complexity <strong>of</strong> <strong>the</strong> joint geometry and thus results in major saving in <strong>the</strong> execution<br />
costs.<br />
REFERENCES<br />
1. Eurocode 3, Design <strong>of</strong> Steel Structures, Part 1.1: General Rules and Rules for Buildings,<br />
European Prestandard ENV 1993-1-1, Brussels: CEN, 1992.<br />
2. Revised Annex J <strong>of</strong> Eurocode 3. Joints in Building Frames, European Prestandard<br />
ENV 1993-1-1: 1992/A2, Brussels: CEN, 1998.<br />
3. Eurocode 3, Design <strong>of</strong> Steel Structures, Part 1.8: Design <strong>of</strong> Joints, European Prestandard<br />
prEN 1993-1-8, Final Draft, Brussels: CEN, 2001.<br />
4. Maquoi, R.; Chabrolin, B.: Frame Design Including Joint Behaviour, ECSC Report<br />
18563, Luxembourg, Ofce for Ofcial Publications <strong>of</strong> <strong>the</strong> European Communities,<br />
1998.<br />
5. Ne<strong>the</strong>rcot, D.; Zandonini, R.: Methods <strong>of</strong> prediction <strong>of</strong> joint behaviour – Beam to<br />
column connections, In: Naranayan, R. (ed.) Structural Connections – Stability and<br />
Strength, Elsevier Applied Science, 1989., 23-62<br />
6. European Convention for Constructional Steelwork, Analysis Design <strong>of</strong> Steel Frames<br />
with Semi-rigid Joints, Publication 67, Brussels: ECCS, 1992.<br />
7. Zoetemeijer, P.: A design method for <strong>the</strong> tension side <strong>of</strong> statically-loaded bolted beamto-column<br />
joints, Heron 1974; 1-59<br />
8. Weynand, K.; Jaspart, J.-P.; Steenhuis, M.: The stiffnes model <strong>of</strong> revised Annex J <strong>of</strong><br />
Eurocode 3, In: Bjorhovde, R.; Colson, A.; Zandonini, R., (editors): Connections in<br />
steel structures, Trento, 1995., 441-452<br />
9. Huber, G.; Tschemmernegg, F.: Modelling <strong>of</strong> steel connections, Journal <strong>of</strong> Constructional<br />
Steel Research 45 (1998) 2, 199-216<br />
10. Weynand, K.; Jaspart, J.-P.; Steenhuis, M.: Economy Studies <strong>of</strong> Steel Building Frames<br />
with Semi-Rigid Joints, Journal <strong>of</strong> Constructional Steel Research 46:1-3 (1998)<br />
11. Giro Coelho, A. M.; Bijlaard, F. S. K.; da Silva, L. S.:Experimental assessment <strong>of</strong> <strong>the</strong><br />
ductility <strong>of</strong> extended end plateconnections. Engineering Structures 26 (2004), 1185-<br />
1206.<br />
12. da Silva, L. S.; Lima, L.; Vellasco, S.; Andrade, S.: Experimental behaviour <strong>of</strong> endplate<br />
beam-to-column joints under bending and axial force, Database reporting and<br />
discussion <strong>of</strong> results. Report on ECCS-TC10, Meeting in Ljubljana, April 2002.<br />
13. Faella, C.; Piluso, V.; Rizzano, G.: Structural Steel Semirigid Connections, Theory,<br />
Design and S<strong>of</strong>tware, CRC Press LLC, 2000 N.W. Corporate Blvd., Boca Raton, Florida<br />
33431, 2000.<br />
14. Cruz, P. J.S; da Silva, L. S.; Rodrigues, D. S; Simes, R. A. D.: Database for <strong>the</strong><br />
semi-rigid behaviour <strong>of</strong> beam-to-column connections in seismic regions, Journal <strong>of</strong><br />
Constructional Steel Research 46 (1998) 1-3, 233-234<br />
15. Robot Millennium Program - Version 17.5: User's Manual, RoboBAT, June 2004.
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 121<br />
Development strategy <strong>of</strong><br />
ECOINDUSTRAL PARK RAŠA – EIPR<br />
Pr<strong>of</strong>.dr.sc. Juraj Božičević<br />
Elitech Ltd.<br />
Trg kralja Tomislava 18 Zagreb 10000, Croatia<br />
juraj.bozicevic1@zg.t-com.hr, Tel. 00385 1 4922264<br />
Dr.sc. Marijan Andrašec<br />
Ecooleum Ltd.<br />
Trg kralja Tomislava 18, Zagreb 10000, Croatia<br />
marijan.andrasec@zg.t-com.hr, Tel. 00385 98 404755<br />
Abstract<br />
The pith <strong>of</strong> <strong>the</strong> strategy <strong>of</strong> development <strong>of</strong> <strong>the</strong> Ecoindustrial Park Raša is presented. The<br />
Park is situated in <strong>the</strong> bay <strong>of</strong> Raša, unique but improperly used and maintained Croatian<br />
fjord. The starting point in conceiving <strong>of</strong> <strong>the</strong> Ecoindustrial Park Raša is synergy <strong>of</strong> service<br />
and production activities, successful cooperation <strong>of</strong> companies aiming to achieve results on<br />
<strong>the</strong> following principle: unity is more successful than <strong>the</strong> amount <strong>of</strong> successfully combined<br />
activity <strong>of</strong> singular companies, units <strong>of</strong> <strong>the</strong> system. The contribution will be <strong>the</strong> care about<br />
industrial ecology, streng<strong>the</strong>ning <strong>of</strong> <strong>the</strong> integration <strong>of</strong> economic raise and protection <strong>of</strong><br />
environment as well as industrial symbiosis that will improve <strong>the</strong> economy and <strong>the</strong> quality<br />
<strong>of</strong> life in <strong>the</strong> Community <strong>of</strong> Raša.<br />
Key words: ecoindustrial park, sustainable development, Port Raša, Community <strong>of</strong> Raša<br />
Contents<br />
1.0 Introduction<br />
2.0 The Community <strong>of</strong> Raša<br />
2.1 Geography<br />
2.2 Present situation<br />
3.0 Ecoindustrial Park Raša – EIPR - conceptualization<br />
3.1 What is an Ecoindustrial Park – EIP?<br />
3.2 Basics for conceptualization <strong>of</strong> EIPR<br />
3.3 Projects in EIPR proposed for realization<br />
3.4 Ecoindustrial Park and <strong>the</strong> Community <strong>of</strong> Raša<br />
3.5 Stakeholders<br />
3.6 Knowledge and experience on which venture relies<br />
4.0 Ecoindustrial Park Raša – EIPR - realization<br />
4.1 The way <strong>of</strong> achieving realization <strong>of</strong> EIPR<br />
4.2 Economy <strong>of</strong> <strong>the</strong> project EIPR<br />
4.3 Risks<br />
4.4 Development effects achieved with EIPR<br />
4.5 Priority tasks
122<br />
5.0 Conclusion<br />
6.0 References<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
1 Introduction<br />
When, during <strong>the</strong> early 1990s, more precisely, in 1993, Croatian System Society has organized<br />
<strong>the</strong> rst conference called The Port as a Complex System, our guiding idea was<br />
<strong>the</strong> promotion <strong>of</strong> knowledge about <strong>the</strong> importance <strong>of</strong> <strong>the</strong> modernization <strong>of</strong> <strong>the</strong> Croatian<br />
ports and <strong>the</strong> strategic consideration about <strong>the</strong> role <strong>of</strong> <strong>the</strong> ports in <strong>the</strong> development <strong>of</strong> <strong>the</strong><br />
economy. At that time <strong>the</strong> rst author was, as <strong>the</strong> Associate Minister for higher education<br />
in <strong>the</strong> Ministry <strong>of</strong> Education and Sports, participated at <strong>the</strong> rst Croatian – Ne<strong>the</strong>rlands negotiations<br />
about <strong>the</strong> cooperation in education area... so he took <strong>the</strong> opportunity to visit <strong>the</strong><br />
Rotterdam port and learn on higher maritime education, as well as on <strong>the</strong> training equipment<br />
at <strong>the</strong> High Maritime College, with <strong>the</strong> intention to transfer gained experiences to<br />
<strong>the</strong> Maritime Faculty in Rijeka. He was accompanied by a host, a Dutchman, who was <strong>the</strong><br />
member <strong>of</strong> <strong>the</strong> management board <strong>of</strong> <strong>the</strong> Rotterdam port and who presented him numerous<br />
data which illustrated <strong>the</strong> importance <strong>of</strong> <strong>the</strong> port and its industrial area for <strong>the</strong> economy <strong>of</strong><br />
<strong>the</strong> Ne<strong>the</strong>rlands. The host was especially interested in Croatia’s plans – in which way will<br />
Croatia, after has gained independence, use its maritime position, and how Croatian strategic<br />
plan <strong>of</strong> development <strong>of</strong> economy and <strong>the</strong> ports will be conceived. We shall quote <strong>the</strong><br />
private communication <strong>of</strong> <strong>the</strong> rst author (3): The Dutchman emphasized <strong>the</strong> extraordinary<br />
geographical position and maritime features <strong>of</strong> <strong>the</strong> Port <strong>of</strong> Rijeka and neighboring ports<br />
within its jurisdiction. He had excellent knowledge <strong>of</strong> Croatian possibilities, and, since I<br />
was preparing for <strong>the</strong> conference „Port as a complex system“ in Rijeka, I presented him<br />
ambitiously <strong>the</strong> vision <strong>of</strong> <strong>the</strong> Conference Program board and our idea about ports development<br />
in Croatia... He asked me if we really have full understanding <strong>of</strong> <strong>the</strong> geo-strategic importance<br />
<strong>of</strong> <strong>the</strong> Port <strong>of</strong> Rijeka and its possible competitiveness in relation to <strong>the</strong> Rotterdam<br />
port. When we shaked hands during <strong>the</strong> parting, he said to me, half-jokingly, something<br />
which stunned me: «I will kill myself if <strong>the</strong> Port <strong>of</strong> Rijeka takes even 1% <strong>of</strong> <strong>the</strong> middle-<br />
European traf c from <strong>the</strong> Port <strong>of</strong> Rotterdam.»<br />
That was <strong>the</strong> war time in Croatia. The Port <strong>of</strong> Rijeka was destination <strong>of</strong> just a few ships.<br />
When I talked about this event to my closest cooperate in <strong>the</strong> project „Port as <strong>the</strong> Complex<br />
System“, dr. Juraj Maari, he suggested me not to talk about <strong>the</strong> Dutchman’s statement,<br />
because no one will understand... However, he proposed that we talk about this in inner<br />
circle as a very serious topic, and that <strong>the</strong> program <strong>of</strong> our future conferences should be<br />
gradually developed in such way that, in <strong>the</strong> sphere <strong>of</strong> <strong>the</strong> economic decision making, we<br />
could develop <strong>the</strong> awareness about <strong>the</strong> need for <strong>the</strong> strategy <strong>of</strong> <strong>the</strong> economic development<br />
which is leaning upon <strong>the</strong> maritime orientation.<br />
The fourth conference ”Port as a Complex System”“ held in 1999, was dedicated to strategic<br />
topics:<br />
„Rijeka – <strong>the</strong> Main Croatian Port and joined Rijeka and Zagreb – Croatian and European<br />
economic Focus“. (4, 5) However, since 2000, <strong>the</strong>re was no more interest for <strong>the</strong> Conferences....<br />
I was introduced to <strong>the</strong> operations <strong>of</strong> <strong>the</strong> Port <strong>of</strong> Rijeka ve years later, when I started to<br />
cooperate with <strong>the</strong> Community <strong>of</strong> Raša. The talks with <strong>the</strong> mayor <strong>of</strong> <strong>the</strong> Community encouraged<br />
me to think about <strong>the</strong> sustainable development <strong>of</strong> <strong>the</strong> port area <strong>of</strong> Raša, which is in<br />
concession <strong>of</strong> <strong>the</strong> Port <strong>of</strong> Rijeka. Really, about very strategic Task.<br />
The mayor <strong>of</strong> Raša accepted <strong>the</strong> idea <strong>of</strong> consulting company Elitech Ltd. to make <strong>the</strong> development<br />
concept <strong>of</strong> <strong>the</strong> Community <strong>of</strong> Raša as socially responsible community, <strong>the</strong> basis<br />
<strong>of</strong> <strong>the</strong> wise exploration <strong>of</strong> its natural resources, revitalisation and fur<strong>the</strong>r development <strong>of</strong>
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 123<br />
its economy, based upon <strong>the</strong> idea <strong>of</strong> eco-industrial society.<br />
The following agreement was accepted as <strong>the</strong> frame work: The target <strong>of</strong> sustainable development<br />
is to facilitate and fulll <strong>of</strong> basic necessities and quality lives to all <strong>the</strong> citizens <strong>of</strong><br />
<strong>the</strong> Community, having in mind not to risk <strong>the</strong> quality <strong>of</strong> life <strong>of</strong> future generations. If we<br />
don’t take care about sustainable development we could face an insecure future. Therefore<br />
<strong>the</strong> sustainable development is our best long-term interest.<br />
We started with work in autumn 2005. We had number <strong>of</strong> meetings with <strong>the</strong> mayor and <strong>the</strong><br />
head <strong>of</strong> <strong>the</strong> County <strong>of</strong> Istria, who suggested that <strong>the</strong> Community <strong>of</strong> Raša, in cooperation<br />
with consulting company Elitech Ltd., should establish <strong>the</strong> Working group for strategic<br />
development <strong>of</strong> port area and sign an agreement upon it.<br />
With <strong>the</strong> acceptance <strong>of</strong> Letter <strong>of</strong> Intent, signed by Community <strong>of</strong> Raša, County <strong>of</strong> Istria,<br />
Port <strong>of</strong> Rijeka, Port Authority <strong>of</strong> Rijeka and Elitech Ltd. on 15th December 2006, Working<br />
group started with meetings and discussions in early 2007.<br />
As a basis for discussions has been <strong>the</strong> study „Development <strong>of</strong> Eco-energy Valley <strong>of</strong> Raša<br />
and Port area Raša based upon <strong>the</strong> production <strong>of</strong> bi<strong>of</strong>uel”, made by Naval Research Institute<br />
in Zagreb in cooperation with <strong>the</strong> company Elitech Ltd. and <strong>the</strong> company Ecooleum<br />
Ltd., which joined <strong>the</strong> Working group. Ecooleum Ltd was established with <strong>the</strong> purpose to<br />
realize conceived projects.<br />
The idea <strong>of</strong> technological platform Ecoenergy Valley was developed and presented to <strong>the</strong><br />
Croatian Institute for Technology as a possible EU-project in order to nance <strong>the</strong> project.<br />
Aside to eco-energy production, <strong>the</strong> particular attention was focused upon trafc infrastructure,<br />
especially in revitalizing <strong>of</strong> <strong>the</strong> railway as important support to port activity.<br />
In <strong>the</strong> period from 2008 to <strong>2010</strong> <strong>the</strong> idea <strong>of</strong> Eco-industrial society has been developed in<br />
nal form, and <strong>the</strong> basic conception is presented in this paper.<br />
In 2008. <strong>the</strong> Forum for Social Responsible Activity was founded to facilitate communication<br />
with local community. The companies Holcim and Murexin - Istarska tvornica vapna<br />
joined <strong>the</strong> Forum, and <strong>the</strong> consultations were held upon <strong>the</strong> Sustainable Development <strong>of</strong><br />
Community <strong>of</strong> Raša. Initially, <strong>the</strong> important contribute was made by Croatian Systems<br />
Society and <strong>the</strong>ir study Ecoproduct areas <strong>of</strong> Croatia and <strong>the</strong>n with <strong>the</strong>ir project Systemic<br />
considerations <strong>of</strong> social responsibility in <strong>the</strong> living area (2008-9), nanced by Ministry <strong>of</strong><br />
Environment Protection, Space Planning and Construction.<br />
In <strong>the</strong> case <strong>of</strong> Ecoindustrial park <strong>of</strong> Raša it has been conceived <strong>the</strong> interaction <strong>of</strong> portrelated,<br />
industrial, energy, agricultural, maricultural and touristic activities that rely on<br />
restructured and highly functional communal and trafc infrastructure – port, road and<br />
railway. We would like to emphasize <strong>the</strong> part <strong>of</strong> activities in management <strong>of</strong> eco-industrial<br />
park that will be very important for successful achievement <strong>of</strong> sustainable development<br />
and industrial symbiosis, for <strong>the</strong> reduction <strong>of</strong> costs, common use <strong>of</strong> information, services,<br />
energy and byproducts as raw material for number <strong>of</strong> users.<br />
From <strong>the</strong> point <strong>of</strong> view <strong>of</strong> <strong>the</strong> Community <strong>of</strong> Raša, it was particularly important to motivate<br />
enterprising spirit in <strong>the</strong> development <strong>of</strong> port activity. It is presumable that <strong>the</strong> building <strong>of</strong><br />
<strong>the</strong> Park will be especially benecial and a big impulse on port activity, and <strong>the</strong> Port <strong>of</strong><br />
Rijeka, which is <strong>the</strong> concessionaire, should be encouraged to show more responsibility on<br />
that valuable national resource.
124<br />
2 The Community <strong>of</strong> Raša<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
2.1 Geography<br />
The Community <strong>of</strong> Raša is situated on south-eastern side <strong>of</strong> Istrian peninsula. Its surface is<br />
7.954 ha, it contains 23 settlements with 3535 inhabitants and <strong>the</strong> shore length <strong>of</strong> 48 km.<br />
It is part <strong>of</strong> Region <strong>of</strong> Istria.<br />
The Community <strong>of</strong> Raša have got high quality land for agricultural production that is partly<br />
owned by private persons, and partly is property <strong>of</strong> Republic <strong>of</strong> Croatia and <strong>of</strong> <strong>the</strong> Community.<br />
It is planted with fodder, vegetable in lowlands, and with grapes, fruits and medicinal<br />
cultures and olives on <strong>the</strong> sunny positions. Forests are mostly <strong>of</strong> private property, and <strong>the</strong><br />
rest <strong>of</strong> it is property <strong>of</strong> Republic <strong>of</strong> Croatia and are managed by Croatian Forests.<br />
The forests are not important for exploitation but are very important for ecological balance,<br />
prevention <strong>of</strong> erosion, improvement <strong>of</strong> microclimatic conditions and development <strong>of</strong><br />
agrotourismus and recreation.<br />
Climatic conditions in Community <strong>of</strong> Raša are characterized with bigger quantity <strong>of</strong> rains<br />
than <strong>the</strong> west side <strong>of</strong> Istria. Rainy period is at its peak in October and minimum in July.<br />
The most signicant winds are nor<strong>the</strong>astern, especially bura in winter period and summer’s<br />
night wind in most part <strong>of</strong> shore territory, sou<strong>the</strong>astern – especially jugo (sirocco) in spring<br />
and autumn, northwestern – especially during summer period.<br />
Soils are shallow and stony, and smaller areas <strong>of</strong> deep soil we can nd in typical karstic<br />
forms – short lanes and karstic heights, predominantly red and brown soil. In <strong>the</strong> area <strong>of</strong><br />
<strong>the</strong> river Raša we nd hidromorc (gley) soil, and on <strong>the</strong> edges aluvial-coluvial soils that<br />
are salinized on <strong>the</strong> estuary <strong>of</strong> river. The river <strong>of</strong> Raša is <strong>the</strong> main and <strong>the</strong> most important<br />
water-current in water-supply-system “Raša-Boljunica“. Its characteristic is <strong>the</strong> frequent<br />
torrents, so at times extremely high water is to be expected. The main water sources are<br />
Fonte Gaj and Kokoti. The town <strong>of</strong> Raša with 1800 inhabitants does not have waste water<br />
purication facility.<br />
Across <strong>the</strong> Community <strong>of</strong> Raša <strong>the</strong>re is state road D21 Pula-Labin, number <strong>of</strong> regional and<br />
local roads. Numbers <strong>of</strong> unclassied roads are under community’s jurisdiction.<br />
Railroad Lupoglav-Štalije is used exclusively for transport; it’s not electried, it’s in group<br />
<strong>of</strong> mountain railways and is qualied for average speed <strong>of</strong> 60 km/h. Railway station is in<br />
Most Raša.<br />
There are several ports on <strong>the</strong> territory <strong>of</strong> <strong>the</strong> Community <strong>of</strong> Raša. The port Bršica is <strong>the</strong><br />
only one classied for international sea shipping (state-level port). The port <strong>of</strong> Koromano<br />
is special purpose port – industrial port. The ports Sveta Marina, Trget and Tunarica are <strong>of</strong><br />
local importance (shery).<br />
The main direction <strong>of</strong> magistral gas pipeline (for international transport) <strong>of</strong> national importance<br />
Casal Borsetti – Karlovac DN700, transits along eastern shore <strong>of</strong> Istria and is passing<br />
through a part <strong>of</strong> Community’s territory.<br />
2.2 Present situation<br />
Once industrially developed and supported by mining (coal mines, coal power plant) and<br />
by production <strong>of</strong> machine-tools (Prvomajska factory), <strong>the</strong> Community <strong>of</strong> Raša became<br />
poor with <strong>the</strong> closure <strong>of</strong> coal mine and failure <strong>of</strong> Prvomajska, leaning nowadays on companies<br />
Holcim, ITV Murexin and hardly active Port.<br />
The territory <strong>of</strong> <strong>the</strong> Port <strong>of</strong> Raša, <strong>the</strong> most valuable industrial zone in Istria, has long tradition<br />
in energetic sector (coal), and <strong>the</strong> industrial culture <strong>of</strong> inhabitants was developed. In
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 125<br />
<strong>the</strong> past it was mostly used to support local industry, and <strong>the</strong>refore never had signicant<br />
value as transit port for Middle-European area, although it was connected by railroad to<br />
Trieste and fur<strong>the</strong>r to North Italy and Middle Europe.<br />
The port is used for stone and timber transport. Port installations and operative quay are in<br />
dilapidated conditions so <strong>the</strong> new one should be made, o<strong>the</strong>r trafc, energy and water infrastructure<br />
should be made too and operative and warehousing surface should be widened.<br />
The bottom <strong>of</strong> <strong>the</strong> bay should be cleaned from <strong>the</strong> mud.<br />
On <strong>the</strong> location <strong>of</strong> Bršica a stone pit existed, that was closed in 1978, so this area should<br />
be revitalized by removing <strong>the</strong> part <strong>of</strong> <strong>the</strong> hill between microlocations Bršica and Štalije.<br />
Agricultural land in river Raša valley and along <strong>the</strong> channels are poorly used, because <strong>of</strong><br />
<strong>the</strong> irresponsible maintenance <strong>of</strong> <strong>the</strong> channels, non resolved water overow protection and<br />
salinization.<br />
The territory <strong>of</strong> Raša is rich with high quality lime stone, so <strong>the</strong> stone exploitation, as well<br />
as in o<strong>the</strong>r parts <strong>of</strong> Istria, is a part <strong>of</strong> traditional exploitation. On that basis <strong>the</strong> production <strong>of</strong><br />
lime (ITV – Istrian Lime Factory) and cement (Holcim) were established, so <strong>the</strong> fur<strong>the</strong>r establishment<br />
<strong>of</strong> complementary production with higher grade <strong>of</strong> material processing would<br />
be natural development <strong>of</strong> this industry.<br />
The project <strong>of</strong> waste water drainage should be replaced with improved project with which<br />
should be projected high quality equipment for depuration and obtained technical water for<br />
agricultural production on uncultivated agricultural land.<br />
The urban plan <strong>of</strong> Community <strong>of</strong> Raša in <strong>the</strong> past did not provide for appropriate development<br />
<strong>of</strong> <strong>the</strong> Community <strong>of</strong> Raša, as well as <strong>of</strong> <strong>the</strong> port-area. It should be upgraded in <strong>the</strong><br />
new urban plan.<br />
3 Ecoindustrial Park Raša – EIPR - conceptualization<br />
3.1 What is an Ecoindustrial Park – EIP?<br />
An Ecoindustrial Park is unity <strong>of</strong> production and service activities that cultivate cooperation<br />
and with emphasized care about environment and materials, harmonizing energy,<br />
water and materials ows. Acting toge<strong>the</strong>r, <strong>the</strong> members <strong>of</strong> business units take care <strong>of</strong> <strong>the</strong><br />
common prot, which is bigger than amount <strong>of</strong> single prot that <strong>the</strong> single company would<br />
achieve optimizing proper activity independently. In that way economic success <strong>of</strong> <strong>the</strong><br />
companies increases and minimizes its inuence on environment. That is a new concept<br />
<strong>of</strong> sustainable development <strong>of</strong> <strong>the</strong> local community which protects <strong>the</strong> economy <strong>of</strong> local<br />
society from undesired inuences <strong>of</strong> globalization.<br />
EIP has rich and different possibilities <strong>of</strong> development including diversity <strong>of</strong> environment<br />
cultivation, infrastructure, single outt and resources, common services etc.<br />
Since <strong>the</strong> beginning <strong>of</strong> 90s, a number <strong>of</strong> ecoindustrial parks were established all over <strong>the</strong><br />
world, and <strong>the</strong>ir activities proved that public and private sector is interested in such concept<br />
<strong>of</strong> sustainable development. (8, 9, 10, 11, 12)<br />
Ecoindustrial parks are developed in three basic forms:<br />
Transformation <strong>of</strong> an existing industrial park,<br />
Revitalization <strong>of</strong> previously used sites (brownelds development).<br />
New industrial site development.
126<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
Often <strong>the</strong>se appear in combination. When existing industrial parks undergo expansion, this<br />
can be developed as an eco-industrial park.<br />
The form <strong>of</strong> ElP development brings certain preconditions that affect <strong>the</strong> whole planning<br />
process: Existing industrial parks have <strong>the</strong> companies for potential synergies, but <strong>of</strong>ten <strong>the</strong><br />
interaction between <strong>the</strong> companies is not developed, making it difcult to get to know companies<br />
and <strong>the</strong>ir potential in detail. Park management and local authorities have to invest<br />
time in community interaction among <strong>the</strong> companies.<br />
As in developing traditional industrial parks, similar strategies can be used to establish ecoindustrial<br />
parks. The most commonly practiced or researched strategies are:<br />
Anchor tenant; This involves having a large industrial user, <strong>of</strong>ten a power plant, sugar<br />
renery or o<strong>the</strong>r types <strong>of</strong> operation with large-scale material ows, which <strong>the</strong> industrial<br />
park will be developed around.<br />
Materials or By-product exchange; This can occur ei<strong>the</strong>r in an eco-industrial park or in a<br />
regional network <strong>of</strong> businesses, including industrial parks. This involves using one industry’s<br />
waste or “by products” as ano<strong>the</strong>r industry’s raw materials.<br />
Resource Recovery System; is an expanded concept <strong>of</strong> <strong>the</strong> byproduct exchange and includes<br />
all forms <strong>of</strong> material and product recovery in an industrial complex (waste management,<br />
by-product exchange, recycling and remanufacturing).<br />
Energy connections; involves maximizing energy efciency through design or rehabilitation,<br />
co-generation, and energy cascading. It can also include <strong>the</strong> utilization <strong>of</strong> renewable<br />
energy.<br />
Thematic Park; a number <strong>of</strong> ElPs focus on co-locating a specic type <strong>of</strong> industry. Examples<br />
<strong>of</strong> such <strong>the</strong>matic parks are:<br />
Agriculture-based EIP or Agro-ElP; provides support for sustainable farming and<br />
food processing and includes several basic types <strong>of</strong> rms and agencies which may be recruited<br />
as tenants. Farms and food processing industries have a high ow <strong>of</strong> organic material<br />
that can easily be fur<strong>the</strong>r processed and maximized in its resource value.<br />
Chemical EIP; <strong>the</strong> chemical industry uses co-location <strong>of</strong> production units and<br />
upstream-downstream synergies as <strong>the</strong>ir core business.<br />
Sector-specic EIP (Textile, Electronics etc.); a number <strong>of</strong> regions have historical<br />
industrial areas with an agglomeration <strong>of</strong> sector industries, <strong>of</strong>ten as industrial clusters, but<br />
also as industrial parks. Companies located <strong>the</strong>re usually have similar problems and less<br />
potential for synergies, but strong links between companies (associations) can stimulate<br />
joint problem solving and innovation concepts.<br />
Resource Recovery EIP; companies involved in recycling, reuse, remanufacturing,<br />
niche collection <strong>of</strong> materials, and manufacturing from recovered materials form a cluster<br />
with good synergies among <strong>the</strong>m.<br />
Environmental Technology Parks/EIP; <strong>the</strong> concept focuses on <strong>the</strong> promotion <strong>of</strong><br />
environmental technology industries. Aside from possible innovation synergies, <strong>the</strong>ir colocation<br />
does not necessarily provide any benet in industrial symbiosis. Many environmental<br />
technology companies produce like ordinary manufacturing businesses.<br />
3.2 Basics for conceptualization <strong>of</strong> EIPR<br />
Considering <strong>the</strong> past experiences and <strong>the</strong> present situation it is natural wish <strong>of</strong> Community’s<br />
management to initiate development on new basis and use extraordinary natural<br />
resources, especially port-area, with <strong>the</strong> special attention on sustainable development approach.<br />
The proposed study should bring <strong>the</strong> strategy <strong>of</strong> development <strong>of</strong> Ecoindustrial Park <strong>of</strong>
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 127<br />
Raša as <strong>the</strong> basis <strong>of</strong> sustainable development <strong>of</strong> <strong>the</strong> Community <strong>of</strong> Raša.<br />
The conditions relevant for conceptualization <strong>of</strong> EIPR are:<br />
<strong>the</strong>re are existing old industrial zone and economically weakly developed port,<br />
not connected all subjects in space and do not cooperate, so <strong>the</strong>y should be brought to create<br />
new relations through new ideas,<br />
local community is interested in forming EIPR, because it should solve its infrastructural<br />
and ecological difculties and economical development,<br />
present entities could be netted with new projects,<br />
location is important also on Croatian level (state port!), and at North Adriatic area is <strong>the</strong><br />
only free and high quality industrial space.<br />
Extraordinary convenient trafc settlement: port, road and railway should be natural encouragement<br />
for <strong>the</strong> logistic support <strong>of</strong> activities in EIPR. With planning and restoring<br />
<strong>of</strong> <strong>the</strong> railway and in <strong>the</strong> future with <strong>the</strong> tunneling <strong>of</strong> <strong>the</strong> mountain Uka, EIPR will have<br />
additional value.<br />
The activities on settling <strong>of</strong> <strong>the</strong> port area, removal <strong>of</strong> ruins <strong>of</strong> <strong>the</strong> abandoned separation <strong>of</strong><br />
<strong>the</strong> coal and <strong>the</strong> formation <strong>of</strong> initial industrial plateaus will be important initial activities,<br />
followed by necessary reconstruction <strong>of</strong> quay and removal <strong>of</strong> mud from <strong>the</strong> bay.<br />
One <strong>of</strong> <strong>the</strong> rst and important steps in realization <strong>of</strong> this unique Croatian project is settlement<br />
<strong>of</strong> <strong>the</strong> neglected port area and forming <strong>of</strong> landscape and horticulturally shaped industrial<br />
area and reconstruction <strong>of</strong> quay.<br />
In order to manage <strong>the</strong> activities in Park, a social, non-prot Company for environment<br />
economy should be established, against <strong>the</strong> model <strong>of</strong> public-private partnership, as a common<br />
model <strong>of</strong> <strong>the</strong> management in eco-industrial parks.<br />
Besides institutional co-property <strong>of</strong> local community, <strong>the</strong> idea is that local inhabitants<br />
should become shareholders, and this should favor acceptance <strong>of</strong> <strong>the</strong> project and increase<br />
<strong>the</strong> possibilities for long-term successful and sustainable development <strong>of</strong> port, transport<br />
and industrial activities, as well as provide attractive benets for internal and foreign investors.<br />
The company for management <strong>of</strong> <strong>the</strong> Park will take care about <strong>the</strong> development and building<br />
<strong>of</strong> EIPR, so acting in concordance with <strong>the</strong> targets dened with this strategy will implement<br />
<strong>the</strong> model <strong>of</strong> <strong>the</strong> Park and will select <strong>the</strong> participating companies, investors and act as<br />
coordinator, manage and coordinate closely with <strong>the</strong> Community <strong>of</strong> Raša. For <strong>the</strong> success<br />
<strong>of</strong> <strong>the</strong> project it is important to clearly dene WHAT and HOW a support <strong>of</strong> <strong>the</strong> citizens<br />
should give <strong>the</strong> impulse for general development.<br />
We have to establish multidisciplinary team in <strong>the</strong> company for <strong>the</strong> management <strong>of</strong> EIPR,<br />
in which <strong>the</strong>re would participate, toge<strong>the</strong>r with <strong>the</strong> experts in branches <strong>of</strong> business, people<br />
from <strong>the</strong> executive authorities. Such expert team should, in accordance with <strong>the</strong> rules <strong>of</strong><br />
business, start working on <strong>the</strong> plan <strong>of</strong> development beginning with <strong>the</strong> denition <strong>of</strong> development<br />
targets, available human resources through <strong>the</strong> ways <strong>of</strong> its executing, from planning<br />
<strong>the</strong> future development till <strong>the</strong> preparation <strong>of</strong> <strong>the</strong> conditions for its achievement.<br />
Large amount <strong>of</strong> funds should be invested in port space design, but it should be emphasized<br />
that this project should, with wisely designed project <strong>of</strong> environment design and use<br />
<strong>of</strong> obtained material, bring prot to use in fur<strong>the</strong>r investments in space and infrastructure<br />
development <strong>of</strong> <strong>the</strong> Park. So <strong>the</strong> conditions to attract investors should be created gradually,<br />
and <strong>the</strong> interest for <strong>the</strong> project has been already shown from <strong>the</strong> investors from Austria,<br />
Italy, Sweden, Denmark and Malaysia.<br />
It should be taken into account that <strong>the</strong> Croatian economy is in big difculties: huge public<br />
debt, unemployment, <strong>the</strong> deprivation <strong>of</strong> quality <strong>of</strong> life, but also <strong>the</strong> leak <strong>of</strong> <strong>the</strong> current<br />
knowledge and skills. In such circumstances <strong>the</strong> development should be conceived considering<br />
existence and fast initialization with new economic initiatives.
128<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
Also, it should be kept in mind that <strong>the</strong> business conditions have changed in last ten years<br />
both in Croatia and in <strong>the</strong> world. There is number <strong>of</strong> barriers in business development<br />
and creative and highly innovative business penetrations will be needed as well as wellconceived<br />
marketing, and organized care for maintaining <strong>the</strong> human resources above all.<br />
Elaboration <strong>of</strong> <strong>the</strong> program <strong>of</strong> development should be <strong>the</strong> beginning. Well-conceived projects<br />
should attract investors. It is important to emphasize that projects cannot be accomplished<br />
just through <strong>the</strong> preparation <strong>of</strong> <strong>the</strong> area <strong>of</strong> industrial zone and calling <strong>the</strong> investors<br />
to come and fulll <strong>the</strong>ir interest, which is <strong>the</strong> present Croatian practice. We should act<br />
completely different from <strong>the</strong> way it’s done actually.<br />
3.3 Projects in EIPR proposed for realization<br />
Territory <strong>of</strong> EIPR covers <strong>the</strong> port <strong>of</strong> Bršica with port area with particular accentuation on<br />
industrial microlocations: Bršica, Štalije and Vlaška, and areas surrounding port area along<br />
<strong>the</strong> river Raša and lime factory, areas along <strong>the</strong> channels to <strong>the</strong> industrial location Vlaška<br />
and <strong>the</strong> city <strong>of</strong> Raša (see picture).<br />
In <strong>the</strong> period from <strong>the</strong> establishment <strong>of</strong> Working group and presentation <strong>of</strong> <strong>the</strong> basic ideas<br />
about settlement <strong>of</strong> <strong>the</strong> eco production in port territory, <strong>the</strong> concept <strong>of</strong> <strong>the</strong> Park was gradually<br />
built. During <strong>the</strong> study based on industrial ecology as <strong>the</strong> basis <strong>of</strong> <strong>the</strong> project, different<br />
industrial, agricultural and service activities came to <strong>the</strong> surface, which could be connected<br />
in <strong>the</strong> community on <strong>the</strong> principles <strong>of</strong> modern eco industrial parks. Special<br />
attention was paid on bi<strong>of</strong>uel production based on imported raw materials, and on <strong>the</strong><br />
o<strong>the</strong>r hand <strong>the</strong> development <strong>of</strong> agricultural production on poorly tended land and processing<br />
agricultural products, as well as on waste water purication and reuse.<br />
There were various ideas about energy, production <strong>of</strong> biogas, wind power station Goli<br />
project and o<strong>the</strong>r useful and complementary ventures, in order to settle operative port and<br />
industrial synergic system.<br />
We will present <strong>the</strong> most important tasks:<br />
The building-up <strong>of</strong> <strong>the</strong> operative port and industrial surface, which will be cleaned <strong>of</strong> <strong>the</strong><br />
ruins from previous industrial period, and <strong>the</strong>n <strong>the</strong> hill between ex quarry Bršica and Štalije<br />
will be removed. For that task <strong>the</strong> Community <strong>of</strong> Raša and Port authority <strong>of</strong> Rijeka should<br />
jointly prepare <strong>the</strong> tende. Stone exploitation as building material with fur<strong>the</strong>r processing<br />
should provide important initial income for development <strong>of</strong> EIPR.<br />
Production <strong>of</strong> bi<strong>of</strong>uel on <strong>the</strong> bio-material basis<br />
Biodiesel as a modern ecological fuel guarantees good positioning <strong>of</strong> <strong>the</strong> Park and connection<br />
with o<strong>the</strong>r economies in Croatia and Middle Europe.<br />
The idea <strong>of</strong> this clue project is supported by knowledge and experience in Biodiesel production<br />
in Croatia. Croatian and foreign experts and business partners were consulted and<br />
<strong>the</strong> whole width was taken in consideration: technical, technological, microeconomics, energetic,<br />
political, organisational and markets knowledge relevant for <strong>the</strong> project realisation,<br />
Croatian, European and global situation and trends have been considered.<br />
With <strong>the</strong> increase <strong>of</strong> bioethanol use, which is mostly imported in Europe from overseas<br />
countries, it opens <strong>the</strong> possibility <strong>of</strong> logistic activity in that direction. Bioenergy business<br />
means creating nal products or semi-products, and could be realized in Štalije location,<br />
and means higher (and more protable) phase <strong>of</strong> technological development in that eld.
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 129<br />
Industrial power plant on bi<strong>of</strong>uels<br />
Industrial location with big production plants should have proper power plant that brings<br />
supplying security, but uses by-products from energy production plant (hot water, steam,<br />
energy power..). Considering great delicacy <strong>of</strong> <strong>the</strong> relation between electro-energy projects<br />
and environment, <strong>the</strong> particular attention was paid to <strong>the</strong> power station on bi<strong>of</strong>uel. The<br />
basic idea is that it should be exclusively in function for local society:<br />
use <strong>of</strong> biodiesel and bio-oils; „green energy“<br />
use <strong>of</strong> biogas as a product in cleaning <strong>of</strong> waste water from Raša and Labin<br />
high level <strong>of</strong> work exibility: industrial power plant<br />
netting with local heat and electric energy consumers; for example food production (greenhouses)<br />
and food processing (dryers, refrigeration plants…)<br />
Water management<br />
Raša and all Labin territory have a large quantity <strong>of</strong> water from <strong>the</strong> coal-mine galleries that<br />
could be used as technical water for irrigation in agriculture. In case it will be shown that<br />
this water could be supplied continuously in necessary clean level, it could be used as an<br />
export product for Adriatic islands and South Italy.<br />
Labin and Raša do have a problem <strong>of</strong> waste water processing, but with <strong>the</strong> use <strong>of</strong> modern<br />
technology where waste is used to produce biogas, and puried water for agricultural irrigation,<br />
we would create a system that should economically and ecologically friendly bring<br />
several benets: purication <strong>of</strong> waste water, produced energy could be used for pumping<br />
water for irrigation in agriculture, puried water could be used for irrigation, ecologically<br />
contaminated channels in valley could be cleaned and used for aquacultural production.<br />
Processing <strong>of</strong> mineral raw materials<br />
Some <strong>of</strong> Istria producers <strong>of</strong> white cement and o<strong>the</strong>r products stopped <strong>the</strong> production or<br />
have uncertain future, because <strong>of</strong> <strong>the</strong> location or lousy industrial politics. The location <strong>of</strong><br />
EIPR is by all means favorable for development <strong>of</strong> complementary production with higher<br />
level <strong>of</strong> processing <strong>of</strong> raw minerals, and means natural development through cooperation<br />
with existing factories.<br />
The micro- and nano-technology allow a wide range <strong>of</strong> products from lime-stone mineral<br />
basis and usefull in processing <strong>of</strong> plastic in food, in metallurgy, in agriculture, in medicine<br />
etc. Building development strategy in that segment is very important task in EIPR project.<br />
Agricultural production and processing<br />
There is a huge potential in vegetable and ower cultivation through modern technology.<br />
To get economically successful project <strong>the</strong> following is important: production direction and<br />
market availability, <strong>the</strong> use <strong>of</strong> new cultivating technologies, surface, quantity <strong>of</strong> production,<br />
irrigation water and irrigation system availability, costs <strong>of</strong> energy for winter period<br />
heating and for product processing, for example drying. On <strong>the</strong> o<strong>the</strong>r hand, <strong>the</strong> production<br />
and processing <strong>of</strong> industrial cultures as hemp might be very protable.<br />
The availability <strong>of</strong> waste heating from energy station might be good condition for economic<br />
success <strong>of</strong> <strong>the</strong> project. In Raša valley <strong>the</strong> conditions: water availability, as well as<br />
possibility <strong>of</strong> <strong>the</strong> building <strong>of</strong> <strong>the</strong> processing facilities for agricultural products are very<br />
good. The port and railway facilitate transportation on <strong>the</strong> market.<br />
3.4 Ecoindustrial Park and <strong>the</strong> Community <strong>of</strong> Raša<br />
Every industrial park is in interaction with its living space (surroundings) and depends on it<br />
from <strong>the</strong> point <strong>of</strong> view <strong>of</strong> natural resources, services and social development and values. It<br />
is extremely important that <strong>the</strong> cooperation and understanding <strong>of</strong> <strong>the</strong> Community’s leadership<br />
and EIPR management is established since <strong>the</strong> initial phase <strong>of</strong> EIPR. The inhabitants
130<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
are included from <strong>the</strong> beginning through <strong>the</strong> public discussions in <strong>the</strong> project and its inuence<br />
on living space, and number <strong>of</strong> benets: contributions to <strong>the</strong> economical and social<br />
development, improvement <strong>of</strong> infrastructure, improvement <strong>of</strong> life quality are accepted as<br />
logical result.<br />
For that purpose <strong>the</strong> consultations are organized, and Forum for socially responsible teritory,<br />
which is stimulating discussions about actual problems <strong>of</strong> sustainable development <strong>of</strong><br />
<strong>the</strong> Community, has been established.<br />
EIPR will be <strong>the</strong> impulse for <strong>the</strong> development <strong>of</strong> educational institutions, and will stimulate<br />
employment and attract working force from <strong>the</strong> surroundings and <strong>the</strong>refore stimulate<br />
residential building. It leads on streng<strong>the</strong>ning and diversication <strong>of</strong> existing infrastructure<br />
and services. Social activities will be developed, cultural institutions will be built, and it is<br />
to be expected that Raša becomes a friendly place for <strong>the</strong> start <strong>of</strong> new businesses, a place<br />
with high quality <strong>of</strong> life.<br />
Following are <strong>the</strong> goals <strong>of</strong> <strong>the</strong> Project<br />
Long term goals<br />
Realisation <strong>of</strong> highly efcient and economically sustainable and efcient ecoindustrial<br />
park <strong>of</strong> manufacturing and service industries, which would be based on principles <strong>of</strong> industrial<br />
symbiosis and realizing benets for <strong>the</strong> inhabitants <strong>of</strong> Raša and Labinština.<br />
Short term goals<br />
Enterprise management in <strong>the</strong> life space and management <strong>of</strong> EIPR<br />
Environment design <strong>of</strong> EIPR, landscape and horticultural design<br />
Renewal <strong>of</strong> <strong>the</strong> transport infrastructure<br />
Renewal and equipment <strong>of</strong> port area Trget – Bršica – Štalije<br />
Management <strong>of</strong> storm water and inundation protection<br />
Ecologically efcient management with waste water from Raša and Labin and with production<br />
<strong>of</strong> biogas<br />
Processing <strong>of</strong> <strong>the</strong> mineral raw material<br />
Agricultural production and processing <strong>of</strong> agricultural products<br />
Production <strong>of</strong> bi<strong>of</strong>uel and management <strong>of</strong> bioenergy<br />
Mariculture<br />
Water from deserted mine galleries as market product<br />
Building <strong>of</strong> Marina and Wind power station Goli as collaborative projects will support ef-<br />
ciency <strong>of</strong> EIPR, so <strong>the</strong>y are considered as <strong>the</strong> additional targets<br />
The development <strong>of</strong> tourism should be designed in harmony with EIPR and leaning on<br />
original ideas and innovation as separately important target<br />
3.5 Stakeholders<br />
We will point out initial statement <strong>of</strong> <strong>the</strong> major <strong>of</strong> Raša in <strong>the</strong> publication Prosperity towards<br />
sustainable development (7):<br />
A set <strong>of</strong> synergy related projects based on production <strong>of</strong> bi<strong>of</strong>uel, management <strong>of</strong> water<br />
and agricultural production, logistic support, commodity business tailored to <strong>the</strong> needs <strong>of</strong><br />
development <strong>of</strong> <strong>the</strong> Community <strong>of</strong> Raša, Istria and Croatia is in complete spirit <strong>of</strong> strategy<br />
<strong>of</strong> EU and Croatia on realisation <strong>of</strong> sustainable development, energy ef ciency and protection<br />
<strong>of</strong> nature with reduction <strong>of</strong> emission. The program will be realized with emphasis on<br />
practice <strong>of</strong> public – private partnership and is consistent with Croatian strategy <strong>of</strong> export<br />
support, and is its important contribution. Local community, Labin area and Istria are<br />
naturally interested for good solutions in activating <strong>the</strong> port <strong>of</strong> Raša, for use <strong>of</strong> industrial<br />
areas and for <strong>the</strong> prosperity <strong>of</strong> <strong>the</strong> economy in that space.
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 131<br />
Local entities that are leaning on available resources (raw material, water, energy, transport)<br />
with this project will gain great developing impulse and support, and location and<br />
market advantages that ensure <strong>the</strong> future prosperity.<br />
Republic <strong>of</strong> Croatia with this project will gain a rounded conception <strong>of</strong> development <strong>of</strong> <strong>the</strong><br />
ports in basin <strong>of</strong> Rijeka, <strong>the</strong> solution for neglected port <strong>of</strong> Raša and impulses for <strong>the</strong> development<br />
<strong>of</strong> <strong>the</strong> rail transport route on <strong>the</strong> eastern part <strong>of</strong> Uka-mountain.<br />
In Middle European area <strong>the</strong> interest for port <strong>of</strong> Raša exists as <strong>the</strong> port specialized for<br />
biomaterial with <strong>the</strong> possibility for <strong>the</strong>ir processing in area settled for industrial plants.<br />
The main reason for such thinking is <strong>the</strong> fact that in o<strong>the</strong>r north Adriatic ports <strong>the</strong>re is no<br />
possibility to accept additional liquid cargo and to supply big biodiesel plants in middle<br />
Europe with vegetable oils, <strong>the</strong>re is a need for sea transport, as well as for supplement <strong>of</strong><br />
bi<strong>of</strong>uel (biodiesel, ethanol). The interest for <strong>the</strong> port <strong>of</strong> Raša has been shown from <strong>the</strong><br />
entrepreneurs from Italy and several Asiatic countries, especially from Malaysia and India.<br />
3.6 Knowledge and experience on which venture relies<br />
The suggested strategy is based on <strong>the</strong> synergy <strong>of</strong> knowledge and experience <strong>of</strong> <strong>the</strong>ir authors,<br />
who participated as individuals in number <strong>of</strong> projects and solved number <strong>of</strong> <strong>engineering</strong><br />
and economic development tasks.<br />
We should emphasize <strong>the</strong> knowledge and experience by which <strong>the</strong> mayor <strong>of</strong> Raša supported<br />
<strong>the</strong> designing <strong>of</strong> <strong>the</strong> strategy, because knowing local society and its wide environment is<br />
especially important for strategy development.<br />
All this knowledge and experience has been implemented in <strong>the</strong> idea <strong>of</strong> this strategy. There<br />
exist huge literature dedicated to development and building <strong>of</strong> ecoindustrial parks, but it<br />
can be understood and implemented just with wide experience and understanding <strong>of</strong> systemic<br />
thinking.<br />
4 Ecoindustrial Park Raša – EIPR - realization<br />
4.1 The way <strong>of</strong> achieving realization <strong>of</strong> EIPR<br />
Good cooperation <strong>of</strong> interested institutions, economic entities and management through<br />
<strong>the</strong> activity <strong>of</strong> Working group, Forum for socially responsible activity and public presentations<br />
and discussions, as well as good cooperation with Croatian System Society, form <strong>the</strong><br />
basis <strong>of</strong> friendly surroundings for realization <strong>of</strong> <strong>the</strong> project EIPR.<br />
Everything mentioned above, toge<strong>the</strong>r with learned experiences, particular knowledge and<br />
information, as well as ga<strong>the</strong>red circle <strong>of</strong> experts, constitutes <strong>the</strong> intellectual capital without<br />
which would it be impossible to initiate realization <strong>of</strong> <strong>the</strong> project EIPR on <strong>the</strong> current<br />
cognitive and project level.<br />
The idea is for <strong>the</strong> Working group to continue its activity as a part <strong>of</strong> expert advisory panel<br />
<strong>of</strong> <strong>the</strong> company for <strong>the</strong> economy in life space and EIPR management.<br />
The rst development task is appointment <strong>of</strong> project manager and establishment <strong>of</strong> <strong>the</strong><br />
management <strong>of</strong> Ecoindustrial Park <strong>of</strong> Raša. This rst in order <strong>of</strong> priority tasks will realize<br />
<strong>the</strong> Working group for strategy <strong>of</strong> development in cooperation with <strong>the</strong><br />
Community <strong>of</strong> Raša and will create a multidisciplinary body – Expert working group or<br />
Expert advisory council <strong>of</strong> appropriate size and competence. It will be a support to <strong>the</strong> project<br />
manager (leader) and will jointly establish and operate <strong>the</strong> Park management. Expert<br />
working group will continue with technical work, dene methodology <strong>of</strong> work, and prepare<br />
necessary documentation and detailed operational plan. Toge<strong>the</strong>r with <strong>the</strong> Forum for<br />
socially responsible activity will promote <strong>the</strong> idea <strong>of</strong> EIPR and report to <strong>the</strong> public about
132<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
progress <strong>of</strong> activities. With <strong>the</strong> progress and raise <strong>of</strong> EIPR it will be established consortium<br />
<strong>of</strong> companies, members <strong>of</strong> Park, which will be important support to <strong>the</strong> activity <strong>of</strong> Park<br />
management.<br />
4.2 Economy <strong>of</strong> <strong>the</strong> project EIPR<br />
Generally it can be predicted that potential <strong>of</strong> <strong>the</strong> space and suggested projects expressed<br />
as an <strong>annual</strong> income is around 200 million €.<br />
In different phases <strong>of</strong> realization <strong>of</strong> EIPR <strong>the</strong> projects need different amount <strong>of</strong> funds, in<br />
relation with actual structure <strong>of</strong> project realization, and are provided from public and private<br />
sources.<br />
In initial activities <strong>the</strong> needs for funds generally are:<br />
execution <strong>of</strong> Preliminary feasibility studies and preliminary <strong>engineering</strong> studies<br />
execution <strong>of</strong> EIP’s infrastructure and marketing activities for attraction <strong>of</strong> investors<br />
execution <strong>of</strong> possible pilot-plant and building <strong>of</strong> objects for initial settlement <strong>of</strong> entrepreneurs<br />
in rent<br />
establishment <strong>of</strong> <strong>the</strong> institutions that support development <strong>of</strong> EIP, as business incubators<br />
and development centers<br />
activities with which local / wider community supports <strong>the</strong> development <strong>of</strong> EIP and acts in<br />
accordance with projected targets<br />
The idea <strong>of</strong> EIPR could be realized if foreign investors with capital are attracted, that will<br />
be driving force <strong>of</strong> <strong>the</strong> project. It should also be considered nancing from EU and Croatian<br />
funds.<br />
4.3 Risks<br />
The hardest thing is to predict, minimize and solve potential risks that could endanger<br />
project. But risk is a part <strong>of</strong> every day’s life and <strong>the</strong> causes should be individuated and separated.<br />
The suggested project is complex and includes number <strong>of</strong> subprojects and <strong>the</strong>refore<br />
different risk causes. We are listing <strong>the</strong> more important possible causes, and in more detailed<br />
way <strong>the</strong> causes will be analyzed during <strong>the</strong> elaboration <strong>of</strong> <strong>the</strong> singular project tasks:<br />
<strong>the</strong> success <strong>of</strong> <strong>the</strong> management <strong>of</strong> <strong>the</strong> EIPR<br />
careful choice <strong>of</strong> collaborative companies<br />
non favorable economic circumstances<br />
competition inuence<br />
Investors are loath to <strong>the</strong> risks, so it is important to decrease <strong>the</strong>ir exposure to <strong>the</strong> risks<br />
through wisely dened conditions <strong>of</strong> investments. Let’s emphasize that, for example, public<br />
private partnership cannot signicantly decrease exposure to risks, but <strong>the</strong>re are several<br />
o<strong>the</strong>r useful approaches <strong>of</strong> decreasing risks that should be examined.<br />
Thinking <strong>the</strong> past and <strong>the</strong> present we cannot think good enough and live <strong>the</strong> already present<br />
future: informatic, biotechnological, nanotechnological that has already begun in highly<br />
developed countries greatly progressed and moved away. Is it possible to think and organize<br />
overcoming <strong>of</strong> <strong>the</strong> technical gap?<br />
Therefore positive thinking should be promoted, which is focused to <strong>the</strong> future, creativeness<br />
and permanent innovation, <strong>the</strong> understanding <strong>of</strong> our exible and creative abilities; it<br />
should wisely be open to <strong>the</strong> most developed world and its permanently changeable life.<br />
Let’s point that <strong>the</strong> old saying which states that mental blockade is <strong>the</strong> biggest obstacle on<br />
progress towards <strong>the</strong> future is true that is so called „embedded human resistance against<br />
anything new“, and <strong>the</strong> Europeans are <strong>the</strong> biggest slaves <strong>of</strong> such resistance and <strong>the</strong>ir lag<br />
behind Americans and Japanese is partly caused because <strong>of</strong> it. We are Europeans as well,
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 133<br />
and we also carry this defect, not so much when spending is concerned, but true when it<br />
concerns work and production. We hardly see <strong>the</strong> embryo <strong>of</strong> big changes, and <strong>the</strong>refore we<br />
defer too late. As a useful example we should examine <strong>the</strong> unsuccessful Lisabon strategy<br />
because <strong>of</strong> which EU Council in 2000 thought EU as „<strong>the</strong> most dynamic and competitive<br />
economy on <strong>the</strong> world…“ and <strong>the</strong> majority <strong>of</strong> targets was not reached till <strong>the</strong> predicted<br />
data, <strong>2010</strong> year.<br />
European culture is one <strong>of</strong> <strong>the</strong> basic risks to <strong>the</strong> possible success <strong>of</strong> our project, and we<br />
can list number <strong>of</strong> o<strong>the</strong>rs. In <strong>the</strong> rst line <strong>the</strong>re is <strong>the</strong> possibility <strong>of</strong> initial nancing <strong>of</strong> <strong>the</strong><br />
venture, understanding <strong>of</strong> <strong>the</strong> surroundings, choice <strong>of</strong> <strong>the</strong> project management, <strong>the</strong>n understanding<br />
<strong>of</strong> <strong>the</strong> surrounding community about <strong>the</strong> need <strong>of</strong> its achievement.<br />
4.4 Development effects achieved with EIPR<br />
On macro plan - Excellently organized community <strong>of</strong> producing agricultural and service<br />
activities taking care <strong>of</strong> industrial ecology by principles <strong>of</strong> ecoindustrial park, dependent<br />
on highly developed infrastructure: sea, road, rail and air.<br />
On micro plan - Support to <strong>the</strong> sustainable development <strong>of</strong> <strong>the</strong> living space and economy<br />
<strong>of</strong> <strong>the</strong> Community <strong>of</strong> Raša, employment and rich social life, work and family life, achieving<br />
reputation <strong>of</strong> <strong>the</strong> place <strong>of</strong> possible business success and quality <strong>of</strong> life.<br />
4.5 Priority tasks<br />
Realization <strong>of</strong> <strong>the</strong> whole EIPR project, as well as partial project courses inside <strong>of</strong> <strong>the</strong> system<br />
starts with <strong>the</strong> realization <strong>of</strong> project documentation and forming <strong>of</strong> singular project<br />
courses with all <strong>the</strong> necessary elements explained that are needed for initiate <strong>the</strong> realization,<br />
as it is used in pr<strong>of</strong>ession. We are mentioning here <strong>the</strong> tasks that should be executed<br />
in short term period, in order to initiate <strong>the</strong> project:<br />
Preliminary project management with life space EIPR and coordination <strong>of</strong> activities<br />
Conceptual space design planning in EIPR<br />
Preliminary project <strong>of</strong> restoring and furnishing <strong>of</strong> <strong>the</strong> Port <strong>of</strong> Raša<br />
Preliminary project <strong>of</strong> activity and management with waste water from Labin and Raša<br />
Preliminary project <strong>of</strong> Cluster for <strong>the</strong> processing <strong>of</strong> <strong>the</strong> mineral raw material<br />
Preliminary project <strong>of</strong> agricultural production<br />
5 Conclusion<br />
The above described idea <strong>of</strong> sustainable development <strong>of</strong> <strong>the</strong> Community <strong>of</strong> Raša and<br />
preparations for development <strong>of</strong> <strong>the</strong> Ecoindustrial Park <strong>of</strong> Raša, as well as description <strong>of</strong><br />
circumstances in which <strong>the</strong> development tasks should be accomplished, tell us about a vision<br />
<strong>of</strong> a Park as rst such Croatian venture: with wise management it can become widely<br />
recognized and accepted successful business model.<br />
During <strong>the</strong> deliberation <strong>of</strong> <strong>the</strong> Strategy <strong>of</strong> development <strong>of</strong> port area Raša points <strong>of</strong> view<br />
were dened, that are starting points <strong>of</strong> sustainable development and with which development<br />
projects and activities should harmonize:<br />
System <strong>of</strong> projects thought on <strong>the</strong> basis <strong>of</strong> space, natural, infrastructural and technological<br />
resources<br />
Activating <strong>of</strong> <strong>the</strong> port <strong>of</strong> Raša and port area <strong>of</strong> Raša and its surroundings, as „anchorage“<br />
<strong>of</strong> EIPR, that will through specializing operations have advantage to <strong>the</strong> o<strong>the</strong>r ports <strong>of</strong><br />
Nor<strong>the</strong>rn Adriatic.<br />
Planned activities in port area won’t disturb o<strong>the</strong>r development projects in spaces or tour-
134<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
istic programs<br />
The area <strong>of</strong> EIPR will be designed by cleaning <strong>of</strong> <strong>the</strong> remained ruins from previous industrial<br />
periods and o<strong>the</strong>r activities (Bršica, Štalije, Vlaška, agricultural land along channel3)<br />
and by removing <strong>the</strong> stones (hill) between <strong>the</strong> deserted stone mine Bršica and Štalije area<br />
operational and industrial space will be formed.<br />
The basis is created for <strong>the</strong> development <strong>of</strong> strategically important and clean activities, <strong>the</strong><br />
production that will in good way place Raša in regional environment (transport, bioenergy,<br />
food, water).<br />
Realization <strong>of</strong> <strong>the</strong> exemplary management with water: storm water and inundation protection,<br />
using <strong>the</strong> water from mine galleries from Raša and Labin, and <strong>the</strong> water from purication<br />
<strong>of</strong> waste water.<br />
It predicts netted system <strong>of</strong> energy management with emphasis on renewable sources <strong>of</strong><br />
energy (biodiesel, biogas, wind power stations)<br />
Revitalisation and realization <strong>of</strong> ecological agricultural production in <strong>the</strong> valley <strong>of</strong> Raša,<br />
considering good climatic conditions, availability <strong>of</strong> water, energetic and o<strong>the</strong>r support <strong>of</strong><br />
industrial systems<br />
Compliance to <strong>the</strong> rigorous norms <strong>of</strong> landscape conservation should harmonize with touristic<br />
activity and harmony with landscape beauty in nearby area.<br />
Realization <strong>of</strong> EIPR project follows usual procedure and phases <strong>of</strong> development for such<br />
project with predicted tasks and subprojects have to be realized:<br />
Company for space management (PGP) and management <strong>of</strong> EIPR<br />
Designing <strong>the</strong> space <strong>of</strong> EIPR: functional, landscape and horticultural design<br />
Renewal <strong>of</strong> trafc infrastructure<br />
Renewal and furnishing <strong>of</strong> <strong>the</strong> port-area Bršica-Štalije<br />
Management <strong>of</strong> storm water and inundation protection<br />
Ecologically efcient treatment <strong>of</strong> waste water from Labin and Raša with biogas production<br />
Basis for forming <strong>of</strong> a Cluster for mineral raw material processing and strategy <strong>of</strong> development<br />
in that segment<br />
Agricultural production and processing <strong>of</strong> agricultural products<br />
Production <strong>of</strong> bi<strong>of</strong>uel and management <strong>of</strong> bio energy<br />
Mariculture<br />
Water from deserted mine galleries as market product<br />
Building <strong>of</strong> marina and wind power station Goli as collaborative projects that would be<br />
supportive to <strong>the</strong> success <strong>of</strong> EIPR, so are emphasized as possible additional targets<br />
Realization <strong>of</strong> <strong>the</strong> whole project EIPR as well as singular units inside <strong>the</strong> Park will start<br />
with design <strong>of</strong> project documentation.<br />
6 References<br />
1. Andrašec M. (2009), Tehnologijske platforme, hrvatski razvojni projekti u EU<br />
okruženju. In: Božievi J. (eds) Inovacijska kultura i tehnologijski razvoj, CROSS,<br />
Zagreb, pp 109-114<br />
2. Baas L. (2008), Industrial Symbiosis in <strong>the</strong> Rotterdam Harbour and Industrial Complex,<br />
Bus. Strat. Env 17, pp 330-340<br />
3. Božicevic J. (1993), Private communication, Zagreb<br />
4. Božievi J. (ed) (1999), Luka kao složen sustav, Rijeka – glavna hrvatska luka,<br />
<strong>HATZ</strong>-CROSS, Zagreb
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 135<br />
5. Božievi J. (ed) (1999), Povezani Rijeka i Zagreb hrvatsko i europsko gospodarsko<br />
žarište, <strong>HATZ</strong>, Zagreb<br />
6. Božievi J., Andrašec M. (2009), Opina Raša – Mudra zajednica, Opina Raša,<br />
Raša<br />
7. Božievi J., Andrašec M. (2009), Napredak prema održivom razvoju, Opina Raša,<br />
Raša<br />
8. Doyle B. et al. (1996), Eco-industrial Parks, TRI, Research Triangle Park, NC, USA<br />
9. Fleig A. K. (2000), Eco industrial Parks, Deutsche Gesellschaft fur Technische Zusamanarbeit<br />
GmbH, Eschborn, Deutschland<br />
10. Lowe E. A. (2001), Eco-industrial Park Handbook for Asian Developing Countries,<br />
Report to Asian Development Bank, Indigo Development, Hidden Valley Lake, CA,<br />
USA<br />
11. Koenig A.W. (ed.) (2005), The Eco-Industrial Park Development, A Guide for Chinese<br />
Government Ofcials and Industrial Park Managers, EU-China Environmental<br />
Management Cooperation Programme - Industry Development, Beijing<br />
12. Tarantini M., Paolo A., Dominici A., Peruzzi A., Dell’Isola M. (2007), Guidelines for<br />
<strong>the</strong> Settlement and <strong>the</strong> Management <strong>of</strong> <strong>the</strong> Sustainable Industrial Areas, The Experience<br />
<strong>of</strong> <strong>the</strong> LIFE SIAM Project, Bologna
136<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
POSIBLE MODEL OF MATHEMATICAL EVALUATION OF<br />
BEHAVIOUR DISORDER THERAPY<br />
Pr<strong>of</strong>. Dr. sc. Marijan Bošnjak, retired;<br />
member emeritus <strong>of</strong> Croatian Academy <strong>of</strong> Engineering (<strong>HATZ</strong>),<br />
HR-10000 Zagreb, Croatia;<br />
Summary<br />
Because <strong>of</strong> enormous complexity <strong>of</strong> mental and behavioural disorders and illnesses <strong>the</strong>re<br />
is still continuous need for fur<strong>the</strong>r advances in <strong>the</strong>ir treatment. Treatment frequently asks<br />
<strong>the</strong> application <strong>of</strong> chemo<strong>the</strong>rapeutic agents in addition to o<strong>the</strong>r conventional <strong>the</strong>rapeutic<br />
actions. The effects <strong>of</strong> treatments should be adequately evaluated. In this work special<br />
emphasis is given to medical treatments when chemo<strong>the</strong>rapeutic agents are applied.<br />
Therefore, <strong>the</strong> major importance was given to chemo<strong>the</strong>rapeutic agent pharmacokinetics.<br />
Consequently, <strong>the</strong> ma<strong>the</strong>matical model composed from more than twenty differential<br />
equations was applied to describe pharmacokinetics <strong>of</strong> chemo<strong>the</strong>rapeutic agent distribution<br />
in human body, kinetics <strong>of</strong> development <strong>of</strong> positive and negative effects <strong>of</strong> applied<br />
chemo<strong>the</strong>rapeutic treatment, effects on body viability and development <strong>of</strong> compounds<br />
inhibiting <strong>the</strong> effects <strong>of</strong> present chemo-<strong>the</strong>rapeutic agent and causing <strong>the</strong> organism<br />
dependence on applied substances. Computer simulation was applied to test ma<strong>the</strong>matical<br />
model convenience. Obtained computer simulation data con rmed <strong>the</strong> model adequacy<br />
suggesting its application in explaining real cases <strong>of</strong> combined psycho<strong>the</strong>rapy treatments.<br />
Special convenience <strong>of</strong> proposed ma<strong>the</strong>matical model type showed to be its con rmation<br />
<strong>of</strong> necessity for controlling chemo<strong>the</strong>rapeutic substance distribution in bodies <strong>of</strong> treated<br />
patients as a prerequisite for more reliable successful patient treatments with reduced<br />
accompanied negative side effects.<br />
Key words: Psycho<strong>the</strong>rapy pharmacokinetics. Ma<strong>the</strong>matical modelling. Computer<br />
simulation application.<br />
1 Introduction<br />
Despite <strong>the</strong> evident advances in discovering <strong>the</strong> causes <strong>of</strong> mental and behavioural disorders<br />
and illnesses as well as in increasing <strong>the</strong> frequency <strong>of</strong> successful patient disorder and<br />
illness treatments, resulting in patient recoveries, <strong>the</strong>re is still continuous need for fur<strong>the</strong>r<br />
advances in this area <strong>of</strong> medicine, because <strong>of</strong> enormous complexity <strong>of</strong> problems asking to<br />
be successfully solved. One <strong>of</strong> important prerequisites for fur<strong>the</strong>r advances appears to be<br />
<strong>the</strong> application <strong>of</strong> appropriate approaches to <strong>the</strong> evaluation <strong>of</strong> results <strong>of</strong> medical treatments.<br />
Some consequences (positive and/or negative) <strong>of</strong> applied medical treatment can appear<br />
even very fast after its start, <strong>the</strong> o<strong>the</strong>rs one can observe much later, whereas it can also<br />
happen that <strong>the</strong> desired effects cannot appear at all. Of course, <strong>the</strong> best situation can be<br />
considered that when desired positive effects appear already during <strong>the</strong> beginning phase <strong>of</strong><br />
medical treatment.<br />
Since, in addition to direct communications <strong>of</strong> psychiatrists and employed personnel with<br />
patients <strong>the</strong> psycho<strong>the</strong>rapy includes also different psychophysical activities and frequently
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 137<br />
an application <strong>of</strong> chemo<strong>the</strong>rapeutic agents, <strong>the</strong> particular effects <strong>of</strong> all mentioned <strong>the</strong>rapeutic<br />
actions should be appropriately evaluated prior to <strong>the</strong> conclusions referring to <strong>the</strong> answer<br />
on <strong>the</strong> question whe<strong>the</strong>r <strong>the</strong> applied <strong>the</strong>rapy was adequate and whe<strong>the</strong>r it can be considered<br />
as successful with respect to patient recovery and his health improvement.<br />
This work refers to an attempt <strong>of</strong> advancing medical treatment by developing <strong>the</strong> one <strong>of</strong><br />
possible models for ma<strong>the</strong>matical evaluation <strong>of</strong> results <strong>of</strong> applied psycho<strong>the</strong>rapy. Although<br />
<strong>the</strong> present approach to <strong>the</strong> development <strong>of</strong> ma<strong>the</strong>matical model does not neglect a decisive<br />
role <strong>of</strong> engaged psychiatrists and o<strong>the</strong>r pr<strong>of</strong>essional persons, <strong>the</strong> emphasis is given to <strong>the</strong><br />
role <strong>of</strong> applied chemo<strong>the</strong>rapeutic agents and <strong>the</strong>ir effects on patients subjected to <strong>the</strong><br />
<strong>the</strong>rapy. The impression based on <strong>the</strong> real practice <strong>of</strong> neglecting <strong>the</strong> relevance for assaying<br />
or estimating amounts <strong>of</strong> chemo<strong>the</strong>rapeutic agents in particular compartments <strong>of</strong> patient<br />
body (mainly caused by nancial reasons) provoked such an orientation to ma<strong>the</strong>matical<br />
modelling.<br />
2 Ma<strong>the</strong>matical modelling<br />
2.1 Development <strong>of</strong> fundamental model<br />
As already long time known 1, 2, 10, 14, different approaches to <strong>the</strong> ma<strong>the</strong>matical<br />
modelling <strong>of</strong> biochemical reaction systems can be applied. In <strong>the</strong> case <strong>of</strong> biological<br />
systems <strong>the</strong> frequent practice is to describe <strong>the</strong>m by systems <strong>of</strong> differential equations 1-<br />
10, 12-15. These can be relatively simple (composed from 2 – 3 differential equations)<br />
or complex ones characterised with series <strong>of</strong> differential equations. Commonly, <strong>the</strong> simple<br />
systems <strong>of</strong> differential equations can be solved analytically or by applying computer<br />
simulation, whereas for complex systems <strong>the</strong> best choice appears to be <strong>the</strong> application<br />
<strong>of</strong> computer simulations. As pointed out in recent publication 2A, <strong>the</strong> whole system <strong>of</strong><br />
human organism certainly is so complex that one can consider illusory to expect that one<br />
can perfectly describe it by adequate ma<strong>the</strong>matical model. It is also pointed out that <strong>the</strong><br />
accepted practice became <strong>the</strong> description only <strong>of</strong> parts <strong>of</strong> realty and mainly more or less<br />
approximately. Consequently, process events in live organisms can be described by series<br />
<strong>of</strong> appropriate differential equations, as demonstrated with reference to microbial cells<br />
and/or microbial biomass 2-6, 8-10, 12-15 and specically with reference to mammalian<br />
bodies 2, 8, 9, 15. The data published recently 2 could be considered to be <strong>the</strong> most<br />
relevant ones in forming <strong>the</strong> ma<strong>the</strong>matical model <strong>of</strong> this work.<br />
Prior to <strong>the</strong> beginning <strong>of</strong> possible ma<strong>the</strong>matical model formation one should select and<br />
group <strong>the</strong> most relevant patient body process events assumed to happen during <strong>the</strong> applied<br />
<strong>the</strong>rapy and to indicate (at least roughly) <strong>the</strong> most probable <strong>the</strong>rapy consequences. In<br />
any case <strong>the</strong> <strong>the</strong>rapy effects on patient survival and viability maintenance should not be<br />
neglected in forming a ma<strong>the</strong>matical model. The emphasis given to <strong>the</strong> evaluation <strong>of</strong> applied<br />
chemo<strong>the</strong>rapeutic agent effects asks a more detailed consideration <strong>of</strong> chemo<strong>the</strong>rapeutic<br />
agent pharmacokinetics. Data and explanations expressed in <strong>the</strong> recent work 2A suggest<br />
a similar approach in this work.<br />
Usual practice is to apply chemo<strong>the</strong>rapeutic agents orally, although in some cases a<br />
parenteral application can be recommended. When applying chemo<strong>the</strong>rapeutic agents<br />
orally one <strong>of</strong> possibilities is to consider that ventricular-intestinal organism compartment<br />
is composed from two sub-compartments. Then one can start ma<strong>the</strong>matical modelling by<br />
putting differential equations:<br />
dF ve<br />
/dt =-k 1<br />
F ve<br />
/1/
138<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
dF it<br />
/dt =k 1<br />
F ve<br />
–k 2<br />
F it<br />
–k 3<br />
F it<br />
/2/<br />
if neglecting chemo<strong>the</strong>rapeutic drug resorption rate value in ventricular part (Such<br />
oversimplication cannot be recommended in every case), and if supposing that some part<br />
<strong>of</strong> drug would be excreted through faeces, i. e. if one can apply <strong>the</strong> equation<br />
dF fc<br />
/dt =k 2<br />
F it<br />
/3/<br />
If, however, a chemo<strong>the</strong>rapeutic agent resorption rate is enough signicant already in<br />
ventricular part, <strong>the</strong>n instead <strong>of</strong> equation /1/ <strong>the</strong> following equation should be applied:<br />
dF ve<br />
/dt =-k 1<br />
F ve<br />
–aF ve<br />
/1a/<br />
All organism parts are interconnected adequately and <strong>the</strong>refore communication between<br />
<strong>the</strong>m is well established when one considers normal viable organism. However, with<br />
respect to organism survival and communications between organism parts <strong>the</strong> main role<br />
could be assigned to <strong>the</strong> cardio-vascular system and its connection with respiratory system,<br />
if respecting <strong>the</strong> control role <strong>of</strong> cerebral-neural system.<br />
Since <strong>the</strong> blood communicates with all organism compartments <strong>the</strong> rate <strong>of</strong> changes <strong>of</strong><br />
chemo<strong>the</strong>rapeutic drug mass in <strong>the</strong> blood could be expressed by equation<br />
dF sng<br />
/dt =k 3<br />
F it<br />
–(k 4<br />
+k 6<br />
+k 7<br />
+k 9<br />
+k k<br />
) F sng<br />
+k 5<br />
F cr<br />
+k 8<br />
F hp<br />
+k a<br />
F t<br />
+<br />
+k i<br />
F men<br />
/4/<br />
or by equation<br />
dF sng<br />
/dt =a F ve<br />
+k 3<br />
F it<br />
–(k 4<br />
+k 6<br />
+k 7<br />
+k 9<br />
+k k<br />
) F sng<br />
+k 5<br />
F cr<br />
+k 8<br />
F hp<br />
+k a<br />
F t<br />
+<br />
+k i<br />
F men<br />
/4a/<br />
Communication with nephritic-urinary system is mainly irreversible and <strong>the</strong>refore drug<br />
transfer rate towards this system can be presented by applying <strong>the</strong> equation<br />
dF nu<br />
/dt =k 6<br />
F sng<br />
/5/<br />
The liver has important role and communicates reversibly with blood. The equation<br />
dF hp<br />
/dt =k 7<br />
F sng<br />
–k 8<br />
F hp<br />
/6/<br />
is considered to be ma<strong>the</strong>matically adequate.<br />
For <strong>the</strong> main part <strong>of</strong> tissue one can apply <strong>the</strong> equations:<br />
dF t<br />
/dt =k 9<br />
F sng<br />
–(k a<br />
+k d<br />
) F t<br />
+k e<br />
F tr<br />
/7/<br />
dF tr<br />
/dt =k d<br />
F t<br />
–k e<br />
F tr<br />
/8/<br />
if drug transfer to tissue receptor happens reversibly.<br />
The hearth communicates reversibly with <strong>the</strong> blood, as <strong>the</strong> engine <strong>of</strong> blood circulation<br />
and as <strong>the</strong> main part <strong>of</strong> cardio-vascular system. Its reversible communication with <strong>the</strong>
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 139<br />
respiratory system is necessary and is a prerequisite <strong>of</strong> body survival. The equations<br />
and<br />
dF cr<br />
/dt =k 4<br />
F sng<br />
–(k 5<br />
+k b<br />
) F cr<br />
+k c<br />
F rsp<br />
/9/<br />
dF rsp<br />
/dt =k b<br />
F cr<br />
–k c<br />
F rsp<br />
/10/<br />
appear to be adequate for application.<br />
Cerebral-neural and consequently mental system should be obligatorily considered. The<br />
rate <strong>of</strong> changes <strong>of</strong> drug amount in this system can be represented by equation<br />
dF men<br />
/dt =k k<br />
F sng<br />
–k i<br />
F men<br />
/11/<br />
Presence <strong>of</strong> chemo<strong>the</strong>rapeutic substance in <strong>the</strong> cerebral-neural system can induce positive<br />
and/or negative effects inuencing <strong>the</strong>ir intensities. One <strong>of</strong> possibilities to express <strong>the</strong>se<br />
consequences could be by applying <strong>the</strong> equations<br />
dMenp/dt =k u<br />
F men<br />
(1-Menp/Mpm) /12/<br />
dMenn/dt =k v<br />
F men<br />
(1-Menn/Mnm) /13/<br />
Equations /12/ and /13/ determine <strong>the</strong> maximal values which can be attained during<br />
medication, if neglecting <strong>the</strong> probable inhibitory effects <strong>of</strong> increased F men<br />
quantities. Since<br />
it is much more probable that <strong>the</strong> inhibitory effects would reect ra<strong>the</strong>r on positive than<br />
negative consequences, it seems to be sufcient to modify <strong>the</strong> equation /12/ with taking<br />
into account inhibitory effects <strong>of</strong> increasing F men<br />
quantities. The equation,<br />
dMenp/dt =k u<br />
F men<br />
(1-fF men<br />
-Menp/Mpm)<br />
/12a/<br />
can be considered adequate one for an explanation <strong>of</strong> expected consequences.<br />
The experience and experts recommendations concerning <strong>the</strong> application <strong>of</strong> chemo<strong>the</strong>rapeutic<br />
agents in psycho<strong>the</strong>rapy suggest us to be cautious, especially in cases <strong>of</strong> longterm<br />
<strong>the</strong>ir application. As mentioned recently 7, none <strong>of</strong> known antipsychotics used to<br />
treat patients is quite adequate. Those called <strong>the</strong> typical antipsychotics are regarded as<br />
efcacious at ameliorating positive symptoms, but generally ineffective against negative<br />
symptoms and with accompanied displaying <strong>of</strong> extra-pyramidal side effects. Atypical<br />
antipsychotics are also ineffective against negative symptoms. Therefore, one should<br />
consider possible negative effects <strong>of</strong> applied chemical substances on organism body <strong>of</strong><br />
patients subjected to psycho<strong>the</strong>rapy. One <strong>of</strong> ways <strong>of</strong> evaluating <strong>the</strong> consequences is by<br />
estimating values <strong>of</strong> arbitrarily dened body viability, Bv, and body non-viability, Bnv. In<br />
present approach it is supposed that differential equations<br />
dB v<br />
/dt =-k s<br />
.B v<br />
F sng<br />
+k r<br />
S rv<br />
B nv<br />
/14/<br />
dB nv<br />
/dt =k s<br />
B v<br />
F sng<br />
–k r<br />
S rv<br />
B nv<br />
/15/<br />
dS rv<br />
/dt =-k r<br />
S rv<br />
B nv<br />
+k rv<br />
(1-S rv<br />
) /16/
140<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
can serve adequately to such a purpose, if one considers that live organism by his metabolism<br />
supplies continuously <strong>the</strong> own body with energy and appropriate substances in necessary<br />
amounts for survival, and that “revitalization substrate”, S rv<br />
, as <strong>the</strong> notion represents <strong>the</strong>m.<br />
The effects <strong>of</strong> psycho<strong>the</strong>rapy on patient recovery and his health improvement should be<br />
evaluated. However, one cannot say that actually reliable and exactly dened methods are<br />
applied systematically. Therefore, every attempt in approaching to <strong>the</strong> problem solution<br />
can be useful. One <strong>of</strong> possibilities seems to be by proposing <strong>the</strong> ma<strong>the</strong>matical expression<br />
in which <strong>the</strong> recognised effects are considered with respect to <strong>the</strong> participation <strong>of</strong><br />
particular factors <strong>of</strong> inuence. No doubt, both <strong>the</strong> efciency <strong>of</strong> psychiatrists and employed<br />
pr<strong>of</strong>essional personnel and effects <strong>of</strong> applied chemo<strong>the</strong>rapeutic substances should be<br />
evaluated on appropriate ways. As <strong>the</strong> rst step in searching an acceptable ma<strong>the</strong>matical<br />
model could be <strong>the</strong> investigation <strong>of</strong> application convenience <strong>of</strong> equations<br />
dPs/dt =z(1-Ps) /17/<br />
dUt/dt =wPs+xMenp-yMenn /18/<br />
Since <strong>the</strong> decision on <strong>the</strong> choice and dosing psycho<strong>the</strong>rapeutic agents essentially depends<br />
on psychiatrists, perhaps it is more adequate instead <strong>of</strong> equation /17/ to apply <strong>the</strong> equation<br />
dPs/dt =z(1-Ps)(1+Menp)/(1+Menn)<br />
/17a/<br />
The idea to propose equations /17/, /17a/ and /18/ is based on <strong>the</strong> assumption that engaged<br />
psychiatrists and pr<strong>of</strong>essional personnel are ethically and by <strong>the</strong>ir education and experience<br />
quite adequate to act efciently in positive sense. It is also supposed that both positive and<br />
negative consequences should be evaluated, primarily without an investigation whe<strong>the</strong>r <strong>the</strong><br />
applied chemo<strong>the</strong>rapeutic agent acts as reaction substance in biochemical processes or as<br />
signal substance inducing <strong>the</strong> biochemical processes which lead to desired patient behaviour<br />
and health improvements. Although <strong>the</strong>re is series <strong>of</strong> o<strong>the</strong>r factors playing important roles<br />
in people <strong>the</strong>rapies (religion, meteorological factors, environmental conditions in general,<br />
etc.), <strong>the</strong> evaluation <strong>of</strong> consequences <strong>of</strong> <strong>the</strong>ir inuence cannot be recommended for detailed<br />
consideration in this work, because <strong>of</strong> actual insufcient knowledge to dene and estimate<br />
exactly <strong>the</strong> effects <strong>of</strong> <strong>the</strong>ir inuence.<br />
2.2 Modelling <strong>of</strong> <strong>the</strong> development <strong>of</strong> organism dependence on applied drugs<br />
Frequently, organism dependence on applied drug can be induced to develop. Then it<br />
can be produced even with danger and irreversible consequences. Various mechanisms<br />
<strong>of</strong> induction and development <strong>of</strong> organism dependence on drugs can be supposed. With<br />
respect to ma<strong>the</strong>matical modelling one <strong>of</strong> possibilities is to suppose that a dependence<br />
product, Dp, is produced in proportion to <strong>the</strong> rate <strong>of</strong> negative effects, Menn production<br />
and that it acts as an inhibitor <strong>of</strong> positive effects and/or as a drug inactivation factor. If,<br />
supposedly, one could consider that induction and development <strong>of</strong> dependence product<br />
formation mainly happen in <strong>the</strong> cerebral-neural compartment, and if roughly it could be<br />
sufcient to focus our interest to events taking place in cerebral-neural system, <strong>the</strong>n <strong>the</strong><br />
following approach could be applied in transforming <strong>the</strong> discussed ma<strong>the</strong>matical model:<br />
Let suppose that <strong>the</strong> dependence product, Dp inactivates <strong>the</strong> applied drug to high extent<br />
irreversibly, i.e. in accordance to <strong>the</strong> scheme:
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 141<br />
Dp + Fmen Fmni<br />
/Sh-1/<br />
If <strong>the</strong> expressed reduction rate <strong>of</strong> positive effects becomes proportional to Dp amount, <strong>the</strong>n<br />
<strong>the</strong> applicability <strong>of</strong> following equations can be expected:<br />
dD p<br />
/dt =b dMenn/dt –c D p<br />
F men<br />
/19/<br />
dF mni<br />
/dt =c D p<br />
F men<br />
/20/<br />
dF men<br />
/dt =k k<br />
F sng<br />
–k i<br />
F men<br />
- c D p<br />
F men<br />
/21/<br />
2.3 When one can stop to apply chemo<strong>the</strong>rapeutic agent?<br />
If one observes that positive effects do not show a tendency to decrease, <strong>the</strong>n one can<br />
recommend a starting <strong>of</strong> <strong>the</strong> procedure <strong>of</strong> decreasing drug doses, or even <strong>of</strong> stopping a<br />
given drug application, especially if with stopping drug application negative effects would<br />
start to reduce. Therefore, it appears that instead <strong>of</strong> equation /12a/ <strong>the</strong> equation /22/ can be<br />
recommended for application, i.e.<br />
dMenp/dt =k u<br />
F men<br />
(1-fF men<br />
-Menp/M pm<br />
) –d D p<br />
/22/<br />
Such a consideration does not ask detailed discussion referring to <strong>the</strong> kinetics <strong>of</strong> inactivated<br />
drug, F mni<br />
. If both positive and negative effects will spontaneously reduce in proportion to<br />
<strong>the</strong>ir specic rates, <strong>the</strong>n instead <strong>of</strong> equation /22/ and equation /13/ one can propose <strong>the</strong><br />
equations:<br />
dMenp/dt =k u<br />
F men<br />
(1-fF men<br />
-Menp/M pm<br />
) –d D p<br />
–g Menp /23/<br />
dMenn/dt =k v<br />
F men<br />
(1-Menn/M nm<br />
) –h Menn /24/<br />
Equations /23/ and /24/ suggest <strong>the</strong> drugs with higher h/g values should be preferred for<br />
application under a condition <strong>the</strong>y do not differ with respect to o<strong>the</strong>r <strong>the</strong>ir properties.<br />
To facilitate <strong>the</strong> understanding <strong>of</strong> properties <strong>of</strong> transformed ma<strong>the</strong>matical model some<br />
simplications can be introduced in order to have more convenient nal model. It seems<br />
that lumping <strong>of</strong> <strong>the</strong> terms F ve<br />
and F it<br />
into common F it<br />
, as well as <strong>of</strong> F hp<br />
, F tr<br />
and F t<br />
into<br />
common term F t<br />
could markedly correspond to <strong>the</strong> desired goal. Then <strong>the</strong> equations /1/,<br />
/1a/, /6/ and /8/ could be neglected, whereas instead <strong>of</strong> equations /2/, /4/, /4a/ and /7/ <strong>the</strong><br />
corresponding equations /25/, /26/ and /27/ could be applied:<br />
dF it<br />
/dt =-k 1<br />
.F it<br />
–k 2<br />
.F it<br />
/25/<br />
dF sng<br />
/dt =k 1<br />
F it<br />
–(k 4<br />
+k 6<br />
+k 9<br />
+k k<br />
) F sng<br />
+k 5<br />
F cr<br />
+k a<br />
F t<br />
+k i<br />
F men<br />
/26/<br />
dF t<br />
/dt =k 9<br />
F sng<br />
–k a<br />
F t<br />
/27/<br />
2.4 Modelling <strong>of</strong> <strong>the</strong> psycho<strong>the</strong>rapeutic substance transformation<br />
The applied psycho<strong>the</strong>rapeutic agent, depending on its chemical structure and its
142<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
instantaneous position in organism body, can be and usually is transformed into degradation<br />
products, at least partly. Degree and <strong>the</strong> rate <strong>of</strong> transformation depend on both substance<br />
characteristics and biodegradative capability <strong>of</strong> particular organism, especially <strong>of</strong> organism<br />
corresponding parts. The formed degradation products can be <strong>the</strong>rapeutically inactive,<br />
more or less toxic, <strong>of</strong> higher or lower <strong>the</strong>rapeutic effects, when compared with original<br />
substance, etc. Commonly, signicant or even <strong>the</strong> main part <strong>of</strong> applied substance mass<br />
is eliminated from organism body as <strong>the</strong> substance <strong>of</strong> original structure, while <strong>the</strong> o<strong>the</strong>r<br />
part is eliminated as applied substance degradation products. In many cases, <strong>the</strong> applied<br />
substance is primarily converted into <strong>the</strong> product (metabolite) <strong>of</strong> similar <strong>the</strong>rapeutic effects<br />
as original substance. Of course, <strong>the</strong>re is spectrum <strong>of</strong> efciency which could be expressed<br />
by formed metabolites. In this work, pharmacokinetics <strong>of</strong> formed <strong>the</strong>rapeutically active<br />
product is considered to demonstrate how some specic situations can be described by<br />
corresponding ma<strong>the</strong>matical model. Substance transformation can be considered as to<br />
happen in organism body in general, or as to happen mainly in <strong>the</strong> specic tissue (e.g.<br />
in liver), or in body tissue in general. Since cardiovascular system communicate with<br />
every particular body compartment, one can consider <strong>the</strong> substance transformation quickly<br />
reected to <strong>the</strong> state in cardiovascular system. Therefore, in <strong>the</strong> extended ma<strong>the</strong>matical<br />
model, instead <strong>of</strong> equation /26/ <strong>the</strong> equations<br />
dF sng<br />
/dt =k 1<br />
Fit-(k 4<br />
+k 6<br />
+k 9<br />
+k k<br />
+j)F sng<br />
+k 5<br />
F cr<br />
+k a<br />
F t<br />
+k i<br />
F men<br />
/26a/<br />
dF tsng<br />
/dt =jF sng<br />
-(k 6<br />
+k k<br />
)F tsng<br />
+k a<br />
F tt<br />
+k i<br />
F mt<br />
/28/<br />
should be applied, if neglecting <strong>the</strong> formed product afnity to <strong>the</strong> hearth and if supposing<br />
its transfers concerning o<strong>the</strong>r body compartments happen with same specic rates as<br />
those referring to transfers <strong>of</strong> <strong>the</strong> original substance. In addition, <strong>the</strong> corresponding new<br />
equations should be also included:<br />
dF nut<br />
/dt =k 6<br />
F tsng<br />
/29/<br />
dF tt<br />
/dt =k 9<br />
F tsng<br />
-k a<br />
F tt<br />
/30/<br />
dF mt<br />
/dt =k k<br />
F tsng<br />
-k i<br />
F mt<br />
/31/<br />
It is supposed that formed product inuences also positive and negative effects <strong>of</strong> patient<br />
behaviour as well as body viability. Therefore, instead <strong>of</strong> equations /12/, /12a/, /13/, /14/<br />
and /15/ <strong>the</strong> following equations should be applied:<br />
dMenp/dt =k u<br />
(F men<br />
+F mt<br />
)(1-fF men<br />
-Menp/M pm<br />
) /32/<br />
dMenn/dt =k v<br />
(F men<br />
+F mt<br />
)(1-Menn/M nm<br />
) /33/<br />
dB v<br />
/dt =-k s<br />
B v<br />
(F sng<br />
+F tsng<br />
)+k r<br />
S rv<br />
B nv<br />
/34/<br />
dB nv<br />
/dt =k s<br />
B v<br />
(F sng<br />
+F tsng<br />
)-k r<br />
S rv<br />
B nv<br />
/35/<br />
Of course, <strong>the</strong> o<strong>the</strong>r possibilities could be supposed. However, <strong>the</strong>re is no need to consider<br />
<strong>the</strong>m now.
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 143<br />
2.5 Relevance <strong>of</strong> drug application method and pharmaceutical product form<br />
It is known that different drug application methods are practised. Parenteral application<br />
is preferred when fast effects are desired and/or when active substance is sensitive to <strong>the</strong><br />
action <strong>of</strong> active factors present in ventricular-intestinal compartment. Active substance<br />
stability and pharmaceutical drug form properties play important role in applied drug<br />
convenience and <strong>the</strong>rapeutic efciency. Active substance half-life time is considered to<br />
be one <strong>of</strong> very relevant pharmacokinetics parameters, especially when different active<br />
substances should be applied during <strong>the</strong>rapy. In <strong>the</strong> case <strong>of</strong> per oral drug application <strong>the</strong><br />
resorption rate <strong>of</strong> active substance markedly affects its pharmacokinetics and <strong>the</strong>refore<br />
<strong>the</strong>rapeutic consequences. The fast resorption rate is desired when <strong>the</strong> fast <strong>the</strong>rapeutic<br />
effects should result (e.g. when applying hypnotics). However, for longer <strong>the</strong>rapy periods<br />
and for maintenance <strong>of</strong> more stable active substance concentrations with minimal<br />
oscillations <strong>of</strong> active substance concentrations in corresponding organism compartments<br />
<strong>the</strong> slower active substance resorption rates could be recommended to be realised. Some <strong>of</strong><br />
computer simulation data demonstrating <strong>the</strong> effects <strong>of</strong> different substance resorption rates<br />
are shown in Tables 2 – 3 and Figs 3a,b.<br />
3 Computer simulations<br />
Computer programme Scientist enables a solving <strong>of</strong> number <strong>of</strong> different scientic problems.<br />
Based on previous and especially <strong>of</strong> recent experience 2, 5, 6 it was applied also in this<br />
work.<br />
4 Results and discussion<br />
4.1 Fundamental comments<br />
Chosen examples <strong>of</strong> computer simulations are presented in Fig. 1 to Fig. 3. To see how <strong>the</strong><br />
estimated consequences relate to quantities in particular organism body compartment <strong>the</strong><br />
calculation results are presented in Table 1.<br />
Table 1<br />
Effects <strong>of</strong> some psycho<strong>the</strong>rapeutic parameters (k u<br />
, k v<br />
, w, x, y, z) on <strong>the</strong> psycho<strong>the</strong>rapy<br />
efciency (Ut). Ut values refer to <strong>the</strong> 50 th hour <strong>of</strong> <strong>the</strong> computer simulation process.<br />
Parameter<br />
name<br />
Values<br />
ku kv w x y z Ut<br />
0.02 0.015 0.001 0.001 0.001 0.001 0.00492<br />
0.022 0.025 0.001 0.001 0.001 0.001 -0.000525<br />
0.02 0.02 0.001 0.001 0.001 0.001 0.00123<br />
0.02 0.02 0.001 0.002 0.001 0.001 0.0218<br />
0.02 0.02 0.001 0.001 0.002 0.001 -0.01934<br />
0.02 0.02 0.002 0.001 0.002 0.001 -0.01811<br />
0.02 0.02 0.002 0.001 0.002 0.002 -0.01573<br />
0.02 0.01 0.002 0.001 0.002 0.002 0.000668<br />
0.02 0.01 0.002 0.002 0.001 0.002 0.033608<br />
It is known that distribution <strong>of</strong> pharmacokinetic substances depends on substance properties,
144<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
way <strong>of</strong> its application and its applied amount, afnity <strong>of</strong> organism compartments to accept<br />
it and to retain its quantity, particular organism compartment mass and physiological state,<br />
whole organism viability and activity, etc. In this work, values <strong>of</strong> computer simulation<br />
parameters were chosen arbitrarily in order to demonstrate acceptability <strong>of</strong> proposed<br />
approach and possible applicability <strong>of</strong> supposed ma<strong>the</strong>matical model, expecting for both<br />
to be conditionally adequate. Chosen computer simulation examples demonstrated in Fig.<br />
1 to Fig. 3 show how parameter values reect on distribution <strong>of</strong> pharmacokinetic substance<br />
in treated patient organism. Psycho<strong>the</strong>rapy experts know well that rarely one can expect<br />
evident patient health improvement during very short time <strong>of</strong> treatment. Therefore, one can<br />
suggest considering <strong>the</strong> consequences <strong>of</strong> longer patient treatment periods. Some insight<br />
one can obtain by observing Fig. 1.<br />
Y (mg)<br />
Psycho<strong>the</strong>rapy pharmacokinetics<br />
10.0<br />
1.00 0.200<br />
9.5<br />
9.0<br />
8.5<br />
8.0<br />
7.5<br />
0.75 0.149<br />
7.0<br />
6.5<br />
6.0<br />
5.5<br />
5.0<br />
0.50 0.098<br />
4.5<br />
4.0<br />
3.5<br />
3.0<br />
2.5<br />
0.25 0.046<br />
2.0<br />
1.5<br />
1.0<br />
0.5<br />
0.0<br />
0.00 -0.005<br />
0 10 20 30 40 50<br />
Time (h)<br />
FVE_CALC vs T<br />
FIT_CALC vs T<br />
FFC_CALC vs T<br />
FSNG_CALC vs T<br />
FCR_CALC vs T<br />
FNU_CALC vs T<br />
FHP_CALC vs T<br />
FT_CALC vs T<br />
FTR_CALC vs T<br />
FMEN_CALC vs T<br />
FPL_CALC vs T<br />
MENP_CALC vs T<br />
MENN_CALC vs T<br />
BV_CALC vs T<br />
BNV_CALC vs T<br />
SRV_CALC vs T<br />
UT_CALC vs T<br />
PS_CALC vs T<br />
Bv,Bnv,Srv,Menp,Menn (nor.units)<br />
Ut,Ps (units)<br />
Y (mg)<br />
10.0<br />
9.5<br />
9.0<br />
8.5<br />
8.0<br />
7.5<br />
7.0<br />
6.5<br />
6.0<br />
5.5<br />
5.0<br />
4.5<br />
4.0<br />
3.5<br />
3.0<br />
2.5<br />
2.0<br />
1.5<br />
1.0<br />
0.5<br />
0.0<br />
Psycho<strong>the</strong>rapy pharmacokinetics<br />
0 10 20 30 40 50<br />
Time (h)<br />
1.00<br />
0.65<br />
0.30<br />
-0.05<br />
-0.40<br />
Bv,Bnv,Srv,Menp,Menn (nor.units)<br />
0.200<br />
0.138<br />
0.075<br />
0.012<br />
-0.050<br />
Ut,Ps (units)<br />
FVE_CALC vs T<br />
FIT_CALC vs T<br />
FFC_CALC vs T<br />
FSNG_CALC vs T<br />
FCR_CALC vs T<br />
FNU_CALC vs T<br />
FHP_CALC vs T<br />
FT_CALC vs T<br />
FTR_CALC vs T<br />
FMEN_CALC vs T<br />
FPL_CALC vs T<br />
MENP_CALC vs T<br />
MENN_CALC vs T<br />
BV_CALC vs T<br />
BNV_CALC vs T<br />
SRV_CALC vs T<br />
UT_CALC vs T<br />
PS_CALC vs T<br />
a) b)<br />
Fig. 1a,b<br />
Simulated psycho<strong>the</strong>rapy course for applied same computer simulation parameters during<br />
two consecutive cycles: k 1<br />
=0.3; k 2<br />
=0.01; k 3<br />
=0.8; k 4<br />
=0.1; k 5<br />
=0.5; k 6<br />
=0.02; k 7<br />
=0.05; k 8<br />
=0.08;<br />
k 9<br />
=0.1; k a<br />
=0.1; k b<br />
=0.05; k c<br />
=0.03; k d<br />
=0.1; k e<br />
=0.05; K k<br />
=0.1; k i<br />
=0.15; k u<br />
=0.02; k v<br />
=0.01;<br />
M pm<br />
=1.0; M nm<br />
=1.0; k s<br />
=0.005; k r<br />
=0.1; k rv<br />
=0.5; w=0.002; x=0.002; y=0.001; z=0.002; f =0.65.<br />
Initial value for F ve<br />
was <strong>the</strong> same for both a) and b) cycles; o<strong>the</strong>r initial independent variable<br />
quantities <strong>of</strong> <strong>the</strong> 2 nd cycle (b) were those at <strong>the</strong> 50 th hour <strong>of</strong> <strong>the</strong> simulated rst (a) cycle;<br />
As shown, some relative improvement one can expect if chosen substance can induce<br />
or cause an improvement and if it cannot lead to negative consequences <strong>of</strong> applied<br />
chemo<strong>the</strong>rapeutic treatment. Results suggest <strong>the</strong> giving <strong>of</strong> more importance to <strong>the</strong> role <strong>of</strong><br />
experts whenever <strong>the</strong> experts can inuence <strong>the</strong> patients by inducing positive consequences<br />
at <strong>the</strong>m. Data presented in Table 1 can explain <strong>the</strong> reasons why one can recommend <strong>the</strong><br />
estimations and/or assaying <strong>of</strong> chemo<strong>the</strong>rapeutic quantities in particular patient organism
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 145<br />
compartments. Fig. 1a, 1b demonstrate that equation /12a/ can be applied in explaining<br />
negative consequences <strong>of</strong> increasing <strong>the</strong> amounts <strong>of</strong> pharmacokinetic substances in<br />
cerebral-neural tissues. Evidently, data in Fig. 1a, 1b support <strong>the</strong> reasons why one should<br />
recommend <strong>the</strong> assaying and/or estimating <strong>of</strong> quantity distribution <strong>of</strong> chemo<strong>the</strong>rapeutic<br />
agents in particular compartments <strong>of</strong> human organism body. The organism dependence<br />
on applied drugs can be explained if applying <strong>the</strong> transformed ma<strong>the</strong>matical models. One<br />
example <strong>of</strong> computer simulation is demonstrated in Fig. 2a, 2b.<br />
Y (mg)<br />
Psycho- <strong>the</strong>rapy pharmacokinetics<br />
6.0<br />
1.00 0.050<br />
5.7<br />
5.4<br />
5.1<br />
4.8<br />
4.5<br />
0.75 0.038<br />
4.2<br />
3.9<br />
3.6<br />
3.3<br />
3.0<br />
0.50 0.025<br />
2.7<br />
2.4<br />
2.1<br />
1.8<br />
1.5<br />
0.25 0.013<br />
1.2<br />
0.9<br />
0.6<br />
0.3<br />
0.0<br />
0.00 0.000<br />
0.0 4.8 9.6 14.4 19.2 24.0<br />
FIT_CALC vs T<br />
Time (h)<br />
FF_CALC vs T<br />
FSG_CALC vs T<br />
FC_CALC vs T<br />
FTSG_CALC vs T<br />
FNU_CALC vs T<br />
FT_CALC vs T<br />
FM_CALC vs T<br />
FR_CALC vs T<br />
MEP_CALC vs T<br />
MEN_CALC vs T<br />
BV_CALC vs T<br />
BNV_CALC vs T<br />
SRV_CALC vs T<br />
U_CALC vs T<br />
PS_CALC vs T<br />
DP_CALC vs T<br />
FMI_CALC vs T<br />
FTT_CALC vs T<br />
FMT_CALC vs T<br />
FUT_CALC vs T<br />
Mep,Men,Bv,Srv (nor.units)<br />
Y3-U,Ps,Dp (arb.units)<br />
Y (mg)<br />
6.0<br />
5.7<br />
5.4<br />
5.1<br />
4.8<br />
4.5<br />
4.2<br />
3.9<br />
3.6<br />
3.3<br />
3.0<br />
2.7<br />
2.4<br />
2.1<br />
1.8<br />
1.5<br />
1.2<br />
0.9<br />
0.6<br />
0.3<br />
0.0<br />
Psycho- <strong>the</strong>rapy pharmacokinetics<br />
0.0 4.8 9.6 14.4 19.2 24.0<br />
Time (h)<br />
1.00<br />
0.75<br />
0.50<br />
0.25<br />
0.00<br />
Mep,Men,Bv,Srv (nor.units)<br />
0.100<br />
0.075<br />
0.050<br />
0.025<br />
0.000<br />
Y3-U,Ps,Dp (arb.units)<br />
FIT_CALC vs T<br />
FF_CALC vs T<br />
FSG_CALC vs T<br />
FC_CALC vs T<br />
FTSG_CALC vs T<br />
FNU_CALC vs T<br />
FT_CALC vs T<br />
FM_CALC vs T<br />
FR_CALC vs T<br />
MEP_CALC vs T<br />
MEN_CALC vs T<br />
BV_CALC vs T<br />
BNV_CALC vs T<br />
SRV_CALC vs T<br />
U_CALC vs T<br />
PS_CALC vs T<br />
DP_CALC vs T<br />
FMI_CALC vs T<br />
FTT_CALC vs T<br />
FMT_CALC vs T<br />
FUT_CALC vs T<br />
a) b)<br />
Fig. 2a,b:<br />
Simulation <strong>of</strong> active degradation product formation by applied drug transformation.<br />
Computer simulation based on transformed extended ma<strong>the</strong>matical model. Applied values<br />
<strong>of</strong> computer simulation parameters: k 1<br />
=0.8; k 2<br />
=0.01; k 4<br />
=0.1; k 5<br />
=0.5; k 6<br />
=0.02; k 9<br />
=0.1;<br />
K k<br />
=0.1; k a<br />
=0.1; b=0.01; c=0.03; d=0.0001; k i<br />
=0.15; k u<br />
=0.02; f=0.5; k v<br />
=0.01; g=0.0001;<br />
h=0.1; M pm<br />
=1.0; M nm<br />
=1.0; k s<br />
=0.005; k r<br />
=0.1; k rv<br />
=0.5; w=0.002; x=0.002; y=0.001; z=0.002;<br />
New parameter j=0.12 was applied. Initial drug quantity for both a) and b) was <strong>the</strong> same;<br />
<strong>the</strong> o<strong>the</strong>r quantities applied for b) were those nal for a);<br />
In practice, patients are frequently classied as belonging to different groups or types,<br />
based on <strong>the</strong> criteria <strong>of</strong> debatable reliability and <strong>the</strong>refore sometimes wrongly or even with<br />
danger consequences. Here, perhaps is reasonable to mention <strong>the</strong> Latina sentence: Quot<br />
homines tot sententiae, which refers to all people in general and <strong>the</strong>refore it certainly can<br />
be applicable to patients as well. Therefore, it follows that <strong>the</strong> best patient classication<br />
criterion seems to be by considering every patient as individual asking specic treatment<br />
and strong control <strong>of</strong> drug quantity distribution during <strong>the</strong>rapy. Data in Tables 2 – 3 and<br />
Figs. 3a,b show one <strong>of</strong> possible ways for <strong>the</strong>rapy improvements or even optimisation.
146<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
As demonstrated, slowed resorption leads to benecial consequences. Therefore, <strong>the</strong>re<br />
is reason for production <strong>of</strong> pharmaceutical products <strong>of</strong> properties which correspond to<br />
desired active substance resorption rate.<br />
Table 2<br />
Effects <strong>of</strong> active substance resorption rate (value <strong>of</strong> resorption rate constant k 1<br />
) on values<br />
<strong>of</strong> pharmacokinetics parameters at 10th hour after drug application.<br />
Parameter name Initial Value at 10th hour for given k 1<br />
value<br />
value<br />
k 1<br />
=0.8/h k 1<br />
=0.08/h<br />
FIT 100.0 0.030354 40.657<br />
FF 0.0 1.2342 6.5937<br />
FSG 0.0 16.699* 15.413**<br />
FC 0.0 3.8810 2.8076<br />
FTSG 0.0 17.915 8.5609<br />
FNU 0.0 6.2270 2.5503<br />
FT 0.0 18.144 8.6117<br />
FM 0.0 14.235 7.2429<br />
FR 0.0 2.1823 0.82074<br />
MEP 0.0 -5.8849 -1.1067<br />
MEN 0.0 0.59911 0.28760<br />
BV 0.99 0.33840 0.57298<br />
BNV 0.01 0.66160 0.42702<br />
SRV 1.0 0.88626 0.93331<br />
U 0.0 -0.058650 -0.005600<br />
PS 0.0 -0.022766 0.014274<br />
DP 0.0 0.0011259 0.0016591<br />
FMI 0.0 0.0048652 0.0012169<br />
FTT 0.0 9.0919 3.1410<br />
FMT 0.0 7.7490 2.75743<br />
FUT 0.0 2.6062 0.84128<br />
*<br />
**<br />
maximum attained during <strong>the</strong> 2nd<br />
hour,<br />
maximum attained during <strong>the</strong> 8th hour<br />
Table 3<br />
Effect <strong>of</strong> active substance repeated per-oral additions on pharmacokinetics quantities.<br />
Application <strong>of</strong> drug form adapted for active substance resorption rate dened with k 1<br />
=<br />
0.08/h.
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 147<br />
Quantity<br />
name<br />
Initial<br />
value<br />
Simulated value<br />
after 8 th hour<br />
Simulated value<br />
after 25 mg active<br />
substance addition<br />
at 8 th hour and after<br />
16 hours<br />
Simulated values after<br />
50 mg active substance<br />
additions at 8 th and at 16 th<br />
hour<br />
Value after<br />
16 th hour<br />
Value after<br />
24 th hour<br />
FIT 100.0 48.675 35.861 48.030 47.716<br />
FF 0.0 5.7028 9.9043 11.330 16.920<br />
FSG 0.0 15.600 17.548 21.448 25.419<br />
FC 0.0 2.6998 3.3293 4.0042 4.8585<br />
FTSG 0.0 7.0583 13.705 15.469 24.329<br />
FNU 0.0 1.9287 4.7860 5.2682 9.3740<br />
FT 0.0 7.0789 13.049 14.819 20.868<br />
FM 0.0 6.1546 10.210 11.749 15.583<br />
FR 0.0 0.58643 1.5909 1.7375 3.0097<br />
MEP 0.0 -0.54687 -3.2256 -3.7939 -6.4175<br />
MEN 0.0 0.19988 0.51099 0.53363 0.69866<br />
BV 0.99 0.63995 0.42220 0.38628 0.27897<br />
BNV 0.01 0.36005 0.57780 0.61372 0.72103<br />
SRV 1.0 0.94798 0.90219 0.89739 0.87648<br />
U 0.0 -0.0018992 -0.034447 -0.037620 -0.12630<br />
PS 0.0 0.013684 0.0042820 0.0022673 -0.039698<br />
DP 0.0 0.0014112 0.0013068 0.0012736 0.00039878<br />
FMI 0.0 0.00058767 0.0038032 0.0040628 0.0065879<br />
FTT 0.0 2.0975 7.0013 7.5256 14.736<br />
FMT 0.0 1.8886 5.7933 6.2655 11.627<br />
FUT 0.0 0.52811 2.2175 2.3495 5.5524
148<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
Y (mg)<br />
100<br />
95<br />
90<br />
85<br />
80<br />
75<br />
70<br />
65<br />
60<br />
55<br />
50<br />
45<br />
40<br />
35<br />
30<br />
25<br />
20<br />
15<br />
10<br />
5<br />
0<br />
Psycho- <strong>the</strong>rapy pharmacokinetics<br />
0 2 4 6 8 10<br />
Time (h)<br />
1.0<br />
0.8<br />
0.5<br />
0.3<br />
0.0<br />
Mep,Men,Bv,Srv (nor.units)<br />
0.100<br />
0.075<br />
0.050<br />
0.025<br />
0.000<br />
Y3-U,Ps,Dp (arb.units)<br />
FIT_CALC vs T<br />
FF_CALC vs T<br />
FSG_CALC vs T<br />
FC_CALC vs T<br />
FTSG_CALC vs T<br />
FNU_CALC vs T<br />
FT_CALC vs T<br />
FM_CALC vs T<br />
FR_CALC vs T<br />
MEP_CALC vs T<br />
MEN_CALC vs T<br />
BV_CALC vs T<br />
BNV_CALC vs T<br />
SRV_CALC vs T<br />
U_CALC vs T<br />
PS_CALC vs T<br />
DP_CALC vs T<br />
FMI_CALC vs T<br />
FTT_CALC vs T<br />
FMT_CALC vs T<br />
FUT_CALC vs T<br />
Y (mg)<br />
100<br />
95<br />
90<br />
85<br />
80<br />
75<br />
70<br />
65<br />
60<br />
55<br />
50<br />
45<br />
40<br />
35<br />
30<br />
25<br />
20<br />
15<br />
10<br />
5<br />
0<br />
Psycho- <strong>the</strong>rapy pharmacokinetics<br />
0 2 4 6 8 10<br />
Time (h)<br />
1.00<br />
0.75<br />
0.50<br />
0.25<br />
0.00<br />
Mep,Men,Bv,Srv (nor.units)<br />
0.100<br />
0.075<br />
0.050<br />
0.025<br />
0.000<br />
Y3-U,Ps,Dp (arb.units)<br />
FIT_CALC vs T<br />
FF_CALC vs T<br />
FSG_CALC vs T<br />
FC_CALC vs T<br />
FTSG_CALC vs T<br />
FNU_CALC vs T<br />
FT_CALC vs T<br />
FM_CALC vs T<br />
FR_CALC vs T<br />
MEP_CALC vs T<br />
MEN_CALC vs T<br />
BV_CALC vs T<br />
BNV_CALC vs T<br />
SRV_CALC vs T<br />
U_CALC vs T<br />
PS_CALC vs T<br />
DP_CALC vs T<br />
FMI_CALC vs T<br />
FTT_CALC vs T<br />
FMT_CALC vs T<br />
FUT_CALC vs T<br />
a) b)<br />
Fig. 3a,b.<br />
Effect <strong>of</strong> active substance resorption rate on pharmacokinetics quantities: a) k 1<br />
=0.8; b)<br />
k 1<br />
=0.08;<br />
Concerning a chemo<strong>the</strong>rapeutic substance distribution in organism body one should point<br />
out that in <strong>the</strong> proposed ma<strong>the</strong>matical model <strong>the</strong> chemo<strong>the</strong>rapeutic substance quantities<br />
represent chemo<strong>the</strong>rapeutic substance masses in particular compartments, not <strong>the</strong>ir<br />
mass concentrations in corresponding compartments. Such an approach to expressing<br />
chemo<strong>the</strong>rapeutic substance distribution can be considered as more convenient than that<br />
referring to chemo<strong>the</strong>rapeutic substance mass concentrations, since at given time <strong>the</strong><br />
corresponding substance mass concentration cannot be considered to be constant with<br />
reference to every compartment part.<br />
4.2 Speculative comments (conclusions)<br />
Enormous complexity <strong>of</strong> psycho<strong>the</strong>rapy asks serious comments. Some questions one can<br />
put here: Are we capable to recognise, in every case, which part <strong>of</strong> patient behaviour can be<br />
and should be changed, and how to proceed in order to realise desired positive changes?!<br />
Are we always sure that chosen chemo<strong>the</strong>rapeutic agent is adequate one?! Whe<strong>the</strong>r at<br />
all <strong>the</strong> application <strong>of</strong> any chemo<strong>the</strong>rapeutic agent can be recommended?! Is it reasonable<br />
a neglecting <strong>of</strong> prayers?! Religious and spiritual factors in general certainly should be<br />
taken into account with giving <strong>the</strong>m adequate relevance. Which degree <strong>of</strong> tranquilisation<br />
<strong>of</strong> particular patient could be considered as optimal one for his health?! How to enhance<br />
<strong>the</strong> endogenous production <strong>of</strong> tranquilisation factors (endorphin, e.g.) in improving <strong>the</strong>
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 149<br />
dreaming and sleeping in order to reduce <strong>the</strong> need for an application <strong>of</strong> hypnotics?!<br />
Why not to consider that particular effect <strong>of</strong> any factor can be quantitatively evaluated,<br />
at least approximately?! Is it possible to dene <strong>the</strong> position <strong>of</strong> psychiatrists, since <strong>the</strong><br />
most frequent situation is that <strong>the</strong>y are simultaneously engaged in treating relatively<br />
large number <strong>of</strong> patients?! One <strong>of</strong> answers on all mentioned questions certainly should<br />
be that exact information on chemo<strong>the</strong>rapeutic substance distribution in organism body<br />
is <strong>of</strong> essential importance for achieving <strong>of</strong> positive results <strong>of</strong> psycho<strong>the</strong>rapeutic treatment<br />
which includes <strong>the</strong> application <strong>of</strong> chemo<strong>the</strong>rapeutic substances. Therefore, methods <strong>of</strong><br />
assaying and estimating <strong>of</strong> chemo<strong>the</strong>rapeutic substance contents in particular human<br />
organism compartments should be available for application whenever one estimates that<br />
such information could be helpful.<br />
Application <strong>of</strong> chemo<strong>the</strong>rapeutic substances in general can induce or cause different<br />
undesired side effects. These can refer to particular organism body compartments or to <strong>the</strong><br />
whole organism. Live organism body can be induced, and frequently is capable to degrade<br />
and/or transform <strong>the</strong> used substance. As a result <strong>of</strong> organism body answer could be <strong>the</strong><br />
overproduction <strong>of</strong> some products asking <strong>the</strong>rapeutic action as well. Therefore, one suggests<br />
a serious consideration <strong>of</strong> all <strong>the</strong>se effects and possible <strong>the</strong>ir description. The extended<br />
ma<strong>the</strong>matical model can enable a more exact expression <strong>of</strong> particular side effect causes<br />
and consequences. The approach similar to that applied for cerebral-neural compartment<br />
can be also applied for any considered compartment. Then, particular positive and/or<br />
undesired negative effects could be evaluated with reference to <strong>the</strong> whole organism body<br />
or to particular its organs or tissues.<br />
As shown recently 2A by expressing roughly <strong>the</strong> decisive role <strong>of</strong> cardio-vascular and<br />
respiratory compartments in affecting <strong>the</strong> quantity <strong>of</strong> “revitalisation substrate”, Srv, in this<br />
work one can focus onto evaluation <strong>of</strong> consequences referring to vital functions <strong>of</strong> liver, or<br />
<strong>of</strong> nephritic-urinary and o<strong>the</strong>r organism body compartments.<br />
Concerning nephritic-urinary system it should be pointed out that <strong>the</strong> total amounts <strong>of</strong><br />
pharmacokinetic agent transferred to it differ from those present in it at given time, because<br />
<strong>of</strong> periodic withdrawals <strong>of</strong> urine from nephritic-urinary system. Therefore, <strong>the</strong> amounts<br />
<strong>of</strong> pharmacokinetic substances actually present in <strong>the</strong> nephritic-urinary system represent<br />
<strong>the</strong> difference between <strong>the</strong> total amounts transferred into <strong>the</strong> system and <strong>the</strong> total amounts<br />
withdrawn from <strong>the</strong> system. Similar consideration as for nephritic-urinary system can also<br />
be applied concerning <strong>the</strong> pharmacokinetic substance quantity present in <strong>the</strong> faeces <strong>of</strong> large<br />
intestine.<br />
The advantages <strong>of</strong> developed ma<strong>the</strong>matical model application can become especially<br />
evident if one tends to estimate consequences <strong>of</strong> simultaneous application <strong>of</strong> different<br />
drugs in <strong>the</strong> <strong>the</strong>rapy. Based on <strong>the</strong> kinetic relationships established for particular substance<br />
one can predict <strong>the</strong> consequences <strong>of</strong> simultaneous applications <strong>of</strong> two or more different<br />
substances during <strong>the</strong> planned psycho<strong>the</strong>rapy treatments, if applying computer simulations.<br />
These could be performed with reference to every patient. It is important to point out that<br />
our possibilities to represent <strong>the</strong> events in organism body by ma<strong>the</strong>matical models are not<br />
limited only on models presented here. Our knowledge referring to o<strong>the</strong>r events can also be<br />
translated into corresponding ma<strong>the</strong>matical models, and <strong>the</strong>refore in those convenient for<br />
<strong>the</strong> use in performing desired computer simulations. By <strong>the</strong> way, it is important to point out<br />
that presented ma<strong>the</strong>matical model can simply be transformed into much simpler models<br />
by neglecting relevance <strong>of</strong> those events which should not be obligatory considered for<br />
evaluation <strong>of</strong> consequences <strong>of</strong> more relevant events.<br />
Finally, <strong>the</strong>re are questions which should not be avoided. One <strong>of</strong> <strong>the</strong>m is how to estimate<br />
normalised values <strong>of</strong> quantities <strong>of</strong> B v<br />
, B nv<br />
, S rv<br />
, Menp, Menn, etc. No doubt, a reliability
150<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
<strong>of</strong> estimations will markedly depend on <strong>the</strong> experience <strong>of</strong> psycho<strong>the</strong>rapy experts and on<br />
<strong>the</strong> availability <strong>of</strong> acceptable evaluation methods. However, <strong>the</strong> experts certainly can and<br />
know dene, at least approximately, <strong>the</strong> maximal specic quantity values, and compare<br />
<strong>the</strong>m with values estimated during <strong>the</strong> patient observation and treatments. Therefore,<br />
<strong>the</strong>re is a certain possibility for at least slight advances in evaluating <strong>the</strong> efciency <strong>of</strong><br />
psycho<strong>the</strong>rapy treatments. Of course, evaluations should be performed in <strong>the</strong> right time.<br />
One cannot recommend e.g. a postponing <strong>of</strong> B nv<br />
evaluations till a situation when <strong>the</strong><br />
pathologist opinion would become <strong>the</strong> most relevant. On <strong>the</strong> o<strong>the</strong>r hand, we should always<br />
take into account <strong>the</strong> fact that <strong>the</strong> frequency <strong>of</strong> successful <strong>the</strong>rapy cases shows positive<br />
trend. Moreover, relevant pharmaceutical companies (e.g. LILLY USA, LLC. 11) give<br />
enough data for optimal approach to <strong>the</strong> use <strong>of</strong> psycho<strong>the</strong>rapeutic agents. Therefore, <strong>the</strong>re<br />
is no doubt for a need <strong>of</strong> controlling drug distribution in human body during <strong>the</strong>rapy.<br />
Already long time introduced practice <strong>of</strong> determination <strong>of</strong> alcohol concentrations in blood<br />
samples <strong>of</strong> persons causing trafc accidents justify <strong>the</strong> giving adequate relevance to <strong>the</strong><br />
control <strong>of</strong> psycho<strong>the</strong>rapeutic drugs distribution in <strong>the</strong> body <strong>of</strong> every patient subjected to<br />
psycho<strong>the</strong>rapy which includes <strong>the</strong> drug application.<br />
Symbols<br />
a kinetic constant, h -1 (T -1 )<br />
B v<br />
body vitality, dimensionless<br />
B nv<br />
body non-vitality, dimensionless<br />
b kinetic constant, dimensionless<br />
c kinetic constant, h -1 (T -1 )<br />
Dp dependence product normalised mass, dimensionless<br />
d kinetic constant, h -1 (T -1 )<br />
F cr<br />
pharmacokinetic substance mass in hearth, mg (M)<br />
F fc<br />
pharmacokinetic substance mass in faeces, mg (M)<br />
F hp<br />
pharmacokinetic substance mass in liver, mg (M)<br />
F men<br />
pharmacokinetic substance mass in neural-brain system, mg (M)<br />
F mni<br />
inhibited F men<br />
, mg (M)<br />
F mt<br />
transformed pharmacokinetic substance mass in neural-brain system, mg (M)<br />
F nu<br />
pharmacokinetic substance mass in nephritic–urinary system, mg (M)<br />
F nut<br />
transformed pharmacokinetic substance mass in nephritic–urinary system, mg (M)<br />
F rsp<br />
, F plm<br />
pharmacokinetic substance mass in respiratory system, mg<br />
(M)<br />
F sng<br />
pharmacokinetic substance mass in blood, mg (M)<br />
F tsng<br />
transformed pharmacokinetic substance mass in blood, mg (M)<br />
F t<br />
pharmacokinetic substance mass in tissue, mg (M)<br />
F tt<br />
transformed pharmacokinetic substance mass in tissue, mg (M)<br />
F tr<br />
pharmacokinetic substance mass in tissue receptor, mg (M)<br />
F ve,<br />
F it<br />
pharmacokinetic substance mass in ventricular (F ve<br />
)-intestinal (F it<br />
) compartment,<br />
mg<br />
(M)<br />
g kinetic constant, h -1 (T -1 )<br />
h kinetic constant, h -1 (T -1 )<br />
j kinetic constant, h -1 (T -1 )<br />
k 1<br />
kinetic constant with reference to F ve<br />
, h -1 (T -1 )<br />
k 2<br />
kinetic constant with reference to F it<br />
, h -1 (T -1 )<br />
k 3<br />
kinetic constant with reference to Fit, h -1 (T -1 )<br />
k 4<br />
kinetic constant with reference to F sng<br />
, h -1 (T -1 )<br />
k 5<br />
kinetic constant with reference to F cr<br />
, h -1 (T -1 )
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 151<br />
k 6<br />
kinetic constant with reference to F sng<br />
, h -1 (T -1 )<br />
k 7<br />
kinetic constant with reference to F sng<br />
, h -1 (T -1 )<br />
k 8<br />
kinetic constant with reference to F hp<br />
, h -1 (T -1 )<br />
k 9<br />
kinetic constant with reference to F sng<br />
, , h -1 (T -1 )<br />
k a<br />
kinetic constant with reference to Ft, h -1 (T -1 )<br />
k b<br />
kinetic constant with reference to F cr<br />
, h -1 (T -1 )<br />
k c<br />
kinetic constant with reference to F rsp<br />
, h -1 (T -1 )<br />
k d<br />
kinetic constant with reference to Ft, h -1 (T -1 )<br />
ke kinetic constant with reference to F tr,<br />
h -1 (T -1 )<br />
ki kinetic constant with reference to F men<br />
,h -1 (T -1 )<br />
k k<br />
kinetic constant with reference to F sng,<br />
,h -1 (T -1 )<br />
k r<br />
kinetic constant with reference to B nv<br />
and Srv, h -1 (T -1 )<br />
k rv<br />
kinetic constant with reference to Srv, h -1 (T -1 )<br />
k s<br />
kinetic constant with reference to Bv and F sng<br />
, mg -1 h -1 (M -1 T -1 )<br />
k u<br />
kinetic constant with reference to F men<br />
and Menp, mg -1 h -1 (M -1 T -1 )<br />
k v<br />
kinetic constant with reference to F men<br />
and Menn, mg -1 h -1 (M -1 T -1 )<br />
Menn normalised quantity <strong>of</strong> negative effects caused by F men<br />
, dimensionless<br />
Menp normalised quantity <strong>of</strong> positive effects caused by F men<br />
, dimensionless<br />
Mnm maximal value <strong>of</strong> Menn, dimensionless<br />
Mpm maximal value <strong>of</strong> Menp, dimensionless<br />
Ps normalised effect <strong>of</strong> psychiatrist and personnel, dimensionless<br />
S rv<br />
normalised quantity <strong>of</strong> revitalisation substrate, dimensionless<br />
Ut normalised psycho<strong>the</strong>rapy efciency, dimensionless<br />
w kinetic constant with reference to Ps, h -1 (T -1 )<br />
x kinetic constant with reference to Menp, h -1 (T -1 )<br />
y kinetic constant with reference to Menn, h -1 (T -1 )<br />
z kinetic constant with reference to Ps, h -1 (T -1 )<br />
References<br />
1. Bošnjak M. (1982) Ma<strong>the</strong>matical modelling <strong>of</strong> biochemical reaction systems (in<br />
Croatian) Kem. Ind. 31 545 -559<br />
2. Bošnjak M. (2009) Introduction to Kinetics <strong>of</strong> Microbial Processes ( in Croatian:<br />
Uvod u kinetiku mikrobnih procesa), Graphis, Zagreb, (2A= pp. 275-306, Appendix)<br />
3. Bošnjak M., Topolovec V., Vrana M. (1978), Growth kinetics <strong>of</strong> Streptomyces<br />
erythreus during erythromycin biosyn<strong>the</strong>sis, J. Appl. Chem. Biotechnol. 28 791- 798<br />
4. Bošnjak M., Topolovec V., Johanides V. (1979), Growth kinetics and antibiotic<br />
syn<strong>the</strong>sis during <strong>the</strong> repeated fed batch culture <strong>of</strong> streptomycetes, In: Armiger W. B.<br />
(ed.), Computer Applications in Fermentation Technology, Biotechnol. Bioeng. Symp.<br />
No. 9 155 -165, Wiley & Sons<br />
5. Bošnjak M., Bago Joksovi A., Pigac J., Bošnjak Cihlar Ž., Hranueli D., (2006)<br />
Applicability <strong>of</strong> ma<strong>the</strong>matical models in dening <strong>the</strong> behaviour kinetics distinction<br />
among microbial strains, Chem. Biochem. Eng. Q. 20 375 -388<br />
6. Bošnjak M., Udikovi Koli N., Petri I., Cihlar D., Hršak D., (<strong>2010</strong>) Integrated<br />
approach to ma<strong>the</strong>matical modelling <strong>of</strong> atrazine degradation in different reaction<br />
systems, Food Technol. Biotechnol.48 392-403<br />
7. Ci<strong>of</strong> Ch. L., Liu Sh., Wolf M. A., (<strong>2010</strong>) Recent development in glycine transporter-1<br />
inhibitors, In: J. E. Macor (Ed.), Annual Reports in Medicinal Chemistry,Vol.45, pp.<br />
19-35, Academic Press, Burlington, MA 01803, USA<br />
8. Karba R., Mrhar A., Bremšak F., Kozjak F., (1977) Estimation <strong>of</strong> pharmacokinetics
152<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
parameters <strong>of</strong> amoxicillin and ampicilline by identication (in Slovenian: Doloevanje<br />
farmakokinetinih parametrov amoksicilina in ampicilina s pomojo identikacije),<br />
In: Informatica 77, Proc. Yug. Intern. Symp. 6 218, Slov. Soc. Informatika, Ljubljana<br />
9. van der Kleijin E., Guelen P. J. M., van Wijk C., Baars I., (1975) Clinical<br />
pharmacokinetics in monitoring chronic medication with anti-epileptic drugs, In:<br />
Schneider H., Janz D., Gardner-Thorpe C., Meinardi H., Shervin A. L. (eds.), Clinical<br />
Pharmacology <strong>of</strong> Anti-Epileptic Drugs, Springer, Berlin- Heidelberg –New York, pp.<br />
11 -13<br />
10. Kossen N. W., (1979) Ma<strong>the</strong>matical modelling <strong>of</strong> fermentation processes: Scope and<br />
limitation, In: Bull A. T., Elwood D. C., Ratledge C. (eds.) Microbial Technology:<br />
Current State, Future Prospects, 29 th Symp. Soc. Gen. Microb., Cambridge, University<br />
Press, Cambridge, , pp. 327 -35<br />
11. LILLY USA, LLC Internet sites.<br />
12. Li R. C., Nix D. E., Schentag J. J., (2006) Pharmacodynamic modelling <strong>of</strong> bacterial<br />
kinetics: -Lactam antibiotics against Escherichia coli, J. Pharm. Sciences 83 970<br />
-975<br />
13. Nikolaou M., Schilling A. N., Vo Giao, Chang Kai-tai, Tam V. H., (2007) Modeling<br />
<strong>of</strong> microbial population responses to time-periodic concentrations <strong>of</strong> antimicrobial<br />
agents, Anuals Biomed. Eng. 35 1458 -1470<br />
14. Ramkrishna D., (1979) Statistical models <strong>of</strong> cell populations, In: Ghose T. K., Fiechter<br />
A., Blakebrough N. (eds.), Advances in Biochemical Engineering, Vol. 11, Springer,<br />
Berlin-Heidelberg-New York, pp. 1 -47<br />
15. Roepke H. und Riemann J.,(1979) Analog Computer in Chemie und Biologie, Springer,<br />
Berlin, 1969<br />
Appendix<br />
Spirituality based <strong>the</strong>rapy and its possible relevance with reference to behaviour disorder<br />
treatments<br />
There are cases <strong>of</strong> health improvements observed after patients being subjected to<br />
spirituality based <strong>the</strong>rapy. Therefore, its relevance should not be neglected regardless<br />
whe<strong>the</strong>r somebody is religious or not. However, <strong>the</strong> evaluation <strong>of</strong> effects based on such a<br />
<strong>the</strong>rapy probably asks <strong>the</strong> specic approach and explanations. Certainly <strong>the</strong>re is no need to<br />
explain why <strong>the</strong> health <strong>of</strong> those religious was improved, while <strong>the</strong> cases <strong>of</strong> those declared<br />
as not being religious ask some comments. In this work <strong>the</strong> effects <strong>of</strong> signals were not<br />
discussed. The source <strong>of</strong> signals can be <strong>of</strong> different characters, but possible explanation<br />
<strong>of</strong> <strong>the</strong>ir effects could be similar. One <strong>of</strong> explanation possibilities is to suppose that signals<br />
activate some biochemical mechanisms capable to repair <strong>the</strong> organism normal behaviour.<br />
Applied drug dose titration<br />
In psycho<strong>the</strong>rapy, depending on particular patient, one or more different drugs can be<br />
applied. In cases when applied drug reliably showed adequate positive effects with absence<br />
or with minimal negative side effects <strong>the</strong> estimation and maintenance <strong>of</strong> optimal drug<br />
dosing do not represent problem, if necessary information on drug properties is available.<br />
More complicate situation appears when <strong>the</strong> quite adequate drug for particular patient is<br />
not available, or when patient behaviour asks <strong>the</strong> searching <strong>of</strong> adequacy <strong>of</strong> known and<br />
available different drugs. Then it can happen that none <strong>of</strong> available known drugs is quite<br />
adequate and that <strong>the</strong> combination <strong>of</strong> different available drugs should be applied. If <strong>the</strong> most
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 153<br />
convenient <strong>of</strong> available drugs is chosen for starting treatment is applied on appropriate way,<br />
<strong>the</strong>n it is recommendable to monitor <strong>the</strong> consequences to decide how to continue, taking<br />
into account drug properties referring to drug half-life time and resorption rate as well as<br />
character and intensity <strong>of</strong> observed consequences. The observed side effects could suggest<br />
additional application <strong>of</strong> ano<strong>the</strong>r drug suitable to suppress observed negative effects. If<br />
such a drug can induce side effects reducing <strong>the</strong> positive effects <strong>of</strong> applied starting drug,<br />
<strong>the</strong>n careful titration become obligate. If this would not be done adequately, <strong>the</strong>n it can<br />
happen that due to overdosing instead <strong>of</strong> health improvement <strong>the</strong> health aggravation could<br />
result. Therefore, in addition to commonly applied trial and error methods in optimizing<br />
chemo<strong>the</strong>rapeutic treatment <strong>the</strong> computer simulation <strong>of</strong> applied drugs pharmacokinetics<br />
and supplying some necessary experimental data could be recommended in order to<br />
optimise <strong>the</strong>rapy.<br />
Information on <strong>the</strong> author:<br />
Pr<strong>of</strong>. Dr. sc. Marijan Bošnjak, retired; member emeritus <strong>of</strong> Croatian Academy <strong>of</strong><br />
Engineering (<strong>HATZ</strong>), HR-10000 Zagreb, Croatia;<br />
E-mail: marijan.bosnjak@hatz.hr<br />
Home address:<br />
HR-10000 Zagreb, Slovenska 19, Croatia<br />
Tel. (385)-1-3702507
154<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
DICES: Distributed Component-based Embedded<br />
S<strong>of</strong>tware Systems<br />
Mario Žagar<br />
Faculty <strong>of</strong> Electrical Engineering and Computing<br />
University <strong>of</strong> Zagreb, Zagreb, Croatia<br />
mario.zagar@fer.hr<br />
Ivica Crnković<br />
Mälardalen Real-Time Research Centre<br />
Mälerdalen University, Västerås, Sweden<br />
ivica.crnkovic@mdh.se<br />
Darko Stipaničev<br />
Faculty <strong>of</strong> Electrical Engineering, Mechanical Engineering and Naval Architecture<br />
University <strong>of</strong> Split, Split, Croatia<br />
darko.stipanicev@fesb.hr<br />
Maja Štula<br />
Faculty <strong>of</strong> Electrical Engineering, Mechanical Engineering and Naval Architecture<br />
University <strong>of</strong> Split, Split, Croatia<br />
maja.stula@fesb.hr<br />
Juraj Feljan<br />
Mälardalen Real-Time Research Centre<br />
Mälerdalen University, Västerås, Sweden<br />
juraj.feljan@mdh.se<br />
Luka Lednicki<br />
Faculty <strong>of</strong> Electrical Engineering and Computing<br />
University <strong>of</strong> Zagreb, Zagreb, Croatia<br />
luka.lednicki@fer.hr<br />
Josip Maras<br />
Faculty <strong>of</strong> Electrical Engineering, Mechanical Engineering and Naval Architecture<br />
University <strong>of</strong> Split, Split, Croatia<br />
josip.maras@fesb.hr<br />
Ana Petričić<br />
Faculty <strong>of</strong> Electrical Engineering and Computing<br />
University <strong>of</strong> Zagreb, Zagreb, Croatia<br />
ana.petricic@fer.hr<br />
Abstract<br />
This article gives a short overview <strong>of</strong> <strong>the</strong> contribution <strong>of</strong> DICES project. The goal <strong>of</strong> <strong>the</strong><br />
project is to advance <strong>the</strong>ories and technologies used in development <strong>of</strong> distributed embedded<br />
systems. Three examples <strong>of</strong> <strong>the</strong> contributions are presented: a) reverse <strong>engineering</strong> <strong>of</strong> webbased<br />
applications, design extraction and extraction <strong>of</strong> reusable user-interface controls,
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 155<br />
b) a fretwork for building systems that use UPnP devices and treat <strong>the</strong>m as components<br />
in <strong>the</strong> same way as s<strong>of</strong>tware components, and c) PRIDE – development environment for<br />
designing, modeling and developing embedded systems, based on ProCom technology.<br />
Keywords: S<strong>of</strong>tware components, embedded systems, web-based applications, reverse<br />
<strong>engineering</strong><br />
1 Introduction<br />
DICES (Distributed Component-based Embedded S<strong>of</strong>tware Systems) is a project<br />
funded by UKF (Unity Through Knowledge Fund) with contributions from Faculty <strong>of</strong><br />
Electrical Engineering and Computing (FER) - University <strong>of</strong> Zagreb, Faculty <strong>of</strong> Electrical<br />
Engineering, Mechanical Engineering and Naval Architecture (FESB), University <strong>of</strong> Split,<br />
School <strong>of</strong> Innovation, Design and Engineering - Mälardalen University (MDU) in Sweden,<br />
and Ericsson Nikola Tesla. The project was performed during three years, from 2007 to<br />
<strong>2011</strong>.<br />
The DICES project goal is to advance development <strong>of</strong> distributed embedded s<strong>of</strong>tware<br />
systems with emphasis on s<strong>of</strong>tware reusability and predictability <strong>of</strong> s<strong>of</strong>tware quality.<br />
The aim <strong>of</strong> <strong>the</strong> project is increasing <strong>the</strong> s<strong>of</strong>tware development efciency and quality by<br />
applying service-oriented and component-based approaches.<br />
The overall presence <strong>of</strong> distributed embedded systems in <strong>the</strong> modern society is a fact.<br />
Examples <strong>of</strong> such systems are telecommunication systems, grid systems, control and<br />
information systems <strong>of</strong> vehicular systems (cars, trains), environmental monitoring<br />
systems. Embedded systems development is one <strong>of</strong> <strong>the</strong> strategic research areas <strong>of</strong> EU-FP7<br />
programmes. It is also <strong>of</strong> signicant importance in Croatia, since many leading companies<br />
in Croatia ei<strong>the</strong>r produce such systems (e.g. Konar, Ericsson Nikola Tesla) or use such<br />
systems (e.g. Pliva, or many small companies).<br />
DICES is addressing efcient reusability <strong>of</strong> s<strong>of</strong>tware components and prediction <strong>of</strong> <strong>the</strong><br />
important properties for embedded systems: resource utilization, and performance,<br />
by applying <strong>the</strong> service-oriented s<strong>of</strong>tware <strong>engineering</strong> and component-based s<strong>of</strong>tware<br />
<strong>engineering</strong> methods and technologies.<br />
The main contribution <strong>of</strong> DICES is provided in two directions: a) analysis <strong>of</strong> legacy webbased<br />
systems, extraction <strong>of</strong> its design and its automatic modeling, and extraction <strong>of</strong> user<br />
interface controls and packaging <strong>the</strong>m as reusable units, and development <strong>of</strong> componentbased<br />
technology appropriate for design <strong>of</strong> embedded systems.<br />
The reverse <strong>engineering</strong>, architecture recovery and extraction <strong>of</strong> reusable UI units is applied<br />
on “iForestFire - Intelligent Forest Fire System” system developed at FESB Split, which<br />
enables thorough validation <strong>of</strong> <strong>the</strong> approach and provides input for fur<strong>the</strong>r development <strong>of</strong><br />
this system and possible commercialization <strong>of</strong> <strong>the</strong> improved product.<br />
Development <strong>of</strong> component-based technology is done in cooperation with Swedish Strategic<br />
Research Centre, PROGRESS that has developed <strong>the</strong>ories and methods for modeling<br />
component-based embedded systems. DICES contribution was in development <strong>of</strong> PRIDE,<br />
PROGRESS Integrated Development Environment, a tool modeling component-based<br />
embedded systems, simulation and analysis <strong>of</strong> resource utilization.<br />
In this article we give a short overview <strong>of</strong> <strong>the</strong>se contributions as follows. In section 2 we<br />
describe Componentizing Web Information Systems, including phpModeller, a tool for<br />
design extraction and its presentation in UML. Section 3 shows a component model that<br />
includes both, s<strong>of</strong>tware and hardware components based on UPnP devices and enables a<br />
unique treatment for s<strong>of</strong>tware and hardware components. Section 4 describes PRIDE.
156<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
2 Analysis and Componenitization <strong>of</strong> Web-based systems<br />
Web-based systems are systems that use <strong>the</strong> Internet as its core infrastructure and are<br />
accessible from anywhere in <strong>the</strong> world via <strong>the</strong> Web. Often <strong>the</strong>se systems extensively<br />
communicate with databases and third-party information sources. In some cases <strong>the</strong>y even<br />
communicate with embedded devices such as cameras or meteo-stations in order to provide<br />
real-time information about <strong>the</strong> environment. One <strong>of</strong> <strong>the</strong>se systems is <strong>the</strong> iForestFire system<br />
developed at <strong>the</strong> University <strong>of</strong> Split, Croatia. iForestFire is an intelligent and integrated<br />
video based monitoring system for early detection <strong>of</strong> forest re. The main idea is that<br />
forest res are detected in incipient stage using advanced image processing and image<br />
analyses methods. The system is based on eld units and central processing unit. The eld<br />
units are embedded devices and include day & night, pan/tilt/zoom controlled IP based<br />
video cameras and IP based mini meteorological stations connected by wired or wireless<br />
LAN to <strong>the</strong> central processing unit where all analysis, calculation, presentation, image<br />
and data archiving is done. In <strong>the</strong> scope <strong>of</strong> <strong>the</strong> DICES project, iForestFire was chosen as<br />
a case study application in order to test <strong>the</strong> ideas <strong>of</strong> componentization <strong>of</strong> Web Information<br />
Systems that extensively communicate with embedded devices.<br />
Since one <strong>of</strong> <strong>the</strong> goals <strong>of</strong> <strong>the</strong> DICES project was <strong>the</strong> analysis <strong>of</strong> legacy web-based systems,<br />
extraction <strong>of</strong> its design and its automatic modeling, extraction <strong>of</strong> User Interface controls<br />
and packaging <strong>the</strong>m as reusable units in this section, we will describe two approaches that<br />
we have developed in order to achieve those goals: phpModeler – an approach for reverse<br />
<strong>engineering</strong> <strong>of</strong> legacy web applications, and Firecrow, an approach for extracting reusable<br />
Web User Interface controls.<br />
2.1 PhpModeler – Reverse Engineering Legacy Web Applications<br />
In order to componentize web information systems, rst a correct assessment <strong>of</strong> <strong>the</strong> current<br />
state is necessary. Web information systems, especially ones that are communicating with<br />
complex embedded devices can be hard to understand because a single behavior can be<br />
realized by code executed on a large number <strong>of</strong> nodes (embedded devices, web servers,<br />
clients, etc.). Also, <strong>the</strong>re usually exist many inter-dependencies between certain parts <strong>of</strong><br />
<strong>the</strong> system that are hard to track and manage manually, and naturally <strong>the</strong> complexity <strong>of</strong><br />
<strong>the</strong>se inter-dependencies grows with <strong>the</strong> size <strong>of</strong> <strong>the</strong> web application. In order to tackle this<br />
problem and to improve development and maintenance efciency, <strong>the</strong>se applications need<br />
to be modeled on a high level <strong>of</strong> abstraction. One <strong>of</strong> <strong>the</strong> ways to achieve this goal is to use<br />
a process called Reverse Engineering (RE).<br />
Reverse Engineering (RE) is used for information extraction from source artifacts (primarily<br />
source code) and <strong>the</strong>ir transformation to easily understandable abstract representations (e.g.<br />
standard UML diagrams). Even though <strong>the</strong>re exists a certain number <strong>of</strong> already available<br />
tools for reverse <strong>engineering</strong> Web Information Systems (WARE [6], WebUml [3], ReWeb<br />
[10], etc.), none <strong>of</strong> <strong>the</strong>m have <strong>the</strong> following functionalities:<br />
• Static analysis <strong>of</strong> web application source code which as an end result produces UML<br />
diagrams that can be used for architecture recovery<br />
• Generates dependency models that show inter-dependence between web application<br />
elements<br />
• Visualizes Web application evolution.<br />
In order to tackle this problem we have developed a plugin for <strong>the</strong> Eclipse IDE (Integrated<br />
Development Environment) that facilitates modeling and web application architecture<br />
recovery called phpModeler. It has three main features: page modeling, dependency<br />
modeling and model comparison.
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 157<br />
In order to be able to build page UML models, <strong>the</strong> information important for those models<br />
has to be extracted. phpModeler ga<strong>the</strong>rs this information by parsing PHP, HTML, and<br />
JavaScript code responsible for <strong>the</strong> behavior <strong>of</strong> <strong>the</strong> web application. Once this information<br />
has been collected fur<strong>the</strong>r analysis techniques can be used to establish dependencies<br />
between parts <strong>of</strong> <strong>the</strong> system. For every web page entity (JavaScript library, database<br />
table, le, PHP library, web page), <strong>the</strong> analysis nds all o<strong>the</strong>r entities dependent on it.<br />
For example, for every database table this means locating all web pages that access <strong>the</strong>m<br />
and for function libraries all web pages that use those functions. Naturally, all information<br />
ga<strong>the</strong>red in <strong>the</strong> analysis process can be generated to UML diagrams.<br />
Figure 1: Reverse Engineering Web applications with phpModeler.<br />
Web information systems evolve with <strong>the</strong> addition <strong>of</strong> new functionalities. This usually<br />
means that <strong>the</strong> existing code is modied to support <strong>the</strong> new user requirements. At different<br />
steps <strong>of</strong> <strong>the</strong> WIS evolution <strong>the</strong> system is represented with different models, so phpModeler<br />
provides a way to track <strong>the</strong> differences between two existing models. This gives a simpler,<br />
and more visual representation <strong>of</strong> <strong>the</strong> evolution <strong>of</strong> <strong>the</strong> target system. The whole process is<br />
shown in Figure 1.<br />
We used phpModeler in <strong>the</strong> reverse <strong>engineering</strong> process <strong>of</strong> <strong>the</strong> iForestFire system in order<br />
to recover <strong>the</strong> architecture and gain better understanding <strong>of</strong> <strong>the</strong> system.<br />
2.2 Firecrow – reusing Web User Interface controls<br />
Important part <strong>of</strong> <strong>the</strong> web application code is <strong>the</strong> code that is used to realize <strong>the</strong> userinterface<br />
<strong>of</strong> <strong>the</strong> web application. Web application user-interface (UI) is <strong>of</strong>ten composed<br />
<strong>of</strong> distinctive UI elements, <strong>the</strong> so called UI controls. Similar controls are <strong>of</strong>ten used in<br />
different parts <strong>of</strong> web applications (and even in different web applications) and facilitating<br />
<strong>the</strong>ir reuse could lead to faster development. Unfortunately, preparing code for reuse is a<br />
slow process which is <strong>of</strong>ten not a priority.<br />
Often, when developers encounter problems that have already been solved in <strong>the</strong> past,<br />
ra<strong>the</strong>r than re-inventing <strong>the</strong> wheel, or spending time componentizing, <strong>the</strong>y reuse <strong>the</strong> code<br />
that is already functioning in ano<strong>the</strong>r context [4]. Reuse tasks are complex and prone to<br />
errors, primarily because it is difcult to establish <strong>the</strong> minimum amount <strong>of</strong> code responsible
158<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
for implementing <strong>the</strong> desired behavior [8].<br />
In <strong>the</strong> web application domain, reuse is especially difcult: <strong>the</strong>re is no trivial mapping<br />
between source code and <strong>the</strong> page displayed in <strong>the</strong> browser, responsible code is <strong>of</strong>ten<br />
scattered between several les and is found in between code that is not important from <strong>the</strong><br />
reuse perspective. Next, <strong>the</strong> developer has to locate and download <strong>the</strong> necessary resources,<br />
source code, and adjust for <strong>the</strong> changes so that <strong>the</strong> component can be easily integrated into<br />
an already existing system.<br />
The structure <strong>of</strong> a web page is dened by HTML code, <strong>the</strong> presentation by CSS (Cascading<br />
Style Sheets) code, and <strong>the</strong> behavior by JavaScript code. In addition, a web page usually<br />
contains various resources such as images or fonts. The interplay <strong>of</strong> <strong>the</strong>se four basic elements<br />
produces <strong>the</strong> end result displayed in <strong>the</strong> user’s web browser. Visually and behaviorally a<br />
web page can be viewed as a collection <strong>of</strong> UI controls, where each control is dened by a<br />
combination <strong>of</strong> HTML, CSS, JavaScript and resources (images, videos, fonts, etc.) that are<br />
intermixed with code and resources dening o<strong>the</strong>r parts <strong>of</strong> <strong>the</strong> web page.<br />
In order to reuse a web UI control, we have to extract all that is necessary for <strong>the</strong> control<br />
to be visually and functionally autonomous. This means extracting all HTML, CSS,<br />
JavaScript and resources that are used in <strong>the</strong> visual presentation and <strong>the</strong> desired behavior<br />
<strong>of</strong> <strong>the</strong> control. The process can be separated into three phases: 1) Interaction recording, 2)<br />
Resource extraction, and 3) UI control reuse (Figure 2).<br />
Figure 2: Extracting Web User-Interface controls.<br />
The rst step <strong>of</strong> <strong>the</strong> Interaction recording phase is to select <strong>the</strong> HTML node that denes<br />
<strong>the</strong> chosen UI control. Next, <strong>the</strong> user performs a series <strong>of</strong> interactions that represent <strong>the</strong><br />
behavior <strong>of</strong> <strong>the</strong> control. The purpose <strong>of</strong> this phase is to ga<strong>the</strong>r a log <strong>of</strong> all resources required<br />
for replicating visual and behavioral aspects <strong>of</strong> <strong>the</strong> control. This is done by logging all<br />
executed code, all CSS styles and all resources used in <strong>the</strong> lifecycle <strong>of</strong> <strong>the</strong> control.
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 159<br />
When <strong>the</strong> user chooses to end <strong>the</strong> recording, <strong>the</strong> process enters <strong>the</strong> Resource extraction<br />
phase, where code models for all code les (HTML, CSS, and JavaScript) are built. Based<br />
on those models and logs ga<strong>the</strong>red during <strong>the</strong> recording phase, <strong>the</strong> code necessary for<br />
replicating <strong>the</strong> visuals and <strong>the</strong> demonstrated behavior is extracted.<br />
After <strong>the</strong> extraction phase is completed, <strong>the</strong> user can enter <strong>the</strong> Reuse phase and automatically<br />
integrate <strong>the</strong> extracted control in an already existing web page, ei<strong>the</strong>r by replacing, or by<br />
embedding it inside an existing node. With this, a full cycle is completed: from seeing <strong>the</strong><br />
potential for reuse, through control extraction, all <strong>the</strong> way to actual reuse and gaining new<br />
functionalities on <strong>the</strong> target web application.<br />
The whole process is currently supported by <strong>the</strong> Firecrow tool, which is an extension for <strong>the</strong><br />
Firebug web debugger. Currently, <strong>the</strong> tool can be used from <strong>the</strong> Firefox web browser, but<br />
it can be ported to any o<strong>the</strong>r web browser that provides communication with a JavaScript<br />
debugger, and a DOM explorer (e.g. IE, Chrome, Opera). The interaction recording phase<br />
is <strong>the</strong> only part <strong>of</strong> <strong>the</strong> process that is browser dependent; building source models, extracting<br />
code, downloading resources, merging code and resources are all functionalities that are all<br />
encapsulated in a library that can be called from any browser on any operating system. The<br />
whole source <strong>of</strong> <strong>the</strong> program can be downloaded from [18].<br />
3 Hardware and S<strong>of</strong>tware Components<br />
The main purpose <strong>of</strong> embedded systems is to monitor or control processes in <strong>the</strong>ir<br />
environment. This interaction <strong>of</strong> s<strong>of</strong>tware with <strong>the</strong> real world requires integration <strong>of</strong><br />
s<strong>of</strong>tware components with hardware components such as sensors and actuators. However,<br />
coping with <strong>the</strong>se two different parts <strong>of</strong> embedded systems simultaneously is still a<br />
challenge. Current component models for embedded systems mostly focus on s<strong>of</strong>tware<br />
components and rarely try to provide extensive support hardware component [7]. In DICES<br />
project we have also addressed this aspect <strong>of</strong> component-based s<strong>of</strong>tware development. One<br />
<strong>of</strong> <strong>the</strong> results <strong>of</strong> our research is UComp component model and supporting technology [9].<br />
3.1 UComp – component-based development for hardware and s<strong>of</strong>tware components<br />
The focus <strong>of</strong> UComp is on distributed systems whose functionality is implemented using<br />
various devices connected to a computer network. These devices may ei<strong>the</strong>r be physical, i.e.<br />
realized using hardware, or virtual, i.e. realized using s<strong>of</strong>tware applications. Main goal <strong>of</strong><br />
UComp is to utilise <strong>the</strong> component-based approach and manage <strong>the</strong> hardware and s<strong>of</strong>tware<br />
components in a uniform way. Fur<strong>the</strong>r, our goal is to apply <strong>the</strong> component-based approach<br />
during whole life-cycle <strong>of</strong> a system, including run-time phase. By having <strong>the</strong> components<br />
available at run-time, systems can be extremely exible because any modications can be<br />
done while <strong>the</strong> system is running and <strong>the</strong> embedded devices are deployed. In addition, this<br />
would allow easier late deployment <strong>of</strong> new devices (i.e. components) or replacement <strong>of</strong><br />
existing ones during run-time.<br />
3.1.1 The UComp component model<br />
UComp distinguishes between two types <strong>of</strong> components: Device components and s<strong>of</strong>tware<br />
components. Device components are used to communicate with hardware or virtual network<br />
devices and provide <strong>the</strong>ir functionalities in a component-based manner. Functionality <strong>of</strong><br />
s<strong>of</strong>tware components is fully implemented in program code, and <strong>the</strong>se components are not<br />
associated with any devices. They are used for additional computations in order to avoid<br />
<strong>the</strong> need for “glue code”.
160<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
Component interfaces consist <strong>of</strong> input and output ports. Ports can be considered as<br />
component access points, used for a component to exchange data and control (triggering)<br />
signals. System execution follows <strong>the</strong> pipes and lter pattern. Data and triggering signals<br />
from output port <strong>of</strong> one component can be directed to input ports <strong>of</strong> one or more components.<br />
When an input port receives a signal from <strong>the</strong> output port it is connected to, it becomes<br />
active. A component can be congured to start it’s execution when ei<strong>the</strong>r all or just selected<br />
ports are activated. This is done by selecting activation types for ports. There are three<br />
activation types for input ports:<br />
• Trigger. A component is activated if all input ports with activation type set to trigger<br />
are active.<br />
• Priority trigger. A component is activated if any <strong>of</strong> its input port with activation type<br />
set to priority trigger is active.<br />
• Data. If port’s activation type is set to data, it is only used to receive data, and does not<br />
affect <strong>the</strong> triggering <strong>of</strong> <strong>the</strong> component.<br />
The graphical representation <strong>of</strong> output ports and all input port types can be seen in Figure<br />
3. The gure shows an instance <strong>of</strong> Component A having input ports a (data port), b (trigger<br />
port) and c (priority trigger port), and an output port out.<br />
Figure 3: Graphical representation <strong>of</strong> an UComp component and its input and output ports.<br />
3.1.2 Device Components<br />
Device components are <strong>the</strong> base for accomplishing <strong>the</strong> uniform treatment <strong>of</strong> components <strong>of</strong><br />
a system, which is not dependant on component realization (hardware or s<strong>of</strong>tware). They<br />
represent hardware (physical) and virtual (realized using s<strong>of</strong>tware applications) network<br />
devices.<br />
Device components, toge<strong>the</strong>r with <strong>the</strong>ir input and output ports, are automatically generated<br />
by <strong>the</strong> UComp framework using device descriptions. Automatic generation <strong>of</strong> devices<br />
and <strong>the</strong>ir ports eliminates need for specialized drivers or manual conguration <strong>of</strong> such<br />
components by <strong>the</strong> developer <strong>of</strong> <strong>the</strong> system.<br />
Every device component signals if <strong>the</strong> actual device is available on <strong>the</strong> network. This<br />
information can be very useful in distributed systems where connection between<br />
components is not reliable and some <strong>of</strong> <strong>the</strong>m can be temporarily unavailable.<br />
Device components can represent actions (synchronous request-response communication)<br />
or events (asynchronous publish-subscribe messaging) <strong>of</strong> network devices. Therefore, we<br />
have dened two types <strong>of</strong> device components: action components and event components.<br />
Action Components. Action components are designed to wrap around synchronous action<br />
invocations or data queries <strong>of</strong> a device connected to <strong>the</strong> network.<br />
When an action component is triggered, action invocation is performed, using values <strong>of</strong><br />
its input ports as arguments. When <strong>the</strong> invocation nishes, results <strong>of</strong> <strong>the</strong> invocation are
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 161<br />
propagated to output ports <strong>of</strong> <strong>the</strong> components.<br />
Event Components. Event components allow receiving asynchronous messages from<br />
devices. These messages may signal data changes or o<strong>the</strong>r events that device may provide.<br />
Interfaces <strong>of</strong> event components have only output ports, which obtain <strong>the</strong>ir values from<br />
event notications provided by <strong>the</strong> network device.<br />
3.1.3 S<strong>of</strong>tware Components<br />
Functionality <strong>of</strong> s<strong>of</strong>tware components is fully implemented by program code, and <strong>the</strong>y are<br />
not bound to any hardware elements. They are used to process <strong>the</strong> data received from, or<br />
sent to, device components or manipulate <strong>the</strong> execution <strong>of</strong> components. Their function can<br />
vary from very simple (for example addition <strong>of</strong> two numbers) to complex data processing.<br />
Execution Semantics<br />
Initially, all components in <strong>the</strong> system are in an idle state waiting to be activated for<br />
execution. Activation can be caused ei<strong>the</strong>r by <strong>the</strong> triggering signals received at <strong>the</strong> input<br />
ports <strong>of</strong> <strong>the</strong> component, or by its internal events. Action components and most s<strong>of</strong>tware<br />
components are passive, meaning that <strong>the</strong>y execute only when <strong>the</strong>y are triggered by signals<br />
received from o<strong>the</strong>r components, while event components and some s<strong>of</strong>tware components<br />
are active and thus may start <strong>the</strong>ir execution by an internal event.<br />
Realisation <strong>of</strong> UComp Component Model<br />
The UComp architecture, shown in Figure 4, is realised as Java application that implements<br />
<strong>the</strong> Universal Plug and Play (UpnP) [16] technology to control devices available on <strong>the</strong><br />
network, process <strong>the</strong>ir data, and relay data between <strong>the</strong>m. The application communicates<br />
with devices through a single UPnP control point implemented using CyberLink UPnP<br />
stack [11]. Centralized architecture allows us to process data received from a device by<br />
UComp application before it is forwarded to o<strong>the</strong>r devices, making <strong>the</strong> system much more<br />
exible and eliminating <strong>the</strong> need to change <strong>the</strong> code <strong>of</strong> devices to adapt <strong>the</strong>m to <strong>the</strong> needs<br />
<strong>of</strong> <strong>the</strong> developed system. Also, run-time modication <strong>of</strong> systems is much easier. System’s<br />
behaviour can be modied by simple changes in <strong>the</strong> interconnection <strong>of</strong> components (or by<br />
changing <strong>the</strong> components <strong>the</strong>mselves) in <strong>the</strong> central application.<br />
4 Development environment<br />
Figure 4: The UComp architecture.
162<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
To facilitate <strong>the</strong> development we have created a tool for visual development <strong>of</strong> UComp<br />
systems - UComp Developer. The UComp Developer enables browsing available device<br />
and s<strong>of</strong>tware components using a tree structure, visual representation <strong>of</strong> components on a<br />
development panel, modifying connections between <strong>the</strong>m, setting <strong>the</strong>ir properties and <strong>the</strong><br />
properties <strong>of</strong> <strong>the</strong>ir ports, and starting and stopping <strong>the</strong> execution <strong>of</strong> <strong>the</strong> developed system.<br />
Systems developed with this tool are saved or restored from XML les.<br />
A screen-shot <strong>of</strong> UComp Developer is shown in Figure 5.<br />
5 An overview <strong>of</strong> <strong>the</strong> PRIDE tool<br />
Figure 5: Screenshot <strong>of</strong> UComp Developer tool.<br />
PROGRESS-IDE (PRIDE) is an integrated development environment used for s<strong>of</strong>tware<br />
development <strong>of</strong> distributed embedded systems (ES), primarily in <strong>the</strong> automotive,<br />
automation and telecom domains. The grand vision <strong>of</strong> Progress-IDE is to cover <strong>the</strong> whole<br />
s<strong>of</strong>tware development cycle <strong>of</strong> distributed embedded systems, from early system design<br />
to deployment and syn<strong>the</strong>sis. The development process <strong>of</strong> ES requires a strong emphasis<br />
on analysis, verication and validation in order to ensure <strong>the</strong> necessary quality <strong>of</strong> <strong>the</strong><br />
delivered product, <strong>the</strong>refore PRIDE integrates various analysis tools. Compared to <strong>the</strong><br />
majority <strong>of</strong> existing IDEs that focus mainly on <strong>the</strong> programming aspect, PRIDE uses <strong>the</strong><br />
notion <strong>of</strong> component as a rst-class concept allowing <strong>the</strong> manipulation and modelling <strong>of</strong><br />
components, with reusability as one <strong>of</strong> <strong>the</strong> key concerns and supporting:<br />
• Component and system design using <strong>the</strong> ProCom component model [13],<br />
• System analysis (worst-case execution time, model checking <strong>of</strong> behavioural models<br />
and fault propagation and transformation calculus),<br />
• Deployment modelling<br />
• Code syn<strong>the</strong>sis.
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 163<br />
In <strong>the</strong> following list we give some overall reections that apply to Progress-IDE.<br />
• Progress-IDE covers <strong>the</strong> whole system development process,<br />
• Progress-IDE is developed as a stand-alone application on top <strong>of</strong> Eclipse RCP,<br />
• Progress-IDE consists <strong>of</strong> a number <strong>of</strong> editors: architectural editors, attribute editor, node<br />
description (virtual/physical), allocation editor, code generation, various analysis editors,<br />
• Progress-IDE is user-friendly by providing a user interface respecting <strong>the</strong> usability<br />
features, partial auto-completion features, default values etc.<br />
5.1 The PRIDE approach<br />
In <strong>the</strong> recent years ES development has changed signicantly due to <strong>the</strong> rapid increase <strong>of</strong><br />
s<strong>of</strong>tware in <strong>the</strong>se systems and s<strong>of</strong>tware becoming as complex as in conventional systems.<br />
In non-embedded domains, new approaches such as model-based, component-based, and<br />
service-oriented development have been proposed to manage s<strong>of</strong>tware complexity and<br />
<strong>the</strong>re is a trend to apply <strong>the</strong>se approaches also in ES development. ES correctness is strongly<br />
correlated to specic extra-functional properties (EFPs) such as timing (e.g. execution and<br />
response time) or dependability (e.g. reliability and safety) under constrained resources<br />
such as memory, energy, or computation speed. This calls for additional domain-specic<br />
technologies that provide support not only for functional development but also for analysis<br />
and verication <strong>of</strong> EFPs.<br />
As a possible solution, we have been developing a new component-based approach [5]<br />
built around a two-layer component model - ProCom, which addresses <strong>the</strong> particularity<br />
<strong>of</strong> ES development from big complex functionalities, to small, close to control loop<br />
functionalities. This approach requires specic tool support that enables:<br />
Efcient system design by using existing components,<br />
Seamless integration <strong>of</strong> different tools to provide <strong>the</strong> analysis and verication required for<br />
system correctness, and<br />
Ef cient EFP management <strong>of</strong> components and systems.<br />
PRIDE uses reusable s<strong>of</strong>tware components as <strong>the</strong> central development units, and as a<br />
means to support and aggregate various analysis and verication techniques. In difference<br />
to similar approaches [1, 2, 17], PRIDE puts emphasis on EFPs during <strong>the</strong> entire lifecycle<br />
- from early specication to deployment and syn<strong>the</strong>sis.<br />
PRIDE has been designed to support four design strategies that are especially important to<br />
consider for having an efcient component-based development <strong>of</strong> ES.<br />
Levels <strong>of</strong> abstraction<br />
Using components throughout <strong>the</strong> whole development process implies that <strong>the</strong> component<br />
concept spans a wide range <strong>of</strong> abstractions, from a vague and incomplete early specication,<br />
to very “concrete” with a xed specication, a corresponding implementation and<br />
information about <strong>the</strong>ir EFPs. This means that components at different levels <strong>of</strong> abstraction<br />
must be able to co-exist within <strong>the</strong> same model.<br />
Component granularity<br />
In distributed ES, components span a large variety in size and complexity; <strong>the</strong> larger<br />
components are typically active (i.e. with <strong>the</strong>ir own thread <strong>of</strong> execution) with an<br />
asynchronous message passing communication style, whereas <strong>the</strong> smaller components are<br />
responsible for a part <strong>of</strong> control functionality with a strong synchronization. For an efcient<br />
development, a support for handling different types <strong>of</strong> components must be provided.<br />
Component vs. system development<br />
The common distinction between component development and system development brings
164<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
issues in ES development, where <strong>the</strong> coupling between <strong>the</strong> hardware platform and <strong>the</strong><br />
s<strong>of</strong>tware is particularly tight. As a consequence, component development needs some<br />
knowledge <strong>of</strong> where <strong>the</strong> components are to be deployed. This requires support to handle<br />
<strong>the</strong> coupling between components, system and target platform, while still allowing separate<br />
development <strong>of</strong> components and systems.<br />
Extra-functional properties<br />
The correctness <strong>of</strong> ES is ascertained based on both <strong>the</strong> functional and extra-functional<br />
aspects. However, many EFPs typically encountered in ES are assessed through different<br />
methods during <strong>the</strong> development lifecycle (from early estimation to precise measurements)<br />
and different values may be obtained according to <strong>the</strong> characteristics <strong>of</strong> <strong>the</strong> resources <strong>of</strong><br />
<strong>the</strong> platform on which <strong>the</strong> components are to be deployed. For this reason, an efcient<br />
ES development should provide a means for specifying, managing and verifying <strong>the</strong>se<br />
properties with respect to <strong>the</strong> context in which <strong>the</strong>ir values are provided.<br />
To comply with <strong>the</strong> design strategies, <strong>the</strong> following requirements have been identied as<br />
principles that guided <strong>the</strong> design and development <strong>of</strong> PRIDE:<br />
Allowing to move freely between any development stages,<br />
Displaying <strong>the</strong> consequences <strong>of</strong> a change in <strong>the</strong> system or within a component,<br />
Supporting <strong>the</strong> coupling with <strong>the</strong> hardware platform, and<br />
Enabling and enforcing <strong>the</strong> analysis, validation and verication steps.<br />
In addition, a central requirement relates to <strong>the</strong> notion <strong>of</strong> component. Components are <strong>the</strong><br />
main units <strong>of</strong> development and seen as rich-design artifacts that exist throughout <strong>the</strong> whole<br />
development lifecycle, from early design stage, in which little information about <strong>the</strong>m<br />
exists, to deployment and syn<strong>the</strong>sis stages, in which <strong>the</strong>y are fully implemented. PRIDE<br />
views a component as a collection <strong>of</strong> all <strong>the</strong> development artefacts (requirements, models,<br />
EFPs, documentation, tests, source code, etc.), and enables <strong>the</strong>ir manipulation in a uniform<br />
way.<br />
Driven by <strong>the</strong> aforementioned principles, several tools have been developed and tightly<br />
integrated into PRIDE. Since <strong>the</strong> PRIDE is built as an Eclipse RCP application, it can be<br />
easily extended with addition <strong>of</strong> new plugins.<br />
5.2 Implementation <strong>of</strong> PRIDE<br />
The Eclipse platform is chosen as <strong>the</strong> supporting architecture for PRIDE. In order to reduce<br />
<strong>the</strong> overhead which exists when using Eclipse directly as <strong>the</strong> main integration environment,<br />
it has been decided that <strong>the</strong> PRIDE will be developed as a stand-alone application built on<br />
top <strong>of</strong> <strong>the</strong> Eclipse Rich Client Platform (Eclipse RCP). This decision is driven by <strong>the</strong> will<br />
to keep <strong>the</strong> control over <strong>the</strong> features present in <strong>the</strong> environment such as, for example,<br />
avoiding <strong>the</strong> presence <strong>of</strong> menus not directly related to <strong>the</strong> particular use <strong>of</strong> <strong>the</strong> PRIDE.<br />
When using Eclipse RCP, only <strong>the</strong> features explicitly added to <strong>the</strong> environment are present.<br />
In o<strong>the</strong>r words, <strong>the</strong> layout and function <strong>of</strong> <strong>the</strong> environment are fully controlled by <strong>the</strong><br />
plugin developers.<br />
The component architecture design part <strong>of</strong> <strong>the</strong> environment was implemented using EMF<br />
[14] and GMF [15]. EMF is a modelling framework and code generation facility for<br />
building tools based on a structured data model. It <strong>of</strong>fers a graphical editor for describing<br />
metamodels. With <strong>the</strong> aid <strong>of</strong> this, <strong>the</strong> ProCom metamodel (model <strong>of</strong> <strong>the</strong> ProCom model)<br />
was dened. From this graphical description <strong>of</strong> <strong>the</strong> metamodel, EMF automatically<br />
generates Java classes representing <strong>the</strong> model, classes used for modifying <strong>the</strong> model and<br />
textual tree editors for <strong>the</strong> model. GEF <strong>the</strong>n takes this existing application model (i.e.<br />
<strong>the</strong> model generated using EMF) and quickly creates a rich graphical editor. However,<br />
this automatically generated editor has rudimentary functionality and needs to be fur<strong>the</strong>r
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 165<br />
manually modied and tweaked in order to achieve <strong>the</strong> desired functionality.<br />
S<strong>of</strong>tware architecture modelling and analysis<br />
As already mentioned, PRIDE uses component-based approach which allows components<br />
to be developed independently and reused in different contexts. Reusability is one <strong>of</strong><br />
<strong>the</strong> main concepts in PRIDE aiming to signicantly shorten development time. The tool<br />
makes a distinction between a component type and a component instance. Each reusage<br />
<strong>of</strong> a component creates a component instance <strong>of</strong> <strong>the</strong> given component type. By editing <strong>the</strong><br />
component all <strong>of</strong> its instances are affected.<br />
In PRIDE components are rich design entities encapsulating a collection <strong>of</strong> development<br />
artefacts; requirements, various models (e.g. architectural, behavioural, resource usage<br />
etc.), EFPs, documentation, tests and source code.<br />
Figure 6 shows a screenshot from PRIDE with main parts highlighted. PRIDE’s modelling<br />
part consists <strong>of</strong> a component explorer and component editors.<br />
Figure 6: A screenshot from PRIDE showing a) <strong>the</strong> component explorer; b) <strong>the</strong> component editor;<br />
c) <strong>the</strong> code editor; d) <strong>the</strong> repository browser; and e) <strong>the</strong> attribute framework.<br />
Component Explorer<br />
Enables browsing <strong>the</strong> list <strong>of</strong> <strong>the</strong> components available in <strong>the</strong> current development project.<br />
In it a component owns a predened and extensible information structure corresponding<br />
to a rich component concept. It also provides a support for component versioning and<br />
importing and exporting from a project to a component repository making <strong>the</strong>m available<br />
to o<strong>the</strong>r projects thus facilitating <strong>the</strong> component reusability.<br />
Component Editors<br />
Although <strong>the</strong> ProCom component model distinguishes between two different types <strong>of</strong><br />
components (ProSave and ProSys), in <strong>the</strong> component editors all components are treated<br />
in a uniform way. Component editors provide two independent views on a component,<br />
external and internal view. The external view manages <strong>the</strong> component specication<br />
and interface such as <strong>the</strong> information about <strong>the</strong> component name, its interface (services<br />
and ports) and EFPs. The internal view focus on component’s internal structure i.e. its<br />
realization and implementation and it depends in <strong>the</strong> component realization type. For<br />
primitive components, <strong>the</strong> internal view is linked to <strong>the</strong> component implementation and
166<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
<strong>the</strong> source code is displayed. For composite components, <strong>the</strong> internal view corresponds<br />
to an interconnection <strong>of</strong> subcomponent instances and a graphical form is made available<br />
allowing to make modications in this inner structure (addition/deletion <strong>of</strong> component<br />
instances, connectors, change in <strong>the</strong> connections, etc.).<br />
The analysis support in PRIDE is based on two main parts, an attribute framework and an<br />
analysis framework.<br />
Attribute Framework<br />
The purpose <strong>of</strong> <strong>the</strong> attribute framework [12] is to provide a uniform and user-friendly<br />
structure to seamlessly manage EFPs in a systematic way. The attribute framework enables<br />
attaching extra-functional properties to any architectural element such as a specic port,<br />
service or <strong>the</strong> component as a whole. Attributes are dened by attribute types, and include<br />
attribute values with metadata and <strong>the</strong> specication <strong>of</strong> <strong>the</strong> conditions under which <strong>the</strong><br />
attribute value is valid. One key feature is that <strong>the</strong> attribute framework allows an attribute<br />
to be given additional values during <strong>the</strong> development without replacing old values.<br />
This allows dening <strong>of</strong> early estimates for EFPs even before <strong>the</strong> component has been<br />
implemented and use it for analysis in early stages <strong>of</strong> system development. New, userdened<br />
attribute types can also be added to <strong>the</strong> model.<br />
Analysis framework<br />
The analysis framework provides a common platform for integrating in a consistent<br />
way various analysis techniques, ranging from simple constraint checking and attribute<br />
derivation (e.g., propagating port type information over connections), to complex external<br />
analysis tools. Analysis results can ei<strong>the</strong>r be presented to <strong>the</strong> user directly, or stored as<br />
component attributes. They are also added to a common analysis result log, allowing <strong>the</strong><br />
user easy access to earlier analysis results. PRIDE also allows to easily integrate new<br />
analysis techniques toge<strong>the</strong>r with <strong>the</strong>ir associated EFPs.<br />
6 Conclusion<br />
In this paper we gave an overview <strong>of</strong> some <strong>of</strong> <strong>the</strong> results <strong>of</strong> DICES project. The project<br />
focus was on distributed embedded systems and <strong>the</strong> overview shows that a large variety<br />
<strong>of</strong> technologies and approaches are needed for an efcient development and maintenance<br />
<strong>of</strong> embedded systems. The project is focused on several phases in <strong>the</strong> lifecycle: a) reverse<br />
<strong>engineering</strong> and analysis <strong>of</strong> <strong>the</strong> existing legacy code that helps in reusability in development<br />
<strong>of</strong> new systems from existing components; b) modelling and designing embedded systems<br />
that include both s<strong>of</strong>tware and hardware components, and ability to perform different<br />
types <strong>of</strong> analysis important for embedded systems (e.g. resources utilization, performance,<br />
timing characteristics). The work done shows that <strong>the</strong> embedded systems <strong>of</strong> today are<br />
complex system, in <strong>the</strong> rst hand with complex s<strong>of</strong>tware that is approaching in size and<br />
complexity s<strong>of</strong>tware from o<strong>the</strong>r domains.<br />
Acknowledgment<br />
This work was supported by <strong>the</strong> Swedish Foundation for Strategic Research via <strong>the</strong><br />
strategic research centre Progress, Croatian Ministry <strong>of</strong> science, education and sports via<br />
<strong>the</strong> research project S<strong>of</strong>tware <strong>engineering</strong> in ubiquitous computing, and <strong>the</strong> Unity Through<br />
Knowledge Fund via <strong>the</strong> Dices project.
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 167<br />
References<br />
1. Akerholm M., Carlson J., Fredriksson J., Hansson H., Hakansson J., Möller A., Pettersson<br />
P., Tivoli M. (2007) The SAVE Approach to Component-Based Development <strong>of</strong> Vehicular<br />
Systems, In: Journal <strong>of</strong> Systems and S<strong>of</strong>tware, Elsevier Science, pp. 655-667<br />
2. Articus Systems, Rubus S<strong>of</strong>tware Components, http://www.arcticus-systems.com ,<br />
Accessed 7 March <strong>2011</strong><br />
3. Bellettini C., Marchetto A., Trentini A. (2004) WebUml: reverse <strong>engineering</strong> <strong>of</strong> web<br />
applications, In: Symposium on Applied computing, ACM New York, NY, USA, 1662-<br />
1669<br />
4. Brandt J., Guo J.P., Lewenstein J., Klemmer S.R. (2008) Opportunistic programming:<br />
How rapid ideation and prototyping occur in practice, In: Workshop on End-user s<strong>of</strong>tware<br />
<strong>engineering</strong>, ACM New York, NY, USA, pp. 1-5<br />
5. Bureš T., Carlson J., Sentilles S., Vulgarakis A. (2008) A Component Model Family<br />
for Vehicular Embedded Systems, In: The Third International Conference on S<strong>of</strong>tware<br />
Engineering Advances, IEEE Computer Society Washington, DC, USA, pp. 437-444<br />
6. Di Lucca G., Fasolino A.r., Tramontana P. (2004) Reverse <strong>engineering</strong> Web applications:<br />
<strong>the</strong> WARE approach, In: Journal <strong>of</strong> S<strong>of</strong>tware Maintenance and Evolution: Research and<br />
Practice - Special issue: Web site evolution Volume 16 Issue 1-2, John Wiley & Sons, Inc.<br />
New York, NY, USA, pp. 71-101<br />
7. Feljan J., Lednicki L., Maras J., Petricic A., Crnkovi I. (2009) Classication and survey <strong>of</strong><br />
component models, MRTC report ISSN 1404-3041 ISRN MDH-MRTC-242/2009-1-SE,<br />
Mälardalen Real-Time Research Centre, Mälardalen University<br />
8. Holmes R.(2008) Pragmatic S<strong>of</strong>tware Reuse, Ph.D. Dissertation, University <strong>of</strong> Calgary<br />
9. Lednicki, L. and Petricic, A. and Zagar, M. (2009) A Component-Based Technology<br />
for Hardware and S<strong>of</strong>tware Components, In: S<strong>of</strong>tware Engineering and Advanced<br />
Applications, 2009. SEAA ‘09. 35th Euromicro Conference on, IEEE Computer Society<br />
Washington, DC, USA, pp. 450 -453<br />
10. Ricca F., Tonella P. (2001) Understanding and Restructuring Web Sites with ReWeb, IEEE<br />
Multimedia, Vol 8, Issue 2, pp. 40-51<br />
11. Satoshi Konno, CyberLink for Java, http://www.cybergarage.org/twiki/bin/view/Main/<br />
CyberLinkForJava , Accessed 7 March <strong>2011</strong><br />
12. Sentilles S., Štpán P., Carlson J., Crnkovi I. (2009) Integration <strong>of</strong> Extra-Functional<br />
Properties in Component Models, In: Proceedings <strong>of</strong> <strong>the</strong> 12th International Symposium on<br />
Component-Based S<strong>of</strong>tware Engineering (CBSE ‘09), Springer-Verlag Berlin, Heidelberg,<br />
pp. 173-190<br />
13. Sentilles S., Vulgarakis A., Bureš T., Carlson J., Crnkovi I. (2008) A Component Model<br />
for Control-Intensive Distributed Embedded Systems, In: Proceedings <strong>of</strong> <strong>the</strong> 11th<br />
International Symposium on Component-Based S<strong>of</strong>tware Engineering (CBSE ‘08),<br />
Springer-Verlag Berlin, Heidelberg, pp. 310-317<br />
14. The Eclipse Foundation, Eclipse Modeling Framework Project, http://www.eclipse.org/<br />
emf/ , Accessed 7 March <strong>2011</strong><br />
15. The Eclipse Fundation, Graphical Modeling Framework, http://http://www.eclipse.org/<br />
gmf , Accessed 7 March <strong>2011</strong><br />
16. UPnP forum, Universal Plug and Play, http://www.upnp.org , Accessed 7 March <strong>2011</strong><br />
17. van Ommering R., van der Linden F., Kramer J., Magee J. (2000) The Koala component<br />
model for consumer electronics s<strong>of</strong>tware, In: IEEE Computer, , IEEE Computer Society<br />
Washington, DC, USA, pp. 78-85<br />
18. Maras J., Štula M., Carlson J, Crnkovi I. (<strong>2011</strong>). Firecrow web page, http://www.fesb.<br />
hr/~jomaras/?id=Firecrow, Accessed 7 March <strong>2011</strong>
168<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
Framework for 3D Motion Field Estimation and<br />
Reconstruction<br />
Martin Žagar, Hrvoje Mlinarić, Josip Knezović<br />
Department <strong>of</strong> Control and Computer Engineering, Faculty <strong>of</strong> Electrical Engineering<br />
and Computing, Unska 3, HR-10000 Zagreb, Croatia<br />
Abstract:<br />
We propose <strong>the</strong> framework for <strong>the</strong> motion estimation <strong>of</strong> 3D objects based on <strong>the</strong> motion<br />
vectors that form motion elds and <strong>the</strong> motion reconstruction based on <strong>the</strong> 3D rotations<br />
and <strong>the</strong> factorization method. The classical two-dimensional motion eld approach is<br />
extended to three dimensions, e.g. on <strong>the</strong> volumetric objects that are moving in time. When<br />
dealing with <strong>the</strong> real world multiple moving objects and <strong>the</strong> complex scenes, lots <strong>of</strong> objects<br />
are moving with different motions, both in space and in time. In this context, an object<br />
can be described as a part <strong>of</strong> a scene that moves with a coherent motion and <strong>the</strong> scene can<br />
be broken down into a number <strong>of</strong> regions, each <strong>of</strong> which can be well approximated by its<br />
own motion. The rst part <strong>of</strong> this paper describes <strong>the</strong> motion vectors based techniques<br />
which are used for <strong>the</strong> motion estimation, and <strong>the</strong> second part addresses problems with <strong>the</strong><br />
reconstruction <strong>of</strong> <strong>the</strong> 3D motion and structure. Proposed methods estimate reconstruction<br />
from a sparse set <strong>of</strong> <strong>the</strong> matched volume features on <strong>the</strong> 3D neuroimage in NIfTI format.<br />
We evaluate <strong>the</strong>se techniques according to <strong>the</strong> different problems <strong>the</strong>y address.<br />
Keywords: 3D motion estimation, differential techniques, matching techniques, motion<br />
and structure reconstruction, factorization method.<br />
1 INTRODUCTION<br />
The motion estimation is used to eliminate a large amount <strong>of</strong> temporal and spatial<br />
redundancy that exists in sequences <strong>of</strong> 3D data. Most video encoders perform <strong>the</strong> motion<br />
estimation in video sequences to search for <strong>the</strong> motion information in sequential video<br />
frames that can improve compression. The difference between <strong>the</strong> current frame and <strong>the</strong><br />
predicted frame in motion estimation and motion reconstruction (based on previous and/or<br />
future frames) is coded and transmitted. The better <strong>the</strong> prediction is, <strong>the</strong> smaller <strong>the</strong> error<br />
and hence <strong>the</strong> transmission bit rates are. If a scene is still, <strong>the</strong>n a good prediction for <strong>the</strong><br />
particular object in <strong>the</strong> current frame is <strong>the</strong> same as object in <strong>the</strong> previous frame and <strong>the</strong><br />
error is close to zero. However, when <strong>the</strong>re is a motion in a sequence, an estimation <strong>of</strong> <strong>the</strong><br />
predicted position <strong>of</strong> <strong>the</strong> object in <strong>the</strong> current frame can be made with <strong>the</strong> motion vectors<br />
[7], [10]. Many techniques based on <strong>the</strong> motion vectors that form motion elds can be<br />
roughly divided into two major classes: differential techniques and feature-based matching<br />
techniques [9]. The differential techniques are based on <strong>the</strong> spatial and temporal variations<br />
<strong>of</strong> <strong>the</strong> volume brightness at all voxels, and can be regarded as methods for computing<br />
optical ow. The matching techniques, instead, estimate <strong>the</strong> disparity <strong>of</strong> special volume<br />
points (features) between frames. Proposed methods are evaluated on 3D neuroimages<br />
in NIfTI (Neuroimaging Informatics Technology Initiative) format which is designed to
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 169<br />
maximize <strong>the</strong> usability <strong>of</strong> neuroimaging computing tools and provide a common resource<br />
for researches in magnetic resonance imaging (MRI). It is used for <strong>the</strong> detailed visualization<br />
<strong>of</strong> internal structure and functions <strong>of</strong> <strong>the</strong> human body. The main purpose <strong>of</strong> determining<br />
motion in MRI is to reduce a vast amount <strong>of</strong> acquired data in coding process. For different<br />
types <strong>of</strong> MRI, different motion techniques should be used for motion estimation and motion<br />
reconstruction. Estimation <strong>of</strong> 3D motion eld is based on <strong>the</strong> motion vectors and is used to<br />
predict and to describe motion between <strong>the</strong> current and <strong>the</strong> next frame. With coding only<br />
motion vectors instead <strong>of</strong> whole dataset, higher compression ratio can be achieved. On <strong>the</strong><br />
o<strong>the</strong>r hand, in <strong>the</strong> motion reconstruction, 3D motion is reconstructed from a sparse set <strong>of</strong><br />
matched volume features.<br />
The paper is organized as follows. Our proposed differential and matching techniques are<br />
described in Section 2. We adapted motion vectors based techniques as one <strong>of</strong> appropriate<br />
techniques for motion estimation and factorization method for 3D motion reconstruction.<br />
Factorization method is described in Section 3.2, while <strong>the</strong> beginning <strong>of</strong> Section 3 presents<br />
our 3D edge detection method for segmentation, based on Canny 2D edge detection [11].<br />
The segmentation is necessary when dealing with multiple moving objects to detect and<br />
to segment objects that are moving with different motions in space in time. Experimental<br />
results <strong>of</strong> proposed methods and techniques are presented in Section 4 and concluded in<br />
Section 5.<br />
2 MOTION VECTORS BASED TECHNIQUES<br />
2.1 Differential Techniques based on Optical Flow<br />
A large number <strong>of</strong> differential techniques for computing <strong>the</strong> optical ow have been proposed<br />
[1], [3], [8]. Some <strong>of</strong> <strong>the</strong>m require <strong>the</strong> solution <strong>of</strong> a system <strong>of</strong> partial differential equations,<br />
o<strong>the</strong>rs <strong>the</strong> computation <strong>of</strong> second and higher-order derivatives <strong>of</strong> <strong>the</strong> volume brightness<br />
and some <strong>of</strong> <strong>the</strong>m require <strong>the</strong> least-squares estimates <strong>of</strong> <strong>the</strong> parameters characterizing <strong>the</strong><br />
optical ow. Methods in <strong>the</strong> latter class have at least two advantages over those in <strong>the</strong> rst<br />
two:<br />
• They are not iterative; <strong>the</strong>refore <strong>the</strong>y are genuinely local and less biased by possible<br />
discontinuities <strong>of</strong> <strong>the</strong> motion eld than iterative methods.<br />
• They do not involve derivatives <strong>of</strong> order higher than <strong>the</strong> rst; <strong>the</strong>refore, <strong>the</strong>y are less<br />
sensitive to noise than methods requiring higher-order derivatives.<br />
2.1.1 Optical flow<br />
Our proposed differential technique is <strong>the</strong> extension <strong>of</strong> 2D motion eld estimation ([3],<br />
[11]) to 3D. It is based on <strong>the</strong> least-squares estimates <strong>of</strong> <strong>the</strong> parameters characterizing<br />
<strong>the</strong> optical ow. The basic assumption is that <strong>the</strong> motion eld is well approximated by a<br />
constant vector eld velocity v r<br />
within any small region <strong>of</strong> <strong>the</strong> volume plane. The optical<br />
ow is <strong>the</strong> approximation <strong>of</strong> <strong>the</strong> motion eld which can be computed from time-varying<br />
volume sequences. For each point p i<br />
within a small patch Q <strong>of</strong> size N x N x N, it can be<br />
written<br />
∂E<br />
( ∇E) T r<br />
v + = 0<br />
(1)<br />
∂t<br />
where E = E(x,y,z,t) is volumetric brightness for voxel at point p(x,y,z) at time t. The spatial<br />
and <strong>the</strong> temporal derivatives <strong>of</strong> <strong>the</strong> volumetric brightness are computed at points (voxels)
170<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
p , p , p ∈<br />
1 2<br />
K , 3 Q . Therefore, <strong>the</strong> optical ow can be estimated within Q as<br />
N<br />
<strong>the</strong> constant vector, v r that minimizes <strong>the</strong> functional<br />
2<br />
T E<br />
<br />
v E<br />
v . (2)<br />
<br />
p Q<br />
t<br />
<br />
i<br />
The solution <strong>of</strong> <strong>the</strong> least squares problem can be found by solving <strong>the</strong> linear system<br />
T<br />
T<br />
A Av = A b [3]. The i-th row <strong>of</strong> <strong>the</strong> N 3 x 3 matrix A is <strong>the</strong> spatial gradient evaluated<br />
at point p i<br />
:<br />
⎡∇E<br />
⎢<br />
⎢<br />
∇E<br />
A =<br />
⎢M<br />
⎢<br />
⎣∇E<br />
( p1<br />
)<br />
( p )<br />
2<br />
⎤<br />
( p ) ⎥⎥⎥⎥ NxNxN ⎦<br />
and b is <strong>the</strong> N 3 -dimensional vector <strong>of</strong> partial temporal derivatives <strong>of</strong> <strong>the</strong> volumetric<br />
brightness, evaluated at p , p2 , K , p 3 after a sign change:<br />
1 N<br />
∂E<br />
⎤<br />
( ) ( ) T<br />
p ,<br />
⎡∂E<br />
b = −<br />
⎢ 1<br />
K , p NxNxN<br />
⎣ ∂t<br />
∂t<br />
⎥ . (4)<br />
⎦<br />
The least squares solution <strong>of</strong> <strong>the</strong> constrained system can be obtained as<br />
r<br />
v =<br />
T −1<br />
T<br />
( A A) A b<br />
(3)<br />
. (5)<br />
ψ [] v r is <strong>the</strong> optical ow (<strong>the</strong> estimate <strong>of</strong> <strong>the</strong> motion eld) at <strong>the</strong> centre <strong>of</strong> patch Q. By<br />
repeating this procedure for all cubic patches, a dense optical ow is obtained. This<br />
algorithm can be summarized as follows. The input is a time-varying sequence <strong>of</strong> n volumes<br />
E 1<br />
, E 2<br />
, …, E n<br />
. Let Q be a sub-volume region <strong>of</strong> N x N x N voxels.<br />
1. Filter each image <strong>of</strong> <strong>the</strong> sequence along each spatial dimension with a Gaussian lter<br />
<strong>of</strong> standard deviation equal to σ s .<br />
2. Filter each image <strong>of</strong> <strong>the</strong> sequence along <strong>the</strong> temporal dimension with a Gaussian<br />
lter <strong>of</strong> standard deviation σ t .<br />
3. For each voxel <strong>of</strong> each volume <strong>of</strong> <strong>the</strong> sequence:<br />
(a) Compute <strong>the</strong> matrix A and <strong>the</strong> vector b using (3) and (4)<br />
(b) Compute <strong>the</strong> optical ow using (5)<br />
Values σ<br />
s and σ<br />
t are set to 1,5 voxels and 1,5 frames respectively, because <strong>of</strong> <strong>the</strong><br />
smoothness. The output is <strong>the</strong> optical ow computed in <strong>the</strong> last step.<br />
2.2 Feature-based Matching Techniques<br />
The second class <strong>of</strong> methods for estimating <strong>the</strong> motion eld is formed by <strong>the</strong> so-called<br />
matching techniques, which estimate <strong>the</strong> motion eld at feature points only. The result is<br />
a sparse motion eld [7]. One proposed solution is a two-frame analysis (nding feature<br />
disparities between consecutive volumetric frames). The robustness <strong>of</strong> frame-to-frame<br />
matching can be improved by motion tracking <strong>of</strong> a feature across a long volumetric
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 171<br />
sequence. We will now describe and adapt two different feature-based matching techniques<br />
from 2D to 3D: <strong>the</strong> constant ow algorithm and <strong>the</strong> feature tracking tracking.<br />
2.2.1 Constant Flow Algorithm<br />
Features can be dened as centres <strong>of</strong> those regions for which <strong>the</strong> smallest eigenvalue <strong>of</strong> A T A<br />
computed over small, cubic regions, is greater than a threshold. The idea <strong>of</strong> <strong>the</strong> matching<br />
method is simple; <strong>the</strong> displacement <strong>of</strong> such feature points can be computed by iterating <strong>the</strong><br />
constant ow algorithm [4].<br />
The input is formed by two frames <strong>of</strong> a volumetric sequence, I 1<br />
and I 2<br />
and a set <strong>of</strong><br />
corresponding feature points in <strong>the</strong> two frames. Let Q 1<br />
, Q 2<br />
, and Q’ be three N x N x N cubic<br />
regions. Let d be <strong>the</strong> unknown displacement between I 1<br />
and I 2<br />
<strong>of</strong> a feature point p on which<br />
Q 1<br />
is centred. Feature point p is centre point <strong>of</strong> <strong>the</strong> shape segmented and dened by <strong>the</strong><br />
edge and corner detection described in Section 3.1. The procedure consists <strong>of</strong> three steps.<br />
First, <strong>the</strong> uniform displacement <strong>of</strong> <strong>the</strong> cube region Q 2<br />
is estimated through constant ow,<br />
and added to <strong>the</strong> current displacement estimate (initially set to 0). Second, <strong>the</strong> patch Q 2<br />
is<br />
wrapped according to <strong>the</strong> estimated ow. This means that Q 2<br />
is displaced according to <strong>the</strong><br />
estimated ow, and <strong>the</strong> resulting patch, Q’, is restored in <strong>the</strong> voxel grid <strong>of</strong> frame I 2<br />
. If <strong>the</strong><br />
estimated ow normalized in time equals (v x<br />
, v y<br />
, v z<br />
), <strong>the</strong> grey value at voxel (i, j, k) <strong>of</strong> Q’<br />
can be obtained from <strong>the</strong> values <strong>of</strong> <strong>the</strong> voxels <strong>of</strong> Q 2<br />
close to (i – v x<br />
, j – v y<br />
, k – v z<br />
). Third, <strong>the</strong><br />
rst and second steps are iterated until a stopping criterion is met. For all feature points p:<br />
1. Set d = 0 and centre Q 1<br />
on p.<br />
2. Estimate <strong>the</strong> displacement d 0<br />
<strong>of</strong> p, centre <strong>of</strong> Q 1<br />
, and let d = d + d 0<br />
.<br />
Let Q’ be <strong>the</strong> patch obtained by warping Q1 according to d0. Compute S, <strong>the</strong> sum <strong>of</strong><br />
<strong>the</strong> squared differences between <strong>the</strong> new patch Q’ and <strong>the</strong> corresponding patch Q2<br />
in <strong>the</strong> frame I2. If S > τ , where τ is <strong>the</strong> threshold, set Q1 = Q’ and go to step 1;<br />
o<strong>the</strong>rwise exit.<br />
The output is an estimate <strong>of</strong> d for all feature points. In reality, long volumetric sequences<br />
ra<strong>the</strong>r than just pairs <strong>of</strong> frames have to be analyzed. The motion <strong>of</strong> feature points is expected<br />
to be continuous, and <strong>the</strong>refore predictable, in most cases. The disparities computed between<br />
frames I i-1<br />
and I i-2<br />
, I i-2<br />
and I i-3<br />
and so on, can be used to make predictions on <strong>the</strong> disparities<br />
between I i-1<br />
and I i<br />
, before observing frame I i<br />
. This algorithm gives satisfying results only<br />
for continuous, predictable motions which can be computed based on disparities between<br />
pairs <strong>of</strong> frames only. To improve <strong>the</strong> algorithm results for long volumetric sequences we<br />
propose to use <strong>the</strong> feature tracking.<br />
2.2.2 Feature Tracking<br />
The feature tracking solves <strong>the</strong> problem <strong>of</strong> matching features from frame to frame in a<br />
long sequence <strong>of</strong> volumes. Our approach is based on [2] and [5] and extended on 3D.<br />
First, it is necessary to formalize <strong>the</strong> tracking problem. A new frame <strong>of</strong> <strong>the</strong> volumetric<br />
sequence is acquired and processed at each instant ti<br />
= ti<br />
0<br />
+ kΔT<br />
, where k is a natural<br />
number and ΔT = 1 is denoted as a sampling interval. Discrete equally-spaced time instants<br />
can be indicated with t i assumed that, for simplicity, ΔT = 1 is small enough to capture<br />
<strong>the</strong> system’s dynamics. The state does not change much between consecutive time instants,<br />
and a linear system model is an adequate approximation <strong>of</strong> <strong>the</strong> state change within ΔT = 1<br />
, i.e. <strong>the</strong> motion <strong>of</strong> feature points from frame to frame is considered to be linear. If only one<br />
feature point, [ ] T<br />
p = x , y , z is considered, in <strong>the</strong> frame acquired at instant t i<br />
, moving<br />
i<br />
i<br />
i<br />
i
172<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
with vector eld velocity v [ ] T<br />
i<br />
= vx, i<br />
, v<br />
y,<br />
i<br />
, v , <strong>the</strong> motion on <strong>the</strong> cubic plane is described<br />
z,<br />
i<br />
with <strong>the</strong> state vector s [ ] T<br />
i<br />
= xi<br />
, yi<br />
, zi<br />
, vx, i<br />
, vy,<br />
i<br />
, v .<br />
z,<br />
i<br />
i<br />
A physical system is modelled by a state vector s , <strong>of</strong>ten called simply <strong>the</strong> state, and a set<br />
s<br />
<strong>of</strong> equations, called <strong>the</strong> system model. The state is a time-dependent vector, i<br />
is stated<br />
as s ( t i<br />
) , <strong>the</strong> components <strong>of</strong> which are system variables, in a sufcient number to capture<br />
<strong>the</strong> dynamic properties <strong>of</strong> <strong>the</strong> system. The system model is a vector equation describing<br />
<strong>the</strong> evolution <strong>of</strong> <strong>the</strong> state in time. Assuming a sufciently small sampling interval (and<br />
<strong>the</strong>refore constant feature velocity between <strong>the</strong> frames), <strong>the</strong> system model can be written as<br />
p ξ (6)<br />
i<br />
= pi− 1<br />
+ vi−<br />
1<br />
+<br />
i−1<br />
v η (7)<br />
= +<br />
i<br />
v i − 1 i−1<br />
where ξi−<br />
1 and η<br />
i−1<br />
are zero-mean, random, vector-modelling additive system noise. In<br />
terms <strong>of</strong> <strong>the</strong> state vector s<br />
i rewrites si<br />
= φ<br />
i− 1si−1<br />
+ wi−<br />
1 with<br />
⎡1<br />
0 0 1 0 0⎤<br />
⎢<br />
⎥<br />
⎢<br />
0 1 0 0 1 0<br />
⎥<br />
⎢0<br />
0 1 0 0 1⎥<br />
φ<br />
i−1<br />
= ⎢<br />
⎥<br />
(8)<br />
⎢0<br />
0 0 1 0 0⎥<br />
⎢0<br />
0 0 0 1 0⎥<br />
⎢<br />
⎥<br />
⎣0<br />
0 0 0 0 1<br />
⎦<br />
and<br />
w i −1<br />
⎡ξ<br />
i−1<br />
⎤<br />
= ⎢ ⎥<br />
⎣η<br />
i−1<br />
⎦<br />
. (9)<br />
The difference between <strong>the</strong> predicted state and <strong>the</strong> real state is measured with <strong>the</strong> state<br />
correction via <strong>the</strong> lter gain factor K t<br />
, known from Kalman ltering. Estimation step, a<br />
priori estimation, estimation error variance and correction step are calculated based on [6].<br />
3 3D MOTION RECONSTRUCTION<br />
When dealing with <strong>the</strong> real world multiple moving objects and complex scenes, lots <strong>of</strong><br />
objects are moving with <strong>the</strong> different motions. In this context, an object can be described<br />
as a part <strong>of</strong> a scene that moves with a coherent motion and <strong>the</strong> scene can be broken down<br />
into a number <strong>of</strong> regions, each <strong>of</strong> which can be well approximated by its own motion.<br />
Segmenting objects against xed background, based on <strong>the</strong> edge detection, as a solution <strong>of</strong><br />
nding <strong>the</strong> regions <strong>of</strong> <strong>the</strong> object corresponding to <strong>the</strong> different moving objects is presented<br />
in Section 3.1. The 3D motion can be reconstructed from a sparse set <strong>of</strong> matched features.<br />
If <strong>the</strong> average disparity between consecutive frames is small, <strong>the</strong> reconstruction can gain in<br />
stability and robustness from <strong>the</strong> time integration <strong>of</strong> long sequences <strong>of</strong> frames. Moreover, in<br />
this case, motion can be dened as a combination <strong>of</strong> small 3D rotations around coordinate<br />
axes and axis and angle. This method is described in Section 3.2. If, on <strong>the</strong> contrary, <strong>the</strong>
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 173<br />
average disparity between frames is large, this problem can be dealt with by using a stereolike<br />
algorithm, applied to a pair <strong>of</strong> frames. The factorization method as a typical example<br />
<strong>of</strong> stereo-like algorithm is described in Section 3.3.<br />
3.1 Segmentation based on <strong>the</strong> Edge Detection<br />
Relax <strong>the</strong> assumption that <strong>the</strong> object motion is described by a single general 3D motion<br />
to deal with <strong>the</strong> problem <strong>of</strong> multiple motions. Identifying moving objects can be seen as<br />
a problem <strong>of</strong> detecting and segmenting objects against a xed background. The goal is to<br />
nd <strong>the</strong> regions <strong>of</strong> <strong>the</strong> volume corresponding to <strong>the</strong> different moving objects. This problem<br />
can be thought <strong>of</strong> as a classication problem. One has to classify <strong>the</strong> voxels <strong>of</strong> each frame<br />
within objects that are detected. The Canny edge detector is, at <strong>the</strong> moment, <strong>the</strong> most widely<br />
used edge detection algorithm in multimedia systems. Constructing a Canny detector<br />
requires <strong>the</strong> formulation <strong>of</strong> a ma<strong>the</strong>matical model <strong>of</strong> <strong>the</strong> edges and <strong>the</strong> corners. The edges<br />
<strong>of</strong> <strong>the</strong> intensity volumes can be modelled according to <strong>the</strong>ir intensity values and proles.<br />
The edge detection operator returns a value for <strong>the</strong> rst derivative in <strong>the</strong> horizontal and <strong>the</strong><br />
vertical direction. The depth dimension (<strong>the</strong> third dimension) can be achieved by modifying<br />
a regular Canny 2D detector [11]. The edge gradient <strong>of</strong> a 3D volumetric object G can be<br />
∂G<br />
determined by computing <strong>the</strong> magnitude gradient components G x =<br />
∂G<br />
∂x<br />
, G y =<br />
∂y<br />
∂G<br />
and G z =<br />
∂z<br />
for each voxel point p(x,y,z) and can be displayed as an volume which<br />
intensity levels are proportional to <strong>the</strong> magnitude <strong>of</strong> <strong>the</strong> local intensity changes. The second<br />
step is to estimate <strong>the</strong> edge strength with<br />
2 2 2<br />
d ( x,<br />
y,<br />
z)<br />
Gx<br />
+ Gy<br />
Gz<br />
e = + as well as <strong>the</strong><br />
orientation <strong>of</strong> <strong>the</strong> edge normal.<br />
Figure 1 Orientation <strong>of</strong> edge normal.<br />
As shown in Figure 1, <strong>the</strong> orientation <strong>of</strong> <strong>the</strong> edge normal is specied by angles and ϕ
174<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
that can be computed by equations<br />
G y<br />
tan θ =<br />
(10)<br />
Gx<br />
and<br />
Gz<br />
tanϕ =<br />
. (11)<br />
( G ) 2 + ( G ) 2<br />
x<br />
y<br />
Weighted summations <strong>of</strong> <strong>the</strong> voxel intensities in local neighbourhoods can be listed as a<br />
numerical array. Intensity gradients which are large are more likely to correspond to edges<br />
than if <strong>the</strong>y are small. Making <strong>the</strong> assumption that <strong>the</strong> important edges should be along<br />
continuous curves in <strong>the</strong> volume allows us to follow a faint section <strong>of</strong> a given line and to<br />
discard a few noisy pixels that do not constitute a line but have produced large gradients.<br />
This marks out <strong>the</strong> edges we can be fairly sure are genuine. Starting from <strong>the</strong>se, using <strong>the</strong><br />
directional information derived earlier, edges can be traced through <strong>the</strong> frames. The output<br />
<strong>of</strong> gradient based edge detection is a binary volume indicating where <strong>the</strong> edges are in order<br />
to decide whe<strong>the</strong>r an edge has been found. From complementary output from <strong>the</strong> edge<br />
tracing step, <strong>the</strong> binary edge map obtained in this way can also be treated as a set <strong>of</strong> edge<br />
curves.<br />
3.2 3D Rotations<br />
This section addresses 3D rotations as a base for reconstruction <strong>of</strong> 3D motions. In most<br />
cases it is useful to represent a rotation by means <strong>of</strong> more natural parameterization. Two<br />
parameterizations are proposed: rotations around <strong>the</strong> coordinate axes and axis and angle.<br />
Rotations around <strong>the</strong> coordinate axes can be expressed as <strong>the</strong> result <strong>of</strong> three consecutive<br />
rotations around <strong>the</strong> coordinate axes, e 1<br />
, e 2<br />
and e 3<br />
, by angles , and respectively. The<br />
angles are <strong>the</strong>n <strong>the</strong> three free parameters <strong>of</strong> linear transformation represented by a 3x3<br />
matrix R and each rotation is expressed as a rotation matrix, R j<br />
, rotating vectors around e j<br />
,<br />
that is,<br />
⎡1<br />
0 0 ⎤<br />
R ( ⎢<br />
⎥<br />
1<br />
α)<br />
= 0 cosα<br />
− sinα<br />
⎢<br />
⎥<br />
(12)<br />
⎢⎣<br />
0 sinα<br />
cosα<br />
⎥⎦<br />
⎡ cos β 0 sin β ⎤<br />
R<br />
⎢<br />
⎥<br />
2( β ) = 0 1 0<br />
(13)<br />
⎢<br />
⎥<br />
⎢⎣<br />
− sin β 0 cos β ⎥⎦<br />
⎡cosγ<br />
− sin γ 0⎤<br />
R =<br />
⎢<br />
⎥<br />
3(<br />
γ ) sin γ cosγ<br />
0 . (14)<br />
⎢<br />
⎥<br />
⎢⎣<br />
0 0 1⎥⎦
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 175<br />
The matrix R describing <strong>the</strong> overall rotation is <strong>the</strong> product <strong>of</strong> <strong>the</strong> R j<br />
(<strong>the</strong> order <strong>of</strong> overall<br />
multiplications matters; different sequences give different results with <strong>the</strong> same triplet <strong>of</strong><br />
angles):<br />
R = R R<br />
1 2R3<br />
=<br />
⎡ cos β cosγ<br />
=<br />
⎢<br />
sinα<br />
sin β cosγ<br />
+ cosα<br />
sinγ<br />
⎢<br />
⎢⎣<br />
− cosα<br />
sin β cosγ<br />
+ sinα<br />
sinγ<br />
− cos β sinγ<br />
− sinα<br />
sin β sinγ<br />
+ cosα<br />
cosγ<br />
cosα<br />
sin β sinγ<br />
+ sinα<br />
cosγ<br />
sin β ⎤<br />
− sinα<br />
cos β<br />
⎥<br />
⎥<br />
cosα<br />
cosγ<br />
⎥⎦<br />
(15)<br />
Second parameterization is as follows. According to Euler’s <strong>the</strong>orem, any 3D rotation<br />
can be described as a rotation by an angle, , around an axis identied by a unit vector<br />
r<br />
n = [ n n n ] T . The corresponding rotation matrix R can <strong>the</strong>n be obtained in terms<br />
1 2 3<br />
<strong>of</strong> and <strong>the</strong> components <strong>of</strong> n which gives a total <strong>of</strong> four parameters. The redundancy <strong>of</strong> this<br />
parameterization (four parameters for three degrees <strong>of</strong> freedom) is eliminated by adding<br />
2 2 2<br />
<strong>the</strong> constraint that n has a unit norm, that is by dividing each n i<br />
by n<br />
1<br />
+ n2<br />
+ n3<br />
[5].<br />
The matrix R in terms <strong>of</strong> and n is given by<br />
2<br />
⎡ n1<br />
n1n2<br />
n1n3<br />
⎤ ⎡ 0 − n3<br />
n2<br />
⎤<br />
⎢<br />
2 ⎥<br />
R = I cosθ + ( 1 − cosθ<br />
)<br />
+<br />
⎢<br />
−<br />
⎥<br />
⎢<br />
n2n1<br />
n2<br />
n2n3⎥<br />
sinθ<br />
n 0<br />
⎢<br />
3<br />
n1⎥<br />
(16)<br />
2<br />
⎢<br />
⎥<br />
⎣n3n1<br />
n3n2<br />
n3<br />
⎦ ⎢⎣<br />
− n2<br />
n1<br />
0 ⎥⎦<br />
Conversely, both and n can be obtained from <strong>the</strong> eigenvalues and <strong>the</strong> eigenvectors <strong>of</strong> R.<br />
The three eigenvalues <strong>of</strong> R are 1, cosθ<br />
+ i sin θ and cosθ<br />
− i sin θ , where i is <strong>the</strong><br />
imaginary unit. The unit vector n is proportional to <strong>the</strong> eigenvector <strong>of</strong> R, corresponding<br />
<strong>the</strong> eigenvalue 1; <strong>the</strong> angle can be obtained from ei<strong>the</strong>r <strong>of</strong> <strong>the</strong> two complex eigenvalues.<br />
3.3 Factorization Method for Motion Reconstruction<br />
Of <strong>the</strong> many methods proposed in <strong>the</strong> literature for <strong>the</strong> case <strong>of</strong> large average disparity<br />
between frames, <strong>the</strong> most used is <strong>the</strong> factorization method, which is simple to implement<br />
and gives very good (and numerically stable) results for objects viewed from ra<strong>the</strong>r large<br />
distances [8]. In this paper it is adopted to three dimensions. The necessary assumption is<br />
that <strong>the</strong> position <strong>of</strong> n volume points corresponding to <strong>the</strong> feature points p 1<br />
, p 2<br />
, ..., p n<br />
, has<br />
been tracked in N frames, with N > 3. This is equivalent to acquiring <strong>the</strong> entire sequence<br />
before starting any processing. Tracking method is described in Section 2.2.2.<br />
If p [ ] T<br />
ij<br />
= xij<br />
, yij<br />
, z denote <strong>the</strong> j-th voxel (j = 1, ..., n) at <strong>the</strong> i-th frame (i = 1, ..., N),<br />
ij<br />
and think <strong>of</strong> <strong>the</strong> x ij<br />
, y ij<br />
and z ij<br />
as entries <strong>of</strong> three N x n matrices, X, Y and Z, respectively, <strong>the</strong><br />
measurement matrix 3N x n can be formed as<br />
⎡X<br />
⎤<br />
W =<br />
⎢<br />
Y<br />
⎥<br />
⎢ ⎥ . (17)<br />
⎢⎣<br />
Z ⎥⎦
176<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
For use in <strong>the</strong> rank <strong>the</strong>orem, <strong>the</strong> mean <strong>of</strong> <strong>the</strong> entries on <strong>the</strong> same row is subtracted from<br />
each x ij<br />
, y ij<br />
and z ij<br />
:<br />
~ xij<br />
= xij<br />
− x<br />
(18)<br />
~ i<br />
yij<br />
= yij<br />
− y<br />
(19)<br />
~ i<br />
(20)<br />
z<br />
ij<br />
= z<br />
ij<br />
− z<br />
i<br />
where<br />
1<br />
x =<br />
n<br />
∑<br />
i<br />
x ij<br />
n j=<br />
1<br />
1<br />
y =<br />
n<br />
∑<br />
i<br />
y ij<br />
n j=<br />
1<br />
(21)<br />
(22)<br />
1<br />
z =<br />
n<br />
∑<br />
i<br />
z ij<br />
n j=<br />
1<br />
are <strong>the</strong> coordinates <strong>of</strong> p<br />
i , <strong>the</strong> centroid <strong>of</strong> <strong>the</strong> voxels in <strong>the</strong> i-th frame. Fur<strong>the</strong>rmore, x~ ,<br />
ij<br />
y~<br />
ij<br />
and z~ can be denoted as entries <strong>of</strong> three N x n matrices, X ~ , Y ~ and Z ~ , and form <strong>the</strong> 3N x<br />
ij<br />
n matrix W, called <strong>the</strong> registered measurement matrix:<br />
(23)<br />
~<br />
⎡X<br />
⎤<br />
~ ⎢ ~ ⎥<br />
W = ⎢Y<br />
⎥<br />
⎢ ~ ⎥<br />
⎣Z<br />
⎦<br />
. (24)<br />
The factorization method is based on <strong>the</strong> pro<strong>of</strong> <strong>of</strong> a simple but fundamental result that <strong>the</strong><br />
registered measurement matrix without noise has at most <strong>the</strong> rank 3 [8]. The pro<strong>of</strong> is based<br />
on <strong>the</strong> decomposition (factorization <strong>of</strong> W ~ ) into <strong>the</strong> product <strong>of</strong> a 3N x 3 matrix, R, and a<br />
3 x n matrix, C. R describes <strong>the</strong> frame-to-frame rotation with respect to <strong>the</strong> points p i<br />
. C<br />
describes <strong>the</strong> structure <strong>of</strong> points (coordinates).<br />
Figure 2 Factorization method.<br />
Consider all quantities expressed in an object-centred reference frame with <strong>the</strong> origin in
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 177<br />
<strong>the</strong> centroid <strong>of</strong> p 1<br />
, ..., p n<br />
(Figure 2), and let i i<br />
, j i<br />
and k i<br />
denote <strong>the</strong> unit vectors <strong>of</strong> <strong>the</strong> i-th<br />
frame (at time instant i). It can be seen from Figure 2 that<br />
x<br />
y<br />
z<br />
ij<br />
ij<br />
ij<br />
T<br />
i<br />
( p − d )<br />
= i<br />
(25)<br />
T<br />
i<br />
j<br />
( p − d )<br />
j<br />
i<br />
= j<br />
(26)<br />
T<br />
i<br />
( p − d )<br />
j<br />
i<br />
i<br />
= k<br />
(27)<br />
where d i<br />
is <strong>the</strong> vector from <strong>the</strong> original reference frame to <strong>the</strong> origin <strong>of</strong> <strong>the</strong> i-th frame;<br />
moreover as <strong>the</strong> origin is in <strong>the</strong> centroid <strong>of</strong> <strong>the</strong> points,<br />
n<br />
1<br />
∑ p j<br />
= 0 . (28)<br />
n 1<br />
j=<br />
Now, it can be rewritten<br />
n<br />
1 T<br />
T<br />
( p − d ) − i ( p − d ) ii<br />
p<br />
j<br />
~ x =<br />
T<br />
ij<br />
= ii<br />
j i ∑ i<br />
n m=1<br />
n<br />
1 T<br />
T<br />
( p<br />
j<br />
− di<br />
) − ji<br />
( pm<br />
− di<br />
) ji<br />
p<br />
j<br />
~ y =<br />
T<br />
ij<br />
= ji<br />
∑<br />
n m=1<br />
n<br />
1 T<br />
T<br />
( p<br />
j<br />
− di<br />
) − ki<br />
( pm<br />
− di<br />
) ki<br />
p<br />
j<br />
~ z =<br />
T<br />
ij<br />
= ki<br />
∑<br />
n m=1<br />
Therefore, <strong>the</strong> rotation matrix R size <strong>of</strong> 3N x 3 can be dened as<br />
m<br />
i<br />
(29)<br />
(30)<br />
. (31)<br />
T<br />
⎡i<br />
⎤<br />
1<br />
⎢ ⎥<br />
⎢ M ⎥<br />
⎢<br />
T<br />
i ⎥<br />
n<br />
⎢ T ⎥<br />
⎢<br />
j1<br />
⎥<br />
R = ⎢ M ⎥<br />
⎢ T ⎥<br />
⎢ jn<br />
⎥<br />
T<br />
⎢k<br />
⎥<br />
1<br />
⎢ ⎥<br />
⎢ M ⎥<br />
⎢<br />
T<br />
⎥<br />
⎣kn<br />
⎦<br />
(32)<br />
and a 3 x n shape matrix C as<br />
[ p K ]<br />
C = (33)<br />
so it can be written that<br />
1<br />
p n<br />
~<br />
W =<br />
RC .
178<br />
4 EXPERIMENTAL EVALUATION<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
This section gives <strong>the</strong> experimental details <strong>of</strong> <strong>the</strong> implemented methods and algorithms<br />
on <strong>the</strong> four-dimensional data sets (3D objects that change in time) collected from an MRI<br />
scanner (i.e. MRI brain volumes). Each dataset is representing <strong>the</strong> brain motion in an MRI<br />
scanner. The framework for <strong>the</strong> estimation <strong>of</strong> different 3D motions is represented in Figure<br />
3. Different types <strong>of</strong> <strong>the</strong> 3D motion are estimated with <strong>the</strong> different techniques. For <strong>the</strong><br />
typical linear motion, such as translation, we propose use <strong>of</strong> <strong>the</strong> constant ow algorithm.<br />
Rotations in space in time are approximated with <strong>the</strong> 3D rotation methods. The optical<br />
ow and <strong>the</strong> feature tracking are proposed for estimation <strong>of</strong> a long sequence <strong>of</strong> data or<br />
accelerating motion. In case <strong>of</strong> large disparity between frames we propose <strong>the</strong> factorization<br />
method. The whole process is semi-automated because a medical expert should know what<br />
kind <strong>of</strong> <strong>the</strong> input data collected from MRI scanner is analyzed, and based on that, <strong>the</strong><br />
system will select an appropriate technique for given 3D motion.<br />
Figure 3 Proposed framework for 3D motion.<br />
Three consecutive frames <strong>of</strong> <strong>the</strong> original experimental dataset (size <strong>of</strong> 429 kB) are<br />
shown in Figure 4. For motion estimation between two frames we use <strong>the</strong> constant<br />
ow algorithm (Figure 4). The resulting displacement between <strong>the</strong> rst and <strong>the</strong><br />
second frame is computed as d = [ ( i − vx<br />
), ( j − vy<br />
)( , k − vz<br />
) ] = [ 153,2,0]<br />
2<br />
2<br />
2<br />
d ( i − v ) + ( j − v ) + ( k − v ) = 153, 013<br />
=<br />
x<br />
y<br />
z<br />
used to estimate <strong>the</strong> movement between <strong>the</strong> second and <strong>the</strong> third frame.<br />
and<br />
. The resulting displacement d is<br />
Figure 4 Example <strong>of</strong> motion between three consecutive frames. Motion between <strong>the</strong> second and <strong>the</strong><br />
third frame is determined by <strong>the</strong> constant ow algorithm.
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 179<br />
Figure 5 Example <strong>of</strong> motion between three consecutive frames. Motion between <strong>the</strong> second and <strong>the</strong><br />
third frame is determined by <strong>the</strong> optical ow.<br />
In Figure 5 <strong>the</strong> same procedure <strong>of</strong> motion estimation between <strong>the</strong> second and <strong>the</strong> third frame<br />
is repeated by using <strong>of</strong> <strong>the</strong> optical ow algorithm. There is a signicant visual difference<br />
between <strong>the</strong> estimated and real position <strong>of</strong> 3D object. Also, in case <strong>of</strong> <strong>the</strong> constant ow<br />
algorithm, we have to code only <strong>the</strong> displacement d which is signicantly less (for three<br />
consecutive frames 151 kB) than coding <strong>the</strong> whole optical ow for <strong>the</strong> each voxel (386<br />
kB).<br />
Figure 6 Motion in long sequence <strong>of</strong> volumes.<br />
Second example is a long sequence <strong>of</strong> volumes (shown in Figure 6). In this case <strong>the</strong><br />
feature tracking is <strong>the</strong> most appropriate technique for <strong>the</strong> motion estimation, because <strong>the</strong><br />
displacement in a long sequence <strong>of</strong> data is not a constant value. In this example feature<br />
tracking is described with p = [ 517 ,38, 92] T<br />
and [ ] T<br />
i<br />
v = − 5,0, 1 , <strong>the</strong> motion on <strong>the</strong><br />
volume plane is described with <strong>the</strong> state vector [ ] T<br />
i<br />
i<br />
s = 517 ,38,92, −5,0,<br />
1 . Estimated
180<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
state vector s<br />
i+ 1 , with assumed = 1<br />
si<br />
1<br />
= 512 ,38,94, −52,6,<br />
−14<br />
Estimated state vectors for consecutive 8 frames are stated below:<br />
ΔT , gives values [ ] T<br />
[ 512 ,38,94, 52,6, ] T<br />
+ 1<br />
= − s = [ 460,44,108, −37,2,<br />
1] T<br />
,<br />
s 14<br />
i<br />
i+<br />
2<br />
−<br />
[ ] T<br />
, s 13<br />
i+ 3<br />
= 423,46,107, −11,<br />
−12,<br />
s [ ] T<br />
,<br />
i+4 = 412 ,34,120,1,0, 2<br />
[ ] T<br />
, s 1<br />
i+ 5<br />
= 413 ,34,122,9,6, − s [ ] T<br />
,<br />
i 6<br />
= 422,40,121,17 ,8, 5<br />
s [ ] T<br />
i+ 7<br />
= 439,48,126,6,24, −2<br />
s [ ] T<br />
,<br />
i 8<br />
= 445,72,124,0,0, 0<br />
+<br />
,<br />
+<br />
.<br />
+<br />
.<br />
To recreate motion on <strong>the</strong> syn<strong>the</strong>sized 3D-models, it is necessary to nd a ma<strong>the</strong>matical<br />
solution to tie <strong>the</strong> analysis to <strong>the</strong> syn<strong>the</strong>sis. To extract motion information from <strong>the</strong> specic<br />
features <strong>of</strong> multiple moving objects, <strong>the</strong> parameters <strong>of</strong> <strong>the</strong> visualization semantics <strong>of</strong> <strong>the</strong><br />
system that will syn<strong>the</strong>size <strong>the</strong> motion, should be known. To do this we rst separate <strong>the</strong><br />
objects with methods described in Section 3.1. Results are shown in Figure 7.<br />
Figure 7 Example <strong>of</strong> segmentation; upper left is <strong>the</strong> original frame; upper right <strong>the</strong> segmented<br />
frame based on <strong>the</strong> edge and <strong>the</strong> corner detection <strong>of</strong> <strong>the</strong> outer shape; lower left all edges detected<br />
based on <strong>the</strong> methods described in Section 3.1; lower right <strong>the</strong> nal segmentation <strong>of</strong> <strong>the</strong> brain with<br />
<strong>the</strong> inner cavities.<br />
It is also necessary to relate <strong>the</strong>se parameters to <strong>the</strong> actions that must be applied to <strong>the</strong><br />
3D-model in order to recreate motion. These motion models translate <strong>the</strong> results into an<br />
object-visualization parameters building motion elds <strong>of</strong> <strong>the</strong> sub-objects in a 3D rotation
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 181<br />
and <strong>the</strong> factorization method for <strong>the</strong> motion reconstruction.<br />
Figure 8 3D rotation.<br />
Example shown in Figure 8 represents a rigid motion eld that is simply described by a<br />
rotation matrix R<br />
R = R R<br />
1 2R3<br />
=<br />
cos( −13°<br />
) cos( 29°<br />
) − cos( −13°<br />
) sin( 29°<br />
) sin( −13°<br />
)<br />
( 8°<br />
) sin( −13°<br />
) cos( 29°<br />
) + cos( 8°<br />
) sin( 29°<br />
) − sin( 8°<br />
) sin( −13°<br />
) sin( 29°<br />
) + cos( 8°<br />
) cos( 29°<br />
) − sin( 8°<br />
) cos( −13°<br />
)<br />
( 8°<br />
) sin( −13°<br />
) cos( 29°<br />
) + sin( 8°<br />
) sin( 29°<br />
) cos( 8°<br />
) sin( −13°<br />
) sin( 29°<br />
) + sin( 8°<br />
) cos( 29°<br />
) cos( 8°<br />
) cos( 29°<br />
)<br />
⎡<br />
⎢<br />
⎢<br />
sin<br />
⎢⎣<br />
− cos<br />
⎡0,1478<br />
− 0,1478<br />
=<br />
⎢<br />
⎢<br />
0,8509 0,8813<br />
⎢⎣<br />
0,2623 − 0,0015<br />
− 0,225 ⎤<br />
− 0,1356<br />
⎥<br />
⎥<br />
0,866 ⎥⎦<br />
⎤<br />
⎥<br />
⎥<br />
⎥⎦<br />
Example <strong>of</strong> <strong>the</strong> factorization method is shown in Figure 9. Aim is to form a registration<br />
measurement matrix. In this example, <strong>the</strong>re is a large average disparity between <strong>the</strong> current<br />
and <strong>the</strong> reference frame. The registered measurement matrix, <strong>the</strong> rotation matrix and <strong>the</strong><br />
shape matrix are<br />
T<br />
⎡−<br />
2,84 ⎤ ⎡i<br />
⎤<br />
1 ⎡−<br />
0,27 − 0,76 0,9 ⎤<br />
⎢ ⎥ ⎢ T ⎥ ⎢<br />
⎥<br />
⎢<br />
−1,13<br />
⎥ ⎢ j1<br />
⎥ ⎢<br />
− 0,73 0,58 − 0,24<br />
⎥<br />
⎢−<br />
2,13 ⎥ ⎢<br />
T<br />
k ⎥<br />
1 ⎢−<br />
0,82 0,24 0,31 ⎥<br />
⎢ ⎥ ⎢ ⎥<br />
T ⎢<br />
⎥<br />
⎢−<br />
3,065⎥<br />
⎢i2<br />
⎥ ⎢−<br />
0,35 − 0,69 0,75 ⎥<br />
⎡ 4 ⎤<br />
~<br />
W = ⎢ − 0,55 ⎥ , ⎢ T<br />
R = j<br />
⎥<br />
= ⎢−<br />
0,69 0,72 − 0,31 ⎥ and C =<br />
⎢ ⎥<br />
⎢ ⎥ ⎢<br />
2<br />
⎥ ⎢<br />
⎥<br />
⎢<br />
3,5<br />
⎥.<br />
T<br />
⎢−<br />
2,875 ⎥ ⎢k<br />
⎥ ⎢−<br />
0,91 0,15 0,24<br />
⎢<br />
2<br />
⎥<br />
⎣ 1 ⎥⎦<br />
⎢−<br />
3,515 ⎥ ⎢<br />
T<br />
⎥ ⎢−<br />
−<br />
⎥<br />
⎢ ⎥ ⎢i<br />
⎥ 0,49 0,61 0,58<br />
3<br />
⎢<br />
⎥<br />
⎢ 0,015 ⎥ ⎢ T<br />
j ⎥ ⎢−<br />
0,59 0,79 − 0,39<br />
3<br />
⎥<br />
⎢ ⎥ ⎢ ⎥<br />
T<br />
⎣−<br />
3,515<br />
⎢<br />
⎥<br />
⎦ ⎢ ⎥ ⎣−<br />
0,97 0,07 0,12<br />
⎣k3<br />
⎦<br />
⎦
182<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
Figure 9 Example <strong>of</strong> factorization method.<br />
The factorization <strong>of</strong> <strong>the</strong> registered measurement matrix W ~ as <strong>the</strong> product <strong>of</strong> R and C<br />
suggests a method for <strong>the</strong> reconstruction <strong>of</strong> structure and motion from a sequence <strong>of</strong><br />
tracked voxels.<br />
5 CONCLUSION<br />
In this paper we have applied different algorithms for <strong>the</strong> 3D motion estimation and<br />
reconstruction. For <strong>the</strong> estimation <strong>of</strong> <strong>the</strong> constant motion eld between only two consecutive<br />
frames, <strong>the</strong> constant ow algorithm is satisfying enough, and it is not computational<br />
extensive as <strong>the</strong> differential techniques because it is oriented only on a sparse set <strong>of</strong> <strong>the</strong><br />
volume points, denoted as <strong>the</strong> feature points. For estimating motion in long sequence <strong>of</strong><br />
3D objects it is more appropriate to use <strong>the</strong> feature tracking. The main purpose <strong>of</strong> this<br />
framework for <strong>the</strong> 3D motion eld estimation in 3D MRI data is to reduce a vast amount <strong>of</strong><br />
acquired data, in coding process. For different types <strong>of</strong> MRI, different motion techniques<br />
should be used for <strong>the</strong> motion estimation and <strong>the</strong> motion reconstruction. By coding only<br />
a sparse set <strong>of</strong> information about <strong>the</strong> volume motion, higher compression ratios can be<br />
achieved. For example, size <strong>of</strong> <strong>the</strong> original dataset is 429 kB, and with coding only <strong>the</strong><br />
reference frame and <strong>the</strong> motion eld based on <strong>the</strong> combination <strong>of</strong> proposed algorithms<br />
for different kind <strong>of</strong> 3D motions, <strong>the</strong> size can be reduced on only 56 kB which gives <strong>the</strong><br />
compression ratio <strong>of</strong> 1 : 8.
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 183<br />
6 REFERENCES<br />
1. Aubert G, Deriche R, Kornprobst P (1999) Computing Optical Flow via Variational<br />
Techniques. SIAM Journal <strong>of</strong> Applied Ma<strong>the</strong>matics, Vol. 60, Issue 1: 156-182<br />
2. Black MJ, Anandan P (1996) The robust estimation <strong>of</strong> multiple motions: parametric<br />
and piecewise-smooth ow elds. Computer Vision and Image Understanding, Vol.<br />
63, Issue 1: 75-104<br />
3. Fashandi H, Fazel-Rezai R, Pistorius S (2007) Optical Flow and Total Least Squares<br />
Solution for Multi-scale Data in an Over-Determined System. Lecture Notes in<br />
Computer Science, Advances in Visual Computing, Springer Berlin/Heidelberg, Vol.<br />
4842: 33-42<br />
4. Favaro P, Soatto S (2007) 3-D Shape Estimation and Image Restoration: Exploiting<br />
Defocus and Motion Blur. Springer-Verlag, London, pp 240-292<br />
5. Furht B, Greenberg J, Westwater R (1997) Motion Estimation Algorithms for Video<br />
Compression. Kluwer Academic Publishers, Boston, pp 120-156<br />
6. Messelodi S, Modena CM, Segata N, Zanin M (2005) A Kalman Filter Based<br />
Background Updating Algorithm Robust to Sharp Illumination Changes. Lecture<br />
Notes in Computer Science, Vol. 3617: 163–170<br />
7. Su Y, Sun MT, Hsu V (2005) Global motion estimation from coarsely sampled motion<br />
vector eld and <strong>the</strong> applications. Circuits and Systems for Video Technology, IEEE<br />
Transactions on, Vol. 15, Issue 2: 232-242<br />
8. Ueshiba T, Tomita F (2000) A factorization method for multiple perspective views via<br />
iterative depth estimation. Systems and Computers in Japan, Vol. 31, Issue 13:87-95<br />
9. Yuille AL, Hallinan PW, Cohen DS (1992) Feature extraction from faces using<br />
deformable templates. International Journal <strong>of</strong> Computer Vision, Springer Ne<strong>the</strong>rlands,<br />
Vol. 8, Issue 2: 99-111<br />
10. Wills J, Agarwal S, Belongie S (2006) A Feature-based Approach for Dense<br />
Segmentation and Estimation <strong>of</strong> Large Disparity Motion. International Journal <strong>of</strong><br />
Computer Vision, Springer Ne<strong>the</strong>rlands, Vol. 68, Issue 2: 125-143<br />
11. Žagar M (2009) 4D Medical Data Compression Architecture. PhD Thesis, University<br />
<strong>of</strong> Zagreb, Zagreb, Croatia
184<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
Laboratory Evaluation <strong>of</strong> Mud Differential Sticking<br />
Tendency and Spotting Fluid Effectiveness<br />
Gaurina-Međimurec Nediljka<br />
University <strong>of</strong> Zagreb, Faculty <strong>of</strong> Mining, Geology and Petroleum Engineering<br />
Pierottijeva 6, Zagreb, Croatia<br />
nediljka.gaurina-medjimurec@rgn.hr, Tel. +385 1 5535 825<br />
Pašic Borivoje<br />
University <strong>of</strong> Zagreb, Faculty <strong>of</strong> Mining, Geology and Petroleum Engineering<br />
Pierottijeva 6, Zagreb, Croatia<br />
borivoje.pasic@rgn.hr, Tel. +385 1 5535 840<br />
Matanović Davorin<br />
University <strong>of</strong> Zagreb, Faculty <strong>of</strong> Mining, Geology and Petroleum Engineering<br />
Pierottijeva 6, Zagreb, Croatia<br />
davorin.matanovic@rgn.hr, Tel. +385 1 5535 835<br />
Abstract<br />
Differential-pressure pipe sticking occurs when a portion <strong>of</strong> <strong>the</strong> drill string becomes<br />
embedded in a lter cake that creates on <strong>the</strong> wall <strong>of</strong> a permeable formation during drilling.<br />
The dominant force is usually associated with a difference in pressures between <strong>the</strong><br />
hydrostatic pressure <strong>of</strong> <strong>the</strong> mud and <strong>the</strong> pore pressure in <strong>the</strong> contact area, though adhesion<br />
and cohesion may also contribute some resistance to pipe movement. Some level <strong>of</strong> sticking<br />
occurs routinely in drilling operations and differential-pressure-pipe-sticking problems<br />
may not be totally prevented. These events only became problematic if <strong>the</strong> force required<br />
to pull <strong>the</strong> pipe free exceeds <strong>the</strong> pipe strength.<br />
If sticking does occur, several methods can be used for freeing <strong>the</strong> stuck pipe but <strong>the</strong> most<br />
common approach is to place a small volume <strong>of</strong> oil, or special spotting uid in a wellbore<br />
annulus to free differentially stuck pipe. Spotting uid reduces <strong>the</strong> stuck area and <strong>the</strong> forces<br />
<strong>of</strong> cohesion between <strong>the</strong> pipe and cake, and allowing pipe to be pulled free.<br />
Differential sticking tendency <strong>of</strong> different drilling uids has been determined in laboratory<br />
using sticking tester as well as inuence <strong>of</strong> spotting uids on freeing differentially stuck<br />
pipe. Results <strong>of</strong> <strong>the</strong> testing are presented in <strong>the</strong> paper.<br />
Key words: differential sticking, stuck pipe, spotting uid, lter cake, drilling<br />
1 INTRODUCTION<br />
Pipe sticking is, for most drilling organizations, <strong>the</strong> greatest drilling problem worldwide.<br />
It results in a signicant amount <strong>of</strong> non-productive time and ends up as one <strong>of</strong> <strong>the</strong> major<br />
causes <strong>of</strong> increased well costs (Reid et al., 2000; Pal et al., 2000). Pipe sticking may result<br />
in abandonment <strong>of</strong> <strong>the</strong> current hole and force a sidetrack, It is estimated that <strong>the</strong> cost <strong>of</strong><br />
stuck pipe in deep oil and gas wells can be approximately 25% <strong>of</strong> <strong>the</strong> overall budget.<br />
In some areas, events related to differentially stuck pipe can be responsible for as much<br />
as 40% <strong>of</strong> <strong>the</strong> total well cost (Reid et al, 2000). Causes <strong>of</strong> stuck pipe can be devided in
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 185<br />
two categories (Isambourg, et al., 1999): (a) mechanical (key seating, formation-related<br />
wellbore instability, wellbore geometry (deviation and ledges), inadequate hole cleaning,<br />
junk in hole or collapsed casing, cement related) and (b) differential pressure (wall sticking).<br />
Differential pressure sticking is usually indicated when <strong>the</strong> drill string cannot be rotated,<br />
raised or lowered, but full circulation at normal pressure can be established (Bushnell-<br />
Watson and Panesar, 1991). The force required to pull <strong>the</strong> pipe free can exceed <strong>the</strong> strength<br />
<strong>of</strong> <strong>the</strong> pipe. Usually, even if <strong>the</strong> stuck condition starts with <strong>the</strong> possibility <strong>of</strong> limited pipe<br />
rotation or vertical movement, it will degrade to <strong>the</strong> inability to move <strong>the</strong> pipe at all.<br />
The risk <strong>of</strong> differentially stuck pipe increases when drilling depleted reservoirs and avoids<br />
when drilling underbalanced. Stuck pipe causes and methods for <strong>the</strong>ir freeing are shown<br />
in Figure 1 (BHI, 1998).<br />
STUCK PIPE CAUSE<br />
MECHANICAL<br />
DIFFERENTIAL STICKING<br />
KEY SEATING<br />
WELLBORE<br />
GEOMETRY<br />
INADEQUATE HOLE<br />
CLEANING<br />
JUNK OR COLLAPSED CASING<br />
CEMENT RELATED<br />
DRILLSTRING<br />
JAMMED<br />
DRILLSTRING<br />
JAMMED<br />
HOLE PACKED<br />
OFF<br />
DRILLSTRING<br />
JAMMED<br />
DRILLSTRING<br />
JAMMED<br />
WORK STRING<br />
DOWN AND<br />
ROTATE<br />
WORK STRING UP IF<br />
RIH.<br />
WORK STRING<br />
DOWN IF P.O.O.H.<br />
WORK STRING<br />
DOWN TO<br />
ESTABLISH OR<br />
IMPROVE<br />
CIRCULATION<br />
WORK STRING UP AND DOWN<br />
WORK STRING UP<br />
AND DOWN. PUMP<br />
ACID IF AVAILABLE<br />
DRILLSTRING<br />
JAMMED<br />
SLUMP STRING AND ROTATE. REDUCE<br />
FLUID WEIGHT. UTILIZE SPOTTING<br />
FLUID<br />
Figure 1. Stuck pipe causes and methods for <strong>the</strong>ir freeing (BHI, 1998.)<br />
2 MECHANISMS OF DIFFERENTIAL STICKING<br />
Differential sticking by denition is a situation in which <strong>the</strong> drilling assembly (pipe, drill<br />
collars and bottomhole assembly) is stuck in lter cake that was previously deposited on a<br />
permeable zone (Figure 2a). A signicant overbalance must also exist and <strong>the</strong> pipe must be<br />
stationary, or almost stationary, to allow <strong>the</strong> bond between <strong>the</strong> cake and <strong>the</strong> pipe to develop.
186<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
The pipe is held in <strong>the</strong> cake by a difference in pressures (P) between <strong>the</strong> hydrostatic<br />
pressure <strong>of</strong> <strong>the</strong> mud (P m<br />
) and <strong>the</strong> formation (pore) pressure (P f<br />
) in <strong>the</strong> permeable zone.<br />
This pressure (P) acts upon <strong>the</strong> area <strong>of</strong> <strong>the</strong> pipe in contact with lter cake and isolated from<br />
<strong>the</strong> pressure <strong>of</strong> <strong>the</strong> mud in <strong>the</strong> hole by new lter cake (Figure 2b). There must be a minimum<br />
penetration <strong>of</strong> <strong>the</strong> drill pipe before a signicant change <strong>of</strong> pressure at pipe/cake interface<br />
can occur. Therefore if lter cake can be made thin enough, sticking can be avoided. Three<br />
differential forces have been used to free <strong>the</strong> pipe: axial (equivalent to working <strong>the</strong> pipe up<br />
or down), radial (equivalent to pulling <strong>the</strong> pipe across <strong>the</strong> wellbore) and torque (rotation <strong>of</strong><br />
<strong>the</strong> pipe) (Figure 2c). In practice it is likely to be a combination <strong>of</strong> an axial force and torque<br />
which is used to free <strong>the</strong> pipe. The force required to free differentially stuck pipe must<br />
overcome adhesion and <strong>the</strong> differential pressure exerted by <strong>the</strong> mud. It has tendency to<br />
increase with time until all water is expelled from <strong>the</strong> lter cake. Once sticking is established,<br />
a signicant force is required to free <strong>the</strong> pipe, even if <strong>the</strong> mud overbalance is removed.<br />
At best, several hours <strong>of</strong> rig time can be spent in various freeing operations. In more<br />
serious cases, <strong>the</strong> pipe cannot be freed and <strong>the</strong> well has to be sidetracked or abondoned.<br />
Pullout force needed to free a stuck pipe is equal to:<br />
F = P·A· (Eq. 1)<br />
where<br />
F - pullout force, N<br />
P - differential pressure, Pa<br />
A - contacta area, m 2<br />
- coefcient <strong>of</strong> friction (adhesion) between <strong>the</strong> collars and <strong>the</strong> cake<br />
Value <strong>of</strong> F is also increased with compressibility and thickness <strong>of</strong> <strong>the</strong> lter cake, hole<br />
deviation, and diameter <strong>of</strong> <strong>the</strong> drill collars. It is decreased with increase in diametar <strong>of</strong> <strong>the</strong><br />
hole. The pullout force is time-dependent since contact area (A) and coefcient <strong>of</strong> friction<br />
() increases with time.<br />
Figure 2. Pipe before stuck (a), stuck pipe (b) and forces to free <strong>the</strong> pipe (c)
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 187<br />
3 DIFFERENTIAL STICKING CAUSES<br />
Differential sticking causes could be: (a) relatively high differential pressure and (b)<br />
mud cake characteristics (thickness, permeability, lubricity). In <strong>the</strong> situations when is not<br />
possible to reduce <strong>the</strong> differential pressure by reducing <strong>the</strong> mud weight <strong>the</strong> option is to act<br />
on <strong>the</strong> mud cake (Outmans, 1974). In situation when pipe is rotating, dynamic lter cake<br />
is formed and drill collar penetrates only a short distance into mud lter cake (Figure 3a)<br />
but when pipe is stationary, static mud lter cake is formed and drill collar is pushed into<br />
mud lter cake by differential pressure (Figure 3b). In highly deviated wellbores, with pipe<br />
being stationary, pressure between mud lter cake and drill collar varies from zero to P<br />
(Figure 3c).<br />
Figure 3. Situation in wellbore with pipe rotating (a) and pipe stationary (b and c)<br />
3.1 Methods <strong>of</strong> reducing <strong>the</strong> risk <strong>of</strong> differential sticking<br />
Differential sticking tendencies <strong>of</strong> mud depend on mud lter cake properties: thickness,<br />
shear strenght, and lubricity. These lter cake properties are inuenced by a combination<br />
<strong>of</strong> variables such as: mud overbalance, solids content <strong>of</strong> <strong>the</strong> mud (both high-gravity and<br />
low-gravity solids), mud type (e.g., oil-based, polymer water-based, gel water-based),<br />
specic mud composition, and uid loss. Early detection <strong>of</strong> differential pressure sticking<br />
risks could be made through observation <strong>of</strong> torque and drag levels while drilling to detect<br />
any sign <strong>of</strong> deviation from a normal trend for <strong>the</strong> well. The objective <strong>of</strong> <strong>the</strong> stuck pipe<br />
avoidance practices is to ensure conditions are maintained at all times that allow pulling<br />
force to exceed sticking force (shear strenght) (Dupriest et al, <strong>2010</strong>). To mitigate differential<br />
pressure sticking events, operators <strong>of</strong>ten (Montgomery et al., 2007, Simon et al., 2005):<br />
• Minimize <strong>the</strong> overbalance (by decreasing mud weight),<br />
• Minimize stationary time (drill string rotates at all times),<br />
• Minimize drilled length through low pressure formations,<br />
• Increase drill collar and drill string stabilization,<br />
• Optimize uid properties in attempts to minimize <strong>the</strong> risk <strong>of</strong> sticking, and<br />
• Select a drilling uid that will yield smooth lter cake with low coefcient <strong>of</strong> friction.<br />
The addition <strong>of</strong> certain lubricants to water- and oil- based muds will reduce <strong>the</strong> risk <strong>of</strong><br />
differential sticking and, should sticking still occur, reduce <strong>the</strong> force needed to free <strong>the</strong><br />
stuck pipe or tool. Lubricants are designed to reduce <strong>the</strong> coefcient <strong>of</strong> friction <strong>of</strong> drilling<br />
uid which decreases torque and drag. Depending on <strong>the</strong>ir chemical composition and state<br />
<strong>of</strong> dispersion or solubility in <strong>the</strong> base mud, lubricants can: (a) coat metal surfaces, reducing
188<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
<strong>the</strong> adhesion <strong>of</strong> steel to <strong>the</strong> mud cake, (b) be incorporated into <strong>the</strong> lter cake and provide<br />
better uid-loss control (resulting in thinner cakes), and (c) be incorporated into <strong>the</strong> lter<br />
cake to reduce <strong>the</strong> yield stress <strong>of</strong> <strong>the</strong> cake.<br />
However, despite <strong>the</strong> best efforts <strong>of</strong> operators a differential pressure sticking event may<br />
still occur (Ayers et al., 1989). The best cure for differential sticking is to prevent it by use<br />
<strong>of</strong> drill-collar stabilizers and, more important, conscientiouslly shortening <strong>the</strong> intervals <strong>of</strong><br />
rest when pipe is opposite permeable formations (Dupriest et al., <strong>2010</strong>).<br />
3.2 Methods <strong>of</strong> freeing <strong>the</strong> stuck pipe<br />
Methods used to get <strong>the</strong> pipe free, in addition to pulling and torquing <strong>the</strong> pipe, include:<br />
(a) lowering hydrostatic pressure in <strong>the</strong> wellbore (by reducing <strong>the</strong> mud weight; this will<br />
reduce <strong>the</strong> differential pressure; should not be used if well control is a problem), (b) placing<br />
a spotting uid next to <strong>the</strong> stuck zone and (c) applying shock force just above <strong>the</strong> stuck<br />
point by mechanical jarring, or (d) all <strong>the</strong> above. The most common approach, however,<br />
to getting pipe free is to pump a chemical spotting uid in a wellbore annulus. Before<br />
placement <strong>of</strong> spotting uid, <strong>the</strong> depth (free point) to where <strong>the</strong> drill string is free and where<br />
sticking starts must be determined. This free point can be calculated using measurements<br />
taken on <strong>the</strong> rig oor. Knowing <strong>the</strong> stretch L and <strong>the</strong> forces applied F 1<br />
and F 2<br />
, Hooke’s<br />
law, <strong>the</strong> length <strong>of</strong> <strong>the</strong> drill string from <strong>the</strong> surface to <strong>the</strong> free point (L f<br />
) is equal to:<br />
(Eq. 2)<br />
where<br />
• E is <strong>the</strong> Elastic Modulus (Young’s Modulus) <strong>of</strong> steel (i.e., 200 GPa),<br />
• A is <strong>the</strong> cross-sectional area <strong>of</strong> <strong>the</strong> pipe body (m 2 ),<br />
• L is <strong>the</strong> stretch distance (elastical stretch <strong>of</strong> <strong>the</strong> free portion <strong>of</strong> <strong>the</strong> drill string (m),<br />
• F1 is <strong>the</strong> force to place <strong>the</strong> entire drill string in tension (N),<br />
• F2 is a force greater than F1 but less that <strong>the</strong> force limited by <strong>the</strong> yield stress <strong>of</strong> <strong>the</strong><br />
pipe grade (N).<br />
3.3 Spotting fluid<br />
A spotting uid (spot) is a small volume or pill <strong>of</strong> any substance, oil or water base, that is<br />
positioned in <strong>the</strong> wellbore to achieve a specic purpose. Most service companies provide<br />
multiple spotting uid options. The purpose <strong>of</strong> <strong>the</strong> spotting uid is to dissolve or break<br />
down <strong>the</strong> lter cake so <strong>the</strong> pipe can be freed. Possible mechanisms <strong>of</strong> spotting uid action<br />
to help free differentially stuck pipe are (Montgomery et al., 2007):<br />
• Breaks <strong>the</strong> capillary forces that hold <strong>the</strong> drillstring against <strong>the</strong> wellbore wall,<br />
• Penetrates, dehydrates and cracks (breaks up) <strong>the</strong> mud lter cake,<br />
• Reduces <strong>the</strong> contact (stuck) area between pipe and wall,<br />
• Reduces <strong>the</strong> forces needed to work <strong>the</strong> pipe free,<br />
• Increases drillstring lubricity throughout stuck zone.<br />
• Allows pipe to be pulled free.<br />
Spotting uids need to be in place as quickly as possible (within six hours after pipe<br />
becoming stuck). Figure 4 shows free drill collar in <strong>the</strong> center <strong>of</strong> hole (a), initial (b) and nal<br />
(c) position <strong>of</strong> drill collar in contact with wall, before spotting uid. When pipe sticking
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 189<br />
occurs, pupming <strong>of</strong> spotting uid has to be as quickly as possible to minimize continued<br />
lter cake buildup which leads to higher contact angle and pull free forces (Fig. 4c).<br />
Mud lter cake shrunk after spotting uid (Fig. 4d) and drill collar is freed (Fig. 4e).<br />
Figure 4. Position <strong>of</strong> drill collar in <strong>the</strong> hole before and after spotting uid<br />
Spotting uids could be devided into four main categories: (a) water-based spotting uids,<br />
(b) diesel-based spotting uids, (c) syn<strong>the</strong>tic-based spotting uids, and (d) acid based<br />
spotting uid. They could be unweighted or weighted (spotting uid, viscosier and<br />
weighting material) and consist <strong>of</strong> detergents, soaps, oils, surfactants and o<strong>the</strong>r chemicals<br />
(wall cake cracking material). Oil-based mud is <strong>the</strong> traditional spotting uid (Krol, 1981;<br />
Ayers et al., 1989). Because <strong>of</strong> concern about mud disposal, spotting uids used <strong>of</strong>fshore<br />
are ei<strong>the</strong>r syn<strong>the</strong>tic-based uids or benign water-based uids (such as drill-in uids<br />
and salt solutions) (Kercheville et al.,1986). Drill-in uids are low and ultra-low solids<br />
uids; <strong>the</strong> sealing mechanism is generated inside <strong>the</strong> rock, so <strong>the</strong>y are leaving just a thin<br />
lm on <strong>the</strong> borehole walls. Salt solutions with a low activity coefcient combined with<br />
environmentally-safe lubricants (two-phase spot) produce low torque levels.<br />
4 LABORATORY RESEARCH<br />
Laboratory tests were carried out to determine <strong>the</strong> effectiveness <strong>of</strong> mud system additives<br />
in minimizing differential pressure sticking <strong>of</strong> drill pipe. Differential sticking tendency <strong>of</strong><br />
<strong>the</strong> tested lignosulfonate mud was evaluated using differential sticking tester marketed by<br />
OFI Testing Equipment International (Figure 5). The test device consists <strong>of</strong> ltration cell<br />
capable <strong>of</strong> holding 200 mL <strong>of</strong> uid, perforated bottom capable <strong>of</strong> holding lter paper and<br />
screen, plate (on a plunger) and torque wrench. Torque necessary to break <strong>the</strong> plate free is<br />
measured. Tests were carried out at room temperature and pressure <strong>of</strong>: 3 292.2 kPa (477.5<br />
PSI). Differential sticking tester measures <strong>the</strong> stuck pipe tendency coefcient <strong>of</strong> drilling<br />
uids, and also determines how effective lubricants or treatments might be with any given<br />
drilling uid.
190<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
Figure 5. OFITE Differential sticking tester<br />
This coefcient takes into account both <strong>the</strong> friction, or “stickiness”, <strong>of</strong> <strong>the</strong> lter cake, as<br />
well as <strong>the</strong> amount <strong>of</strong> cake building that must occur in order to freeze or stick <strong>the</strong> pipe in <strong>the</strong><br />
hole. By measuring <strong>the</strong> area <strong>of</strong> cake building during a test, <strong>the</strong> bulk sticking coefcient is<br />
obtained and read directly at <strong>the</strong> end <strong>of</strong> <strong>the</strong> test. How likely a given uid will be to produce<br />
a stuck pipe situation and how effective a given treatment may be, can be immediately<br />
determined on-site. The bulk sticking coefcient (K sc<br />
) is calculated by dividing <strong>the</strong> sliding<br />
force (F s<br />
) by <strong>the</strong> normal force (F n<br />
):<br />
K sc<br />
= F s<br />
/ F n<br />
(Eq. 3)<br />
For radius <strong>of</strong> plate r = 25.4 mm (1”):<br />
The sliding force (F s<br />
) <strong>of</strong> <strong>the</strong> plate is a function <strong>of</strong> <strong>the</strong> measured torque (T u<br />
).<br />
F s<br />
= 1.5 × T u<br />
(Eq. 4)<br />
The normal force (F n<br />
) on <strong>the</strong> plate is determined by multiplying <strong>the</strong> area by <strong>the</strong> differential<br />
pressure (This assumes that a pressure <strong>of</strong> 477.5 PSI was used during <strong>the</strong> test) (Eq. 5)<br />
Fn = 1 500 × r 2 (Eq. 5)<br />
Finally, <strong>the</strong> bulk sticking coefcient (Ksc) is equal to:<br />
K sc<br />
= 0.001 × T n<br />
(Eq. 6)
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 191<br />
Where T u<br />
is torque reading (lb f<br />
-inch or 0.1129 N·m).<br />
The stuck tendency coefcient (K st<br />
) is equal to <strong>the</strong> bulk sticking coefcient (K sc<br />
) multiplied<br />
by <strong>the</strong> variable stuck area (Eq. 7):<br />
K st<br />
= K sc<br />
× (Variable Stuck Area) (Eq. 7)<br />
Laboratory tests were run to evaluate <strong>the</strong> effectiveness <strong>of</strong> mud system additives:<br />
Carboxymethylcellulose - CMC (ltration control additive) and lubricant on differential<br />
sticking tendency <strong>of</strong> <strong>the</strong> tested uids (Gaurina-Meimurec et al., <strong>2010</strong>.). Formulations<br />
<strong>of</strong> tested lignosulphonate mud is shown in Table 1. To determine inuence <strong>of</strong> CMC and<br />
lubricant on mud properties, especially on differential sticking tendency, amount <strong>of</strong> CMC<br />
and lubricant (Lube 167) in base lignosulphonate mud was varied. The numbers included<br />
in <strong>the</strong> mud sign implies CMC concentration in grams per 1 L <strong>of</strong> <strong>the</strong> mud (C0, C3, C5)<br />
and lubricant concentration in ml per 1 L <strong>of</strong> <strong>the</strong> mud (L0, L20, L40). Mud properties such<br />
as API uid loss, cake thickness, rheological properties and pH value were determined<br />
according to API RP13B (API Recommended Practice Standard Procedure for Testing<br />
Drilling Fluids). Spotting uid was composed <strong>of</strong> 620 ml Diesel oil, 80 ml Pipe Lax W, 280<br />
ml water and 73 g barite. Spotting time was 16 hours.<br />
Composition Units C0L0 C3L0 C5L0 C0L20 C3L20 C5L20 C0L40 C3L40 C5L40<br />
Water ml 1000 1000 1000 1000 1000 1000 1000 1000 1000<br />
Bentonite g L -1 80 80 80 80 80 80 80 80 80<br />
FCL g L -1 20 20 20 20 20 20 20 20 20<br />
NaOH g L -1 3 3 3 3 3 3 3 3 3<br />
CMC g L -1 0 3 5 0 3 5 0 3 5<br />
Viscosifer g L -1 5 5 5 5 5 5 5 5 5<br />
Biocide g L -1 0,6 0,6 0,6 0,6 0,6 0,6 0,6 0,6 0,6<br />
Lubricant ml L -1 0 0 0 20 20 20 40 40 40<br />
Defoamer ml L -1 1 1 1 1 1 1 1 1 1<br />
Barite g L -1 500 500 500 500 500 500 500 500 500<br />
Table 1. Formulation <strong>of</strong> <strong>the</strong> tested lignosulphonate mud<br />
Laboratory results are shown in gures from 6 to 15. Torque increases with time regardless<br />
<strong>of</strong> concentration <strong>of</strong> CMC and lubricant, but for specied time decreases with increasing<br />
concentration <strong>of</strong> CMC and lubricant. In addition, torque decreases after placement <strong>of</strong><br />
spotting uid for16-hours (Fig. 6).
192<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
Figure 6. Effect <strong>of</strong> time and additives on torque<br />
Bulk sticking coefcient after 300 min sticking time is decreased with increasing<br />
concentration <strong>of</strong> CMC and lubricant. For instance, <strong>the</strong> value <strong>of</strong> bulk sticking coefcient <strong>of</strong><br />
C5L40 mud is 35,3 % lesser than <strong>of</strong> C0L0 mud, and for C5L40 bulk sticking coefcient is<br />
2,3 times less after 16 hours spotting time (Fig. 7).<br />
Figure 7. Effect <strong>of</strong> CMC and lubricant on bulk sticking coefcient<br />
Bulk sticking coefcient is decreased with increasing concentration <strong>of</strong> CMC but increased<br />
with time regardless <strong>of</strong> concentration <strong>of</strong> CMC (mud without lubricant) (Fig 8).
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 193<br />
Figure 8. Effect <strong>of</strong> time and concentration <strong>of</strong> CMC on bulk sticking coefcient<br />
Bulk sticking coefcient <strong>of</strong> mud without lubricant and with 5 g CMC after 300 min is 9,2<br />
% less than without CMC but 2,4 times higher than after 60 min (Fig. 9). The bulk sticking<br />
coefcient <strong>of</strong> mud with 5 g CMC and with 4% lubricant after 300 min test is 29 % less than<br />
without lubricant.<br />
Figure 9. Effect <strong>of</strong> lubricant on bulk sticking coefcient<br />
The corelation between torque after 60 min and mud properties (uid loss, cake thicknes,<br />
plastic viscosity, gel strengts and pH value) is observed. The lower uid loss value <strong>the</strong><br />
lower value <strong>of</strong> torque (Fig. 10). In addition, API Fluid loss decreases with addition <strong>of</strong> CMC
194<br />
and lubricant.<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
Figure 10. The corelation between torque and uid loss <strong>of</strong> tested muds<br />
Cake thickness and torque decreases with increasing concentration <strong>of</strong> CMC and lubricant.<br />
The thiner cake <strong>the</strong> lower torque (Fig. 11).<br />
Figure 11. The corelation between torque and cake thickness <strong>of</strong> tested muds<br />
Plastic viscosity and yield point increase with increasing concentration <strong>of</strong> CMC and
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 195<br />
lubricant (Fig. 12 and 13).<br />
Figure 12. The corelation between torque and plastic viscosity <strong>of</strong> tested muds<br />
Figure 13. The corelation between torque yield point <strong>of</strong> tested muds<br />
Gel strenght increases and pH value decreases with increasing concentration <strong>of</strong> CMC and<br />
lubricant (Fig. 14 and 15).
196<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
Figure 14. The corelation between torque and gel strenght <strong>of</strong> tested muds<br />
Figure 15. The corelation between torque and pH value <strong>of</strong> tested muds<br />
5 CONCLUSIONS<br />
The sticking tendency <strong>of</strong> mud can be estimated using differential sticking tester. Use <strong>of</strong><br />
<strong>the</strong> test at <strong>the</strong> well-site allows <strong>the</strong> mud engineer to identify <strong>the</strong> problem before <strong>the</strong> sticking
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 197<br />
occurs, and to select and implement <strong>the</strong> most effective treatment option.<br />
According to performed tests torque and bulk sticking coecient constantly increase with<br />
sticking time regardless <strong>of</strong> concentration <strong>of</strong> CMC and lubricant in tested lignosulphonate<br />
mud, but for any specied sticking time (60 min, 120 min, 180 min, 240 min or 300 min)<br />
<strong>the</strong>ir values decrease with increasing concentration <strong>of</strong> CMC and lubricant. After placement<br />
<strong>of</strong> spotting uid for16-hours torque signicantly decreases. In addition, <strong>the</strong> corelation between<br />
mud composition, torque and mud properties is observed. API uid loss, lter cake<br />
thickness and pH value decrease but plastic viscosity, yield point and gel strenght increase<br />
with increasing concentration <strong>of</strong> CMC and lubricant. Fluid loss and lter cake thickness<br />
have a direct inuence on values <strong>of</strong> <strong>the</strong> torque and sticking coefcient (reducing sticking<br />
area). The lower uid loss and <strong>the</strong> thiner cake <strong>the</strong> lower values <strong>of</strong> <strong>the</strong> torque and <strong>the</strong> bulk<br />
sticking coefcient. Therefore using <strong>the</strong> lignosulphonate mud signed C5L40 with 5 gL-1<br />
CMC and 40 mlL-1 lubricant sticking pipe can be avoided or if it still occurs <strong>the</strong> force<br />
required to free differentially stuck pipe will be less.<br />
6 REFERENCES<br />
1. Ayers, R.C., O´Reilly, J.E., Henry, L.R. (1989.) Offshore Operators Committee Gulf<br />
<strong>of</strong> Mexico Spotting Fluid Survey, paper SPE/IADC 18683, presented at <strong>the</strong> 1989 SPE/<br />
IADC Drilling Conference, New Orleans, Louisiana, February 28-March 3, 517-519.<br />
2. Bushnell-Watson, Y.M., Panesar, S.S. (1991) Differential Sticking Laboratory Tasts<br />
Can Improve Mud Design, paper SPE 22549 presented at <strong>the</strong> 66th Annual Technical<br />
Conference and Exhibition <strong>of</strong> <strong>the</strong> Society <strong>of</strong> Petroleum Engineers, Dallas, TX,<br />
October 6-9.147-156.<br />
3. Dupriest, F.E., Elks Jr., W.C., Ottesen, S. (<strong>2010</strong>.) Design Methodology and Operational<br />
Practices Eliminate Differential Sticking, paper IADC/SPE 128129, presented at<br />
<strong>the</strong> <strong>2010</strong> IADC/SPE Drilling Conference and Exhibition, New Orleans, Louisiana,<br />
February 2-4. 5-13.<br />
4. Gaurina-Meimurec, N., Paši, B., Matanovi, D. (<strong>2010</strong>.) Spotting Fluids for<br />
Freeing Differentially Stuck Pipe, The 1st Annual Congress <strong>of</strong> Oil Field Chemicals,<br />
Conference Proceedings, Beijing, China, November 15-17, 57.<br />
5. Isambourg, P., Ottesen, S., Benaissa, S., Marti, J. (1999.) Down-Hole Simulation Cell<br />
for Measurement <strong>of</strong> Lubricity and Differential Pressure Sticking, paper SPE/IADC<br />
52816, presented at <strong>the</strong> 1999 SPE/IADC Drilling Conference, Amsterdam, Holland,<br />
March 9-11, 1-12.<br />
6. Kercheville, J.D., Hinds, A.A., Clements, W.R. (1986.) Comparison <strong>of</strong> Environmentally<br />
Acceptable Materials with Diesel Oil for Drilling Mud Lubricity and Spotting Fluid<br />
Formulations, paper IADC/SPE 14797, presented at <strong>the</strong> 1986 IADC/SPE Drilling<br />
Conference, Dallas, Texas, February 10-12, 611-615.<br />
7. Krol, A.D. (1981.) Laboratory Evaluation <strong>of</strong> Stuck Pipe Spotting Fluid Effectiveness,<br />
paper SPE 10096, presented at <strong>the</strong> 56th Annual Fall Technical Conference and<br />
Exhibition <strong>of</strong> Society <strong>of</strong> Petroleum Engineers <strong>of</strong> AIME, San Antonio, Texas, October<br />
5-7.1-13.<br />
8. Montgomery, J.K., Keller, S.R., Krahel, N., Smith, M.V. (2007) Improved Method for<br />
Use <strong>of</strong> Chelation to Free Stuck Pipe and Enhance Treatment <strong>of</strong> Lost Returns, paper<br />
SPE/IADC 105567 presented at <strong>the</strong> 2007 SPE/IADC Drilling Conference,Amsterdam,<br />
The Ne<strong>the</strong>rlands, 20-22 February, pp 1-7.<br />
9. Outmans, H.D. (1974.) Spot Fluid Quickly to Free Differentially Stuck Pipe, The Oil<br />
and Gas Journal, July 15, 65-68.<br />
10. Pal, K., Mangla, V.K., Joshi, N.P., Tewari, P.P., Bose, U.N. (2000.) A New Approach
198<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
Solves Differential Sticking Problems in Geleki and Ahmedabad Fields, paper SPE<br />
63055, presented at <strong>the</strong> 2000 SPE Annual Technical Conference and Exhibition,<br />
Dallas, Texas, October 1-4.1-8.<br />
11. Reid, P.I., Meeten, G.H., Way, P.W., Clark, P., Chambers, B.D., Gilmour, A., Sanders,<br />
M.W. (2000.) Differential-Sticking Mechanisms and a Simple Wellsite Test for<br />
Monitoring and Optimizing Drilling Mud Properties, SPE Drilling & Completion,<br />
Vol. 15, No. 2, June, 97-104.<br />
12. Simon, K., Gaurina-Meimurec, N.; Paši, B. (2005) Drilling Fluids Differential<br />
Sticking Tendency Determination, Rudarsko-geološko-naftni zbornik, Vol.17, Faculty<br />
<strong>of</strong> Mining, Geology and Petroleum Engineering, Zagreb, 31-35.<br />
13. ***, (1998) FLUID Facts – Engineering Handbook, Baker Hughes Inteq, Houston,<br />
TX, March 1998.
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 199<br />
Use <strong>of</strong> Plasma Technology for Modification <strong>of</strong> Textiles<br />
Pr<strong>of</strong>. dr. Ružica Čunko<br />
Faculty <strong>of</strong> Textile Technology, University <strong>of</strong> Zagreb<br />
Prilaz baruna Filipovića 28a, Zagreb<br />
rcunko@ttf.hr<br />
tel. 01/3712-523<br />
Dr. Sanja Ercegović Ražić<br />
Faculty <strong>of</strong> Textile Technology, Univesity <strong>of</strong> Zagreb<br />
Prilaz baruna Filipovića 28a, Zagreb<br />
Abstract<br />
The use <strong>of</strong> non-<strong>the</strong>rmal, oxygen and argon low-presser plasmas for modication <strong>of</strong> cellulose<br />
textiles is discussed. The emphasis is given to <strong>the</strong> characterization <strong>of</strong> <strong>the</strong> complex micromorphology<br />
and physical-chemical changes <strong>of</strong> <strong>the</strong> textile surface after plasma treatments,<br />
as well as on dening <strong>the</strong> impacts <strong>of</strong> <strong>the</strong>se changes on <strong>the</strong> obtained nal properties <strong>of</strong><br />
textile materials treated in this way. SEM and AFM results conrmed <strong>the</strong> <strong>the</strong>sis that using<br />
oxygen and argon plasmas, two, essentially different, processes <strong>of</strong> textile surface ablation<br />
occurred; <strong>the</strong> rst is chemical etching and <strong>the</strong> second physical sputtering. Using <strong>the</strong> AFM,<br />
<strong>the</strong> bre surface morphology was estimated at nanoscale. The measurements <strong>of</strong> <strong>the</strong> vertical<br />
rise <strong>of</strong> water in fabric sample and <strong>the</strong> results <strong>of</strong> <strong>the</strong> wetting time (vertical test and drop<br />
test) indicate <strong>the</strong> improvement <strong>of</strong> hydrophilic properties <strong>of</strong> all tested samples after lowpressure<br />
oxygen plasma treatment. The presented results conrm that <strong>the</strong> low-pressure<br />
plasma treatment is an acceptable and appropriate method <strong>of</strong> textile surface modication<br />
and an eco-friendly method at <strong>the</strong> same time.<br />
Keywords: low-pressure plasma, surface functionalization, cellulose fabric modication.<br />
1 Introduction<br />
The textile and clothing industries in Europe, USA, Japan and some o<strong>the</strong>r developed<br />
countries are facing some big challenges today, largely because <strong>of</strong> <strong>the</strong> globalization process.<br />
Therefore, <strong>the</strong> shift to high- and multifunctional, added value, luxury and technical textiles<br />
is deemed to be essential for <strong>the</strong>ir sustainable growth. The growing environmental and<br />
energy-saving conditions will also lead to <strong>the</strong> gradual replacement <strong>of</strong> many traditional wet<br />
chemistry-based textile processing, using large amounts <strong>of</strong> water, energy and efuents,<br />
by various forms <strong>of</strong> low-liquor and dry-nishing processes. Plasma technology, when<br />
developed at a commercially viable level, has strong potential to <strong>of</strong>fer in an attractive way<br />
achievement <strong>of</strong> new functionalities in textiles.<br />
The signicant research work on introduction <strong>of</strong> plasma technology in <strong>the</strong> textile eld has<br />
been going since <strong>the</strong> early 1980s and recently <strong>the</strong> application <strong>of</strong> plasma treatments for<br />
textile modication has become more and more signicant. In many laboratories across<br />
<strong>the</strong> world <strong>the</strong> researches are dealing with plasma treatments <strong>of</strong> variety <strong>of</strong> brous materials<br />
in order to improve <strong>the</strong>ir functional properties. A variety <strong>of</strong> low-pressure and atmospheric<br />
plasma machines, mostly in prototype form, have been <strong>of</strong>fered for industrial processing <strong>of</strong>
200<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
textiles. In recent years, considerable efforts have been made by many plasma technology<br />
suppliers, as well as researches, to impart broad range <strong>of</strong> new functionalities and textile<br />
properties by plasma treatment [2, 7, 10, 23, 24, 27, 28].<br />
In <strong>the</strong> recent point <strong>of</strong> view, a desired ber surface characteristics and textiles modication<br />
by plasma seem very promising. Plasma treatments can be used both in substitutions<br />
<strong>of</strong> conventional processes and for <strong>the</strong> production <strong>of</strong> innovative textile materials with<br />
properties that cannot be achieved via wet processing. They are applicable to different<br />
textile substrates, even to those that cannot be modied by conventional methods. It is<br />
important to note, that plasma agency is limited to <strong>the</strong> top surface layer and no signicant<br />
alteration <strong>of</strong> bulk properties <strong>of</strong> <strong>the</strong> ber is produced. Plasma treatments are fast and<br />
extremely gentle, as well as environmentally friendly, being dry processes characterized<br />
by low consumption <strong>of</strong> chemicals and energy [3, 15]. If plasmas are used as pre-treatment,<br />
<strong>the</strong>y can reduce <strong>the</strong> amount <strong>of</strong> chemicals required by <strong>the</strong> conventional process and <strong>the</strong><br />
concentration <strong>of</strong> pollutants in <strong>the</strong> efuents.<br />
The rst investigation <strong>of</strong> possibilities and phenomenology <strong>of</strong> textiles modication by<br />
plasma treatment in Croatia was carried out within <strong>the</strong> scope <strong>of</strong> <strong>the</strong> research project<br />
Multifunctional textile materials for personal protection (Faculty <strong>of</strong> Textile Technology,<br />
Department <strong>of</strong> materials, bres and textile testing, 1997-<strong>2011</strong>). The results <strong>of</strong> a part <strong>of</strong> this<br />
research are presented in this paper.<br />
2 Possibilities and phenomenology <strong>of</strong> cellulose textiles modification<br />
by plasma treatment<br />
2.1 What is plasma?<br />
Plasma was rst identied (as “radiant matter”) by Sir W. Crookes in 1879; Sir J. J. Thomson<br />
identied <strong>the</strong> nature <strong>of</strong> <strong>the</strong> matter in 1897, and in 1928. I. Langmuir has introduced <strong>the</strong><br />
term “plasma”. In physics plasma is dened as distinct phase <strong>of</strong> matter, separate from<br />
<strong>the</strong> traditional solids, liquids, and gases. It comprises a dynamic mix <strong>of</strong> ions, electrons,<br />
neutrons, photons, free radicals, meta-stable excited species and molecular fragments.<br />
Since <strong>the</strong> particles in plasma are electrically charged (generally by being stripped <strong>of</strong><br />
electrons), it is frequently described as an “ionized gas.” Energy is needed to strip electrons<br />
from atoms to make plasma. The energy can be <strong>of</strong> various origins: <strong>the</strong>rmal, electrical, or<br />
light (ultraviolet light or intense visible light from a laser). With insufcient sustaining<br />
power, plasmas recombine into neutral gas.<br />
Plasma is electrically conductive and can be manipulated by magnetic elds. It is odd to<br />
consider that plasma is actually <strong>the</strong> most common phase <strong>of</strong> matter, especially since it was<br />
<strong>the</strong> last one discovered. Flame, lightning, interstellar nebulae, stars, and even <strong>the</strong> empty<br />
vastness <strong>of</strong> space are all examples <strong>of</strong> <strong>the</strong> plasma state <strong>of</strong> matter.
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 201<br />
Figure 1 Plasma lamp [13]<br />
The colors are a result <strong>of</strong> relaxation <strong>of</strong> electrons in excited states to lower energy states<br />
after <strong>the</strong>y have recombined with ions. These processes emit light in<br />
a spectrum characteristic <strong>of</strong> <strong>the</strong> gas being excited.<br />
2.2 Principles <strong>of</strong> plasma processes and application <strong>of</strong> plasma<br />
in <strong>the</strong> field <strong>of</strong> textiles<br />
Plasmas are generally classied as <strong>the</strong>rmal and non-<strong>the</strong>rmal. Thermal (“hot”) plasmas are<br />
characterized by a condition <strong>of</strong> <strong>the</strong>rmal equilibrium between different species contained<br />
in <strong>the</strong> gas and <strong>the</strong>ir temperature raise <strong>of</strong> several thousand degrees. It is clear that <strong>the</strong>se<br />
plasmas due to <strong>the</strong>ir destructive nature, are not suitable for textile treatment.<br />
Contrary to <strong>the</strong>rmal plasmas, non-<strong>the</strong>rmal plasmas are “cold” plasmas and are produced<br />
at room temperature or little above room temperature. In this case, electrons have higher<br />
energies than ions and molecules (<strong>the</strong>ir energies ranging from 0.1 to some electron volts),<br />
and, due to <strong>the</strong> low density <strong>of</strong> <strong>the</strong> gas, collisions with <strong>the</strong> o<strong>the</strong>r species are relatively rare<br />
and <strong>the</strong>rmal equilibrium is not reached. The bulk temperature <strong>of</strong> <strong>the</strong> gas is comparable to<br />
room temperature. Electron collisions with neutral species produce additional electrons<br />
and ions. Thanks to <strong>the</strong> low operating temperatures, cold plasmas are suitable for textile<br />
treatment. The most important for this purpose are low-pressure plasma and atmospheric<br />
pressure plasma [11, 17, 20, 22, 23].<br />
Using <strong>of</strong> high vacuum pumps, <strong>the</strong> low-pressure plasma reactors are working in <strong>the</strong> pressure<br />
range <strong>of</strong> 10 -2 to 10 -3 mbar (vacuum vessel). The gas which is <strong>the</strong>n introduced in <strong>the</strong> vessel<br />
is ionized with <strong>the</strong> help <strong>of</strong> high frequency generator. For radi<strong>of</strong>requency range (typically<br />
40 KHz or 13, 56 MHz), normally <strong>the</strong> working gas pressure is kept in <strong>the</strong> lower 0.1 mbar<br />
range, whereas for micro-wave sources (915 MHz or 2.45 GHz), a working pressure<br />
between 0.5 and 1 mbar is <strong>of</strong>ten used.<br />
The interaction between <strong>the</strong> very active chemical species and photons present in <strong>the</strong> plasma<br />
gas and a textile substrate is <strong>the</strong> basis <strong>of</strong> dissimilar industrial applications. As a consequence<br />
<strong>of</strong> <strong>the</strong> very complex and non-equilibrium nature <strong>of</strong> cold plasmas, a multiplicity <strong>of</strong> very<br />
different phenomena can occur, depending on <strong>the</strong> art <strong>of</strong> <strong>the</strong> gas and <strong>the</strong> operating conditions<br />
[11, 12, 17, 18, 28]. The most important plasma/textile interaction processes are: cleaning<br />
or etching, surface activation, grafting, polymerization and deposition (Fig. 2). All <strong>the</strong>se<br />
phenomena occur at <strong>the</strong> bre surface and are limited to <strong>the</strong> most external layer <strong>of</strong> <strong>the</strong><br />
substrate. Normally, <strong>the</strong> effects do not involve layers deeper than 10 -100 nm.
202<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
Figure 2 Cold plasma processes [19]<br />
2.3 Mechanisms <strong>of</strong> interactions between plasmas and cellulose fibres<br />
Cellulose is <strong>the</strong> most widely known brous polysaccharides and <strong>the</strong> basic building polymer<br />
<strong>of</strong> agricultural bres (cotton, ax, hemp) as well as <strong>of</strong> various man-made bres (viscose,<br />
modal, lyocell, bamboo). The basic structural repeating unit for cellulose is <strong>the</strong> cellobiose<br />
unit, composed <strong>of</strong> two anhydroglucose rings e.g. anhydroglucose units. Each unit contains<br />
three alcohol hydroxyls groups (-OH). These hydroxyls form hydrogen bonds inside<br />
<strong>the</strong> macromolecule itself (intra-molecular) and between o<strong>the</strong>r cellulose macromolecules<br />
(intermolecular). Cellulose in bres is highly crystalline polymer with tight packing <strong>of</strong><br />
ordered chains wherein <strong>the</strong> hydrogen bonding is very strong. Cellulose bre also contains<br />
amorphous regions in which <strong>the</strong>re can be a variation from complete disorder <strong>of</strong> <strong>the</strong><br />
cellulose chains to some order. Owing to <strong>the</strong> hydroxyl groups, cellulose can be traditionally<br />
modied through reactions <strong>of</strong> esterication and e<strong>the</strong>rication. It has been determined that<br />
<strong>the</strong> accessibility <strong>of</strong> <strong>the</strong> -OH groups is controlled by <strong>the</strong> relative ratios <strong>of</strong> crystalline and<br />
amorphous regions <strong>of</strong> <strong>the</strong> cellulose bre superstructure. Amorphous cellulose can be<br />
functionalized much more readily than <strong>the</strong>ir crystalline counterparts.<br />
Cellulose modication using cold plasma differs signicantly from <strong>the</strong> traditional<br />
esterication and e<strong>the</strong>rication reaction. Reactive plasma gas components (particles)<br />
can react directly with <strong>the</strong> anhydroglucose units by abstraction <strong>of</strong> hydrogen atoms from<br />
ei<strong>the</strong>r carbon or oxygen atoms in <strong>the</strong> plasma reactor. Thus, cellulose can be oxidized,<br />
reduced and/or substituted in new ways. XPS data collected from plasma treated cellulose<br />
samples indicate that new functionalities are formed compared with unmodied cellulose.<br />
In addition to C-OH and C-O-C, <strong>the</strong> existence <strong>of</strong> O=C=O, O-CO-O, O=C-O, C=O and<br />
O-CO-O functionalities are detected [6, 27].<br />
The interaction <strong>of</strong> <strong>the</strong> active species <strong>of</strong> plasma with cellulose in bre surface, involves<br />
several electron-mediated processes, as well as positive ion-induced reactions. Some <strong>of</strong><br />
<strong>the</strong>se reactions can promote hemolytic cleavages leading to <strong>the</strong> formation <strong>of</strong> free-radical
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 203<br />
sites. These reactive centers can lead to a variety <strong>of</strong> functionalization mechanisms.<br />
Cellulose can be grafted by monomer introduced in plasma gas and <strong>the</strong> grafting process<br />
is restricted to <strong>the</strong> surface layer [2, 4, 12, 21, 26]. By using specic molecules, a process<br />
known as plasma-enhanced chemical vapor deposition (PE-CVD) may occur, as well as<br />
metal nanoparticles can be deposed on bre surface using plasma gas. All <strong>the</strong>se processes<br />
are by-product free. The reactions are not accompanied by <strong>the</strong> deposition <strong>of</strong> undesired<br />
macromolecular layers on <strong>the</strong> reactor walls. Plasma treatments usually do not create<br />
signicant environmental problems.<br />
3 Experimental, materials and methods<br />
The emphasis in this investigation is given to <strong>the</strong> characterization <strong>of</strong> <strong>the</strong> complex micromorphology<br />
and physical-chemical changes <strong>of</strong> <strong>the</strong> textile surface after plasma treatments,<br />
as well as on dening <strong>the</strong> impacts <strong>of</strong> <strong>the</strong>se changes on <strong>the</strong> obtained nal properties <strong>of</strong><br />
textile materials treated in this way.<br />
3.1 Materials<br />
The textile substrate used was <strong>the</strong> cellulose fabrics woven in plain wave, using threads<br />
spun from lyocell and modal man-made cellulose bres and from cotton. The structural<br />
characteristics <strong>of</strong> sample fabrics are presented in Table 1. Before plasma treatment, <strong>the</strong> raw<br />
cotton fabric was undergoes to desizing and scouring treatments according to <strong>the</strong> industrial<br />
process conditions (sample CO sc).<br />
Table 1: Basic characteristic <strong>of</strong> tested fabric samples woven in plain weave<br />
Samples<br />
Yarn<br />
neness<br />
(warp/<br />
weft) / tex<br />
Numb. <strong>of</strong> threads<br />
(warp/weft) / cm<br />
Volume mass / gcm -3<br />
Mass per unit area /<br />
gm -2<br />
UT O 2<br />
-p Ar-p UT O 2<br />
-p Ar-p<br />
Cotton<br />
(CO sc)<br />
Lyocell<br />
(CLY I)<br />
Modal<br />
(CMD I)<br />
29.4/25.0 25/20 0.369 0.367 0.362 130.6 128.1 128.3<br />
31.3/25.0<br />
21/19<br />
0.409 0.431 0.444 120.1 126.8 129.8<br />
31.3/25.0 23/19 0.420 0.429 0.444 121.8 123.9 126.0<br />
UT = untreated sample, O 2<br />
-p = oxygen plasma treated, Ar-p = argon plasma treated sample<br />
3.2 Plasma treatment <strong>of</strong> fabric samples<br />
Plasma system<br />
Low-pressure (LP) - plasma NANO LF laboratory system (Diener Electronic GmbH,<br />
Germany, Fig 3) <strong>of</strong> 40 kHz low-frequency generator and maximum power <strong>of</strong> 300 W<br />
was used for fabric treatments. One electrode and <strong>the</strong> four trays were placed inside <strong>the</strong><br />
cylindrical vacuum chamber <strong>of</strong> 24 L volume. A standard part <strong>of</strong> <strong>the</strong> plasma system was a<br />
conventional rotary vane pump, type D8B, suction power <strong>of</strong> <strong>the</strong> approx. 8 m 3 /hour, with an<br />
electromagnetic valve, which prevented <strong>the</strong> return <strong>of</strong> <strong>the</strong> oil vapor into <strong>the</strong> vacuum chamber.<br />
Brass needle valves (max. ow rate <strong>of</strong> 400 sccm), with pressure reducers specied for each<br />
type <strong>of</strong> gases applied, provided constant gas ow rate during plasma processes performed.<br />
The plasma treatments were controlled fully automatically.
204<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
Figure 3 Plasma system LP-Nano LF-40kHz, Diener Electronic GmbH<br />
Surface activation with oxygen and argon plasma<br />
Before plasma treatment, <strong>the</strong> fabric samples (dimensions 340x200 mm) were dried to<br />
remove <strong>the</strong> moisture and afterwards treated (separately with oxygen and argon plasma)<br />
under <strong>the</strong> optimum process conditions, determined by preliminary work [6]: <strong>the</strong> gas ow<br />
rate <strong>of</strong> 40 sccm, working pressure <strong>of</strong> 0.34 - 0.4 mbar, operating power <strong>of</strong> 300 W and<br />
treatment time <strong>of</strong> 5 min.<br />
3.3 Method <strong>of</strong> analysis<br />
Surface Analysis<br />
The bre surface micro-morphology characteristics <strong>of</strong> untreated and plasma treated fabrics<br />
were investigated using JEOL 6060LV Scanning Electron Microscopy (SEM) technique.<br />
Before analysis all <strong>the</strong> samples were coated with carbon and a 90% Au/10%Pd alloy layer.<br />
Magnication <strong>of</strong> 1000x and 7000x was used for analysis <strong>of</strong> surface changes depending on<br />
<strong>the</strong> tested bre type and applied plasma gas.<br />
The topography <strong>of</strong> bre surface roughness caused by oxygen and argon plasma treatment<br />
was analyzed using <strong>the</strong> most recent Atomic force microscopy (AFM). The concept on<br />
which AFM is based is <strong>the</strong> generation <strong>of</strong> <strong>the</strong> images <strong>of</strong> <strong>the</strong> bre surface by measuring <strong>the</strong><br />
physical interactions (forces) between a sharp tip <strong>of</strong> AFM cantilever and <strong>the</strong> sample (Fig<br />
4). When <strong>the</strong> tip is brought into proximity <strong>of</strong> a sample surface, forces between <strong>the</strong> tip and<br />
<strong>the</strong> sample lead to a deection <strong>of</strong> <strong>the</strong> cantilever according to Hooke´s law. Depending on<br />
<strong>the</strong> situation and surface topography, forces that are measured in AFM include mechanical<br />
contact force, Van der Waals forces, capillary forces, chemical bonding, electrostatic forces,<br />
magnetic forces, etc. Registered values <strong>of</strong> cantilever deection are electronically converted<br />
into pseudo three-dimensional (3D) image <strong>of</strong> a sample.
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 205<br />
Figure 4 Basic schematic principle <strong>of</strong> <strong>the</strong> AFM in contact mode [14]<br />
An important feature <strong>of</strong> <strong>the</strong> AFM technique is based on <strong>the</strong> fact that, apart from <strong>the</strong><br />
qualitative characterization, it also allows <strong>the</strong> quantitative assessment <strong>of</strong> bre surface<br />
topography at <strong>the</strong> nano-scale, where <strong>the</strong> changes caused by plasma action are expected.<br />
AFM imaging was performed in contact mode, using a Multimode AFM with Nanoscope<br />
IIIa controller (Veeco Instruments, Santa Barbara, CA) with a vertical engagement (JV)<br />
125 m scanner. Contact mode imaging was performed using silicon-nitride tips (NP-<br />
20, Veeco, nom. freq. 56 KHz, nom. spring constant <strong>of</strong> 0.32 N/m); with <strong>the</strong> highest scan<br />
resolution <strong>of</strong> 512 samples per line. Processing and analysis <strong>of</strong> images were carried out<br />
using <strong>the</strong> NanoScope s<strong>of</strong>tware (Digital Instruments, version V531r1).<br />
Wet ability properties<br />
Wet ability properties were analyzed using vertical test and drop test.<br />
The Vertical test [16] was used to measure <strong>the</strong> rate <strong>of</strong> vertical capillary rise <strong>of</strong> distilled<br />
water in a woven specimen strip suspended in <strong>the</strong> distilled water (at 152 mm below <strong>the</strong><br />
water surface). The vertical rise <strong>of</strong> distilled water in fabric sample was measured in weftand<br />
warp direction. In order to achieve <strong>the</strong> better visibility <strong>of</strong> <strong>the</strong> water front rise, <strong>the</strong><br />
methylene-blue as a water colorant was used. All test specimens were conditioned at <strong>the</strong><br />
standard atmosphere, according to HRN ISO 139. Test was performed in accordance with<br />
EN ISO 9073-6:2000. In order to improve reproducibility and reliability <strong>of</strong> measurement<br />
results, all <strong>the</strong> measurements were recorded by digital video camera recorder SONY DCR-<br />
DVD304E.<br />
The Drop test was carried out as a measure <strong>of</strong> absorbency <strong>of</strong> fabric samples. The absorbency<br />
is dened as <strong>the</strong> propensity <strong>of</strong> a material to take in and retain a liquid, usually water, in <strong>the</strong><br />
pores and interstices <strong>of</strong> <strong>the</strong> material.<br />
Test was performed in accordance with AATCC e.g. TEGEWA prescription [1, 25]. A drop<br />
<strong>of</strong> water <strong>of</strong> a dened volume is allowed to fall from a xed height onto <strong>the</strong> taut surface <strong>of</strong><br />
a test specimen. The time required for full absorption <strong>of</strong> water drop (e. g. time to disappear<br />
<strong>of</strong> drop) is measured and recorded as wetting time (seconds, stopwatch).<br />
The shorter <strong>the</strong> time, <strong>the</strong> more absorbent is <strong>the</strong> textile material. Five seconds or less is<br />
generally considered to represent adequate absorbency.
206<br />
4 Results and discussion<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
Surface analysis by Scanning Electron Microscopy<br />
SEM analysis <strong>of</strong> <strong>the</strong> lyocell and modal cellulose bres in <strong>the</strong> fabric sample surface were<br />
carried out with magnication <strong>of</strong> 1000x and 7000x. A part <strong>of</strong> obtained results [9] are<br />
presented in Fig. 5 and 6.<br />
Untreated lyocell bres<br />
Oxygen-plasma treated lyocell<br />
bres<br />
Argon-plasma treated lyocell<br />
bres<br />
Figure 5 SEM photographs <strong>of</strong> untreated, oxygen and argon plasma treated lyocell bres<br />
Untreated modal bres Oxygen treated modal bres Argon treated modal bres<br />
Figure 6 SEM photographs <strong>of</strong> untreated, oxygen and argon plasma treated modal bres<br />
The obtained results with SEM technique indicate that <strong>the</strong> oxygen plasma treatment<br />
certainly clean <strong>the</strong> bres surface, removing <strong>the</strong> organic contaminants from surface. The<br />
treatment with argon plasma results in sputtering <strong>of</strong> <strong>the</strong> bre surface layer increasing <strong>the</strong><br />
roughness and brillations-effect (especially by lyocell) <strong>of</strong> <strong>the</strong> bres. Thereby <strong>the</strong> specic<br />
area <strong>of</strong> bres increases and <strong>the</strong> size <strong>of</strong> bre area accessible to reactions increases to.<br />
Surface topography analysis by atomic force microscopy (AFM)<br />
The examples <strong>of</strong> bre surface topography images <strong>of</strong> untreated, as well as oxygen- and<br />
argon-plasma treated lyocell bres, observed by AFM [6, 8], are shown in Fig. 7 - 9. The<br />
results are presented in a standard form, e.g. as 3D plot <strong>of</strong> bre surface (a), <strong>the</strong> height data<br />
image (b), and section analysis <strong>of</strong> bre along marked lines (c). The vertical distance <strong>of</strong><br />
scanned topographic elements, are given as a measure <strong>of</strong> roughness <strong>of</strong> imaged surface area<br />
<strong>of</strong> bre.
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 207<br />
a<br />
Figure 7 The AFM <strong>of</strong> untreated lyocell bres:<br />
a - Surface plot - 3D topographic image; b - Section analysis along <strong>the</strong> marked line vertical to <strong>the</strong><br />
axis <strong>of</strong> <strong>the</strong> tested bre (imaged area is 25 m × 25 m with vertical scale <strong>of</strong> 3000 nm).<br />
Vertical distance marked on <strong>the</strong> c: approx. 122 nm<br />
b<br />
a<br />
Figure 8 The AFM <strong>of</strong> oxygen plasma treated lyocell bres:<br />
a - Surface plot - 3D topographic image; b - Section analysis along <strong>the</strong> marked line vertical to <strong>the</strong><br />
axis <strong>of</strong> tested bre (imaged area is 15 m x 15 m with vertical scale <strong>of</strong> 1500 nm)<br />
Vertical distance marked on <strong>the</strong> c: approx. 24 nm<br />
b<br />
a<br />
b<br />
Figure 9 The AFM <strong>of</strong> argon plasma treated lyocell bres:<br />
a - Surface plot - 3D topographic image; b - Section analysis along <strong>the</strong> marked line vertical to <strong>the</strong><br />
axis <strong>of</strong> tested bre (imaged area is 10 m × 10 m with vertical scale <strong>of</strong> 2000 nm)<br />
Vertical distance marked on <strong>the</strong> c: size range from approx. 23 nm to 59 nm
208<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
The results <strong>of</strong> surface roughness noticeable on <strong>the</strong> surface plots (Fig. 7 - 9) and <strong>the</strong> results<br />
obtaining by section analysis point to a different effect <strong>of</strong> oxygen and argon plasma on <strong>the</strong><br />
tested bres, as compared to <strong>the</strong> untreated ones. Surface roughness <strong>of</strong> <strong>the</strong> oxygen plasma<br />
treated bres was considerably lower (approx. 24 nm) compared to <strong>the</strong> untreated samples,<br />
where bre surface roughness was approximately 122 nm. These results additionally prove<br />
that oxygen, as a reactive gas, changes <strong>the</strong> outermost bre surface layers, resulting in <strong>the</strong><br />
removal <strong>of</strong> surface impurities and leading to smoo<strong>the</strong>r surface layer, as can be seen in Fig. 8.<br />
As opposed to oxygen, argon, as an inert gas, caused increased bre surface roughness<br />
and ruptures along <strong>the</strong> longitudinal axis <strong>of</strong> <strong>the</strong> tested bres, as can be seen in Fig. 9. The<br />
surface <strong>of</strong> <strong>the</strong> argon plasma treated sample was altered and ruptures appeared, visible on<br />
<strong>the</strong> surface and oriented in <strong>the</strong> direction <strong>of</strong> bre axis. Their size was in <strong>the</strong> range from 23<br />
to 59 nm at <strong>the</strong> imaged area as well.<br />
AFM results conrmed <strong>the</strong> <strong>the</strong>sis that using oxygen and argon plasmas, two essentially<br />
different processes <strong>of</strong> textile surface ablation occurred; <strong>the</strong> rst is chemical etching and <strong>the</strong><br />
second physical sputtering.<br />
Capillary rise <strong>of</strong> water by vertical test<br />
The capillarity method measures <strong>the</strong> rate <strong>of</strong> vertical capillary rise in a specimen strip<br />
suspended in <strong>the</strong> distilled water. The part <strong>of</strong> obtained results [5] is presented on <strong>the</strong> graphs,<br />
Fig. 10 – 12.<br />
Height <strong>of</strong> <strong>the</strong> water front, h [mm]<br />
150<br />
140<br />
130<br />
120<br />
110<br />
100<br />
90<br />
80<br />
70<br />
60<br />
50<br />
40<br />
CLY I ut (warp)<br />
CLY I O2 plasma, 5 min (warp)<br />
30<br />
CLY I O2 plasma, 5 min (weft) CLY I Ar plasma, 5 min (warp)<br />
20<br />
CLY I Ar plasma, 5 min (weft) CLY I ut (weft)<br />
10<br />
0<br />
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30<br />
Time, t [min]<br />
Figure 10 Results <strong>of</strong> <strong>the</strong> capillary rise <strong>of</strong> water measured on untreated and plasma treated lyocell<br />
fabrics in warp and weft direction
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 209<br />
Height <strong>of</strong> <strong>the</strong> water front, h [mm]<br />
150<br />
140<br />
130<br />
120<br />
110<br />
100<br />
90<br />
80<br />
70<br />
60<br />
50<br />
40<br />
30<br />
20<br />
10<br />
CMD I ut (weft)<br />
CMD I Ar plasma, 5 min (warp)<br />
CMD I O2 plasma, 5 min (weft)<br />
0<br />
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30<br />
Time, t [min]<br />
CMD I O2 plasma, 5 min (warp)<br />
CMD I Ar plasma, 5 min (weft)<br />
CMD I ut (warp)<br />
Figure 11 Results <strong>of</strong> <strong>the</strong> capillary rise <strong>of</strong> water measured on untreated and plasma treated modal<br />
fabrics in warp and weft direction<br />
Height <strong>of</strong> <strong>the</strong> water front, h [mm]<br />
150<br />
140<br />
130<br />
120<br />
110<br />
100<br />
90<br />
80<br />
70<br />
60<br />
50<br />
40<br />
CO I O2 plasma, 5min (warp) CO I Ar plasma, 5min (warp)<br />
30<br />
CO I ut (warp)<br />
CO I ut (weft)<br />
20<br />
CO I O2, 5min (weft)<br />
CO I Ar plasma, 5min (weft)<br />
10<br />
0<br />
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30<br />
Time, t [min]<br />
Figure 12 Results <strong>of</strong> <strong>the</strong> capillary rise <strong>of</strong> water measured on untreated and plasma treated cotton<br />
fabrics in warp and weft direction<br />
It is important to emphasize that several <strong>of</strong> aspects affect <strong>the</strong> wettability <strong>of</strong> treated samples,<br />
such as: bre superstructure, bre diameter, bre surface irregularity and surface chemical<br />
composition. Micro- and macrostructure <strong>of</strong> fabric is very important to. In this respect, <strong>the</strong><br />
increase <strong>of</strong> fabric structure density during plasma treatment cud be reason <strong>of</strong> <strong>the</strong> capillary<br />
force raising and better vertically rise <strong>of</strong> water through fabric.
210<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
Absorption time <strong>of</strong> water drop by dropt test<br />
Results <strong>of</strong> <strong>the</strong> water drop absorption time <strong>of</strong> tested fabrics are presented in Tab. 2.<br />
Table 2: Results <strong>of</strong> <strong>the</strong> water drops absorption time [s] <strong>of</strong> tested fabrics<br />
Sample<br />
Treatment<br />
Cotton<br />
(Co raw)<br />
Cotton<br />
(Co sc)<br />
Lyocell<br />
(CLY I)<br />
Modal<br />
(CMD I)<br />
Untreated > 3600 5.1 2 1.2<br />
O 2<br />
plasma, 5 min 10.0 0.8 1 0.8<br />
Ar plasma, 5 min > 3600 37.8 0.8 0.8<br />
Obtained results presented in Table 2 indicate <strong>the</strong> improvement <strong>of</strong> wetting and wicking<br />
properties <strong>of</strong> all tested samples by low-pressure oxygen plasma treatment. The oxygen<br />
plasma treatment is especially effective for raw cotton fabric. In this case <strong>the</strong> oxygen<br />
plasma treatment could be recommended as eco-friendly substitution for <strong>the</strong> wet scouring<br />
process, in order to achieve good soak properties (wettability).<br />
Plasma treatments with argon as an inert gas are not enough effective to improve wettability<br />
<strong>of</strong> cotton fabrics; <strong>the</strong> drop absorption time (s) determined for raw cotton fabric is extremely<br />
high (characteristic for hydrophobic surface). Hydrophilic property <strong>of</strong> scoured cotton<br />
fabric after argon plasma treatment still decreases (drop absorption time was increased<br />
from 5.1 to 37.8 s). Could be supposed that supercial primary wall <strong>of</strong> cotton bres is thick<br />
enough to protect bers <strong>of</strong> plasma ionized particles attacks. Contrary, <strong>the</strong> low-pressure<br />
argon plasma treatments <strong>of</strong> lyocell and modal fabrics additionally improve <strong>the</strong>ir original<br />
high hydrophilic properties.<br />
5 Conclusions<br />
Presented results are a part <strong>of</strong> investigation <strong>of</strong> <strong>the</strong> nature <strong>of</strong> low-pressure plasma inuence<br />
on <strong>the</strong> bre surface characteristics and <strong>the</strong>ir effects on <strong>the</strong> properties modication <strong>of</strong><br />
cellulose based textile materials.<br />
• The results obtained using <strong>the</strong> SEM technique indicate that <strong>the</strong> oxygen plasma<br />
treatment leads to <strong>the</strong> cleaning and etching <strong>of</strong> <strong>the</strong> bre surface, in contrary to <strong>the</strong> argon<br />
plasma treatment which resulting in sputtering <strong>of</strong> <strong>the</strong> cellulose bre surface layer. This<br />
results in increasing <strong>the</strong> roughness <strong>of</strong> <strong>the</strong> bres and contributes to <strong>the</strong> enhancement <strong>of</strong><br />
bre specic area (ratio: surface area/volume).<br />
• The AFM results conrmed <strong>the</strong> <strong>the</strong>sis that using oxygen and argon plasmas, two<br />
essentially different processes <strong>of</strong> textile surface ablation occurred; <strong>the</strong> rst is chemical<br />
etching and <strong>the</strong> second physical sputtering. The estimated vertical altitude <strong>of</strong> surface<br />
topographic elements <strong>of</strong> <strong>the</strong> oxygen plasma treated bres was considerably lower<br />
compared to <strong>the</strong> untreated samples (i.e. 24 nm in relation to 122 nm for untreated<br />
sample). The surface <strong>of</strong> <strong>the</strong> argon plasma treated sample was altered and <strong>the</strong> visible<br />
ruptures appeared on <strong>the</strong> surface. The cracks at <strong>the</strong> imaged area are oriented in <strong>the</strong><br />
direction <strong>of</strong> bre axis and <strong>the</strong>ir vertical altitude range <strong>of</strong> 22 to 59 nm as well.<br />
• The results <strong>of</strong> <strong>the</strong> vertical test and <strong>the</strong> drop test conrm <strong>the</strong> improvement <strong>of</strong> hydrophilic<br />
properties <strong>of</strong> all tested samples after low-pressure oxygen plasma treatment. Plasma<br />
treatment with inert gas argon, at <strong>the</strong> given process parameters, was not enough effective<br />
to improve wettability <strong>of</strong> cotton fabrics. Contrary, <strong>the</strong> argon plasma treatments <strong>of</strong><br />
lyocell and modal fabrics additionally improve <strong>the</strong>ir original high hydrophilic surface
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 211<br />
properties.<br />
• It could be concluded that presented investigation results conrm that <strong>the</strong> plasma<br />
treatment is an acceptable and appropriate method <strong>of</strong> textile surface modication,<br />
avoiding bre damage, and an eco-friendly method at <strong>the</strong> same time.<br />
References<br />
1. AATCC Test Method 79:2000 - Drop test<br />
2. Chen P, Wang J, Wang B, Li W, Zhang C, Li H, Sun B (2009) Improvement <strong>of</strong><br />
interfacial adhesion for plasma-treated aramid ber-reinforced poly(phthalazinone<br />
e<strong>the</strong>r sulfone ketone) composite and ber surface aging effects. Surface and Interface<br />
Analysis 41 (1) pp 38-43<br />
3. Demir A, Özdogan E, Özdil N, Gürel A (<strong>2011</strong>) Ecological materials and methods in <strong>the</strong><br />
textile industry: Atmospheric-plasma treatments <strong>of</strong> naturally colored cotton. Journal<br />
<strong>of</strong> Applied Polymer Science 119 (3) pp 1410-1416<br />
4. Denes F, Young RA (1998) Surface modication <strong>of</strong> polysaccharides under coldplasma<br />
conditions. In: Dumitriu S (Ed.) Structural Diversity and Functional Versatility <strong>of</strong><br />
Polysaccharides, New York, Basel, Hong Kong, Marcel Dekker Inc<br />
5. Ercegovi Raži S, unko R, Bautista L, Mota J, Crespo L (<strong>2010</strong>) Ageing Effect<br />
on Wettability Properties <strong>of</strong> Low-Pressure Plasma-Treated Cellulosic Fabrics. In:<br />
Dragevi Z (Ed) Proceedings <strong>of</strong> <strong>the</strong> 5th International Textile, Clothing & Design<br />
Conference - Magic World <strong>of</strong> Textiles, Faculty <strong>of</strong> Textile Technology, University <strong>of</strong><br />
Zagreb, Zagreb, pp 570-575<br />
6. Ercegovi Raži S (<strong>2010</strong>) Ciljana modikacija svojstava tekstilija primjenom plazme<br />
i metalnih spojeva. PhD Thesis, University <strong>of</strong> Zagreb, Faculty <strong>of</strong> Textile Technology,<br />
Zagreb, Croatia<br />
7. Ercegovi Raži S, unko R (2009) Modikacija svojstava tekstilija primjenom<br />
plazme. Tekstil 58 (3) pp 55-74<br />
8. Ercegovi Raži S, unko R, Svetlii V, Šegota S (<strong>2011</strong>) Application <strong>of</strong> AFM<br />
Microscopy for Identication <strong>of</strong> Fibres Surface Changes after Plasma Treatments.<br />
Materials Technology (accepted for publishing)<br />
9. Ercegovi Raži S, unko R, Bukošek V, Rolich T (210) Hydrophilicity Improvement<br />
<strong>of</strong> Cellulose Based Materials by Plasma. In: Proceedindgs <strong>of</strong> <strong>the</strong> 41th International<br />
Symposium on Novelties in Textiles, Faculty <strong>of</strong> Natural Sciences and Engineering,<br />
University <strong>of</strong> Ljubljana, Ljubljana, pp 344-350<br />
10. Guimond S, Hanselmann B, Amberg M, Hegemann D (<strong>2010</strong>) Plasma processing <strong>of</strong><br />
textiles: Perspectives. Melliand International 16 (4) pp 182-183<br />
11. Hauser P J, El-Shafei A (<strong>2011</strong>) Atmospheric pressure plasma treatments for repellent<br />
textiles. AATCC Review 11 (1) pp 70-74<br />
12. Horrocks AR, Nazaré S, Masood R, Kandola B, Price D (<strong>2011</strong>) Surface modication<br />
<strong>of</strong> fabrics for improved ash-re resistance using atmospheric pressure plasma<br />
in <strong>the</strong> presence <strong>of</strong> a functionalized clay and polysiloxane. Polymers for Advanced<br />
Technologies 22 (1) pp 22-29<br />
13. http://en.wikipedia.org/wiki/File:Plasma-lamp_2.jpg (visited 14.01.<strong>2011</strong>.)<br />
14. http://www.scientec.fr/include/surface/agilent/options/EN-mode_contact_friction.<br />
html (visited 14.01.<strong>2011</strong>.)<br />
15. Ibrahim NA, Hashem MM, Eid MA, Refai R, El-Hossamy M, Eid BM (<strong>2010</strong>) Ec<strong>of</strong>riendly<br />
plasma treatment <strong>of</strong> linen-containing fabrics. Journal <strong>of</strong> <strong>the</strong> textile Institute101<br />
(12) pp 1035-1049<br />
16. ISO 9073-6:200 Textiles – Test methods for nonwovens – Part 6: Absorption (Liquid
212<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
wicking rate)<br />
17. Wei QF (2005) Functinalization <strong>of</strong> textile bres using plasma-based technology.<br />
Technical Textiles int. 2005 International Newsletter pp 27-29<br />
18. Jimenez M, Bellayer S, Duquesne S, Bourbigot S (<strong>2010</strong>) Improvement <strong>of</strong> heat<br />
resistance <strong>of</strong> high performance bers using a cold plasma polymerization process.<br />
Surface and Coatings Technology 205 (3) pp 745-758<br />
19. Marchandalli B, Riccardi C (2007) Plasma treatments <strong>of</strong> bres and textiles. In: Shishoo<br />
R (Ed) Plasma technologies for textiles Woodhead Publishing Ltd, Cambidge, pp 282-<br />
298<br />
20. Mehta R (<strong>2010</strong>) Plasma Treatment in <strong>the</strong> textile industry. Colourage 57 (7) pp 45-48<br />
21. Morent R, De Geyter N, Vershuren J, De Clerck K, Kickenes P, Leys C (2008) Non<strong>the</strong>rmal<br />
plasma treatment <strong>of</strong> textiles. Surface &Coatings Technology 202 pp 3427<br />
-3449<br />
22. Nath K (<strong>2010</strong>) Atmospheric pressure plasma systems in textile application.<br />
International Dyer 195 (9) pp 25-31<br />
23. Severich B (2008) Atmospheric pressure plasma - A new technology for modifying<br />
textile fabrics. Melliand International 14 (2), pp 120-121<br />
24. Shahidi S, Rashidi A, Ghoranneviss M, Anvari A, Wiener J (<strong>2010</strong>) Plasma effects on<br />
anti-felting properties <strong>of</strong> wool fabrics. Surface and Coatings Technology 205 (SUPPL.<br />
1) pp S349-S354<br />
25. TEGEWA Bericht (1987) Tropftest - eine Methode zur schnellen Bestimung der<br />
Saugfähigkeit an textile Flächengebilden Melliand Textilberichte 68 pp 581-583<br />
26. Tseng HJ, Hsu SH, Wu MW, Hsueh TH, Tu PC (2009) Nylon textiles grafted with<br />
chitosan by open air plasma and <strong>the</strong>ir antimicrobial effect. Fibers and Polymers 10<br />
(1) pp 53-59<br />
27. Vesel A (2008) XPS study <strong>of</strong> surface modication <strong>of</strong> different polymer materials by<br />
oxygen plasma treatment. Informacije MIDEM 38 (4) pp 257-265<br />
28. Vesel A, Mozetic M, Strnad S, Peršin Z, Kleinschek S, Hauptman N (2009) Plasma<br />
modication <strong>of</strong> viscose textile. Vacuum 84 (1) pp 79-82
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 213<br />
Importance <strong>of</strong> salt content reduction<br />
in bakery products<br />
Žaneta Ugarčić-Hardi<br />
Josip Juraj Strossmayer University <strong>of</strong> Osijek, Faculty <strong>of</strong> Food Technology<br />
F. Kuhača 20, Osijek<br />
zaneta.ugarcic-hardi@ptfos.hr<br />
Abstract<br />
Studies have shown that <strong>the</strong> excessive intake <strong>of</strong> sodium/salt has a negative effect on health,<br />
primarily on <strong>the</strong> cardiovascular system.The average daily salt intakes <strong>of</strong> <strong>the</strong> European<br />
population are high (10-20 g salt/day) and surpass quantity <strong>of</strong> physiological requirements<br />
(5-6 g salt/day). Therefore, many international and national bodies set an aim to reduce<br />
<strong>the</strong> salt intake in diet. Since <strong>the</strong> average daily salt intake <strong>of</strong> Croatian population is 13-<br />
16g/day, <strong>the</strong> Croatian Academy <strong>of</strong> Medical Science launched national programme for<br />
reducing salt intake. The rst task <strong>of</strong> Croatian programme is to obtain <strong>the</strong> exact data on<br />
salt consumption. It is estimated that 70-75% <strong>of</strong> total salt comes from manufactured foods.<br />
Since <strong>the</strong> salt intake from bread is signicant (~30%), <strong>the</strong> bakery industry is joining <strong>the</strong><br />
national initiative to salt reduction. The aim <strong>of</strong> this paper was to estimate <strong>the</strong> salt content in<br />
23 samples <strong>of</strong> commercial bread and 20 samples <strong>of</strong> different rolls taken from <strong>the</strong> market in<br />
Osijek-Baranja County (East Croatia). The obtained results showed that <strong>the</strong> salt content in<br />
bread varied in <strong>the</strong> range from 0.96 to 2.05%. This share is even higher in products strewed<br />
with salt (2.04 to extremely 4.76).<br />
Key words: salt reduction, bakery product, salt content in bakery product<br />
1 Introduction<br />
Cardiovascular diseases including heart disease and stroke cause more than 50% <strong>of</strong> all<br />
deaths in <strong>the</strong> world. The major risk factors include high blood pressure, high cholesterol<br />
level obesity and smoking (Skupnjak, 2007, Reiner, 2008, Drenjanevi-Peri, <strong>2010</strong>,<br />
EFSA 2005).<br />
High salt consumption can raise blood pressure and lead to cardiovascular diseases. Too<br />
much salt in food carriees also <strong>the</strong> risk <strong>of</strong> kidney disorders, osteoporoses or stomach<br />
cancer. Studies indicate that it is possible to lower blood pressure by cutting sodium/salt<br />
intake in diet. Salt and sodium are currently one <strong>of</strong> <strong>the</strong> top issues for healthy reformulation<br />
efforts by food and beverage manufactures (Heidolph et al, <strong>2011</strong>). On a global basis, <strong>the</strong><br />
population has a dietary intake <strong>of</strong> sodium that exceeds <strong>the</strong> required level. The current daily<br />
dietary guideline for sodium is 2,300 mg/day (5-6g/day salt) (IOM, 2004). This amount<br />
<strong>of</strong> salt intake should be much lower for children depending on age. In most countries <strong>the</strong><br />
average daily sodium consumption has increased to 10-20 g salt (MacGregor, 2008). Males<br />
in general, consume more sodium than females (Dietary Guidelines Advisory Committee,<br />
<strong>2010</strong>). In Croatian population <strong>the</strong> average daily salt intake accounts 13-16 g/day. Average<br />
daily intakes <strong>of</strong> salt are 13.3 ± 4.3 g for men and 10.2 ± 4.2 g for women (Premuži et al,<br />
<strong>2010</strong>, Miškulin et al, <strong>2010</strong>).
214<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
1.1 Salt sources<br />
10% from total daily consumption <strong>of</strong> sodium is coming from natural food that contains salt<br />
in <strong>the</strong>ir origin (meat, sh, eggs, vegetables). 15-20% salt in our daily diet is originates from<br />
cooked food or meal consumption (salt shaker).<br />
70-75% <strong>of</strong> <strong>the</strong> sodium in <strong>the</strong> diet in Europe, 77% in <strong>the</strong> U.S. and 88% in Australia (Heidolph<br />
et al <strong>2011</strong>), comes from processed foods („Hidden salt“), for which <strong>the</strong> consumers have<br />
little control. Salt intake from bread and bakery products, according to some investigations<br />
accounts for 30-35%.<br />
Table1. Share <strong>of</strong> salt intake from food<br />
Products %<br />
Bread and bakery products 34<br />
Meat and meat products 28<br />
Cheese, cream, eggs 10<br />
Fish and sh products 7<br />
Milk and milk products 5<br />
Fruits and products 5<br />
Fat, confectionary, drinks 11<br />
Source: Bundesforschungsanstalt für Getreide–und Kart<strong>of</strong>felverarbeitung (1989)<br />
As <strong>the</strong> cost <strong>of</strong> health care rises, with cardiovascular disease, health experts, health agencies,<br />
and medical societies are looking for ways <strong>of</strong> improving health <strong>of</strong> <strong>the</strong>ir nations. It is<br />
believed that reduction in sodium may help in diminishing health care costs by reducing<br />
<strong>the</strong> risks <strong>of</strong> cardiovascular-related diseases and o<strong>the</strong>r sodium-impacted illnesses, such as<br />
kidney damage, stomach cancer, and osteoporosis. Although <strong>the</strong> direct cause and effect<br />
with regard to high blood pressure and sodium is not clear, <strong>the</strong>re are many international and<br />
national bodies that set an aim to reduce <strong>the</strong> salt intake in diet.<br />
Among <strong>the</strong>m, Consensus Action on Salt and Health – CASH, provides recommendations,<br />
follows developments, looking for consensus among medical pr<strong>of</strong>essionals and <strong>the</strong> public.<br />
The sodium-reduction strategies have been adopted in Finland, <strong>the</strong> United Kingdom,<br />
Ireland, France, <strong>the</strong> European Union, Switzerland, Canada and USA and <strong>the</strong>y emerge<br />
globally (IOM 2004, EFSA 2005).<br />
Finland is <strong>the</strong> only country that has im plemented a sodium-reduction policy that is<br />
considered to be “mandatory“. In 1993, Finland adopted a more aggressive sodiumreduction<br />
strategy which established mandatory labeling requirements for food products.<br />
Currently, <strong>the</strong> most comprehensive such “voluntary” program is being implemented in <strong>the</strong><br />
United Kingdom, and employs a strategy consisting <strong>of</strong>:<br />
a) programs designed to encourage <strong>the</strong> food industry to reformulate products to<br />
reduce sodium levels on a voluntary basis,<br />
b) a public relations campaign aimed at in creasing public awareness concerning <strong>the</strong><br />
relationship between sodium intake and HTN risk and ways to reduce sodium in<br />
take, and<br />
c) “voluntary” front-<strong>of</strong>-package “trafc light” labeling, which employs a colorcoded<br />
signaling system to inform consumers whe<strong>the</strong>r <strong>the</strong> sodium levels in food<br />
are deemed to be “high” (red), “medium” (amber), or “low” (green) by public<br />
health <strong>of</strong>cials (Heidolph et al, <strong>2011</strong>)
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 215<br />
In <strong>the</strong> UK salt intake has already fallen from 9.5 to 8.6 g/day (i.e. 10% reduction) from when<br />
salt reduction rst started in 2004 to <strong>the</strong> end <strong>of</strong> 2006 (MacGregor ,2008).<br />
In 2006 Croatian Action on Salt and Health (CRASH) was established by Croatian Academy<br />
<strong>of</strong> medical Sciences and an alliance <strong>of</strong> societies including Croatian Hypertension Society,<br />
Croatian A<strong>the</strong>rosclerosis Society and Croatian Cardiac Society, launching an initiative to<br />
reduce salt intake in Croatian population and developing a National program for salt intake<br />
reduction ( Reiner, 2008, Jelakovi, 2008).<br />
Croatian Ministry <strong>of</strong> Health and Social Care supports <strong>the</strong>se activities, and CRASH is<br />
included in <strong>the</strong> action <strong>of</strong> <strong>the</strong> World Health Organization on mapping sodium intake in<br />
European countries.<br />
Since <strong>the</strong> salt intake from bread is signicant (~30%), <strong>the</strong> bakery industry is joining <strong>the</strong><br />
national initiative for salt reduction. The bakery manufacturers have been <strong>the</strong> rst that<br />
joined this action gaining information on <strong>the</strong> “Flour-Bread ‘07” Congress in October<br />
2007 (Ugari-Hardi et al, 2007). In order to encourage and educate <strong>the</strong> manufacturers to<br />
produce food and meals with low or reduced salt content, Croatian Food Agency, Faculty<br />
<strong>of</strong> Food Technology in Osijek, Medical Faculty and Croatian National Institute <strong>of</strong> public<br />
Health in Osijek have organized seminars for bakers and public.<br />
The levels <strong>of</strong> salt used in baking are variable and it is very difcult to obtain absolute levels<br />
<strong>of</strong> salt in bread. The rst task <strong>of</strong> <strong>the</strong> Croatian programme is to obtain <strong>the</strong> exact data on salt<br />
consumption in different parts <strong>of</strong> Croatia (Reiner, 2008). Therefore, <strong>the</strong> aim <strong>of</strong> this paper<br />
was to estimate <strong>the</strong> salt content in bakery products in Osijek-Baranja County in Croatia.<br />
2 Why is salt added to bread dough?<br />
Salt content in bakery products varies from country to country from 1 – 2%. In Croatia salt<br />
addition amounts from 2 to 2.5%, (20g/kg our = 13g/kg bread; 5g Na/kg bread).<br />
The primary reason to salt addition is to improve <strong>the</strong> taste. Unsalted bread is tasteless.<br />
Salt has also a signicant technical effect on <strong>the</strong> properties <strong>of</strong> wheat gluten by streng<strong>the</strong>ning<br />
<strong>the</strong> gluten network in dough. It inuences gluten development, <strong>the</strong> rheology <strong>of</strong> dough<br />
and <strong>the</strong> speed <strong>of</strong> fermentation. (Slumier, 2005). Salt increases <strong>the</strong> resistance, extensibility<br />
and elasticity <strong>of</strong> <strong>the</strong> wheat gluten (Hlinka, 1962, He et al, 1992, Fisher et al, 1994.). Salt<br />
decreases <strong>the</strong> amount <strong>of</strong> water required to produce dough <strong>of</strong> xed consistency (Cauvain<br />
and Young, 1998). The investigation conducted by <strong>the</strong> Ger man Research Center for Food<br />
Chemistry and <strong>the</strong> Technical University <strong>of</strong> Munich shows that salt appears to increase<br />
<strong>the</strong> osmotic pressure leading to a reduction in yeast activity and a decrease in <strong>the</strong> crumb<br />
pore size. Salt also increased <strong>the</strong> mixing time and dough resistance. A maximum loaf<br />
volume was achieved at 0.5% salt per 100 g <strong>of</strong> our. Sensory considerations show that salt<br />
affects <strong>the</strong> overall avor pro le <strong>of</strong> wheat bread. Sodium recognition strongly depends on<br />
salt content (Heidolph et al, <strong>2011</strong>).<br />
Although <strong>the</strong> salt addition has an impact on sensory and technological properties,<br />
investigations showed that salt addition in bakery products can be reduced up to 25%<br />
without negative effect on quality (Forschung hilft dem Backgewerbe, 1989).<br />
Currently <strong>the</strong> investigation shows that bread and bread products can be manufactured<br />
satisfactorily at levels <strong>of</strong> 1% salt on a our basis. (Linko et al, 1984, EFSA, 2005).<br />
2.1 Sodium replacement<br />
Replacement <strong>of</strong> sodium chloride is <strong>the</strong> next best method to reduce salt in a prod uct. Cations<br />
similar to sodium such as potassium work best. However, a bitter or metallic <strong>of</strong>f-avor<br />
is <strong>of</strong>ten sensed when potassium chloride is used. O<strong>the</strong>r mineral salts can partially replace
216<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
salt in a cost-effective manner as well, but, <strong>the</strong>y <strong>of</strong>ten deliver <strong>of</strong>f-avors which can limit<br />
<strong>the</strong>ir use. Replacement with NH 3<br />
Cl tastes <strong>of</strong> plant licorice, and replacement with CaCl<br />
has a negative effect on dough rheology. Although articial salt does not exist, <strong>the</strong> most<br />
common strategy is to replace sodium salts by o<strong>the</strong>r inorganic salts such as potassium<br />
chloride. It has been demonstrated that replacing 20% <strong>of</strong> sodium by potassium results in<br />
bread with an accept-able taste, whereas 40% replacement re sults in unacceptable <strong>of</strong>f-taste<br />
(Salovaara, 1982). As a consequence, sodium replacement is limited. With <strong>the</strong> growing<br />
market <strong>of</strong> salt reduced food products, <strong>the</strong>re is also an increasing demand for salt substitutes<br />
or salt taste enhancing agents. In <strong>the</strong> recent years a lot <strong>of</strong> commercial mixtures <strong>of</strong> NaCl and<br />
KCl can be found on <strong>the</strong> market. Such a commercial products are “Pansalt”, “Sub4salt”and<br />
“Sanusol”. They should give 25-50% reduction in sodium.<br />
3 Material and methods<br />
The salt content was estimated in different bakery products taken from <strong>the</strong> market in<br />
“Osijek-Baranja” County, (East Croatia): 23 samples <strong>of</strong> commercial bread and 20 samples<br />
<strong>of</strong> different rolls strewed with salt. Chloride ion was determined by Volhard’s titration.<br />
This method determines <strong>the</strong> salinity <strong>of</strong> foods based on <strong>the</strong> concentration <strong>of</strong> <strong>the</strong> chloride ion<br />
titrated with silver nitrate solution.<br />
The analyses were provided in <strong>the</strong> Croatian National Institute <strong>of</strong> Public Health in Osijek<br />
and <strong>the</strong> Faculty <strong>of</strong> Food Technology in Osijek.<br />
4 Results and discussion<br />
Table 2. Salt content in different bread types<br />
Bread types Salt content (%)<br />
Wheat French bread (White baguette) 1.75<br />
Wheat Easter braided bread 1.41<br />
“Sovital” bread 1.79<br />
Wheat water bread 1.49<br />
“Krunovit” bread 1.57<br />
“Drava vital” bread 1.31<br />
Family bread 1.30<br />
“Monastery sun” bread I 1.90<br />
“Monastery sun” bread II 2.01<br />
“Grandmo<strong>the</strong>r mix” bread 1.01<br />
Bread with onion 0.96<br />
Corn baguette 1.76<br />
White bread with inulin 1.29<br />
Extra white bread 1.52<br />
White ciabatta 1.73<br />
White bread I 1.58<br />
Roll bread 2.02<br />
White bread II 1.58<br />
White bread III 1.03<br />
White bread IV 2.05<br />
Brown bread 1.61<br />
Baguette 1.84<br />
Water bread 1.29<br />
Average 1.56
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 217<br />
The obtained results showed that <strong>the</strong> salt content in bread in <strong>the</strong> Osijek-Baranja County<br />
varied in <strong>the</strong> range from 0.96 to 2.05%, (Table 2). The average is 15.6 g /kg bread, which<br />
is higher than <strong>the</strong> average value for salt content in bread in EU. It means that eating only<br />
2 slices <strong>of</strong> bread (100 g), salt intake amounts to 1.56 g. According to some investigations<br />
<strong>the</strong> most frequently consumed type <strong>of</strong> bread is <strong>the</strong> white wheat one (>4 slices), (Pucarin-<br />
Cvetkovi, et al, 2008).<br />
This share <strong>of</strong> salt is even higher in products strewed with salt (from 2.04 to extremely<br />
4.76%, average value 2.84), (Table 3). Consumption <strong>of</strong> 1 roll bread (70 g) gives <strong>the</strong><br />
salt intake amounts <strong>of</strong> 2.09 g, which is 1/3 <strong>of</strong> <strong>the</strong> total recommended daily intake (5-6 g).<br />
Most consumers <strong>of</strong> those products (breakfast rolls) are younger, who get a habit <strong>of</strong> eating<br />
salty food from childhood The investigation <strong>of</strong> <strong>the</strong> diet habits <strong>of</strong> school children in Osijek<br />
has shown that 32.0% <strong>of</strong> school children population has a habit to eat daily some <strong>of</strong> <strong>the</strong><br />
bread products, 28.3% (154/545) girls and 35.9% (191/532) boys. Sample was stratied by<br />
settlements population, size 1077 pupils, 545 girls and 532 boys (Miškulin et al, 2009). The<br />
high salt intake in childhood leads to higher blood pressure and an increased risk <strong>of</strong> stroke<br />
and heart disease in later life.<br />
Table 3. Salt content in rolls strewed with salt<br />
Type <strong>of</strong> bakery products strewed with salt Salt content (%)<br />
Finger stick (Baguette ngers) 4.76<br />
Pretzel 3.04<br />
Salt stick 5.98<br />
Finger stick (Baguette ngers) 2.20<br />
Pretzel 2.41<br />
Salt stick 2.60<br />
Roll (Baguette roll) 2.04<br />
Finger stick (Baguette ngers) 2.08<br />
Pretzel 2.14<br />
Salt stick 2.51<br />
Salt stick 2.28<br />
Pretzel 3.44<br />
Finger stick (Baguette ngers) 2.45<br />
Roll (Baguette roll) 2.40<br />
Pretzel 4.57<br />
Finger stick (Baguette ngers) 2.17<br />
Salt stick 2.53<br />
Roll (Baguette roll) 2.34<br />
Pretzel 2.26<br />
Roll (Baguette roll) 2.71<br />
Average 2.84<br />
Very high salt content was determined in snack products, up to 5%. Average salt content<br />
<strong>of</strong> <strong>the</strong>se products is 2.81%. Eating 100 g (1 package) <strong>of</strong> such products, 1/2 <strong>of</strong> total salt<br />
quantity <strong>of</strong> physiological requirements will be taken, without being conscious <strong>of</strong> that fact<br />
(Ugari-Hardi et al, <strong>2010</strong>). Therefore, it is important to raise consumer awareness abut<br />
<strong>the</strong> association between salt consumption and health in order for people to assume more<br />
responsibility for <strong>the</strong>mselves. One <strong>of</strong> <strong>the</strong> preconditions for this is <strong>the</strong> improved labelling <strong>of</strong><br />
foods with nutrition and health information, including details <strong>of</strong> salt content. Salt labelling
218<br />
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering<br />
should detail <strong>the</strong> salt content per serving with <strong>the</strong> recommended intake per day. Some<br />
products in Croatia are labelling on this way. Fig.1.<br />
Fig.1. Recommended daily intake <strong>of</strong> nutrition’s and percentage <strong>of</strong> each nutrition in product<br />
The bakery and o<strong>the</strong>r food industries are invited to support <strong>the</strong> national action on salt<br />
reducing. The bakery industry is recommended to provide that gradually in two steps. The<br />
rst step should be 10% salt reduction, which means addition <strong>of</strong> 1.8 g salt/100 g our.<br />
In <strong>the</strong> second step <strong>the</strong> reduction should be up to 25% (1.5 g salt/100 g our). 25% salt<br />
reduction would not affect <strong>the</strong> taste and technical properties. This target could be achieved<br />
across <strong>the</strong> next three years.<br />
5 Conclusion<br />
The results on salt content in different bakery products show that reduction on salt amounts<br />
is necessary. The salt content in bread in <strong>the</strong> Osijek-Baranja County varied in <strong>the</strong> range<br />
from 0.96 to 2.05%, and in special products from 2.04 to extremely 4.76%. Although <strong>the</strong><br />
salt addition has an impact on sensory and technological properties, investigation showed<br />
that salt addition in bakery products can be reduced up to 25% without negative effect on<br />
quality.<br />
It may be possible to set an average target <strong>of</strong> salt reduction in bakery products gradually<br />
from 10 to 25% in <strong>the</strong> next three years, with <strong>the</strong> aim <strong>of</strong> getting used <strong>the</strong> food with less salt,<br />
which should contribute to long-term prevention <strong>of</strong> cardiovascular diseases.<br />
Discussions with <strong>the</strong> bakery products industry and <strong>the</strong> government would be needed to<br />
dene a way to achieve <strong>the</strong> reduction in salt level consistent with technological and sensory<br />
needs.<br />
6 References<br />
1. Cauvain SP, Young LS (1998): Technology <strong>of</strong> Breadmaking. Blackie Academic and<br />
Pr<strong>of</strong>esionals, London<br />
2. Dietary Guidelines Advisory Committee Draft Report (<strong>2010</strong>)<br />
3. Drenjanevi-Peri I (<strong>2010</strong>) Prekomjeran unos soli kao imbenik rizika za razvoj<br />
kardiovaskularnih bolesti, Znanstveno - struni skup „Sol i zdravlje“, Faculty <strong>of</strong> Food<br />
Technology in Osijek, Osijek, 28. 1. <strong>2010</strong>.<br />
4. EFSA- European Public Health Alliance (2005) EFSA highlights <strong>the</strong> dangers <strong>of</strong> eating<br />
too much salt, Published online at www.epha.org/a/1912./ Accessed 3 March <strong>2011</strong><br />
5. Fisher MH, Aitken TR, Anderson JA (1994) Effects <strong>of</strong> mixing, salt and consistency on<br />
extensograms. Cereal Chemistry 26, 81–97.<br />
6. Forschung hilft dem Backgewerbe (1989): Nutzen und Möglichkeiten einer<br />
natriumreduzierte Ernährung, Bundesforschungsanstalt für Getreide–und<br />
Kart<strong>of</strong>felverarbeitung,, Detmold, pp 10 – 23.
Annual <strong>2010</strong>/<strong>2011</strong> <strong>of</strong> <strong>the</strong> Croatian Academy <strong>of</strong> Engineering 219<br />
7. FSAI (<strong>2010</strong>) Salt and health: Review <strong>of</strong> <strong>the</strong> scientic evidence and recommendations<br />
for public policy in Irland<br />
8. He H, Roach RR, Hoseney RC (1992) Effect <strong>of</strong> nonchaotropic salts on our breadmaking<br />
properties. Cereal Chemistry 69 (4), 366–371<br />
9. Heidolph B B, Ray DK, Koehler P, Weiber, Slocum S, Noort MW (<strong>2011</strong>): Looking<br />
for my lost shaker <strong>of</strong> salt…replacer: avor, funktion, future, Cereal, Food World, Vol.<br />
56, No.1, 5-19.<br />
10. Hlynka I (1962) Inuence <strong>of</strong> temperature, speed <strong>of</strong> mixing, and salt on some rheological<br />
properties <strong>of</strong> dough in <strong>the</strong> farinograph, Cereal Chemistry 39, 286–303.<br />
11. IOM-Institute <strong>of</strong> Medicine (<strong>2010</strong>) Strategies to reduce sodium intake in <strong>the</strong> United<br />
States, Appendix C: International efforts to reduce sodium consumption<br />
12. IOM- Institute <strong>of</strong> Medicine (2004) Dietary reference intakes for water, potassium,<br />
sodium, chloride and sulfate, United States Institute <strong>of</strong> Medicine.<br />
13. Jelakovi B (2008) Salt and hypertension, Znanstveni skup, Kardiovaskularno zdravlje,<br />
Prehrana i sol, Akademija medicinskih znanosti, 31, Zagreb, 21.11.2008, Abstract, 31<br />
14. Linko P, Harkonen H, Linko YY (1984) Effects <strong>of</strong> sodium chloride in <strong>the</strong> processing <strong>of</strong><br />
bread baked from wheat, rye and barley ours. Journal <strong>of</strong> Cereal Science 2 (1), 53–62<br />
15. MacGregor GA (2008) Salt-from evidence to implementation, Znanstveni skup,<br />
Kardiovaskularno zdravlje, Prehrana i sol, Akademija medicinskih znanosti, 3, Zagreb,<br />
21.11.2008. Abstract, 3<br />
16. Miškulin M, Periš D, Ugari-Hardi Ž, Dumani G (<strong>2010</strong>) Skrivena sol u prehrani<br />
djece školske dobi - ima li mjesta zabrinutost?, 34. struni skup s meunarodnim<br />
sudjelovanjem, “Zdravstvena ekologija u praksi”, Zagreb, 21.-23. 4. <strong>2010</strong> Proceedings,<br />
pp 103-109<br />
17. Premuži V, Erceg I, Jovanovi A, Reiner Ž, Jelakovi B (<strong>2010</strong>): Unos soli u odrasloj<br />
populaciji, Hrvatski asopis za javno zdravstvo HCJZ, On.line 6, (21),<br />
18. Pucarin-Cvetkovi J, Polašek O, Kern J, Vuleti S, (2008) Regional features <strong>of</strong><br />
Croatian nutrition, Znanstveni skup, Kardiovaskularno zdravlje,Prehrana i sol,<br />
Akademija medicinskih znanosti, Zagreb, 21.11.2008. Abstract, 8<br />
19. Reiner Ž (2008) National programme for <strong>the</strong> reduction <strong>of</strong> salt intake in<br />
Croatia,Znanstveni skup, Kardiovaskularno zdravlje, Prehrana i sol, Akademija<br />
medicinskih znanosti, Zagreb, 21.11.2008. Abstract, 5.<br />
20. Salovaara H (1982) Sensory limitations to replacement <strong>of</strong> sodium with potassium and<br />
magnesium in bread. Cereal Chemistry 59(5):427<br />
21. Skupnjak B(2007): Inicijativa za nacionalnu kampanju smanjenja utroška soli,<br />
Hrvatski asopis za javno zdravstvo HCJZ, Online, Vol 3, 10.<br />
22. Slumier P (2005) Principles <strong>of</strong> breadmaking, Amercan Association <strong>of</strong> Cereal Chemists,<br />
Inc. St. Paul, Minnesota, USA, pp46-47.<br />
23. Ugari-Hardi Ž (2007) Salt reduction in Croatian baked products: Yes or no?<br />
4th International Congrss „Flour-Bread 07“ and 6th Croatian Congress <strong>of</strong> Cereal<br />
Technologists, Faculty <strong>of</strong> Food Technology in Osijek, Josip Juraj Strossmayer<br />
University <strong>of</strong> Osijek, Opatija, October 24-27, 2007, Abstract, 31<br />
24. Ugari-Hardi Ž, Dumani G, Pitlik N, Koceva Komleni D, Juki M, Kuleš A,<br />
Sabo M (<strong>2010</strong>) The salt content in bakery products in Osjecko-baranjska County,<br />
Proceedings <strong>of</strong> 5th International Congress „Flour-Bread 09“ and 7th Croatian Congress<br />
<strong>of</strong> Cereal Technologists, Ugari-Hardi Ž., Faculty <strong>of</strong> Food Technology Osijek, Josip<br />
Juraj Strossmayer ty <strong>of</strong> Osijek, ty <strong>of</strong> Osijek, 551-556.
Podupirajui lanovi:<br />
Arhitektonski fakultet, Zagreb, www.arhitekt.hr<br />
Centar za vozila Hrvatske d. d., www.cvh.hr<br />
Elektrotehniki fakultet, Osijek, www.etfos.hr<br />
Energetski institut Hrvoje Požar, Zagreb, www.eihp.hr<br />
Fakultet elektrotehnike i raunarstva, Zagreb, www.fer.unizg.hr<br />
Fakultet elektrotehnike, strojarstva i brodogradnje, Split, www.fesb.hr<br />
Fakultet kemijskog inženjerstva i tehnologije, Zagreb, www.fkit.hr<br />
Fakultet prometnih znanosti, Zagreb, www.fpz.hr<br />
Fakultet strojarstva i brodogradnje, Zagreb, www.fsb.hr<br />
Geodetski fakultet, Zagreb, www.ge<strong>of</strong>.hr<br />
Graevinski fakultet, Osijek, www.gfos.hr<br />
Graevinski fakultet, Rijeka, www.gradri.hr<br />
Fakultet graevinarstva, arhitekture i geodezije, Split, www.gradst.hr<br />
Graevinski fakultet, Zagreb, www.grad.hr<br />
Graki fakultet, Zagreb, www.grf.hr<br />
Hrvatska elektroprivreda d. d., Zagreb, www.hep.hr<br />
Hrvatska zajednica tehnike kulture, Zagreb , www.hztk.hr<br />
Kemijsko-tehnološki fakultet, Split , www.ktf-split.hr<br />
Konar - Institut za elektrotehniku d. d., Zagreb , www.koncar-institut.hr<br />
Odašiljai i veze d.o.o., Zagreb, www.oiv.hr<br />
Pliva Hrvatska d.d., Zagreb, www.pliva.hr<br />
Pomorski fakultet, Rijeka, www.pfri.hr<br />
Prehrambeno-biotehnološki fakultet, Zagreb, www.pbf.hr<br />
Prehrambeno-tehnološki fakultet, Osijek, www.ptfos.hr<br />
Rudarsko-geološko-naftni fakultet, Zagreb, www.rgn.hr<br />
Strojarski fakultet, Slavonski Brod, www.sfsb.hr<br />
Supra Net, d.o.o., www.supranet.hr<br />
Sveuilište u Dubrovniku, www.unidu.hr<br />
Sveuilište J. J. Strossmayera u Osijeku<br />
Tehniki fakultet, Rijeka, www.riteh.hr<br />
Tehniko veleuilište u Zagrebu, www.tvz.hr<br />
Tekstilno-tehnološki fakultet, Zagreb, www.ttf.hr<br />
Uljanik brodogradilište d. d., Pula, www.uljanik.hr