Volume-3 English - Colleges - King Saud University

IN THE NAME OF ALLAH, 

MOST MERCIFUL, MOST GRACIOUS

Proceedings of the Seventh 

Saudi Engineering Conference 

Riyadh 2-5, December 2007 

� � 

Volume III 

Research and development to serve the industry and upgrade its services 

Civil Engineering 

General Engineering

PREFACE 

The Seventh Saudi Engineering Conference comes to complement the 

series of Saudi engineering conferences which started in 1402H and have 

been hosted successively by different colleges of engineering of the Saudi 

universities. The College of Engineering at King Saud University is honored 

to host the conference for the second time. 

These conferences have greatly contributed to the resettlement of 

technology, the dissemination and exchange of experiences between 

engineering professionals, and have helped to promote the scientific 

research besides advancing innovation and excellence. 

At a time of the advanced technology and the availability of information 

in various ways, nations have become closer and the world is turning into a 

small village, the economy has become the prime engine of the world. It is 

necessary for all nations to work hard to cope with this technical progress 

and benefit from it, and moreover create appropriate conditions to deal with 

this tremendous development and competition as much as possible. It is 

incumbent upon all professionals in general and engineers in particular to 

work hard to provide the proper environment in such circumstances. 

As a consequence, The Seventh Saudi Engineering Conference 

discusses an important and vital theme for researchers, engineers and 

industrialists. The theme is to provide an Engineering Environment to merge 

in a Competitive Global Economy in an open and boundary-less economy 

and profession. This conference is trying to answer this question through 

well-formulated seven topics. 

Conference topics discuss multiple issues related to engineering 

profession and engineering firm, engineering environment through 

education and labor market requirements, engineering rehabilitation, 

preservation of the environment, rationalization of resource consumption, 

Saudi construction code, development of the engineering sector to 

diversify sources of national income, and research and development to 

service the industry and upgrade its services.

PREFACE 

The conference proceedings contain 168 refereed scientific research 

papers which are distributed into a number of volumes, and each volume 

contains one or more topic. A separate volume for paper abstract is also 

published in addition to electronic proceedings that includes all papers 

accepted in the conference. These proceedings will be a scientific reference 

for engineers in the Kingdom and the worldwide. 

Finally, thanks to Almighty God for his help in completing of this work 

and deep thanks for all members of the Conference Committees for their 

efforts, and special thanks to members of the Scientific Committee for their 

efforts to have this documentation of the huge scientific research, which is 

an important reference for researchers and engineers. Thanks also for 

authors and experts who have contributed their ideas, their research to the 

success of the conference. 

Proceedings of the 7 th Saudi Engineering Conference, KSU, Riyadh, 2007 

Thanks 

Chair of organizing committee 

Prof. Abdulaziz A. Alhamid 

VI

INTRODUCTION 

Under the high patronage of his Royal Highness the Prince Sultan Ibn 

Abdulaziz the crown prince and minister of defense, aviation and inspector 

general, the College of engineering at the King Saud university hosted the 

Seventh Saudi Engineering Conference during the period 22 to 25 Dhu 

Alqeeda 1428 corresponding to 2-5 December 2007. The theme issue of 

the conference is “Towards An Engineering Environment Competitive to 

the Economics of Globalization”. 

The response to contribute in the conference has been most 

encouraging. A large number of abstracts were received. After a 

thorough peer-review process for evaluating the submitted papers, the 

scientific committee has selected a total of 168 papers, presented by 300 

researchers. The conference has drawn participants from the different 

kingdom universities, colleges, institutes and technical education 

establishments as well as governmental and national companies. The 

conference has also attracted international participation from universities 

and institutes of United Arab Emirates, Egypt, Sudan, Algeria, Tunisia, 

Malaysia, India, Great Britain, Germany, France, Deutschland, Canada, 

Japan and United States of America. 

One of the main objectives of the conference was to contribute to the 

review and development of important aspects of the engineering sector both 

public and private. The topics of the conference were chosen to tackle the 

challenges that engineering education and its outputs are facing. In 

addition, the themes also emphasized on the contribution of the engineers 

to the development of the country. The conference themes were as follows: 

• Engineering qualification and its role in the strategy of Saudization 

• Engineering specialties as viewed from the educational 

establishments and the job market requirements 

• Engineering sector contribution to resources conservation 

• Engineering and environmental protection 

VII 

Proceedings of the 7 th Saudi Engineering Conference, KSU, Riyadh, 2007

INTRODUCTION 

• The Saudi building code 

• Development of the engineering sector for diversification of income 

resources 

• Research and development in the service of industry and for the 

improvement of services 

In addition to the specialized scientific papers that covered the above 

mentioned themes the conference also hosted a number of plenary lectures 

and discussion forums that attracted the participation of key policy makers 

as well as academics and economic parties. 

The selected abstracts and papers have been documented in the 

proceedings which comprise of six volumes in accordance with the 

conference themes. The papers are also documented in CDs. 

Before concluding I would like to express my gratitude to all members of 

the Scientific Committee for their efforts and active participation to the 

success of the conference. Thanks are also due to the referees who have 

been of great help in selecting high quality papers for the conference. The 

support provided by the secretarial and technical staff of the college of 

engineering is also thankfully acknowledged. 

Finally on my own behalf and behalf of the Scientific Committee I would 

like to record our appreciation and sincere thanks to His Excellency the 

rector of the King Saud University and the Dean of College of engineering, 

the chairman of the organizing committee for their continued support and 

valuable guidance, We are all hopeful that this scientific conference will be 

of a support for recruiting engineering specialties on a larger scale and 

contribute to the growth and prosperity of our country. May Allah Almighty 

accept our sincere efforts. 


Chairman of the Scientific Committee 

Prof. Khalid Ibrahem Alhumaizi 

VIII

Contents 

Topic 7 

Research and development to serve the industry and 

upgrade its services 

• Civil Engineering 

EVALUATION OF FOAMED ASPHALT FOR ROAD BASE 

STABILIZATION 

Hamad I. Al-Abdul Wahhab , Mirza G. Baig , Isam A. Mahmoud ,and Hisham M. 

Kattan 

EVALUATION OF SULFUR-ASPHALT TECHNOLOGY FOR 

LOCAL APPLICATIONS 

Hamad I. Al-Abdul Wahhab and Mirza G. Baig 

APPLICATION OF ARTIFICIAL NEURAL NETWORKS FOR 

THE PREDICTION OF STRENGTH AND PERMEABILITY OF 

HIGH PERFORMANCE CONCRETE 

Mohammad Iqbal Khan 

A NOVEL METHOD FOR MEASURING POROSITY OF HIGH 

STRENGTH CONCRETE 


FIELD STUDY FOR THE EVALUATION OF STEEL BRIDGES 

IN RIYADH CITY 

Khalid A. AlSaif , Alaa Malaika , Mohamed M. ElMadany 

PERFORMANCE OF CONCRETE UTILIZING THE NATURAL 

POZZOLAN AVAILABLE IN THE KINGDOM OF SAUDI 

ARABIA 

M. I. Khan , A. M. Alhozaimy and A. M. Al-Saleem 

QUALITY OF CONCRETE USED IN HOUSING 

CONSTRUCTION IN RIYADH AND THE ROLE OF THE 

FORTHCOMING SAUDI BUILDING CODE 

Abdulaziz I. Al-Negheimish , Abdulrahman Alhozaimy , Saleh Al-Sulaiman 

, Saud Al-Swaida , and Said Shahran 


Page No. 

1 

3 

23 

37 

49 

63 

81 

97

INFLUENCE OF HORIZONTAL CONSTRUCTION JOINT ON THE 

FLEXURAL BEHAVIOUR OF REINFORCED CONCRETE SLABS 

Ibrahim M. Metwally , Mohamed S. Issa 

BUILDING MORE DURABLE ASPHALT PAVEMENT IN THE 

KINGDOM OF SAUDI ARABIA USING SUPERPAVE SYSTEM 

Mohamed S. Aazam , Al-Hosain M. Ali , Saleh Alswailmi 

USING ACCESS MANAGEMENT TO REDUCE TRAFFIC 

CONFLICT 

Mohammed T. Mallah 

EFFECT OF POLYPROPYLENE FIBERS ON MECHANICAL 

PROPERTIES OF HIGH-STRENGTH CONCRETE 

A.A.F. Shaheen , A.A.Emam , and I.M.Metwally 

THEORETICAL INVESTIGATION ON THE BEHAVIOR OF 

HYBRID STEEL GIRDERS UNDER PURE MOMENT 

Abdel-Lateef, T.H , Tohamy, S.A. , and Sadeek, A.B 

REMOTE SENSING AND GIS FOR LAND SUITABILITY: AN 

APPLICATION FOR ECOTOURISM 

Nedal, A. Mohammad , Sharifah Mastura S. A. , Johari Mat Akhir 

EVALUATION OF RAP AS A CEMENT-TREATED BASE FOR 

HIGHWAYS CONSTRUCTION 

Mohamed A. Elshabrawy , Al-Hosain M. Ali , Alaa R. Gaber 

DYNAMIC MEASUREMENTS FOR UPDATING CABLE- 

STAYED BRIDGE MODEL- CASE STUDY 

Shehab Mourad, Mohamed Fayed, Ayman Khalil, and Hosam Abozeid 

ANALYSIS OF RECTANGULAR HOLLOW SECTIONS 

UNDER UNAXIAL COMPRESSIVE LOADS 

Sedky Abd Allah Tohmay 

SEISMIC PERFORMANCE OF SHEAR DEFICIENT EXTERIOR 

RC BEAM-COLUMN JOINTS REPAIRED USING CFRP 

COMPOSITES 

Y. A. Al-Salloum, S. H. Alsayed, T. H. Almusallam and N. A. Siddiqui 


Topic 7 

Research and development to serve the 

industry and upgrade its services 

• Civil Engineering

EVALUATION OF FOAMED ASPHALT FOR ROAD BASE 

STABILIZATION 

Hamad I. Al-Abdul Wahhab 1 , Mirza G. Baig 1 , Isam A. Mahmoud 1 , 

and Hisham M. Kattan 2 

1: Department of Civil Engineering, King Fahd University of Petroleum & 

Minerals, Dhahran 31261, Saudi Arabia, Email: hawahab@kfupm.edu.sa 

2: Roads Department, Saudi Aramco, Dhahran, Saudi Arabia 

ABSTRACT 

This paper summarizes the research carried out in the area of Foamed Asphalt 

Technology that was planned to compare the performance of foamed asphalt 

pavement mixes with conventional aggregate of road bases. The research work 

focused on the investigation and evaluation of the feasible use of foamed asphalt 

technology for local roads using marginal quality construction materials, marl, and 

Reclaimed Asphalt Pavement (RAP) materials for local applications. Materials 

included Ministry of Transport (MOT) granular base class A and B, subbase 

material class B, and reclaimed asphalt pavement (RAP) material. Foamed asphalt 

mixes were designed for subbase class B (SB) and RAP material utilizing low 

percentage of Portland cement. Foamed asphalt mixes were optimized to meet dry 

and wet ITS requirements. Designed mixes in addition to granular base class A and 

B were evaluated for shear strength, angle of internal friction, and dynamic resilient 

modulus at 25°C. Results indicate that Portland cement was effective in reducing 

stability loss. Base class A achieved the highest shear strength followed by base 

class B and foamed SB then RAP mix. Resilient modulus testing indicated that SB 

mix has behavior comparable to base class A. RAP mix has shown the best 

behavior. Saturation has reduced the resilient modulus of all mixes significantly. 

Foamed asphalt technology can be used successfully to construct road bases from 

locally available marginal or recycled materials. 

KEY WORDS 

Foamed asphalt, Recycling, Stabilization, Granular Base, Subbase. 

BACKGROUND 

In eastern Saudi Arabia, there is a scarcity of good quality construction 

materials. Portland cement or other stabilizers are required to enhance the marginal 

materials available thereby making the improvement uneconomical and sometimes 

impractical. Other means of stabilization techniques are needed to upgrade their 

performance in order to use them for the construction of base or subbase layers in 

the harsh arid desert climate. 


EVALUATION OF FOAMED ASPHALT FOR ROAD BASE STABILIZATION 

Foamed asphalt epitomizes the asphalt industry’s drive towards energy 

efficient, environmentally friendly, and cost-effective solutions for road 

building. Foamed asphalt mix refers to asphalt concrete mixtures of roadbuilding 

aggregates and foamed asphalt binder (Figure 1). Although this 

technology was developed more than 30 years ago and lauded by researchers the 

world over, it did not gain much acceptance or implementation after its 

development, mainly because there were no equipments available at that time to 

produce or apply the product on a commercial scale. 

Water to foam the 

bitumen 

Milling and mixing rotor 

Proceedings of the 7 th 4 

Saudi Engineering Conference, KSU, Riyadh, 2007 

Injection hot bitumen 

Injecting water for 

compaction 

Figure 1. Schematic drawing showing the process in WR 2500, as it accurately 

meters and sprays asphalt, water and air 

In recent years, foamed asphalt technology has increasingly gained acceptance 

as an effective and economical construction materials improvement and stabilization 

technique mainly because of its improved aggregate penetration, coating 

capabilities, handling, and compaction characteristics. Evidence of renewed interest 

in this technology was observed worldwide recently, largely because new paving 

equipment has been designed specifically for its application and is easily available. 

The technology has been successfully employed in Europe, Africa, and the Middle 

East since the late 1980s and is being increasingly adopted in the U.S., Canada, and 

Australia as its benefits become widely known (A.A. Loudon & Partners, 1996a, 

1996b; Bowering and Martin, 1976; Csanyi, 1957, 1962; Kendall et al., 2001; Lee, 

1981; Mobil Oil, 1973; Muthen, 1998 a and b; Soter International, 1994). 

Foamed asphalt does not behave like regular asphalt during the application 

process. Foamed asphalt mixture acts like soil treatment rather than asphalt concrete 

mix. It is very moisture sensitive. This moisture sensitivity is the primary reason


Hamad I. Al-Abdul Wahhab, et al 

why foamed asphalt is not appropriate for every road. The technique did not work on 

trial roads because the water table is high or springs flow underneath these roads. 

The existence of the water table softens and weakens the subgrade which has 

negative effect on the top layer. However, as long as the subgrade is good, foamed 

asphalt technology will work well (Focus, 2003). Recently, the use of foamed 

asphalt in cold recycling has gained more and more acceptance in Europe, South 

Africa, and Asia (Chiu-Te et al., 2002). 

Asi et al. (1998, 1999, 2002) carried out several laboratory research programs 

to investigate the feasible use of foamed asphalt technology in Saudi Arabia to 

improve the prevalent dune sands for possible use as a base or subbase material. 

Several variables were investigated to evaluate the relative improvement of dune 

sand as well as to permit the development of design procedures for the future use of 

foamed asphalt technology in the harsh climatic conditions of eastern Saudi Arabia. 

Statistical analysis of the results was employed to verify the effects of emulsified 

asphalt and foamed asphalt treatment, with and without the addition of Portland 

cement, on the strength characteristics of the treated mixes. The results displayed a 

significant improvement in the performance of dune sand-foamed asphalt mixes, as 

compared to that of the emulsified asphalt mixes. 

The first use of the foamed asphalt application in Saudi Arabia was in 1997 in 

Shaybah road. The road was constructed, planed, and the top 200 mm of the existing 

marl was mixed with foamed bitumen and cement before being laid onto the 

remainder of the marl road (subbase). The recycled road base was rolled and graded 

to the required profile to produce the foamed asphalt pavement. The foamed asphalt 

was subsequently surfaced with a slurry seal. The overall assessment of the road’s 

condition and its performance from the data recorded is that the road and the foamed 

asphalt, in particular, performed remarkably well under the heavy traffic loading and 

harsh conditions for the intended design life. It has exceeded its original design life 

as a construction access road for the Shaybah development. The level of 

deterioration, from recent visual surveys, shows incompetent slurry seal surfacing 

and it is increasing rapidly and severe in some places (Figure 2), which indicated the 

need to protect the stabilized layer with a layer of hot asphalt concrete mix. 

Nearly all local roads were built according to the old Ministry of Transport 

(MOT) specifications utilizing aggregate base type A or B and subbase type B (SB). 

Most of these roads will now need major maintenance or reconstruction. The 

utilization of these materials in the reconstruction will result in a major cost saving. 

In this study, laboratory tests were conducted to compare the performance of foambased 

layers with the conventional aggregate-based layers mixes. Old MOT base 

class A, B and subbase class B were compared to foamed asphalt mixes made with 

MOT class B subbase and with recycled asphalt pavement and aggregate (RAP). 

5


EXPERIMENTAL WORK 

Material Selection 

Figure 2. Shaybah access road 

The selected materials used for this study are MOT class B subbase and MOT 

class A and B base as per old MOT specifications. The materials were tested for 

gradation and physical characteristics to assure their conformity to MOT standards. 

Mix Design Procedure 

Laboratory mix design procedure for foamed asphalt treated mix was carried 

out based on the cited literature (Wirtgen, 2004) and summarized in Figure 3. 

Performance Testing 

In this study, designed foamed asphalt mixes that include class B subbase and 

recycled asphalt pavement and aggregate (RAP) in addition to the virgin materials 

that include class A aggregate, class B aggregate and class B subbase (SB) were 

subjected to indirect tensile strength (ASTM D 4867), California bearing ratio 

(CBR), dynamic resilient modulus (MR), and static triaxial to evaluate their 

engineering properties. The study was conducted in three interdependent phases. 

These test phases included: 


Saudi Engineering Conference, KSU, Riyadh, 2007

1. CBR test for 

- Class A Aggregate 

- Class B Aggregate 

- MOT Class B Subbase (SB) 

- Recycled Asphalt Pavement and Aggregate (RAP) 

2. Split tensile strength test for 

- Foamed SB 

- Foamed RAP 

3. Resilient modulus and static triaxial test for 

- Class A Aggregate 

- Class B Aggregate 

- Foamed SB 

- Foamed RAP 



7


Select Asphalt 

Determination of 

Foaming Characteristics 

Expansion Ratio > 8 times 

Half-Life > 6 sec. 

Indirect Tensile Strength 

Dry at 25 o C 

Wadd = 1 + (0.5 WOMC – Wair-dry) 

blend 5 samples of 10kg with 

different foam contents, add the 

remaining water for OMC, and 

compact using standard Marshall 

Compaction 

Extrude samples from the 

mold and Cure samples in 

oven at 40 o C for 72 hours 

Plot Data and Select Optimum Foamed 

Asphalt Content Based on: 

Minimum Dry-ITS of 200 kPa 

Minimum Soaked-ITS of 100 kPa 

Maximum Loss in ITS of 20% (For 

high water table) 



Select Aggregate 

Gradation 

Conform to 

Gradation? 

Determination of 

Optimum Moisture 

Content (OMC) 

Indirect Tensile Strength 

After 24 hours Soaking at 25 o C 

Figure 3. Summary of mix design procedure for foamed asphalt treated mix

RESULTS 

Design of Foaming Characteristic 



Laboratory scale WLB 10 foamed asphalt plant built by Wirtgen Gmbh 

Company was used to carry out laboratory mix designs and produce foamed asphalt 

that closely simulates full-scale production, as shown in Figure 4. The unit 

essentially consists of a kettle to heat the asphalt and calibrated systems for asphalt, 

water, and air. It enables predetermined volumes of asphalt, water, and air to be 

injected into the expansion chamber where the foam is formed and is then 

discharged through a nozzle. The expansion ratio and half-life of the foam can be 

manipulated by altering the proportion of water that is added to the asphalt and the 

optimum addition of water determined. Once the design of the foam has been 

completed, the required volume of foamed asphalt is discharged directly into a 

sample of aggregate, while it is being agitated in a laboratory mixer. Normally five 

samples are produced in this way, with varying asphalt contents. Prior to mixing in 

the foamed asphalt, water is added to bring the material to Wadded of its optimum 

moisture content for compaction, where water added (Wadded) is determined as 

follows (Wirtgen, 2004): 

where 

Wadded = 1 + (0.5 WOMC − Wair-dry) (1) 

Wadded = pre-mixing water to be added to the sample 

WOMC = optimum moisture content 

Wair-dry = water in air dried sample 

If required, cement is also added to the aggregate before it is mixed with the 

foamed asphalt. After mixing with foamed asphalt, water is added to bring the total 

water content to the optimum moisture content of the aggregate. Both new 

aggregates and recycled asphalt pavement (RAP) were obtained from the field. New 

aggregates include MOT class A and B base and class B subbase (SB). The RAP 

materials were blended together with SB at a ratio of 50/50 by volume, to produce a 

combined gradation. 

9


Figure 4. Laboratory scale foamed asphalt plant WLB 10 

Two different foamed asphalt mixes were used in this study, SB and RAP + 

MOT class B base. Portland cement was evaluated at the ratios of 2 and 4%. 

Laboratory work carried out on the treated mixes consisted of: 

• Plant calibration and optimum foam production. 

• Establishing the optimum moisture/fluid content of the two different 

aggregate mixes. 

• Mixing varying percentages of foamed asphalt and water with samples of 

each mix type. 

• Compacting briquette specimens and curing the briquettes prior to testing. 

• Carrying out the required engineering tests. 

Foamed asphalt was produced using the laboratory foaming machine WLB 10. 

Calibration of the machine was needed to determine the flow rate of the asphalt at 

different temperatures and water ratios at a specific pressure (Wirtgen, 2004). The 

asphalt flow rate was measured at different temperatures ranging from 150°C to 180°C 

since the asphalt flow rate increases with the temperature. The amount of foaming water 

was varied at each temperature and the expansion ratio and half-life were measured for 

each water content. Figure 5 shows the variation of the expansion ratio and half-life at 

temperature of 180°C. It was found that the expansion ratio and half-life increase with 

increase in temperature. By comparing the foam characteristics at the three temperatures, 

it was found that the asphalt at 180°C produced the best foaming characteristic. The 

optimum water content was selected to provide the minimum expansion ratio of eight 

times and minimum half-life of 6 sec as explained in Figure 6 (Wirtgen, 2004). The 

optimum water content was found to be 3.3% at 180 o C. 



Expansion Ratio (times), Half-Life (sec) 

20 

16 

12 

8 

4 

0 

Expansion Ratio 

Half-Life 



2 3 4 5 

Water Content (%) 

Figure 5. Expansion ratio and half-life at 180°C 

Figure 6. Determination of optimum foaming water content (Wirtgen, 2004) 

11


Aggregate Gradation and Testing 

The aggregate gradation for the supplied materials was selected to meet MOT 

specifications shown in Table 1 for base material class A and B and SB. Aggregate 

envelope was selected at the mid-point of the upper and lower limits of the 

specifications. Aggregate size greater than 19 mm was eliminated since the 

recommended maximum aggregate size for the foamed asphalt mix design is 19 mm 

and therefore the aggregate gradation was adjusted accordingly. 

Sieve No. 

Table 1. Selected aggregate gradation 

Base Material 

Class A 

Base Material 

Class B 



Subbase Material 

Class B 

1 100 100 100 

¾ 87 96 93 

½ 69 86 60 

⅜ 56 75 78 

# 4 44 64 63 

# 10 29 50 48 

# 40 16 30 27 

# 200 8 15 13 

Aggregate materials were tested to determine the California bearing ratio 

(CBR) and optimum moisture content (OMC). The OMCs for all aggregates and 

RAP + MOT class B base blend are determined by using AASHTO T-180. Test 

results are shown in Figures 7 and 8 and summarized in Table 2. Determination of 

OMC is a prerequisite for foamed asphalt mix design. Experience has indicated that 

foamed asphalt and cold asphalt mixes will yield the highest density when 

compacted at OMC.



Figure 7. Relation between dry density and moisture content 

Figure 8. Relation between CBR and moisture content 

13


Table 2. OMC, maximum dry density and CBR for the received materials 

Specimen Type 

Optimum 

Moisture Content 

(%) 



Maximum Dry 

Density (g/cm 3 ) 

CBR (%) 

Base Class A 7.54 2.15 128 

Base Class B 7.53 2.11 217 

Subbase Class B 14.57 1.89 105 

Base Class B + RAP 7.59 2.01 51 

Optimum Foamed Asphalt Content 

The selected aggregate gradation for SB was mixed with different foamed 

asphalt percentages to find the optimum foamed asphalt content. Aggregate 

gradation of base class B was mixed with recycled material (RAP) by a ratio of 

50:50 by volume. Portland cement was added at the ratios of 2% and 4%. The 

blended mixtures were compacted by using 75-blow Marshall hammer into standard 

four-inch samples. The compacted specimens were then tested using indirect tensile 

strength (ITS @ 25°C) for dry and soaked conditions (24 hrs @ 25°C). The results 

are presented in Figure 9 for the dry specimens. 

Recommendations from experts in the Australian Road Research Board (ARRB) 

are that the dry indirect tensile strength should be at least 29 psi (200 kPa) and the 

soaked tensile strength be at least 14.5 psi (100 kPa) (SABITA, 1998). The design 

binder content should be selected as the binder content at which the soaked indirect 

tensile strength is at the maximum. For use as a base course layer in the pavement 

structure, where the water table is close to the surface, it is believed that an ITS of 

more than 29 psi (200 kPa) together with more than 80% retained strength of ITS will 

perform adequately (SABITA, 1998). The results show that all mix combinations 

passed the requirements at the dry condition as shown in Figure 9. Mixes without 

cement appear to be very sensitive to the effect of water as the ITS after soaking 

dropped below the target limit. Soaked SB at low-foamed asphalt content and 0% 

cement content showed complete collapse when soaked for 24 hrs. The addition of 

Portland cement significantly improved the soaked ITS; using lower cement content 

would be more economical. Therefore, 2% cement with 4% foamed asphalt resulted in 

an ITS of 26 psi, while the addition of 2% cement with 4% foamed asphalt to the RAP 

resulted in an ITS of 40 psi. It is noted that the retained ITS of 80% was not possible 

for SB nor economical. A summary of the ITS results at the optimum foamed asphalt 

content and different percentages is shown in Figure 10.

ITS (psi) 

100 

80 

60 

40 

ITS, psi 

SB+0% Cement 

SB+2% Cement 

SB+4% Cement 

RAP+0% Cement 

RAP+2% Cement 

RAP+4% Cement 

3 4 5 6 

Foamed Asphalt Content (%) 

Figure 9. Dry ITS at different cement contents 

120 

100 

80 

60 

40 

20 

0 

0% 

Cement 

2% 

Cement 

4% 

Cement 

Dry Soaked 

0% 

Cement 

2% 

Cement 

SB Class B RAP 

Mix Type 



4% 

Cement 

Figure 10. ITS at optimum foamed asphalt content and different cement contents 

15


Static Triaxial Shear Strength 

The triaxial shear strength (ASTM D 2850) was determined for the four 

aggregate mixes that include: base material class A, base material class B, SB + 4% 

foamed asphalt + 2% cement, and RAP + 4% foamed asphalt + 2% cement. The 

specimen size was 4 in. diameter and 8 in. height, prepared at the optimum water and 

foamed asphalt content compacted to the optimum density using Proctor compactor. 

Samples were cured for 72 hrs at 40°C and then tested for the static triaxial test. The 

test was performed at four different confined pressures up to failure. The test results 

were used to construct the Mohr-Coulomb failure envelope as shown in Figure 11. The 

angle of internal friction and cohesion were measured and are summarized in Figure 

11. It appears from the results that the highest shear strength is achieved by base class 

A followed by base class B and SB mix and finally by RAP mix. Upon soaking, the 

cohesion and the angle of internal friction dropped. The highest shear strength is 

achieved by base class A followed by base class B then by RAP and finally by SB 

mix. The highest cohesion is given by the RAP mix. 

Figure 11. Mohr-Coulomb failure envelope for tested materials 





Dynamic Resilient Modulus 

Dynamic resilient modulus (AASHTO T-307) was measured using the 

dynamic triaxial test at 25°C for four mixes including: base material class A (Figure 

12), base material class B (Figure 13), SB + 4% foamed asphalt + 2% cement 

(Figure 14), and RAP + 4% foamed asphalt + 2% cement (Figure 15). The samples 

were prepared similar to those of the static triaxial test. The specimens were tested 

under different levels of confined pressure (5-20 psi) and deviator stress (10-35 psi) 

to simulate the traffic loading that the granular base and subbase materials are 

subjected to in the road. Results fit a logarithmic relation between the bulk stress 

(deviator + 2x confining) and the resilient modulus with good regression correlation 

as shown in each figure. The results show that the addition of foamed asphalt to the 

base class B with RAP improves the MR significantly as shown in Figure 15. 

Moreover, granular base class B shows resilient modulus behavior similar to that of 

granular base class A. SB mix has provided an MR of 22 to 45 ksi, depending on the 

applied bulk stress. All the mixes showed a comparable rate of increase in the MR. 

RAP material treated with foamed asphalt has the best performance, followed by 

base course type B and A, and finally by SB treated with foamed asphalt. Upon 

soaking, the resilient modulus values dropped by a magnitude of about 50% of the 

dry resilient modulus. 

Resilient Modulus (ksi) 

100 

10 

Dry Specimens 

ln(Y) = 0.468*ln(X) + 2.000, R 2 = 0.79 

Soaked Specimens 

ln(Y) = 0.536*ln(X) + 0.899, R 2 = 0.75 

10 100 

Bulk Stress (psi) 

Figure 12. Variation of MR with bulk stress for base class A 

17



100 

10 

Dry Specimens 

ln(Y) = 0.558 * ln(X) + 1.723, R2 = 0.43 


ln(Y) = 0.803 * ln(X) + 0.205, R2 = 0.87 

10 100 


Figure 13. Variation of MR with bulk stress for base class B 


100 

10 

Dry Specimens 

ln(Y) = 0.476* ln(X) + 1.770, R2 = 0.77 


ln(Y) = 0.421* ln(X) + 1.564, R2 = 0.95 

10 100 


Figure 14. Variation of MR with bulk stress for foamed asphalt 

treated subbase class B 




100 

CONCLUSIONS 

10 

Dry Specimens 

ln(Y) = 0.536* ln(X) + 1.974, R2 = 0.66 


ln(Y) = 0.362* ln(X) + 1.685, R2 = 0.60 



10 100 


Figure 15. Variation of MR with bulk stress for foamed asphalt 

treated RAP material 

The intent of the study was to evaluate RAP and SB materials treated with 

foam asphalt and compare their behavior to that of granular base class A and B. 

Observation on the results can be summarized in the following points: 

1- Portland cement was effective in reducing stability loss for foamed asphalt 

mixes. 2% Portland cement was required to reduce stability loss to an 

acceptable value of ITS for roads that are not in a high water table area. 

2- Shear strength of base class A was the highest followed by base class B and 

foamed SB then RAP. Cohesion of foamed asphalt mixes was similar or less 

than that of un-stabilized mixes, indicating that foamed asphalt stabilization is 

not like emulsified/cutback asphalt stabilization. Foamed asphalt primarily 

binds fine particles. Saturation of mixes resulted in great reduction in the 

cohesion. 

3- Resilient modulus testing indicated that SB mix has behavior comparable to base 

class A. RAP mix has shown the best behavior. Saturation has reduced resilient 

modulus of all mixes significantly. 

4- Foamed asphalt can be used successfully to improve the quality of subbase 

materials and recycled road materials for road base construction. 

19


REFERENCES 

A.A. Loudon & Partners, 1996a, “In-Place Recycling with Foamed Bitumen”, Series 

of Technical Bulletins, South Africa. 

A.A. Loudon & Partners, 1996b, “Foamed Bitumen Mix Design Procedures”, Series 

of Technical Bulletins, South Africa. 

Asi, I.M., H.I. Al-Abdul Wahhab, M.I. Khan, and Z.U. Siddiqui, 1998, 

“Stabilization of Dune Sand for Base and Sub-base Applications Utilizing 

Foamed Asphalt Technology”, 1 st International Conference on Performance of 

Roads, Bridges and Airport Pavements in Hot Climates, Dubai, April 28-29, 

1998. 

Asi, I.M., Z.U. Siddiqui, H.I. Al-Abdul Wahhab, and M.I. Khan, 1999, 

“Stabilization of Sabkha Soil for Road Bases Using Foamed Asphalt 

Technology”, Fifth Saudi Engineering Conference, March 1999. 

Asi, I.M., H.I. Al-Abdul Wahhab, O.S.B. Al-Amoudi, and Z.U. Siddiqui, 2002, 

“Stabilization of Dune Sand Using Foamed Asphalt”, Geotechnical Testing 

Journal, ASTM, Vol. 5, Issue 2. 

Bowering, R.H. and C.L. Martin, 1976, “Performance of Newly Constructed Full 

Depth Foamed Bitumen Pavements”, Proceedings of the 8th Australian Road 

Research Board Conference, Perth, Australia. 

Chui-Te Chiu, M.U. Huang, and R.C. Lu, 2002, “A Study on Application of Foamed 

Bitumen Treated Base in Taiwan”, Final Report, Department of Civil 

Engineering, Chung Hua University, March. 

Csanyi, L.H., 1957, “Foamed Asphalt in Bituminous Paving Mixtures”, Highway 

Research Board Bulletin, No. 160, p. 108. 

Csanyi, L.H., 1962, “Foamed Asphalt for Economical Road Construction”, Civil 

Engineering, Vol. 32, No. 6, p. 62. 

Focus, 2003, “Foamed Asphalt – A Success on Federal Lands Highway”, U.S. 

Department of Transportation, FHWA, August. 

Kendall, M., B. Baker, P. Evans, and J. Ramanujan, 2001, “Foamed Bitumen 

Stabilization: The Queensland Experience”, 20 th Australian Road Research 

Conference, March 2001. 





Lee, D.Y., 1981, “Treating Marginal Aggregates and Soils with Foamed Asphalt”, 

Proceedings, Association of Asphalt Paving Technologists, Vol. 50, p. 211. 

Mobil Oil Australia Ltd., 1973, “Foamed Bitumen Cold Dustless Mixtures”, 

Technical Bulletin Bitumen, No. 5. 

Muthen, K.M., 1998a, “Foamed Asphalt Mixes: Mix Design Procedure”, Contract 

Report CR-98/077, Dec. 1998 © 1998 SABITA Ltd. & CSIR Transportek 

http://foamasph.csir.co.za:81/ 

Muthen, K.M., 1998b, “Foamed Asphalt Mix Design Procedure”, Technical Report, 

CSIR Transportek & SABITA Ltd., South Africa. 

SABITA Ltd. and CSIR Transportek, 1998, “Foamed Asphalt Mixes: Mix Design 

Procedure”, Contract Report CR-98/Draft, February. 

Soter International, 1994, “Foamed Asphalt Technology Transfer to the Kingdom of 

Saudi Arabia”, Preliminary Proposal, Canada. 

Wirtgen, GmbH, 2004, Cold Recycling Manual, Windhaven, Germany, ISBN 3-00- 

936215-05-7, 2nd edition. 

ACKNOWLEDGMENTS 

The authors would like to acknowledge the support provided by King Fahd 

University of Petroleum & Minerals (KFUPM) and Saudi Aramco for the execution 

of this research. 

21




EVALUATION OF SULFUR-ASPHALT TECHNOLOGY 

FOR LOCAL APPLICATIONS 

Hamad I. Al-Abdul Wahhab and Mirza G. Baig 

Department of Civil Engineering, King Fahd University of Petroleum & 

Minerals, Dhahran 31261, Saudi Arabia, Email: hawahab@kfupm.edu.sa 

ABSTRACT 

This research work was planned with the objective of exploring the possibility 

of utilizing sulfur as a binder for making sulfur asphalt in light of the new 

developments in this area in recent years. The study on sulfur-asphalt concrete 

consisted of testing Devco sulfur-extended asphalt modifier (SEAM), with local 

asphalt-concrete mixes and to assess the effect of sulfur and modified sulfur 

materials by comparing with the performance of local plain and sulfur extended 

asphalt-concrete mixes. Results indicated that SEAM modified asphalt concrete can 

be produced, hauled, placed and compacted easily with conventional methods and 

equipment. There will be no constructability problem with the use of SEAM binder. 

SEAM additive has provided asphalt-concrete mixes a superior stability compared to 

the conventional asphalt mix. Also, it was noted that the stability result (14.4 kN) 

obtained with sulfur 30/70 mix is higher than the value (11.1 kN) reported in 

previous local studies for the same type of mix with 1% cement added. Use of 

SEAM material at the volume replacement ratio of R=1.5 proved to be economical 

as compared to regular asphalt as the amount of required asphalt is reduced in 

proportion to the SEAM percentages added. The tests on assessing the 

environmental impact of this sulfur-asphalt technology show that there is no longterm 

hazard for mixes as indicated by acceptable values of emission of hazardous 

gases such as H2S and SO2 (

EVALUATION OF SULFUR-ASPHALT TECHNOLOGY FOR LOCAL APPLICATIONS 

BACKGROUND 

Asphalt binder is a thermoplastic material that behaves as an elastic solid at 

low service temperatures or during rapid loading and behaves as a viscous liquid at 

high temperatures or slow loading. This double behavior creates a need to improve 

the performance of an asphalt binder in order to minimize stress cracking, which 

occurs at low temperatures, and permanent deformation, which occurs at high inservice 

temperatures. Daily and seasonal temperature variations plus the growth in 

truck traffic volume, tire pressure and loading have increased stresses on asphalt 

pavements. Local asphalt pavement temperatures range between −10°C in winter and 

73°C in summer (Al-Abdul Wahhab et al., 1997). This has led to an increased 

demand to modify asphalt binders. Different methods have been used to upgrade the 

properties of asphalt binders. One of the most commonly used procedures is the 

modification of asphalt binders by the addition of modifiers. 

Sulfur’s ability to modify and enhance the properties of construction materials 

has been extensively researched and exploited over the past four decades. Most of 

the research work has come to a halt as a result of the historic market collapse about 

two decades ago. Since then, the amount of sulfur produced from oil and gas has 

increased resulting in a drop of market prices. For the last five years, world sulfur 

production has exceeded consumption by one to two million tons per year. Some 

forecasts have predicted continued overproduction ranging from one to three million 

tons per year through 2010 (Weber and McBee, 2000). In the last decade, the 

availability of sulfur has considerably grown in many countries. This is mainly due 

to the current environmental restrictions regarding the petroleum and gas refining 

processes, which limit the maximum quantity of sulfur present at combustibles. 

Extremely large quantities of sulfur are thus obtained as a by-product of these 

processes, together with coal processing and refining of copper in the mining sector. 

The development of new applications for sulfur becomes fundamental. Sulfur 

asphalt used for paving roads is one of the prominent among its applications. Using 

sulfur to enhance or rejuvenate asphalt could consume significant amount of sulfur, 

even if sulfur captured only a conservative 5% of the current asphalt market, it 

would represent a market for nearly one million ton of sulfur annually and can help 

alleviate the oncoming sulfur surplus. Sulfur-asphalt concrete has a relatively simple 

composition and manufacture, and very interesting characteristics and properties. Its 

extremely high corrosion resistance, mechanical strength and fast hardening make it 

a high performance material suitable for several applications, especially the ones in 

which other materials may not suffice (Gracia and Vazquez, 2003). 

In 1978, the Metrology, Standards and Materials Division of the Research 

Institute at King Fahd University of Petroleum and Minerals (KFUPM/RI) launched 

an in-house research study on sulfur-asphalt pavement development. Among the 

various available techniques of substituting asphalt with sulfur, the sulfur-extended 

asphalt (SEA) paving technology developed by Gulf Canada was considered to be 




Hamad I. Al-Abdul Wahhab and Mirza G. Baigl 

the closest to practical applications. Three SEA test roads were laid in the Eastern 

Province in cooperation with Gulf Canada and the Ministry of Communications 

(MOC), Saudi Arabia as part of the ongoing road development program of MOC. A 

sulfur/asphalt ratio of 30/70 by weight was used in Test Road 1 (Kuwait Diversion) 

and Test Road 3 (KFUPM), whereas a higher percentage of 45/55 was used in Test 

Road 2 (Abu-Hadriyah Expressway). Performance of the test roads was monitored 

from time to time. For each test road, the control section using normal asphalt 

concrete showed better performance than the SEA sections (Arora et al., 1994). The 

SEA sections developed mostly longitudinal/ transverse cracking in Test Road 1; 

alligator and block cracking in Test Road 2; and block and longitudinal/transverse 

cracking in Test Road 3. On the control AC sections, the most predominant distress 

types were found to be longitudinal/traverse cracking and polishing of aggregate. 

The inherently stiffer SEA mix of Test Road 2, where the sulfur/asphalt ratio was 

45/55, resulted in earlier cracking of SEA pavement, particularly in a thinner section 

where the thickness was intentionally reduced by 20 percent. 

Fatani and Sultan (1982) conducted a study to determine the feasibility of 

using dune sand in asphalt-concrete pavement in hot, desert like climates through the 

use of one-size crushed aggregates. Dense-graded aggregate and powdered sulfur 

were used in the sand-asphalt mixes. Engineering properties, including Marshall 

design parameters, compressive strength, tensile strength, modulus of rupture, and 

dynamic modulus of elasticity were evaluated. Results indicated that a mixture of 

dune sand and asphalt is weak, unstable, easily deformed under light loads, and 

therefore unacceptable for pavement construction in desert like environment. The 

use of powdered sulfur and sand-asphalt mixes reduces the optimum asphalt content, 

increases considerably the qualities of the mix even under severe environmental 

conditions, and reduces the pavement thickness. 

Arora and Abdul-Rahman (1985) have explored the use of sulfur as a 

rejuvenation agent in recycling reclaimed asphalt pavement from a typical failed 

segment of Dammam-Abu-Hadriyah Expressway. They indicated that the addition 

of sulfur, at mixing temperature, would lower the viscosity of the aged asphalt. 

Upon cooling, recrystallization of sulfur is known to occur, which improves the 

strength of the mix. Properties like Marshall stability, resilient modulus and fatigue 

behavior of sulfur-recycled mix are compared with those of the conventional 

asphalt-concrete mix. The addition of sulfur results in higher Marshall stability 

without significant loss in flow values, higher retained strength index, and higher 

MR and tensile strength, indicating superior engineering properties of the recycled 

mixture over the conventional asphalt hot mix. The above properties are particularly 

advantageous to the hot region of the Arab World since they provide adequate 

resistance to wheel track rutting otherwise associated with conventional asphaltconcrete 

mixtures. 

25


Akili (1985) carried out an extensive laboratory testing program designed to 

measure improvements in engineering properties of sulfur-asphalt-sand (SAS) mixes 

attributable to the presence of sulfur in the mix, considering locally available sands 

and prevailing environmental conditions in eastern Saudi Arabia. The laboratory 

characterization data include: Marshall design parameters, resilient moduli, and 

permanent strain characteristics under repeated triaxial loading. The results, in 

general, showed improvements in Marshall stability, resilient modulus values and 

reduced permanent deformation of SAS mixes in comparison to conventional sandasphalt 

mixes. From Spring 2001 through February 2002, about 42 lane miles of 

roads containing sulfur were built in the southwest U.S. These projects incorporated 

a formed, solid sulfur product that was added directly to existing hot mix plant 

equipment. Following mixing, the sulfur asphalt was hauled to the project location 

using conventional dump trucks, road paving, and compaction equipment. An 

additional 104 lane miles of roads containing sulfur are planned in the southwest 

U.S., and other road projects incorporating sulfur are also being considered in China, 

Kazakhstan, and Egypt. The use of a formed, solid material and the direct mixing 

method minimizes hot asphalt mix plant modifications and associated costs. Also, 

solid sulfur can be shipped freely without regulation; whereas, liquid sulfur requires 

special shipping considerations (Weber, 2002). 

The use of sulfur as an additive to extend or replace asphalt has been 

demonstrated successfully in both laboratory tests and actual construction. The 

current availability and low cost of sulfur offer the potential to reduce paving 

material costs by as much as 21 percent. Binder cost reductions as high as 32 percent 

are feasible (Weber and McBee, 2000). 

Sulfur asphalt is enjoying a resurgence of interest worldwide. The Sulfur 

Institute (TSI) has been actively working with the US Federal Highway 

Administration and through other channels to promote the utilization of sulfur 

asphalt, and at least one world-class oil and gas company is seriously interested in 

this outlet for its recovered sulfur. At the same time, Devco Company has already 

moved to penetrate the market. Using proprietary innovations to produce easily 

handled forms of pre-blended sulfur asphalt, the company has been constructing 

roads in Nevada and California for the last two to three years, along with parking 

lots and other paving projects. 

King Fahd University of Petroleum and Minerals has conducted major sulfur 

research in the late 70’s and early 80’s in cooperation with the Ministry of Transport 

where sulfur was incorporated in the construction of major road sections, some of 

which are still functioning. Sulfur is added to asphalt to overcome temperature 

susceptibility (dependency) of local binders, thereby reducing or eliminating rutting 

tendency of local mixes. By mid 1980’s, sulfur research had been stopped due to the 

disadvantages of sulfur mixes. Several agencies such as The Sulfur Institute (TSI), 

Devco Company and others have continued to improve sulfur mixes to eliminate 





their disadvantages. Research on development of new construction materials such as 

sulfur polymer cement concrete, plasticized sulfur and polymer asphalt 

extender/replacement may be found to be fruitful. 

This study was carried out to study the effect of sulfur and sulfur-extended 

asphalt modifier (SEAM) as compared to conventional local asphalt mixes and 

verify their adequacies for safe use locally. 

EXPERIMENTAL DESIGN 

Material Selection 

The locally available aggregate in the Eastern Province was selected for 

this study. The Ministry of Transport (MOT) specification BWC-1 for wearing 

course layer was selected and followed in all mix designs, and is shown in Table 

1. Asphalt cement of grade 60/70 was obtained from Ras Tanurah refinery. The 

materials were tested for gradation and physical characteristics to assure their 

conformity to MOT standards. 

Table 1. Aggregate gradation and job mix formula (JMF) as per MOT BWC-1 

specification 

Sieve Size % Passing Limits JMF 

¾″ 100 100 

½″ 75–90 82.5 

3/8″ 64–79 71.5 

#4 41–56 48.5 

#10 23–37 30 

#40 7–20 13.5 

#80 5–13 9 

#200 3–8 5.5 

The yellow sulfur in pellet form, used in the study was obtained from Saudi 

Aramco. Sulfur-extended asphalt modifier (SEAM), which was manufactured by 

Rock Binder/Shell Co., was obtained from Shell (Canada). This material was 

selected to compare the improvements in engineering properties obtained by using it 

over the sulfur-asphalt mixes and conventional asphalt mix. 

27


Sulfur-Asphalt and SEAM-Asphalt Blends 

In addition to the conventional asphalt mix, six blends of modified asphalts 

were prepared by adding elemental sulfur or SEAM, each with ratios of 30/70, 40/60 

and 50/50 by weight of binder. Each blend was prepared by heating the plain asphalt 

to 140°C and slowly adding the required amount of modifier (sulfur or SEAM) as 

per the weight ratios using a shear-blender. The mixing was carried out for 10 

minutes in order to get a well-homogenized blend. 

Mix Design 

Seven different asphalt-concrete mixes were designed using the standard 

Marshall Mix design method as per ASTM D 1559 test method with 75 blows, and 

the optimum asphalt content (OAC) was determined. The first mix was the control 

mix having plain asphalt as binder. Three other mixes were prepared with sulfur 

modified asphalt, in which the sulfur/asphalt ratios were 30/70, 40/60 and 50/50. 

Similarly, the 

last three mixes were prepared with SEAM modified asphalt, in which the 

SEAM/asphalt ratios were kept the same as 30/70, 40/60 and 50/50 (as 

recommended by Shell (Canada)). In all these seven mixes, the equal volume 

replacement of asphalt by sulfur or SEAM was used. This corresponds to a sulfur 

substitution ratio R = 1.93, where R for this case corresponds to the ratio of specific 

gravity of sulfur to that of the asphalt. All the mixes were compacted at 145°C, and 

care has been taken not to exceed 150°C temperature limits for sulfur/SEAM 

modified mixes. 

Performance Testing 

In order to evaluate the basic engineering properties and environmental impact 

of SEAM and sulfur modified mixes, the following standard tests were carried out 

on each blend type and on the control mix: 

1. Marshall stability loss test at 60°C. 

2. Resilient modulus test at 25°C. 

3. Laboratory and field environmental pollution monitoring. 

The environmental pollution monitoring is done by using gas analyzer to 

measure the gases that are expected to be emitted by asphalt concrete mixes 

prepared using sulfur or SEAM as modifier. The measurements were carried out in 

the lab environment and field. Gases emitted directly from source (using a sample 

probe at the point of origin) were taken at different temperatures that simulate 

construction and in-service temperatures. The emission of two most toxic gases, 

sulfur dioxide and hydrogen sulfide, was expected due to heating and mixing of 





sulfur or SEAM at high temperature of construction (145°C) with asphalt and 

aggregate and at the high service temperature of 76°C. Therefore, two temperatures 

were used in this test, 145°C and 76°C. 

Asphalt was heated first and mixed with sulfur or SEAM at certain 

percentages by weight of asphalt using a mixer. The resulting blend was added to the 

hot aggregates (2400 gm) and mixed thoroughly at 145°C. Five different asphaltconcrete 

mixtures prepared and tested for gaseous emissions are as follows: 

1. Plain asphalt mixture as control mix (CT) 

2. SEAM: Asphalt is 30:70 by wt. (SE-30) 



5. Sulfur: Asphalt is 30:70 by wt. (SU-30) 

The prepared mixtures were immediately placed in a closed oven at 145°C. 

One and a half inch diameter outlet was used as opening for gaseous emission to 

move out from the oven. This oven outlet was used for inserting the probe tube for 

sampling gaseous sample and was taken through a sampling line to the designated 

analyzer. The gaseous emission of the flue gas was monitored using a gas analyzer. 

The sample was also continuously swept by a steady flow of air. 

RESULTS 

Properties of Sulfur and SEAM Modified Asphalt Blends 

Viscosity test at 135°C using rotational viscometer and specific gravity test as 

per standard ASTM D 70 test method was carried out on plain asphalt and all the six 

modified asphalt blends. The results obtained are shown in Table 2. The results 

indicate that the addition of SEAM or sulfur has reduced the viscosity of the original 

asphalt and this will aid in mixing and compaction of mixes at a lower temperature. 

Table 2. Specific gravity and viscosity of asphalt blends 

Blend Type Average Specific Gravity Viscosity at 135°C, (cP) 

Plain asphalt (Control) 1.026 343 

Sulfur-asphalt 30/70 

1.195 

145 

" " 40/60 

1.215 

190 

" " 50/50 

1.338 

263 

SEAM-asphalt 30/70 

1.126 

173 

" " 40/60 

1.177 

215 

" " 50/50 

1.334 

287 

29


Optimum Asphalt Content 

Using sulfur or SEAM modified asphalt blends as binder, seven mix 

designs were carried out as per the standard Marshall method and MOT 

specifications. It was noted that the stability result (14.4 kN) obtained in this 

study with sulfur 30/70 mix is higher than the value (11.1 kN) reported in 

previous local studies for the same type of mix with 1% cement. Also, the air 

voids in the compacted modified mixes increased with the amount of SEAM or 

sulfur added, which is higher in the case of both modifiers compared to the 

control mix and is in accordance with previous studies by Akili and Arora 

(1985). The description of each mix type along with the assigned mix ID and the 

optimum asphalt content (OAC) obtained are shown in Table 3. 

Table 3. Description and OAC for each designed mix 

Mix ID Description OAC, % 

CT Control mix 

(having plain asphalt as binder) 

4.5 

SU-30 Sulfur-asphalt 30/70 mix 

5.0 

SU-40 " " 40/60 mix 

5.5 

SU-50 " " 50/50 mix 

5.8 

SE-30 SEAM-asphalt 30/70 mix 

4.8 

SE-40 " " 40/60 mix 

5.3 

SE-50 " " 50/50 mix 

5.7 

Marshall Stability Loss 

Six Marshall specimens were prepared at the OAC for each blend type. These 

specimens were then tested for Marshall stability loss at 60°C as per the standard 

test procedure ASTM D 1559. The results obtained are shown in Figure 1. The 

results indicate that SU-30, SE-30 and SE-40 mixes have a stability loss of less than 

20 percent (maximum 20% allowed as per MOT specifications). In the other mixes, 

the stability loss is more; it means these mixes will require additives to control the 

damage due to water and, hence, the stability loss. 



Percent loss 

40 

30 

20 

10 

0 

Marshall Stability Loss 



CT SU-30 SU-40 SU-50 SE-30 SE-40 SE-50 

Mix Type 

Figure 1. Comparison of Marshall stability loss for each mix type 

Resilient Modulus (MR) Test 

Three Marshall specimens from each blend type were subjected to resilient 

modulus test at 25°C using a repeated load testing system. A sinusoidal axial 

compression stress is applied to each specimen at a loading frequency of 0.1 second. The 

resulting recoverable strain response of the specimen is measured. The resilient modulus 

is calculated as the ratio of stress to strain. The results obtained are shown in Figure 2 as 

average resilient modulus (MR) for each blend type. Higher values of MR are found for 

both SEAM and sulfur modified blends compared to the control mix. The values have 

increased as the percentage of SEAM or sulfur added is increased. It indicates that the 

modified mixes are getting stiffer with the addition of SEAM or sulfur. 

Environmental Pollution Monitoring 

Measurement of gases that are expected to be emitted by modified asphalt 

concrete mixes prepared using sulfur or SEAM as modifier was evaluated in the lab 

and field. In the laboratory, a bench scale study using a standard gas analyzer at two 

different temperatures that simulate construction (145°C) and high in-service 

temperatures (76°C) was performed. The results of gaseous measurements obtained 

are shown in Figure 3. The analysis results for CO and NO were all below 2 ppm 

even at mix preparation temperature of 145ºC and are, therefore, not included. In the 

case of plain asphalt concrete blend (control mix), the gaseous emission at 145°C 

was not noticeable as indicated in Figure 3. 

31


Resilient Modulus, ksi 

Emission in ppm 

1200 

1000 

800 

600 

400 

200 

0 

Resilient Modulus (MR) at 25 o C. 

CT SU-30 SU-40 SU-50 

Mix Type 

SE-30 SE-40 SE-50 

Figure 2. Comparison of resilient modulus for each mix type 

1000 

100 

10 

1 

0.1 

SO2 

H2S 

Gas Emission from Source at 145 O C. 

Control (CT) SE-30 SE-40 SE-50 SU-30 

Blend Type 

Figure 3. Gas emission for each mix type in the lab 





In the modified mixtures prepared using SEAM (SE-30, SE-40 and SE-50) or 

sulfur (SU-30), it was found that at 145°C the average source concentration of SO2 

(sulfur dioxide) was in the order of 60 to 70 ppm as shown in Figure 3, which 

exceeds the permissible limit of sulfur dioxide (2.0 ppm) in the lab environment. It 

may be noted that the source emission in the lab environment does not simulate the 

conditions in actual practice, where the mixing is carried out in an open space with 

much easy dilution of the emission gases from the source into the nearby 

environment. When these measurements were carried out in the field, it was found 

that the maximum values of SO2 concentrations ranged from 0.0 to 12 ppm when 

measured close to the source (20−40 cm above the auger). The maximum measured 

values were 8, 12 and 8 ppm for 30/70 sulfur, 40/60 SEAM and 30/70 SEAM mixes, 

respectively. The measured values of SO2 were found to be within the acceptable 

limits when measured at the foreman and driver levels ranging from 0.3 to 0.4. The 

measured values of hydrogen sulfide ranged from 0.0 to 3.17 ppm close to the 

source and 0.47 to 0.51 at the foreman and driver levels. It was also noted that the 

variation of temperature (124−147°C) during spreading and compaction did not 

significantly affect the concentration of fumes. 

Apart from the measured SO2 emission close to the source above the auger 

screw which continuously agitates asphalt-concrete mix and releases trapped fumes, 

all measured gases concentrations are within the acceptable limits. Workers should 

not stand in downwind direction of the paver close to the auger; otherwise, special 

safety precautions should be taken. Nevertheless, full precautions should be taken 

during mixing and construction phase of a road project involving SEAM/sulfur 

modified asphalt concrete. This can be done by using appropriate gas masks. 

The emission value for test samples at 76°C simulates emission from the road 

surface under maximum temperature in Saudi Arabia. Under these conditions, the 

emissions were well below acceptable values, as confirmed by the low results 

obtained after 24 hours at 76°C. 

CONCLUSIONS 

Sulfur has been used previously worldwide and also in Saudi Arabia as an 

asphalt modifier. But due to environmental concerns and higher market price of 

sulfur, its use was discontinued. However, there has been considerable progress 

achieved in the polymer modified asphalt additives and binder technology to prepare 

more stable and durable sulfur-asphalt mixes. Also, due to unbalance of 

supply/demand situation in the sulfur industry, the sulfur market price has also 

dropped considerably (from $120-150/ton to about only $40-60/ton of sulfur). Both 

of these factors have renewed worldwide interest to examine and develop sulfur 

asphalt into utilizable technology. 

33


The following conclusions were drawn based on this research: 

1. SEAM modified asphalt concrete can be produced, hauled, placed and 

compacted easily with conventional methods and equipment. There will be 

no constructability problem with the use of SEAM binder. 

2. SEAM additive has provided asphalt-concrete mixes a superior stability 

compared to the conventional asphalt mix. Also, it is noted that the stability 

result (14.4 kN) obtained with sulfur 30/70 mix is higher than the value 

(11.1 kN) reported in previous study for the same type of mix with 1% 

cement added. 

3. Both the SEAM and sulfur modified mixes have improved the resilient 

modulus at each percentage as compared to the control mix. 

4. There is no short-term environmental hazard for sulfur-asphalt mixes 

prepared using SEAM as indicated by the low emission of hazardous gases 

at road surface temperature as high as 76ºC prevailing in Saudi Arabia nor 

during field construction if workers do not stand in downwind direction of 

the paver close to the auger. Precautions should be taken during mixing and 

construction phase of a road project involving SEAM modified asphalt 

concrete. This can be done by using appropriate gas masks. 

5. Use of SEAM material can show economical advantage as well as savings 

of asphalt. The actual economic advantage and the optimum sulfur/asphalt 

ratio to be used will depend on the market price of SEAM, the amount of 

binder in the mix as well as on the asphalt replacement ratio. 

REFERENCES 

Akili, W., 1985, “Sulphur-Asphalt-Sands for Pavement Applications in Eastern 

Saudi Arabia”, Proceedings of the 2 nd Arab Regional Conference on Sulphur 

and Its Usage, Vol. III, pp. 55-71. 

Akili, W. and M.G. Arora, 1985, “Post Construction Evaluation of Sulphur- 

Extended Asphalt (SEA) Pavement Sections in Dammam-Dhahran Area”, 

Proceedings of the 2 nd Arab Regional Conference on Sulphur and Its Usage, 

Vol. III, pp. 19-39. 

Al-Abdul Wahhab, H.I., I.M. Asi, I.A. Al-Dubabe, and M.F. Ali, 1997, 

“Development of Performance-Based Bitumen Specifications for the Gulf 

Countries”, Construction and Building Materials Journal, Vol. 11, No. 1, pp. 

15-22. 





Arora, M.G. and K.M. Abdul-Rahman, 1985, “Sulphur in Recyling Old Asphalt 

Pavements,” Proceedings of the 2 nd Arab Regional Conference on Sulphur and 

Its Usage, Vol. III, pp. 37-53. 

Arora, M.G., A.I. Al-Mana, A.J. Al-Tayyib, R.H. Ramadhan, and Z.A. Khan, 1994, 

“Long-Term Pavement Performance History of Sulfur-extended Asphalt Test 

Roads in Eastern Province of Saudi Arabia”, Transportation Research Record 

14315, Transportation Research Board, pp. 77-85. 

Fatani, M.N. and H.A. Sultan, 1982, “Dune Sand-Aggregate Mixes and Dune Sand- 

Sulfur Mixes for Asphalt Concrete Pavements”, Transportation Research 

Record No. 843, National Academy of Science, Washington, D.C., pp. 72-80. 

Gracia, V. and E. Vazquez, 2003, “Utilization of By-produce Sulfur for the 

Manufacture of Unmodified Sulfur Concrete”, Technical Report, Department of 

Construction Engineering, University of Tecnica Federico, Santa Maria, Spain. 

U.S Department of Labor, “Guide on Occupational Safety & Health Administration, 

Limits for Air Contaminants 29 CFR 1910.1000”, Washington DC, 

www.osha.gov. 

Weber, H.H., 2002, “Market Opportunities for Sulphur Asphalt Road Paving 

Materials”, presented at The Sulfur Institute’s Eighth International Symposium 

‘Sulphur Markets – Today and Tomorrow’, Amsterdam, The Netherlands, 

March 10-12, 2002. 

Weber, H.H. and W.C. McBee, 2000, “New Market Opportunities for Sulfur 

Asphalt”, presented at The Sulfur Institute’s Seventh International Symposium 

‘Sulfur Markets – Today and Tomorrow’, Washington, D.C., March 28, 2000. 


The authors would like to acknowledge the support provided by King Fahd 

University of Petroleum & Minerals (KFUPM) and Saudi Aramco for the execution 

of this research. 

35




APPLICATION OF ARTIFICIAL NEURAL NETWORKS FOR THE 

PREDICTION OF STRENGTH AND PERMEABILITY OF HIGH 

PERFORMANCE CONCRETE 

ABSTRACT 


Department of Civil Engineering, College of Engineering 

King Saud University, Kingdom of Saudi Arabia 

Email: miqbal@ksu.edu.sa 

This paper presents the application of artificial neural networks for the 

prediction of strength and permeability of high performance concrete. Compressive 

strength and permeability of concrete incorporating cementitious materials as partial 

cement replacement prepared with various water-binder ratios are reported. Based 

on the experimentally obtained results, the applicability of artificial neural network 

for the prediction of strength and permeability has been established. The predicted 

values obtained using artificial neural networks have a good correlation between the 

experimentally obtained values. Therefore, it is possible to predict strength and 

permeability of high performance concrete using artificial neural networks. 

KEYWORDS 

Artificial neural networks, high performance concrete, permeability, strength. 

INTRODUCTION 

High-performance concrete (HPC) is relatively a new terminology used in the 

concrete construction industry. HPC is designed to give optimized performance 

characteristics for the given set of materials, usage and exposure conditions, 

consistent with requirements of cost, service life and durability. The ACI Committee 

[1] defines HPC as, “Concrete meeting special performance requirements 

which cannot always be achieved routinely using any conventional constituents and 

normal mixing, placing and curing practices. These requirements may involve 

enhancements of the following: ease of placement without segregation, long-term 

mechanical properties, early-age strength, toughness, volume stability and life in 

severe environments”. Therefore, high performance describes a concrete which is 

superior to ordinary concrete with respect to particular design properties because it 

has been tailored and optimized for every special application. HPC should have both 

high strength and high durability properties pertinent to an application. 


APPLICATION OF ARTIFICIAL NEURAL NETWORKS FOR THE PREDICTION OF STRENGTH 

In order to produce HPC a very dense homogeneous concrete microstructure 

especially in the interface region between hydrated paste and aggregate is required 

[2-4]. This is generally achieved through the use of low water-binder ratio between 

0.20 and 0.30 with the help of superplasticizers that can produce slumps ranging 

from 70 to 130 mm. Additional densification and homogeneity of the interfacial 

region are achieved through the incorporation of mineral admixtures which improve 

concrete microstructure. Therefore, in addition to the three basic ingredients in 

conventional concrete, i.e., Portland cement, fine and coarse aggregates, and water, 

the making of HPC needs to incorporate supplementary cementitious materials, such 

as fly ash (FA), silica fume (SF) and/or blast furnace slag, and chemical admixture, 

such as superplasticizer. HPC can be manufactured involving up to 10 different 

ingredients whilst having to consider durability properties in addition to strength. 

Since the number of ingredients and the number of properties of HPC, which 

needs to be considered in its design, are more than those for ordinary concrete. 

Therefore, it is difficult to predict the properties of this type of concrete using 

statistical empirical relationship. An alternative approach is to use an artificial neural 

network (ANN). The ANN approach is good for modelling non-linear systems. A 

neural network model is a computer model whose architecture essentially mimics 

the learning capability of the human brain. 

In this investigation, ANN based on the radial basis function (RBF) have been 

used. A RBF neural network is a layered network consisting of an input layer, an 

output layer and at least one layer of nonlinear processing elements known as hidden 

layer. The input layer of the neural network receives signals from the external 

environment. The hidden layer receives signals from the input layer and transmits an 

output signal based on a transfer function to the subsequent layer. 

NONLINEAR MODELING 

In this investigation, a nonlinear autoregressive model with exogenous inputs 

(NARX) [5], which provides a concise representation for a wide class of nonlinear 

systems, was used. An RBF network was employed to model the input-output 

relationship described as follows. 

Radial Basis Function 

An RBF network can be regarded as a special two-layer network which is 

linear in the parameters provided all the RBF centres are prefixed. Given fixed 

centers i.e. no adjustable parameters the first layer or the hidden layer performs a 

fixed nonlinear transformation, which maps the input space onto a new space. The 

output layer then implements a linear combiner on this new space and the only 





adjustable parameters are the weights of this linear combiner. These parameters can 

therefore be determined using the linear least square method, which is an important 

advantage of this approach. A schematic of the RBF network with n inputs and a 

scalar output is shown in Fig. 1. Such a network could be represented as 

∧ 

n 

∑ 

( ) 

yt ( ) = w + wf xt ( ) −c 

0 

i= 

1 

i i 

where: 

y(t) ∧ 

is the network predicted output; 

x(t) is the network’s input vector and presented as 

xt ( ) = [( yt ( −1),... yt ( −n), ut ( −1 ),... ut ( −n)] 

y u T 

wi are the weights or parameters; 

ci are known as RBF centres 

nr is the number of centers or the hidden neurons. 

x1(t) 

x2(t) 

xn(t) 

Hidden Layer 

( ( ) i c t x f − 

) 

wi 

w0 

Σ 

Fig. 1: Radial basis function network. 

1 

Linear Combinator 

Nonlinear Transformation 

(1) 

y(t) ∧ 

39


The nonlinear functional form f( . ) in the RBF expansion, used in this study is 

the Guassian function. The orthogonal least square provides an elegant method for 

determination of model structure as well as parameter estimation [6]. Once the 

functional form f(.) and the centres ci are fixed, and the set of input x(t) and the 

corresponding desired output vector (y(t) in this study) provided, the weights wi can 

be determined using the linear least squares method. Assuming the RBF network in 

Eq. 2 as a special case of the linear regression model is presented as follows: 

M 

∑ 

i = 1 

y( t) = p ( t) θ + ε ( t) 

(2) 

i 



i 

where: 

y(t) is the desired output; 

pi are known regressors, which are some nonlinear functions of lagged outputs and 

inputs. 

A constant term (w0 in Fig. 1) can be included in Eq. 2 by setting the corresponding 

term pi (t) = 1. The residual ε (t) is assumed to be uncorrelated with the regressors pi 

(t). It is clear that a given centre ci with a given nonlinear function f( . ) corresponds 

to pi (t) in Eq. 2. 

EXPERIMENTAL PROGRAMME 

Type-I cement complying with ASTM C150, fly ash and silica fume were 

used throughout the investigation. The SF was obtained in slurry form with solids to 

water ratio of 50/50 by weight. A sulphonated naphthalene formaldehyde 

condensate superplasticizer was used to disperse the slurry. Fine aggregate (quarry 

sand) and coarse aggregate (uncrushed gravel) of 10 mm nominal size, were used. 

The fine aggregate was of medium grading in accordance with standard 

specifications. The aggregates were air-dried before use, and allowance was made 

for absorption when calculating batch weights. Mixes containing 0 – 40% FA were 

made to which 0 – 15% SF replacement levels were incorporated to make various 

binary and ternary cementitious combinations. Various water-binder ratios were 

used. The slump for all the mixes investigated was maintained at 125±10 mm using 

the superplasticizer. The water contents of superplasticizer and SF slurry were taken 

into account when calculating the batch weights for mixing. 

Cube compressive strength (100 mm cubes) was carried out in accordance 

with BS 1881: 1983. The sample preparation and testing procedure for permeability 

was similar to those used in other investigation [7]. All the specimens were cast and 

compacted in accordance with standard specifications. After casting, the samples 

were covered under damp burlap and polyethylene sheets for 24 hours. The samples 

were demoulded the following day and then immediately kept in a mist room at 

20±2 o C and 98±2% RH prior to testing.

RESULTS AND DISCUSSION 

Compressive Strength 



The change in compressive strength of concrete at various ages caused by the 

interactive effects of FA and SF contents is demonstrated in Fig. 2. It can be seen 

that compressive strength decreased with an increase in FA content for all ages 

investigated (Fig. 2). SF affected the strength of FA mixes and this seems to be 

related to the FA content. An increase of strength is registered for FA levels lower 

than 10% when SF is incorporated, however the results suggest that at higher FA 

levels (>30%) the incorporation of SF results in a reduction in strength. At 28 days, 

up to 10% SF increased the strength for all levels of FA replacements, whilst SF 

above 10% did not result in any advantage in improving the strength. Therefore, 

10% SF seems to be the optimal replacement level. This optimal level of 10% SF is 

shown in Fig. 3. This figure shows the interactive isoresponse for 28-day 

compressive strength of concrete prepared with w/b ratio 0.27. At 90 and 180 days, 

only a modest improvement in strength has resulted from SF incorporation and this 

was evident for low levels of FA (


Permeability 

Compressive Strength, MPa 

105 

95 

85 

75 

0 

10 

20 

30 

PFA Content, % 

Fig. 3: 3D Isoresponse for 28-day compressive strength of concrete 

The influence on permeability of concrete at various ages is shown in Fig. 4. It 

is evident from this figure that the incorporation of FA had a very limited influence 

on permeability at ages up to 90 days but at 180 days slight reductions in oxygen 

permeability, associated with the level of replacements, were recorded for SF levels 

up to 5%. The inclusion of SF, however, resulted in more significant reductions in 

permeability for mixes with and without FA at all ages investigated. The reduction 

in the permeability was greater when SF was incorporated at up to 10% replacement 

level, above that the reduction was marginal. It can be seen that 8 to 12% SF 

incorporation seems to be optimal level for the reduction of oxygen permeability. 

The optimum replacement level's of SF is demonstrated in three-dimensional 

diagram in Fig. 5. This figure shows the interactive isoresponse for 28-day 

permeability of concrete. 



40 

15 

10 

5 

0 

SF Content, %

Oxygen Permeability, x 10-16, m2 

3.0 

2.5 

2.0 

1.5 

1.0 

0.5 

Permeability, x 10^-16, m^2 


w/b ratio: 0.50 

Control 

30% FA 

40% FA 

30%FA+10%SF 

40%FA+10%SF 


0 50 100 

Age, days 

150 200 

0.4 

Fig. 4: Oxygen permeability of concrete 

0.3 

0.2 

0.1 

0 

10 

20 

30 

PFA Content, % 

40 

Fig. 5: 3D Isoresponse for 28-day oxygen permeability of concrete 

15 

10 

5 

SF Content, % 

0 

43


The lower permeability values are as a result of micro-filling and pore 

refinement in these mixes as shown in Fig. 6. From this figure it can be seen that 

the incorporation of SF demonstrated significant improvement on pore 

refinement as compared to the control mix. Pore size distribution was carried out 

using mercury intrusion porosimetry. A detailed discussion on pore structure is 

published elsewhere [8]. 

Cumulative Pore Volume, mL/g 

0.10 

0.08 

0.06 

0.04 

0.02 

0.00 

Control 

40% FA+10% SF 

20% FA+10% SF 

0.001 0.01 0.1 1 10 100 1000 

Pore Diameter, µm 

Fig. 6: Pore size distribution of mortar at 28 days, w/b ratio 0.27 

ARTIFICIAL NEURAL NETWORK SOLUTION 

As mentioned above, this investigation RBF network was employed. The 

network developed in this investigation has eight units in the input layer and two 

units in the output layer. The experimentally obtained data have been divided into 

two sets, one for the network learning called learning set, and the other for testing 

the network called testing set. Each set is composed of dozens of pairs of input 

vectors and output vectors (vectors in the input layer called input vectors, and in the 

output layer called output vectors). The experimental data used for ANN analysis is 

used from this investigation and from other reference [9]. An input vector consists of 

8 components which influence the output vectors are compressive strength and 

permeability (Fig. 7). 





The predicted values obtained using ANN for the compressive strength and 

permeability of concrete have been plotted against their respective experimentally 

obtained values as shown in Figs. 8 and 9, respectively. It can be seen from these 

figures that there is a good correlation between experimental values and those 

predicted using ANN. Therefore, it is possible to predict the compressive strength 

and permeability of concrete using artificial neural networks. It is interesting to note 

that the compressive strength represents a wide range of values from 30 to 115 MPa. 

1 ( ) t x 

2 ( ) t x 

3 ( ) t x 

( ) x 

4 t 

x 5 ( t) 

x 

( 

6 t 

) 

( ) x 

7 t 

x8 

( t) 

x 1 = Cement (kg/m 3 ) 

x 2 = FA (kg/m 3 ); 

x 3 = SF (kg/m 3 ) 

x 4 = water (kg/m 3 ) 

x 5 = superplasticizer (kg/m 3 ) 

x 6 = fine agg. (kg/m 3 ) 

x 7 = coarse agg. (kg/m 3 ) 

x = age of testing (days) 

8 

Fig. 7: Schematic diagram of ANN solution 

Output 

45


Predicted values, MPa 

120 

100 

80 

60 

40 



R 2 = 0.97 

40 60 80 100 120 

Experimental values, MPa 

Fig. 8: Experimental values vs predicted values of compressive strength 

Predicted Permeability Values 

2.8 

2.4 

2.0 

1.6 

1.2 

0.8 

0.4 

0.0 

R 2 = 0.98 

0.0 0.4 0.8 1.2 1.6 2.0 2.4 2.8 

Experimental Permeability Values 

Fig. 9: Experimental values vs predicted values of permeability (×10 -16 , m 2 )

CONCLUSIONS 

Based on the results of this investigation the conclusions are as follows: 



• The incorporation of SF content increased the early-age strength for all 

mixes, compensating for the early-age strength loss as a result of FA 

inclusion. It is worth noting that all the mixes were adjusted to equal 

workability by varying the amount of superplasticizer in each mix. 

• Concrete mixes containing 30% FA and above with or without SF were not 

able to achieve the strength of control. However, these systems are viable 

given the level of performance achieved when economical and environmental 

benefits are concerned. 

• The incorporation of FA resulted in a marginal reduction in the oxygen 

permeability values, especially at later ages. The inclusion of SF up to 10% 

significantly reduced both oxygen permeability of concrete for all FA 

replacement levels; above 12% incorporation the reductions were marginal. 

• Based on the experimentally obtained results, ANN has been used to 

establish its applicability for the prediction of concrete strength and 

permeability of concrete. It was demonstrated that strength and permeability 

of concrete can be predicted using ANN. 

REFERENCES 

1. ACI Committee, 1993, “An Essential Program for America and its 

Infrastructure”, Technical Report No. 93-5011, Planning Committee for the 

National-Coordinated Program on High Performance Concrete and Steel, High 

Performance Construction Materials and Systems, USA. 

2. Mehta, R.K. and O.E., Gjorv, 1982, “Properties of Portland Cement Concrete 

Containing Fly Ash and Condensed Silica Fume”, Report STF65 A82030, 

Norwegian Institute of Technology, Trondheim, Norway. 

3. Aitcin, P.C. and A. Neville, 1993, “High-Performance Concrete Demystified”, 

Concrete International, Vol. 15, pp. 21-26. 

4. Gjorv, O.E., 1994, “High Strength Concrete”, Advances in Concrete 

Technology, (Malhotra, V. M. Ed.), CANMET, Energy, Mines and Resources, 

Canada, pp. 19-82. 

47


5. Leontaritis, I.J. and S.A. Billings, 1985, “Input-output parametric models for 

nonlinear systems part-1: deterministic nonlinear systems”, International 

Journal of Control, Vol. 41, pp. 303-328. 

6. Chen, S., C.F.N., Cowan and P.M. Grant, 1991, “Orthogonal Least Squares 

Learning Algorithm for Radial Basis Function Networks”, IEEE Trans. on 

Neural Networks, Vol. 2(2), pp. 302-309. 

7. Khan, M.I., 2003, “Permeation of High Performance Concrete” Journal of 

Materials in Civil Engineering, ASCE, Vol. 15, pp. 84-92. 

8. Khan, M.I., Lynsdale, C.J. and Choo, B.S., 2001, “Pore Structure of High 

Strength Cement Mortar containing Fly Ash and Microsilica.” Proc., Int. Symp. 

Sustainable Development and Concrete Technology, ACI-CANMET, 

supplementary volume, San Francisco, USA. 

9. Khan, M.I., C.J. Lynsdale, 2002, “Strength, Permeability and Carbonation of 

High-performance Concrete”, Cement and Concrete Research, Vol. 32, pp. 

125-133. 



ABSTRACT 

A NOVEL METHOD FOR MEASURING POROSITY 

OF HIGH STRENGTH CONCRETE 


Department of Civil Engineering, College of Engineering 

King Saud University, Kingdom of Saudi Arabia 

Email: miqbal@ksu.edu.sa 

This paper describes design and development of a novel apparatus for the 

measurement of porosity for normal and high strength mortar and concrete using 

pressure saturation method. A detailed description of design, test procedure and 

evaluation of results of the pressure saturation apparatus is presented. Porosity 

values obtained using the apparatus are highly repeatable and reproducible. The 

apparatus is sensitive to the age, influence of water-binder ratio and replacements of 

cement by cementitious materials. The apparatus yielded full saturation for high 

strength and dense mortar and concrete. The porosity values for normal strength 

concrete using the pressure saturation apparatus are in good agreement with the 

results obtained using vacuum saturation method proposed by RILEM. 

KEYWORDS 

High strength concrete, porosity, pressure saturation, vacuum saturation. 

INTRODUCTION 

Porosity is one of the major parameters which influence the strength and 

durability of concrete. The porosity of a porous material, such as cement paste, 

mortar, concrete and other porous material can be determined by measuring any of 

two quantities; bulk volume, pore volume or solid volume. The porosity is the 

fraction of the bulk volume of the material occupied by voids. There are numerous 

techniques being employed for the measurement of porosity. Most commonly used 

techniques for mortar, concrete and other porous material are helium pycnometry, 

mercury intrusion porosimetry and saturation method. 

Among all methods, saturation method is probably the best method as it 

facilitates the real situation encountered by concrete structures. In order to study the 

durability of concrete it is essential to know the maximum porosity which can be 


A NOVEL METHOD FOR MEASURING POROSITY OF HIGH STRENGTH CONCRETE 

penetrated by water. The total porosity or total water absorption for concrete can be 

determined using various methods. The total pore volume, porosity (% by volume) 

and apparent volume can be estimated using methanol saturation method [1]. In this 

method the samples are d-dried (i.e. evacuated over a dry ice trap) with pure 

methanol and weighed in air and then again in methanol. However, it has been 

reported that methanol reacts chemically with cement [2]. 

The simplest way to measure the porosity of porous materials is by water 

absorption. The water absorption can be determined by immersing the sample in 

water for prolonged periods or by boiling the sample in water for several hours. 

However, cement paste, mortar and concrete are more complicated, due to the 

presence of large number of very fine pores, as compared to rocks. In addition, some 

of these pores have bottlenecks which makes it worst. It is, therefore, doubtful that 

these methods achieve total or full saturation. Hence in order to fill all accessible 

pores in the sample with water, it is necessary to empty these pores first from air and 

water before allowing water to penetrate them. RILEM [3] recommended a test 

method based on evacuation of air from oven dried samples then allowing the water 

to fill in the pores under vacuum to reach full saturation. However, RILEM [3] 

vacuum saturation method did not yield full saturation for high strength and dense 

mortar and concrete. Therefore, in order to achieve full saturation for these high 

strength and dense mortar and concrete specimen, a pressure apparatus capable of 20 

bars pressure was designed and developed. This paper describes the detailed design, 

test procedure and evaluation of results of a novel pressure apparatus for measuring 

porosity for normal and high strength mortar and concrete. 


Porosity measurements were carried out for mortar and concrete using 50 mm 

cubes at 28, 90 and 180 days. All the specimens were cast and compacted in 

accordance with standard specifications. After casting, the samples were covered 

under damp hessain and polyethylene sheets for 24 hours. The samples were 

demoulded the following day and then kept in a mist room at 20±2 o C, 95±5% RH 

for curing prior to testing. The samples were taken out of the curing environment at 

the required testing age. The samples were then dried in an oven at the temperature 

of 105±5 o C for approximately 24 hours, until constant weight was reached. Prior to 

testing of porosity, the samples were kept in a dessicator for cooling for another 24 

hours. Triplicate samples were tested for each age and the mean of three is reported 

as result. Three w/b ratios 0.27, 0.40 and 0.50 were used. 

In order to investigate the influence of cementitious materials, blended cement 

systems containing pulverised fuel ash (PFA) and silica fume (SF) were used as 

partial cement replacements on a weight to weight basis. PFA was used up to 40% 

and to these blends 10% SF was incorporated. 



Porosity Measurement 



Vacuum saturation method described by RILEM [3] was used for the 

measurement of porosity of all mortar and concrete specimens. The procedure for 

the measurement is as follows: 

The samples were oven dried for 24 hours at 105 o C and then cooled in a 

desiccator for the next 24 hours and weighed. The samples were then kept in a 

desiccator under 1 bar of vacuum, for 24 hours. The desiccator was then filled with 

de-aired and de-ionised water so that samples are fully submerged in water. Then the 

samples were kept under vacuum for 24 hours and allowed to equilibrate for the next 

24 hours. The samples were then weighed in air and water. In order to ensure the full 

saturation, the samples were broken in to two halves. It was found that none of the 

mortar or concrete specimens prepared with w/b ratio of 0.27 and some concrete 

samples prepared with w/b ratio of 0.40 achieved full saturation or total absorption 

using the vacuum saturation method. In order to estimate the porosity the specimen 

must reach total absorption. 

It was felt that if pressure saturation is applied instead of vacuum saturation 

the dense mortar and concrete samples could be fully saturated. Therefore, an 

attempt was made to design and develop an experimental set-up capable of 

measuring porosity for high strength and dense mortar and concrete using rapid 

pressure saturation. 

DEVELOPMENT OF PRESSURE SATURATION APPARATUS 

The aim of the development of pressure saturation apparatus was to achieve 

total absorption or full saturation of dense mortar and concrete to enable the 

estimation of porosity. The design of the pressure saturation apparatus consisted of a 

pressure vessel capable of withstanding 20 bar of pressure, a vacuum pump, water 

tank, pressure supply source (nitrogen gas cylinder), pressure gauge and vacuum 

gauge. In order to apply the pressure without causing air entrainment, the pressure 

must not be applied directly to the water. To achieve this, an additional pressure cell 

with rubber membrane was used preceded by the original pressure vessel. In this 

case pressure is applied to the rubber membrane which inflates and pushes the water 

at the same pressure rate. The schematic diagram of pressure saturation apparatus is 

shown as in Fig. 1 and general view of the apparatus is shown in Fig. 2. It consists 

of the following components:- 

Pressure Cell 

The cell consists of perspex acrylic cylinder internal diameter 110 mm, height 

220 mm and 25 mm thick. The cylinder is fitted in with three stainless steel platens; 

two at the top and one at the bottom. Two platens are provided at the top to facilitate 

51


the rubber membrane fitting and air and water vents. The platens are connected 

together with 4 stainless steel rods screwed with knurled nuts at the both ends. The 

cell including the rubber membrane can withstand 20 bar of pressure. The top platen 

is provided with an opening which is connected to the pressure supply (nitrogen gas) 

and air vent. Another top platen provides the facility of water vent. Bottom platen is 

provided with two openings; one is connected to water supply from water tank and 

another is connected to the pressure output for the pressure vessel. 

trap 

V4 

vacuum 

gauge 

vacuum 

pump 

water pressure 

gauge 

V2 

Sv2 

vacuumpressure 

vessel 

water 

tank 

pressure 

cell 



V3 

O 

V6 

V1 

Wv 

V7 

air pressure 

gauge 

Av = air vent ; M = rubber membrane; O = rubber O-ring; 

Sv = pressure safety valve; V = valve; Wv = water vent 

Fig. 1: Schematic diagram of pressure saturation apparatus 

V5 

Not to the scale 

Sv1 

M 

V8 

pressure 

regulator 

Av 

pressure 

supply

trap 

vacuum 

pump 

vacuu 

m 

water 

pressure 

vacuumpressure 

vessel 

water supply 

air pressure 

gauge 

pressure 

Fig. 2: General view of pressure saturation apparatus 



pressure 

regulator 

pressure 

supply 

Vacuum-Pressure Vessel 

The vessel consists of acrylic perspex cylinder internal diameter 180 mm, height 

330 mm and 25 mm thick. The total volume of the vessel is approximately 8.4 litres. The 

vessel has been designed to withstand 20 bar of pressure and 1 bar of vacuum. The 

perspex cylinder is fitted with top and bottom stainless steel platens, 300 mm diameter 

and 40 mm thick. In order to ensure the air-tight system, rubber O-rings were provided 

between both platens and the cylinder. The top and the bottom platens are connected 

together by 6 stainless steel rods screwed with knurled nuts at both ends. The top platen 

is provided with water pressure gauge, pressure safety valve and two openings; one 

opening is connected to a water tank and another to a vacuum pump via vacuum trap 

assembly. The bottom platen has an opening in the centre, which facilitates the pressure 

supply and draining of the vessel. The bottom platen is slanted towards the opening to 

facilitate the proper cleaning of the vessel. 

Pressure Supply 

Nitrogen gas is used as pressure provided in a cylinder. The cylinder is 

provided with a regulator connected to the pressure cell through a pressure safety 

valve and a pressure gauge. 

53


Vacuum Pump 

A rotary vacuum pump is used capable of evacuation of 1 bar vacuum. In 

order to stop any water or moisture into the pump a vacuum trap assembly is provide 

between the vessel and the pump. 

Water Tank 

A plastic water tank of 10 litres capacity. 

Other accessories 

Brass valves, pressure safety valves, air pressure gauge, water pressure gauge, 

vacuum gauge, connectors and 8 mm nylon tubing. 

Test Procedure 

The test procedure and the operation of the apparatus is as follows: 

1. Place oven dried samples in the vacuum-pressure vessel after taking their 

weights. 

2. Keep the samples under vacuum for 3 to 4 hours until rising of air bubbles will 

stop. 

3. Fill both vacuum-pressure vessel and pressure cell with water and are ready for 

the application of pressure. 

4. Set pressure at 20 bars and open apply pressure. The membrane M will start 

inflating. Allows the pressure to enter into the pressure vessel. 

5. Ensure that air pressure gauge and water pressure gauge reads as 20 bars and 

leave it on for 24 hours. At this stage both vacuum-pressure vessel as well as 

pressure cell are at the applied pressure of 20 bars. 

6. After 24 hours, open vents gently and release the applied pressure and the 

whole system goes back to atmospheric pressure. 

7. Leave the samples in the vessel for the next 24 hours to reach the equilibrium 

state. 

8. Take weight measurements of samples in air and in water. 

The amount of water penetrated into the sample is a measure of the porosity 

and is calculated as follows: 

B − A 

P = × 100 

(1) 

B − C 

where; P is porosity (%); A is oven-dry weight; B is saturated surface dry weight; C 

is saturated submerged weight. 



REPRODUCIBILITY AND REPEATABILITY 



The reproducibility and repeatability of the porosity results obtained using the 

pressure apparatus were determined for various concrete mixes. The reproducibility 

was determined by measuring the porosity of a number of specimens taken from the 

same concrete mix. Various mixes were selected and from each mix five specimen 

were tested for the porosity and the results obtained are presented in Table 1. The 

results demonstrated very low variability as can be seen from coefficient of 

variation, which was between 0.27 and 0.56. 

The repeatability was evaluated by repeating the test three times for the same 

specimen. The test for each specimen was repeated three times and the results 

obtained are presented in Table 2. The coefficient of variation of reproducibility was 

also low between 0.29% and 0.74%. However, the coefficient of repeatability is 

slightly higher than that of coefficient of reproducibility. The reason for higher 

coefficient of repeatability is probably associated with variation of testing age. Since 

same specimen was tested thrice and each test takes three days to finish, therefore 

the time difference between first test and the last test is about 9 days. During this 

period some hydration of the specimens is expected. Although the specimen used 

were oven dried, the test specimens were in water after the second day which might 

have restarted the hydration process. 

Table 1: Statistical analysis of reproducibility of porosity results 

Sample No. 

Porosity, % 

Mix 1 Mix 2 Mix 3 Mix 4 Mix 5 Mix 6 

1 12.99 14.00 14.54 14.80 16.19 16.87 

2 13.01 13.90 14.50 14.85 16.29 17.04 

3 12.95 13.97 14.62 14.80 16.22 16.93 

4 12.97 14.03 14.58 14.92 16.26 16.99 

5 13.04 13.93 14.60 14.90 16.20 16.95 

Mean 12.99 13.97 14.57 14.85 16.23 16.96 

Standard deviation 0.035 0.052 0.048 0.055 0.042 0.064 

Coefficient of variation 0.269 0.374 0.331 0.374 0.259 0.377 

55


Table 2: Statistical analysis of repeatability of porosity results 

Test run 

1 st test 

2 nd test 

3 rd test 

Porosity, % 

Mix 1 Mix 2 Mix 3 Mix 4 Mix 5 Mix 6 

12.95 13.97 14.62 14.80 16.22 16.93 

12.97 14.03 14.67 14.92 16.26 16.99 

13.02 14.07 14.62 14.90 16.30 17.07 

13.05 14.05 14.75 14.95 16.34 17.05 

13.10 14.12 14.70 14.96 16.32 17.12 

13.09 14.14 14.75 15.00 16.34 17.15 

Mean 13.03 14.06 14.69 14.92 16.30 17.05 

Standard deviation 0.062 0.062 0.059 0.069 0.048 0.082 

Coefficient of variation 0.473 0.440 0.401 0.461 0.295 0.478 

EVALUATION OF RESULTS 

The porosity values measured using the pressure saturation apparatus for 

concrete made with w/b ratio of 0.27 at various ages is shown in Fig. 3. This 

figure demonstrates the change in porosity of concrete at 28, 90 and 180 days. 

The results indicate that the incorporation of PFA had a very small influence on 

porosity at all ages investigated. At 28 days, it is evident that the incorporation 

of PFA had a small effect on porosity. At 90 and 180 days a slight reduction in 

porosity, up to 20% PFA replacement. The inclusion of SF resulted in a 

reduction in porosity for all mixes with and without PFA at all ages. However, 

the reduction in the porosity of concrete was greater when 10% SF was 

incorporated and this reduction was exhibited at all ages. 



Porosity, % 

15 

14 

13 

12 

11 

10 

0 50 100 150 200 

Age, days 



OPC 

20%PFA 

30%PFA 

40%PFA 

10%SF 

Fig. 3: Porosity of concrete using pressure saturation method 

20%PFA+10%SF 

30%PFA+10%SF 

40%PFA+10%SF 

The influence of w/b ratio on the porosity of concrete at various ages 

investigated is demonstrated in Fig. 4. It can be seen that the porosity increased with 

an increase in w/b ratio for all ages, as to be expected. From Fig. 4, the influence of 

PFA replacement was not noticed at 28 days whilst at 90 and 180 days the 

incorporation of PFA had little effect on porosity but this effect was limited to low 

w/b ratios. The incorporation of 10% SF decreased the overall range of porosity 

values for all ages investigated. The slight reductions in porosity as a result of PFA 

incorporation was seen at 28 days. At 90 and 180 days, the reduction in porosity by 

incorporation of PFA is significant for lower w/b ratios. 

57


Porosity, % 

Porosity, % 

Porosity, % 

20 

18 

16 

14 

12 

10 

20 

18 

16 

14 

12 

10 

20 

18 

16 

14 

12 

10 

Control 

Control 

Control 

w/b 0.27 w/b 0.4 w/b 0.5 

20%PFA 

20%PFA 

30%PFA 

30%PFA 

40%PFA 

40%PFA 

(a) Age: 28 days 

10%SF 

20%PFA+10%SF 

(b) Age: 90 days 

w/b 0.27 w/b 0.4 w/b 0.5 

20%PFA 

30%PFA 

40%PFA 

10%SF 

20%PFA+10%SF 

(c) Age: 180 days 

w/b 0.27 w/b 0.4 w/b 0.5 

10%SF 

20%PFA+10%SF 

30%PFA+10%SF 

30%PFA+10%SF 

30%PFA+10%SF 



40%PFA+10%SF 

40%PFA+10%SF 

40%PFA+10%SF 

Fig. 4: Porosity of concrete at various w/b ratios using pressure saturation



The results indicated that the incorporation of PFA in mortar and 

concrete demonstrated very little or no reduction in porosity. The 

incorporation of SF resulted in reduction in porosity in all mixes of mortar and 

concrete with or without PFA. 

It has been reported that the inclusion 10% SF as cement replacement 

reduces total porosity, using mercury porosimetry, prepared with w/b ratios 

between 0.20 and 0.30 [4]. Since porosity test is not a standard test, hence 

researchers use different methods for porosity measurement. The test method 

used has a great influence on test results therefore, it is difficult to compare 

the results with other researchers. However, there is general consensus that the 

incorporation of SF as cement replacement reduces the total porosity. But 

incorporation of SF is more efficient in refining the pore structure than in 

reducing the total porosity. 

It is worth noting here that the pressure saturation porosity values using 

pressure saturation apparatus clearly demonstrated sensitivity to the age, 

influence of water-binder ratio and replacements of cement by cementitious 

materials such as PFA and SF. The porosity results using pressure apparatus 

are reported in other investigation [5] are in well agreement with the results of 

the present investigation. 

RELATIONSHIP BETWEEN PRESSURE SATURATION AND VACUUM 

SATURATION RESULTS 

The experimental results of porosity of mortar and concrete obtained using 

pressure saturation method (using pressure saturation apparatus) have been plotted 

against their respective results obtained using vacuum saturation method (using 

RILEM recommendations). The comparison of porosity results of mortar [6], 

prepared with w/b ratio of 0.27, and concrete prepared with w/b ratios of 0.27, 0.40 

and 0.50 are demonstrated in Fig. 5. 

59


Porosity (pressure saturation), % 

18 

16 

14 

12 

10 

8 

6 

Fig. 5: Relationship between experimentally obtained results using vacuum 

saturation and pressure saturation 

It can been seen from this figure that as the vacuum saturation porosity values 

increases the difference between vacuum saturation porosity values and pressure 

saturation porosity values narrows down. For example, the porosity values of mortar 

and concrete prepared with w/b ratio of 0.27, using vacuum saturation method were 

between 7.0 and 13.0% as compared to their pressure saturation porosity values 

between 11.3 and 15.7%. Whilst concrete, prepared with w/b ratio of 0.50, exhibited 

vacuum saturation porosity values between 15.4 and 17% as compared to their 

pressure saturation porosity values between 15.5 and 17.2%. This difference can be 

appreciated while comparing the regression lines of these results as shown in Fig. 5. 

The regression line for normal strength concrete prepared with w/b ratio of 0.50 falls 

very close to the line of equality. The results demonstrated high correlation between 

vacuum saturation porosity values and pressure saturation porosity values. It is 

worth noting here that all concrete specimens prepared with w/b ratio of 0.50 and all 

paste specimens (w/b ratio 0.30) were fully saturated using vacuum saturation. 

CONCLUSIONS 

Line of equality 

6 8 10 12 14 16 18 

Porosity (vacuum saturation), % 

Pressure saturation apparatus yielded full saturation for high strength and 

dense mortar and concrete. The apparatus is sensitive to the age, influence of waterbinder 

ratio and replacements of cement by cementitious materials. Porosity values 

obtained using pressure saturation apparatus are highly repeatable and reproducible. 



w/b 0.27 (conc) 

R^2 = 0.86 

w/b 0.40 (conc) 

R^2 = 0.91 

w/b 0.50 (conc) 

R^2 = 0.97 

w/b 0.27 (mor) 

R^2 = 0.92



The porosity values for normal strength concrete using the pressure saturation 

apparatus are in good agreement with the results obtained using vacuum saturation 

method proposed by RILEM. 

REFERENCES 

1. Day, R.L., 1981, "Reactions between Methanol and Portland Cement Paste",. 

Cement and Concrete Research, Vol. 11, pp. 341-349. 

2. Feldman, R.F., 1972, "Density and Porosity studies of Hydrated Portland 

Cement", Cement Technology, Vol. 3, pp. 5-14. 

3. RILEM Recommendations, 1984, "Absorption of Water by Immersion under 

Vacuum", Materials and Structures, RILEM CPC 11.3, Vol. 101, pp. 393-394. 

4. Zhang, M.H. and O.E., Gjorv,1991, "Effect of Silica Fume on Pore Structure 

and Chloride Diffusivity of Low Porosity Cement Pastes", Cement and 

Concrete Research; Vol. 21, pp. 1006-1014. 



6. Khan, M.I., C.J. Lynsdale, 2003, "Isoresponses for Strength, Permeability and 

Porosity of High Strength Blended Mortar", Building and Environment, Vol. 38, 

pp. 1051-1056. 

61




FIELD STUDY FOR THE EVALUATION OF STEEL BRIDGES 

IN RIYADH CITY 

Khalid A. AlSaif 1 , Alaa Malaika 2 , Mohamed M. ElMadany 1 

1: Mechanical Engineering Department, 2: Civil Engineering Department 

King Saud University, P.O. Box 800, Riyadh 11421, mmadany@ksu.edu.sa 

ABSTRACT 

In this study, a comprehensive database on the steel bridges in the city of 

Riyadh has been constructed using Access 2002 software package developed by 

Microsoft Corporation. The information contained in the bridge database can be 

searched and indexed in a variety of ways to produce the desired reports. A selected 

bridge was instrumented with various instrumentations and an intensive 

measurements program was conducted. Several accelerometers were placed along 

the girder at points with expected maximum response. All critical members of the 

bridge were considered (girders, cap beams and column piers) during the test. 

Dynamic bridge testing was used as a means of evaluating the selected bridge. The 

tests were carried out to obtain the vibration frequencies, and levels of dynamic 

responses, and were based on normal traffic-induced vibrations. The collected field 

testing data were analyzed both in time and frequency domains. Statistical analysis 

was utilized to estimate some statistical measures of the vibration levels at the 

critical points. 

In addition, guidelines for field testing of bridges are presented. This is to outline the 

basic principles and procedures for field testing of highway bridges. 

KEYWORDS 

Bridge, Dynamic bridge testing, Guidelines for field testing 

INTRODUCTION 

In the last four decades, there have been significant developments in the 

transportation systems in the kingdom of Saudi Arabia. Most of the roadways and 

bridges of Saudi Arabia’s transportation system were constructed during this 

construction boom. 

A bridge is considered to be a key element in any transportation system. Not 

only it controls the capacity of the system, but is expensive to construct and costs 

much more than that of approach roads. Hence, bridges are the highest cost per 

kilometer in any transportation system. Also, if a bridge fails or goes out of service 

the whole system could fail. 


FIELD STUDY FOR THE EVALUATION OF STEEL BRIDGES IN RIYADH CITY 

The complexities and cost associated with preserving the nation’s bridge 

infrastructure demand innovative approaches to collecting inventory and inspection 

data, analysis of gathered data, and prediction of current and future bridge 

preservation actions. These needs, coupled with the availability of high quality and 

precession instrumentation, modern simulation techniques, and high-speed 

computers may lead to the development of means for evaluating and ensuring safety 

and reducing the risks associated with aging highway bridge structures. 

Bridges are critical components in the flow of traffic and the costs of 

replacement, construction of new ones and/or closure may often be prohibitively 

high. Therefore, the increase of the service life of the bridges is an important 

objective. Proper identification of a deficient bridge and its subsequent rehabilitation 

or repair require a sufficient understanding of existing bridge conditions and are 

imperative from both a safety and economic standpoint. 

A major contributing factor to the deterioration of the bridges is overloads. 

Heavy trucks or some other reason could cause these overloads. A very useful tool, 

from a bridge management perspective and as a tool for helping to make the most 

efficient use of bridge inspection resources, would be testing bridges which would 

measure the loads experienced by a bridge. 

An alternative to the use of field testing is to carry out complex dynamic 

analysis, using finite element methods that account for the bridge-vehicle interaction 

model to determine the bridge response under traffic loading [1-3]. Such approach 

relies heavily on the dynamic characteristics of the vehicle, the surface roughness, 

and soil-structure interaction. Besides, to obtain a meaningful statistical evaluation, 

the dynamic analysis must be carried out for a large number of vehicles, each having 

different dynamic properties, with different roadway surface irregularities leading to 

different bridge vehicle interaction characteristics. This can be prohibitive at the 

present time. 

On the other hand, field tests under normal traffic provide a fast and costeffective 

way to evaluate the existing bridges. Interest in this method is growing, as 

affordable data acquisition systems and real-time signal processing hardware are 

now available. These tests are performed for a variety of reasons including studies 

of the aerodynamic response of bridges, correlation of numerical models with 

measured data, seismic assessment of the bridges, bridge condition monitoring, and 

studies related to the development of dynamic impact factors for design of the 

bridges. These tests can also be used to evaluate structural degradation by 

monitoring changes in dynamic properties during repeated tests. In the course of 

these studies many different types of excitation methods have been applied to bridge 

structures. 

The experimental procedures which have been used to yield bridge dynamic 

properties and bridge evaluation include (i) impact tests; (ii) use of eccentric or 

reactional mass exciters; (iii) use of test vehicles (controlled traffic); (iv) use of 

normal traffic; (v) wind-induced vibrations; (vi) sudden release of static load or 

imposed displacement; and (vii) vibrations generated by braking vehicles. 




Khalid A. AlSaif , et al 

Paultra et al. [2] presented a literature review involving dynamic testing of 

highway bridges and analytical studies dealing with the evaluation of existing 

bridges. Signal processing techniques, experimental evaluation and dynamic 

properties of bridges were also presented. In a series of publications, Paultra et al. 

[3-5] used dynamic bridge testing techniques to obtain an evaluation of the dynamic 

amplification factor for existing bridges. The vibration frequencies and mode 

shapes were calculated from a frequency analysis of the measured data, and used to 

calibrate finite element models for each structure. In Reference [6] the history and 

present use of bridge testing were reviewed. Steps for testing transducers, signal 

conditioning and amplification, and data recording and processing were described. 

Tests, which evaluate cracks and hardness, were examined. Bakht and Jaeger [7] 

emphasized the importance of carrying bridge field-testing due to a number of 

significant surprises they encountered during testing of highway bridges in Ontario. 

They have shown that in some cases the bridge behavior may be so deceptive that it 

can be understood only through a field test. For such cases, no amount of 

refinement in the theoretical analysis can yield the true load-carrying capacity of the 

bridge. 

A large number of bridges have been field tested in Switzerland [8,9], and it 

has been common practice there for many years to test new bridges prior to putting 

them in service, and to perform field tests on bridges after several years in service to 

further check behavior. Moses et al. [10] explored bridge testing as a specific part 

of the bridge evaluation, especially for cases in which a safe capacity evaluation or 

bridge rating ought to be carried out. Their work examined testing from the point of 

view of assessing in the safety checking and in particular how the test results might 

influence the bridge evaluation and rating. 

Some researchers [7, 11] have tested older bridges and controlled extreme 

loads to verify capacity. Many other researchers [e.g., 12, 13] have used bridge tests 

at loads much below the service loads and which were intended to diagnose bridge 

behavior and especially the distribution of the forces within the structure. On the 

other hand, some bridge tests have been performed under normal traffic loadings. 

Such tests have usually been oriented towards measuring stresses either to infer the 

traffic load characteristics, the force distributions and dynamic response or to obtain 

spectra for fatigue estimates [14, 15]. 

Boothby and Craig [16] have conducted a field test of a complicated statically 

indeterminate floor system on a historic steel truss bridge to obtain a load rating. 

They have shown that the experimental determination of wheel load distributions 

has provided an inexpensive means of intelligently allocating limited country 

resources for the maintenance and rehabilitation of an inventory of aging bridges. 

Application of field data to condition assessment and prediction of service life of 

highway bridges has been shown by Mohammadi et al. [17]. The field stress range 

data compiled for each bridge was used along with a probabilistic method to 

estimate fatigue life. The results were then used to investigate the significance of 

truck weight increase and traffic growth on fatigue life. 

65


BRIDGE DATABASE 

The first step in the construction of the Bridge Management Program is the 

development of comprehensive databases for critical steel bridge structures. The use 

of the bridge database to be developed could include the following: disaster relief 

planning, development of geographic information systems, extrapolation of field and 

computer analysis results, bridge design research, estimation of bridge replacement 

costs, maintaining an updated record on the condition and status of the steel bridges, 

management of bridge inspection and maintenance, etc. 

The elements of a typical bridge structure can be classified into two primary 

components, the substructure and the superstructure. The substructure refers to the 

elements of the bridge that transfer the loads from the bridge deck to the ground, 

such as abutments and piers. The superstructure refers to the elements of the bridge 

above the substructure, including the deck, floor system (beams or stringers), 

supporting members (beams, trusses, frames, girders, arches, or cables), and bracing. 

Other bridge elements, which are subject to corrosion, include guard railing and 

culverts. Bridge construction materials that are subject to corrosion include 

conventional reinforced concrete, prestressed concrete, and steel. Of these three 

bridge types, steel has the highest percentage of structurally deficient structures, 

followed by conventionally reinforced concrete and prestressed concrete. 

In this study, a comprehensive database for the ten steel bridges in Riyadh 

city has been developed. The data gathered and recorded for each bridge included: 

bridge location code and name, number of spans, typical span length, total length, 

number of carriageways, number of lanes in each carriageway, and vertical 

clearance; super structure including width of deck, width of bridge, number of 

girders across deck, lengths of steel expansion joints, rubber expansion joints, 

guardrails, handrails, and side sheets; substructure including type of girder support, 

numbers of bearings, piers, and cap beams. The database has been constructed, for 

our preliminary investigation, using one record per bridge. The information 

contained in the bridge database can be searched and indexed in a variety of ways to 

produce the desired reports. 

Field Inventory and Bridge Access Database 

The sets of information for the different steel bridges in the city of Riyadh 

were gathered during several field trips. The data gathered were partly taken from 

the municipality of Riyadh city. For each field trip, several field inventory sheets 

were developed. These inventory sheets were used in various combinations to record 

all the necessary information for each bridge. 

A database for the steel bridges in the city of Riyadh has been developed. The 

data recorded for each bridge will assist future database users in estimating bridge 

rehabilitation (or replacement) costs, and in other research and engineering 

activities. It could also be a part of National Bridge Inventory and Inspection 

Program. The Riyadh steel bridges database was constructed using Access 2002 

software package developed by Microsoft Corporation. 





The structure of the bridge database was developed in this initial stage so 

that the information for each bridge could be stored in one record. Each bridge 

record contains data such as bridge location code and name, number of spans, 

typical span length as discussed earlier. A total of 28 fields are used in this stage 

for the general inventory information about any bridge. Figure 1 shows a typical 

form used for data-entry in the bridge database showing all the fields used for 

storing the bridge’s inventory. 

INSTRUMENTATION AND MEASUREMENT ANALYSIS 

Bridge Selection for Field Testing 

There are 10 steel bridges in Riyadh area. However, bridge S01, located in Shemysi 

area; Al Imam Abdulaziz Ibn Mohd.Ibn Saud st.Amro Ibn Al A’as St.(TV 

Street), is selected for field testing. This selection is based on the following reasons; 

1) Project time constraint (project schedule) and 2) The bridge is located in a heavily 

populated area and the traffic for this bridge is high. 

This steel bridge S-1 (S01) is a two-lane dual carriageway bridge having 10 spans in 

all, with a typical span length of 24 m. The total length of bridge is 238.60 m. Each 

of the two main carriageways is formed by four main girders. The two carriageways 

are separated by means of a central row of posts on which the guard rails are fixed 

on each face. All the eight girders of the superstructure are supported by a single 

capping beam of built up section through elastomeric bearing pads. The steel 

capping beams are supported by two columns on either end. The columns are made 

up of steel built up sections. The vertical columns are connected to the cap beams by 

bolted connection and the bottom connected to the base plate by bolts and nuts. 

Figure 1 Typical Access data-entry for bridge database – location code S01. 

67


Dynamic Testing Under Traffic Loads 

Dynamic tests of the bridge with controlled and normal traffic excitations can 

provide very useful and cost effective way to estimate the dynamic response of the 

bridge to moving loads. In this study a normal traffic excitations is used as a source 

of disturbing the bridge structure. No attempt is made to perform a controlled 

excitation source (i.e. single heavy truck at a predetermined speed and weight). The 

root mean square as well as the maximum amplitude of the ensuing dynamic 

oscillations of the bridge, during the passage of vehicles, is used as measures of the 

response strength. 

The following tests are conducted under normal traffic conditions; a) The 

acceleration response of the bridge at specific critical points, b) The maximum strain 

and stress at the girder flanges, and c) The first few natural frequencies of the bridge 

structure. 

Instrumentation and Experimental Setup 

The experimental setup is shown schematically in Figures 2 and 3, and in a 

physical form in Figure 4. The setup consists of accelerometers and strain gages, 

signal conditioning modules, filters and A/D converter with a laptop. The strain 

gages are connected to a strain meter HBM type. The accelerometers are low 

frequency type and have high sensitivity (1 v/g). The strain gages are dynamic type 

and can withstand the ambient conditions during the measurements. 

The proper locations for the sensors are determined after intensive preliminary 

testing. The signal output from the sensor (accelerometer or strain gage) is filtered 

using anti-aliasing filters (with a corner frequency between 10 Hz to 200Hz) and 

amplified before entering the data acquisition module where the analog signal is 

digitized with sampling frequency at 200 Hz. The digitization module has 12 bit 

accuracy and a maximum throughput of 1 MHz. The signals can be displayed in real 

time and stored simultaneously. The stored data is post-processed to calculate rms 

values, maximum acceleration, and frequency response 

Measurement locations 

After several preliminary measurements at different points on the bridge 

structure, it was decided to use 12 critical points (points in mid span of girders) 

to conduct the dynamic response measurements. Figure 5 shows a top view of 

the portion of the bridge indicating the accelerometers and strain gages 

locations. The accelerometers are attached to points (1s, 2s, 3n, 1L, 3R, 2L, 3L) 

as indicated in the figure and the strain gages are placed at points (0st, 1st, 2st) 

to measure the longitudinal strain. Figure 6 shows the physical locations. The 

accelerometers are screwed in an interface circular plate, and the plate is 

attached to the girder using an adhesive scotch weld type. The strain gages are 

attached to a thin aluminum plate and the plate is placed on the bottom flange of 

the girder with the same adhesive material. 



Vibration 

measurement 

direction 

ACCELEROMETERS 

AND STRAIN GAGES 

Wbk- 

Dynamic signal conditioning 

module 8 channels 

filters and amplifiers 

Main girders 

Supporting piers 

Wave book data acquisition 

8 channels, 500 hZ 

SAMPLING PER 

CHANNEL 


BRIDGE CROSS SECTION 

Figure 2 Initial testing stage- experimental setup 

Figure 3 Schematics of the experimental setup 


Laptop-PC1 

TIME AND 

FREQUENCY DOMAIN 

ANALYSES 

69


Figure 4 Experimental apparatus - physical setup 

Figure 5 Top view of a portion of the bridge – measurement points 



Figure 6 Underneath bridge showing mid girders and cap beam 

EXPERIMENTAL RESULTS 



Acceleration Measurements 

During normal traffic bridge excitations, the data from the accelerometers and 

strain-gages are logged. The duration of each measurement is about 25 s. Several 

time histories for each measurement point is averaged and stored. The sampling 

frequency is fixed at 200 Hz (i.e. the range of natural frequencies of interest is below 

50 Hz). An anti-aliasing filter is used with a corner frequency of 100 Hz. 

71


The time histories are then transformed to the frequency domain using Fast Fourier 

transform. The maximum number of points used to calculate the transform is about 

4096 points. The Haning window type is used to avoid leakage in the frequency 

spectrum. 

Sample time histories at points, 1 L and 2 L are shown in Figures 8`and 9, 

respectively. 

The root mean square and the maximum acceleration response are calculated for 

each wave form and the results show that the rms accelerations vary between 0.028t 

g and 0.09 g, while the maximum accelerations vary between 0.1 g to 0.5 g. It is 

believed that this acceleration magnitude is not severe and does not indicate a 

serious malfunction or defects in the girders and their interfaces with the cap beams. 

acceleration gx10 

acceleration gx10 

6 

4 

2 

0 

-2 

-4 

-6 

0 

0.17 

0.34 

0.51 

0.68 

0.85 

1.02 

1.19 

1.36 

1.53 

1.7 

1.87 

2.04 

2.21 

2.38 

2.55 

2.72 

2.89 

3.06 

3.23 

3.4 

3.57 

3.74 

3.91 

4.08 

4.25 

4.42 

4.59 

4.76 

4.93 

time sec 

Figure 7 Time history of point 1 L (left girder) 

6 

5 

4 

3 

2 

1 

0 

-1 

-2 

0 

0.17 

0.34 

0.51 

0.68 

0.85 

1.02 

1.19 

1.36 

1.53 

1.7 

1.87 

2.04 

2.21 

2.38 

2.55 

2.72 

2.89 

3.06 

3.23 

3.4 

3.57 

3.74 

3.91 

4.08 

4.25 

4.42 

4.59 

4.76 

4.93 

time sec 

Figure 8 Time history of point 2 L (left girder) 

Samples of the frequency spectra at points 2L, 3L, 1S, 2S, 3S are depicted in 

Figures 9 to 11. It should be noted that the large peaks appear in the spectrum at 60 

Hz and its sub-harmonics (20, 30, 40, etc) are due to power line frequency (60 Hz) 

and must be excluded from the analysis. 



level 

0.18 

0.16 

0.14 

0.12 

0.1 

0.08 

0.06 

0.04 

0.02 

0 

point 2 (solid line) & 3 (dotted line) ( f3l) 

0 

4.64 

9.28 

13.9 

18.6 

23.2 

27.8 

32.5 

37.1 

41.7 

46.4 

51 

55.7 

60.3 

64.9 

69.6 

74.2 

78.9 

83.5 

88.1 

92.8 

97.4 

102 

107 


hz 


Figure 9 Frequency spectra - points 2L (solid line) and 3L (dotted line) 

vibration level 

0.009 

0.008 

0.007 

0.006 

0.005 

0.004 

0.003 

0.002 

0.001 

0 

frequency spectrum -point 1( dotted) , 2 (solid line ) south 

0 

2.93 

5.86 

8.79 

11.7 

14.6 

17.6 

20.5 

23.4 

26.4 

29.3 

32.2 

35.2 

38.1 

41 

43.9 

46.9 

49.8 

52.7 

55.7 

58.6 

61.5 

64.5 

67.4 

70.3 

73.2 

76.2 

79.1 

82 

85 

frequency Hz 

Figure 10 Frequency spectra - points 1S (dotted liine) and 2S (solid line) 

Some of the peaks in the spectrums represent local natural frequencies of the 

bridge substructure. For instance, Figure 11 shows a peak at 14.16 Hz which is a 

local frequency. This component is not common among all the measured points. 

These local peaks can suppress the global components. To obtain the global natural 

frequencies of the bridge, the cross correlation function is implemented. A brief 

description of the function is given. 

level (volts) 

0.012 

0.01 

0.008 

0.006 

0.004 

0.002 

0 

0 

5.13 

10.3 

15.4 

20.5 

25.6 

30.8 

35.9 

41 

point 3 s 

46.1 

51.3 

56.4 

61.5 

66.7 

71.8 

76.9 

82 

87.2 

92.3 

97.4 

103 

Figure 11 Frequency spectrum –point 3s 

Hz 

108 

113 

118 

73


Cross correlation 

While the auto correlation function of a wave form is concerned with 

averaging the product of the wave and a time-shifted version of itself, the cross 

correlation is concerned with averaging the product of two different waveforms one 

of which is time shifted. The averaging process for the two waves is performed over 

a specified interval called the window, the time T is known as the width of the 

window. 

By performing the cross correlation (converting the integral to summation, 

since the wave forms are in digital format) between each pair of wave forms (fi, gj), 

the major components in the cross spectra between the different channels are 

identified. From the results of the cross correlation together with the frequency 

spectrum plots depicted in Figures 9-11, the most probable global natural 

frequencies (first few modes) can be given as follows 

Natural frequencies: {6 Hz, 7.08 Hz, 15.3 Hz} 

It should be noted that these frequencies include the dynamic interaction 

between the passing vehicles and the bridge. 

Strain measurements 

To measure the maximum normal stress in the girder bottom flange during 

normal traffic conditions, dynamic strain gages are attached at the points that are 

expected to have large deformations (0st, 1st, 2st as indicated in Figure 5). The 

gages have a gage factor of 2 supplied by the manufacturer. The dynamic signal is 

recorded and displayed in the frequency domain as shown in Figures 12 and 13. It 

can be readily seen that the frequency components (7.1 and 15.3 Hz from Figure 

11) picked up by the strain gages are similar to the natural frequencies obtained 

using the accelerometers signals presented in the last section. The 40 Hz component 

shown in Figure 11 is believed to be sub-harmonic of the power line frequency (60 

Hz) indicated previously. 

signal level 

0.1 

0.09 

0.08 

0.07 

0.06 

0.05 

0.04 

0.03 

0.02 

0.01 

0 

frequency spectrum -strain measurement - point 2 E 

0 

2.44 

4.88 

7.32 

9.77 

12.2 

14.6 

17.1 

19.5 

22 

24.4 

26.9 

29.3 

31.7 

34.2 

36.6 

39.1 

41.5 

43.9 

46.4 

48.8 

51.3 

53.7 

56.2 

Figure 12 Strain spectrum at point 2st 



Hz

level 

0.07 

0.06 

0.05 

0.04 

0.03 

0.02 

0.01 

0 

2.44 

0 

7.32 

4.88 

12.2 

9.77 

17.1 

14.6 

strain spectrum-point 0st 

22 

19.5 

26.9 

24.4 

31.7 

29.3 

36.6 

34.2 

41.5 

39.1 

46.4 

43.9 


51.3 

48.8 

Figure 13 Strain spectrum at point 0st 

GUIDELINES FOR FIELD TESTING OF BRIDGES 

hz 

56.2 

53.7 


The intent of these guidelines is to outline the basic principles and procedures 

for field testing of highway bridges. While general in nature, and related primarily to 

bridge structural performance and observed response behavior, the procedures 

described in this report are geared toward the capabilities and functionality provided 

by the available data acquisition system, analysis software package and data 

processing software specifically designed for bridge testing. 

Outline of Procedures 

1. Define goals of field test. 

2. Develop instrumentation and test plan. 

3. Test the bridge under normal traffic operation. 

4. Acquire the test results and the measured responses 

5. Examine and make qualitative assessment of the data. 

6. Assess results and determine what is reliable and reasonable for presentation. 

Goals of Load Test 

The first step in any load test is to first determine the test goals and to outline 

specifically what information is desired. Execution of this step has two main 

prerequisites. First, a thorough knowledge of bridge rating is necessary so that the 

engineer can assess what the critical components are and determine if there are any 

potential gains to be obtained by a more accurate analysis. 

Secondly, goal definition requires field test experience and knowledge of what 

“can” be determined from a field test. In most cases, it is sufficient to simply 

measure strains, compare them with a theoretical value or code provision and 

interpolate or extrapolate results. Another objective is to determine “why” the 

measurements are different than expected and to verify the mechanisms causing the 

discrepancies by reproducing the measured results with a representative analysis. 

The Instrumentation and Test Plan 

Many factors influence the design of an instrumentation plan. Because of the 

wide variety of bridge structure types, construction details, accessibility, bridge 

structural conditions, and so on, it is impossible to define an exact set of rules for 

sensors placement. Bridge structure type and test goals are the main factors in 

58.6 

75


developing the initial plan, but items such as accessibility and limitations of the 

testing hardware often control the final placement. In many cases, the 

instrumentation plan will be “fine tuned” in the field because there will be some 

reason why a gage cannot be placed where it was originally intended to be. 

One of the factors influencing the selection of the measuring points and 

amount of instrumentation is the condition of the bridge structure. For example, it 

may be sufficient to instrument one or two spans on a multi-span structure if there 

appears to be minimal deterioration and each span has similar beam support 

conditions. On the other hand, if the bridge has significant deterioration and 

consistency among spans or beams cannot be assumed, then the most critical spans 

should be instrumented and the instrumentation on each span should be more 

extensive (each beam instrumented at three or more locations along the span). All 

gage locations must be referenced from obvious work points such as center of beam 

bearing, center of pier, etc. It is also important to clearly indicate where on the beam 

cross-section that each gage is mounted. 

For accuracy purposes, it is generally desirable to obtain measurements where 

live-load responses are reasonably large such as near midspan. However, 

measurements must also be made in less critical regions to help define the shape of 

the flexural responses and to define stress distributions within member crosssections. 

Hardware considerations 

Typical hardware used for dynamic testing of bridges are accelerometers, 

displacement sensors and strain gages. The selection of accelerometers is based on 

the frequency response expected for the structure. For structures such as bridges, 

high gain accelerometers (high sensitivity 1 v/ g) with low resonant frequency are 

recommended. The method of attachment of the sensor is very crucial for obtaining 

meaningful data which represent the actual dynamic response of the structure being 

measured. For instance, certain type of adhesives if applied with large quantity may 

develop an elastic behavior which will interact with sensor movement causing errors 

in the readings. When measuring the strain, the dynamic strain gages should be used 

rather than the static gages. Temperature compensation scheme should be used to 

eliminate strains caused by temperature variations at the point of measurement. The 

signal conditioning circuit of the strain gages should extract any noise associated 

with the signal before the digitizing process performed by A/D converter. In 

addition, an anti-aliasing filter should be used after receiving all analog signals from 

the sensors (accelerometers , strain gages) and before feeding the signals to the A/D 

converter (use sampling frequency which exceeds twice the cut-off frequency of the 

filter) in order to avoid anti-aliasing problem. 

Since long cables are necessary in this type of field testing, low noise, 

miniature coaxial cables should be used to connect the sensors (ICP sensors) to the 

data acquisition unit. 





Finally, brief observations of field data, at the time of the test, are also useful 

for detecting equipment problems such as poor transducer attachment, a 

malfunctioning transducer, or faulty load position equipment. Well trained personnel 

in data assessment can detect these problems and solve them in the field rather than 

bringing flawed data back to the office. 

Data Reduction and Analysis 

The following data processing functions and techniques are recommended in 

treating large amount of data such as the data acquired from the bridge structure; 

• RMS and average magnitudes rather than the analog signal itself. 

• Digital fast Fourier transform to emphasize the major components of the 

response and to show the contribution at each frequency 

• Cross-correlation function to enhance the common frequency components 

which are embedded in each measured signal with different strength. 

• Use dimension analysis techniques and symmetry properties in order to 

minimize the number of experiments and measuring stations. 

CONCLUDING REMARKS AND RECOMMENDATIONS 

The following concluding remarks may be stated: 

1. A selected bridge was instrumented with the various instrumentations 

and an intensive measurements program was conducted. Several 

accelerometers were placed along the girder at points with expected 

maximum response. All critical members of the bridge were 

considered (girders, cap beams and column piers) during the test. The 

proper locations for the sensors were determined after intensive 

preliminary testing and data analysis. 

2. Dynamic bridge testing was used as a means of evaluating the selected 

bridge. The tests were carried out to obtain the vibration frequencies, 

and levels of dynamic responses, and were based on normal trafficinduced 

vibrations. 

3. The collected field testing data were analyzed both in time and 

frequency domains. Statistical analysis was utilized to estimate some 

statistical measures of the vibration levels at the critical points. 

4. It is believed, based on the measurements and the post processing of 

the data, that the dynamic response of the bridge during normal traffic 

is considered normal. The max bending stress measured is shown to 

stay within the elastic limit of the girders material. 

5. It is observed that the bridge structure is well maintained and therefore, 

no local defects were noticed during the field measurements which 

extended to about 3 months. 

77


6. Guidelines for field testing of bridges are presented. This is to outline 

the basic principles and procedures for field testing of highway 

bridges. While general in nature, and related primarily to bridge 

structural performance and observed response behavior, the procedures 

described are geared toward the capabilities and functionality provided 

by the available data acquisition system, analysis software package and 

data processing software specifically designed for bridge testing. 

7. The followings can be recommended: 

1. A controlled test, where a specific vehicle with a given weight 

traveling at specific speed, can reduce the random nature of the 

response and the natural frequencies can be more pronounced in 

the frequency spectrum 

2. A large hammer test can also be of great improvement to the 

modal analysis and can provide better natural response since no 

interaction between bridge structure and the vehicles. 

ACKNOWLEDGEMENT 

The authors would like to acknowledge and thank the Deanship of Scientific 

Research of King Saud University for the financial support through the grant 

number DSR-AR-39 rovided to carry out this work. 

REFERENCES 

1- Fafard, M., Mallikarjuna, S.M., Dynamics of bridge-vehicle interactions", Proc., 

EURODYN'93, Struct. Dynamics, Balkema, Rotterdam, The Netherlands, Vol. 

2, 1993, pp. 951-960. 

2- Paultre, P., Chaalal, O., and Prouix, J., "Bridge dynamics and dynamic 

amplification factors-a review of analytical and experimental findings", Can. J. 

Civ. Engrg., 19(2), 1992, pp. 260-278. 

3- Paultre, P., Prouix, J., and Talbot, M., "Dynamic testing procedures for highway 

bridges using traffic loads", Journal of Structural Engineering, Vol. 121, No. 2, 

1995, pp. 362-376. 

4- Paultra, P., D'Aoust, I., Hebert, D., and Prouix, J., "Dynamic testing of Omervill 

Bridge", Res. Rep. SMS-93/04, Dept. of Civil Engineering, Faculty of Applied 

Science, University of Sherbrooke, Sherbrooke, Quebec, 1993. 

5- Paultre, P., Hebert, D., and Prouix, J., "Dynamic testing of Grand-Mere Bridge", 

Res. Rep. SMS-93/04, Department of Civil Engineering, Faculty of Applied 

Sciences, University of Sherbrooke, Sherbrooke, Quebec, 1993. 





6- "A guide for the field testing of bridges", Working Committee on the Safety of 

Bridges, ASCE, 1980, Rehabilitation, Konig and Nowak, eds., Ernst & Sohn 

Pub., Darmstadt, Germany, 1992. 

7- Bakht, B., and Jaeger, L.G., "Bridge testing - a surprise every time", Journal of 

Structural Engineering, Vol. 116, No. 5, 1990, pp. 1370-1383. 

8- Markey, I., "Load testing of Swiss bridges", Steel Const. Today, 5(1), 1991, pp. 

15-20. 

9- Favre, R., Hassan, M., and Markey, I., "Bridge behavior drawn from load tests", 

3 rd Int. Workshop on Bridge Rehabilitation, Koaiig and Nowak, eds., Ennst & 

Sohn Pub., Darmstadt, Germany, 1992. 

10- Moses, F., Lebet, J.P., and Bez, R., "Applications of field testing to bridge 

evaluation", Journal of Structural, Vol. 120, No. 6, 1994, pp. 1745-1762. 

11- Bakht, B., and Pinjarkar, S.G., "Review of dynamic testing of highway bridges", 

Struct. Res. Rep. SRR-89-01, Ministry of Transportation of Ontario, 

Dowonsview, Ontario, Canada, 1989. 

12- Dauenheimer, E.G., Schuring, J.R., and Lichtenstein, A.G., "Stress, strain and 

historic bridges", Proc. Int. Bridge, Conf., Pittsburgh, Pa., 1989. 

13- Ghosn, M., and Moses, F., "A reliability calibration of a bridge design code", J. 

Struct. Engrg., ASCE, 112(4), 1986. 

14- Synder, R., and Moses, F., "Application of in-motion weighing using 

instrumented bridges", Transp. Res. Record 1048, Washington, D.C., 1985. 

15- Pietraszek, T.T., "Full scale testing of steel railway bridges", 2 nd U.S. European 

Conf. on Bridge Evaluation, Repair and Rehabilitation, A.S. Nowak ed., Kluwer 

Academic Publishers, Amsterdam, The Netherlands, 1990. 

16- Boothby, T.E., and Graig, R.J., "Experimental load rating study of a historic 

truss bridge", Journal of Bridge Engineering, Vol. 2, No. 1, 1997, pp. 18-26. 

17- Mohammadi, J., and Guralnick, S.A., "Bridge fatigue life estimation from field 

data", Practice Periodical on Structural Design and Construction, Vol. 3, No. 3, 

1998, pp. 120-133. 

79




PERFORMANCE OF CONCRETE UTILIZING THE NATURAL 

POZZOLAN AVAILABLE IN THE KINGDOM OF SAUDI ARABIA 

M. I. Khan 1 , A. M. Alhozaimy 1 and A. M. Al-Saleem 2 

1: Department of Civil Engineering, King Saud University, Saudi Arabia 

2: Public Security, Ministry of Interior, Saudi Arabia 

ABSTRACT 

The results presented in this paper form part of an investigation into the 

optimisation of locally available natural pozzolanic material for the development of 

durable and sustainable concrete. Local natural pozzolan was incorporated in 

concrete as partial cement replacement to study the effect of replacement level, 

pozzolan fineness and its source. Chemical and physical properties of local 

pozzolan, properties of fresh concrete, compressive strength development and 

chloride permeability of concrete incorporating local natural pozzolanic material as 

partial cement replacement is presented. XRD analysis for the pozzolanic material is 

also reported. It has been observed that irrespective of its different sources, chemical 

and physical properties of local pozzolan are similar and conforms to the 

requirements of ASTM C 618, Class N. The inclusion of natural pozzolan in the 

concrete as partial cement replacement was not detrimental for the properties of 

fresh concrete. The incorporation of this pozzolan in the concrete did not benefit in 

the strength development. 

KEYWORDS 

Natural pozolan, chemical composition, fresh properties, compressive 

strength, chloride permeability, X-ray diffraction. 

INTRODUCTION 

Cement consumption in the Kingdom of Saudi Arabia exceeds 25 million tons 

per year, which is one of the highest per capita consumption in the world. In the 

Kingdom of Saudi Arabia, concrete structures are more prone to deterioration due to 

its hot and harsh environmental conditions particularly in the Eastern and Western 

regions. This deterioration of concrete can be minimized by using natural or byproduct 

pozzolanic materials as partial cement replacement. In the Eastern and 

Western regions of the kingdom pozzolanic material is used in the most projects to 

help in preventing the deterioration of concrete. In addition, pozzolanic material 

have lower heat of hydration than Portland cement, consequently its use reduces the 


PERFORMANCE OF CONCRETE UTILIZING THE NATURAL POZZOLAN AVAILABLE 

amount of heat generation in the concrete structures [1, 2]. This reduction in 

temperature rise is especially beneficial in concrete used for massive structures such 

as concrete dams etc. The reduction in temperature rise in the concrete is very 

beneficial for hot weather like in the Kingdom of Saudi Arabia. However, in the 

kingdom the pozzolanic material is imported which makes concrete costly. 

Natural pozzolanic material is available in Saudi Arabia from basalt plateaus 

(Harrat) spread within the “Edge of Arabian Shield” [3, 4]. The area of these 

plateaus is about 90000 sq. km in the east of the escarpment onwards to the coast of 

the Red Sea. Out of this area Harrat al Hutaymah, Harrat Lunayyir, midwest Harrat 

Rahat, Harrat al Birk and Harrat Khaybar contains pozzolanic material which can be 

used as a partial cement replacement for the production of the concrete. Its use 

conserves energy and has environmental benefits (reduced emission of carbon 

dioxide) as a result of reduction in manufacture of Portland cement. In addition, the 

use of this local pozzolanic material will be economical. 

Since past two decades the use of pozzolanic material in concrete is gaining 

impetus because of its benefits [1, 5-9]. Most of the research has been concentrated 

on by-product pozzolanic materials and little effort was dedicated to the natural 

pozzolanic materials [2, 10]. Regarding Saudi natural pozzolanic material; no 

information pertaining to engineering and durability related properties is available. 

Therefore, there is a need to investigate and explore the potential of this material for 

the use in concrete as partial cement replacement. 

The results presented in this paper form part of an investigation into the 

optimisation of locally available natural pozzolanic material for the development of 

durable and sustainable concrete. Chemical and physical properties of local 

pozzolan, properties of fresh concrete, compressive strength development and 

chloride permeability of concrete incorporating local natural pozzolanic material as 

partial cement replacement is presented. XRD analysis for the pozzolanic material is 

also reported. 

MATERIALS 

Cement 

Cement Type I complies with the requirements of the ASTM C150 was used in 

this investigation. The chemical and physical properties of cement are given in Table 1. 

Pozzolan 

Pozzolan in the form of pellets was collected from three different sources and 

ground to powder in the laboratory using pulverisette planetary type grinder. The 

fineness achieved using labortary grinding was in the range of 1800 cm 2 /g (passing 

through 75µm sieve) to 3400 cm 2 /g (passing through 45µm sieve). Pozzolan, from 

another source was procured, in powdered form with fineness of 3750 cm 2 /g. The 

82 



M. I. Khan , et al 

specific gravity was measured as 2.65. The geological classification of local natural 

pozzolan is shown in Table 2. A detailed chemical and physical properties of this 

pozzolan are presented in the section of results and discussion. 

Aggregates 

Fine and coarse aggregates available in the laboratory were used for this 

investigation. In order to meet the ASTM C 33 grading limits, 60% silica sand and 

40% crushed sand was used as fine aggregate. Crushed coarse aggregate comprising 

of 80% of 20 mm and 20% of 10 mm was used. 

Table 1: Chemical and physical analysis of Cement Type-I 

SiO2 (%) 19.97 

Al2O3 (%) 5.85 

Fe2O3 (%) 3.43 

CaO (%) 64.13 

MgO (%) 0.60 

SO3 (%) 2.80 

Loss on ignition (%) 1.60 

Fineness - Blaine (cm 2 /g) 3148 

Table 2: Geological classification of local natural pozzolan 

Category Structure / Appearance 

Volcanic tuffs (Scoria) 

Striations due to flow structures. 

Very light weight due to numerous vesicles. 

83


METHODOLOGY 

Sample Preparation 

Water to cementitious (w/c) ratio was fixed at 0.55 fo all mixes. Mixing was done 

in revolving drum mixer in accordance with ASTM C 192. The pozzolan replacements 

were selected at 17% and 25% as a partial cement replacement by weight of cement 

content. The selection of 17% replacement was on the basis of information that blended 

pozzolan cement is being produced by some local cement factories and marketed in 

Saudi Arabia which contains 17% local pozzolan. Specimens were cast and compacted 

in two layers by external vibration in accordance to the ASTM specifications. After 

casting, the samples were covered with damp hessain and polyethylene sheets for 24 

hours. The samples were demoulded the following day and then kept in the standard 

curing at 22±2 o C prior to testing. The slump test and setting time were carried out in 

accordance to ASTM C 143 and ASTM C C 403, respectively. Compresive strength was 

measured using 150 mm cubes in accordance BS 1881. The permeability test was 

performed as per the ASTM C 1202. 

X-Ray Diffraction Technique 

X-ray diffraction (XRD) is a technique to investigate the change of the 

amount of crystalline phases, during the hydration process of cement. The 

schematical configuration of the x-ray diffractometer is shown in Fig. 1. In this 

investigation, XRD was carried out for pozzolanic paste samples to monitor its 

change of the amount of crystalline phases. Pozzolan having fineness similar to 

cement was used in this paste. The paste sample was ground to form desired 

powder and compacted in a glass slide. A metal cap is placed on the back of the 

slide allowing one face to be exposed. The slide is then loaded into the 

diffractometer and the diffraction runs are controlled and logged by a computer. 

The peaks on each trace can be automatically pinpointed and the identities of 

compounds determined using database software. 

84 

Fig. 1: Schematic diagram of X-ray Diffraction 



Properties of Pozzolan 

Chemical composition 



Pozzolan was procured from four different sources. The pozzolan obtained 

from three sources (Yanbu, Rabog and Tabuk) was in the form of pellets. The 

powdered pozzolan was obtained from a Factory, Jeddah. The chemical composition 

of pozzolan was measured for all four sources as shown in the Table 3. It can be 

seen from the results that the variation in chemical composition of all four sources is 

not significant. The SiO2 content in Yanbu, Rabog and Jeddeh are close within the 

range of 42.09 to 42.75% while as for Tabuk its value is slightly low (41.14%). 

Al2O3 and Fe2O3 are within the ranges of 16.04 to 17.20% and 14.40 to 17.25%, 

respectively. The CaO content for Yanbu and Jeddah are 9.51 and 8.95% while as 

for Rabog and Tabuk values are 10.55 and 11.15%, respectively. 

Comparison between local pozzolan and Class-N of ASTM Standards 

The pozzolan used in this investigation was compared with the requirements 

of pozzolan outlined in ASTM C 618. The ASTM C 618 is the standard 

specification for coal fly ash and raw or calcined natural pozzolan for use as a 

mineral admixture in concrete where cementitious or pozzolanic action, or both, is 

desired. Raw or calcined natural pozzolans that comply with the applicable 

requirements for the Class-N are volcanic ashes or pumicites, tuffs and calcined or 

uncalcined, opaline cherts and shales, and some diatomaceous earths. 

Table 4 presents the comparison of local pozzolan with Class-N of ASTM C 

618. In this table the comparison of all sources of pozzolan is presented. It can be 

seen that the total content of silicon, aluminum and iron oxides are in the range from 

73.19 to 74.83% which is higher than the minimum requirement prescribed in 

ASTM C 618 for Class-N. SO3, alkalies and loss on ignition are much below than 

the upper limit of ASTM. It is clearly evident from the Table 4 that the local 

pozzolan conforms with the requirements of ASTM C 618. 

The strength activity index was conducted at the ages of 7 and 28 days in 

accordance to ASTM C 311. The ASTM requirement for strength index of 

cementitious/pozzolanic material is that mortar prepared in accordance with ASTM 

procedure must have 75% (0.75) compressive strength of its companion control mix 

at 7 and 28 days. Mortar mixes of control mix and mortar with 17% pozzolan were 

prepared in accordance to the ASTM C 311 specifications. The compressive strength 

results of pozzolan mix at 7 and 28 days were 77% and 78% of the control mix, 

respectively. These results demonstrate that the both 7-day and 28-day strength 

comply with the specification of ASTM C 618. 

85


86 

Table 3: Chemical composition of natural pozzolan 

Chemical 

Composition, % by mass 

constituent Source-I Source-II Source-III Source-IV 

SiO2 42.09 42.69 41.14 42.75 

Al2O3 16.43 17.20 16.44 16.04 

Fe2O3 14.97 14.88 17.25 14.40 

CaO 9.51 10.55 11.15 8.95 

MgO 3.47 2.92 1.22 3.71 

SO3 0.19 0.39 0.15 0.15 

K2O 0.81 0.29 0.95 0.85 

Na2O 0.15 0.11 0.11 0.22 

Loss on Ignition 2.70 2.60 3.20 2.90 

Table 4: Comparison of local pozzolan with Class-N of ASTM C 618 

Composition, % 

Description 

Range of local Requirement as per 

pozzolan ASTM for Class-N 

Silicon dioxide (SiO2), Aluminum 

oxide (Al2O3), Iron oxide (Fe2O3) 

73.19 – 74.83 Min. 70.0 

Sulfur trioxide (SO3) 0.15 – 0.39 Max. 4.0 

Available alkalies, as equivalent as 

Na2O 

0.28 - 0.73 Max. 1.5 

Loss of ignition 2.60 – 3.20 Max. 10.0 


Properties of Fresh Concrete 



Properties of fresh concrete such as initial slump, plastic density, air content, 

initial and final setting times for concrete containing natural pozzolan as partial 

cement replacement are presented in Table 5. 

The initial and final setting times for concrete with and without pozzolan were 

measured as shown in Table 5. The results demonstrate that the initial and final 

setting times for pozzolanic concrete mix (17% pozzolan, fineness 3750 cm 2 /g) and 

control mix are similar. In general, the pozzolanic material tends not to increase the 

air content in the given mixes. The air content of pozzolanic concrete mix (17% 

pozzolan, fineness 3750 cm 2 /g) was recorded as 1.5% which is less then control mix 

(Table 6). Therefore, pozzolan is not venerable in terms of increase in air content in 

a given mix. 

The initial slump of all mixes was within the range of 120±15 mm. The slump 

loss of pozzolanic concrete as compared to the control is shown in Fig. 2. This 

figure demonstrates that there is no significant variation in slump loss of pozzolanic 

concrete mix (17% pozzolan, fineness 3750 cm 2 /g) and the control mix. However, 

pozzolanic concrete mixture starts with slightly higher slump and drops slightly 

steeper as compared to the control mix. 

Mix 

Table 5: Properties of fresh concrete 

Initial 

slump 

mm 

Air content 

% 

Initial setting 

time 

hrs 

Final setting 

time 

hrs 

Control 120 1.9 3.5 4.5 

Pozzo-17% 

(fineness 3750 cm 2 /g) 

130 1.5 3.5 4.0 

87


88 

Slump, mm 

140 

120 

100 

80 

60 

40 

20 

0 

0 20 40 60 80 100 120 140 160 180 200 

Elapsed Time, min 


Control 

Pozo-17% 

Fig. 2: Effect pozzolanic material on the slump loss of concrete 

Compressive Strength 

Compressive strength was measured on concrete containing various 

combinations of pozzolanic material with different finenesses and sources. Influence 

of pozzolan replacement level, pozzolan fineness and pozzolan source on the 

Strength development was studied in detail. 

Influence of replacement level 

The influence of pozzolanic replacement level on the compressive strength can 

be seen is shown in Fig. 3. This figure shows the compressive strength of concrete 

prepared with 17% and 25% pozzolanic material as partail cement replacement with 

similar fineness and source. As mentioned earlier, the selection of 17% replacement 

was on the basis of information that blended pozzolan cement produced by local 

cement factories is marketed which contains 17% local pozzolan. The natural 

pozzolan having fineness 1800 cm 2 /g was incorporated at 17% and 25% levels. The 

strength of concrete with 17% pozzolan demonstrated significantly low compressive 

strength as compared to that of control mix at all ages investigated. Initially, it was 

expected that pozzolanic concrete might show some enhancement in strength at later 

ages but no improvement was recorded. As expected, mix with 25% replacement



showed lower strength as compared to the mix containing 17% pozzolan at all ages. 

At this stage, it was decided to increase the fineness of pozzolan at least equal to the 

cement fineness. Eventually, the pozzolan was further ground up to the fineness of 

3400 cm 2 /g and incorporated at 17% replacement level. 

Influence of fineness 

The influence of pozzolan fineness on the compressive strength of 

concrete is shown in Fig. 4. This figure shows the compressive strength of 

concrete prepared with 17% only having pozzolan fineness as 1800 and 3400 

cm 2 /g. It is evident from the results that the increase in fineness did not improve 

strength at any age. Mix having pozzolan fineness 3400 cm 2 /g demonstrated 

similar strength to its corresponding mix having pozzolan fineness 1800 cm 2 /g 

up to the age of 28 days. At the later ages there is insignificant increase in the 

strength. It evident from the results that there is no improvement in the strength 

in pozzolanic concrete having fineness similar to cement. It may be possible to 

improve strength development by using pozzolan having higher fineness. This 

area of resaerch is still under investigation. 


50 

40 

30 

20 

10 

0 50 100 150 200 

Age, days 

Control 

Pozo-17% 

Pozo-25% 

Fig. 3: Effect of pozzolanic replacement on the compressive strength of concrete 

(replacement level 17% and 25%; pozzolan fineness 1800 cm 2 /g) 

89


90 


50 

40 

30 

20 

10 

0 50 100 150 200 

Age, days 


Control 

Pozo-1800 

Pozo-3400 

Fig. 4: Effect of pozzolan fineness on the compressive strength of concrete 

(replacement level 17%; pozzolan fineness 1800 and 3400 cm 2 /g) 

Influence of source 

Fig. 5 demonstrates the influence of source of pozzolan on the compressive 

strength of concrete. This figure shows the compressive strength of concrete 

incorporated 17% replacement of pozzolan from two different sources. It is evident 

from the results that effect of different source did not reflect in the strength. The 

strength of concrete having pozzolan from one source are similar to the strength of 

its companion mix having different source of pozzolan, at all ages. It may be 

mentioned here that the pozzolan fineness of source-2 (3750 cm 2 /g) is slightly 

higher than that of source-1 (3400 cm 2 /g) but this slight change in fineness was 

considered negligible for the net compressive strength. Especially, keeping in view 

that pozzolan fineness in this range has no effect on the compressive strength (Fig. 

4). It is worth while to mention here that the chemical analysis of pozzolan from 

various sources did not show significant variations in their constituents (Table 4). 

Therefore, it is expected that strength of pozzolan concrete from various sources will 

remain unchanged as pozzolan's chemical composition is similar.



It is worth noting here that the pozzolan inclusion is behaving in similar 

manner irrespective of fineness, source and age as is eveident from Figs. 3 to 5. 

Therefore, it clear that the pozzolanic reaction activation is very slow. 


50 

40 

30 

20 

10 

0 25 50 75 100 

Age, days 

Control 

Source-1 

Source-2 

Fig. 5: Effect of source of pozzolan on the compressive strength of concrete 

(replacement level 17%; pozzolan fineness 3400 and 3750 cm 2 /g) 

Chloride Permeability 

The values of chloride permeability of pozzolanic concrete at the age of 

28 and 90 days are presented in Table 6. For chloride permeability test, 

pozzolan having fineness similar to cement was used. The results show that at 

90 days, the chloride permeability decreased significantly, the reduction from 

28 to 91 days was about 30%. Therefore, it is clear that the pozzolanic 

material is efficient in the refinement of pore size distribution which reflected 

in the reduction chloride permeability. These results were confirmed using 

XRD technique discussed below. 

91


92 

Table 6: Chloride permeability of pozzolanic concrete 

Chloride permeability, Coulombs 

Description 

28-day 90-day 

Pozzolanic Concrete 5503 3279 

X-Ray Diffraction Analysis 

XRD is a useful technique for the study of crystalline materials [11]. The 

reactivity of a natural pozzolan is usually determined by the amount of amorphous 

material. The constituents of a natural pozzolan are in amorphous form (reactive 

materials) to crystalline products that will react either slowly or not at all. The 

amount of amorphous materials can be determined using XRD technique. The X-ray 

diffraction for paste samples with pozzolan at 28 and 90 days, was investigated to 

identify the change in the crystalline phases. Due to pozzolanic activity the calcium 

hydroxide changes to amorphous phase of calcium silicate hydrate. Fig. 6, 

demonstrates the peaks of crystalline phases of pozzolanic paste at the age of 28 

days. In this figure, the high peaks demonstrate the presence of crystalline phase of 

pozzolan mix. At the age of 90 days, these crystalline peaks are slightly lower at the 

given conditions as shown in Fig. 7. This suggests that there is slight change of the 

crystalline phase into amorphous phase due the pozzolanic reaction. These results 

are in agreement with results of chloride permeability. 

Fig. 6: XRD of pozzolanic paste at 28 days 


CONCLUSIONS 

Fig. 7: XRD of pozzolanic paste at 90 days 

The main conclusions drawn from this research project are: 



• The local pozzolan obtained from different sources in the Kingdom showed 

similar chemical and physical properties. 

• The pozzolan conforms with the requirements of ASTM C618 and can be 

designated as Class N. Strength activity index for both 7-day and 28-day 

comply with the specification of ASTM C 618. 

• The incorporation of pozzolan showed similar slump loss as compared to 

Portland cement concrete. Whereas pozzolan slightly increased initial and 

final setting times. 

• The compressive strength development of pozzolanic concrete was low as 

compared to Portland cement concrete upto investigated age. Pozzolan 

fineness and source did not effect the compressive strength. 

93


• The chloride permeability of pozzolanic concrete demonstrated better 

performance at 90 days which was confirmed using XRD analysis. XRD 

analysis shows that there is change of the crystalline phase into amorphous 

phase due the pozzolanic reaction. 

It is evdient from this investigation that the pozzolanic reaction of local natural 

pozzolan is slower in the concrete at the normal conditions. However, its pozzolanic 

reactivity could be improved or modified after employed the appropriate 

activation/treatment. The results presented in this paper form part of an investigation 

which is underway. Further findings would be published in near future in relevant 

conferences/journals. 

ACKNOWLEDEMENT 

Authors are thankful to SABIC for funding this investigation. 

REFERENCES 

1. Malhotra, V.M. and P.K., Mehta, “Pozzolanic and Cementitious Materials – 

Advances in Concrete Technology,” Vol. I, Gordon and Breach Publishers, 

Amsterdam, Netherlands, 1996. 

2. Cook, D.J. “Natural Pozzolana,” Cement Replacement Material Vol. 3, Editor, 

R.N. Swamy, Surry Press, 1986, UK. 

3. Atlas of Industrial Minerals, “Pozzolan and Basalt,” Ministry of Petroleum 

and Mineral Resources, Directorate General of Mineral Resources, Kingdom 

of Saudi Arabia. 

4. Information Bulliten No. 13, Ministry of Petroleum and Mineral Resources, 

Directorate General of Mineral Resources, Kingdom of Saudi Arabia, 

(01.07.1419H). In Arabic. 

5. Hooton, R.D., “Permeability and Pore Structure of Cement Pastes Containing 

Fly Ash, Slag, and Silica Fume,” Blended Cements, ASTM STP 897, G. 

Frohnsdorff, Ed., American Society for Testing and Materials, Philadelphia, 

1986, pp. 128-143. 

94 




6. Alhoziamy, A., P., Soroushian and F., Mirza, “Effects of Curing Condition 

and Age on Chloride Permeabilty of Fly Ash Mortar”, ACI Material Journal, 

Vol. 93, No. 1, Jan-Feb 1996, pp. 87-95. 

7. Khan, M.I. and C.J., Lynsdale, “Strength, Permeability and Carbonation of 

High-performance Concrete,” Cement and Concrete Research, USA, Vol. 32, 

No. 1, February 2002, pp 123-131. 



9. FIP Concrete Report. “Condensed silica fume in concrete,” FIP state-of-art 

report, fip Commission of Thomas Telford House, London, 1988. 

10. Mehta, P.K., “Studies on Blended Portland Cement Containing Santorin 

Earth,” Cement and Concrete Research, Vol. 11, 1981, pp. 507-518. 

11. Scrivener K. L., T. Füllmann, E. Gallucci, G. Walenta and E. Bermejo, 

"Quantitative study of Portland cement hydration by X-ray 

diffraction/Rietveld analysis and independent methods," Cement and Concrete 

Research, Vol. 34, 2004, pp. 1541-1547. 

95


96 


ABSTRACT 

QUALITY OF CONCRETE USED IN HOUSING 

CONSTRUCTION IN RIYADH AND THE ROLE OF THE 

FORTHCOMING SAUDI BUILDING CODE 

Abdulaziz I. Al-Negheimish 1 , Abdulrahman Alhozaimy 1 , 

Saleh Al-Sulaiman 2 , Saud Al-Swaida 2 , and Said Shahrani 2 

1 Civil Engineering Department, King Saud University 

2 General Directorate of Operation and Maintenance, Municipality 

of Riyadh 

A quality scheme for ready-mixed concrete (RMC) was started by the 

Municipality of Riyadh in 1995. Results of random samples collected at the RMC 

plants as part of the scheme’s continuous monitoring activities, show acceptable 

quality of concrete is being produced by most plants. Yet, complaints of bad quality 

concrete especially on small construction sites still persist. In this paper, the actual 

quality of concrete used in housing construction is investigated and the adverse 

effect of adding water to the delivery truck prior to concreting work and its impact 

on the compressive strength is quantified. In addition, data on the quality of sitemixed 

concrete (SMC) which are still used occasionally on some sites and for 

critical elements such as columns are presented. The results show serious 

shortcomings in the construction practice and the quality of concrete used in 

housing. To ensure the quality of concrete used in housing construction, the use of 

SMC for structural applications should be prohibited. Also, effective site 

supervision and acceptance testing stipulated by SBC 304 should be implemented 

for all buildings including small housing projects. 

KEY WORDS 

Ready-mixed concrete, Site-mixed concrete, Quality control, Housing 

construction, Saudi Arabia. 


QUALITY OF CONCRETE USED IN HOUSING CONSTRUCTION IN RIYADH AND THE ROLE 

INTRODUCTION 

Site-mixed concrete (SMC) was used for all housing construction in Riyadh in 

the 1970’s and the early 1980’s resulting in concrete construction of substandard 

quality. The introduction of ready-mixed concrete (RMC) into Saudi Arabia during 

the construction boom of the mid 1970’s was significant milestone toward better 

quality of concrete construction. However, RMC usage was limited to large 

projects due to its high cost. It was during the 1990’s that the use of RMC became 

dominant in all segments of the construction industry including housing but 

complaints and doubts about the quality of concrete from RMC plants persisted [1- 

4]. 

To improve the performance of RMC plants and enhance concrete quality, a 

quality scheme for RMC was started by the Municipality of Riyadh in 1995. The 

Scheme resulted in significant improvement in all aspects of RMC operations. As 

part of this scheme, random checking of the quality of concrete is done at the plant. 

The results from the Municipality program show that concrete with acceptable 

quality is produced by most RMC plants [5,6]. 

According to results from extensive study by the Al-Negheimish and 

Alhozaimy [7], the compressive strength at plant and site should be similar provided 

that no additional water is added to the delivery trucks during delivery or at the job 

site. Slump at site is expected to loose 30 to 40% of its initial value (slump 

measured at plant) during delivery. Concrete temperature should increase by an 

average of 2°C during delivery due to heat of hydration of cement, solar radiation 

and thermal gain from the hot environment. However, the wide spread practice of 

adding water to the delivery trucks during concrete placement is certain to impact 

the measured properties of concrete at site. This practice is still one of the major 

risks to concrete quality which needs to be addressed as part of the forthcoming 

implementation of the Saudi Building Code. 

This study was designed to evaluate the actual quality of concrete used on 

small construction sites and to assess chances of non-compliance with the 

forthcoming Saudi Building Code for Concrete Construction (SBC304) [8]. 

Specifically, the adverse effect of adding water to the delivery truck at site and its 

impact on the compressive strength is evaluated. 

METHODOLOGY 

This study involved taken random samples of concrete from RMC delivery 

trucks during discharge of concrete at the job site during hot weather months of 

May to October. For each sample, slump, temperature of fresh concrete at the time 




Abdulaziz I. Al-Negheimish, et al 

of delivery and compressive strength at 7 days and 28 days were measured. The 

properties of concrete at site were compared to the properties of concrete measured 

at the plant by the Municipality of Riyadh as part of their Quality Scheme to 

monitor the quality of RMC production in Riyadh. 

The test sample is collected as concrete being discharged into the pump at a job 

site. Sufficient amount of concrete is collected using plastic container for 

performing the slump test and for casting 5 cubes (150 mm x 150 mm x 150 mm). 

The use of small plastic container instead of wheel barrow was done to improve 

mobility and to make surprise inspection of the concrete work in progress easier. At 

the time of sampling, the delivery ticket issued by the RMC was checked and 

relevant data about the mix and the time of loading and arrival at the construction 

site were recorded. In addition, relevant information such as ambient temperature, 

type of construction and other observations of concreting work were recorded. 

These include the addition of water to the delivery truck at the job site and site 

supervision by the owner or his representative. 

The sample collected was remixed thoroughly and concrete temperature was 

measured and the slump test was carried out in accordance with ASTM C143. Also, 

5 cubes were cast, as per BS 1881, in 3 layers by tamping each layer 36 times. The 

specimens were kept in the shade and covered properly to prevent evaporation and 

loss of water. The following day, the cubes were collected and transported to the 

Concrete Laboratory at King Saud University, where they were demold and placed 

in curing tank until the time of testing at the age of 7 and 28 days. 

The results of slump, concrete temperature and compressive strength for 

samples collected at the job site were compared to data from the same plants 

obtained by the Municipality of Riyadh for samples collected at the plants during 

the same period. The differences in slump, concrete temperature and compressive 

strength of concrete are related to observations made at the job site, especially, the 

practice of uncontrolled addition of water to the delivery truck. 

For comparison, SMC samples were collected from several construction sites 

where mixing was done with or without mechanical mixers. Sampling and testing 

procedures was similar to that outlined above. Extra care was taken to ensure that 

samples are representative of the actual quality of concrete used on the construction 

site. The Quality of SMC was compared to that of RMC. 

99


RESULTS AND DISCUSSIONS 

READY-MIXED CONCRETE 

A total of 82 samples were collected from RMC delivery trucks at job sites 

during the hot summer months. The majority of the samples (about 86.6%) had 

cement content of 350kg/m 3 . Site supervision during concreting work was found to 

be limited or non-existent. Also, adding water to the delivery trucks at the 

construction site to increase slump was found to be wide-spread with the 

documented cases of water addition exceeding 63.4% of all sites. 

Data on concrete temperature, slump and compressive strength of samples with 

350kg/m 3 were analyzed and compared with data collected from the same plants and 

for the same grade by the Municipality Quality Scheme during the same period. 

Results of these comparisons between site samples and samples collected at the 

plant show the changes in the properties of concrete and reveal the impact of adding 

water at the job site on these properties. 

Overall Comparisons 

Concrete temperature 

Comparison between the concrete temperature measured at the plant and that 

measured at site is shown as a histogram in Figure 1. Concrete temperature at both 

plant and site follows normal distribution with a mean value of 32.8 °C and 35.4 °C, 

respectively. The increase in concrete temperature of 2.6 °C is slightly higher than 

that reported earlier by Al-Negheimish and Alhozaimy [7]. These results clearly 

show the need for better control of temperature during summer in order to satisfy 

the forthcoming Saudi building code requirements for concrete structures (SBC304). 

According to SBC304 [8], temperature of concrete delivered to site shall not exceed 

35°C. Data from this study show this limit is likely to be exceeded on many sites. 

This means that non-compliance with code provisions for maximum temperature is 

going to be a serious problem during the initial implementation stage of the code. 

Slump 

Concrete delivered to job site shall have adequate workability to allow for 

adequate consolidation and finishing. According to the requirements of the Quality 

Scheme for RMC of the Municipality of Riyadh, the slump should be in the range of 75- 

125 for footings and 100-150mm for slabs and other shallow elements. At plants, 

slump should be 50% higher than these values to allow for the slump loss during 

delivery. 





Comparison between the slump at the plant and slump measured at site is 

shown as a histogram in Figure 2. The distribution of the measured slump at the 

plant follows normal distribution with a mean value of 152 mm. However, the 

statistical characteristics of slump at site are totally different due to addition of water 

at the job site,. The average slump at site increased to 184.5mm and the distribution 

is severely skewed to the right. Slump for 52% of samples exceeded 180mm. 

Instead of the normal reduction in the initial slump (or slump loss) which takes 

place with the passage of time, the average slump at site increased by 21.4% as a 

result of addition of water at the job site. The slump measured at site is significantly 

higher than the average slump of 100 and 125 mm recommended by the Quality 

Scheme for RMC for footings and slabs, respectively. The increased slump (due to 

the addition of water) should result in proportional decrease in compressive strength 

and may cause segregation.. Also, the increased w/c ratio as a result of adding 

water is certain to impact the durability of hardened concrete. 

Compressive strength 

Comparison between the 28-day compressive strength of concrete from the 

Municipality random samples collected at the plant and those taken at site is shown 

as a histogram in Figure 3. The compressive strength of samples collected at the 

plants follows normal distribution; however, the distributions of site samples show 

clear shift to the left. The average compressive strength at the plant is 38.4 MPa 

compared to 33.8 MPa at site. The reduction of 4.6 MPa corresponds to a drop of 

12% in the compressive strength. The reduction in compressive strength is due to 

the addition of water to concrete on many sites. The practice has significant impact 

on the chances of non-compliance with strength requirements of Saudi Building 

Code. Based on Figure 3, the chances of falling below 30 MPa is 7%, a modest 

increase of 2% over the approximately 5% chances stipulated by the code. By 

contrast, for samples taken at site, the percentage is 29% which means serious noncompliance 

problem with strength provisions of SBC304 unless the issue of adding 

water to concrete is solved through effective site supervision. The use of 

superplasticizers instead of water to adjust workability, when needed, should be 

made mandatory for all RMC plants in Riyadh. 

Comparisons for Individual Plants 

The deterioration in the quality of concrete at site is examined further by 

comparing the average temperature, slump and compressive strength at plant and 

site for six plants having four or more samples at both plant and site. The results of 

these comparisons are presented in Figures 4 through 6. 

Concrete temperature 

Figure 4 shows comparison of average concrete temperature between plant and 

site for the six plants. For most plants, the figure shows substantial increase in 

concrete temperature at site compared to plant. The average concrete temperature at 

101


plant ranges from a low of 28.5 °C to a high of 34.7 °C. By comparison, the lowest 

average concrete temperature at site was 32.6 °C and the highest 39.6 °C. Based on 

these data, only plant No. 5 is expected to meet the 35 °C code limit on concrete 

temperature at site without difficulties. Other plants need to enhance their 

temperature control capabilities in order to satisfy the code limit consistently. Other 

means to avoid violating the code limits is to reschedule concreting work during 

summer to cooler hours such as at night and early hours of the morning. 

Slump 

Figure 5 shows comparison of average slump between plant and site for the six 

plants. The average slump at plant is in the expected range, with the highest slump 

being 160.7 mm for plant No. 3 and the lowest 138.9 for plant No. 6. However, the 

slump at site is higher than that measured at plant for all plants ranging from a 

maximum of 213 mm for plant No. 1 to 161.8 mm for plant No. 3. For four out of 

the six plants, the average slump is well above 180 mm. These results clearly show 

that all plants are in the habit of increasing the slump at site; however, some are 

worst than others in such practice. The underlying cause for such practice is the 

demand by contractors on site for increased slump to speed up placing operations 

coupled by almost total lack of engineering supervision and control at the job site 

during concreting work. 


Figure 6 shows comparison of average compressive strength between plant and 

site for the six plants. The figure shows consistently lower strength for site 

compared to plant. The highest average strength at plant was 47.3 MPa for plant 

No. 5 and the lowest was 36.6 MPa for plant No.4. By comparison, the highest 

strength achieved at site was 39 MPa for plant No. 2 and the lowest was 28 MPa for 

plant No. 1. The obvious reason for the reduction of compressive strength at the site 

in comparison with the strength achieved at the plant is the wide-spread practice of 

adding water to concrete at the construction site. 

Impact of adding water on the compressive strength 

Adding water to concrete is known to have negative impact on the compressive 

strength. In a recent study by Alhozaimy [9], it was found that the negative impact 

on strength is related to the slump. In cases where water is added to restore the 

slump within the specification’s limits (100 ± 25 mm), the decrease in strength is 

below 10 percent. However, when water is added to increase slump beyond these 

limits the decrease in strength is in the range of 25-35%. The rate of the reduction 

in compressive strength is approximately 2% for each 10 mm increase in slump. 

Figure 7 depicts the relationship between compressive strength and slump at plant 

and at site based on data from thirteen RMC plants covered in this study. At the plant, 

the compressive strength shows large variation among plants but does not have any 





correlation with slump. However, at site a clear correlation between compressive 

strength and slump exists with high slump corresponding to lower compressive strength. 

The reduction in compressive strength of 1.5 MPa for each 10 mm increase in slump is 

much higher than the reduction due to retempering alone cited by Alhozaimy [9]. It 

appears that a combination of long waiting time and maintenance of higher slump by 

retempering with water more than one time is the reason for the higher loss of strength. 

SITE-MIXED CONCRETE 

A total of 16 samples of site-mixed concrete (SMC) were collected from 

construction sites in several districts in Riyadh over a period of few weeks. Slump 

and compressive strength for each sample are given in Table 1. The table includes, 

also, location, type of structural element and mixing method used. Slump was very 

high ranging from 180 mm to collapse. Typical consistence of SMC used for footing 

on one of the site to avoid compaction is shown in figure 8. The 28-day strength is 

very low ranging from 4.8 to 14.0 MPa. These values are extremely low compared 

to the strength of RMC documented in this study and much lower than the 21.2 MPa 

and 17.0 MPa for mechanical and manual mixing reported previously by Arafah etal 

[10]. The reason for these low values is high water content, the use of non-graded 

aggregates and the apparently low cement content used. According to SBC 304, the 

minimum design strength allowed by the code is 25 MPa based on cube (20 MPa 

based on cylinder). Therefore, the extremely low quality of SMC precludes its use 

for structural applications designed and constructed in accordance with SBC304. 

SUMMARY AND CONCLUSIONS 

This study was conducted to investigate the actual quality of concrete used in 

housing construction in Riyadh and to quantify changes of properties of concrete 

between plant and job site. Both the quality of RMC and SMC at site were 

documented. Based on the results of this study, the following conclusions are made: 

1. Addition of water to concrete at construction sites to increase slump is 

wide-spread with the documented cases of water addition exceeding 

63.4%. 

2. The average compressive strength at site is 33.8 MPa compared to 38.4 

MPa at the plant. The drop of 12% in the compressive strength is due to 

addition of water to concrete at the job site. 

3. Effective site supervision during concreting work is presently lax or nonexistent 

on most construction sites. 

4. The use of superplasticizers instead of water to adjust workability, when 

needed, should be made mandatory for all RMC plants in Riyadh. 

103


5. The quality of SMC is very low and its use for structural applications 

should be prohibited. 

6. Effective site supervision and acceptance testing stipulated in the SBC 304 

are keys to ensuring quality of concrete used for housing construction. 

REFERENCE 

1. ﻞﺠ�ﺳ " ،ﺎ�ﻬﺗدﻮﺠﻟ نﺎﻤ�ﺿ ةﺰهﺎ�ﺠﻟا ﺔﻧﺎ�ﺳﺮﺨﻟا ﻊﻧﺎﺼ�ﻣ ﻲ�ﻓ ﺔ�ﻴﺗاﺬﻟا ﺔﺒﻗاﺮﻤﻟا ﻦﻣ ﻖﺛﻮﺘﻟا" 

،ﺐﻴﺒﺣ ،ﻦﻳﺪﺑﺎﻌﻟا ﻦﻳز 

نﺎﺒﻌ��ﺷ ،ﺔﻳدﻮﻌﺴ��ﻟا ﺔ��ﻴﺑﺮﻌﻟا ﺔ��ﻜﻠﻤﻤﻟا ،ضﺎ��ﻳﺮﻟا ،ﺎ��ﻬﺘﺠﻟﺎﻌﻣ ﺔ��ﻴﻔﻴآ و ﻲ��ﺑﺮﻌﻟا ﻢﻟﺎ��ﻌﻟﺎﺑ ﻲﻧﺎ��ﺒﻤﻟا تﺎﻋﺪﺼ��ﺗ ةوﺪ��ﻧ 

٢٢٧ -٢١٩ 

ص ،ـه١٤١٢ 

2. ﻢﻟﺎ�ﻌﻟﺎﺑ ﻲﻧﺎ�ﺒﻤﻟا تﺎﻋﺪﺼ�ﺗ ةوﺪ�ﻧ ﻞﺠ�ﺳ " ، ﺔﻧﺎﺳﺮﺨﻟا 

ةدﻮﺟ ﺔﺒﻗاﺮﻤﻟ لﺎﻌﻓ مﺎﻈﻧ ﺮﻳﻮﻄﺗ " ،ﺐﻴﺒﺣ ،ﻦﻳﺪﺑﺎﻌﻟا ﻦﻳز 

. ٢١٨-٢٠٧ 

ص ،ـه١٤١٢ 

نﺎﺒﻌﺷ ،ﺔﻳدﻮﻌﺴﻟا ﺔﻴﺑﺮﻌﻟا ﺔﻜﻠﻤﻤﻟا ،ضﺎﻳﺮﻟا ،ﺎﻬﺘﺠﻟﺎﻌﻣ ﺔﻴﻔﻴآ و ﻲﺑﺮﻌﻟا 

3. Al-Negheimish, A.I., 1994, " Ready Mixed Concrete Practice in Developing 

Countries-Case Study," Proceedings of the 5th International Colloquium on 

Concrete in Developing Countries, Cairo, Egypt, Jan. 1994, pp.1047-1057. 

4. Medallah, K , 1996, “Towards establishing a Quality Assurance Program for 

Saudi Readymix Concrete Industry,” Saudi Commerce & Economic Review, 

No. 25, May 1996, pp 30-31 

5. Alhozaimy, A, Al-Negheimish A I., 1999, “Introducing and Managing Quality 

Scheme for RMC Industry in Saudi Arabia,” J. Construction Engineering and 

Management, ASCE, V.125, No. 4, pp. 249-255. 

6. ﻲﻧﺎ�ﻄﺤﻘﻟا ﻦﻳﺪ�ﻟا ﻲ�ﺤﻣ و ،ءاﺪﻳﻮﺴ�ﻟا 

دﻮﻌ�ﺳ ،نﺎﻤﻴﻠﺴﻟا 

ﺢﻟﺎﺻ ، ﺶﻤﻴﻐﻨﻟا ﺰﻳﺰﻌﻟا ﺪﺒﻋ ، ﻦﻤﺣﺮﻟاﺪﺒﻋ ،ﻲﻤﻳﺰﺤﻟا 

،" 

ﺔﻴﻧﺎﺳﺮﺨﻟا تاءﺎﺸﻨﻤﻠﻟ يدﻮﻌﺴﻟا ءﺎﻨﺒﻟا دﻮآ ﻖﻴﺒﻄﺗ حﺎﺠﻧإ ﻲﻓ ةﺰهﺎﺠﻟا ﺔﻧﺎﺳﺮﺨﻟا ةدﻮﺟ ﺔﺒﻗاﺮﻣ مﺎﻈﻧ ﺔﻴﻤهأ" 

. دﻮﻌﺳ 

ﻚﻠﻤﻟا ﺔﻌﻣﺎﺟ ،ﻊﺑﺎﺴﻟا يدﻮﻌﺴﻟا ﻲﺳﺪﻨﻬﻟا ﺮﻤﺗﺆﻤﻟا ﻲﻓ ﺮﺸﻨﻠﻟ ﺔﻟﻮﺒﻘﻣ ﺔﻗرو 

7. Al-Negheimish A. and Alhozaimy, A., 2001, “Ready-mixed Concrete Problems 

Associated with Hot and Dry Weather in Riyadh,” Final Report, King Abdulaziz 

City for Science and Technology, Grant No. LG-1-54, Riyadh, Saudi Arabia. 

8. SBC 304, 2005, "Saudi Building Code for Concrete Structures," Part 3: 

Structural Design, Saudi Building Code National Committee, Riyadh. 

9. Alhozaimy, A, 2007, “Effect of retempering on the compressive strength of 

ready-mixed concrete in hot-dry environments,” Journal of Cement and Concrete 

Composites, Vol. 29, pp. 124-127. 

10. ﻲ�ﻓ ﺔﻧﺎ�ﺳﺮﺨﻟا ﺔﻋﺎﻨﺻ رﻮﻄﺗ" 

ﻲﻤﻳﺰﺤﻟا ﻦﻤﺣﺮﻟاﺪﺒﻋ و ﺪﻳﺰﻟا ﺢﺟار ،ﺶﻤﻴﻐﻨﻟا ﺰﻳﺰﻌﻟا ﺪﺒﻋ ،ﻢﻴﺣﺮﻟاﺪﺒﻋ ﺔﻓﺮﻋ 

-١١ 

،ضﺎﻳﺮﻟا ،دﻮﻌﺳ ﻚﻠﻤﻟا ﺔﻌﻣﺎﺟ ،ﻲﺳﺪﻨﻬﻟا رﻮﺤﻤﻟا ،ىﺮﺒﻜﻟا ﺔﻴﻌﻣﺎﺠﻟا ةوﺪﻨﻟا " ﺔﻳدﻮﻌﺴﻟا ﺔﻴﺑﺮﻌﻟا ﺔﻜﻠﻤﻤﻟا 

. 

ـه١٤٢٠ 

،ﺐﺟر ١٢ 



Sample 

No. 

Table 1: Data and properties of site-mixed concrete 

Location Method of 

mixing 

Slump (mm) 



28 day 

compressive 

strength (MPa) 

1 Alhamra mechanical 180 10.3 

2 Yarmouk mechanical collapse 7.8 

3 Ashbilia mechanical collapse 7.0 


5 Ashbilia mechanical 180 6.6 

6 Ashbilia mechanical 190 14.0 


8 Nazeem manual collapse 4.8 

9 Nazeem manual collapse 6.4 

10 Nazeem mechanical collapse 12.3 

11 Wadi Laban mechanical collapse 10.0 






Percentage of samples 

30.0 

25.0 

20.0 

15.0 

10.0 

5.0 

0.0 

Plant 

Site 

40 

Concrete Temperature, C 

Figure 1: Distribution of concrete temperature at site vs. plant 

105


Percentage of Samples 

Percentage of Samples 

40.0 

35.0 

30.0 

25.0 

20.0 

15.0 

10.0 

5.0 

0.0 

40.0 

35.0 

30.0 

25.0 

20.0 

15.0 

10.0 

5.0 

0.0 

Plant 

Site 

٦٠> ٨٠-٦٠ ١٠٠-٨٠ -١٠٠ 

١٢٠ 

-١٢٠ 

١٤٠ 

-١٤٠ 

١٦٠ 

Slump, mm 

-١٦٠ 

١٨٠ 



-١٨٠ 

٢٠٠ 

-٢٠٠ 

٢٢٠ 

Figure 2: Distribution of slump at site vs. plant 

Plant 

Site 

-٢٢٠ 

٢٤٠ 

١٥> ٢٠-١٥ ٢٥-٢٠ ٣٠-٢٥ ٣٥-٣٠ ٤٠-٣٥ ٤٥-٤٠ ٥٠ -٤٥ 

٥٠< 


Figure 3: Distribution of compressive strength at site vs. plant

Concrete Temperature,C 

Slump, mm 

50 

45 

40 

35 

30 

25 

20 

15 

10 

5 

0 

250 

225 

200 

175 

150 

125 

100 

75 

50 

25 

0 

31.0 

34.1 

39.6 

34.7 34.0 34.4 35.1 

33.9 



32.6 

28.5 

31.1 

١ ٢ ٣ ٤ ٥ ٦ 

Plant Code 

Plant 

Site 

Figure 4: Comparison of concrete temperature at site vs. plant 

213 

144 141 

183 

161 162 

176 

168 

١ ٢ ٣ ٤ ٥ ٦ 

Plant Code 

143 

Plant 

Site 

Figure 5: Comparison of slump at site vs. plant 

187 

139 

203 

36.8 

107



50 

40 

30 

20 

10 

0 

C0mpressive Strength, MPa 

50 

45 

40 

35 

30 

25 

20 

15 

10 

5 

0 

40.3 

28 

44.5 

Figure 6: Comparison of compressive strength at site vs. plant 

Site 

Plant 

ﻲﻄﺧ 

(Si ) 

39 39.1 38.7 

y = -0.0254x + 42.44 

R 2 = 0.0113 

36.6 

32.7 

y = -0.15x + 61.017 

R 2 = 0.5639 

100 110 120 130 140 150 160 170 180 190 200 210 220 230 240 250 

Slump, mm 

Figure 7: Correlation of compressive strength with slump 

47.3 



34.3 

37.6 

1 2 3 4 5 6 

Plant Code 

33.3 

Plant 

Site

a) Loading aggregates into the mixer 



b) Pouring concrete into footing 

Figure 8: Preparation and casting of site-mixed concrete. 

109




STRENGTH DEVELOPMENT OF CONCRETE MADE OF TYPE I AND 

TYPE V CEMENTS PRODUCED IN SAUDI ARABIA 

Abdulrahman Alhozaimy 1 and Abdulaziz Al-Negheimish 2 

1. Civil Engineering Department, King Saud University, P.O. BOX 800, Riyadh 

1421, alhozimy@ksu.sa 

2. Civil Engineering Department, King Saud University, P.O. BOX 800, Riyadh 

1421, negaimsh@ksu.sa 

ABSTRACT 

Current cement consumption in the Kingdom exceeds 25 million tons per 

year, which is one of the highest per capita consumption in the world. There are 

eight cement companies operating in the Kingdom with total production capacity of 

about 22 million tons per year. New plants and major expansion of existing plants 

are expected to increase the production capacity of cement to more than 45 million 

in the near future. 

In this paper, the variability of chemical composition of cements produced in 

the Kingdom and their effect on strength development of concrete are assessed. Test 

results showed that C3S and C2S compounds of Type I produced in the Kingdom are 

similar to Type V with the major difference in C3A content. Both cement types can 

be classified as low-alkali cements based on ASTM C150. The results of 

compressive strength of concrete made with Type I cement were 2-4 MPa higher 

than corresponding concrete made with Type V at all investigated ages (up to 5 

years). 

KEYWORD 

Cement, Saudi Arabia, Chemical Composition, Concrete Strength 

INTRODUCTION 

Portland cement concrete is the dominant construction material in the 

Kingdom of Saudi Arabia and other countries in the Middle East. The first cement 

factory in the Kingdom was established in Jeddah in 1959 with production capacity 

of 100,000 tons per year. The number of cement factories increased gradually and 

cement production reached about 700 thousand tons per year in 1970. Cement 


STRENGTH DEVELOPMENT OF CONCRETE MADE OF TYPE I AND TYPE V CEMENTS 

production vs. consumption in the Kingdom during the periods 1970-2005 is shown 

in Figure 1. Cement consumption reached the highest level during the construction 

boom of the mid-seventies; the total cement consumption reached about 12.9 million 

tons in 1980, while the production was only 2.9 million tons. Cement consumption 

continued to increase until 1983 when it touched 24 million tons then decreased 

gradually till it reached about 10.9 million tons in 1990. In the mean time, cement 

production continued to grow steadily and became almost equal with its 

consumption in 1990. To meet the consumption demand, the kingdom has imported 

around 10-15 million tons of cement per year early 1980's. In the last 15 years, both 

production and consumption grew steadily reaching 22 million tons per year, which 

is one of the highest per capita consumption in the world [1,2]. There are eight 

cement factories geographically spread all over the kingdom with total production 

capacity of 22 million tons per year in 2005. Most cement factories produce only 

Type I and Type V cement. Currently, there are new plants and major expansion of 

existing plants. It is expected that the cement production in the Kingdom will reach 

45 million tons per year in the coming years. 

The quality of cement is vital for the production of good concrete. The 

manufacturing of cement requires stringent control and conformance to the approved 

standards and specifications. The properties of local cement comply with 

international standards. However, the variability of physical and chemical properties 

of cement from these factories is not well documented. The issue of strength 

development of concrete made with locally cement and its variation with existing 

literature has not been addressed. 

This paper presents the results of experimental program conducted to examine 

both chemical and physical properties of local cement. Cement samples for both 

Type I and Type V used in this study were obtained from four cement factories in 

the Kingdom. These samples were also used to study the strength development of 

concrete made using local cement and to establish the ratio of 7-day to 28-day 

strength of concrete. 

112 


Quantity (million of Tons) 

25 

20 

15 

10 

5 

0 

Consumption 

Production 

EXPERIMENTAL PROGRAM 


Abdulrahman Alhozaimy and Abdulaziz Al-Negheimish 

1970 1975 1980 1985 1990 1995 2000 2005 

Year 

Fig. 1. Cement production vs. consumption 

The experimental program was designed to investigate the cement properties 

and the effect of local cement types and sources on the strength development of 

concrete. Cement samples for both Type I and Type V were collected from four 

factories In addition, data-sheet showing the chemical and physical properties of 

each type of cement were obtained from factories. X-ray fluorescence analysis 

(XRF) was conducted on each cement to determine the elemental compositions of 

cement form the various sources. In addition, 8 laboratory mixes were prepared 

using the two cement types from the four sources. For each mix, a total of 32 cubes 

(150 x 150 x 150 mm) were prepared and cured in lime-saturated water in the 

laboratory until the testing age. Compressive strength testing was performed at the 

age of 3, 7, 28, 90, 180 days, 1, 2, and 5 years. 

113


MATERIALS AND MIX PROPORTIONS 

The cement used was Type I and Type V obtained for this study. Coarse 

aggregates were a blend of crushed limestone with aggregate sizes of 20 mm and 10 

mm. The specific gravity based on saturated and surface dry condition (SSD) of 20mm 

coarse aggregate was 2.65, and its absorption capacity was 1.69%, while the 

specific gravity of 10-mm coarse aggregate was 2.7, and its absorption capacity was 

1.94%. The fine aggregate with absorption capacity of 1.2% was used. The 

gradation of coarse and fine aggregates met the ASTM C 33 requirements. The mix 

proportions used in this investigation are shown below: 

114 

Cement 350 kg/m 3 

Water 192.5 kg/ m 3 

20 mm Agg. 730 kg/m 3 

10 mm Agg. 390 kg/m 3 

Sand 715kg/m 3 

MIXING AND CURING PROCEDURES 

Specimens for the compressive strength were cast from eight separate batches. 

Each batch was mixed in a concrete mixer of 0.4 cubic meter capacity. The mixing 

procedures followed basically ASTM C-192. After mixing, the concrete temperature 

and slump tests were performed for each batch according to ASTM C1064 and 

ASTM C 143, respectively. The initial concrete temperatures of all mixes were in 

the range of 17 – 20 o C. The effect of cement source and type on the workability of 

concrete was limited with slump values of all eight batches being in the range of 125 

± 15 mm. 

Cube specimens (150 x 150 x 150 mm) were cast in steel molds according to 

BS 1881 Part 116. The cube specimens were consolidated by external vibration in 

two layers. After casting, the specimens were covered with wet burlap and plastic 

sheets for 24 hours. After demolding, the specimens were cured in lime saturated 

water at temperature of 21 ± 2 o C until the testing age. 


Chemical Composition 

In the Kingdom all cement factories produce cement in accordance with 

ASTM C 150. Independent chemical analysis of cement samples for the purpose of 

comparison with test data provided by the cement factories was performed; the 

results are shown in Table 1. The results of the chemical analysis are similar to the 




values provided by cement factories in their test certificates. The difference was 

very limited within the range of 0 to 1% except for CaO and loss on ignition values, 

which differed by up to 2%. These results confirm that the results provided by the 

cement factories in the test certificate are accurate and reliable. 

The chemical compound composition, calculated by Bogue equations for both 

Type I and Type V are given in Table 2. It can be observed that the chemical 

composition of the cement from the various sources is similar with small difference 

between the maximum and minimum values for both types of cement. The mean 

values of compound composition are comparable to the typical values of Type I and 

Type V cements published in the literature [3-5] with the exception of C3S and C2S 

for Type V. 

In Portland cement, C3S and C2S are the most important compounds, which 

constitute about 75% of the cement by weight and are responsible for the strength of 

hydrated cement paste. As shown in Table 2, the mean values of C3S and C2S are 54 

and 18% for Type I cement, and 59% and 16% for Type V cement. By comparison, 

typical values from the published literature [3-5] for C3S and C2S are 49-55% and 

19-25% for Type I cement and 38-43% and 36-43% for Type V cements, 

respectively. It is obvious that Type V cement produced by cement factories in the 

Kingdom is different from typical values reported in the literature. It is interesting 

to observe that the C3S and C2S compounds of Type V cement in the Kingdom are 

similar to Type I cement. Therefore, the C3A and C4AF are the main compounds 

which differ between Type I and Type V cement. 

Another potentially important component in cement composition is alkalis. 

Although Na2O and K2O are minor component in the cement, they are important 

because they have been found to react with some aggregate, known as alkaliaggregate 

reaction. In addition, it is reported that compressive strength has strong 

correlation with alkali content (in particular K2O content), indicating that cements 

with a high alkali content produced concrete with lower compressive strength [6]. It 

seems, however, that it is difficult to produce low alkali-cement economically in 

most countries. Not only source of low alkali cements becoming more scarce, but 

also modern process tend to concentrate the alkali in clinker during the 

manufacturing of cement [4, 5]. According to ASTM C150, cement is considered 

low alkali cement when alkalis (equivalent = Na2O +0.66 K2O) is below 0.6%. The 

values of alkali equivalent for local cements are within the range of 0.21 to 0.54 

(Table 1). By comparison, typical values from the published literature are in the 

range from 0.3 to 1.2 % [5, 7]. The results from this study clearly show that all 

cement produced in the Kingdom can be classified as low-alkali cement. 

115


Table 1: Chemical Analysis of All Cements. 

Chemical 

Component 

s% 

116 

Source of Cement 

“A” “B” “C” “D” 

Cement Type Cement Type Cement Type Cement Type 

I V I V I V I V 

SiO2 19.81 20.94 21.78 22.28 20.10 20.34 20.27 21.06 

Al2O3 6.05 3.92 4.66 3.51 5.26 4.24 6.11 4.32 

Fe2O3 3.46 4.77 2.78 3.83 4.14 5.13 3.31 5.45 

CaO 63.86 63.74 64.21 64.82 63.67 63.49 61.27 62.61 

MgO 0.81 0.73 2.08 2.10 1.31 1.27 2.79 1.37 

SO3 2.16 1.88 2.17 1.86 2.11 2.14 2.12 1.92 

K2O 0.08 0.09 0.21 0.18 0.28 0.13 0.12 0.10 

Na2O 0.16 0.16 0.32 0.28 0.36 0.21 0.18 0.16 

Alkali 

Equivalent 

as Na2O 

Loss on 

Ignition 

0.21 0.22 0.46 0.40 0.54 0.30 0.26 0.23 

3.35 3.56 1.46 1.24 2.76 2.62 2.56 2.74 

Table 2: Chemical compound composition of all cements (calculated by Bogue 

equations) 

Chemical 

Compo-nents 

% 

Source 

A 

Source 

B 

Type I Cement Type V Cement 

Source 

C 

Source 

D 

Average Source 

A 


Source 

B 

Source 

C 

Source 

D 

Average 

C3S 57.63 54.35 59.13 43.50 53.65 61.78 60.14 61.90 52.50 59.08 

C2S 13.32 21.43 13.02 25.29 18.27 13.42 18.50 11.60 20.77 16.07 

C3A 10.18 7.63 6.95 10.59 8.84 2.32 2.82 2.56 2.23 2.48 

C4AF 10.52 8.47 12.59 10.08 10.41 14.50 11.66 15.61 16.59 14.59

Cement Fineness and Setting Time 



The fineness values of both cement types from the various sources are shown 

in Table 3, The values are within the range of 282 to 333 m 2 /kg with a mean value of 

305 m 2 /kg for Type I and in the range of 287 to 336 m 2 /kg with a mean value of 306 

m 2 /kg for Type V. These values are less than range of 350-380 m 2 /kg for similar 

cements reported in the literature [3, 4]. However, the fineness of Saudi cement 

satisfy the minimum requirement of 280 m 2 /kg stipulated by the ASTM C 150 for 

both types of cement. The results of the setting time performed on the samples from 

the four factories are shown in Table 3. These results are within the recommended 

range of > 45 minutes and < 375 minutes. 

Table 3: Results of Fineness and Setting Time of All Cement Sources 

Source 

of 

Fineness of Cement 

m 2 /kg 

Cement Type I Type V 

Setting Time (minuities) 

Type I Type V 

Initial Final Initial Final 

A 299 310 110 155 155 200 

B 333 336 155 200 205 250 

C 304 287 120 160 105 150 

D 282 289 115 155 155 205 

Effect of Cement Source and Type on Compressive Strength Development 

The development of strength of concrete made with both Type I and Type V 

from the four sources are shown in Figures 2 and 3, respectively. These figures 

show that the compressive strength increases with age. Figure 2 shows that the 

variation in compressive strength of Type I among the four sources is limited at 

early age; however, the variation increases with time. For example, at 3 days, the 

range of variation in compressive strength of concrete made from all sources was 2.8 

MPa compared to 9.2 MPa at 1 year. The same trend was also observed for Type V 

cement as can be seen from Figure 3. The cause of the variation in strength among 

the four sources being more pronounced at later age is not clear. Some possible 

causes of this phenomenon could be the raw materials and the manufacturing 

process [5, 6]. 

117


A Comparison of average values of compressive strength between Type I and 

Type V from all cement sources for periods up to 5 years is shown in Figure 4. It 

can be seen from the figure that the compressive strength for Type I cements is 

higher than Type V cements by about 2 to 4 MPa at all ages up to 5 years. It is 

established from the literature that Type V cement gain strength more slowly than 

Type I because of the lower C3A content, but at later age, strength is the same or 

higher. Saudi cement showed Type I to maintain at least 2 MPa advantage even at 5 

years. The explanation for this unexpected behavior can be attributed to chemical 

composition of Saudi cement. As discussed earlier, the C3S and C2S contents for 

both Type I and Type V were similar with the major difference being the C3A, 

which is higher for Type I. Therefore, the higher early strength of Type I compared 

to Type V was sustained with age up to 5 years. 

118 


60 

50 

40 

30 

20 

10 

0 

Source A 

Source B 

Source C 

Source D 

Average 

3 7 28 90 180 

Days 

Age (log scale) 

1 2 5 10 

years 

Fig 2. Strength development of concrete made with type I cement 



60 

50 

40 

30 

20 

10 

0 

Source A 

Source B 

Source C 

Source D 

Average 



3 7 28 90 180 

Days 


1 2 5 10 

years 

Fig 3. Strength development of concrete made with type V cement 


60 

50 

40 

30 

20 

10 

0 

Type 1 

Type V 

3 7 28 90 180 

Days 


1 2 5 10 

years 

Fig. 4. Compressive strength of Type I vs. Type V 

119


The strength development of concrete as affected by cement source and type 

were examined. Table 4 shows compressive strength at all considered ages as a 

percentage of the 28-day strength. The results show the ratios of 3 to 28-day strength 

are in the range of 0.49 to 0.58 with a mean value of 0.54 for Type I and 0.45 to 0.55 

with a mean value of 0.49 for Type V. The ratios of 7 to 28-day strength show that 

the ratios fall in the range of 0.72 to 0.81 with a mean value of 0.75 for Type I and 

in the range of 0.68 to 0.74 with a mean value of 0.70 for Type V. These results 

show clearly that Type I gained strength faster than Type V at early age which is 

consistent with expectations based on the literature [2]. Also, the ratio of 7 to 28day 

strength is similar to that reported by [5] for modern cement. It should be 

emphasized that these values are based on mixes with w/c ratio of 0.55 and plain 

cement without admixtures. 

The strength gain beyond 28 days is affected by cement source and type. Strength 

gain from 28-day to 1 year was in the range of 29 to 50% with a mean value of 39% for 

Type I and 35 to 60% with a mean value of 45% for Type V. The variability within each 

type is the effect of cement source. Also, the strength gain beyond 28-day for Type V is 

higher than for Type I. Based on these results it is clear that the strength gain of Saudi 

cement is similar to cement in other countries [4, 5]. 

Table 4: Strength changes as a percentage with respect to 28-days strength 

Age at 

Type I cement Type I cement 

Test Source Source Source Source Source Source Source Source 

Average 

A B C D A B C D Average 

3 days 54.6 49.3 54.7 58.1 54.2 49.7 55.1 45.5 47.4 49.4 

7 days 76.5 71.9 72.8 80.6 75.4 68.0 73.7 68.1 68.3 69.5 

28 

days 

90 

days 

180 

days 

120 

100 100 100 100 100 100 100 100 100 100 

113.1 122.9 112.1 124.1 118.0 125.8 126.9 116.9 129.4 124.7 

122.1 137.0 122.5 132.1 128.4 135.0 147.2 127.6 130.1 134.9 

1 year 139.9 150.4 128.7 137.5 139.1 151.3 160.1 134.6 135.6 145.4 

2 

years 

5 

years 

144.3 154.7 136.1 142.5 144.4 162.1 169.3 140.9 153.6 156.5 

157.4 168.5 142.3 159.4 156.9 167.3 176.3 147.2 158.8 162.4 


CONCLUSIONS 



1. The C3S and C2S compounds of Type V cement produced in the Kingdom are 

similar to Type I cement. Based on ASTM C150 classification, both cement 

types can be classified as low-alkali cements. 

2. Compressive strength of concrete made from Type I cement gives 2-4 MPa 

higher strength as compared to Type V at all ages up to 5 years. The higher 

strength of Type I can be attributed to the fact that C3S and C2S contents for 

both Type I and Type V were similar with the major difference being the 

C3A, which is higher for Type I. 

3. Cement type and source plays an important role in the variation in the 

strength development of concrete. Type I gained strength faster than Type V 

at early age. The ratios of 7-day to 28-day strength are in the range of 0.72 to 

0.81 with a mean value of 0.75 for cement Type I and 0.68 to 0.74 with a 

mean value of 0.70 for Type V. 

4. Strength gain of Saudi cements is similar to cement in other countries. The 

strength gain from 28-day to 365-day is in the range of 29 to 50% with a 

mean value of 39% for Type I and 35 to 60% with a mean value of 45% for 

Type V. 


This study was funded by KACST under Grant No. LGP-1-54 and their 

financial support is highly appreciated. Testing was done at the Concrete 

Laboratory, King Saud University. The help and commitment of everybody 

involved are gratefully acknowledged. 

REFERENCES 

[1] Alhozaimy, A, Al-Negheimish A.,1999, “Introducing and Managing 

Quality Scheme for RMC Industry in Saudi Arabia,” J. Construction 

Engineering and Management, ASCE, V.125, No. 4, pp. 249-255. 

[2] Tuncalp, S and Alibrahim, A. C.,1990, "Development of cement Industry in 

Saudi Arabia" Industrial Cooperation in Arabian Gulf Journal, No, 42, pp. 

3-25. 

[3] Kosmatka, S. and Panarese W., 1992, Design and Control of Concrete 

Mixtures, 13 th Ed., PCA, Skokie, Illinois, 

121


[4] Mindess and Young, 1981, Concrete, Prentice-Hal Inc., New Jersey. 

[5] Neville, A. M., 1995, Properties of Concrete, 4th ed., Longman, England. 

[6] Rose, K., et al., 1989, “Factors Affecting Strength and Durability of 

Concrete Made with various Cements,” Transportation Research Record, 

No. 1234, pp 13-23 

[7] Mehta, P.,K., and Monteiro. P.J., 1993, Concrete: structure, properties, and 

materials 2 nd Edition, Prentice-Hall Inc, New Jersey. 

122 


PERFORMANCE OF PUSHOVER PROCEDURE IN EVALUATING THE SEISMIC 

ADEQUACY OF REINFORCED CONCRETE FRAMES 

ABSTRACT 

A. Shuraim , and A. Charif 

Civil Engineering Dept., King Saud University 

ashuraim@gmail.com 

The nonlinear static analytical procedure (Pushover) as introduced by ATC-40 

was applied for the evaluation of existing design of a reinforced concrete frame, in 

order to examine the applicability of the pushover for evaluating design of new 

buildings. Potential structural deficiencies in the frame were assessed by the code 

seismic-resistant design and pushover approaches, for the sake of comparison. In the 

first approach, the potential deficiencies were determined by redesigning under one 

selected seismic combination in order to show which members would require 

additional reinforcement. In the second approach, a pushover analysis was 

conducted to assess the seismic performance of the frame and detect the locations of 

the plastic hinges. The paper shows that vulnerability locations revealed from the 

two procedures are significantly different, where the latter procedure tends to 

overestimate column strength, consequently, concealing earlier detection of column 

weaknesses. The paper provides rational explanations for the apparent discrepancy 

that can be taken into consideration in order to make pushover methodology 

applicable when designing or evaluating existing design of new buildings. 

KEYWORDS 

building codes, structural, coding, seismic design. 


PERFORMANCE OF PUSHOVER PROCEDURE IN EVALUATING THE SEISMIC ADEQUACY 

INTRODUCTION 

The generalized nonlinear static analytical procedure (Pushover) is a key 

element in the methodology introduced by ATC-40 for the seismic evaluation and 

retrofit design of existing buildings which represents a fundamental change for the 

structural engineering profession. The methodology is performance-based where the 

design criteria are expressed as performance objectives, which define desired levels 

of seismic performance when the building is subjected to specified levels of seismic 

ground motion. The generalized nonlinear static analytical procedure incorporated in 

the methodology has three primary elements ( as shown in 

Figure 1 shows a capacity curve of a structure by use of a static pushover 

analysis, a method to determine displacement demand by use of reduced demand 

spectra, and the resulting identification of the performance point and the subsequent 

check for acceptable performance. Plastic hinges observed prior to the performance 

point reveals the locations of the potential deficiencies as well as the damage extent. 

ATC-40 asserts that although the methodology is not intended for the design of new 

buildings, the analytical procedures are applicable. 

Over the previous decade, pushover analysis has been carried out for either 

user-defined nonlinear hinge properties or default-hinge properties, available in 

some programs based on the ATC-40[1] and FEMA-356 [2] guidelines. In the 

implementation of pushover analysis, modeling is one of the important steps where 

all material plasticity is lumped at appropriately located hinges. The capacity 

response of the structure is closely related to the plastic properties assigned to the 

various members (beams and columns). Such a model requires the determination of 

the nonlinear properties of each component in the structure that are quantified by 

strength and deformation capacities. Some programs (i.e. SAP2000 [3, 4]) have 

already implemented these default nonlinear properties. The use of this 

implementation is very common among the structural engineering profession and 

researchers. In the past several years, many researchers [5-10] have discussed the 

underlying assumptions and limitations of pushover analysis. The pushover has been 

utilized to investigate the seismic performance of reinforced concrete frames with 

emphasis on the suitability of default material assumptions [11-12]. 



Spectral Acceleration 

Capacity Curve 

Demand Spectrum 


Performance point 

Spectral Displacement 

Figure 1: schematic representation of ATC-40 method. 

A. Shuraim , A. Charif 

The ATC-40 [1] assertion that the analytical procedure is applicable to new 

buildings raises some questions in view of the known significant differences in the 

underlying assumptions of the pushover procedure in comparison to those in design 

codes [13-17] for new buildings. In new building design, the code always maintains 

certain factor of safety that comes from load factors, materials reduction factors, and 

ignoring some post yielding characteristics (hardening). In the modeling assumptions of 

ATC-40, reduction factor is assumed to be one, and hardening is to be taken into 

consideration. Therefore, it is important to evaluate pushover accuracy and reliability in 

comparison to code design criteria that are relevant to seismic design. 

The paper addresses the applicability of the pushover to the design of new 

buildings, through a 2D RC frame case study having been designed for gravity load 

only. The adequacy/deficiency of the existing longitudinal reinforcement in the 

frame for resisting moderate seismic forces will be assessed by the code seismicresistant 

design and pushover approaches. In the former approach, new longitudinal 

reinforcements for the frame members will be compared with the existing 

125


reinforcement where the increase in the reinforcement represents potential 

deficiency in the original design. In the latter approach, the capacity curve, spectrum 

demand and performance point will be determined on the basis of the existing 

gravity design. Plastic hinges formed prior to the performance point represent the 

locations of the deficiency in the original design. Locations of deficiency from the 

two approaches will be evaluated and discussed. It is always believed that 

verifications of design/analysis procedures and their underlying assumptions are an 

essential step for safety and economic considerations. 

CASE STUDY 

The case study is a typical regular 2D frame with three bays and three stories 

All column and beams sections are 300 x 500 mm but internal columns are 

used in the weak direction ( b = 500 mm and h = 300 mm) deliberately to make them 

more vulnerable than external columns (see Figure 2) . The ground beams were 

included in the model with a much lower loading and the base is assumed fixed. 

The frame was first analyzed and designed with SAP2000 [1] using a standard 

linear analysis combining dead and live loading. 

Figure 3 shows the various reinforcement percentages for the gravity load 

combination (U1= 1.4 D + 1.7 L). A unique steel percentage is given for columns 

whereas for beams, top and bottom values are given at both ends as well as at mid-span. 

Figure 2: frame labeling numbers 





Figure 3: Steel reinforcement percentages in design under combination U1 

SEISMIC-RESISTANT DESIGN 

Earthquake-induced inertia forces depend on the response characteristics of 

the structure and the intensity of ground motion at the site. The latter depends 

primarily on three factors: the distance between the source and the site, the 

magnitude of the earthquake, and the type of soil at the site. Different individual 

structures shaken by the same earthquake respond differently. One important 

characteristic is the fundamental period of vibration of the structure. Shape or 

configuration is another important characteristic that affects structure response. 

It is generally uneconomical and unnecessary to design a structure to respond in 

the elastic range to the maximum earthquake-induced inertia forces. Thus, the design 

seismic horizontal forces prescribed in the seismic codes (UBC 97[13], IBC 2003[14], 

ASCE-7[15], SBC 301[16]), are generally less than the elastic response inertia forces 

induced by the design earthquake. Acceptable performance can be achieved by structures 

elastically designed for reduced forces, if suitable structural systems are selected, and 

structures are detailed with appropriate levels of ductility, regularity, and continuity. 

Accordingly, structural systems are expected to undergo fairly large deformations, 

allowing inelastic energy dissipation, when subjected to a major earthquake. Some 

structural and nonstructural damage can be expected due to large deformations. 

Therefore, seismic provisions regulate both strength and lateral drift. 

127


In this paper, base shear was computed and distributed vertically in 

accordance with UBC97 seismic provisions, assuming Ca = Cv =0.2 (Zone 2B soil 

SB). The total base shear computed was modified slightly (V =428.1 kN) in order to 

permit comparison with one of the critical pushover cases. 

Table 1: Summary of un-factored loads on the frame 

Load Case Global Global Global 

FX, kN FY, kN FZ, kN 

DEAD 0 0 5218.848 

Live 0 0 850.56 

Ex 428.1 0 0 

Following the same provisions, there are a number of load combinations that need 

to be considered. However, for the sake of this study, the frame was designed under one 

load combination only (termed seismic 1), namely: U2 = 1.1( 1.2 D +1 L+ 1 E). 

Based on the above load combination, the RC frame was designed where the 

longitudinal reinforcement ratios are presented in 

In Figure 4 the reinforcement from U1 and U2 combinations are presented in 

Table (2) and Table (3) for columns and beams, respectively. The positive percentage of 

change indicates the deficiency in a member if subjected to seismic loading under U2 

combination. The findings from this procedure can be summarized as: 

1. Twelve columns out of the sixteen columns require additional reinforcement. Six 

of them need more than one-hundred percent increase, indicating their deficiency 

under U2. Column 26 exhibits the highest increase, where the original 

reinforcement under gravity amounts to only 1500 mm 2 , and becomes 6060 mm 2 , 

under U2 and thus would require strengthening if the frame is part of an existing 

building. It should be pointed out that negative percentage of change should not 

be interpreted as a need for decreasing reinforcement. 

2. Four beams out of the sixteen beams require additional top reinforcement 

exceeding one-hundred percent, at their right end. Other beams show 

modest increase. Beam 21 exhibits the highest increase, where the original 

top reinforcement under gravity amounts to only 351 mm 2 , and becomes 

998 mm 2 , under U2 and thus would require strengthening if the frame is 

part of an existing building. 

3. Overall, columns in this case study need more strengthening than beams. 



Figure 4: code design under seismic 1 (U2) only 



129


Table 2: columns longitudinal reinforcment (Gravity versus seismic1) 

Longitudinal reinforcement, mm 2 Change (%) 

Column No. Section 

Gravity Seismic 1 

1 EXC 1500 4018.817 168% 

2 EXC 1500 1984.007 32% 

3 EXC 1588.138 1500 -6% 

4 EXC 3285.352 1500 -54% 

9 INC 5667.08 7267.426 28% 

10 INC 4295.008 8423.157 96% 

11 INC 1500 4702.885 214% 

12 INC 1500 1500 0% 

17 INC 5667.08 6737.943 19% 

18 INC 4295.008 7167.375 67% 

19 INC 1500 4188.63 179% 

20 INC 1500 1500 0% 

25 EXC 1500 6310.764 321% 

26 EXC 1500 6060.629 304% 

27 EXC 1588.138 5211.875 228% 

28 EXC 3285.352 4395.174 34% 





Table 3: Beams longitudinal reinforcement (Gravity versus seismic1 (U2)) 

Beam 

No. 

Section 

Longitudinal 

reinforcement, mm 2 

Gravity Seismic 1 

Mid 

section 

End 

section 

Longitudinal 

reinforcement, mm 2 

Mid 

section 

End 

section 

5 GDB 0 268.644 0 285.632 6% 

Chang 

e (%) 

5 GDB 345.58 0 933.419 0 170% 

6 FLB 2302.381 0 3306.89 246.005 44% 

7 FLB 2230.003 0 2915.75 0 31% 

8 RFB 2071.969 0 2328.786 0 12% 

13 GDB 0 269.283 0 288.923 7% 

13 GDB 354.61 0 874.452 0 147% 

14 FLB 2227.354 0 2651.251 0 19% 

15 FLB 2173.2 0 2468.543 0 14% 

16 RFB 2005.8 0 2099.135 0 5% 

21 GDB 351.118 0 998.118 0 184% 

22 FLB 1318.269 0 3100.768 28.898 135% 

23 FLB 1509.525 0 2688.267 0 78% 

24 RFB 1139.567 0 1624.229 0 43% 

131


PUSHOVER ANALYSIS AND STRUCTURE CAPACITY 

The design results (reinforcement sections) obtained previously, were 

used to assess the model performance as an existing structure designed under 

gravity loading only. It was aimed to assess its seismic response in a typical 

earthquake zone with seismic coefficients Ca = Cv = 0.2. The static nonlinear 

analysis combined the application of the dead load followed by the application 

of the lateral seismic forces which were increased up to failure under 

displacement control. The SAP2000 default hinge properties were used for 

beams and columns. Hinges were assigned at both ends of each element and 

for beams a mid-span hinge was also assigned to track possible span hinges 

because of uniform loading. Pure bending hinges were used for beams whereas 

for columns, the hinges were related to the axial force – bending moment P-M 

interaction curve. 

Figure 5 shows the structure capacity response up to failure. The 

structure developed a maximum base shear of about 612 kN and the ultimate 

roof lateral displacement is about 141 mm. The nonlinear pushover curve 

illustrates the successive formation and evolution of plastic hinges. These 

were formed in seven different steps. Figure 6 shows the plastic hinge patterns 

at the third step of loading, corresponding to a base shear value of 475 kN. 

The figures show the locations of the hinges as well as their state illustrated by 

appropriate colors. 

This nonlinear response curve illustrates also the first yielding and may be 

used to quantify the ductility of the structure. It cannot however serve to evaluate the 

actual seismic performance of the structure unless it is compared to the actual 

demand of the seismic action. 



Figure 5: Capacity curve 



133


Figure 6: Plastic hinges from nonlinear analysis, at the third step, 

base shear = 475 kN. 

SEISMIC DEMAND AND PERFORMANCE POINT 

The seismic demand on a structure is usually expressed in the form of a design 

spectrum according to the prevailing seismic code and including all structural and 

zoning parameters. The seismic demand is also related to the nonlinear behavior of 

the structure and is obtained iteratively. 

Figure 7 shows the ATC-40 [1] demand spectrum using the initial seismic 

coefficients (Ca = Cv = 0.2) for a standard 5% damping ratio. The intersection of the 

demand spectrum with the nonlinear pushover response is called “Performance 

Point”. It corresponds to the state the structure is expected to reach under the 

considered earthquake. Depending on the position and state of the performance point 

(with respect to the actual pushover curve), the analyst may decide on how safe or 

vulnerable the structure is and where possible strengthening should be performed. 

For our particular example the performance point is indicated by a base shear value 

of 484 kN corresponding to a roof displacement of 75 mm. 



The forgoing results provide a number of interesting observations. 



1. The capacity-demand intersection suggests that the frame is expected to 

withstand the assumed moderate seismic shaking of 2B soil SB, even 

though it was not designed for any earthquake forces. 

2. All beams have reached their nominal yield capacity at one or more points 

and some redistribution of moments have taken place. Plasticity 

promulgation at the right end of Beam 6 is illustrated in 

3. Figure 8: The sectional moment is 463 kN-m which is higher than its 

nominal yield flexural capacity of 448 kN-m, computed on the basis of 

SBC 304 assumptions without reduction factor. The excess moment beyond 

the yield is assumed to be carried by the contribution of hardening that is 

not considered in the design. 

4. Only one column has reached its nominal yield capacity, namely Column 

26. This is illustrated by the interaction diagram in Figure 9. 

5. The column is subjected to a demand point (M= 272 kN-m and an axial 

load of 1327 kN) which is shown to be located approximately on the 

perimeter of the nominal interaction diagram (Figure 9). 

135


Figure 7: Capacity, Demand and Performance point according to ATC- 40 

Figure 8: plasticity level at the end of step 3, in member 6. 



Axial load, kN 

4500 

4000 

3500 

3000 

2500 

2000 

1500 

1000 

500 

0 

-500 

P_hinge P_a 

0 50 100 150 200 250 300 

Column Moment , kN-m 



Figure 9: M-P interaction diagram of column 26 and demand point. 

WHY NO PLASTIC HINGES IN OTHER COLUMNS 

To understand the apparent discrepancy between the first linear code method 

which shows that most columns are deficient and the pushover procedure which 

shows deficiency in only one column, it is important to investigate a typical case 

such as column 10. Based on the code procedure column 10 is subjected to a 

moment of 175 kN-m and an axial load of 2300 kN at its end. Based on the former 

code procedure, the column is required to additional reinforcement in order to 

withstand the seismic demand prescribed by case U2. While based on the latter 

pushover procedure, it is capable of carrying the demand. 

Figure 10 shows the demand point and two interaction diagrams. The outer 

diagram is the nominal diagram assumed in the pushover while the inner diagram 

represents the reduced diagram on the basis of code provisions of column design. 

The demand point is shown between the two curves. From the perspective of code 

provisions, the demand point is outside the inner design curve and thus the column is 

deemed deficient and requires additional reinforcement as indicated by the linear 

design shown earlier. On the other hand, the pushover default assumptions ignore 

these provisions and do not incorporate the code reductions factors and upper limits 

137


imposed by the code provisions. Therefore, the discrepancy is attributed to the 

assumptions existed in the two procedures, and engineering judgment should be 

exercised when conducting pushover analysis. 

If the procedure is to be used for designing new structures, code provisions 

should be fully observed regardless of the method used. On the other hand, for an 

existing building some reduction of nominal capacity should be imposed based on 

the actual conditions and the level of safety deemed acceptable. 

CONCLUSIONS 

Axial load, kN 

6000 

5000 

4000 

3000 

2000 

1000 

0 

-1000 

-2000 

P_hinge φP_n P_a 

0 50 100 150 200 250 300 

Column Moment , kN-m 

Figure 10: column No 10 interaction diagram. 

The nonlinear static analytical procedure (Pushover) as introduced by ATC-40 

has been utilized for the evaluation of existing design of a new reinforced concrete 

frame, in order to examine its applicability. Potential structural deficiencies in RC 

frame, when subjected to a moderate seismic loading, were estimated by the code 

seismic-resistant design and pushover approaches. In the first method the design was 

evaluated by redesigning under one selected seismic combination in order to show 

which members would require additional reinforcement. It was shown that most 

columns required significant additional reinforcement, indicating their vulnerability 

if subjected to seismic forces. On the other hand, the nonlinear pushover procedure 

shows that the frame is capable of withstanding the presumed seismic force with 





some significant yielding at all beams and one column. Vulnerability locations from 

the two procedures are significantly different. 

The paper has discussed the reasons behind the apparent discrepancy which is 

mainly due to the default assumptions of the method as implemented by the software 

versus the code assumptions regarding reduction factors and maximum permissible 

limits. In new building design, the code always maintains certain factor of safety that 

comes from load factors, materials reduction factors, and ignoring some post 

yielding characteristics (hardening). In the modeling assumptions of ATC-40, 

reduction factor is assumed to be one, and hardening is to be taken into 

consideration. 

Hence, the paper suggests that engineering judgment should be exercised 

prudently when using the pushover analysis and that engineer should follow the 

code limits when designing new buildings and impose certain reductions and limits 

in case of existing buildings depending on their conditions. In short software should 

not substitute for code provisions and engineering judgment. 

REFERENCES 

1. Applied Technology Council, 1996, ATC-40: Seismic Evaluation and 

Retrofit of Concrete Buildings, vols. 1 and 2. California. 

2. Federal Emergency Management Agency, 2000, FEMA-356: Prestandard 

and Commentary for Seismic Rehabilitation of Buildings, Washington, DC. 

3. CSI. SAP2000 V-10. Integrated finite element analysis and design of 

structures basic analysis reference manual. Berkeley (CA, USA): 

Computers and Structures Inc; 2006. 

4. Habibullah, A., Pyle, S., 1998, “Practical Three Dimensional Nonlinear 

Static Pushover Analysis”, Structure Magazine, Winter, 1998. 

5. Krawinkler, H., Seneviratna, G.D. , 1998, “Pros and Cons of a Pushover 

Analysis of Seismic Performance Evaluation”, ASCE, Journal of Structural 

Engineering, Vol. 20, pp. 452-464. 

6. Naeim, F., Lobo, R. M., 1998, “Common Pitfalls in Pushover Analysis.” 

Proceedings of the SEAOC Annual Convention, Reno, Nevada. 

7. Kim, B., D’Amore, E., 1999, “Pushover Analysis Procedure in Earthquake 

Engineering.” Earthquake Spectra, Vol. 13(2), pp. 417-434. 

139


8. Elnashai, A. S., 2001, “Advanced Inelastic Static (Pushover) Analysis for 

Earthquake Applications”, Structural Engineering and Mechanics,Vol. 

12(1), pp. 51-69. 

9. Fajfar, P., “Structural Analysis in Earthquake Engineering—A 

Breakthrough of Simplified Non-Linear Method”, Paper Reference 843, 

Proceedings of the 12th European Conference on Earthquake Engineering, 

London,. 

10. Chopra, A. K., 2004, “Estimating Seismic Demands for Performance-Based 

Engineering Of Buildings”, Paper No. 5007, 13th World Conference on 

Earthquake Engineering, Vancouver, B.C., Canada. 

11. Lee, H-S., Woo, S-W, 2002, “Seismic Performance of a 3-Story RC Frame 

in a Low-Seismicity Region”, Engineering Structures, Vol. 24, pp. 719– 

734. 

12. Inel, M., Ozmen, H. B., 2006, “Effects of plastic hinge properties in 

nonlinear analysis of reinforced concrete buildings”, Engineering 

Structures, Vol. 28, pp. 1494–1502. 

13. ICBO, et al. “Uniform Building Code (UBC),” by International Conference 

of Building Officials (ICBO), Whittier, California; 1997. 

14. International Code Council, Inc., International Building Code, 2003. 

15. American Society of Civil Engineers, Minimum Design Loads for Buildings 

and Other Structures, SEI/ASCE 7-02, Reston, Virginia, 2002. 

16. SBC 301, “Design loads for Building and Structures”, draft of Saudi 

Building Code, 2007. 

17. SBC 304, “Concrete Structures”, draft of Saudi Building Code, 2007. 



SELECTION AND EFFICIENCY OF CORROSION INHIBITORS FOR 

CORROSION PROTECTION IN FERROCEMENT COMPOSITES 

S. Akhtar 1 , M. Arif 2 , and M.A. Quraishi 3 

1: Department of Civil Engineering, Aligarh Muslim University, Aligarh- 202 002 

(India), sabihced@yahoo.co.in 

2: Department of Civil Engineering, College of Engineering, P. O. Box 1299, 

Qassim University, Buraidah (KSA), marifamu@gmail.com 

3: Department of Applied Chemistry, Institute of Technology, B.H.U., Varanasi 

(India), maquraishi@rediffmail.com 

ABSTRACT 

Reinforcement corrosion continues to be a matter of serious concern 

since it governs the durability of composites. The issue assumes greater 

relevance in context to the climatic conditions of Saudi Arabia. The 

phenomenon is especially important for fibre reinforced composites mainly 

ferrocement, since the diameter of the wire meshes used in ferrocement are 

much smaller as compared to the conventional reinforced cement concrete. 

The problem of corrosion of meshes in ferrocement becomes more damaging 

under saline environment. An attempt has been made to improve the corrosion 

resistance of the metallic wire meshes used in ferrocement by using corrosion 

inhibitors. Two corrosion inhibitors namely Calcium Nitrite and Tannic Acid 

were used. The dose of inhibitors was kept at 5%. Tests were undertaken 

under normal (tap) water, artificial saline water (normal water mixed with 4% 

NaCl), tap water with inhibitors and artificial saline water with inhibitors. 

Excellent to good corrosion inhibition has been observed for both of these 

inhibitors. The results have been further validated using Potentio-dynamic 

polarization tests. 

Key words 

Corrosion inhibitors, corrosion inhibition efficiency, ferrocement, fibre 

composites, wire meshes, durability. 


SELECTION AND EFFICIENCY OF CORROSION INHIBITORS 

INTRODUCTION 

Mild steel wire meshes having diameter usually in the range of 0.5 to 1.5 mm 

are used as reinforcement in ferrocement composites [1]. Corrosion protection is 

generally achieved through the use of galvanized wire mesh or by making dense 

mortar with the use of additives such as fly ash, silica fumes and blast furnace slag [2- 

4]. Increasing the effective cover has also been reported in literature as one of the way 

of controlling corrosion of reinforcement [5]. These methods suggested by various 

researchers have been proven to be ineffective with the passage of time thereby 

reducing the effective diameter and the bond between wire meshes and the matrix, 

leading to the reduction in strength and durability of the ferrocement components [6]. 

ACI-549R strongly recommends that studies be undertaken to suggest durable and 

long term anti corrosion techniques to prevent penetration of water and salts that could 

lead to the corrosion of reinforcing wire mesh [7]. It has been clearly established that 

corrosion inhibitors are extremely effective in controlling/delaying onset of corrosion 

in protecting rebar in concrete [8-9]. The present study is aimed at the assessment of 

the effectiveness of two corrosion inhibitors namely Calcium Nitrite (Type-I) and 

Tannic Acid (Type-II) in protecting mild steel plate and welded steel wire meshes 

under normal (tap) water, saline water, tap water with Type I and II inhibitors and 

saline water with Type I and II inhibitors. 


Three types of metallic specimen viz. mild steel plate, welded steel wire mesh 

and cement slurry coated welded steel wire mesh were taken for the present 

investigation. A total of six exposure mediums were considered using normal water 

and artificial saline water; both with and without corrosion inhibitors. 

Immersion Test 

Specimens 

For the immersion test (weight loss test) mild steel plate specimen of size 

25×18.7×0.25 mm and galvanized welded steel wire mesh specimen of size of 

25×20 mm with 1.42 mm diameter and 12.5 mm opening size were chosen. 

Exposure Condition 

A total of six exposure mediums namely normal (tap) water, artificial 

saline water (distilled water mixed with 4% NaCl), tap water with Type I and II 

inhibitors and saline water with Type I and II inhibitors were undertaken for the 

present investigation. The concentration of NaCl for saline water was kept as 

4% which is very near to sea water NaCl content. The exposure time was 1 

month at room temperature. 




S. Akhtar, M. Arif , and M.A. Quraishi 

Corrosion Inhibitors 

Two corrosion inhibitors namely Calcium Nitrite (Type-I) and Tannic Acid 

(Type-II) were chosen for the present study. The dose of inhibitors was kept at 1, 3 

and 5%. 

Details of the investigation scheme for immersion test are given in Table-1. 

The efficiency of corrosion inhibitors on the basis of weight loss study was 

calculated as: 

% Efficiency = [(Wi – W) / Wi] 3100 % (1) 

where, Wi = initial weight and W = weight loss of the specimen. 

Corrosion rate was calculated using the following equation [10] 

Corrosion rate (mpy) = [(534W)/(DAT)] (2) 

where, W = weight loss of the specimen in mg, D = density of the specimen in 

gm/cm 3 , 

A = surface area of the specimen in sq. inch, T = exposure time in hours 

Efficiency and Corrosion rate so calculated have been presented in the Table-2. 

Potentio-dynamic Polarization Test 

Potentio-dynamic polarization studies were carried out using an EG & G 

(PAR model 173) potentiostat/galvanostat, a model 175 universal programmer and a 

model RE 0089 X-Y recorder. A platinum foil was used as the auxiliary electrode, a 

saturated calomel electrode was used as reference electrode and mild steel was used 

as working electrode. The experiments were carried out at a constant temperature of 

28±2 °C and a scan rate of 1mVs -1 versus O.C.P. A time interval of about 10–15 

minutes was given for each system to attain a steady state and the open circuit 

potential (OCP) was noted. The polarization curves were obtained after immersion 

of the electrode in the solution until a steady state was reached. The polarization 

curves are shown in Fig. 1. Data collected from these curves are given in Table-3. 

RESULTS AND DISCUSSIONS 

It has been observed that the inhibitor Type-I not only delays the onset of 

corrosion in all the exposure conditions but also reduces the corrosion even in saline 

environment. With the varying dose of Type-I inhibitor, the efficiency rate 

calculated on the basis of immersion test, has been found to be ranging between 

99.68-100%, 38.39-88.91%, 23.35-82.98% for mild steel plate, naked and slurry 

143


coated welded wire meshes respectively under tap water condition, whereas the 

same ranges between 97.78-99.68%, 24.96-81.25% and 21.63-78.21% respectively 

under saline water condition. For the Type-II inhibitor with the varying doses, the 

efficiency ranges between 33.38-78.62%, 26.61-83.76%, 29.76-79.58% for mild 

steel plate, naked and slurry coated welded wire meshes respectively under tap 

water condition, whereas the same ranges between 23.26-56.83%, 35.73-67.35% 

and 30.13-69.42% respectively under saline water condition. 

Type-I inhibitor delays the onset of corrosion which leads to the reduction of 

corrosion rate. In line with the efficiency results, the corrosion rate has been found 

to be lowest for Type-I inhibitor in tap water medium for all the three types of 

specimen [Figs. 2-4]. The rate of corrosion decreases with the increase of dose of 

inhibitor from 1 to 5 %. Even for the saline water medium with the addition of 

Type-I inhibitor, the corrosion rate has been as low as of 0.0061, 0.296 and 0.222 

mpy for MS plate, naked wire mesh and slurry coated wire mesh specimen 

respectively. For the Type-II inhibitor a reasonably good reduction in corrosion rate 

has been observed with the increase of the dose of the inhibitor. 

The efficiency of inhibitor Type-I has been found to be excellent for the mild 

steel plate under tap water as well as saline water conditions. In meshes the 

efficiency is good at 3% of the inhibitor dose and it increases as the dose is 

increased to 5%, which may improve further after the decay of the galvanising 

layer. Perhaps the galvanised surface of the wire mesh prevents the formation of 

protective film, which in turn reduces the efficiency. It is expected that for longer 

duration of immersion, after the removal of galvanised coating layer, the formation 

of strong passivating film will take place, which will protect any further decay of 

steel wire meshes. Though the inhibitor Type-II also shows quite acceptable 

efficiency at 3 and 5% of the inhibitor dose, but apparent reason behind it not being 

so effective at 1% dose, seems to be the weak film of iron tenate over the surface of 

the mild steel plate and welded steel wire mesh. 

The value of Icorr has been determined manually using the abscissa of the 

intersection point of the anodic and cathodic slopes of the polarisation curves. The 

efficiency of corrosion inhibitors of Type-I and Type-II using these curves is found 

to be 91.18% and 63.63% respectively, which confirms the results obtained from 

weight loss studies. 

The present investigation has been carried out using artificial saline water 

containing 4% NaCl water solution. However for actual simulation the real sea 

water runs are required and long time exposure studies are needed for monitoring 

the performance stability of the inhibition effect of these inhibitors. The authors are 

currently engaged in similar investigations and the results will be published in due 

course of time. 



CONCLUSIONS 



On the basis of this limited investigation it may be concluded that the 

efficiency exhibited by Type-I inhibitor for all the specimens is excellent for all of 

the exposure conditions. The Type-II inhibitor also exhibits reasonably good 

efficiency thereby a lower corrosion rate, however it is lower than the efficiency 

exhibited by Type-I inhibitor. Potentio-dynamic studies also indicate excellent 

efficiency for Type-I inhibitor and moderate efficiency for Type-II inhibitor. 

REFERENCES 

1. International Ferrocement Society, 2001, Ferrocement Model Code, IFS 

Committee 10-01, Asian Institute of Technology, Bangkok, Thailand. 

2. Vickridge, I.G., Ranjabar, M.M., 1998, “The Effect of Aggressive Environment 

on the Flexural Performance of Ferrocement”, Proceedings of 6 th International 

Symposium on Ferrocement, University of Michigan, Ann Arbor, USA., 1998, 

pp. 313-328. 

3. Dotto, J.M.R, de Abreu, A.G., Dal Molin, D.C.C. and Muller, I.L., 2004, 

“Influence of Silica Fume Addition on Concrete Physical Properties and on 

Corrosion Behaviour of Reinforcing Bars” Cement and Concrete Composites, 

Vol. 26, pp. 31-39. 

4. Vickridge, I.G., Nakassa, A.S. and Turner, H., 1998, “High Durability 

Ferrocement”, Proceedingsof 6 th International Symposium on Ferrocement, 

University of Michigan, Ann Arbor, USA., 1998, pp. 297-312. 

5. Ramesht, H., 1995, “Effect of Corrosion on Flexural Behaviour of Ferrocement 

in Corrosive Environment”, Journal of Ferrocement, Vol. 27(1), pp. 7-18. 

6. Masood, A., Arif, M., Akhtar, S. and Haquie, M., 2003, “Performance of 

Ferrocement Panels in Different Environments”, Cement and Concrete 

Research, Vol. 33(4), pp. 555-562. 

7. American Concrete Institute, 2000, State of Art Report on Ferrocement ACI 

549R-97, 2000, ACI Committee 549, American Concrete Institute. 

8. Giadis, J.M., 2004, “Chemistry of Corrosion Inhibitors”, Cement and Concrete 

Composites, Vol. 26, pp. 181-189. 

9. Berke, N.S. and Hicks, M.C., 2004, “Predicting Long Term Durability of Steel 

Reinforced Concrete with Calcium Nitrite Corrosion Inhibitor”, Cement and 

Concrete Composites, Vol. 26, pp. 191-198. 

145


10. Fontana, M.G., 1987, Corrosion Engineering, Mc-Graw-Hill book Company, 

New York 

Table- 1 Investigation Scheme for Immersion Test 

Exposure Medium Dose of 

Inhibitor 

(%) 

Designation 



Number of 

samples 

Tap/Normal Water (NW) 0 NW 3 

Tap Water with Inhibitor 

Type-I (NW-I) 

Tap Water with Inhibitor 

Type-II (NW-II) 

1 NW-I1 3 

3 NW-I3 3 

5 NW-I5 3 

1 NW-II1 3 

3 NW-II3 3 

5 NW-II5 3 

Saline Water (SW) 0 SW 3 

Saline Water with Inhibitor 

Type-I (SW-I) 

Saline Water with Inhibitor 

Type-II (SW-II) 

1 SW-I1 3 

3 SW-I3 3 

5 SW-I5 3 

1 SW-II1 3 

3 SW-II3 3 

5 SW-II5 3

Table- 2 Efficiency and Corrosion Rate 



Mild Steel Plate Welded Steel Wire Mesh 

Slurry Coated Welded Wire 

Mesh 

System 

Av. 

Weigh 

t loss 

(mg) 

Efficiency 

(%) 

Corrosio 

n Rate 

(mpy) 

Av. 

Weigh 

t loss 

(mg) 

Efficienc 

y (%) 

Corrosio 

n Rate 

(mpy) 

Av. 

Weight 

loss 

(mg) 

Efficiency 

(%) 

Corrosio 

n Rate 

(mpy) 

Tap 51.430 -- 3.143 11.90 -- 1.164 9.610 -- 0.940 

Tap + 

1% I 

Tap + 

3% I 

Tap + 

5% I 

Tap 

+1%II 

Tap + 

3% II 

Tap + 

5% II 

0.033 99.68 0.002 7.331 38.39 0.247 7.366 23.35 0.720 

0.166 99.93 0.010 2.524 78.79 0.129 2.861 70.23 0.280 

0.000 100.0 0.000 1.320 88.91 0.072 1.636 82.98 0.160 

39.466 33.38 2.412 8.733 26.61 0.855 6.750 29.76 0.661 

24.166 67.55 1.472 3.500 70.59 0.340 3.061 68.15 0.299 

22.200 78.62 1.357 1.933 83.76 0.189 1.962 79.58 0.192 

Saline 73.430 -- 4.487 16.130 -- 1.578 10.400 -- 1.018 

Saline 

+ 1 % I 

Saline 

+ 3 % I 

Saline 

+ 5 % I 

Saline 

+ 1 % 

II 

Saline 

+ 3 % 

II 

Saline 

+ 5 % 

II 

1.633 97.78 0.099 12.104 24.96 1.184 8.150 21.63 0.797 

0.200 99.73 0.012 4.868 69.82 0.476 3.558 65.79 0.348 

0.100 99.86 0.0061 3.024 81.25 0.296 2.266 78.21 0.222 

48.600 23.26 2.970 10.366 35.73 1.014 7.266 30.13 0.711 

23.830 53.01 1.456 6.500 59.70 0.636 4.507 56.66 0.441 

15.700 56.83 0.959 5.266 67.35 0.515 3.180 69.42 0.311 

147

Potential (mV) 


Table -3. Electrochemical Polarization Parameters for the Corrosion of 

Mild Steel in 4% NaCl in Presence of Calcium Nitrite and Tannic 

Acid 

Sl. 

No. 

System Ecorr (mv) Icorr (mA/cm 2 ) 

1. Saline Water (4% NaCl) –570 0.11 - 

2. 

3. 

Saline Water and Inhibitor Type- 

I (4% NaCl and 5% Calcium 

Nitrite) 

Saline Water and Inhibitor Type- 

II (4% NaCl and 5% Tannic 

Acid) 



Efficienc 

y (%) 

–526 0.002 98.18 

–582 0.04 63.63 

Log Current (mA) 

Inhibitor Type-I Inhibitor Type-II 

Blank 

Fig. 1 Potentio-dynamic Polarization Curves of Mild steel in NaCl 

Containing 5% Calcium Nitrite and Tannic Acid

Corrosion Rate (mpy) 


5 

4 

3 

2 

1 

0 

2 

1.5 

1 

0.5 

0 

0 1 2 3 

Inhibitor Dose 

4 5 6 

Fig.2 Corrosion Rate for Mild Steel Plate 



0 1 2 3 4 5 


Fig.3 Corrosion Rate for Welded Steel Wire Mesh 

NW-I 

NW-II 

SW-I 

SW-II 

NW 

SW 

149 

NW-I 

NW-II 

SW-I 

SW-II 

NW 

SW



1.2 

0.8 

0.4 

0 

0 1 2 3 4 5 


Fig.4 Corrosion Rate for Slurry Coated Welded Steel 

Wire Mesh 



NW-I 

NW-II 

SW-I 

SW-II 

NW 

SW

PHYSICAL, MECHANICAL, AND TRANSFER PROPERTIES OF 

MORTAR: EFFECT OF LIMESTONE FILLER ADDITION. 

H. Houari 1 , Y. Benachour 2 and 3 , F. Skoczylas 3 . 

1 Laboratoire de Matériaux et Durabilité des Constructions, Route Ain El bey, 

University Mentouri of Constantine, Algeria. hhouarilmdc@Yahoo.fr 

2 Laboratoire de Génie Géologique, University of Jijel, BP 98, Ouled Aissa, Algeria 

benachour.yacine@ec-lille.fr 

3 Laboratoire de Mécanique de Lille, CNRS UMR 8107, Ecole Centrale de Lille, BP 

48, 59651 Villeneuve d'Ascq Cedex, France frederic.skoczylas@ec-lille.fr 

ABSTRACT: 

The influence of limestone filler (0/100 µm) addition on mortar was studied. Its 

main goal is to utilize sands containing high rates of fillers available in large 

quantity in Algeria and to determine the maximum proportion to be introduced 

without altering the quality of concrete. The study consists in making mortars 

containing sands with a filler-to-sand ratio of 0, 15, 25, 35 and 45%, (dry weight 

basis). The variation was studied by maintaining a constant consistency. The 

physical, mechanical, rheological properties and durability were studied. The results 

obtained showed that mortars are very influenced by the filler addition and these 

results highlight the existence of an optimum rate of performance. The addition of a 

high percentage of fines (up to 35%) seems to little affect the mechanical and 

hydraulic characteristics of mortar. 

KEY WORDS 

filler, mortar, valorisation, properties, performance 

1- INTRODUCTION 

Industrial by-products such as fly-ash silica, fume and the furnace slag, are 

widely used in cement industry owing to their hydraulic and pozzolanic properties. 

Their use is generalized as additives in durable concretes and high performance 

concrete production (De Larard, 1988., Langan and al, 2002., Pala and al, 2007). 

The cost of these additives is relatively high compared to that of fillers, defined in 

the European Standard ENV 206 as materials with finely divided minerals that can 

be added to concrete in order to improve some of its properties. Their rates vary 

from one country to another (see table 1). 


PHYSICAL, MECHANICAL, AND TRANSFER PROPERTIES OF MORTAR: EFFECT 

Table. 1. Admitted crushed filler percentage according to various countries 

(With respect to sand mass %). 

Germany 

DIN 

4226 

1971 

Canada 

CSA 

A231 

1973 

< 4 % < 3,5 

% 

Italy 

Uni 

7 163 

1972 

Belgium 

NBN 

589-102 

1969 


H. Houari, et al 

aim of this study is to identify the maximum proportions which one can introduce 

without altering the concrete quality, and to highlight possible beneficial effects. 

2. EXPERIMENTAL PROCESS 

2.1 Materials 

Additions of type 1 were used in this study (EN 206-1 2004). They are 0/100 

µm limestone fillers of Bolted, Rinxent careers, France. Their main characteristics 

provided by the producer are given in table 2. In figure 1, the results of a laser 

granulometric analysis are presented. The cement used (type CEM I-52.5), confirms 

to the EN standard 197-1. 

Table. 2. Physicochemical composition of filler 

CaCO3 

(%) 

Blaine 

fineness 

value 

(m 2 /kg) 

Methylene 

blue value 

MBF 

(g/l) 

Water 

content Wf 

(%) 

Fe 

(%) 

MgCl 

Mean 98.50 375.97 0.58 0.00 0.00 0.35 0.01 

Standard 

deviation 

0.36 30.45 0.10 0.00 0.00 0.17 - 

Cumulative passing (%) 

100 

90 

80 

70 

60 

50 

40 

30 

20 

10 

0 

0.1 1 10 100 

Particule diameter (µm) 

CEM I cement Filler 

Fig. 1. Grain size distribution for CEM 1 cement and filler powder obtained by 

laser granulometry 

(%) 

153 

Cl 

(%)


2.2 Method 

The principle is to compare various mortars with different substitution rates 

of filler 0, 15, 25, 35 and 45 % (dry weigh basis). For each series the characteristics 

of fresh and hardened mortar were measured. In table 3 are synthesized the 

proportion adapted in this study. 

2.3 Preparation of samples and test protocol 

The mixtures were cast in the same day, to make beams of (50x15x15 cm 3 ) 

and prisms of (4x4x16 cm 3 ). They were protected by plastic sheet in order to avoid 

any desiccation for 24 hours. Then the beams and the prisms were immersed in 

water at 20 °C ± 1 during 28 days. Samples were cored cut and rectified to obtain 

the parallelism of surfaces. Mechanical and hydraulic tests were carried out on 

cylinders of 37 mm diameter and 74 cm height. A three point bending test was 

carried out on the prisms and compressive test was undertaken on cubs of 4x4x4 

cm 3 . Porosity and density were measured on discs of 37 mm diameter and a height 

ranging between 10 and 60 mm. 

2.3.1 Mechanical tests 

The bending test was undertaken using a three point bending device 

(EN 196-1). Measurements of compressive strength were carried out using the 

device (EN 196-1). The displacement speed was fixed at 0.12 mm/mn. For Young’s 

modulus, the longitudinal deformations were measured on cylindrical specimens 

using a LVDT device (figure 2.) designed by our laboratory (Agostini and al., 2006). 

An average value of 4 LVDT sensors fixed between two rings at a distance of 30 

mm was measured. Preliminary results showed that the values obtained with LVDT 

sensors are comparable to those measured by gauges 





Table. 3. Composition of Mortars added with filler (expressed per m 3 of mortar) 

Filler Addition Normalised sand CEM I 52,5 Cement Water B –type Maniabilimeter 

EN 196-1 EN 197-1 NF P 18-452 

(%) (kg/m 3 ) (kg/m 3 ) (kg/m 3 ) (l/m 3 ) (sec) 

0 0 

1350 450 216 

15 202.50 

1147,50 450 210 

25 337.50 

1012 ,48 450 231 

25 ± 1 

35 472.50 

877,50 450 275 

45 607.50 

742,5O 450 302 

2.3.1 Permeability tests 

Cylindrical samples (Ø=37 mm, h=74mm) were first oven dried at 65 °C to 

evacuate water present into the pores. Measurements of permeability were carried 

out using a hydrostatic cell. The sample, protected by a jacket is placed on the lower 

base plate of the hydrostatic cell. The test sample was subjected to a confining 

pressure of 4.2 MPa using hydraulic oil (HF95Y Enerpac). The gas used is pure 

argon (inert gas), with a viscosity of 2.2 10 -5 Pa.s at 20 °C. It is injected at pressure 

Pi = 1.8 MPa through the lower face of the sample. The upper face is drained at 

atmospheric pressure until a steady state is achieved due to pressure gradient 

(Loosveldt H et al 2002). 

Fig. 2. Uniaxial compressive test: Axial Fig.3. Sketch of a triaxial cell displaying the 

displacement is measured using four LVDT associated permeability measurement set up . 

sensors placed between two rings 

155


3. RESULTS AND DISCUSSIONS 

3.1 Variation of density and porosity 

They are measured on 4 discs of 37 mm diameter and a height ranging 

between 20 and 60 mm The pore volume was determined by difference between the 

saturated mass and dry mass at 65 °C (until constant mass). Volume is simply 

calculated starting from dimensions of the samples. Figures 4 and 5 respectively 

present the variation of density and porosity according to the variation of filler rate. 

It is noted that the density and porosity are very influenced by the fine addition. 

Beyond 15% filler, we observe a progressive fall of the density by stage from 2.7 to 

2.9 % and an increase in porosity until reaching twice the initial value. Fine 

limestone manages to fill part of porosity while releasing water. The filler excess 

supports the formation of a volume of paste (water + cement +filler) more important 

compared to the solid matter constituents of sand (see figure 6) since the filler 

addition is made by substitute of sand. Water cement ratio W/C varies little for 

additions up to 25% (figure. 6). 

Density ( kg/m 3 ) 

2200 

2100 

2000 

1900 

0 5 10 15 20 25 30 35 40 45 50 

Filler mass percentage (%) 



Porosity (%) 

12 

10 

8 

6 

4 

2 

0 

0 10 20 30 40 50 


Fig. 4. Evolution of density as a function Fig. 5. Evolution of porosity as a function 

of filler rate (%: dry weigh basis) of filler rate (%: dry weigh basis) 

Ratio 

0,8 

0,7 

0,6 

0,5 

0,4 

0,3 

0,2 

0,1 

0 

W/C W/C+F 

0 15 25 35 45 

Filler percentage (%) 

Sand density (Kg/m 3 ) 

1400 

1200 

1000 

800 

600 

400 

400 600 800 1000 1200 1400 

Ciment paste density (Kg/m 3 C+F C+F+W 

) 

C= Cement, F=Filler, W= Water 

Fig. 6. Evolution of W/C and W/(C+F) ratios Fig. 7. Evolution of mass proportion betwee as a 

function of filler rate (%: dry weigh basis) cement paste and sand as a function of filler rate (%: dry 

weigh basis).

3.2 Variation of strength and modulus of elasticity. 



The compressive strength was measured at 28 days on the broken parts of 

three prisms initially subjected to the 3 point bending test. The ends of the prisms 

were intact after failure under bending (ASTM 116-90 and EN 196-1). To determine 

Young’s modulus, elastic strains were measured on three cylindrical test-tubes. The 

results at 28 days according to the filler rate are represented in figure 8. We can 

observe that the modulus decreases with increase in the filler rate and thus in 

porosity. The mortars made with 15% of filler addition improve of the modulus by 

12.16 % compared to that of the reference mortar. Beyond that, a progressive 

reduction is observed varying from 8.57 to 16.3 %. Compressive strength values 

(figures 9 and 10) are higher than those of the mortar of reference for filler rates up 

to 35%. For the optimum value, an increase of 18 % (14 MPa) and 23 % (1.76 MPa) 

was obtained in compressive and tensile strength respectively. Strength of mortars 

decreases little whereas the increase in porosity is important. Thus there is no direct 

relation between porosity and strength. Improvement of strength independently of 

the filler effect is due the heterogeneous nucleation which produces an excess of 

formed hydrates. That is due to the fact that the reaction between water and cement 

is faster in the presence of calcareous powder (Nonat, 1994., Baron 1996). It was 

shown that the effect of the smoothness is beneficial for specific surfaces lower than 

500 m 2 /kg (Lawrence and al, 2005). Improvement of strength can be explained by 

the formation of aluminates and carboaluminates due to the chemical reaction 

between calcium carbonate and tricalcium aluminates (Soroka and Stern, 1976). 

3.3 Variation of permeability 

As previously mentioned, the permeability was measured on oven dried 

samples. The results obtained (Figure 11) were calculated from the average of three 

tests. It was observed that the permeability varies little (8.79 10 -18 to 2.51 10 -17 m 2 ) 

independently of the variation of the porosity observed which is much more 

important. This indicates that the volume of the hydrates formed by the binder is an 

indicator of the degree of interconnection of the porous network. The less permeable 

mortars were observed for 15 and 25 % filler rates. 

157


Young's Modulus (GPa) 

50 

40 

30 

20 

10 

0 

0 5 10 15 20 25 30 35 40 45 50 

Filler mass percentage 



Ultimate compressive stress 

σucs (MPa) 

90 

80 

70 

60 

50 

40 

30 

20 

10 

0 

0 5 10 15 20 25 30 35 40 45 50 


Fig. 10. Evolution of Young’s modulus as Fig. 11. Evolution of ultimate compressive stress σucs 

a function of filler rate (%: dry weigh basis). as a function of filler rate (%: dry weigh basis). 

Ultimate Flexural stress σufs 

(MPa) 

10 

8 

6 

4 

2 

0 

0 5 10 15 20 25 30 35 40 45 50 


Apparent Permeabiltiy Kapp 

(10 -17 m 2 ) 

90 

80 

70 

60 

50 

40 

30 

20 

10 

0 

0 5 10 15 20 25 30 35 40 45 50 


Fig. 11. Evolution of ultimate flexural stress σufs Fig. 13. Evolution of apparent permeability 

Kapp as a function of filler rate (%: dry weigh basis) as a function filler rate (%: dry weigh 

basis) 

The manifestation of the physicochemical contribution of the addition on the 

structuring of the cementing matrix is probably the cause of this phenomenon. That 

depends on the way in which the hydrates are assembled, of their arrangement in 

space and their connection (Baron 1996). Remove (Uchikawa and Hanehara, 1996) 

showed that the mineral powder addition in the concrete reduces the size of the 

products of hydrate, prevent the Ca (OH) 2 deposit under the terms of their filler role 

and consequently decrease the size of the pores. This results in obtaining not very 

permeable mortars for limestone additions up to 45%.

4. CONCLUSION 



The mortars compound by introducing various limestone filler rates (up to 

45%) by substitution with sand were compared in order to use sands containing high 

filler rates. The rheological, physical, mechanical properties and durability were 

studied for mortars having the same consistence. The results show that the optimal 

content of fillers of 15% (being equivalent to 45% compared to the weight of 

introduced cement) led to improved performances (increase in strength of 18 % 

compared to the mortar of reference). However, strength remained higher compared 

to the mortar of reference for additions up to 35% (equivalent to 105% compared to 

the weight of introduced cement), in spite of the considerable increase in porosity 

with high filler addition. This can be explained by the physical effect of 

heterogeneous nucleation and the formation of aluminates and carboaluminates. On 

the other hand the modulus of elasticity and the density decrease considerably with 

the increase in porosity. The permeability which is an indicator of durability varies 

little. The physicochemical contribution of the addition on the structuring of the 

cementing matrix is probably responsible of the phenomena observed. This depends 

on the way, in which the hydrates are assembled, of their arrangement in space and 

their connection. The filler addition reduces the size of the hydration products 

prevents the Ca (OH) 2 deposit leading to reduction in pore size. However a 

physicochemical and microstructural study would certainly bring clearer answers to 

these phenomena. It will be also interesting to study the shrinkage and freezing and 

thawing resistance of mortars containing high filler contents. Given the obtained 

results we can suggest the use of sands filler rates up to 35% in mass of calcareous 

filler powder are already possible without high mechanical and durability 

performance losses, which is economically important, particularly in Algeria, but 

also for major European countries. 

5- REFERENCES 

Agostini, F., Lafhaj, Z., Skoczylas, F. and Loodsveldt, H., 2007, “Experimental 

study of accelerated leaching on hollow cylinders of mortar”, Cem Concr Res, 

Vol 37, pp 71- 78. 

Baron, J., 1996, “Les additions normalisées pour le Béton”, in : Les bétons - Bases 

et données pour leur formulation, Association technique - industrie des liants 

hydrauliques, Eyrolles Ed., Paris, pp 47-57. 

159


Benachour, Y., 1992, “Analyse de l’influence du sable de mer et du sable de carrière 

sur les caractéristiques du béton ”, Magister thesis, University Mentouri of 

Constantine, Algeria. 

Benachour, Y., 1996, “Contribution a l’étude de l’influence des fillers sur les 

caractéristiques du béton”, Proceedings of the National Seminar « Qualité des 

bétons de construction » organised by PUBLITECH, Alger, Algeria, pp 89-99. 

Bonavetti, V., Donza, H., Menendez, G., Cabrera, O. and Irassar, E.F., 2003, 

“Limestone filler cement in low w/c concrete: a rational use of energy”, Cem 

Concr Res, Vol 33, pp 865–871 

Celik T. and Mamer T.K., 1996, “Effects of crushed stone dust on some properties 

for concrete”, Cem Concr Res Vol 26, pp 1121-1130. 

De Larrard, F., 1988, “ Formulation des bétons et propriétés des bétons à très hautes 

performances ”, LCPC internal research report, Vol n° 149. 

Dreux, G., 1969, “Contribution à l’étude de l’influence de la finesse des sables sur 

diverses qualités des bétons”, Annales de l’ITPB, Vol n° 261. 

Kara, A.R., 2001, “Influence des additions minérales sur le besoin en eau et les 

résistances mécaniques des mélanges cimentaires”, PhD thesis, University of 

Cergy Pontoise, France. 

Koutchoukali, N., Adjali S. and Benachour, Y., 1992 “Reconnaissance des sables de 

concassage de la région de Constantine”, Proceedings of the National Seminar 

organised by ERCEst and CTC Est, Bejaia, Algeria, pp 99-110. 

Lawrence, P., Cyr, M. and Ringot, E., 2005, “Mineral admixtures in mortars: effect 

of type, amount and fineness of fine constituents on compressive strength”, Cem 

Concr Res, Vol 35, pp 1092-1105. 

Langan, B.W., Weng K. and Ward M.A., 2002 “Effect of silica fume and fly ash on 

heat of hydration of Portland cement”, Cem Concr Res, Vol 32, pp 1045-1051 

Le Roux, A. and Unikowski, Z, 1980, “Mise en évidence des fines argileuses dans 

les granulats à béton”, Bull de liaison du LCPC, Vol n° 110. 

Loosveldt, H, Lafhaj, Z., Skoczylas, F, 2002, “Experimental study of gas and liquid 

permeability of a mortar”, Cem Concr Res, Vol 32, pp 1357-1363. 

Moosberg-Bustnes, H., Lagerblad, B. and Fotsberg, E., 2004, “The function of 

fillers in concrete”, Materials and Structures, Vol 37, pp 74-81. 





Nonat, A., 1994, “Interactions between chemical evolution (hydration) and physical 

evolution (setting) in the case of tricalcium silicate”, Materials and Structures, 

Vol 27, pp 187-195. 

Pala, M., Özbay, M., Özta, A. and Yuce, I.M., 2007, “Appraisal of long-term effects 

of fly ash and silica fume on compressive strength of concrete by neural 

networks”, Construction and Building Materials, Vol 21, pp 384-394. 

Soroka, I. and Stern, N., 1976, “Calcareous fillers and compressive strength of 

Portland cement”, Cem Concr Res, Vol 6, pp 367-376. 

Topçu, I.B. and Ugulru, A., 2003, “Effect of the use of mineral fillers in the 

properties of concrete”, Cem Concr Res, Vol 33, pp 1071-1075. 

Trang, N.L., 1980, “L’essai au bleu de méthylène, un progrès dans la mesure et le 

contrôle de la propreté des granulats ”, Bull de liaison du LCPC, Vol n° 107. 

Uchikawa, H., Hanehara, S. and Hirao. H., 1996, “Influence of microstructure on the 

physical properties of concrete prepared by substituting mineral powder for part 

of fine aggregate”, Cem Concr Res, Vol 26 pp 101-111. 

161




CONTRIBUTION TO THE STUDY OF THE BEHAVIOUR 

OF THE STEEL BEAMS WITH WEB OPENINGS 

Benyagoub DJEBLI 1 

, Djamel ELDdine KERDAL 2 

1: Département de Génie Civil, Faculté des sciences et de la technologie 

Université Ibn khaldoun, Tiaret -14000- ALGERIA, djebli_mb@yahoo.fr 

2: Département de Génie Civil, Faculté d’architecture et de génie civil 

Université des sciences et de la technologie Mohamed Boudiaf, Oran -31000- 

ALGERIA 

ABSTRACT 

Steel beams with web opening are used in high buildings to allow the passage 

of the technical equipments through these openings, therefore making it possible to 

decrease the height of the floors and thus to reduce the weight of the superstructure. 

However, the openings in the beam webs have an influence upon their resistance. 

The work presented herein is devoted to the study of the behaviour in flexure 

of these beams in the elastic range and the plastic range, by using the software 

CAST3M which is a program based on the finite element method. 

The element chosen to conduct this study is the iso-parametric eight-noded 

quadratic elements (quadratic element with 8 nodes). Simply supported steel beams 

with web openings of circular and rectangular shape were studied. 

A comprehensive parametric study was carried out to investigate and 

compare the behaviour of steel beams with web openings of various shapes 

(rectangular, circular), sizes and number. The results obtained show a good 

agreement with the results found in the literature. 

KEY WORDS 

Steel beams, openings, height, resistance, deflection. 

1. INTRODUCTION 

Modern multi-storey buildings have always stringent requirements on 

headroom. 

In order to accommodate building services within the constructional depth of 

a floor, it is common practice to provide web openings in structural floor beams for 

passage of services. But the presence of web openings in steel beams reduces both 

the global and the local resistance of these beams. Indeed, it is considered that the 


CONTRIBUTION TO THE STUDY OF THE BEHAVIOUR OF THE STEEL BEAMS WITH WEB OPENINGS 

collapse of this type of beams can be reached if one exceeds one of following 

resistances [1, 2]: 

- Local resistance of the web posts between openings and the tee-sections 

above and below the web opening. 

- Global resistance to the bending (flexure), shear and the lateral buckling. 

The modes of failure [1, 3] met are generally the following: 

-The “vierendeel” mechanism, as shown in Fig. 1, due to the formation of 

four plastic hinges in the tee-sections above and below the web opening caused by 

the transfer the of lateral shear forces across a web opening (vierendeel action). 

-The buckling of the web posts between openings, as shown in Fig. 2. 

Fig.1. Vierendeel mechanism around 

the circular web opening. 

In general, both the shear and the moment capacities of the perforated 

sections may be readily assessed. However, the moment capacities of the teesections 

above and below the web openings under local moments are relatively 

difficult to evaluate in the presence of co-existing axial and shear forces due to 

global bending action. 

Moreover, it is necessary to use plastic design to incorporate the formation of 

four plastic hinges in the tee-sections for an improved prediction of the load carrying 

capacity of the beams. 

For beams with multiple web openings, buckling of web posts may be critical 

when the openings are closely spaced. Moreover, additional deflection due to the 

presence of web openings should also be considered. Due to the complexity of the 

structural problems, it should be noted that Eurocode3: Part 1.1 Annex N [4] has 

been drafted to present design principles and application rules for steel beams with 

web openings. 

There are a number of design recommendations [5 to 12] available in the 

literature for steel beams with web openings. Deformation and stress analyses are 

presented for both, elastic range [13 to 16] and plastic range [17 to 20]. 

Deflection calculations are also presented in a number of design recommendations 

[21, 22] 



Fig.2. Buckling of the web posts 

between openings.

2. DESIGN CRITERIA 


Benyagoub DJEBLI, Djamel ELDdine KERDAL 

The design criteria presented in this section are based on both theoretical 

considerations and experimental observations. Many of the criteria were developed 

for non-composite beams [23] and extended as appropriate to composite beams [24]. 

The criteria helped insure that the limit states can be obtained upon which the design 

formulas are based. 

In this paragraph, only the criteria developed for non-composite beams with 

web openings and without reinforcement are presented as shown in Fig. 3. 

- Web buckling 

Based on experimental test results, Redwood and Uenoya [25] have defined 

the opening parameter Po that should be lower than the value of 5.6 to prevent web 

buckling. 

a0 d0 

P0 = + 6. 

(1) 

d0 

d 

- Opening size 

The opening length and height should not exceed 2d and 0.7d respectively. 

- Tee proportions 

The limitations on the depths of the top and bottom tees are based on the need 

to transfer some load over the opening. However on the basis of the lack of test data 

for shallower tees the following limitations must be respected: 

a 

S 

t 

do 

ao 

0 St 

≤12 and >0.15 (2) 

d 

- Concentrated loads 

The limitations on the locations of concentrated loads near openings to 

prevent web crippling are based on an extension by Darwin [16] of the criteria 

presented by Redwood and Shrivastava [17]: 

St 

Fig. 3. Geometric notation. 

d 

tw 

b 

b 

tf 

165


d - 2t 

t 

w 

d - 2t 

t 

w 

1100 

f 

f ≤ ; 

f 

y 

1365 

f ≤ ; 

y 

b 140 d 

≤ ⇒ (3a) 

tf 

f 2 

y 

b 170 

≤ ⇒ d (3b) 

tf 

f 



y 

- Placement (position) of opening 

The condition that openings should be placed no closer than a distance d (the 

beam depth) of a support limits the horizontal shear stresses that must be transferred 

by the web between the opening and the support. 

- Spacing between openings 

The distance between adjacent openings should not be less than beam depth. 

3. SCOPE OF THE STUDY PRESENTED 

The investigation carried out in the study presented may be divided into two 

parts: 

In the first part, we compare the moment (Mo,Rd ) and shear (Vo,Rd ) capacities 

of perforated 

sections obtained from the finite element modelling and from the formula 

given in the annex N of Eurocode3, which will therefore presented first. 

Then in the second part a parametric study is performed using a finite 

element model for steel beams with different opening shapes (rectangular and 

circular) and sizes. Both material and geometrical linearity or material and 

geometrical non-linearity are considered. 

Based on the results of the finite element models, the behaviour of 

different steel beams with rectangular and circular shapes and sizes is presented. 

The proposed design criteria given in section (2) are respected in order to predict 

this behaviour. Only the case of unreinforced openings is considered. 

In the two parts; an iso-parametric eight-noded quadratic element is used 

to discretize the beam.



4. CALIBRATION AGAINST RESULTS OBTAINED IN ACCORDANCE 

WITH ANNEX. N OF THE EUROCODE3 

In this study, an IPE300 of class 1 is used. Two web openings situated at the 

mid-height of the section with a diameter equal to (2/3)d, where d is the depth of the 

section. These openings are placed symmetrically about the middle of the span of 

the beam. No reinforcement is considered. The formulation is presented in 

accordance with Annex N of the Eurocode3 for easy reference. It should be noted 

that both the moment resistance, Mo,Rd, and the plastic shear resistance, Vo,Rd, of the 

perforated section are evaluated at the centre of the web openings. 

4.1. Formulation in accordance with Eurocode3: 

In order to provide simple rules for practical design of steel beams with 

circular web openings, approximate design expressions of ultimate resistances are 

proposed by Eurocode3 as follows: 

The geometric terms used are given in fig. 4. 

4.1.1. Ultimate shear resistance: 

d0 

For circular openings the shear resistance Vo,Rd shall be taken as follows: 

Vo,Rd = min (Vo, pl, Rd ; Vo, ba, Rd) (4) 

0,9.d . 0 t 

Vo, pl, Rd = (V pl, Rd − 

γ . 

M0 

w 

. f 

3 

yw 

) 

η 

1+ 

η 

(5) 

d 0 

Vo, ba, Rd = V ba, Rd (1− ) 

(6) 

d 

d1 

d2 

Fig. 4. Notation for « circular opening » 

e 

tw 

d 

167


( ) 

( ) ⎥ ⎥ 

2⎤ 

0 

⎡ 

2 

d - 0,9.d + 

Where: η = 3,7 ⎢ 

0 4.e 

⎢ d 0 d − 0, 

9. 

d 

⎣ 

0 ⎦ 

Vpl,Rd; Vo,pl,Rd : plastic shear resistances of unperforated and perforated crosssection 

respectively 

Vba,Rd; Vo,ba,Rd : buckling shear resistances of unperforated and perforated crosssection 

respectively 

fyw : yield strength of the web 

γMO = 1.0 for steel 

4.1.2. Ultimate bending moment resistance + shear: 

The checking of the openings is done according to the classes' of the sections 

proposed by Eurocode3. 

• For class 1 and 2 sections: 

⎡ 0,225.t . 

Mo,c,Rd = w d 0 

Mpl,Rd ⎢1 

- 

⎢⎣ 

w 

⎡ 0,25.t . ( + ) ⎤ 

w d 0 d 0 4e0 

⎢1 

- 

⎥ 

⎢⎣ 

w pl, y ⎥⎦ 

Where : 



2 

( 0,9.d + ) 

pl, y 

0 

4e 

0 

− 

V 

V 

⎤ 

⎥ 

⎥⎦ 

χ 1 

o, sd 

o, Rd 

≤ Mpl,Rd 

(7) 

(8) 

⎡ ⎛ d 0 ⎞ d 0⎤ 

. 2 0,25.t d 1 1, 

35 0, 

7 0, 

9 . 

w ⎢ + ⎜ − ⎟ ⎥ 

⎣ ⎝ d ⎠ d ⎦ 

χ = 

(9) 

1 

w 

• For class 3sections: 

Mo,c,Rd = Mel,Rd 

⎡ 

. ⎢ 

⎢ 

⎣ 

t 

- 

⎡ 3 ( ) ⎤ 

. ⎢ 

t w. 

d 0 + 2.e0 

1- 

⎥ 

⎢ 12.I y ⎥ 

⎣ 

⎦ 

pl, y 

( 0,9.d + ) 

V 

⎤ 

o, sd 

− χ . ⎥. 

χ 

1 

Vo, 

Rd ⎥ 

⎦ 

3 

w. 

0 2.e0 

1 ≤ Mel,Rd 

2 

12.I y 

(10)

χ 1 

if 

2 = 

d 

= 1, 

126 − 0, 

03. 

t 

0 yw 

χ . 

if d0 > 

2 

w E 

4.2. Comparison of resistances values: 


f 

d 

0 


E 

≤ 4, 

2. 

t w. 

(11a) 

f 

yw 

4, 

2. 

t 

. 

E 

w (11b) 

f yw 

The values of resistances Mo,c,Rd and Vo,Rd were obtained according to the 

determination of the normal and shearing stresses respectively at the opening. 

• Details of the beam studied are shown in the figure 5: 

Fig. 5. Beam studied 

• The distribution of stresses obtained from the finite element investigation are 

presented in Fig.6: 

• 

a- Normal stress (σxx) b- Shearing stress (τxy) 

Fig. 6. Distribution of the stresses 

• The values of stresses and the deduction of the internal efforts at the opening 

obtained from the program are presented in the table.1. 

169


Table1. Stresses and internal efforts results 

Stress (KN/m 2 ) Internal efforts 

Normal 

2.02 

10 5 

Shearing Shear (KN) 

Teesections 

above 

the 

web 

opening 

Teesections 

below 

the 

web 

opening 

Bending 

Moment 

(KN.m) 

V1 

Teesections 

above 

the web 

opening 



V2 

Teesections 

below 

the web 

opening 

3.33 10 4 2.75 10 4 118.11 65.83 54.50 

120.33 

The comparison between the results can be made as follows: 

Plastic shear resistance: Moment resistance: 

According to Eurocode3: Vo,Rd = According to Eurocode3: Mo,c,Rd = 

120.4 KN 120KN.m 

According to Cast3m: Vo,pl = According to Cast3m: Mo,pl = 

120.33 KN 118.11KN.m 

According to the results obtained, it is noticed that the resistances values have 

almost the same order of magnitude in both methods. 

5. PARAMETRIC STUDY WITH FINITE ELEMENT MODELS 

In order to provide information for practical design of steel beams with web 

openings, a parametric study was performed to assess the structural performance of 

simply supported steel beams. 

Firstly, two studies were made in the elastic range. The first one to make a 

comparison between both deflections at mid-span of the beam and under the opening 

respectively and the second one, in order to understand the effect of shapes 

(rectangular or circular) of web openings at different locations along the beams; we 

studied the normal stress and the variation of the deflection at the corner of the 

opening and at mid-span of beam. 

Secondly, we are going to study the beams behaviour with rectangular 

openings of different sizes under the effect of great displacements and the behaviour 

of a rectangular section with an opening at the mid-span under the effect of an 

imposed rotation. 

The geometric conditions for both elastic and plastic ranges are respected.

5.1. Parametric study in elastic range 



• In the first study, the variation of the deflection of three beams with two 

openings placed symmetrically about the centre of the beam under the 

effect of different types of loading are presented in Fig. 7. 

7. 50 

6. 00 

4. 50 

3. 00 

1. 50 

L/6 q 

L/6 

Force (KN) 

7. 74 

6. 19 

4. 64 

3. 09 

1. 55 

0. 00 

L 

Beam (b) 

L/6 

Force (KN) 

L/2 L/2 

Beam (a) 

L/6 

Fig. 7. Variation of the deflection 

P 

0.0 0.1 0.2 0.3 0.4 0.5 0.6 

7. 37 

5. 90 

4. 42 

2. 95 

L/6 L/6 L/3 L/6 L/6 

Force (KN) 

Beam (c) 

(δc) 

(δo) 

1. 47 

Displacement (mm) 

0. 00 

0.0 0.1 0.2 0.3 0.4 0.5 0.6 

171


According to the Fig. 7, it is noticed that the deflection at mid-span of the 

beam is larger than the deflection under the opening. 

• In the second study, the variation of the stresses in 5 beams with a number 

of openings varying from 1 to 5 under the effect of two types of loading α 

(two symmetrical concentrated loadings) and β (uniformly distributed 

load) are presented in Fig.8. 

α : Cases of two symmetrical concentrated loadings of 7.5 KN each 

one 

β : Cases of uniformly distributed load of 3.334 KN each on 

Fig. 8. Concentration of stress in the beams 

The higher stresses are found at the corners (levels A and B) of the openings 

nearest to the supports (Fig. 9.). 

A 

B 

Geometric 

al 

Fig. 9. Concentration of stresses at corners A and B 



h 0 ≤ 

A 

B



The values of stresses at the corners A and B of the opening and at mid-span 

of the beams obtained from the program are presented in the table2. 

Table2. Values of stress 

Beams 

Stress (Mpa) 

rectangular opening circular opening 

types level A level B Mid length level A level B Mid length 

α β α β α β α β α β α β 

a-1 +173 +113 -159 -104 +13.6 +14.1 +38 +25 -35 -24 +13.6 +13.5 

a-2 +174 +117 -160 -108 +13.8 +14.4 +35 +23 -32 -22 +10.6 +10.9 

a-3 +173 +111 -159 -103 +14.2 +14.1 +33 +21 -30 -19 +11.1 +10.9 

a-4 +174 +116 -160 -107 +13.7 +14 +33 +23 -32 -23 +13.7 +13.5 

a-5 +174 +114 -160 -105 +14.4 +14.4 +30 +20 -27 -18 +09.8 +09.8 

It is noticed that the values of the stresses in the beams with rectangular 

openings are larger than those in the beams with circular openings. The stresses at 

corners A and B are greater than the stresses at middle span of the beam and the 

maximum stress in the beam is located at corner A. 

For a better analysis on resistance, the variation of the stress of the type of 

beam “a-2” (with two openings) under the effect of the type of loading (α) is 

presented in Fig. 10. 

+250 

+200 

+150 

+100 

+50 

0.0 

-50 

-100 

-150 

-200 

σx (Mpa) σx (Mpa) 

σx (Mpa) 

Force 

(KN) 

0 1 2 3 4 5 6 7 8 9 

Level A 

Level B 

Mid-length 

of the beam 

+25 

+20 

+15 

+10 

+5 

0.0 

-5 

-10 

Fig. 10. Variation of stresses 

173 

Force 

(KN) 

0 1 2 3 4 5 6 7 8 9


According to Fig. 9, one notes that the stresses at the mid-span of the beam 

have the same order of magnitude. On the other hand, the stresses on levels A and B 

in the case of the rectangular openings are larger than for the case where they are of 

circular shape. That is caused by the additional bending moment which is caused by 

the vertical displacement between the corners of the opening (the vierendeel effect). 

In order to show the importance of the beams deformations with web 

openings, the deflection of perforated beams of the type of beam “a-2” (with two 

openings) under the effect of the type of loading (α) is presented in Fig. 11. 

The presence of the openings has been found to cause additional deformations 

in the beam. Such effect is more significant in steel beams with rectangular web 

openings than in beams with circular openings. 

5.2. Parametric study in plastic range 

Fig. 11. Deflection of perforated beams 

The effect of the presence of the opening on the behaviour of the steel beams 

is analysed by considering two cases. The first one is on steel I-beam with one 

rectangular opening of various sizes positioned at both the mid-height and mid-span 

of the beam under the effect of great displacements and the second one is on steel 

cross-section of rectangular shape perforated in the middle. 

The force-displacement curves concerning the first study and moment-rotation 

evolution concerning the second study are presented in Fig. 12 (a and b 

respectively). 



14 

12 

10 

8 10 5 

6 10 5 

4 10 5 

Force (N) 

a0 = 2 d 

unperforated 

beam section 

h0 = 0.5 d 

h0 = 07d 



According to the Fig. 12, it should be noted that despite the variation in 

the sizes of the web openings, all the curves (force-displacement) show in 

general that an increase in the opening depth, always reduces the ultimate load 

which causes the maximum deflection at mid-span of the beam. However, with 

an increase in the opening length, the maximum deflection at mid-span of the 

beam is reached under the effect of the same ultimate load. 

For the case of the behaviour of a rectangular cross-section with a rectangular 

opening at mid-span under the effect of an imposed rotation, one illustrates the 

curves (moment-rotation) of three sections types (one full and the two others with 

different opening depths) in Fig. 13. 

14 

12 

10 

8 10 5 

6 10 5 

4 10 5 

Force (N) 

h0 = 0.7 d 

unperforated 

beam section 

a0 = 1 d 

a0 = 2 d 

2 10 5 

2 10 

00 

5 

Displacement (mm) 

00 

Displacement 

00 5 10 15 20 25 30 

a- opening depth effect 

00 5 10 15 20 25 30 

b- opening length 

Fig.12. Force-displacement curves obtained from finite element 

X.10 6 

2. 00 

1. 80 

1. 60 

1. 40 

1. 20 

1. 00 

0. 80 

0. 60 

0. 40 

Moment (N.m) 

Unperforated 

beam section 

h0=0.5d 

h0=0.7d 

0. 20 

Rotation 

0. 00 

0.00 0.02 0.04 0.06 0.08 0.10 

Fig.13. Moment-Rotation curves obtained from finite 

element investigation. 

175


According to the Fig. 13, an increase in the opening depth, always reduces the 

resistant moment, therefore, it is to be confirmed that the size of the opening has an 

influence on the resistance. 

6. CONCLUSIONS 

Numerical investigations of the behaviour of steel beams with web openings 

of various shapes (rectangular and circular) and sizes in both elastic and plastic 

ranges are presented in detail. It should be noted that: 

1. The deflection at mid-span of the beam is larger than the deflection relating to the 

opening. 

2. Beams of same sections with circular openings are more resistant than the beams 

with rectangular openings. 

3. Due to the “Vierendeel effect” the maximum stresses in the beam occurs at the 

opening 

4. The deflection increases following an increase in the opening depth. 

5. The deflection increases following an increase in the opening length. 

6. The ultimate load of perforated sections decreases following an increase in the 

opening depth, do, and thus, deformation “failures” of perforated sections are 

primarily controlled by the magnitude of do. However, with an increase in the 

opening length, ao, the ultimate load depends only on part of the sections above 

and below the hole, and thus promotes the “Vierendeel” mechanism in perforated 

sections. Consequently, for web openings with same values of do but with 

different values of ao, the load capacities of the perforated sections are inversely 

proportional to the values of ao. 

7. The size of web opening has an effect on the resistance of the beam section. 

8. The opening depth has more influence on the resistance than the opening length. 

REFERENCES 

1. Redwood, RG., 1968, “Plastic behaviour and design of beams with web 

openings”, Proceeding, 1 st Canadian Structural Engineering. Conference, 

Canadian Steel Industry Construction Council, p. 127-138. 

2. Redwood, RG. , 1969, “The strength of steel beams with unreinforced web 

holes”, Civil Engng Public Works Rev, London,64(755):559–62. 

3. Redwood, RG., 1983, “Design of I- beams with web perforation”, Beams and 

beam columns, stability and strength,(Editor:R.Narayanan), applied Science 

Publishers, London, p. 559–62. 

4. ENV 1993-1-3, Eurocode 3: Design of steel structures: Part 1.1. General rules 

and rules for buildings,1992, and Amendment A2 of Eurocode 3: Annex N 

‘Openings in webs’. British Standards Institution,1998. 





5. Redwood, RG., 1968, “Ultimate strength design of beams with multiple 

openings”, Preprint N°757, ASCE: Structural Engineering Conference, 

Pittsburg (Oct). 

6. Redwood, RG., 1971, “Stresses in webs with circular openings”, Final Report 

to the Canadian Steel Industries Construction Council, Research Project N° 

695, (Dec). 

7. Chan, PW. , Redwood, RG. , 1974, “Stresses in beams with circular eccentric 

web holes”, J Struct Div, Proc ASCE, 100(ST1):231–48. 

8. Knowles, PR., 1985, “Design of castellated beams for use with BS5950 and 

BS449”, Constrado 

9. Ward, JK. , 1990, “Design of composite and non-composite cellular beams”, 

The Steel Construction Institute, publication 100. 

10. Redwood, RG., Cho SH. 1993, “Design of steel and composite beams with 

web openings”, J.Construct Steel Res; 25: 23–41. 

11. Ko, CH., Chung KF. 2000, “A comparative study on existing design rules for 

steel beams with circular web Openings” In: Yang YB, Leu LL, Hsieh SH, 

editors, Proceedings of the First International Conference on Structural 

Stability and Dynamics, Taipei.. p. 733-8 

12. Chung, KF. , Liu, TCH. , and Ko, ACH. , 2001, “Investigation on Vierendeel 

mechanism in steel beams with circular web openings”, J Construct Steel 

Res;57(5):467–90. 

13. Savin, GN. , 1961, “Stress concentration around holes”, Pergamon Press, 

Oxford. 

14. Bower, JE. , 1966, “Elastic stresses around holes in wide-flange beams”, J 

Struct Div, Proc ASCE, 92(ST2):85–101. 

15. Bower, JE. , 1968, “Design of beams with web openings”, J Struct Div, Proc 

ASCE, 94(ST3):783–807, See also US Steel, Building Design Data, ADUSS 

27-3500-01 (Apr.1968) and preprint N° 493, ASCE: Structural Engineering 

Conference (May.1967). 

16. Chan, P., 1971, “Approximate methods to calculate stresses around circular 

holes”, Fourth Progress Report to the Canadian Steel Industries Construction 

Council, Project N° 695, (Nov). 

17. Bower, JE. , 1968, “Ultimate strength of beams with rectangular holes”, J 

Struct Div, Proc ASCE, 94(ST6):1315–37. 

18. Redwood, RG. , McCutcheon, JO. , 1968, “Beam tests with un-reinforced 

web openings”, J.Struct Div, Proc ASCE; 94(ST1):1–17. 

19. Redwood, RG., 1978, “Analyse et dimensionnement des poutres ayant les 

ouvertures dans lesâmes”, Construction Métallique N° 3, 15-25. 

20. Liu, TCH., Chung, KF., 1999, “Practical design of universal steel beams with 

single web openings of different Shapes”, In, Proceedings of the Second 

European Conference on Steel Structures ‘Eurosteel 99’, Prague, May 1999, 

p. 59–62. 

177


21. McCormick, MM., 1972, “Open web openings: behaviour analyses and 

design”, BHP Melbourne Research Laboratories, Report 17.18. Melbourne, 

Australia. 

22. Dougherty BK. 1980, “Elastic deformation of beams with web openings”, J 

Struct Div, Proc ASCE; 106(ST1):301–312. 

23. Redwood, RG., shrivastava, CS., 1980, “Design recommendations for steel 

beams with web holes”, Canadian Journal Civil Engineering; 7(4), 642-650. 

24. Darwin, D., 1990, “Design of steel and composite beams with web openings”, 

In: Steel design guide series 2. American Institute of Steel Construction,. 

25. Redwood, RG., Uenoya, M., 1979, “Critical loads for webs with holes”, J 

Struct Div, Proc ASCE; 105(ST1):2053 –2067. 



THE PREVENTION OF VORTICES AND SWIRL AT PUMP INTAKES 

Samir Ali Ead 1 

1: Associate Professor, Dept. of Civil Engineering, King Saud Univ., Riyadh, KSA; 

and Former Research Scientist, Northwest Hydraulic Consultants, Edmonton, Alberta, 

Canada, E-mail: Samir@ualberta.net 

ABSTRACT 

Experiments were conducted on a physical hydraulic model of a circulating 

water pump sump structure. The cooling water intake structure consists of two 

circulating water pumps and two auxiliary water pumps withdrawing flow from one end 

of a cooling tower basin. The objectives of the hydraulic model study were to evaluate 

the performance of the initial design of the pump sump and to develop modifications to 

improve the approach flow hydraulics to the circulating water pumps. The initial design 

of the intake structure developed strong surface vortices in the cooling tower and in the 

vicinity of the circulating water pumps. As a result, constant surface and sub-surface 

vortices were observed entering the circulating pumps. Modifications were developed in 

the model to reduce the level of the vortex activity and improve the flow conditions 

entering the pump. With these modifications installed in the model, flow pre-swirl, 

surface and sub-surface vortices, and the velocity distribution at the pump impeller 

location were all within acceptable limits for the range of expected pump discharge and 

water level operating combinations examined in the model. It was found that there is a 

minimum water level for the cooling tower basin when two circulating water pumps and 

two auxiliary water pumps are operating. The crest of the slope leading to the sump 

becomes a control point at low water levels that causes the flow to go through critical. 

Once this occurs, the water level in the cooling tower basin cannot drop below this 

level. If the plant tries to operate below this minimum level, the water level in the 

cooling tower will not change and instead, the water level in the sump will decrease and 

may dewater the sump. 

KEY WORDS 

Vortices; Swirl; Pump sump; Circulating water pumps; Auxiliary water pumps; 

Cooling tower. 



INTRODUCTION 

The pump intake is perhaps the most important element in the structure of the 

pumping station. Unless it is properly located, designed, and sized, the flow conditions 

within could have an adverse effect on the operation of the pump. There are many 

variations in sump arrangements that are acceptable; however, best results are obtained 

when the sumps or sump bays are oriented parallel to the line of flow. Flows 

approaching from an angle create dead spots and high local velocities, which result in 

the formation of vortices, non-uniform entrance velocities, and an increase in entrance 

losses. The flow to any pump should not be required to pass another pump before 

reaching the pump it is meant for. When sumps or sump bays are normal to the direction 

of flow the distance between the sump or sump bay entrance and the pump must be 

sufficient for the flow to straighten itself out before reaching the pump. 

Adverse hydraulic conditions that can affect pump performance include 

formation of surface and sub-surface vortices, swirl approaching the pump impeller, 

flow separation at the pump bell, and a non-uniform axial velocity distribution at the 

suction inlet (see Tullis (1979) and Durgin (1978)). Strong free-surface vortices that are 

capable of entraining air could result in a reduction of the pump capacity and potentially 

the loss of prime (see Murakami (1969)). In addition, the low pressure at the core of the 

vortex could reduce the local pressure at the impeller to the point where localized 

cavitation damage develops. Any of these problems can adversely affect pump 

performance by causing cavitation, vibrations, and/or loss of efficiency (see Denny and 

Young (1957)). Usually there is more than a single reason for these problems, and the 

extent of the combined effects are seldom predictable by mathematical modeling. 

Formation of vortices, for example, even though dependent on suction pipe velocity and 

submergence, is strongly influenced by added circulation from vorticity sources, such as 

a non-uniform approach flow resulting from intake and approach channel geometries, 

rotational wakes shed from obstructions, such as columns or piers; and the velocity 

gradients resulting from boundary layers at the walls and floor. The circulation 

contributed by these vorticity sources is unpredictable and strongly dependent on intake 

design and operating conditions, especially for large pumping units with multiple bays 

fed by a common approach channel. In these cases, physical modeling is the only way 

of predicting the behavior of the prototype with a reasonable degree of reliability. 

The ideal conditions and assumptions upon which the geometry and dimensions 

recommended for rectangular intake structures are based, are that the structure draws 

flow so that there are no cross-flows in the vicinity of the intake structure that create 

asymmetric flow patterns approaching any of the pumps, and the structure is oriented so 

that the supply boundary is symmetrical with respect to the centerline of the structure. 

As a general guide, cross-flow velocities are significant if they exceed 50% of the pump 

bay entrance velocity. If multiple pumps are installed in a single intake structure, 

dividing walls placed between the pumps result in more favorable flow conditions than 




Samir Ali Ead 

that found in open sumps. Adverse flow patterns can frequently occur if dividing walls 

are not used. For pumps with design flows greater than 5,000 gpm dividing walls 

between pumps are required. 

Severe free surface vortices may be broken up and effectively suppressed by 

arranging baffles and vanes to correct the rotational flow field due to the approach 

flow distribution. Relocation of pumps, using breaker pipes, introduction of 

horizontal grids below the water surface, changing of wall and floor clearances, 

improvements in approach channel configurations, and changes in lengths and 

spacing of piers are some of the common techniques for reducing vortex activity. 

Surface vortices can also be controlled by installing a curtain wall immediately 

upstream of the pump. Submerged vortices are usually eliminated by installing 

splitter vanes or floor cones under the bell. Variations in floor and wall clearances 

can also be effective and should be tried in the model. 

Given these potential impacts on pump performance, the following performance 

criteria have been developed from the Hydraulic Institute’s (HI) Pump Intake Design 

Standard (1998) and Flowserve’s Test Standard (2002), and were utilized in evaluating 

the performance of the circulating water pump sump design for the current study: 

• Free surface and sub-surface vortices entering the pump must be less severe 

than vortices with diffusive dye cores (Type 2). Figure 1 illustrates the 

Vortex Classification system utilized for the study. 

• The average swirl angle should be less than 2.5 degrees. The swirl meter 

rotation should be reasonably steady, with no abrupt changes in direction 

when rotating near the maximum allowable rate (angle). 

• Time-averaged velocities, V, at points in the throat of the bell or at the pump 

suction in a piping system shall be within 10 percent of the cross-sectional 

area average velocity. 

• Time-varying fluctuations at a point shall produce a standard deviation from 

the time-averaged signal of less than 10 percent. 

181


Figure 1 Vortex classification system 

PROPOSED PUMP STATION DESCRIPTION 

The physical model was constructed to assist in the design of water pump sump 

of the prototype (Hillabee Energy Center Project located in Tallapoosa County, 

Alabama). The cooling tower intake structure consists of two circulating water pumps 

and two auxiliary water pumps withdrawing flow from one end of a 648’ long by 66’ 

wide cooling tower basin. The cooling water sump is approximately 63’-0” long by 49’- 

8” wide and is divided into two circulating water pump bays and two auxiliary water 

pump bays. There is an 11:1 sloping transition from the cooling tower basin floor at El. 

568’-0” to the pump sump floor at El. 553’-0”. Also, there is a transition in the cooling 

tower walls from the full width of the cooling tower basin (66’-0”) to the sump width 

(49’-8”). The circulating water pump bays are 12’-10” wide by 46’-3” long, while the 

auxiliary pump bays are 6’-0” wide by 46’-3” long. The pump bays are separated with 

4’-0” wide walls that are tapered at their upstream ends. 

The circulating water pumps have 77-inch diameter bells and are rated at 

99,000 gallons per minute (gpm), with a runout discharge of 115,000 gpm. The 




Samir Ali Ead 

auxiliary water pumps have 36-inch diameter bells and are rated at 13,000 gpm. Both 

circulating and auxiliary water pumps are 50 percent capacity pumps. The cooling tower 

basin low water level is El. 569’-0” and the high water level is El. 573’-0”. Because of 

symmetry about the sump centerline, performance data were collected for only one 

circulating water pump (Circ. 1). 

DIMENSIONAL ANALYSIS 

Scale hydraulic models require that the force relationships in the model and 

prototype are dynamically similar. To achieve complete similarity, the ratio of the 

inertia to the gravitational, viscous, and surface tension forces must be the same 

between model and prototype. Only a 1:1 scale model can achieve all these criteria. 

Modeling at a reduced scale involves identifying the primary force relationship to best 

simulate prototype conditions, then selecting a model scale to minimize any resulting 

scale effects. For free-surface flow conditions of the type being examined in the present 

investigation, the gravitational force is the dominant force that defines the 

hydrodynamic flow conditions, and the ratio of the inertial to gravitational forces, 

represented by the Froude number F must be equal in the model and prototype. That is, 

F 

r 

F 

= 

F 

where, 

M 

P 

= 1 

F M = Froude number in the model 

= 

Inertial Force 

= 

Gravitational 

Force 

F P = Froude number in the prototype = 

U P 

g L 

and, U = characteristic flow velocity, 

g = gravitational acceleration, and 

L = characteristic length 

M = model values 

P = prototype values 

P 

U M 

g L 

In modeling flow in a pump sump to evaluate the potential for the formation of 

vortices, the geometric scale is selected to minimize viscous and surface tension scale 

effects. Also, the model should be large enough to allow flow visualization, accurate 

measurements of flow pre-swirl and velocities, and sufficient dimensional control. The 

Reynolds number defines viscous effects and the Weber number defines surface tension 

effects as defined below: 

183 

M


R e 

W e 

UL Inertial Force 

= Reynolds number = = 

v Viscous Force 

= Weber number = 

U Inertial Force 

= 

σ Surface Tension Force 

ρL 

Where σ is the surface tension coefficient, ρ is the water mass density, and ν is 

the kinematic viscosity. Based on the available literature, the influence of viscous and 

surface tension forces is negligible if the model bell entrance Re and We are above 6 x 

10 4 and 240, respectively. At a model scale of 1:8.93, the circulating water pumps met 

the necessary criterion with a bell entrance Reynolds number and Weber number of 1.4 

× 10 5 and 1,437, respectively. 

At a scale of 1:8.93, adherence to the Froude criterion for dynamic similarity 

leads to the following scale ratios: 

Parameter 

Length 

Velocity 

Discharge 

PHYSICAL MODEL 

Model Scale Relationships 

Relation 



Lr 

Lr 1/2 

Lr 5/2 

Ratio 

1: 8.93 

1: 2.99 

1: 238.30 

The physical model used reproduced the four pump bays and a portion of the 

cooling tower basin adjacent to the intake structure in accordance with the model design 

drawings as shown in Figures 2 & 3 and photos 1 & 2. The intake structure was built on 

a raised platform and was constructed using a combination of marine plywood and 

transparent plastic for viewing of flow patterns. At the selected scale of 1:8.93, the 

model circulating water pump bell diameter was 8.625 inches. Circulating pump bells 

were made from transparent plastic and connected to a cast acrylic tube representing the 

pump column (see Figure 4). Flow was withdrawn through the two auxiliary pumps to 

ensure accurate simulation of the flow patterns approaching the circulating water pump 

bays, but performance data was not collected for the auxiliary pumps.


Samir Ali Ead 

The model used a self-contained circulation system to produce the model flows, 

which were controlled with a butterfly valve installed in the pipe supplying flow to the 

head box located at the end of the model cooling tower basin. The model circulating 

water and auxiliary water pumps were connected through a discharge manifold to the 

suction end of a centrifugal laboratory pump. Individual pump flow rates were adjusted 

with butterfly valves in each of the model pump suction lines. Water levels in the 

cooling tower basin were varied by changing the amount of water in the model. The 

flow requirements in the model ranged from 1.3 cfs for a single circulating water pump 

operating with two auxiliary water pumps to 2.1 cfs for the operation of all four pumps. 

Figure 2 Model layout (section A-A) 

185


NOTES: 

AUX 1 

CIRC 1 

CIRC 2 

AUX 2 

Figure 3 Model layout (sump details) 




Samir Ali Ead 

Photo 1 View looking downstream at the four pump bays, the sloping forebay 

area, and transition of the sidewalls. 

187


Photo 2 Close-up view of the model pump bell for Circ 1 pump showing sump details 

(fillets and splitters) with swirl meter and high-speed velocity probe installed. 



Figure 4 Model pump bell geometry 


Samir Ali Ead 

189


MODEL MEASUREMENTS AND INSTRUMENTATION 

The following measurements were required for the model 

Flow Rates - The total model flow rate was measured using orifice-plate flow 

meter, with air-water manometer used to measure the pressure differentials. An 

orifice plate was installed in accordance with ASME Test Code Standards. 

Individual pump flows were measured using elbow meters that measured the 

pressure difference between the inside and outside of a 90 degree bend installed in 

the discharge line for each pump. These elbow meters were calibrated in-place 

using the orifice-plate flow meters. 

Free-Surface Vortices - Free-surface vortices were measured by visual 

observation using dye injection and were based on the Hydraulic Institute's free 

surface vortex strength scale of Type 1 to Type 5 (Figure 1). 

Sub-Surface Vortices - Sub-surface vortices were measured by visual observation 

and based on the Hydraulic Institute's sub-surface vortex strength scale of Type 1 

to Type 5, as shown in Figure 1. As with the free-surface vortex tests, dye was 

used to identify these vortices. 

Flow Pre-swirl – Swirl meters installed in circulating model pump bells, as shown 

in Figure 4 and Photo Plate 2, provided a measure of average intensity of swirl 

angle, θ, according to the following equation: 

θ = 

tan 

− 1 

π dn 

( 

u 

where: d = diameter of the pipe at the swirl meter 

n = revolutions/second of the swirl meter 

u = average axial velocity at the swirl meter 

Water Levels - Water levels were measured using staff gages in the cooling tower 

basin and the individual pump bays. 

Approach Flow Patterns - Visual aids such as colored dye and neutral-density 

particles were used to document the flow patterns in the cooling tower basin and 

individual pump bays. 

Velocity Distribution at the Throat of the Pump Suction Bell - Pump inlet 

velocities were measured with a miniature propeller velocity probe installed at the 

throat of Circ 1 pump (see Figure 4). Velocities were measured at eight locations 

at a constant radius from the pump axis around the circumference of the pump 



)


Samir Ali Ead 

throat. Statistical values such as maximum, minimum and maximum temporal 

fluctuation were computed from the velocity probe data. 

Photographs and Video - Still photographs were taken throughout the test 

program to provide a visual documentation of the model study progress and key 

results. Relevant photographs have been included in this paper. 

MODEL TESTING 

The test program consisted of the following three phases: 

Initial Design Testing - Tests conducted to evaluate the performance of the initial 

design and determine the most severe operating condition(s). Two initial design tests 

were conducted at the lowest expected water level. 

Modification Testing - Tests conducted to develop modifications for the initial design 

to improve the performance of the intake operating under the most severe flow 

condition(s) that were identified in the Initial Design Testing. 

Documentation Testing – Tests to confirm that the final adopted design will meet the 

specific operational criteria over the full range of operating conditions. Four 

documentation tests were conducted with the proposed final modifications installed in 

the model. 

It should be noted, that given the fact that the cooling tower and sump are 

symmetrical about the cooling tower centerline, data was only collected for one of the 

circulating water pumps (Circ 1). The results obtained for this pump are also applicable 

to the second circulating water pump (Circ 2). 

TEST RESULTS 

Initial Design Testing 

Two evaluation tests were conducted with the initial design (Geometry 1) 

installed in the model and at the lowest attainable water level in the cooling tower basin. 

The results of these tests are presented in Table 1 and representative flow patterns are 

illustrated in Photos 3-5. Tests included a single circulating water pump operating at run 

out flow (115,000 gpm) and two circulating water pumps operating at design flow 

(99,000 gpm). Both tests were run with two auxiliary water pumps operating at 

13,000 gpm, as outlined in the following table. 

191


Test 

No. 

Table 1 Initial Design Test Program 

Water Level 

Circ 1 

Operating Pump Discharge (gpm) 

Circ 2 Aux 1 Aux 2 

1-1 569’-0” 115,000 x 13,000 13,000 

1-2 569’-0” 99,000 99,000 13,000 13,000 

It is important to note that the specified water level of El 569’-0” resulted in a 

flow depth of 1.0 ft in the cooling tower basin. This is near or at “critical” flow depth in 

the basin which is computed as 1.2 ft when both circulating water pumps and both 

auxiliary water pumps are operating (total discharge = 224,000 gpm), and 0.9 ft when 

one circulating water pump is operating at runout discharge in combination with both 

auxiliary pumps (total discharge = 141,000 gpm). In the prototype, the water level in the 

cooling tower basin will not fall below these critical flow depths. However, there will 

be the potential for the water levels in the sump to fall well below the upstream level, 

which could lead to excessive vortex activity, flow pre-swirl, air entrainment, and 

possibly dewatering of the sump. 

In the model, critical flow conditions were observed in the cooling tower basin, and 

a transfer from supercritical flow to subcritical flow occurred at the upstream end of the 

sloping floor transition in the form of a hydraulic jump (see photo 3). The hydraulic jump 

generated a turbulent water surface in the cooling tower and in the vicinity of the 

circulating water pumps. The flow entering the pump sump separated from the dividing 

wall between the circulating water pump bays. This flow separation generated a moderate 

variation in the flow velocity across the bay width, which resulted in circulation in the 

pump bay and flow pre-swirl entering the pump. The maximum flow pre-swirl recorded in 

the initial design testing was 2 degrees (14 rpm). Strong surface vortex activity of type 

4(C), with air bubbles in the vortex core, was observed during both initial design tests (see 

photos 4 and 5). Also, constant Type 1 sub-surface vortices were observed originating 

from the backwall of the circulating water pump bays, adjacent to the vertical backwall 

splitter, for all operating conditions. The maximum deviation from the average flow 

velocity within the throat of Circ 1 was 10%, and the temporal fluctuations produced a 

maximum deviation of 9% from the time-averaged signal. 

In summary, the surface vortex activity was found to exceed the HI-specified 

performance criterion while all other parameters (sub-surface vortex activity, pre-swirl 

and throat velocities) were within the specified criteria limits. 




Samir Ali Ead 

Photo 3 View of hydraulic jump forming on the transition slope with both circulating pumps 

operating at a design discharge of 99,000 gpm and auxiliary water pumps at 13,000 

gpm. 

Photo 5 View of constant Type 4 vortex forming at the 

water surface of Circ 1 pump bay with both pumps 

operating at a design discharge of 99,000 gpm. 

Photo 4 View of constant Type 4 vortex forming at the 

water surface of Circ 1 pump bay with pump 

operating at a runout discharge of 115,000 gpm. 

193


Modification Testing 

Following the initial design testing, a series of design modification tests was 

conducted. The objective of these tests was to develop and implement changes to the 

initial design that would reduce or eliminate the adverse flow conditions observed for 

the initial geometry. Modification testing was conducted at the minimum water level 

attainable in the cooling tower basin. 

A total of thirteen different geometrical arrangements were investigated in the 

model as part of modification testing. Some of the modifications were examined 

independently, while others were examined in combination with other modifications. 

Modifications examined in the model included the following 

• installation of a curtain wall at various depths, locations, and inclinations 

within the circulating water pump bays; 

• installation of flow straightening bars at various locations; 

• removal of vertical corner fillets and backwall splitter; 

• installation of surface vortex breaker pipes. 

Curtain Wall 

Pump sump designs frequently require the installation of a curtain wall to reduce 

or eliminate the level of surface vortex activity and reduce flow pre-swirl. For the 

present study, testing was conducted with a curtain wall installed in the circulating 

water pump bays at alternative locations, elevations, and inclinations. It was found that 

the placement of the curtain wall was influential on reducing surface vortex activity. For 

example, by setting the bottom of the curtain wall at El. 565’-9.5” and locating the wall 

17’-2” from the backwall of the sump, the surface vortex activity was reduced to 

constant Type 2 vortices observed both upstream and downstream of the curtain wall. 

Based on these results, the installation of a curtain wall located 17’-2” from the 

backwall and set at El. 565’-9.5” was recommended. With the curtain wall in the 

recommended location, flow pre-swirl entering the pump decreased from 2 degrees (14 

rpm) to 1 degree (10 rpm). 

Flow Straightening Bars 

Several tests were conducted to examine the use of flow-straightening bars, in 

combination with the curtain wall, at various locations within the circulating water 

pump bays. Installation of flow straightening bars created a more uniform flow 

distribution within the pump bay. Also, installation of flow-straightening bars 

eliminated the surface vortex activity downstream of the curtain wall. However, 

intermittent Type 2 surface vortices were observed between the bar rack and the 

curtain wall. Generally, these vortices were dissipated upon flowing under the curtain 

wall but some could be observed reaching the circulating water pump bell. Constant 




Samir Ali Ead 

Type 2 sub-surface vortices were observed along the sump backwall in some tests, 

and flow pre-swirl increased to 2 degrees (15 rpm). Although the use of flow 

straightening bars provided some improvements in the flow distribution within the 

pump bay, the improvement was relatively insignificant and the installation of a bar 

rack was not recommended for this project. 

Removal of Corner Fillets and Backwall Splitter 

Subsequent to the installation of flow-straightening bars and a curtain wall, the 

back wall splitter and vertical corner fillets were removed from both pump bays to 

examine their impact on the flow approaching the pumps. The removal of the backwall 

splitter reduced the strength of the sub-surface vortex activity downstream of the pump 

column from constant Type 2 to constant Type 1. However, removal of the backwall 

splitter also resulted in a minor increase in the flow pre-swirl entering the pumps 

(increased from 1 degree to 2 degrees). In addition, removal of the corner fillets caused 

constant Type 1 vortices to appear near the downstream corners of the pump bays. 

Based on these tests, the removal of the backwall splitter was recommended whereas the 

existence of the corner fillets was required. Although vortex activity was acceptable, it 

was recommended to keep the corner fillets in the design because flow approaching the 

pumps was more uniform. 

Installation of Surface Vortex Breaker Pipes 

Several tests were conducted to examine the use of surface vortex breaker pipes 

installed upstream of the curtain wall as a means of dissipating the surface vortex 

activity observed in this area. Five pipes (9” diameter and 13.5” center to center) were 

installed flush with the bottom elevation of the curtain wall (El. 565’-9.5”). With the 

vortex breaker pipes installed, the surface vortices upstream of the curtain wall were 

effectively dissipated as they passed through the pipes. Also, flow distribution within 

the pump bay was more uniform. Based on these results, the installation of surface 

vortex breaker pipes was recommended. 

Proposed Modifications 

On the basis of the above modification testing, the proposed design 

modifications for both circulating water pump bays included 

� installation of a curtain wall 17’-2” from the backwall with a bottom set at 

El. 565’-9.5”; 

� installation of five 9-inch diameter surface vortex breaker pipes spaced at 

13.5 centers upstream of the curtain wall and flush with the bottom 

elevation of the curtain wall ; and 

� removal of backwall splitter. 

With one circulating water pump operating at runout discharge of 115,000 gpm 

and both auxiliary pumps operating at 13,000 gpm for the lowest attainable water level, 

195


the flow pre-swirl angle was 2 degree (13 rpm). No vortex activity was observed at the 

surface, sidewall or on the floor. Constant Type 1 sub-surface vortices were observed 

near the backwall as a result of flow separation from the pump column. The spatial 

variation of velocities within the throat of the bell (Circ 1) was within 2% and the 

temporal velocity fluctuations were 5% (both are within the specified criteria of ±10%). 

Documentation Testing 

The objective of the documentation testing was to ensure that the modifications 

to the intake would perform satisfactorily for a range of operating conditions. Photos 6 

& 7 illustrate the modifications to the sump, which include 


El. 565’-9.5”; 


13.5-inch centers upstream of the curtain wall and flush with the bottom 

elevation of the curtain wall ; and 


The four detailed documentation tests listed below were conducted with the above 

modifications installed in the model. 

Test 

No. 

Water Level 

Table 2 Documentation Testing 

Operating Pump Discharges (gpm) 

Circ 1 Circ 2 Aux 1 Aux 2 

1 569’0” 115,000 x 13,000 13,000 

2 569’-0” 99,000 99,000 13,000 13,000 

3 573’-0” 115,000 x 13,000 13,000 

4 573’-0” 99,000 99,000 13,000 13,000 



Comments 

Runout discharge 

(Circ 1)-lowest 

attainable water level 

Design dischargelowest 

attainable 

water level 

Runout discharge 

(Circ 1)-high water 

level 

Design dischargehigh 

water level


Samir Ali Ead 

Table 2 summarizes the results from these tests. In all cases, the 

recommended modifications to the circulating water pump bays were successful in 

providing acceptable flow to the pumps. With the installation of a curtain wall, 

the vortex breaker pipes and the removal of the backwall splitter, the surface and 

sub-surface vortex activity in the vicinity of the pump bells was all Type 1 or 

less, and the flow pre-swirl entering the pumps did not exceed 1 degree (12 rpm). 

Velocities recorded at the throat of the circulating water pump (Circ 1) 

demonstrated that the maximum deviation in the average flow velocity was 4%, 

and the maximum temporal fluctuation recorded was 4%. 

In addition, to the above tests, two supplemental tests were conducted to 

determine the minimum water levels within the cooling tower basin required to 

eliminate the formation of a hydraulic jump and the associated turbulent flow 

conditions within the basin for the two operating scenarios (one pump at runout 

discharge, or two pumps at design discharge). These water levels were found to be 

569’-6” and 570’-0” for one and two circulating water pumps operating, respectively. 

As indicated earlier, the specified low water level of 569’-0” is close to or less than 

critical flow depth and will likely not be achievable in the prototype. Furthermore, 

operating at or near the critical flow depth could lead to potential pump performance 

problems in the sump, including possible dewatering of the sump. 

197


Photo 6 Close up view of the curtain wall and vortex breaker pipes. 




Samir Ali Ead 

Photo 7 View looking upstream towards Circ 1 pump bay. The removal of the 

backwall splitter reduces the sub-surface vortex activity and improves 

flow movement to the pumps. 

CONCLUSIONS 

Testing conducted for the initial design of the cooling tower intake structure 

demonstrated that the specified low water level of 569’-0” will lead to a the 

formation of a hydraulic jump near the upstream end of the sloping floor leading to 

the sump. This water level is very close to the critical flow depth in the cooling 

tower basin and could result in adverse hydraulics in the sump, including the 

possibility of dewatering of the sump. 

With the initial design, constant Type 4 surface vortices formed and entered the 

circulating water pumps. Sub-surface vortex activity, flow pre-swirl and the velocity 

measurements at the throat of the circulating water pump (Circ 1) were all within the 

Hydraulic Institute performance criteria for the initial design. 

199


Several modifications to the initial design were developed in the model to reduce 

the surface vortex activity such that the hydraulic conditions in the sump would meet 

the specified performance criteria. The modifications developed in the model included 

the following 


El. 565’-9.5”; 


13.5-inch centers upstream of the curtain wall and flush with the bottom 

elevation of the curtain wall; and 


With these modifications installed, flow pre-swirl, surface and sub-surface 

vortex activity, and the velocity distribution at the pump impeller location were all 

within acceptable limits for the range of expected pump and water level operating 

combinations examined in the model. The minimum recommended water level in the 

cooling tower basin to eliminate the formation of a hydraulic jump and avoid the 

potential for adverse hydraulics and degradation in pump performance in the sump is El. 

570’-0”. 

REFERENCES 

Denny, D. F., G.A.J. Young (1957) “The Prevention of Vortices and Swirl at Intakes.” 

Paper No. C1 in Proceedings of IAHR 7 th Congress, Lisbon. 

Durgin, W. W., and Hecker, G. E. (1978) “The Modeling of Vortices at Intake 

Structures.” In Proceedings of the ASCE, IAHR, and ASME Joint Symposium on 

Design and Operation of Fluid Machinery, Colorado State University, Fort 

Collins, Colo., June, Vol. 1, p. 381. 

Flowserve Pump Division (2002) “Test Standards for Pump Model Intakes.” 

Murakami, M. (1969) “Flow of Entrained Air in Centrifugal Pumps.” In Proceedings of 

the 13 th Congress IAHR, Kyoto, Japan, Vol. 2, p. 71. 

The Hydraulic Institute (1998) “American “National Standard for Pump Intake Design.” 

Parsippany, New Jersey. 

Tullis, J. P. (1979) “Modeling in Design of Pumping Pits.” J. Hydraulic Div. ASCE, 

September, p. 1053. 



NOTATION 

D = Bell diameter; 

d = Diameter of the pipe at the swirl meter; 

F = Froude number equals U/(gL) 0.5 ; 

g = Acceleration due to gravity; 

L = Characteristic length; 

n = revolutions/second of swirl meter; 

R = The Reynolds number equals Ud/ν; 

u = average axial velocity at the swirl meter; 

U = Characteristic flow velocity; 

V = Throat velocity; 

W = The Weber number equals ρU 2 L/σ; 

ν = Kinemetic viscosity; 

θ = Swirl angle equals tan -1 (πdn/u); 

ρ = Density; 

σ = Surface tension coefficient. 


Samir Ali Ead 

201




STRENGTHENING OF RC BEAMS USING HONEYCOMB PLATES 

REINFORCED BY FRP COMPOSITES 

T. H. Almusallam 

Department of Civil Engineering, KSU, PO Box 800, Riyadh, Saudi Arabia, 

musallam@ksu.edu.sa 

ABSTRACT 

In the present study, performance and efficiency of concrete beams upgraded 

using honeycomb material reinforced with Carbon or Glass Fiber Reinforced 

Polymers (CFRP or GFRP) has been studied. The performance and efficiency of 

beams are studied in terms of load carrying capacity, vertical deflections and 

ductility. For this purpose, 9 concrete beams, reinforced with SABIC steel, were 

cast. The beams were divided into three groups. The first group consists of RC 

beams without any strengthening and this group was considered as group of control 

samples. The second and third groups of beams were the same as first group but 

strengthened with honeycomb plates reinforced with glass and carbon fiber 

reinforced polymers, respectively. These beams were tested in flexure, and load 

carrying capacity, deflection, ductility, and modes of failure of beams were 

investigated. The results, in general, indicated that the use of GFRP or CFRP 

strengthening increases the load carrying capacity and decreases the deflection 

substantially. 

KEY WORDS 

Honeycomb, GFRP, CFRP, strengthening, repair, composites. 

INTRODUCTION 

Corrosion of steel reinforcement is considered as one of the biggest problem 

for structures standing near shores of Kingdom of Saudi Arabia. Over the last few 

decades substantial research has been conducted in order to find solutions for the 

corrosion problem of steel reinforced concrete. As a result, methods such as 

galvanization, the use of stainless steel bars, cathodic protection and epoxy coatings 

have been tried. None of these remedies has proved to be completely efficient. The 

excellent properties of fiber-reinforced polymers (FRP) suggested that these 

materials may be the solution for corrosion problems of reinforcing concrete 

members. In addition to high corrosion resistance, FRP also possesses high strengthto-weight 

ratio, high tensile strength, light weight and high fatigue resistance. 


STRENGTHENING OF RC BEAMS USING HONEYCOMB PLATES REINFORCED BY FRP COMPOSITES 

In the last three decades several research papers have been published on FRP 

strengthened RC beams. Ritchie et al. [1] investigated behavior of concrete beams 

strengthened by bonding FRP- (glass, carbon, and aramid) plates to the tension zone. 

Experimental results showed that FRP reinforcement increased beam stiffness by 

17-79% and beam ultimate-strength by 40-97%. They also developed an analytical 

method, based on strain compatibility, to predict the strength and stiffness of the 

plated beams. Sharif et al. [2] studied strengthening of pre-loaded reinforcedconcrete 

beams using GFRP plates and concluded additional failure mechanisms. 

Hag-Elsafi et al. [3] discussed application of FRP materials in retrofitting reinforced 

concrete bridge members. Hag-Elsafi and Alampalli [4] investigated similar 

applications for prestressed concrete bridge member. Trinatafillou and Plevris [5] 

showed that a relationship exists between plates sheet area fraction and energy 

absorption capacity for a standard RC beam. They illustrated that at any load level, 

higher CFRP plate fraction area lowers energy absorption capacity and failure type 

is governed by a very low energy absorption capacity at ultimate load. They 

illustrated that load carrying capacity increases by 50% with a reduction in energy 

absorption capacity of 100% at ultimate load. Alsayed et al. [6] tested thirty-six 

simply supported RC beams to study the effectiveness of GFRP laminates in 

improving the flexural capacity of beams GFRP laminates were externally bonded to 

the tension side of the beams. The behavior was presented in terms of loaddeflection, 

load-strain, failure patterns and structural ductility. All beams showed 

considerable increase in ultimate load capacity with a good energy absorption 

capacity. They also observed that strengthening of existing concrete beams is 

particularly effective when the internal steel reinforcement ratio is relatively small. 

Saadatmanesh and Ehsani [7] tested four simply supported beams that were 

internally reinforced by a single 9.5 mm diameter steel rebar and externally 

reinforced by epoxy bonded FRP laminates. It was noticed that by using a tough 

rubber epoxy with consistent cement paste, the test specimens reach to an increase 

load of 100% over the control beam and no cracks were observed up to 70% of the 

ultimate load. The failure was the result of delamination of a strip of concrete just 

above the bond line, along the full length of the beam. Al-Sulaimani et al. [8] tested 

sixteen simply supported reinforced concrete beams to study the effectiveness of the 

use of FRP laminates in improving the shear capacity of beams. FRP laminates were 

externally bonded to the webs and the tension side of the beams. The beam 

specimens were damaged to have a deficiency in shear capacity. Prior to 

strengthening, the beams were damaged to a predetermined load level. Test results 

show that the use of composite laminates not only effectively restored the stiffness 

of the degraded beam but also increased the capacity beyond the pre-loading stage. 

AlMusallam and Al-Salloum [9] used GFRP and CFRP laminates to strengthen 

eighteen simply supported RC beams with ρ = 0.384ρmax. In their study considerable 

increase in ultimate load capacity (100 to 200%) was recorded. 





A detailed review of literature shows that a limited study is reported on 

strengthening of RC beams using honeycomb plates reinforced with fiber 

polymers. Keeping this scope in view, in this paper effectiveness and efficiency of 

concrete beams upgraded using honeycomb material reinforced with Carbon or 

Glass Fiber Reinforced Polymers (CFRP or GFRP) has been studied. The response 

of strengthened beams is studied in terms of load carrying capacity and vertical 

deflections. Effect of honeycomb plate on ductility is also studied. For this 

purpose, 6-9 concrete beams, reinforced with steel, were cast. The beams were 

divided into three groups. The first group consists of RC beams without any 

strengthening and this group was considered as group of control samples. The 

second and third groups of beams were the same as first group but strengthened 

with honeycomb plates reinforced with glass and carbon fiber reinforced polymers, 

respectively. These beams were tested in flexure, and load carrying capacity, 

deflection, ductility, and modes of failure of beams were investigated. 


Beam Details and Test Setup 

The test program consists of nine beams. All beams had a span length of 

1200 mm and cross section of 150 × 150 mm. 2 φ10 mm steel bars were used in the 

tension side (bottom). The reinforcement ratio used was ρ = 0.392ρmax. All beams 

were provided with φ 8 mm steel stirrup @ 70 mm center to center. Fig. 1 shows the 

beam setup and dimension. The cross section of beams is shown in Fig. 2. 

425 mm 

P 

P/2 P/2 

150 mm 

425 mm 

100 

h 

FRP 

b LVDT 

1000 mm 

1200 mm 

Strain 

100 mm 

Fig. 1: Schematic diagram of test setup. 

205


Beam Groups 

d = 117 mm 

φ 10 mm 

steel bars 

Stirrup φ 8 mm @ 70 mm 

b=150m 

h = 150 mm 

Fig. 2: The cross-section of all beams. 

The beams considered in the present study are divided into following three groups. 

These are: 

a) Group “Con”: This group represents the control group (without FRP) as shown 

in Fig 3. 

b) Group “GB-G”: This group represents beams strengthened with (GFRP) 

honeycomb plates at the bottom tension face of the beams as shown in Fig. 3. 

c) Group “GB-C”: This group of beams represent beams strengthened with (CFRP) 

honeycomb plates at the bottom tension face of the beams as shown in Fig. 3. 

Group Con Group GB-G Group GB-C 

Fig. 3: The cross-section of the three groups. 



FRP 

honeycomb

Material Properties 



Concrete 

The compressive strength of concrete was determined after testing concrete 

cylinders for all beams at the day of testing (120 days after casting). The average 

compressive strength of concrete was found to be 44 MPa. 

Steel 

The yield of steel bars was determined by performing the tensile test on three 

specimens of each bar diameter. The average tensile strength of steel was observed 

as 425 MPa. 

FRP Honeycomb Plates 

The average tensile strength of GFRP and CFRP plates were found to be 268 

and 488 MPa respectively. The average moduli of elasticity of the same plates were 

obtained as 18 and 37 MPa respectively. 

Preparation of Test Specimens 

All the specimens were prepared in the Civil Engineering Structure Lab of 

King Saud University, Riyadh, Saudi Arabia. To prepare the specimens first steel 

was cut to desired length and molds were formed to cast RC beams. The design mix 

was prepared and poured into the form work for casting. After casting the concrete 

and curing it for specified time, measurements were taken to make holes in the 

specimens for the nails which give better bonding between FRP and surface of 

concrete. The concrete surface was cleaned through sand-blasting followed by 

acetone cleaning. The beam surface and FRP surface were then saturated by epoxy. 

Having completed the saturation, FRP was placed over the surface of concrete and 

fixed together through nails. To make sure that there is no air between the FRP and 

surface, pressure was applied on FRP plate via gribs (Fig. 4). 

207


Fig. 4: Specimen preparation using honeycomb plate reinforced with GFRP. 

Test Procedure 

All specimens were loaded by two concentrated loads until failure. The mid-span 

deflections of all beams were measured with LVDT gauge placed at the mid-span of 

the specimen as shown in Fig. 1. The strains in the steel bars were also measured 

using two strain gauges attached to each steel bars placed at the center of the middle 

bar. Strains in FRP plates were measured using two strain gauges fixed at the center 

of the span. Strains in compression zone concrete were measured by strain gauge 

fixed in the middle of span. 

DISCUSSION OF TEST RESULTS 

Load-Deflection Behavior 

The load versus mid span deflection for all the specimens of strengthened groups 

(group 2 and 3) are shown in Figs. 5 and 6. One curve representing the average of 

three specimens for each group was selected to represent the load-deflection 

relationship for that group. Therefore, the load-deflection for the three groups is 

shown in Fig. 7, and the results are summarized in Table 1. This table shows that for 

beams strengthened with CFRP, the ultimate load was increased by 89.4% over the 

control sample, while beams strengthened with GFRP, the ultimate load was 

increased by 63.48%.The above increase in ultimate load shows effectiveness of 

FRP plates in improving load carrying capacity of beams. 



Load (KN) 

100 

90 

80 

70 

60 

50 

40 

30 

20 

10 

0 

0 10 20 30 40 50 

Deflection (mm) 


GB-G-1 

GB-G-2 

GB-G-3 


Fig. 5:Load deflection relationship for specimens of strengthened group GB-G. 

Load (KN) 

120 

100 

80 

60 

40 

20 

0 

GB-C-1 

GB-C-2 

GB-C-3 

0 5 10 


15 20 

Fig. 6: The load-deflection relationship for specimens of strengthened group GB-C. 

209


Load (KN) 

120 

100 

80 

60 

40 

20 

0 

0 10 20 30 40 50 


Fig. 7: The load-deflection relationship for the three groups of beams. 



Control 

GFRP 

CFRP

Yield 

Stage 

Ultimate 

Stage 

Service 

Stage 

Ductility 

Index 

Ductility 

Table 1: Summary of results 

Con 

(control) 

GB-G 

Load 

Gain 


GB-C 


Load 

Gain 

Py – (kN) 40.92 38.06 - 43.32 5.9 % 

∆y – 

(mm) 

2.17 1.805 1.845 

Pu – (kN) 52.19 85.32 63.5 % 98.85 89.4 % 

∆u – 

(mm) 

18.27 10.415 9.02 

P - (kN) 18.26 23.00 26.0 % 25.19 38.0 % 

∆ – (mm) 0.79 0.79 0.79 

MDy 8.42 5.77 - 4.89 - 

In the present study ductility was obtained in terms of deflection ductility 

index which is defined as, 

∆ 

µ y = 

∆ 

u 

y 

Mid - span deflection at ultimate load ( ∆ u ) 

= (1) 

Mid - span deflection at first yielding of steel ( ∆ ) 

The values of ductility indices on deflection are given in Table 1. This table 

clearly indicates that the control beams are more ductile than the strengthened beams 

with GFRP and CFRP honeycomb plates. It can be also noted that the values of 

ductility indices are reduced as the stiffness of the beam increased due to FRP 

reinforced honeycomb plates as external reinforcement. Therefore, in order to 

achieve a desired capacity as well as desired ductility for any strengthened RC 

beam, quantity of FRP should be selected with great care. 

y 

211


Modes of Failure 

It was observed that the mode failure of beams strengthened with GFRP 

plates was tensile failure in the plates, while the failure mode in the CFRP plates 

was observed as a shear failure in concrete as shown in the Figs. 8 to 10, 

respectively. Both failures did not show any type of delamination of FRP plates. 

Fig. 8(a): Failure mode of control beams. 

Fig. 9(a): Failure mode of GFRP plates. 



Fig. 8(b): A close view of failure. 

Fig. 9(b): A close view of failure.

Fig. 10(a): Failure mode of CFRP plates. 

CONCLUSIONS 



Fig. 10(b): A close view of failure. 

Following conclusion have been drawn from the results of present study: 

• Strengthening of RC beams with CFRP & GFRP honeycomb plates increases 

their service, yield, and ultimate strengths. 

• The gain in service and ultimate loads for all beams, strengthened with GFRP 

or CFRP plates, are substantially higher than that of control beams. 

• Strengthening of beams with both GFRP and CFRP plates increases the 

stiffness of the concrete beams. 

• Strengthening with GFRP and CFRP plates does affect the ductility of beams. 

• Strengthening of RC beams with GFRP or CFRP plates significantly reduces 

the deflection. 

ACKNOWLEDGEMENTS 

Author acknowledges the financial support provided by Saudi Basic 

Industries Corporation (SABIC) under grant number 34/426. 

213


REFERENCES 

[1] Ritchie, P.A., Thomas, P.A. Lu, L.W. and Connely, G.M., 1991, “External 

reinforcement of concrete beams using fiber reinforced plastics,” ACI Struct. 

J Vol. 88, pp. 490–500. 

[2] Sharif A., Al-Sulaimani G.J,, Basunbul I.A., Ghaleb, B.N., 1994, 

“Strengthening of initially loaded reinforced concrete beams using FRP 

plates,” ACI Struct. J., pp. 160–68. 

[3] Hag-Elsafi O., Alampalli S., Kunin J., Lund R., 2000, “Application of FRP 

materials in bridge retrofit,” In: Proceedings of the Seventh International 

Conference on Composites Engineering, Denver, CO, pp. 305–306. 

[4] Hag-Elsafi O., and Alampalli S., 2000, “Strengthening prestressed-concrete 

beams using FRP laminates,” In: Sreenivas Alampalli, editor. Structural 

materials technology. Atlantic City, NJ: An NDT Conference, pp. 287–292. 

[5] Traintafillou, T.C. and Plevris N., 1992, “Strengthening of RC beams with 

epoxy bonded fiber composite materials,” Materials and Structures, Vol. 25, 

pp. 201-211. 

[6] Alsayad, S.H., Al-Salloum, Y.A., Al-Musallam, T.H., 2002 “Rehabilitations 

of the infrastructure Using Como site Fabrics,” Final Report-project AR16- 

52, King Abdulaziz City for Science and Technology (KACST), Riyadh, 

Saudi Arabia. 

[7] Seadatmanesh, H. and Ehsani, M.R., 1999, “Fiber composite plates can 

strengthen beams,” Concrete international, Vol.12, No. 3, pp. 65-71. 

[8] Al-Sulaimani, G.J., Sharif, A., Basunbul I.A., Baluch, M.H. and Ghaleb, 

B.N." Shear Repair for renforced concrete by Fiber glass Plate" 

[9] Almusallam, T.H. and Al-Sallom, Y.A., 2001, “Ultimate strength Prediction 

for RC beams Externally strengthened by composite,” Materials part B 

engineering Vol. 32, pp. 609-61. 



INVESTMENT MAPPING USING GIS AND MULTI CRITERIA 

TECHNIQUES 

Ranya Fadlalla Abdalla ElSheikh 1 

, Noordin bin Ahmad 2 

, Sani Yahaya 3 

1: Geographic Information System, Survey Department, Sudan University for 

Science and Technology, P.O. Box 72, Eastren Daims, Khartoum, Sudan 

rania58@gmail.com 

2: GeoInfo Services Sdn Bhd, 30 Jalan Bandar 2, Taman Melawati, P.O. Box 53100, 

Kuala Lumpur ,Malaysia 

noordin@geoinfo.com.my 

ABSTRACT 

Economic Development organizations now can have the ability to benefit 

from GIS solutions by leveraging this technology to attract and retain business from 

world-wide sources. This paper is designed to produce investment classification map 

which shows the potential of investment in agriculture field. Government agencies 

can use GIS to access to information regarding the potential of investments, and 

minimize investment risks. This study will perform spatial multi-criteria analysis in 

order to rank and display potential places of investment in agriculture. The study 

area selected is the state of Sinnar in Sudan. Whilst the study tries to explore the 

utilization of GIS to map potential investment area, it does not cover a 

comprehensive analysis of all factors that influence the investment of agriculture. 

The analysis is limited to criteria that have spatial reference. The investment criteria 

of spatial analysis are defined from the guidance provided by the Ministry of 

Investment. Even with the shortcomings of getting good data, it was found that the 

result obtained is very encouraging which provides clear indicative areas for 

agricultural investment in Sinnar. 

KEY WORDS 

GIS, Multi criteria, Spatial, Agriculture, Investment 


Although there is great potential in the field of agriculture, development and 

reaping maximum benefit from this sector needs more effort to move the wheel of 

production forward towards improvement and progress (Act, 2003). In 2003 the 


PRODUCTION OF MAGNESIA FROM LOCAL DOLOMITE ORES 

Ministry of Investment in Sudan began to do an investment map for Sudan but the 

project is still under progress. Until now they are still gathering data related to 

investment in GIS software but there is no any analysis done for this data to define 

which the area that have great potential for investment with the minimum risk. One 

of the ways to encourage investment is to provide the investor with good and high 

quality information because wrong information can be costly in the investment. This 

paper shall provide the potential areas in agriculture investment by evaluating them 

based on several types of criteria using Geographic Information System (GIS). 

There are many factors contributing to the success of decision making to yield right 

investment. Two of these factors are: 

1.1 Information 

The correct information is the main condition for any project to success; 

wrong information can cost a lot. In the past people need to search and consume 

their time and money behind finding the answers for many questions such as: Who 

is the owner of the land? Is it suitable for agriculture or not? After that they can face 

many problems such as: 

- Lack of correct data when needed. 

- Maps and data are out of date. 

- Different data sets and maps have incompatible formats, definitions, and 

scale. 

- Combining data and maps for a study is time-consuming and difficult. 

1.2 Technology 

The deal with compatible technology with right people who can use it can 

over come to solve many problems. GIS offers the potential to minimize the abovementioned 

problems and to generate many benefits through its flexibility, speed, 

availability, and processing power. 

1.3 GIS and Multi Criteria application 

GIS application on the technical area in the construction industry is 

evident, while its usefulness to the non-technical area (e.g., business, 

economics) is being explored (Cheng et al, 2005). One of the previous studies 

was to presents utility for shopping mall location selection, which is one of the 

core business activities or in other word improvement of the investment in Hong 

Kong. In the current, study a project is demonstrated to create features 

associated with household incomes, demand points, etc. Queries are then created 

for finding solutions for four location problems: (1) minimum distance, (2) 

maximum demands coverage, (3) maximum incomes coverage, and (4) optimal 

center by using GIS software then the potential location is determined for any 

city by using the mathematical equation to present the geographic center as a 




Ranya Fadlalla Abdalla ElSheikh, et al 

potential location for any local demand density area after determined the 

potential location in different cities the data for any area around the potential 

location is prepared this data include existing of shopping mall. Another 

mathematical expression is set to assume that if a potential location is in close 

proximity to another existing super shopping mall (i.e., within the same district), 

the potential demand for this location will be half. The result of this study 

indicates that Tsimshatsui is the geographical center of Hong Kong with respect 

to all local major demand density areas. Also Kennedy Town is the best site with 

respect to Maximum incomes coverage problem and Tuen Mun is the best 

location as it has the maximum demands coverage. From this, the limitation of 

this study that the results depend on one factor with out consider the selection of 

the best mall through analysis of all factors and give these factors different 

weights. Other study was to perform spatial multi-criteria analysis in order to 

rank and display marketability of thirty-two pay pond businesses in West 

Virginia (Moldovanyi, 2003), The result of this study comparing with the other 

that mention above is more reasonable because it depend on the Evaluation 

Criteria of Spatial data and it consider factors that influence marketability. 

2. METHODOLOGY 

The selection of the criteria was restricted by the available data. The criteria 

that have spatial reference were Land use, Road, Rail way and Water resources. The 

stability of the result will be tested by sensitivity analysis and tradeoffs methods. 

2.1 Ranking Method 

In Ranking Method, every criterion under consideration is ranked in the order 

of the decision maker’s preference. To generate criterion values for each evaluation 

unit, each factor was weighted according to the estimated significance for 

agriculture investment project. The inverse ranking was applied to these factors. 

2.2 Reclassification 

Reclassifying data means replacing input cell values with new output cell 

values. 

The most common reasons for reclassifying landuse are: 

1- To replace values based on new information. 

2- To group certain values together. 

3- To reclassify values to a common scale. 

4- To set specific values to NoData or to set NoData cells to a value. 

Reclassification procedure and ranking Judgments were made in consultation 

with the experts on agriculture for landuse. 

217


2.3 Buffering 

Buffer Wizard was used and the straight line distance function of the Spatial 

Analyst within GIS to obtain values for evaluation criteria. The function Buffer 

Wizard was created for road, railway and water. 

2.4 Map Calculator 

Map calculator provides user with a powerful tool for performing multiple 

tasks. Input can be raster layers, grid datasets, shape files, coverages and tables. 

Mathematical processes are applied to the criteria with Map Calculator. 

2.5 Model Builder 

The Model Builder allows a user to build a model of repetitive spatial 

analyses or data management tasks and save that process as a repeatable tool in GIS. 

A model was created to summarize all the important steps that were been done 

(Figure 2.1). 

Figure 2.1 Model builder 



2.6 Classification Map 



Finally, classification map, which shown the potential of investment in 

agriculture field, was produced. This map will help the investor to choose between 

alternatives and reduce the doubt around it. 

3. CASE STUDY: 

In this paper the agriculture investment opportunities in Sinnar was taken as 

the case study. Sinnar extended over most of the eastern part of the present Sudan. 

Sinnar form a triangular-shaped territory between the White and Blue Niles (Figure 

3.1). The soil, mainly alluvial, is very fertile, and wherever cultivated yields 

abundant crops. 

.. 

SUDAN 

Sinnar 

Figure 3.1 Map showing the study area (Sinnar). 

219


Ranking method was used to evaluate criteria. The first result depends on 

four types of spatial criteria (road, landuse, railway and water recourses). The 

evaluation result concludes that the closer the agriculture project is to a major road, 

railway and water sources the greater is its investment potential. Areas that are less 

or not suitable for agriculture project such as bare rock and shifting sand have the 

lowest investment potential. The results of the analysis indicate that the darkest area 

in figure 3.2 is the most suitable place for investment however, the suitability for 

investment decreases as the area becomes lighter. 

Water resources 

Roads and railways 

Figure 3.2 The potential projects with road and water resources 



3.1 Sensitivity Analysis 



Subsequent to obtaining ranking of alternatives, sensitivity analysis was 

performed to determine the robustness. To be more specific it aims at identifying the 

effect of changing the inputs (weights and criterion scores). The result above shows 

that the change does not significantly affect the outputs, thus ranking is considered 

to be robust. 

3.2 Changing the Criteria to test Stability 

To determine the consistency of the available solutions the criteria was been 

changed to numbers of farmer, facilities (transportation, communication, security 

and electricity) and crop rotation. For this group the number of potential area for 

agriculture investment will be reduced due to lack of attributes data. The evaluation 

result concludes that the existence of facilities and high numbers of farmer beside 

the high level of crop rotation will yield to high potential of agriculture investment 

(Figure 3.3). 

Figure 3.3 Potential area of investment for the second group of criteria 

221


4. CONCLUSION 

The potential areas of agriculture investment in the study area were evaluated 

in four classes. The subjective numbers in the weights and the values of the criteria 

can be changed according to the study area characteristics and experts opinions. The 

conclusions made from this study are: 

When performing the sensitivity analysis on all the criterion weights, it has 

shown that the accuracy in estimating weights needs to be examined carefully. 

Sensitivity analysis helps to see the role of attribute and weight. 

The classification map of agriculture investment projects can be produced by 

using GIS and multi criteria techniques. This map can give planners and investor 

the tool for assessing and minimizing investment risk. 

REFERENCES 

Act. , 2003 The Investment Encouragement, Ministry of Investment, from the World 

Wide Web www.sudaninvest.gov.sd 

Cheng, E., Li , H., and Yu, L., 2005, A GIS approach to shopping mall location 

selection , from the World Wide Web: 

http://www.sciencedirect.com/science 

Florent, J., Marius, T., and MUSY, A., 2001, Using GIS and outranking multicriteria 

analysis for land-use suitability assessment. Geographical information science 

vol. 15, no. 2, 153± 174 

Malczewski, J., 2004, GIS and MultiCriteria Decision Analysis. Department of 

Gegraphy, University of Westren Ontario. 

Moldovanyi, A., 2003 December, GIS and Multi-Criteria Decision Making to 

Determine Marketability of Pay Pond Businesses in West Virginia, West 

Virginia University, Division of Forestry 



COUPLING OF MIKE 11 HYDRODYNAMIC MODEL AND GIS FOR URBAN 

FLOOD MAPPING: A CASE STUDY OF KUALA LUMPUR CITY, MALAYSIA 

A’kif Al-Fuagara 1 , Thamer Ahmed 2 , Abdul Halim Ghazali 3 , 

Salmah Zakaria 4 , Shattri B. Mansor 4 , Ahmad Rodzi Mahmud 6 , 

1: PhD Candidate, 2, 3, 6: Associate professor: Department of Civil Engineering, 

Faculty of Engineering Universiti Putra Malaysia , Malaysia, akifmohd@yahoo.com, 

thamer@eng.upm.edu.my, abdhalim@eng.upm.edu.my, arm@eng.upm.edu.my, 

4: Dr, National Hydraulic Research Institute, Malaysia, salmah@nahrim.gov.my 

5: Prof. Dr, Spatial & Numerical Modeling Laboratory, Institute of Advanced 

Technology, Universiti Putra Malaysia, Malaysia, shattri@eng.upm.edu.my 

ABSTRACT 

Despite Flooding recurrent nature in Malaysia, it is clear that Flood-prone 

areas are still under concentrated development, also there is no holistic effort for digital 

flood risk mapping for an appropriate platform level of floodplain management planning. 

This paper involves the application of flood modeling that integrates hydraulic modeling 

(MIKE 11 hydrodynamic model) and geographic information systems (GIS) for urban 

floodplain inundation. The Kuala Lumpur city is the domain of the study; it is the most 

densely populated area in the country. The form of flooding frequently experienced in 

the study area is the flash flood and it can happen several times each year. In this study 

the flood plain and river geometry of Klang river basin is developed using MIKE 11 

hydrodynamic model oriented GIS. The river network system is incorporated in the 

DEM for hydrodynamic modeling of water level and discharge of 100-year return period 

storm design. 3D-GIS and spatial analytical techniques together with hydraulic data 

processing are performed on ArcView GIS platform to enhance the visualization and 

display techniques for visual presentation and generation of flood inundation maps. 

KEYWORDS 

GIS, Mike11 Hydrodynamic Model, DEM, Flood map. 

INTRODUCTION 

Kuala Lumpur is the capital city of Malaysia. It has been built along the Klang 

and Gombak River’s corridors. Therefore the drainage system is very crucial to channel 

the water out of the city. As the development took root and kept changing the face of the 

city coupling with high intensity monsoon rainfall storms (DID, 2002), existing storm 


COUPLING OF MIKE 11 HYDRODYNAMIC MODEL AND GIS FOR URBAN FLOOD MAPPING: 

drainage infrastructures came regularly under pressure to service demands that were way 

beyond their design limits, as a result, flooding has become very frequent and severe in 

the recent years that have caused great economical losses and disruption to social 

activities (Keizul & Azuhan, 1998; Chan, 2002). A National Water Resources Study in 

1982 estimated the average annual flood damage at RM 100 million (JICA, 1982). 

However, the rapid socio-economic development has led to significant urban expansion 

and escalation of land and property prices, recent studies estimate current annual flood 

damage to be about RM 2.7 billion (DID, 2003). 

Although variety of flood mitigation measures has had different levels of 

success, flooding continues to plague many areas of Klang River Basin. It is 

however becoming clear to planners that it is neither possible nor desirable to 

control floods completely through the structural measures due to technical and 

economic reasons (Billa et al, 2004). A vital element of any floodplain management 

program is the identification and delineation of the flood-prone areas. However, 

management of flood-prone areas is the result of complex decision-making process 

to define all those measures that can compensate the effects of the progressive 

urbanization over hydrological processes. Estimation and Prediction of flood 

inundation is not straightforward since the extended flood inundation depends highly 

on topography, which plays an important role in water flux, distribution and energy 

within natural landscapes. 

For years, floodplain management studies have been expensive and tedious 

task. Recently, in addition to the development of improved computational 

capabilities, the availability of high sophisticated 3D-GIS software continues to 

expand new possibilities for engineers to perform flood inundation analysis in 

conjunction with hydraulic models to represent water surface elevations generated 

from hydrologic and hydraulic models in a three-dimensional terrain model. 

Visualization helps to bridge the gap between the engineers and decision-makers by 

providing a method for exploring, analyzing, and verifying hypothesis from large 

quantities of spatially referenced data (Tate et al., 1999). GIS is being used in a 

variety of ways to plan, prepare, respond, and evaluate flood events, (Sinnakaudan et 

al., 2003; Anrysiak, 2000). On the other hand, water resources management 

applications are inherently spatial and generate large amounts of data relating to 

physical phenomenon in space and time (Koussoulakou, 1994), Thereby, The 

multidimensional nature of this data makes it quite natural to couple hydraulics 

modeling and GIS environment through an appropriate interface. Through the 

hydraulic analyses, modeling simulations and the visualization of floodplain 

inundation decision-makers can obtain improved understanding and make more 

informed management decisions (Bates, 2004). 



2. STUDY AREA LOCATION 


A’kif Al-Fuagara, et al 

The city of Kuala Lumpur (KL) is located at the confluence of the Klang 

and the Gombak Rivers (Figure 1). Basically Kuala Lumpur and the merging Klang 

Valley conurbation are located in a bowl-like topography with an opening into the 

Port Klang coastline to the west. The rest is surrounded by relatively hilly 

topography reaching the foothills of the Main Range to the east. The Study area 

climate is governed by the northeast and southwest monsoons. The northeast 

monsoon blows from December to March and the southwest monsoon from June to 

September. These two main monsoon seasons are separated by two relatively short 

inter-monsoon seasons, which usually recorded heavy rainfall. The annual rainfalls 

vary between 2,000 mm and 2,500 mm and the mean monthly rainfall between 133 

mm and 259 mm. Studies conducted in Klang River showed that the confluence of 

Klang-Ampang rivers and Klang-Gombak in Kuala Lumpur City are the most 

critical and prone to floods. It is further worsened by having a bridge namely Tun 

Perak Bridge (near Masjid Jamek) at the city heart , which causes its surrounding 

areas to be prone to floods as well. 

Figure 1: Study Area Location of Klang River Basin 

225


3. DATA PROCESSING 

To execute floodplain simulation using the integration of hydraulic 

model Mike 11 and GIS, two basic types of data were used to estimate an extent 

of flooding map: spatial and non-spatial data. The spatial Geometry data 

comprise the description of the terrain related to the stream channel, including 

the entire river network, terrain cross-sections along the streams, stream bed 

resistance, and hydraulic structures such as culverts and bridges, where the 

Digital terrain model was produced from interpolation of combination of 

contour heights at interval of 1meter, mass points of spot height and field 

surveyed data of floodplain and river cross section elevation points. On other 

hand, the non-spatial data include the flow data include the values of the stage 

and discharges. These values were imported directly from the Runoff simulation 

was carried out using the MIKE11 NAM model for rainfall-runoff modelling. 

Real-time data can be collected from the study area via the existing hydrometeorological 

networks of Selangor State and Federal Territory of Kuala 

Lumpur. All telemetry stations transmit via radio communication in a real-time 

mode to the master station at DID Federal Territory of Kuala Lumpur. From the 

earlier studies, river system in Klang River Basin only can cater heavy rainfall 

up to 2 to 3 hours (report, did). The major design storm standard recommended 

by “Malaysian Urban Drainage Manual” is 50-years ARI, except for major city 

centres where it is 100 years ARI. Therefore in this study, 3 hours of 100-year 

ARI Rainfall Runoff data has been used to forecast the discharge and water level 

in Klang River Basin according to the requirements of the Malaysian Urban 

Drainage Manual (DID, 2000). 

4. METHODOLOGIES AND MODELING 

4.1 Hydraulic model: 

Mike-11 is a widely used one-dimensional flow routing modelling, developed 

in 1987 by the Danish Hydraulic Institute, which has been used widely to simulate water 

levels and flow in river systems (Danish Hydraulic Institute, 2000; Mishra et al., 2001; 

Hammersmark, 2002). The core to this model is a hydrodynamic module, which solves 

the Saint-Venant equations (non-linear equations of open channel flow), and the 

kinematic wave or diffuse wave simplifications to simulate the flows within branched 

and looped river networks. The hydraulic model requires as input the output hydrographs 

from The MIKE 11/NAM model; the discrimination for each subbasin is provided 

through boundary flow time series, cross-sections, and contraction and expansion 

coefficients, Roughness coefficients, which represent a surface’s resistance to flow and 

are integral parameters for calculating water depth. These can be defined either globally 





or locally for each cross-section. The quality of the results is dependant upon the distance 

between the cross-sections. Due to the regional scale of the model, channel geometry 

was considered only for the major tributaries in the network: of which the Klang, 

Gombak, Batu, Ampang, Kerayong, Kuyoh, Bunus and Damansara. The floodplain 

maps project required a common projection to spatially correspond the unsteady flow 

model results to the GIS environment. Since the Mike 11 unsteady flow model uses an 

XYZ coordinate system, a Cartesian coordinate system is required. The Selangor State 

Cassini Soldner Projection was used throughout the study. To develop a MIKE 11 

hydrodynamic model, five files are necessary: a River Network file, a Cross-section file, 

a Boundary file, a Hydrodynamic Parameter file, and a Simulation file (Figure 2). 

[meter] 1-1-2020 22:00:00 

44.0 

0.0 500.0 1000.0 1500.0 2000.0 2500.0 3000.0 3500.0 4000.0 4500.0 5000.0 5500.0 6000.0 6500.0 7000.0 7500.0 8000.0 

[m] 

Figure 2: Network preparation and water level and discharge Simulation in MIKE 11 

42.0 

40.0 

38.0 

36.0 

34.0 

32.0 

30.0 

28.0 

26.0 

24.0 

22.0 

20.0 

18.0 

16.0 

14.0 

23285 

21169 

20739 

20280 

20046 

19859 

19681 

19561 

19342 

19163 

18982 

SGKLANG 23285 - 16384 

18679 

18431 

18249 

18063 

17717 17745 

17577 

17449 

17385 

17206 17166 

17052 

16860 

16695 

16452 

16384 16425 

16334 

16145 16173 16177 

15899 

15733 

15747 

SGKLANG CELL B03 

[meter] 

24.6 

24.4 

24.2 

24.0 

23.8 

23.6 

23.4 

23.2 

23.0 

22.8 

22.6 

22.4 

22.2 

22.0 

21.8 

21.6 

21.4 

21.2 

21.0 

20.8 

20.6 

20.4 

20.2 

20.0 

19.8 

19.6 

19.4 

19.2 

19.0 

18.8 

18.6 

18.4 

18.2 

18.0 

17.8 

17.6 

SGKLANG 20.739 1/1/2020 11:46:59 PM 

0.0 5.0 10.0 15.0 20.0 25.0 30.0 35.0 40.0 45.0 50.0 55.0 60.0 

[meter] 

500 

0 

227


Another improvement has been added to the hydraulic models inclusion of 

bridges and other hydraulic structures such as weirs and pumps into the flow model. 

Based on the modeled characteristics of the study area’s flood plains, adding bridges 

and hydraulic structures into the model would amplify flood inundation through 

backwater effects from piers and abutments. A lower-resolution of time step 

simulation was performed over the Klang River Basin sub basins. Initial and 

boundary conditions were based upon the inputs for Mike 11 operational analysis. 

Sub-Hourly data for rainfall, water level, and flow were created into compatible 

MIKE 11 time series in a separate file as input to the parameter editors. 

4.2 Terrain Development for Floodplain Delineation 

ArcView GIS 3.2 was used to develop the digital elevation model (DEM) 

for the study area. The integrated MIKE 11 GIS provided extensive tools for 

thorough channel and surface geometry modeling that included floodplain 

processing. Using the DEM module requires that data be processed in MIKE 11 GIS 

established steps, where contour data is converted from shape theme to XYZ file and 

corrected area definitions assigned before DEM is generated. Additional Arcview 

GIS themes were used to further improve the terrain data. Thus, the building and 

street themes of the study area were also provided. 

Figure 3: Flood plain & river geometry 





Existing buildings that lie within a flood plain can create a physical barrier to 

high stage flows; it can divert water flow that may have an impact on the flow model’s 

results. Where, streets can assist an observer unfamiliar with the study area by 

providing a spatial reference for the terrain model. MIKE 11 GIS provides a bidirectional 

interchange between MIKE 11 and Arcview. MIKE 11 GIS can extract 

cross-sectional profiles of the river channels and area-elevation curves from a Digital 

Elevation Model (DEM) and export that data into a MIKE 11 cross-sectional database 

and vice versa. The cross-section point elevations for the main tributes of the study 

area Rivers were collected through various land surveys. Hence, the accuracy of the 

Mike 11 stream geometry data is greater within the channel than in the over bank areas 

(Figure 3). MIKE 11 GIS creates a grid-based water surface, compares it with the 

generated DEM, and produces floodmaps. MIKE 11 simulation results can also be 

displayed as graphs and longitudinal profiles using the MIKE 11 GIS GUI. 

5. FLOOD MAP DEVELOPMENT 

Flood inundation maps were developed using the Flood Management (FM) 

module, which is apart of the MIKE 11 GIS (DHI, 2000). MIKE 11 simulation results 

and network branches were imported to the ArcView GIS and linked to the processed 

DEM. In the MIKE 11 GIS module, it is possible to form flood maps corresponding to 

any time, hour or day during the flooding. At the end, the module gives the highest water 

depths (Hmax) in the riverbed during the flood. Long section profiles are derived from 

flood maps at a point located on the Klang River. The long section profile obtained from 

the results of the predicted 100-year flood simulation is shown in (Figure 2) in the format 

of software output. Flood map of the study area of the simulation period was generated 

in stages that could be loaded in the 3D viewer. (Figure 4) illustrates the 3D visualization 

of flood extent of the 100-year return period storm of 3-hours simulation period. A 

valuable by-product of the flood maps generated with MIKE 11 GIS is that the flood 

depth can be determined base on the color difference. A Flyby simulator in the 3D 

viewer also allows for interactive observation of inundated area. 

In flood studies, not only the extent of flooded areas but also the depth of 

water in this area should be determined to help predict the damage that water will 

cause to both land property . Determining the inundated area in the flood map 

involves the subtraction of flood surface elevation model from the land surface 

elevation model at each location, resulting in negative values wherever the flood 

elevation is greater than the land elevation. In (Figure 5) roads, buildings, or other 

features have been generated to make it easy to assess the flood extent and also to 

determine whether these cultural features are in a flood area. For the current study, 

inundation map are obtained for the River Basin at the end of the simulation made 

for the 100-year flood. The values of highest water depths in different cross-sections 

can be used to determine the heights of embankments that may need to be built 

around the riverbed for protection of the surrounding areas. 

229


Figure 4: Flood map generation in MIKE 11GIS 

Figure 5: Inundated houses and roads in MIKE 11GIS 



6. CONCLUSION 



Floods are uncontrollable natural events causing loss of lives and damage 

to public property. Flood maps produced by GIS allow users to overlay additional 

digital information in the form of roads, buildings, and critical facilities to give as 

much information as possible to land-use planners, local authorities, emergency 

services and the people who may be affected. MIKE 11 hydrodynamic model and an 

integrated ArcView GIS were used to simulate water level, rainfall-runoff and 

generate flood map for the highest water depths obtained in the riverbed for the 

forecasted 100-year floods. In addition, the flooded areas corresponding to the 

maximum water depth (Hmax) is obtained and presented in digital format for visual 

interpretation and further analysis. Generally the study revealed for the strength and 

potential of a hydrological oriented GIS as digital support system for flood 

forecasting and mapping. Results can also be used for the determination of 

economic losses. 

REFERENCES: 

Anrysiak, P.B., (2000). “Visual floodplain modeling with Geographic Information 

Systems (GIS)”. Master Thesis, University Of Texas at Austin, Austin. 

Bates, P.D., (2004). “Remote sensing and flood inundation modeling”. Hydrological 

Processes 18, 2593–2597. 

Billa, L., Shattri, M., Rodzi, A., Noardia, A., (2004). “Spatial information 

technology in flood early warning systems: an overview of theory, application 

and latest developments in Malaysia”. Disaster Prevention and Management, 

13, 5, 22-29. 

Chan, N. W. (2002). “Flood Hazards and Disasters in Malaysia: Causes, Impacts and 

Solutions With Respect To River Floods”. IN NW Chan (Editor) “Rivers: 

Towards Sustainable Development”. University Sains Malaysia, Peneng, 114-127. 

DHI (Danish Hydraulic Institute), (2000). “MIKE 11 Reference Manual”. Appendix 

A: Scientific background, Danish Hydraulic Institute. 

DID (Department of Irrigation and Drainage), (2003). “Final Report on Updating of 

Condition of Flooding in Malaysia”. 

DID (Department of Irrigation and Drainage), (2002). “Magnitude and frequency of 

flood in peninsular Malaysia”. Hydrological Procedure, No. 4, Department of 

Irrigation and Drainage, Kuala Lumpur, available at: 

http://agrolink.moa.my/did/hydro/man_hp/HP4.pdf (accessed March 23, 2007). 

231


DID (Department of Irrigation and Drainage), (2000). “Urban Stormwater 

Management for Malaysia “Drainage Manual”. Department of Irrigation and 

Drainage, Kuala Lumpur, Malaysia. 

Hammersmark, C.T., (2002). “Hydrodynamic modeling and GIS analysis of the 

habitat potential and flood control benefits of the restoration of a Leveed Delta 

Island”. Master Thesis. University of California, Davis, California. 

JICA. (1982). “Water resources development and use plan. National Water Resources 

Study, Malaysia”. Main report. Government of Malaysia. Japan International 

Cooperation Agency. Kuala Lumpur, Malaysia, Volume 2, pp 183. 

Karlsson, K., Thoss, A., Dybbroe, A., (1999). "High resolution cloud product from 

NOAA AVHRR and AMSU". Swedish Meteorological and Hydrological 

Institute (SMHI), Norrköping, available at: 

http://produkter.smhi.se/saf/INM_workshop.pdf (accessed March 19, 2007). 

Keizrul Abdullah and Azuhan Mohamed (1998). “Water – A Situation Appraisal 

and Possible Actions at the Community Level”. Seminar on local communities 

and the environment II, Environmental Protection Society of Malaysia, 

Petaling Jaya, Malaysia. 

Koussoulakou, A., (1994). “Spatio-temporal analysis of urban air pollution. In: A. 

MacEachren and F. Taylor, eds”. Visualisation in Modern Cartography. 

Oxford: Elsevier Science, 243-67. 

Mishra, A. Anand, R. Singh and N.S. Raghuwanshi (2001). “Hydraulic modeling of 

kangsbati main canal for performance assessment”. J. Irrig. Drain. Eng. 127 

(1), pp. 27–34. 

Noman, Nawajish S., E. James Nelson, and Alan K. Zundel, (2001). “A Review of 

Automated Flood Plain Delineation from Digital Terrain Models”. ASCE 

Journal of Water Resources Planning and Management, Vol. 127, No. 6, pp. 

394-402. 

Sinnakaudan, S., Ab Ghani, A., S. Ahmad, M. S., & Zakaria, N, A. (2003). “Flood 

Risk Mapping for Pari River Incorporating Sediment Transport”. Journal of 

Environmental Modeling and Software, Elsevier Science, Vol. 18, No. 2, pp. 

119-130. 

Tate, E.C., Olivera, F., Maidment, D., (1999). “Floodplain Mapping Using HEC- 

RAS and ArcView GIS”. Master Thesis, University Of Texas at Austin, 

Austin. 



INFLUENCE OF HORIZONTAL CONSTRUCTION JOINT ON THE 

FLEXURAL BEHAVIOUR OF REINFORCED CONCRETE SLABS 

ABSTRACT 

Dr. Ibrahim M. Metwally 1 , Dr. Mohamed S. Issa 2 

1: Ass. Prof., Reinforced Concrete Dept., Building Research Centre, 

87, El-Tahrir St., Dokki, Giza, Egypt, E-Mail: im2aa@yahoo.com 

2: Ass. Prof., Reinforced Concrete Dept., Building Research Centre, 

87, El-Tahrir St., Dokki, Giza, Egypt 

Some concrete casting contractors fall in a certain mistake during casting of 

reinforced concrete slabs in field by casting them in two layers. The first layer is 

usually casted up to cover the reinforcement steel mesh to allow the concrete trolley 

(which carries the fresh concrete) to move easily above the rough surface of the slab 

(after hardening) for casting the second layer without any destruction of the 

arrangement of steel mesh. This technique of casting creates a horizontal joint which 

is considered as a plane of weakness which may be subjected to leakage, 

deterioration, and possible failure due to tensile or shear stresses. This paper 

concentrates on the effect of the horizontal joint and the type of the different 

bonding materials (which are used to bond the two layers) on the flexural response 

of reinforced concrete slab. Test results show many surprises as being of horizontal 

joint in reinforced concrete slabs do not affect the performance of slabs in flexure. 

Moreover, jointed slabs with and without bonding materials recorded higher values 

of ultimate load carrying capacities, stiffness, ductility, flexural toughness, and less 

deflection compared with the reference slab (solid slab without joint). 

KEY WORDS 

Horizontal Joint; Slab; Adhesive; Flexural Behavior; Stiffness 

1-INTRODUCTION 

Horizontal lift joints in roller-compacted concrete structures as shown in 

Fig. 1 are planes of weakness subject to leakage, deterioration, and possible failure 

from tensile or shear stresses[1]. 


INFLUENCE OF HORIZONTAL CONSTRUCTION JOINT ON THE FLEXURAL BEHAVIOUR 

Fig. 1- Construction of horizontal joint [1] 

1.1- Codes Requirements for Joint Treatment 

1.1.1-American Concrete Institute: ACI 318-89R [2] 

The requirements of the 1977 code for the use of neat cement on construction 

joint have been removed, since it is rarely practical and can be detrimental where deep 

forms and steel congestion prevent proper access. Often wet blasting and other 

procedures are more appropriate. Since the code set only minimum requirements, the 

engineer may have to specify special procedures if condition warrant. The degree to 

which mortar batches are needed at the start of concrete placement depended on 

concrete properties, congestion of steel, vibrator access, and other factors. ACI Code 

recommended that the concrete surface of construction joint shall be cleaned and 

laitance removed. Immediately before new concrete is placed, all construction joints 

shall be wetted and standing water removed. 

1.1.2- British Standard: BS 8110[3] 

It is necessary for a joint to transfer tensile or shear stresses, the surface of 

the first pour should be roughened to increase the bond strength and to provide 

aggregate interlock. With horizontal joints, the joint surface should, if possible, be 

roughened, without disturbing the coarse aggregate particles, by spraying the joint 

surface, approximately 2-4 hours after the concrete is placed, with a fine spray of 

water and / or brushing with stiff brush. 

1.1.3- Australian Standard: AS 1480-1982[4] 

Before fresh concrete is placed against hardened concrete at a construction 

joint, the joint surface of hardened concrete shall be thoroughly roughened and 

cleaned, so that all loose or soft material, free water, foreign matter and laitance are 

removed. At the time of placement of the fresh concrete, the joint surface of the 

hardened concrete shall be damp but there shall be no free water. 





1.1.4- Indian Standard: IS 456:1978[5] 

For horizontal joints, the surface shall be covered with a layer of mortar 

about 10 to 15 mm thick composed of cement and sand in the same ratio as the 

cement and sand in concrete mix. This layer of cement slurry or mortar shall be 

freshly mixed and applied immediately before placing the concrete. 

1.1.5- Egyptian Code 1990 [6] 

When the work has to be resumed on the horizontal construction joint 

(after more than one day), the surface of the hardened concrete shall be carefully 

scrabbled to expose the coarse aggregate. The surface shall be cleaned and loose or 

soft material removed. The surface shall be then thoroughly wetted. A layer of 

water-cement grout or bonding new to old concrete paint shall be sprayed. 

The aims of this research are to show and discussion the different 

techniques of bonding the horizontal joint in reinforced concrete slab by use of 

several bonding materials, and to explain their effects on the flexural performance of 

slabs. In the present paper, the new used materials for this purpose are: 

1- Combination of water glass, and Portland cement 

Water glass is consisted of several compounds containing sodium oxide 

(Na2O), and silica (SiO2). The solution is strongly alkaline and viscid. 

Thompson et al. [7] concluded that sealing concrete surface with soluble 

sodium silicate may improve surface properties such as hardness, permeability, 

chemical durability, and abrasion resistance. 

Sodium silicate (water glass) is unique in that it can undergo four very 

distinct chemical reactions. These reactions have been defined as: hydration / 

dehydration, gelatin, precipitation, and surface charge modification. These reactions 

allow silicate to act as a: film binder, matrix binder, and chemical binder [8]. 

It is the very quick reaction with Ca +2 that allows silicate to be used as a 

cement accelerator. A problem commonly encountered with using Portland cement 

as a matrix binder is the achievement of sufficient green bond strength. The 

incorporation of silicate into a cement- based formulation will accelerate the set of 

the cement [8]. 

2- Combination of water glass, silica fume, and Portland cement 

Silica fume is a by-product material resulting from the production of 

ferrosilicon alloys. 

Silica fume reacts chemically and pozzolanically with liberated lime during 

cement hydration at ordinary temperature and in the presence of moisture to form a 

strong compound material (calcium silicate hydrates, gel), which has an adhesive 

characteristic [7]. 

235


The benefits of using silica fume as a matrix binder are that it tends to be 

more absorbent of liquids and produce a highly durable product [8]. 

In order to achieve the necessary green strength, silica fume must be 

activated either by heat or chemically by alkali. Silicates (water glass) are widely 

accepted as the best material to activate pozzolans (as silica fume) in general. The 

alkali serves to activate the siliceous material present in pozzolans, and the silica 

portion contributes to the formation of calcium silicate hydrate; a cementitious phase 

that is the matrix binder [7]. 

3- Epoxy Resin 

Polymer adhesives provide a better bond of plastic concrete to hardened 

concrete than can be obtained by relying on the cement itself or on a cement slurry, 

because polymer adhesives shrink less than cement paste upon curing, and because 

they tolerated a wider range of moisture conditions in the plastic concrete and the 

hardened substrate. The primary use of all types of water-borne adhesives with 

concrete is to bond plastic concrete to hardened concrete. The only solvent-free 

adhesives used for bonding plastic concrete to hardened concrete are epoxy 

adhesives because, unlike other solvent-free adhesives, they can be readily 

formulated to cure and bond in the presence of water [9]. 

To ensure adequate bonding using epoxy adhesives, the following 

requirements should be met: [10] 

A-Prepared surface should be strong, dense, and sound. 

B-Prepared surface should be clean and free from such contaminants. 

Epoxy adhesives are supplied as a two-part system, one containing the 

epoxy resin and the other containing the hardened or curing agent. Prior to 

combining the two components, it is recommended that each component is 

thoroughly mixed to ensure uniformity [10]. 

Although the epoxy adhesives provides satisfactory adhesion if the freshly 

mixed concrete is placed immediately after applying the adhesive, the contractor 

should wait for five to ten minutes so that the adhesive can wet the existing surface 

prior to contact with the freshly mixed concrete. The freshly mixed concrete must be 

placed while the adhesive is tacky [10]. 

Six reinforced concrete slabs were tested to investigate the effect of the 

two-step casting procedure and the bonding materials on strength, deflection, 

stiffness, toughness, and failure mode. As well as, making a comparison between the 

effects one-step and two-step casting procedures of reinforced concrete slabs. 



2-EXPERIMENTAL PROGRAM 



2.1- Concrete Materials 

The concrete mix was designed to produce a 28-day cube compressive 

strength of 30 MPa. The mix proportion of 1.0 m 3 was 1224 kg of gravel, 612 kg of 

sand, 350 kg of ordinary Portland cement, and 154 liters of water (w/c=0.44). The 

slabs were reinforced with one bottom smooth steel mesh 6ø8/m (yield strength = 

330MPa) for two directions. 

2.2- Specimens 

The test slab specimens were square with 1100 mm side length and 60 mm 

thickness. The test specimens were simply supported along two edges. Fig. 2 shows 

reinforcement details and dimensions of the slabs. 

Fig. 2- Details of a typical test specimen 

2.3- Casting and Joint Treatment 

All the specimens were cast horizontal. In the one-step procedure, the slabs 

were cast by the traditional method to the full depth of 60 mm. In contrast, the other 

slabs were cast by the two-step procedure; the bottom portion of 30 mm was cast 

first and the top portion of 30 mm was cast in the following day. The bottom layer 

was casted under the direct sun and exposed to the direct sun till casting the top 

layer in the second day as a field simulation. The surface of the first fresh concrete 

layer was kept rough without any leveling to create the bond stresses between the 

two layers. During the first seven days after hardening, all concrete slabs were cured 

by water spraying twice every day. Then the slabs were left under the sun for six 

months till testing. 

237


This paper discusses also the effect of different bonding materials for the two 

layers on the slab performance as follow: 

SS: "Control Slab": Slab was cast to the full depth (without horizontal joint) 

S0: Slab was cast in two layers without binder 

In the following next slabs, different binders were coated between two layers as 

follow: 

S1: Using cement slurry (water: cement; 1:1 by weight) as a binder; 

S2: Using epoxy resin; 

S3: Using a mixture of water glass, cement and water (0.1:1:1 by weight); 

S4: Using a mixture of water glass, cement, water, and silica fume (0.1:0.9:1:0.1 by 

weight). 

2.4- Test Setup 

All the specimens were tested under two line loads using hydraulic jack until 

the failure. The slabs were instrumented with LVDT under midspan (the slab center) 

to monitor deflection. During testing, cracks were marked. Midspan deflection and 

load were recorded for each slab. Fig. 3 shows the test setup. 

Fig. 3- Test Setup. 



3- TEST RESULTS & DISCUSSION 



3.1- Load-Deflection Response 

The load-deflection curves were obtained using LVDT measurements. The 

whole curves were monitored automatically by means of the X-Y plotter connected 

to the testing machine. The applied two line loads versus the deflection at the center 

of the slab for all test specimens are shown in Figures 4, 5, 6, 7, 8, and 9. 

3.1.1-Ultimate Load 

The ultimate load carrying capacities (Pu) of the studied slabs are shown in Table 1 

The maximum ultimate load was obtained by slab S0 (jointed slab without 

adhesive). It attained a noticeable increase in Pu equal 17.5% over the reference slab 

SS (without joint). 

The most useful adhesive bonding material for the jointed slabs were 

developed by using a mixture of water glass , cement, and water as in slab S3 and a 

mixture of water glass , cement, water , and silica fume as in slab S4. They 

recorded an increase in the ultimate load carrying capacity of 13.3% and 12.6% 

respectively compared with the reference slab SS. 

In general, from the above results analysis, it can be estimated that being of 

horizontal joint in R. C. slabs (whether with adhesive or without) improve the 

ultimate load clearly. This is due to that the slab horizontal shear resistance at the 

joint was enhanced (by the friction forces between two layers) compared with the 

slab without joint. 

3.1.2- Deflection 

It is known that, the deflection of R. C. slab is considered as a measure of its 

rigidity; with decrease of deflection, the rigidity increases. 

The deflection values of the slabs in Table 1 show that all the jointed slabs 

deflected less than the reference slab (SS). The drops of deflection of the jointed 

slabs were very clear. For example, at ultimate load, S3 slab recorded the maximum 

decrease in the deflection. It deflected 0.33 times the reference slab deflection. 

3.2- Stiffness 

From the load-deflection curves (Fig. 4 to 9), two values of the stiffness of the 

tested slabs were obtained. The uncracked stiffness Ki is indicated by the slope of the line 

at a value less than the first crack load (1 mm deflection), and the ultimate stiffness Ku is 

measured by the slope of the line at about 90% of the ultimate load[11]. 

Stiffness degradation is defined as the ratio between the ultimate stiffness and the 

uncracked stiffness [11]. Values of Ki, Ku and stiffness degradation are given in Table 1. 

From the deflection curves, it can be seen that the slope of S3 slab is very 

steep. It recorded the maximum value of Ki and Ku with increments of 175 % and 

29 % compared with the reference slab SS. 

Stiffness degradation is considered as a measure of the ductility. As the 

stiffness degradation increases, the slab specimen indicates lower ductility [11]. 

239


The results in Table 1 show that all the jointed slabs have small values of stiffness 

degradation than the reference slab. This lead to more ductile behavior than the reference 

slab SS, which agrees with the experimental results of Mo and Lai [12]. 

The lowest stiffness degradations were obtained by S1 and S3. Therefore, it 

can be concluded that the S1 and S3 slabs have the higher ductilities. For example, 

S3 slab has nearly increase in ductile response by about 53 % than the reference slab 

SS. This phenomenon may be explained as the horizontal construction joints create 

the springy effect between the two layers of the R. C. slabs resulted in improvement 

of ductility. 

Load, t 

Load, ton 

2.5 

2 

1.5 

1 

0.5 

SS Slab 

Pu = 1.947 t 

Def. at Pu = 27.904 mm 

0 

0 5 10 15 20 25 30 35 40 45 50 

Central deflection, mm 

Fig. 4- Load-Deflection Curve of Slab SS 

2.5 

2 

1.5 

1 

0.5 

S0 Slab 

Pu = 2.288 t 

Def. at Pu = 14.069 mm 

0 

0 5 10 15 20 25 30 35 40 


Fig. 5- Load-Deflection Curve of Slab S0 



Load, t 

Load, ton 

2.5 

2 

1.5 

1 

0.5 

2.5 

1.5 

0.5 

S1 Slab 

Pu = 2.015 t 

Def. at Pu =11.274 mm 



0 

0 5 10 15 20 25 30 35 40 


2 

1 


S2 Slab 

Pu = 2.159 t 

Def. at Pu = 10.752 mm 

0 

0 5 10 15 20 25 30 



241


Load, ton 

2.5 

2 

1.5 

1 

0.5 

S3 Slab 

Pu = 2.206 t 

Def. at Pu = 9.331mm 

0 

0 5 10 15 20 25 30 35 40 45 50 


Load, ton 


2.5 

2 

1.5 

1 

0.5 

S4 Slab 

Pu = 2.193 t 

Def at Pu = 13.524 mm 

0 

0 5 10 15 20 25 30 35 40 



3.3- Toughness 

Flexural toughness or energy absorption is defined as the area under the 

load-deflection curve up to a deflection equal to the span length divided by 150[13]. 

The flexural toughness values for various slabs calculated at the designated 

deflection of 7 mm are tabulated and shown in Table 1 and Fig.10 respectively. 

From these, it can be concluded that the jointed slabs have the biggest toughness 

values. The maximum value was obtained by S3 slab and equals 8.1 t.mm. It 

attained 54 % increase in toughness compared with the reference slab. 



Table 1- Results of the tested R. C. Slabs 

Flexural Toughness, t.mm 

Slab 

Code 

SS 

S0 

S1 

S2 

S3 

S4 

9 

8 

7 

6 

5 

4 

3 

2 

1 

0 

Ultimate 

Load, 

Pu 

(ton) 

1.947 

2.288 

2.015 

2.159 

2.206 

2.193 

Ultimate 

Central 

Deflection, 

Initial 

Stiffness, 

Ultimate 

Stiffness, 


∆u 

(mm) 

27.904 

14.069 

11.274 

10.752 

9.331 

13.524 

Ki 

(t / mm) 

0.20 

0.29 

0.45 

0.33 

0.55 

0.37 


Ku 

(t / mm) 

0.21 

0.22 

0.21 

0.24 

0.27 

0.20 

Stiffness 

Degradation, 

Ku / Ki 

1.05 

0.76 

0.47 

0.73 

0.49 

0.54 

SS S0 S1 S2 S3 S4 

Slab Code 

Fig. 10- Flexural Toughness of the Tested Slabs 

Flexural 

Toughness, 

243 

ψ 

( t. mm) 

5.25 

6.35 

6.27 

7.64 

8.1 

6.25


3.4- Cracking and Mode of Failure 

At low level of load, the slab behavior was linear elastic with no crack 

occurrence. As load level increases, the extreme fiber concrete stress reaches its 

limiting concrete tensile stress and hair fine flexure cracks occur. The first crack 

was nearly under the position of the loaded area. As the load increases above the 

first cracking load, the cracks seem to widen and start to propagate, and 

generally extend (to be initiated) and deviate towards the slab boundaries in all 

directions up to failure. 

As the load increases, all the flexural cracks propagate upwards with 

extensive cracking parallel to the longitudinal tensile reinforcing bars which 

develop and extend diagonally toward slab corners. The cracking pattern and 

mode of failure for all tested slabs are shown in Fig.11. From the above 

analysis, it can be concluded that, the mode of failure of solid slab (without 

joint) and jointed slabs are nearly the same and almost all slabs failed due to 

flexural type mode. 

Fig.11 – Crack Pattern at Failure of the Tested Slabs 



CONCLUSIONS 



Conclusions related to the influence of horizontal construction joint on flexural 

behavior of reinforced concrete slab: 

1- Being of horizontal construction joint inside the reinforced concrete slab 

improves its ultimate load carrying capacity compared with reference slab 

without joint. 

2- R. C. slabs with horizontal joints attained a clear reduction in deflection. 

3- R. C. slabs with horizontal joints exhibited stiffer than the slab without joint. 

4- Jointed slabs have more ductile behavior than the reference one. 

5- Horizontal joint in R. C. slab is very effective in improving the toughness. 

6- Mode of failure of all slabs with and without horizontal joints is nearly the same. 

Conclusions related to the influence of bonding materials type upon the flexural 

behavior of jointed slabs: 

1- A mixture of water glass, cement, and water (0.1:1:1 by weight) as in slab 

S3 is the best adhesive for bonding the two horizontal layers of the jointed slab. This 

adhesive technique attained the superior flexural response of R. C. jointed slab 

regarding minimum deflection and maximum ultimate load, stiffness, ductility and 

toughness. 

RECOMMENDATIONS 

If the R. C. slabs can not be directly cast to the full depth in the field due to 

construction requirements or problems. It can be cast by the two-step procedure; the 

half bottom portion can cast first and the top portion will cast in the following day 

with using a mixture of water glass, cement, and water (0.1:1:1 by weight) as an 

adhesive coating compound between the two layers. 

FUTURE RESEARCH TOPICS 

It is recommended that, some studies must be carried out regarding: 

1- Effect of horizontal construction joint with using the recommended adhesive 

compound on the behavior of R. C. slab exposed to: high temperatures, chemical 

attack, and drying and wetting cycles. 

2- Effect of horizontal construction joint with using the recommended adhesive 

compound on the flexural and shear behavior of R. C. beams. 

245


REFERENCES 

1- Hess, J. R., Jan. 2002, "RCC Lift-Joint Strength", Concrete International, pp. 50-56. 

2- ACI Committee 318R-89, 1989, "Commentary on Building Code Requirements 

for Reinforced Concrete". 

3- British Standard: BS 8110:1985, "Structural Use of Concrete", Part 1, Code of 

Practice for Design and Construction of Construction Joint. 

4- Australian Standard: AS: 1480-1982. 

5-Indian Standard: IS: 456-1978. 

6-Egyptia Code of Practice for Reinforced Concrete Structures: Construction and 

Design, 1990. 

7-Thompson, L. R.; Silsbee, M. R.; Gill, P. M.; and Scheetz, B. E., 1997, 

"Characterization of Silicate Sealers on Concrete", Cement and Concrete 

Research, Vol. 27, No. 10, , pp. 1561-1567. 

8-McDonald, M.; and Thompson, L. R., 2005 "Sodium Silicate as a Binder for the 21st 

Century", Report of The PQ Corporation: Industrial Chemicals Division, 6pp. 

9-ACI Committee 503.5R, 1989,"Giude for the Selection of Polymer Adhesives 

with Concrete", 15pp. 

10- ACI Committee 503.6R, 1997, "Guide for the Application of Epoxy and Latex 

Adhesives for Bonding Freshly Mixed and Hardened Concretes", 4pp. 

11- Osman, M.; Marzouk, H.; and Helmy, S., May-June 2000, "Behavior of High- 

Strength Lightweight Concrete Slabs under Punching Loads", ACI Structural 

Journal, V. 97, No. 3, pp. 492-498. 

12- Mo, Y. L.; and Lai, H. C., July-August 1995," Effect of Casting and Slump on 

Ductilities of Reinforced Concrete Beams", ACI Structural Journal, V. 92, No. 

4, pp. 419-424. 

13- Japanese Concrete Institute, 1984," JCI Standard for Test Methods of Fiber 

Reinforced Concrete," Report No. JCI-SF-1984, 68pp. 



BUILDING MORE DURABLE ASPHALT PAVEMENT IN THE KINGDOM 

OF SAUDI ARABIA USING SUPERPAVE SYSTEM 

Mohamed S. Aazam ١ , Al-Hosain M. Ali ٢ , Saleh Alswailmi 3 

1. Research Director, Materials & Research Dept., MOT, P.O. Box 28696, Riyadh 

11147, KSA, aazam@mot.gov.sa 

2. Assistant Prof. of Highway Eng., Dept. of Civil Eng., Mansoura Univ., EGYPT 

35116. Superpave Committee Member, MOT, KSA, aali0418@yahoo.com. 

3. Chairman of Gulf Eng. House, Riyadh, KSA, Swailmi@daralkhalij.com 

ABSTRACT 

Interest in the Superpave performance-based mix design and analysis system, 

developed through the asphalt research program of the Strategic Highway Research 

Program (SHRP), is growing throughout the Kingdom of Saudi Arabia (KSA). 

Ministry of Transport (MOT) is actively gearing up for Superpave implementation. 

Since 2005, projects are being paved with Superpave Hot Mixtures Asphalt (HMA) 

across the Kingdom including Dualization of Jeddah-Jazan highway (sections 9, 10, 

and 12) and Dammam–Abu Hedreya Expressway. This paper analyzes results for all 

Quality Control and Quality Assurance (QC/QA) procedures conducted during 

mixture production and pavements construction of the above projects. Analysis 

included volumetric characteristics and criteria. Also, mixtures aggregate structure 

and restricted zones were examined. 

One year after opening the above projects to traffic, Superpave HMA was 

evaluated by testing core and slab samples taken from pavements in both wheel path 

and lane center. Volumetric and dynamic characteristics for field samples were 

evaluated. Based on highway field conditions and results, pavements were 

performing as would be expected. However, mixture gradation, Superpave design 

volumetric criteria in terms of design air void, and the need for Simple Performance 

Test (SPT) are critical aspects in Superpave system design and constructed 

pavement performance. 

KEY WORDS 

Superpave, Performance, Restricted zone, Pavement conditions 



INTRODUCTION 

The use of the Superpave process and technology is rapidly growing 

with many projects planned in KSA. As with any new technology during the 

implementation phase, fine-tuning the technical elements and results demands 

constant attention and technical evaluation. The MOT has paved its first 

Superpave projects in Makkah, Jazan and Eastern regions early 2005. This 

article summarizes the technical analysis of those projects based on collected 

data during construction and one year later as well. A summary of the 

Superpave technology is given next. 

In 1986, the American Association of State Highway and 

Transportation Officials (AASHTO) provided research plans for a program of 

strategic research on asphalt mixture design methods. In 1987, the U.S. 

Congress funded the Strategic Highway Research Program (SHRP). In 1993, 

SHRP was completed with the release of 130 research products for 

implementation. Superpave (Superior Performing asphalt Pavement), a 

principal product of SHRP, is a system of standard specifications, test 

methods, and engineering practices that enable the appropriate materials 

selection and mixture design of hot mix asphalt to meet the climatic and traffic 

conditions of specific roadway paving projects. Superpave considers the three 

major pavement distresses; rutting; fatigue cracking and low temperature 

cracking. 

The concept behind the Superpave mix design system is 

straightforward: use available materials to prepare a mix design that achieves a 

level of performance appropriate to the demands of traffic, environment, 

structure, and reliability on the pavement. The Superpave mix design system 

contains three different levels of design, termed level 1, level 2, and level 3. 

The principal tool of volumetric mix design (level 1) is the Superpave gyratory 

compactor (SGC), Figure 1. A satisfactory mix design is one that meets exact 

volumetric requirements at initial and design levels of gyrations (Nini and 

Ndesign). These levels are determined by the total traffic, expressed in 

Equivalent Single Axle Load (ESAL), expected on the pavement over its 

projected service life. The SGC operates by applying constant vertical pressure 

of 600 Kpa, angle of gyration of 1.25 degree, and rotational speed of 30 rpm to 

an asphalt mixture during the compaction process. Superpave volumetric 

design procedure is detailed in AASHTO MP2. 




Mohamed S. Aazam, et al 

Figure 1 Superpave Gyratory Compactor (SGC) & Compaction Mechanism 

PROJECTS LOCATION AND DESCRIPTION 

The Jeddah-Jazan four-lane divided highway is located in Makkah region. 

Superpave HMA 12R24YC (as per the MOT Superpave HMA coding system) 

surface course with 5 cm thickness was used on top of 7 cm Marshall HMA base 

layer for sections 9, 10, and 12. Sections 9 and 10 were located in the south bound 

travel way. Total length for section 9 was 70 kilometers (station 00+000 to 70+000). 

Section 10 has a total length of 70 kilometers (station 70+000 to 140+000). Section 

12 extended in the north bound travel way from station 512+000 to 567+000 with a 

total length of 55 kilometers. 

Dammam-Abu Hedreyaa 6-lane divided Expressway is located in the Eastern 

Region. Superpave 12R55YC top course HMA with a thickness of 5 cm was used on 

top of a 7 cm Superpave 25R55YC HMA base layer in a major rehabilitation project 

between stations 371+500 and 430+000. Layouts of pavement cross sections for 

investigated projects are illustrated in Figure 2. 

5 cm 12R24YC Superpave HMA 5cm 

7 cm Marshall Base HMA 

7 cm 

20 cm 

Aggregate Base Course (ABC) 

12R55YC Superpave HMA 

25R55YC Superpave HMA 

20 cm 

Aggregate Base Course (ABC) 

30 cm Subbase 30 cm Subbase 

Jeddah-Jazan (Sections 9, 10, 12) Dammam–Abu Hedreyaa 

Figure 2 Pavement Cross Sections Layout for Superpave Projects 

249


MATERIALS USED AND TESTING PROGRAM 

Materials used in those projects were obtained from approved sources and 

meet all requirements of MOT specifications. Performance Grade binder, PG76-10, 

was obtained by blending virgin bitumen (PG64-10) with 3.0% SBS polymer for 

Jeddah-Jazan projects and with 5% Eva polymer for Dammam-Abu Hedreyaa project. 

Following the MOT specifications, Superpave 12R24YC and 12R55YC were 

designed by MOT and King Fahd University of Petroleum and Minerals, 

respectively. This article analyzes all QC/QA results obtained during production and 

laydown of Superpave Surface Mixtures for Jeddah-Jazan and Dammam-Abu 

Hedreyaa projects (12R24YC and 12R55YC) including; volumetric properties 

(%AV, %VMA, % VFB, and %Gmm at Ninitial, Ndesign, and Nmaximum); mixtures 

gradation; and binder contents. Field densities that were evaluated after pavement 

laydown and compaction have been examined. 

Great attention was paid to those early MOT Superpave projects. 

Pavements were closely monitored and observations were properly documented. A 

comprehensive testing program was carried out one year after projects were opened 

to traffic. Samples were randomly collected from pavements. Cores from wheel path 

and lane centerline were obtained (Photo 1). Also, slab samples were cut and tested. 

Table 1 shows collected samples types and statistics. Slab tests included; extraction 

and PG content evaluation (ASTM D-2172), gradation of extracted aggregates 

(ASTM C-136), and maximum theoretical specific gravity (ASTM D-2041). 

Collected field cores evaluation tools involved; compacted specimen thickness 

(ASTM D-3549), bulk specific gravity and density of compacted mixtures (ASTM 

D-2726), determining degree of pavement compaction (AASHTO T230-1), and 

determining the resilient modulus using indirect tension (AASHTO TP31-96). 

Project 

Table 1 Field Samples after one Year of Construction 

Number of Samples 

Section Wheel 

Path 

Core 

Lane 

Center 



40 cm x 40cm Slab 

Lane Center 

Projec 

t 

Lengt 

h (km) 

Jeddah-Jazan 

9 

10 

10 

8 

10 

8 

10 

8 

70 

70 

12 8 8 8 55 

Dammam-Abu Hedreyaa 10 10 2 58.5 

Total 36 36 28

RESULTS AND ANALYSIS 



Results for QC/QA process conducted during Superpave HMA production 

and laydown were obtained and analyzed. Also, results for filed samples taken one 

year after pavements were opened for traffic were examined. Results include: 

a. Mixture Gradation and Binder Content 

Standard Deviation Distribution (AASHTO R 9-97) was used to 

statistically analyze results uniformity by determining its Percent Within Limit 

(PWL). Gradation and binder content results for QC/QA process conducted 

during production and laydown proved uniformity and compliance with 

specifications with PWL of 100 percent. However, field samples result one year 

after pavements were opened for traffic were, to some extent at some parts, 

inconsistent with QC/QA results. Collected field slabs gradation and binder 

content were evaluated. Average gradations were within required specification 

control points and pass below the restricted zone as shown in Figure 3. Based on 

Job Mix Formula (JMF) tolerances, mixture gradations PWL are given in Table 

2. Gradation inconsistency has been mostly occurred within the coarse portion 

(retained on sieve #8). Section 12 of Jeddah-Jazan highway gave the lowest 

PWL for extracted aggregate mixture gradations. Binder content PWL values for 

sections 9, 10, and 12 were 86, 76, and 92, respectively. Higher PWL for binder 

content and gradation of Dammam-Abu Hedreyaa Superpave mixtures might be 

due to fewer tested samples. 

b. Volumetric Characteristics 

Production and laydown QC/QA results including Superpave mixtures 

volumetric characteristics were evaluated and presented in Table 3. Calculated 

parameters include; %Gmm at variable gyration levels (Ninitial, Ndesign, and Nmaximum); 

% AV (Air Voids); % VMA (Voids in Mineral Aggregates; and %VFA (Voids 

Filled with Asphalt). Average values for volumetric characteristics meet 

specification limits. Relatively lower PWL for percent air void was due to higher 

variation in results with individual measurements close and exceeding upper 

production specification limit (5.5%). inconsistency in mixture gradations and 

binder contents gave variable air void contents. Also, variation in VFA values gave 

lower PWL for section 12 of Jeddah-Jazan highway. PWL for individual volumetric 

parameters are shown in Table 3. 

251


% Passing 

100 

90 

80 

70 

60 

50 

40 

30 

20 

10 

0 

Specification Control Points 

Specification Restricted Zone 

Jizan-Jeddah, Section 9 



Dammam-Abu Hedreyaa 

0.01 0.1 1 

Sieve Size-Log Scale-mm 

10 100 

Figure 3 Average Mixture Gradations versus Superpave Criteria 





Table 2 JMF, Average Gradation, and PWL Based on Production Tolerances 

Road Section 

Jeddah-Jazan 

9 

10 

12 

Dammam 

Abu 

Hedreyaa 

BC, 

% 

3/4" 1/2" 3/8" 

# 

4 

Percent Passing 

JMF 5.50 100 91.1 76.5 51.5 35.0 23.5 15.8 11.1 8.3 5.6 

1 Yr. 5.17 100 91.7 78.8 53.9 35.0 22.9 16.0 11.3 8.0 5.8 

PWL 68 100 97 81 73 92 100 100 100 100 100 

JMF 5.50 100 91.1 76.5 51.5 35.0 23.5 15.8 11.1 8.3 5.6 

1 Yr. 5.20 100 91.4 79.1 52.9 34.7 23.2 16.4 11.6 8.3 6.1 

PWL 76 100 100 100 98 100 100 100 100 100 100 

JMF 4.71 100 89.5 75.7 50.3 31.5 21.0 15.3 10.6 7.3 5.4 

1 Yr. 4.63 100 94.5 83.5 49.2 31.2 21.5 15.9 10.8 9.1 6.6 

PWL 92 100 100 72 82 94 100 100 59 93 50 

JMF 4.20 100 100 83.8 43.7 30.1 20.1 16.2 12.3 9.2 5.0 

1 Yr. 4.26 100 100 82.1 46.5 31.4 23.0 18.0 12.3 7.5 5.2 

PWL 100 100 100 94 100 100 100 100 100 100 100 

Table 3 QC/QA Production Volumetric Characteristics and PWL 

Road Section @ 

Jeddah-Jazan 

9 

10 

12 

Dammam-Abu 

Hedreyaa 

Nini 

% Gmm 

@ 

Ndes 

# 

8 

@ 

Nmax 

# 

16 

# 

30 

# 

50 

253 

# 

100 

Volumetric Properties 

% AV 

% 

VMA 

% 

VFA 

JMF 87.5 96.0 97.1 4.2 15.1 73.7 

QC/QA 86.7 95.1 96.4 4.9 15.8 68.9 

PWL 100 92 100 92 100 100 

JMF 87.5 96.0 97.1 4.2 15.1 73.7 

QC/QA 86.9 95.6 96.7 4.4 15.1 71.0 

PWL 99 100 100 100 100 100 

JMF 87.9 96.0 97.1 4.0 14.2 72.3 

QC/QA 86.6 95.0 96.2 5.0 15.1 66.6 

PWL 100 93 100 94 100 98 

JMF 86.3 95.9 97.2 4.1 15.2 73.3 

QC/QA 87.3 95.8 96.9 4.2 15.4 72.6 

PWL 100 100 100 100 100 91 

# 

200


Improving the consistency of Superpave HMA requires close monitoring 

and analysis of the AV%. The air void is a critical volumetric parameter and is 

significantly affected by the portion of fine materials (Passing #8 sieve) in the 

aggregate mixture. Production QC/QA results were analyzed by drawing 

relationship between produced air void and passing #8 sieve, Figure 4. Based on 

results, higher percent passing sieve #8 gave lower air void contents. Mixtures were 

produced with air void contents close to the upper specification limits (5.0%) and 

that may increase the pavement tendency for raveling and higher permeability. Since 

excessive infiltration of water into the pavements can deteriorate properties of both 

surface and subsurface layers, water permeability is an important factor in design 

and construction of Superpave pavements. Since the introduction of coarse graded 

Superpave mixes, there has been some concern regarding excessive permeability of 

HMA. This concern has primarily been due to the presence of relatively higher 

amounts of interconnected air voids in coarse graded Superpave mixes. In addition, 

segregation of coarse and fine aggregates during placement of asphalt mixture 

increases the permeability and may threaten the durability of pavement. 

c. Resilient Modulus Results 

Superpave mix design procedures do not include any testing to evaluate a 

design asphalt mixture in terms of performance. Without performance-related tests 

in a Superpave level 1 system, users became increasingly uncomfortable, relying 

strictly on mixture volumetric properties and component material requirements. A 

simple performance test may be considered to enhance Superpave design level 1. 

The repeated load indirect tensile test simulates the state of stress in the lower 

position of the asphalt layer, (i.e., tension zone). The resilient modulus may be used to 

evaluate mixture properties that are directly related to distresses. Field cores taken 

from wheel paths, one year after construction, were evaluated for resilient modulus 

following AASHTO TP31-96. Testing temperature was 25 o C. A haversine repeated 

load with amplitude of 1770 N and frequency of 1 Hertz (load duration of 0.1 second) 

was used during testing. Data were acquired using a data acquisition system and total 

resilient modulus was determined (Table 4). Average total resilient modulus for 

sections 9, 10, and 12 were consistent and in the neighborhood of 5400 Mpa. 



% Air Voids 

6.0 

5.5 

5.0 

4.5 

4.0 

3.5 

3.0 

2.5 

2.0 

Figure 4 Air Voids (%) versus Percent Passing Sieve #8 



16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 

% Passing Sieve #8 

Table 4 Total Resilient Modulus for Jeddah-Jazan Superpave surface Mixtures 

Section Average Total Resilient Modulus (Mpa) 

9 5252 

10 5364 

12 5573 

The relationship between the total resilient modulus and mixture air voids 

is illustrated in Figure 5. As air void increases, the total resilient modulus appears to 

decrease. Wide variability of individual measured resilient modulus values was 

obtained and might be due to inconsistency in mixture gradations and binder 

contents. 

255


d. Field Compaction 

Pavement compaction was evaluated right after construction and one year later 

after opening for traffic. Pavement densification due to traffic was evaluated along the 

wheel path and lane centerline. MOT percent compaction specifications range is 92-95 

(percent by the Gmm). Right after construction field compaction QC/QA results proved 

PWL of 100 percent. However, after one year cores proved inconsistent PWL values 

compared to those of QC/QA outputs. Wheel-path cores showed slightly higher bulk 

specific gravities than those for cores in lane centerline (Table 5). For Jeddah-Jazan 

highway, overall average bulk specific gravities of pavement in wheel-path and lane 

centerline were 2.433 and 2.416, respectively. In other words, Superpave mixtures in 

wheel-path and lane centerline have been densified by percentages of 1.80 and 1.09 

(percent by average field bulk specific gravity right after construction), respectively. 

Lower PWL may be due to inconsistent pavement compaction and incompliance with 

MOT specification range (92-95%). Dammam-Abu Hedreyaa expressway had relatively 

low average compaction values. This may be due to the use of different raw materials 

including mineral and steel slag aggregates with variable proportions in Superpave 

designs. Without appropriate quality control and existing variability in steel slag and 

mineral aggregate specific gravities and proportions, misleading and inappropriate 

compaction results may be obtained. Photo 1 shows different types of aggregates used 

(from left to right: 1-steel slag, 2-mineral aggregate, and 3-mineral aggregate). First two 

types were used in Dammam-Abu Hedreyaa Expressway while the third aggregate type 

was used in Jeddah-Jazan highway. 

Total Resilient Modulud, Mpa 

9000 

8000 

7000 

6000 

5000 

4000 

3000 

2000 

3.0 3.5 4.0 4.5 

% Air Voids 

5.0 5.5 6.0 

Figure 5 Total Resilient Modulus versus Air Voids Content 





Table 5 Statistics of Field Bulk Gs and Compaction One Year after Construction 

Road Section Parameter 

Jeddah-Jazan 

9 

10 

12 

Dammam-Abu 

Hedreyaa 

Values and Statistics 

QC/QA Wheel path Lane center 

Bulk Gs 2.390 2.429 2.380 

% Compaction 92.8 94.1 92.2 

PWL 100 46.0 43 

Bulk Gs 2.389 2.404 2.415 

% Compaction 92.3 93.2 93.6 

PWL 100 63.0 62 

Bulk Gs N/A 2.466 2.454 

% Compaction N/A 92.5 92.1 

PWL N/A 77.0 94 

Bulk Gs 2.840 2.824 2.454 

% Compaction 92.4 91.9 93.8 

PWL 74 59.0 50 

e. Pavement and Field Performance 

The pavement was closely monitored and observations were properly 

documented. Photo 2 shows pavement conditions after one year of construction. The 

Superpave pavement was performing as would be expected for 1-year old overlays. 

No sign for any distresses was observed. Dammam-Abu Hedreyaa expressway 

pavement might face future raveling due to lower compaction and binder content. 

a. Field Sampling b. Cores & Aggregate Types 

Photo 1 Pavements Sampling and Aggregate Types 

257


a. Dammam- Abu Hedreyaa b. Jeddah-Jazan 

Photo 2 Pavement Conditions One Year after being Open to Traffic 

CONCLUSIONS 

Since 2005 projects are being paved with Superpave HMA across the 

Kingdom including Dualization of Jeddah-Jazan highway (sections 9, 10, and 12) 

and Dammam–Abu Hedreya Expressway. QC/QA results were analyzed and proved 

compliance with specifications. Pavement conditions were evaluated after one year 

of opening to traffic. Core and slab samples were taken from pavements in both 

wheel path and lane center. Cores and slab samples results were to some extent 

inconsistent with those of QC/QA procedures. Future filed performance monitoring 

should include sufficient number of samples for better representation and PWL 

calculation. Testing program for quality control and acceptance of Superpave 

mixtures should be investigated and implemented. Superpave design volumetric 

criteria in terms of design air voids, and the need for Simple Performance Test (SPT) 

are critical aspects in Superpave system design and constructed pavement 

performance. 

LIST OF REFERENCES 

1. Kingdom of Saudi Arabia, Ministry of Transport, General Directorate for 

Materials and Research, 2006 “Hot Asphalt Mix Design System, Based on 

Superpave System as Detailed in Asphalt Institute SP-2/2001”. 

2. Al-Hosain M. Ali and Robert L. Harding, 2006 “Evaluation of Superpave 

Asphalt Mixtures: Case Study” 5 th International Engineering Conference, Sharm 

El-Sheik, March 27-31, Egypt. 





3. Goulias D.G., Al-Hosain M. Ali, and J. Crauss, 1997 "Moisture Susceptibility 

of Asphalt Rubber Modified Mixtures and Comparison with Conventional 

Asphalt Mixtures," In RILEM Mechanical Tests for Bituminous Materials, pp. 

429-436, A.A. Belkema, Rotterdam, Netherlands. 

4. National Academy of Science, 1995 “Strategic Highway Research Program 

(SHRP) Library” Reports A-304, A407, A-408, and A-410. 

5. W. S. Mogawer, 2002 “Evaluation of Permeability of Superpave Mixes” the 

New England Transportation Consortium, NETCR 34 Project No. NETC 00-2. 

6. D. Larsen, 2003 “Demonstration and Evaluation of Superpave Technologies: 

Final Evaluation Report for Connecticut Route 2” Report No. FHWA-CT-RD- 

2219-F-02-7, Connecticut Department of Transportation, USA. 

259




USING ACCESS MANAGEMENT TO REDUCE TRAFFIC CONFLICT 

ABSTRACT 


Department of Civil and Architectural Technology 

Riyadh College Technology 

P.O. Box 90359 Riyadh 11613, Kingdom of Saudi Arabia 

mtmallah@yahoo.com 

Traffic accidents, congestions, delays and safety are the main issues in Saudi 

Arabia and in the most parts of the world as well. Many agencies in the government 

or in the privet sector are working hard to find a way to solve those problems. The 

main goal is to increase safety and to decrease crashes and delays. There are many 

ways and tools to reach this goal. One of the tools is access management; where the 

technique of access management can be applied with low cost and high efficiency. 

Using Access management techniques can increase safety for more than 50% in 

some locations. Access management is much more implemented than driveway 

regulation; it is the systematic control of the location, spacing, design and operation 

of driveways, median openings, interchanges, and street connections. It also 

encompasses roadway design treatments such as medians and auxiliary lanes, and 

the appropriate spacing of traffic signals. Introduction of traffic conflict types and 

access management techniques will be discussed in this paper with examples of the 

application of the access management techniques. Finally, the locations and the 

proper techniques will be presented in this paper. 

KEY WORDS 

Access Management, Traffic Conflict, Safety, Intersections. 

INTRODUCTION 

The majority of intersections in the Kingdom of Saudi Arabia and in the most 

parts of the world are unsignalized and the traffic accidents mostly occurred at those 

intersections. Recognizing the safety problems and the importance of reducing 

accidents, highway planners and engineers work very hard to get all data that lead to 

reaching the safety goal (Sayed, Brown and Navin, 1994). Traffic conflicts which 

cause accidents often happen at road intersections where drivers are required to 

perform more driving maneuvers within a shorter time period than are necessary on 

roadway segments between intersections. Parker and Zeeger defined traffic conflict 



as an event involving two or more road users in which the action of one user causes 

the other user to make an evasive maneuver to avoid a collision. Most of the traffic 

conflicts occurred due to conflicting maneuvers at unsignalized intersections where 

design and operation is comprised of a complex set of parameters including number 

of traffic lanes, traffic volumes, spacing between intersections, medians, speed and 

turning movements. Unsignalized intersections are complex to analyze because their 

capacity and delay depend on driver, vehicle, roadway characteristics and 

environmental conditions. They are a major source of vehicular conflict resulting in 

delay, congestion, and accidents (Katamine and Hamarneh, 1998). Improving the 

design and operation of unsignalized intersections can reduce user cost in delays and 

accidents. Improvement in design and operation largely depend on applying the 

suitable access management techniques. The safety benefits of access management 

techniques have been attributed to reduction in traffic conflict points, improved 

access design, and higher driver response time to potential conflicts. 

Unsignalized intersections are junctions were a driver could across through or 

make one of the three known maneuvers: right turn, left turn or u-turn. These 

intersections are usually controlled by stop signs. At unsignalized intersections, left 

turn movements from a driveway to a major road pose many problems and a large 

number of conflict points. The higher the number of conflicts, the greater the 

potential for accidents and a related goal is to reduce these accidents by reducing the 

number of conflicts. Because a driver can only handle one conflict at a time, access 

management is used to separate the remaining conflict points that cannot be 

eliminated. According to the roadway safety and operations problem, specific access 

management treatments are applied according to many variables such as location, 

volume, speed, geometric design and land use. 

Of all the traffic accidents that occurred in Saudi Arabia, more than 35 

percent occurred at an intersection or were intersection-related (Traffic Road 

Department Report 1426). One possible reason for this is that at an intersection, 

drivers must be aware of many conflicts, since there are many possible maneuvers 

that might be made by other drivers. An intersection is basically an access point that 

allows people to ingress to or egress from a road. 

Much progress has been made in reducing the number of deaths and serious 

injuries on the United States highways using access management techniques. In 

2005, the fatality rate per 100 million vehicle miles of travel fell to a new historic 

low of 1.4, down from 1.6, the rate from 1997 to 1999. The1990 rate was 2.1 per 

100 million vehicle miles traveled The cost in 2005 consists of the fact that 39,715 

people were killed in the estimated 6,394,000 police reported motor vehicle traffic 

crashes, 3,189,000 people were injured, and 4,286,000 crashes involved property 

damage only (FARS, GES and NHTSA, 2006). 



TABLE 1 Numbers of Crashes Related to Location, USA, 2000. 

Relation to Junction None 

Traffic Control Device 

Traffic 

Signal 

Stop 

Sign 



Other/ 

Unknown 

Total 

Non--Junction 25594 42 209 1011 26856 

Junction -- Intersection 1822 2254 2928 174 7178 

Junction -- Intersection 

Related 

574 449 223 50 1296 

Other/Unknown 1375 40 64 600 2079 

Total 29365 2785 3424 1835 37409 

The growth of cities and towns generated more traffic and changed land use 

patterns as new developments and businesses have been built, which makes it more 

likely for a driver to be involved in an accident, especially at access points. At 

unsignalized intersections many other operational problems occur including 

congestion, travel time, and delay. 

TRAFFIC CONFLICT 

A conflict is an observable situation in which two or more road users 

approach each other in space and time for such an extent that there is a risk of 

collision if their movements remain unchanged (Amundson and Hyden, 1977). 

Traffic conflicts are measures of accident potential and operational problems at a 

highway location. Many highway agencies are now using traffic conflict techniques 

to complement the limited accident data found in accidental records. The use of 

traffic conflict techniques has, to date, been primarily limited to intersections. Poor 

signing and inadequate geometric design often cause erratic maneuvers. A variety of 

observation methods have been developed to measure traffic conflicts (Allen and 

Cooper, 1978). One of the most important aspects to consider when using conflict 

data is the reliability of data collected by observer 

263


ACCESS MANAGEMENT 

Access management is balancing the competing needs of traffic movement 

and land access. Access management is a relatively new method to solve the 

problems of congestion, capacity loss, and accidents along the national roadways. It 

is the process that provides or manages access to land development while 

simultaneously preserving the flow of traffic on the surrounding road systems in 

terms of travel times, speed and safety. It applies traffic engineering principles to the 

location, design and operation of access drives serving activities along the highway 

(Stover, 1994). 

The safety benefits of access management techniques have been attributed to 

reduction in traffic conflict points, improved access design, and higher driver 

response time to potential conflicts. Many access management techniques were 

implemented to improve the traffic operation and safety along the arterials. Access 

management can improve safety and traffic operations using one of its techniques. 

PROBLEM STATEMENT 

Traffic accidents are the most direct measure of safety for a highway location. 

A traffic conflict is an observable situation in which two or more road users 

approach each other in space and time to such an extent that there is a risk of 

collision if their movements remain unchanged. Because the situations are 

observable and happen at high frequency, relative to that of accidents, conflicts are 

an enticing traffic measure. Despite such differences, most traffic engineers and 

analysts believe that traffic conflicts are of value in describing or identifying safety 

problems at unsignalized intersections. 

Traffic conflict studies have been accepted because of several reasons. First, 

data can be collected in a short period of time, which impacts the ethical issue that 

an engineer does not have to wait for the occurrence of several accidents to improve 

the conditions of a site (Parker and Zeeger, 1989, Chin and Quek, 1997, Katamine, 

2000). Second, the effectiveness of a treatment can be evaluated in a short period of 

time and if this fails to correct the problem, then the countermeasure can be changed 

again in a very short time. Third, traffic conflicts include the human factor because 

the behavior of drivers can directly be observed in the field (Brown, 1994). Fourth, 

traffic conflict data can be used with or without crash data since each type of conflict 

is associated to a particular type of crash. Finally, traffic conflict data provides 

information about traffic volumes, routine conflicts, moderate conflicts, erratic 

maneuvers, severe conflicts or near-miss accidents, and other minor accidents, while 

crash data can only give information of property damage, injury accidents and fatal 

accidents. The main objective of this paper is to clarify the most types of conflicts 





and to suggest the correct decision may be made and a suitable access management 

technique may be applied and implemented. 

ACCESS MANAGEMENT TECHNIQUES 

A variety of access management, location and design practices and policies 

can be used to improve safety and operations of the roadway within a state’s, city’s 

or county’s jurisdiction (Williams, 2000). These are some of the access management 

techniques; 

1- Limiting the Number of Conflict Points. 

2- Separating Conflict Areas. 

3- Remove Turning Vehicles from the Through Lanes. 

4- Reduce the Number of Turning Movements. 

5- Improve Traffic Operations on the Access Drive or Intersection Local Street. 

6- Improve Traffic Operations on the Roadway. 

Figure 1 One example of using Access technique 

271


CONFLICT TYPES 

Adescription of all conflicts that could be possibly occurring at signalized or 

unsignalized intersections was provided in Traffic Conflict Technique for Safety and 

Operation Engineers Guide (Parker and Zeeger, 1989). However, not all of these 

conflicts can be considered as important for the maneuvers evaluated in this paper, 

only the most significant and common conflicts were considered. This resulted in 

the selection and evaluation of nine types of conflicts including: 

1. Conflict Type 1 (C1) Right-Turn Out of the Driveway Conflict, 

2. Conflict Type 2 (C2) Slow-Vehicle Same-Direction Conflict, 

3. Conflict Type 3 (C3) Lane Change Conflict, 

4. Conflict Type 4 (C4) U-turn Conflict, 

5. Conflict Type 5 (C5) Left-Turns Out of Driveway Conflict, 

6. Conflict Type 6 (C6) Direct-Left Turn and Left-Turn in From-Right Conflict, 

7. Conflict Type 7 (C7) Direct-Left-Turn and Left-Turn in From-Left Conflict, 

8. Conflict Type 8 (C8) Left-Turn Out of Driveway Conflict, 

9. Conflict Type 9 (C9) Slow U-turn Vehicle Same-Direction Conflict. 

CONCLUSION 

Conflict prediction will make conflict data more helpful, not only safety-wise 

but also for future traffic planning. The benefits of using the conflict data are 

obviously clear on the safety concerns. The first benefit is that the conflict data can 

be used for the compression of the maneuvers such as comparing direct left turn 

with right turn plus u-turn. The second is that the conflict data can be used as a 

measuring tool for the medians treatments. With high conflict rates, medians can be 

either closed or changed into directional opining medians. Using the conflict data 

will help planers and engineers to take right decisions. In addition, conflict data will 

be the tools in their hands to support those decisions. Finally, the conflict data could 

then be used as a good measuring tool for safety, and according to that suitable 

access management treatment could be applied. 

RECOMMENDATIONS 

This paper shows that we need to apply Access Management Technique in 

most of the major roads in Saudi Arabia. Median treatment is one of the most 

techniques that can be used for separating conflict. Sharing access is also another 

important issue along the major roads. The location of the signals before and after 

the driveway is critical and that will affect the arrivals of the incoming vehicles on 

the major road both directions. For that reason more consideration of upstream and 

downstream distance and signal timing is recommended. Finally, reviewing all the 

traffic operations and safety at signalizing and unsignalized intersection is very 

important before applying any access management technique, were those solutions 

can be fast and affective with lower cost than any other solutions. 



REFERENCES 



AASHTO, 1994. "American Association of State Highway and Transportation 

Officials", A Policy on Geometric Design of Highway and Streets, USA. 

Allen, B. L., Cooper, P. J. 1978, "Analysis of Traffic Conflicts and Collisions", 

Transportation Research Record, N. R.C., Washington D.C., pp.67-73. 

Amundsen, F., Hyden, C. 1977, "Proceedings of first workshop on traffic conflicts", 

Institute of Transport Economics, Oslo. 

Brown, G.R., 1994, " Traffic Conflicts for Roads Safety Studies", Canadian Journal 

of Civil Engineering, Vol. 21, No.1, pp.1-15, pp. 185-194. 

Chin, H.C., Quek, S.T., 1997, "Measurement of traffic conflicts", Accident Analysis 

and Preventation, Vol.26, No.3, PP. 169-185. 

Florida Department of Transportation. , 1997, "Median Handbook", FDOT, Florida. 

Highway Capacity Manual , 2000, "Unsignalized Intersections", Chapter 17. 

Katamine, N.M., Hamarneh, I.M., 1998, "Use of the traffic conflicts technique to 

identify hazardous intersections", Australian Road Research Board, Road and 

Transportation Research, Vol. 7, No. 3, pp. 17-35. 

Lu, J.J., Dissanayake, S., Castillo, N., Williams, K.M., 2001 October, "Safety 

Evaluation of Right Turns Followed by U-turns as an Alternative to Direct Left 

Turns – Conflict Data Analysis", Report submitted to Florida department of 

Transportation. Tallahassee, FL. 

National Highway Traffic Safety Administration, 2003 January, "A compilation of 

motor vehicle crash data from the fatality analysis reporting system and the 

general estimates system", http://www.nhtsa.dot.gov. 

NCHRP Report 420, 1999, "Impact of Access Management Techniques", US. 

Parker, M.R., Zeeger, C.V., 1989, "Traffic Conflict Technique for Safety and 

Operation Engineers Guide", Report FHWA-IP-88-026. FHWA,U.S. 

Department of Transportation. Washington, D.C. 

273


Sayed, T., Brown, G., Navin, F., 1994, "Simulation of Traffic Conflicts at 

Unsignalized Intersections with TSC-Sim", Accident Analysis and Prevention, 

Vol. 26, No. 5, pp. 593 – 607. 

Stover, V. , 1994, "Median Access Management and Design", prepared for the 

Florida Department of Transportation, Center for Urban Transportation 

Research, Florida, USA. 

United States Department of Transportation/ Federal Highway Administration, 

1996, 2000, "Proceedings of the Second and the Fourth National Conference 

on Access Management", Vail, Colorado and Portland, Oregon. 

Williams, K.M., 2000, "Access Management And Corridor presentation", Transportation 

Research Board, National Research Council. NCHRP Synthesis 233. 

Zegeer, C.V., Deen, R.C., 1978, "Traffic conflicts as a Diagnostic Tool in Highway 

Safety", Transportation Research Record 667, Washington D.C., pp. 48-55. 



ABSTRACT 

EFFECT OF POLYPROPYLENE FIBERS ON MECHANICAL 

PROPERTIES OF HIGH-STRENGTH CONCRETE 

A.A.F. Shaheen 1 A.A.Emam 1 , and I.M.Metwally 2 

1 2 

Strength of Material Institute Reinforced Concrete Institute 

Housing & Building National Research Center, 

87 Tahreer Street, Dokki, Giza, P.O. Box 1770, Egypt. 

Email: amro_fawzy@hotmil.com and Email: dr_aemam@yahoo.com 

Current trends in construction projects require the use of high strength 

concrete. Achieving such task mandates the use of proportionally high cement 

content, which comes with a band of problems such as high rate of plastic shrinkage 

and rapid development of micro-cracks. The focus of the present investigation is the 

differences in mechanical properties of high-strength concrete and ordinary strength 

concrete and the effect of using polypropylene fibers in the manufacturing of highstrength 

concrete. Three mixtures of concrete were cast to be representative of 

ordinary concrete, high strength concrete with and without polypropylene fibers. 

The samples were tested for compressive strength at 7, 28 and 90 days and splitting 

tensile strength at 7 and 28 days. Plastic shrinkage was monitored for 24 hours 

immediately after casting and impact resistance was determined after 28 days. 

The results indicate that high strength concrete with and without 

polypropylene fibers has better mechanical properties than ordinary concrete 

especially for compressive strength, tensile strength and impact. The plastic 

shrinkage for high strength without polypropylene fibers was higher than that of 

ordinary concrete. The study signifies that using polypropylene fibers can improve 

plastic shrinkage and reduce the formation of micro-cracking of high strength 

concrete. Moreover, the fibers significantly improve concrete impact resistance. 

KEYWORDS 

Mechanical properties; high-strength concrete; polypropylene fibers; impact; 

plastic shrinkage. 


The current technology makes it possible to produce high-strength concrete 

with compressive strength that meets the demands of the recent structural design 

practice. To produce concretes with high strength requires using ultra-fine 

cementitious materials such as silica fume [1], and a low water to binder ratio (the 


EFFECT OF POLYPROPYLENE FIBERS ON MECHANICAL PROPERTIES OF HIGH-STRENGTH CONCRETE 

binder being defined as the Portland cement together with any supplementary 

cementing material). Generally, for high strength concrete, the binder content is 

higher than that for normal-strength concrete. Moreover, the use of super plasticizer 

is required to provide adequate workability [2-7]. 

The main objective of this study is to investigate the influence of the 

condensed silica fume dust and polypropylene fibers on compressive strength, 

splitting tensile strength, plastic shrinkage and impact resistance of high strength 

concrete. Of meticulous interest herein is the monitoring of plastic shrinkage of highstrength 

concrete with and without polypropylene fibers. In this regard, the results 

presented by Bruno et al [8] indicated that very high values of plastic shrinkage can 

exist leading to drastic micro cracking in high strength concrete. To reduce plastic 

shrinkage cracking, the most widely accepted method is the use of randomly 

distributed fibers, particularly fine synthetic fibers in reasonable volume fraction. 

The most commonly used fibers to control plastic shrinkage cracking are steel, nylon, 

polyester, cellulose fibers, PVA, polypropylene, and polyolefin [9]. 

It has been demonstrated that fibrillated polypropylene fibers significantly 

reduce the plastic shrinkage of concrete. The mechanisms by which polypropylene 

fibers reduce plastic shrinkage can be summarized as follows: 

1. Increasing tensile strain capacity of fresh concrete beyond the tensile strains 

caused by shrinkage as shown schematically in Figure (1). In the study 

detailed in reference [10], fresh concrete was cast around a threaded steel 

rod and strain was applied to the rod. Crack initiation within fresh concrete 

was observed to determine its strain capacity; 

2. The regulation of bleed water rise by the reduction of coarse aggregate 

consolidation. Therefore, higher water retention by fresh concrete is 

achieved, resulting in better resistance to shrinkage [11]. Polypropylene 

fibers cause fresh concrete to bleed less and set faster, thereby reducing its 

time of exposure to plastic shrinkage. With these effects, the quantity of 

harmful capillaries formed by bleeding is reduced and intergranular friction 

is increased [12] ;and 

3. Arresting of micro cracks and capillaries, thereby restraining their potential 

expansion and development into plastic shrinkage cracks [13]. 

The uniform dispersion of fibers within fresh concrete is essential to their role 

in plastic shrinkage crack reduction. Achieving enhanced mixture uniformity with the 

application of fibrillated polypropylene fiber prevents localized expansion and 

development of capillaries and micro cracks into shrinkage cracks. Two effects of 

fibrillated polypropylene fibers upon plastic shrinkage can be distinguished: (i) Their 

influence in reducing free plastic shrinkage strain; and (ii) Their effect in resisting the 

resultant plastic shrinkage cracking strain/stress [14]. 




A.A.F. Shaheen, et al 

Figure (1): Effect of polypropylene fibers on strain capacity of fresh concrete, [10]. 

Polypropylene fibers have significant influence in enhancing other 

mechanical properties of concrete as reported by Soroushian et al.[15], they 

concluded that Polypropylene fibers increased the flexural and impact strengths, and 

decreased the compressive strength, of concrete materials; with 1% fiber volume 

fraction of polypropylene fibers, the average flexural and impact strengths were 1.21 

and 5.8 times greater, respectively, and the average compressive strength was 

reduced by 23% in comparison with plain concrete without fibers. 

The present research focuses on making a comparison among the three 

concrete grades, with respect to plastic shrinkage, compressive strength, tensile 

strength, and impact resistance. The first grade is traditional conventional concrete 

mix without any admixtures with an average compressive strength of 300 kg/cm 2 , 

the second is high strength concrete mix with silica fume, and the third is high 

strength concrete mix with silica fume and polypropylene fibers. 

2. EXPERIMENTAL PROGRAM 

Three mixes were cast to evaluate the advantages of adopted approach of 

adding silica fume and polypropylene fibers on the compressive strength, splitting 

tensile strength, impact and plastic shrinkage of concrete. Proportions of all the 

mixes and experimental program are outlined in Table (1). 

2.1 Materials 

Ordinary Portland cement complying with the Egyptian Standard 

Specification 4756- 1/2005 is used through out the present work. The Pozzolanic 

material was silica fume, which has been used in the high strength mixes in a ratio 

15% of cement weight as an additive material to improve the properties of 

concrete. It must be noted that a ratio of 15% SF in high-strength concrete is in line 

with literature e.g. Ref. [16] and [17]. 

271


Table (1): Proportions of concrete mixes and experimental program 

Designation Type 

A1 

A2 

A3 

Mix Mixture constituents 

Control 

High strength 

High strength 

with 

polypropylene 

fibers 

Cement 

kg/m 3 

350 

600 

600 

Silica 

fume 

kg/m 3 

- 

90 

90 

Sand/stone 

1:1.5 

1:1.5 

1:1.5 

Superplastizer % 

of cementitious 

materials 

Water/ 

cementitious 

materials 



- 

4 

4 

0.61 

0.19 

0.19 

polypropylene 

fibers (kg/m 3 ) 

- 

- 

2.0 

Tests 


(7,28, 90 days) 

Splitting tensile 

strength (7, 28 days) 

Beam impact test 

Plastic shrinkage 

The coarse aggregate was crushed stone with maximum grain size of 20 

mm, and the fine one was natural siliceous sand with a fineness modulus of 2.6, their 

relative proportions were 1.5:1, respectively. The water/cement ratio was taken 0.61 

for the control mix (A1), and then the ratio was reduced to 0.19 of binder cement 

materials for the second and third mixes (A2 and A3). This mandated the use of BVF 

as a superplastizer additive to reduce water content in the mix to get the desired 

strength for the concrete mix. Polypropylene fibers were added by an amount of 2 

kg/m 3 to concrete in the third mix to improve the plastic shrinkage of concrete. 

2.2 Mix Proportions 

In the first mix, neither pozzolanic admixture, nor superplastizer additive 

or polypropylene fibers was used to produce the control mix. In the other two mixes 

high cement content, silica fume, superplastizer additive were employed for 

achieving a high strength concrete. The third mix is typically the same as the 

second one except for adding polypropylene fibers to investigate its influence on 

concrete properties, especially plastic shrinkage and impact resistance. 

2.3 Specimens Preparation and Testing 

The three mixes were cast according to their proportion given in Table (1). 

After casting of each mix, the slump test was performed and then, oiled steel molds 

of different specimens were filled and compacted. After 24 hours, the specimens 

were demoulded and immersed in water at room temperature for curing until



testing. Nine cubes 150 mm were prepared to be tested for compressive strength 

after 7, 28 and 90 days. Six cylinders 150 mm in diameter and 300 mm in height 

were cast to be tested for splitting tensile strength after 7 and 28 days. Three beams 

with square cross-section (100x100 mm) and 500 mm span were prepared for 

impact testing after 28 days. Three cylinders of diameter 100 mm and 200 mm in 

height were cast for plastic shrinkage monitoring immediately after casting and for 

a period of 24 hours. Plastic shrinkage test setup is shown in Figure (2). 

Figure (2): Plastic Shrinkage Monitoring setup 

3. TEST RESULTS AND DISCUSSION 

3.1 Compressive Strength 

As a material parameter, the concrete compressive strength governs almost 

all other mechanical properties. Commercial production of concrete with high and 

ultra-high strength became a necessity to meet the never ending demand for urban 

expansion and development. In the current study a moderate high strength concrete 

(80 MPa.) was developed by manipulating the mix ingredients and using additives. 

This was achieved by increasing the cement content to 600 kg/m 3 (compared to 350 

kg/m 3 for the control mix) and adding 15% silica fume along with 4% 

superplastizer. The additives reduced water to cementitious materials ratio (w/cm) 

to 0.19, while maintain a reasonable workability (slump about 100 mm.). 

Comparing the compressive strength for the developed three mixes (A1, A2, and 

A3), indicates that the gain in the compressive strength for A2 and A3 is over 100% 

as depicted in Figure (3). Moreover, the figures show that adding the polypropylene 

fibers has no significant influence on the compressive strength. 

273


3.2 Tensile Strength 

The same trend outlined above is observed for splitting tensile strength. 

Figures (4) shows that the gain in the tensile strength for High Strength Concrete 

(HSC) is over 150% relative to ordinary concrete. The influence of adding 

polypropylene fibers is insignificant for tensile strength (comparing A3 and A2) as 

it improves the splitting tensile strength by only 5%. These results are expected 

and very much inline with literature [11,12]. It is worth mentioning that many 

concrete code of practice under estimates the concrete tensile strength for HSC 

such empirical relations are mainly intended for ordinary concrete, e.g. ACI 

, 

estimates tensile strength as 7. 5 f cu and the new trend for HSC is to use 

, 

11 f cu in psi units. 

3.3 Impact 

Drop weight impact test was carried out on simple concrete beam structural 

configuration. The total impact weight was about 2 kg.with semispherical impactor 

end. The impact energy is the weight of the impactor multiplied by the sum of drop 

heights until beam failure. Then, a dimensionless impact factor was calculated as the 

ratio of impact energy of A2 and A3 relative to the control specimen A1. As 

illustrated in Figure (5), the impact resistance of HSC mixes A2 and A3 has been 

improved by 6 and 13 times; respectively relative to the control mix A1. Moreover, 

adding polypropylene fibers doubled the impact resistance. Despite the very limited 

number of test specimens, this result is very significant and shows a general trend 

that polypropylene fibers can greatly improve absorbed impact energy. As a note of 

caution, a more statistically significant number of samples should be considered 

before drawing a definite conclusion regarding the cited results. 

3.4 Plastic Shrinkage 

As it has been mentioned earlier, adding polypropylene fibers causes fresh 

concrete bleed less which leads to reducing capillaries formed by bleeding and 

hence a reduction in micro-cracks. Moreover, the mix sets faster and the end 

result is better resistance to plastic shrinkage. In the current study, plastic 

shrinkage readings have been recorded over a period of 24 hours immediately 

after casting. At each time point, the reading is the average of three samples for 

each mix. Then, a plastic shrinkage dimensionless factor is calculated using mix 

A1 as a reference. As depicted in figures (6) and (7), it can be seen that the first 

four hours after casting is the most influential period at which most plastic 

shrinkage occurs. Comparing the results of mix A2 (cement content of 600 kg/m 3 ) 

with mix A1 (cement content 350 kg/m 3 ), an increase in plastic shrinkage between 

20% and 60% is observed. On the other hand, adding polypropylene fibers (mix 





A3) reduces shrinkage by about 20% relative to A2 ( i.e. the same cement 

content). It is worthy to mention that after 4 hours of casting the mix A3 gives the 

same result as mix A1 (with just over half the cement content). This result is a 

clear reflection of the mechanism created by adding the fibers which results in 

reducing plastic shrinkage and hence, micro-cracking of concrete. 

Compressive Strength (KN/mm 2 ) 

90 

80 

70 

60 

50 

40 

30 

20 

10 

0 

A1(control) 

A2 

A3 

0 14 28 42 56 70 84 98 

Concrete Age (days) 

Figure (3): Variation of Compressive Strength for Different Mixes over Time 

275


Splitting Tensile Strength 

(KN/mm 2 ) 

6 

5 

4 

3 

2 

1 

0 

0 7 14 21 28 35 

Concrete Age (days) 



A1(control) 

A2 

A3 

Figure (4): Variation of Tensile Strength for Different Mixes over Time 

Impact Factor 

14 

12 

10 

8 

6 

4 

2 

0 

A1(control) A2 

Mix Designation 

A3 

Figure (5): Variation of Impact Factor for Different Mixes

Plastic Shrinkage Dimensionless 

Factor 

1.8 

1.6 

1.4 

1.2 

1 

0.8 

0.6 

0.4 

0.2 

0 

A2/A1 

A3/A1 

0 5 10 15 

Time (hours) 

20 25 30 



Figure (6): Variation of Dimensionless Plastic Shrinkage Factor over Time 

Relative Plastic Shrinkage 

Factor 

1.8 

1.6 

1.4 

1.2 

1 

0.8 

0.6 

0.4 

0.2 

0 

4. CONCLUSIONS 

A2/A1 A3/A1 

Mix Designation 

0.5 hr. 

2 hrs 

24 hrs. 

Figure (7): Dimensionless Plastic Shrinkage Factor Chart 

In the current study, a moderate HSC has been developed with compressive 

strength of about 80 MPa. using crushed stone aggregate and some admixtures to 

reduce w/cm ratio and improve workability. The influence of adding polypropylene 

fibers to HSC on impact resistance as well as plastic shrinkage has been 

investigated. The study results indicate that the fibers have no significance on either 

the compressive strength or the splitting tensile strength. On the other hand, the 

277


fibers influence on both the impact resistance and plastic shrinkage is remarkable. 

The absorbed impact energy during drop weight test is more than doubled. 

Moreover, the concrete plastic shrinkage is reduced by over 20%, which results in 

less micro-cracking. It is also worthy to mention that adding the fibers compensates 

for the resulting increase of plastic shrinkage due to almost doubling the cement 

content needed for HSC. 

REFERENCES 

1. Roles, S.; M., Mbessa; J., Ambroisa; and J., Pera, 1999, " Influence of 

Ultra-Fine Particle Type on Properties of Very-High-Strength Concrete", 

SP 186-39, High Performance Concrete, ACI, pp.671-685 

2. Zeghib, R., and M., Nacer-Bey, 1997, "Study and Formulation of High 

Performance Concrete with Various Ultra-Fine Admixtures", Fifth 

CANMET/ACI International Conference on Superplasticizers and 

Chemical Admixtures in Concrete", Supplementary papers, Rome, pp. 286- 

293. 

3. Lang, E., and J., Geisseler, 1996, " Use of Blast Furnace Slag Cement with 

High Slag Content for High-Performance Concrete", 4 th International 

Symposium on Utilization of High-Strength/High-Performance Concrete, 

Paris, pp. 213-222. 

4. Novokshchenov, V., 1992, "Factors Controlling the Compressive Strength 

of Silica Fume Concrete in the Range 100-150 MPa", Magazine of 

Concrete Research, Vol. 44, No. 158, pp. 53-61. 

5. Shah, S. P., and S.H., Ahmed, 1994, "High Performance Concrete: 

Properties and Application", Mc Graw-Hill, Inc, London, 403 p. 

6. Gjorv, O.E., 1994, "High-Strength Concrete", In: Advances in Concrete 

Technology, 2 nd Edition, CANMET, Editor: V. M. Mlhotra, pp. 19-82. 

7. De Larrard, F., 1988, " Formulation et proprietes des beton a tres hautes 

performances", Rapport de recherché du laboatoire Central des Ponts et 

Chaussees, Paris, No. 149, 350p. 

8. Bruno, V., M., Chabannet, J., Ambroise, and J., Pera, 1991, " Betons Tres 

Haute Performance (BTHP): Microfissuration et durabilite", 2 nd 

CANMET/ACI International Conference on Durability of Concrete, 

Supplementary Papers, Montreal, pp. 175-193. 





9. Naaman, A.E. , T., Wongtanakitcharoen, and G., Hauser, 2005, " Influence 

of Different Fibers on Plastic Shrinkage Cracking of Concrete", ACI 

Materials Journal, Vol.102, No. 1, Jan-Feb. 2005, pp.49-58. 

10. Webster, T., 2001, personal communication, Webster Engineering 

Associates, Independence, Ohio, Mar. 2001. 

11. Zollo, R.; J., Ilter; and G., Bouchacourt, 1986, “Plastic and Drying 

Shrinkage in Concrete Containing Collated Fibrillated Polypropylene 

Fiber,” Third International Symposium on Developments in Fiber 

Reinforced Cement and Concrete, RILEM Symposium FRC 86, V. 1, July 

1986. 

12. Soroushian, P.; F., Mirza; and A., Alhozaimy, 1995, “Plastic Shrinkage 

Cracking of Polypropylene Fiber Reinforced Concrete,” ACI Materials 

Journal, Vol. 92, No. 5, Sept.-Oct. 1995, pp. 553-560. 

13. Bayasi, Z., and G., Zeng, 1994, “Application of Polypropylene Fibers for 

Reduction of Shrinkage Cracking of Concrete,” International Conference 

on Reinforced Concrete Materials in Hot Climates, United Arab Emirates 

University, Nov. 1994, pp. 441-452. 

14. Bayasi, Z., and M., McIntyre, 2002, " Application of Fibrillated 

Polypropylene Fibers for Restraint of Plastic Shrinkage Cracking in Silica 

Fume Concrete", ACI Materials Journal, Vol. 99, No. 4, July-August 2002, 

pp.337-344. 

15. Soroushian, P., A., Khan, and J., Hsu, 1992, " Mechanical Properties of 

Concrete Materials Reinforced with Polypropylene or Polyethylene fibers" 

, ACI Materials Journal, Vol. 89, No. 6, Nov. – Dec. 1992, pp. 535-540. 

16. Wei, S., P., Ganghua and D., Dajun, 1997, " Effect of the Combined Use of 

Ultra-Fine Fly Ash and Silica Fume on Strength of HPC", SP 172-16, High 

Performance Concrete, ACI, pp.299-312. 

17. Sabir, B.B., 1995," High-Strength Condensed Silica Fume Concrete", 

Magazine of Concrete Research, Vol. 47, No. 172, pp. 219-226. 

279




THEORETICAL INVESTIGATION ON THE BEHAVIOR OF HYBRID 

STEEL GIRDERS UNDER PURE MOMENT 

Abdel-Lateef, T.H 1 ,Tohamy, S.A. 2 , and Sadeek, A.B. 3 

1: Civil Engineering Department. El Minia University, El Minia-Egypt, Email, 

TH_abdellateef@yahoo.com 

2: Civil Engineering Department. El Minia University, El Minia-Egypt, Email, 

Sedky_t2000@yahoo.com 

3: Structural Engineering Department. El Azhar University, Qena-Egypt, 

Email, Amr_bakr2003@yahoo.com 

1-Abstract 

I-shaped beams and girders are flexural structural members that carry 

transverse loads perpendicular to their longitudinal axis primarily in a 

combination of bending and shear. Bending resistance is achieved through the 

action of a compression and tension force inducing a couple resisting the 

externally applied moment. The compression element (flange) of the cross 

section is integrally braced perpendicular to its plane through its attachment to a 

stable tension flange by means of a web. A hybrid steel girder is a welded girder 

with different steel grades in flanges and web. Usually, the flanges are made of 

high strength steel (HSS) and the web of a lower grade. Such girders are more 

economical than homogenous girders. 

In this study, theoretical analysis is performed using the finite element 

program (ANSYS 5.4) on twelve dimensional finite element models to clarify the 

influence of sectional compactness, aspect ratios of web panel and initial 

imperfections of webs on the estimation of the critical load of plate girder under 

pure moment. It is assumed that the flanges have higher strength than the web for 

six models and the same for other homogenous models. The procedure adopted in 

the paper is directed to suggest some accurate and simple procedures from which a 

fairly good estimate for the critical moment can be obtained, taking into account the 

real boundary conditions between the web and the flanges. This procedure is 

concerned with the estimation of the critical moment from a reasonably accurate and 

simple stress distribution on the cross section. Also, the comparison between 

models, show that compact flange effect on the degree of restraint between the web 

and the flanges that, the connection between the web and the flanges behave as 

partially fixed connection. It can be noticed from results that, the effect of using web 

with less yield stress than flanges is small on the relation between applied loads and 

out-of-plane web deflection. 


Theoretical Investigation on the behavior of Hybrid Steel Girders under Pure Moment 

Key Words 

Critical Load- Yield Stress- Buckling –Web -Flange 

2-Introduction 

A hybrid steel girder is a welded girder with different steel grades in flanges 

and web [1]. Usually, the flanges are made of high strength steel (HSS) like S690 

and the web of a lower grade say S355 but combinations like S460 and S355 are 

also used. Such girders are more economical than homogenous girders. Hybrid 

girders have been used in the US since long but they are not commonly used in 

Europe. Some examples of the use of hybrid girders in Sweden are presented 

together with economic comparisons. Design rules for hybrid girders are presented 

together with justifications. Typically, hybrid girders are of cross-section class 4 

according to Euro code 3. The resistance in bending in ultimate limit state is 

influenced by the local yielding of the web, which limits the stresses in the web and 

affects the effective width of the web as well. For hybrid girders, it is shown that 

this restriction applies to the yield strength of the flanges and that yielding of the 

web does not influence the fatigue strength. The cross-section class of the flanges is 

determined as usual e.g. according to Ref. [2-3]. The cross-section class of the web 

should however be determined using the yield strength of the compression flange. 

The primary function of the web plate is to resist the applied shear force. In 

practice, plate girder structures can be classified into two main categories according 

to the behavior of the compression flange: 

a- Compression flange prevented or restrained against lateral 

tensional buckling as in case of deck bridges. 

b- Compression flange is free or partially restrained from lateral 

tensional buckling as demonstrated in through bridges. 

It is obvious that case (a) is more common which encouraged us to 

investigate the buckling of web in such case under pure bending conditions. 

Graciano and Johansson [4] presented a design procedure for the determination of 

the ultimate resistance of longitudinally stiffened girder webs to concentrated loads. 

The elastic critical buckling stress is estimated using the load vs lateral displacement 

squared method, as described by Venkatarmaish and Roorda [5] .The platebuckling 

coefficient k for a beam-column web hinged on its four edge will lie 

between the value k=23.9 for pure bending and k=4 for uniform compression. 

Values for various combinations of moment and axial force have been determined 

[6-9]. An analysis of the I-flange under compression, taking into account rotational 

restraint from the web, showed that this value is increased by only 2% or 3% 

percent by such restraint [10]. The 1997 Edition of the Eurocode 3 (EC3) Part 1.5 




Abdel-Lateef, T.H, et al 

[11] includes a check of the buckling resistance of girder webs to concentrated 

transverse loads or patch loading at ultimate limit load state, which is relevant, e.g. 

for bridge launching [12]. The methodology is based on the research conducted by 

Lagerqvist and Johansson [13], which makes use of yield resistance (σY), 

slenderness parameter σ λ / = , where σcr is the elastic critical stress. 

3-Theoretical analysis 

Y cr σ 

3-1 The analytical model 

The analytical model used in the current study consists of a simply supported 

symmetric girder subjected to two point load at third points as shown in Fig. (1). 

Fig. (1) Analytical model of simply supported symmetric girder 

1-Three cases of web slenderness ratios were considered : 

(a)- hw/tw = 200 (b)- hw/tw = 167 (c)- hw/tw = 143 

2-Two cases of flanges configuration as compact flange denoted by CM., and noncompact 

flanges denoted by NC. 

(a)- Compact (CM with tf=2cm) 

(b)- Non-Compact (NC with tf=1.2cm) 

Where, bf constant at 24cm. 

3- Aspect ratios of web panel φ = 0.6 

4- Initial imperfections of web wo/tw = 40% 

Figure (2) shows the basic cross section , and Fig. (3) shows the plate girders 

used in the analysis with different aspect ratios for pure bending parts. 

283


Fig. (2) Basic cross-section of 

theoretical study 

Fig (3) Plate girder with different aspect 

ratio for web panels 

Table (1) lists the dimensions of the cross sections of six hybrid specimens 

with yield stresses for web and flanges equal to 2.4 t/cm 2 and 3.6 t/cm 2 respectively. 

Also, in this table, the classification of these specimens according to Egyptian Code 

of Practice for Steel Construction and Bridges [16]. The six homogenous specimens 

are with the same cross section and classification with yield stresses for web and 

flanges equal to 3.6 t/cm 2 . 

Table (1) Classification of hybrid plate girders models 

Dimension of Cross Sections and Classification of Web & Flanges 

Specimen h (cm) tw (cm) bf (cm) tf (cm) h/tw c/tf 



Slenderness of 

Web 

Slenderness of 

Flange 

1 100 0.5 24 2 200 5.87 Slender Compact 

2 100 0.5 24 1.2 200 9.79 Slender Non-Compact 





3-2 Finite Element Modeling 

The ANSYS models are used to study the behavior of slender plate girders 

subjected to bending by incorporating all the nodes, element, material properties, 

dimensions and boundary conditions. The geometric boundaries of the plate girder 

such as flange width, web height and girder length, etc., should be defined at the 

first step in solid modeling.



Soild45 element was used to model the plate girder in the finite element 

analysis. This element is defined with eight nodes each having 3 degrees of freedom 

; translations in the nodal x, y and z directions. The Soild45 element was chosen 

mainly because it has plasticity, large deflection and strain capabilities. This will 

give an accurate representation of the actual spread of plasticity and yielding 

behavior of the girder model. The geometry, node locations, and the coordinate 

system for this element are shown in Fig. (4). The bearing and transverse stiffeners 

are present along the longitudinal axis of the girder analytical model on both sides 

of the web. The stiffener dimensions remain constant throughout the entire study. 

All stiffeners are provided for the entire height of the web and are braced against 

out-of-plane translation in an idealized way in order to reduce the number of 

parameters observed in the current study .All the elements comprising the girder 

plate components are rectangular and square shaped mesh with a ratio not exceeding 

1:20, this limit is recommended for the finite element program. The planes of the 

mesh surfaces correspond with the middle surfaces of the constituent cross-sectional 

plate components of the girder as shown in Fig. (5). 

Fig. (4) Properties of element Solid45 Fig. (5) Illustration of mesh 

surface planes 

The size of a finite element model can be reduced significantly by using 

symmetry in the body being analyzed. There was symmetry along the length of 

the beam (the longitudinal axis, or the axis-1 in the finite element model). The 

loading was symmetric about the middle of the beam. The support conditions are 

almost is symmetric about the middle cross section of the beam, but only one 

support provides restraint against axis-1 translation. It was possible to consider 

only half the length of the beam, and apply the boundary conditions, as shown in 

Fig. (6), to all nodes at the middle section of the beam. axis-1 translation was 

prevented at the middle cross section. 

285


Fig. (6) Longitudinal symmetry of finite element model 

To include the initial imperfection of the component plates, a simulation of 

the buckling shape by sine wave is performed by a small concentrated force at the 

mid height of panels; then the geometry of girder is updated to the desire value of 

initial imperfection. To investigate the effect of large displacements on the behavior, 

a nonlinear geometric analysis was performed, a bilinear elasto-plastic response of 

the material was considered. 

3-3 Results of finite element analysis 

The analysis was performed with a certain value of the load less than the 

ultimate load for the perfect beam. This load was automatically divided into equal or 

unequal steps (Force Control or Displacement Control) until the program stops due 

to the control of load or displacement or the convergence of solution. A series of 

finite element analyses were performed to obtain the data of load-lateral 

displacement relation for calculating the critical load of plate girder under pure 

moment for the different cases illustrated before . 

3-3-1 Results of applying initial imperfection on the relation between load and 

lateral displacement. 

Both of, relation between load and lateral displacement at the mid height of 

panel under the action of pure bending moment for a compact flanges and the output 

of finite element program of final failure load are shown in Figs (7), (8) and (9). All 

of the previous models have web initial imperfection equal to 40% of tw at the 

panels under pure moment. 



Load (ton) 

72 

64 

56 

48 

40 

32 

24 

16 

8 

0 

0.2 0.22 0.24 0.26 0.28 0.3 0.32 0.34 

Web Buckling (cm) 

hw =100cm 

bf =24cm 

tf =2cm 

tw =0.5cm 

h/a =0.6 

Figure (7a) Load-lateral displacement curve 

(ACM, hw/tw=200, φ=0.6) 

Load (ton) 


(BCM, hw/tw=167, φ=0.6) 

wo/tw=40% 

72 

64 

56 

hw =100cm 

48 

bf =24cm 

tf =2cm 

40 

32 

24 

16 

8 

0 

tw =0.6cm 

h/a =0.6 

wo/tw=40% 

0.24 0.28 0.32 0.36 0.4 0.44 0.48 0.52 




Figure (7b) Buckling pattern for hybrid model 

(ACM, hw/tw=200, φ=0.6) 


(BCM, hw/tw=167, φ=0.6) 

287


Load (ton) 

72 

64 

56 

48 

40 

32 

24 

16 

8 

0 

0.28 0.32 0.36 0.4 0.44 0.48 0.52 0.56 0.6 


hw =100cm 

bf =24cm 

tf =2cm 

tw =0.7cm 

h/a =0.6 

wo/tw=40% 


(CCM, hw/tw=143, φ=0.6) 




(CCM, hw/tw=143, φ=0.6) 

Also, the relation between load and lateral displacement at the mid height 

of panel under the action of pure bending moment for a non-compact flanges and 

the output of finite element program of final failure load are shown in Figs (10), 

(11) and (12). 

Load (ton) 

72 

64 

hw =100cm 

bf =24cm 

tf =1.2cm 

56 

tw =0.5cm 

48 

40 

32 

24 

h/a =0.6 

16 

8 

0 

wo/tw=40% 

0.2 0.22 0.24 0.26 0.28 0.3 0.32 0.34 



(ANC, hw/tw=200, φ=0.6) 


(ANC, hw/tw=200, φ=0.6)

Load (ton) 

Load (ton) 

72 

64 

hw =100cm 

bf =24cm 

tf =1.2cm 

56 

tw =0.6cm 

48 

40 

32 

24 

h/a =0.6 

16 

8 

0 

wo/tw=40% 

0.24 0.28 0.32 0.36 


0.4 0.44 0.48 


(BNC, hw/tw=167, φ=0.6) 

72 

64 

hw =100cm 

bf =24cm 

tf =1.2cm 

56 

tw =0.7cm 

48 

40 

32 

24 

h/a =0.6 

16 

8 

0 

wo/tw=40% 

0.28 0.32 0.36 0.4 0.44 0.48 0.52 0.56 



(CNC, hw/tw=167, φ=0.6) 




(BNC, hw/tw=167, φ=0.6) 


(CNC, hw/tw=167, φ=0.6) 

Based on the results presented herein for the homogenous and hybrid cross 

section, a comparison between the load-lateral displacement curves for each case of 

compact and non compact flanges are shown in Figs (13), (14) and (15). 

289

Load (ton) 


72 

64 

56 

48 

40 

32 

24 

16 

8 

0 

0.2 0.22 0.24 0.26 0.28 0.3 0.32 0.34 0.36 0.38 


hw =100cm 

bf =24cm 

tf =2cm 

tw =0.5cm 

h/a =0.6 


(ACM, hw/tw=200, φ=0.6) 

Load (ton) 


(BCM, hw/tw=167, φ=0.6) 

Figure (13b) Load-lateral displacement curve 

(ANC2, hw/tw=200, φ=0.6) 


(BNC, hw/tw=167, φ=0.6) 



Hybrid 

Homogenous 

80 

72 

64 

56 

hw =100cm 

bf =24cm 

48 

tf =2cm 

tw =0.6cm 

40 

32 

24 

h/a =0.6 

16 

Hybrid 

8 

0 

Homogenous 

0.24 0.28 0.32 0.36 0.4 0.44 0.48 0.52 0.56 


Load (ton) 

72 

64 

hw =100cm 

bf =24cm 

tf =1.2cm 

56 

tw =0.5cm 

48 

40 

32 

h/a =0.6 

24 

Hybrid 

16 

8 

0 

Homogenous 

0.2 0.22 0.24 0.26 0.28 0.3 0.32 0.34 0.36 0.38 0.4 


Load (ton) 

80 

72 

64 

56 

48 

40 

32 

24 

16 

8 

0 

0.24 0.28 0.32 0.36 0.4 0.44 0.48 


hw =100cm 

bf =24cm 

tf =1.2cm 

tw =0.6cm 

h/a =0.6 

Hybrid 

Homogenous

Load (ton) 

80 

72 

64 

56 

48 

40 

32 

24 

16 

8 

0 

0.28 0.32 0.36 0.4 0.44 0.48 0.52 0.56 0.6 


hw =100cm 

bf =24cm 

tf =2cm 

tw =0.7cm 

h/a =0.6 

Figure (15a) Load-lateral displacement 

curve 

(CCM, hw/tw=143, φ=0.6) 

Hybrid 

Homogenous 



80 

72 

hw =100cm 

bf =24cm 

64 

tf =1.2cm 

tw =0.7cm 

56 

48 

40 

32 

h/a =0.6 

24 

Hybrid 

16 

8 

0 

Homogenous 

0.28 0.32 0.36 0.4 0.44 0.48 0.52 0.56 



(CNC, hw/tw=143, φ=0.6) 

It can be noticed from these results that, the effect of using web with less 

yield stress than flanges is small on the relation between applied loads and out-ofplane 

web deflection. Accordingly, a reasonably small effect will be introduced to 

the buckling load of the whole cross section. This reflect the well known fact that 

the flanges provide the majority of the bending resistance for bending. 

4- Calculation of critical load from the appropriate stress distribution on cross 

section 

As the web is slender in all cases, then the cross section should be considered 

as non-compact, However, by careful inspection of the value of the stress at the 

junction between the flange and the web, we can assume that when this value 

approaches the critical value of the plate in bending, it could be said that the 

bending moment at this instant is the critical bending [15]. Figures (16) shows the 

suggested distribution at this stage 

2 

Where kπ 

E 

σ cr = (1) 

2 2 

12 ( 1−υ 

)( b / t ) 

And k represent the buckling coefficient according to the boundary 

conditions between the web and the flanges. Currently, this value is taken equal to 

23.9 for web-flange junction as simple. However, if this junction is considered as 

built-in, we may consider k=39 where, values of buckling coefficients are depend 

on the aspect ratio defined before. 

Accordingly, we may have two different situations as follows; 

(i) A simply supported end condition for the web along the unloaded edges. 

(ii) A built-in end condition for the web along the unloaded edges. 

Load (ton) 

291


Fig (16a) Stress distribution at the 

level of critical load Pcr 



Fig (16b) Suggested stress distribution 

Hence, two different values for the critical bending moment can be calculated 

according to Fig. (16b) by the following equations:- 

In the case of simple edge:t 

f hw 

1 

2 hw 

M = x σ t b ( + ) + ( σ −σ 

) t h ( t + ) (2) 

f ( simple) 

[ simple f f 

f simple w w f ] 

2 2 2 2 

3 2 

M = ⎡1 σ 

2 

h ) ⎤ 

w( 

simple) 

2* 

t h ( 

⎢⎣ 2 

simple w w 

3 

w 

(3) 

⎥⎦ 

Mcr( 

simple) 

Mcr 

= Mf 

+ Mw 

⇒ Pcr 

= 

(4) 

L/ 

3 

In the case of fixed edge:- 

⎛ 

t f h w 

⎞ 

= ⎜σ 

+ + 

1 

σ − σ 

2 h w 

Mf 

( fixed) 

2 fixed tf 

bf 

( ) ( f fixed). 

t w hw 

( tf 

+ ) ⎟ 

⎝ 

2 2 2 

3 2 ⎠ 

(5) 

[ 1 t h ( 2 h ) ] 

Mw( fixed) 

2 σ 

2 fixed w w 3 w 

Mcr( 

fixed) 

Mcr 

Mf 

+ Mw 

⇒ Pcr 

= 

L/ 

3 

= (6) 

= (7) 

It is clear that the value of σf can be determined from the finite element analysis 

given before, which is corresponding to the stress on the outer fiber of the flange once 

the maximum stress on the web reaches σsimple or σfixed as the case may be. 

Hence for each model we shall have two different values for Mcr, one based on 

Eq. (4) and the other on Eq. (7).



Before calculating these values and make use of them, the critical load is 

also estimated using L.F.R.D. The design flexural strength for plate girder with 

slender webs shall be φb*Mn where φb =0.85 and Mn is the lower value obtained 

according to the limit states of tension flange yield and compression flange 

buckling as follows [16]; 

M 

nt 

= S xt R e σ 

( For tension flange) (8) 

Y t 

M nc = S xc R PG R e σ cr ( For compression flange) (9) 

R PG 

a r h c 

= 1 − 

( − 

222 

) ≤ 1 

(10) 

1200 + 300 a r t w σ 

If σcr =σY for a doubly symmetric shape, Sxc = Sxt = Sx 

M n = S x R PG σ 

Yt 

(11) 

Equation (11) is suitable for the case of study herein as the lateral constraint 

prevent any lateral torsional buckling in the compression flange. 

Tables (2) shows the different values of Mcr and Pcr for each model calculated 

on the previous bases i.e., 

(i) Simply supported web edges, Eq. (4) 

(ii) Built-in web edges Eq. (7) 

(iii) L.R.F.D. Eq. (11) 

Table (2a) Values of critical load produced for hybrid models 

Model 

cr 

Stresses Critical Moment Critical Load 

σf σsimple σfixed M simple M fixed M LRFD Psimple Pfixed PLRFD 

ACM 3.0 2.0 1.2 11315.4 13825.2 18603.5 37.7 46.1 62.0 

ANC 2.6 2.0 1.2 6499.8 8256.1 11780.9 21.7 27.5 39.3 

BCM 3.3 2.4 1.7 13871.0 16245.1 19423.4 46.2 54.2 64.7 

BNC 2.8 2.4 1.7 8342.9 10038.7 12598.2 27.8 33.5 42.0 

CCM 3.4 2.4 2.3 16769.7 17039.4 20232.7 55.9 56.8 67.4 

CNC 3.0 2.4 2.3 10432.7 10628.9 13402.2 34.8 35.4 44.7 

293


Table (2b) Values of critical load produced for homogenous models 

Model 

Stresses Critical Moment Critical Load 

σf σsimple σfixed M simple M fixed M LRFD Psimple Pfixed PLRFD 

ACM 3.3 2.0 1.2 12079.3 14589.1 19020.4 40.3 48.6 63.5 

Where 

ANC 3.0 2.0 1.2 7099.7 8856.0 12186.0 23.7 29.5 40.6 

BCM 3.5 2.8 1.7 14487.1 18285.7 19932.5 48.3 61.0 66.5 

BNC 3.1 2.8 1.7 8694.0 11407.5 13098.4 29.0 38.0 43.7 

CCM 3.6 3.4 2.3 17114.7 21018.7 20836.3 57.0 70.1 69.5 

CNC 3.2 3.1 2.3 10725.3 12728.0 

M LRFD 

σ f Stress on flnges σ simpl 

Moment result from LRFD 

formula. 



e 

σ fixed 

14000.1 35.8 42.4 46.7 

Critical stress in case of simply 

supported edges 

Critical stress in case of fixed 

supported edges 

5- Comparison of results 

Finally, a comprehensive comparison of results is introduced in Figs.(17), (18), 

and (19) for web thickness equal to 0.5cm, 0.6cm, and 0.7cm respectively with 

aspect ratios of web panel equal to 0.6. 

Load 

70 

60 

50 

40 

30 

20 

10 

0 

Hybrid Model 

2 1.2 

Flange Thickness (cm) 

Simply Supported Load LRFD Load Yield Load Fixed Supported Load 

Fig (17a) Comparison between different 

methods for estimating critical 

load (hw/tw=200, φ=0.6). 

h w =100cm 

b f =24cm 

t w =0.5cm 

φ =0.6 

Load 

80 

70 

60 

50 

40 

30 

20 

10 

0 

Homogenous Model 

2 1.2 



Fig (17b) Comparison between 

different methods for estimating 

critical 

load (hw/tw=200, φ=0.6). 

h w =100cm 

b f =24cm 

t w =0.5cm 

φ =0.6

Load 

Load 

70 

60 

50 

40 

30 

20 

10 

0 

Hybrid Model 



load (h w/t w=167, φ=0.6). 

80 

70 

60 

50 

40 

30 

20 

10 

0 

2 1.2 



Hybrid Model 

2 1.2 



h w =100cm 

b f =24cm 

t w =0.6cm 

φ =0.6 

h w =100cm 

b f =24cm 

t w =0.7cm 

φ =0.6 



load (hw/tw=143, =0.6). 






critical load (h w/t w=167, φ=0.6). 



critical load (hw/tw=143, =0.6). 

It can be seen that in almost all cases, a quite good agreement is obtained 

when considering the web plate to be built-in with the flanges. The assumption of 

simply supported web to flange gives lower values for the critical load in all cases 

considered. However, in cases of more thick webs (hw/tw =143), Figs. (19) 

approximately all methods give equal critical load as the web becomes thicker and 

able to resist stresses close to the yield stress without being buckled. 

It is clear that for the cases when λweb=200,167,143 the estimation of the 

buckling load through the suggested stress distribution as the web is built-in gives 

more reasonable values. Thus the relation h w / t w ≥ 120 might be more accurate 

and economical in proposing the web buckling criterion , when calculating the full 

critical bending moment in panels under pure bending. 

Load 

Load 

80 

70 

60 

50 

40 

30 

20 

10 

0 

80 

70 

60 

50 

40 

30 

20 

10 

0 

2 1.2 




2 1.2 



h w =100cm 

b f =24cm 

t w =0.6cm 

φ =0.6 

h w =100cm 

b f =24cm 

t w =0.7cm 

φ =0.6 

295


6- Conclusions 

The design of plate girders, like any other structure, is linked with many 

important parameters. One of these important aspect is the interaction between the 

flange and web which determines the capacity of the girder under loadings. The 

current research deals with the determination of the critical moment for both of, 

hybrid plate girders and homogenous plate girders whose compression flanges is 

adequately supported against lateral torsional buckling by studying the panels under 

pure moment. 

As has been proved through the analysis of the results, a reasonable stress 

distribution can be assumed for the section of pure bending moment at the stage of 

web buckling. This stress distribution depends on the approved assumption state 

that; the line junctions between web and flanges are built-in edges. This procedure 

can directly gives a good estimate for the critical bending moment. This research has 

also demonstrated that, the influence of the flange compactness on calculating the 

critical moment is very small so far as the compression flanges are laterally 

supported. 

References 

1- M. Veljkovic and B. Johansson "Design of hybrid steel girders", Journal of 

Constructional Steel Research, 60 (2004) 535–547. 

2- R. Frost and C. Schilling "Behavior of hybrid beams subjected to static loads", 

Journal of Structural Division, Proceeding of the American Society of Civil 

Engineers, 1964; 90(ST3). 

3- C. L. Pan and W. W. Yu "Bending strength of hybrid cold-formed steel beams", 

Journal of Thin-Walled Structures, 40 (2002), 399–414. 

4- C. Graciano and B. Johansson "Resistance of longitudinally stiffened I-girders 

subjected to concentrated loads", Journal of Constructional Steel Research 

2002. 

5- K. R. Venkatarmaish and J. Roorda "Analysis of local plate buckling data", 

Proceedings of the 6th International Specialty Conference on Cold-formed 

Steel Structures, St. Louis, 45–74, 1982. 

6- M. R. Bambach and K. J. R. Rasmussen "Tests of unstiffened elements under 

combined compression and bending" Research Report No R818, May 2002. 

7- T. H. Abdel-Lateef, S. A. Tohamy "Bending of isotropic plates under transverse 

line loading " 3rd Minia International Conference, 3-5 April 2005,28-48. 

8- T. H. Abdel-Lateef, "Buckling of stiffened plates subject to combined shear and 

compression load " Ph.D., University College London, November 1982. 

9- M. A. Dabaon and G. M. Atia “Buckling analysis of steel plates with stiffened 

openings subjected to in-plane combined stresses”, 5th Al-Azhar Conference, 

2003. 

10- M. Lay "Flange local buckling in wide-flange shapes", Journal of Structural 

Division, ASCE, December 1965. 





11- Euro Code 3 (EC3). “Design of steel structures”, Part 1-5, General rules 

Supplementary rules for planar plated structures without transverse loading. 

ENV 1993-1-5, 1997. 

12- Euro Code 3 (EC3). “Design of steel structures”, Part 2, Steel bridges, February, 

ENV 1993-2, 1997. 

13- O. Lagerqvist and B. Johansson "Resistance of I-girders to concentrated loads", 

Journal of Constructional Steel Research 1996; 39(2): 87–119. 

14- S. Walbridge and J. P. Lebet "Patch loading tests of bridge girders with 

longitudinal web stiffeners" Rapport d’essais E´cole Polytechnique Fe´de´rale 

de Laussane, ICOM 447, 2001. 

15- Sadeek, A. B. “Buckling of web plate in plate girder bridge” PhD., El-Minia 

University , 2006. 

16- Egyptian Code of Practice for Steel Construction and Bridges, L.F.R.D, 

Research Center for Housing and Physical Planning, Cairo, Under Publication. 

Symbols 

A Cross sectional area E Modulus of elasticity of steel 

bf Flange width Pcr Critical Load 

tf Flange thickness I Second moment of area 

d Overall depth of girder ν Possion’s ratio 

h Web height k Buckling Coefficient 

tw Web thickness w Deflections functions of plate 

φ panel aspect ratio wo Initial imperfection functions 

a length of one-bay λ Slenderness parameter 

b spacing between two adjacent σcr 

longitudinal stiffeners 

Critical 

= ( ) 

stress 

2 

k π 

2 

Ε 

12 ( 1 − ν 

2 

) 

t 

b 

σyf Yield Stress for Flange. σyw Yield Stress for Web. 

Mf Moment resisted by flange Mw Moment resisted by web 

297




REMOTE SENSING AND GIS FOR LAND SUITABILITY: AN 

APPLICATION FOR ECOTOURISM 

Nedal, A. Mohammad 1 , Sharifah Mastura S. A. 2 , Johari Mat Akhir 1 

1 School of Environmental and Natural Resource Sciences 

2 School of Social, Development and Environmental Science 

Universiti Kebangsaan Malaysia 

ABSTRACT 

This study employs the integration of remote sensing and GIS database with 

Multicriteria Spatial Decision Support System (MC-SDSS) for the identification of 

the potential ecotourism sites based on the available natural resources so as to reach 

an optimum sustainability. The study area is in Kuala Selangor district, located 

northwest of Selangor state in Malaysia. In identifying potential ecotourism sites for 

ecotourism development; several maps are taken as parameters. These maps are 

landuse/landcover (using Satellite Remote Sensing data from SPOT 2005), 

vegetation density (using NVDI approach), soil suitability, erosion risks and 

transportation accessibility. The analysis has been done using Erdas Imagine 8.4, 

ArcGIS 9 and Expert Choice softwares. Weighted approach using Pairwise 

Comparison Method and Analytical Hierarchy Process (AHP) of nine-degree to 

produce ecotourism suitability scenario 1 and AHP of three-degree to produce 

ecotourism suitability scenario 2 was also used to identify the areas having 

ecotourism potential. The multicriteria spatial decision support model has been 

prepared and the weighted layers are overlayed in ArcGIS 9 to identify the 

ecotourism potential sites. Subsequently, this study produces 2 scenarios for a 

suitability map showing five ecotourism potential classes viz. very high, high, 

moderate, low and very low for ecotourism planning purposes. The results can be 

presented to relevant authorities so that further action on planning and managing the 

environment along with ecotourism of Kuala Selangor can be advanced and 

sustainability can be accomplished 

Key Words 

Remote Sensing, GIS, Ecotourism Suitability, Multicriteria Spatial 

Decision Support System 


REMOTE SENSING AND GIS FOR LAND SUITABILITY: AN APPLICATION FOR ECOTOURISM 

INTRODUCTION 

Ecotourism is increasing throughout the world, particularly in species-rich 

tropical countries. It is seen as a tool for conservation as well as ecologically 

sustainable development. In Malaysia, ecotourism is the fastest growing form of 

tourism, averaging 35 % per year, which currently makes up about 10 percent of the 

country’s tourism revenue. Nevertheless, it is imperative that the level, type, and 

management of ecotourism are appropriate so that it will not result in the 

deterioration of Malaysia’s environment or culture. 

Ecotourism is a concept that evolved over the last 25 years as the 

conservation of community for people living and around the protected areas, and as 

the travel industry witnessed a boom in nature tourism the realization of their mutual 

interests directed its growth (Drumm & Moore 2002). According to experts the 

emergence of ecotourism were contributed by two key components. Firstly, 

ecotourism is connected to the significant environmental movement of the 1970’s 

and 1980’s. Secondly, ecotourism was driven by a great dissatisfaction with mass 

tourism and its overdevelopment, environmental pollution, and the invasion of 

culturally insensitive as well as economically disruptive foreigners (Honey 1999; 

Orams 1995). The global drive towards ecotourism is due to its promise of 

achieving conservation goals, improving the well-being of local communities and 

generating new business-promising a rare win-win-win situation. Ecotourism has 

emerged as a successful platform to establish partnerships and to jointly guide the 

path of tourists seeking to experience and learn about natural areas and diverse 

cultures. 

Roberts & Thanos (2003) state that currently, ecotourism is the fastest growing 

sector of the global tourism industry. Honey (1999) corroborates this statement by 

pointing out that most estimates indicate that the demand for ecotourism is on the 

rise at an annual rate of 10 to 30 percent. Even though ecotourism is not the ultimate 

universal remedy; but it holds huge potential to promote sustainable development 

and this definitely merits considerable attention. 

According to Weaver (2005) from the various efforts made to define ecotourism, 

there are three core elements of ecotourism can be extracted. 

1. Ecotourism is an essentially a form of nature-based tourism 

2. Ecotourism emphasises learning as a result of the interaction between 

ecotourists and the natural environment 

3. Ecotourism has product planning and management that is conducive to 

sustainability 



THE EMERGENCE OF ECOTOURISM IN MALAYSIA 


Nedal, A. Mohammad, et al 

Until the 1970s tourism was not regarded as an important economic activity 

in Malaysia. Though set up in 1972 with statutory responsibility to act as a 

development authority, the Tourism Development Corporation of Malaysia (TDC) 

was given low priority. Since the 1980s, tourism has developed into an ever more 

important industry worldwide with investments in new facilities and capital 

equipment reaching US$ 350 million per year. As a result of such phenomenon, the 

tourism industry swiftly became one of the largest earners of foreign exchange in 

Malaysia, second only to the manufacturing industry. Therefore, like many other 

developing nations that yearn for progress and economic development, Malaysia 

was quick to embrace tourism. This is evident in the figures 1 and 2: 

tourist Arrivals (Million) 

18 

16 

14 

12 

10 

8 

6 

4 

2 

0 

7.93 

10.22 

12.78 

Figure 1: International Tourists Arrival in Malaysia (1999 – 2005) 

Source Data: http://www.apec-tourism.org/tin.php, Tourism Malaysia 

13.29 

10.58 

1999 2000 2001 2002 2003 2004 2005 

Year 

15.7 

16.43 

301


Tourist Receipts (Million) 

35,000.00 

30,000.00 

25,000.00 

20,000.00 

15,000.00 

10,000.00 

5,000.00 

0.00 

12,321.30 

17,335.40 

24,221.50 

Figure 2: International Tourists Receipts in Malaysia (1999 – 2005) 

Source Data: http://www.apec-tourism.org/tin.php, Tourism Malaysia 

Parallel to development of global tourism trend and intense international 

promotional strategies by TDC, Malaysia also received its wave of Mass Tourism in 

the late 80s and early 90s – the number of international tourist arrival had leaped up 

to 53.6% with a revenue growth of 60.5%. The decline in the number of 

international tourist arrival in Malaysia in the years 1991, 1999 and 2003 were 

actually upshots from global incidents. 1991 witnessed the Gulf War, 1999 was the 

aftermath of a world economic crisis and 2003 was the result of SARS epidemic in 

Southeast Asia. Nonetheless, the number of arrivals continues to accelerate once the 

adversity was overcame. 

Without doubt, with the advent of mass tourism, cities in Malaysia rapidly became 

overpopulated and towns became crowded. Fuelled by the desire to escape the 

bustling, crowded and stressful city life, many tourists or travellers, international and 

local, resorted to natural and relatively undisturbed areas in quest of peace and 

solitude as well as to study or seek adventure in nature. 

For the past 30 years the number of visitors coming to Malaysia to see its wildlife, 

scenery, forests and beaches has steadily been increasing. Places such as Kampung 

25,781.10 

21,291.10 



29,651.40 

1999 2000 2001 2002 2003 2004 2005 

Year 

31,954.10



Kuantan fireflies, Rantau Abang, Niah Cave, Sipadan Island, Taman Negara, 

Kinabalu Park, Fraser’s Hill and Cameron Highland have long enjoyed streams of 

international and local visitors carrying out activities that could be defined as 

ecotourism. Nevertheless, awareness to this market component from tourism sector 

itself has been insignificant and irregular. 

In realising that Malaysia has a wide range of natural and cultural assets – 

undulating verdant virgin jungles, sun drenched pristine powdery beaches, an 

abundance of unique flora and fauna – the Malaysian Ministry of Culture, Arts and 

Tourism (MOCAT) officially open its doors to ecotourism in the late 1990s. 

Abiding by IUCN, Malaysia’s definition of ecotourism equally stresses: 

• low impact on the environment and on local culture 

• encompassing nature as well as culture 

• averting damage as much as possible 

• restoring inevitable damage 

• benefits for the people of the area 

By 1996, MOCAT outlined a 6-part-strategy called Malaysian National Ecotourism 

Plan which consists of 25 guidelines, some of which covers the aspects of: 

• Categorising Sites and Activities 

• Carrying Capacity and Limits of Acceptable Change 

• Marine Parks and Islands 

• National Parks and Reserves 

• Interpretation, Education and Guide Training 

• Use of Local Accommodations 

• Accreditations of Eco-Tourism Products 

• Visitors’ Roles and Responsibilities 

• Contributing to Conservation Programmes 

• Ensuring Local Community Participation 

Source: Malaysian National Ecotourism Plan 1996. 

Among the entire tourism industry in Malaysia, the progress of ecotourism is fast 

and remarkable with an average growth of 35% per annum. The National Ecotourism Plan, 

drawn up in 1996, was aimed at providing the framework for the development of 

ecotourism, as highlighted in the seventh Malaysian plan (1996 – 2000). During such plan 

period, several significant ecotourism projects were implemented such as Wang Kelian 

State Park in Perlis, Tasik Bera in Pahang and Taman Hidupan Liar Hilir Kinabatangan in 

Sabah. Also, in the 8th Malaysian Plan (2001 – 2005), the Malaysian government 

conveyed ecotourism as having the ability to improve rural economy. 

By the year 2000, 1.25 million overseas tourists were estimated to visit 

Malaysia for observing and enjoying nature. These groups brought in more than RM 

1.57 billion in the same year. This represented approximately 10% of revenue from 

Malaysia’s tourism industry. 

303


National Ecotourism Plan 1996 states current survey indicated Malaysia 

already has 54 protected areas of more than 1,000 hectares or about 4.5% of the 

country’s land surface. Using the IUCN World Conservation Union classification, 

these protected areas include 16 national parks, nine managed nature 

reserves/wildlife sanctuaries, one protected landscape, and 28 strict nature reserves 

(see Malaysian National Ecotourism Plan: Part 4, 1996). 

It is evident that Malaysia stands to benefit economically, socially and 

environmentally from a properly implemented ecotourism. In the long term, these 

promising benefits include: 

• augmented revenue and work opportunities in the industry from increased 

number of tourists 

• economic benefits to rural communities neighbouring an ecotourism 

location, via job creation, improved infrastructure and communication 

facilities, as well as consequential developments 

• support for conservation and environmental management 

• an enhanced tourism, and consequently, national image 

THE STUDY AREA 

The chosen study area is Kuala Selangor, one of the northern districts of 

Selangor, 64 km northwest of the city of Kuala Lumpur (see map in figure 3). The 

study area lies between 3° 10´ N to 3° 34´ N and between 101° 6´ E to 101° 30´ E 

with an area of 1192.9 km². The development of Selangor state has extended into its 

surrounding districts, whose natural resources are coming under increasing pressure. 

The chosen area still has several patches of upland forest, swamp forest as well as 

some mangroves. The other major land use in the study area is agriculture, 

especially oil palm, paddy and coconut 

Sungai Selangor (Selangor River) is one of the main rivers in the state of 

Selangor. The headwater of Sungai Selangor originates from the highland area of the 

Titiwangsa Range, the backbone of Peninsular Malaysia. It is an important source of 

water supply for domestic and agriculture use as well as fishing industries for 

communities along the riverbanks. The Selangor River is still in a pristine and natural 

state in most places especially in the upstream reaches. It also provides recreational 

opportunities as well as posting challenges to the intrepid travellers, ecotourists and the 

adventure-seekers (i.e. river tubing, whitewater rafting and kayaking). 

Kuala Selangor’s mangrove forests are identified as having a potentially high 

value for wildlife conservation and ecotourism. Interestingly, this riparian mangrove is 

also home to large colonies of fireflies that glow in the dark thus providing a brilliant 

illumination at night. Its nature park is renowned for a variety of flag species of fauna 

and flora such as the Silverleaf Monkeys, rare species of lizards and birds. 



100°40'34"E 

3°42'40"N 

3°24'40"N 

3°6'40"N 

2°30'40"N 2°48'40"N 

METHODOLOGY 

100°52'34"E 

101°4'34"E 

Kuala 

Selangor 

Legend 

101°16'34"E 

101°28'34"E 

STATE 

NEGERI SEMBILAN 

PAHANG 

PERAK 

SELANGOR 

SELAT MELAKA 

KUALA LUMPUR 

0 5 10 20 30 40 

Kilometers 

Figure 3: Location of the Study Area 


101°40'34"E 


101°52'34"E 

0 5 10 20 30 40 

Kilometers 

102°4'34"E 

Satellite Remote Sensing data (SPOT XS 2005 DATA) purchased from 

Earth Observation Center (EOC), Geography Department, University 

Kebangsaan Malaysia (UKM) have been utilized to produce landuse/cover map. 

NDVI map has been generated for better distinction of forest and other landuse 

classes. Land use/cover classes are very much useful for the identification of 

ecotourism potential sites. There are in all nine and the weights have been given 

on the basis of the relative importance of land use/cover classes from the point 

of view of ecotourism development. Soil map has been utilized prepared by 

Agriculture Department, Malaysia. According to this map three types of soil 

found in this area. Erosion risk map prepared as erosion is a problem of high 

305 

3°42'40"N 

3°24'40"N 

3°6'40"N 

2°48'40"N


significance occurring on land dedicated to forestry, transport and recreation. 

Erosion leads to environmental damage through pollution, sedimentation and 

increased flooding (Morgan 2005). The accessibility map also prepared 

according to transportation map of Malaysia 2003. These maps are taken as the 

parameters to identify the potential ecotourism sites. 

The analysis has been done using Erdas Imagine 8.4, ArcGIS 9 and 

Expert Choice softwares. Weighted approach using Paiwise Comparison Method 

and Analytical Hierarchy Process (AHP) of nine-degree and AHP of threedegree 

was used to calculate the evaluation criteria weights. This has been done 

to assign relative weights in accordance to its influence/importance for the 

identification of potential ecotourism sites.Expert Choice software also used to 

calculatethe weights of the evaluation criteria. The multicriteria spatial decision 

support model has been prepared and the weighted layers are overlayed in 

ArcGIS 9 to identify the ecotourism potential sites. Subsequently, this study 

produces 2 scenarios for a suitability map showing five ecotourism potential 

classes viz. very high, high, moderate, low and very low for ecotourism planning 

purposes. Figure 4 showing the methodology flowchart. AHP of nine-degree 

used to produce ecotourism suitability scenario 1 and AHP of three-degree used 

to produce ecotourism suitability scenario 2. 

Figure 4: Methodology Framework. 




Scores of Suitability Factors 



Scoring factors used in the identification potential ecotourism sites can be 

measured using interval or ratio values (scores) or ordinal values (score classes). 

The choice of either scores or score classes can distinctly impact the efficiency of 

scoring factors. There is little consistency in the use of terminology in the suitability 

analysis literature, and often the term ‘score’ is used to refer to interval /ratio and 

ordinal values. 

Ratio values are the highest level, which have ordinal and interval 

properties, as well as the property of ratios. Just as the interval values, 

corresponding ratios on different parts of a ratio scale have similar meaning. 

For example, the ratio between two objects with values of ‘10’ and ‘5’ is 

equivalent to the ratio between two objects with values of ‘200’ and ‘100’. 

From the above opinions it can be showed that ratio values are more 

compatible for scoring suitability factors in this suitability analysis. The 

description of suitability classes within factors and corresponding scores are 

listed in Table 1. The scores of ‘5, 4, 3, 2, 1’ are used to distinguish the 

differences among very high suitability, high suitability, moderate suitability, 

low suitability and very low suitability. The maps of suitability classes and 

scores in every factor (single factor map) are presented from figure 5 to figure 

9. They have been converted into vector format compatible for the operation in 

ArcView Model Builder. 

307


Table 1: The description of suitability classes within factors and corresponding scores 

Factor (Criteria) Suitability Class Score 

Landuse/Landcover Mangrove forest 5 

Peat Swamp Forest 5 

Upland Forest 5 

Water Bodies 4 

Mixed Horticulture 3 

Oil Palm Plantation 1 

Coconut Plantation 1 

Paddy Fields Restricted 

Urban and Associated Areas 1 

Vegetation Density (NDVI) Very High 5 

High 4 

Moderate 3 

Low 2 

Non Vegetation 1 

Erosion Risk Very High 1 

High 2 

Moderately High 3 

Moderate 4 

Low 5 

Soil Suitability Class 2 (S 2) 4 

Class 4 (S 4) 2 

Class 5 (S 5) 1 

Accessibility (Road network) 0 – 300 m 1 

300 – 3000 m 5 

3000 – 4000 m 4 

4000 – 5000 m 3 

5000 – 6000 m 2 





309






Criterion Weight Using Analytical Hierarchy Process (AHP) of nine-degree 

The method used in this study originates from the technique developed by 

Saaty. It is a parwise comparison in the context of decision-making process known 

as Analytical Hierarchy Process (AHP). Though first developed for complex 

decision making for business and corporate leaders (Saaty 2006), this technique has 

since gained popularity in spatial analysis for example Eastman (1995). 

Saaty proposed for matrix to be used for pairwise comparison as it is a 

simple and well established tool that offers a framework for testing consistency, 

gaining additional information through making all possible comparisons and 

analysing the sensitivity of overall priorities to changes in judgment. The basis of 

this technique is to compare one element with the others, one at a time. Therefore, 

AHP is a logical framework that basically extracts a relative weight for each 

objective by computing an eigenvector for that objective. The eigenvector is 

computed from a square matrix that is populated with pairwise comparisons of the 

relative importance of each objective to the overall goal – meaning the user has to 

provide these judgments for the model to be evaluated (Beedasy & Whyatt 1999). 

The actual extraction of data can either be verbal, graphical or numerical. 

Saaty has developed the scale for this pairwise comparison ranging from 1 – 9, with 

1 being “equal importance of both elements” and 9 is “absolute importance of one 

element over another. Table 2 shows the verbal scale. The method used to calculate 

the factors or criteria weights is shown in figure 10 below. 

311


Table 2: The Verbal Scale for Pairwise Comparison Method 

Intensity of importance Definition 

1 Equal importance 

2 Equal to moderate importance 

3 Moderate importance 

4 Moderate to strong importance 

5 Strong importance 

6 Strong to very strong importance 

7 Very strong importance 

8 Very to extremely strong importance 

9 Extreme importance 

Source: Saaty (1980) in Malczewski, (1999) 

Evaluation Criteria for the 

Ecotourism Potential Sites 

Development of the Pairwise Comparison Matrix 

The method employs an understanding scale with values from 1 to 9 to rate the 

relative preferences for two criteria 

Computation of the Criterion Weights 

This step involves the following operations: 

1. Sum the values in each column of the pairwise comparison matrix 

2. Divide each element in the matrix by its column total ( the resulting matrix 

is referred to as the normalized pairwise comparison matrix) 

3. compute the average of the elements in each row of the normalised matrix 

Estimation of the Consistency Ratio 

It involves the following operations: 

1. Determine the weighted sum vector 

2. Determine the consistency vector 

Figure 10: The Pairwise Comparison Method Scheme as adapted from Malczewski (1999) 





The final criteria weights calculated from the pairwise comparison method 

using AHP is shown in table 3. In this study, Expert choice 11 software is also used 

to calculate the criterion weights. The results of the calculation from the software are 

as in figure 11. 

Table 3: Criterion Weight for Potential Ecotourism Sites Using the Pairwise 

Comparison of Nine-degree 

Factors (Evaluation Criteria) Factor or Criterion Weight 

Landuse/landcover 41.9 

Vegetation Density 31.0 

Soil Suitability 16.0 

Accessibility (Road Network) 6.9 

Erosion Risk 4.2 

Figure 11: Criterion Weight for Potential Ecotourism Sites Using Expert Choice 11 

Beedasy & Whyatt (1999) concluded that the AHP is essentially a ratiobased 

pairwise comparison, meaning that if objective X is considered twice as 

important as objective Y, then it is understood that objective Y is held to be half as 

important as objective X. While this facilitate consistency among a pair of 

objectives, it does not guarantee consistency among a set of objectives (i.e., if X is 

twice as important as Y, which is three times as important as Z, then a comparison 

between X and Z should yield a factor of 6 - which may not always emerge for the 

user). To counter this, the AHP develops and reports an inconsistency index. For an 

313


index greater than 0.1, it is recommended to reconsider the matrix. The smaller the 

index, the less inconsistent is the preference matrix. 

The acceptable consistency ratio (CR) range varies according to the size of 

matrix – i.e. 0.05 for a 3 by 3 matrix, 0.08 for a 4 by 4 matrix and 0.1 for all larger 

matrices, n more than or equal 5 (Saaty 2000). If the value of CR is equal to or less 

than that value, it means that the evaluation within the matrix is acceptable or shows 

a good level of consistency in the comparative judgments represented in the matrix. 

Therefore, this matrix is acceptable because the CR = 0.03, which is less than 0.10 

for matrix of 5 by 5. 

Criterion Weight Using Analytical Hierarchy Process (AHP) of Three-degree 

The analysis above shows one of the disadvantages of hierarchic analysis of 

nine-degree, which is relatively difficult to convert the comparison judgments into 

numerical values according to the nine degree scale. It is easier to only judge the 

more important, same important and less important these three classes in the 

pairwise comparison. Thus the method of hierarchic analysis of three-degree is 

developed. The first step of hierarchic analysis of three-degree is to make a 

comparison matrix based on the three-degree pairwise comparison. Then a structural 

judgment matrix is generated and after the normalization, all the factors weights will 

be represented by the eigenvector of the largest eigenvalue derived from the 

structural judgment matrix. Table 4 shows the criterion weight for potential 

ecotourism sites using the pairwise comparison of three-degree 

Table 4: Criterion Weight for Potential Ecotourism Sites Using the Pairwise 

Comparison of Three Degree 

Factors (Evaluation Criteria) Criterion Weights 

Landuse and Landcover 0.5167 

Vegetation Density 0.2584 

Soil Suitability 0.1296 

Accessibility (Road Network) 0.0632 

Erosion Risk 0.0311 

Multicriteria Spatial Decision Support Model: Model Building 

The potential ecotourism sites suitability model was created using ArcView 

3.2 Spatial Analyst Extension and Model Builder. The model was designed to be 

flexible so input variables could be added or removed, and parameters within data 

layers could be rescaled. Figure 12 is an overview of the potential ecotourism sites 

suitability model, which because of its graphical nature, allows the GIS analyst to 





see every initial data input (blue ovals) and intermediate calculation (yellow 

rectangles). The final column of intermediate output files (green ovals) are the input 

predictor variables for the Weighted Overlay tool, which executes the suitability 

analysis. The Weighted Overlay tool requires all input data to be in vector format 

and represent categorical data. The Weighted Overlay tool allows the analyst to: 

1. Set the relative importance that every input data layer has on the suitability 

analysis by assigning weights (or “percent of influence”); and 

2. To scale (or “rank”) the categorical values within data layers. 

Figure 13 shows the interface of the Weighted Overlay tool where weights 

(% Influence) and scaled values are altered between and within input data layers, 

respectively. Finally, the “overlay” procedure multiplies across all given weights 

and score values to produce a suitability map of five ecotourism potential classes 

viz. very high, high, moderate, low and very low created for ecotourism planning 

purposes and ecotourism development for Kuala Selangor (figure 14). 

Figure 12: The potential ecotourism sites suitability graphical model 

developed using Model Builder. All input data are on the left (blue ovals), followed 

by a series of intermediate data manipulations and calculations (yellow rectangles) 

and temporary output files (green ovals) that converge into the Weighted Overlay 

tool to derive the map output for the suitability analysis (final green oval). 

315


Figure 13: The interface for the Weighted Overlay tool is used to assign 

weights (% Influence) between each input raster data file and to scale (or rank) the 

values within the data files. All input raster files received an equal percent of 

influence in the model. 

Results of the ecotourism suitability scenario 1 figure 15 shows that most of 

the study area 46.3% comes under highly potential ecotourism spot, followed by 

37.6% under moderately potential and 10.7% under low potential. It is noted that the 

very high ecotourism potential covers 4.2% of the study area which located on the 

mangrove forest around the Selangor River, and peat swamp forest. 1.2% of the 

study area comes under very low potential is suitable for development of ecotourism 

infrastructure. It is imperative that the level, type and management of ecotourism are 

appropriate. Otherwise, it could result in the deterioration of the environment or 

culture around which it revolves, and ultimately destroy itself. Therefore, it is 

important to pay attention to carrying capacity, limits of acceptable change (LAC) 

and have synergy among all relevant parties. 



Area (km2) 

600 

500 

400 

300 

200 

100 

0 

Very Low 

Siutability 

Low Suitability Moderate 

Suitability 

Ecotourism Suitability 

High Suitable Very High 

Suitable 



Restricted 

Figure 14: Results of Potential Ecotourism Sites (Suitability Scenario 1) 

317


101°4'55"E 

3°35'10"N 

3°26'10"N 

3°17'10"N 

3°8'10"N 

Paddy 

Fields 

Cocon 

Mangrove 

Forest 

Urban 

Areas 

Oil Palm 

Plantation 

Legend 

101°12'43"E 

Very High Suitable 

High Suitable 

Moderate Suitability 

101°12'43"E 

101°20'31"E 

Peat Swamp 

Forest 

101°20'31"E 

Figure 15 Ecotourism Potential Sites Suitability Map for Kuala Selangor istrict, 

Malaysia. Results from Suitability Scenario 1 



101°28'19"E 

Low Suitablity 

Very Low Suitability 

Restricted 0 2 4 8 12 16 

Kilometers 

101°28'19"E 

3°35'10"N 

3°26'10"N 

3°17'10"N 

3°8'10"N



In ecotourism suitability scenario 1, it is visible that the mangrove forest is 

divided into two parts. The part around the Selangor River is deemed very highly 

suitable for ecotourism covering an area of 10.29 km 2 , while the part near the beach 

of the Straits of Malacca covering 18.06 km 2 is deemed highly suitable for 

ecotourism. The reason for variance in suitability is the type of soil available in the 

area examined. The type of soil in the mangrove of the beach of the Straits of 

Malacca is a class 5 soil, which is soil with at least one very serious limitation (high 

salinity) to crop growth. The area of mangroves in tropical countries in general and 

in Malaysia in particular is on a continuous decline. This decline is even more 

visible in Selangor, being the most modern and industrialized state in Malaysia. The 

best suggested practice to manage this decline is to allow it to continue under 

primary or regeneration forest. 

The development of Kuala Lumpur and Selangor state has extended into the 

surrounding districts, whose natural resources are coming under increasing pressure. 

This pressure imposed by Kuala Lumpur and surrounding areas made it harder to 

maintain the mangrove as many factors here affect it. These factors include 

urbanization, infrastructure development, population growth, the irresponsibility of 

the local community having to made any use of it’s natural resources. Furthermore it 

is hard to protect the mangrove from illegal logging. The best action to be taken in 

this case is to open some of the mangrove areas as ecotourism destinations. The 

revenue generated from ecotourism can be used for conservation efforts for this 

forest and increase the local community awareness to protect their natural resources, 

as the local community living in and around the area will be economically benefited 

from the revenue generated by ecotourism. 

The same case can be noticed for the peat swamp forest. Some areas were 

deemed very highly ecotourism suitable covering an area of 36.61km 2 , other areas 

where deemed highly ecotourism suitable covering 325.72 km 2 . The reason for this 

inconsistency is the accessibility. Most of the oil palm and coconut plantations come 

under moderate ecotourism suitability. The urban and associated areas have low 

potential for ecotourism and are suitable for ecotourism infrastructure. 

The importance of existing landuse/landcover for any macro or micro 

development plan stem from the weight which we got through pairwise comparison 

method. The weight gained through that method has roughly 41.65% of the total 

weight, making it the most important factor (evaluation criteria). Landuse/landcover 

is intersected with ecotourism scenario 1 because of its importance. 

Results of the ecotourism suitability scenario 2 figure 16 and table 5 shows 

that most of the study area 36.8% comes under moderately potential ecotourism 

spot, followed by 29.1% under high potential and 20.4% under low potential. It is 

noted that the very high ecotourism potential covers 6.7% of the study area which 

319


located on the mangrove forest around the Selangor River, and peat swamp forest. 

1.6% of the study area comes under very low potential which is suitable for 

development of ecotourism infrastructure. 

Table 5: Results of Potential Ecotourism Sites (Suitability Scenario 2) 

Suitability Classes Area (km 2 ) Area Percentage (%) 

Very Low Suitability 18.77 1.6 

Low Suitability 244.43 20.4 

Moderate Suitability 439.23 36.8 

High Suitability 347.27 29.1 

Very High Suitability 78.44 6.7 

Restricted (Not 

Suitable) 

64.48 5.4 

Area (km2) 

450 

400 

350 

300 

250 

200 

150 

100 

50 

0 

Very Low 

Siutability 

Low 

Suitability 

Moderate 

Suitability 

Ecotourism Suitability 

High Suitable Very High 

Suitable 

Figure 16: Results of Potential Ecotourism Sites (Suitability Scenario 2) 

In ecotourism suitability scenario 2, it is visible that the area of the very high 

ecotourism suitability parts of the forest increased when compared to the previous scenario 

as shown in table 4.27 and figure 4.39. This is the result of using pairwise comparison 

methods of three-degree which gives the existing landuse/landcover factor a higher weight 

than that given in the pairwise comparison method of nine-degree. 



Restricted

101°4'55"E 

3°35'10"N 

3°26'10"N 

3°17'10"N 

3°8'10"N 

± 

Legend 

101°12'43"E 

Very High Suitable 

High Suitable 

Moderate Suitability 

Low Suitability 

Very Low Suitability 

Restricted 

101°12'43"E 

101°20'31"E 

Ecotourism Suitability Scenario 

101°20'31"E 



101°28'19"E 

0 2 4 8 12 16 

Kilometers 

101°28'19"E 

Figure 17: Ecotourism Potential Sites Suitability Map for Kuala Selangor 

District, Malaysia. Rtesults from Scenario 2 

321 

3°35'10"N 

3°26'10"N 

3°17'10"N 

3°8'10"N


CONCLUSION 

In order to optimise the benefits of ecotourism and prevent, or at least, 

mitigate any problems that might be generated, good planning and careful 

management of the ecotourism is essential. More generally, planning for tourism is 

as important as planning for any type of development in order for it to be successful 

without creating problems. The ecotourism sector objectives can be accomplished 

more effectively if they are carefully planned and prudently integrated into the 

country’s total development plan and program. 

With the aid of remote sensing and GIS applications, coupled with MCDM, 

the result for the chosen research area – Kuala Selangor – is obtained more 

promptly, accurately and visually. The results (as mentioned earlier) can now be 

presented to relevant authorities so that further action on planning and managing the 

environment along with ecotourism of Kuala Selangor can be advanced and 

sustainability can be accomplished. 

This research stresses that it is imperative that the level, type and 

management of ecotourism are appropriate. Otherwise, it could result in the 

deterioration of the environment or culture around which it revolves, and ultimately 

destroying itself. Therefore, it is imperative to pay attention to carrying capacity and 

limits of acceptable change (LAC) in addition to having a synergy among all 

relevant parties. 

REFERENCES 

1. Beedasy, J. & Whyatt, D. (1999). Diverting the tourists: A spatial decisionsupport 

system for tourism planning on a developing island: ITC-Journal, 

(3-4), 163-174. 

2. Drumm, A. & Moore, A. in Singer, A (Ed). (2002). Ecotourism 

Development A Manual for Conservation Planners and Managers. Volume 

1. The Nature Conservancy, Arlington, Virginia, USA. 

3. Eagles,P.J. (1997)."International Ecotourism Management : Using 

Australia and Africa as Case Studies, Protected Areas in the 21st Century: 

From Islands to Networks, Albany, Australia, Nov.1997"- Paper prepared 

for the IUCN World Commission on Protected Areas, Protected Areas in 

the 21st Century:From Islands to Networks,Albany, Australia 





4. Eastman, J. et al. (1995). Raster procedures for multi-criteria/multi 

objective decisions. Photogrammetric Engineering & Remote Sensing. 

61(1995):539 – 547. 

5. Honey, M. (1999). Ecotourism and Sustainable Development: Who Owns 

Paradise? Washington, D.C. Covelo, CA: Island Press. 

6. http://www.apec-tourism.org/tin.php. Online document. Accessed on 12 

June 2006. 

7. Morgan. R. (2005). Soil Erosion and Conservation. Third Edition. 

Blackwell Puplishing Ltd. UK. 

8. National Ecotourism Plan, Malaysia. (1996). Ministry of Culture, Arts, and 

Tourism. Government of Malaysia. 

9. Orams, B. (1995). Towards a more desirable form of ecotourism. Tourism 

Management. 16(1), 3-8. 

10. Roberts, T. & Thanos, N. (2003). Trouble in Paradise: Globalization and 

Environmental Crises in Latin America. New York: Routledge. 

11. Saaty, T. (2006). Rank from comparisons and from ratings in the analytic 

hierarchy/network processes. European Journal of Operational Research. 

168 (2006) 557–570. 

12. Weaver, D. (2005). Comprehensive and minimalist dimensions of 

ecotourism. Annals of Tourism Research, Vol. 32, No. 2, pp. 439–455. 

323




EVALUATION OF RAP AS A CEMENT-TREATED BASE FOR 

HIGHWAYS CONSTRUCTION 

Mohamed A. Elshabrawy 1 , Al-Hosain M. Ali 2 , Alaa R. Gaber 3 

1. Dean, faculty of Eng., Mansoura Univ., EGYPT 35116., 

mshabrawy@mans.edu.eg 

2. Assistant Prof., Dept. of Civil Eng., Mansoura Univ., EGYPT 35116. 

aali0418@yahoo.com. 

3. Researcher, Dept. of Civil Eng., Mansoura Univ., EGYPT 35116. 

eng_alaa1400@yahoo.com 

ABSTRACT 

Pavement rehabilitation and reconstruction generate large quantities of 

Reclaimed Asphalt Pavement (RAP). Recycling of RAP into new asphalt mixtures 

requires significant adjustments to asphalt plants, equipment, and field operations. 

RAP acceptance in road bases and subbases has still been limited because of the lack 

of laboratory and field performance data. RAP Cement-Treated Base (RAPCTB) is a 

new dimension in highway construction. 

This paper presents the results of a laboratory study for RAPCTB 

characterization. Factorial experiments include various evaluation tools, variable 

RAP percentages, and different sample conditioning times. Among evaluation 

tools were Unconfined Compressive Strength, California Bearing Ratio (CBR) 

and Static Modulus of Elasticity. RAP was used at different levels including 0, 

30, 50, 70, and 100 by total aggregate weight. RAPCTB samples were prepared 

using two Portland cement (Type I) contents, 150 and 300 Kg/m 3 (6.5 and 13.0 

percentages by RAPCTB total weight). Different conditioning times including 3, 

7, 28 and 56 days were performed. Allowable Equivalent total Single Axle Load 

(ESAL) for pavements with RAPCTB was found to be 2 and 10 times higher 

than that for control pavement in case of Portland cement contents of 150 and 

300 Kg/m 3 , respectively. 

KEY WORDS 

RAP, Cement-Treated Base, Pavement, Recycling 


EVALUATION OF RAP AS A CEMENT-TREATED BASE FOR HIGHWAYS CONSTRUCTION 

INTRODUCTION 

Reconstruction of highways produce large amount of asphalt material wastes 

that require disposal and also causing environmental imbalance and other health 

hazard. Furthermore, the disposal of waste materials is an expensive operation. This 

problem is more alarming in developing countries, like Egypt, where suitable 

technologies of disposal are not fully applied. RAPCTB is defined as a mixture of 

RAP and measured amount of hydraulic binder (Portland cement) and water that 

hardens after compaction and curing to form a durable paving material. It could be 

used as a base course for either flexible or rigid pavements. 

Recycling pavement materials has proved to be a feasible process to 

rehabilitate worn-out pavements. Recycling will be cost-effective when 

reconstructing or rehabilitating existing pavements and other layers such as base. 

Recently the use of recycled pavement materials in highway reconstruction as 

RAPCTB has gained attention for the following reasons; conservation of resources, 

preservation of the environment, and decrease the costs by reusing pavement 

materials. A previous study conducted at Sultan Qaboos University indicated that 

RAP-virgin aggregate mixtures could be utilized in the base layer. Taha and Al- 

Harthy ,2002, evaluated the use of cement stabilized RAP and RAP-virgin aggregate 

blends as base materials. Results indicated that the optimum moisture content, 

maximum dry density, and strength of RAP will generally increase with the addition 

of virgin aggregate and cement. 

Lim and Zollinger, 2003, evaluated the development of strength and 

modulus of elasticity of Cement-Treated Aggregate Base (CTAB) materials. 

Test results indicated that the relationship between the compressive strength and 

elastic modulus of CTAB materials could be expressed in a single equation 

regardless of aggregate type and mixture proportions. Sayed et al. 1993, 

evaluated the use of RAP as a base material for paved shoulders. Results proved 

that RAP is a well-graded material, and its maximum dry density is comparable 

to those of other conventional granular materials. Laboratory and field 

evaluations of the use of RAP in road base and subbase applications were 

conducted by Maher and Popp, 1997. Results showed that RAP has a slightly 

higher resilient modulus and field elastic modulus than the dense-graded 

aggregate. Field performance was comparable to that of a crushed stone base. 

Early studies recommendations for RAPCTB gradation are shown in Table 1. 




Mohamed A. Elshabrawy, et al 

Table 1Recommendations for RAPCTB Gradation 

Percent Passing (RAPCTB Maximum Size) 

Sieve Size 

1” (25.0 mm) 1 ½” (37.5 mm) 2” (50.0 mm) 

2 in (50.0 mm) 100 

1–1/2 in (37.5 mm) 100 

1 in (25.0 mm) 100 70–95 55–85 

3/4 in (19.0 mm) 70–100 55–85 50–80 

No. 4 (4.75 mm) 35–65 30–60 30–60 

No. 40 (0.450 mm) 14–30 10–30 10–30 

No. 200 (0.075 mm) 0–15 0–15 0–15 

MATERIALS USED 

RAP was obtained from a major highway with 12 years old pavement. 

Aggregate Base course of that old pavement which included crushed limestone 

coarse aggregates and sand were sampled and used in this investigation. Preliminary 

testing of recycled aggregates included gradation (AASHTO T-27), abrasion 

(AASHTO T-96), specific gravity, and water absorption (AASHTO T-85). Portland 

cement (Type I) conforming the requirements of ASTM C-150 was used. Water 

used in mixing and/or curing met the requirements of AASHTO T- 26. 

MATERIALS PROPERTIES 

Physical characteristics of raw materials and prepared RAPCTB mixtures 

were tested. Laboratory maximum dry density and corresponding optimum water 

contents (AASHTO T-180, Method C) for tested mixtures were evaluated. 

Unconfined compressive strength, California Bearing Ratio (CBR), and Static 

Modulus of Elasticity (E) were among evaluated properties. 

Sieve analysis results for RAP and recycled base aggregates are illustrated in 

Figure 1. Based on gradation results, RAP was classified as well-graded material 

(GW) with uniformity coefficient (Cu) of 10.8, and coefficient of curvature (Cc) of 

1.4. Recycled base aggregate was classified as well-graded gravel (GW) with 

uniformity coefficient of 120, and coefficient of curvature was 1.00. Results for Los 

Angles (LA) abrasion, specific gravity and absorption for both RAP extracted 

327


aggregate and base course recycled aggregate are shown in Table 2. Lower water 

absorption of extracted aggregate from RAP may be due to the presence of old 

absorbed bitumen inside aggregate porous. LA abrasion for RAP extracted 

aggregate (30.7%) was higher than that of recycled base aggregate (26.2) could be 

due to differences in aggregate source, durability and specific gravities. 

Aggregat 

e 

RAP 

Recycled 

Table 2 LA Abrasion, Specific Gravity and Absorption Results 

LA Abrasion 

Specific Gravity 

(%) Bulk SSD 1 % Water 

Apparent Absorption 

30.7 2.153 2.176 2.203 1.064 

26.2 2.516 2.552 2.610 1.439 

1 = Saturated Surface Dry 

RAPCTB SAMPLES COMPACTION 

Two RAPCTB blends were prepared using 150 and 300 Kg/m 3 contents of 

Portland cement, Type (I), for RAP percentages of 100, 70, 50, 30, and 0.0 % (By 

total aggregate weight). Samples were compacted according to AASHTO T-180 

standards. Table 3 shows maximum dry densities (γdmax) and corresponding 

Optimum Moisture Contents (OMC) for compacted RAPCTB blends. Results 

proved that as Portland cement content and recycled aggregate increased, the 

optimum moisture content and maximum dry density values were slightly increased. 

At higher RAP content, samples that were compacted at lower moisture contents 

would not remain intact upon removal from the mold. The addition of more 

aggregate and Portland cement made compaction and handling much easier. The 

moisture-holding capability of RAP is negligible, because there is little minus No. 

200 fraction, and most RAP aggregates are coated with asphalt. Recycled aggregate 

and Portland cement made compaction more effective by filling more fines among 

RAP coarse particles. Lack of fine materials in compacted RAP mixture resulted in a 

higher permeability that made water to drain out at the bottom of mold during 

compaction. It appears from the results that the content of coarse aggregates was the 

most influencing factor for the OMC and density of the RAPCTB mixtures. Thus, in 

field applications, the addition of recycled aggregates and cement would make it 

much easier to compact RAP as a base material. 





Table 3 Maximum Dry Density and Optimum Moisture Content Results 

Portland Cement 

(Kg/m 3 ) 

RAP, % γdmax, t/m³ 

OMC, % 

100 

2.011 

8.10 

70 

2.100 

7.00 

150 

50 

2.120 

6.60 

30 

2.160 

6.00 

0 

2.290 

7.10 

100 

2.083 

8.40 

70 

2.160 

8.20 

300 

50 

2.160 

6.80 

30 

2.220 

6.40 

0 

2.300 

7.50 

UNCONFINED COMPRESSIVE STRENGTH 

Sample preparation and compaction procedures for unconfined compressive 

strength testing were in accordance with ASTM D-558. Prepared samples were 

cylindrical in shape with diameter of 15 cm and height of 30 cm (Photo 1). Samples 

were wet cured for 3, 7, 28 and 56 days. Compression test was performed in 

accordance with ASTM D-1633 using a 100 ton testing machine with a loading rate 

of 0.3 ton per second. Samples vertical strain was monitored against applied 

compressive load. Figures 2 and 3 illustrate the unconfined compressive strength 

results versus curing time at Portland cement contents of 150 and 300 Kg/m 3 , 

respectively. Results indicated that as aggregate, Portland cement content, and 

curing period increased, RAPCTB compressive strength increased. 

At Portland cement content of 150 Kg/m 3 , compressive strengths at RAP 

content of 30, 50, and 70% were not significantly different. At RAP content of 

100%, compressive strength was significantly low due to lower density and 

higher voids content of RAPCTB. Highest compressive strength was achieved at 

0.0% RAP content. Raising the Portland cement content to 300 Kg/m 3 has 

increased the RAPCTB compressive strength at various RAP contents. At 

different RAP contents, overall average of compressive strength after 28 days 

curing time has been increased by about 68% when added Portland cement has 

been increased from 150 to 300 Kg/m 3 . 

329


Photo 1 RAPCTB Cylinder 

Samples for Compression Test 

CALIFORNIA BEARING RATIO (CBR) 

RAPCTB sample were prepared and compacted as discussed earlier and 

tested in accordance with ASTM D-1883. Samples size were 150 mm in diameter by 

115 mm in 

height. Two samples were prepared and tested for each RAPCTB mixture. 

Samples were dry cured for 1, 2, 6, 18 and 24 hours. Figures 4 and 5 present the 

CBR results versus curing time at Portland cement contents of 150 and 300 Kg/m 3 , 

respectively. CBR values have increased as curing time increased for variable RAP 

contents. RAPCTB with higher levels of RAP contents showed lower CBR values 

for different Portland cement contents. As Portland cement content increased from 

150 Kg/m 3 to 300 Kg/m 3 , the resulted CBR values of RAPCTB (after 24 hours 

curing time) were increased by percentages of 85, 38, 26, 22, 19 for blends with 

RAP contents of 100, 70, 50, 30, and 0.0%, , respectively. CBR values were 

increased by average rates of 4.2% and 5.4% per curing hour (for different RAP 

contents) at Portland cement contents of 150 and 300 Kg/m 3 , respectively. 

STATIC MODULUS OF ELASTICITY (E) 

Elastic modulus of RAPCTB was investigated using the stress-strain 

relationships developed during compression testing. It was determined as the initial 

secant modulus at 33 percent of the ultimate stress. Table 4 shows the calculated E 

for RAPCTB mixtures. It was noted that as aggregate and Portland cement contents 

increased, the elastic modulus value increased. 



Cement 

Content 

(Kg/m 3 ) 



Table 4 Initial Modulus of Elasticity (Curing Time = 7 Days) 

Initial Modulus of Elasticity (10 6 KN/m 2 ) at RAP Content (%) 

100 70 50 30 0.0 

150 3.80 4.70 5.90 7.00 9.80 

300 7.0 

8.50 

11.00 

12.00 

STRUCTURAL ANALYSIS OF RAPCTB AS A BASE COURSE 

331 

13.30 

Typical highway cross section in Egypt is shown in Figure 6 and consists of 

5 cm HMA surface course, 7 cm HMA base layer, and 25 cm aggregate base course 

on top of subgrade. Considering the recycling of a typical cross section into 

RAPCTB, RAP percentage will represent 30% by total RAPCTB weight. Based on 

filed bulk density of old pavement (before milling) and RAPCTB compaction results 

(Table 3), recycling the whole typical section will produce about 35 cm of a 

RAPCTB course. RAPCTB seven days’ compressive strength, at RAP content of 

30%, were 450 and 940 Kpa for Portland cement contents of 150 and 300 Kg/m 3 , 

respectively (Figures 2 and 3). Using correlation charts for estimating resilient 

modulus of bases based on compressive strength after seven days curing (Van Til et 

al), RAPCTB resilient modulus were found to be 4.5x10 6 and 6.2x10 6 Kpa for mixes 

with Portland cement contents of 150 and 300 Kg/m 3 , respectively. Layers Poisson 

ratios were assumed as shown in Figure 6. Single axle load with 40 KN wheel load 

and contact radius of 15 cm were used in this analysis. 

Horizontal tensile strain at the bottom of asphalt layer (εr) and vertical 

compressive stress on top of subgrade (σv) were investigated using KENPAVE 

software (Yang H. Huang). Results are given in Table 5. Since resilient modulus for 

RAPCTB has been found higher than that for HMA course, the resulted εr was 

compressive. In other words the probability for asphalt layer on top of RAPCTB to 

experience fatigue cracks will be significantly diminished compared with the control 

section that has tensile εr. Also, σv with RAPCTB presence were 12.2 and 10.3 Kpa 

for Portland cement contents of 150 and 300 Kg/m 3 , respectively. Compared with 

control section, RAPCTB with lower σv will experience lower pavement distress 

and rutting that might result from subgrade failure.


Table 5 Stress and Strain Results using KENPAVE Software 

Typical Cross Section with RAPCTB as Base Course 

Parameter Section with ABC Cement = 150 

(Control) Kg/m 3 

Cement = 300 

Kg/m 3 

εr (mm/mm) - 3.40x10 -4 +2.57x10 -5 + 2.40x10 -5 

σv (Kpa) 49.1 12.2 10.3 

TRAFFIC ANALYSIS OF FLEXIBLE PAVEMENT CONTAINING RAPCTB 

American Association of State Highway and Transportation Officials 

(AASHTO) modified equation (Yang H. Huang) given below was used to calculate 

the total allowable number of the 18-kips (80 kN) Equivalent Single Axle Load 

(ESAL) application (W18) for pavement sections illustrated in Figure 6. Resilient 

moduli shown in Figure 6 were used to determine pavement layers coefficient using 

correlation charts (Yang H. Huang). Structural Number (SN) was calculated using 

layer coefficients and pavement thicknesses following AASHTO method for flexible 

pavement design (Yang H. Huang). Calculated SN for control pavement section was 

2.98 while SN for pavement sections with RAPCTB were 3.31 and 4.13 for Portland 

cement contents of 150 and 300 Kg/m 3 , respectively. Present Serviceability Index 

loss (∆PSI) was assumed 2.5 and the Standard Deviation So of 0.45 was used. 

Design reliability was assumed to be 90% (Standard Normal Deviate ZR is -1.282). 

log[∆PSI/(4.2-1.5)] 

0.4 + 1094/(SN+1) 5.19 

log W18 = ZRSo + 9.36 log (SN+1) - 0.20 + + 2.32 log MR - 8.07 

Calculated allowable total numbers of ESAL are given in Table 6. With the 

same design period, structural capacity for pavements with RAPCTB in terms of 

allowable ESAL cumulative number has significantly increased compared to that for 

control pavement. Allowable ESAL for pavements with RAPCTB were found to be 

2 and 10 times higher than that for control pavement in case of Portland cement 

contents were used at 150 and 300 Kg/m 3 , respectively. 

Table 6 Allowable Total Number of ESAL using Modified AASHTO Equation 

Typical Section Section with RAPCTB as Base Course 

with ABC Cement = 150 

Kg/m 3 

Cement = 300 

Kg/m 3 

Total Number of 

ESAL (million) 

1.06 2.16 10.69 



CONCLUSIONS 



Reclaimed Asphalt Pavement (RAP) and recycled base aggregates were 

obtained from a 12 years old pavement and used in this study. Samples were 

prepared at RAP contents of 0, 30, 50, 70, and 100% by total aggregate weight. 

Portland cement, type I, was used at levels of 150 and 300 Kg/m 3 . Unconfined 

compression test, CBR, and static Modulus of Elasticity were among testing tools in 

RAPCTB characterization. Unconfined compressive strength, CBR, and modulus of 

elasticity have increased as aggregate and cement contents increased. Stress and 

strain analysis showed that pavement with RAPCTB, as a base course, gave lower 

σv that will reduce pavement distress and rutting due to subgrade failure. Allowable 

ESAL for pavements with RAPCTB were found to be 2 and 10 times higher than 

that for control pavement in case of Portland cement contents were used at 150 and 

300 Kg/m 3 , respectively. Finally, RAPCTB seems to be a viable alternative to 

dense-graded aggregate used in road base construction. 

REFERENCES 

1. Chesner, W. H., R. J. Collins, and M. H. MacKay, 1998, "Users Guidelines for 

Waste and By-Product Materials in Pavement Construction" Report No. 

FHWA-RD-97-148. 

2. Garg, N., and Thompson, M. R., 1996, "Lincoln Avenue reclaimed asphalt 

pavement base project." Transportation Research Record 1547. 

3. Maher, M. H., and Popp, W., Jr., 1997, "Recycled asphalt pavement as a base 

and subbase material." Journal of ASTM STP 1275. 

4. R.Taha, and A. Al-Harthy, 2002 "Cement Stabilization of Reclaimed Asphalt 

Pavement Aggregate for Road Bases and Subbases" Journal of Materials in 

Civil Engineering, Vol. 14. 

5. Yang H. Huang, 2004, Pavement Analysis and Design, Second edition, Pearson 

Prentice Hall, Upper Saddle River, New Jersey 07458. 

333


Compressive Strength (KN/m 2 ) 

% Passing 

100 

90 

80 

70 

60 

50 

40 

30 

20 

10 

Figure1 Particle size distribution for RAPCTB at Variable RAP Contents 

12000 

10000 

8000 

6000 

4000 

2000 

0 

0.01 0.10 1.00 10.00 100.00 

Sieve Size (mm) Log-Scale 

0 

0.0% RAP 

30% RAP 

50% RAP 

70% RAP 

100% RAP 

Specification UL 

Specification LL 

RAP = 100% 

RAP = 70% 

RAP = 50% 

RAP = 30% 

RAP = 0% 

0 10 20 30 40 50 60 

Age (Day) 

Figure 2 Compressive Strength versus Time (Cement Content = 150 Kg/m 3 ) 



CBR (%) 

Compressive Strength (KN/m 2 ) 



Figure 3 Compressive Strength versus Time (Cement Content = 300 Kg/m 3 ) 

250 

200 

150 

100 

50 

0 

20000 

16000 

12000 

8000 

4000 

0 

RAP = 100% 

RAP = 70% 

RAP = 50% 

RAP = 30% 

RAP = 0% 

0 10 20 30 40 50 60 

Age (Day) 

RAP = 100% 

RAP = 70% 

RAP = 50% 

RAP = 30% 

RAP = 0% 

0 5 10 15 20 25 30 

Age (Hour) 

Figure 4 CBR versus Time (Cement Content = 150 Kg/m 3 ) 

335


CBR (%) 

250 

200 

150 

100 

50 

0 

RAP = 100% 

RAP = 70% 

RAP = 50% 

RAP = 30% 

RAP = 0% 

0 5 10 15 20 25 30 

Age (Hour) 

Figure 5 CBR versus Time (Cement Content = 300 Kg/m 3 ) 

Wheel Load = 40 KN Wheel Load = 40 KN 

r = 15 cm r = 15 cm 

5 cm, HMA Top Mr = 3.0x10 5 cm, HMA Top 

7 cm, HMA Base 6 

Mr = 2.5x10 Kpa ν = 0.35 

6 Kpa ν = 0.35 Mr = 3.0x10 6 Kpa ν = 0.35 

35 cm, RAPCTB & ν = 0.30 

25 cm, Aggregate Base Course (ABC) 7 days Mr=4.5x10 6 Kpa & Cement =150 Kg/m 3 

Mr = 1.5x10 5 Kpa & ν = 0.35 7 days Mr=6.2x10 6 Kpa & Cement =300 Kg/m3 

Subgrade, Mr = 7.0x10 4 Kpa ν = 0.45 

a. Control Pavement Section b. Pavement Section with 

RAPCTB 

Figure 6 Layouts with Parameters for Flexible Pavement Stress and Strain Analysis 



Subgrade, Mr = 7.0x10 4 Kpa ν = 0.45

DYNAMIC MEASUREMENTS FOR UPDATING 

CABLE-STAYED BRIDGE MODEL- CASE STUDY 

Shehab Mourad 1 , Mohamed Fayed 2 , Ayman Khalil 3 , and Hosam Abozeid 4 . 

1 Associate Professor, Department of Civil Engineering, King Saud University, Riyadh, KSA, 

smourad@ksu.edu.sa 

2 Professor, Department of Structural Engineering, Ain Shams University, Cairo, Egypt, 

mnourf@yahoo.com 

3 Associate Professor, Department of Structural Engineering, Ain Shams University, Cairo, Egypt, 

akhalil@darcairo.com 

4- Master Student, Department of Structural Engineering, Ain Shams University, Cairo, Egypt, 

hossam_abozeid@yahoo.com 

ABSTRACT 

Structural Health Monitoring “SHM” and damage detection at the earliest 

possible stage is considered one of the most interesting issues of the civil 

engineering community, especially for those structures with long design life, lifesafety 

implications and high capital expenditures like cable-stayed bridges. The 

Experimental Modal Analysis “EMA” gives the required measurements that used in 

model update and needed for those damage detection techniques that based on 

changes in modal properties. This paper presents the technology of determining the 

structural properties of the cable-stayed bridges using dynamic measurements. A 

proposed computerized analysis tool is introduced that uses the EMA output data to 

extract the modal properties of the measured structure. The Suez-Canal cable-stayed 

bridge is assumed as a case study to perform the model update operation using the 

dynamic measurements. The paper addresses the EMA test setup, the modal 

parameters extracting technique and the model update strategy. 

KEYWORDS 

Experimental Modal Analysis, Mode shapes, Natural Frequencies, 

Model Update. 

INTRODUCTION 

Commonly in design process, a mathematical model is needed for 

structural analysis of the designed structure. The mathematical model is usually 

created with assumed dimensions and material properties that certainly differ 


Dynamic Measurements for Updating Cable-Stayed Bridge Model- Case Study 

from those of the real structure after construction. In order to obtain a realistic 

mathematical model to be used for design, prediction, simulation, diagnosis and 

monitoring, a heavily 

computer-based technology, with emphasis on using computer models is 

needed to predict the performance of the structures in question. Experimental testing 

could play a major and vital role in design process, especially when it is properly 

integrated with analytical processes. Experimentation serves two important 

functions in such design activities. The first is to obtain measured data with which to 

check the accuracy of theoretical predictions and the second is to check their 

completeness. Experience gained in recent years recommends the Experimental 

Modal Analysis “EMA” as the most economic, accurate and effective nondestructive 

tool for inspection and health monitoring of the existing structures. [1] In 

an EMA test, the dynamic response of the structure is measured using special 

sensors “accelerometers” that record the response versus time intervals, which called 

time-domain response. Using a special signals analyzer and Fast Fourier Transform 

[2] “FFT” technique, the time-domain response is transferred to frequency-domain 

response “response versus the frequency” as shown in Figure (1). Some of the 

appeared peaks in the frequency-domain response for the different tested joints may 

represent one of the natural frequencies of the bridge, so they should be checked. 

The assumed peaks may be chosen by guidance of the theoretical modal analysis 

results “finite elements analysis”. The peak is considered one of the natural 

frequencies, when it appears in all frequency-domain responses of the whole 

observed joints, in addition it corresponds to one of the tested structure mode 

shape [3,4] , so the accelerometers stations should be verified accurately [5,6] . Therefore, 

an appropriate method of excitation should be chosen according to the type of the 

tested structure, the frequency resolution and the required mode shapes [7] . 

For bridges, two main major methods of excitation may be used. The first 

is the ambient excitation techniques like traffic, wind and earthquakes. The 

second is the measured input techniques like hammer, shaker and step relaxation 

[8] . The extracted dynamic properties of a structure could be established in terms 

of modal parameters “mode shapes and natural frequencies” that are used to 

update the mathematical model. There are two major techniques for model 

update. The first is the direct matrix methods in which, the individual elements 

in the system matrices are adjusted directly from comparison between test data 

and initial analytical model prediction using especial and advanced softwares. 

The second is the indirect, physical property adjustment methods in which 

changes are made to specific physical or elemental properties in the model 

(using the engineering sense) in a search for an adjustment, which brings 

measured and predicted data closer together [9] . 




SHEHAB MOURAD, MOHAMED, ET AL 

Fig (1): The Time-Domain response and the frequency-domain Response 

THE PROPOSED ANALYSIS TECHNIQUE 

Usually the number of accelerometers is limited with respect to the 

large scale of the bridge like in this case study, so the test may be carried out 

through several subtests. Also the measurement stations may be taken for a 

quarter of the bridge at the cables line on the deck. In addition extra two 

stations should be taken. The first is assumed mirror of the nearest one to mid 

span to distinguish between symmetric and non symmetric modes. The second 

is opposite to the first in the transverse direction to distinguish between 

torsion and bending modes. The test output data gives the acceleration 

response versus time “Time Domain” in the three directions in text file, and 

the graph of acceleration response versus frequency that is known as 

Frequency Response Function “FRF” is gotten using any Fast Fourier 

Transform “FFT” analyzer as shown in figure (2) for subtest 7 in vertical 

direction that includes the response of points 16, 17 and 18 on the Suez-Canal 

cable-stayed bridge that will be explained later. Then the compatible peaks 

that appear in all graphs of the whole stations of the same subtest should be 

selected. The mode shapes could be plotted using the response of the all 

stations for each frequency value. Then they should be compared with those 

extracted from the finite element analysis to find out the natural frequencies 

and the corresponding mode shapes of the bridge. A proposed data analysis 

tool was created using Microsoft Excel to simplify the extraction operation of 

modal data “mode shapes and corresponding natural frequencies”. 

339


Fig (2): FRF plot in Z-direction subtest (7) for frequency range 0 - 0.5 Hz 

THE ANALYSIS TOOL 

Microcal Origin Version 6.0 was used as FFT analyzer, and the FFT analysis 

was applied for each test to transfer data from time-domain form to frequency 

domain form. Part of the output data obtained from FFT analyzer for each frequency 

is the real response component (R) and the phase angle (Ф) for each tested point. 

The data is imported from the FFT analyzer to the Excel File, each subtest in a 

separate sheet as shown in Table (1) for subtest 7 that includes the responses of 

points 16, 17 and 18 on the Suez-Canal cable-stayed bridge that will be explained 

later. The analysis tool performs two main functions, the first is peaks extraction and 

the second is mode shapes plotting [10] . 

Table (1): FFT sample data for subtest (7) 

Point (16) Point (17) 

Freq. 

Hz. 

Point (18) Criteria (2) Criteria (1) 

R16 

m/sec 2 

Ф16 R17 

deg. m/sec 2 

Ф 17 

deg. 

R18 

m/sec 2 

Ф 18 

R16 R17 R18 S2 R16 R17 R18 S1 

deg. 

0.000 4.6E-06 360 1.4E-05 0 1.0E-05 360 0 0 0 0 0 0 0 0 

0.012 2.2E-05 240 4.7E-05 -133 4.4E-05 231 1 1 1 3 1 1 1 3 

0.024 1.5E-05 401 3.2E-05 39 3.6E-05 402 0 0 0 0 0 0 0 0 

0.037 8.0E-06 203 8.1E-06 169 1.1E-05 213 0 0 0 0 0 0 0 0 

0.049 4.2E-06 367 1.3E-05 279 8.2E-06 281 0 0 0 0 0 0 0 0 

0.061 2.1E-06 465 1.5E-05 446 1.4E-05 453 0 1 1 2 0 1 1 2 

0.073 2.2E-06 596 1.1E-05 296 8.8E-06 297 0 0 0 0 0 0 0 0 

0.085 2.2E-06 484 1.3E-05 146 7.5E-06 150 0 0 0 0 0 0 0 0 

0.098 6.1E-06 259 1.5E-05 264 1.4E-05 254 0 0 0 0 0 0 0 0 

0.110 1.1E-05 422 2.2E-05 419 2.4E-05 417 1 1 1 3 1 1 1 3 



1- Peaks Selection Criteria 



Since any peak value is bigger than the value before and the value after, so it 

is needed to compare the response amplitude of each record with the two adjacent 

values (one after and one before) as criteria (1) or with the four adjacent values (two 

after and two before) as criteria (2). 

By the help of Microsoft Excel computational functions, MAX function used 

for this comparison on a separate column for each point, and with IF function, if the 

check is true a flag value (1) appear in this cell, else flag value (0) appear, and by 

using extra one column for each criteria containing the sum of the three flag values, 

then if the sum is equal to the number of points, it means that this record 

corresponds to a natural frequency as shown in Table (1), and by using the data filter 

tool in the sum column equal to the number of points we can get all records of the 

compatible peaks and corresponding natural frequencies, By repeating the same 

operations for all subtests, all peaks of all points were extracted. 

Another check was done to assure that the extracted frequency is considered 

as one of the natural frequencies of the bridge, by comparing the corresponding 

plotted mode shape of the studied direction with the obtained one from the 

mathematical model. If the plotted mode has a logical shape with respect to the 

studied direction and the mathematical model analysis results then the corresponding 

frequency is considered as one the bridge natural frequencies. 

2- Plotting of Mode Shapes 

Plotting of any mode shape corresponding to a certain frequency is very 

important to check the reality of this frequency as one of the natural frequencies of 

the bridge, so it is needed to find out an automated, efficient and fast technique to be 

used in plotting of the different mode shapes corresponding to the extracted large 

number of peaks frequencies. A Microsoft Excel macro is used to do the following 

operations: 

1. Get the response amplitude and the phase angle of all points in the all subtests 

to a table form for the selected frequency value. 

2. Normalize or rescale the response of the all joints with respect to the subtest (1) 

using the relative response values of the overlap joints. 

3. Calculate the relative sign of the joint response for all joints with respect to the 

subtest (1) using the relative sin value of the phase angles of the overlap joints. 

4. Check the mode symmetry by comparing the response sign of joint 16 and joint 17. 

5. Check the mode type (bending or torsion) by comparing the response sign of 

joint 17 and joint 18. 

6. Draw the normalized mode shape ordinates of the bridge along its longitudinal 

axis using the symmetry check in step 4. 

341


All the above operations is being done automatically by just selecting a 

frequency value from those extracted in the previous step in the calculation sheet, 

three files containing the same technique are created for the three response 

directions X, Y and Z. 

MATHEMATICAL MODEL UPDATE STRATEGY 

The mathematical model update operation follows one of two major 

methods. The first is the direct matrix method, which needs especial and advanced 

software's that are not available in the local market. The second is “the indirect 

physical property adjustment method” [9] , in which some changes are made to 

specific elemental properties in the finite element model (using the engineering 

sense) searching for an adjustment, which brings Finite Element Analysis “FEA” 

and Experimental Modal Analysis “EMA” modal data closer together. The second 

method was used for the bridge FEA model update as follow: 

ω 1 K 

f = = 

(1) 

2π 

2π 

m 

Where: f is the natural frequency, ω is the angular frequency, K is the stiffness and 

m is the mass. 

And: 

EI 

k = λ 

L 

(2) 

w 

m = 

g 

(3) 

Where: λ is a factor depends on boundary conditions of the member, E is the 

modulus of elasticity, I is the inertia of the member and L is the length of the 

member. 

Therefore, to get a higher value of frequency, it should to either increase the 

stiffness or decrease the mass. Any changes in masses affect the member’s own 

weights and any load changes makes the model lose its reality especially for static 

analysis. For the steel main girder and cables own weights, they were perfectly 

calculated according to the physical statistical data of the main bridge and, but for 

concrete elements own weights, the concrete specific unit weight may be changed 

slightly within normal figures because it was cast in site. It is preferable mainly to 

update the stiffness to compensate any neglected factors or elements of the boundary 

conditions of the different members like variations in materials properties, variations 

in sections dimensions, variations in supports and fixation conditions and local 

stiffeners of the main girder. To change the stiffness according to equation (2), it is 

no way to change the member length and also it is to hard to find out the required 

parameters that governs the changes of λ coefficient and member inertia due to the 





large number of members, so the only choice is to use the modulus of elasticity to 

change the stiffness. The engineering sense governs the selection of the updated 

elements to change the natural frequency of a certain mode due to of being those 

elements the main resistance elements to the motion described by this mode. For the 

longitudinal modes “sway modes”, they are mainly affected by the stiffness of 

towers and girder bearing elements. The transverse bending modes are affected 

mainly by main girder stiffness (especially the upper and lower decks). The vertical 

bending modes are affected mainly by the stiffness of both the main girder and the 

cables. So the model update strategy depends on changing the modulus of elasticity 

value of the different materials that assigned to the different bridge elements to 

change their stiffnesses by trials and errors to adjust the model to give the required 

natural frequencies values similar to those of the EMA selected mode shapes. 

CASE STUDY 

The Suez-Canal cable-stayed bridge is assumed as a case study to implement 

the proposed technique. An EMA test had been done for the bridge and the data was 

analyzed and the analysis results used to update the finite element model of the 

bridge as follow. 

The Suez-Canal Cable Stayed Bridge 

The Suez-Canal cable stayed bridge links Africa and Eurasia, crossing the 

Suez-Canal in Qantara city, which is located about 50-km south of the 

Mediterranean Sea. The bridge construction was totally completed in autumn of 

2001. The bridge has total length of about 3900 meters with two lanes for each 

direction and with maximum vertical grade is 3.3% for smooth traffic flow. The 

main bridge is a steel cable-stayed bridge with girder length of 730 meters and 

central span of 404 meters “clear span is 384 meters” and two side spans of 163 

meters with a vertical clearance of 70 meters above high water level to assure free 

navigation on the canal. The concrete pylons are H-shaped R.C with height of about 

160 meters. The tower cross-section is a variable box section of dimensions 7.6 m x 

7.8 m x 0.7 m at base level that reduce gradually with average slope 1:35 till reach 

2.5 m x 4.5 m x 0.5 m at top level. The stay cables are 128 cables of 16 steps in 

double plane for each side, and for each plan the cables were arranged using four 

types in cross-section, where the first two cables at tower side are of 27.51 cm 2 , the 

next three cables are of 42.51 cm 2 , the next two cables are of 54.79 cm 2 and the last 

nine cables are of 67.06 cm 2 as shown in Figure (3). The main girder is a single-cell 

steel box girder of 20.8m wide “4 traffic lanes, 0.8m sidewalks and 1.2m median 

strip” and variable depth of 1.2m at edges and 2.6m at the middle, the upper and 

lower decks are made of orthotropic plates with closed ribs, the outer sides and the 

two longitudinal stiffeners are made of solid plates of thickness 16mm and 11mm 

respectively, solid steel cross diaphragms of 10mm thickness are added at cables 

locations and in mid-distance between cables to strengthen the section as shown in 

343


Figure (4). The reinforced concrete used for pylons and piers has a cubic 

compressive strength of 50 MPa, the steel reinforcement is of grade 36/52 with 

Young’s modulus of elasticity 210 GPa, the stay cables ultimate strength is 1800 

MPa, with Young’s modulus of elasticity 200 GPa. 

Fig (3): Schematic elevation and side view of The Suez-Canal Bridge 

Fig (4): Main Girder Cross Section 



Mathematical Model of the Bridge 



A finite elements model was created using “SAP2000” by the help of the 

original bridge documentations [11,12] . The model was verified also for both static and 

dynamic responses using “Staad Pro 2003”. Pylons, piers, cables and bearings 

members were modeled using frame elements while the main girder was modeled 

using the shells elements with equivalent thicknesses that gives the same properties 

of the original section. All Pylons and piers supports were assumed totally fixed. 

Main girder bearings at piers were released for displacement and rotation in the 

longitudinal direction in addition to rotation about their axes (torsion). Main girder 

vertical bearings at towers were released for displacement and rotation in transverse 

direction in addition to torsion, but were partially restrained for displacement in the 

longitudinal direction with a spring stiffness of 10000 kN/m). Main girder horizontal 

bearings at towers were restrained only for axial displacement (transverse direction). 

The compression limit of cables was assumed zero and their post tension forces 

were modeled using p-delta initial force program option. Masses were modeled by 

the program as lumped masses at joints using the different materials mass density. 

The total mass of the model was checked by that reported in the original bridge 

documentation. 

Modal Analysis Results 

Geometric non-linear analysis with ten iteration steps assuming large 

deformation into consideration were performed using SAP2000 Model of the bridge and 

the resultant stiffness matrix was used for modal analysis to extract the first thirty mode 

shapes and their corresponding natural frequencies. Figure (5) shows the first modal 

mode shape in each global direction X, Y and Z respectively of the bridge. 

1 St longitudinal mode 1 St transverse mode 1 St vertical mode 

F=0.160 Hz F=0.216 Hz F=0.361 Hz 

Fig (5): The first modal mode shape in the three directions 

345


The Bridge EMA Test 

An Experimental Modal Analysis “EMA” test was carried out on the 

bridge using ambient excitation by a truck. The test consists of seven subtests, 

for each, three or four accelerometers were used to record the bridge response. 

Only response of half of the bridge was recorded due to limited number of 

accelerations. Figure (6) shows the subtest layout and locations of the used 

accelerometers, where 1, 2, 3 and 4 are the accelerometer reference number. 

All accelerometers were located at cables connections with the deck in the 

southern side of the bridge “right side” except the accelerometer no 4 of 

subtest 7 was located at the northern side “left side” of the bridge to 

distinguish between torsion and bending mode shapes. The accelerometer no 3 

of subtest 7 that located at the west side indicates the mode shape symmetry. 

The subtests were carried out by a sampling rate ∆t =0.02 sec for the half of 

the bridge only due to the lack of accelerometers depending on the symmetry 

of the bridge. 

Test Records Analysis 

The EMA test measurements were recorded in seven text files for the 

seven subtests each contains the acceleration response versus time intervals in 

the three directions. The locations of the accelerometers were renumbered and 

rearranged as shown in Figure (7), in order to simplify the coordination 

between the accelerometers locations and their response, hence to draw the 

mode shapes easily, taking into consideration that the bold circled points are 

representing the subtest overlap locations. 

A FFT analysis were performed using the Microcal Origin version 6.0 

program using Hanning windowing technique to transfer data from timedomain 

form to the frequency-domain form. The FFT output records of the 

different subtests have 0.01 Hz frequency resolution (the minimum accurately 

resolved frequency shift) and 25 Hz folding frequency range (Nyquist 

Frequency- the maximum accurately resolved frequency), which is enough to 

get the master mode shapes within the available excitation conditions. 





Fig (6): Accelerometers locations for the different subtests 

Fig (7): Accelerometers Locations reference numbers for the different subtests 

347


The FFT analyzer gives the opportunity to plot the FFR (Frequency 

Response Function) of each joint to select some of the appeared peaks to be 

checked. To minimize the peaks selection, it is better to plot the FFR of the whole 

joints of each subtest in one graph to select the compatible peaks that appear in the 

all FFR S as shown in Figure (2). It should be noted that each graph represents the 

response of only one point during the excitation, and some the resultant peaks may 

appear due to resonance representing natural frequencies of the bridge, and some 

others may appear due to noise or error in sampling at the observed point, which is 

not needed, so to find out a peak corresponding to a natural frequency, it should be 

appeared in all graphs of all points of the same subtest regardless the response value. 

Normally, the FRF peaks is extracted using visual judgment for the function graph, 

which is impractical due to the large numbers of recorded peaks and the required 

complicated calculations and procedures to draw the corresponding mode shape, so 

it had to use the proposed automated analysis tool that can perform those procedures 

automatically to give the opportunity to check the large number of peaks accurately 

and quickly. 

Data Analysis Results 

For such structures like cable-stayed bridges it is enough to study the FRF 

up to 1 Hz for global mode shapes, while the higher frequency values represent 

some members’ local mode shapes, so the studied peaks were selected within this 

range. The extracted theoretical modes from Finite Element Analysis (FEA) guided 

the selected peaks. From the bridge EMA test data analysis, six mode shapes and 

their corresponding natural frequencies were extracted using the proposed analysis 

tool as shown in Table (2) and in Figures (8), (9) and (10). 

Table (2): FEA and EMA natural frequencies and mode shapes 

Mode 

No. 

Mode 

Direction 

Mode Description 

FEA Freq. 

(Hz.) 

EMA Freq. 

(Hz) 

1 Longitudinal Symmetric Pylon Bending “Sway” 0.160 0.256 

2 Transverse 

Symmetric Girder/Pylon Bending 

“1 st Transverse Bending of Girder” 

0.216 0.269 

3 Transverse 

Anti-symmetric Pylon Bending 

“2 nd Transverse Bending of Girder” 

0.254 0.281 

N.A. Transverse “3 rd Transverse Bending of Girder” - 0.293 

5 Vertical 

Symmetric Girder Bending 

“1 st Vertical Bending of Girder” 

0.361 0.403 

22 Vertical 

Symmetric Girder/Pylon Bending 

“3 rd Vertical Bending of Girder” 

0.570 0.830 





Fig (8): The 1 st Longitudinal “Sway” mode and the corresponding vertical mode shape 

FEA Results with Respect to EMA Results 

By comparing the extracted modal parameters using the FEA model with 

those extracted from the EMA test results, it was found some differences in the 

natural frequencies values, where the mathematical values are less than the 

experimental values that means that the assumed bridge stiffness in the FEA model 

is less than it should be. Table (2) shows the FEA modal frequencies values versus 

those of the EMA. Therefore, the FEA model should be updated to match the EMA 

model as possible as it could be and gives the same modal parameters of the EMA 

test. This updating operation will introduce the FEA model as a good representative 

mathematical model for the bridge to be used for further studies. The low frequency 

modes (early modes) are always representing the global behavior of the structure, 

while the high frequency modes are representing the local behavior of the different 

elements. The first mode in each direction is enough to represent the motion in this 

direction, so only the first three EMA modes that were used in FEA model update. 

The selected modes are the shaded modes in Table (2). 

349


Fig (9): The 1 st , 2 nd and 3 rd transverse bending modes respectively 





Fig (10): The 1 st and 3 rd vertical bending modes respectively 

Update Operation Details 

The FEA model update operation passed over than fifty trials of model 

modifications until adjustment was achieved by applying the following: 

- The bridge bearing elements on piers and towers were partially restrained 

as shown in Table (3) to increase the stiffness of the bridge against the 

sway mode. 

- The modulus of elasticity of some concrete elements (towers and piers) and 

steel elements (main girder elements and cables) were increased as shown 

in Table (4) to increase the stiffness of towers and bridge consequently 

against both longitudinal and transverse directions. 

- In order to account for effect of non-structural elements, composite actions, 

different connection stiffnesses, and other effects that are hardly to be 

evaluated, which in turn increase the bridge stiffness, an equivalent reduced 

unite weight of some concrete elements can be adopted and selected as 

given in Table 4, which in turns increase the bridge natural frequencies. 

351


Table (3): Bearing partial restraining values in the longitudinal direction 

Bearing 

Element 

Elements 

Number 

Original V2 Value 

(kN/m.) 

Modified V2 Value 

(kN/m.) 

Towers 

Bearings 

4 10000 15000 

Piers Bearings 12 0 3000 

Material 

Name 

Table (4) Original and modified properties of the used materials 

Material 

Description 

Unit Mass 

Ton-s 2 /m 4 

Unit Weight 

Ton/m 3 



E 

Ton/m 2 

Original Mod. Original Mod. Original Mod. 

CONC Towers Concrete 0.25 0.20 2.50 2.00 2100000 3350000 

Conc_P Piers Concrete 0.25 0.22 2.50 2.20 2100000 2100000 

STEEL Diaphragms Steel 1.10 1.10 10.76 10.76 21000000 27500000 

Steel_C Cables Steel 0.80 0.80 7.85 7.85 21000000 25000000 

STEEL_LD Lower Deck Steel 1.10 1.10 10.76 10.76 21000000 27500000 

STEEL_UD Upper Deck Steel 1.16 1.16 11.43 11.43 21000000 27500000 

STEEL_WB Webs Steel 1.10 1.10 10.76 10.76 21000000 27500000 

Z_Conc Zero Wt. Concrete 0.00 0.00 0.00 0.00 2100000 2100000 

Z_Steel Zero Wt. Steel 0.00 0.00 0.00 0.00 21000000 21000000 

CONCLUSION 

The EMA could be considered the best available non-destructive testing tool 

for cable-stayed bridges, by which the bridge modal properties could be extracted 

that be used powerfully in the bridge model update. Then the updated model is 

simulating the real bridge and could be used in monitoring its structural health by 

repeating the EMA test and using the appropriate structure health monitoring 

techniques. A proposed computerized analysis tool was introduced that could be 

used powerfully in analysis of any EMA test results especially incase of using a 

limited number of accelerometers. For the case-study of the Suez-Canal cable-stayed 

bridge, the experimentally extracted mode shapes and corresponding natural 

frequencies were very near to those extracted theoretically using FE analysis. The 

extracted modal properties were used to update the FE model of the bridge using the 

indirect physical property adjustment method that used later to study the appropriate 

structure health monitoring techniques for cable-stayed bridges.



REFERENCES 

1- Ewins, D., 1984, Modal Testing: Theory and Practice, John Wiley and 

Sons, Inc., New York, 

2- Morrison, N., 1994, Introduction to Fourier analysis, John Wiley and Sons, 

Inc., New York. 

3- Bendat, J.S.; Piersol, A.G., 1971, Random Data: Analysis and 

Measurement Procedures, John Wiley and Sons, Inc., New York. 

4- Ren, W., X. and Z. H. Zong, 2004,“Output-Only Modal Parameter 

Identification of Civil Engineering Structures”, Structural Engineering and 

Mechanics, Vol. 17, No. 3-4. 

5- Ewins, D., (2000), Basics and state-of-the-art of modal testing, Sadhana ، 

Vol. 25, Part 3, pp. 207-220, India, June 2000. 

6- DTA 1996 Hand Book of Best Practice, Vol. 3, Model testing 

7- Khalil, A., (1998), Aspects in Nondestructive Evaluation of Steel Plate 

Girder Bridges, PH. D. Dissertation, Iowa State University, Iowa, USA, 

1998. 

8- Charles R. Farrar, Thomas A. Duffey, Phillip J. Cornwell, Scott W. 

Doebling, (1999), “Excitation Methods for Bridge Structures”, in 

Proceedings of the 17 th International Modal Analysis Conference 

Kissimmee, FL, Feb 1999. 

9- Ewins, D., (2000), Adjustment or Updating of Models , Sadhana ،Vol. 25, 

Part 3, pp. 235-245, India, June 2000. 

10- Abozeid, H., (2005), Dynamic Response of Cable-Stayed Bridges, M. 

Sc.Dissertation, Ain Shams University, Cairo, Egypt, 2005. 

11- Pacific Consultants International, Chodai Co., LTD., (1997), “Final Design 

Drawings of The Suez-Canal Bridge, The Shop Drawings of The Main 

Bridge”, August 1997. 

12- Kajema-NKK-Nippon Steel Consortium (2001), “Suez-Canal Bridge – 

Construction brochure”, October 2001. 

353



Many thanks for all team members for test data supply .The EMA test team 

supervisors consisted of Dr. Ayman Hussein for the test plan, Dr. Adel El-Attar and 

Dr. Mashhour Ghuniem for providing and execution of the test equipments. 



ANALYSIS OF RECTANGULAR HOLLOW SECTIONS UNDER 

UNAXIAL COMPRESSIVE LOADS 

Dr.Prof. Sedky Abd Allah Tohmay 

Prof. , Department of Civil Engineering ,EL-Minia University –Egypt. 

E-mail: sedky_T2000@yahoo.com 

ABSTRACT 

This paper presents a theoretical analysis for rectangular hollow sections 

in order to find the buckling coefficient of the section under compressive loads . 

The critical buckling load for rectangular hollow sections is affected by the 

dimensions of the wide and narrow faces of the sections. The effect of interaction 

between adjacent elements has been investigated for rectangular hollow section 

(RHS). The interaction was accounted for by determining the plate buckling 

coefficients for the wide and narrow faces from a rational buckling analysis . A 

simple and convenient method is proposed by using the energy method in order to 

formulate the necessary equations from which results can be obtained . 

Buckling loads for rectangular hollow sections are given in explicit 

expressions which provide simple and accurate design procedures . A group of 

design charts is also presented form which the buckling stresses can be directly 

calculated for a wide range of rectangular hollow section. A comparison between the 

buckling coefficient as obtained by the energy method ( using one term and two 

terms analysis in the deflection function ) and the finite element method through 

ANSYS program is also performed . 

INTRODUCTION 

Rectangular hollow sections ( RHS ) Fig.( l ) are used in wide range of 

applications in many fields of structural engineering . For example, in space truss 

systems , roof sheeting , bridges , beams and columns , wall cladding and tanks . 

Box girders consist of steel plates for the wide and narrow faces. The critical 

buckling siress of steel plate is given by : 

2 

2 

E 0 π ⎛ t ⎞ 

σ cr = K 

2 ⎜ ⎟ 

…………. ( 1 ) 

12 

( 1 − ν 


) 

⎝ 

b 

⎠

ANALYSIS OF RECTANGULAR HOLLOW SECTIONS UNDER UNAXIAL COMPRESSIVE LOADS 

In equation (1) E0 is the initial elastic modulus, ν is Poisson’s ratio, t 

and b are the plate thickness and width respectively and K is the plate buckling 

coefficient . The strength of plate supported along both longitudinal edges (stiffened 

plate) can be determined with reasonable accuracy by substituting elastic modulus 

E0 by E 0E1 

〈 E 0 in the plate buckling stress Bleich [1] . The plate 

buckling coefficient K is to be obtained from a rational buckling analysis . This 

method requires the cross section to be treated as a plate assembly taking the 

interaction between the component plates into account .The critical buckling load for 

a rectangular hollow section is affected by the dimensions of the plate of the wide 

and narrow faces of the rectangular hollow section . Abdel-Lateef [2] carried out a 

theoretical and experimental investigations on the analysis of stiffened panels 

subjected to combined loading of shear and compression . Abd el-Sayed [3] studied 

the basic equations for the stresses analysis of orthotropic plates by using the theory 

of elasticity approach , El-Aghoury [4] carried out a parametric study using a finite 

element approach on the elastic buckling of stiffened rectangular plates in order to 

determine the buckling loads for a variety of stiffeners arrangements. Harris [5,6] 

investigated the stiffness after buckling of orthotropic plates under biaxial loading . 

Numerous researchers [7,8,9,10,11] focused on determining the compressive 

buckling stresses for rectangular plates. Sedky A.T [12] investigated the behavior of 

perforated isotropic rectangular plates under shearing and compressive loads. 

The theoretical work described in this paper is devoted to the calculation of 

the coefficient of buckling stresses for wide and narrow faces of rectangular hollow 

sections by using the minimum potential energy method and to study the effect of 

the dimensions of the plate wide face (a, d) and narrow face (a, b) , and to 

determine the relation between the coefficients of buckling of wide and narrow 

faces. Plate buckling problems may be introduced in many techniques: conservation 

of energy using Fourier Series, Rayleigh method, virtual displacement method and 

minimum potential energy procedure. In this paper the minimum potential energy 

was used to find the buckling loads for of rectangular hollow section ( RHS ). 

THEORETICAL ANALYSIS 

Consider the rectangular hollow section shown in fig (1); the corners are 

ignored in the present analysis . Symbols d and b are used the denote the outside 

dimensions of the wide and narrow faces of the rectangular hollow section . A simple 

energy analysis may be obtained by assuming the plate buckling deflection surfaces 

of the wide and narrow faces are both sinusoidal 




Prof. Sedky Abd Allah Tohmay 

Fig.(1) Rectangular hollow section under compressive load 

m π 

n π 

∑ ∑ A sin x sin y 

a 

b 

∞ ∞ 

= ….(2) 

w mn 

m = 1 n = 1 

Wide Face 

(axd) 

d 

Compressive Load 

Narrow Face 

(axb) 

b 

To simplify the mathematical calculation , it was found that using one 

term of the double Fourier series may give accurate results . This function obeys the 

boundary conditions of the plate since each term of Eq.( 2 ) clearly satisfies the 

boundary conditions of zero bending moment along the four edges. 

a 

Rectangular Hollow Section 

a 

b 

Narrow face under compressive load 

Wide face under compressive load 

a 

d 

t 

d 

Cross Section 

b 

357


So “ w b ” the plate buckling deflection of the narrow face b can be considered as : 

w b 

A 1 π π 

= sin x sin y . 

(3) 

ϕ a b 

and “ w d ” the plate buckling deflection of the wide face d can be obtained as : 

w d 

π π 

= A1 

sin x sin y . 

( 4 ) 

a d 

The strain energy due to bending and twisting of the narraw face plate can be found as :- 

⎡ 

a b 

D w b w b 

w b w b 

w 

U b ∫∫ v 

x y 

x y 

x y ⎥ ⎥ 

/ 2 

2 

2 

2 2 

2 

2 

⎛ ∂ ⎞ ⎛ ∂ ⎞ ∂ ∂ 

⎛ ∂ ⎞ 

= ⎢⎜ 

⎟ + ⎜ 

⎟ + 2 

− 2 

2 

2 

2 2 

2 ⎢ 

∂ 

∂ ∂ 

⎜ ∂ ∂ ⎟ 

0 0 ⎝ ∂ ⎠ 

U b 

⎣ 

⎝ 

Substituting Eq. (3) in Eqs. (5) we get :- 

= 

D A 

2 

2 

1 

4 ⎛ α sin 

⎜ 

⎜ 

⎜ 2 

⎜ 2να 

π 

⎝ 

Then , U b can be found as : 

⎠ 

2 

2 4 

2 

( α x ) sin ( β y ) α sin ( α x ) sin ( β y ) 

2 

η 

sin 

2 



+ 

b ( 1 − ν ) ⎜ ⎟ d x d y 

⎝ 

⎠ 

⎤ 

⎦ 

…..(5) 

( ) ( ) ( ) ( ) ( ) 

⎟ ⎟⎟⎟ 

2 

2 

2 2 

2 

2 

α x sin β y 2 1 − ν α π cos α x cos β y 

b 

2 

− 

2 

2 2 

( a b ) 

2 4 

D A π + 

U b 2 3 3 

16 η a b 

= (6) 

The strain energy due to bending and twisting of the narrow face plate can be found as :- 

ad 

D wd 

wd 

wd 

wd 

wd 

U d ∫∫ v 

( ) d xd 

y 

x y x y 

x y ⎥ ⎥ 

/ 2 ⎡ 2 

2 

2 2 

2 

2 

⎛ ∂ ⎞ ⎛ ∂ ⎞ ∂ ∂ 

⎛ ∂ ⎞ ⎤ 

= ⎢⎜ 

⎟ + ⎜ 

⎟ + 2 

− 2 1−ν 

⎜ 

⎟ 

2 

2 

2 2 

2 ⎢ 

∂ ∂ 

⎣ 

⎝ ∂ ⎠ ⎝ ∂ ⎠ 

⎝ ∂ ∂ 

0 0 

⎠ ⎦ 

(7) 

b 

2 

2 

b 

+ 

2 

⎞ 

⎠

Substituting Eq. (4) in Eqs. (7) we get :- 

U d 

= 

D 

A 

3 

16 a d 

2 

2 2 

( a + d ) 


2 

π 

4 

3 


Then, the total strain energy for the rectangular hollow section ( Ut )can be found as : 

U 

t 

D 

= 

2 

D 

2 

a d 

∫∫ 

0 

U t 

0 

⎡ 

2 

2 

2 

⎛ ∂w 

b ⎞ ⎛ ∂w 

b ⎞ ∂ w b 

⎢⎜ 

⎟ + 

v 

x 

⎜ 

y 

⎟ + 2 

2 

2 

2 

⎢ 

x 

⎣ 

⎝ ∂ ⎠ ⎝ ∂ ⎠ ∂ 

2 

∂ w b 

2 

∂y 

− 2 

⎡ 

2 

2 

2 

⎛ ∂w 

d ⎞ ⎛ ∂w 

d ⎞ ∂ w d 

⎢⎜ 

⎟ + + v 

⎢ x 

⎜ 

y 

⎟ 2 

2 

2 

2 

∂x 

⎣ 

⎝ ∂ ⎠ ⎝ ∂ ⎠ 

2 

∂ w d 

2 

∂y 

− 2 

a b / 2 

∫∫ 

0 

/ 2 

0 

= 

D 

A 

4 

π 

η 

2 

( 1 − ν ) 

2 ⎛ ∂ w ⎞ b 

⎜ 

x y ⎟ 

⎝ ∂ ∂ ⎠ 

2 ⎛ ∂ w 

2 

⎞ ⎤ 

⎜ x y ⎟ 

⎝ ∂ ∂ ⎠ ⎥ 

⎦ 

(9) 

⎤ 

⎥d 

⎥ 

⎦ 

359 

(8) 

d ( 1 − ν ) ⎜ ⎟ ⎥d 

x d y 

2 2 2 

2 4 2 2 

( a + b ) D A π ( a + d ) 

2 3 3 

3 3 

16 a b 16 a d 

or U t can be written as : 

+ 

2 2 

2 2 2 

( φ + 1) 

η + ( φ + η ) ) 

2 4 

A π D b 

U = 

(11) 

t 3 3 

16 η φ 

1-Work done of the rectangular hollow section : 

The compressive force acts through the centroid of cross section in which the 

cross section is fully effective . The work done by external loading system in wide 

and narrow faces can be determined by using the form : 

a b 

a d 

t w b 

t w d 

T t ∫∫ d x d y 

d x d y 

x 

∫∫ x ⎟ 2 

2 

σ ⎛ ∂ ⎞ 

σ ⎛ ∂ ⎞ 

= ⎜ 

⎟ + ⎜ 

(12) 

2 

2 

2 ⎝ ∂ ⎠ 

2 ⎝ ∂ 

0 0 

0 0 ⎠ 

Substituting Eqs. (3 & 4) in Eqs. (12) the work done in narraw and wide faces 

and thus the total work done can be obtained :- 

2 2 

1 A π 

T b = t σ 

(13) 

2 

16 η φ 

T d 

2 

2 

x 

d 

y 

(10) 

2 2 

1 A π η 

= t σ 

(14) 

16 φ 

+


T t 

2 2 

2 2 

1 A π 

1 A π η 

= t σ + 

t σ 

(15) 

2 

16 η φ 16 φ 

3 ( 1 + η ) 

2 2 

1 A π 

T t = t σ 

(16) 

2 

16 η φ 

The concept of minimum potential energy can be used in buckling analysis .This 

technique depends mainly on assuming a deflection shape for the structure under 

investigation which necessarily satisfies the boundary conditions , then the 

equilibrium equation of the structure can be also satisfied by using the technique that 

the total potential energy (V ) is the sum of the strain energy ( U ) and the work 

done ( T ) by the external loading system . The potential energy (V) of the hollow 

rectangular section can be expressed by : 

V = U − T 

t 

t 

t 

By substituting Eqs.(11 and 16 ) into equation (17) and carrying out the integrations 

and reducing the results the hollow rectangular section buckling equations can be 

obtained as : 

2 4 

A π D b 

V = 

3 3 

16 η φ 

2 [ ( φ + 

2 

1 ) η 

2 

+ ( φ + 

2 2 

η ) ] − 

(18) 

1 

16 

2 2 

A π 

2 

η φ 

t σ ( 1 + 

3 

η ) 

The potential energy (V) is a minimum for the stable structure in which the 

deformed shape is compatible with the support conditions so that: 

∂V 

= 0. 

0 

∂A 

(19) 

The critical buckling stress of the hollow section can be found as: 

σ cr H 

2 

D π 

2 

= 

( ) ( φ 

2 2 

3 

t φ η b 1 + η 

2 

+ 1) 

η + 

2 ( φ 

2 2 

+ η ) ) 

(20) 

σ cr 

From 

= k 

2 

Eπ 

⎛ t ⎞ 

2 ⎜ ⎟ 

12( 

1−ν 

) ⎝ b ⎠ 

3 

Et 

D = 

2 

12( 

1−ν 

) 

and 

The general equation of critical buckling can be written as: 

σ 

crH 

= 

2 2 

2 2 

( φ + 1) 

η + ( φ + η ) 

2 

3 

φ η ( 1 + η ) 



2 

2 

E π 

12 ( 1 − v 

2 

2 

) 

⎛ 

⎜ 

⎝ 

t 

b 

⎞ 

⎟ 

⎠ 

2 

(17) 

(21)

Or can be written in the form : 

σ 

crH 

= 

K H 

12 ( 1 

2 

E π 

− v 

2 

) 


⎛ 

⎜ 

⎝ 

t 

b 

⎞ 

⎟ 

⎠ 

2 


(22) 

Where KH is the buckling coefficient of the rectangular hollow section 

and can be obtained as 

H 

= 

2 2 

2 2 

( φ + 1) 

η + ( φ + η ) 

2 

3 

φ η ( 1 + η ) 

2 

K (23) 

Fig (2) shows the buckling coefficient of the rectangular hollow section for 

different aspect ratios φ = a/b from 1 to 5 and η = d / b from 1 to 1.5 . 

2-The buckling coefficient for wide and narrow faces 

The potential energy (V) of narrow and wide plate faces can be expressed by 

:- 

Vb = U b − Tb 

(24) 

and 

Vd = U d − Td 

(25) 

The coefficients of buckling kb and kd for wide and narrow faces can be 

determined by substituting Eqs.(6,8,13 and 14 ) into equation ( 24 , 25 ) and from 

the minimum potential energy . 

∂Vb 

∂V 

d 

= 0. 

0 

and = 0. 

0 

∂A 

∂A 

the critical buckling stresses can be obtained in wide and narrow faces . 

2 

Eπ 

⎛ t ⎞ 

σ d = kd 2 ⎜ ⎟ 

and 

12( 

1−ν 

) ⎝ d ⎠ 

2 

2 

2 

Eπ 

⎛ t ⎞ 

σ b = kb 2 ⎜ ⎟ 

12( 

1−ν 

) ⎝ b ⎠ 

The buckling coefficients kb and kd for wide and narrow faces can be found as : 

d 

η + 

= 2 

2 

η ( 1 − η + η ) 

η 

2 ( 1 − η + η ) 

2 

+ η ( 1 − η + η ) 

2 

η ( 1 − η + η ) 

k (26) 

and 

η 

= 2 

k (27) 

b 

361


From the equations (26 )and( 27) the buckling coefficients of rectangular 

hallow section can be determined . Figures ( 3) and (4) shown the buckling 

coefficient Kb ,Kd for wide and narrow faces of RHS for different aspect ratios η = 

d/b form 1 to 3 . The plate buckling coefficients are minimum values corresponding 

1 

2 

= η −η 

3 

η where φ = a / b 

φ + 

to : ( ) 4 

Buckling coefficient K H 

16 

14 

12 

10 

8 

6 

4 

2 

η=1 

η=1.1 

η=1.2 

η=1.3 

η=1.4 

η=1.5 

1 2 3 4 5 

φ= a/b 

( 2 ) Buckling Coefficient KH for rectangular 

hollow section under compressive load 

32 

28 

24 

20 

16 

12 

8 

4 

Aspect ratio η= d/b 

η= 1 η=1.2 η=1.4 η=1.6 

η=1.8 η=2 η=2.5 η=3 

0 

1.0 1.2 1.4 1.6 1.8 2.0 

Fig. (4 ) Buckling coefficient for wide face Kd 

for different ratios of η = d/b . 

φ 



k b 

4.4 

4.0 

3.6 

3.2 

2.8 

2.4 

2.0 

1.6 

Aspect ratio η=d/b 

η= 1 η=1.2 η=1.4 η=1.6 

η=1.8 η=2 η=2.5 η=3 

1.2 

1.0 1.2 1.4 1.6 1.8 2.0 

( 3 ) Buckling coefficients for narrow face Kb 

for different ratios of η = d/b . 

coefficient of buckling k 

6 

5 

4 

3 

2 

1 

kd 

Coeff. of buckling in wide face 

kb 

Coeff. of buckling in narrow face 

0 

1.0 1.5 2.0 2.5 

η = d/b 

3.0 3.5 4.0 

Fig. ( 5 ) Comparison between the buckling 

oefficients of wide face Kd and narrow face Kb 

for different aspect ratio η = d/b form 1 to 4 . 

φ

3-Two term analyses of the double Fourier series 



To improve the accuracy of the plate buckling coefficients for large aspect 

ratios ,say η ≥ 1. 

5 the displacement field for wd is augmented by a function that 

allows reverse curvature of the wide face 

π x ⎛ π y ⎛ 2 π ⎞ ⎞ 

w d = sin ⎜ A1 

sin + A 2 ⎜ 1 − cos y . ⎟ ⎟ 

a ⎝ d ⎝ d ⎠ ⎠ 

(28) 

And “ w b ” the plate buckling deflection of the narrow face b can be considered 

A1 

π π 

as: w b = sin x sin y . 

(29) 

ϕ a b 

The augmented displacement field maintains compatibility of rotations ( ∂ w) 

∂y 

at the corners and allows a restraining moment (My) to be developed on the edges 

of the wide face . However ,it does not satisfy equilibrium of moments at the corners 

, since (My) remains zero at the edges of the narrow face . Thus the equilibrium 

equation of the structure can be also satisfied by using the technique the total 

potential energy ( V ) is the sum of its strain energy ( U ) and the work done (T) by 

the external loading system . 

V = U t + Tt 

(30) 

The total strain energy U t due to bending and twisting of the plate of the wide 

and narrow faces can be found as :- 

U t 

= 

1 

16 

1 

64 

2 

A 1 D 

3 2 

a d b 

D π 

3 

a d 

4 

3 

π 

⎛ 4 A 

⎜ 

⎜ 

⎜ 

⎝ A 

2 

1 

2 

2 

4 

2 2 2 

( b + a ) 

4 4 

2 2 

( a + d + 2 d a ) 

4 

4 

2 2 

( 3 d + 16 a + 8 d a ) 

4 4 

2 2 

( d + a + 2 a d ) 

1 

3 

D π A 1 A 2 

3 

d 

3 3 

a 

The total work done by the external loading system of wide and narrow 

faces can be determined by using the form : 

+ 

+ 

⎞ 

⎟ 

⎟ 

⎟ 

⎠ 

+ 

363 

(31)


T t 

= 

1 

16 

1 

3 

2 

A 1 π 

ad 

2 

A 

1 

2 

A 2 

ad 

π 

2 

b 

3 

t σ + 

d t σ 

2 2 ( 4 A + A ) 



1 

64 

π 

2 

d t σ 

a 

The potential energy (V) of the section can be expressed be eq.(17) : 

2 

1 A1 

D 

= 3 2 

16 a d b 

4 2 2 

π ( b + a 

2 

) + 

V t 

⎛ 4A 

4 

1 Dπ 

⎜ 

⎜ 3 3 

64 a d ⎜ 

⎝ A 

2 2 

1 A1 

π 

b 2 

16 ad 

2 4 4 2 2 

1 ( a + d + 2d 

a ) + ⎞ 

⎟ 

3 4 4 2 2 

1 Dπ 

A1 

A2 

( d + a + 2a 

d ) 

⎟ + 

3 3 

d a 

2 4 4 2 2 ⎟ 3 

2 ( 3d 

+ 16a 

+ 8d 

a ) ⎠ 

3 

t σ − 

2 

π d t σ 2 2 ( 4A 

+ A ) − 

1 A1 

A2 

π 

d t σ 

1 

64 

a 

1 

2 

The potential energy (V) is a minimum for the stable structure in which deformed 

shape is compatible with the support conditions so that: 

and 

∂ v 

∂A 

∂ v 

∂A 

1 

= 

= 

0. 

0 

0. 

0 

2 

The plate buckling coefficients ( Kd and Kb ) for the wide and narrow 

faces can be obtained using the algebraic program Mathematic. The minimum 

values of Kd and Kb do not have a closed solution thus require the minimum value to 

be found as the envelope to families of Kd and Kb functions obtained for discrete 

values of φ . The envelope can be approximated by the function : 

and 

d 

⎛ η − . 955 ⎞ 

= 6 . 97 ⎜ ⎟ 

⎝ η + 20 ⎠ 

0 . 09 

3 

ad 

2 

1 

2 

− 

+ 

(32) 

(33) 

(34) 

(35) 

k (36) 

0 . 09 

6 . 97 ⎛ η − . 955 ⎞ 

k b = ⎜ ⎟ 

2 

(37) 

η ⎝ η + 20 ⎠

4- Finite Element Model 



The ANSYS models are used to study the behavior of rectangular hollow 

sections under compressive load by incorporating all the nodes, element, material 

properties, dimensions and boundary conditions. The geometric boundaries of the 

HRS such width , length and height, etc., should be defined at the first step in solid 

modeling. Soild45 element is used for the analysis of square and rectangular hollow 

sections to determine the buckling coefficients for the wide and narrow faces. This 

element is defined with eight nodes each having 3 degrees of freedom: translations 

in the nodal x-, y- and z-directions. 

In the current study, nonlinear buckling analysis is used, taking into 

account both of the nonlinear response of the material and the nonlinear behavior of 

the structure. The analysis was started with a certain value of load not less than the 

ultimate load for perfect column. This load was automatically divided into equal or 

unequal steps (Force Control or Displacement Control) until the program stops due 

to the control of load or displacement or the convergence of solution. 

Buckling can be defined as the sudden deformation which occurs when the 

stored membrane energy is converted into bending energy with no change in the 

externally applied load. Buckling occurs when the total stiffness matrix becomes 

singular. 

The formulation of a linear static problem for solution by the displacement 

method is fully described by the matrix equation: 

[ K ]{ U} 

= { F} 

= { Fa} 

+ { Fc} 

The buckling problem is formulated as an eigenvalue problem: 

( [ K ] [ ] ) { ψ } = { 0} 

λ 

Where 

+ i i S 

[K] Structural stiffness matrix. {Fc } Reaction (or single point constraint) 

{U } Vector of unknown nodal 

displacements 

λi 

forces vector 

eigenvalue (used to multiply the 

loads which generated [S]) 

{F } Load vector [S ] stress stiffness matrix. 

{Fa } Applied nodal loads vector {ψ } i 

eigenvector of displacements 

365 

(38) 

(39)


The computation of buckling load factors is an eigenvalue problem. The 

computation of natural frequencies and mode shapes is known as modal or normal 

modes analysis. Figs (6,7) shows a computer generated plot of the finite 

element model which had dimensions of square and rectangular hollow sections. 

Fig (6 ) The square hollow section nodes 

and elements ( aspect ratio η = d/b =1). 



Fig ( 7) Nodes and elements for rectangular 

hollow section (aspect ratio η = d/b =1.5). 

The deformed shapes of square and rectangular hollow sections resulting from the finite 

element model using ANSYS are shown in Figs (8) and (9) respectively.

Fig ( 8-a) The deformed shape of square 

hollow cross section kb = kd = 4 

Fig.(9-a) The deformed shape of 

rectangular hollow cross section ( aspect 

ratio η = d/b =1.5) 




Fig ( 8-b ) The behavior of square hollow 

section in wide and narrow faces 

Fig ( 9-b) The Behavior of rectangular 

hollow section in wide and narrow faces 

The plate buckling coefficient (k) can to be obtained from a rational 

buckling analysis. This method requires the cross-section to be treated as a plate 

assembly taking the interaction between component plates into account. It uses the 

standard values of k of 4 and 0.5 for stiffened and unstiffened elements respectively. 

367


It can be seen from the given results that the use of figures ( 2,3,4 and 5 ) and 

formulae (23,26 and 27) reduces considerably the effort required to directly calculate the 

critical buckling stress in wide and narrow faces. From figures (10,11 and 12 ) the 

comparison of the buckling coefficients for wide and narrow faces is shown as obtained 

from the different methods (energy method and finite element method ) and with values 

obtained from a finite element , the energy method is accurate within 3% for the values 

of two term , while the one term analysis by 18% . 

Direct calculations also allow the designer to incorporate the available 

information in the design stage. The effect of interaction between adjacent elements 

has been investigated for rectangular hollow sections. The interaction was accounted 

for by determining the plate buckling coefficients Kd and Kb for the wide and 

narrow faces from a rational buckling analysis. 

K d 

6.0 

5.5 

5.0 

4.5 

4.0 

One term 

Two term 

Finite element 

1.0 1.1 1.2 1.3 1.4 1.5 

Fig (10 ) Comparison between the buckling coefficients of wide face K d using single term,two 

term analysis and finite element method 

K d 

7 

6 

5 

4 

3 

One term 

Two term 


zero error 

0% 

error 18% 



1.2% 

η 

18% 

2.1% 

18% 

3% error 

18% 

1.0 1.5 2.0 2.5 3.0 

Fig ( 11 ) Comparison between the buckling coefficients of wide face K d using single 

term,two term analysis and finite element method 

η 

3%

K b 



Fig (12) Comparison between the buckling coefficient of narraw face K b using single 

term , two term analyses and finite element method 

CONCLUSIONS 

The theoretical analysis presented herein has shown to a certain degree of 

accuracy the behaviour of rectangular hollow sections under compressive loads . 

Rectangular hollow sections can be analyzed directly by using the minimum 

potential energy technique. From the previes study the following conclusions can 

be drown :- 

1- Plate geometry and rigidity are important development in the analysis and in the 

application of hollow section results . 

2- The plate buckling coefficients are minimum values corresponding to a half 

wavelength for 

4.0 

3.5 

3.0 

2.5 

2.0 

1.0 1.1 1.2 1.3 1.4 1.5 

( ) 4 

1 

2 3 

η −η 

η 

η 

One term 

Two term 


a 

φ = + 

where φ = 

b 

3-The effect of interaction between adjacent elements has been investigated for 

rectangular hollow sections. 

4-The interaction was accounted for by determining the plate buckling coefficients 

for the wide and narrow faces from a rational buckling analysis. 

5-It was assumed that the section strength can be more accurately determined by 

using the actual plate buckling coefficients than the classical value of k=4. 

6-The results indicate that while the section strength increases at small plate 

slenderness values, it decreases at high slenderness ratios when based on a 

rational buckling analysis. 

369


7- The results of the analysis were directed to obtain a design approach for square 

and 

rectangular hollow sections . 

8-The results of the study have been cast into forms and graphs which can 

be easily used in design. 

LIST OF SYMBOLS 

U b : Strain energy for narrow face 

U d : Strain energy for wide face 

U : total strain energy 

t 

t : thickness of hollow section 

a, b : dimension of narrow face 

a, d : dimension of wide face 

η : d / b 

α : π / a 

β : π / b 

φ : a / b 

3 

Et 

D : 

2 

12( 

1−ν 

) 

ν : Poisson’s ratio. 

E : Modulus of elasticity. 

REFERENCE 

ω ω′ 

′ 



ω : Deflection function. 

′, ,= 

2 

∂ω ∂ ω 

: , 2 

∂x 

∂x 

,………….. 

.. .. 

ω ,ω ,= : 

2 

∂ω ∂ ω 

, ,…………… 

2 

∂y 

∂y 

T : Work done due to loads 

V : Total potential energy 

σ x : Applied Compressive stress 

σ crb : Critical Compressive stress 

for narrow face. 

σ crd : Critical Compressive stress 

for wide face. 

[1] Bleich F “ Buckling Strength of Metal Structures.” New York, NY, 

McGraw-Hill 1952. 

[2] Abdel-Lateaf,T.H.,” Buckling of stiffened plates Subjected To 

Combined Shear and Compression Loading. “ Ph..D Thesis 

University College , London 1982 

[3] Abd el-Sayed , M. ASCE “ Effective width of thin plates in 

compression “ Journal of the Structural Div ., vol .95 , No. ST10 , oct 

1969.



[4] El-Aghoury ,M.A.K. ,” Stability of steel plates axially loaded stiffened 

plates “ Master Thesis , Ain Shams university ,1980 . 

[5] Harris ,G .Z “ Buckling and post buckling of orthotropic plates : AIAA 

Journal, . vol .14 , No. 11 , Nov. 1976. 

[6] Harris ,H.G. and Pifco ,A.B “ Elastic – Plastic buckling of stiffened 

rectangular plates” Proc. Symp. App. Finite Element Method in Civ. 

Eng . Vanderbilt University , Nov . 1976. 

[7] Korashy, A.A., Mokhtar, A.A and M. A., “ Structural Analysis of 

Rectangular Plate Having Rectangular Hole by Using Finite 

Difference and S.O.R Method.”4th Arabic Civil Engineering 

Conference, 18-21 November, 1991, Vol. 2. 

[8] Graham H. Powell “ Analysis of orthotropic steel plate bridge decks ” 

, Journal of the struct. Division ASCE Vol 95, No. ST5 , May , 1969 

. 

[9] Bulson, P.S., “ The Stability of Flat Plates.”, Chatto and Windus, 

London, 1970 

[10] Dabaon , M. A. and Atia, G. M. “ Buckling analysis of Steel Plates 

with stiffened Opening subjected to in plane combined stresses .”, 

Al-Azhar Engineering 7 th International Conference Cairo 7-10 

April 2003 

[11] Abdel-Lateaf,T.H., Sedky A.T , Afaf ,A.M. , Omer ,M.A “ buckling 

loading for stiffened panels with holes “ Fourth Alexandria Int. Con. 

on Structural and Geotechnical Engineering 2-4 April 2001 

[12] Sedky . A.T “ Influence of Stiffener on Buckling of Perforated 

Rectangular Plates under Unixail Compressive Load “ 10 th 

International Colloquium on Structural and Geotechnical Engineering 

Ain Shams University Cairo 22-24 April 2003 . 

371




SEISMIC PERFORMANCE OF SHEAR DEFICIENT EXTERIOR RC 

BEAM-COLUMN JOINTS REPAIRED USING CFRP COMPOSITES 

Y. A. Al-Salloum, S. H. Alsayed, T. H. Almusallam and N. A. Siddiqui 

Department of Civil Engineering, KSU, P.O. Box 800, Riyadh, Saudi Arabia 

ABSTRACT 

In this paper, efficiency and effectiveness of using Carbon Fiber Reinforced 

Polymers (CFRP) sheets in repairing and upgrading the shear strength and ductility 

of seismically deficient exterior beam-column joint has been studied. For this 

purpose, a reinforced concrete exterior beam-column sub-assemblage was 

constructed with non-optimal design parameters (inadequate joint shear strength 

with no transverse reinforcement) representing pre-seismic code design construction 

practice of joints and encompassing the vast majority of existing beam-column 

connections. The specimen was subjected to cyclic lateral load histories so as to 

provide the equivalent of severe earthquake damage. The damaged specimen was 

repaired using CFRP sheets and then subjected to the similar cyclic lateral load 

history and its response history was obtained. Response histories of the specimen 

before and after repair were then compared. The results were compared through 

hysteretic loops, load-displacement envelops, ductility and stiffness degradation. 

The comparison shows that CFRP sheets improve shear resistance and ductility of 

the joint substantially. 

KEYWORDS 

Beam-column joints, exterior joint, CFRP, seismic, retrofitting, repair, 

cyclic loads. 

INTRODUCTION 

Majority of the pre-1970 constructed reinforced concrete (RC) frame 

buildings existing across the world and within the Kingdom of Saudi Arabia are 

shear deficient as they were constructed before the introduction of Seismic Code for 

construction. Recent earthquakes have illustrated that inadequate shear 

reinforcement in the existing beam-column joints, especially exterior ones (Figs.1a 

and 1b), is the prime cause of failure/collapse of moment resisting RC frame 

buildings. Hence, effective and economical rehabilitation techniques to upgrade 


SEISMIC PERFORMANCE OF SHEAR DEFICIENT EXTERIOR RC BEAM-COLUMN 

joint shear-resistance in existing structures are needed. In the past, variety of 

techniques have been employed to upgrade shear capacity and ductility of RC joints, 

with the most common being construction of RC or steel jackets. Plain or corrugated 

steel plates have also been tried. These techniques cause various difficulties in 

practical implementation at the joint, namely intensive labor, artful detailing, 

increased dimensions, corrosion protection and special attachments. To overcome 

the difficulties associated with these techniques, recent research efforts have 

focused on the use of epoxy-bonded FRP sheets or strips with fibers oriented 

properly so as to carry tension forces due to shear. 

Exterior joint 

Figure1a. RC building frame showing an exterior joint. 

Figure 1b. A failed exterior joint (Turkey Earthquake, August 1999). 




Y. A. Al-Salloum, et al 

In the last four decades several research papers have been published on the 

effect of seismic loads on poorly detailed reinforced concrete beam-column joints, 

typical of pre-seismic code designed moment resisting frames. Pantazopoulou and 

Bonacci [1], Cheung et al. [2], Pantazopoulou and Bonacci [3], Hakuto et al. [4], 

Hwang and Lee [5], Baglin and Scott [6] are some of the important contributions. 

The research papers, however, on FRP repaired/upgraded beam-column joints are 

limited. Antonopoulous and Triantafillou [7] conducted a comprehensive 

experimental program through 2/3-scale testing of 18 exterior joints. Their study 

demonstrated the role of various parameters, e.g. area fraction of FRP, distribution 

of FRP etc, on shear strength of exterior joints. They also highlighted the importance 

of mechanical anchorages in limiting premature debonding. Ghobarah and Said [8], 

El-Amoury and Ghobarah [9] developed an effective selective rehabilitation scheme 

for RC beam-column joints using advanced composite materials. Mukherjee and 

Joshi [10] studied experimentally the effect of FRP in improving shear strength and 

ductility of RC beam-column joints under simulated seismic forces. Ghobarah and 

El-Amoury [11] developed effective rehabilitation systems to upgrade the resistance 

to bond-slip of the bottom steel bars anchored in the joint zone and to upgrade the 

shear resistance of beam-column joints. 

A detailed review of literature shows that systematic studies to determine 

the behavior of the FRP repaired/upgraded members under cyclic loading are still 

limited. Moreover, the behavior of seismically excited FRP repaired beam-column 

joints is not well established at various stages of response e.g. before and after 

yielding of reinforcements, crushing of concrete, fiber fracture or debonding. The 

present paper is also an effort in the same direction. In this paper, efficiency and 

effectiveness of Carbon fiber reinforced polymers (CFRP) in repairing and 

upgrading the shear strength and ductility of seismically deficient exterior beamcolumn 

joint has been studied. For this purpose, a reinforced concrete exterior beamcolumn 

sub-assemblage was constructed with non-optimal design parameters 

(inadequate joint shear strength with no transverse reinforcement) representing preseismic 

code design construction practice of joints and encompassing the vast 

majority of existing beam-column connections. The specimen was subjected to 

cyclic lateral load histories so as to provide the equivalent of severe earthquake 

damage. The damaged specimen was then repaired using CFRP sheets. This repaired 

specimen was subjected to the similar cyclic lateral load history and its response 

history was obtained. Response histories of the specimen before and after repair 

were then compared. The results were compared through hysteretic loops, loaddisplacement 

envelops, ductility and stiffness degradation. 

375



Test Specimens 

In finding out the size of exterior joint specimen, first a prototype member 

size was chosen and then a crude analysis was carried out to come up with the most 

reasonable scale for the test specimen that comply with the available testing facility 

and equipment. Half-scale beam-column joint was found to be the most convenient. 

The dimensions and details of the half-scale test specimen are shown in Fig. 2a. The 

specimen was constructed with no transverse reinforcement (Fig. 2b), representing 

pre-seismic code design construction practice of joints and encompassing the vast 

majority of existing beam-column connections. 

Having decided the size of the test specimen, a reinforced concrete joint 

specimen was cast. The specimen was then subjected to cyclic lateral load histories 

so as to provide the equivalent of severe earthquake damage. The damaged specimen 

was then repaired through injecting epoxy into the cracks and externally bonding the 

specimens with CFRP sheets, as shown in Figs. 3a and 3b. 

4 PVC Pipes 

4 PVC Pipes 

35 cm 

60 cm 

Column 

16 x 30 cm 

40 cm 

30 cm 

60 cm 

60 cm 



6-cm Slab 

Column 16 x 30 cm 

PL 40 x 40 x 4 cm 

Beam 16 x 35 cm 

Top Box 60 x 60 x 30 cm 

Figure 2a. Schematic diagram of exterior joint specimen.

Test Setup 

ACTUATOR SIDE 

6 Threaded Rods 

(Dia = 2.5 cm; L = 29 cm) 

φ6 Hoops 

φ6 Stirrups 

No shear reinforcement 

in the joint region 

Figure 2b. Reinforcement details of the specimen. 



φ10 Stirrups 

The specimens were tested using the testing apparatus designed and installed in 

the Structural Test Hall, Department of Civil Engineering, King Saud University, Saudi 

Arabia. To apply the simulated seismic type cyclic load on the specimen, a 500-kN 

servo-controlled hydraulic actuator was connected to a reaction steel frame as shown in 

Fig. 4. The bottom of the column surface was attached to a base pivot using 4 high 

strength threaded rods. The base pivot, in turn, was fastened to a strong steel I-beam. 

The latter was post-tensioned to the lab floor using high strength post-tensioning rods. 

The rigid end of the concrete beam was tied to rigid link through steel pivots. 

Testing Procedure 

To test the specimens horizontal-loading regime was used. The said loading 

was based on the conventional guidelines of quasi-static type testing as followed by 

most researchers in simulating seismic forces to test reinforced concrete structures. 

The loading cycles were controlled by the peak displacement until failure. For each 

displacement level, three fully reversed cycles were completed. It is important to 

note that the frequency of applied load (or induced displacement) was maintained 

constant throughout the test program; it was picked up to be around one cycle per 

minute, which corresponds to a frequency of 0.0167 hertz. All cycles were started 

with the pull direction first then went into the push direction. 

377


Top Box 

(1) Layer of CFRP Sheet all 

around the Column Section 

(Length =105cm & Width =30 

) 

(1) Layer of CFRP Sheet all 

around the Column Section 

(Length = 105 cm & Width = 30 cm) 

R/C Column 

Figure 3a. Schematic representation of FRP repaired specimen 

Figure 3b. Picture showing FRP repaired specimen. 



30 

30 

R/C Slab 

R/C Beam 

Rigid Beam

3 cm 

300 cm 

93 cm 

100 cm 

150 cm 

300 cm 

INSTRON 

Actuator 


Hydraulic Jack 

Annular Load Cell 


Strong floor of 

concrete 

(Thick. = 93 cm) 

Figure 4. Schematic diagram showing the test set up for exterior connection specimens. 

DISCUSSION OF TEST RESULTS 

In the present section, through various experimental results, the effectiveness 

of CFRP in improving the as-built joint shear strength and ductility has been studied. 

The results are presented and discussed under the head of general behavior, 

hysteretic loops, load-displacement envelopes and stiffness degradation. 

General Behavior 

Figure 5 shows the general response of joint specimen under lateral cyclic 

loading. This figures shows that during the displacement controlled loading stages; 

significant X-shear cracks appeared in the specimen almost symmetrically on both 

faces of the joint. The shear cracks initiated in diagonal directions and propagated 

towards the ends of joint. This may be attributed to diagonal tension caused due to 

excessive shear stresses in the joint. 

Having completed the tests on above specimen, the damaged specimen was 

repaired through injecting epoxy into the cracks and externally bonding the 

specimens with CFRP sheets (Fig. 3). Externally bonded CFRP sheets were 

expected to provide shear resistance to the joint which in turn may add strength and 

ductility to the joint. It was observed that use of CFRP sheets delayed shear failure 

379


of the joint significantly and failure was primarily due to debonding and tearing of 

CFRP sheets in the beam region (Fig. 6). This is due to the fact that at higher stages 

of loading, there was significant yielding in beam reinforcing steel bars that allowed 

cracks to widen in the beam region which in turn ultimately tore the CFRP sheets. 

Figures 5. Development of cracks in the exterior RC beam-column joint specimen. 

Tearing and 

de-bonding 

Figure 6. Failed FRP repaired exterior beam-column joint specimen. 



Hysteretic Behavior 



The hysteretic behavior of exterior joints was examined in terms of shear 

strength (measured in terms of ultimate load) and deformation capacity. The loaddisplacement 

relationships for specimens before and after the repair are shown as 

hysteretic curves in Figs. 7 and 8. Fig. 8 shows that the ultimate load for repaired 

specimen is substantially higher than its corresponding original (before repair) 

specimen (Figs. 7 and 8). This is primarily due to the increased confinement of joint 

resulting from externally bonded CFRP sheets. A further comparison of deformation 

capacity of repaired specimens with the original (i.e. before repair) specimen 

illustrates that the use of CFRP increases the deformation capacity of repaired 

specimens considerably. 

Lateral load (kN) 

90 

70 

50 

30 

10 

-10 

-30 

-50 

-70 

-90 

-55 -45 -35 -25 -15 -5 5 15 25 35 45 55 

Lateral displacement (mm) 

Figure 7. Load-displacement hysteretic plot for original (before repair) specimen. 

381



90 

70 

50 

30 

10 

-10 

-30 

-50 

-70 

-90 

-55 -45 -35 -25 -15 -5 5 15 25 35 45 55 


Figure 8. Load-displacement hysteretic plot for repaired specimen. 

Load-Displacement Envelopes 

In order to study load carrying capacity and ductility of original (before 

repair) and repaired exterior joint specimens, envelopes of load-displacement 

hysteretic curves for these two specimens are plotted and shown in Figure 9. Using 

these envelopes the peak load, ultimate displacements, and ductility for the 

specimens are obtained and listed in Table 1. The second column of Table 1 shows 

the average peak load (i.e. average of peak push and pull values) and third column 

shows the displacement corresponding to first yield of steel bars. This displacement 

is required to calculate ductility of the specimens. The estimated ductility, an 

important parameter for earthquake resistant construction, is shown in the last 

column of Table 1. The ductility is computed as the ratio of ultimate displacement to 

the displacement at first yield of internal steel. For computation, the ultimate 

displacement was set at a displacement corresponding to 20% drops of peak load. 

The values of ductility clearly show that the application of CFRP sheets has 

improved the ductility of repaired specimen significantly. This increase in the 

ductility is up to 39% with respect to the before repaired specimen. 



Specimen 

Before 

repair 

Table 1: Peak test load and maximum ductility 

Peak load 

(Average) 

kN 

Disp. at first 

yield of 

steel, 

∆ (mm) 


y 

Disp. at 20% 

drop of peak 

load, ∆ 20 

(mm) 


Ductility 

Factor 

∆ / 

20 

47.08 18.67 30.0 1.61 

After repair 81.79 18.67 * 

*Taken same as respective “before repair” values. 

40.7 2.24 

Table 1 illustrates that the increase in average peak load for repaired 

specimen is substantially higher than the original (before repair) as-built specimen 

by 74%. This is highly encouraging trend and may be attributed to excellent 

performance of CFRP sheets attached to the damaged specimen. 


90 

70 

50 

30 

10 

-10 

-55 -45 -35 -25 -15 -5 5 15 25 35 45 55 

-30 

-50 

-70 

-90 


Figure 9. Envelope of hysteretic loops. 

Before repair 

After repair 

∆ 

383 

y


The ductility, an important parameter for earthquake resistant construction, 

is shown in the last column of Table 1. The ductility is computed as the ratio of 

ultimate displacement to the displacement at first yield of internal steel. For 

computation, the ultimate displacement was set at a displacement corresponding to 

20% drops of peak load. This table shows that the application of CFRP sheets has 

improved the ductility of repaired specimen significantly. Magnitude-wise the 

increase in ductility for repaired specimen is up to 39% with respect to its original 

(before repair) specimen. 

Stiffness Degradation 

The beam-column joint stiffness is estimated by computing the slope of the 

peak-to-peak line in each loop [9]. Figure 10 shows the stiffness degradation with 

lateral displacement. This degradation can be attributed to concrete non-linear 

deformations, flexural and shear cracking, distortion of the joint panel, slippage of 

reinforcement, loss of cover, debonding or delamination of CFRP etc. A comparison 

of repaired specimen curve with before repair curve shows that the initial stiffness of 

repaired specimen is significantly higher than before repair specimen. This high 

initial stiffness for repaired specimen may be attributed to external bonding of CFRP 

sheets on beams, joint and column regions. Figure 10 also reveals that, in CFRP 

repaired specimen, the degradation of stiffness with lateral movement are slow 

compared to before repair specimen. This is a desirable property in earthquake like 

situations. It was observed, in the past earthquakes, most of the RC structures failed 

(or collapsed) due to sudden loss of stiffness of structural joints with increasing 

lateral movement of the structure. 

Stiffness (kN/mm) . 

10.00 

8.00 

6.00 

4.00 

2.00 

0.00 

Before repair 

After repair 

0 10 20 30 40 50 60 


Figure 10. Stiffness degradation in the specimens. 



CONCLUSIONS 



The results of the experimental program, presented in this paper, establish 

the effectiveness of CFRP sheets in repairing and upgrading deficient exterior beamcolumn 

joints. The results of CFRP repaired specimen was compared with its 

corresponding before repair specimen and, in general, it was observed that CFRP 

sheets improve the shear resistance and ductility of the RC joint to a great extent. 

The failure of CFRP repaired exterior joint due to debonding was examined and it 

was observed that at higher stages of loading there was significant yielding in beam 

reinforcing bars that allowed cracks to widen in the beam region which in turn 

ultimately tore the CFRP sheets. 


Authors acknowledge the financial support provided by King Abdualaziz 

City for Science and Technology (KACST) under grant Number AR-21-40. 

REFERENCES 

1. Pantazopoulou, S. and Bonacci, J. 1992. “Consideration of questions about 

beam-column joints,” ACI Structural Journal, 89(1): 27-36. 

2. Cheung, P.C., Paulay, T., and Park, R. 1993. “Behavior of beam-column joints 

in seismically loaded R.C. frames,” Journal of The Structural Engineer, 71(8): 

129-138. 

3. Pantazopoulou, S.J., and Bonacci, J.F. 1994. “On earthquake-resistant 

reinforced concrete frame connections,” Canadian Journal of Civil 

Engineering, 21: 307-328. 

4. Hakuto, S., Park, R. and Tanaka, H. 2000. “Seismic load tests on interior and 

exterior beam-column joints with substandard reinforcing details,” ACI 

Structural Journal, 97(1): 11-25. 

5. Hwang, S. and Lee, H. 2000. “Analytical model for predicting shear strengths 

of interior reinforced concrete beam-column joints for seismic resistance,” ACI 

Structural Journal, 97(1): 35-44. 

6. Baglin, P.S., and Scott, R.H. 2000. “Finite element modeling of reinforced 

concrete beam-column connections,” ACI Structural Journal, 886-894. 

385


7. Antonopoulos, C., and Triantafillou, T.C. 2003. “Experimental investigation of 

FRP-strengthened RC beam-column joints,” Journal of Composites for 

Construction, ASCE, 7(1): 39-49. 

8. Ghobarah, A., and Said, A. 2001. “Seismic rehabilitation of beam-column joints 

using FRP laminates,” Journal of Earthquake Engineering, 5(1): 113-129. 

9. El-Amoury, T. and Ghobarah, A. 2002. “Seismic rehabilitation of beam-column 

joint using GFRP sheets,” Engineering Structures, 24:1397-1407. 

10. Mukherjee, A. and Joshi, M. 2005. “FRPC reinforced concrete beam-column 

joints under cyclic excitation,” Composite Structures, 17: 185-199. 

11. Ghobarah, A. and El-Amoury, T. 2005. Seismic rehabilitation of deficient 

exterior concrete frame joints. Journal of Composites for Construction, ASCE, 

9(1): 408-416. 



COMAPATIVE STUDY ON FOUR METHODS FOR THE 

DETERMINATIONS OF THE COEFFICIENT OF CONSOLIDATION 

Abdulhafiz Omar S. Alshenawy 

Associate Professor, Civil Engineering Department, King Saud University, 

P.O.Box 800, Riydh 11421, shenawi@gmail.com 

ABSTRACT 

Four available methods are used for determining the coefficient of 

consolidation, cv: Casagrande's logarithm of time method, Taylor's root of time 

method, the improved rectangular hyperbola method and the early stage of log-t 

method. The first two methods are commonly and widely used in practice. However, 

some difficulties are experienced in using them to determine the value of cv. 

Casagrande’s method in general gives the lowest values of cv , and Taylor’s method 

gives higher values of cv than that of Casagrande’s method especially at low applied 

pressure. At higher values of applied pressure, the cv values of both methods are 

close to each other. The early stage of log-t method gives higher values than 

Taylor’s and Casagrande’s methods and lower values than the improved rectangular 

hyperbola method. The improved rectangular hyperbola method is fast, easy and 

gives more consistent results than the others give. This method is suggested for 

determining the cv value. The empirical equations for the compression index and 

swell index versus liquid limit are formulated and presented. In addition, the 

empirical equations for the swell index versus compression index and the plasticity 

index versus clay percent are presented. 

KEY WORDS 

Coefficient of consolidation, compression index, swell index, plasticity index. 

INTRODUCTION 

The assessment of time rate of settlement of a soil undergoing deformation 

under loading requires, in the range of the primary consolidation, the knowledge of 

the coefficient of consolidation, cv, of the soil. It must be determined correctly to 

predict the true time rate of settlement. Generally, the value of cv is obtained from 

the one dimensional consolidation test by means of curve fitting procedures and 

based on Terzaghi’s one-dimensional consolidation theory. This value may not be 


COMAPATIVE STUDY ON FOUR METHODS FOR THE DETERMINATIONS OF THE COEFFICIENT 

representative of the soil that experiences the deformation in the field due to various 

factors such as the inherent assumptions involved in the consolidation theory, in the 

methods for evaluation of cv from the laboratory consolidation test data and the 

differences between laboratory testing conditions and those in the field. In situ 

settlements are affected by many variables, principally drainage condition, thickness 

of soil layer, and macrostructure characteristics of the layer. These aspects are not 

taken into account during laboratory tests, therefore differences between predicted 

and real behavior of soils are expected. Attempts have been made in the past to use 

in situ testing techniques in evaluating the field time rate parameter, but they have 

not succeeded due to complicated testing procedures and interpretation techniques. 

The coefficient of consolidation depends on the permeability and the 

compressibility of the soil. Both of them decrease as consolidation pressure 

increases and accordingly the coefficient of consolidation changes. However, the 

coefficient of consolidation is usually assumed constant. This assumption creates 

limitations that are [1]: 

1. The coefficient of permeability remains constant during consolidation. 

2. The relation between void ratio and effective stress is linear. 

3. The relation between void ratio and effective vertical stress is independent of time. 

Selecting a single value of cv to represent the entire clay layer throughout a range of 

pressures at in-situ condition is not a straightforward matter. Factors that 

complicate the selection of a single cv value include [2]: 

1. The value of cv is larger at pressures below the preconsolidation pressure than at 

pressures above the preconsolidation pressure. 

2. The drainage path length decreases significantly as the pressure increases. Values of cv 

calculated using the initial value of drainage will be significantly different from values 

of cv calculated using values of drainage path after compression. 

In practice, Engineers use laboratory test data to estimate the rate of settlement of a soil 

and has to decide which of the methods needed to be used. So the main objective is to study 

the variation of the coefficient of consolidation values using different methods on local 

soils in order to draw a conclusion on which of those methods is more realistic and 

reasonable to predict the time rate of settlement in the range of the primary consolidation. 

Four available methods are used for determining the coefficient of consolidation: 

Casagrande's logarithm of time method, Taylor's root of time method, the improved 

rectangular hyperbola method and the early stage of log-t method. The values of the 

coefficient of consolidation using these methods are determined and compared. The effects 

of applied pressures and the mechanical properties of soils on the coefficient of 

consolidation are also investigated. The coefficient of consolidation has been related to the 

liquid limit and other index properties. The second important issue is the relationship 

between the coefficient of consolidation and the applied pressure that is also presented. In 

addition, other index properties such as the compression index cc and the swelling index cs 

were correlated with the plasticity index Ip and liquid limit. 



LITERATURE REVIEW 



In the literature there are many different methods used to determine the 

coefficient of consolidation. Most of the methods of obtaining cv from the laboratory 

time-delta data are graphical and are based on Terzaghi’s consolidation theory. The 

very existence of so many procedures itself is an indication that all methods may not 

be applicable under all circumstances, As the values of cv obtained by different 

methods vary widely, it is difficult to decide which value of cv is a reasonable 

estimate for the soil; at least in the laboratory testing conditions, though not the field 

behavior. 

Four different methods are used to determine the coefficient of 

consolidation and to compare among these values. Two of them are commonly and 

widely used in practice, the first is called the logarithm of time method proposed by 

Casagrande and Fadum [3] and the second method is called the square root of time 

method given by Taylor [4]. Both methods have the advantage of giving a visual 

picture of the consolidation process as the test proceeds. Casagrande method is 

affected by the secondary compression part of the consolidation curve and a lower 

value of cv is expected. However Taylor’s method is affected by the initial 

compression part of the consolidation curve which increases the value of cv. Taylor’s 

method yields to cv values larger than those obtained from Casagrande method. If 

the shapes of laboratory time curves are exactly similar to the theoretical shape, 

Casagrande and Taylor methods would result in the same value of cv. However, due 

to the fact that clay compressibility varies with effective stress and stress rate of 

strain, and due perhaps to other effects as well, the actual shapes differ from the 

theoretical shapes [2]. In spite of their wide acceptance, some difficulties are 

experienced in using the Casagrande’s and the Taylor’s procedures. They are [5]: 

1. Apart from the standard or ideal δ versus log t curve given by Casagrande, other 

curves shapes could be obtained. For different types of soils and loading 

conditions, the shapes of curves vary considerably. In all these cases, the 

determination of t100, (time corresponding to 100% primary consolidation) and 

cv poses a serious problem. 

2. There are cases in which the initial portion of the log t-compression curve is 

rather erratic. In such cases, .R0, and t0 are erroneous, and this affects t50 as well. 

3. Cases are also quite common where the t ½ compression curve is a continuously 

curved line with no specific initial straight line portion or where the initial 

straight line portion of the t ½ -compression when extended backwards gives a 

negative intercept for R0, on the δ-axis. 

In most cases, it is found that the rates of consolidation estimated using 

laboratory test data and conventional consolidation theory are slower than the actual 

rates of settlement observed in the field. So due to the effect of initial and secondary 

compressions on Casagrande and Taylor methods, other methods were proposed to 

389


determine the coefficient of consolidation taking into account reducing or 

eliminating the initial and secondary compressions effects. Thus if the procedure for 

cv determination is least affected by initial and/or secondary compression, then the 

value of cv obtained from such a procedure may be considered a reasonable estimate 

for the soil. 

Sridharan and Parakash [6] proposed the improved rectangular hyperbola 

method. They utilized the rectangular hyperbola method [5] with a simplified and 

accurate procedure that has been evolved and recommended. The rectangular 

hyperbola method assumes the theoretical time factor versus degree of 

consolidation, as obtained from Terzaghi’s equation, to be a rectangular hyperbola. 

The improved rectangular hyperbola method is simple to use and give good results 

for the average degree of consolidation in the range 60% ≤ U ≤ 90%. This method 

could be used for soils for which the conventional methods, namely, Casagrande and 

Taylor methods could not be used. It requires only the identification of the straightline 

portion beyond 60% consolidation in the t/δ versus t plot. 

The forth method was proposed by Robinson and Allam [7] called the early 

stage of log-t method. This method uses the earlier part of the consolidation curve 

and the values of cv are less affected by secondary compression than the 

conventional methods. As the influence of secondary compression is reduced, the 

value obtained by this method is greater than that yielded by both conventional 

methods. In this method, a laboratory time t and deflection δ data for a shorter 

duration are adequate for the determination of cv. 

The four methods mentioned above were used to determine the coefficient 

of consolidation. However, in the literature there are many other graphical methods 

proposed to estimate the value of cv. For the same data of a test, these methods may 

estimate different values of cv. Some of these methods would be presented here that 

are not used in this paper. Olson [8] suggested to obtain the coefficient of 

permeability (k) and the coefficient of volume compressibility (mv) and substitute 

them in the equation cv = k/mv γw. 

Sridharan and Prakash [9] proposed the one point method that is expressed 

in the form of log10 (H 2 /t) vs. U curves. It could be noticed that the effects of 

primary and secondary compressions are least in the range 40% U< 60%, and hence, 

the cv values obtained in that range may be considered to represent the soil 

reasonably well. The very observation that the experimental behaviors of soil 

without correction for initial and secondary compression effects match well with the 

theory in the range 40%



Even though the values of cv obtained by Taylor and Casagrande methods vary 

widely, their average values of cv appears to have near one to one match with the 

results from the one point method. The values of cv obtained by the one point 

method are slightly more than those given by the rectangular hyperbola method. 

Based on this idea, Narasimha, et al. [10] estimated the value of cv for 

vertical flow of normally consolidated, saturated uncemented soils using the stress 

state permeability relationships without carrying out the consolidation test. 

Comparing their results with the graphical methods used, it showed that the 

graphical methods give higher cv values than the estimated values. 

Cortellazzo [1] examined four methods to determine cv values from the 

laboratory and compared them with in situ values by measuring the settlement of the 

corresponding layer using borehole extensometers for three different Italian sites. 

These four methods are Casagrande method, Taylor method, the log (H 2 /t)-U 

method, and the log δ- log t method. The difference in the laboratory cv values is 

sometimes large and depends on the method used. The laboratory results showed 

that Casagrande method, in some cases, may not be usable or determined values and 

not comparable with those from other methods. The other methods showed more 

consistent evaluation of cv. The higher values of cv are generally from Taylor’s 

method and the log(H 2 /t)-U method, and the lower values from the log δ - log t 

method at U=88.3%. The findings showed large differences between in situ and 

laboratory values of cv for a clay layer thickness greater than 10 m and in good 

agreement for a thickness less than 5 m. 

Feng and Lee [11] proposed a simplified version of the Taylor’s 

method that using method by only drawing a straight line passing through the 

linear portion of the measured √t consolidation curve. The point at which the 

consolidation curve deviates from this straight line gives the time of 60% 

consolidation. Ten natural soft clay samples with liquid limits ranging from 40 

to 152% were used to carry out conventional oedometer tests with 

consolidation increments in the recompression range, spanning the 

preconsolidation pressure, and in the compression range. Based on the 

oedometer test results, the effects of secondary compression on the shape of 

the t 1/2 consolidation curve are evaluated and found limited between 60% and 

90% consolidation. In the proposed simplified t 1/2 method, the time of 60% 

consolidation is recognized from the lower end of the linear segment and is 

used together with the Terzaghi theoretical time factor of 0.286 and the 

maximum drainage distance of the oedometer specimen to determine the 

coefficient of consolidation. A large amount of the oedometer coefficient of 

consolidation data obtained from the simplified t 1/2 method are in good 

agreement with those from the Taylor t 1/2 method and are within one to two 

times those from the Casagrande logarithm of t method. 

391


Mesri et al. [12] proposed the inflection point method to determine cv from 

the experimental data of the tangent slope (m=∆δ/∆log t) with log t. Visual 

identification of the inflection point can be performed easily and cv determined. The 

values of cv obtained by this method are quite similar to those from Casagrade 

method. The advantage of this method is that it does require the definition of the 

beginning and end of the primary consolidation stage that are required by other 

methods. Another advantage is that the inflection point at average degree of 

consolidation of 70% is within the midrange of the compression curve and is least 

effected by the initial and secondary compressions. 

EXPERIMENTAL WORKS AND RESULTS 

The methodology is based on the experimental works that have to be 

clearly identified to accomplish the objectives. Therefore, different types of soil 

samples were collected from different regions of Saudi Arabia. The soils samples 

were classified experimentally by carrying out the fundamental tests such as the 

grain size analysis test, hydrometer test, specific gravity test; Atterberg limits tests, 

and compaction tests. The results of these tests are presented in Table 1. The 

consolidation tests were carried out using the odometer apparatus. Based on the 

standard compaction test results, the consolidation tests were carried out at 95% of 

their maximum dry densities with their corresponding optimum moisture contents at 

their dry sides. The coefficients of consolidation were determined at different 

applied pressures by different methods, namely: 

1. The logarithm of time method. 

2. The square root of time method. 

3. The improved rectangular hyperbola method. 

4. The early stage log-t method. 

To ensure the test results, two or three samples for each type of soils 

were used. 

Consolidation test 

The consolidation properties determined from the consolidation test 

used to estimate the magnitude and the rate of both primary and secondary 

consolidation settlements of a structure or an earth fill. Estimates of this type 

are of key importance in the design of engineered structures and the evaluation 

of their performance. The consolidation tests were carried out according to 

ASTM D2435. 



Table 1: Characteristics of tested soils. 



Property Nasim Dalam Dariyah Madinah Khurais Assalihyah 

L.L [%] 23.3 25.4 18.4 22.9 41.3 55.5 

P.L. [%] 20.1 23 15.9 18 24.2 35 

P.I. [%] 3.2 2.4 2.5 4.9 17.1 20.5 

Gs 2.72 2.74 2.75 2.69 2.76 2.8 

Gravel [%] - - 1.5 0.6 - - 

Sand [%] 61.7 49.8 60.2 54.6 20.1 13 

Silt [%] 30.8 45.2 25.3 28.9 23.4 40 

Clay [%] 7.5 5 13 16 56.5 47 

Unified SM ML SM SC-SM CL MH 

AASHTO A-4 A-4 A-4 A-4 A-7-6 A-7-5 

γ d ( max) 

[kN/m 3 ] 

17.95 17.64 19.3 18.45 15.45 14.84 

OMC [%] 12.5 17.36 10.2 13 21 25.3 

The samples of 70 mm diameter and 19 mm thickness were used. The 

specimens were prepared at 95 % of the maximum dry density and the 

corresponding water content on their dry side of the curves obtained from the 

standard compaction tests. Following the normal procedure, the specimens were 

loaded double the pressure at each new stage. The sequence of loading was 100, 

200, 400, 800, and 1600 kPa. Because the readings for the 24 hours did not show a 

393


flattening out from the steep part of the curve to a straight line which less steeply 

inclined in the deformation-log time graph, the duration of each load increment 

throughout the tests was increased to either 2 or 3 days. For each pressure increment, 

the readings of the deformation and time were recorded and the coefficient of 

consolidation cv was determined using the four methods. The specimens were then 

unloaded in a series of decrements and the final readings at each pressure decrement 

were recorded. The void ratio at the end of primary consolidation for each pressure 

increment was calculated and the log pressure versus void ratio was plotted. Based 

on this plot, compression index cc, and swell index cs were determined. 

For a given load increment of the consolidation test, the specimen 

deformation-time data are used to determine the coefficient of consolidation by the 

different four methods. The only difference among these methods is how to present 

the data to determine the coefficient of consolidation. Casagrande’s method refers to 

a typical t-δ curve on a semilogarithmic plot. In Taylor’s method, the experimental 

data are plotted as deformations versus square root of time. The improved 

rectangular hyperbola method is simple compared to the other method's procedures. 

The oedometer test results are plotted in the form of t/δ versus t plot. In the early 

stage log-t method, the experimental data are plotted in log scale. The corrected zero 

dial reading is obtained by assuming the initial portion as a parabola as in 

Casagrande’s method. The point of intersection between the tangent passing through 

the inflection point and the line passing through the corrected zero, which is parallel 

to the time axis, gives the time t22.14 . 

Consolidation tests results 

Consolidation tests were carried out on six different soils samples from 

different regions and with wide variation in their plasticity characteristics (Plasticity 

index ranging from 2.4 to 20.5). Consolidation test results were drawn for the four 

methods used to determine the coefficient of consolidation for each soil sample at all 

levels of applied pressures. The values of coefficient of consolidation were 

determined for all methods at all levels of applied pressures for each soil type. The 

best curve fitting fit the values of the coefficients of consolidation for each soil type 

determined by each method. This procedure gives the general behavior trend for 

each method. In some tests, it is difficult to use Casagrande’s method because the Sshape 

curve could not be obtained as in the theoretical procedure. This means that 

the t100 could not be determined and therefore the coefficient would not be 

determined. This is considered as one of the disadvantages of this method. Figs 1 to 

6 show the fitting curves for the four methods for each soil type. 

The compression index and the swell index were then determined from the 

void ratio-applied pressure curve for each soil type. The compression index versus 

liquid limit for all soils is shown in Fig. 7. As shown in this figure that the 

relationship is linear and could be written as: 





cc = a1 (LL + b1) (1) 

where a1 = 0.0037 and b1 = 23.35. The swell index vs. the liquid limit 

relationship was obtained from the data as shown in Fig. 8. The data was fitted to the 

linear relationship that could be written as: 

cs = a2 (LL + b2) (2) 

where a2 = 0.0014 and b2 = 9.3. The swell index vs. the compression index 

relationship was obtained from curve fitting of the data as shown in Fig. 9. The 

equation could be written as: 

cs = a3 cc 2 + b3 cc + c3 (3) 

where a3 = 2.04, b3 = -0.55 and c3 = 0.0514. Finally, as shown in Fig. 10, the 

plasticity index vs. clay percent (CP) was fitted with the following linear curve as: 

PI = a5 (CP + b5) (4) 

where a4 = 0.364 and b4 = -0.763. 

c v (mm/min) 

30 

25 

20 

15 

10 

5 

Casagrande 

Hyperbola 

Taylor 

Early 

0 

0 500 1000 

Pressure (kPa) 

1500 2000 

Fig. 1. Comparison among the four methods for Madinah soil. 

395


c v (mm/min) 

10 

8 

6 

4 

2 

0 

0 500 1000 


1500 2000 



Casagrande 

Hyperbola 

Taylor 

Early 

Fig. 2. Comparison among the four methods for Khurais soil. 

c v (mm/min) 

50 

40 

30 

20 

10 

Casagrande 

Hyperbola 

Taylor 

Early 

0 

0 500 1000 


1500 2000 

Fig. 3. Comparison among the four methods for Nasim soil.

c v (mm/min) 

40 

30 

20 

10 



Casagrande 

Hyperbola 

Taylor 

Early 

0 

0 500 1000 


1500 2000 

Fig. 4. Comparison among the four methods for Dariyah soil. 

c v (mm/min) 

30 

25 

20 

15 

10 

5 

Hyperbola 

Taylor 

Early 

0 

0 500 1000 


1500 2000 

Fig. 5. Comparison among the four methods for Dalam soil. 

397


c v (mm/min) 

5 

4 

3 

2 

1 

0 

0 500 1000 


1500 2000 



Casagrande 

Hyperbola 

Taylor 

Early 

Fig. 6. Comparison among the four methods for Assalihyah soil. 

Compression Index, Cc 

0.35 

0.3 

0.25 

0.2 

0.15 

0.1 

0.05 

y = 0.0037x + 0.0864 

R 2 = 0.8235 

0 

10 20 30 40 50 60 

Liquid Limit [%] 

Fig. 7. Relationship between compression index and liquid limit for all tested soils.

Swell Index, Cs 

0.07 

0.06 

0.05 

0.04 

0.03 

0.02 

0.01 

y = 0.0014x - 0.013 

R 2 = 0.968 



0 

10 20 30 40 50 60 

Liquid Limit [%] 

Fig. 8 Relationship between swell index and liquid limit for all tested soils. 

Swell Index, Cs 

0.07 

0.06 

0.05 

0.04 

0.03 

0.02 

0.01 

y = 2.0388x 2 - 0.5459x + 0.0514 

R 2 = 0.9071 

0 

0.1 0.15 0.2 

Compression Index, Cc 

0.25 0.3 

Fig. 9 Relationship between compression index and swell index for all tested soils. 

399


Plasticity Index [%] 

30 

25 

20 

15 

10 

5 

0 

y = 0.3642x - 0.2778 

R 2 = 0.9218 

0 10 20 30 

Clay Percent 

40 50 60 

Fig. 10 Relationship between plasticity index and clay percent index for all tested soils. 

DISCUSSIONS AND CONCLUSIONS 

Based on Terzaghi’s one dimensional consolidation theory, many curvefitting 

methods are available to interpret the laboratory oedometer compression for 

the determination of the coefficient of consolidation. Four different methods for the 

determination of cv have been evaluated and the values compared. The difference in 

the values of cv determined from the laboratory test is sometime large and depends 

on the methods used. The laboratory results show that the Casagrande’s method in 

some cases may not be used or determined values not comparable with those from 

other methods. 

As mentioned in the literature review that the rate of settlement in the field is 

faster than that evaluated experimentally. This is because of the factors affecting the 

rate of settlement in the field that are not considered experimentally. Therefore, it is 

better to choose a method that gives the highest value of cv. This method is usually 

not affected either by the initial compression part of the consolidation curve or by 

the secondary compression part of the consolidation curve. 





From the fitting curves for the four methods for each soil type, the following 

conclusions could be drawn: 

1. In some cases, the standard curve of δ versus log t for Casagrande method 

could not be obtained, other curves shapes obtained without the secondary 

compression part of the curve. Therefore, the determination of t100 poses a 

serious problem. 

2. In general, Casagrande method gives the lowest values of the coefficient of 

consolidation. 

3. In some cases, the compression curve of the Taylor method is continuously 

curved line with no specific initial straight-line portion. 

4. Taylor method gives in general higher values of cv than that of Casagrande 

method especially at low applied pressure. At higher values of applied 

pressure, the cv values of both methods are very close together. However, 

the early stage of log-t method gives higher values than Taylor and 

Casagrande methods and lower values than the improved rectangular 

hyperbola method. 

5. The improved rectangular hyperbola method gives in general the highest 

values of the coefficient of consolidation. 

6. Most of the methods used, the values of the coefficient of consolidation are 

decreasing with increasing applied pressures. 

7. For most of the methods, the differences in the values of cv at low applied 

pressures are much higher than that at high applied pressures. 

8. Casagrande and Taylor methods are sometimes difficult in determining the 

cv value. However, the improved rectangular hyperbola method is fast, easy 

and gives more consistent results than the others give. This method is the 

most suggested method for determining the cv value. 

9. The empirical equations for the compression index and swell index versus 

liquid limit were obtained. In addition, empirical equations for the swell 

index with compression index and the plasticity index with clay percent 

were obtained. 

REFERENCES 

[1] Cortellazzo, G., “Comparison Between Laboratory And In Situ Values of 

the Coefficient of Primary Consolidation cv” Canadian Geotechnical 

Journal, 39, pp. 103-110, 2002. 

[2] Duncan, M., “Limitations of Conventional Analysis of Consolidation 

Settlement” J. of Geot. Eng., Vol. 119, N 9, pp. 1333-1359, 1993. 

[3] Casagrande, A. and Fadum, R. E. "Notes on Soil Testing for Engineering 

Purposes," Harvard Graduate School of Engineering, Soil Mechanics 

Series, No. 8, 1940. 

[4] Taylor, D.W. "Fundamentals of Soil Mechanics" John Wiley and Sons, 

New York, 1948. 

401


[5] Sridharan, A. and Sreepada Rao, A., “Rectangular Hyperbola Fitting 

Method for One Dimensional Consolidation”, Ceotechnical Testing Journal 

GTJOD3, Vol. 4, No. 4, Dec. 1981, pp. 161-168. 

[6] Sridharan, A. and Prakash, K. "Improved Rectangular Hyperbola Method 

for Determination of Coefficient of Consolidation".Geotechnical Testing 

Journal, March 1985, ,pp37-40. 

[7] Robinson, R.G. and Allam, M.M., "Determination of Coefficient of 

Consolidation from Early Stage of Log t Plot". Geotechnical Testing 

Journal, September 1996, Vol.19, No.3, 316-320. 

[8] Olson, T. E. "State of the art: Consolidation Testing," Consolidation of 

Soils: Testing and Evaluation, ASTM STP 892, R. N. Yong and F. 

C. Townsend, Eds., American Society for Testing Materials, Philadelphia, 

pp. 7-70. 

[9] A Sridharan and Prakash K, " Determination of Coefficient of 

Consolidation : a user friendly approach ".Ground Engineering, February 

1998. 

[10] Narasimha Raju, P. S. R., Pandian, N. S. and Nagaraj, T. S. "Analysis and 

Estimation of the Coefficient of Consolidation", Geotechnical Testing 

Journal, GTJODJ, Vol. 18, No. 2, June 1995, pp. 252-258. 

[11] Feng, T. W. and Lee, Y. J., " Coefficient of consolidation from linear 

segment of the t curve " , Can. Geotech. J., 2001, 38 : 901-909. 

[12] Mesri, G., Feng, T. W., and Shahien, M., “Coefficient of consolidation by 

inflection point method”, J. Geotech. And Geoenv. Eng., 1999, Vol. 125, 

No. 8, pp. 717-718. 


I would like to thank and acknowledge the financial support of the 

Research Center in the College of Engineering, Grant No. 42/426. In addition, I 

would like to thank Eng. Jalal Mahjoub and Eng. Abduldattar Alget for their 

assistances in gathering the samples and laboratory testing. 



KHALEEJ MARDOMAH RESEARCH STATION – AN OVERVIEW OF THE 

FIVE-YEAR DURABILITY PERFORMANCE RESULTS 

Mesfer M. Al-Zahran 1 

, Mohammad Shameem 2 

, Mohammad Ibrahim 2 

, 

Mohammad Rizwan 2 

, Mohammad Barry 2 

1: King Fahd University of Petroleum and 

Minerals, Civil Engineering Department, 

Dhahran 31261, Saudi Arabia, 

mesferma@kfupm.edu.sa 

2: King Fahd University of Petroleum and 

Minerals, Research Institute, Dhahran 31261, 

Saudi Arabia 

ABSTRACT 

In the severe environmental exposure conditions that prevail in the Arabian 

Gulf region, durability is still the most important factor for the service life of most 

concrete structures, both in terms of economy and safety. Of the various types of 

deterioration mechanisms, reinforcement corrosion is at present considered to be the 

main cause of premature failure of reinforced concrete structures in most countries. 

Field exposure station has been established at Khaleej Mardomah in Jubail Industrial 

City to gain practical knowledge on how to produce and protect concrete structures 

under natural and long term exposure conditions. This field station is the first of its 

kind in the region. Exposure conditions include aggressive ground-water, marine 

environment (submerged, splash zone and zone above sea water), and outdoor 

exposure in air. The research program involves casting of different concrete 

specimens using different design and protection parameters, including cement type, 

water/cement ratio, pozzolanic admixtures, reinforcement type, water type, curing 

condition, corrosion inhibitors, and concrete coatings. 

The paper will present part of the recently obtained results after five years 

of exposure in the field station and highlight issues concerning durability of concrete 

structures exposed to various exposure conditions. 

KEY WORDS 

concrete, field exposure, corrosion, marine environment, pozzolanic 

admixtures. 


KHALEEJ MARDOMAH RESEARCH STATION – AN OVERVIEW 

INTRODUCTION 

The production and placement of high-quality durable concrete has always 

been a major concern for the construction industry in the Arabian Gulf region. The 

primary reasons for this concern are related to the extensive use of concrete as a 

major construction material and the severity of the environment in this region, which 

causes concrete deterioration to occur in a relatively short period of time unless 

proper corrosion protection is provided. Of the various types of deteriorating 

mechanisms, chloride induced reinforcement corrosion is at present considered the 

main cause of premature deterioration in most countries. Other common 

deteriorating mechanisms are chemical attack and carbonation. 

The most reliable way of assessing the durability of a concrete structure is 

to study the performance under natural conditions. By varying the main parameters 

such as production parameters, materials quality, reinforcement type and 

environment, more valuable information can be obtained from field investigations, 

within the order of a decade than in accelerated laboratory tests. Field exposure 

stations have been used all over the world to develop durable concrete structures to 

withstand various severe conditions for at least a century. 

The purpose of establishing a field research station is to gain practical 

knowledge on how to develop durable concrete and protect concrete structures 

against deterioration under various natural conditions. Exposure conditions may 

include aggressive ground water, marine environment (submerged, splash zone, and 

zone above sea water), and outdoor exposure in air. Therefore, a field site was 

selected in coordination with Royal Commission for Jubail and Yanbu (RCJY) to 

establish a major research field station. The site is located in Khaleej Mardomah at 

Jubail Industrial City. The preparation of the field site started in 2000 which 

included several activities, such as site grading, access road paving, and setting up of 

two portable offices. The research field site was prepared as a long-term exposure 

station. The exposure of the prepared specimens started in 2001. Detailed 

information about the preparation and establishment of the research field site was 

presented in the final report of the Corrosion Research Contract No. 560-T08 Project 

No. CER 2209. 

This paper will share part of the recently obtained results after five years of 

exposure in the field station and highlight issues concerning durability of concrete 

structures under various exposure conditions. Continuing the research on the 

existing exposure station is an efficient way to gain reliable documentation of long 

time durability data of reinforced concrete structures under local environment. 



OBJECTIVES 


Mesfer M. Al-Zahran, et al 

Kaheleej Mardumah research station was established to gain practical 

knowledge on how to develop durable concrete and protect concrete structures 

against deterioration under various natural conditions. Exposure conditions included 

aggressive ground water, marine environment (tidal and splash zones), and outdoor 

exposure in air. 

The specific objectives of establishing such a station are as following: 

1. Documenting the long term performance of properly produced and placed 

concrete in the local marine environment represented by Madinat Al-Jubail Al- 

Sinaiyah. 

2. Identifying variables that are important to control early deterioration in 

concrete. 

3. Establishing the degree of deterioration that may be expected from good quality 

concrete. 

The results generated from investigation in such a field research station will be 

very important, both for practical recommendations concerning durability under 

various exposure conditions, and for the benefit of future research. Continuing the 

research at the existing exposure station is an efficient way to gain reliable 

information on long term durability of reinforced concrete structures under our local 

environment. The results can also be shared with other field stations at the regional, 

as well as international, levels. 

SCOPE OF WORK 

This field station is the first of its kind in the region. The principal work 

under this project involved field and laboratory work with the objective to determine 

the long term durability performance of properly mixed and placed concrete in the 

local aggressive environment of Madinat Al-Jubail Al-Sinaiyah, identify variables 

that are important to control early deterioration in concrete, and establish the degree 

of deterioration that may be expected in good quality concrete. The reinforced 

concrete specimens were prepared using high quality concrete and considering 

different variables, including corrosion inhibitors, type of reinforcement, pozzolanic 

additives, water/cement ratio, water type, curing conditions, and concrete coatings. 

The research field station is divided into four zones representing the most 

aggressive and harsh conditions that prevail in the Arabian Gulf region and attack 

concrete structures. The field exposure zones are tidal zone (zone 1), splash zone 

(zone 2) partially buried and below ground zone (zone 3), and above ground zone 

(zone 4). 

405


SPECIMEN PREPARATION 

High-quality concrete specimens were cast as per RCJY specification [03347, 

2000]. For reference purposes, unreinforced concrete specimens were cast 

representing the main concrete mixes and cured in water under laboratory condition 

for up to 90 days. 

Variables and Materials 

The prepared concrete mixes consist of twenty two different mixes 

covering a wide range of materials and variables as summarized in Table 1. The 

control concrete mix (M1) used is represented by mix J 25 b according to RCJY 

specifications 033347 as described in Table 1. The following sections describe 

some of the materials and variables used in this study as follow: 

Cement type: Cement types I and V as per ASTM C 150 was used. 

Cement content: Two cement contents were considered in the plain cement 

concrete mixes, namely 370 kg/m 3 and 450kg/m 3 . 

Mixing water type: Two types of water were used namely potable water and 

reclaimed water. 

W/C ratio: Two effective water to cementitious materials ratios were used 

namely 0.40 and 0.30. 

Reinforcement type: 

1. Regular deformed steel reinforcement conforming to ASTM A 515M. 

2. Rusted deformed steel reinforcement. 

3. Fusion bonded epoxy coated (FBEC) reinforcement conforming to 

ASTM A 775. 

4. Hot rolled stainless steel (SS) reinforcement from the austenitic group 

with 18 % chromium, 10 % nickel and 2 % molybdenum (grade 

1.4401/AISI 316) was used to reinforce concrete columns exposed to the 

splash zone. 

5. Glass Fiber Reinforced Plastic (GFRP) reinforcement bars known as C- 

Bar. 

6. GFRP reinforcement bars known as Aslan 100. 

Pozzolanic admixtures: 

1. Silica fume (SF). 

2. Fly ash (FA). 

3. Highly fine and reactive alumino-silicate fly ash (HFA). 

4. Ground Granulated Blast-furnace Slag (GGBS). 

Additive: Polypropylene (PP) fibrillated fibers (The length of these fibers 

was 19 mm, and they were added to the concrete mix at a dosage of 0.2% by volume 

of concrete). 





CuringMethods: 

1. Wet burlap using potable water. 

2. Wet burlap using reclaimed water. 

3. Membrane curing compound. 

Concrete coating: Coal tar epoxy coating was applied on clean and dry 

concrete surfaces after proper curing. 

Corrosion inhibitors 

1. A green liquid concrete admixture (C1) mixed with water and added to 

the concrete mixer. 

2. A blend of organic and inorganic corrosion inhibitors. A transparent 

emulsion type impregnation liquid for concrete (C2). Three coats were 

applied on concrete surface using a brush. 

Table 1. Summary of the concrete variables*. 

Mix No. 

Control 

M1 

Cement 

Content 

kg 

Water 

Type 

370 Potable 

Reinforcement 

Steel 

Additives 

Curing 

Coating 

Corrosion 

Inhibitors 

Exposure zone No. 

- Normal - - 1, 2, 3, 4 

M2 370 Potable 

Steel 

- - - - 1, 3, 4 

M3 450 Potable Steel - - - - 1, 3, 4 

M4 405 Potable Steel 

36 kg 

SF 

- - - 1, 3, 4 

M5 370 Reclaimed Steel - 

0.2 % 

- - - 1, 3, 4 


Steel 

PP 

Fibers 

Normal - - 1, 3, 4 


30 kg 

SF 

Normal - - 1, 2, 3, 4 


Steel 

70 kg 

FA 

Normal 

Curing 

- - 1, 2, 3, 4 


Steel 

- Compoun 

d 

- - 1, 3, 4 

M10 370 Potable Steel - Normal 

Coal tar 

epoxy 

- 1, 3, 4 


Steel 

- Normal - C1 1, 3, 4 


Steel 

- Normal - C2 1, 3, 4 

M13 370 Potable FBEC - Normal - - 1, 3 

407







GFRP 

C-Bar 

Steel 

Rusted 

steel 

Bars 

Steel 

GFRP 

(Aslan) 



- Normal - - 1, 3 

37 kg 

HFA 

Normal - - 1, 3, 4 

- Normal - - 1, 3 

105 kg 

FA 

M19 370 Potable Steel - 



Stainle 

ss steel 

M22 - 360 otable Steel 

Normal - - 1, 3, 4 

- Normal - - 1, 3 

266 kg 

GGBS 

Reclaimed 

Water 

- - 1, 3, 4 

Normal - - 1, 2, 3, 4 

- Normal - - 2 

90 

HFA 

Normal - - 1 

* Cement type I for all mixes except M2 where type V was used. 

W/C ratio of 0.4 was used for all mixes except M4 where 0.30 was used. 

- M22 is Self Compacting concrete: started field exposure in August 2004. 

Types of Specimens 

Reinforced concrete specimens 

Total of seventy columns and sixty beams were prepared and placed at the 

exposure site for field monitoring as described below: 

1. Columns with 35x35x300 cm dimensions were installed in the partially 

buried zone (zone 3), and splash zone (zone 2). The reinforced columns are 

utilized for visual observation and corrosion monitoring since the 

commencement of field exposure. 

2. Beams with 25x30x75 cm dimensions were installed in the tidal zone (zone 

1). The reinforced beams are used for visual observation and corrosion 

monitoring since the commencement of field exposure. 

Unreinforced concrete specimens 

1. A total of more than 1150 cubes with 15x15x15 cm dimensions were placed 

at the tidal zone (zone 1), below ground zone (zone 3), and above ground 

zone (zone 4), and for laboratory conditioning. The cubic specimens are used 

for conducting the water permeability test. 

2. A total of more than 2600 cylinders measuring 7.5 cm in diameter and 15 cm 

in height were placed at the tidal zone (zone 1), below ground zone (zone 3), 

and above ground zone (zone 4), and for laboratory conditioning. The 

cylindrical specimens are used to conduct chloride permeability, electrical 

resistivity, water absorption, compressive strength, and sulfate resistance 

tests.

Placement of the specimens in the field exposure 



After proper curing, the specimens were kept in the casing ground to reach 

a minimum age of 28 days prior transportation and placement in the field exposure 

zones. By May 2001 all specimens were placed in their particular zones of the field 

station. 

MONITORING AND TESTING 

Field monitoring 

Visual observation in field 

All the exposed specimens have been monitored and observed for any 

noticeable signs of deterioration every six months throughout the duration of 

the study. 

Corrosion monitoring in field 

The corrosion potential of the steel reinforced columns and beams in zones 

1, 2 and 3 is monitored according to [ASTM C 876, 1996], and readings were taken 

every six months. For this purpose, thirty two points were identified for corrosion 

potentials on two faces of each beam specimen in zone 1. Forty eight and forty 

points were identified for corrosion potentials on two faces of each column 

specimen in zones 2 and 3, respectively. 

Testing of Field Specimens in Laboratory 

Cylindrical and cubic concrete specimens were retrieved from the fieldcured 

specimens after 28 days of casting (unexposed), and after 1, 2 and 4 years of 

exposure in their respective exposure zones. Similarly, concrete core samples were 

retrieved from the field exposed columns and beams after 1, 2, and 5 years of 

exposure to monitor changes in the physical and mechanical properties of the 

concrete mixes. 

The following tests are being conducted in the laboratory on these specimens: 

1. Chloride, sulfate, and alkalinity (OH¯) profiles by analyzing core samples 

collected at several depths from the columns and beams [Vogel, 1985]. 

2. Water permeability according to the standard test, [DIN 1048, 1990]. 

3. Chloride permeability according to [AASHTO T277, 1999]. 

4. Electrical resistivity using an SGABEM Terrameter SAS 330 C precision 

digital electrical resistance meter. 

5. Water absorption according to [ASTM C 642, 1996]. 

409


6. Compressive strength according to [ASTM C 39, 1997]. 

7. Carbonation depth by spraying phenolphthalein indicator on the freshly 

cored samples from the columns and beams. 

PRELIMINARY RESULTS 

Visual Observation in Field 

After five years, the visual observation of all the field-exposed specimens 

did not show any major signs of deterioration were detected under all exposure 

conditions. In the tidal zone, deposition of fungus was noticed in almost all the 

specimens due to which the color of concrete changed to greenish black to black in 

some cases as shown in Figure 1. Minor to major deposits of shells were noticed on 

the specimens with high density of shells found on the side facing the sea. Shell 

deposits were not noted on the top face of the beam specimens. 

Figure 1. A typical beam specimen placed in tidal zone. 





Carbonation 

After five years of exposure, carbonation in the tidal zone is reported 

only in mixes M8 (8% fly ash), M9 (curing compound), M13 (standard mix) 

and M17 (30% fly ash) with depths ranging from 1.5 to 2.5 mm which is very 

marginal. In the below ground exposure, negligible carbonation depth is 

reported in mixes M1 (Type I cement), M8 (8% fly ash), M9 (curing 

compound) and M13 and M16 (standard mix). For the upper portions of the 

columns exposed to atmosphere, the carbonation was noticed in all the 

concrete mixes except mixes AM4 (High performance concrete with low w/c 

ratio), AM7 (8% Silica Fume), AM10 (coal tar epoxy) and AM15 (10% super 

pozz). The other mixes showed carbonation depths ranging from less than 1 

mm to as high as 10 mm. The carbonation depth values noticed in the above 

ground portions of the columns were higher than those observed in the tidal 

and below ground zones due to the continuous and direct exposure to the 

atmosphere. 

Chloride Concentration at the Rebar Level 

The chloride concentrations at the rebar level (i.e. 75 mm) in most of 

the specimens exposed for five years in the tidal zone as shown in Figure 2. It 

was observed that fifteen of the concrete mixtures crossed the threshold 

chloride value (0.025% by weight of concrete established by ACI 318 for 

concrete prior to service exposure to prevent corrosion initiation of reinforcing 

steel), indicating the harsh exposure condition. The data indicate the superior 

performance of blended cements. Concretes prepared with silica fume and fly 

ash (Mixes M7 and M8, respectively) exhibited better performance as 

compared with all the other mixtures. In the case of the buried portion of the 

columns, the average chloride concentration at the rebar level in eleven of the 

twenty mixes exceeded the threshold chloride value of 0.025% by concrete 

weight (i.e. 0.15% by cement weight). In general the results indicate that the 

above ground exposure is less aggressive as compared with the below ground 

exposure. 

411


Chloride - % by wt. of Concrete 

0.4 

0.35 

0.3 

0.25 

0.2 

0.15 

0.1 

0.05 

0 

TM1 

Tidal - 5 Years 

0-5 10-15 25-30 72-77 95-100 

TM2 

TM3 

TM4 

TM5 

TM6 

TM7 

TM8 

TM9 

TM10 

TM11 

TM12 

TM13 

TM14 

TM15 

TM16 

TM17 

TM18 

TM19 

TM20 

Mix No. 

Figure 2. Chloride concentration profiles in beams after five year exposure in 

the tidal zone (TM19 and TM20 after 3 year exposure). 

Corrosion Performance 

The assessment of corrosion tendency of the reinforcing steel bars in the 

beams and columns was determined through their corrosion potentials measured 

against Cu/CuSO4 reference electrode (CSE). The corrosion potentials were 

measured for about 60 months, from May 2001 to May 2006. According to 

recommended criteria in ASTM C876 [5], corrosion potentials more positive than - 

200 mV CSE, indicate 90% probability of no corrosion state on steel bars in 

concrete with 90% probability. Potentials between -200 and -350 mV CSE indicate 

uncertain corrosion of the steel bar. Corrosion potentials more negative than -350 

mV CSE indicate 90% probability of active corrosion. 

SPECIMENS AT TIDAL ZONE 

All the beams, except in specimens M9 (curing compound), M12 (SIKA 

903), and M16 (rusted bars), the average corrosion potentials remained numerically 

below -350 mV CSE. The corrosion potentials of all the other beams with Type-I 

cement, 20%FA, 30% FA cement replacement concretes, with high cement content, 

low w/c ratio, SF, Super-Pozz, GGBS, SIKA 901, 0.2% PP fibers, reclaimed water, 

and FBEC reinforcement showed corrosion potential values within the range of 

uncertain corrosion state. 



SPECIMENS AT SPLASH ZONE 



The corrosion potentials of the columns with control concrete and SF 

indicated active corrosion in lower portion and uncertain corrosion state in upper 

portion. The addition of the FA did show a significant effect on the potential 

corrosion. The upper and lower portions of the FA columns showed no active 

corrosion state with much lower corrosion potential values as compared to control 

mix. Similarly, the columns with GGBS indicated state of no active corrosion. 

According to the results, GGBS addition showed beneficial effect on corrosion 

potentials with time in both portions of the column under the splash zone conditions. 

The corrosion potential values of the stainless steel (SS) reinforcement were in 

uncertain state of corrosion in the entire column throughout the monitoring period. 

SPECIMENS AT PARTIALLY BURIED ZONE 

The corrosion potential values in all the columns, except for the columns 

with PP fibers, curing compound, SIKA 901, SIKA 903, pre-rusted steel 

reinforcement, and curing with reclaimed water, remained numerically below -350 

mV CSE during the reported exposure period in both the upper and lower portions, 

indicating the columns are not yet in active corrosion state. In all the columns, the 

potentials measured on the above ground portion exhibited less negative behavior 

compared to the potentials measured on the interface and below ground portion. 

The corrosion potentials in the interface and below ground portion of the 

columns with control mix, high Type I cement content, reclaimed water for mixing, 

and coal tar epoxy coating remained within the uncertain corrosion range, whereas 

the corresponding potentials of the above ground portion maintained its state of no 

active corrosion, as expected. 

In the columns with low w/(c+sf) ratio, silica fume (SF) cements, Type V 

cement, FA, Super-Pozz, GGBS, and FBEC reinforcement, the corrosion potentials 

both in the interface and below ground and above ground portions indicated state of 

no active corrosion as shown in Figure 3. The corrosion potentials in the columns 

cured with reclaimed water and the columns with PP fibers, and pre-rusted 

reinforcement indicated state of high probability of corrosion in the interface and 

below ground portion of the column. 

413


Corrosion Potential - mV 

-600 

-550 

-500 

-450 

-400 

-350 

-300 

-250 

-200 

-150 

-100 

-50 

0 

High probability of corrosion 



BM1 

BM2 

BBM3 

BM4 

BM5 

BM6 

BM7 

BM8 

BM9 

BM10 

BM11 

BM12 

BM13 

BM15 

BM16 

BM17 

Mix No. 

Figure 3. Average variation of corrosion potential on steel in below ground 

portion of partially buried columns after 5-years of exposure. 

CONCLUDING REMARKS 

Based on the obtained results of all properties considered in the study of the 

sixteen main concrete mixes after five years of field exposure (including mixes M19 

and M20 which were exposed for three years and excluding mix M21), the overall 

ranking showed that the mixes with pozzolanic admixtures including the mix coated 

with Coal Tar epoxy have shown the best performance followed by the high cement 

content and control mix. The following are all the concrete mixes listed in 

descending order based on the overall field performance: 

1. M4 : 8 % SF; w/c=0.3 

2. M7 : 8 % SF 

3. M15 : 10 % Super-Pozz 

4. M8 : 20 % FA 

5. M10 : Coal Tar Epoxy 

6. M20 : 70% GGBS 

7. M17 : 30 % FA 

8. M3 : 450 kg/m 3 (Type I Cement) 

9. M1 : Control (Type I Cement) 

10. M11 : Mixing Corrosion Inhibitor (SIKA 901) 

11. M2 : Type V Cement 

12. M5 : Mixing water: reclaimed

13. M19 : Curing: reclaimed water 

14. M9 : Curing Compound 

15. M6 : 0.2 % PP fibers 

16. M12 : Applied Corrosion Inhibitor (SIKA 903) 




The authors acknowledge the support provided by King Fahd University of 

Petroleum and Minerals and Royal Commission for Jubail and Yanbu, Jubail, Saudi 

Arabia, and effort exerted by all researchers, engineers and technicians in the 

Research Institute of King Fahd University of Petroleum and Minerals 

REFERENCES 

AASHTO T277, 1999, Standard test method for rapid determination of the chloride 

permeability of concrete. 

ASTM C39, 1997, Standard test method for compressive strength measurement in 

concrete, Annual Book of ASTM Standards, v. 4.05, American Society for 

Testing and Materials, Philadelphia. 

ASTM C642, 1996, Standard test method for specific gravity, absorption, and voids 

in hardened concrete, Annual Book of ASTM Standards, v. 4.02, American 

Society for Testing and Materials, Philadelphia. 

ASTM C 876, 1996, Standard test method for half-cell potentials of uncoated 

reinforcing steel in concrete, Annual Book of ASTM Standards, v. 4.02, 

American Society for Testing and Materials, Philadelphia. 

DIN 1048, 1990 Concrete according to German standards. Ed. Berrihard Dartsch, 

Beton – Vertag Gmbh. 

Guideline Specifications for Portland Cement Concrete Section 03347, 2000, Royal 

Commission for Jubail and Yanbu, Director General for Jubail Project, Jubail 

Industrial City, Saudi Arabia. 

Research Institute (2003). Long-term Durability Investigation of Concrete in the 

Arabian Gulf Environment. Research Institute, King Fahd University of 

Petroleum and Minerals, Report Project No. CER 2209, p. 303. 

Vogel, I, A., 1985. Quantitative Inorganic Analysis, 5th Edition, Longman, London, 

pp. 754-755. 

415





CAURSAT FUNCTIONS FOR A PROBLEM OF AN ISOTROPIC PLATE WITH A CURVILINEAR HOLE 




S.A.ASEEN 

419





S.A.ASEEN 

421





S.A.ASEEN 

423





S.A.ASEEN 

425





S.A.ASEEN 

427





S.A.ASEEN 

429





S.A.ASEEN 

431




Topic 7 

Research and development to serve the 

industry and upgrade its services 

• General Engineering

EFFECT OF VIRTUAL REALITY ON BREAST CANCER PATIENTS 

(REVIEW ARTICLE) 

Rasha A. Al-meleak 1 

, Wadee S. Al-halabi 

1: Umm Al-Qura University, Computer Scince, , Department of Computer Science 

rose1405_2@hotmail.com 

2: University of Miami , Department of Electrical and Computer Engineering 

7820 Camino Real Apt J320, Miami, FL 33143-6877, USA, 

w.alhalabi@umiami.edu 

ABSTRACT 

Most of women with breast cancer are treated with one or more of the 

standard therapies such as chemotherapy, radiation therapy or surgery. Women 

receiving these therapies often suffer from side effects such as nausea, vomiting, 

severe pain, stress and anxiety. 

This article discusses the effect of virtual reality (VR) as one of the guided 

imagery techniques which distracts patients from stress and discomfort often 

associated with chemotherapy. The paper shows how side effects such as nausea, 

vomiting, stress and severe pain are alleviated. 

KEY WORDS 

breast cancer, guided imagery, virtual reality , chemotherapy. 

INTRODUCTION 

Breast cancer is a common malignancy disease occurs mostly in women 

among aged 30 – 50 [1]. Treatment of breast cancer often requires several different 

treatment modalities including surgery, radiation therapy, chemotherapy and 

hormone therapy. 

Chemotherapy treatment may cause severe side effects such as nausea, 

anxiety and fatigue. 

The purpose of this study is to explore the use of VR as a distraction 

intervention to reduce fatigue and anxiety in women receiving chemotherapy for 

breast cancer treatment. 

Proceedings of the 7 th Saudi Engineering Conference, KSU, Riyadh, 26-28 Nov, 2007

EFFECT OF VIRTUAL REALITY ON BREAST CANCER PATIENTS (REVIEW ARTICLE) 

This paper addresses the following problems: 

1- The impact of VR on breast cancer patients. Effectiveness of VR to alleviate 

pain associated with the disease for patients under chemotherapy in the first 

year of thetreatment. 

2- Can VR improve the lifestyle of cancer patients. 

VIRTUAL REALITY THERAPY 

Virtual Reality Therapy is one of the emerging and most effective 

applications of VR technology, where patients are exposed to stimuli in fully 

controllable environments. 

VR therapy has been successfully implemented in various diseases, such as 

cancer (Figure 1,2),dentist (Figure 3) and phobia (Figure 4). Patients become 

immersed within a virtual environment, because he or she will have fewer resources 

to focus on during the clinical examination or surgical operation. The idea is to 

recreate a believable artificial environment that stimulates physical responses similar 

to those of a real environment which can be individually controlled, replicated, and 

tailored to the patient’s experiences. In some applications, patients see themselves 

engaging in various activities with virtual people. In virtual reality, a user interacts 

with a computer-generated three-dimensional world through a head-mounted 

display, stereo-earphones, a position sensor, a controller such as a joystick or speed 

wheel and output devices. 

Figure 1: Chemotherapy patient and virtual reality user 


Saudi Engineering Conference, KSU, Riyadh, 26-28 Nov, 2007

Rasha A. Al-meleak , Wadee S. Al-halabi 

Figure 2: 9-year-old leukemia patient William Bugbee navigates the virtual 

reality environment. 

Figure 3: Virtual Reality at the Dentist 

Proceedings of the 7 th Saudi Engineering Conference, KSU, Riyadh, 26-28 Nov, 2007 

437


In the application in Figure 3, a child is looking through the goggles and 

manipulating the scenes that he sees with a game controller. In Figure 4, the UW 

researcher Hunter Hoffman leads a participant through Spider World. A virtual 

reality program that addresses spider phobia. 

Figure 4: Spider phobia 

The main tool to alleviate pain is to use palliative drugs, which often are not 

enough to do so. Moreover, in most cases various this type of drug causes sideeffects. 

The virtual reality technique is used to distract concentration on the disease. 

This method was proven to be effective and this was clearly addressed in earlier 

literature. Another advantage of using VR eliminates or reduces side-effects of 

chemotherapy such as vomiting, nausea and fatigue. The reduction in pain is 

measured using pain assessment plan (1 to 10), Figure (5). 

PERFORMANCE CRETERIA 

Measuring level of pain for patients after each VR session which is called a 

"Measure of Assessing Pain", this is a global measure of ten degrees which 

describes the pain intensity, start at 0 for ( no pain ) until 10 for (pain in the worst 

case) refer to Figure 5. 



BACKGROUND 

Figure 5: Pain Assessment Scale 


David and Alex [1] presented a computer vision and virtual reality 

application which provides the tools for cancer patients to allow for direct 

visualization and enables the user to easily control the environment at a time when 

he or she is losing control of their lifes. The author reported that visualization and 

imagery boost the immune system to defend the body. David and Alex system which 

was dubbed "Staying Alive" was able to provide some relaxation to patients which 

is an advantage in favor of the immune system. Stress has been reported by 

scientists to retard the immune system [7]. Feldman and Salzberg [9] shows that 

imagery reduces the stress and anxiety caused by cancer treatment. Similar paper 

was presented by Decker and others [8] which supports Feldman and Salzberg in 

their conclusion. In his paper [5], Siegal reported a great growing in the believe 

among people that visualization and self-imagery technique influence the mind to 


439


encourage the body to make it realty and truly defend the body from disease. 

Another fact was reported by Little in his work [6] that an increase of 56% in the 

survival rate for cancer patient was achieved. "Staying alive" presents a scenario in 

which patients can fight malignant cells and clear the environment while they are in 

the blood stream. 

Schneider and others [10] reported that when women use virtual reality 

intervention, they have significant decreases in symptom distress, fatigue and 

anxiety. The authors conclude an advantage of ease implementation in the clinical 

setting. In their work [11], Schneider and Hood randomly selected their patients to 

receive the VR destruction intervention in one of the chemotherapy treatment. 82% 

of the patients would like to use the VR system again. Patients feel that the 

treatments were shorter which indicates a tolerable treatment. Schneider and others 

[12] studies the VR intervention for older women with breast cancer. They study 

selected its participants randomly to receive the VR distraction intervention during a 

chemotherapy treatment. The study indicates that 100% of the participants were 

satisfied and would use the VR again. Participants experienced no cybersickness. In 

another study Schneider and Workman [13] selected older children to test the VR as 

a distraction intervention in chemotherapy. The main reason in attempting this 

experiment was to describe the perceived effectiveness and feasibility of using VR 

as a distraction intervention for older children in the age of 10-17, receiving 

chemotherapy. The study conclude that 82% of the children pointed out that 

chemotherapy treatment with VR was better than previous chemotherapy treatment. 

100% of the participants would like to use the VR again. The ease use of the system 

which does not require previous experience or skill in any kind encourages patients 

to participate the experiments. The study indicates that VR has the potential to 

enhance positive clinical outcome. 

Durham,n.c [2] described how chemotherapy patients eased their fatigue and 

discomfort by solving a mystery and touring an art gallery. The goal of the study 

was to reduce anxiety, fatigue and distress symptoms. In the study, women were 

asked to complete surveys about their symptoms including distress, anxiety and 

fatigue level. The results showed that women who used virtual reality during 

chemotherapy treatments reported significant decreases in symptom distress and 

fatigue level immediately following the treatments. 

Moore R, Spiegel D in their study [3] explains that guided imagery therapy is 

clearly not a cure for breast cancer, however it helps patients to feel as if they have 

some sense of control over the progression of the disease in a situation that feels out 

of control. It relieves pain as well. When VR is used with conventional medical 

treatment, imagery therapy provides a noninvasive, inexpensive and necessary 

means of repairing this ruptured relationship between the self, the body and the 

woman's social world by improving her lifestyle. 




Susan and others [4] examined the effects of a virtual reality distraction 

intervention on chemotherapy related distress levels to answer the following two 

questions :(1) Is virtual reality an effective distraction intervention for reducing 

chemotherapy related distress levels in older women with breast cancer? (2) Does 

virtual reality have a lasting effect? 

Sixteen women aged 50 and older received intravenous chemotherapy as a 

part of their treatment plan. Results of the tests showed that when participants used 

virtual reality, there is a significant decrease in the State Anxiety Inventory (SAI) 

scores immediately following chemotherapy treatment. This means that virtual 

reality distraction intervention can be used with older women since it reduced 

anxiety in the population. The study reported that 94% of women could use headset 

with no difficulty. 

THE IMAGERY WAY 

In our virtual reality project at Umm Alqura University (UQU), we have 

developed a virtual reality software to enhance the immune system and elevate pain. 

The system was dubbed "Imagery Way". It can be used for children undergoing 

chemotherapy treatment as well as women suffering from breast cancer. Imagery 

Way immerses patients in the environment and expected to attract children for the 

chemotherapy treatment, because it changed the treatment session into an adventure. 

Figure (6, 7 and 8) give screenshot for the Imagery Way. 

Figure 6: The player is trying to kill a cancer cell 


441


Figure 7: Cancer cells in all five stages 

Figure 8: The player is collecting points 



CONCLUSION 


Virtual reality(VR) helps patients cope with chemotherapy, it's defined as a 

way to immersed mind and body into a computer-generated environment in a 

naturalistic fashion. The distraction of a virtual world significantly reduces the 

adverse effects of chemotherapy treatment for breast cancer patients. Our experience 

with Quest3d program to design a virtual world demonstrates the power of VR and 

its world in breast cancer patients. This project is under construction meanwhile, 

images of relaxation techniques and deep imagination are required for this 

environment. 

REFERENCES 

[1] Davide A.Becker & Alex Pentland, April 1997, ”Using A virtual Enviroment to 

Teach Cancer Patients T'ai Chi, Relaxation and Self-Imagery”, MIT Media 

Laboratory. 

[2] DURHAM,N.C, ”Virtual Reality Helps Breast Cancer Patients Cope with 

Chemotherapy”, ,Jan 29,2004, Duke University School of Nursing and Case West 

Reserve Comprehensive Cancer Center. 

[3] Moore R, Spiegel D, 1999, "Uses of Guided Imagery for Pain Control by African- 

American and White Women with Metastatic Breast Cancer", Integrative Medicine 

Consult. 1999; 2(2/3): 115-126. 

[4] Susan M. Schneider, PhD, RN, AOCN®, Maryjo Prince-Paul, MSN, RN, CRNH, 

Mary Jo Allen, BSN, RN, Paula Silverman, MD, and Deborah Talaba, MSN, RN, 

”Virtual Reality as a Distraction Intervention for Women Receiving 

Chemotherapy”. 

[5] Bernie Siegal. Love, Medicine, and Miracles: lessons learned about self-healing from 

a surgeon's experience with exceptional patients. Harper and Row, 1986. 

[6] Bill Little. Eight Ways to Take an Active Role in Your Health. Harold Shaw 

Publishing, 1994. 

[7] S. E. Sims. Relaxation training as a technique for helping patients cope with the 

experience of cancer: a selective review of literature. Journal of Advanced Nursing, 

Sep 1987. 


443


[8] T. W. Decker, J. Cline-Elsen, and M. Gallagher. Relaxation therapy as an adjunct in 

radiation ontology. Journal of Clinical Psychology, May 1992. 

[9] H.C. Salzberg C. S. Feldman. The role of imagery in the hypnotic treatment of 

adverse reactions to cancer therapy. J. S. C. Medical Association, May 1990. 

[10] Schneider SM, Prince-Paul M, Allen MJ, Silverman P, Talaba D, " Virtual reality 

as a distraction intervention for women receiving chemotherapy." Oncology 

Nursing Forum – 2004 Feb – Vol. 31, No 1,; 81-8. 

[11] Schneider, S. M. & Hood, L. E, "Virtual reality: a distraction intervention for 

chemotherapy." Oncology Nursing Forum, v2007 Feb -Vol. 34, no. 1, pp. 39-46. 

[12] Susan M. Schneider, Mathew Ellis, William T. Coombs, Erin L. Shonkwiler, Linda 

C. Folsom, " Virtual Reality Intervention for Older Women with Breast Cancer" 

Mary Ann Liebert, Inc Publisher, 03, 6(3): 301-307 

[13] Susan M. Schneider, M.L. Workman, "Virtual Reality as a Distraction Intervention 

for Older Children Receiving Chemotherapy." Pediatric Nursing, Nov-Dec, 2000: 

593-597 



IMPACT OF VIRTUAL REALITY AND COMPUTER GAME ON 

IMAGERY THERAPY FOR CANCER PATIENTS 

Sahar Ali Aseeri 1 , Wadee S. Al-halabi 2 

1: Umm Al-Qura University, Department of Computer science, Makkah, Saudi 

Arabia, P.O.BOX 6425, saharaseeri@yahoo.com 

2: University of Miami, Department of Electrical and Computer Engineering, 

7820 Camino Real Apt J320, Miami, FL 33143-6877, USA, 

w.alhalabi@umiami.edu 

ABSTRACT 

Treatment of cancer disease depends heavily on patient’s psychology as 

well as clinical therapy. Cancer's patients usually experience some common 

symptoms when they are exposed to chemotherapy treatments as a part of its side 

effects. These symptoms could be depression, suspense, vomiting, nausea and many 

others. Earlier literature reported that patients can control these symptoms if he or 

she believes so. Imagery therapy is a well-known technique used to enhance 

patients' believe and boost his or her self-esteem and confidence. This result in 

tremendous improvement in patient’s mind and inner feeling. Which clinically 

improves the immune system. Imagery therapy is considered as a complementary 

psychological intervention used in conjunction with conventional therapy. This kind 

of therapy creates a virtual world for cancer patients to cope with their diseases and 

experience better life. Virtual Reality is a professional laboratory proven method 

used as a tool to enable physicians and psychiatrists to lead their patients through 

huge obstacles of depression during the treatment. Many VR environments and 

computer games were developed and tested to enable health specialist help cancer 

patients in the treatment. 

KEYWORDS 

Psychological Therapy, Guided Imagery, Virtual Reality, Computer 

Games. 

Proceedings of the 7 th Saudi Engineering Conference, KSU, Riyadh, 26-28 Nov, 2007

IMPACT OF VIRTUAL REALITY AND COMPUTER GAME ON IMAGERY THERAPY FOR CANCER PATIENTS 

INTRODUCTION 

Cancer's patients experienced depression, vomiting, nausea and suspense as 

a side effects for being exposed to chemotherapy treatments. Consequently, 

physicians refer them to psychiatrists for an assistance to create a vivid optimistic 

mental picture. This picture improves their inner feeling and transfers patients to an 

enjoyable active lifestyle. We can split cancer treatments into two major categories: 

conventional therapies and complementary therapies. The latter consists of different 

therapeutic techniques one of them is imagery therapy. Imagination is the language 

used in human mind and therapists believe that this healing instrument develop the 

ability of imagination which have a grate influence on the disease. That is way 

imagination is used in imagery therapy. 

Many of cancer patients try to intercept with their inner feeling and inner 

body. By different means and measures. This interception could be achieved through 

a positive mental image. This method effectively control pain and other symptoms. 

An example of that is the imagination of a huge army of white blood cells fighting 

cancer cells and beating them [2]. Many studies in this field confirm that imagery 

therapy improves patient’s psychology and inner feeling, consequently boost the 

overall patient health. In his study [1] Oyama reported that patients attempted to use 

imagery therapy as a supplementary treatment beside cancer conventional therapies 

to support their immune system. The author reported that VR technology is used to 

help patients accept their disease. The mental state of cancer's patients changes step 

by step from denial of cancer to hope for a new cancer's treatment. Sometimes 

patients retreat to suspicion of medical treatment and uneasiness regarding their 

future life which result in irritation, depression then changed to acceptance and 

optimization. In our work at Umm Alqura University (UQU), we attempt to develop 

a new type of counseling system in medical cyberspace to provide mental care. The 

system which we are developing focuses on children who suffer from Acute 

Lymphoblastic Leukemia (ALL). 

The concept beyond imagery therapy is that it guides the human mind to 

gain comfort, consequently enhances cure instead of fear and worry. All what 

patients has to do is to recollect images that promote cure and lose of concentration 

on the disease. Virtual Reality and computer games are considered as a tool to 

implement the imagery therapy effectively. It immerses the patient in an 

environment full of attractions to gain concentration, this would provide patients 

with emotional support and encourage them to practice an active lifestyle against 

cancer. Cancer patients who are practicing an active positive lifestyle live longer 

than patients having negative lifestyle [1]. So, virtual reality and games are 

considered as an essential treatment for cancer patients, helping to fix patient’s 

mental state and treating his or her neurosis [1] and [5]. 




Sahar Ali Aseeri , Wadee S. Al-halabi 

BACKGROUND 

We have conducted extensive investigating on several studies focus on the 

effects of psychological therapy on cancer patients. These studies have been evaluated 

according to their research methods and psychological interventions. These studies 

which we have examined presented experimental results. The results have proven that 

psychological therapy for cancer patients have an effective role in improving patient’s 

improvement rate. Trijsburg and others [4] conducted a critical review on the effects of 

psychological treatment on cancer patients. The authors reported that customized 

counseling has been shown to be effective with respect to distress, self-concept, 

fatigue, and sexual problems. In her study [6], Jacqueline Stenson reported that 70% of 

cancer victims use alternative therapies including dietary supplements, acupuncture, 

hypnotism, massage therapy, guided imagery, magnets and biofeedback. The author 

added that a recent study on more than 350 cancer patients indicates that breast cancer 

group were the most likely to use imagery therapy and most of the respondents have 

improved their health and mentality. The author reported that alternative medicine is 

widely used by people with life-threatening diseases. She refers to a study conducted 

in Ohio State University-Columbus which indicates that there is a strong sign to 

support that using imagery therapy improves the life quality and in some situations it 

extend life expectations. Thus, cancer patients who used imagery therapy when taking 

chemotherapy sessions felt that they became more relaxed, more ready for the 

treatments and more positive than those who didn’t use this method. Rudolph [3] 

reported that she has conducted a study on 80 women having an advanced breast 

cancer. The author used imagery therapy to treat her patients and the result indicated 

that imagery therapy succeeded in improving patient’s lifestyle and increase the 

number of cancer fighting cells. 

Many educative computer games for cancer patients were introduced such 

as Ben's game, Re-mission's game, Cancer's game and Staying Alive. The idea 

beyond the Ben Games developed by a little boy called Ben Duskin who lives in the 

United States, California. At the age of five, Ben was diagnosed with intense 

leukemia. He kept on fighting the disease more than three years. With a painful and 

stressful treatment Ben used to play video games to help him go through the pain as 

well as other chemotherapy treatment symptoms. So, when he got to the age of nine, 

Ben thought of designing a video game to help other kids with cancer. Ben says “my 

mom used to let me imagine the medicine inside my body as Pac-man moving 

through my veins eating all cancer cells,” from here became the idea of “Ben's 

game”. The game converted fighting cancer with chemotherapy into an exiting game 

for children diagnosed with cancer and explaining to them the steps of the treatment 

and its advantages. Children used to forget pain caused by the chemotherapy and to 

the total lose of sensation due to the effect of computer game. The main character in 

the game is the “little Ben” who slide on the screen through a skateboard and jumps 

on the infected cells to catch seven shields to save himself from chemotherapy side 

effects such as vomiting, fever and hair loses, please refer to Figure 1[7]. 

447


Figure 1: Ben Duskin 

HopeLab organization have developed the “Re-mission's game” for 

different types of cancer patients and specifically targeting teenagers. It consists of 

20 levels and works on destroying cancer cells and fighting extreme side effects. 375 

cancer patients between the age of 13-29 enrolled in 34 medical center in the United 

States, Canada and Australia. Each of them were assigned to an individual computer 

game loaded with popular games in addition to re-mission game. The research 

resulted in making cancer patients have a full information on cancer and what 

happens inside their bodies, also patients states have been improved clearly over 

time, please refer to Figure 2 [8]. 

Figure 2: Re-Mission game 





The “Cancer game” was developed by Oda and Kristula. In this game 

patients visualize the elimination of cancer cells. The game came with an idea that 

instead of displays “Game Over” on the screen when a player loses, a white cell is 

displayed on the screen saying “I’m tired, let’s play again later,” please refer to 

Figure 3 [11]. 

Figure 3: Yuko Oda, assistant professor of digital media, and David Kristula 

Staying Alive is a visualization tool for cancer victims to enhance 

imagination and supports relaxation. This game allows for direct visualization as 

well as provides the user with an easy control on the VR environment at the same 

time. Figure 4 shows a participant using this game [10]. 

"Medical News Today" [9] presented a study which was conducted on 20 

cancer patients using virtual reality technique along with chemotherapy. The report 

found that VR technique succeeded in reducing stressful and distracting patient’s 

attention from pain during/after chemotherapy. All participants preferred to use 

virtual reality through chemotherapy rather than not using any thing at all, 95% of 

them would like to use VR again. They also said that VR helped on passing time 

while taking the chemotherapy. 

449


Figure 4: User interacting with Virtual Teacher 

Moore and Spiegel indicated in their study [12] that it is possible to use 

virtual environment as a powerful agency to distract user concentration. VR has 

the advantages of elevating pain and other side effects related chemotherapy. 

The author added that VR provides relaxation which lowers heartbeats according 

to these who did not use this reality. Another study conducted by Hoffman and 

others [13] supports the earlier research and proved that VR reduces pain 

because it reduces the activity of centers which are responsible for sense of pain 

in the brain. This effect could also be achieved by creating positive images. 

Schneider and Workman [14] conducted an experiment on children at the age of 

10-17. The authors reported that virtual reality is an effective method to distract 

patient concentration, thus reduces pain and worry during chemotherapy 

treatment. The children who participated in the experiment suffer from 

leukemia. They choose virtual environment suitable to their age and needs. Kids 

used to wear Head mounted display (HMD) during the chemotherapy treatment 

session. In some cases this last between 40-120 minutes. The result shows that 

82% of participants did not find troubles using the kits. 18% of them did not feel 

much difference in using VR. However, all participants prefer to use virtual 

reality through chemotherapy treatments. 

Connecting patients with the outside world is a concern of Dr.Oyama 

[1]. The author reported that treatment of cancer disease requires a long stay in 

the hospitals. This situation takes the patient away from his family, relatives and 

friends. In this case, the patient may suffer from depression and tumultuous. The 

author reported that we could use VR as a psychological treatment to connect 





patients with the outside world. In the same paper the author added that the 

result of using the virtual environment in the Kansi project in the nationality 

center hospital for tumors in Japan indicate that patients who watch movies on 

different shows felt more relaxed. These patients use to forget reality of being 

inside the hospital. It is also reduces side effects related to chemotherapy 

treatment such as nausea, vomiting and stress. 

CONCLUSIONS 

According to earlier literature, using imagery therapy and virtual reality as a 

complementary therapy along with chemotherapy is an effective method to distract 

patient’s concentration on symptoms and pain. Therefore, it improves patient’s inner 

feeling, which is reflected on their overall improvement. we conclude that VR 

technology can be applied to palliative medicine (1) to support communication between 

the patient and other people, (2) to provide psychological support for patients to treat 

neurosis and help to stabilize the patient's mental state, and (3) to truly treat cancer. 

REFERENCES 

[1] Hiroshi Oyama, Virtual reality for palliative medicine, Ios Press: Amsterdam, 

Netherlands, 1998. 

[2] Spiegel D, Moore R, Imagery and hypnosis in the treatment of cancer patients, 

Oncology (Williston Park), Aug;11(8):1179-89, 1997. 

[3] Cindy Ann Rudolph, A study of the benefits of cancer patients engaging in 

complementary therapies, December, 2002. 

[4] R. W. TRIJSBURG, F. C. E. VAN KNIPPENBERG, S. E. RIJPMA, Effects of 

Psychological Treatment on Cancer Patients: A Critical Review, Psychosomatic 

Medicine 54:489-517, 1992. 

[5] Oyama H, Virtual reality for the palliative care of cancer, Stud Health Technol 

Inform, 44:87-94, 1997. 

[6] Jacqueline Stenson, http://www.chelationtherapyonline.com/PreventCancer/p55.htm 

[7] http://express.howstuffworks.com/ep-ben-eric.htm 

[8] Re-Mission Outcomes Study: A Research Trial of a Video Game Shows 

Improvement in Health-Related Outcomes for Young People with Cancer, 101 

University Avenue, Suite 220 - Palo Alto, CA 94301-1638 hopelab.org. 

451


[9] http://www.ihealthbeat.org 

[10] David A.Becker, Alex Pentland, Staying Alive:A Virtual Reality Visualization 

Tool for Cancer Pationts, 20 Ames St.-Cambridge.MA 02139. 

[11] http://www.albright.edu/reporter/fall2003/games3.html 

[12] Moore R, Spiegel D, Uses of Guided Imagery for Pain Control by African- 

American and White Women with Metastatic Breast Cancer, Integrative 

Medicine Consult, 2(2/3): 115-126, 1999. 

[13] Hoffman HG, Richards TL, Coda B, Bills AR, Blough D, Richards AL, Sharar, 

of thermal painrelated brain activity with virtual reality: evidence from fMRI, 

Neuroreport, 15:1245-1248, 2004. 

[14] Schneider SM, Workman ML, Virtual reality as a distraction intervention for 

children receiving chemotherapy, Pediatr Nurs. Nov-Dec;26(6):593-7, 2000. 



INDEX 

(A) 

Proceedings of the 7 th Saudi Engineering Conference, KSU, Riyadh,2007 

Volume / Page 

A. A.Al-Zahrani V1/173 

A. Almaroo V4/315 

A. Al-Negheimish V2/159,V2/201,V2/131,V3/97,V3/111 

A. Al-Samhan V5/355 

A. Bokhary V1/157 

A. E. Abasaeed V2/27,V4/263 

A. Es.Nassef V5/307 

A. Gari V1/157 

A. Haddad V1/15 

A. M. Alhozaimy V3/81,V3/97,V3/111 

A. M. Al-Saleem V3/81 

A. M. Shibl V5/83 

A. M. Zaki V4/219 

A. S. Al-Mozini V2/27 

A. Shuraim V2/201,V2/159,V2185,V3/123 

A. Telba V4/315 

A.A. Al-Arainy V1/129,V4/231,V4/243,V/257 

A.A. Algarni V2/215 

A.A. Al-Ohaly V2/119 

A.A. Al-Tuwaijri V1/33 

A.A. Ibrahim V4/73,V4/103 

A.A. Mansour V4/219 

A.A.Emam V3/269 

A.A.F. Shaheen V3/269 

A.Abo El-Naser V5/307 

A.Al-Ayiashi, V5/99 

A.H.M.A.Rahim V5/33

Index 

A.Ibrahim 


V5/111 

A.M. Al-Mujahid V1/281 

A’kif Al-Fuagara V3/223 

Aaron Phoenix V4/131 

Ab Rahamn Ahmad V5/455 

Abdallah O. Bafail V1/143 

Abdel Fattah Sheta V4/403 

Abdelhamid Charif V2/145,V3/123 

Abdel-Lateef, T.H V3/281 

Abdessalem bsissa V4/379 

Abdul Halim Ghazali V3/223 

Abdul Rashid Mohamed Shariff V1/115,V4/299 

Abdulaziz Alhamid V1/129 

Abdulaziz Efaien Alsolami V5/397 

Abdulghaffar Aljawi V5/185 

Abdulhafiz Omar S. Alshenawy V3/387 

Abdulhakim A. Almajid V1/99,129 

Abdullah Alqahtani V4/181 

Abdullah M. Mohorjy V2/53 

Abdullah Mohammad AL-Garni V5/3 

Abdullah Omar Baz V5/409 

Abdulmohsen Alheraish V4/269 

Abdul-Rahman M. Al-Ahmari V5/367 

Adnan H. Zahed V1/143 

Ahmad A. Fayed V5/125 

Ahmad Rodzi Mahmud V1/115,V3/223 

Ahmed K. Abdellatif V5/3 

Ahmed KADACHI V2/41 

Ahmed Mohameden V4/369 

Ahmed S. Hagazy V5/199 

Proceedings of the 7 th Saudi Engineering Conference, KSU, Riyadh,2007

Alaa Malaika 


V3/63 

Alaa R. Gaber V3/325 

Al-Fatish Ahemed S. V4/15,73 

Al-Hosain M. Ali V3/325,347 

Ali M. Al-Bahi V1/143 

Ali Muhammad Rushdi V4/389,V5/397,409 

Ali S. Hennache V1/87 

AlNashef I. M V2/3 

Aly. A. Emam V5/283 

Amer. A. F. Shaheen V5/283 

Ashraf ElAshaal V2/99 

Ashraf Mohra V4/323 

Ayed Algarni, V4/357 

Ayman Khalil V3/337 

(B) 

B. Almashary V4/315,V4/333 

B.M. Alshammari V1/3 

Basharat Salim V5/155, V5/173 

Bassam A. AlBassam V5/125 

Ben Betlem V4/143 

Benyagoub DJEBLI V3/163 

Brain Roffel V4/143 

Brant Peppley V4/131 

V5/199 

(D) 

Djamel ELDdine KERDAL V3/163 

(E) 

E. E. Abu-Elzahab V4/219 

Ela Halliop V4/131 


Index

Index 

El-Monishawy N. 


V5/247 

ElSayed Elbeheiry V5/433 

Essam E. Hassan V1/71 

(F) 

F. Skoczylas V3/151 

F.M. Trabzuni V1/293 

Fakeeha Anis V1/259,V4/15,47 

Farag Abd El Salam Abd El Aleem V1/317 

Faruk Yigit V5/231,V5/321 

Foudil Abdessemed V1/75 

(G) 

G. E. Y. I. Abou Raya V5/307 

G. Zaki V1/157 

Günter Weickert V4/143 

(H) 

H. Bassindowa V5/99 

H. Houari V3/151 

H.A. Abaoud V2/229 

Hamad I. Al-Abdul Wahhab V3/3, V3/23 

Hanafy M. Omar V5/219 

Hany Al-Ansary V5/17,V5/69 

Hasan K. Atiyeh V4/131,169 

Helmi Zulhaidi Mohd Shafri V1/115, V4/299 

Hisham M. Kattan V3/3 

Hosam Abozeid V3/337 

(I) 

I. S. Al-Mutaz V2/27 

I.M.Metwally V3/269 


Iain Cumming 


V4/181 

Ibrahim A.A V4/15 

Ibrahim Elshafiey V4/357 

Ibrahim M. Metwally V3/233 

Inas AlNashef V4/193 

Isam A. Mahmoud V3/3 

Ismail K. Warad V4/205 

(J) 

James Warner V4/3 

Jamil Jarallah Al-Bagawi V1/21 

Johari Mat Akhir V3/299 

Jon Pharoah V4/131 

(K) 

K. Al-Nawad V4/119 

K. Al-Sheref V2/159,V2/185 

K. Bhattacharya V2/215 

K.B. Ramachandran V4/159 

K.Koerner V5/111 

K.M.Wagialla V2/241 

Khaled M. Shokry V5/283 

Khalid A. AlSaif V3/63.V5/83 

Khalid Al-Osaimi V4/269 

Klaus Hellgardt V4/181 

Kunal Karan V4/131 

(M) 

M. A. Ali V4/119 

M. A. Daous V1/305 

M. A. El Ela V4/315 

M. A. Soliman V1/259,V4/15,V4/35 


Index

Index 

M. Al-Haddad 


V2/185 

M. Al-Shamrani V2/159 

M. Arif V3/141 

M. E. Seliman V5/455 

M. F. Zedan V1/215 

M. H. Abdel-Majeed V1/173 

M. K. Al-Mesfer V4/63 

M. Mahmood V5/335 

M. Moustafa V2/159 

M. S. Anam V4/257 

M. Siddique V1/157 

M.A. Bahafzallah V5/99 

M.A. Quraishi V3/141 

M.Ahsanul Alam V5/33 

M.Boumaza V5/141 

M.F. Kandlawala V5/33 

M.F. Zedan V1/237 

M.F. Zedan V5/83 

M.I. Qureshi V4/231,V4/243,V4/257 

M.M. Al-Edini, V5/99 

M.S. Al-Saud V4/285 

M.S. Owayedh V1/59 

Magdy Massoud V4/403 

Maged H. Hussein V2/53 

Maher Alodan V1/129 

Mahmoud S. Soliman V5/367 

Majeed A. Alkanhal V4/323,V4/357 

Mansour AlHazzaa V4/193 

Matthews, M. A V2/3 

Meftah Salem M Alfatni V4/299 

Mehmet Akyurt V5/185 

Mesfer M. Al-Zahrani V3/403 




Mirza G. Baig V3/3,V3/23 

Mohamad Mahmoud Al-Rahhal V1/201 

Mohamed A. Elshabrawy V3/325 

Mohamed A. M. I V5/247, V5/265 

Mohamed E. Ali V5/17, V5/69, V5/293 

Mohamed Elkady V1/3,33,59,129,133,189,281,V2/15,109,21 

5,229,V4/285,V5/383 

Mohamed Fayed V3/337 

Mohamed ElMadany V1/99,V1/129,V1/133,V3/63,V5/125,V5/199 

Mohamed S. Aazam V3/247 

Mohamed S. Issa V3/233 

MohammadAL-ESHAIKH V2/41 

Mohammad Abuzaed V4/299 

Mohammad Al-haj Ali V4/143 

Mohammad Asif V4/23 

Mohammad Barry V3/403 

Mohammad Ibrahim, V3/403 

MohammadIqbal Khan V1/293,V3/36, V3/49,V3/81 

Mohammad Rizwan V3/403 

Mohammad Shameem V3/403 

Mohammed M. Amro V1/267 

Mohammed T. Mallah V3/261 

Mona N. Eskander V4/343 

Mostafa A. Hamed V5/3 

Muhammad A. Al-Zahrani V2/111 

Muhammad Taher Abuelma’atti V1/47 

(N) 

N. A. Siddiqui V3/373 

N. M. Al-Abbadi V5/383 

N.H. Malik V4/231,V4/243,V3/257 

Naeem Akhtar V2/111 

Index

Index 

Nahid Maraie 


V2/99 

Nedal, A. Mohammad V3/299 

Nedim Turkmen V1/157,V5/185 

Noordin Ahmad V1/115,V2/83 

Noordin bin Ahmad V3/215 

(O) 

O. A. Maghgoub V4/219 

O. Zeitoun V5/69 

O.M. Al-Rabghi V1/157,V5,99 

Omar Mohammed Ba-Rukab V4/389 

Omar M. Eshanta V4/299 

Osama A. Mahgoub V4/343 

Osama Al-Gahtani V4/3 

Osama M. Arafa V4/343 

Osama M. Ben Saaed V4/299 

(P) 

Paul Charpentier V4/87 

(R) 

R. Mirza V2/185 

R. A. Khaled V1/157 

R.D. Findlay V4/285 

R.Singer V5/111 

Rajeh Z. Al-Zaid V2/131,185 

Ranya Fadlalla Abdalla V2/83,V3/215 

Rasha A. Al-meleak V3/435 

Reda M.S. Abdel Aal V1/143 

Richard Holdich V4/181 

(S) 

S. Al-Faidi V5/99 


S. A. Ali 


V4/119 

S. Akhtar V3/141 

S. Al-Khattaf V4/119 

S. M. Darwish V5/355,V5/367 

S.A. Al-Ajlan V2/229 

S.A. Al-Rumaih V1/3 

S.A.Aseeri V3/417 

S.I. Al- Mayman V4/103 

S.M. Al-Ghuwainem V2/119 

Saad Haj Bakry V4/269 

Sadeek, A.B V3/281 

Saeed Alzahrani V1/129,V4/63 

Sahar Ali Aseeri V3/445 

Said Shahran V3/97 

Salaheddine Bendak V5/419 

Saleh Alshebeili V4/323 

Saleh Al-Sulaiman V3/97 

Saleh Alswailmi V3/247 

Salmah Zakaria V3/223 

Sami A. Al-Sanea V1/215,V1/237 

Samir Ali Ead V3/179 

Sani Yahaya V2/83,V3/215 

Saud Al-Swaida V3/97 

Sedky Abd Allah Tohmay V3/355 

Sharifah Mastura S. A. V3/299 

Shattri B. Mansor V1/115,V3/223 

Shehab Mourad V3/337 

Sherif ElKholy V2/99 

Shuja Ahmad Abbasi V4/333 

Siti Nor Alia Roslan V1/115 


Index

Index 

Sohrab Rohani 


V4/87 

Souid Souad V4/379 

Suhial Kiwan V5/293 

Sulaiman Al-Zuhair V4/159 

Syed Manzoor Qasim V4/333 

(T) 

T.C. Al-Smari V4/103 

T.H. Almusallam V1/293,V3/203,V3/373 

Thamer Ahmed V3/223 

Tohamy, S.A. V3/281 

Toufik Zebbiche V5/47 

V4/299 

(W) 

W.A. Almasy V1/293,V2/15,V4/323 

Wadee S. Al-halabi V3/435,V3/445 

Waleed a. Al-Rumaih V1/339 

Waleed M. Zahid V2/67 

Weidner , J. J V2/3 

(Y) 

Y. A. Al-Salloum V2/201,V3/373 

Y. Benachour V3/151 

Y.S. Al-Sugayer V1/293,V4/103 

Yousef Bakhbakhi V4/87

Volume-3 English - Colleges - King Saud University

You also want an ePaper? Increase the reach of your titles

Delete template?

Save as template?