You are on page 1of 264

Behaviour of Glass FRP Composite Tubes

Under Repeated Impact for Piling Application
By
Ernesto Jusayan Guades

Supervised by
Prof. Thiru Aravinthan
Dr. Mainul Islam

A dissertation submitted for the award of
DOCTOR OF PHILOSOPHY

Centre of Excellence in Engineered Fibre Composites
Faculty of Engineering and Surveying
University of Southern Queensland
Toowoomba, Queensland, Australia
May 2013

EJ Guades

Abstract
Fibre composites have been a viable option in replacing traditional pile materials
such as concrete, steel and timber in harsh environmental conditions. On the other
hand, the emergence of fibre reinforced polymer (FRP) composite tubes as a
structural component and their corrosion-resistant characteristics made these
materials potential in piling application. Driving these piles, however, requires more
careful consideration due to their relatively low stiffness and thin walls. The
possibility of damaging the fibre composite materials during the process of impact
driving is always a concern. Research has therefore focused in understanding the
impact behaviour of these materials in order for them to be safely and effectively
driven into the ground.
This study investigated the behaviour of composite tubes subjected by
repeated axial impact. The effects of impact event (incident energy and number of
impact) on the instantaneous response and the residual properties of composite tubes
were examined. Tubes made of glass/vinyl ester, glass/polyester, and glass/epoxy
materials of different cross sections were considered. The impact behaviour of the
tubes was experimentally and analytically investigated.
An experimental study on the repeated impact behaviour of square composite
tube was conducted. The result showed that the dominant failure mode of the tube
repeatedly impacted was characterised by progressive crushing at the upper end. This
failure was manifested by inter and intra laminar cracking and glass fibre ruptures
with simultaneous development of axial splits along its corners. It was found that the
drop mass and impact velocity (or drop height) have pronounced effects on the
collapse of the tubes at lower incident energies. Their effects, however, gradually
decrease at relatively higher energies. The result also indicated that the incident
energy is the major damage factor in the failure of tubes for lower number of
impacts. On the contrary, the number of impacts becomes the key reason as soon as
the value of incident energy decreases.
The effects of the damage factors such as the level of impact energy, the
impact repetitions, and the mass impactor on the residual (post-impact) properties
were also examined. The result of the investigation revealed that these factors
significantly influenced the residual strength degradation of the impacted tubes. In
contrast, the residual modulus was found to be less affected by these factors since the

Behaviour of glass FRP composite tubes under repeated impact for piling application

i

EJ Guades

damage brought by them is localised in most of the cases. The maximum reduction
on the residual moduli is roughly 5%. On the other hand, the residual strengths
degraded by up to 10%. The flexural strength of the tube was the most severely
affected by the impact damage than its compressive and tensile strengths. This result
was due to the fact that the impact damage on matrix and fibre both contributed on
the flexural strength degradation. Moreover, the presence of matrix cracks or
delamination lead to an increase in buckling instability during the flexural test,
resulting to a much higher degradation compared to the other strengths. The
comparison of the residual compressive strengths sourced at different locations along
the height of the tube revealed that the strength reduction varied with its location.
The degradation of the compressive strength of the impacted tube decreased when its
location from the top of the tube increased. This result indicated that the influence of
impact damage on the degradation of residual compressive strength of the tube is
concentrated only in region closer to the impact point.
Finally, theoretical prediction using the basic energy principle was performed
to gain additional understanding on the damage evolution behaviour of composite
tubes subjected by repeated axial impact. The damage evolution model was verified
through experimental investigation on a 100 mm square pultruded tube. The model
was applied to composite tubes of different cross sections and materials made from
vinyl ester/polyester/epoxy matrix reinforced with glass fibres. It was found that the
experimental results on a 100 mm square pultruded tube and the proposed damage
model agreed well with each other. The variation is less than 10% indicating that the
model predicted reasonably the damage evolution of the tube subjected by repeated
impact loading. It was also found that the energies describing the low cycle, high
cycle, and endurance fatigue regions of the composite tubes are largely dependent on
their corresponding critical energy Ec. The higher the Ec values, the higher the range
of energies characterising these regions. The repeated impact curves (or Ec) of tubes
made from glass/epoxy is higher compared to the other matrix materials. Similarly,
circular tubes have greater Ec values of comparable square and rectangular tubes.
From this study, an improved understanding of the behaviour of glass fibre
FRP composite tubes under repeated axial impact can be achieved. The information
provided in this study will help in developing efficient techniques and guidelines in
driving composites piles.

Behaviour of glass FRP composite tubes under repeated impact for piling application

ii

EJ Guades

Certification of Dissertation

I certify that the ideas, experimental work, results, analysis and conclusions reported
in this dissertation are entirely my own effort, except where otherwise
acknowledged. I also certify that the work is original and has not been previously
submitted for any award, except where otherwise acknowledged.

/

/

/

/

/

/

Signature of Candidate

Endorsed:

Signature of Supervisor/s

Signature of Supervisor/s

Behaviour of glass FRP composite tubes under repeated impact for piling application

iii

EJ Guades

Acknowledgements
With humble gratitude I must acknowledge the following that have in one way or the
other contributed to the successful completion of this dissertation.
Prof. Thiru Aravinthan, my Principal Supervisor, for giving me the
opportunity to do a PhD at the University of Southern Queensland (USQ). I am
grateful to him for coaching me and willingly providing invaluable input and
direction. I learned a great deal of things from him in my entire journey of PhD. I am
also indebted to Dr. Mainul Islam, my Associate Supervisor, for sharing his time and
ideas to make this dissertation a success. I greatly appreciate Dr. Allan Manalo for
his support in my application to study at USQ. His technical suggestions and
assistance were indispensable in improving the quality of this research. The
generosity he extended to me during my study is greatly appreciated.
I would like to acknowledge the people behind USQ who provided the Post
graduate Scholarship Grant. I thank the supports of the Faculty of Engineering and
Surveying and the Centre of Excellence in Engineered Fibre Composites (CEEFC)
for making this research possible. My thanks to Assoc. Prof. Karu Karunasena, Dr.
Jay Epaarachchi, Dr. Francisco Cardona for all the useful discussions and
suggestions. I owe an appreciation for the technical and administrative support from
Martin Geach, Wayne Crowell, Atul Sakhiya, and Mohan Trada. Thanks to CEEFC
staff and postgraduate students for the support and friendship. I especially thank
Michael Kemp and all the staff of Wagners Composite Fibre Technology for
providing the precious test samples. Thanks are expressed to the administration and
staff of Northwest Samar State University for the Study Grant that would pave the
way for my travel to Australia in pursuit of another academic achievement.
My unending recognition to Myla, who always, in all ways, was there for me.
I am grateful to her for unselfishly setting aside her personal needs to give way to my
personal dreams and aspiration. Very special thanks to my family who have been a
source of encouragement and inspiration throughout my life. My appreciation to the
Inocentes family, Jen, and the Filipino community of Toowoomba for welcoming me
into their homes. Their incredible hospitality and generosity helped me overcome my
homesickness. Above all, I am thanking the Almighty God for guiding me all
throughout this endeavour. To those whom I missed to mention but have been a great
part of my study, thank you very much.

Behaviour of glass FRP composite tubes under repeated impact for piling application

iv

J. T.J. Japan. Aravinthan. Toowoomba. 237-242 4. Manalo (2012). Proceedings of the Southern Region Engineering Conference. Aravinthan. 2. May issue . Behaviour of glass FRP composite tubes under repeated impact for piling application v .C. and A. 2010. Islam. May issue. Islam (2013). Islam. M. Damage modelling of repeatedly impacted square fibre-reinforced polymer composite tube. Stiffness degradation of FRP pultruded tubes under repeated axial impacts.com/science/article/pii/S0261306912008801 Conference Papers/Poster Presentation 1. Aravinthan. E. March issue. Sydney. E. p 1932-1942.J. p 211-221. Proceedings of the 22nd Australasian Conference on th Mechanics of Structures and Materials (ACMSM22). E.sciencedirect. Experimental investigation on the behaviour of square FRP composite tubes under repeated axial impact. and A.M. T. Volume 95. Guades. M. Volume 47. Aravinthan (2013). A. Residual properties of square FRP composite tubes subjected to repeated axial impact. and M. Guades. Driveability of composite piles. New South Wales. Manalo.com/science/article/pii/S0263822312000451 2. Aravinthan. T. T. Proceedings of the 1st International Postgraduate Conference on Engineering. p 354-365. Islam. and M. and M. T. An overview on the application of FRP composites in piling system. Composite Structures. Australia. Composite Structures. Aravinthan. http://www.J. Islam. M. QUT.sciencedirect. Aravinthan. April 27-29.EJ Guades Associated Publications Journal 1. Designing and Developing the Built Environment for Sustainable Wellbeing. Journal of Materials and Design. and M. Paper no T3-4. Brisbane. Proceedings of the 3rd Asia-Pacific Conference on FRP in Structures.sciencedirect.M.M. (2010). Islam. T.com/science/article/pii/S0263822312004072 4. http://www.M.C. Guades. 3. 11-14 December. http://www. Islam (2013). Manalo. Australia. Manalo (2012).sciencedirect.J. T. Guades and T.C. Guades.M. Paper no F1B05.M. Composite Structures.C. Hokkaido. Volume 94.C. A review on the driving performance of FRP composite piles. Guades. and A. Effects of energy levels on the impact fatigue behaviour and post-impact flexural properties of square FRP pultruded tubes. E.M. A. Guades. p. Aravinthan. Guades.4. http://www. E.com/science/article/pii/S0263822312005296 3. p 687-697. E. E. Australia. January issue. February 2. E.J. November 11-12.J. Volume 97. Manalo (2012).J. (2011).

November 15. T. Australia. Proceedings of the 22nd Australasian Conference on the Mechanics of Structures and Materials (ACMSM21).M. Poster presentation during the USQ Community Engaged Research Evening. (2010). Aravinthan. Queensland. T. Behaviour of composite pile under axial compression load. Australia. and M. Islam (2011). November 10.M. Melbourne. Guades. Sirimanna. Impact behavior of pultruded tubes as hollow FRP piles. Guades. Islam (2010). Australia.J. Behaviour of glass FRP composite tubes under repeated impact for piling application vi . Sacred Heart Church Function Room.J. Aravinthan. Islam. 7. 457-462. Aravinthan & M. E. Towoomba. C. Poster presentation during the USQ Community Engaged Research Evening. Towoomba. and M. Queensland. E.M.EJ Guades 5. T. Application and impact behavior of pultruded tube as FRP composite pile. USQ Refectory. p. S. 6. E. December 7-10.J. Guades.

5 Study on the impact behaviour of FRP composite tubes as a research needs………………………………………… 35 2..3 The ability of the pile to transfer driving stresses………………………………………. 1 1. 10 Chapter 2 Review of composite piles and their driving performance 2..……………. 8 1.9 Outline of the thesis…………………………………………… 9 1. ……..2.5 Challenges in using hollow FRP pipe piles……. 2 1.4 Recent developments on hollow FRP pipe piles………………… 30 2....2 Resistance to driving offered by the soil……………. 4 1. 5 1..2 Types of composite piles………………………………………… 11 2.EJ Guades Table of Contents List of figures List of tables Notations xii xvii xx Chapter 1 Introduction 1.3.6 Behaviour of FRP composite plates/laminates repeatedly impacted or tubes under repeated transverse impact…….1 Steel pipe core piles………………………………….7 FRP sheet piles……………………………………….1 Types of driving hammer and its effect……………… 18 2.. 15 2..... 23 2.2.2.8 Scope of the thesis.2.5 Fibreglass reinforced plastic piles……………………. 16 2. 14 2. 1 1. 39 Behaviour of glass FRP composite tubes under repeated impact for piling application vii .1 General…………………………………………………………. 12 2..7 Objectives……………………………………………………… 7 1. 18 2..2 Structurally reinforced plastic piles………………….. ………35 2..4 FRP tubes as composite piles…………………………………..3 Driving performance of composite piles……………………….. 17 2.2 Background…………………………………………….3.10 Summary………………………………………………………...3.2...……………………………………………...7 Conclusions …………………………………………….4 Fibreglass pultruded piles……………………………..2. 11 2..2.4 Strength of the pile to resist driving stresses………… 25 2. 11 2... 13 2..3 Fibre composites as an alternative in piling applications………...1 General…………………………………………………………...6 Research needs related to their driving performance………… 6 1.3 Concrete-filled FRP pipe piles………………………. 20 2..6 Hollow FRP pipe piles………………………………...3.

1 General…………………………………………………………... 41 3.......1 Test specimen and repeated impact testing ………….1 Mode of damage……………………………. 80 4.9 Conclusions……………………………………………………… 71 Chapter 4 Investigation on the behaviour of square FRP composite tubes under repeated axial impact 4..3 Experimental results and discussion……………………………..2 Progressive failure pattern…………………………… 80 4.2.1 Introduction……………………………………………………… 72 4...5..1 Compressive test.. …………… 63 3. ………78 4. 99 5. 42 3...3.3..2..2 Residual properties testing …………………………..7.1 Mode of damage…………………………….4 Impact energy…………………………………………87 4.3 Data processing……………………………….3 Manufacturing of tubes using pultrusion process……….2 Test set-up and procedure……………………. 51 3. ……… 60 3.. 69 3. 54 3.. 101 5...8 Summary of the mechanical properties of composite tubes……... 80 4.EJ Guades Chapter 3 Characterisation of the properties of composite tubes 3....2 Tensile test…………………………………………… 47 3. 96 Chapter 5 Residual properties of square FRP composite tubes subjected to repeated axial impact 5...3..... 106 Behaviour of glass FRP composite tubes under repeated impact for piling application viii ....2.1 Introduction……………………………………………………… 98 5.2 Summary of coupon test results……………. 59 3.7 Finite element (FE) analysis on full scale specimen ……………. 92 4.3 Experimental results and discussion……………………………..... 45 3...5...4 Glass fibre content……………………………………………….1 FE simulation on the compressive behaviour.2 FE simulation on the flexural behaviour. 106 5..5.1 Test specimen…………………………………………73 4.2.2.5 Impact damage tolerance limit………………………... 49 3..2 Flexural test…………………………………………..3....5 Coupon tests……………………………………………………..2 Experimental program…………………………………………… 73 4.6..7.3.3.4 Conclusions ……………………………………………………. 45 3........ ……...2 Experimental program…………………………………………… 99 5.. 41 3. ……………………………………..3..3 Flexural test…………………………………. 43 3.....1 Compressive test…………………………………….2 Composite tubes under study……………………………………..6. ………73 4..3 Impact load……………………………………………83 4. 106 5..6 Full scale tests…………………………………………………… 51 3...

1 Minimum number of impacts to failure of the tube.3.6.3...... 146 6..5 Evaluation of damage using parameter D………………………. 132 6...3 Determination of (Ec)Quasi-static using quasi-static compressive test……………………………………… 138 6.1 Square and rectangular FRP composite tubes...1 Introduction……………………………………………………… 128 6.1 Specimen and testing………………………………… 131 6.3... 141 6. 155 7. 128 6.3 Effects of impact energy……………………. 136 6. 141 6....8 Variations of residual compressive strength with the height of the tube…………………………… 124 5. 146 6.4 Conclusions……………………………………………………… 125 Chapter 6 Damage modelling of repeatedly impacted FRP composite tube 6..1 Introduction……………………………………………………...2. 131 6. 140 6.. 153 Chapter 7 Application of the damage evolution model to other types of composite tubes 7. Nf …..3 Quasi-static compressive test…………………………………….3...6.3..1 Verification of the repeated impact curve…………….. 108 5..9. 134 6..………….7 Comparison with the experimental data………………………….3.…………………….EJ Guades 5..6.. …….5 Effects of mass of the impactor……………………… 116 5.3 Epoxy resin……………………………………………157 Behaviour of glass FRP composite tubes under repeated impact for piling application ix .7.3. 134 6. 142 6..2 Theoretical prediction methods…………………………………. 156 7.9 Application of the model to FRP composite tubes with square and rectangular cross sections …………………… …….6 Proposed damage response model………………………………..7 Residual strength versus modulus…………………….4 Effects of impact repetitions …………….…………………………………….1 Vinyl ester resin……………………………………… 156 7..4 Solving b value……………….2 Polyester resin…………………………………………157 7.…………………………………… 144 6.2.7. tensile and flexural properties……………………………….. 122 5..2 Validation of the proposed model…………………….2 Background on the constituents of composite tubes used in the model……………………………………………….6 Comparison between compressive.6..2 Minimum incident energy to fail the tube for one impact (critical energy).4 Repeated impact test results…………………………………….2.8 Summary of procedure in establishing the damage evolution curve ………………………. 112 5.10 Conclusions……………………………………………………. 120 5.... Ec ………………………136 6.

6 Flexural test on full scale specimen…………………………….1 Analytical study on the variation of acceleration data…………..8 Conclusions……………………………………………………… 181 Chapter 8 Conclusions 8. A-3 A. B-3 B...3 Apparatus used in the micro observation of damage…………… B-4 Appendix C Variation of acceleration data and impact stress with the height of the tube C..4 Conclusions…………………………………………………….5 Compressive test on full scale specimen………………………. 182 8...3 Glass/vinyl ester composite tubes……………………………….EJ Guades 7..1 Summary………………………………………………………… 182 8. 182 8. A-7 A.4 Glass/polyester composite tubes………………………………… 163 7.3 Tensile test on coupon specimen……………………………….3 Prediction on the damage evolution of composite tubes………………………………..2. 178 7.1 Behaviour of composite tubes subjected by impact loading……………………………………... 183 8. C-1 C.6 Discussion on the repeated impact and damage evolution curves of FRP composite tubes…………………………………. C-13 C..2 Compressive test on coupon specimen…………………………..2 Effects of impact loading on the residual properties of composite tubes………………..7 Discussion on the application of FRP composite tubes in piling system ……………………………………………………..3 Recommendations for future study……………………………… 185 References 186 Appendix A Summary of results of the coupon and full scale tests on CT1 and CT2 specimens A.. C-5 C..2 Main conclusions from the study………………………………...5 Glass/epoxy composite tubes……………………………………. 170 7.. B-1 B.1 Fibre fraction test………………………………………………. A-2 A. A-1 A.2..2 Finite element modelling………………………………………..... A-9 Appendix B Summary of specimen dimension and snapshots of the machine/apparatus used in repeated impact test B... ……... 184 8. 158 7.. 176 7..3 Finite element analysis results and discussion………………….. A-5 A..4 Flexural test on coupon specimen……………………………….2 Repeated impact testing set-up and specimen snapshots………. C-21 Behaviour of glass FRP composite tubes under repeated impact for piling application x ..1 Summary on the details of the specimens……………………….2.

D-1 D.4 Summary of results of coupon flexural test…………………….. D-1 D. D-8 Behaviour of glass FRP composite tubes under repeated impact for piling application xi .1 Summary of the details of the tubes…………………………….2 Summary of results of coupon compressive test………………..EJ Guades Appendix D Summary of specimen dimension and results in residual properties testing D.3 Summary of results of coupon tensile test……………………… D-6 D.

.17 3.19 3..8 2.15 3. 62 Compressive failure mode of the tested tube …………………………….8 3.………………… 60 Lamina lay-up arrangement used in FE model …………………………..7 2. 16 Geometry of hollow FRP pipe piles used in the application…………….10 2.1 Problems of traditional piles installed in harsh environments Chapter 2 Figure 2...16 3.……… 58 Flexural load-strain relationships (4-point bending test)….……… 57 Flexural load-displacement relationship (4-point bending test)….5 3. 56 Flexural load-strain relationship (3-point bending test)………..12 3.…………….1 3. 46 Compressive stress-strain relationship…………………………………… 47 Compressive failure mode and condition of the specimens after the test…47 Tensile test set-up on coupons…………………………………………… 48 Tensile stress-strain relationship………………………………………….21 3. 50 Flexural stress-strain relationship………………………………………… 51 Flexural failure mode and condition of the specimens after the test……. .13 2.5 2. 33 Impact driving of 475 mm diameter hollow FRP pipe pile……………… 34 Chapter 3 Figure 3. 61 Compressive stress-strain relationships……..6 3.23 3.26 Page 2 Characterisation of the properties of composite tubes Figure caption Page Oblique view of the composite tubes…………………………………….………….20 3. 42 The basic pultrusion process concept…………………………………….1 2.3 3.4 2. 43 Coupon specimens and residue showing the fibre glass orientation…….10 3.. 55 Flexural load-displacement relationship (3-point bending test)…………..22 3.……… 59 Material modelling of the composite tube …………….11 2. 44 Compressive test set-up on coupons…………………………………….9 3..12 2.EJ Guades List of Figures Chapter 1 Introduction Figure Figure caption 1..25 3. 51 Compressive test set-up on full scale specimens………………………… 52 Compressive stress-strain relationship of full scale specimens………….14 3... 49 Tensile failure mode and condition of the specimens after the test……… 49 Flexural test set-up on coupons………………………………………….11 3.6 2.………………………….2 3. 57 Typical failure modes for in 3-point bending tests……………….7 3...4 3..15 Review of composite piles and their driving performance Figure caption Page Steel pipe core piles……………………………………………………… 12 Structurally reinforced plastic piles……………………………………… 13 Concrete-filled FRP pipe piles…………………………………………… 14 Fibreglass pultruded piles………………………………………………… 15 Fibreglass reinforced plastic piles……………………………………….13 3. 63 Behaviour of glass FRP composite tubes under repeated impact for piling application xii ..... 28 Composite piles driven near Route 351 Bridge………………………….3 2.. 29 Hollow FRP pipe piles replacing deteriorated timber piles……………… 31 Pultruded composite tubes used in shoring-up boardwalks……………… 32 Impact driving of 125 mm square pultruded tubes………………………..18 3.2 2. 53 Compressive failure mode and condition of the full scale specimens…… 54 Flexural test on full scale specimens…………………………………….14 2. 16 FRP sheet piles…………………………………………………………… 17 Condition of the composite piles after driving…………………………… 26 Condition of the composite piles after driving…………………………… 27 Composite pile installed in Route 40 Bridge……………………………. 59 Typical failure modes for in 3-point bending tests……………….9 2.24 3.

20 Investigation on the behaviour of square FRP composite tubes under repeated axial impact Residual properties of square FRP composite tubes subjected to repeated axial impact Figure caption Page Conditions of the tubes after impact test ………………………………… 101 Cutting plan of coupons used in residual properties testing…. …………..8 J……………....6 5.28 3....... 94 Nf vs...4 5.2 5. 68 Flexural failure mode in 4-point bending test…………..33 3......8 4.....……….... 74 Typical acceleration-displacement curves in impact testing …………… 80 Condition of the tubes after impact test…………………………………....impact energy relationships … 111 Enlarged view: Residual flexural modulus-impact energy relationships… 112 Residual strength-number of impacts relationships ……....1 5...…….9 5.......34 Actual tube (length varies from 1.... 111 Enlarged view: Residual tensile modulus.3 4............. 81 Damage progressions of collapsed tube impacted by 476...19 5.... 102 Compressive test specimens……………………………………………… 103 Tensile test specimens …………………………………………………… 104 Flexural test specimens …………………………………………………... L=1......12 Figure caption Page Impact testing set-up….....2 m to 1.....17 5.2 4.... 66 Flexural failure mode in 3-point bending test…………….1 4..69 Chapter 4 Figure 4..... 85 Peak load progressions of repeatedly impacted tubes…………………… 87 Typical energy curves……………………………………………………...27 3. 64 Support condition during flexural test (both ends)……………………….9 4...12 5..32 3..........10 4.…….. 67 Flexural load-displacement relationships (4-point bending)……….....29 3..13 5............. 65 Flexural load-displacement relationships (3-point bending)………..... 115 Enlarged view: Residual compressive modulus-number of impacts relationships……………………………………………………………… 115 Behaviour of glass FRP composite tubes under repeated impact for piling application xiii .4 4.14 5.......5 5.3 5........……...11 5.... 90 Comparison of the damage degree curves of repeatedly impacted tubes… 91 Incident energy vs. 63 FE model (3-point bending..8 5.....…………..……. 88 Impact energy histories of repeatedly impacted composite tubes……….....31 3.15 5...EJ Guades 3. 109 Enlarged view: Residual tensile strength-impact energy relationships…....impact energy relationships....5m)…………………………………….6 4... 83 Impact load histories of repeatedly impacted composite tubes………….. drop mass curve of repeatedly impacted tubes……………………....7 5.5 m)…………………………. 110 Enlarged view: Residual flexural strength-impact energy relationships… 110 Residual modulus-impact energy relationships …………... Nf curve of repeatedly impacted tubes………………...... 105 Scanned images showing typical micro-cracks on the surface of the tubes 106 Residual strength and impact energy relationships ………………………109 Enlarged view: Residual compressive strength-impact energy relationships……………………………………………………….7 4........ 113 Enlarged view: Residual compressive strength-number of impacts relationships … …………………………………………………………...... L=1.2 m)…………… ……………………… 64 FE model (4-point bending.10 5..... 111 Enlarged view: Residual compressive modulus..18 5..16 5.... 95 Nf vs.....……… 114 Residual modulus-number of impacts relationship …….11 4.………………………………………………….....30 3.. 113 Enlarged view: Residual tensile strength-number of impacts relationships 114 Enlarged view: Residual flexural strength-number of impacts relationships……………………………………………………….5 4.. impact velocity curve of repeatedly impacted tubes……………… 96 Chapter 5 Figure 5.

14 6.. 132 Normalised energy and number of impacts relationship………………… 133 D vs. N curve of the representative composite tube…………………….....…………..………...2 6...... 140 Comparison between the experimental data and repeated impact curve … 141 Proposed model vs....15 6.3 6..147 Crushed composite tubes…………………………….......4 6....... experimental data for non-collapsed tubes ………...16 6.19 6......….9 6.5 6... 121 Comparison of residual compressive....………….3 Damage modelling of repeatedly impacted FRP composite tubes Application of the damage evolution model to other types of composite tubes Figure caption Page Repeated impact curves of glass/vinyl ester tubes………………………..... tensile..30 5..............…....145 Square and rectangular composite tubes ……………......7 6.....21 5...28 5..... 161 Damage evolution curves of GV-C tube ………………………………… 162 Damage evolution curves of GV-S tube ………………………………… 162 Behaviour of glass FRP composite tubes under repeated impact for piling application xiv ..............22 5.13 6.29 5.27 5......…................EJ Guades 5... 118 Enlarged view: Residual flexural strength-drop mass relationships at different energy levels and number of impacts ……....2 7........ tensile...20 Figure caption Page Quasi-static compressive test……………………………………………....... 118 Residual modulus-drop mass relationships at different energy levels and number of impacts………………………………………………… …119 Enlarged view: Residual compressive modulus-drop mass relationships at different energy levels and number of impacts……......10 6...24 5.. and flexural strengths……... 134 Idealised lifetime response curve of the repeatedly impacted tube……… 135 Typical curve described by Ein = aNf-b………………………………….... 148 Load-displacement curves of S125 specimen ……………..34 Enlarged view: Residual tensile modulus-number of impacts relationships……………………………………………………………… 115 Enlarged view: Residual flexural modulus-number of impacts relationships …….… 144 Flow chart in establishing the damage evolution curve ………...8 6.………….........6 6........18 6.......33 5.148 Load-displacement curves of R75x100 specimen …….....…………..... 119 Enlarged view: Residual tensile modulus-drop mass relationships at different energy levels and number of impacts ……..………….1 7...25 5. 119 Enlarged view: Residual flexural modulus-drop mass relationships at different energy levels and number of impacts ……. and flexural moduli……… 122 Strength and modulus curves plotted at increasing impact energy levels...…………...... 136 Variation of the correlation β of glass/vinyl ester composite tubes ………137 Data points with the fitting line showing β and α relationship…………… 138 Typical load-displacement curves from quasi-static compressive test…… 139 Schematic diagram used in computing (Ec)Quasi-static ……………………… 139 b values using Excel 2010 “Solver” function……………………………........11 6.......17 6...... 151 Damage evolution curves of square and rectangular tubes ….. 116 Residual strength-drop mass relationships at different energy levels and number of impacts…………………………………………………… 117 Enlarged view: Residual compressive strength-drop mass relationships at different energy levels and number of impacts………………………… 117 Enlarged view: Residual tensile strength-drop mass relationships at different energy levels and number of impacts……....12 6.23 5.... 123 Variation of residual compressive strengths with the height of the tube… 125 Chapter 6 Figure 6.........31 5...26 5....…………...... 152 Chapter 7 Figure 7.. experimental data for collapsed tubes ……………… 143 Proposed model vs.......……….......... 120 Comparison of residual compressive.... 149 Repeated impact curves of the square and rectangular tubes…...32 5...1 6.

168 Damage evolution curves of GP-C2 tube ……………………………….5 7.19 Damage evolution curves of GV-H tube ………………………………… 162 Data points with the fitting line showing β and α relationship of glass/polyester tubes ………………………………… ………………... A-10 Schematic plan of 4-point bending test ………………………………….3)... B-4 Condition of the specimen after impact test (Test matrix from Table 4..15 7... 174 Damage evolution curves of GE-C3 tube ……………………... A-10 Schematic plan of 3-point bending test …………………………………........... A-7 Compressive load-displacement relationship of full scale specimens (CT1......4 Figure caption Page Repeated impact testing set-up data logger and fixtures…………………......4 A.10 A.17 7. 175 Damage evolution curves of GE-C5 tube …………………….. 169 Damage evolution curves of GP-S3 tube ……………………………….5 A... 169 Damage evolution curves of GP-S2 tube ……………………………….4 7.12 A.14 A... 169 Damage evolution curves of GP-S1 tube ……………………………….....11 A. 175 Damage evolution curves of GE-C4 tube ……………………..6 A. L=200 mm)…………………………………………………………A-9 Compressive load-displacement relationship of full scale specimens (CT2...... L=100 mm)…………………………………………………………A-9 Specimen cross section lay-out …...2 A...13 7..10 7. B-4 MOTIC® SMZ 168 Series stereo zoom microscope…………………….....6 7..... L=100 mm)…………………………………………………………A-8 Compressive load-displacement relationship of full scale specimens (CT1....... 175 Damage evolution curves of GE-S1 tube …………………….11 7... A-10 Flexural stress-displacement relationship (3-point bending test) of CT1… A-11 Flexural stress-displacement relationship (3-point bending test) of CT2…A-12 Flexural stress-strain relationship (4-point bending test) of CT1………… A-12 Appendix B Summary of specimen dimension and snapshots of the machine/apparatus used in repeated impact test Figure B.9 A..7 7.2)...1 B.. 179 Repeated impact curves of glass/epoxy tubes …………………….8 7...............2 B...EJ Guades 7....... B-4 Behaviour of glass FRP composite tubes under repeated impact for piling application xv ...3 B.. A-4 Tensile load-displacement relationship of coupon specimens (CT2)……...8 A. A-5 Flexural load-midspan deflection relationship of coupon specimens (CT1)……………………………………………… A-6 Flexural load-midspan deflection relationship of coupon specimens (CT2)……………………………………………… A-6 Simplified cross section of the tube………………………………………..... B-3 Condition of the specimen after impact test (Test matrix from Table 4...12 7.....16 7. 176 Appendix A Summary of results of the coupon and full scale tests on CT1 and CT2 specimens Figure A....1 A.....3 A.. 164 Repeated impact curves of glass/polyester tubes …………………………167 Damage evolution curves of GP-C1 tube ………………………………..………………………………………...... 168 Damage evolution curves of GP-C3 tube ………………………………...16 Figure caption Page Compressive load-displacement relationship of coupon specimens (CT1) A-3 Compressive load-displacement relationship of coupon specimens (CT2) A-3 Tensile load-displacement relationship of coupon specimens (CT1)……...9 7. 173 Damage evolution curves of GE-C1 tube ……………………...7 A..15 A. 174 Damage evolution curves of GE-C2 tube ……………………..18 7.14 7.13 A.....

C-2 Comparison of aL/2 and a1 values at varying impact mass ………….... time table simulating repeated impact loading (E480)……….. time table for the impulse period of 0.. C-19 Damaged length simulation using FE analysis…………………………… C-20 Behaviour of glass FRP composite tubes under repeated impact for piling application xvi .15 C. time table simulating material degradation (E480)…………… C-12 Comparison of time steps for E630……………………………………… C-13 Variation of peak axial stress in longitudinal direction………………….14 C.7 C.8 C. time table simulating repeated impact loading (E630)………. C-7 Lamina lay-up arrangement used in FE model…………………………… C-7 Factor vs.2 C.10 C.01 second……………… C-9 Variation of the static load case with the measured acceleration………… C-9 Factor vs..6 C. C-10 Factor vs.3 C.EJ Guades Appendix C Variation of impact stress with the height of the tube using finite element (FE) analysis Figure C.. C-10 Factor vs.1 C.16 C.18 C..9 C.5 C. time table simulating material degradation (E480)…………… C-12 Factor vs.13 C. time table simulating repeated impact loading (E420)……….. C-11 Factor vs. C-16 Variation of peak axial strength degradation with number of impacts…… C-17 Absolute peak axial strength degradation at failure……………………… C-18 Comparison of the simulated damaged length at the start of failure…….12 C.11 C. C-2 Material modelling of the composite tube……………………………….19 Figure caption Page Schematic view of the impacted tube and the idealised model………….17 C. time table simulating material degradation (E630)…………… C-12 Factor vs. C-14 Variation of peak axial stress in transverse direction…………………….4 C.

…………………………………………… 73 Test matrix used in defining the impact behaviour……………………….7 Investigation on the behaviour of square FRP composite tubes under repeated axial impact Residual properties of square FRP composite tubes subjected to repeated axial impact Table caption Page Details of the specimen…………………………………………………… 99 Repeated impact test matrix……………………………………………… 100 Details of the specimen for coupon tests…………………………………. 107 Chapter 6 Table 6... GV-S.4 Table caption Comparison of pile impedance………………………………………….2 5. and GV-H tubes ………………………………..1 6..3 5.4 Review of composite piles and their driving performance Damage modelling of repeatedly impacted FRP composite tubes Table caption Page Details of the specimen used in quasi-static compressive test…………… 131 Summary of (Ec)Quasi-static values…………………………………………….5 3.1 3..6 3. 158 7...1 4. 140 Comparison of incident energies at different Nf…………………………. 142 Properties of S125 and R75x100 specimens . 70 Summary of mechanical properties from full scale tests…………………. 149 Chapter 7 Application of the damage evolution model to other types of composite tubes Table Table caption Page 7. 61 Summary of mechanical properties from coupon tests …………………. 142 Comparison of incident energies at average Nf………………………….………………………….... 102 Summary of compression test results…………………………………….3 3.2 3.4 6.2 2. List of applications of hollow FRP pipe piles…………………………… Mechanical properties of the 125 mm square tube……………………… Summary of recent experimental studies on repeated impact test……….....2 4. 159 Behaviour of glass FRP composite tubes under repeated impact for piling application xvii .3 4. 147 Summary of parametric values of square and rectangular tubes.4 5.1 5..3 6.5 6... 107 Summary of tensile and flexural tests results…………………………….2 6.……………………….. 78 Test matrix used in defining the impact damage tolerance……………… 78 Summary of Nf values…………………………………………………….1 2... 44 Details of the specimen for coupon tests ………………………………… 45 Material properties of the tube wall laminate ply ………………………. 42 Details of the specimen for fibre fraction test …………………………… 44 Summary of glass fibre content of each ply …….EJ Guades List of Tables Chapter 2 Table 2. 70 Chapter 4 Table 4..3 2..4 3.6 Characterisation of the properties of composite tubes Table caption Page Details of the specimen …….5 Page 25 31 32 36 Table caption Page Section properties of the 100 mm square tube………………………….1 Details of GV-C... Chapter 3 Table 3.2 Summary of (Ec)Quasi-static and β values of glass/vinyl ester tubes ……….……. 92 Chapter 5 Table 5.

..2 A.... A-6 Summary of results of full scale compressive test for CT1 (L = 100 mm).1 Summary of the dimension of the tubes………………………………….. B-1 Dimension of specimen E320……………………………………………..6 B.... A-2 Summary of results of coupon compressive test for CT2…………………A-2 Summary of results of coupon tensile test for CT1……………………… A-4 Summary of results of coupon tensile test for CT2……………………… A-4 Summary of results of coupon flexural test for CT1……………………..7 A. B-3 Dimension of specimen E480-2………………………………………….10 A.8 7...1 Material properties of the tube wall laminate ply………………………… C-7 C.. 160 Details of glass/polyester tubes (circular cross section)…......8 A.5 A.13 A....2 Summary of applied static load cases used in FE analysis………………... 172 Summary of the repeated impact equation of glass/polyester tubes…. B-1 Dimension of specimen E480…………………………………………….........3 7.....9 B.3 B.. 165 Details of glass/epoxy tubes (circular cross section)…………………...8 B... C-10 Appendix D Summary of specimen dimension and results in residual properties testing Table Table caption Page D.11 Summary of the repeated impact equation of glass/vinyl ester tubes ….10 7.12 A.......6 A....1 A....11 A.. A-1 Summary of results of coupon compressive test for CT1………………. B-2 Dimension of specimen E160…………………………………………….........4 7... B-3 Dimension of specimen E420-1…………………………………………. A-5 Summary of results of coupon flexural test for CT2……………………...3 A.5 7. 171 Summary of (Ec)Quasi-static and β values of glass/epoxy tubes…….1 B....10 Table caption Page Dimension of specimen E630……………………………………………..6 7... A-7 Summary of results of full scale compressive test for CT1 (L = 200 mm)....14 Table caption Page Summary of results of fibre fraction test for CT1……………………….. 163 Details of glass/polyester tubes (square cross section)…..9 A....... 170 Details of glass/epoxy tubes (circular and square cross sections)……..... B-2 Dimension of specimen E210…………………………………………….. B-1 Dimension of specimen E420…………………………………………….. D-1 Behaviour of glass FRP composite tubes under repeated impact for piling application xviii .......4 B..2 B. B-3 Appendix C Variation of impact stress with the height of the tube using finite element (FE) analysis Table Table caption Page C. 172 Appendix A Summary of results of the coupon and full scale tests on CT1 and CT2 specimens Table A........ B-2 Dimension of specimen E630-1…………………………………………... A-1 Summary of results of fibre fraction test for CT2………………………..7 7.... 165 Summary of the repeated impact equation of glass/polyester tubes ….EJ Guades 7.4 A. A-8 Summary of results of full scale compressive test for CT2 (L = 100 mm). A-8 Summary of results of full scale flexural test (3-point loading) for CT1… A-11 Summary of results of full scale flexural test (3-point loading) for CT2… A-11 Summary of results of full scale flexural test (4-point loading) for CT1… A-11 Appendix B Summary of specimen dimension and snapshots of the machine/apparatus used in repeated impact test Table B.. B-2 Dimension of specimen E480-1…………………………………………....9 7... 163 Summary of (Ec)Quasi-static and β values of glass/polyester tubes..5 B...7 B..

D-4 Coupon dimension and compressive test result for E480-10 (Bottom)….32 D. D-3 Coupon dimension and compressive test result for E480-40 (Middle)….. D-2 Coupon dimension and compressive test result for E320-80 (Middle)…..34 D... D-6 Coupon dimension and compressive test result for E740-10 (Bottom)….16 D.30 D..29 D.31 D. D-9 Coupon dimension and flexural test result for E630-30………………….. D-2 Coupon dimension and compressive test result for E480-10 (Middle)….18 D.37 Coupon dimension and compressive test result for E160-80 (Top)……… D-1 Coupon dimension and compressive test result for E320-80 (Top)……… D-1 Coupon dimension and compressive test result for E480-10 (Top)……… D-2 Coupon dimension and compressive test result for E630-10 (Top)……… D-2 Coupon dimension and compressive test result for E160-80 (Middle)…. D-8 Coupon dimension and flexural test result for E320-80………………….28 D.2 D.12 D.EJ Guades D.23 D.9 D.33 D.3 D.10 D.. D-5 Coupon dimension and compressive test result for E480-80 (Bottom)…..25 D. D-5 Coupon dimension and compressive test result for E630-30 (Bottom)…. D-10 Behaviour of glass FRP composite tubes under repeated impact for piling application xix .7 D. D-6 Coupon dimension and tensile test result for E160-80…………………… D-6 Coupon dimension and tensile test result for E320-80…………………… D-6 Coupon dimension and tensile test result for E480-10…………………… D-7 Coupon dimension and tensile test result for E630-10…………………… D-7 Coupon dimension and tensile test result for E480-40…………………… D-7 Coupon dimension and tensile test result for E480-80…………………… D-7 Coupon dimension and tensile test result for E630-30…………………… D-8 Coupon dimension and tensile test result for E740-10…………………… D-8 Coupon dimension and flexural test result for E160-80………………….. D-3 Coupon dimension and compressive test result for E480-80 (Middle)….11 D. D-9 Coupon dimension and flexural test result for E480-80………………….13 D.5 D.26 D.6 D.4 D..27 D.15 D. D-9 Coupon dimension and flexural test result for E480-40…………………. D-3 Coupon dimension and compressive test result for E630-10 (Middle)…... D-10 Coupon dimension and flexural test result for E740-10…………………. D-5 Coupon dimension and compressive test result for E630-10 (Bottom)…. D-3 Coupon dimension and compressive test result for E630-30 (Middle)…. D-4 Coupon dimension and compressive test result for E320-80 (Bottom)…. D-4 Coupon dimension and compressive test result for E740-10 (Middle)…. D-5 Coupon dimension and compressive test result for E480-40 (Bottom)….36 D. D-4 Coupon dimension and compressive test result for E160-80 (Bottom)….24 D.. D-8 Coupon dimension and flexural test result for E480-10………………….21 D.. D-9 Coupon dimension and flexural test result for E630-10…………………..17 D.14 D.19 D..8 D.20 D.35 D.22 D.

parametric constant.EJ Guades Notations Roman alphabets Notation Description A Cross-sectional area of tube/coupon specimen a distance between one of the end supports and the nearest applied load. acceleration at Acceleration as a function of time or at present time increment at-1 Acceleration at previous time increment b Width of the tube/coupon specimen or parametric constant c Neutral axis depth of the tube or parametric constant cw Compression wave velocity D Damage parameter d Depth of the tube E Modulus of elasticity Eabs Absorbed energy Ec Critical energy (energy causing the failure of tube at one impact) Ecomp Compressive elastic modulus of tube/coupon specimen (Ec)Dynamic Critical energy obtained from dynamic (impact) test Ef Flexural elastic modulus Eim Impact energy Ein Incident energy EK Kinetic energy EP Potential energy (Ec) Quasi-static Critical energy obtained from quasi-static compressive test Esat Saturation energy Et Tensile elastic modulus ET Total energy Ews Energy as a function of displacement Ewt Energy as a function of time Fs Load at present displacement increment Fs-1 Load at previous displacement increment Ft Impact load as a function of time g Acceleration due to gravity h Drop height h0 Drop height (used in Appendix C) j Inner depth of the tube k Inner width of the tube l Length of the tube /coupon specimen ls Test span in flexure L Length of the tube (used in Appendix C) Mg Fibre glass content in mass percentage m Mass of the impactor mc Critical impact mass Behaviour of glass FRP composite tubes under repeated impact for piling application xx .

0025 Life duration Behaviour of glass FRP composite tubes under repeated impact for piling application xxi .0005 stress measured at the strain values ε2 = 0.EJ Guades m0 m1 m2 m3 mm N Nf Nmax Ppc Ppf Ppt (Pm)0 (Pm)N I Ix Iy t R(Nf) ri re sm st t t–1 v vff vm v0 vt vt-1 z Initial mass of the specimen used in fibre fraction test Initial mass of the dry crucible used in fibre fraction test Initial mass of the dry crucible plus dried specimen used in fibre fraction test Final mass of the crucible plus residue used in fibre fraction test Equivalent mass at the mth point (used in Appendix C) Number of impact Number of impacts to initiate failure/collapse of the tube Maximum number of impact Peak compressive load of tube/coupon specimen Peak flexural load of tube/coupon specimen Peak tensile load Maximum load at the 1st impact Maximum load at the Nth impact Moment of inertia Moment of inertia along the x-axis Moment of inertia along the y-axis Thickness of the coupon specimen Reliability of Nf Internal radius of the chamfered corner of the rectangular tube External radius of the chamfered corner of the rectangular tube Travelled distance by the wave at the mth point (used in Appendix C) Displacement as a function of time Present time increment Previous time increment Impact velocity Volume of the specimen used in fibre fraction test Wave velocity at the mth point (used in Appendix C) Initial velocity of the impactor before hitting the target Velocity as a function of time or at present time increment Velocity at previous time increment Pile impedance Greek letters Notation α β εpc ρ ρt σpc σpf σpt σ1 σ2 θ Description Ratio of the loading rates between quasi-static compressive and impact tests Correlation factor Peak compressive strain of tube or coupon specimen Mass density/specific mass Mass density of the tube (used in Appendix C) Peak compressive stress of tube or coupon specimen Peak compressive stress Peak tensile stress stress measured at the strain values ε1 = 0.

The damage evolution model was verified through experimental investigation on a 100 mm square pultruded tube. An improved understanding of the behaviour of glass fibre FRP composite tubes under repeated axial impact is expected from this study. The model was applied to composite tubes of different cross sections and materials made from vinyl ester/polyester/epoxy matrix reinforced with glass fibres. However. The mechanical properties of the tubes used in investigating the impact behaviour of the tubes and their residual properties were obtained experimentally and using finite element (FE) analysis. steel and timber reduces their structural capacities. The costs associated with the repair and rehabilitation Behaviour of glass FRP composite tubes under repeated impact for piling application 1 . Theoretical prediction using the basic energy principle was performed to gain additional understanding on the damage evolution behaviour of composite tubes subjected by repeated axial impact. The deterioration of concrete. 1. 2002).. there are problems associated with their use especially when installed in corrosive and marine environments. and marine borer attack or deterioration of timber piles. which may ultimately result in damage or failure of the structure (Iskander and Stachula. steel and timber as pile foundations. Examples of deteriorated traditional piles in harsh environments are shown in Figure 1. steel corrosion.2 Background Pile foundations are generally used to support structural loads in situations where soil settlement is a major concern or where shallow foundations cannot provide the required bearing capacity (Sakr et al. 2004).Chapter 1 – Introduction EJ Guades Chapter 1 Introduction 1. Piling industry has historically involved the use of traditional materials such as concrete.1.1 General This thesis presents the results conducted to investigate the behaviour of fibre reinforced polymer (FRP) composite tubes under repeated axial impact loading. These include concrete degradation. The effects of impact event (incident energy and number of impact) on the instantaneous response and the residual properties of composite tubes were examined.

au) Figure 1. and with disruption to the public’s use of facility.S. (a) Degraded concrete pile (www. there is a growing concern in the environmental and health impact of using treated pile materials. Lampo et al. steel and timber piles costs the U. Conclusively.Chapter 1 – Introduction EJ Guades of deteriorated piles. These problems coupled on the use of traditional piles led researchers around the world to look for viable alternative materials that are suitable in harsh environments. 1998). Similarly.1 Problems of traditional piles installed in harsh environments Behaviour of glass FRP composite tubes under repeated impact for piling application 2 .substructure. 2003). (1997) estimated that the deterioration of concrete.watimas.majorprojects.gov. Aside from the cost.. Creosote and Copper Chromium Arsenic (CCA) treated timber pose a threat to marine life and the workers who handled during manufacturing and installation are in potential health risk (Iskander et al. military and civilian marine and waterfront communities nearly $2 billion a year.com) (b) Corroded steel piles (www. can be very high and in certain circumstances often exceed the original of the construction cost (Neff.. 2007).com) (c) Deteriorated timber piles (www. using same material in the rehabilitation and replacement of these deteriorated traditional piles is not an optimum solution as apparently the cycle of inherent problems of their usage will just be repeated. steel treated using sandblasting or painted with solvent and heavy-metal containing coatings are potentially harmful to the environment and are increasingly being regulated (Lampo et al.vic.

pipelines. and low maintenance cost (Sakr et al. 2002). The application of “composite piles” was first recorded in the late 1980’s (Iskander and Hassan. the cost of labour and use of equipment necessary for construction work may be lower due to their lighter weight. rapid advances in construction materials technology have enabled civil engineers to achieve impressive gains in safety. Composite piles refer to alternative pile foundations composed of fibre reinforced polymers (FRP). (2006) suggested that not only should costs be compared on a total installed first-cost basis but also on a reasonable total life cycle cost basis. This disadvantage. Iskander and Hassan (1998) reported that manufacturers claim their composite piles may last twice as long as treated wooden piles. corrosion resistance. mainly due to their tailorable performance characteristics. chemical and environmental resistance. although the cost of FRP composite materials may be higher. their initial cost is generally expensive compared to traditional pile materials. 2005). high specific strength. bridges. Pando et al. and low cycle costs (Einde et al. however. Composite pile materials may exhibit large deformations in excess of Behaviour of glass FRP composite tubes under repeated impact for piling application 3 . Their application is of most importance in the renewal of constructed facilities infrastructure such as buildings. as the overall cost is expected to decrease as composite piles gain wider penetration in the civil engineering industry. 2006).. is relative. 1994). 2002). Most composite piling manufacturers believe their products may be competitive when compared to the life cycle cost of traditional piles in some applications (Ballinger. high durability. and functionality of structures built to serve the common needs of the society (Bakis et al. The application of FRP composite materials in piling system is relatively new compared to other civil engineering applications.. 2003). They are considered viable alternatives due to their inherent advantages over traditional piles. economy.. First.3 Fibre composites as an alternative in piling applications In the last 200 years. etc. Their advantages include light weight. On the other hand. recycled plastics or hybrid materials that are placed into the ground to support axial and/or lateral loads (Pando et al.. For instance.Chapter 1 – Introduction EJ Guades 1. Ballinger (1994) emphasised that. The second drawback of using composite piles is due to their inherent low modulus. Their use has also increased in the rehabilitation of concrete structures. there are also potential drawbacks of using composite piles. These include the consideration and application of fibre reinforced polymer (FRP) composite materials in civil engineering. Iskander and Hassan (1998) enumerated four disadvantages of using composite piles. ease of application.

An attempt has already been performed by Pando et al. thus. however. and hollow FRP pipe pile are generally considered to be potentially suitable in load-bearing applications. Steel pipe core piles have some integrity issue since cracking on the plastic shell is imminent after it was installed (Lampo et al.Chapter 1 – Introduction EJ Guades the settlement permitted by the codes. The third drawback of using composite piles is that their long-term performance under increasingly larger structural loads is not yet well defined.S. Traditional piles are considered too stiff for fendering application. 2002). The detailed discussion on the impedance properties of composite piles is presented in Chapter 2. Although composite materials present a number of disadvantages related to their application in piling system. The fourth disadvantage of their use is that composite piles are generally less efficient to drive than traditional piles. their low modulus property provides an added advantage of their use. the use of composite piles is still an alternative that will eliminate deterioration problems of traditional piling materials in waterfront environments and aggressive soils (Iskander et al. They can also dampen seismic forces transferred to the structure through the foundation and they may reduce moments in piled rafts (Iskander and Hassan. 1998). 1998). Moreover. Ashford and Jakrapiyanun... In some situations. 2000. Their poor driving performance was attributed by their inherent low impedance property (Mirmiran et al. U. steel pipe core pile. Impedance is associated to the ability of the pile to transmit the energy imparted by the driving hammer into the ground (Pando et al. (2006) to monitor the long-term performance of two composite piles located in Route 351 Bridge in Virginia. structurally reinforced plastic (SRP) pile.. concrete-filled FRP pipe pile. 2005. 2001). Behaviour of glass FRP composite tubes under repeated impact for piling application 4 . 2006) although techniques are being developed to minimise its occurrence (Baxter et al. 2001). Common problems of these piles are debonding between the component materials (Mirmiran et al. fender. Previous studies conducted on the first three types of composite piles.. however. Among composite piles. making composite piles an ideal potential use. Pando et al.A. and bearing piles for light structures (Iskander and Hassan. showed some concerns of their use in this application.. 1.4 FRP tubes as composite piles Composite piles have been used in ports and harbours primarily as waterfront barriers. 2002. their low modulus (or stiffness) property may cause problems during installation and handling. 2006).. Fam and Rizkalla.

The cost of transportation and installation is also lower due to their lighter weight. Compared to concretefilled FRP pipe pile.. On the other hand. 2002). pultruded composite tubes were used as fibre composite bridge decking unit. The impedance characteristics of hollow FRP pipe piles are inherently material-dependent and therefore increasing it may not be simple.. thus more cost efficient. their integrity and post-impact performance is in question. This challenge is attributed to the techniques on how they are being placed into the ground. doubling the wall thickness would essentially double the impedance.Chapter 1 – Introduction EJ Guades 1998)...e. For instance. hollow FRP pipe piles are commonly installed using impact driving. Unfortunately. As a result. 1998) and structural performance due to its inherent excessive deformation behaviour.5 Challenges in using hollow FRP pipe piles One of the main challenges in the efficient use of composite piles is to ensure that they can carry the intended design loads and be installed to the necessary depth. In this installation technique. the cross-sectional area of the hollow FRP pile can be increased. hollow FRP pipe pile can be readily installed without the intricacy of placing concrete infill using additional equipment. For instance. 2002). as transmission line cross arms. Additionally. This method drives a pile by raising a weight between guideposts and dropping it on the head of the pile. and as a major structural component of a fibre composite bridge girder (QDMR. delamination between FRP shell and concrete core in the case of concrete-filled FRP pipe pile) is not an impending issue on the use of this pile. 2006). Due to this rupture. SRP piles have issues on handling and installation (Iskander et al. 1. the last type of composite piles is considered a comparably good option in piling application and is the focus of the present study. square-shape composite tubes bonded to an FRP plate are used as a structural decking in a flooring system (Bakis et al. Like other types of composite piles. The emergence of FRP composite tubes as a structural component provided the industry to consider these materials as a potential composite load-bearing pile type since they can carry design load. hollow FRP pipe piles were found to exhibit poor driving performance due to their low impedance. In Australia. since fibre Behaviour of glass FRP composite tubes under repeated impact for piling application 5 . bond failure (i. Field test results showed that their thin-walled section generally shatters under high driving stresses when encountering sand layer or boulders (Mirmiran et al.

Chapter 1 – Introduction EJ Guades materials are the primary cost in the manufacture of hollow FRP piles (Ashford and Jakrapiyanun. the feasibility of adopting these alternative driving techniques to hollow FRP piles has not been implemented yet in actual field condition. The elastic modulus can be varied by the fibre orientation. Increasing the impedance by working on the material parameters such as specific mass. 1.. the effect of the specific mass on the impedance of hollow FRP pile is not straightforward. Characterisation of the impact damage behaviour of fibre composite materials is highly important as they exhibit distinctive damage characteristics compared to traditional materials.. 2001). Recently. 2001). 2001). there is still a need to confirm this method using field tests in various subsurface conditions. Since the result is based on experimental investigation in a laboratory facility. Few recommended installation techniques include using steel mandrel to essentially drag the pile into place or to use high-frequency vibratory driver (Mirmiran et al. The fibre composite materials of the hollow FRP pipe piles are susceptible to impact damage using the current installation method. driving them requires more careful consideration due to their relatively low stiffness and thin walls. Sakr et al. Similarly. analytical study showed that varying the fibre orientation still not sufficient to increase significantly the modulus leading to the increase of impedance (Ashford and Jakrapiyanun. Ashford and Jakrapiyanun. The compression wave velocity of the hollow FRP piles is directly related to their modulus of elasticity (Iskander et al. However. 2002. Aside from the fact that it is difficult to increase due to their inherent lightweight characteristics. elasticity and area is not an optimum solution to enhance the driving performance of hollow FRP piles. (2004) developed a driving technique called toe driving to install the hollow FRP piles into granular soils. So far. 2001). This driving method was carried in a laboratory where the large-scale model hollow FRP pile was driven in dense dry sand enclosed in a pressure chamber. When they are subjected to impact loading. doubling the wall thickness could also nearly double the cost. Working on some aspects such as driving installations may also found to improve their driving performance. On the other hand. increasing it would results to only minimal contribution as this parameter will also reduce the wave velocity.6 Research needs related to their driving performance Hollow FRP pipe piles have poor driving performance due to their inherent low impedance. there Behaviour of glass FRP composite tubes under repeated impact for piling application 6 .

This damage may have an adverse effect on the structural integrity and post-impact performance of the composite materials. (b) Investigate the residual (after-impact) properties behaviour of repeatedly impacted square GFRP composite tubes. Previous studies related to the driving performance of hollow FRP piles only include superficial consideration of the impact behaviour of the fibre composite materials and does not systematically describe their impact strength. issues related to the determination and prevention of impact induced damage on fibre composite materials become more important. the effects of the damage parameters such as impact energies and number of impacts on the behaviour of fibre composite materials have not been fully investigated. drop height/velocity. Consequently. These studies described the impact behaviour of the fibre composite materials through the observed damage mechanisms only. Literature revealed that the studies on the behaviour of FRP composite materials subject to repeated impact are limited only to composite laminates/panels or tubes under transverse impact. impact mass. and there is a need to develop an understanding of damage phenomena at the materials level. The aim of this study is to investigate the behaviour of glass fibre reinforced polymer (GFRP) composite tubes under repeated axial impact. a need to conduct a study on the behaviour of composite tubes under repeated axial impact is of prime priority.Chapter 1 – Introduction EJ Guades might be no damage indication on the surfaces by visual evaluation but internal damage may have occurred (Zhang and Richardson. Therefore. The effects of these parameters should be clearly understood to determine whether by varying their magnitude results in significant changes on their impact strength and damage behaviour. 1. 2007). and impact repetitions on the impact behaviour of square GFRP composite tubes experimentally. Moreover. and Behaviour of glass FRP composite tubes under repeated impact for piling application 7 . The main objectives of the study are the following: (a) Characterise the effects of energy. (c) Develop prediction model on the impact damage evolution of square GFRP composite tubes.7 Objectives The evaluation of the impact behaviour of fibre composite materials is significant to describe their driving and post-impact performance.

matrix material). however. The composite tubes used in this study are commercially manufactured using pultrusion process.e. a general idea on pultrusion process referenced from the literature is presented. and  Application of the proposed model to these FRP composite tubes. The following are considered during the progress of the study. On the other hand. the manufacturing process of these pultruded tubes are not publicised. Similarly.  Testing and evaluation of the residual properties behaviour of the repeatedly impacted square GFRP composite tubes. Behaviour of glass FRP composite tubes under repeated impact for piling application 8 . geometry. Due to commercial sensitivity.  Characterisation of the material properties of square GFRP composite tubes. and impact studies on FRP composite materials.Chapter 1 – Introduction EJ Guades (d) Investigate the potential application of the proposed damage model to other GFRP composite tubes with different properties (i.  Review on the driving performance of composite piles and recent development on hollow FRP pipe piles. further information on the details of fibre and matrix cannot be revealed. 1.  Testing and evaluation of the behaviour and failure mechanisms of square GFRP composite tubes under repeated axial impact.8 Scope of the thesis The study focused on understanding the behaviour of GFRP composite tubes subject to repeated impact for piling application. the following are beyond the scope of this study and considered potential areas of research in the near future:  Evaluation of the behaviour of GFRP composite tubes under actual pile driving considering the effect of soil..  Development of an energy-based model predicting the impact damage behaviour of square GFRP composite tubes.  Investigation of the energy absorption behaviour of other FRP composite tube materials (from quasi-static compressive tests and directly from the literature).  Comparison of the results from experiment and proposed damage model.  Characterisation of the behaviour of GFRP composite tubes under lateral impact.

 Chapter 1 gives an introduction and objectives of this study. the residual properties of composite tubes were characterised by determining the residual properties of the coupons cut from the impacted tube for each impact condition. As this work emphasised the use of hollow FRP pipe piles in load-bearing applications.Chapter 1 – Introduction EJ Guades 1. This also discussed the significance of the study relative to the use of FRP composite tubes in piling application.  Chapter 7 discusses the potential application of the proposed model on GFRP composite tubes with different properties. drop mass. The studies worldwide on the impact behaviour of FRP materials as a research need related to their driving performance are highlighted. The manufacture of composite tubes using pultrusion process is discussed. and the number of impacts are highlighted in this chapter. This chapter highlights the existing problems and the motivation of conducting an impact study on FRP composite materials.  Chapter 2 provides an overview on composite pile technologies and their behaviour under impact driving. In this study. This model adopts an energy-based approach in simulating the damage evolution curve.  Chapter 5 emphasises on the characterisation of the residual properties of the square FRP tubes under repeated axial impact.  Chapter 4 characterises the behaviour of the square GFRP composite tubes through experimental investigation.  Chapter 3 presents the characterisation of the mechanical properties of square GFRP composite tubes used in this study. the recent development of their applications is also presented. The effects of impact energy. These include tubes with Behaviour of glass FRP composite tubes under repeated impact for piling application 9 . drop height/velocity.  Chapter 6 covers the development of a predictive model to characterise the damage evolution of a repeatedly impacted square GFRP composite tubes. A finite element (FE) analysis on the compressive and flexural behaviours of full scale specimens was also included in the discussion.9 Outline of the thesis This thesis contains 8 chapters in which each describes the different investigations conducted in this study.

Therefore. and circular sections. Behaviour of glass FRP composite tubes under repeated impact for piling application 10 . rectangular. Driving these piles. 1. The possibility of damaging the fibre composite materials during the process of impact driving is always a concern. This motivated the author to conduct this study.  Chapter 8 presents the main conclusions of the research and recommendations for future work. and tubes with matrix material made of polyester and epoxy. tubes with bigger dimension. there is a need to understand the impact behaviour of these materials in order for them to be safely and effectively driven into the ground. however.Chapter 1 – Introduction EJ Guades square. steel and timber in harsh environmental conditions.10 Summary Hollow FRP pipe piles have been a viable option in replacing traditional pile materials such as concrete. One of the main factors that affect the driving performance of these piles and needs special attention is the impact strength of the fibre composite materials. requires more careful consideration due to their relatively low stiffness and thin walls.

These piles were composed of steel pipe core encased by recycled plastic shell and used for fendering applications.1 General This chapter provides an overview on the types of composite pile and their driving performance used in replacing traditional piles. The pile was formed by 6 m segments each connected by a threaded coupling. fibreglass reinforced plastic piles. The description and applications of each type of composite piles are presented in the following subsections. 2.1 Steel pipe core piles Steel pipe core piles consist of two layers.2.2 Types of composite piles The application of composite piles was first recorded in the United States (US) when they were used in April 1987 at Berth 120 in the port of Los Angeles (Horeczko. This type of pile is available in 200 to 600 mm Behaviour of glass FRP composite tubes under repeated impact for piling application 11 . To date.Chapter 2 – Review of composite piles and their driving performance EJ Guades Chapter 2 Review of composite piles and their driving performance 2. there are seven types of composite piles. and FRP sheet piles. The studies worldwide on the impact behaviour of FRP materials as a research need related to their driving performance are highlighted. is currently the manufacturer of this type of composite pile in the US (Iskander and Stachula. The cross sectional view of this pile is shown in Figure 2. 2002). As this work emphasised the use of hollow FRP pipe piles in load-bearing applications. fibreglass pultruded piles. hollow FRP pipe piles.1a. structurally reinforced plastic piles. 2. These include steel pipe core piles. 1995). The 18 m long pile has a 330 mm diameter recycled plastics and 125 mm diameter steel pipe core. The inner layer provides the structural strength while the outer shell (commonly made from high density polyethylene (HDPE)) is used to protect the steel from corrosion.1). Plastic Piling Inc. an inner steel layer and thick outer plastic shell (Figure 2. concrete-filled FRP pipe piles. the recent development of their application is also presented.

steel pipe core piles are also considered potentially suitable for load-bearing applications. Steel core HDPE plastic (a) Cross section (Baxter et al. The recycled materials are usually from waste plastic such as plastic milk jugs.1b shows the application of steel pipe core piles in this environment. The structural pipe cores range from 100 to 400 mm outer diameter. soap bottles and juice containers (Lindsay. (2006).2 Structurally reinforced plastic piles Structurally reinforced plastic (SRP) piles are composed of extruded recycled plastic matrix reinforced with fibreglass rods or steel rebar (Figure 2.2..Chapter 2 – Review of composite piles and their driving performance EJ Guades outer diameter and up to 23 m long.. USA (Lampo et al.1 Steel pipe core piles 2. 1998) Figure 2. Figure 2. These piles were observed to have cracks at the plastic shell surface a year after they were installed (Lampo et al. According to Pando et al. There was relatively little need for further research on this kind of pile since the design procedure of steel pipe piles is well established.2). The most common use of this type of pile is in fendering applications in region with marine influence and change of the tide. the design procedure of this type of composite pile would be essentially the same as for the traditional steel pipe pile if the plastic shell is used only in the upper portion of the pile that is exposed to water. 1998).. 2005) (b) Application in Tiffany Street Pier. Figure 2. NY. with wall thicknesses between 6 and 40 mm. However. The piles are available in diameters between 254 and 430 mm Behaviour of glass FRP composite tubes under repeated impact for piling application 12 . Early applications of this product suffered from delamination of the steel core from the plastic shell due to the difference in thermal stresses (Iskander and Hassan. 1996). 2002). SRP piles are produced using continuous extrusion process which allows manufacturing of up to 32 m long.2a shows the cross section of the pile.

This type of piles exhibits larger deflections under axial and lateral load (Pando et al..Chapter 2 – Review of composite piles and their driving performance EJ Guades and are reinforced with 6 to 16 pieces of FRP or steel reinforcing rods of diameters ranging from 25 to 35 mm (Baxter et al. 2006). On the other hand..2 Structurally reinforced plastic piles 2. 2006) and causes problem during installation and handling due to their excessive deformation (Iskander and Hassan. SRP piles are mainly used in fendering applications and regarded as potential load-bearing piles. Figure 2. Fibreglass rods HDPE plastic (a) Cross section (Baxter et al. 1996). gives confinement to concrete in compression. 1998) Figure 2.3a shows the cross section of these piles. One version of this pile is structurally reinforced by a steel cage with the rebars welded to a continuous steel spiral (Pando et al. Typically.3 Concrete-filled FRP pipe piles Concrete-filled FRP piles are comprised of an outer FRP shell with unreinforced concrete infill (Figure 2... 2002).3).. Figure 2.2b illustrates the use of SRP piles in fendering application. USA (Lampo et al. Behaviour of glass FRP composite tubes under repeated impact for piling application 13 . 2000). 2002). NJ.2. 2005). This structural system found to perform better than the equivalent prestressed and reinforced concrete structural members under combined axial and flexural loads (Mirmiran et al. 2005) (b) Application in Port Newark.. and protects the concrete from severe environmental effects (Mirmiran and Shahawy. acts as non-corrosive reinforcement. Problems associated with these piles include the possibility of debonding of the reinforcing FRP rods and high creep rate related with the high polymeric content. the concrete infill offers the internal resistance in the compression zone and increases the stiffness of the member and prevents local buckling of the FRP tube (Fam and Rizkalla. The FRP shell provides a stay-in-place structural formwork for the concrete infill.

These include the roughening of inside shell surface by applying thin layer of epoxy sprayed with course silica (Fam and Rizkalla.2. FRP shell Concrete infill (a) Cross section (Baxter et al. 2005).. USA (Pando et al. In August 3.4a. 2006) Figure 2. Recently. VA.6 to 9. New Jersey (Iskander and Hassan.Chapter 2 – Review of composite piles and their driving performance EJ Guades concrete-filled FRP piles are available in diameters ranging from 203 to 610 mm. 2000). 1996..4)..4b (Lampo et al.. USA (Pando et al. 2006).. Fibreglass pultruded piles were also used in Tiffany Street Pier project as shown in Figure 2. plastic lumber.1 mm (Pando et al. Figure 2.3b illustrates concrete-filled FRP piles being adopted in a bridge rehabilitation projects in Virginia.. 2002). The grid inserts are sometimes filled with HDPE. or polyethylene foam fills. The grid consists of two sets of orthogonal plates joined at four intersecting points and forms a tic-tac-toe pattern. The cross sectional view of fibreglass pultruded pile is presented in Figure 2. 2005) (b) Application in Route 40 Bridge. 2002) and application of bonding agents (Baxter et al. These piles were used as fender piles in 1996 in a demonstration project at Berth 7 in Port Newark. with wall thicknesses ranging between 4.3 Concrete-filled FRP pipe piles 2.. techniques and fabrication process were developed to minimise the occurrence of delamination.4 Fibreglass pultruded piles Fibreglass pultruded piles are composed of outer fibreglass sheet fitted with a fibreglass grid to provide structural strength (Figure 2. 2006). 1998). a major fire occurred Behaviour of glass FRP composite tubes under repeated impact for piling application 14 . The HDPE shell and fibreglass inserts are used to absorb vessel impact in fendering applications. These piles are suitable for both fendering and load-bearing applications. An impending concern in using these piles is the interface bonding and delamination problem between FRP shell and concrete core (Mirmiran et al.

5a displays the cross section of fibreglass reinforced plastic piles. sound barriers. Figure 2. NY.Chapter 2 – Review of composite piles and their driving performance EJ Guades on the Tiffany Street Pier. FRP shell FRP grids Plastic inserts (a) Cross section (Baxter et al.5 Fibreglass reinforced plastic piles Fibreglass reinforced plastic piles consists of recycled plastic matrix with randomly distributed fibreglass reinforcement (Figure 2. (1998) reported that the plastic lumber inserts and the polymer matrix material in the tic-tac-toe profile section were consumed in the fire. railings and fender piles.. Figure 2. 2002)... 2007) and US Plastics (US Plastics. Various additives can be mixed with the plastic materials to enhance the performance of the structural member. 1998) Figure 2. 2007) are manufacturer of this product consisting of high density extruded recycled polyethylene reinforced with approximately 20% fiberglass. The suitability of using fibre reinforced plastic piles in load-bearing applications in this project has not been studied since they did not undergo testing for bearing piles (Lampo.2. fungicides and compatibilizers. Lampo et al.5 m. UV protectors. Piles are available in 250 mm diameter with a standard length of 7. USA (Lampo et al. car stops. These additives include antioxidants. 2005) (b) Application in Tiffany Street Pier.4 Fibreglass pultruded piles 2. The dense solid outer shell is bonded to the peripheral surface of the inner plastic core which is foam-filled to reduce weight. walkways. 1998) Behaviour of glass FRP composite tubes under repeated impact for piling application 15 . Trimax produce a variety of structural members that conform to lumber industry standards (Iskander and Hassan. Fibre reinforced plastic piles are commonly applied as retaining walls. 1998). colorants.5). et al. Trimax (Trimax Building.5b shows tapering of Trimax piles used in the construction of the Tiffany Street Pier in New York City (Lampo et al.

the significant features and issues of their usage are discussed in details in Section 2.5 Fibreglass reinforced plastic piles 2. 2002). pultrusion. Hollow FRP pipe piles are typically consist of a thermosetting matrix reinforced with glass fibres forming a tubular section made either by filament winding. USA (Lampo et al. Hollow FRP pipe piles are considered potentially suitable in load-bearing applications. or resin transfer moulding process.Chapter 2 – Review of composite piles and their driving performance EJ Guades Plastic matrix w/ fibreglass (a) Cross section Trimax Building. NY.6 Hollow FRP pipe piles Hollow FRP pipe piles are an outer shell component of a concrete-filled FRP composite system.. Figure 2.2.6 shows the different sections of the piles used in the application. The diameter and wall thickness of these piles can be varied up to 460 mm and 22 mm. As this paper gives emphasis on this type of composite piles. 1998) Figure 2. UV and chemical attacks (Iskander and Hassan. (a) Circular section (b) Square section Figure 2. respectively. Some versions of these piles are coated with acrylic to protect against abrasion.6 Geometry of hollow FRP pipe piles used in the application Behaviour of glass FRP composite tubes under repeated impact for piling application 16 .4. 2007 (b) Application in Tiffany Street Pier.

2006) (b) FRP pultruded profiles used as retaining walls (www. widths from 400 to 460 mm. (a) Cross section (Shao. 2007).apeesmarlan.7a.7).. The available products on the market have section depths of 100 to 350 mm. Earlier study on composite sheet piles includes recycled HDPE in tongue-and-groove profile reinforced with chopped glass fibres as potential material (Lampo et al. 2002).com) Figure 2. 1998). and wall thicknesses from 4 to 12 mm (Shao..Chapter 2 – Review of composite piles and their driving performance EJ Guades 2. The cross sectional view of FRP sheet piles is presented in Figure 2. The problem associated in using FRP sheet piles includes possible damage at their corners caused by ice impact and rubbing if installed in cold regions (Dutta and Davinder. the asymmetrical shapes typically seen for FRP sheet piles make the testing of these materials more difficult than for many other commonly produced structural shapes (Lackey et al. 1997). The single unit corrugated profile is composed of a symmetric double Zcross section.7 FRP sheet piles FRP sheet piles are typically made of FRP pultruded sections with corrugated-shape profile (Figure 2. FRP sheet piles in general is used for a wall that resists horizontal loads (Figure 2. As opposed to the other type of composite piles which carry vertical axial load.7 FRP sheet piles Behaviour of glass FRP composite tubes under repeated impact for piling application 17 . Additionally. 2006). FRP sheet piles found to be increasingly used as waterfront retaining structures for both new installations and rehabilitations (Marsh.2.7b).

composite piles used in load bearing and fendering applications are commonly installed using impact driving. namely: (1) the energy delivered to the pile by the pile driving hammer. and (4) the strength of the pile to resist driving stresses (Pando et al..Chapter 2 – Review of composite piles and their driving performance EJ Guades The information of the reported studies on composite piles suggested that composite piles are mostly used in fendering and in retaining-structure applications. As this paper emphasised impact driving performance.1 Types of driving hammer and its effect Driving hammers play a significant role in successfully driving composite piles. Good driving occurs when the hammer effectively transmits energy to the pile and the induced stress wave develops a force in the pile sufficient to overcome the soil resistance. These studies also showed that most composite piles are suitable in load-bearing application as proven by some bridge rehabilitation and replacement projects. Pile driveability refers to the ability of a pile to be safely and economically driven to support the required bearing capacity and possibly to a minimum required penetration depth (Hussein et al. the transmission of waves will not be effective if the stresses induced by the hammer during driving are higher than the impact strength of the pile as damage will be created. However. (2) the resistance to driving offered by the soil.3 Driving performance of composite piles One of the main challenges in the efficient use of composite piles is to ensure that they can carry the intended design loads and be installed to the necessary depth. FRP sheet piles can be placed into the ground using several methods (www..pdf). other installation methods such as vibratory hammer and water jetting equipment which are normally used in driving FRP sheet piles are not included in the discussions. (3) the ability of the pile to transfer driving stresses to the pile tip. 2006). This challenge is attributed to the techniques on how they are being placed into the ground. 2.com/LitLibrary/sheetpile/sheetpileinstalls.pultrude. However. pile driveability depends on four significant factors.3. In general. It is therefore a requirement for effective driving to Behaviour of glass FRP composite tubes under repeated impact for piling application 18 . The role of these factors in the driving performance of composite piles will be discussed in the following subsections. Research studies are still undergoing to support the full utilisation of load-bearing composite piles. 2. 2006).

This study revealed that both hammers showed similar effect on the driving performance of composite piles when initially driven in medium stiff clay. SRP and hollow FRP composite piles) was studied by Iskander et al..89.e. and driving system using a system of concentrated weights that are connected by linear elastic springs (Iskander and Stachula. Key finding of this study is that single-acting steam hammer is more efficient than the diesel hammer as it can drive the composite piles deeper at same number of blows. single-acting steam and open-ended diesel hammers.55.e. medium or heavy hammer in each of the three types of soils. cushion and pile properties.86 kJ energy output. their influence was apparent as the piles reach the medium dense sand layer. 30.Chapter 2 – Review of composite piles and their driving performance EJ Guades choose driving hammers not only suitable for the soil conditions but also should be appropriate for the specific pile materials. Result of their study showed that both hollow and concrete-filled FRP piles can be driven by heavier hammers to a higher depth. 2002). Behaviour of glass FRP composite tubes under repeated impact for piling application 19 . (2001) using wave equation analysis. however. (2002) conducted a parametric study to determine the effect of light. the former cannot attain more than 40-50% of the capacity of the latter. The variation of their driveability becomes more pronounced under heavier hammers. This analysis incorporates the effects of hammer weight and velocity. The effect of the types of hammers in driving composite piles (i. Mirmiran et al. Each pile was theoretically driven with light. however. medium and heavy (i. two soil profiles and at the two driving depths for different magnitude of soil resistance using a software program Microwave. soil. 73. Study on the effect of driving hammers on the driving performance of composite piles is rather sparse because of their novelty. still considered heavier hammers to be more efficient than light hammers in driving as they can drive these composite piles deeper at the same blow count. it was found that heavier hammers induce much larger stresses compared to light hammer. The analysis is based on discretising the pile. The significance of the types of hammer is more apparent in driving SRP piles compared to hollow FRP piles.. Nonetheless. 1986). Two types of driving hammers were considered in this study. respectively) single-acting diesel hammer on the driveability of hollow and concrete-filled FRP piles. When driving concrete-filled FRP piles. and 158. as compared to light hammer. The composite piles were virtually driven in a soil profile composed of two layers of medium stiff clay and medium dense sand. Mirmiran et al. and the dynamic behaviour of the soil during driving (Fenske and Hirsch.

there is more substantial difference in friction and end-bearing conditions for concrete-filled FRP and concrete piles. steel sheets have the axial capacity to support the hammer weight and effectively transfer energy through the pile for penetration. This indicates that the installation procedure and pile driving machine for steel sheet piles can be used successfully with the FRP sheet pile. (2002) using wave equation analysis of piles (WEAP) program. However. It is important that the resistance of the hammer-pile-soil system should overcome the resistance that the soil can offer in order to achieve effective driving.Chapter 2 – Review of composite piles and their driving performance EJ Guades Impact hammers are suitable to break or shear the skin friction bond between the pile and soil especially for cohesive soil. (2008) revealed that FRP sheet piles possess similar dynamic response to that of their steel counterpart. sand and silt) and two embedment depths were considered in their study. The effect of the side friction and end bearing resistance in different soils during composite pile driving was studied by Mirmiran et al. Moreover. two soil profiles (with 90% of the total capacity is provided by end bearing and the rest contributed by friction in a triangular distribution along the length of the pile. It is also noteworthy that the effect of driving hammer should be associated with the impact strength of the composite pile materials as it contributes on effective driving.3. Behaviour of glass FRP composite tubes under repeated impact for piling application 20 . Whilst the area considered in side friction is a function of embedded pile length. In this case. and the other with 10% of the capacity is afforded by the end bearing) were adopted. their emphasis is directed more on the transmissibility of stress waves induced by the hammer. The resistance of the soil in driving are attributed by the components of the static pile capacity which are the frictional resistance on the side and the end bearing on the tip of the pile. However.2 Resistance to driving offered by the soil Pile driving constitutes substantial penetration of piles to dense sand layers or other strong soils. Three types of soils (clay. 2. Results from the entire spectrum of their study showed that there is no significant difference between the driveability of the hollow FRP piles in different soil profiles. Boscato et al. end bearing resistance utilised the cross section of the pile as the effective gross area. The result of the studies provided valuable information on the influence of driving hammers on the driving performance of composite piles.

simple shear test. A number of studies characterising the interface behaviour between FRP materials and soils using interface shear test are already available. but also on the resistance in driving. 1999). and silica powder in the IST apparatus. The shear test was performed at a horizontal displacement rate between 0. The FRP composite tubes were fabricated using different material constituents and manufacturing techniques. torsion or ring shear test.08 mm/min. This study also revealed that the angularity of sand particles was seen to be influential on the behaviour of interfaces as angular materials have higher interface friction coefficient than rounded materials. glass beads. For FRP materials and soils.Chapter 2 – Review of composite piles and their driving performance EJ Guades The side frictional resistance at the interface between the pile material and the surrounding soil represents a considerable element not only on the compression and uplift capacity of the pile. FRP exhibited similar relationships between the peak interface friction coefficients and the relative roughness for a given granular material. the friction coefficient increased linearly with increasing relative roughness. 1999).25 to 5. Their study involves testing of these materials with sands (Valdosta and Ottawa). Outcome of the study showed that the interface friction coefficient between FRP and sand decreases as the mean grain size increases. This finding implies that large particles have lower friction angle than smaller grains with the same mineralogy when a mass particle slides on identical rough surfaces. This resistance can be experimentally obtained using direct shear test. In comparison with steel materials. On the other hand.1 mm. whilst FRP composite tube 2 (glass/vinyl ester) has an outside diameter of 612 mm and wall thickness of 9. Interface shear test was performed by Pando et al.2 mm. the determination of shear resistance between them is generally obtained using interface shear test (IST). and pull-out test (Frost and Han. On the other hand. FRP composite tube 1 is made from glass/polyester materials with an outside diameter of 629 mm and wall thickness of 7. silica with sub-rounded to rounded grains) and model sand Behaviour of glass FRP composite tubes under repeated impact for piling application 21 . (2002) in investigating the frictional resistance among sand and two commercially available FRP materials. IST refers to tests using a modified direct shear apparatus to study the shearing of granular materials on the surface of the FRP or steel materials (Frost and Han. Normal stresses ranging from about 25 to 175 kPa were used in the shear test. Frost and Han (1999) assessed the friction between sand and FRP and steel materials. This test was also performed on a prestressed concrete pile for comparison with the FRP materials. density sand (fine to medium grained.

the outside diameter of FRP composite tube 2 is 178 mm with wall thickness of 7. This finding made them to conclude that the economical interface shear test can be used efficiently to capture the skin friction characteristics of FRP piles driven in granular soils. Whilst in interface shear test it utilises only a coupon cut from the FRP tube. In comparison with the material types. The interface shear test adopted an applied normal stresses between 23 to 200 kPa. Test results indicated that the relative roughness parameters and angularity of the soil significantly influence the interface friction coefficient as previously found out by Frost and Han (1999).35 mm thick cylindrical steel open-ended tube having a diameter of 168. the interface is more constraint so that the values of the interface coefficient are higher and the shear failure tends to occur by ploughing of the soil grains along the material surface. Unlike the two previous studies. when soil grains do penetrate into the contact material. found a result similar to that obtained by Frost and Han (1999) and Pando et al. (2005) studied the interface friction of FRP materials and fine subround to round air-dried sand. The shells of the FRP composite tubes were both made of glass/epoxy materials and manufactured using filament winding technique. They reported that the pile capacity obtained from uplift loading test compared reasonably well with those calculated from interface shear test. which produces more complete interlock of the soil grains with its surface as compared to the FRP materials.4 mm with wall thickness of 6 mm. this study characterised the interface friction using interface shear and uplift pile load tests. Sakr et al. Pando et al. On the other hand. On the other hand. emphasised that the concrete material has the highest value of the interface friction angle because of its rougher surface topography.3 mm was also tested to serve as a reference for comparison of result. They commented that shear failure at the interface tends to occur by sliding of the soil grains along the material surface when the soil does not penetrate. (2002) that the relative roughness of the FRP composite material has a significant effect on the FRP/sand interaction behaviour.8 mm. the whole tube undergoes testing in uplift load test. A 6.Chapter 2 – Review of composite piles and their driving performance EJ Guades (consisted of fine grained sand with sub-angular to angular grains) were used as a granular soil. The values of the peak interface friction angle for the two FRP materials/dense Behaviour of glass FRP composite tubes under repeated impact for piling application 22 . This study also showed that surface hardness found to have significant effects on the interface friction values for a relatively smooth FRP surfaces. Sakr et al. FRP composite tube 1 has an outside diameter of 162.

3 m long. The greater the impedance of the pile. the concrete-filled FRP pipe pile (315 mm outside diameter) and steel pipe core pile (390 mm outside diameter) are 16. it is apparent that not only the impedance has the direct influence on the ability of pile to transfer driving stresses but also other parameters such as the mass density. 2001). Alternatively.3. according to Sakr et al. cross sectional area. Behaviour of glass FRP composite tubes under repeated impact for piling application 23 . A is the cross sectional area.1) where z is the impedance. cw can be calculated using Equation 2.Chapter 2 – Review of composite piles and their driving performance EJ Guades sand (26 and 310.2) where E is the composite modulus of elasticity. respectively. 2. and compressional wave velocity of the pile materials. cw = (E/ρ)1/2 (2. for FRP composite tube 1 and 2. ρ is the mass density. made the use of FRP materials in deep foundation more advantageous due to their increased shaft frictional resistance in addition to their resistance to degradation. Iskander & Stachula (2002) predicted the effect of modulus of elasticity on the driveability of three types of composite piles using WEAP. These results. the greater is the force that will be transmitted by the pile into the soil.2 (Rausche et al. show that section with high composite modulus is easier to drive than the lower ones. However.1 and 2. The result showed that the variation of modulus of elasticity has virtually no effect on the driveability of the concrete-filled FRP pipe pile.. and cw is the compression wave velocity. if not higher than. On the other hand.1 (Ashford and Jakrapiyanun.5 m. respectively) were similar to. The 400 mm diameter SRP pile has a length of 27.2. In the case of piles with relatively lower modulus. according to them. modulus of elasticity. Literature shows a number of studies conducted on the effects of these parameters on the driving stress transferring mechanism of composite piles. the friction angle for the steel/dense sand (26. z = ρAcw (2. The details of the other input data on these composite piles are outlined in their paper. Pile impedance can be defined mathematically by Equation 2. As seen in Equations 2.8 and 18.60). the modulus has large influence on the driving performance of both SRP and steel pipe core piles.3 The ability of the pile to transfer driving stresses The capability of the pile to transmit the energy imparted by the driving hammer into the ground is associated to its impedance or dynamic stiffness.. This result. 1988).

they reported that buckling may happen when the surrounding soil is very soft or when a large portion of the pile extends above the ground. 2000).1 that among composite piles. It is apparent from Table 2. which may result in large shear deformations. The sensitivity of unit weights on the driveability of composite piles was also characterised in the study of Iskander and Stachula (2002). In comparison with concrete-filled FRP pipe piles. respectively. Its value is approximately similar with prestressed concrete pile and significantly higher than the other two traditional piles. the steel pipe core and the SRP piles impedance values are about 65% and 38%. Damping ratio has no effect on concrete-filled FRP piles but slightly influenced the driveability of SRP and steel pipe core piles.1 summarises the typical impedance values of three traditional piles and four selected composite piles with approximately similar outside diameters. Relatively. On the other hand. damping ratio has no major influence on the driveability of composite piles compared with the modulus of elasticity and unit weight. Table 2. The deflection of the pile is always larger when the shear deformation is considered (Han and Frost.Chapter 2 – Review of composite piles and their driving performance EJ Guades buckling is imminent under extreme loading conditions during their installation by driving or when they are subjected to superstructure loads (Han and Frost. concretefilled FRP pipe pile has the highest impedance value. The lowest impedance value Behaviour of glass FRP composite tubes under repeated impact for piling application 24 . FRP composites generally have anisotropic material properties and high elastic to shear modulus. (2006) and Mirmiran et al. It should be noted that the impedance values indicated in the table are calculated using Equation 2. Note that FRP sheet piles are not included in the table due to their totally different geometric configuration and application compared to other composite piles. In a parametric study on the effect of shear deformation on buckling of vertically loaded FRP composite piles conducted by Han and Frost (1999). (2002). unit weight has little influence on the driveability of concrete-filled FRP pipe piles as their weight is well-defined just like traditional piles. it is expected that their driving performance will behave similarly. The result showed that the unit weight of the pile is a major factor in driving SRP and steel pipe core piles and further highlights the significance of quality control during manufacturing. 1999). This expectation was confirmed experimentally in some studies conducted by Pando et al. As the impedance of both concrete-filled FRP pipe and prestressed concrete piles is comparable. Another parameter that is directly related to impedance is the damping characteristics of pile materials.1.

900 ρ (kg/m3) 2.4 Strength of the pile to resist driving stresses The strength of the composite piles in resisting driving stresses is attributed to their axial impact response characteristics and energy absorption behaviour.900 7.787 z (kg/s)x103 710 340 mm ø steel pipe pile (9.1 Comparison of pile impedance A (mm2) 77.406a E (GPa) 34.652 692 254 mm ø steel pipe core pile 11. Table 2.644 265 1. impact Behaviour of glass FRP composite tubes under repeated impact for piling application 25 . Filling the empty pipe by a denser material such as concrete would provide extra mass and cross sectional area.4a 2. In general. For composite piles.000 pile (13 mm wall thickness) a values from Iskander and Stachula (2002) b values from Ashford and Jakrapiyanun (2001) 2.5a c (m/s) 3.000 2. (2002). the low impedance value of hollow FRP piles caused these piles to endure much higher stresses and to get damaged as observed in their field tests.3. It should be noted that these characteristics are associated to the impact fatigue response of the composite piles since they are repeatedly impacted.5 mm wall thickness) 9.048 392 356 mm ø timber pile 99.048 448 406 mm ø SRP pile (reinforced with FRP tendons) 129.500 815a 13. However. composite piles have low impedance values than traditional piles.455 93 Pile type 315 mm ø prestressed concrete pile 356 mm ø hollow FRP pipe 14.8a 4.240b 31b 3.849b 200b 5. According to Mirmiran et al.114 334 325 mm ø concrete-filled FRP pipe pile 83.849b 200b 5.Chapter 2 – Review of composite piles and their driving performance EJ Guades corresponds to the hollow FRP pipe pile with 13% that of the concrete-filled FRP pipe piles.300 7. although making the pile costly and heavier for transportation.927b 23b 3.500 770a 5. Composite materials are inherently characterised by their low mass densities that would be rather difficult to increase substantially. their impedance can be improved by increasing the mass density and the cross-sectional area.

Mirmiran et al. (a) Concrete-filled FRP pipe pile (b) Hollow FRP pipe pile Figure 2.3 m without damage at the top. (2005) to investigate their behaviour under impact driving. The concrete-filled FRP pipe pile has an outside diameter of 250 mm with an FRP shell thickness of 20 mm. concrete-filled FRP pipe and SRP piles were tested by Baxter et al. Unlike traditional piles. It was believed that the ruptures began when the pile encountered sand layer. the concrete core and FRP shell worked in composite action and the integrity of the system was not compromised.g.8b). Their objective is to compare the behaviour of the two composite piles under actual field driving impacts.85 m long open-ended single acting diesel hammer. This indicated that despite of impact repetitions induced on the pile.1 m long concrete-filled FRP pipe pile were successfully driven to depths at about 7.5 m (Figure 2.8 Condition of the composite piles after driving (Mirmiran et al. and no separation between the concrete and FRP shell (Figure 2. The 9. the 337 mm diameter SRP pile was Behaviour of glass FRP composite tubes under repeated impact for piling application 26 . Mirmiran et al. FRP shell and concrete infill in the case of concrete-filled FRP pipe pile). the impact fatigue response of composite piles is not yet clearly defined. 2002) In another field study. (2002) field-driven hollow and concrete-filled FRP pipe piles using a 3.. (2002) observed that approximately 1 m of the tube at the top crumbled and peeled off. Formation of fronds and vertical cracks at the top of the pile is also apparent from the observed damage. the top of the hollow FRP piles was found to be damaged after it was driven to a depth of 3.Chapter 2 – Review of composite piles and their driving performance EJ Guades stresses imparted by the driving hammer are generally resisted by the composite action between the component materials (e. The FRP tubes adopted for the composite piles had an outside diameter of 348 mm with a wall thickness of 14 mm. On the other hand.8a). On the other hand..

9b). Upon inspection.Chapter 2 – Review of composite piles and their driving performance EJ Guades fabricated from recycled plastics and is reinforced by 8-25 mm diameter steel rebar.3 m and were driven using hydraulic hammer with a rated energy of 7. The damage was characterised by cracking and spalling of concrete core at the top.35 mm. USA.4 kJ.2 kJ. Another observation they reported is that the diameter at the top of the pile was significantly increased from 337 mm to 368 mm. They also noticed that the steel reinforcement at the pile tip was exposed as a result of damage on the plastic material. Baxter and his associates observed that the top of the pile was visibly broken at the end of driving. (2006) in a two separate projects in Virginia.5 m and recorded a blow count of 6 blows per 25 mm at the end of driving. Driving of SRP pile runs smoothly until embedment depth of 1. and disintegration of some portion of FRP shell (Figure 2.1 m long concretefilled FRP pipe pile has an outside diameter of 625 mm with an FRP shell thickness of 7. Unlike the result reported by Mirmiran et al. (a) Concrete-filled FRP pipe pile (b) Structurally reinforced plastic pile Figure 2. The Behaviour of glass FRP composite tubes under repeated impact for piling application 27 . However. The 13. This pile was driven by a hydraulic impact hammer with a rated energy of 85..9a).9 Condition of the composite piles after driving (Baxter et al. The concrete-filled FRP pipe pile was successfully driven to a depth of 8. Both composite piles have a length of 7. 2005) Composite piles were driven by Pando et al.8 m. This damage according to them was attributed by the energy imparted by the hammer or by the generation of heat from the driving equipment itself. the driving was eventually halted at an approximate depth of 2 m when no advances were observed. on concrete-filled FRP pipe piles. The first project involves driving of concrete-filled FRP pipe pile to replace the damaged concrete piles in Route 40 Bridge. they observed that the top 1 m of the pile bent out of vertical alignment by slightly more than 3 degrees (Figure 2.

(a) Driving process (b) Driven concrete-filled FRP Pipe pile Figure 2. The damage on the tip of the SRP pile.10b). The SRP pile was driven to a depth of up to 17. The second project also involves driving of “enhanced” concrete-filled FRP pipe pile. This enhanced composite pile has an outside diameter of 622 mm and an FRP wall thickness of 10. On the other hand.27 m below the original ground level. A single acting diesel hammer with a maximum energy rating of 106.5 m embedment depth did not cause any significant damage on the pile.10 Composite pile installed in Route 40 Bridge (Pando et al. however. on the top portion of the SRP pile after driving is not present as evidenced by Figure 2.8 kJ was used in the driving process.11b).11a.. The 592 mm diameter and 18. This enhanced concrete-filled FRP pipe pile was successfully driven to a depth of 7. This study revealed that driving of this composite pile up to 8. Neither cracking/spalling of concrete infill nor rupture on the FRP shell were observed during the driving process (Figure 2. Enhancement of this composite pile was achieved by providing additional 14-25 mm diameter reinforcement bars aside from the FRP composite shell. 2006) Behaviour of glass FRP composite tubes under repeated impact for piling application 28 . was not investigated as no extraction was undertaken after driving.7 mm. The result of the study indicated that the damage observed by Baxter et al. The information on the length of the pile and the hammer used in driving of the enhanced composite pile are similar to that of the SRP pile.3 m long composite pile is made from medium density polyethylene material and reinforced by steel cage (2425 mm diameter rebars).35 mm without significant damage on its component materials (Figure 2. the second project includes driving of SRP pile as test pile near Route 351 Bridge.10a.Chapter 2 – Review of composite piles and their driving performance EJ Guades driving process is shown in Figure 2.

it was found that the dominant damage behaviour of steel pipe core piles when repeatedly impacted was delamination between the steel core and the plastic shell. The influence of impact energy and impact cycles needs to be considered as they are significant not only in their driveability but also in their post-impact performance characterisation. The effects of impact energy.Chapter 2 – Review of composite piles and their driving performance (a) SRP piles EJ Guades (b) Enhanced concrete-filled FRP Pipe pile Figure 2. have not been investigated in detail. substantial amount of research are needed in this area. and FRP sheet piles) were not available. 1998). Even so. For the past decade. Noticeably. FRP sheet piles were observed to be susceptible to damage from transverse stresses that hammers induced. as well as the impact cycles.. fibreglass reinforced plastic. General finding of the studies on the driving performance of composites piles suggests that they are less efficient to drive than traditional piles. This poor driving performance affects their integrity and limits their application. Similarly. records of either a static or dynamic load test on these piles were not reported (Lampo et al. studies on composite piles had been mostly focused on their use in load-bearing Behaviour of glass FRP composite tubes under repeated impact for piling application 29 . Nevertheless.. The reported studies described the impact behaviour of composite piles through the observed damage mechanisms only. 2006) Although steel pipe core piles have been used in many locations.11 Composite piles driven near Route 351 Bridge (Pando et al. the driving records of the other types of composite piles (fibreglass pultruded.

the driving performance of these composite piles was not investigated since the installation technique does not require the method of impact driving. (4) jacking the pile to the underside of the headstock. 2. In this study. The studies conducted on hollow FRP pipe piles are summarised in Table 2. Behaviour of glass FRP composite tubes under repeated impact for piling application 30 . 1998). This project used an innovative technique for the repair of damaged timber piles in Shorncliffe Pier in Brisbane (Figure 2.2. 2011). The replacement project is a collaborative effort between the Centre of Excellence in Engineered Fibre Composites (CEEFC) of the University of Southern Queensland and BAC Technologies Pty. The procedures of this concept can be summarised as follows and can be found from this link (http://www. Ltd. and concrete-filled FRP piles since they are considered potentially suited for load-bearing applications Lampo et al.Chapter 2 – Review of composite piles and their driving performance EJ Guades applications. These piles provide a solution in this particular application with the added advantages over other potential load-bearing composite piles. The FRP tube wall is consists of 26 layers constituting an overall thickness of 22 mm. they were adopted in replacing damage timber piles and as bearing pile for light structures in Australia. The outside diameter of this pile is between 300 and 470 mm and was manufactured using resin infusion method. (Sirimanna. The aim of this technique is to replace the deteriorated upper portion of a pile in an existing bridge or pier by hollow FRP pipe pile without the need to remove or modify the bridge superstructure.4 Recent developments on hollow FRP pipe piles Most of the applications of hollow FRP pipe piles are as test piles or in theoretical studies.html): (1) identifying the serviceable section of the existing pile and removing the upper portion. The complete description and material characteristics of the composite tube used is presented in the work of Sirimanna (2011).net. SRP.bac. Recent developments on hollow FRP piles for various structural applications suggest their high potential as load-bearing piles. (2) locating an FRP tubular connector on the pile stump. and (5) injection of epoxy grout or other fasteners to complete the installation. (3) inserting the hollow FRP pipe pile into the connector. The recent developments and research needs related in understanding the driving performance of these piles in load-bearing applications are discussed in the following section.12).. The hollow FRP pipe pile is made of vinyl ester resin reinforced by glass fibres and XF Soric. Recently.au/futurepile. These studies mainly discussed steel pipe core.

2 Field test Test piles b 6. hIskander and Stachula (2001) Figure 2.html.5 4 Field test Support for light structures c 294 22 7...bac.2 Analytical test n/a g Circular 356 7. These projects utilised composite Behaviour of glass FRP composite tubes under repeated impact for piling application 31 .13).http://www. Ltd.3 Undriven/ field test Load bearing piles d Circular 162 5 1.Chapter 2 – Review of composite piles and their driving performance EJ Guades Table 2. Queensland. g Ashford and Jakrapiyanun (2001).12 Hollow FRP pipe piles replacing deteriorated timber piles (Courtesy of BAC Tech.au/futurepile. eSakr et al.2 18&27 Analytical n/a h test a 125 mm square section. cAravinthan and Manalo (2012). Pty.net. fMirmiran et al. (2004). 2002. Australia) Hollow FRP pipe piles were also adopted to shore up boardwalks located in New South Wales and Queensland (Figure 2.2 Laboratory test Test piles e Circular 348 14 7.9 Field test Test piles f Circular 356 13 12. dSirimanna (2011).2 List of applications of hollow FRP pipe piles Circular Outside diameter (mm) 475 Wall thickness (mm) 22 Square n/aa Circular Section geometry Length (m) Type of test Nature of application Sources 9. bwww.

3 shows the mechanical properties of the pultruded tube.Chapter 2 – Review of composite piles and their driving performance EJ Guades tubes manufactured by Wagners Composite Fibre Technology (WCFT). transverse (MPa) 104 Shear strength (MPa) 84 Modulus of elasticity.13 Pultruded composite tubes used in shoring-up boardwalks (Courtesy of WCFT.50 mm. longitudinal (MPa) 650 Tensile strength. Queensland.400 Modulus of elasticity. transverse (MPa) 41 Compressive strength. Starting from the exterior of the wall. Queensland Figure 2. Australia using pultrusion process. The information on the installation process and their behaviour under impact driving is presented in the next paragraph. Mackay.900 Field driving of square hollow FRP pipe piles were lately undertaken in Australia. The 125 mm square pultruded tubes were driven to support an elevated Behaviour of glass FRP composite tubes under repeated impact for piling application 32 . Tweed Heads. the stacking sequence of the plies is in the form of [00/+450/00/-450/00/-450/00/+450/00]. (a) Boardwalk project under construction. New South Wales (b) Finished boardwalk project. The tube wall is consisted of nine plies with a total thickness of 6. transverse (MPa) a WCFT Product Specifications 12. Table 2. Australia) Table 2. The 125 mm square pultruded tubes are made from E-glass fibres and vinyl ester resin. where the 00 direction coincides with the longitudinal axis of the tube. longitudinal (MPa) 35. longitudinal (MPa) 550 Compressive strength.3 Mechanical properties of the 125 mm square tubea Tensile strength.

While attempts have been conducted to demonstrate the driveability of hollow FRP pipe piles made of pultruded square tubes.14 Impact driving of 125 mm square pultruded tubes (Courtesy of Wagners CFT. It was observed that most of the tubes were successfully driven without damage. Queensland. On the other hand.Chapter 2 – Review of composite piles and their driving performance EJ Guades walkway in Tweed Heads. The 4 m long pultruded tubes were driven to a depth of 2.14a displays the installation process of the 125 mm square pultruded tubes using impact driving. (2002) when this type of composite pile of circular cross section was impact-driven. and formation of vertical cracks at the corners (Figure 2. It should be noted that the presence of ±450 glass fibre reinforcement on the tube provided a stronger structural resistance along the transverse direction. These cracks were noticed to be concentrated only on the portion in contact with the impact mass.13). no geotechnical data was obtained on the sites where the field tests were carried out and no instrumentation was considered. In this test. however. if not suffered minimal damage only in a form of cracks along their wall (Figure 2. is generally common to hollow FRP pipe piles as this was also the observation of Mirmiran et al. no systematic study has been conducted so far that will provide a general understanding on their behaviour under impact driving.14c). This damage induced during impact driving. Australia) Behaviour of glass FRP composite tubes under repeated impact for piling application 33 . (a) Driving rig (b) Undamaged tubes (c) Crushed tubes Figure 2. This unique property made this tube suitable for structural application particularly as hollow FRP pipe pile.5 to 3 m using a 1-ton impact hammer. The damage at the top of the tube was characterised by lamina splitting. New South Wales (shown in Figure 2. it was observed that the head of few driven tubes were crashed during impact driving.14b). fibre breakage. Figure 2.

Figure 2. The final stages of pile driving involved the maximum energy from the pile rig with the hammer dropping at 800 mm. it was noticed that during this test regime.au/futurepile.15 illustrates the impact driving procedure of the composite pile.7 m. Queensland. however. Pty.bac. the pile would bow like a string every time the hammer strikes the top of the pile. did not attach any instrumentation on the pile that will provide additional information on its behaviour during impact driving.. The 9. Following test driving. However.html) shows that the pile was driven through very stiff to sandy clay with an SPT N value of >50. Driving resistance started to develop when the bottom end of the pile reached a depth between 4 to 4.15 Impact driving of 475 mm diameter hollow FRP pipe pile (Courtesy of BAC Tech.2 m long composite pile with a wall thickness of 22 mm was made of vinyl-ester resin reinforced by glass fibres. Information from the manufacturer (http://www.net. the pile was visually inspected with no significant damage identified. Driving of this pile runs smoothly until embedment depth of 4 m. At this stage. respectively).162 kN (815 and 1. Figure 2. Queensland. timber ply cushion was broken although it was observed that no sign of damage on the top of the pile.Chapter 2 – Review of composite piles and their driving performance EJ Guades Field driving of a 475 mm diameter circular hollow FRP pipe pile was undertaken in Wilkie Creek. Australia) Behaviour of glass FRP composite tubes under repeated impact for piling application 34 . The CAPWAP method was used for the analysis and determined the geotechnical capacity of the foundation to be 2. Ltd.347 kN shaft and toe resistance. This pile was effectively driven to a depth of 6 m using a 9 tonne impact hammer. This test. The pile was left to settle for approximately 48 hours before dynamic pile driving analysis was completed to check the capacity of the foundation and also identify any structural damage.

The laminate was subjected by a hemispherical head impactor by impact energy between 1. They stated that the impact fatigue life exceeds 500 impacts up to 13 J. These studies. A brief description of these studies and their corresponding key results are presented in the next paragraphs. and not through plastic deformation (Mamalis et al. are focused on composite plate/laminates or tubes which are transversely impacted. The characterisation of the impact behaviour of fibre composite materials is definitely of great importance to define the driveability and post-impact performance of hollow FRP pipe piles. Table 2. (2011) investigated the repeated impact behaviour of selfreinforced polypropylene composite using instrumented falling mass tests. Additionally. 1996). unlocking this information may also yield an opportunity to improve their poor driving performance and their optimum use.49 J. On the other hand..Chapter 2 – Review of composite piles and their driving performance EJ Guades 2.5 Study on the impact behaviour of FRP composite tubes as a research needs relative to their driving performance The optimum use of hollow FRP pipe piles is being challenged due to their poor driving performance and lack of design guidelines of their installation. large amount of energy may be absorbed (Richardson and Wisheart. they emphasised that the strain Behaviour of glass FRP composite tubes under repeated impact for piling application 35 . The summary of these studies are presented in Section 2. but drops sharply for 14 J. however. Furthermore. Aurrekoetxea et al. the behaviour of the FRP composite materials under impact driving has not been fully characterised since the studies related to their driving performance only described their impact behaviour through the observed damaged mechanisms only. The result indicated that the nature of the laminate is highly anisotropic with strain hardening failure.6.4 shows the summary of these studies. Composite materials on the other hand are brittle and can only absorb energy in elastic deformation and through damage mechanism. owing to the ductile behaviour of the materials. 2. Impact damage is generally not considered to be an issue in metal structures because.6 Behaviour of FRP composite plates/laminates repeatedly impacted or tubes under repeated transverse impact Studies on the behaviour of FRP composite plates/laminates subject to impact repetitions or tubes under repeated transverse impact using experimental investigation have already been reported. Research on the behaviour of FRP composite materials under repeated impact has been extensive. 2006).

Table 2.13 Glass/vinyl ester Glass/polyester Falling mass 392 40 David-West et al. They found that the curves of damage evolution against the number of repeated impacts to fracture the composites revealed three distinct zones: fibre micro buckling and shear fracture Behaviour of glass FRP composite tubes under repeated impact for piling application 36 . (2011) Laminate Composite thickness material (mm) 2. (2010) 6. 6 ø rod Glass/vinyl ester Pendulum 0. bno data on matrix material..4 Summary of recent experimental studies on repeated impact test Researcher/s Aurrekoetxea et al.1 (2004) Glass/epoxy Carbon/epoxy Kevlar/epoxy Falling mass 14. ccomposite rod 10.2 Self-reinforced polypropylene Type of Impact No.Chapter 2 – Review of composite piles and their driving performance EJ Guades hardening is the origin of the trend of peak load increase and plastic deformation decrease with impact events.16 10. (2008) 3 Carbon prepegsb Falling mass 5. (2005) 4.18 Found and 0. (2003) Glass/epoxy Falling mass 50 40 Carbon/epoxy Falling mass 0.70 98 Hosur et al.65 3. (2009).98 c (2001b) 180 long type a maximum value..000 3. (2008) 10. of a impact test energy (J) impactsa Falling mass 20 500 Sevkat et al.580 Belingardi et al.35 Glass/epoxy Graphite/epoxy Falling mass 32 69 Coban et al. (2009) 2 Carbon/polyethe rimide Pendulum type 2.000 Roy et al.27 Carbon/epoxy Falling mass 7.8 Howard (1995) Roy et al. (2001a) 4 ø rod Carbon/vinyl 180 longc ester The deformation characteristics of thermoplastic matrix composites during repeated impact loading were investigated by Coban et al.50 1500 Sugun and Rao 2.87 20 De Morais et al.93 100 Pendulum type 0.

5 and 1 m.16 to 2. Belingardi et al. reported that no significant differences existed in the force and energy curves in which no perforation happened. irrespective of the reinforcing fibre used.Chapter 2 – Review of composite piles and their driving performance EJ Guades of fibres (1st region).42 mm in which they are subjected by a 765 g impactor dropped from a height between 0. carbon and Kevlar composites using a range of impact energies. initiation and propagation of delamination and matrix deformations (2nd region). Under this condition. 2010) and can be explained as a result of compaction process (Wyrick and Adams. In another study. The thickness of the laminates adopted in their study ranges from 1. They also reported that the rate of damage progression in the event was characterised by an equation from the energy profile that correlates the propagation energy and time.. the experimental points of all tested laminates fall in a single curve. the number of impacts to perforation varied in harmonic series. They reported that the symmetric plate with different ply directions proved to have the best resistance to impact. Sugun and Rao (2004) characterised the impact fatigue behaviour of glass. Coban et al. anti-symmetric. and propagation of delamination and fibre cracking and pull out especially in tensile area (3rd region). 2004). They also Behaviour of glass FRP composite tubes under repeated impact for piling application 37 . carbon and aramid fabrics reinforced composites for two levels of energy impacts. (2008) investigated the response of glass reinforced laminates under repeated impacts using impact energy of up to 392 J. the cross section of the laminate is the most relevant variable that determines the impact resistance. and asymmetric) under repeated low energy level impacts was characterised by DavidWest et al. As the incident energy was varied in arithmetic progression. reported that the intensive deformation observed in compression region during impact-fatigue loading is due to lower compressive strength of composites as compared to their tensile strength. The result showed that the maximum peak force sustained by the laminate is usually not reached in the first impact. De Morais et al. They reported a numerical relationship between the impact energy and the number of impacts to perforation. The behaviour of a balanced laminates (symmetric. The results obtained from their tests show that below a certain energy level. Belingardi et al. This phenomenon has been reported by other researchers as well (Sevkat et al. They found that the impulsive force was influenced by stacking sequence and the crack path through the laminate. 1998) or change of dominant damage mode (Liu. (2005) evaluated the influence of laminate thickness on the resistance to repeated low energy impacts of glass. (2008).

93 J. laminates were subjected to repeated impact loading up to maximum of 40 impacts at energy levels ranging from 10 to 50 J. and the number of impacts affect the behaviour of composite laminates or tubes which are transversely impacted.Chapter 2 – Review of composite piles and their driving performance EJ Guades emphasised that the peak load decreases while the total energy increases until the perforation of the composite laminates. explain the impact fatigue behaviour of the studied composite tubes. Under this study.000 cycles with an energy level between 0. incident energy (or drop mass/height/velocity). The outcome of their study revealed that the damage caused by repeated impacts at energies of 0. respectively. This study was conducted by impacting the tubes up to 10.73 J did not produce changes in the peak impact force. This mode of failure was also observed by Roy et al.5 m whilst the mass was varied to produce a wide range of impact energies between 0. a second impact at 0. Furthermore on the impacted carbon fibre reinforced tubes. (2003) investigated the damage resistance of stitched/unstitched S2-glass/epoxy composites. However. Found and Howard (1995) performed repeated impact tests on carbon FRP laminates using drop-weight impact rig. respectively. They also pointed out that the absorbed energy showed similar trend with respect to number of impacts. Their analysis on the fractured surface of the tube revealed debonding. (2001a).54 and 0. They reported a sudden drop of peak load after a certain number of impacts at an energy level between 40 and 50 J. (2001b) when they subjected fibreglass-reinforced composite tubes under repeated lateral impacts. Impact tests were conducted from a height of 0. It would be equally Behaviour of glass FRP composite tubes under repeated impact for piling application 38 . they noted that the presence of few macro-cracks and an increased volume of microcracks in the matrix with damaged fibres at the high and low impact endurance regions. The result indicated an existence of a plateau region in the impact fatigue behaviour curve between 10 and 100 cycles immediately below the single cycle impact strength. The behaviour of high and medium strength carbon fibre-vinyl ester composite tubes under repeated transverse impacts was studied by Roy et al. Hosur et al. Literature review shows that parameters such as impact load. This was followed by a progressive endurance with decreasing impact loads terminating at an endurance limit at about 71% and 85% of the single impact strength for high and medium strength composite tubes.93 J produced a significant reduction in the peak force and an increase in impact duration.06 to 0.16 J using pendulum-type impact apparatus. fibre breakage and pull-out at the tensile zone of the impacted samples.54 and 0.

Consequently. and the impact strength of the pile materials are the main factors that affect the driving performance of composite piles. Literature shows that the studies on the behaviour of FRP materials under repeated impact are mostly focused on composite laminates/panels or tubes under transverse impact. However. the behaviour of FRP composite tubes under repeated axial impact was characterised. hollow FRP piles show high potential in load-bearing applications. Their effect however on the driving performance of hollow FRP piles are not fully investigated. information on the behaviour of FRP composite tubes under repeated axial impact is very limited. the possibility of damaging the fibre composite materials during the process of impact driving is still imminent. and environmental friendly.7 Conclusions Composite piles have longer service life. to conduct a research on their impact behaviour. the characterisation of the impact behaviour of this tube will apparently extend its usage to piling application. the pile impedance.Chapter 2 – Review of composite piles and their driving performance EJ Guades important to know on how these parameters affect the behaviour of composite tubes when they are axially impacted. Therefore there is a need to conduct a study on composite tubes that will characterise their behaviour when they are subjected by repeated axial impact. There is a need. require less maintenance. Just like other types of composite piles. Further research studies on the impact behaviour of this type of composite pile ranging from materials to full-scale levels should be conducted to understand their driving performance. In this research. To date. Behaviour of glass FRP composite tubes under repeated impact for piling application 39 . 2. These piles provided significant advantages in terms of cost efficiency and structural capabilities. resistance offered by the soil. These inherent characteristics made them a viable option in replacing traditional piles in harsh environmental conditions. therefore. the existence of ±450 glass fibre reinforcement contributed to a stronger structural performance of the tube. Specifically. Therefore. The information acquired from the impact damage evaluation of the composite tubes will definitely lead to a better understanding of the impact performance of hollow FRP pipe piles. It was found that the type of driving hammers used. these piles have not yet gained wide acceptance because of the lack of design guidelines especially on their installation techniques. The investigated tube is suitable for structural application since glass fibre reinforcements are provided in several directions.

Therefore. an investigation into the material characteristics of the adopted composite tubes and their manufacturing process is presented.Chapter 2 – Review of composite piles and their driving performance EJ Guades The presence of ±450 glass fibre reinforcement on the investigated tube contributed to a better structural performance. the investigation on the impact behaviour of this tube apparently lengthens its structural usage specifically in piling application. The information provided from this investigation will provide a more systematic understanding on the impact behaviour of fibre composite materials and eventually help researchers and engineers in developing installation guidelines for their optimum use and wider application. Behaviour of glass FRP composite tubes under repeated impact for piling application 40 . In Chapter 3.

. The tube used in the experiments presented in Chapters 4 (i. Basically. the compressive.. however. and the flexural properties of the tubes are investigated. tensile. an idea on this process sourced from the literature is provided. Australia.1 General This chapter presents the characterisation of the properties of the FRP composite tubes adopted in this study. the process of manufacturing of these tubes in the site is not revealed. The 100 mm square tubes are made from vinyl ester resin with E-glass fibre reinforcement and manufactured using the process of pultrusion.e. 3.Chapter 3 – Characterisation of the properties of FRP composite tubes EJ Guades Chapter 3 Characterisation of the properties of FRP composite tubes 3. impact behaviour of pultruded tube) and 6 (i. a finite element analysis was carried to simulate the compressive and flexural behaviour of full-scale specimen. Their Behaviour of glass FRP composite tubes under repeated impact for piling application 41 . prediction model on impact behaviour of pultruded tube) are same and is designated as Composite Tube 1 (CT1).e. CT1 is an older version of the pultruded tubes relative to CT2.2 FRP composite tubes under study The composite tubes tested in this study are manufactured by Wagners Composite Fibre Technology based in Toowoomba. The detailed information on this process is presented in the next section. Both have comparable physical and mechanical properties as evidenced by the test results presented in this chapter. The discussions on the technical description and chemical composition of the glass fibre and the matrix materials are not included due to commercial confidentiality.. This study used two types of 100 mm square pultruded tubes as provided by the manufacturer. residual properties of pultruded tube). Likewise. Moreover. This has been included to demonstrate the feasibility of using FE method in predicting its mechanical behaviour to eliminate the need of repeating costly arrangements for experimental tests.e. On the other hand. Specifically the fibre content. the tube designated as Composite Tube 2 (CT2) is adopted in Chapter 3 (i. Tests on coupons and full scale specimens were undertaken to determine the mechanical properties of the tubes.

1 Oblique view of the composite tubes Table 3.1. CT1 has green colour texture whilst CT2 is white finished (Figure 3. b (mm) 100 Nominal thickness.86 Moment of inertia. 2002).75 External radius. Iy (106 mm4) WCFT Product specification 2.1 Section properties of the 100 mm square tubea b ri re d t Nominal depth.2 illustrates the Behaviour of glass FRP composite tubes under repeated impact for piling application 42 .1).3 Manufacturing of tubes using pultrusion process The process of pultrusion in manufacturing FRP composite tubes provides both product consistency and economy (Bakis et al.932 6 4 Moment of inertia..25 Internal radius.Chapter 3 – Characterisation of the properties of FRP composite tubes EJ Guades only difference is the colour texture. Figure 3. d (mm) 100 Nominal width. t (mm) 5. Ix (10 mm ) 2. (a) Composite tube 1 (CT1) (b) Composite tube 2 (CT2) Figure 3. The section properties of the tubes are shown in Table 3. re (mm) 10 Gross area (mm2) 1. ri (mm) 4.86 a 3.

com/pultrusion) 3. Figure 3. 2002). The process involves pulling these raw materials through a heated steel forming die using a continuous pulling device.. The nominal dimension of the specimen used in the fibre fraction test is shown in Table 3. The summary of the dimensions and results of the test for CT1 and CT2 can be found in Appendix A (Section A. Raw materials are a liquid resin mixture (containing resin. A total of four coupons for each type of pultruded tube (i. Filled thermosetting resins in the polyester and vinyl ester groups are generally used in the pultrusion. Pultrusion is a manufacturing process for producing continuous lengths of reinforced polymer structural shapes with constant cross-sections..1). CT1 and CT2) were tested in accordance with the standard. the gelation.2 The basic pultrusion process concept (www. cured profile is formed that corresponds to the shape of the die. The common fibre-reinforcement in pultruded shapes consists of fibre bundles (called rovings for glass fibre and tows for carbon) fibre. of the resin is initiated by the heat from the die and a rigid. After which the cured product is cut on the desired length by the cut-off saw.4 Glass fibre content The content of the glass fibre in the composite tube was characterised using fibre fraction test. and nonwoven surfacing veils (Bakis et al. fillers and specialized additives) and flexible textile reinforcing fibres. As the reinforcements are saturated with the resin mixture (wet-out) in the resin bath and pulled through the die.e. The reinforcement materials are in continuous forms such as rolls of fiberglass mat and doffs of fiberglass roving. This test was conducted following the standard ISO 1172 (1996). or hardening.strongwell. Coupons measuring approximately 20 x 30 mm were cut from the four sides of the tube. continuous strand mat. It was found that the laminate lay-up and fibre orientation is identical for the two tubes.Chapter 3 – Characterisation of the properties of FRP composite tubes EJ Guades schematic diagram showing the basic concept of pultrusion process.2. Behaviour of glass FRP composite tubes under repeated impact for piling application 43 .

25 Length. where the 00 direction coincides with the longitudinal axis of the tube. Table 3. t (mm) 5.3 shows the sliced coupons and the residue showing the glass fibre orientation of a representative tube (i. Figure 3. l (mm) 30 t 1 1 l 9 2 3 4 5 6 7 8 9 Figure 3.Chapter 3 – Characterisation of the properties of FRP composite tubes EJ Guades Therefore the laminate lay-up and fibre orientation of one tube can already represent both of them. Figure 3. CT 2).3 Summary of glass fibre content of each ply Ply no.2 Details of the specimen for fibre fraction test Type of test Width. 1 Ply orientation 00 Glass content (%) 24 2 +450 4 3 00 12 4 -450 4 5 00 12 6 -450 4 7 00 11 0 8 +45 4 9 00 24 Behaviour of glass FRP composite tubes under repeated impact for piling application 44 . b (mm) 20 Test standard Fibre fraction ISO 1172 a Nominal thickness of the tube Thicknessa.3 Coupon specimens and residue showing the fibre glass orientation Table 3.3 indicates that the stacking sequence of the plies is in the form of [00/+450/00/-450/00/-450/00/+450/00].e.

the fibre content of the two tubes varies from 75. and flexural tests. The average specific mass of the tubes is in the range between 1.25 Type of test Test standard Compressive Tensile 3. hence there is no significant difference occurs between these two properties. This length is decided to be used since a strain gage Behaviour of glass FRP composite tubes under repeated impact for piling application 45 .4 Details of the specimen for coupon tests ASTM D 695:2010 Width.5 Coupon tests Tests on coupons cut from the tubes were undertaken to characterise the mechanical properties of the tubes. with an unsupported length of 20 mm. The summary of the dimensions of the specimens tested and the results of the whole test are presented in Appendix A (Section A. Note that in coupon tests.2–A.84% to 76.Chapter 3 – Characterisation of the properties of FRP composite tubes EJ Guades The summary of the results of the fibre fraction test for the two tubes can be found in Appendix A (Section A.25 Flexural ISO 14125:1998 a Nominal thickness of the tube 15 150 5. This value is comparably small. l (mm) 140 Thicknessa.21%.5 MPa). The experimental characterisation of the coupons has been performed using compressive.3.5.25 ISO 527–1:1996 25 250 5.4).4.934 kg/m3 to 1943 kg/m3. The test was performed in the MTS 810 Servo-hydraulic testing machine (100 kN capacity). On the other hand. all calculated values are the mechanical properties of the tubes along their longitudinal direction. side supported (gripping pressure of 8. 3. The details of the nominal dimension of the specimen and the standards used in the coupon tests for CT1 and CT2 specimens are shown in Table 3. t (mm) 5. The content of the glass fibre of each ply is summarised in Table 3.50 Length.5 mm were loaded using an end-loaded. The difference of the average specific mass and fibre content between the two tubes is less than 1%. The next subsections present the details of each coupon test performed and their results.1 Compressive test The compressive test was conducted using the procedure defined in ASTM D 695 (2010). Compressive test coupons with nominal width b of 12. Table 3.1). The test results of CT2 specimen obtained in this section served as the baseline values in comparison with the residual properties of the impacted tubes discussed in Chapter 5. tensile. b (mm) 12.

4 Compressive test set-up on coupons Figure 3. the value is roughly around 13 in this study). The nominal overall length of the coupon taken from CT1 is 140 mm whilst 115 for CT2.Chapter 3 – Characterisation of the properties of FRP composite tubes EJ Guades needs to be attached on the specimen. the unsupported lengths of the specimens from CT1 and CT2 are the same. A total of 5 specimens were tested for each tube in which at least one was taken from its four sides.2). Note that this length is still in the range recommended in the standard (slenderness ratio from 11 to 16. Slicing of the coupons was carefully done by using a wet saw machine. Figure 3. The test was conducted at a loading rate of 1. it was observed that the specimen Behaviour of glass FRP composite tubes under repeated impact for piling application 46 . The compressive modulus was then established from the linear fit of the stress-strain curve between 500 and 2500 microstrains. Recording of data for compressive test was generated using Systems 5000 data logger. It should be noted that the values of the stress and the strain of the curve displayed in the figure are the mean values of the specimens with strain gages.5 shows the typical compressive stress-strain relationship of the tested coupons taken from CT1 and CT2. Two of the 5 specimens were instrumented by a 6mm long uniaxial strain gage attached on the 20 mm unsupported length using a super glue or epoxy adhesive. Figure 3.5 mm/min until failure. The results of the whole test are presented in Appendix A (Section A.4 shows the test set-up used in performing the compressive test. Nevertheless. For both CT1 and CT2 specimens. The compressive stress was calculated by dividing the applied load with the cross-sectional area of the specimen (b x t) whilst the strain was determined using uniaxial strain gage attached to the specimen. The overall length adopted in CT2 is comparatively lower than CT1 as the former is used in comparison with the residual properties of impacted tubes as presented in Chapter 5.

6 illustrates the typical failure mode of the specimens tested under compressive loading and their conditions at the end of the test.500 to 10. Inter-laminar failure along the unsupported length was observed during the test.5. It was observed that the strain gages failed earlier than the specimen.2 Tensile test The tensile test was performed in a 100 kN capacity MTS Insight Electro-mechanical testing machine using a crosshead speed of 2 mm/min. the maximum compressive stress calculated for CT2 specimen ranges from 430 to 450 MPa with an estimated strain at about 8400 to 9300 microstrains. On the other hand.6 Compressive failure mode and condition of the specimens after the test 3.000 microstrains. The tensile test specimens have nominal Behaviour of glass FRP composite tubes under repeated impact for piling application 47 .Chapter 3 – Characterisation of the properties of FRP composite tubes EJ Guades tested under compressive loading behaved linearly elastic up to failure. the estimated strain with the mentioned failure stress is in the range from 8.5 Compressive stress-strain relationship CT1 CT2 Figure 3. The test was conducted in accordance with standard ISO 527-1 (1996). Figure 3. however. 500 Stress (MPa) 400 300 Failure of strain gage 200 CT1 CT2 100 0 0 2000 4000 6000 8000 10000 Strain (microstrain) Figure 3. The CT1 specimen was observed to fail at a compressive stress between 420 to 485 MPa.

The experimental set-up used in conducting the tensile test is shown in Figure 3. the strain was determined using a 20 mm gage length strain gage attached on the specimen. the maximum tensile stress calculated is in the range of 570 to 650 MPa. the maximum calculated Behaviour of glass FRP composite tubes under repeated impact for piling application 48 .8. the strain gage attached on CT1 failed before the specimen.300 to 16. the values of the stress and the strain in the curve are the average values of the specimens with attached strain gages.3). Two of the specimens were instrumented by a 20 mm gage length uniaxial strain gage. A total of 5 coupons were cut from each tube using a wet saw machine and tested. The length l of coupons taken from CT1 is 250 mm whilst for CT2.7.To determine the tensile stress. On the other hand.900 microstrains.7 Tensile test set-up on coupons The longitudinal stress-strain curves of CT1 and CT2 specimens tested under tensile loading is displayed in Figure 3. the length is around 230 mm. the applied load was divided by the cross sectional area of the specimen. Just like in compressive test. After which the tensile modulus was obtained from the linear portion of the stress-strain curve at a strain between 500 and 2500 microstrains. For CT1 specimen. The 230 mm length of CT2 is used as these specimens were adopted as the baseline in comparison for the residual tensile properties of the impacted tubes (Chapter 5). Figure 3. The estimated strain at this failure stress is about 14.Chapter 3 – Characterisation of the properties of FRP composite tubes EJ Guades width b of 25 mm.8 that CT1 and CT2 specimens both exhibited an elastic behaviour up to failure. It can be observed from Figure 3. On the other hand. All the data were recorded using Systems 5000 data acquisition machine. The results of the whole test are presented in Appendix A (Section A. In this test. A 50 mm long gripping tabs (same material as the tested specimens) were attached to both ends of the specimen using Techniglue CA epoxy adhesive.

The specimen was held and pressed. A total of 5 specimens were taken from each composite tube and tested. The test was performed in a 10 kN capacity MTS Insight Electro-mechanical testing machine with a loading rate of 3 mm/min until failure.Chapter 3 – Characterisation of the properties of FRP composite tubes EJ Guades tensile stress for CT2 specimen varies from 570 to 640 MPa at a strain values between 14. flexural test specimens were cut from the pultruded tubes using a wet saw machine. The typical failure observed during the tensile test was glass fibre rupture along the gage length (Figure 3.3 Flexural test The 15 x 150 mm (width b and total length l.5.9 Tensile failure mode and condition of the specimens after the test 3. respectively) specimen was tested under three-point static bending using the standard procedure defined in ISO 14125 (1998). 600 Stress (MPa) 500 400 Failure of strain gage 300 200 CT1 CT2 100 0 0 5000 10000 15000 Strain (microstrain) Figure 3.8 Tensile stress-strain relationship CT1 CT2 Figure 3. Similar with the compressive and tensile tests specimens. by two fixed supports and loading steel cylinders having a diameter of Behaviour of glass FRP composite tubes under repeated impact for piling application 49 . respectively. according to the standard.900 microstrains. A span ls of 84 mm was selected giving a span to depth ratio of 16:1.800 to 14.9).

Figure 3. strain. however.10 demonstrates the test set-up used in performing the flexural test.12 displays the failure mode of the specimens and their condition after the flexural test.900 to 27. as well as the actual set-up during the flexural tests. Similar with the other two tests. For CT2 specimen.11 indicates that CT1 and CT2 specimens remain elastic throughout the test. Figure 3.3). It was found that the maximum calculated flexural stress of CT1 specimen ranges from 1.3). Behaviour of glass FRP composite tubes under repeated impact for piling application 50 .070 MPa with a maximum strain at around 25. The data were recorded via Systems 5000 data acquisition machine. Figure 3.000 microstrains.Chapter 3 – Characterisation of the properties of FRP composite tubes EJ Guades 10 mm. the values used to plot the curves in Figure 3.11 shows the curve that relates the stress and the strain of the CT1 and CT2 specimens. The figure indicates that the typical type of failure on the specimens tested under flexure is fracture of the fibre at the tensile side of the specimen below the point of loading.700 to 27. and modulus were calculated using the equations indicated in the test standard (see Appendix A.000 to 1.10 Flexural test set-up on coupons Figure 3.060 MPa having a failure strain at about 23. Figure 3. The values of the flexural stress. Section A. In flexural test.300 microstrains. these values were achieved from the test results of 5 specimens as compared to 2 specimens for compressive and tensile tests. The results of the whole test are presented in Appendix A (Section A.11 are the mean values. Some of the specimens tested using three point bending also showed inter-laminar shear fractures. the maximum flexural stress based from the calculation is between 940 to 1. This figure indicates the schematic illustration.

(1995) in performing compressive test on composite tubes was considered. tests on full scale specimens were also undertaken to characterise their mechanical properties.Chapter 3 – Characterisation of the properties of FRP composite tubes EJ Guades 1200 Stress (MPa) 1000 800 600 CT1 CT2 400 200 0 0 5000 10000 15000 20000 25000 30000 Strain (microstrain) Figure 3. The following subsections discuss the details of the tests. the procedures made available from the literature were considered as a guide in conducting the test. 1995). In the present study. The characterisation of their properties was carried out through experimentation using compressive and flexural tests.12 Flexural failure mode and condition of the specimens after the test 3. the method adopted by Guess et al. As a result.1 Compressive test There is currently no standard method in performing compressive test on composite tubes.6. The results obtained from the tests on full scale specimens provide additional information on the properties of the studied tubes. Behaviour of glass FRP composite tubes under repeated impact for piling application 51 .6 Full scale tests Aside from the coupon tests that have been performed on the two tubes. Specifically. This specimen length provides a slenderness ratio of around 2.11 Flexural stress-strain relationship CT1 CT2 Figure 3. the adopted length of the specimen is 100 mm. 3.6 (slightly below to that used by Guess et al.

The compressive test was performed in the 2000 kN capacity servohydraulic compression testing machine. CT1 CT2 Figure 3. the peak stress. Since the test results on 100 mm long specimen provided the basis in characterising the compressive properties of the tube. All test specimens were compressed at a rate of 1. Aside from testing a specimen having a length of 100 mm. as well as the deformation behaviour is the main concern of testing the 200 mm long specimen. The outcomes of the compressive test on a 200 mm long specimen. is only performed on CT1 specimen (total of 3 replicates). are presented in Appendix A (Section A. Consequently. however.13 displays the test set-up and the specimen used in conducting the compressive test.10).5 mm/min up to failure. Figure 3. Snapshots were taken during and after the test on the specimens to document their mode of failure.13 Compressive test set-up on full scale specimens Behaviour of glass FRP composite tubes under repeated impact for piling application 52 . the results on longer specimen is not discussed in this section. Recording of data was generated using Systems 5000 data logger. compressive test on a 200 mm long specimen was also undertaken.5).Chapter 3 – Characterisation of the properties of FRP composite tubes EJ Guades A summary of the details of the 5 specimens tested is presented in Appendix A (Section A. A total of 5 replicates for each type of tube were tested in which two of them were instrumented by a strain gage. The 20 mm gage length uniaxial strain gage was attached on one of the sides of the tube positioned along its mid-height to record the strain values. The method used in performing compressive test on longer specimen is similar to that of the shorter one except that it is tested without attached strain gage. The main reason of including a longer specimen in the test is simply to get additional information whether by using a length of up to 2d or 2b (where d and b are the sides of the tube) provides no significant change in its compressive strength. The compressive test on longer tube.5. however. Table A.

14 Compressive stress-strain relationship of full scale specimens Behaviour of glass FRP composite tubes under repeated impact for piling application 53 .14. the compressive property of one tube can be used in representing the property of the other.5.Chapter 3 – Characterisation of the properties of FRP composite tubes EJ Guades The stress-strain curves of CT1 and CT2 tested under compressive loading is shown in Figure 3. delamination along the wall.14 that CT1 and CT2 tubes subjected by compressive load remained linearly elastic although some of the strain gages failed earlier than the specimens. The results of the whole test are summarised in Appendix A (Section A. 300 Stress (MPa) 250 200 150 Failure of strain gage 100 CT1 50 CT2 0 0 2000 4000 6000 8000 Strain (microstrain) Figure 3. Consequently. It can be observed from Figure 3. glass fibre rupture. It was observed that the common type of damage is buckling bulge (inside and outside). the modulus was established from the linear fit of the stress-strain curve between 500 and 2500 microstrains.800 to 7.400 to 6. The linearly-elastic behaviour of the tested tubes is also found in testing coupons specimen as reported earlier in Section 3. Figure 3. The values of the stress and the strain in the figure indicate the mean values of the specimens having strain gages.500 microstrains. It was also noticed that few of the tested tubes manifest “brooming” on their top and bottom ends. CT2 specimen exhibited a failure stress between 253 to 289 MPa with a maximum strain ranging from 6.1. and matrix cracking.15 illustrates the damage mode of the specimens tested under compressive loading and their conditions after the test.000 microstrains. The calculated maximum compressive stress for CT1 specimen is in the range of 268 to 294 MPa with a strain at about 6. The results obtained from the compressive tests on the full scale specimens indicate that their compressive strengths are comparable. The compressive stress was calculated by dividing the applied load with the cross-sectional area of the tube whilst the strain was obtained based from the data recorded by the attached strain gage. Just like the compressive test on coupon specimens.5). On the other hand.

The middle top (compression) and bottom (tension) sides of the tube were Behaviour of glass FRP composite tubes under repeated impact for piling application 54 . respectively. The specimens were loaded at a constant rate of 3 mm/min until failure. however.200 mm long specimen was centred on the lower machine supports (the support is 1000 mm apart from each other).15 Typical failure mode and condition of the full scale specimens 3. A summary of the details of the specimens (CT1 and CT2) tested is presented in Appendix A (Section A. total specimen length is 1. the flexural properties of the composite tubes were characterised using 3-point bending test. The 3-point bending test was performed in the 2000 kN capacity servo-hydraulic compression testing machine. This test was also undertaken to have a comparison with the results in 3-point bending. Data was collected using Systems 5000 data acquisition machine and the data acquisition system of the compressive testing machine. The load is applied using the 60 mm diameter steel rod backed up with a 150 mm x 100 mm x 12 mm flat steel plate placed between the rod and specimen to help distributing the applied load. A relatively longer span (ls = 1200 mm. Figures 3. In addition to 3-point bending test. The 1.2 Flexural test In this study. flexural test using 4-point loading was also performed on the composite tube to get additional information especially on the flexural strength.16a and 3.6). is only conducted on CT1 specimen (three replicates). The testing machine used in the 4-point bending test is similar to that in the 3-point bending.6.500 mm) was used in testing the specimen under 4-point loading. used in conducting the 3point bending test.16c show the test set-up and the specimens. The 4-point bending tests. A total of 3 replicates were tested for each type of tube. All specimens were instrumented by a 20 mm gage length uniaxial strain gage attached on the bottom face (tensile side) along the mid-span of the tube.Chapter 3 – Characterisation of the properties of FRP composite tubes EJ Guades CT1 CT2 Figure 3.

The details of the specimen used in 4-point bending test are presented in Appendix A (Section A. P d 500mm 500mm ls = 1000mm l (a) Actual set-up and schematic illustration of 3-point bending test P P 200mm d 500mm 500mm ls = 1200mm l (b) Actual set-up and schematic illustration of 4-point bending test CT1 CT2 3-point bending test CT1 4-point bending test (c) Specimens with attached strain gage Figure 3.16 Flexural tests on full scale specimens Behaviour of glass FRP composite tubes under repeated impact for piling application 55 .16c. respectively.Chapter 3 – Characterisation of the properties of FRP composite tubes EJ Guades instrumented by a 20 mm length uniaxial strain gage.6).16b and 3. The test setup and specimen used in conducting the 4-point bending test are displayed in Figures 3.

17 Flexural load-displacement relationship (3-point bending test) The curve describing the load-strain relationship of the tested tubes is displayed in Figure 3. respectively).18. At this point. It should be noted that these values were calculated using Equation A. The maximum calculated flexural stress for CT1 specimen ranges from 125 to 131 MPa. After this point. the peak flexural stress of CT2 specimen is between 127 to 143 MPa. It should be noted that the curves shown in the figure are curves of the representative tubes. One can notice that the peak load found to be affected by the initiation of the initial crack. however. Note that the strain gage was only attached on the bottom Behaviour of glass FRP composite tubes under repeated impact for piling application 56 .17 shows the relationship between the load and displacement (midspan) of the specimens tested under 3-point loading. initial cracks on the surface of the tested tube in contact with the loading plate were observed. It was suspected that local crushing on the contact point between the loading plate and the compressive zone makes the load steady. The earlier is the occurrence. This result is expected since the occurrence of premature failure reduced the bending stiffness of the tested tubes. On the other hand. however. When they deflected by around 7 or 9 mm (CT1 and CT2. It is worth noting that the stiffness before the occurrence of the initial cracks is comparably higher than after the manifestation. the value of the load increases with increasing displacement until failure.16 (Appendix A). the value of the load tends to become steady. The load initiating these cracks is about 18 kN and 23 kN for CT1 and CT2 specimens respectively.Chapter 3 – Characterisation of the properties of FRP composite tubes EJ Guades Figure 3. It can be observed from the figure that initially the curve of the tested specimen exhibits an elastic behaviour. the lower is the peak load. 40 Load (kN) 30 20 10 CT1 CT2 0 0 5 10 15 20 25 Displacement (mm) Figure 3. The loaddisplacement relationships of the entire specimens under flexural test (3-point and 4point) are presented in Appendix A.

The figure indicates the curves of the three replicates (i.20 demonstrates that the load increases continuously with increasing displacement until failure.Chapter 3 – Characterisation of the properties of FRP composite tubes EJ Guades side of the tube.19). It was observed that all specimens failed by crushing on the compression side of the tube at the loading point (Figure 3.e. Unlike the load-displacement curve during 3-point loading test.. respectively).18 Flexural load-strain relationship (3-point bending test) Figure 3. CT1 specimen) tested. 40 Load (kN) 30 20 Failure of strain gage 10 CT1 CT2 0 0 1000 2000 3000 4000 Strain (micro) Figure 3. The specimen failed at a range between 40 Behaviour of glass FRP composite tubes under repeated impact for piling application 57 . the curve in Figure 3. This is because no sign of premature failure occurred during the 4-point bending test. The curve in Figure 3.19 Typical failure modes in 3-point bending tests Figure 3.20 shows a typical load-displacement curve of the flexural test under 4-point loading.18 indicates that it is linearly elastic up to the start of initial cracks. This trend continues after a certain point (initial cracks) whereby the tube becomes stable and is able to carry additional load by as much as 24 kN and 30 kN (CT1 and CT2. The attached strain gage on the tensile side (mid-span) was able to record the strains up to the failure of the tube.

the top (midspan) surface of the tube goes back to its local undeformed position (neutral) when the load reached to around 27 kN. the trend of the strain on the top is different to that in the bottom. By comparing this value. all tubes failed by crushing on their compression side at the loading point (see Figure 3. A further load increase provided the top to be in tension. however. After some point.21 that the loadstrain relationship at the bottom of the tube is linearly elastic up to failure.21 demonstrates the relationship of the load and strain (top and bottom) of the tubes tested under 4-point loading. Just like the dominant failure mode observed in 3-point bending.Chapter 3 – Characterisation of the properties of FRP composite tubes EJ Guades kN to 41 kN. 50 Specimen 1 Load (kN) 40 Specimen 2 Specimen 3 30 20 10 0 0 10 20 Displacement (mm) 30 40 Figure 3. On the other hand. This was also the case observed in 3-point bending whereby in all instances the bottom part is in tension all throughout the test. The maximum calculated flexural stress for CT1 specimen ranges from 166 to 173 MPa. mid-span) is shifting from being compressed to under tension. these values tend to become positive demonstrating that the tube (top. The attached gages recorded the strains up to the failure of the tube.19b). it follows that they underestimate the values in the 3-point bending test by roughly 25% which is primarily caused by premature failure that had been observed during the 3-point bending test. This phenomenon can be explained by the following. As a result. the loading rams (spaced at 200 from each other) pushed the surface in contact with them creating a concave surface (see Figure 3.22).20 Flexural load-displacement relationship (4-point bending test) Figure 3. As can be observed in the figure. Initially these values are negative indicating that the tube is compressed. It can be observed from Figure 3. However when the load increases. this triggers to push the Behaviour of glass FRP composite tubes under repeated impact for piling application 58 . At the initial stage of the test. the top (midspan) surface of the tube is compressed. Note that these values were calculated using Equation A.17 (Appendix A).

the information from this analysis can be used in understanding the mechanical properties and structural performance of this FRP composite tube for various applications. The results obtained from the FE method would be beneficial since this composite tube has been used for other structural applications such as power pole cross arms.21 Flexural load-strain relationships (4-point bending test) Figure 3.1. Consequently. the primary objective of its inclusion is to exhibit the feasibility of using FE method in predicting the mechanical behaviour in aid of minimising the need of performing a relatively expensive experimental test.34b).7 Finite element (FE) analysis on full scale specimen Numerical simulations were carried out to compare with the experimental measurements of the compressive and flexural behaviour of the tubes.Chapter 3 – Characterisation of the properties of FRP composite tubes EJ Guades initially compressed surface to go back to its undeformed position (neutral line) and finally in tension or forming a convex shape (see Figure 3.22 Typical failure modes in 4-point bending tests 3. Finite element (FE) analysis has been included to demonstrate its feasibility in predicting the mechanical behaviour to eliminate the need of repeating Behaviour of glass FRP composite tubes under repeated impact for piling application 59 . As highlighted in Section 3. 50 Bottom Top Load (kN) 40 Failure of strain gage 30 20 10 0 -2000 -1000 0 1000 2000 3000 4000 5000 6000 Strain (micro) Figure 3.

Finite element method was carried out simulating the specimen and the loading set-up in the actual experimental conditions to have a reliable result. Figure 3. As this is the interest in the analysis.23 shows the material model of the 100 mm square pultruded tube with a wall thickness of 5. the material model is characterised by the initial linear part of the stress-strain curve represented by the elastic modulus along the longitudinal direction of the FRP composite tube.24. This FE analysis considered the elastic linear behaviour of the FRP composite material in comparing with the experimental results. Laminate properties were adopted as property attributes of plate elements.5. 3. The investigation was carried out using the Strand7 finite element analysis commercial package (Strand7.475 x 5 mm (sides) and 1. the 100 mm square tube with a length of 100 mm was modelled which is comprised of 1.Chapter 3 – Characterisation of the properties of FRP composite tubes EJ Guades costly arrangements for experimental tests.23 Material modelling of the composite tube Behaviour of glass FRP composite tubes under repeated impact for piling application 60 . with meshes of 4. (a) Actual tube (b) FE model Figure 3. The simulation of the compressive and flexural behaviour of the full scale tubes using finite FE method is discussed in the next subsections.25 mm and a length of 100 mm.7. The laminate was modelled as a stack of several plies as shown in Figure 3. 2012).840 plate elements.4 x 5 mm (corners). The ply properties adopted in modelling the laminate is summarised in Table 3. The model was developed whereby the property inputs are based from the material properties derived from coupon tests. This value has been inputted in the analysis to predict its mechanical behaviour. The figure also displays the simulated composite tube.1 FE simulation on the compressive behaviour In this study.932 nodes and 1.

900b MPa Poisson ratio υ12 0. A linear static solver was used to investigate the compressive behaviour of the tube (Strand7. Tensile coupon test with extensometer conducted on CT1 specimen for the use in FE analysis In the conducted experiment. Initially. Fraction of this load was then used in the analysis to provide several load values in aid of plotting the load relationship. Behaviour of glass FRP composite tubes under repeated impact for piling application 61 .5 Material properties of the tube wall laminate ply Material property Density Symbol ρ Property value 1. the composite tube was in contact with stiff loading plates at the two ends. Therefore.6.970a Unit kg/m3 Thickness t 0.Chapter 3 – Characterisation of the properties of FRP composite tubes EJ Guades Figure 3. Even if the support condition may emerge to be close to a simply-supported condition.5833 mm a Elastic modulus (longitudinal direction) E11 39. 2006).24 Lamina lay-up arrangement used in FE model Table 3.234 MPa Elastic modulus (transverse direction) E22 12.35c a b c Table 3. the clamped-end condition is more appropriate for this model. This value was chosen arbitrarily as this is more or less the peak load recorded during the experiment. To adopt such support condition. previous research conducted showed a much closer value to the experiment results if a “clamped support condition” is adopted (Teng and Hu. 2012). a 284 MPa uniform pressure load (equivalent to 550 kN) was applied on the top edge of the model. the two ends were fully fixed in all direction except that the axial displacement of the top end was left unrestrained to allow the application of axial loading. A uniformly distributed pressure on the top of the model was applied to properly simulate the loading condition. WCFT Product specification.

Similarly. This value is 4. The experimental result shows linear stress-strain relationship up to final failure and is in good agreement with the predicted stressstrain relation based from FE method.26a. In the figure. Moreover. Unlike in Figure 3. The simulated failure of the tube reveals that bucking bulge happened at its four corners (Figure 3. bulging is also imminent at the sides of the tube.Chapter 3 – Characterisation of the properties of FRP composite tubes EJ Guades Figure 3. The difference of the values is comparably small and therefore the values used in the inputs. The typical failure mode observed in the experiment is buckling bulge at the corner and at the sides of the tube (Figure 3.26b). Though this was not apparent in Figure 3. In the FE analysis.1% lower to that of the actual failure stress. the predicted failure stress using FE method at same strain is around 257 MPa (496 kN).26. simulated cracks (white-coloured portion) found to be happening at the corners. 300 Stress (MPa) 250 200 150 Experiment 100 FEM 50 0 0 2000 4000 6000 8000 Strain (micro) Figure 3. as well as assumptions used in modelling. On the other hand. The actual failure stress of the tube using experimental investigation is 268 MPa (equivalent to 520 kN failure) at a failure strain of 7000 microstrains. the experimental data are from one of the CT1 specimen tested.26a). are considered acceptable as it predicts the experimental values reasonably.25 shows the longitudinal stress-strain relationship using experimental and FE investigation.25 Compressive stress-strain relationships The comparison between the failure modes of the tube obtained from experiment and FE analysis is shown in Figure 3. The experimental results show that cracking is also Behaviour of glass FRP composite tubes under repeated impact for piling application 62 . delamination and matrix cracks at several locations including the corners of the tubes were present during the compressive test as shown in the figure. the simulated cracks occurred on the top and the bottom corners of the tube.26a. it was observed that some of the tested tubes revealed matrix cracking on both top and bottom regions.

a relatively bigger numbers of nodes and plates are employed. The simulated failure mode did not apparently have this kind of failure. respectively.500 mm. In flexural simulation however.27 to 3. Figures 3. The lengths of the tube simulated in the 3-point and 4-point bending tests are 1.5 m) Behaviour of glass FRP composite tubes under repeated impact for piling application 63 . the ply properties used in the former is similar to the latter.200 mm and 1.29 show the simulated and FE models used in flexural behaviour investigation.26 Compressive failure mode of the tested tube 3. Figure 3. it is clear that stress concentration in this area is highlighted indicating that cracks are imminent in this region.27 Actual tube (length varies from 1.2 FE simulation on the flexural behaviour The material model used in the FE analysis of the behaviour of the tube under 3point and 4-point flexural tests is similar to that used in simulating its compressive behaviour. Buckling bulge Buckling bulge Corner cracking Corner cracking (a) Experiment (b) FE analysis Figure 3.7.Chapter 3 – Characterisation of the properties of FRP composite tubes EJ Guades transpiring at the mid-length along the corners of the tube.2 m to 1. Consequently. However.

29 FE model (4-point bending. Figure 3. Therefore an area load (pressure load) is suitable to be used in simulating the loading condition in the FE analysis. Figure 3. the ends of the tube rest on two steel cylindrical supports of the testing machine as shown in Figure 3.Chapter 3 – Characterisation of the properties of FRP composite tubes EJ Guades Figure 3. a constraint support that will resist the translation along its longitudinal was provided to at least one of its end support. L=1.30 indicates that the condition is close to a rollerroller support.30. L=1.30 displays the actual support conditions used in the experimental study. As a result. However for the purpose of stability requirement.2 m) Figure 3. It was found that the contact area of the steel plate to the tube is 80 mm x 100 mm Behaviour of glass FRP composite tubes under repeated impact for piling application 64 . This supports condition shows that the tube maybe allowed translating in its longitudinal direction since the contact area between the steel cylinder and the tube is quite small (can be assumed as line support). the support constraint in the FE analysis was idealised as a simply supported condition. The applied load in the experiment was transmitted from the loading rams to the specimen through a 12 mm thick flat steel plate.28 FE model (3-point bending.5m) During flexural tests (3-point and 4-point).

occurrence of initial crack) on the test. an area load of 3. For 3-point bending behaviour simulation.31 shows the load-deformation relationship using experimental and FE investigations.48 MPa (20 kN) was applied on two points of loading applications in 4-point bending simulation (see Figure 3. the method used in comparing the peak loads obtained from the experiment and from FE analysis is based on the a similar strain condition whereby the reference is the strain obtained from the former corresponding to the failure load. It should be noted that the comparison between the experiment and FEM results are up to the initial peak load (i. In this case. Behaviour of glass FRP composite tubes under repeated impact for piling application 65 . However. Just like the compressive behaviour simulation. the comparison is only up to 17. Figure 3. a fraction of this load was also used in the analysis in aid of plotting the load-displacement relationship.16b for the loading location). This scheme is adopted as apparently the FEM result may not be able to predict reasonably due to the significant reduction on the failure load contributed by the initial crack.6 kN load. The experimental result in the figure was obtained from one of the tested tubes for the three tubes (CT1) testing under 3-point loading condition.30 Support condition during flexural test (both ends) Figure 3. It should be noted that they are selected as an initial loading values since they are considered the maximum peak values in the two corresponding tests.10 MPa (25 kN) was initially applied on the midspan on the simulated tube.. Moreover.Chapter 3 – Characterisation of the properties of FRP composite tubes EJ Guades (8000 mm2). the linear static solver technique was used in simulating the flexural behaviour of the tube.e. an area load of around 2. On the other hand.

For FEM.31. A 2. This was characterised by matrix crushing at the corners. If we consider the failure load of the specimen tested under 3-point bending (26 kN at a displacement of 15.32b. The observed failure of the specimen under 3-point bending test is by crushing on the compression side of the tube at the loading point (Figure 3.3 mm.8% difference indicates that the experimental behaviour up to the first peak load can be fairly simulated using FE method.32a).32b). the crushed edges are represented by a whitecoloured area. Behaviour of glass FRP composite tubes under repeated impact for piling application 66 . it was found that the calculated load at 6. The simulated failure of the tube reveals that crushing on the compression side of the tube is the dominant failure mode.1 kN. the comparison up to the first peak load can be considered reasonable.3 mm displacement is 17. Therefore in this condition where premature failure (local crushing) is imminent to occur. Moreover. 50 Load (kN) 40 30 20 10 Experiment FEM 0 0 5 10 15 20 Displacement (mm) Figure 3.6 kN with a displacement at about 6.8% higher than that predicted using the FE analysis.Chapter 3 – Characterisation of the properties of FRP composite tubes EJ Guades From Figure 3. it follows that the FE result underestimates the experimental result by around 38%.31 Flexural load-displacement relationships (3-point bending) Figure 3.32 displays the failure modes obtained from 3-point bending test and from the simulation. Crushing of the two edges and the formation of the indented (concave) surface were the manifestations of the failure (Figure 3. the failure was manifested through indentation of the loaded area forming a concave surface. The actual peak load (initial) is 2. From this result and from the comparison of the load-displacement curves we can infer that the flexural behaviour obtained from FE analysis predicts well the actual flexural behaviour of the tube up to the initial linear part. the peak load obtained from the experiment is around 17.5 mm). It should be noted that in Figure 3.

Chapter 3 – Characterisation of the properties of FRP composite tubes

EJ Guades

Indented (concave) surface

Crushed edge

(a) Experiment

Indented (concave) surface

Crushed edge

(b) FE analysis
Figure 3.32 Flexural failure mode in 3-point bending test

Figure 3.33 shows the load-deformation relationship obtained from both the
experiment and FE simulation. The comparison between the results of the two
methods involves only the initial linear part of the curve derived from the
experiment. A complete load-displacement curve of this specimen (i.e., CT1) is
previously displayed in Figure 3.20 (Specimen 1). The initial linear part of this curve
extends up to 9.1 mm at a corresponding load of 19 kN. From Figure 3.33, it was
observed that the calculated load from FE analysis at 9.1 mm displacement is 19.7
kN. In this case, the load value predicted from FE analysis is 3.6% higher than the
experimental value. This value is relatively small indicating that the FE analysis
predicted the flexural behaviour of the FRP composite tube up to the initial linear
part.
Supposing we consider to compare the experimental and FE analysis values
up to actual failure (40.8 kN at a corresponding displacement of 25 mm), it shows

Behaviour of glass FRP composite tubes under repeated impact for piling application

67

Chapter 3 – Characterisation of the properties of FRP composite tubes

EJ Guades

that FE value is higher than the experimental by 24%. This difference in the peak
load is contributed by the nonlinearity behaviour of the tube especially in its
deformation behaviour. Apparently, the FE analysis provided a good estimate of
linear flexural behaviour of the FRP tube but seems did not deliver a reasonable
estimation value when reaching the non-linear part. From this result and from the
comparison of the load-displacement curves, we can infer that the flexural behaviour
obtained from FE analysis predicts fairly the actual flexural behaviour of the tube up
to the initial linear part.

60

Load (kN)

50
40
30
20

Experiment
FEM

10

0
0

5

10

15

20

25

30

Displacement (mm)

Figure 3.33 Flexural load-displacement relationships (4-point bending)

The failure pattern of tubes tested under 4-point bending and from FE
analysis is revealed in Figure 3.34. The failure mode observed in 3-point bending test
was also present in 4-point bending (see Figure 3.34a). This was characterised by
crushing at the compression area in direct contact with the loading rams. Similarly,
an indented region (concave) was also noticeable in the failed tubes under 4-point
bending test. These manifestations can also be observed in the simulated failure
mode (Figure 3.34b). It is apparent from the simulated failure that aside from the
mentioned failure patterns, crushing on the midspan area (initially compression zone)
is imminent. In the figure, the crushed portion is represented by a white-coloured
area. It is worth noting that whilst the surface in contact with the loading rams
provided a concave shape, the middle area produces a convex line. This simulation
confirms the results obtained from the load-strain relationship (Figure 3.21) that
while this region is compressed during the initial loading, the increase of loading
until failure shifted the surface into tension mode.

Behaviour of glass FRP composite tubes under repeated impact for piling application

68

Chapter 3 – Characterisation of the properties of FRP composite tubes

EJ Guades

Indented (concave) surface

Crushed edge

(a) Experiment

Convex lines
Indented (concave) surface
Crushed edge

(b) FE analysis
Figure 3.34 Flexural failure mode in 4-point bending test
3.8 Summary of the mechanical properties of composite tubes
Tables 3.6 and 3.7 summarise the average value of the properties of the composite
tubes determined from the different coupon and full scale tests. Note that in coupon
tests, all calculated values are the mechanical properties of the tubes along their
longitudinal direction. As shown in Table 3.6, the peak compressive stress derived
from coupon test of CT1 specimen is 459.14 MPa whilst 441.55 MPa for CT2. The
strength value of the former is slightly higher than the latter by 3.9%. On the other
hand, the average elastic modulus of CT1 specimen subjected under compressive
loading is 51,081 MPa. This value is 2.7% higher compared to the value of CT2
specimen. Table 3.6 also shows that the peak stress of CT1 and CT2 specimens
under tension using coupon test are 618.48 and 603.20 MPa, respectively. These
values suggest a difference of about 2.5% relative to the other value. It was found
that the tensile elastic modulus of CT1 specimen is 39,233 MPa whereas for CT2, the
value is 40,698 MPa. The value of the former underestimates the latter by roughly

Behaviour of glass FRP composite tubes under repeated impact for piling application

69

Chapter 3 – Characterisation of the properties of FRP composite tubes

EJ Guades

3.7%. The average flexural peak stress of CT1 and CT2 specimens are 1,037.54 and
994.44 MPa, respectively. The value of CT1 specimen is marginally higher than CT2
by 4.2%. The values of their flexural modulus, on the other hand, are 36,092 and
38,534 MPa respectively. These values indicate that the difference of the modulus
between CT1 and CT2 specimens is about 6.8%. For coupon tests, the difference of
the strain at peak values for both specimens under compressive, tensile, and flexural
loading is 4.4, 5.1, and 0.4%; respectively.
The results indicated in Table 3.7 shows that for full scale test, the peak
compressive stress of CT1 is 284.14 MPa. This value is 4.8% higher than that of
CT2. The difference between their compressive elastic modulus and strain at peak is
1.9 and 5.8%, respectively. On the other hand, the difference between the peak
flexural stress of CT1 and CT2 specimens is 5.4%. By comparing the values
generated from the coupon and full scale tests, it follows that the value of the former
is relatively higher than the latter regardless of the type of the tested tubes.

Table 3.6 Summary of mechanical properties from coupon tests
Properties
Compressive, Peak stress (MPa)

CT1
459.14

CT2
441.45

Difference (%)
3.9

Compressive, Elastic modulus (MPa)

51,081

49,690

2.7

Compressive, Strain at peak (%)

0.92

0.88

4.4

Tensile, Peak stress (MPa)

618.48

603.20

2.5

Tensile, Elastic modulus (MPa)

39,234

40,698

3.7

Tensile, Strain at peak (%)

1.56

1.48

5.1

Flexural, Peak stress (MPa)

1,037.54

994.44

4.2

Flexural, Elastic modulus (MPa)

36,092

38,534

6.8

Flexural, Strain at peak (%)

2.61

2.60

0.4

Table 3.7 Summary of mechanical properties from full scale tests
Properties
Compressive, Peak stress (MPa)

CT1
284.14

CT2
270.41

Difference (%)
4.8

Compressive, Elastic modulus (MPa)

39,970

39,215

1.9

Compressive, Strain at peak (%)

0.69

0.65

5.8

Flexural, Peak stress (MPa)a

128.64

135.63

5.4

Flexural, Peak stress (MPa)b
169.75
b
From 3-point bending test, from 4-point bending test

-

a

Behaviour of glass FRP composite tubes under repeated impact for piling application

70

Chapter 3 – Characterisation of the properties of FRP composite tubes

EJ Guades

3.9 Conclusions
The mechanical properties of the composite tubes were characterised using
experimental investigation. Two types of tubes were tested; designated as CT1 and
CT2. CT1 is adopted in studies presented in Chapters 4 and 6, whilst CT2 is used in a
study discussed in Chapter 5. The tests were performed on coupons and full scale
specimens. The result showed that generally, CT1 and CT2 specimens exhibited
linearly elastic up to failure. For coupon test, it was observed that the flexural
strength is comparably higher than its corresponding compressive and tensile
strengths. The maximum variation of the experimental data (fibre fraction, specific
mass, peak stress, and elastic modulus) is less than 5%. This result indicates that the
reproducibility of the test is quite reasonable which verifies that the manufacturing
process of the composite tubes is consistent. This result also indicates that the
experimental procedures were conducted within the acceptable margin of error. The
comparison of the values of the mechanical properties between CT1 and CT2
specimens revealed that the difference is less than 6%. It was also revealed that both
tubes have similar plies lay-up and glass fibre content. Also, no significant difference
on the properties occurs between the two composite tubes.
The compressive and flexural behaviours of FRP composite tube were
investigated using experiment and FE methods. The result demonstrated that the
flexural stress of the tube obtained from 4-point bending test is relatively higher than
from 3-point bending due to the presence of pre-mature failure on the latter. As a
result, it is recommended that a 4-point bending test can be used in characterising the
flexural behaviour of the FRP composite tube. The comparison between the
compressive peak load values using experiment and FE methods revealed that their
difference is less than 5%. On the other hand, it was found that the variation of the
compared load values describing the flexural behaviour up to the initial linear part of
the load-displacement curves is 4%. Though the FE method did not provide a good
estimation of the ultimate moment capacity of the tube, it is apparent that the both
compressive and flexure failure modes were fairly simulated. These results indicated
that FE analysis predicted reasonably up to the initial linear part of the actual
compressive and flexural behaviours of the FRP composite tubes.
In Chapter 4, an investigation on the behaviour of composite tube under
repeated axial impact is presented.

Behaviour of glass FRP composite tubes under repeated impact for piling application

71

One of the main factors that affect their driving performance is the impact strength of the fibre composite materials. requires more careful consideration due to their relatively low stiffness and thin walls. the damage Behaviour of glass FRP composite tubes under repeated impact for piling application 72 . however. The results of these studies revealed that parameters such as impact load (or mass). This chapter presents an experimental investigation on the behaviour of a 100 mm square FRP pultruded tube under repeated axial impact. incident energy. However. it is considered suitable to characterise the impact behaviour of a full-scale hollow FRP pipe piles used in piling application. The behaviour of FRP composite materials under repeated impact is commonly characterised using experimental investigation. Therefore. Although FRP composite tubes with a relatively smaller section (100x100 mm square) were used in the experimental investigation. The main interest of the study is to characterise the impact behaviour of the FRP material itself and therefore a possibility of scaling down the size of the tube is reasonable. It would be equally important to know on how these parameters affect the behaviour of composite tubes when they are axially impacted. As the cross section of the tubes increases. Driving them. The experimental studies that investigate the impact behaviour. and the number of impacts affect the impact behaviour.1 Introduction The high corrosion-resistant characteristic of FRP composite tubes and their emergence as a structural component made them suitable alternatives for piling application in harsh marine environment. the impact energy (or impact load) required during the test to collapse or fail them also increases.Chapter 4 – Investigation on the impact behaviour of FRP tubes EJ Guades Chapter 4 Investigation on the behaviour of square FRP composite tubes under repeated axial impact 4. however. The possibility of damaging the fibre composite materials during the process of impact driving is always a concern. mostly focused on composite laminates or tubes which are transversely impacted. there is a need to understand the impact behaviour of these materials in order for them to be safely and effectively driven into the ground.

2. The maximum drop mass (mass of the impactor and added weights) that can be attained from the set-up is 25 kg. On the other hand.49 Length. The investigated tube has ±450 glass fibre reinforcement that provides better performance making it suitable for structural application.1c) and the instrumentation and data acquisition methods were designed by the author and his supervisors as part of the test methods.1 Details of the specimen b d t Dimension Depth.1 shows the cross sectional dimension of the specimen. A total of 20 specimens were tested following the test matrices presented in Section 4. Table 4. b (mm) 100. drop mass. It should be noted that the values in the table are the mean values of the 20 specimens. d (mm) Value 100.2 Test set-up and procedure Repeated impact test was performed using an un-instrumented drop weight impact testing machine defined in AS 4132.2.1 Test specimen The composite tube used in the investigation presented in this chapter has mechanical properties similar to that of CT1..3 (1993) with some modifications on the steel clamping frame to suit for the testing condition of the specimen (Figure 4. failure mode) between FRP composite tubes with smaller and bigger geometrical sections can be associated. t (mm) 5. l (mm) 375.2 Experimental program 4.2.1).Chapter 4 – Investigation on the impact behaviour of FRP tubes EJ Guades behaviour (e. 4. The effects of parameters such as the incident energy. Table 4.40 Thickness. The impactor is a 135 mm diameter steel cylinder with a Behaviour of glass FRP composite tubes under repeated impact for piling application 73 .2.52 Width. number of impacts.1). the steel clamping frame (Figure 4.22 4. The impact testing machine was readily available (pre-fabricated) in the Centre of Excellence in Engineered Fibre Composites of USQ.g. The details of the dimension of all specimens tested can be found in Appendix B (Section B. and impact velocity (or drop height) on their damage tolerance limit are also presented.

Steel frame to hold the specimen 10. Fixed guide PVC tube 5. 1. Sighting cut-outs at 500 mm intervals 6. Light rope for release and retrieval of impactor 2.1 Impact testing set-up Behaviour of glass FRP composite tubes under repeated impact for piling application 74 . Specimen (375 long mm pultruded tube) 11. Foam to flexibly hold the specimen 12. Improvised gripping/releasing devise 3. Impactor (mass can be varied) 4. 10 mm thick steel plate capping 9.Chapter 4 – Investigation on the impact behaviour of FRP tubes EJ Guades flatted-nose contact surface. The nominal (net or baseline) mass of the impactor is 16 kg. with additional 5 kg steel weights can be attached to the impactor as desired. Main impact testing housing (steel tripod) 8. Solid concrete base 3 4 1 5 7 2 8 6 9 10 11 12 (a) Schematic diagram of drop weight impact apparatus (b) Oblique view of the test set-up (c) Steel frame fixture Figure 4. in which the incident (applied) energy can be varied up to 736 J. Extended movable guide PVC tube 7. The maximum available drop height is 3 m.

2 provided a progressive crushing type of failure. During test. progressive end crashing) usually occurred on a relatively short specimens.e. This length was selected based on the type of failure observed during field driving of composite tubes. It was reported that commonly the damage occurred during impact driving of square composite tubes is end crushing at the top portion. In the present study.75) was selected due to some considerations especially in placing the accelerometer on the Behaviour of glass FRP composite tubes under repeated impact for piling application 75 . whilst g is the gravitational constant. Two replicates with a length of 375 mm for any given incident energies were subjected to a maximum of 130 impacts or up to collapse/failure of the tubes. It is then temporarily held and later released by an improvised clamping devise positioned a distance from the impact apparatus. Moreover. this type of failure was initially considered in selecting the length of the specimen based from this result.Chapter 4 – Investigation on the impact behaviour of FRP tubes EJ Guades The incident energy Ein can be calculated using Equation 4. a relatively longer length of 375 mm (b/l = 3. this type of failure (i.. The rope is caught manually after each impact to avoid bouncing and extraneous impacts on the specimen. A 10 mm thick steel plate was used in capping the top of the tested tube to help in evenly distributing the impact load and to simulate actual pile driving condition. respectively. Ein = mgh (4.1) where m and h are the drop mass and drop height. the result of their study showed that an aspect ratio (b/l. Mamalis et al. where b and l are the sides and axial length of the tubes. Therefore. Steel cap is removed at least every three impacts to check the position of the impactor relative to the contact section of the tube and to ensure that the tup strikes the specimen each time at approximately same location. The steel cap was held by a spring connected to the steel frame to avoid overthrowing during the rebound. (1997a) reported that for square composite tube made of glass fibre and vinyl ester subjected to single impact. This process is repeated until the required number of impacts on the tube is achieved or damage is observed on the specimen. respectively) of up to 3. the impactor is raised manually to the desired drop height through an attached rope. The present study considered this “worst scenario” during the conduct of the impact tests on FRP composite tubes.1. The damage was observed to be imminent at the top of the pile (end crushing) with not much more on mid-height collapse (buckling failure).

Chapter 4 – Investigation on the impact behaviour of FRP tubes specimen EJ Guades and the specimen length the current impact testing set-up can accommodate.5 J). The results obtained using the baseline values suggest that at 130 impacts. These two conditions are considered important and used in defining the behaviour of composite tubes subjected to repeated impact.2 and 4. From this preliminary test.g.e. It should be noted that the specimen identification in Table 4.g. Chapter 2). 130) was chosen based on the initial result of impacting the tube using the minimum drop mass (i.2 kg) at a drop height of 3 m. it was adopted as it is found suitable in rupturing a 100 mm square FRP composite tube. The adopted drop masses shown in the tables are the minimum and maximum values that can be attained from the testing set-up and the intermediate mass is determined by attaching a 5 kg steel weight on the impactor. The maximum number of impact (i. the impact load (or impact mass) needed to fail the FRP composite tube increases with increasing cross section..2 and 4.. Tables 4. a trial test (repeated impact) was performed first on the 100 mm square tube without attaching an instrumentation to have a little bit of an idea whether a 16. As highlighted in Section 4. As an example. the specimen identification in Table 4. the tube was physically observed to rupture after a certain impact repetitions (around 100 impacts).. The load (or mass) used in the experiment is apparently not the typical load used in actual pile driving.3 show the detailed test matrix for the impact test adopted in this study.2 is referred from the incident energy (e. It should be noted that this mass is the baseline (net) mass of the impactor without attaching additional weights for the current test set-up. 16. Consequently. respectively.2 kg minimum drop mass can rupture the tube at a certain number of impact. a 125 mm square pultruded tube needed a 1000 kg hammer in driving until it ruptures (Section 2. E480-2. this load becomes Behaviour of glass FRP composite tubes under repeated impact for piling application 76 .3 are used in defining the impact behaviour of the tube and its impact damage tolerance limit.4.. This number of impact. E630 ≈ 634. On the other hand. However. impact energy higher than the baseline will fail whilst those with relatively lower value may not induce a significant damage on the tube.1. E480-1. In the present study. drop mass. and E480-3) indicates similar incident energy but with different drop mass and height.e.3 (e. The test matrix presented in Tables 4. and drop height mentioned served as the baseline since it was observed that end crushing on the top portion occurred on the tested specimen.

8b from the top of the tube). 2/3. the author performed a simple analytical modelling study explaining the accuracy of the assumption to use the data obtained at the mid-height of the tube and is presented in Appendix C (Section C. at the head of the tube).5b (where b is the side or diameter of the pile) from the top of the pile. the acceleration history data was post processed to get the energy history curves needed for further analysis. Inc.Chapter 4 – Investigation on the impact behaviour of FRP tubes EJ Guades the minimum drop mass adopted in this study to characterise the impact behaviour of FRP composite tube.2. Behaviour of glass FRP composite tubes under repeated impact for piling application 77 .2 are distributed in the order of 1/3.2 and 4.e.1). As will be presented in Section 4. This study used the acceleration recorded by the shock sensor placed at the mid-height of the tube to represent its impact response. The data acquired by the accelerometer were recorded and saved on a personal computer via LMS SCADAS Mobile data acquisition machine using a sampling rate of 51. Section 4. whilst in Table 4. and 3/3 of the maximum available drop height of 3 m.3.3 highlighted that the value of the calculated energy at the mid-height is closed to the applied (incident) energy indicating that the amplitude of the recorded acceleration history will be likely similar when the sensor was placed relatively nearer to the impact point (i. The entire test specimens used in the impact test and some details on the machine used in impact testing are presented in Appendix B (Section B. A relatively longer distance is selected to provide extra protection on the accelerometer from direct hitting when failure of the tube happened. The drop heights shown in Table 4.2 kHz.2. In pile driving.3) to avoid damage of the accelerometer when rupturing of tube occurred. The results presented in Appendix C shows that the difference of the acceleration values at the mid-height and at the top most portion of the tube is relatively small indicating that the former can be used to represent the impact response of FRP composite tube.2).3. Some specimens were subjected to less than 130 impact repetitions (see Tables 4. they are obtained depending on the corresponding drop mass and the targeted incident energy (column 3). This recommendation was considered in the present study and the accelerometer was mounted on the mid-height of the tube (distance is 1. To support this hypothesis. ASTM D 4945 (2008) recommends that accelerometer should be placed at a distance of at least 1. The specimen was instrumented by an accelerometer with model 350A14 from PCB Piezometrics.

56 3.Xpress. respectively in Table 4. 2012).c.56 Drop height (m) 3.20 3.20 2. NC (non-collapsed tube).20 3.93 476.0 130 (C/F)a E320 16.20 1.20 1.00 476.57 Incident energy (J) 634. Table 4.5 Number of Remarks impacts 30 (C/F) E630-2b 21. The snapshot of this microscope is shown in Appendix B (Section B.3 Test matrix used in defining the impact damage tolerance Specimen ID E630-1 Drop mass (kg) 25.00 423. The microscopic observation was performed using a magnification factor of about x100.00 211.2). the impacted tube was taken and inspected to determine its damage.56 2.56 2.9 130 (NC)a C/F (collapsed/failed tube).8 130 (NC)a E210 21.8 130 (C/F)a E420 21. asee Figure 4. E480.20 Drop height (m) 2.00 634.8 90 (C/F) E480-3c 16. The impact load as a function of time Ft is proportional to the recorded acceleration signal at by the mass impactor m and is calculated using Behaviour of glass FRP composite tubes under repeated impact for piling application 78 .5 45 (C/F) E480-1 25.Chapter 4 – Investigation on the impact behaviour of FRP tubes EJ Guades After the test.3 Table 4. and E420.2 b.3c.00 423. Visual inspection and MOTIC® SMZ 168 Series stereo zoom microscope were used in observing the damage on the surfaces of the impacted tube.3 Data processing From the data acquisition machine. A typical scanned image showing micro-cracks on the top of the tube using this apparatus is displayed in Figure 4.00 158.5 130 (NC)a E160 16.0 60 (C/F) d E420-2 21.d 4.00 Incident energy (J) 634.71 423.20 1.5 Number of Remarks impacts 45 (C/F)a E480 16.56 2. the acceleration-time responses acquired by the accelerometer were then transferred to Excel format using “Report Preview” method (LMS Test.56 1.0 130 (C/F) same specimen as E630.8 130 (C/F) E420-1 25.00 317.2 Test matrix used in defining the impact behaviour Specimen ID E630 Drop mass (kg) 21.8 45 (C/F) E480-2 21.25 476.00 476.2.

and the energy absorbed by the systems in vibration.2. respectively. The incident energy (column 4 in Table 4. respectively. vt is the velocity at present time increment. Fs is the load at present displacement increment. A typical accelerationdisplacement curve is displayed in Figure 4. are not examined in this study. 2005). The energy loss can be estimated as the difference of these two energies.6 to 4. In Equation 4. st. at-1 is the acceleration at previous time increment.8) at the 1st impact to determine the possible energy loss and the reliability of the testing set-up during the impact test. The velocity vt and displacement st as a function of time can be obtained by the first and second integration of the acceleration history. using this method. and Ews (Baxter et al.5. vt-1 is the velocity at previous time increment. t is the present time increment. v0 is the initial velocity which is represented by the impactor velocity just before impact. Equations 4. The calculated energy loss is up to 3% and believed to be contributed by the friction between the guide pipe and the impactor. Equations 4. The details on the effects of these factors on the impact behaviour of pultruded tubes.6) st = ½ (vt + vt-1 ) (t – (t–1)) (4.2.4) Ews = ∫ (4.2) was compared to the measured energy value (Equation 4.3. the energy absorbed by the steel cap and its support (spring).2.3) st = ∫ (4. and (s–1) is the previous displacement increment. Fs-1 is the load at previous displacement increment. this result indicates that an Behaviour of glass FRP composite tubes under repeated impact for piling application 79 . It should be noted that Ews value is obtained from the curve whereby the load is plotted as a function of displacement. s is the present displacement increment. Ft = mat vt = ∫ (4. vt = ½ (at + at-1 ) (t – (t–1)) (4.Chapter 4 – Investigation on the impact behaviour of FRP tubes EJ Guades Equation 4.8) where at is the acceleration at present time increment. however. heat or the support.8 illustrate the relationship in calculating the values of vt.2.2) + v0 (4. respectively. and Es.3 and 4. (t–1) is the previous time increment. The work done or energy Ews (or Ewt) can be calculated using Equation 4.5) A discrete measurement trapezoidal rule was used for integration in finding the values of vt. As reported in Section 4.7) Ews = ½ (Fs + Fs-1 ) (s – (s–1)) (4. st..4 show the integral relationships of vt and st.

30 0.20 0.2 Typical acceleration-displacement curves in impact testing 4.40 0.20 0. As shown in Figure 4.40 0.40 0.40 0. tubes impacted by higher incident energies (423 J or more) ruptured when they were Behaviour of glass FRP composite tubes under repeated impact for piling application 80 .10 0.3 Experimental results and discussion 4.50 0.20 0.Chapter 4 – Investigation on the impact behaviour of FRP tubes EJ Guades acceleration data at the mid-height can be used to represent the response of the whole tube.30 0.00 0.00 0.10 0.30 0.00 0.20 E160 300 1st impact 40 90 130 250 200 150 100 50 0 0.40 0.10 0.30 Displacement (m) Acceleration (m/s2) Acceleration (m/s2) 350 0.10 Displacement (m) 350 E480 300 1st impact 40 90 130 250 200 150 100 50 0 0.30 0. however.20 0. 350 E630 300 1st impact 10 20 40 250 200 150 100 50 Acceleration (m/s2) Acceleration (m/s2) 350 0 E320 300 1st impact 40 90 130 250 200 150 100 50 0 0.1).50 Displacement (m) (a) Collapsed tubes 0.1 Mode of damage Figure 4.3 shows the condition of the collapsed (failed/ruptured) and non-collapsed composite tubes at the end of repeated impact tests.40 0.50 Displacement (m) (b) Non-collapsed tubes Figure 4.50 E210 300 1st impact 40 90 130 250 200 150 100 50 0 0. are not examined.00 0.10 0.3a. This was supported by an analytical study presented in Appendix C (Section C.50 Displacement (m) 100 50 0 Acceleration (m/s2) Acceleration (m/s2) 0.00 0.30 0.10 Displacement (m) 350 350 E420 300 1st impact 40 90 130 250 200 150 0.00 0.3.20 0.50 0. The details on the effects of these factors.

Chapter 4 – Investigation on the impact behaviour of FRP tubes EJ Guades subjected to at least 45 impacts.3 Conditions of the tubes after impact test Behaviour of glass FRP composite tubes under repeated impact for piling application 81 . However. composite tubes impacted by lower incident energies (318 J or less) did not show visible damage even up to 130 impacts as illustrated in Figure 4. The head of the collapsed tubes was observed to be the most severely damaged portion (end crushing) and the damage was manifested by the formation of matrix cracks and glass fibre ruptures.3c). (a) Collapsed/failed tubes (b) Non-collapsed tubes (c) Scanned images showing typical micro-cracks on the tube (E320) Figure 4. Axial splits along the four corners of the tubes were observed and both external and internal fronds curled downwards.3b. On the other hand. microscopically-scanned images showed that micro-cracks have occurred on the top portion of the non-collapsed tubes (Figure 4.

The damage observed on the tube at this point was characterised by the formation of intra. This phenomenon can be associated to the impact damage characteristics of fibre composite materials.3.Chapter 4 – Investigation on the impact behaviour of FRP tubes EJ Guades 4.. This load was obtained from the load history curve of each impact number (e. The higher the number of impact the composite tube is subjected.g.5).3. In spite of the difference in damage intensities that occurred starting from the initiation of collapse to 130th impacts. a clear formation of lamina bundles which bent inwards and outwards due to flexural damage can be noticed. This indicates that the effect of number of impacts in this region is more significant on the severity of the macroscopic physical damage than the peak load response of the repeatedly impacted tubes. However. interestingly the peak load values at the post-collapse region is approximately similar. 1997a).e.. there are no damage indications on surfaces by visual inspection but internal damage (called barely visible impact damage) may have already occurred (Zhang and Richardson. debris wedge of pulverised materials were formed on the surface of the tube. Behaviour of glass FRP composite tubes under repeated impact for piling application 82 .4 are the peak load of each corresponding impact. the impacted tube initially remained intact and no visible damage was sustained up to 40 th impacts.2 Progressive failure pattern Figure 4. This finding was also supported by the results obtained on the peak load versus number of impacts curves of non-collapsed tubes (discussed in Section 4. the massive is the damage brought by the impact event. it can be noticed from the impact load versus number of impact relationship curve that its peak load is apparently reduced. This value apparently indicates the initiation of collapse of the repeatedly impacted composite tube.4 displays the damage progressions of a typical collapsed composite tube (i. 2007). It should be noted that the impact load data shown in Figure 4. E480) repeatedly impacted and its corresponding peak load response. In the damage progression curve. The presence of the internal damage led to the strength degradation of the fibre composite material.3.. Figure 4. The debris formation is a result of the friction between the bent bundles and contact surface of the drop mass (Mamalis et al. The damage on the impacted tube continued to grow in the post-collapse region up to 130th impacts.and inter-laminar cracks with simultaneous development of axial splits along its corners. The damage was visually noticed on the impacted tube at approximately 55th impacts where the peak load reaches to its relatively lowest value. Additionally. This time. When they are subjected to impact loading.2).

the impact energy (drop mass and/or height) has to be increased more in order to collapse the tube up to these impact number. It is worth noting that both 1st and 10th impacts are located in the pre-collapse region of the peak load-number of impacts curve of E630 as can be seen in Figure 4. the micro-cracks developed on the top of the tube caused the load reduction.1 Load histories of the impacted tubes Figure 4.3 Impact load 4.6a. The linearity of the curve up to peak implies that the impacted tube has been damaged only minimally and impact repetitions up to 10th impact did not cause any significant change in the load histories. Apparently.5 shows the load histories of the impacted tubes after the repeated impact tests. This behaviour was also observed in the case of E480 (1st and 40th impacts) and E420 (1st. Figure 4.3.4 Damage progressions of collapsed tube impacted by 476.8 J 4. its effect on the shape of the curves was not seen to be an influential factor. Both curves demonstrated a linear increase at the beginning up to peak and drops until dissipation.3. it can be observed that these Behaviour of glass FRP composite tubes under repeated impact for piling application 83 . and 90th impacts). As emphasised previously.6a. 40th.3.Chapter 4 – Investigation on the impact behaviour of FRP tubes 20th impact 40th 55th 60th EJ Guades 80th 100th 130th 6000 Peak load (N) 5000 4000 3000 2000 1000 0 0 20 40 60 80 Number of impacts 100 120 140 Figure 4. In Figure 4.5a indicates that the load-time curves of the 1st and 10th impacts of E630 exhibited almost identical behaviour except that the peak load value of the latter is slightly reduced. The load histories of the 30th and 40th impacts of E630 showed series of peaks until reaching their maximum values. However.

40th. fibre splitting. E210 and E160). 90th.e. the nature of these load histories pointed out that significant damage has not been introduced on the impacted tube even after the 130th impact (Figure 4. troughs and peaks are imminently formed. and fibre ruptures although the sequence of the fracturing process cannot be distinctively followed due to the dynamic nature of the impact loading. The damaged portion was deflected together with the impactor as a result of a more compliant tube making the contact duration between them longer as expected. it was observed that the mass impactor rebounded consistently upon hitting the tube thereby producing a shorter contact duration between them. the characteristics of the load-time curves of noncollapsed tubes are identical to that of collapsed tubes at the pre-collapse region. the mass impactor moved deeper into the composite.. the load histories of collapsed tubes can be described by either one of the collapsed tubes as all of them exhibited similar load-time curves. Experimental results presented in Figure 4. One notable observation on the distinction of load histories of collapsed tubes between the two regions (i. The similarity of their behaviour is likely to happen as both tubes are in their undamaged conditions.3b). Note that the maximum peak load can be obtained right before unloading happened and usually dependent on the contact duration (related to the maximum downward deflection of the tube) between the mass impactor and the contact surface of the tube.) This indicates that as soon as damage was induced on the tube. Behaviour of glass FRP composite tubes under repeated impact for piling application 84 . when significant damage occurred on the composite tube (post-collapse region). The load-time curves of E480 (90th and 130th impacts) and E420 (130th impact) showed series of peaks up to maximum value just like the load histories of E630 in the post-collapse region.Chapter 4 – Investigation on the impact behaviour of FRP tubes EJ Guades impact numbers are situated in the post-collapse region and conclusively damage is expected to be created (analogous to the post-collapse condition in Figure 4. As discussed earlier.4.e. The time of occurrence in the pre-collapse region is relatively short compared in the post-collapse region. However. The troughs and peaks can be attributed to the different fracturing mechanism of the tube in the forms of cracks. This observation was also valid for specimens impacted by lower incident energies (i.5b show that for E320. During the first few impacts (pre-collapse region). Interestingly. delamination.. Conclusively. and 130th impacts are similar. the loadtime curves of 1st. pre and post-collapse) is on the duration of occurrence of the maximum peak load.

Chapter 4 – Investigation on the impact behaviour of FRP tubes

7000

E630

E320

7000

6000

6000

1st impact
10
20
40

5000
4000

3000
2000

Impact load (N)

Impact load (N)

EJ Guades

1000

1st impact
40
90
130

5000

4000
3000

2000
1000

0

0
0

5

10

15

20

25

0

5

Time (ms)

E480

7000

20

25

6000
1st impact
40
90
130

5000

4000
3000

Impact load (N)

Impact load (N)

15

E210

7000

6000

2000

5000

1st impact
40
90
130

4000
3000
2000
1000

1000

0

0
0

5

10

15

20

0

25

5

7000

10

15

20

25

Time (ms)

Time (ms)

E420

E160

7000

6000

6000
1st impact
40
90
130

5000

4000
3000

Impact load (N)

Impact load (N)

10

Time (ms)

2000

1000

5000

1st impact
40
90
130

4000
3000
2000
1000
0

0
0

5

10

Time (ms)

15

(a) Collapsed tubes

20

25

0

5

10

15

20

25

Time (ms)

(b) Non-collapsed tubes

Figure 4.5 Impact load histories of repeatedly impacted composite tubes

4.3.3.2 Peak load progressions
The peak load progression of tested tubes under repeated impact is shown in Figure
4.6. Note that the impact load data indicated in this figure are the peak load of each
corresponding impact (average of two replicates). The peak load was obtained from
the load history curve of each impact number (e.g., Figure 4.5). It should be noted
that the peak load value in Figure 4.6 is the measured load (via recorded
acceleration) at the mid-height of the tube. This value assumed the overall load
response of the tube and was used in comparing the load value per impact. It was
observed, however, that there was a variation of load (or stress) response along the
height (longitudinal and transverse directions) of the tube when it is subjected by

Behaviour of glass FRP composite tubes under repeated impact for piling application

85

Chapter 4 – Investigation on the impact behaviour of FRP tubes

EJ Guades

impact loading. The discussion on the variation of impact stress with the height of
the tube is presented in Appendix C. The data points in Figure 4.6 show fluctuations
of peak load values which are probably due to the dynamic nature of the test and
different fracturing mechanisms (in the case of collapsed tubes) that occurred.
Nevertheless, a clear trend (using a solid line) can still be followed distinctly on the
peak load evolutions of the impacted tubes. For collapsed tubes (Figure 4.6a), all
three cases had a very similar trend that described their pre- and post-collapse
behaviours. Their trend line suggested that the peak load values initially decreased
(first region) up to the start of collapse and become constant upon reaching the postcollapse region (second region). By closely examining the propagation of peak load
in the second region, one can apparently deduce that the peak load values after 130
impacts are expected to be relatively similar if it would have been continuously
impacted.
The findings obtained by the present study on the peak load evolutions of the
repeatedly impacted tubes in the second region were also observed in previous
studies (Yang et al., 2009; Mamalis et al., 1997a; and Czaplicki et al., 1991). These
earlier studies, however, crushed the composite tubes progressively and described the
post-collapse behaviour in terms of displacement and not on the number of impacts
as adopted in the present study. As observed in the experiment, the number of
impacts is very much associated to the axial displacement at the top of the tube and
both exhibit dependency with one another. This can be evidenced by Figure 4.4 in
which there was an apparent increase of damaged materials at the top of the tube
with increasing number of impacts.
A clear disparity observed between the peak load evolutions of the collapsed
tubes is the location of the start of collapse whereby the specimen impacted by lower
incident energies endured more impacts than the other. The number of impacts
required to commence collapsing the composite tube (Nf) is approximately 20, 57
and 95 for E630, E480 and E420, respectively. By considering these numbers of
impacts, it can be established that the peak load degradation of collapsed tubes is
more rapid if it is impacted by higher incident energy. Unlike collapsed tubes, the
corresponding trend line of non-collapsed tubes (Figure 4.6b) indicated a single-line
peak load value behaviour up to 130th impacts. As emphasised in Section 4.3.2, the
strength degradation is possible even without the manifestation of visible damage on
the fibre composite materials. The peak load value can still be potentially reduced
Behaviour of glass FRP composite tubes under repeated impact for piling application

86

Chapter 4 – Investigation on the impact behaviour of FRP tubes

EJ Guades

with the presence of micro-cracks (Figure 4.3c) as illustrated in Figure 4.6b in the
case of E320. It is worth noting that the nature of the peak load evolutions of the
non-collapsed tubes can be categorised as the peak load response of the collapsed
tubes in the pre-collapse region.
8000

8000

E630

6000

Peak load (N)

Peak load (N)

6000

4000

2000

0
10

20

30

Number of impacts

8000

40

2000

50

0

20

40

60

80

100

120

140

80

100

120

140

100

120

140

Number of impacts

8000

E480

6000

E210

6000

Peak load (N)

Peak load (N)

4000

0
0

4000

2000

0

4000

2000

0
0

20

40

60

80

100

Number of impacts

8000

120

140

0

20

40

60

Number of impacts

8000

E420

E160

6000

Peak load (N)

6000

Peak load (N)

E320

4000

4000

2000

2000

0

0
0

20

40

60

80

100

Number of impacts

(a) Collapsed tubes

120

140

0

20

40

60

80

Number of impacts

(b) Non-collapsed tubes

Figure 4.6 Peak load progressions of repeatedly impacted tubes

4.3.4 Impact energy
Figure 4.7 shows the typical energy history curves during impact test of the FRP
composite materials (Sevkat et al., 2010 and Sugun and Rao, 2004a). This figure
demonstrates the two distinct cases during impact test between the interaction of the
fibre composite materials and the mass impactor. The shape of the curves of the two
cases depends primarily on the energy absorption capability of the impacted

Behaviour of glass FRP composite tubes under repeated impact for piling application

87

Chapter 4 – Investigation on the impact behaviour of FRP tubes

EJ Guades

composite materials. Rebound case (Case 1) is likely to happen if the energy
absorbed by the composites is very small and the impactor tends to bounce back
from the impacted surface. In this condition, all the impact energy is completely
transferred from the projectile to the target where it is stored elastically or absorbed
via creation of damage. Upon unloading, the stored elastic energy (rebound energy)
is used to accelerate the now rebounding impactor. On the other hand, no rebound
will occur if most of the energy is absorbed by the impacted composites in a form of
damage. Once the impact energy is absorbed mostly by the composites, penetration
(or perforation) case (Case 2) usually happens.

Case 2

Rebound energy

Energy

Case 1

Absorbed energy

Absorbed energy ≈
Impact (total) energy

Time

Figure 4.7 Typical energy curves. Rebound and penetration (perforation) cases
4.3.4.1 Energy histories of the impacted tubes
The energy-time records of the composite tubes under drop-weight test are shown in
Figure 4.8. It should be noted that the energy values in the curve were calculated
using Equation 4.8. For collapsed tubes (Figure 4.8a), both rebound and penetration
cases were observed during the test regime. In this study, penetration/perforation
means the start of collapse or end crushing of the tube, as compared to the composite
plates where the perforation is generally characterised by the formation of a hole on
the impacted surface. The energy histories of E630 showed that while initial impacts
(1st and 10th) produced the rebound case, the later impacts (30th and 40th) created the
penetration case. This condition was also noticed in specimens E480 and E420
whereby impact successions provided two distinct energy curves. This indicated that
the impacted tubes had only endured minimal damage (micro-cracks) during the first
few impacts enabling them to develop significant rebound energy. However when
the damage started to increase due to impact repetitions, the rebound energy was

Behaviour of glass FRP composite tubes under repeated impact for piling application

88

Chapter 4 – Investigation on the impact behaviour of FRP tubes

EJ Guades

almost insignificant and all of the impact energy was absorbed by the tube. These
results support the fact that successive impacts enhance the damage of the composite
materials ensuing in an increase in the absorbed energy (Sevkat et al., 2010). It is
worth noting that by comparing the energy histories and the peak load progressions
of the collapsed tubes, the rebound and penetration case occurred in the pre-collapse
and post-collapse region, respectively. The damage was fully introduced in the
collapsed tubes at the second region (shown in Figure 4.4) and as expected the entire
impact energy was absorbed by them. On the other hand, the energy histories of noncollapsed tubes (Figure 4.8b) showed only rebound case regardless of the number of
impact repetitions. This is because the damage introduced to the composite tubes in
the form of micro-cracks up to 130th impacts was not sufficient to cancel out the
rebound energy.
Most of the instrumented drop-weight impact testing machines generally
mounts the recording sensor on the impactor. For repeated impact tests having
uniform applied incident energy, the value of the impact (total) energy (sum of
absorbed and rebound energies) recorded by the sensor is expected to be
approximately similar per impact number (Sevkat et al., 2010). However, it was
observed that the result from the present study is in contrary to the aforesaid
statement. The numerical values of impact energy recorded by the accelerometer
apparently decreased with increasing number of impacts as shown in Figure 4.8. It
should be reminded that in the present study, the sensor was placed on the mid-height
of the tube and not on the impactor itself. This technique of sensor placement
provided significant reductions of impact energy from the 1st impact up to the
maximum number the tube has impacted (i.e., 45 or 130). The impact energy
recorded by the sensor was reduced as a consequence of the damage developed on
the top of the tube that provides as an extra energy absorber. Conclusively, the
applied energy during the test at this point is equivalent to the energy recorded at the
location of the sensor and the energy being absorbed by the top end of the tube due to
damage. In this study, however, the absorbed energy due to the damage at the top of
the tube is not quantified. Instead, the energy calculated at the mid-height assumes to
represent, generally, the energy absorption behaviour of the tube. To avoid confusion
on the rate of energy absorption of the impacted tubes, this study adopted damage
degree variable. This variable was recently proposed by Belingardi et al. (2008) to
account for the damage accumulation in composites. Its value is numerically
Behaviour of glass FRP composite tubes under repeated impact for piling application

89

2 Degree of damage of the impacted tubes Figure 4. respectively. The curves of non-collapsed tubes (Figure 4.. E630 1st impact 10 20 40 500 400 300 200 100 0 10 20 30 60 70 500 400 300 200 100 0 80 10 20 30 200 100 50 60 70 80 50 60 70 80 50 60 70 80 E210 1st impact 40 90 130 500 300 40 Time (ms) 600 Impact energy (J) 400 50 E480 1st impact 40 90 130 500 40 Time (ms) 600 Impact energy (J) 1st impact 40 90 130 0 0 400 300 200 100 0 0 0 10 20 30 40 Time (ms) 50 60 70 0 80 E420 600 500 400 10 20 30 300 200 40 Time (ms) E160 600 1st impact 40 90 130 Impact energy (J) Impact energy (J) E320 600 Impact energy (J) Impact energy (J) 600 1st impact 40 90 130 500 400 300 200 100 100 0 0 0 10 20 30 40 50 60 70 80 Time (ms) (a) Collapsed tubes 0 10 20 30 40 Time (ms) (b) Non-collapsed tubes Figure 4.7.g. the absorbed and impact (total) energies were determined by making use of Figure 4.8 Impact energy histories of repeatedly impacted composite tubes 4. Behaviour of glass FRP composite tubes under repeated impact for piling application 90 .3. The value of the energy for each impact in Figure 4.7.Chapter 4 – Investigation on the impact behaviour of FRP tubes EJ Guades equivalent to Eabs/Eim where Eabs and Eim are the absorbed and impact (total) energies. From the energy history curve.9 was obtained based from its corresponding energy history curve (e.9b) suggests that the rate of energy absorption was higher for tubes impacted by higher incident energies indicating that heavier impacts induced more damage than lighter one.9 shows the damage degree-number of impacts curves of the impacted composite tubes.8). Figure 4. and was defined schematically in Figure 4.4. These tubes apparently absorbed energy very quickly due to their fast damage accumulation.

90 0. One difference that was observed from the current study in comparison with the results from the studies conducted by Belingardi et al.00 0. (2008) is the different magnitude of values of the degree of damage.00 0. However.70 0. delamination.50 J 158.50 0 20 40 60 80 100 120 140 Number of impacts (b) Non-collapsed tubes Figure 4. Eabs/Eim 1.90 0.80 0. and fibre ruptures substantially increased. On the other hand.60 E630 E480 E420 0.70 0.84 J 211. the effect of the incident energy was only seen in the pre-collapse region.9a).50 0 20 40 60 80 100 120 140 Number of impacts (a) Collapsed tubes Damage degree.9 Comparison of the damage degree curves of repeatedly impacted tubes Behaviour of glass FRP composite tubes under repeated impact for piling application 91 .92 J 0. it was observed that a small value of rebound energy though negligible was still recorded in the present study and a value of one was not ultimately reached during the test.Chapter 4 – Investigation on the impact behaviour of FRP tubes EJ Guades However for collapsed tubes (Figure 4. It is interesting to note that the number of impacts did not significantly change the value of the damage degree in the post-collapse region. The rate of energy absorption after the initiation of collapse became similar regardless of the magnitude of the incident energies applied. The value of the damage degree of the latter approached one during complete perforation (no resistance offered by the laminate). it was clear that the accumulated physical damage on the composite tube in the form of matrix cracks.60 317. Damage degree.80 0. Eabs/Eim 1.

. Nf was also chosen as the damage tolerance limit.10 illustrates the incident energy and the corresponding Nf of two different impactor masses (i. 2005).20 1. 25.93 476.5 20 E480-1 25.8 32 E480-2 21. Generally.57 Incident energy (J) 634.4 Summary of Nf values Specimen ID E630-1 Drop mass (kg) 25.20 3..Chapter 4 – Investigation on the impact behaviour of FRP tubes EJ Guades 4.3. damage degree approaches the value of unity).e. The summary of Nf values for different incident energies is displayed in Table 4. the number of drops to failure (Nf) can be used to define their damage tolerance limit (Datta et al. the data points on incident energy-number of impacts curves follow a logarithmic (or exponential) Behaviour of glass FRP composite tubes under repeated impact for piling application 92 . 2004 and Ho et al. These values were obtained from the peak load progression curve (e. number of impacts.00 476. These studies specified Nf as the number of impacts until total perforation of the laminate specimen (i.5. Figure 4. For repeated impacts.0 48 423..00 a average value from 2 replicates 4.1 Effects of incident energy and number of impacts Figure 4.0 95 E420-2 21...3. In the present investigation. 2004).20 Drop height (m) 2.20 1. drop mass.56 2. This index can be defined as the number of impacts to initiate collapse/failure on the impacted tubes which was used to characterise the effects of incident energy.20 kg) is not considered in the plot due to insufficient data points.5 Impact damage tolerance limit The damage tolerance in composite laminates is usually studied by determining the effects of different impact energies on their residual strengths (Sanchez-Saez et al.g. Note that the other mass (16.25 476. It should be noted that Nf is the result of the impact test based from the test matrix shown in Table 4.8 51 E480-3 16.8 57 E420-1 25. Table 4.71 423.56 2.56 kg).6) of each corresponding impact test condition.4.20 and 21.e. Identifying of the Nf values were achieved using the peak load progression curve and visual inspection on the tube during the progress of the test.3. and impact velocity (or drop height).5 Number of impacts to initiate failure Nf a 13 E630-2 21.56 3..00 634.

However. and Azouaoui et al. It is worth noting that at this level of energy. On the other hand. the significance of the mass variation becomes less when the value of the incident energy is above Ec. At this point. This is because all of the applied incident energies higher than Ec will have a corresponding Nf of 1. 2004.e.. this incident energy corresponds to a critical energy Ec that will fail/collapse the composite tube for a single impact. the more the propagation of damage slows down which is the case of high cycle fatigue. Figure 4. 2004. the damage will occur quickly and only a few numbers of repeated impacts will make the FRP composite tubes to collapse for an energy level higher than 600 J for both masses. This was also observed in the present study whereby the points shown in the figure follow a logarithmic curve. It can be observed from the graph that the effect of the variation of the incident energy is significant only up to roughly 1000 J (see intercepts of trend lines). 2007). It is worthwhile to note that the degree of separation between the two curves increased as the number of impact increases.Chapter 4 – Investigation on the impact behaviour of FRP tubes EJ Guades relationship as observed in most of the studies conducted on impact fatigue (Datta et al. 2007). These results conclusively show that energy levels are the major damage factor for lower number of impacts.56 kg) coincide with each other. the curves with the corresponding drop masses (i. This can be considered as a case of low cycle fatigue (Azouaoui et al.10. This certainly implies that the effect of impactor mass is more significant when the number of impacts takes a higher value than its lower counterpart. 2007). The more the energy level decreases. Behaviour of glass FRP composite tubes under repeated impact for piling application 93 .20 and 21. for a range of incident energy between 300 J and 600 J. The collapse of the composite tubes is more imminent if it is impacted by higher incident energy. the degree of damage of the tube was found to be less rapid than previously. On the other hand. As can be seen from Figure 4. For an energy level lower than 300 J. This result is interesting. 25. since this can serve as a basis in prioritisation between them for design purposes.10 indicates that the incident energy varies inversely with Nf. Ho et al. This case was described as the case of endurance fatigue (Azouaoui et al.. the curves showed that the collapse of the tubes was very slow and the slope angle of the curve have a tendency to approach to zero... the number of impacts becomes the dominant factor as soon as the value of incident energy was reduced...

5ln(x) + 1034. This finding was also reported in the study performed by Sugun and Rao (2004b). The slope of the line for each case indicates that the number of impacts to collapse an FRP tube was decreasing with increasing incident energies. Reasonably. The trend of the curves shown in Figure 4.56 kg y = -136.3.1ln(x) + 1053.56 kg 400 25.6 y = -164. The slope of the line with the lowest incident energy (i. Using the equation of the trend line.1 600 21. thus nullifying the effect of drop mass for a higher incident energy. however.e.20 kg 200 0 0 20 40 60 Nf 80 100 120 140 Figure 4. For a near-zero slope case. 423 J) provided the highest value. On the other hand. the drop mass had a significant effect on the composite tubes at lower incident energies leading to earlier collapse.5. This suggests that the effect of drop mass at different energy levels takes into account on the impact damage tolerance limit of the composite tubes.10 Incident energy vs.11 suggests that they tend to meet at a relatively higher drop mass. it is apparent that the values of Nf along this line are approximately similar. This effect. It can be clearly observed from the figure that Nf decreases with increasing impactor mass for all cases. this range of drop masses is considered as the critical drop mass mc whereby the failure of the tube will occur for one impact.. The deviation of the drop mass provides a contribution in the damage tolerance of the impacted tubes when the Behaviour of glass FRP composite tubes under repeated impact for piling application 94 . a near-zero slope value can be observed on the tube impacted by 634 J.2 Effects of drop mass The relationship between the Nf and drop mass for the three incident energies is shown in Figure 4.11. gradually reduced as the incident energy increased. At this point. Nf curve of repeatedly impacted tubes 4. it is expected that the curves will meet approximately at drop mass between 28 to 45 kg at Nf =1.20 kg 21. However.Chapter 4 – Investigation on the impact behaviour of FRP tubes EJ Guades 800 Incident energy (J) 25. their emphasis was concentrated on composite laminates and not on tube as investigated in the present study.

Ein = ½mv2 (4. On the other hand.10).8242x + 37. their associated effects can be inferred logically. This figure shows that.5 J 0 10 15 20 25 30 Drop mass (kg) Figure 4.6535x + 102.77 J y = -2. In this section.462x + 387.e. the curve follows an exponential (or logarithmic) curve just like the Ein –Nf curve (Figure 4.769 Nf 80 423 J 60 476. Therefore we can infer that the effects of impact velocity on the impact damage tolerance of composite tubes are somehow comparable to that of impact energy at a given m.10.51 J y = -0. The curve indicates that the failure or collapse of the tube is quicker under higher level of velocity (7 m/s or above) at a given mass.12 demonstrates the relationship between the Nf and impact velocity for the two drop masses.11 Nf vs.8 J 40 20 634. Figure 4. Nf =1). it is apparent that v is directly related with h.01 J y = -13. Equation 4.35 634.9.3 Effects of impact velocity and drop height The relationship between the impact velocity v and drop height can be defined mathematically using Equation 4. v=√ (4.1 can be written in a form shown in Equation 4. the rate of damage of the tube was found to be to be less rapid under a relatively lower velocity (below 7 Behaviour of glass FRP composite tubes under repeated impact for piling application 95 . 120 423.10. it is clear that the incident energy and the impact velocity are directly related for a given m. Therefore.3.5. the effect of impact velocity was investigated.9) where g is the gravitational constant.23 100 476. in general..Chapter 4 – Investigation on the impact behaviour of FRP tubes EJ Guades value is smaller than mc. its effect becomes less important when the drop mass is higher than mc since the tube impacted by a relatively higher mass will all fail at same impact number (i. drop mass curve of repeatedly impacted tubes 4.10) By virtue of Equation 4. From this equation. On the other hand.

The experimental investigation showed that the failure of the square composite tube subjected to Behaviour of glass FRP composite tubes under repeated impact for piling application 96 . as at this Nf both curves are expected to meet each other.4 Conclusions Repeated impact tests were carried out on square FRP composite tubes over a range of incident energies to determine their impact behaviour. Figure 4.08x 100 Nf 80 60 21.56 kg becomes nearer).56 kg y = 76286e-1. The occurrence of Nf =1 for a given m indicates a critical impact velocity vc that will fail/collapse the tube for one impact. The number of impacts to initiate collapse was used to characterise the effect of incident energy. its influence on the impact damage tolerance limit becomes insignificant since all of these velocity values will have a corresponding Nf value of 1. 120 25.993x 21. drop mass. By using the equation of the trend line of each curve indicated in the figure.56 kg 40 25.12 also shows that the effect of the variation of drop mass on the damage tolerance of tubes is more significant for lower level of impact velocity. This result can be substantiated by Equation 4. When the value of the impact velocity is higher than vc. The effect of the variation of impact velocity is dominant when the value of the impact velocity is less than vc. In fact. the drop mass variation effect becomes zero at Nf =1 (or at vc). Increasing the impact velocity will reduce the effect on the mass variation (curves of 25. however. and impact velocity (or drop height).Chapter 4 – Investigation on the impact behaviour of FRP tubes EJ Guades m/s). the more the propagation of damage slows down (related to high cycle fatigue).20 kg y = 14806e-0.12 Nf vs. number of impacts.10 whereby it shows that m and v are inversely related for a given Ein. The more the velocity decreases.2 and 21. impact velocity curve of repeatedly impacted tubes 4. it follows that the impact velocity will be in the range between 9 to 11 m/s at Nf =1.20 kg 20 0 0 2 4 6 8 10 Impact velocity (m/s) Figure 4.

Moreover. the post-impact mechanical properties (residual properties) of composite tubes subjected by repeated axial impact is discussed. This formation is attributed by the friction between the bent bundles and the contact surface of the impactor. This demonstrates that the effect of the variation of the applied incident energy can be neglected. The repeated impact curve of the failed tubes shows that incident energy is inversely related to the number of impacts. This result indicates that the effect of impact repetitions in the post-collapse region is more significant on the multiplication of physical damage than the peak load response of the repeatedly impacted tubes.Chapter 4 – Investigation on the impact behaviour of FRP tubes EJ Guades repeated impact is generally dominated by crushing at its top end. Moreover. a single test under this condition can already represent the behaviour of the noncollapsed tubes subjected by repeated impact loading. the peak load values remained constant. It is therefore important to study the effect of impact loading on the postimpact performance of the composite tube. Not only that it affected the instantaneous performance of the materials during the impact event but also it affected their bearing capacity. Development of external and internal fronds was also present during the failure of the tube. Thus. micro-structural observation on their surfaces revealed some micro-cracks occurred especially on the portion near the impact point. In Chapter 5. The shape of the load and energy history curves of the non-collapse tubes is approximately similar. This failure was characterised by matrix cracking and breaking of glass fibre reinforcement with simultaneous development of axial splits along its corners. Composite materials are sensitive to impact loading because even minor damage can affect their structural integrity. however. Though there was no visible damage observed on the non-collapsed tubes. The variation. Behaviour of glass FRP composite tubes under repeated impact for piling application 97 . In spite of the difference in damage intensities that occurred on the tested tube from the initiation of failure to the final state. is less important when the impacted tube started to fail. Micro-cracks were considered as the main reason on the peak load degradation of the impacted tubes. debris wedge of pulverised materials were formed on the surface of the tube. the drop mass and impact velocity (or drop height) have a pronounced effects on the damage tolerance limit of composite tubes at a relatively lower incident energy. This result provided the basis in prioritisation between them for consideration in the failure of the impacted composite tubes. It was found that the variation of incident energy and number of impacts are significant on the rate of energy absorption in the pre-collapse region.

2007. 2010... 1998. Impact damage has adverse effect on the load bearing capability of the materials. 2010. Most of these studies. A number of studies characterising the effects of impact events on the postimpact performance of composite tubes are available.... 1997. These micro-cracks are often difficult to detect which can result to premature catastrophic failure due to decreased strength caused by the impact loading. 1995). Mouritz et al. their residual tensile and flexural properties were also investigated (Belingardi et al. This mode of damage might not be the case for the non-collapsed tubes. However. For fibre composite materials. are limited on composite laminates for aerospace and automobile applications. the study on the effect of impact events to their residual properties has been very extensive. 1996).Chapter 5 – Residual properties of FRP composite tubes EJ Guades Chapter 5 Residual properties of FRP composite subjected to repeated axial impact tubes 5. which can significantly reduce their mechanical properties (Im et al. The residual compressive properties of composite laminates subjected to low velocity impact have been reported (Sanchez-Saez et al. 2007). however. it is of vital importance to have better understanding on their structural performance in the presence of impact damage in order to realise their potential.. and Davies et al. and Found and Howard. 2002. Therefore. particularly impact loading. Freitas and Reis.1 Introduction Composite materials have low resistance under dynamic loading. It was reported in Chapter 4 that the typical damage on the impacted tubes appeared in the form of matrix cracking and fibre fracture especially on collapsed or ruptured tubes. Wang et al. Behaviour of glass FRP composite tubes under repeated impact for piling application 98 . however. Santiuste et al. Ambur and Starnes. These materials are especially sensitive to impact loading since even minor damage can cause considerable reduction in structural integrity. 2012. Likewise. 1998. microstructural observation revealed that there was an occurrence of micro-cracks on their surfaces near the impact point. Wyrick and Adams. 1998. 2005. All of these studies. Short et al. referred to as “residual strength” or “strength-after-impact” (Zhang and Richardson.. 2001). Zhang and Richardson..

. 2010.Chapter 5 – Residual properties of FRP composite tubes EJ Guades focused on the residual properties of the tubes under transverse impact (Deniz et al. b (mm) 100. the tube was subjected to repeated impact loading. 2011. t (mm) 5. It was emphasised that. 2001). The details Behaviour of glass FRP composite tubes under repeated impact for piling application 99 . Abdallah et al. Table 5. and Chotard et al.1 Details of the specimen b d t Dimension Depth. the residual properties of square FRP pultruded tubes under repeated axial impact using experimental investigation is presented.. Table 5.1). Common results obtained from the studies on the residual properties of composite laminates or tubes under transverse impact revealed that impact damage significantly affect their post-impact performance.2.23 Series of tests were performed to characterise the residual properties of square composite tubes. The effects of the incident energy. and the drop mass on the residual properties of the tubes are emphasised. In this chapter.2.1 shows the cross sectional dimension (average value) of the tubes. in general. the comparisons between the residual strength and modulus. d (mm) Value 100.43 Length. The details of the dimension of the 9 tubes can be found in Appendix D (Section D. Minak et al. 2012. A total of 9 tubes were used following the test matrix presented in Section 5. 2005. as well as the three testing modes (compressive. After which the impacted tube underwent residual properties testing.. 2005).. First.22 Thickness. It was shown that their strength in a damaged component may have only 40% of that in an undamaged structural element (Sanchez-Saez et al. tensile. and flexural tests) are discussed.2 Experimental program 5..51 Width.2. Gning et al. Moreover. l (mm) 375.. 5.1 Test specimen and repeated impact testing The specimen used in characterising the residual properties of FRP composite tube has the mechanical properties similar to that of CT2. the reduction is largely dependent on the level of impact energy and the number of impacts the composite material was subjected. impact repetitions.

2. it was reported that there was no significant difference occurred on the mechanical properties between CT1 and CT2 specimens.6 10 (C/F)a NC (non-collapsed tube). This method is adopted to get rid of the drilled hole used in mounting the accelerometer which affects the uniformity of the cross section of the coupons used in the tests.2) and therefore the accelerometer may not be needed for the investigation.8 80 (C/F)a E630-30 21.1 shows the condition of the tube at the end of the impact test.56 3 634. Both visual and microscopic inspections were performed in documenting the damage on the impacted tube. the test matrix shown previously in Table 4.1 Behaviour of glass FRP composite tubes under repeated impact for piling application 100 .2 (Chapter 4) except that the tube is no longer instrumented.9 80 (NC)a E320-80 16.20 3 476. collapsed/failed and non-collapsed conditions).8 80 (NC)a E480-10 16. It should be noted that the tube IDs are referred from the incident energy and number of impact. After which the impacted tubes were subjected to residual properties testing.20 3 476.56 3 634.20 2 476.20 3 741. C/F (collapsed/failed tube).20 1 158.5 10 (NC)a E480-40 16.2. The length of the tube shown in Table 5.e. The repeated impact test was conducted following the test matrix shown in Table 5.1 was selected as this is the maximum length the impact testing set-up can accommodate.20 2 317. As a result. Figure 5.8 40 (C/F)a E480-80 16. The results obtained from Chapter 4 is very important as they provided an idea on the damage conditions of the composite tube under repeated axial impact (i. Table 5.Chapter 5 – Residual properties of FRP composite tubes EJ Guades of these tests are discussed in the next sections. The machine and procedure used in repeated impact test are similar to that presented in Section 4.2 (Chapter 4) served as a reference in coming up with the test scheme in Table 5. In Chapter 3. asee Figure 5. Additionally. The tube was then taken out for inspection to determine the extent of the impact damage.5 30 (C/F)a Tube ID E740-10 25.2.8 10 (NC)a E630-10 21.2 Repeated impact test matrix E0-0 Drop mass (kg) 0 Drop height (m) 0 Incident energy (J) 0 Number of Remarks impacts 0 Baseline tubea E160-80 16. the applied impact energy was characterised in terms of the incident energy (column 4 in Table 5.

The coupons that were used in the tests only include portions which are free from visible damage and were considered feasible for testing.1).e. 2012.2 illustrates the cutting plan of coupons used in the tests.Chapter 5 – Residual properties of FRP composite tubes EJ Guades Figure 5.. A total of four specimens were taken from each impacted tubes.2 Residual properties testing In general. It is worth noting that the baseline tube and the tube used in characterising the mechanical properties of CT2 (presented in Chapter 3) Behaviour of glass FRP composite tubes under repeated impact for piling application 101 . the whole portion of the non-collapsed tubes were considered feasible for testing as no sign of visible damage was observed on the entire length (see Figure 5.. E740-10) extended only to approximately 130 mm from the top of the tube. and Mouritz et al.. Wyrick and Adams. (2006) tested a coupon taken from the FRP composite tube to determine the residual tensile strength and to establish the repeated impact curve. this was achieved by cutting and excluding approximately 140 mm of length from the top of the tube.. The test specimens were cut on each face of the tube. For collapsed tubes. In the work presented here. On the other hand. This technique allows in comparing reasonably the residual properties since the coupons used in the tests are taken from one source only.2.e. It should be noted that the visible damage in a form of vertical cracks for the “most damaged tube” (i. the residual properties of composite tubes were characterised by determining the residual properties of the coupons cut from the impacted tube for each impact condition. Five specimens were also cut from the un-impacted tube (i.. On the other hand. 1998. 1997). Slicing of the coupons was carefully done by using a wet saw machine.1 Conditions of the tubes after impact test 5. E0-0) and tested to serve as the baseline value of the composite tubes. (2009) and Helmi et al. 2009. Belingardi et al. coupons cut from the impacted surface of the composite plates or panels are used to characterise their residual properties (Wang et al. Ballere et al. They assumed the overall behaviour of an impacted tube as they are sourced from the four sides of the tube. Figure 5.

only the coupons taken at the middle portion are instrumented.3 Details of the specimen for coupon tests Type of test Compressive Width. tensile. The length of the tensile test specimen is relatively shorter to that recommended by the standard (i.2 Cutting plan of coupons used in residual properties testing The compressive. 117. respectively). Table 5.3 displays the nominal dimension of the specimen used in the coupon tests. The summary of the dimensions of the specimens tested and the results of the whole test are presented in Appendix D (Sections D.Chapter 5 – Residual properties of FRP composite tubes EJ Guades using coupon tests are the same.5 shows the specimens used in the three tests. and flexural tests on coupons were conducted using the test procedures and machines similar to that of CT2 specimen presented in Chapter 3). Nevertheless.50 Length.50 mm (a) Non-collapsed tubes 375 mm (b) Collapsed tubes Figure 5.3 to 5. Middle and bottom coupons 5. its mechanical properties are again presented in this chapter for ease of analysis and discussions.25 Behaviour of glass FRP composite tubes under repeated impact for piling application 102 .25 Tensile 25 230 5.4). l (mm) 140a. Two of the compressive and tensile test specimens for impacted tubes were instrumented by a strain gage (6 and 20 mm long uniaxial strain gage.2 to D. b (mm) 12. t (mm) 5.. Figures 5.50 mm Bottom 117. It should be noted that for compressive test specimens from impacted tubes.e. Table 5. 250 mm) as this is the maximum length that can be obtained from the impacted tube.25 Flexural 15 150 a b Top coupon.5b Thickness. Reference line L1=70 mm Top Compressive Tensile Flexural 140 mm Middle 117.

3 Compressive test specimens Behaviour of glass FRP composite tubes under repeated impact for piling application 103 .Chapter 5 – Residual properties of FRP composite tubes EJ Guades (a) Before the test (b) After the test Figure 5.

4 Tensile test specimens Behaviour of glass FRP composite tubes under repeated impact for piling application 104 .Chapter 5 – Residual properties of FRP composite tubes EJ Guades (a) Before the test (b) After the test Figure 5.

Chapter 5 – Residual properties of FRP composite tubes EJ Guades (a) Before the test (b) After the test Figure 5.5 Flexural test specimens Behaviour of glass FRP composite tubes under repeated impact for piling application 105 .

microscopically-scanned images showed that micro-cracks have occurred on the top portion of the non-collapsed tubes after the test (Figure 5.6 Scanned images showing typical micro-cracks on the impacted tubes 5.3 Experimental results and discussion 5. The head of the collapsed tubes was observed to be the most severely damaged portion.6a). The strength (peak stress) and modulus values shown in in the tables are computed based from the calculations specified in the corresponding test standards (see Section 3. Axial splits along the four corners of the tubes were also observed and both external and internal fronds curled downwards. (a) Non-collapsed (b) Collapsed Figure 5.6b).6 shows the tested tubes (representative) under collapsed and non-collapsed conditions.3. Microcracks were also observed on a portion below the location where the visible vertical crack had occurred on the collapsed tubes (Figure 5.4 and 5. Tubes impacted by higher incident energies (477 J or more) collapsed when they were subjected to at least 40 impact repetitions.1 Mode of damage Figure 5.2 Summary of coupon test results Tables 5.2 to A. On the other hand.5 summarise the results of the residual properties testing of the impacted tubes.4 of Appendix A). All compressive residual properties values discussed in Sections 5.3.3. The values reflected in the tables are the mean value of the tested coupons. However.Chapter 5 – Residual properties of FRP composite tubes EJ Guades 5. composite tubes impacted by lower incident energies (318 J or less) did not show visible damage even up to 80 impacts.7 are Behaviour of glass FRP composite tubes under repeated impact for piling application 106 .3.5 of Chapter 3 and Section A.3 to 5.

38 41.30 37.690 Bottom portion Strength (MPa) - E160-80 425.50 41.2).37 E480-80 - 425.79 37.359 444. The results obtained from these specimens were used to characterise the effects of the incident energy.632 E630-10 601.604 900.06 37.14 441.67 41.543 E160-80 604.390 941.86 41.88 50.474 941.17 39.38 434.84 50.039 895.35 E630-30 - 416.937 E480-10 610.25 50.45 49. the impact repetitions.98 40.944 450.38 434.5 Summary of tensile and flexural tests results E0-0 Tensile properties Strength Modulus (MPa) (MPa) 603.37 49.050 E740-10 602.11 49.026 454.130 E480-80 603.53 41.280 441.39 E320-80 430.4 Summary of compressive test results E0-0 Top portion Strength (MPa) - Middle portion Strength Modulus (MPa) (MPa) 441. This was selected for comparison with other properties since the start of cutting line used at the middle portion is in same level with the tensile and flexural specimens (see Figure 5.33 37. the results acquired from the compressive tests on top and bottom portions were used to examine the variations of the residual strength with the height of the tube.72 41. and the variation of drop mass on the residual properties of the tubes subjected to repeated axial impacts.149 441.90 435.64 50.032 454.226 944.40 37.79 40.Chapter 5 – Residual properties of FRP composite tubes EJ Guades exclusively the values at the middle portion of the impacted tubes.24 E480-10 436.50 37.993 Tube ID Behaviour of glass FRP composite tubes under repeated impact for piling application 107 .649 444.98 E740-10 - 411.70 50.253 918. Table 5.81 Tube ID Table 5.698 Flexural properties Strength Modulus (MPa) (MPa) 994.462 E630-30 601.439 E320-80 606.01 38.707 955.62 49.44 38. On the other hand.55 E480-40 - 432.20 40.802 442.808 E480-40 611.803 899.84 E630-10 431.

and flexural strengths is 6. respectively. 0. 40. This result suggests that the reduction of tensile strength can be neglected.3. Although the samples that were tested in determining the residual properties do not have visible damage. Similarly. The increase. 40.6). Additional discussion as to the reasons why the tensile strength is less affected compared to the other two strengths are presented in Section 5. a clear gap of residual strengths can be noticed between 10 and 30 impacts at 634 J.8 and 5.8%.9).8 to 5.10. residual strengths decrease with increasing impact energy as can be observed in Figure 5. respectively) are connected by straight lines because no tests were run at the intermediate energies.6.8 and 5. The decrease from 10 to 40 impacts at 477 J is relatively lower to that from 10 to 30 impacts at 634 J. In fact.Chapter 5 – Residual properties of FRP composite tubes EJ Guades 5. 2007). This shows that the rate of reduction between increasing impact number becomes rapid when impact energy increases.3.7 shows the relationships of residual strength with increasing impact energies of the tubes subjected by 10. Figures 5. One comment is worthwhile making on the comparison of the strength reductions at these energy levels.7.3%.7 indicates that impact events result in reductions in strengths to varying degrees. 30. Figure 5. Appendix A). This finding can be confirmed by comparing the curves of 10 and 80 Behaviour of glass FRP composite tubes under repeated impact for piling application 108 . the maximum reduction of the tensile residual strength is lower than the standard deviation found in the unimpacted specimen (see Table A.6. and 80 impact repetitions. it follows that the effect of increasing the impact energy significantly reduced both residual compressive and flexural strengths.10 illustrate that the residual strengths of the tubes impacted by 10.3 Effects of impact energy Figure 5. It should be noted that the two data points (477 J and 634 J at 40 and 30 impacts. did not provide considerable reduction on the residual tensile strength of the impacted tubes (Figure 5. An enlarged view of the curves is displayed in Figures 5. and 80 impacts are in a reducing trend at 477 J. In general. however. The presence of the external or internal damage in a form of micro-cracks led to the strength degradation of the fibre composite material (Zhang and Richardson. tensile. microscopic observation revealed that microcracks were present on their surface (see Figure 5.10 to help in the discussions of the results. and 10% of their corresponding baseline strength. The maximum reduction of residual compressive. By examining closely these values and observing the trend of the data points on Figures 5.

6% and 4. respectively.7 Residual strength. The damage imparted by the mentioned impact energies only includes micro cracks along the impact point. respectively. It should be reminded that impacting the tube by 318 J (at 80 impacts) and 634 J (at 10 impacts) did not induced visible damage on the top of tubes as shown in Figure 5. it can be observed that there is a sudden drop of curve after increasing these impact energies to 477 J and 742 J. at 477 J. Residual strength (MPa) 1000 Flexural 800 Tensile 600 400 Compressive 10 impacts 30 40 80 200 0 0 100 200 300 400 500 600 700 800 Impact energy (J) Figure 5. The collapse of the tubes happened only after increasing the impact energy to 477 J and 742 J.1% of their corresponding baseline strength.impact energy relationships Behaviour of glass FRP composite tubes under repeated impact for piling application 109 .Chapter 5 – Residual properties of FRP composite tubes EJ Guades impacts. respectively.10 also demonstrate that the rate of reduction up to 318 J and 634 J at 80 and 10 impacts. Figures 5. However.impact energy relationships Residual strength (MPa) 450 Baseline 440 40 430 80 420 30 410 10 400 0 100 200 300 400 500 600 700 800 Impact energy (J) Figure 5. respectively.8 and 5. Their relative difference in compression and flexure increased from 0 to approximately 3. is relatively slow.8 Enlarged view: Residual compressive strength . This certainly shows that the effect in increasing the impact energy on the residual strength reduction of impacted tubes is more substantial when the tube collapsed. The separation between these curves becomes noticeable as the impact energy increases.1.

Moreover. Behaviour of glass FRP composite tubes under repeated impact for piling application 110 .14.12 and 5. In Figures 5. the residual modulus is slightly reduced with increasing impact energies compared to its strength.. increasing the incident energy does not provide significant difference on the value of residual modulus regardless of the condition of the tubes (i. This result points out the less sensitivity of the residual modulus in the interaction effect as compared to the residual strengths. increase of energy with increasing number of impacts). the tensile modulus is at far slightly above to its corresponding baseline values (Figure 5.4% of their equivalent baseline modulus. 40.13). The figures show that. respectively. the maximum reduction of residual compressive and flexural moduli is 1. from non-collapse to collapse condition).. On the other hand. and 80 impact repetitions is illustrated in Figures 5.10 Enlarged view: Residual flexural strength-impact energy relationships The residual modulus versus impact energy curves of the tubes under 10.3 and 4. generally.11 to 5.9 Enlarged view: Residual tensile strength-impact energy relationships Residual strength (MPa) 1000 Baseline 975 950 925 40 900 30 80 10 875 0 100 200 300 400 500 600 700 800 Impact energy (J) Figure 5. It can be observed from the figures that the residual modulus is less sensitive on the interaction effect of impact energy and number of impacts (i.Chapter 5 – Residual properties of FRP composite tubes EJ Guades 615 Residual strength (MPa) 40 610 605 80 Baseline 10 30 600 595 590 0 100 200 300 400 500 Impact energy (J) 600 700 800 Figure 5. 30.14.e.e.

13 Enlarged view: Residual tensile modulus.impact energy relationships Behaviour of glass FRP composite tubes under repeated impact for piling application 111 .11 Residual modulus-impact energy relationships Residual modulus (MPa) 51000 10 50500 80 40 50000 Baseline 49500 30 49000 48500 0 100 200 300 400 500 600 700 800 Impact energy (J) Figure 5.impact energy relationships Residual modulus (MPa) 42000 30 41500 40 41000 80 10 Baseline 40500 40000 39500 0 100 200 300 400 500 600 700 800 Impact energy (J) Figure 5.12 Enlarged view: Residual compressive modulus.Chapter 5 – Residual properties of FRP composite tubes EJ Guades Residual modulus (MPa) 60000 Compressive 50000 Tensile 40000 Flexural 30000 20000 10 impacts 30 40 80 10000 0 0 100 200 300 400 500 600 700 800 Impact energy (J) Figure 5.

subjecting the tube by 159 J with 80 impact repetitions yielded a 1. when the tube is impacted by 742 J. 0.8.3 and 10% of their corresponding baseline strength. Figure 5. It is apparent that impact repetitions at a specific level of energy enhanced the damage sustained by the FRP material. The maximum loss of residual compressive. respectively. and flexural strengths of the tube when impacted up to 80 repetitions is 6. On the other hand.18 show the relationships of residual strengths with increasing number of impacts at various energy levels. reducing its strength.15 to 5. respectively. The reduction. For instance. respectively. it can be ascertained that the values of the maximum reduction of residual modulus plotted against impact energy will be relatively the same when it is plotted in number of impacts.2 since both impact energy and number of impact are simultaneously considered in the plot.14 Enlarged view: Residual flexural modulus-impact energy relationships 5. One notable observation that can be achieved from the figures is that the effect of impact repetitions on increasing the strength reduction of impacted tubes Behaviour of glass FRP composite tubes under repeated impact for piling application 112 . thus.6 and 3.17 shows that the tensile strength of the impacted tube is not particularly sensitive to the increase of number of impacts. the number of impacts dramatically reduced both compressive and flexural strengths as clearly shown in Figures 5. Similarly.3. Likewise. Note that these values are identical to the values mentioned in Section 5.18.Chapter 5 – Residual properties of FRP composite tubes EJ Guades Residual modulus (MPa) 40000 30 39000 Baseline 80 38000 10 40 37000 36000 35000 0 100 200 300 400 500 600 700 800 Impact energy (J) Figure 5. It can be observed from the figures that the residual strengths of impacted tubes decrease with increasing number of impacts.16 and 5. depends on the magnitude of the applied impact energy. tensile.4 Effects of impact repetitions Figures 5. it needs only 10 impacts to reduce the compressive and flexural strengths by 6 and 10% of their baseline values.9% loss of its compressive and flexural strengths. however.3.

Residual strength (MPa) 1000 Flexural 800 600 Tensile 400 Compressive 200 158.8 J 420 634.8 J 158.9 J 430 476. This indicates that at same number of impacts.number of impacts relationships 450 Residual strength (MPa) Baseline 440 317. it follows that the loss of strength of impacted tubes is significantly contributed due to the increase of impact energy and not much on impact repetitions.5 J 410 741.15 Residual strength.Chapter 5 – Residual properties of FRP composite tubes EJ Guades varies inversely with impact energy. the reduction of strength when the tube is subjected by 742 J at 10 impact repetitions is comparably higher than when it is impacted by 159 J with 80 impacts. Looking on Figures 5.5 J 741. This finding was also found by Wyrick and Adams (1998) when they investigated the effect of repeated impact loading on the residual properties of composite laminate. This result is interesting since one can prioritise in choosing among these two factors for design purposes.6 J 0 0 20 40 60 80 100 Number of impacts Figure 5.6 J 400 0 20 40 60 80 100 Number of impacts Figure 5. the residual strengths after 80 impacts at 159 J suggest that it is approaching a threshold energy below in which significant reductions in strength are not observed. By carefully observing this relation.8 J 634. impacts at higher energy levels induce a greater loss in residual strengths of composite tubes than lighter impacts.9 J 317.18.16 Enlarged view: Residual compressive strength -number of impacts relationships Behaviour of glass FRP composite tubes under repeated impact for piling application 113 .16 and 5.8 J 476. In fact.

number of impacts relationships Residual strength (MPa) 1000 Baseline 975 158. The curves shown on the figures indicate that the residual modulus is slightly degraded when it is subjected by series of impacts.8 J 634.8 J 605 158. This result supports the finding highlighted previously in Section 5.8 J 925 900 476. Furthermore.5 J 600 Baseline 595 590 0 20 40 60 80 100 Number of impacts Figure 5. It can be observed from the figures that the repetition of impact did not significantly alter the value of modulus regardless of the energy levels.7 J 634.6 J 875 0 20 40 60 80 100 Number of impacts Figure 5.19 to 5.5 J 741.9 J 950 317.18 Enlarged view: Residual flexural strength-number of impacts relationships Figures 5.Chapter 5 – Residual properties of FRP composite tubes EJ Guades Residual strength (MPa) 615 610 317. Behaviour of glass FRP composite tubes under repeated impact for piling application 114 . The maximum percentage loss of the residual modulus is at far below than their strength counterpart. the outcome shows that the effect of the damage caused by impact repetitions on the tubes is more concentrated on the strength and not on the modulus. In fact.3.21) is slightly higher than the baseline value.22 show the comparison of residual modulus and number of impacts curves of impacted tubes under different energy levels. the residual tensile modulus of the impacted tube (Figure 5.17 Enlarged view: Residual tensile strength.9 J 741.6 J 476.3 that the modulus property of the impacted tube is less affected by the interaction of the impact energy and number of impact.

8 J 476.20 Enlarged view: Residual compressive modulus-number of impacts relationships Residual modulus (MPa) 42000 634.8 J 50000 158.9 J Baseline 49500 49000 634.8 J 317.6 J 50500 476.Chapter 5 – Residual properties of FRP composite tubes EJ Guades Residual modulus (MPa) 60000 Compressive 50000 Tensile 40000 Flexural 30000 20000 10000 158.8 J 634.9 J 317.19 Residual modulus-number of impacts relationship Residual modulus (MPa) 51000 741.5 J 41500 317.5 J 741.6 J Baseline 158.8 J 41000 476.5 J 48500 0 20 40 60 80 100 Number of impacts Figure 5.6 J 0 0 20 40 60 80 100 Number of impacts Figure 5.8 J 741.9 J 40500 40000 39500 0 20 40 60 Number of impacts 80 100 Figure 5.21 Enlarged view: Residual tensile modulus-number of impacts relationships Behaviour of glass FRP composite tubes under repeated impact for piling application 115 .

whereby increasing the mass intensifies the rate of the strength reduction of the impacted composite tube. This result shows that the relationship between the impactor mass and residual strength is non-linear. 25.6 J 38000 317. the strength reduction rises to 6.5 and 4. The data points plotted in Figures 5.5 Effects of mass of the impactor Figures 5. It is worth noting that incrementally increasing the original impactor mass by 30% up to 25. and 1.4 and 5.23 to 5.24 and 5. is more noticeable in the compressive and flexural strengths of the tubes (see Figures 5.8 J 741.4%.2 kg mass are 0. 21.22 Enlarged view: Residual flexural modulus-number of impacts relationships 5.24 and 5.3..26). In fact.Chapter 5 – Residual properties of FRP composite tubes EJ Guades Residual modulus (MPa) 40000 39000 634.8 and 9. the relationship between drop mass and impact Behaviour of glass FRP composite tubes under repeated impact for piling application 116 . Moreover. The degradation due to the effect of drop mass. the curves plotted in the figures are just a mirror image when it is plotted against impact mass instead of impact energy. The initial reductions of the compressive and flexural strengths of tubes impacted by 16.4 and 5.8 J 158.2 kg). the residual strengths of the tubes decreased significantly when increasing the impact mass.5 J Baseline 476.1% (flexural). however.2 kg provided an equivalent strength loss of 1.5% of their baseline values.02 and 3.9 J 37000 36000 35000 0 20 40 60 80 100 Number of impacts Figure 5. respectively. However. The effect of the drop mass on the reduction of residual strength varies proportionally with impact energy as can be observed in figures.e.e..26 for tubes subjected by 10 impact repetitions indicates that the increase of impactor mass reduced the residual strengths in varying magnitudes. when the initial mass is increased by approximately 60% (i.26 illustrate the plot of residual strength versus drop mass of the impacted tubes at different energy levels and number of impacts.6 kg) increased the loss of strength to 1. As can be seen from the figures. An additional 30% of the initial mass (i.9%. respectively.4% (compressive).

23 Residual strength-drop mass relationships at different energy levels and number of impacts 450 Residual strength (MPa) 16.24 Enlarged view: Residual compressive strength-drop mass relationships at different energy levels and number of impacts Behaviour of glass FRP composite tubes under repeated impact for piling application 117 .2 kg impactor mass.Chapter 5 – Residual properties of FRP composite tubes EJ Guades energy is directly proportional as expressed by equation 4.6 and 25.e. the residual strength of impacted tubes decreases when the number of impacts increases (10. a wider gap between these points can be observed.20 kg Compressive 400 0 100 200 300 400 500 600 700 800 Impact energy (J) Figure 5. from non-collapse to collapse) and their interaction with the number of impact can be inferred logically.26 show that when the tube was subjected by 16. Figures 5.20 kg Baseline 440 21. The associated effect of impact repetitions and mass impactor on residual strength degradation can be observed also at 21. however.20 kg 420 10 impacts 30 40 80 410 21.1 (Chapter 4)..20 kg 430 16. At these mass levels.2 kg. Residual strength (MPa) 1000 Flexural 800 Tensile 600 400 Compressive 10 impacts 30 40 80 200 0 0 100 200 300 400 500 Impact energy (J) 600 700 800 Figure 5.56 kg 16. This can be substantiated by a clear gap between the data points at this mass level.56 kg 25.24 and 5. 40 and 80 impacts). This result suggests that the associated effects of impactor mass and impact repetitions on the residual strength reduction of tubes is more pronounced for heavier mass. Therefore their associated effects such as their effect on the condition of the tube (i.

20 kg 16.20 kg 925 16.20 kg 21.29) is slightly higher than the corresponding baseline value. Conclusively.20 kg Baseline 25.20 kg 21. the curves in the figure point out that the maximum percentage reduction of residual modulus is comparably smaller than their equivalent strength. Behaviour of glass FRP composite tubes under repeated impact for piling application 118 .27 to 5. These figures show that increasing the drop mass does not provide significant difference on the value of residual modulus regardless of the condition of the tubes (i.20 kg 875 0 100 200 300 400 500 600 700 800 Impact energy (J) Figure 5.20 kg 605 16.56 kg 16.e. from non-collapse to collapse condition).56 kg 21..Chapter 5 – Residual properties of FRP composite tubes EJ Guades Residual strength (MPa) 615 610 16.56 kg 16. the residual tensile modulus of the impacted tube (Figure 5. As a matter of fact.25 Enlarged view: Residual tensile strength-drop mass relationships at different energy levels and number of impacts Residual strength (MPa) 1000 Baseline 975 950 21. The associated effects of the drop mass with the impact energy and number of impacts found to be less significant in the residual strength degradation of the impacted tubes.20 kg 10 impacts 30 40 80 900 25.30. the variation of impact mass is more significant on the reduction of residual strength than their corresponding modulus. Moreover.26 Enlarged view: Residual flexural strength-drop mass relationships at different energy levels and number of impacts The increase of impact mass slightly reduced the residual modulus of the impacted tubes as shown in Figures 5.56 kg 600 10 impacts 30 40 80 595 Tensile 590 0 100 200 300 400 500 600 700 800 Impact energy (J) Figure 5.

56 kg Baseline 49500 10 impacts 30 40 80 49000 21.20 kg 21.20 kg Baseline 40500 10 impacts 30 40 80 40000 39500 0 100 200 300 400 500 600 700 800 Impact energy (J) Figure 5.28 Enlarged view: Residual compressive modulus-drop mass relationships at different energy levels and number of impacts Residual modulus (MPa) 42000 41500 16.20 kg 16.Chapter 5 – Residual properties of FRP composite tubes EJ Guades Residual modulus (MPa) 60000 Compressive 50000 Tensile 40000 Flexural 30000 20000 10 impacts 30 40 80 10000 0 0 100 200 300 400 500 600 700 800 Impact energy (J) Figure 5.56 kg 21.20 kg 41000 21.20 kg 50500 16.56 kg 48500 0 100 200 300 400 500 600 700 800 Impact energy (J) Figure 5.20 kg 16.56 kg 25.27 Residual modulus-drop mass relationships at different energy levels and number of impacts 51000 Residual modulus (MPa) 25.20 kg 16.20 kg 50000 16.29 Enlarged view: Residual tensile modulus-drop mass relationships at different energy levels and number of impacts Behaviour of glass FRP composite tubes under repeated impact for piling application 119 .

Chapter 5 – Residual properties of FRP composite tubes

EJ Guades

Residual modulus (MPa)

40000

21.56 kg

39000

Baseline
16.20 kg

38000

25.20 kg

16.20 kg 21.56 kg
16.20 kg

37000

10 impacts
30
40
80

36000

35000
0

100

200

Flexural
300

400

500

600

700

800

Impact energy (J)

Figure 5.30 Enlarged view: Residual flexural modulus-drop mass relationships at
different energy levels and number of impacts
5.3.6 Comparison between compressive, tensile and flexural properties
Figure 5.31 shows the comparison between the compressive, tensile, and flexural
residual properties of the impacted tubes. The comparison was done by plotting their
respective strength (or modulus) retention factor (residual strength divided by
corresponding baseline strength) against increasing incident energy. It should be
noted that the value adopted in plotting the strength (or modulus) retention factor for
477 J and 634 J are the average among the derived values at different impact
numbers (i.e., retention factor at 10, 40, and 80 impacts for 477 J; retention factor at
10 and 30 impacts for 634 J). The compressive, tensile and flexural strength retention
factors at 742 J are 0.93, 0.99, and 0.90; respectively. Based from these values and
the observation we can get from Figure 5.31, it can be deduced that the flexural
strength is the most severely affected by the impact event compared to the other
strengths. The reason of this phenomenon can be explained by the following. During
the impact event, the typical impact damage of FRP composite materials appears in
the form of matrix cracking, delamination, and fibre shear-out and fracture. Whilst
matrix cracking or delamination reduces the compressive strength, fibre shear-out or
fracture decreases the tensile strength of the composites (discussed in the next
paragraph). It should be noted that in flexural testing, the composite material is both
subjected by compressive and tensile forces. Therefore, it is expected that the
reduction of flexural strength is comparatively higher than the other two
corresponding strengths as impact damage on both matrix and fibre affects the
flexural strength of the composites. Moreover, matrix cracks or delamination lead to
an increase in buckling instability (Kinsey et al., 1995) present during the flexural

Behaviour of glass FRP composite tubes under repeated impact for piling application

120

Chapter 5 – Residual properties of FRP composite tubes

EJ Guades

test, resulting to a much higher degradation in the flexural strength as compared to
the compressive strength.

Strength retention factor

1.10

1.00

0.90
Compressive

Tensile

Flexural

0.80
0

100

200

300

400

500

600

700

800

Impact energy (J)

Figure 5.31 Comparison of residual compressive, tensile, and flexural strengths

During flexural loading, the upper portion of the coupon is generally in
compression while the bottom part is in tension. It can therefore be inferred that
characterising both residual compressive and tensile strengths would judiciously
characterise also the flexural strength of the tested coupon. As a result, the discussion
in this section is more focused on the comparison between the residual compressive
and tensile strengths of the impacted tubes. It can be seen from Figure 5.31 that
despite the damage in the form of micro-cracks observed along the surface of the
tested samples (Figure 5.6), the tensile strength is not affected up to impact energy
levels of 477 J and is only slightly reduced by a 742 J impact. In contrary,
compressive strength are markedly affected even by lighter impacts (i.e., 159 J). This
outcome is expected to happen as the impact-induced damage on the surface of the
tested coupons is mainly cracking of matrix or presumably delamination. Matrix
cracking or delamination reduces the compressive strength but has little effect on the
tensile strength, whereas broken fibres have more effect on tensile strength (Behesty
and Harris, 1998). This indicates that during the impact event, tensile strength is less
sensitive compared to the compressive strength resulting to have a much higher
strength retention value. This result was also found in the studies conducted by
Wyrick and Adams (1998) and Behesty and Harris (1998) on composite laminates.
The curves of the modulus plotted against different levels of energy shown in
Figure 5.32 indicate that the modulus values of the impacted tubes are comparable.
The lowest compressive and flexural modulus retention factors during the impact

Behaviour of glass FRP composite tubes under repeated impact for piling application

121

Chapter 5 – Residual properties of FRP composite tubes

EJ Guades

regime are 0.98 and 0.96; respectively. On the other hand, the tensile modulus
retention factor is slightly above the baseline value. It is apparent from Figure 5.32
that during the impact regime, the loading conditions did not significantly affect the
modulus of the impacted tubes. In fact, the data points of compressive and tensile
modulus illustrates that they coincide with each other at some energy levels.

Modulus retention factor

1.10

1.00

0.90
Compressive

Tensile

Flexural

0.80
0

100

200

300

400

500

600

700

800

Impact energy (J)

Figure 5.32 Comparison of residual compressive, tensile, and flexural moduli
5.3.7 Residual strength versus modulus
The comparison between the strength and modulus of the impacted tubes plotted
against increasing energy levels is shown in Figure 5.33. The figure indicates that
impact events result in reductions in residual properties to varying degrees. The
higher the damage, the higher is the degradation of residual properties of the
impacted tube. The reductions, however, are obviously concentrated on the residual
strength rather than in residual modulus. It can be noticed that the variation of their
values is more pronounced in the compressive and flexural properties (see Figures
5.33a and 5.33c). The effect of the damage caused by this impact event on the tensile
properties of the tubes (Figure 5.33b), however, is insignificant. The reason of this
phenomenon was highlighted in Sections 5.3.3 and 5.3.7.
Figure 5.33a and 5.33c show that the residual compressive and flexural
moduli are slightly reduced with increasing impact energies compared to their
corresponding strength. The retention factors at an energy level of 742 J are 1 and
0.98 in compression and flexure, respectively, indicating that the modulus is not
particularly sensitive to the damage presence. Comparatively, strengths are more
severely affected by impact damage which leads to higher reductions. The impacted
tube retained only 93 and 90% of its compressive and flexural strengths,
respectively, when it was impacted by 742 J. This higher sensitivity can be explained

Behaviour of glass FRP composite tubes under repeated impact for piling application

122

Chapter 5 – Residual properties of FRP composite tubes

EJ Guades

by the fact that the damage in the form of micro-cracks induced by the impact events
is localised in most cases and therefore it has less effect on global properties such as
modulus. This result was also found by Zhang and Richardson (2007) when they
evaluated the effect of impact damage on the flexural properties of pultruded glassreinforced composites.

Retention factor

1.10

1.00

0.90
Strength

Modulus

0.80
0

100

200

300

400

500

600

700

800

Impact energy (J)

(a) Residual compressive properties

Retention factor

1.10

1.00

0.90

Strength

Modulus

0.80

0

100

200

300

400

500

600

700

800

700

800

Impact energy (J)

(b) Residual tensile properties

Retention factor

1.10

1.00

0.90

Strength

Modulus

0.80
0

100

200

300

400

500

600

Impact energy (J)

(c) Residual flexural properties
Figure 5.33 Strength and modulus curves plotted at increasing impact energy levels

Behaviour of glass FRP composite tubes under repeated impact for piling application

123

Chapter 5 – Residual properties of FRP composite tubes

EJ Guades

5.3.8 Variations of residual compressive strength with the height of the tube
Figure 5.34 shows the comparison of the residual compressive strengths of the
coupons sourced from different locations on the impacted tubes. It should be noted
that the measured values of the compressive strengths represent the strength of the
entire length of the coupons. In this study, however, the measured strength assumed
to represent the value at the mid-length of the coupons for ease in the comparison of
the strengths. The lay-out illustrating the mid-length location of the tested coupons
below a reference line (top edge of the non-collapsed tube) is shown previously in
Figure 5.2.
Figures 5.34a and 5.34b indicate that the residual compressive strength of
impacted tube varies with its location from the reference line. The residual strength
reduction found to increase when its location becomes nearer to the surface in direct
contact with the impactor. The maximum retention factor of a coupon taken from a
non-collapsed tube (i.e., E160-80) at the initial location L1 of 70 mm is 0.96 as
shown in Figure 5.34a. However when we tested a coupon taken from 2.8 and 4.5
times L1 below the reference line, it was found that the strength retention factor
increased to 0.98 and 1, respectively. This was also the case in the collapsed tubes
(i.e., E740-10) whereby the strength retention factor increases from 0.93 to 1.01 at
2.8 and 4.5 of L1, respectively (see Figure 5.34b). This indicates that the effect of
impact event on the reduction of residual compressive strength is concentrated only
in areas which are relatively near from the source of the impact. Similarly, the
damage that is created in the form of micro-cracks by this event decreases when the
point of location moves away from this source. This result supports the findings on
the residual modulus of the impacted tubes discussed in Section 5.3.7 that impact
damage is localised in most cases.
This outcome is very interesting and noteworthy on the use of FRP tubes as
composite piles. It should be reminded that the testing set-up adopted in conducting
impact test fairly simulate the actual conditions in pile driving. Similarly, the damage
mode observed on the composite tubes at the end of test reflects the condition of the
hollow FRP pile when encountering hard soils or boulders. Although localised
impact damage has adverse effect on the post-impact performance of the FRP
materials, the result of the present study suggests that the load-bearing capacity of
the hollow FRP pile after installation can be improved. This can be achieved by
removing portion of the FRP materials specifically near the pile head in direct
Behaviour of glass FRP composite tubes under repeated impact for piling application

124

98 1. tensile.00 (b) Collapsed tubes Figure 5.8 x L1 4.97 0.98 0.40 0.03 1.03 4.60 0.00 E160-80 E320-80 E480-10 E630-10 (a) Non-collapsed tubes L1 = 70 mm Strength retention factor 1. The damage caused by the impact loading on the composite tubes played an important role in their post-impact bearing performance.5 x L1 0.34 Variation of residual compressive strengths with the height of the tube 5.00 0.5 x L1 1.00 1. number of impacts.99 1.94 0.00 2.Chapter 5 – Residual properties of FRP composite tubes EJ Guades contact with the impact hammer.96 E480-40 E480-80 2. It was found that the levels of impact energy.60 0. The higher their magnitude Behaviour of glass FRP composite tubes under repeated impact for piling application 125 .40 0. and the drop mass is significant on the residual strength reduction of the impacted tubes. The coupons taken from the impacted tubes were then tested statically to determine the residual compressive.99 1.00 1. Initially.80 0.00 0.01 0.4 Conclusions The residual properties of a square composite tube under repeated axial impact were investigated in this chapter.20 1.96 0.80 0.20 0.20 1.02 0. L1 = 70 mm Strength retention factor 1.00 0.20 0.98 1.99 1.93 E630-30 E740-10 0.01 0. the tubes were subjected by repeated impact loading using a range of incident energies. The removal of this “sacrificial length” will apparently restore back the 100% baseline compressive strength of the FRP materials.8 x L1 0. and flexural properties.

This method deems an alternative for a costly and sometimes not straightforward experimental investigation. In Chapter 6. resulting to a much higher degradation compared to the other strengths. Comparing between the reductions of the residual strengths. the presence of matrix cracks or delamination lead to an increase in buckling instability during the flexural test. It was found that the tensile strength of the tube is less sensitive on the damage caused by the impact event. The comparison of the residual compressive strengths sourced at different locations along the height of the tube revealed that the strength reduction varied with its location. It was also found that the reductions of the residual strength values of non-collapsed tubes are lower than the value when the tubes are impacted up to failure. This is because the impact event provided damage on both matrix and fibre damage resulting to a combined effect on the flexural strengths. The maximum reductions of the residual compressive. respectively. The degradation of the compressive strength of the impacted tube decreased when its location from the top of the tube increased. For the full-scale actual piles. the influence of impact damage on the degradation of residual compressive strength of the tube is concentrated only in region closer to the impact point. On the contrary.Chapter 5 – Residual properties of FRP composite tubes EJ Guades increased.8%. Similarly. the flexural strength is severely affected by the impact loading compared to the compressive and tensile strengths. Clearly there is some degradation of residual properties after repeated axial impact for a short specimen. 0. Therefore. the residual properties far away from the impact location may not be affected by the impact damage at the top. In addition to this. the faster the strength of the impacted tubes degraded. tensile and flexural strengths are 6. the damage modelling of repeatedly impacted FRP composite tube is discussed. residual properties testing on a full length pile might be beneficial will provide additional information on the effect of axial impact loading. however. The damage behaviour of the impacted tubes can be characterised in terms of their response during the application of impact loading and/or their postBehaviour of glass FRP composite tubes under repeated impact for piling application 126 .3% and 10%. It is apparent that the impact damage provided significant effect on the performance of the FRP composite materials during the impact event. Supplementary technique such as analytical method provides a significant role in predicting the damage response of the FRP composite tube. On the other hand. the residual modulus property found to be less affected by impact event since the damage brought by them is localised in most of the cases.

delamination. it can be deduced that it has more or less similar meaning with the strength retention ratio that defines the residual properties of the impacted composite tubes. The damage prediction model presented in Chapter is directly related to the former since it illustrates the repeated impact/fatigue (Ein–N) curves of the composite tubes. and fibre ruptures. Specifically. the damage was characterised by a damage index (DI) which is a ratio between the absorbed energy Eabs and Ein. The results obtained from residual properties testings indicated that generally. For both characterisations. The damage was represented by a strength retention factor (ratio between the initial strength and the strength in the damage state) whereby this value decreases with increasing Ein and N values. Behaviour of glass FRP composite tubes under repeated impact for piling application 127 . the strengths of the composite tubes are significantly reduced. it was highlighted that the reduction resulted from damage in a form of matrix cracking. it was found that factors such as the impact energy and number of impacts contributed on the damage response of the FRP composite tube. Since the DI in the prediction model describes the amount of damage.Chapter 5 – Residual properties of FRP composite tubes EJ Guades impact (residual) behaviour. In the damage response modelling.

2 Theoretical prediction methods This section presents a number of analytical models developed to predict the repeated impact damage evolution of FRP composite laminates or transversely Behaviour of glass FRP composite tubes under repeated impact for piling application 128 . Analytical models characterising the damage behaviour of FRP composite materials under repeated impact have been reported (Bora et al.. however. 2001.2. 6. 2004a. Quasi-static compressive test was conducted on the composite tubes to aid in the formulation of the lifetime prediction model. The proposed damage model quantifies the energy absorption response of the impacted tubes when subjected to various impact energy levels.. 2008. Belingardi and Vadori. 2007.1 Introduction The widely used method of determining the response of fibre composite materials subjected by repeated impact loading is by experimental testing (Aurrekoetxea et al. Sugun and Rao. Roy et al... 2009. 2007.. Belingardi et al.. are limited on laminates or tubes which are transversely impacted. Azouaoui et al. The values of the parameters considered in modelling the response of the tube were either obtained experimentally or referenced from the literature.. Azouaoui et al. the high cost of the experimental works and other limitation such as the unavailability of testing machine make the design by means of an analytical method attractive.Chapter 6 –Damage modelling of repeatedly impacted FRP composite tube EJ Guades Chapter 6 Damage modelling of repeatedly impacted FRP composite tube 6.. 1998. Azouaoui et al. Datta et al. 2004. 2008.. In this chapter. 2011. This chapter also presents the application of the model to glass/vinyl ester tubes with different cross sections. and Lhymn. 1997. Belingardi et al. However. 2001. 1992. These models are presented in Section 6. Their applications. Ho et al. 2001.. The proposed model was then verified by comparing it to the results of the experimental work discussed in Chapter 4. 2003. a proposed lifetime prediction model that will characterise the damage evolution of a repeatedly impacted square FRP composite tube is presented. Sevkat et al. Roy et al. 1985). 1995). 2010.. and Found and Howard.. Wyrick and Adams. Jang et al.

the. The relationships of these parameters are shown in Equations 6. These constants control the slope of the second. (2008) by introducing a damage parameter. Several researchers correlated the damage response of the material to a quantifiable parameter to characterise the damage on the FRP composite materials due to repeated impact. Their modelling is based on a non-linear parametric creep relation proposed by Mankowsky with a modification on the denominator function (Equation 6. D = Eabs / Eim up to penetration (6. respectively. The damage progression of a glass-reinforced laminate was studied by Belingardi et al.Chapter 6 –Damage modelling of repeatedly impacted FRP composite tube EJ Guades impacted tubes. absorbed energy Eabs and the saturation energy Esat. respectively). a. 2003). The model. number of impacts curve. and c are experimental constants depending on material properties and incident energy. (2001) presented a lifetime prediction model to determine the damage evolution of a glass/epoxy laminate subject to repeated impact. They reported that a quadratic relationship was found between the rate of damage accumulation and Eim in the initial linear part of the D vs. Azouaoui et al. D = a [θb / ((a+1) – θc)] (6.2) D = Eabs / Esat after penetration (6.1) where D is the damage parameter.4) Behaviour of glass FRP composite tubes under repeated impact for piling application 129 . θ is the life duration (= N/Nf.2 and 6. Some of the notations of the equations used in the referred papers were changed to correlate with the symbols used in other studies. D correlates the values of impact energy Eim. The equations describing the repeated impact curve of the FRP composite materials obtained from the literature are also presented. which is based on an energy principle and twoparameter Weibull distribution function. and the third zone levels of the “S” shape damage curve. b. Eim = a [(–ln R(Nf)1/αf)] [Nf]-1/b (6.4.3) A theoretical lifetime analysis was developed by Lhymn (1985) to predict the damage response of a Polyphenylene sulphide (PPS)/glass composite laminate under repeated impact.3. The derivation of this damage parameter is based on the energy balance concept obtained from the first principle of thermodynamics (Belingardi and Vadori. first. N and Nf being the number of impact and failure impact number. Their result suggested a good agreement between the experimental data points and the proposed model. displays a correlation of the Eim and Nf in Equation 6.1).

7) (6. Lhymn emphasised the scattery of the experimental data and reported that there exist a low limit of Eim below which no impact failure occurs on the laminate. and Ho et al. Ein is the incident energy. (1997). For most epoxy-based composites studied. (1992) evaluated the damage tolerance of a continuous fibrereinforced epoxy laminate under repeated impact. N0 curve.8) where a and b are material constants that define the slope and intercept. N0 curve (N0 being the “location” parameter). Bora et al. Nc is the number of impact in which the first observation of a significant delamination crack occurred. are the slope and intercept of the log(Eim) vs. R(Nf) and αf can be determined by data pooling technique whilst a and b by plotting log(Eim) vs. Common results obtained from these studies showed that Nf is inversely proportional to the incident energy Ein and the fatigue curve follows a simple power function in the form of either Equation 6.8 can also characterise the damage response of a transverselyimpacted composite tubes. a and b.5) (Pm)N / (Pm)0 = (N–Nc)-b if Ein ≥ Ec (6.5 and 6. (2005). logN curve. (2007). (2001) reported that the curve defined by Equation 6. Similarly. respectively.Chapter 6 –Damage modelling of repeatedly impacted FRP composite tube EJ Guades where R(Nf) denotes the reliability of the Nf.8. The damage tolerance and response of composite laminates under repeated impact loading were investigated by Azouaoui et al. respectively.6) where (Pm)N and (Pm)0 are the maximum loads at the Nth and 1st impacts. Jang et al. vs. αf is the “shape” parameter. (2009). respectively.7 or 6. These researchers chose the number of impacts to failure Nf as an index to define the damage tolerance limit. Sugun and Rao (2004a). of the logEin. Roy et al. Datta et al.6 can describe the damage response of the composite laminate: (Pm)N / (Pm)0 = N-b if Ein < Ec (6. Ec is critical energy in which a significant delamination crack occurs in response to a single impact. and b is the slope of the log[(Pm)N / (Pm)0] vs. Nf = a Einb Ein = a Nf -b (6. logNf curve. some of the principles used in developing a prediction model for composite laminates or tubes which are transversely impacted Behaviour of glass FRP composite tubes under repeated impact for piling application 130 . The above-mentioned analytical models combined with experimental verification were proven to satisfactory predict the repeated impact behaviour of composite materials. In the present study. the residual strength models shown in Equations 6.

68 100. Chamfering one end of the tube was done in a rotating sander (Figure 6. 6.18 2 100. the material and method used in the quasi-static compressive test are discussed. t (mm) 5. chamfering of one end (top) is adopted as a failure triggering mechanism. Width. Equation 6. The complete discussion of the role of this parameter in the damage model is presented in Section 6..54 375. Behaviour of glass FRP composite tubes under repeated impact for piling application 131 . Table 6. This failure initiator can reduce the peak load experienced by the specimen without affecting the sustained crushing load (Mamalis et al.00 5.20 3 100.3 of Chapter 4). d (mm) b (mm) 100. particularly implementing the concepts considered in formulating Equation 6. a triggering mechanism is used to promote progressive deformation.21 It was observed that the damage on the impacted tube is rupture on its head or end crushing (detailed discussion was presented in Section 4. Table 6.50 Thickness.1 shows the details of the specimens used in the test.3.Chapter 6 –Damage modelling of repeatedly impacted FRP composite tube EJ Guades were adopted for FRP tubes under repeated axial impact. The proposed damage model used an energy-based approach.1a). In this study. l (mm) 374.2.1b).58 Length. 1997b). To ensure that neither brittle failure nor buckling instability failure will take place during quasi-static compressive test.1 Specimen and testing A total of 3 replicates (composite tube with properties similar to the tubes used in Chapter 4) were tested.65 100.69 100. The specimens used in conducting quasi-static compressive and impact tests are identical except that the top end of the tube adopted in the former was chamfered by 450 (Figure 6.60 375.58 100.6.24 Average 100. 6. which is needed in determining its specific absorbed energy.2. In this section.1 Details of specimen used in quasi-static test b d t Specimen no 1 Depth.43 375.00 5.50 5.3 Quasi-static compressive test The results acquired from quasi-static compressive test were used to determine a parameter needed in developing the proposed model.8 was also adopted in tracing the repeated impact (fatigue) curve of the composite tubes.

1c).1 Quasi-static compressive test 6. Figure 6. the energy absorption behaviour and the impact damage tolerance limit of the impacted tubes are emphasised.e.4 Repeated impact test results The results obtained from the experimental work discussed in Chapter 4 (i.2 suggest that the rate of energy absorption Behaviour of glass FRP composite tubes under repeated impact for piling application 132 . impact behaviour of composite tube) were used to verify the proposed model.e. This figure is again presented in this chapter to aid in the discussion and a brief discussion was included. In particular.Chapter 6 –Damage modelling of repeatedly impacted FRP composite tube EJ Guades The 375 mm long composite tube was crushed by a 2000 kN capacity servohydraulic compressive testing machine at a constant speed of 50 mm/min (Figure 6.9 (Chapter 4) except that trend lines are indicated. (a) Specimen with 450 chamfer (c) Progressive crushing (b) Chamfering process (d) Crushed specimens Figure 6.8). The curves shown in Figure 6..2 shows the energy absorption (normalised energy) evolution of impacted tubes (i. specimens used in comparing with the prediction model shown in Table 4.3 (Figure 6.6.. Note that this figure is almost identical to that of Figure 4. The load-displacement curves of the three specimens tested were recorded using an automated data acquisition system attached on the machine and presented in Section 6. The figure was intentionally changed to suit in coming up with the proposed prediction model.2 of Chapter 4).

70 0.2a demonstrates that the trend of the energy absorption response of the collapsed tubes is similar regardless of the magnitude of the impact energy.51 J 476.00 0. N (a) Collapsed tubes Normalised energy.Chapter 6 –Damage modelling of repeatedly impacted FRP composite tube EJ Guades was higher for tubes impacted by higher incident energies indicating that heavier impacts induced more damage than lighter one. Figure 6.77 J 423. The initial line shows that the absorbed energy increases at a constant rate indicating that rebounding of the impactor is still imminent. the effect of energy loss is not considered in the analysis. the absorbed energy is nearly similar all throughout as illustrated by a zero-slope trend line. Tubes under higher impact energies relatively absorbed energy very quickly due to their fast damage accumulation. Eabs/Eim 1.90 0.50 0 20 40 60 80 100 Number of impacts.2 Normalised energy and number of impacts relationship Behaviour of glass FRP composite tubes under repeated impact for piling application 133 .00 0.70 0.50 0 20 40 60 80 100 120 140 120 140 Number of impacts.80 0. After rupturing. Figure 6. It is worth noting that in the postcollapse region. however. the recorded values of the absorbed energy are slightly less than the incident energy.01 J 0.50 J 158. Eabs/Eim 1.2b illustrates that the absorption energy response of the non-collapsed tubes characterised a single-line trend.90 0.80 0. On the other hand. Normalised energy. Their response can be approximated by a bilinear curve as highlighted by a line. N (b) Non-collapsed tubes Figure 6. Apparently.60 317. the rate of the energy absorption in the first line is largely dependent on point where collapse or failure initiated (Nf).60 634.84 J 211.92 J 0. In the proposed model.

80 0. N curve presented in Figure 6.20 1. collapsed) plotted in increasing number of impacts N. Equation 6. The modification made the intercept of the initial line becomes zero at the 1st impact (see Figure 6.4 and served as a schematic reference in formulating the equation of the proposed model. The equation of the first (initial) line describes the value of D when 1≤ N <Nf.5 Evaluation of damage using parameter D The damage mechanism of the impacted composite tubes is evaluated in terms of their absorbed energy during the test regime (Delfosse and Poursartip.60 0. Damage parameter. The idealised curve is illustrated in Figure 6.3 D vs. It should be noted that the N value of the curve in Figure 6.20 0.e.4 is normalised by the maximum number the tube is subjected. The curve indicates that the damage response of the tube can be defined mathematically by the equations of the two lines. The damage can be characterised by introducing a damage parameter D presented previously in Equation 6. A solid line is drawn to emphasise the flow of the trend. Figure 6.00 0. Eabs/Eim) as expressed in Equation 6. D = [(Eabs)N – (Eabs)1 ] / [(Eim – (Eabs)1] (6. Likewise. 1997). Similarly.2) resulting to a more simplified prediction model.00 0 20 40 60 80 100 120 140 Number of impacts.3. D 1.Chapter 6 –Damage modelling of repeatedly impacted FRP composite tube EJ Guades 6. the equation of the second line characterises Behaviour of glass FRP composite tubes under repeated impact for piling application 134 .9 can be characterised as the evolution of damage of the composite tube relative to the initial damage.3 shows the damage parameter D of the representative tube (Ein = 476.3 allows in approximating the evolution of damage of the tubes using a bilinear curve model.40 Experimental data (476. N curve of the representative composite tube 6. N Figure 6.6 Proposed damage response model The D vs.9.2 but modifying the right-hand side (i.8 J. Nmax.77 J) Trend line 0.9) th where (Eabs)N is the absorbed energy at N impact and (Eabs)1 is the absorbed energy at the 1st impact. an idealised life time curve of the tube under repeated impact can be drawn from Figure 6.

6 1 (Nf -1)/Nmax 0. and Nmax. the value of D at increasing N remains constant. The equations of the proposed damage model of tube subjected by repeated axial impact are shown in Equations 6. Nf. It should be noted that the curve defined by Equation 6. Specifically.10 is considered imaginary when Nf approaches to unity.10 and 6.2 0.11.8 1..2 (Nf -1)/Nmax 1-((Nf -1) /Nmax ) 1. Using the straight line equation in slope-intercept form. The value of D at N = 1 is zero when Nf >1.e. however.4 0.N/Nmax ) 0.0 P1 (D. From the line. N/Nmax Figure 6.Chapter 6 –Damage modelling of repeatedly impacted FRP composite tube EJ Guades the value of D when N > Nf.0 0. the damage response of the impacted tube under this condition falls in a curve defined by Equation 6. all of the impact energy are absorbed by the tube (D = 1) thereby neglecting the energy loss.11) On the other hand. 1. the following are the assumptions adopted in the proposed damage evolution model. On the other hand. Damage parameter.0 Life fraction. The proposed model correlates the D to corresponding N.10 tends to become zero when Nf approaches infinity. 2. the value of D in Equation 6.6 0.8 0.2 (0. Equation 6. D = (N/Nmax)/((Nf – 1)/Nmax) if 1≤ N < Nf (6.11. the equation of the first line can be obtained by considering a point along this line (denoted by P1 in the curve).4 0.0) 0. At the initiation of collapse (N = Nf).e.0 0.10) D=1 if N ≥ Nf (6. the value of D in this range is equivalent to unity. The determination of the parameters used in the proposed predictive model to characterise the damage of axially impacted FRP tubes are described in the following sub-sections. It can be observed that when N > Nf (i. D 1. Likewise. we can get the equation (i.10) describing the damage response of the tube when 1≤ N < Nf. N/Nmax). the coordinate of P1 can be taken as (D.4 Idealised lifetime response curve of the repeatedly impacted tube Behaviour of glass FRP composite tubes under repeated impact for piling application 135 . It is worth noting that the D in the coordinate corresponds to a value at a given N. second line).

2 Minimum incident energy to fail the tube for one impact (critical energy). 2010.8 only includes few parameters. and Ein = Ec. 2001.12. Ec The minimum energy required to fail the composite tubes for one impact can be found through experiment (Palanivelu et al. Mamalis et al. Ein (J) 140 0 120 0 100 0 Ein = a Nf-b 800 600 400 Nf =1 200 0 0 20 40 60 80 100 120 Number of impacts to failure. Yang et al.1 Minimum number of impacts to failure of the tube.. it can be observed that when Nf = 1.5..12) 6. the value of a becomes Ec. and Kim and Arora. will make the proposed predictive model simple.6. Ec) Incident energy. Equation 6. the latter is not straightforward as it requires complex analysis and fine tuning of the model to satisfactory simulate the actual behaviour during Behaviour of glass FRP composite tubes under repeated impact for piling application 136 . Greve et al.... however.. Nf Figure 6. needs expensive testing machine or special testing set-up to follow the crushing process.8 and Figure 6.8 yields Equation 6..5 Typical curve described by Ein = aNf-b From Equation 6. The former. Equation 6. 2010. 2003). Mamalis et al. This basis of computation is supported by the result on impact test of composite materials discussed in Section 2 whereby their repeated impact curve can be traced by a simple power relationship. Xiao.8 was adopted to determine the relationship between the incident energy Ein and number of impacts to collapse Nf. Han et al.. 2009. On the other hand.Chapter 6 –Damage modelling of repeatedly impacted FRP composite tube EJ Guades 6. 2007. Ein = Ec Nf-b (6. This equation was used as a basis of computation since the Ein and Nf responses of the FRP composite materials usually follow a power law. 2006.. Song et al. 2009. 2005. 2002. and if found suitable.5 demonstrates the curve defined by this equation. Nf In the present study. and Mamalis et al. Figure 6. Substituting a to Equation 6. 160 0 (1. 1997a) or using finite element (FE) analysis (Palanivelu et al.6.

(1997a) with various strain rate of loading (Figure 6. Equation 6.. 2009. Thus.13).) shown in Figure 6. (199 6) Thornton (199 0) Figure 6.7). and Thornton. In the case of glass/vinyl ester composite tubes.14 defined a fitting line for an experimental results conducted by Mamalis et al.16 1.Chapter 6 –Damage modelling of repeatedly impacted FRP composite tube EJ Guades impact or dynamic loading.00 Yang et al. 2. Figure 6.. (1997a) Thornton (1990) 1. (2 009) Mamalis et al. however.50 0. Behaviour of glass FRP composite tubes under repeated impact for piling application 137 . To account for this sensitivity. 2002). the present study devises an alternative method in determining the value of Ec.6 shows the summary of β values in bar chart of composite tubes made of glass/vinyl ester obtained from the literature. This was achieved by correlating the value of Ec under dynamic or impact load to its value when loaded statically (Equation 6.16. their dynamic progressive collapse can be idealised as a quasi-static response (Jones.14 in calculating β. 1.00 Yang et al. 1995). adopting the average value of β (i..6 Variation of the correlation β of glass/vinyl ester composite tubes For tubes made from strain-rate-insensitive materials. this method may not be applicable due to their strain-rate sensitivity characteristic (Jacob et al. The quasi-static method of analysis usually neglects the variation of the axial force about the mean value (i. It should be noted that only of this type of composite material was considered since glass/vinyl ester is the material used in the present study.. the present study adopted Equation 6. (2009) Mamalis et al. (1996) Mamalis et al. Mamalis et al.e. 1996. (199 7a) Mamalis et al. As a result.6 may not represent a suitable value for the analysis. Mamalis et al.13) Literature revealed that the value of the correlation factor β for composite tubes made of E-glass fibres and vinyl ester resin is in the range of 0. neglecting the effect of rate of straining) which is caused by the resistance cycle change (Jones.35 (Yang et al.e.90 to 1.00 0.. (Ec)Dynamic = β (Ec)Quasi-static (6. 1990). 1997a. 1989).50 β Average = 1.

8709 α +1.50 0.0210 0. (Ec)Dynamic = 1.14) α = [(Rate of loading)Quasi-static] / [(Rate of loading)Dynamic] (6. (1997a) 1.3 in Chapter 4).225.6.Chapter 6 –Damage modelling of repeatedly impacted FRP composite tube EJ Guades 2.00 Mamalis et al. associated by a shallow serration. On the other hand.0220 α (x10-3) 0. 2004. 2004.15..15) where. we can get α = 0. Al Galib and Limam.3 Determination of (Ec)Quasi-static using quasi-static compressive test Figure 6.16) 6. Bardi et al. This feature can also be observed on tubes made from metallic materials (Hsu and Jones.16.13 results to Equation 6. equivalent to drop height of 3 m).7.225 (Ec)Quasi-static (6. is on their crushing response in terms of Behaviour of glass FRP composite tubes under repeated impact for piling application 138 .32 1. 2003). the (Rate of loading)Dynamic used the value of 7.9 in Chapter 4. It can be noticed from the figure that the main feature of the curve in the post crushing region is the characteristic oscillation from the mean post-crushing load.2 and 4. and Huang and Lu.00 0.6. In this study. β = -0.32 (6.8709α + 1.7 Data points with the fitting line showing β and α relationship.3. From these values.0240 Figure 6. however.14 gives β equals to 1.8 shows the typical load displacement curves of the composite tubes from quasi-static compressive test presented in Section 6. the (Rate of loading)Quasi-static is equal to 50 mm/min as indicated previously in Section 6. Substituting α to Equation 6.0200 0.67 m/s (using Equation 4.108 x10-3 using Equation 6. Experimental data from Mamalis et al. (1997a) From Figure 6.1.3.00 0. The value used in the rate of loading for dynamic (impact) testing was adopted considering that it is the maximum value (see Tables 4. Note that this value is 5% higher than the average value shown in Figure 6.50 β β = -0.0230 0. 2003. The difference. Substituting the value of β to Equation 6.

The average value of (Ec)Quasi-static of the specimens tested is 1. Substituting this value to Equation 6.2 summarises the (Ec)Quasi-static values of the 3 specimens obtained from the quasi-static compressive test. On the other hand.12. the failure of metallic tube is associated with the development of wrinkle or buckle and grows until sufficient external energy has been imparted to complete the formation (Jones.9) was calculated by numerical integration using trapezoidal rule method (see Equation 4. The value of (Ec)Quasi-static (shaded area in Figure 6.9 Schematic diagram used in computing (Ec)Quasi-static Behaviour of glass FRP composite tubes under repeated impact for piling application 139 . The shaded area represents the minimum incident energy required to crash/fail the tube in response to a single impact.44Nf-b (6.44 J. Ein = 1485.17) 250 200 Specimen 1 Load (kN) Specimen 2 150 Specimen 3 100 50 0 0 10 20 30 40 Displacement (mm) Figure 6.1c). 1995).17 can be obtained by replacing (Ec)Dynamic value in Equation 6.8 Typical load-displacement curves from quasi-static compressive test 250 Load (kN) 200 150 F = f(s) 100 E c Quasi  static 50 S1 0 0 S2 10  S2  Fds S1 20 30 40 Displacement (mm) Figure 6.16.8 in Chapter 4).Chapter 6 –Damage modelling of repeatedly impacted FRP composite tube EJ Guades physical damage in which composite tubes exhibited mushrooming failure (see Figure 6. Moreover.60 J. Equation 6. we will get (Ec)Dynamic = 1485. Table 6.212.

6 2 1.10 shows different b values for 20 runs (i. different initial combination values of Ein and Nf). (2001) in finding the material constants in modelling the damage of composite laminates subjected by repeated impact loading.17 was computed using an Excel 2010 “Solver” function. This method was also adopted by Azouaoui et al. This function uses an optimisation technique in calculating the value of b by changing variable cells subject to the applied constraints.6 b baverage = 0.2 Summary of (Ec)Quasi-static values Specimen no 1 (Ec)Quasi-static.0 0.8 0.233.222.10 b values using Excel 2010 “Solver” function Behaviour of glass FRP composite tubes under repeated impact for piling application 140 .9 3 1.230 and 0.181.44Nf-0. Nf should be an integer.291.6 Standard deviation 22. Note that a total of 20 runs were conducted to check the reliability and accuracy of the result.291 (6.6.18).44Nf b – Ein) to zero (as the objective) with the following constraints: 0 < Ein ≤ 1485.4 Solving b value The constant value b from Equation 6.Chapter 6 –Damage modelling of repeatedly impacted FRP composite tube EJ Guades Table 6. Substituting this b value to Equation 6. Ein = 1485.e. 0 ≤ b ≤ 1.44 J.4 0. J 1. Setting the term (1485.17. Nf ≥ 1.291 0. we can get the equation that will characterise the repeated impact (fatigue) curve of the composite tubes (Equation 6.0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Number of runs Figure 6.330. The adopted value of the present study is the average b value equals to 0.212.. Figure 6.2 0.3 Average 1. The b values obtained from this method ranges between 0.18) 1.4 6.

Conclusively.2 of Chapter 4). It should be noted that the experimental points presented in this figure are the results of the test defined by Table 4. This result was already presented in Table 4.69 to 11. Nf may be different. It can be observed from Table 6. Nevertheless.18 is in between the experimental values. This result is expected as different impact masses (or drop heights) for every Ein were used in the impact tests (Section 4.291 EEqn.3 in Chapter 4. It was emphasised that the data points on incident energy-number of impacts curves of composite laminate or tubes under transverse impact follow a power relationship.7. The effect of impact mass is significant in the value of Nf more pronouncedly at lower Ein as discussed in Section 4.3). Incident energy. This result was also found in the present study whereby the points shown in Figure 6. Equation 6. is again shown in Table 6. the value of Nf computed from Equation 6.3. however.2.18 and the experimental data.1 Verification of the repeated impact curve Figure 6.4 shows that when an average experimental Nf is used. Ein (J) 1600 1400 Experimental data 1200 -0. their variation is reduced to less than 3%.18 can be used in developing the damage prediction model of composite tubes under repeated impact..7 Comparison with the experimental data 6. Nf Figure 6. The difference between the Ein from Equation 6. 634.4 (Chapter 4).11 illustrates the Ein-Nf relationship that compares the curve traced by Equation 6.3 for ease of comparison. This verifies the assumption that the equation of repeated impact curve of composite laminates is valid for composite tubes subjected by repeated axial impact.Chapter 6 –Damage modelling of repeatedly impacted FRP composite tube EJ Guades 6.11 indicates that the curve reasonably fit the experimental points.18 and from impact test ranges from -8.3 that even if Ein is same (e.11 exhibit an exponential curve.5 J). 18 in = 1485.54% (Table 6.44N f 1000 800 600 400 200 0 0 20 40 60 80 100 120 Number of impacts to failure. Figure 6. Table 6.5 (Chapter 4).g.11 Comparison between the experimental data and repeated impact curve Behaviour of glass FRP composite tubes under repeated impact for piling application 141 .

7. collapsed/failed tubes).2 72 423.2 from experimental result -8. In Figures 6.12 and 6.Chapter 6 –Damage modelling of repeatedly impacted FRP composite tube EJ Guades Table 6. a value of 200 is uniformly adopted in all cases to extrapolate the lines of the model for clear comparison.4 Comparison of incident energies at average Nf Ein (J) Experiment Equation 6.8 J.2 11.67.8 481.85% 48 423.3 Comparison of incident energies at different Nf 13 Ein (J) Experiment Equation 6.13a to 6.2 Validation of the proposed model Figures 6.7 -4.85.69% Nf a % Difference 9.13 compare the results between the prediction model and experiment using Nmax of 200. Figures 6.5 J) or 130 (Ein = 423 J or lower).8 -1. respectively. On the other hand.8 538. and 95.9 J. 211.12a to 6.5 617. 57. The experimental data shown in the figures are the results of the impact tests conducted based from test matrix presented in Table 4.90% -0. This indicates a good agreement between the experimental data points and the prediction model.9 20 634.0 422.18 17 634.46% 51 476.12c shows an R2 value of at least 0.9 a from experimental result Nf a % Difference 2.18 634. nevertheless.54% 95 423. Note that supposedly Nmax values should be either 45 (Ein = 634.10% 0.5 700. and 2065 when it is impacted by a relatively lower incident energies of 317.12 and 6. and 8. 779.e.2 (Chapter 4).5 11.49% 57 476.69% 32 476.13.97 between the proposed model and the experimental data for Ein = 423 J or higher (i. respectively (see Table 6.69 at Nf of 20. It can Behaviour of glass FRP composite tubes under repeated impact for piling application 142 .5 648.9 -2.01% 6.0 47 476.3).0 478.5 J. and 158.8 454. 4. the R2 value of Ein = 318 J or lower cannot be obtained due to the absence of (Nf)Exp.13c)..48% a Table 6.0 389.18) predicted that failure of the composite tubes will occur at an Nf of 194.8 469. The percentage difference of Ein between the model and experiment is 2. The model (or Equation 6. however. (Nf)Predicted are indicated (see Figures 6.

971 0. N/Nmax (a) Ein = 634.00 R2 = 0. N/Nmax (c) Ein = 423 J Figure 6.40 0.20 Prediction model 0.00 0.80 0.80 1.00 0.5 J Damage parameter. This is possibly due to the dynamic nature of the test whereby the D values during the first few impacts are serrated. Damage parameter.983 0.20 Life fraction.20 Damage parameter.40 Experimental data 0. however.13a to 6.80 1.13c that the D values are less than 1 at initial N/Nmax.40 0.20 1.60 0. experimental data for collapsed tubes Behaviour of glass FRP composite tubes under repeated impact for piling application 143 .60 0.20 0.20 0.40 Experimental data 0.60 0. D 1.60 0.60 0.00 R2 = 0.20 Prediction model 0.12 Proposed model vs. N/Nmax (b) Ein = 476.Chapter 6 –Damage modelling of repeatedly impacted FRP composite tube EJ Guades be observed from Figures 6.00 -0.00 0.00 0.20 1.80 1.8 J 1. D 1.978 0.00 0.00 Life fraction.00 R2 = 0.20 Life fraction.20 0. becomes apparent at relatively higher N/Nmax values.40 0.00 0. The trend.20 -0.20 Prediction model 0. D 1.80 0.40 Experimental data 0.80 0.60 0.00 -0.

N/Nmax (b) Ein = 211. N/Nmax (c) Ein = 158.00 -0.20 Life fraction.40 0.60 (Nf)predicted = 779 0.20 Life fraction.00 0. experimental data for non-collapsed tubes 6.00 0. N/Nmax (a) Ein = 317.80 Prediction model (Nf)predicted = 2065 0.20 0.60 0.20 Life fraction.13 Proposed model vs.20 0.80 Prediction model (Nf)predicted = 194 0.80 1.00 0.60 0.9 J Figure 6.40 0.00 Experimental data 0.80 1.00 0.40 0.20 Damage parameter.00 Experimental data 0.00 0.20 1.40 0.8 Summary of procedure in establishing the damage evolution curve Figure 6.80 Prediction model 0.5 J 1. This chart is explained and summarised by the following: Behaviour of glass FRP composite tubes under repeated impact for piling application 144 .00 0.Chapter 6 –Damage modelling of repeatedly impacted FRP composite tube EJ Guades Damage parameter. D 1.60 -0.20 0.00 0.00 0. D 1.60 0.8 J 1.20 0.14 shows the flow chart that describes the procedures in establishing the damage evolution curve of the impacted tube.60 -0.00 Experimental data 0.20 Damage parameter.20 0.20 0.40 0.80 1.40 0. D 1.

5. 3. The curve can be defined by either Equation 6. Ec can be obtained from either dynamic (impact) or quasi-static compressive tests. the values of Ein are calculated by assigning values to Nf. This curve provides the relationship between Ein and Nf.10 or Equation 6. The repeated impact (fatigue) curve can then be drawn by substituting the values of Ec and b to Equation 6. Nf and Nmax. and the type of the matrix material of the composite tube.14 Flow chart in establishing the damage evolution curve Behaviour of glass FRP composite tubes under repeated impact for piling application 145 . The damage evolution curve (D versus N/Nmax) can now be established by substituting the value of Nf with its corresponding Ein. its value is chosen to be greater than Nf. Preassigning values to Ec in finding Nf might result to a non-integer Nf values.Chapter 6 –Damage modelling of repeatedly impacted FRP composite tube EJ Guades 1. Calculate Ec and b Ec (using dynamic or quasi static tests) b (using Excel “Solver” function) Repeated impact curve (Ein-Nf relationship) Identify Nmax Damage evolution curve Figure 6.11. use Equation 6. Consequently. Calculate the values of Ec and b. Note that it is more suitable to pre-assign values to Nf instead the other way around since Nf is required to be an integer. If using quasi-static compressive test. The value of β depends on the rate of loading used in the quasi-static compressive test.12. or their combination. depending on the value of N.13 and find the correlation factor β. 2. the impact velocity during the impact test. 4. Several trials can be run and finally choose the average value of b. 6. The value of Nmax can be selected based from the value of Nf. Generally. The value of b can be obtained using Excel “Solver” function.

For the two available tubes. Yang et al. Its application to other types of composite tubes. 2008. One of the parameters identified in the damage model is the absorbed energy of the composite tube during impact loading. It is worth noting that the Behaviour of glass FRP composite tubes under repeated impact for piling application 146 . circular.. These tubes are designated as S125 and R75x100 (Figure 6. rectangular) are approximately similar (Palanivelu et al. It should be noted that the energy at this point corresponds to the energy absorbed as a result of the progressive crushing of the tube.8 provided a guide in establishing the curve for each corresponding tubes. square.e. It is therefore reasonable that the model can be used to characterise the damage evolution curve to tubes with different cross sections subjected by repeated axial impact loading.1 Square and rectangular FRP composite tubes In this section. on the other hand. The procedure presented in Section 6..9 Application of model to FRP composite tubes with square and rectangular cross sections In the preceding sections... The model predicted reasonably the damage and failure response of a 100 mm square composite tube. the prediction model characterising the damage evolution of a repeatedly impacted FRP composite tubes was discussed. Two tubes (square and rectangular shapes) of different sizes made of glass fibre and vinyl ester resin are made available.. Melo et al. 2001).12) and the damage evolution curves (Equations 6.10 or/and 6.9. 6. 1997b).11) of the glass fibre/vinyl ester composite tubes using the model are discussed. needs to be investigated.. and Schultz and Hyer. impact or quasi static) for different geometries (i.e. Literature revealed that the shape of the load-deformation curves of composite tubes under similar test (i. the repeated impact and the damage evolution curves of the previously studied tube (100x100mm) was included in the discussions in comparison with the other tubes. the repeated impact curve (Equation 6. 2010.15). The composite tubes are manufactured using the process of pultrusion. Similarly. It should be noted that in this section. a suitable parameters need to be established for the other types of composite tubes in correctly predicting their damage response using the model.Chapter 6 –Damage modelling of repeatedly impacted FRP composite tube EJ Guades 6. 2009. the critical absorbed energy used in the model was obtained experimentally. The absorbed energy of the composite tube can be obtained from its load-deformation curve derived from either impact or quasi static compressive tests (Mamalis et al.

5 Properties of S125 and R75x100 specimens Properties Depth.) will no longer be presented in this section.32 R75x100 100.49 Width. mm) 125. their repeated impact curves (Equation 6.990 1. respectively. Table 6.6.15 Square and rectangular composite tubes Table 6.51 Thickness. etc. l (mm) 374. ρ (kg/m ) 1.36 6. Likewise. whilst the number indicates the nominal dimensions of the tubes. b (outer.44 74.33 375.80 79.Chapter 6 –Damage modelling of repeatedly impacted FRP composite tube EJ Guades letter S and R in the designation represents square and rectangular sections. The dimensions presented in the table are the average of the three specimens.3 except that the tube Behaviour of glass FRP composite tubes under repeated impact for piling application 147 .26 Specific mass.91 Length. (a) S125 specimen (b) R75x100 specimen Figure 6.965 Glass content (%) 79. where the 00 direction coincides with the longitudinal axis of the tube. the 100x100mm square tube is designated as S100 although some of its details (properties. mm) S125 125. a To obtain the damage evolution curve of the composite tubes using the model.12) are determined first. The value of Ec for S125 and R75x100 tubes was determined from the result of the quasi-static compressive tests. The methods and testing machine used in this test is similar to that used for 100x100 pultruded section presented in Section 6. d (outer.26 5. t (mm) 3 Fibre lay-up a a 0 0 0 0 0 0 0 0 0 [0 /+45 /0 /-45 /0 /-45 /0 /+45 /0 ].5 shows the properties of the tubes.

16 Crushed composite tubes Figures 6. The height of the tested tubes was 375 mm.18 illustrate the load-displacement curves of S125 and R75x100 specimens. It should be noted that the curve displayed in the figures are for the three replicates. The area in the curve used in obtaining (Ec)Quasi-static is the area corresponding to the displacement where the crushing load started to become stable. Figure 6. 400 Specimen 1 Specimen 2 Load (kN) 300 Specimen 3 200 100 0 0 10 20 Displacement (mm) 30 40 Figure 6.9. From the curve. 1997a). respectively.Chapter 6 –Damage modelling of repeatedly impacted FRP composite tube EJ Guades was crushed at 10 mm/min. An example of the area used in calculating the (Ec)Quasi-static was presented in Figure 6.17 Load-displacement curves of S125 specimen Behaviour of glass FRP composite tubes under repeated impact for piling application 148 . (Ec)Quasi-static was calculated by numerical integration using trapezoidal rule method.15).. It should be noted that the axial height of the tubes subjected by axial loading and collapsing in a progressive manner does not affect their energy absorbing capacity (Mamalis et al.16 shows the crushed tubes at the end of the quasi-static compressive test.17 and 6. (a) S125 specimen (b) R75x100 specimen Figure 6. A total of three replicates were used for each test. Both tubes are tested using a 450 chamfer on their top end (Figure 6.

052. After getting the two parameters (i.6).291 Ein = 1485. The average b value for S125 and R75x100 specimens is 0..9 Nf-0.5 Nf-0.6 summarises the values of (Ec)Dynamic.301.13 with β value of 1. The value of Ein in the curve was obtained by pre-assigning value of Nf in the repeated impact Behaviour of glass FRP composite tubes under repeated impact for piling application 149 .368.6 for S125 and R75x100 specimens was computed using an Excel 2010 “Solver” function similar to that of S100 (see Section 6.205 and 0.316 Figure 6.8 J.6 Summary of parametric values of square and rectangular tubes Tube S125 (Ec)Dynamic (J) 3.316. Table 6. the average value of (Ec)Quasi-static of specimen R75x100 is 1. It can be observed that the value of S125 is comparably higher than R75x100 due to the higher geometric size of the former. Table 6.368.472. The table also includes the values for the S100 specimen. b.5 b 0.4Nf-0. respectively. On the other hand.205 Repeated impact equation Ein = 3.217.205 S100 1.316 Ein = 1.4 0.Chapter 6 –Damage modelling of repeatedly impacted FRP composite tube EJ Guades 400 Specimen 1 Specimen 2 Load (kN) 300 Specimen 3 200 100 0 0 10 20 Displacement (mm) 30 40 Figure 6. and the equation of the repeated impact curves of the two composite tubes. S100 specimen) except that the value of Ec was changed to 3. It should be noted that the applied constraints for the two tubes are similar to that of a 100x100 section (i.18 Load-displacement curves of R75x100 specimen The average value of (Ec)Quasi-static of S125 specimen is 2.217.5 J and 1..e. respectively.485. The value of b in Table 6.e.1 J.6. the equations defining the repeated impact curve of S125 and R75x100 specimens are now determined by substituting these values (Column 4 of Table 6.368.19 illustrates the repeated impact curve of the tubes.4). Ec and b).9 0.291 R75x100 1.217.9 J for S125 and R75x100 specimens. Note that the (Ec)Dynamic values are calculated using Equation 6.

The endurance fatigue of R75x100 specimen under repeated impact loading occurs below 300 J and 290 J. respectively. This indicates that the effect of the increase of incident energy in this region is minimal. it can be observed that the impact energies ranging between 300 J and 600 J. for an energy between 1400 J and 1. and 290 J and 400 J define the high cycle fatigue behaviour of S100 and R75x100 specimens. In this region. The curve indicates that no failure would likely to occur if the number of impacts at a corresponding level of energy falls below this curve. On the other hand.400 J.6 times the specific energy absorption of comparable square specimen. The significance of the repeated impact curve shown in the figure is that this provides an idea as to the required impact repetitions for a certain level of energy to initiate collapse or failure of the tube. This is expected since the cross section of the S125 specimen is higher than S100 and R75x100 specimens.000 J. For S100 and R75x100 specimens. respectively. Moreover. This region corresponds to the high cycle fatigue behaviour of S125 specimen. For an energy level lower than 1.000 J. This indicates that the range of the energies constituting the low and high cycle fatigue regions increases with increasing Ec. the minimum value of energy at the high cycle fatigue region of S125 is greater than S100 and R75x100 specimens relative to their Ec. the curve of S125 specimen demonstrated that the rate of failure was very slow. Behaviour of glass FRP composite tubes under repeated impact for piling application 150 . This range of impact energy with its corresponding number of impact defines the low cycle fatigue behaviour of S125 specimen. Mamalis et al.Chapter 6 –Damage modelling of repeatedly impacted FRP composite tube EJ Guades equation of each corresponding tube. the rate of reduction is less rapid compared to the rate in the low cycle fatigue region. On the other hand. In fact.19. Similarly. one can observe that the Ec of S125 specimen is relatively higher than the other two specimens. From Figure 6. (1997a) and Kindervater (1990) reported that rectangular cross section tube has 0. it can be observed that the initiation of failure for S125 specimen is rapid if it is impacted by incident energy higher than 1. By comparing the three specimens. This region can be described as region of the endurance fatigue of S125 specimen. The lowest level of energy under low cycle fatigue region of S125 is comparably higher than that of the two specimens. the curve along this region tends to become parallel to the x-axis. it can be observed that the minimum energy constituting the low cycle fatigue region is around 600 J and 400 J. respectively. a smaller increment of impact energy would already provide a higher value of Nf as compared to the low cycle fatigue region with same impact energy increment.

e.570 J and 1. On the other hand. the curve traced by the damage evolution model provides the degree of damage in the non-collapsed region. The reduction rate is clearer when the tube is impacted by 20% of Ec (i. S100 and R75x100 specimens using the model (Equations 6.20 shows the damage evolution curves of S125.265.Chapter 6 –Damage modelling of repeatedly impacted FRP composite tube EJ Guades Incident energy. 235.e.20a). 2. This curve illustrates the quantitative damage due to repeated impact loading on the tube from its non-damage to a fully damaged state (collapse or failure). the reduction rate of the slope is not constant relative to the magnitude of the applied energy. and 150 for S125. S100 and R75x100 specimens. It should be noted that the slope of the curve is inversely proportional to the N value. % of Ec) up to failure of the tubes. For instance. respectively) fall in the energy constituting the low cycle fatigue of the tube.930 J. the change in slope of 40% and 20% of Ec relative to the slopes at the low cycle fatigue region is more evident. This result is expected to happen since the incident energies of 80% and 60% of Ec (2.. it can be noticed that the slope of the curve in the non-collapse region decreases with decreasing applied incident energy. the slopes in damage evolution curve under this region will be comparable. It should be noted that the Nmax adopted for each curve corresponds to the N value at the start of collapse when impacted by 20% of Ec (i.10 or/and 6. Unlike the repeated impact curve. Ein (J) 3500 3000 S125 2500 S100 2000 R75x100 1500 1000 500 0 0 100 200 300 400 500 600 700 800 900 1000 Number of impacts to failure. the slope of the curve for 80% and 60% of Ec almost coincides with each other. For S125 specimen (Figure 6.. respectively). the failure of the tube is influenced by the applied incident energy with not much on the number of impact. 660 J). This incident energy is within the energy range describing the region of the endurance Behaviour of glass FRP composite tubes under repeated impact for piling application 151 .11). In the low cycle fatigue region..e. As a result. Moreover. Nf Figure 6.19 Repeated impact curves of the square and rectangular tubes Figure 6. The curve shown in the figure illustrates the damage evolution when impacted by different levels of energy (i.

60 0.80 Ec 0.00 0.60 0. Damage parameter.00 Life fraction.00 0.20 0.20 x Ec 0.80 x Ec 0.00 0.60 x Ec 0.00 Life fraction. It is therefore expected that the change in slope will be faster since in this region the failure of the tube is controlled by the number of impact rather than the incident energy.20 x Ec 0.60 0.40 0.00 0. N/Nmax (a) S125 1.40 0.80 Ec 0.80 1.Chapter 6 –Damage modelling of repeatedly impacted FRP composite tube EJ Guades fatigue of S125 specimen.20 0.60 0.00 Life fraction.80 x Ec 0.20 1.60 x Ec 0.20 Damage parameter.80 Ec 0.20 Damage evolution curves of square and rectangular tubes Behaviour of glass FRP composite tubes under repeated impact for piling application 152 .00 0.40 0.40 x Ec 0.40 0.20 1.80 1.20 0.40 0.60 0.60 0.20 0.60 x Ec 0.00 0.80 x Ec 0.00 0. D 1. N/Nmax (c) R75x100 Figure 6.40 0.80 1.20 0.40 x Ec 0. D 1.00 0. D 1.40 x Ec 0.20 0.20 x Ec 0.00 0. N/Nmax (b) S100 Damage parameter.

The variation is less than 10% indicating that the model predicted reasonably the damage evolution of the tube subjected by repeated impact loading. the repeated impact (fatigue) curve of the composite tubes follows a power function correlation. the experimental data points and the fatigue curve fairly agreed with each other. Moreover. This was evidenced by both the experimental data points and the repeated impact equation. The results showed that the energy obtained from quasistatic compressive test can be used in determining the dynamic critical energy by carefully selecting a suitable value of the correlation factor. This is because the range of the number of impact constituting the regions in the damage evolution curve decreases with decreasing Ec values of the tubes. 6. The repeated impact curve provides an idea as to the required impact repetitions for a certain level of energy to initiate collapse or failure of the tube.Chapter 6 –Damage modelling of repeatedly impacted FRP composite tube EJ Guades The trend of the change in slope observed in S125 specimen can be observed also in S100 and R75x100 specimens (Figures 6. On the other hand. Impact and quasi-static compressive tests on composite tubes were undertaken in determining the parametric values and in validating the proposed model. One comment is worthwhile in comparing their damage evolution curves. some of the parameters used in the model were directly obtained from the literature. it indicates that no failure would likely to occur if the number of impacts at a corresponding level of energy falls below this curve. Just like the composite laminates. The application of the damage evolution model was extended to available glass fibre/vinyl ester tubes having square and rectangular cross sections. It can be seen that the decrease of the rate of D with respect to N/Nmax is faster in specimen R75x100 compared to S125 and S100 specimens. The damage was characterised using a damage parameter D based from the energy principle.20b and 6. The correlation factor is a function of the rate of loading used in static and dynamic tests. It was Behaviour of glass FRP composite tubes under repeated impact for piling application 153 .10 Conclusions This chapter presented a lifetime prediction model that determines the damage response of square FRP composite tube subjected by repeated axial impact loading.20c). This indicated that a simple static compressive test can substitute a relatively expensive and complex dynamic (impact) test in finding the critical energy. Consequently. It was found that the experimental results and the proposed damage model agreed well with each other.

the application of the proposed damage model to other composite tubes of different matrix material (i. Behaviour of glass FRP composite tubes under repeated impact for piling application 154 .Chapter 6 –Damage modelling of repeatedly impacted FRP composite tube EJ Guades revealed that S125 specimen has a higher value of energy that describes the repeated impact regions compared to S100 and R75 specimens.. polyester and epoxy) and vinyl ester (with cross sections other than presented in this chapter) that are referenced from the literature is discussed. This result indicated that the range of energies under these regions is highly dependent on the value of Ec.e. The application of the model found to be suitable to vinyl ester tubes reinforced by glass fibres. It would also be worthy to characterise its usage to other matrix materials such as polyester and epoxy. In Chapter 7.

or from different matrix materials (i. Note that the application of the model is extended to other composite tubes made of polyester or epoxy matrices since they are also commonly used as matrix materials in FRP composite tubes. 2010.g. It was reported that the shape of the loaddeformation curve under progressive crushing test (dynamic or quasi-static tests) for tubes made of glass fibres and vinyl ester/polyester/epoxy matrices exhibit comparable shape (Palanivelu et al.1 Introduction The prediction model has been reported and found that it can characterise the damage evolution response of FRP composite tubes. the model can be used to predict the damage response of the tubes made of polyester and epoxy reinforced by glass fibres or vinyl ester of different cross sections. and rectangular) are emphasised. The value of Ec is mainly dependent on the absorption capacity of the tube and usually obtained from its load-deformation curve during the progressive crushing test. FRP composite tubes made of different geometries (e.. square. 1997a.e. polyester and epoxy)..9. vinyl ester. polyester. This chapter discusses the application of the damage model to other types of composite tubes. rectangular.9. However. Mamalis et al. Moreover. the tubes described in this chapter are all referenced from the literature.. circle. circular etc. and epoxy) is reported..Chapter 7 – Application of the model to other types of composite tubes EJ Guades Chapter 7 Application of the damage evolution model to other types of composite tubes 7. square. Similarly. The model predicted reasonably the damage response of a square composite tube made from glass/vinyl ester material..) have similar load-deformation trend as reported earlier in Section 6. its application can also be extended to composite tubes manufactured from vinyl ester with cross section other than square or rectangular ones. Tubes of different sizes and geometries (e.g. Unlike the tubes presented in Section 6. Likewise.. Consequently. the application of the model to tubes of different matrix materials (i. The parameters used in Behaviour of glass FRP composite tubes under repeated impact for piling application 155 .H. and Thornton P. 1990).e. consideration should be carried out in selecting appropriate parameters including the critical energy Ec to predict soundly their damage response..

2002). and epoxy (Helmi et al. and Mirmiran et al.. the successful structural performance of a composite is greatly dependent on both constituent phases. 2006. Sakr et al... the composite tube that was adopted in deriving the damage evolution model used glass fibres as reinforcement.. polyesters and epoxies. On the other hand. Ashford and Jakrapiyanun.. 2001).1 Vinyl ester resin Vinyl ester resins are unsaturated esters of epoxy resin (Kim. carbon. While the reinforcement fibres are responsible for determining main structural properties such as tensile strength and stiffness in the fibre direction.Chapter 7 – Application of the model to other types of composite tubes EJ Guades the model for this application were derived experimentally or from the information published in the literature. the common fibre reinforcement used in composite tube for piling application is glass fibre (Sirimanna. Mirmiran et al. A summary of information on these resins is presented in the next sections... 1995).e. aramid) is its economic cost.1. 2002. To date. 2005.2.. 2006. 2011. 7. Similarly. Reasonably. The chief advantage of glass fibres compared to other types of reinforcement fibres in civil engineering application (i. In Chapter 6. 2005. glass fibres provide distinctive advantage in terms of their compatibility with broad range of polymer systems such as vinyl esters.. polyester (Pando et al. the common resins that are used in composite tubes for piling application are vinyl ester (Sirimanna. 2011). These resins are often identified as a class of unsaturated polyester thermosetting resins because of the curing and processing similarities. these types of resins (matrices) are the subject of interest in the damage evolution model since they are considered suitable for application as emphasised in Section 7. Additionally. 2006). 2006. Helmi et al. 7. They have similar properties as epoxies and processibility of polyester. the application of the model is limited on composite tubes which are reinforced by glass fibres. Vinyl ester resins are formed by the reaction of epoxy resins with acrylic or methacrylic acid. They have higher elongation and corrosion Behaviour of glass FRP composite tubes under repeated impact for piling application 156 . Vinyl ester resins have been found to offer exceptional chemical resistance characteristics and have been the matrix material of choice in harsh chemical environments. Sakr et al.2 Background on the constituents of composite tubes used in the model The two main constituents of fibre composite materials are the reinforcement fibres and the polymer matrix. Pando et al.

2011). high mechanical properties. such as styrene (Barbero. but maintaining the processing versatility of polyesters (Barbero. which are dissolved in a reactive monomer. have poor resistance to acids. These resins do present several attractive features from a civil engineering perspective. Behaviour of glass FRP composite tubes under repeated impact for piling application 157 . clear liquids based on unsaturated polyesters. In general. Polyester resins have good resistance to chemical attack and have been used for many corrosion-resistant structures (Kim.2. 2011). While epoxy resins are regarded as covering the high end of the polymer matrix performance field and polyester as covering the lower end. providing a transition in properties and cost to the high performance epoxy resins. while it is possible to formulate very high performance polyester resins. Epoxy resins are widely used because of their versatility. such as organic peroxides. results in a cross linking reaction between the unsaturated polymer and unsaturated monomer. 1995). These resins can endure extreme exposure to the elements for a period of more than 30 years. Epoxy resins have higher specific strength and dimensional stability. however.3 Epoxy resin Epoxies are a broad range of products with a common epoxy ring consisting of two carbon atoms single bonded to an oxygen atom (Kim.Chapter 7 – Application of the model to other types of composite tubes EJ Guades properties than polyesters. and better resistance to solvents and alkalis compared to polyester resins.2. these materials can cost as much as or more than competing vinyl ester or epoxy systems. they possess good high temperature resistance and are normally used at temperatures up to 1770 C or as high as 3160 C. converting the low-viscosity solution into a three-dimensional thermoset polymer. 7. Epoxies shrink less than the other two resin materials which help explains their excellent bond characteristics when used as adhesives. 1995). 1995). The addition of heat and a free radical initiator system. glass/epoxy composites maybe of higher quality than glass/polyester composites. However. including a lower cost structure than epoxy. 2011). For civil engineering application. the higher cost and difficulty in processing of epoxy may counterweigh such advantage. although some discoloration and loss of strength may occur (Schweitzer. and high corrosion resistance (Barbero. On the other hand.2 Polyester resin Polyester resins are low velocity. 7. vinyl ester resins hold the middle ground. Epoxy resins. UV and weathering.

Chapter 7 – Application of the model to other types of composite tubes

EJ Guades

In the next sections, the application of the damage evolution model to
composite tubes made from vinyl ester, polyester, and epoxy resins is presented.

7.3 Glass/vinyl ester composite tubes
Table 7.1 summarises the information of the tube from the literature used for
application in the damage evolution model of glass fibre/vinyl ester composites.
Aside from square section, the table indicates circular and hourglass cross sections of
the composite tubes. As emphasised in Section 7.1, the application of the model to
other geometries aside from square section is suitable since the shape of the loaddeformation curve of these tubes are comparable. In relation to the absorbed energy
of different cross section, Mamalis et al (1997a) and Kindervater (1990) stated that
square and rectangular cross-sectioned tubes have, respectively, 0.8 and 0.5 times the
specific energy absorption of comparable circular specimens. They emphasised that
the lower specific absorption of square and rectangular sections is generally
attributed to the fact that the corners act as stress concentrator leading to the
formation of splitting cracks. In Table 7.1, the first two letters in the designations of
the tube (i.e., GV) indicates glass/vinyl ester, whilst the last letter indicates the
geometrical section.

Table 7.1 Details of GV-C, GV-S, and GV-H tubes
Designation

GV-C

GV-S

GV-H

Geometry

Circular

Square

Hourglass

Diameter (outside, mm)

38

-

b

Depth, d (outer, mm)

-

47.7

b

Width, b (outer, mm)

-

47.7

b

Thickness, t (mm)

3

2.5

3.3

220

101.6

76.2

Specific mass, ρ (kg/m )

1,819

1,550

1,550

Glass content (%)

50.4a

33.9a

33.9a

Length, l (mm)
3

Reference

Palanivelu et al. Mamalis et al. Mamalis et al.
(2010)
(1997a)
(1996)
a
by volume, bdetails of the dimension can be found on the indicated reference

Behaviour of glass FRP composite tubes under repeated impact for piling application

158

Chapter 7 – Application of the model to other types of composite tubes

EJ Guades

To obtain the damage evolution curve of the composite tubes using the
model, their repeated impact curves are determined first. The repeated impact curve
(Equation 6.12 in Chapter 6) will again presented for ease of discussion.
Ein = Ec Nf-b

(7.1)

The repeated impact curves of GV-C, GV-S, and GV-H specimens were
obtained based from Equation 7.1. The Ec value of each tube which is used in
Equation 7.1 was derived based from the load-displacement curve of the tube being
considered. Due to the absence of numerical values representing the data points on
the curve, Ec was calculated approximately using the mean sustained crushing load ̅
multiplied by the crushing displacement x. The value of the crushing displacement
was measured up to the point where the crushing load starts to stabilise. The energy
obtained using this procedure (i.e., ̅ x) provides the critical impact energy that will
crush the tube for one impact. It should be noted that the total energy (specific)
absorbed during crushing of the composite tubes under progressive collapse is a
function of ̅ x, specimen cross section, and the material density (Abdewi et al.,
2006). Table 7.2 shows the summary of the (Ec)Quasi-static and β values for GV-S and
GV-H tubes. For GV-C tube, these values are not indicated since the cited reference
employed dynamic (impact) testing and therefore the value of (Ec)Dynamic can be
directly obtained from the curve.
Table 7.2 Summary of (Ec)Quasi-static and β values of glass/vinyl ester tubes
Tube
GV-C

̅ (kN)
38.66

x (mm)
10

(Ec)Quasi-static (J)
a

β
a

GV-S

40.60

5

203

1.302b

GV-H
132.60
5
663
1.302b
a
From dynamic (impact) test, bfrom Equation 6.14 using a drop height of 3.3 m

Table 7.3 summarises the values of (Ec)Dynamic, b, and the equation of the
repeated impact curves of the composite tubes. It should be noted that the (Ec)Dynamic
values of GV-S and GV-H in Table 7.3 were calculated using Equation 6.13 whilst
for GV-C, its value is directly obtained from the load-deformation curve. On the
other hand, the value of b from Equation 7.1 was computed using “Solver” function.
The constraints used in 100x100 mm section were adopted except that the Ec values
were changed to 386.60 J, 264.31 J., and 863.23 J for GV-C, GV-S, and GV-H

Behaviour of glass FRP composite tubes under repeated impact for piling application

159

Chapter 7 – Application of the model to other types of composite tubes

EJ Guades

specimens; respectively. The value of b in the table represents the average value of
20 runs. Substituting the values of Ec and b to Equation 7.1 we can get the repeated
impact equation of the three tubes (Column 4 in Table 7.3)

Table 7.3 Summary of the repeated impact equation of glass/vinyl ester tubes
Tube
GV-C

(Ec)Dynamic (J)
386.6

b
0.459

Repeated impact equation
Ein = 386.6 Nf-0.459

GV-S

264.3

0.470

Ein = 264.3 Nf-0.470

GV-H

863.2

0.363

Ein = 863.2 Nf-0.363

Figure 7.1 shows the repeated impact curve of the three tubes. It
should be noted that the value of Ein in the curve was obtained by pre-assigning value
of Nf in the repeated impact equation shown in Table 7.3. As mentioned in Section
6.9.1, the significance of the repeated impact curve is it provides information as to
the required impact repetitions at a certain level of energy to initiate collapse or
failure of the tube. The repeated impact curve shows that there will be no failure or
rupture will occur if it is subjected by a number of impact at specific energy below
this curve. In the curve, the Nmax adopted corresponds to the N value at the start of
collapse when impacted by 20% of Ec. Similarly, this relationship in obtaining Nmax
was also used for the tubes with polyester or epoxy as the matrix material.
For GV-C tube, the minimum impact energy defining the low cycle fatigue
behaviour is approximately 80 J. This is 21% of the Ec value of the GV-C tube. On
the other hand, the range of incident energies describing the high cycle fatigue region
of GV-C is between 80 J and 25 J. The lower limit of energy under this region is 6%
of Ec. Energies lower than 25 J are considered energies that fall under the region of
the endurance fatigue. For GV-S and GV-H tubes, the lowest level of energy that
characterises the low cycle fatigue behaviour is around 25 J and 150 J, respectively.
A relatively higher value measured for specimen GV-H is expected since its Ec value
is higher compared to that of GV-C and GV-S tubes. These values are 17% and 10%
of their corresponding Ec. The range of energy defining the high cycle fatigue region
for GV-S tube is between 25 J and 10 J whilst between 150 J to 80 J for GV-S tube.
The lowest limit values are 9% and 4% of their equivalent Ec values. Energies lower
than 80 J and 10 J define the endurance fatigue region of tubes GV-S and GV-H,
respectively. A consistent trend can be observed in comparing with the repeated

Behaviour of glass FRP composite tubes under repeated impact for piling application

160

Chapter 7 – Application of the model to other types of composite tubes

EJ Guades

impact curve of GV-C, GV-S, and GV-H tubes. The range of the energies describing
the low and high cycle fatigue regions increases with increasing Ec values.

Incident energy, Ein (J)

1000

800

GV-C
GV-S

600

GV-H

400
200
0
0

100

200

300

400

500

600

700

800

900

1000

Number of impacts to failure, Nf

Figure 7.1 Repeated impact curves of glass/vinyl ester tubes
The damage evolution curve of GV-C, GV-S, and GV-H tubes is displayed in
Figures 7.2 to 7.4. From the figures, one can observe that the slope of the curve
changes with magnitude of the applied impact energy. A trend can be noticed relative
to the change of slope of the curve whereby the rate of decrease becomes faster with
decreasing impact energy. It should be noted that for GV-C tube (Figure 7.2), the
incident energy that will crush the tube for 2 repeated impacts is equivalent to 73%
of Ec. This value was used in comparison since an 80% of Ec cannot be obtained from
this condition unlike the case of other tubes. It can be noticed from Figure 7.2 that
there has been no significant difference occurred between the change of slope from
the initial curve (i.e., impacted by Ec) when it is impacted by 73% and 60% of Ec.
The reason is that the 281 J and 233 J (i.e., 73% and 60% of Ec, respectively) are in
the range of energies describing the low cycle fatigue region. The influence of the
number of impacts (inversely proportional on the slope of the curve in the model) on
the collapse of the tube is negligible. It is therefore expected that the change in slope
between the curves impacted by 73% and 60% of Ec is comparable. In fact, the
relative difference between the number of impacts of these energies is same (i.e., 1
and 2 impacts for 73% and 60% of Ec, respectively). On the other hand, a clear
change of slope (decrease) can be noticed when the tube is impacted by 40% and
20% of Ec. This indicates that the rate of reduction on the slope of curve in the
damage evolution of the tube becomes quicker in the high cycle and endurance
fatigue regions. The result obtained from GV-C tube on the decrease of slope in the

Behaviour of glass FRP composite tubes under repeated impact for piling application

161

Chapter 7 – Application of the model to other types of composite tubes

EJ Guades

low cycle fatigue region can be observed also for GV-S and GV-H tubes (Figures 7.3
and 7.4, respectively). The reduction is not substantial when GV-S and GV-H tubes
are impacted by 80% and 60% of Ec. The decrease, however, becomes apparent
when the tubes are subjected by 40% and 20% of their corresponding Ec.
1.20

Damage parameter, D

1.00
0.80

0.60

Ec
0.73 x Ec

0.40

0.60 x Ec

0.40 x Ec

0.20

0.20 x Ec
0.00
0.00

0.20

0.40

0.60

0.80

1.00

Life fraction, N/Nmax

Figure 7.2 Damage evolution curves of GV-C tube

Damage parameter, D

1.20
1.00
0.80
0.60

Ec
0.80 x Ec

0.40

0.60 x Ec
0.40 x Ec

0.20

0.20 x Ec

0.00
0.00

0.20

0.40

0.60

0.80

1.00

Life fraction, N/Nmax

Figure 7.3 Damage evolution curves of GV-S tube
1.20

Damage parameter, D

1.00
0.80
0.60

Ec
0.80 x Ec

0.40

0.60 x Ec

0.40 x Ec

0.20

0.20 x Ec
0.00
0.00

0.20

0.40

0.60

0.80

1.00

Life fraction, N/Nmax

Figure 7.4 Damage evolution curves of GV-H tube

Behaviour of glass FRP composite tubes under repeated impact for piling application

162

is placed to distinguish the study relative to other studies.6 3 3.490 No data Glass content (%) No data 51. mm) 60 50 50 Width. In the tables. (2011) Palanivelu et al. Table 7.5 summarises the information of the tube used for application in the damage evolution model of glass/polyester composites. l (mm) 128. t (mm) 4. ρ (kg/m3) 1. b (outer. (2010) Velmurugan et al.g. mm) 60 50 50 Thickness. d (outer.5 5 4. (2004) Table 7.4 Glass/polyester composite tubes Tables 7.6 Specific mass.. ρ (kg/m3) 1. They reported that the increase in absorbed energy was due to the better inter-laminar strength and the higher strain to failure of vinyl ester resin. The number. on the other hand. the studies conducted by Palanivelu et al. With regards to the energy absorption capacity of glass/vinyl ester and glass/polyester composite tubes.068 1.3 Length. (2000) Behaviour of glass FRP composite tubes under repeated impact for piling application 163 . Glass/vinyl ester absorbed energy as much as 33% higher than the glass/polyester tube.Chapter 7 – Application of the model to other types of composite tubes EJ Guades 7.3 220 5.5 Details of glass/polyester tubes (square cross section) Designation GP-S1 GP-S2 GP-S3 Geometry Square Square Square Depth.4 Details of glass/polyester tubes (circular cross section) Designation GP-C1 GP-C2 GP-C3 Geometry Circular Circular Circular Diameter (inside.400 Glass content (%) 49. mm) 70 50 70 Thickness. The tables indicate square and circular cross sections as the subject in this section. (2010) revealed that the former absorbed more energy than the latter. (2010) Ismael and Ahmad (2007) Saito et al.2 62. respectively. t (mm) 3. GP-C) indicates glass/polyester and the cross section.2 50 Reference Palanivelu et al.665 1.7 No data Reference Stamenovic et al.4 and 7.3 Length. l (mm) 220 100 80 Specific mass. the first two and the last letter in the designations of the tube (e.812 2.

029 α (x10-3) Figure 7.00 Mamalis et al. For GP-C2 and GP-S1 tubes.50 0. Similarly. GP-C3. Equation 7. and GP-S3.00 0.2 was used in calculating the value of β for glass/polyester composites. Data points from Mamalis et al.023 0.021 0.025 0. on the other hand.027 0. This value was calculated using the mean sustained crushing load ̅ multiplied by the crushing displacement x as emphasised previously in Section 7.028 0. As a result.. (2002) reported that glass/polyester tubes which are axially loaded are strain rate sensitive. This result was also verified by Mamalis et al. (1997a) Behaviour of glass FRP composite tubes under repeated impact for piling application 164 . the repeated impact curves of glass/polyester tubes were obtained using Equation 7. (1997a) whereby the absorbed energy obtained from dynamic testing on square tubes underestimated the values acquired from static testing by as much as 18%.13).5 Data points with the fitting line showing β and α relationship of glass/polyester tubes.00 β = -12.020 0. For GP-C1. (1997a) β 1. β value is less than 1).024 0. (1994) reported that the absorbed energy during dynamic testing is lower than the static testing for fibre glass/polyester circular tubes and frusta (i. Equation 6. This means that a correlation of speed testing adopted in quasi-static test should be associated to the impact velocity from the dynamic test in simulating the reasonable (Ec)Dynamic value. The result of their study was adopted in finding the value of β because. not only that the material used in their study is glass/polyester composites. the value of Ec (or (Ec)Dynamic) was obtained based from the load-displacement curve of the tube being considered.026 0.862α + 1.Chapter 7 – Application of the model to other types of composite tubes EJ Guades Just like the glass/vinyl ester composites.5.50 1. Equation 7. (1997a) as shown in Figure 7.1.022 0.0999 0.. GP-S2.e. β values need to be determine in finding the relationship between (Ec)Dynamic and (Ec)Quasi-static (i. but also the tests were performed using different loading rates (static and dynamic tests). Mamalis et al.3.2 described a fitting line for the experimental results conducted by Mamalis et al. Jacob et al. the (Ec)Dynamic values can be directly obtained from the reference since these studies employed dynamic testing.e. 2.

2) where α is defined by Equation 6.Chapter 7 – Application of the model to other types of composite tubes EJ Guades From Figure 7.336 Ein = 885.0 β 0.8 0.78 15 416.5.862 α +1.390 Ein = 620. Table 7. we can get the repeated impact equation of the glass/polyester tubes (Column 4 in the table).8 Nf-0.13.7.322 GP-C2 566.47 15 922.5 b 0.00 x (mm) 17 (Ec)Quasi-static (J) 1.044b GP-S1 31. For glass/polyester tubes.7 1.426 Ein = 566. The value of b for glass/polyester composite tubes was computed using “Solver” function.14 using a drop height of 3 m The summary of the (Ec)Dynamic and b values.7 is the average of 20 runs.390 GP-S2 793. Just like the b value of the glass/vinyl ester tubes.01 20 a a GP-S2 75.1. Table 7.6 Nf-0.961b a From dynamic (impact) test.433 Ein = 435. the value in Table 7.0 1. the value of (Ec)Dynamic for GP-C2 and GP-S1 were directly obtained from its load-deformation curve while the rests using Equation 6. By substituting the values of Ec and b to Equation 7.7 Summary of the repeated impact equation of glass/polyester tubes Tube GP-C1 (Ec)Dynamic (J) 1.34 20 a a GP-C3 27.0 0.00 10 750.15. is displayed in Table 7.6 0.012.1 Nf-0.433 GP-S1 620.054. bfrom Equation 6.8 0.058b GP-S3 61.6 summarises the (Ec)Quasi-static and β values of the glass/polyester composite tubes. Table 7.2 0.1 0.6 Summary of (Ec)Quasi-static and β values of glass/polyester tubes Tube GP-C1 ̅ (kN) 62.356 Ein = 793.961b GP-C2 28.336 Behaviour of glass FRP composite tubes under repeated impact for piling application 165 .8 Nf-0. Note that in the table.012.356 GP-S3 885.5 Nf-0.426 GP-C3 435.0999 (7.2 Nf-0. the constraints used in finding b for glass/vinyl ester composites are same except that the Ec values in the latter where substituted by the actual Ec values of glass/polyester composites. as well as the equation of the repeated impact curve for the glass/polyester tubes.322 Repeated impact equation Ein = 1. β = -12.

and 14% of their corresponding Ec values. the energy defining the low cycle fatigue regions are approximately 90 J and 120 J. The lowest limit of incident energy under this region is 30% of Ec. The energies that described the endurance fatigue regions of GP-S are those energies lower than 60 J. One can observe that the values of GP-C2 and GP-C3 tubes in this region are comparably lower to that of GP-C1 specimen. extends between 300 J and 150 J. The lowest limit describing the low cycle fatigue region of GP-S1 specimen is around 100 J. The higher the Ec.Chapter 7 – Application of the model to other types of composite tubes EJ Guades Figure 7. These values are 21% and 20% of their corresponding Ec values. These values are. The energy corresponding to the lowermost limit is 15% of its Ec value.e. Behaviour of glass FRP composite tubes under repeated impact for piling application 166 . on the other hand. respectively.. The range of incident energies characterising the high cycle fatigue region. Similar with the comparison of values on the low cycle fatigue region.013 J and 300 J. The endurance fatigue regions of GP-C2 and GP-C3 tubes were defined by incident energies less than 50 J and 30 J. These values are 9% and 7% of their corresponding Ec values. In Figure 7. At these energy levels. respectively. For GP-S2 and GP-S3 tubes. the lowest level of energy of GP-C1 exhibits higher value than the other circular glass/polyester composite tubes in the region of high cycle fatigue. respectively. For the two other circular tubes (i. This result is due to the fact that the Ec value of GP-C1 is much higher than the other circular glass/polyester tubes. the lowest incident energies that cover the regions of low fatigue cycle are approximately 120 J and 85 J.6 (marked with solid lines). the effect of the variation of energy is considered negligible since the trend of the curve is dictated by the number of impact. The repeated impact curves of glass/polyester tubes with square cross section are also displayed in Figure 7. the higher will be the energy characterising the limit at the low cycle fatigue region. it can be observed that the low cycle fatigue region of GP-C1 tube (circular tube. This value is 10% of its corresponding Ec value. respectively.6. 11%. respectively. Incident energies lesser than 150 J constitute the energies that describe the endurance fatigue region of GP-C1 tube. For GPS2 and GP-S3 tubes the values are around 180 J and 230 J.6 shows the repeated impact curve of the glass/polyester composite tubes. GP-C2 and GP-C3). On the other hand. dotted line) occurred with applied incident energy between 1. On the other hand. respectively. these figures show that the lowest level of energy constituting the region of high cycle fatigue of GP-S1 is approximately 60 J. the lowest level of energy that defines the high cycle fatigue area of GP-C2 and GP-C3 tubes is 50 J and 30 J.

respectively) levels of energy are in the range of energies constituting the low cycle fatigue region.7 to 7. Therefore this result is expected since in this region failure of the tube Behaviour of glass FRP composite tubes under repeated impact for piling application 167 .6 Repeated impact curves of glass/polyester tubes Figures 7.7) indicate that no major difference transpired between the slopes of the curve when it is impacted by 80% and 60% of Ec. It is worth noting that 810 J and 600 J (80% and 60% of Ec.Chapter 7 – Application of the model to other types of composite tubes EJ Guades On the other hand. Nf Figure 7. The comparison of the repeated impact curve between glass/vinyl ester and glass/polyester is not straightforward.6 indicated that the lowest energy level at specific region increases with increasing Ec values of the tubes. This is because the tubes have different cross sectional areas and the fibre reinforcements (lay-up and content) vary from one another. it is evident that the repeated impact curve (or level of energy in the three regions) of the glass/vinyl ester will be much higher compared to the glass/polyester tube. Ein (J) 1200 1000 800 600 GP-C1 GP-C2 GP-C3 GP-S1 GP-S2 GP-S3 400 200 0 0 100 200 300 400 500 600 700 800 900 1000 Number of impacts to failure. Incident energy. however. This is because glass/vinyl ester composites exhibit comparably better energy absorption behaviour than the glass/polyester material (Palanivelu et al. energy lesser than 90 J and 120 J describes the endurance fatigue regions of GP-S2 and GP-S3 tubes. The reduction rate. The trend observed on glass/polyester tubes having circular cross section relative to the variation of incident energies on the three regions was also found in the tube with square cross section. It can be seen from the figure that the slope of the curve before the collapse decreases at decreasing applied energy.12 display the damage evolution curves of the glass/polyester composite tubes. 2010).. depends on the magnitude of the energy whereby the rate becomes faster at a relatively smaller energy value. The curves of GP-C1 tube (Figure 7. Figure 7. However.

00 0. D 1.00 0. a huge reduction of the slope can be seen when the tube is impacted by 207 J (i.00 Life fraction.00 0.60 0.7 Damage evolution curves of GP-C1 tube Damage parameter.20 1.40 0. 40% of Ec).40 0.8 to 7.. The result obtained from GP-C1 tube on the reduction of slope in the region of low cycle fatigue can also be seen for other glass/polyester tubes (Figures 7.20 0. D 1.80 Ec 0.Chapter 7 – Application of the model to other types of composite tubes EJ Guades is dominated by the level of energy.20 0.60 0.40 x Ec 0.20 x Ec 0.80 Ec 0.e.20 0. This was also the result observed in glass/vinyl ester tubes whereby no significant change in slope of the curve in this region.20 0. The decrease in the slope of the curve for GP-C1 tube is noticeable when it is subjected by 410 J (i..60 0.12).20 x Ec 0. N/Nmax Figure 7.8 Damage evolution curves of GP-C2 tube Behaviour of glass FRP composite tubes under repeated impact for piling application 168 .80 x Ec 0.e.00 Life fraction.74 x Ec 0. The reduction of the slopes of these tubes is significant when they are subjected by 40% and 20% of their corresponding Ec.00 0.80 1.40 0.80 1. Moreover.40 0.40 x Ec 0.00 0.20 1. The decrease is not significant when these tubes are impacted by incident energies describing the low cycle fatigue region (80% to 60% of Ec).60 x Ec 0. 20% of Ec). N/Nmax Figure 7.00 0.60 x Ec 0.60 0. Damage parameter.

60 x Ec 0.60 x Ec 0.00 0.20 0.60 0.20 x Ec 0. D 1.20 1.00 0.20 1.40 x Ec 0.75 x Ec 0.00 0.60 0.76 x Ec 0.00 Life fraction.9 Damage evolution curves of GP-C3 tube Damage parameter.Chapter 7 – Application of the model to other types of composite tubes EJ Guades Damage parameter.00 0.40 0. N/Nmax Figure 7. D 1.60 0.20 0.20 0.00 0.00 0.00 0.40 x Ec 0.00 Life fraction.40 0.40 x Ec 0.80 Ec 0.20 0.20 x Ec 0. N/Nmax Figure 7.20 1.80 1.60 0.80 Ec 0.20 0.80 1.11 Damage evolution curves of GP-S2 tube Behaviour of glass FRP composite tubes under repeated impact for piling application 169 .80 1.78 x Ec 0.20 x Ec 0.80 0.40 0.00 0.20 0.60 Ec 0.10 Damage evolution curves of GP-S1 tube Damage parameter. N/Nmax Figure 7.60 x Ec 0.40 0.60 0.40 0. D 1.00 0.00 Life fraction.40 0.

60 x Ec 0. the number in the designation is placed to distinguish the study relative to the other. mm) 74. D 1.20 0.20 1.12 Damage evolution curves of GP-S3 tube 7.8 and 7. They emphasise that the changes in matrix stiffness have little effect on the energy absorption capability of composite materials with ductile reinforcement. Table 7. they reported that further studies are essential to understand clearly the role of matrices in the energy absorption capability of the composite material..00 0.g.60 Ec 0. (2010) Ochelski and Gotowicki (2009) Kim et al.00 0. t (mm) 3. Just like in designating glass/polyester tubes.5 Glass/epoxy composite tubes Tables 7. GE-C) indicates glass/epoxy and the cross section. ρ (kg/m3) No data No data No data Glass content (%) No data No data 50.40 0. The geometries considered for application are tubes with square and circular cross sections as indicated in the table.60 0. Moreover.40 0.40 x Ec 0. The first two and the last letter in the designation of the tube (e.70 9 1. respectively.80 0.00 0.9 shows the information of the glass/epoxy composite tubes sourced from the literature used for application in the damage evolution model.30 30 Thickness.50 39.80 x Ec 0. N/Nmax Figure 7.00 Life fraction.20 0. (2009) Behaviour of glass FRP composite tubes under repeated impact for piling application 170 ..1 Reference Muralikannan et al.5 Length. In general. l (mm) 150 91 100 Specific mass.8 Details of glass/epoxy tubes (circular cross section) Designation GE-C1 GE-C2 GE-C3 Geometry Circular Circular Circular Diameter (inside.80 1.20 x Ec 0. glass/epoxy composite tubes absorbed more energy than the glass/polyester tubes (Jacob et al.Chapter 7 – Application of the model to other types of composite tubes EJ Guades Damage parameter. 2002).

(2010) whereby the mean load in impact loading decreases with the increase in impact velocity.5 m/s. l (mm) 300 100 100 Specific mass. As illustrated in Equation 6. ρ (kg/m3) No data No data 2.. Berry and Hull (1984) found that the specific energy of the glass/epoxy composites increases with increasing loading rates up to 8. Just like the glass/vinyl ester and glass/polyester tubes. This result.Chapter 7 – Application of the model to other types of composite tubes EJ Guades Table 7.13.3 Song and Du (2002) Ghasemnejad et al. the present study assumed that the value of (Ec)Quasi-static can be used as a direct value of (Ec)Dynamic (β ≈ 1) for those obtained using quasi static compressive tests. t (mm) 3. This correlation was already illustrated in determining the repeated impact and damage evolution curves of glass/vinyl ester or glass/polyester tubes. Table 7.9 3 Length.35 1. however.35 55 80a Thickness. (2009) Reference Aljibori et al. (2008) a outer width or outer depth (square section) The repeated impact curves of glass/epoxy tubes were obtained using Equation 7. Behaviour of glass FRP composite tubes under repeated impact for piling application 171 . the value of (Ec)Dynamic and (Ec)Quasi-static are correlated by a factor β. there are contradicting remarks with regards to this relationship. As a result.10 shows the summary of (Ec)Quasi-static and β values of the glass/epoxy composite tubes. can be derived using impact testing if desired. the value of Ec (or (Ec)Dynamic) of glass/epoxy composites was obtained based from the loaddisplacement curve of each corresponding tube. however.9 Details of glass/epoxy tubes (circular and square cross sections) Designation GE-C4 GE-C5 GE-S1 Geometry Circular Circular Square Diameter (inside.1. For glass/epoxy composite tubes.100 Glass content (%) No data No data 40. is because of the reduced fracture strength and reduced frictional energy absorption in impact loading. This result was also found by Muralikannan et al. A more accurate value of the (Ec)Dynamic. Schmueser and Wickliffe (1987) reported that the dynamic specific energy of glass/epoxy tube was lower compared to when it is loaded statically. In contrary. according to Muralikannan et al. mm) 103. This value was calculated using the mean sustained crushing load ̅ multiplied by the crushing displacement x.

The equation of the repeated impact curve was obtained by inputting the values of (Ec)Dynamic and b to Equation 7. the value of b from Equation 7. b values.00 10 200 1. Table 7.10 Summary of (Ec)Quasi-static and β values of glass/epoxy tubes Tube GE-C1 ̅ (kN) 50.0 GE-C5 20.13.514 GE-C4 400 0. dotted line). and the repeated impact equation of the glass/epoxy tubes.529 GE-S1 520 0.0 10 520 1.369 GE-C2 1.0 GE-S1 52.1 for glass/epoxy tubes was computed using “Solver” function.0 GE-C3 22. The range of energies that define the region of high cycle fatigue of GE-C1 tube is between 90 J to 150 J.514 Ein = 224 Nf-0. This range of energies characterises the low cycle fatigue region of the tube.321 GE-C3 224 0.177 1.177 Nf-0.13.1.70 10 1. Just like in glass/vinyl ester and glass/polyester composites.00 x (mm) 15 (Ec)Quasi-static (J) a β a GE-C2 117.Chapter 7 – Application of the model to other types of composite tubes EJ Guades Table 7.435 The repeated impact curve of the glass/epoxy composite tubes is displayed in Figure 7.529 Ein = 200 Nf-0.40 10 224 1. For GE-C1 tube (circular tubes. Note that (Ec)Dynamic was calculated using Equation 6.11 displays the summary of the (Ec)Dynamic.321 Ein = 1. The constraints used in finding b for glass/vinyl ester and glass/polyester composites are same except that the Ec values where changed to the actual Ec values of the corresponding glass/epoxy tubes.452 GE-C5 200 0.369 Repeated impact equation Ein = 750 Nf-0.452 Ein = 400 Nf-0.13 by taking into account the corresponding value of β. a slower Behaviour of glass FRP composite tubes under repeated impact for piling application 172 .0 GE-C4 40. As can be seen from Figure 7.177 0.435 Ein = 520 Nf-0. The lowest level of energy under this region is equivalent to 20% of its corresponding Ec.11 Summary of the repeated impact equation of glass/polyester tubes Tube GE-C1 (Ec)Dynamic (J) 750 b 0. it can be observed that the start of failure is quick when the tube is impacted by incident energy between 150 J to 750 J.00 10 400 1.00 a From dynamic (impact) test Table 7.

the lowest level of energies that characterise their high cycle fatigue region is within the range of 5% to 14% of their corresponding Ec value. The lowest level of energy characterising this region is 19% of the corresponding Ec value. On the other hand. 10% and 24%) correspond to GE-C2 and GE-C5. Energies lower than 60 J covers the energy explaining its endurance fatigue region.14 to 7. Incident energy. Just like the other two composite tubes. Incident energies lower than 90 J constitute the energies that describe the endurance fatigue region of GE-C1 tube.13 Repeated impact curves of glass/epoxy tubes The damage evolution curves of the glass/epoxy composite tubes are shown in Figures 7. GE-S1 tube). the lowest level of incident energy characterising the region of high cycle fatigue of GE-S1 tube is approximately 60 J. It should be noted that the values of the other tubes (GE-C3 and GE-C4) lie in between 10% to 24% of their Ec value..19.. respectively. the slope of the DN/Nmax curves decreases with decreasing applied energy. The curve in Figure 7. The extreme values (i. For glass/epoxy square composite tube (i. Nf Figure 7.Chapter 7 – Application of the model to other types of composite tubes EJ Guades rate of failure can be observed when impacted by this level of energies compared to the energies defining the low cycle fatigue region. Ein (J) 1200 1000 800 600 GE-C1 GE-C2 GE-C3 GE-C4 GE-C5 GE-S1 400 200 0 0 100 200 300 400 500 600 700 800 900 1000 Number of impacts to failure. On the other hand.e. The rate of reduction becomes quicker when the tube was impacted by smaller incident energy.e. For the other glass/epoxy composite tubes with circular cross sections. Incident energies lower than these values explain the endurance fatigue region of these tubes. the lowest level of incident energies that describe their low cycle fatigue regions are in between 10% to 24% of their Ec values. the energies that describe the low cycle fatigue region is roughly between 100 J to 520 J.14 shows that no significant deviations on the slope of the damage Behaviour of glass FRP composite tubes under repeated impact for piling application 173 .

This means that a large reduction of incident energy (e.20 1.20 x Ec 0. N/Nmax Figure 7.20 0.20 0.15 to 7.00 Life fraction.Chapter 7 – Application of the model to other types of composite tubes EJ Guades evolution curve of GE-C1 when it is impacted up to collapse by 80% and 60% of its Ec value.00 0.20 Damage parameter.40 x Ec 0.60 x Ec 0. incident energies greater than 60% of Ec are energies considered to be in low cycle fatigue region.40 0.40 0.40 0. The reduction in the slope of the curve for GE-C1 tube is noticeable when it is subjected by 40% and 20% of Ec. a significant reduction of slope can be noticed when they are impacted by 20% of Ec.60 0.80 Ec 0.14 Damage evolution curves of GE-C1 tube Damage parameter.g. From Figure 7.15 Damage evolution curves of GE-C2 tube Behaviour of glass FRP composite tubes under repeated impact for piling application 174 .00 0.00 0.60 0.80 1. from Ec to 60% of Ec) in this region will not significantly increase the rate of slope reduction of the curve. D 1.20 0.00 0.19.14.. The result found from GE-C1 tube on the rate of slope reduction in the region of low cycle fatigue can also be observed for the remaining glass/epoxy tubes.00 0.80 x Ec 0.60 x Ec 0.40 x Ec 0. the difference in slope when they are subjected up to collapse by at least 60% of Ec is relatively small. N/Nmax Figure 7.40 0. In this region. Looking on Figures 7. However.60 0.20 0.00 0. This result was also observed in the glass/vinyl ester and glass/polyester composite tubes wherein the rate of slope reduction is almost similar from Ec to 80% and 60% of Ec.60 Ec 0.80 0. 1. the initiation of failure of GE-C1 tube is highly dependent on the level of energy. D 1.00 Life fraction.80 x Ec 0.80 1.20 x Ec 0.

20 Damage parameter.20 x Ec 0.80 1.00 0. N/Nmax Figure 7.00 0.80 Ec 0.40 0.40 0.18 Damage evolution curves of GE-C5 tube Behaviour of glass FRP composite tubes under repeated impact for piling application 175 .73 x Ec 0.60 0.00 0.60 x Ec 0.00 0.40 x Ec 0.00 0.Chapter 7 – Application of the model to other types of composite tubes EJ Guades Damage parameter.40 0. N/Nmax Figure 7.60 0.20 0.80 Ec 0.40 x Ec 0. D 1.40 x Ec 0.00 Life fraction.40 0.20 0.80 Ec 0.60 0.00 0.00 0.20 0.60 x Ec 0.20 0.00 Life fraction.20 1.20 x Ec 0.70 x Ec 0.80 1. N/Nmax Figure 7.60 x Ec 0.40 0.70 x Ec 0.80 1.60 0.00 0.20 x Ec 0.16 Damage evolution curves of GE-C3 tube 1.40 0.60 0.20 0. D 1.20 1.20 0. D 1.17 Damage evolution curves of GE-C4 tube Damage parameter.60 0.00 0.00 Life fraction.

60 0.19 Damage evolution curves of GE-S1 tube 7.20 x Ec 0. the higher is the value of Ec. From the equation of the repeated impact curve. The result showed that for common Behaviour of glass FRP composite tubes under repeated impact for piling application 176 .60 0. and epoxy reinforced by glass fibre have been characterised.20 0. This indicates that the effect of the increase of energy above Ec value becomes insignificant on the failure of the tube. The curve indicates that no failure would likely to occur if the number of impacts at a corresponding level of energy drops below this curve.Chapter 7 – Application of the model to other types of composite tubes EJ Guades Damage parameter.6 Discussion on the repeated impact and damage evolution curves of FRP composite tubes The repeated impact (fatigue) curve and the damage response of composite tubes made of vinyl ester.00 Life fraction.20 0. D 1.74 x Ec 0.60 x Ec 0. Incident energy with value below Ec needs to be associated to several numbers of impacts to collapse or fail the composite tube. Studying the effect of tube dimensions.80 1. The repeated impact curve traces three regions that define the impact damage tolerance of the impacted tubes.00 0.00 0.40 x Ec 0. On the other hand. the tube is expected to fail at a single impact if the applied impact energy exceeds the value of the critical energy Ec.20 1. N/Nmax Figure 7.40 0.80 Ec 0. The higher is the specific energy absorption value. These regions characterise the rate of damage or failure of the tube which starts from a rapid initiation of failure (low cycle) to the region of endurance whereby the effect of the impact event is almost insignificant.40 0.00 0. it can be inferred that the crush zone fracture mechanisms are influenced by the tube dimensions and these fracture mechanisms determine the overall energy absorption capability of the composite tubes. This curve provided information on the incident energy and number of impact relationship indicating their maximum values when the tube apparently will start to fail. polyester. one can notice that its intercept (at Nf = 1) is mainly dependent on the value of Ec.

An increase in matrix failure strain causes greater energy absorption capabilities in brittle reinforcement such as glass fibres (Jacob et al. glass/epoxy composite tubes absorbed more energy than the glass/polyester tubes. thus provides higher Ec value. the value of the incident energy that falls in the repeated impact curve of the circular tube is much higher compared to the other two sections. Therefore it is expected that tubes made of epoxy matrix has better impact damage resistance due to its higher Ec value than glass/polyester tubes. Moreover. In characterising the energy absorption capability of an FRP composite material. respectively. tube having circular cross section exhibited better energy absorption performance compared to comparable square and rectangular tubes. polyester. As a result. and epoxy with glass fibre reinforcement are the main interest of this study. The density of the glass fibres in the composite tubes has a lot to do with their energy absorption characteristics. respectively to that of the former. 2002). This indicates that the former has better damage tolerance under repeated impact loading than the latter. apparently we can observe that the location of the curve of circular tube in the graph is above relative to the curves of tubes with square and rectangular sections.Chapter 7 – Application of the model to other types of composite tubes EJ Guades geometries considered in this study. In the contrary. the specific energy of the FRP composite tubes increased from lower to a higher value. A much lower specific energy value of the square and rectangular sections is generally credited to the presence of the corners acting as a stress concentrators leading to the formation of splitting cracks. The effect of the fibre content on the energy absorption (or Ec value) of FRP composite tubes is not straightforward since the increase of fibre might not always necessarily Behaviour of glass FRP composite tubes under repeated impact for piling application 177 . If we plot the repeated impact curves of these tubes. tubes reinforced with fibres having higher failure strain result in greater energy absorption. On the other hand. The specific absorption values of the latter can be reduced as much as 20% and 50%. As the density of the fibre is reduced from a higher to lower value. glass/vinyl ester tubes have higher Ec value due to their better energy absorption performance than glass/polyester tubes. In a similar way. one can ascertain that the higher inter-laminar fracture toughness would increase its energy absorption response. changes in matrix stiffness have very little effect on the energy absorption capability of composite materials with ductile fibre reinforcement. Tubes made of vinyl ester. Glass/vinyl ester exhibited much specific absorption energy value compared to glass/polyester tube due to its better inter-laminar strength and higher failure strain..

Relative to the other types of fibre reinforcements (e.e. or rectangular) and the type of matrix materials (vinyl ester. polyester. or epoxy). they also offer better durability characteristic. the change in slope will be faster since in this region the failure of the tube is dominated by the number of impact rather than the incident energy as expected. 20% of Ec) which implies that at this energy level the impact damage is not imminent. As a result.g. The reduction rate. It should be noted that as the fibre fraction increases. the volume of the matrix between the fibres decreases. The deviation becomes apparent only at relatively lower incident energy (i.. is not constant since Ein-Nf relationship found to be non-linear (i. Carbon.e. The reduction contributed to the formation of the inter-laminar cracks at relatively lower impact load resulting in a reduction in the energy absorption capability.. 2002). At a relatively higher incident energy (i. Environmentally. however.e. polyester. Generally the slope of the damage evolution curve in the region where failure of the tube is not observed (N < Nf and Ein < Ec) decreases when the applied incident energy is reduced.e. Similarly. This curve demonstrates the quantitative damage provided by the impact event on the tube from the start of loading up to its failure state.. Kevlar). The damage evolution curve provides an idea on the degree of damage in the non-collapsed region unlike the repeated impact curve. Aside from providing an excellent structural performance. In general. 80% to 60% of Ec). power function). the inter-laminar strength of the composite material will be reduced (Jacob et al. the common FRP composite materials used in this application is a thermosetting matrix (i. the deviation of the reduction rate within this range is minimal regardless of the geometrical sections (circular. glass fibres are the choice of reinforcement due to their relatively minimal cost while still meeting the structural Behaviour of glass FRP composite tubes under repeated impact for piling application 178 . and epoxy) with glass fibre reinforcement.Chapter 7 – Application of the model to other types of composite tubes EJ Guades improve the absorption behaviour. Their application offered an alternative solution for traditional pile materials especially in harsh environmental condition.. these materials refute the need for further chemical treatment or protection due to their inherent anti-corrosive property.7 Discussion on the application of FRP composite tubes in piling system and the practical implication of the results obtained from the present study The application of glass FRP composite tubes in piling system has been reported in the present study. square. 7.. vinyl ester. Although preference as to the suitable matrix materials may varies for each design requirement.

The characterisation of the response of these materials under repeated axial impact loading ascertains their performance when they are driven and especially during the encounter of hard soils or boulders. This allows whether the impact-driven FRP tubes still serve their purpose in supporting design loads or whether their structural integrity is compromised. Generally. Although FRP composite tubes with a relatively smaller section were considered in the experimental investigation (i. The present study considered this “worst scenario” during the conduct of the impact tests on FRP composite tubes.4 (Chapter 2). Issue such as their driving performance. it is considered suitable to characterise the impact behaviour of a full-scale hollow FRP pipe piles used in piling application. however. This investigation also provided insights on the effects of impact stresses on the postimpact bearing capacity of the composite tubes which is vitally important as a supporting structure in bridge construction.e. As highlighted in Section 4. It was emphasised that one of the factors affecting their driving performance is the impact behaviour of the fibre composite materials and was the focus of this study. 100 mm square section). This was also supported by some results on the progressive collapse behaviour whereby the length does not affect the progressive crushing behaviour of the composite tubes. the rupture of the FRP composite materials during impact driving happened when they encounter hard soil or boulders (no penetration).g.. Tubes made of thermosetting matrix with glass fibre reinforcement were the subject of interests in the investigation.. a 375 mm length is reasonable to characterise the impact behaviour of FRP composite tubes.Chapter 7 – Application of the model to other types of composite tubes EJ Guades requirements of the design. The application of the FRP composite tubes in piling system has been the subject of some recent studies. their structural properties have been sufficiently investigated. This study presented the behaviour of FRP composite tubes subjected by repeated axial impact loading. As reported in Section 2. a relatively shorter specimen (375 mm length) was used in this study in characterising the impact behaviour of FRP composite tubes. The damage was observed to be imminent at the top of the pile (end crushing) with not much more on mid-height collapse (buckling failure). . As a result. failure modes) between FRP composite tubes with smaller and bigger geometrical sections are similar.1. Behaviour of glass FRP composite tubes under repeated impact for piling application 179 . Likewise. the damage behaviours (e. needs to be carefully investigated since it affects their optimum use for widespread application.

one can realise their actual response during impact driving when used in piling system. Nf curve of repeatedly impacted tubes” shown in Figure 4. FRP composite tubes having circular.e. The degradation of residual properties caused by the impact event is imminent.Chapter 7 – Application of the model to other types of composite tubes EJ Guades There are results obtained from the present study which have practical significance in driving hollow FRP pipe piles.. Figure 4. the value of the incident energy at a corresponding Nf will be lower than that of a “penetrating” tube due to its relatively lower impact tolerance. in the case of the driven tube) as long as it is impacted up to failure.10). square and rectangular geometries are the typical cross sections used in piling application although the first one is considered the most efficient as it exhibits better energy absorbing performance. By understanding the impact behaviour of FRP composite materials of the tubes. however. the repeated impact (fatigue) curve of mostly impacted composite materials whether it is laminate or tube axially or transversely impacted follows a power function (i. One thing that might be different from these support conditions is that for tubes impacted with fixed support. This assumption. For instance. On the other hand. The present study. Additionally. however. square or rectangular sections have the advantage over the other section as they can be easily connected to the other structural components in the system.10 would be similar when it will be driven and penetrating into the soil until the occurrence of damage. This means that the curve of the former will be located below the curve of the latter when they are plotted at same Nf values. This result might be true whether the tube is impacted at different end support (fixed or penetrating. In general. the shape of the curve “Incident energy vs. A systematic information on the impact behaviour on these tubes leads to Behaviour of glass FRP composite tubes under repeated impact for piling application 180 . It is clear that the impact damage provided significant effect on the performance of the FRP composite materials during the impact event. needs to be verified in actual test and is considered as potential research work in the future. It should be noted that Nf illustrates damage into the impacted composite materials and therefore the tube penetrating into the soil should be continuously driven until the occurrence of damage or failure. is performed on a small-scale specimen and therefore a residual properties testing on a full length pile might be beneficial to provide additional information on the effect of axial impact loading. the understanding on the residual properties of the driven (impacted) composite tubes provided a reference on their structural carrying capacity.

the range of energies describing these regions also increases. tubes of different cross sections are the interests of this study. The critical energy values Ec of the composite tubes were derived using dynamic or quasi-static compressive tests. Similarly. The result showed that the energies describing the low cycle. It was found that the change of slope of the damage evolution curve of glass/vinyl ester composite tubes between Ec and 60% of Ec is comparably small. Behaviour of glass FRP composite tubes under repeated impact for piling application 181 . On the other hand. The application included tubes made from glass/vinyl ester. glass/polyester. The information on the impact behaviour on these tubes leads to an efficient and improved driving performance for a wide acceptance of their application. From this study. 7. circular tubes have greater Ec values of comparable square and rectangular tubes. When using the latter. and endurance fatigue regions of the composite tubes are mainly dependent on their critical energy. high cycle.8 Conclusions The application of the damage evolution model to other composite tubes of different sizes.Chapter 7 – Application of the model to other types of composite tubes EJ Guades an efficient and improved driving performance for a wide acceptance of their application. the rate of failure in the endurance fatigue region was very slow indicating that the effect on the variation of incident energy in this region is minimal. the yield of this study will help in developing efficient techniques and guidelines in driving composites piles. The rate of reduction on the failure of the tubes in the low cycle fatigue region is comparably faster than the other two regions. Similarly. and different types of matrix material was discussed. cross sections. However. the critical energy used in the model was calculated using a correlation factor. This result was also observed in tubes made of glass/polyester and glass/epoxy composite materials. the reduction is apparent when they are impacted until failure by at least 20% of Ec. an understanding of the behaviour of glass fibre FRP composite tubes under repeated axial impact can be obtained. and glass/epoxy composites. Moreover. By carefully selecting suitable parameters. the model was able to demonstrate the damage evolution curves of the composite tubes. As the critical energy increase. The repeated impact curves (or Ec) of tubes made from glass/epoxy is higher compared to the other matrix materials.

the failure mode of square composite tube repeatedly impacted was characterised by progressive crushing at the upper end.  There is no significant difference exists on the shapes of the load and energy curves for non-collapsed tubes. which is the main focus of this study. Particularly. composite tubes) are presented in this chapter. This failure was manifested by inter and intra laminar cracking and glass fibre ruptures. The conclusions gathered from the various studies conducted towards understanding the behaviour of the fibre composite materials (i. The conclusions related to this study are summarised below:  In general. several challenges needed to overcome for hollow FRP composite piles for their optimum use especially on their installation.Chapter 8 – Conclusions EJ Guades Chapter 8 Conclusions 8. This includes lack of information on the behaviour of the fibre composite materials under impact loading.2. this failure shows bunches of lamina splaying into the outside and inside of the tube.1. Moreover. The load and energy responses of the non- Behaviour of glass FRP composite tubes under repeated impact for piling application 182 . however. and timber resulted in deep foundation industry to look for alternative materials suitable in harsh and corrosive environment application. Hollow FRP composite piles provided significant advantages in terms of cost efficiency and structural capabilities.. steel. This work studied the impact behaviour of FRP composite tubes.1 Summary The deterioration of traditional pile materials such as concrete. Additional research studies are suggested to facilitate their acceptance in piling application. 8. There are.e. the effects of impact loading on the instantaneous and post-impact structural performances of the FRP material were investigated.2 Main conclusions from the study 8. Behaviour of composite tubes subjected by impact loading This study has experimentally investigated the behaviour of a square composite tube subjected by repeated axial impact.

 The maximum reductions of residual compressive and flexural strengths are 6.  The peak load evolution of collapsed tubes constitutes two regions.Chapter 8 – Conclusions EJ Guades collapsed tubes are comparably similar to that of the collapsed tubes before the initiation of failure occurs. Effects of impact loading on the residual properties of composite tubes The residual properties of square composite tubes under repeated axial impact was characterised using experimental investigation. The peak load in the pre-collapse region is in decreasing trend. 8. their effects gradually decrease at relatively higher energies. are almost negligible in the residual modulus property.8% and 10%. Based on the results of this investigation. and the mass of the impactor significantly influenced the residual strength degradation of the impacted tubes.2.  The drop mass and impact velocity (or drop height) have pronounced effects on the collapse of tubes at lower incident energies. the maximum reduction in Behaviour of glass FRP composite tubes under repeated impact for piling application 183 .  The rate of residual strength degradation between increasing impact number becomes rapid when impact energy increases. however. the number of impacts becomes the key reason as soon as the value of incident energy decreases.  The number of impacts played an important role on the peak load evolution in the pre-collapse region. number of impacts.2. Their effects. respectively. its effect becomes less significant in the post-collapse region. however.  The effects of incident energy and number of impacts were found to be significant on the rate of energy absorption in the pre-collapse region only and not in the post-collapse region. the following conclusions are drawn:  The levels of impact energy.  The decrease of residual strength values is more substantial when the composite tubes collapsed. whilst they become constant after the initiation of failure. On the other hand. however. however.  Incident energy is the major damage factor in the collapse of tubes for lower number of impacts. The rate of load degradation is more rapid when the tube was impacted by higher incident energy.

 A good agreement was observed between the experimental results and the calculated values using the proposed damage model. Their difference is less than 10%. tubes of circular sections have higher Ec values of comparable square and rectangular tubes.2.3%. Prediction on the damage evolution of composite tubes A prediction model was established in describing the damage evolution of the composite tubes.Chapter 8 – Conclusions EJ Guades the tensile strength is roughly 0. It was also applied to composite tubes with different cross sections. The model was verified through experimental investigation on a 100 mm square pultruded tube. Behaviour of glass FRP composite tubes under repeated impact for piling application 184 . thus.  The effect of impact damage on the reduction of residual compressive strength of the tube is concentrated only in region closer to where the source of impact originates. the repeated impact curves (or Ec) of tubes made from glass/epoxy is comparably higher than the tubes made from glass/vinyl ester and glass/polyester composites. thus can be neglected. The reduction in tensile strength is within the standard deviation of the baseline value. the model can be used in predicting the damage evolution of square composite tubes under repeated impact loading. The model was applied to composite tubes made from vinyl ester/polyester/epoxy matrix reinforced with glass fibres. The higher the Ec values. Moreover. and endurance fatigue regions of the composite tubes are largely dependent on their corresponding Ec. The variation of incident energies Ein between the fitting curve and experimental data points for a 100 mm square specimen loaded up to failure is less than 3%.  The repeated impact (fatigue) curve (Ein-Nf) of the composite tubes subjected by axial impact loading followed a power function relationship. high cycle. The conclusions of the theoretical prediction on the damage behaviour of composite tubes are summarised as follows:  The critical energy Ec obtained from quasi-static compressive test can be used in determining (Ec)Dynamic by carefully selecting a suitable value of the correlation factor β.3. 8.  The energies describing the low cycle. the higher the range of energies characterising these regions.  In general.

 Experimental investigation on the repeated axial impact behaviour of composite tubes made of glass/polyester and glass/epoxy.  A more rigorous FE analysis is needed using sophisticated FE software package to accurately predict the impact behaviour of composite tubes.3 Recommendations for future study The following areas need to be studied in more detail for a widespread acceptance of composite tubes in piling application and to improve their driving performance:  Systematic evaluation of the behaviour of glass FRP composite tubes under actual pile driving considering the effect of soil. Moreover. failure mode.  Finally. This analysis will provide other important information such as the simulated failure mode caused by the impact event.) associated to the depth of penetration into the ground.. This will deliver as an alternative to a more-expensive experimental study in determining the impact behaviour of composite tubes. etc. continuous research and development are essential to develop the market and increase the confidence in using the composite tubes in piling application.  The present study investigated the residual properties of composite tubes using coupon tests. It is also worth to investigate the residual properties of driven hollow FRP piles using full scale specimen. the experiment should include other cross sections aside from square since various geometries are considered in piling application. The development of local and international standards on their installation technique will encourage their adoption worldwide. This will provide a more realistic and reliable output on the effect of impact stress on their load bearing capacities. The information that will be obtained from this test will double check the results in the damage evolution curve using the prediction model.Chapter 8 – Conclusions EJ Guades 8. Behaviour of glass FRP composite tubes under repeated impact for piling application 185 . load evolution curve. This study will characterise the impact behaviour (i.e. The output of this study will provide additional information as to the variation of impact stresses along the height of the tube and its residual properties. The yield will provide design engineers information in choosing suitable safety factors in installing these piles using impact driving. This information is considered important in determining the effect of soil profiles on the failure mode of the impacted tube.

USA. Ambur D. 39th AIAA/ASME/ASCE/AHS/ASC Structures. Philadelphia. Proceedings of the Composites in Civil Engineering. Abdewi E. 394-54. ASTM International. Rechak S. and Limam A. Effect of curvature on the impact damage characteristics and residual strength of composite plates. (2011).References EJ Guades References Abdallah E. Bouvet C. 30(2). Rome.M. Badaruddin A.M.H. and Manalo. Experimental and numerical investigation of static and dynamic axial crushing of circular aluminium tubes. Australian standard on boat and ship design and construction. 42(8). (1964). Benmedakhene S. Mateos M. (2007). 23(10).. Driveability of glass FRP composite piling. Laksimi A. 216-221. Ouali N...... Effect of geometry on the crushing behaviour of laminated corrugated composite tubes. (2008).. and Boukarohba T. (2007). 308(3-5). Journal of Sounds and Vibrations. Rivallant S. Hamouda A. 58-60. (2012). and Mahdi E. Pasadena. ASTM D 695 (2010). T. Barrau J. Journal of Materials Processing Technology. and Pluvinage G. Australian Standards 132. Ouroua Y. Sulaiman S. Polymer Testing. A. Behaviour of glass FRP composite tubes under repeated impact for piling application 186 . Badruddin I.J. Italy. (2004). Aljibori H. Standard test method for high strain dynamic testing of deep foundations. and Chong W. Aurrekoetxea J... Mesbah A. California..3 (1993). International Journal of Fatigue. Azouaoui K. 3(1). and Buchin J.F. Paper no 13-210 Arenz R.K.A. T.. Philadelphia. USA. California Institute of Technology. (1998). California.S. Doctor of Philosophy dissertation.. 1103-1137. Azari Z. 13-15 April. and Aretxabaleta L.J. Standard test method for compressive properties of rigid plastics.. (2006). 877-885. 1925. ASTM International.. Thin Walled Structures. Experimental study of composite structures in automotive applications.. Azouaoui K. Part 3: Fibre-reinforced plastics construction.. Theoretical and experimental studies of wave propagation in viscoelastic materials. AI4IA First International Conference. and Starnes J. ASTM D 4945 (2008).A.. 20-23 April.. Aravinthan. Damage characterisation of glass/polyester composite plates subjected to low-energy impact fatigue. Modelling of damage and failure of glass/epoxy composite plates subject to impact fatigue.S.. Repeated low energy impact behaviour of self-reinforced polypropylene composites. 5(1). USA. Sarrionandia M.A. and Jakrapiyanun W. Ashford S. The influence of composite tubes damage tolerance on compressive residual strength: experimental observations. (2001). Journal of Composites for Construction. 504-513. 47-54.R. Al Galib D. Structural Dynamics and Materials Conference.. International Journal of Mechanical and Materials Engineering. Long Beach. 172(3).. Field applications and case studies of FRP in civil infrastructures: The Australian experience. (2001).

Machida... Belingardi G.C. Repeated impact response and hand lay-up and vacuum infusion thick glass reinforced laminates.J. The Journal: Roads and Bridges. (2011).. C. Boscato G. 35(7). Belingardi G. Field study of composite piles in the marine environment. (2009)..F. Proceedings of the 4th International Conference on FRP Composites in Civil Engineering. and Paolino D. and Russo S. (1998)... 40(12).. Riskalla. USA. Composite Science and Technology.. Taylor and Francis Group.S. A. 61(1-2). Cavatorta M.D. On the axisymmetric progressive crushing of circular tubes under axial compression.H. On vibration of pultruded FRP sheet piles. 27-38. Marinucci. Barbero.536153). A constant-life model of fatigue behaviour for Carbon-fibre composites: The effect of impact damage. Pasquiet J. Coban. Plastics: Determination of tensile properties. 6(2). (1984).. Edinburgh.E. R. T.References EJ Guades Bakis. and Paolino D.. 61. Robertson R. E. (2002). Fibre-reinforced polymer composites for construction: State-of-the-art review.J. M.J. Influence of the laminate thickness in low velocity impact behaviour of composite material plate. and Vadori R. Introduction to composite materials design. (2008).. H. Cosenza. 3137-3155.. Oxford. UK. The 17th International Conference on Composites (ICCM17). Effect of speed on progressive crushing of epoxy-glass cloth tubes. Residual performance of scarf patch-repaired pultruded shapes initially impact damaged.L. Composites poised to make inroads as highway structural materials.J. (2008). Brown. 9-18. Lesko.E.O. and Harris B. Curgul I. Paper no. April 1994.. (1994). (1991). 53(3). 73-87. 31-46. BS EN ISO 527-2 (1996). Single and repeated impact tests in fibre composite laminates: Damage index vs. J. Ballinger. 30(3):628-634. Zurich. 609-619. Composite Structures.H. Composite Structures.C. A.. and Thornton P. 317-331. J. A. 40(1). V.C... Comparison of bevel and tulip triggered pultruded tubes for energy absorption. and Morgan. (2003). Behaviour of glass FRP composite tubes under repeated impact for piling application 187 . (2005).. E. 40-44. O. (URITC No. C. (2009). and Hull D. T. P. International Journal of Solids and Structures.. Cavatorta M. Bradshaw.H. Berry J. V. Bora. L. Davalos. Materials and Design.. S. Switzerland. C.. Sinmazcelik. Composite Science and Technology. International Journal of Impact Engineering. F10:8..D. and Benzeggagh M.P. Behesty M. Belingardi G. Czaplicki M. and Triantafillou. (2003). Baxter. Fracture morphology and deformation characteristics of repeatedly impacted thermoplastic matrix composites. Yun. 27-31 July.J... and Kyriakides S.S. The 3rd International Conference on the Mechanical Properties and High Rates of Strain... 463. Journal of Composites for Construction. 58(1). (2001). 22-24 July. Chotard T. University of Rhode Island Transportation Center. Bank.. and Gunay.L. Bardi F. residual flexural properties.

Composite Structures. (2005).Z. (2001). 88(2). Einde. Behaviour of glass FRP composite tubes under repeated impact for piling application 188 . (1998). 98(4). Pile driveability analysis. Krishna V. Use of FRP composites in civil structural applications.E.. 39(10). 737-745. New York. Flexural behaviour of concrete-filled fibrereinforced polymer circular tubes. (1996).Z. and Hirsch. (2002). Sari M. Confinement model for axially loaded concrete confined by circular fibre-reinforced polymer tubes. Journal of Geotechnical and Geo-environmental Engineering.C.R. and Davies P.. 28(7). Composite Structures. (2003). (1999). (1986). Davies G. Gning P.. L. Ice action on sheet piles made of composite material. Failure mechanisms of composite specimens subjected to compression after impact. S (1998). and Reis L.. 159-163. and Poursartip A. and Rao R. and Seible.R.O.M. and Rizkalla SH.J.Prediction of damage in composite cylinders after impact. Army Cold Regions Research and Engineering Laboratory (CRREL). Journal of Composite Materials. Effect of laminate thickness on the composite strength to repeated low energy impacts.. Freitas M. Composites Part A: Applied Science and Manufacturing. Dutta P.. Single and multiple impact behaviour of a CFRP laminate.M.. (1995).. Van Nostrand Reinhold Company.D. 23(3). (2005). Deniz M.W.. (2009).V. 365-373.B. ACI Structural Journal. Fenske. 17(6-7).S. 1147-1156. Collombet F. Hitchings D. Fam A. 27(12). (2012). 451-461.M.. Journal of Reinforced Plastics and Composites. Ghasemnejad H.. and Icten B. and d’Almeida J. Low velocity impact damage tolerance studies on glass-epoxy laminates – effects of material process and test parameters. 32(1). Zhao. 70(2):223-228. Delfosse D. Blackman B.D. 42(4). L. Frost J. Impact damage and residual strengths of woven fabric glass/polyester laminates. Karakuzu R. Found M. (1997).. and Rizkalla SH.A. 46(6).. and Sudall B. (2004).. Energy-based approach to impact damage in CFRP laminates. Composites Part A: Applied Science and Manufacturing.. Composite Structures. Journal of Composites for Construction. 327345. T.K. W. 389-403. Tarfaoui M.. Draft Technical Report (U. C. Journal of Composite Materials. F. 917-928.. and Howard I. De Morais.A. Composite Structures. Planning and Design of Fixed Offshore Platforms. Hadavinia H.S. Fam A.On the residual compressive strength of the glass-epoxy tubes subjected to transverse impact loading. 633-640. and Devinder. 647-655. 123-132. Construction and Building Materials.. Behaviour of interfaces between fibre-reinforced polymers and sands. 6(2). Monteiro S.N.References EJ Guades Datta D. and Han J. 253-261.A. Experimental studies on fracture characterisation and energy absorption of GFRP composite box structures. and Zhou G. 125(8).

.. International Journal for Numerical and Analytical Methods in Geomechanics. Composite Structures.E.R.. University of Southampton. Han H. Latin American Journal of Solids and Structures. Ho K. (2006). and Stavig M. and Payen F. Load-deflection response of transversely isotropic piles under lateral load.D.L. 8(1):87-95. Mufti A. Atlanta. Journal of Reinforced Plastics and Composites. Hsu. X.J.R. (2006). TU Delft. Han J. Han J. 683-691. (1997). Geo Congress 2006: Geotechnical Engineering in the Information Technology Age. Behaviour of glass FRP composite tubes under repeated impact for piling application 189 . N. A numerical study on the axial crushing response of hybrid pultruded and ±450 braided tubes. International Conference on Advance Polymer Composites in Construction (ACIC).K. L (2002). 1221-1232. April 15-17. Heyder F. Hosur. (1999). International Journal of Reinforced Plastics and Composites. Helmi K. Effects of driving forces and bending fatigue on structural performance of a novel concrete-filled fibre reinforced polymer tube flexural pile. 28-29 June. Karim.. Hwang J. S.. 253-264. Pickett A. and Paulu F (2012). and Lu Y. (2004). 290318. A laminated glass wall will protect Warnemunde from high water. 61(1-2):89-102. Experimental investigations on the response of stitched/unstitched woven S2-glass/SC15 epoxy composites under single and repeated low velocity impact loading. Hussein MH. 33(6). 509-529. American Journal of Physics. and Jones. M. 1(3):277-296. Buckling of vertically loaded fibre-reinforced polymer piles.. International Journal of Crashworthiness. 514-521.. Composite Structures. Pegg N. Characterisation of E-glass polyester woven fabric composite laminates and tubes. (2004). Woerner WA. 18(4).. Impact fatigue of short glass fibre reinforced polycarbonate. (2000). 24(5). Greve L. Composites Part B: Engineering. 39(7-8). G.. Canadian Journal of Civil Engineering. Pile driveability and bearing capacity in high-rebound soils. (1995).References EJ Guades Guess T. Georgia. 72(4). and Hall M.. M. and Lu. Sharp M. Taheri F.V. Fam A. S.S.R. Hanc J. and Jeelani. and Hwang C. Netherlands. Axisymmetric progressive crushing of circular tubes. Huang.. 80(2). (2001). (2003). The development and the future of advanced polymer composites in the civil infrastructures. 16(10). and Frost JD. (2003). Reedy E. Experimental testing and phenomenological modelling of the fragmentation process of braided carbon/epoxy composite tubes under axial oblique impact. Conference on Architectural and Structural Applications of Glass. and Frost JD.C. 903-925. and Doong J... (2007). From conservation of energy to the principle of least action: A story line. and Taylor E. UK.. SANDIA REPORT: SAND95-2352 UC-704. Holloway. Dynamic axial crushing of aluminium alloy 6063-T6 circular tubes.

.H. and Kwon T. 25-28 September. 33(3). United Kingdom. Effects of temperature on impact damages in CFRP composite laminates. Crushing behaviour of pultruded composites.. (1995). Geotechnical News. Ismail A. GA. 107-118. M.. (1995). Behaviour of glass FRP composite tubes under repeated impact for piling application 190 . United Kingdom. Fibre-reinforced plastic composites: Determination of flexural properties. Kim. 4th European Conference on Composite Materials.M. D.E. E and FN Spon. Composites Part B: Engineering.S. Iskander.P. The 17th International Conference on Composites (ICCM17). (2002). Journal of Composites for Construction. ISO 1172 (1996). Journal of Composites for Construction. (2000).S. Structural impact. Atlanta. prepegs. Jones N.Y.G. and Yang I. Georgia Technology. Kim S.J. 32(8). (1992). Energy absorption characterisation of composite tubes for railway application. and Arora J.C. (1999).” MSc Thesis. and Starbuck J. and Hassan M. Yoon H. (2002). M.. Iskander M. State of the practice review in FRP composite piling. and Stachula.H.S. Jurnal Mekanikal. 36(7). (2001).S. Fellers J.S.H. Kim J. 669-682. 116-120. Energy absorption of composites as an aspect of aircraft structural crash resistance. Kindervater C.. (1990). Edinburgh. and Jang B. 6(2).K. UK. 24. moulding compounds and laminates: Determination of the textile-glass and mineral-filler contentCalcination methods. 2(3). Quasi-static analysis of structural impact damage.. Energy absorption in polymer composites for automotive crashworthiness. Development of simplified dynamic models using optimization: Application to crushed tubes.L. 813-850.. and Ahmad A. (1995). Iskander M. Journal of Composite Materials. (2007). (2002). Kim C. M. Stuttgart. 192(16-18). Germany. Jacob G. 2729. 44(2). 27-31 July.F.G. Cha C. Jang B. ISO 527-2 (1996).G. Composites Science and Technology. Jones N. A. ISO 14125 (1998). Impact behaviour and impact-fatigue testing of polymer composites.. 15-31. (2009).M.. Cambridge University Press. and Stachula A. Textile-glass-reinforced plastics. 151-177. Plastics: Determination of tensile properties. Composites structures for civil and architectural engineering. Characterization of geotechnical surfaces via stylus profilometry. Simunovic S.Z. Journal of Constructional Steel Research. Kowbel W. Computational Methods in Applied Mechanical Engineering. Lee H. (2003).G. 2073-2097... Johnson.References EJ Guades Im K. Wave equation analyses of fibre-reinforced polymer composite piling. FRP Composite Polymer Piling: An alternative to timber piling for waterfront applications. 8896.

(2002).P. Reinforced plastics prevail in the waterfront. 77-90.E.B.Xpress?. and Wimbrow W. Army Corps of Engineers (CERL). (1996). Vaughan J. Tampa.Xpress User Guide. and Soutis C... Impact fatigue of PPS/glass composites.. Reinforced Plastics. Demosthenous G..E. Demosthenous G. 46(6). 1996. 27-29 January.G.G. 34(1).A... (1997). Axial collapse of thin-walled fibreglass composite tubular components at elevated strain rates. Bushel J. Review: Crashworthy capability of composite material structures. 12-13. Composite Design and Application (CDA-Summer issue)... 28(4). Manolakos D.Xpress (2012). and Papapostolou D. Examination of mechanical test methods for pultruded sheet piling. Ioannidis M. 38(16). Design and development of FRP composite piling systems. 213-225.B. 1221-1224.References EJ Guades Kinsey A.. and sheet piling systems... (1995).. New York: Dredging and Port Construction. Leuven. Manolakos D..G.. and Odelio R. Lindsay K. (1994). Composites and Polycon Conference. (2006). load-bearing.A. (2004).E. 30-35. Liu D. Composite Structures. 4(6).. Ioannidis M. Lampo R.. Mamalis A. (1997b).B. and Ioannidis M. USACERL TR-98/123..G. Composites Part B: Engineering.. 407-420.A. and Odelio R. (1998) Development and demonstration of FRP composite fender. Lhymn C. 653-677.B. Nashville.G. Manolakos D.. Proceedings of the International Composites Expo. Mamalis A. Composite Structures. 74(2). Journal of Composite Materials. 37(2). LMS Test. and Ioannidis M. Barno D. Demosthenous G. Bushel J.. Saunders D. U. 1719 October. Composites Engineering..B. Mamalis A.G.E. (2005). 26(9).. and Papapostolou D. Lampo R.. 4(10). Composite Structures. Manolakos D. (1985). Post-impact compressive behaviour of low temperature curing woven CFRP laminates. Tennessee. 109-134.. Characterisation of impact properties and damage process of glass/epoxy composite laminates. Maher A.B... and Carruthers J. (1996). Ioannidis M. and Ioannidis M. (2007). Composite piling market heats up. Composites. Manolakos D. 661-667.A. Mamalis A.E.E.. The static and dynamic axial collapse of CFRP square tubes: Finite element modelling..S. 1425-1442.G. Dutta P. Demosthenous G. 69(4)..J. Florida USA. Nosker T. (1997a). On the response of thin-walled CFRP composite tubular components subjected to static and dynamic axial compressive loading: experimental. Manolakos D. Composite Structures. Belgium. How are time traces visualized in LMS Test. Journal of Materials Science Letters.E. LMS Test. Robinson M. Maher A. Behaviour of glass FRP composite tubes under repeated impact for piling application 191 . 439-451. Marsh G. Lackey E. Mamalis A. The static and dynamic axial collapse of fibreglass composite automotive frame rails. Mamalis A.P. The static and dynamic axial crumbling of thin-walled fibreglass composite square tubes.....

References EJ Guades Melo J... T.. Providence.. Flexural strength and interlaminar shear strength of stitched GRP laminates following repeated impacts. Degrieck J.. and Hoppe E. Gallagher J. (2000).D. Degrieck J.V. Florida. Ackeren J.D.J. and Shahawy.. Composite Structures. 29(6). The 9th International Bridge Management Conference. 87(3). Paepegem W.. Mirmiran A.L. Energy absorption characteristics of circular composites in quasi static and compression loading. Abrate S.. Shahawy M. (2003). Filz G.A.. Pando M. Palanivelu S. Composite Science and Technology. (2008). USA. Dove J. 127-135. 12-14 October. 263-268. Analysis and field tests on the performance of composite tubes under pile driving impact. 97(2). Interface shear tests on FRP composite piles.L. Kakogiannis D.. 27(1). QDMR (2006). Wastiels J. Silva A.. and Eswaraprasad G. ASCE International Deep Foundations Congress. (1997). Technical Note 54. Composites Part B: Engineering. 224234.A. Ochelski S. 637-645. Wastiels J.. Mirmiran A... Kakogiannis D. 215-224. Composites Part B: Engineering. Pando M. Rhode Island.. (2010). A laboratory and field study of composite piles for bridge substructures. (2006). FHWA-HRT-04-043... Proceedings of the IMPLAST Conference. Mouritz A. Composite Structures.. and Vantomme J. (2002). Residual torsional strength after impact of CFRP tubes. 622-628. 82(4). Neff.V.. and Vantomme J. 41(8)... Part I: Central delamination and triggering modelling. Orlando.V. 509-522. Federal Highway Publication.. Polymer Testing. Mirmiran A. Hemelrijck D. Muralikannan R. Palanivelu S. ACI Structural Journal. The effect of processing on the energy absorption capability of composite tubes.V. Large beamcolumn tests on concrete-filled composite tubes.. Experimental assessment of energy absorption capability of carbon-epoxy and glass-epoxy composites. 268-277.. Shao Y. and Goodwin AA. (1996). (2010). Polymer Testing. A new concrete-filled hollow FRP composite column.J... March 2006. Minak G. Transportation Research Board.. El Khoury C.. (2010). 57(5). 55(2).S.V. Lesko J. and Villena J. Ghelli D. Predicting the life cycle costs of structures based on accelerated corrosion testing: A framework for incorporating Reliability Concepts.. and Zucchelli A. Ealy C.. Ackeren JV. Fibre composite projects. 729-741. Parametric study of crushing parameters and failure patterns of pultruded composite tubes using cohesive elements and seam. and Gotowicki P.. Velmurugan R.. 29(2). Composite Structures. Florida.P.M. Behaviour of glass FRP composite tubes under repeated impact for piling application 192 .. (2002). and Hoppe E. Experimental study on the axial crushing of pultruded composite tubes. 14-16 February.J. Panciroli R....N. Filz G. (2010).E. and Shahawy M. Orlando.M. Hemelrijck D. (2009). and Naguib W. Paepegem W.

W. Bitech Publishers.J. Characterization of a pultruded FRP sheet pile for waterfront retaining structures... 25-30. Composites Part A: Applied Science and Manufacturing. (2001). Santuiste C. Composites Part B: Engineering. 3rd International Conference on the Application of Stress Wave Theory to Piles. 18(5).. 1487-1495. (2005)..References EJ Guades Rausche.. Effect of repeated impacts on the response of plain-woven hybrid composites. Liaw B. Schweitzer. NY. 41(5). Static energy absorption capacity of graphiteepoxy tubes. 35(19). (2005). 41(1).H. and Pavier M.O. Dekker. International Journal of Crashworthiness.E. (2001).R. and Raju B. H. Bulletin of Material Science..K. and Wisheart M.J. 27(12).. and Bose N. and Wickliffe L. (2010). Short G. Compression after impact of thin composite laminates. Fellenius. 72-77. Journal of Composites for Construction. (2004). 24(1). 21-34. 79-86... Composite Science and Technology. Sanchez-Saez S. and Barbero E.. F. 32(6). Composite Structures. Sanchez-Saez S. Canada. (1988). Shao Y.. (2006). Inai R. 44-55. Sakr M.. Guild F. G. (1996). (1995). 1747-1761. 70-88.. 9(3). Impact fatigue of glass fibre-vinyl ester resin composites. Richardson M. 274-283. 65(13). Composites Part A: Applied Science and Manufacturing. Rana A.K. Crushing of pultruded glass reinforced square tubes.H.. El Naggar M.W. Residual flexural strength after low-velocity impact in glass/polyester composite beams. Roy R. (2001). Sarkar B. Lisbon. Interface characteristics and laboratory constructability tests of novel fibre-reinforced polymer/concrete piles.. Stress Waves. Barbero E. Delale F. Schultz M. and Nehdi M... 33(11). and Hyer M.. and Navarro C. Corrosion resistance table. B.. P. 1911-1919. Schmuesser D. and Hussein. Behaviour of glass FRP composite tubes under repeated impact for piling application 193 . 1123-1131. 403-413. Saito H. Review of low-velocity impact properties of composite materials. Portugal.K.J. Sakr M. Santos A. and Nehdi M. (2010). M. 7(1). Naggar M. (2000). Journal of Materials in Civil Engineering.R. Sarkar B. 871-876. Zaera R. Editor. E. (2002). Canadian Geotechnical Journal. Journal of Composite Materials. 626-633.J. 92(1). Journal of Engineering Materials and Technology. (2008). Vancouver. 109.J.. and Hamada H.A. Impact energy absorption of continuous fibre composite tubes. 8th International Conference on the Application of Stress Wave Theory in Piles. Roy R. Post-impact compressive strength of curved GFRP laminates.R. Likins. Impact fatigue behaviour of carbon fibre-reinforced vinyl ester resin composites. Load transfer of fibre-reinforced polymer (FRP) composite tapered piles in dense sand. Sevkat E. Composites Part A: Applied Science and Manufacturing.W. and Bose N. Pile Integrity by High and Low Strain Impacts. New York.... (1987).

Impactor mass effects in glass/epoxy composites subjected to repeated drop tests. Construction and Building Materials. 43(5). 30(8-9).F.A. Toowoomba.. Modelling energy absorption with a damage mechanics-based composite material model. Journal of Composite Materials. (2010).W. 50(3).strand7. Song H. Specific energy absorption capacity of glass-polyester composite tubes under static compressive loading. 2183-2198.. carbon and Kevlar composites using repeated drop tests. S. Milovic L. 6237-6244. R. Low-velocity and residual tensile strength analysis of carbon fibre composite laminates. Sugihara S.(2011).S. Modelling shock problems in Strand7. (2009). The dynamic study of drive-in racks under horizontal impact load.com.M. and Kim Jr. and Rao R. (2002). Journal of Composite Materials.M.K. Vinh H.. Behaviour of glass FRP composite tubes under repeated impact for piling application 194 . International Journal of Crashworthiness. University of Southern Queensland. and Adams A. Australia.. (2009). and Gunay V. Stamenovic M. Behaviour of FRP-jacketed circular steel tubes and cylindrical shells under axial compression. (2002). 427-444. 1583-1599. Materials and Design.W. Sugun B. 31(1).W. Gupta N. Metalurgija. and Rao R.. and Pavlovic-Krstic (2011). (2006). Xiao X. Arici A. Du H.F. 118-125. Solaimurugan.. The effect of stitching on FRP cylindrical shells under axial compression. Off-axis crushing of GFRP tubes. Australia.. Wu L. Yang Y. Research Report R915.M. 14(5). Behaviour of fibre composite piles for timber pile rehabilitation. Composites Science and Technology. Teng. University of Sydney. Wang S. Journal of Materials Science. 62(15). 23(15).. (2004a). 41(19). Thornton P. The crush behaviour of pultruded tubes at high strain rates. and Elayaperumal A. Zrilic M.827-838. Putic M. 407-418.X. 197-200.S. International Journal of Impact Engineering. Journal of Composite Materials.. J.. Nakai A. 21.Z. Low–velocity impact characterization of glass. J. Journal of Plastics and Composites. Master of engineering dissertation.References EJ Guades Sinmazcelik T. Residual strength of a Carbon/epoxy composite material subjected to repeated impacts. 749-765. 923-938. 24(6). 22(8). 23(14). 36(18).. http://www. Queensland. Y. 1547-1560. Strand7 Web Notes (2012). Wyrick D.. Journal of Reinforced Plastics and Composites. and Hu. Sugun B.. Song H..G. and Du H. Velmurugan R. (2004b). and Ma L.A. and Zhao G.. 594-615.. Impact fatigue behaviour of unidirectional carbon fibre reinforced polyetherimide (PEI) composites. Sirimanna CS (2011). Journal of Composite Materials. Energy absorption capability of multi-axial warp-knitted FRP tubes. (1990). and Hamada H. 2065-2073.W.. (2007).H. (1998). Energy absorption behaviour of double-chamfer triggered glass/epoxy circular tubes. (2004).

.W. and Richardson M. 195-201. Behaviour of glass FRP composite tubes under repeated impact for piling application 195 . (2007). 81(2).References EJ Guades Zhang Z.Y.O. Low velocity impact damage evaluation and its effect on the residual flexural properties of pultruded GRP composites. Composite Structures.

63 31.59 4 22.2 Summary of results of fibre fraction test for CT2 Specimen no 1 Width (mm) 24.83 5.64 5.23 1.1 and A. and m3 is the final mass of the crucible plus residue after calcination.24 1.967 78.95 4 24.69 29.77 5.03 5.61 Length (mm) 31.943 75.944 76.07 0.1 Summary of results of fibre fraction test for CT1 Specimen no 1 Width (mm) 22.65 28. The specific mass ρ and glass content Mg were calculated using Equations A.29 Specific mass (kg/m3) 1.29 Thickness (mm) 5.14 1.932 75. ρ = m0 /vff (A.62 28.929 Glass content (%) 75.74 Average 24.17 1.03 2 24.30 1.65 5.2 show the detailed results of the fibre fraction test performed using ISO 1172 (1996).69 Thickness (mm) 5. A.02 6 0. elastic modulus. m2 is the initial mass of the dry crucible plus dried specimen. respectively.949 75.20 1.75 Average 22.Appendix A EJ Guades Appendix A – Summary of results of the coupon and full scale tests on CT1 and CT2 specimens The details of the result of the entire tests discussed in Chapter 3 are presented here.78 Table A.10 0. glass content.947 Glass content (%) 76.71 28.02 26 1.16 Specific mass (kg/m3) 1.1) Mg = (m3–m1)/(m2–m1) x100 (A.943 75.84 Standard deviation 0.63 5.88 Length (mm) 29.2.2) where m0 is the mass of the specimen.66 31.71 31. Table A.894 73.20 1. peak stress.95 5.52 32.21 Standard deviation 0.77 2 22. The equations used in determining the specific mass. and strain at peak are discussed.64 3 24. vff is the volume of the specimen.1 Fibre fraction test Tables A.24 0.11 5.18 1.20 0.16 Behaviour of glass FRP composite tubes under repeated impact for piling application 1 .934 76. m1 is the initial mass of the dry crucible.61 3 22.1 and A.

40 Peak stress (MPa) 482.0005 and ε2 = 0. A. The peak stress σpc.2 Compressive test on coupon specimen The summary of result of the coupon test under compression loading is displayed in Tables A.4. σ1 and σ2 are the stresses measured at the strain values ε1 = 0.72 5.07 114. Table A. respectively.38 455.90 Thickness (mm) 5. A is the average cross sectional area of the specimen.5) where Ppc is the peak compressive load.80 53.95 5.09 Peak stress (MPa) 432.92 Standard deviation 0.08 - Average 12.123 0.3 Summary of results of coupon compressive test for CT1 Specimen no 1 Width (mm) 12.3) Ecomp = (σ1–σ2)/(ε1– ε2) (A.12 444.567 0. elastic modulus Ecomp.75 5.07 Table A.55 47.39 5.11 0.99 2 12. and strain at peak εpc were calculated using Equations A.44 140.88 Standard deviation 0.Appendix A EJ Guades A.49 140.87 5.03 114.45 49.21 0.31 141.65 2.081 0.05 Behaviour of glass FRP composite tubes under repeated impact for piling application 2 .10 5.2 show the typical load-displacement curves of the specimen tested under compression loading.86 5.3 and A.85 Modulus (MPa) 51.26 5.531 0.45 140.690 0.16 450.0025.84 2 12.84 2.37 0.12 424.47 140.90 5.07 449.77 Modulus (MPa) 48.5.813 Strain at peak (%) 0.07 5.00 5. The compressive test was conducted following ASTM D 695 (2010).06 - Average 12.3.4) εpc = σpc/Ec x 100 (A.67 Length (mm) 140.96 - 4 12.20 - 4 12.83 - 5 12.12 0.13 21.4 Summary of results of coupon compressive test for CT2 Specimen no 1 Width (mm) 12.05 Length (mm) 113. respectively.20 114. Figures A.4 and A.27 441.19 438.551 Strain at peak (%) 0.85 3 12. σpc = Ppc/A (A.93 3 12.73 Thickness (mm) 5.87 - 5 12.14 51.29 459.19 114.612 0.15 441.1 to A.59 114.41 482.

Equations A.7.5 and A.0 0.6 0.8 1. and A.0 1. respectively. and strain at peak εpt.7) εpt = σpt/Et x 100 (A.8 were used to calculate.6 Displacement (mm) Figure A. A.2 0.6 Displacement (mm) (b) CT2 Figure A.2 1.4. elastic modulus Et.Appendix A EJ Guades 40 1 2 Load (kN) 30 3 4 5 20 10 0 0.8) Behaviour of glass FRP composite tubes under repeated impact for piling application 3 .8 1.4 1.4 0.0 0.4 0. the values of the peak stress σpt. σpt = Ppt/A (A.6. In the tables.3 Tensile test on coupon specimen The procedure defined in ISO 527-1 (1996) was adopted in performing the tensile test on coupon specimens.2 0.1 Compressive load-displacement relationship of coupon specimens (CT1) 40 1 2 Load (kN) 30 3 4 5 20 10 0 0.3 to A. The typical load-displacement relationship of the specimen subjected to tensile loading is displayed in Figures A.2 Compressive load-displacement relationship of coupon specimens (CT2) A.4 1.0 1.6 0.2 1. Tables A.6) Et = (σ1–σ2)/(ε1– ε2) (A.6 summarise the results of the coupon tensile test.

698 1.46 230.24 603.625 1.00 5.01 637.50 5.Appendix A EJ Guades where Ppt is the peak tensile load.01 100 1 80 Load (kN) 2 3 60 4 5 40 20 0 0 2 4 6 8 Displacement (mm) 10 12 14 Figure A.69 3 25.09 Peak stress Modulus (MPa) (MPa) 622.36 230.09 619.12 27.16 618. respectively. σ1 and σ2 are the stresses measured at the strain values ε1 = 0.50 5.00 5.55 0.98 - 5 25.20 0.50 5.26 612.50 5.0025.00 5.88 38.20 40.69 249.94 39.31 574.48 Standard deviation 0.58 0.234 1.56 Standard deviation 0.05 Length (mm) 251.44 Length (mm) 231.30 596.6 Summary of results of coupon tensile test for CT2 Specimen no 1 Width (mm) 25.16 251.48 - Average 25.34 651.5 Summary of results of coupon tensile test for CT1 Specimen no 1 Width (mm) 25.0005 and ε2 = 0.60 5.31 585.48 39.09 22. A is the average cross sectional area of the specimen.297 1.59 - 5 24.23 250.21 636.00 Thickness (mm) 5.53 - 4 25.65 229.00 Thickness (mm) 5.3 Tensile load-displacement relationship of coupon specimens (CT1) Behaviour of glass FRP composite tubes under repeated impact for piling application 4 .43 2 25.48 2 25.099 Strain at peak (%) 1.10 0.50 5.401 0. Table A.93 1.50 5.45 230.13 Table A.07 42.49 3 25.09 Peak stress Modulus (MPa) (MPa) 570.88 - 4 25.29 - Average 25.20 251.37 231.842 Strain at peak (%) 1.56 609 0.76 39.07 250.

8 illustrates the summary of results of the coupons subjected to flexural loading following ISO 14125 (1998). respectively.00 923 0.67 Average 14.0025.23 150. The values of the peak stress σpf and elastic modulus Ef were calculated using Equations A.038.32 5.68 35.25 1031.12) where ε1 and ε2 are the strains having values of 0.71 5.091 2.092 2.65 150.25 Peak stress (MPa) 1.44 3 14.0005 and 0.12) respectively.878 2.46 Modulus (MPa) 36.9 and A.65 5 14.84 5.051.4 Flexural test on coupon specimen Tables A.09 Behaviour of glass FRP composite tubes under repeated impact for piling application 5 . t and b are the thickness and width of the specimen.70 4 14.195 2.03 22.56 37. s1 (Equation A.4 Tensile load-displacement relationship of coupon specimens (CT2) A.858 Strain at peak (%) 2.Appendix A EJ Guades 100 1 80 Load (kN) 2 3 60 4 5 40 20 0 0 2 4 6 8 Displacement (mm) 10 12 14 Figure A.85 5. σpf = (3Ppf ls)/(2tb2) (A. s1= (ε1 ls 2)/(6t) (A. σ1 and σ2 are the stresses measured at the deflections.02 35.11) and s2 (Equation A.48 150. Table A.22 1.9) Ef= 500(σ1–σ2) (A.00 Thickness (mm) 5.23 1.16 1065.7 and A.22 1.43 150.57 2 14.11) s2= (ε2 ls 2)/(6t) (A.037.97 35.54 36.15 0. respectively.10). On the other hand.47 Length (mm) 151.000.60 151.10) where Ppf is the peak flexural load.61 Standard deviation 0.440 2.33 0.31 5.10 respectively.7 Summary of results of coupon flexural test for CT1 Specimen no 1 Width (mm) 14. ls is the span length (see Figure 3. the value of the strain at peak εpf was taken directly from the machine.

39 5.61 39.21 952.46 5.043.520 Strain at peak (%) 2.79 39.50 5.01 151.50 0.56 2 15.04 Peak stress Modulus (MPa) (MPa) 944.6 shows the curve that relates the load and the midspan deflection of the coupons tested under flexural loading.17 994.51 5.22 150.90 37.5 Flexural load-midspan deflection relationship of coupon specimens (CT1) 4 1 Load (kN) 3 2 3 4 2 5 1 0 0 2 4 Displacement (mm) 6 8 Figure A.134 2.5 to A.8 Summary of results of coupon flexural test for CT2 Specimen Width no (mm) 1 15.23 1.07 Figures A.61 5 15.12 1.18 975.021 0.00 37.44 38.07 46.51 3 15. 4 1 Load (kN) 3 2 3 4 2 5 1 0 0 2 4 6 8 Displacement (mm) Figure A.52 5.534 2.90 38.73 4 15.18 1.21 147.60 Standard deviation 0.055.195 2.28 Length (mm) 150.975 2.891 2.95 Thickness (mm) 5.Appendix A EJ Guades Table A.36 151.34 1.61 Average 15.22 150.6 Flexural load-midspan deflection relationship of coupon specimens (CT2) Behaviour of glass FRP composite tubes under repeated impact for piling application 6 .

69 Standard deviation 0.6.9 to A.8 to A.88 102.17 290.11.5 Compressive test on full scale specimen The summary of test results on full scale specimen under compression loading is shown in Tables A.9 Summary of results of full scale compressive test for CT 1 (L = 100 mm) Specimen no 1 Depth (mm) 100.13) Ecomp = (σ1–σ2)/(ε1– ε2) (A.51 5. In this study.39 100.366 0.68 2 100.0005 and ε2 = 0.15.40 100.74 38.24 10.18 294. elastic modulus Ecomp.98 102. respectively.Appendix A EJ Guades A. Figures A.70 3 100.31 5. The peak stress σpc.86 - - 4 100.21 283.08 8.15) where Ppc is the peak compressive load.14) εpc = σpc/Ec x 100 (A.29 268.10 displays the load-displacement curves of the full scale specimen subjected to compression loading.75 - - Average 100. σpc = Ppc/A (A. the cross sectional area of the tube was approximated as the area of a simplified square section neglecting the added/subtracted areas due to the chamfered corners of the tube (Figure A.65 100. t t d d b b Figure A.71 5.78 5.93 100.574 Strain at peak (%) 0. σ1 and σ2 are the stresses measured at the strain values ε1 = 0.95 0.604 0.25 284.25 0.0025.18 - - 5 100.7 Simplified cross section of the tube Table A.01 Behaviour of glass FRP composite tubes under repeated impact for piling application 7 . and strain at peak εpc were calculated using Equations A.65 101. Note that the simplified cross section was also used in the calculation of peak flexural stress of the composite tube presented in Section A.17 Modulus (MPa) 41.7).77 100.13. A is the average cross sectional area of the specimen.14 39.85 Width (mm) 100.60 Length (mm) 100.93 100.75 1.970 0.32 100.39 Thickness (mm) 5. A.39 Peak stress (MPa) 281. respectively.19 5.14 and A.

05 Standard deviation 0.59 100.41 39.51 99.22 253.62 0.0 0.30 - - 4 100.5 Displacement (mm) Figure A.83 5.311 Strain at peak (%) 0.49 98.06 0.45 255.26 288.49 100.97 Modulus (MPa) 41.50 Length (mm) 199.50 106.15 Peak stress (MPa) 265.73 2.50 100.0 2.08 Average 100.5 2.120 0.27 254.32 - - Average 100.22 270.23 2.92 37.13 0.81 200.46 Width (mm) 100.50 5. L=100 mm) Behaviour of glass FRP composite tubes under repeated impact for piling application 8 .00 Thickness (mm) 5.19 261.64 3 100.07 266.42 100.215 0.51 103.10 5.87 5.44 5.00 5.11 Summary of results of full scale compressive test for CT2 (L = 100 mm) Specimen no 1 Depth (mm) 100.18 4.01 600 1 500 Load (kN) 2 400 3 4 300 5 200 100 0 0.09 0.53 Length (mm) 102.22 289.54 5.8 Compressive load-displacement relationship of full scale specimens (CT1.0 1.10 Summary of results of full scale compressive test for CT1 (L = 200 mm) Specimen no 1 Depth (mm) 100.50 100.33 3 100.5 1.Appendix A EJ Guades Table A.50 102.75 200.68 199.65 Standard deviation 0.54 100.04 15.58 100.60 100.096 0.37 5.72 Width (mm) 100.41 Table A.65 2 100.31 Thickness (mm) 5.88 0.55 - - 5 100.06 Peak stress (MPa) 261.73 2 100.

0 Displacement (mm) 3.0 4. L=100 mm) A.18).16) σpf = (Ppf ac)/(I) (A.12 to A.9 Compressive load-displacement relationship of full scale specimens (CT1. L=200 mm) 600 1 500 2 Load (kN) 400 3 4 300 5 200 100 0 0. respectively. c is the neutral axis depth of the tube (equals d/2).0 0.17. the results of the full scale CT1 and CT2 specimens subjected to flexural loading.6 Flexural test on full scale specimen Tables A. I is the moment of inertia (Equation A. respectively.13 illustrate the specimen cross section lay-out Behaviour of glass FRP composite tubes under repeated impact for piling application 9 . Figures A.16 and A.14 summarise.5 2.0 1.10 Compressive load-displacement relationship of full scale specimens (CT2. ls is the span length (equals 1000 mm). σpf = (Ppf ls c)/(4I) (A.11 to A.5 1.0 Figure A.Appendix A EJ Guades 600 CT1 500 1 2 Load (kN) 400 3 300 200 100 0 0.0 1. a is the distance between one of the end supports to the location of the nearest applied load.18) where Ppf is the peak flexural load.17) I = (bd3 – jk3)/12 (A.5 Displacement (mm) Figure A.0 2.0 2. The values of the peak flexural stress σpf under 3 and 4-point loading were calculated using Equations A.

respectively.A. The typical flexural stress versus cross-head displacement curve of the full scale specimens tested under 3-point bending is shown in Figures A. j d/2 N. to aid in the calculation.Appendix A EJ Guades and the schematic plan of flexural test (3-point and 4-point loading).16.11 Specimen cross section lay-out P d 500mm 500mm ls = 1000mm l Figure A. k d t b Figure A.14 to A.15. On the other hand.12 Schematic plan of 3-point bending test P P 200mm d a = 500mm a = 500mm ls = 1200mm l Figure A.13 Schematic plan of 4-point bending test Behaviour of glass FRP composite tubes under repeated impact for piling application 10 . the comparison of the stress-strain curve between the middle top (compression) and bottom (tension) sides of the tube using 4-point bending test is illustrated in Figure A.

32 128.195.30 0.09 0.07 0.200.64 Standard deviation 0.00 Thickness (mm) 5.00 Thickness (mm) 5.31 128.18 169.63 Width (mm) 100.46 Length (mm) 1.91 Average 100.39 Length (mm) 1.90 Average 100.00 5.44 1.00 5.14 Summary of results of full scale flexural test (4-point loading) for CT1 Specimen no 1 Depth (mm) 100.27 Peak stress (MPa) 131.13 Summary of results of full scale flexural test (3-point loading) for CT2 Specimen no 1 Depth (mm) 100.00 5.94 0.46 1.60 100.29 Table A.00 5.72 3 100.17 Table A.07 0.06 3.14 Flexural stress-displacement relationship (3-point bending test) of CT1 Behaviour of glass FRP composite tubes under repeated impact for piling application 11 .27 127.203.35 Length (mm) 1.201.52 1.34 125.72 Width (mm) 100.500.35 2 100.51 1.13 143.33 5.60 1.52 100.16 3.99 3 100.00 5.59 1.55 100.199.00 5.90 3 100.04 Peak stress (MPa) 173.12 Summary of results of full scale flexural test (3-point loading) for CT1 Specimen no 1 Depth (mm) 100.502.18 135.500.99 2 100.200.199.14 Peak stress (MPa) 135.04 160 CT1 Stress (MPa) 120 1 2 3 80 40 0 0 5 10 15 20 25 Displacement (mm) Figure A.63 Standard deviation 0.58 Width (mm) 100.40 165.68 100.03 2.30 2 100.00 Thickness (mm) 5.67 5.50 100.00 Average 100.65 1.Appendix A EJ Guades Table A.00 5.75 Standard deviation 0.200.05 0.50 100.33 169.56 100.56 1.53 1.82 0.06 6.58 100.500.13 0.55 100.

15 Flexural stress-displacement relationship (3-point bending test) of CT2 50 40 Load (kN) 30 SP1 .16 Flexural stress-strain relationship (4-point bending test) of CT1 Behaviour of glass FRP composite tubes under repeated impact for piling application 12 .Bottom SP1 .Top SP3 .Top SP2 .Bottom SP3 .Appendix A EJ Guades 160 CT2 Stress (MPa) 120 1 2 3 80 40 0 0 5 10 15 20 25 Displacement (mm) Figure A.Top 20 10 0 -2000 -1000 0 1000 2000 3000 Strain (micro) 4000 5000 6000 Figure A.Bottom SP2 .

51 100.50 5.47 Length (mm) 377. B.18 0.48 100.24 2 100.02 0.00 5.2 Dimension of specimen E480 Specimen no 1 Depth (mm) 100.00 Thickness (mm) 5.3 Dimension of specimen E420 Specimen no 1 Depth (mm) 100. snapshots of the apparatus used in observing the damage on the impacted tubes are also presented.53 375.1 Summary on the details of the specimens Table B.70 100.Appendix B EJ Guades Appendix B – Summary of specimen dimension and snapshots of the machine/apparatus used in the repeated impact test The details of the specimen and the machines discussed in Chapter 4 are presented here.85 375.53 Width (mm) 100.07 Behaviour of glass FRP composite tubes under repeated impact for piling application 1 .50 0.82 100.44 2 100.1 Dimension of specimen E630 Specimen no 1 Depth (mm) 100.04 0. Likewise.37 Standard deviation 0.42 Width (mm) 100.62 100.75 0.75 5.39 Length (mm) 374.19 Average 100.26 Standard deviation 0.02 1.43 374.34 Width (mm) 101.20 0.11 2 100.31 0.25 0.00 Thickness (mm) 5.50 Thickness (mm) 5.16 Length (mm) 374.45 375.37 375.43 374.52 100.02 Table B.30 Average 100.22 Standard deviation 0.75 5.41 Average 100.50 5.15 Table B.50 5.

55 Length (mm) 374.47 376.91 Average 100.70 100.65 Length (mm) 374.00 4.25 0.27 Average 100.32 Width (mm) 100.43 375.41 375.18 2 100.78 375.23 Standard deviation 0.52 100.42 100.27 Standard deviation 0.01 0.40 Length (mm) 376.67 100.24 Average 100.47 Length (mm) 376.75 0.50 5.42 2 100.19 Table B.39 100.68 100.68 Width (mm) 100.46 377.12 0.06 0.6 Dimension of specimen E160 Specimen no 1 Depth (mm) 100.44 100.90 376.01 0.75 5.90 Standard deviation 0.01 Table B.89 2 100.23 Standard deviation 0.10 0.71 Width (mm) 100.50 Thickness (mm) 4.Appendix B EJ Guades Table B.50 Thickness (mm) 5.56 100.94 Width (mm) 100.49 374.04 Average 100.50 Thickness (mm) 5.27 0.00 5.00 Thickness (mm) 5.5 Dimension of specimen E210 Specimen no 1 Depth (mm) 100.25 0.42 375.25 5.01 0.00 5.7 Dimension of specimen E630-1 Specimen no 1 Depth (mm) 100.13 0.03 Table B.30 2 100.05 Behaviour of glass FRP composite tubes under repeated impact for piling application 2 .50 0.00 5.4 Dimension of specimen E320 Specimen no 1 Depth (mm) 100.75 4.

44 Length (mm) 374.09 Standard deviation 0.29 Width (mm) 100.10 Dimension of specimen E420-1 Specimen no 1 Depth (mm) 100.40 376.00 5.05 0.05 2 100.14 2 100.46 100.8 Dimension of specimen E480-1 Specimen no 1 Depth (mm) 100.39 374.64 376.38 100.24 Standard deviation 0.04 Table B.1 Repeated impact testing set-up data logger and fixtures Behaviour of glass FRP composite tubes under repeated impact for piling application 3 .40 2 100.50 Thickness (mm) 5.01 0.43 Length (mm) 374.19 Table B.04 Average 100.Appendix B EJ Guades Table B.32 Average 100.50 376.31 100.32 100.25 5.75 5.05 B.9 Dimension of specimen E480-2 Specimen no 1 Depth (mm) 100.42 375.42 Average 100.14 0.50 Thickness (mm) 5.25 5.57 100.00 5.20 0.02 0.59 100.54 Width (mm) 100.08 0.36 Standard deviation 0.2 Repeated impact testing set-up and specimen snapshots (a) LMS data logging machine (b) Mounted accelerometer (c) Mass impactor Figure B.02 0.25 0.50 Thickness (mm) 5.34 375.54 Width (mm) 100.36 Length (mm) 376.07 0.00 5.

and E420-2 specimens B.2 Condition of the specimen after impact test (Test matrix from Table 4. E480-3.4 MOTIC® SMZ 168 Series stereo zoom microscope Behaviour of glass FRP composite tubes under repeated impact for piling application 4 .3 Apparatus used in the micro observation of damage Figure B.3).2) Figure B. Note: these exclude E630-2.3 Condition of the specimen after impact test (Test matrix from Table 4.Appendix B EJ Guades Figure B.

the data obtained from the mid-height can be used as a valid representation of the impact response of FRP composite tubes. Specifically. Nevertheless. The objective of this analytical modelling is to provide information on the relationship of the accelerations measured in the mid-height and at the top portion of the tube. The variation of the stresses on square FRP pultruded tubes under repeated axial impact using FE analysis is also presented.2. Particularly. the dynamic response of the FRP tube can be analysed one-dimensionally similar to that of the analysis used in elastic wave Behaviour of glass FRP composite tubes under repeated impact for piling application 1 . the analysis of the dynamic behaviour of the FRP composite tube with the present testing set-up can be treated as one of the conventional wave propagation problems in solids. Consequently.e. The results presented in Section 4. The recorded data at the mid-height of the tube by the shock sensor is an acceleration of the wave propagated from the source (impact load) coming into that point. As a result. at the head of the tube). it is reported that the value of the calculated energy at the mid-height is closed to the applied (incident) energy indicating that the amplitude of the recorded acceleration history will be likely similar when the sensor was placed relatively nearer to the impact point (i. to use acceleration at the mid-height point as the response of the impactor) was partly discussed. the acceleration history data was post processed to get the energy history curves needed for further analysis. Moreover.3 on the accuracy of using this assumption (i.1 Analytical study on the variation of acceleration data with the height of the tube This study used the acceleration recorded by the shock sensor placed at the midheight of the tube to represent its impact response.e. the peak axial stress degradation of the tubes impacted by different incident energies and number of impacts are discussed. the author performed a simple analytical modelling study explaining the accuracy of the assumption to use the data obtained at the mid-height of the tube characterising its impact behaviour. For simplicity. The stress variations along the longitudinal and transverse directions are emphasised.. C.Appendix C EJ Guades Appendix C – Variation of acceleration data and impact stress with the height of the tube using analytical modelling and finite element (FE) analysis The comparison between the acceleration data at the mid-height and at the top most portion of the FRP tube using analytical study is presented here.

It was restrained at the bottom by a rigid base and subjected by an impact mass m0 dropped at a height h0 (velocity at the contact point is v0).1) ET = m a s + 0. and a mass density ρt and the model used in the analysis.Appendix C EJ Guades propagation in a rod.1or C. Evidently.. L/2 . Figure C.2) Behaviour of glass FRP composite tubes under repeated impact for piling application 2 . a1.5 m v2 (C.1 shows the schematic view of the impacted FRP tube of length L. v2 sL/2 ET(L/2). . an elastic modulus E. mth Ein s1 ET1. This principle states that the total amount of energy in an isolated system remains constant over time or location (http://hyperphysics. 2008).1 Schematic view of the impacted tube and the idealised model The widely-accepted principle of conservation of energy was used in the formulation of the model. a2. . vL/2 sm ETm. ρt .gsu.2 (Hanc and Taylor.1 m0 vm0 0 s 1 2 L A. 2004). v1 s2 ET2. vm Figure C.phy-astr. ). An example of a relationship obtained from the principle of conservation of energy is shown in Equations C. E. . am.edu). This was adopted as this provides a complete description of the motion of a particle (propagation of waves) and provides relationship of the acceleration (or velocity) of the traveling wave at different location along the tube. The tube was modelled as a series of connected element with corresponding properties as shown in Figure C. constant cross section A. ET = EK + EP (C. the wave propagation analysis used for rod is also applicable to the FRP tube. aL/2. It should be noted that this analysis was also used to investigate the wave propagation in piles including hollow steel tubes due to impact loading (Santos.

. a.3) { } (C. the mass and wave velocity relationships at different location along the tube can be established (Equations C. On the other hand. we can obtain the total energy relationship at different location along the tube as shown in Equation C. 1st.4 and C.4. This principle was applied to an FRP composite tube shown in Figure C. ET is defined mathematically using Equations 2. 2nd. respectively. m. and travelled distance of the propagating wave. (C. Using Equation 1 or 2.Appendix C EJ Guades where the symbol ET represents the total energy and the symbols EK and EP represent the kinetic and potential energy. Alternatively. respectively. v.g. 3rd position) from the top of the tube.4) { } Behaviour of glass FRP composite tubes under repeated impact for piling application 3 . the subscript L/2 and m denote location at the mid-height and at mth distance (e. and s represents the theoretical mass. respectively. From the figure and from dynamics. acceleration. In the equations. velocity.5).1.

it was observed that the input wavelength due to the impact loading on FRP pultruded tube presented in Chapter 4 is higher than its width as shown in Figure 4. ET(L/2)) and at the top (ET1) is equivalent.5 mL/2 vL/22 = m1 a1 s1 + 0.6 results in Equation C.e.9 using the conditions on s and Δv and considered the equation of motion of the wave propagation from the top to the mid-height of the tube.5 m1 v12 ) ( v1 –ΔvL/2)2 = (C... s1 = s2= s3 =sL/2 = sm). mL/2 = m1 + ∑ = m0 + ∑ (C. Δv1 ≈ Δv2 ≈ΔvL/2 ≈ 0).e.e. mL/2 aL/2 sL/2 + 0.9.9) If we divide the tube into equal elements.7 and C.. This implies that at this condition the change in wave velocity from one point to another is approximately zero (i. In the case of the present study. The total energy at the mid-height (i. Arenz (1964) reported that a constant wave propagation speed is expected if the input wavelength is either small or large compared to the thickness of a rod (or tube). This relationship is defined mathematically in Equation C.2 (wavelength >100 mm).3). It should be noted that these equations are derived from Equations C. then we are supposing a uniform distance travelled by the propagating wave from one point to the other (i.10 shows the simplified version of Equation C.5.5) { } The relationship of the acceleration at the mid-height and at the top (approximately at point 1) can be established using the principle of conservation of energy (i.6.5(m0 + ∑ m1 a1 s1 + 0.8) Substituting mL/2 and vL/2 to Equation C. Equation C.4 and C. Behaviour of glass FRP composite tubes under repeated impact for piling application 4 .7) vL/2 = v1 + ΔvL/2 (C.5 m1 v12 (C. (m0 + ∑ ) aL/2sL/2 + 0.e..6) Equations C. Equation C.Appendix C EJ Guades (C.8 show the relationships of the mass and wave velocity at point 1 and at the mid-height of the tube.

2010).2 kg aL/2 /al 0..7%. It can be noticed that the difference of their respective values decreases with increasing impact mass. then s1 is equal to L/100.5 and can be defined as the one-dimensional wave propagation speed of the tube (Schwarz et al.e.60 0.20 1. Figure C. Assuming that the length of the tube is divided into 100 uniform slices. The difference between aL/2 and a1 values is 4.e. aL/2 ≈ [(m0 +Δ m1) s1] / [(m0 +∑ ) s1] a1 (C.5 (E/ρt) [(m0 +Δ m1) – (m0 ) s1] (C. the weight of the tube in the wave propagation analysis (impact) can be neglected as its weight is relatively small compared to the weight of the impacting mass. Since the main interest of the analytical study is to determine the relationship of aL/2 and a1.2 displays the comparison of the aL/2 and a1 values at three different m0 values (i. the weight of the tube (i.11..12. ∑ ) is considered and is equal to (ρt A L/2). however.00 0.5(m0 + ∑ ) (E/ρt) = (m0 +Δ m1) a1 s1 + 0. 1.973 16. Solving the value of aL/2 in term of a1 from Equation C.11) Equation C. In most cases.6 kg and 25.5 (m0 +Δ m1) (E/ρt) (C.2 kg 21. Therefore the relationship of aL/2 and a1 can be approximated using Equation C.80 0. 3.1%. The difference is less than 5% pointing out that the acceleration taken at the mid-height point is valid to be used in the analysis.969 0.2 kg.40 0.2 kg. In the present study. 16.2 Comparison of aL/2 and a1 values at varying impact mass Behaviour of glass FRP composite tubes under repeated impact for piling application 5 .20 0.6 kg 25. aL/2 = [(m0 +Δ m1) s1] / [(m0 +∑ +∑ )] / [(m0 +∑ ) s1] a1 + 0.00 Figure C.10 yields Equation C.. and 2. we can neglect the intercept of the line and consider only its slope.11 can be considered as one of the general equations of a line.10) where v1 is equal to (E/ρt)0.12) The relationship between aL/2 and a1 can now be obtained using Equation C. 21.Appendix C EJ Guades (m0 + ∑ ) aL/2s1 + 0. respectively.1%.959 0.12.

the steel cap is not included in modelling resulting to a more simplified model. with meshes of 4.928 nodes and 4. This was evidenced by the damage mode of the tube whereby fronds were developed on its upper portion after the test. It should be noted that the value of the shear modulus G12 in the table is calculated approximately using Equation C. Table C.4.1) where E11 is the elastic modulus at the longitudinal direction. the maximum amplitude of loading is applied to the tube. The laminate was modelled as a stack of several plies as shown in Figure C. the use of the steel cap is to help in evenly distributing the load from the impactor and did not necessarily restrain the translation of the tube at its axes. The solver is run using a small step to capture the response.840 plate elements.54 x 6. C..1.2. massive concrete) indicating that the displacement along the longitudinal direction is restrained. In the conducted impact test.1 displays the ply properties adopted in modelling the laminate. In this modelling. A nonlinear transient solver was used to investigate the dynamic behaviour of the tube subjected to repeated axial impact. The nonlinear transient solver can calculate the full time history of the response of the structure to impact (Strand7. This indicates that the upper end of the tube is free to translate or rotate along the three axes.82 mm (sides) and 1. As emphasised in Chapter 4.5 E11 / (1+ υ12) (C. υ12 is the Poisson ratio. Figure C. Some of the studies adopting nonlinear transient solver using Strand7 in characterising the behaviour of the structure or material under impact loading are those conducted by Vinh and Kim (2011) and Heyder and Paulu (2012).e. Behaviour of glass FRP composite tubes under repeated impact for piling application 6 . the lower end of the tube was in contact with the rigid surface (i.7 x 6. the upper end of the tube was held by a steel cap. the mesh model comprised of 4. As a result. laminate properties were adopted as property attributes of plate elements. The time history of this load is defined by the input of a load versus time table.3 shows the material model of the 100 mm square pultruded tube with a wall thickness of 5. 2012).1 Material model and support restraints In this study. This table is linked to the appropriate load case in the transient solver panel.25 mm and a length of 375 mm.82 mm (corners). G12 = 0.2 Finite element modelling on the variation of impact stress with the height of the tube The investigation was carried out using the Strand7 finite element analysis software.Appendix C EJ Guades C. In this case. On the other hand.

Figure C. the supports at the two transverse directions at the lower end of the tube were held unrestrained to allow movement whilst all the supports on its upper end remained unrestrained (see Figure C.900 b MPa Poisson ratio υ12 0.35c - Shear modulus G12 14.3).3 Material modelling of the composite tube Figure C.970a Unit kg/m3 Thickness t 0.5.1 Behaviour of glass FRP composite tubes under repeated impact for piling application 7 .4 Lamina lay-up arrangement used in FE model Table C. WCFT Product specification. steel cap held with a spring which was attached to the steel frame). To achieve such support conditions..Appendix C EJ Guades The test set-up (i.531d MPa a b c d Table 3. was considered by properly selecting the maximum amplitude of loading applied on the tube. from extensometer.e. from Equation C.5833 mm Elastic modulus (longitudinal direction) E11 39.1 Material properties of the tube wall laminate ply Material property Density Symbol ρ Property value 1.234a MPa Elastic modulus (transverse direction) E22 12. The details of the applied load are presented in the next section.

51J. the equivalent static load case was determined at the first impact. respectively.77 J.01 J. and E420.2 kHz) used in the experimental study presented in Section 4. total steps of 1500). First. 476.03 sec (i.01 sec. there is a need to determine the equivalent static load for the three incident energies. This indicates that the impact energy during the first impact can be used as the applied energy at increasing number of impacts. Only one impulse period was used in this analysis since it was observed that the occurrence of the peak load in the first impact obtained from the experimental study is similar for the three incident energies (Figure 4. These energies correspond to the combinations of drop masses and heights presented in Table 4.01 second.2 Applied static load case Three incident energies were used in the analysis.5. These level of energies was chosen as it can characterise both the collapsed (ruptured) and non-collapsed conditions of the impacted tubes.2 (Chapter 4). The time step used in the analysis is 0.2.5. The rectangular shape was used since this minimises the total impulse period resulting to a relatively shorter time during running of the solver. the occurrence of peak load is around 0.2. In the repeated impact testing. this static load case was then combined with the factor versus time table in Strand7 to provide an instantaneous impulse force simulating series of unit impact loading. Behaviour of glass FRP composite tubes under repeated impact for piling application 8 . E480..02 sec assuming that the load varies from 0 and increase linearly up to the maximum value (triangular).Appendix C EJ Guades C. C.e.e. see Figure C. and 423.2 (Chapter 4) for specimens E630. namely: 634.3) was applied on the tube depending on the incident energy considered. It should be noted that the time step used in the analysis is equivalent to the sampling rate (i. Second. In the modelling. however. From Figure 4. 51.00002 sec and the response was obtained for a period of 0.1 Determination of static load case at the first impact An iterative process was used in obtaining the static load case equivalent to the incident energy at the first impact..5 shows the typical factor versus time table for the impulse period of 0. An edge load (pressure. the applied energy remains constant (same drop mass and height) all throughout until the failure of the composite tubes. Chapter 4).2. rectangular impulse load was used and therefore the period was chosen as 0. Figure C. Imperatively. Two steps were undertaken to determine the equivalent static load needed in simulating repeated impact loading.2.

6 displays the variation of the static load case with the measured acceleration (longitudinal direction) at the mid-height of the tube using FE analysis.10 0. and E420 is 0.2. respectively. steel capping) are also considered in the FE analysis.50 0.3832 MPa. It is worth noting that the data recorded by the accelerometer resemble the results in conjunction with the current set-up used in the impact testing. time table for the impulse period of 0. Therefore it is imperative that by using this data in finding the appropriate static load case means that the effect of test set-up (i..2748 MPa.01 second Selection of the appropriate static load case was established by comparing the nodal acceleration at the mid-height obtained from the FE analysis to that of the recorded acceleration by the shock sensor in the impact experiment. E480.60 Edge load (MPa) 0. 0.Appendix C EJ Guades Figure C.3521 MPa.6 Variation of the static load case with the measured acceleration Behaviour of glass FRP composite tubes under repeated impact for piling application 9 . 0.40 0.e. the appropriate static load case for E630. The summary of applied static load cases used in FE analysis is shown in Table C. Figure C. From this table. and 0.5 Factor vs.30 0.20 0.00 0 100 200 300 400 500 Acceleration (m/s2) Figure C.

19 E420 0.2748 193. respectively. Figure C. time table simulating repeated impact loading (E630) Figure C. and 120 impact repetitions for E630.9 show the factor versus time table used in combining with the load case. 57. The time interval between successive impacts is chosen as 0. respectively.2. the tube was assumed to be subjected by 45. and 95 impacts.65 Specimen ID C. In the analysis.2 Determination of static load case due to repeated loading The pressure load from Table C.7 to C.8 Factor vs.71 E480 0.3521 248. This number of impact is sufficient to simulate the effect of repeated impact loading since it was found experimentally that ruptures of the tube occurred at 20.3832 Acceleration (m/s2) FE Analysis 270. E480. time table simulating repeated impact loading (E480) Behaviour of glass FRP composite tubes under repeated impact for piling application 10 .7 Factor vs.87 193. 70.2 was then linked to factor versus time table to simulate repeated impact loading.2.34 Acceleration (m/s2) Experiment 270.2 Summary of applied static load cases used in FE analysis E630 Edge load (MPa) 0.Appendix C EJ Guades Table C. Figures C.1 sec.40 249. and E420.

time table simulating repeated impact loading (E420) C. 2010). Linking of the table and the modulus property of the tube was achieved by using the command of Property_Plate_Tables_Time_Modulus vs. In the modelling. In Figure 4. it showed that the peak load (strength/stress or stiffness) of the composite tube under repeated impact loading initially decreased up to failure and become constant upon reaching the post-collapse region. however.9 Factor vs. time. E480.2. Consequently. or separation of the component being designed. The results obtained from Chapter 5 on residual properties of composite tubes revealed that the maximum degradation of modulus for collapsed or crushed tubes is approximately 5%. loss of stiffness. It should be noted that the geometric and mechanical properties of the tubes used in the experiment presented in Chapters 4 and 5 are almost similar. “Failure” being defined as some predetermined crack length. and E420. 57.3 Repeated impact (fatigue) model for FE analysis The stress-life method was used in modelling the strength degradation of the composite tubes subjected by repeated impact loading. This technique used the principle of similitude in determining the number of cycles to failure (Bishop and Sherratt. A factor versus time table was made and combined with the modulus property of the tube. Figures C. In Strand7.Appendix C EJ Guades Figure C. respectively).6 (Chapter 4). The reduction of the strength in the first region can be assumed to be uniformly decreasing with increasing number of impacts until failure (20. this 5% modulus degradation was adopted in the FE analysis for the three incident energies.10 to C.12 demonstrates Behaviour of glass FRP composite tubes under repeated impact for piling application 11 . the degradation was associated to the modulus property of the plate elements of the modelled composite tubes. Therefore the data obtained from the test on each tube is also suitable to be used in the analysis for either of them. the stiffness degradation of the tube was considered by linking it to its material properties. 95 impacts for E630.

time table simulating material degradation (E420) Behaviour of glass FRP composite tubes under repeated impact for piling application 12 . time table simulating material degradation (E480) Figure C.10 Factor vs.12 Factor vs. time table simulating material degradation (E630) Figure C. Figure C.11 Factor vs.Appendix C EJ Guades the factor versus time table used in combination with the property of the composite tube.

1 Variation of peak axial stress along the longitudinal direction Figure C. E480 and E420. However. The stress distribution found to be higher at the extreme top portion and decreases as the location moves away from the top.e.02 -0.001 s 0.000.04 0.14 Time (s) Figure C.06 0. 375 mm from the bottom of the tube). The total steps used for E630. the time step adopted found to be still suitable as it did not produce a significant difference in capturing the peak load compared to a much lower time steps as illustrated in Figure C.005 sec. respectively.0005 s 0.1). This value is relatively higher to that used for single impact (Section 2. and 15 sec.0002 s 1. 2.13 Comparison of time steps for E630 C. 2. It should be noted that the value in the abscissa was obtained by dividing the peak axial stress at each location by the value at the extreme top edge (i. 10 sec. It can be observed that the difference of peak stress between the extreme ends of the tube (i.12 0.3.50 0.50 0.50 0.. and 3.00 0.1 0.50 Axial Load (N/mm) 2.2.Appendix C EJ Guades C.. respectively.13.005 s 1. E480. This was selected to save the time in running the solver. it follows that the trend of the peak stress distribution is same for specimens corresponding to E630.3 Finite element analysis results and discussion C.000. These means that the response of the tube was obtained for a period of 5. It can be observed from the figure that the peak axial stress varies with its location regardless of the level of incident energy and impact repetitions.000.00 0.e. In investigating the effect of impact repetitions.4 Time step and total steps used in the analysis for repeated loading The time step used in the nonlinear transient dynamic analysis for repeated loading is 0.00 0. 0 and 375 mm Behaviour of glass FRP composite tubes under repeated impact for piling application 13 .00002 s 0.08 0.1 sec.14 shows the variation of the normalised peak axial stress in longitudinal direction with its location on the tube for the three incident energies and different number of impacts. and E420 are 1.2.

375 mm from the bottom of the tube) of the tube.996 0.998 1 Peak axial stress (normalised value) c) E480 1st impact 10 300 20 30 200 40 50 60 100 70 80 94 0 0. This is the main reason why the physical damage on the tube after repeated impact test showed that the damage is only concentrated on its upper portion. respectively. however. This trend.14 Variation of peak axial stress in longitudinal direction C.992 1 400 20 200 0.992 0. 57th.994 0. the values of the ordinate were obtained by normalising the peak axial stress at each location by the value at the corner of the tube.Appendix C EJ Guades from the bottom of the tube) increases with increasing number of impacts. It should be noted that the values of the peak axial stress are the values at the top extreme edge (i. Behaviour of glass FRP composite tubes under repeated impact for piling application 14 . Furthermore.994 0. 375 Distance from the bottom (mm) 400 300 200 100 1st impact 5 300 10 15 200 20 100 0 0.998 1 Peak axial stress (normalised value) d) E420 Figure C. and 95th impact.15.992 0. is valid only before the start of rupturing of the tube as the peak stress difference between the extreme ends was reduced after the 20th. the impact stress at the upper portion remains constant whilst the lower portion is getting a decreasing impact stress value..998 b) E630 400 300 0.996 0.e.3. This shows that at increasing number of impacts.2 Variation of peak axial stress along the transverse direction The peak axial stress variation with the location along the transverse direction of the impacted tubes for the three simulated incident energies is displayed in Figure C.994 0 mm a) Distance reference Distance from the bottom (mm) Distance from the bottom (mm) 1st impact 10 30 40 50 100 0 0.996 Peak axial stress (normalised value) 57 0.

respectively.Appendix C EJ Guades Figure C. This implies that the effect of impact repetitions on the stress variation at the middle section (top extreme edge) of the tube is significant for relatively lower incident energies. the normalised peak stress value at the 1st impact is 0. the higher the variations between the values at the middle and at the corner of the tube.99990 and 0. 0 mm distance) decreases with increasing number of impacts.5 times the specific energy absorption of comparable circular specimens. E480. It is worth noting that the stress variation at the middle of the tube between the first impact and the impact number to initiate collapse becomes less when the incident energy increases. In fact. however.15b).e. The more is the impact repetitions. On the other hand. For instance.99984. Mamalis et al. Figure C. For the three incident energies. this phenomenon was emphasised by Mamalis et al. This indicates that stress concentration is likely to happen for the square composite tube when it is subjected by repeated impact loading. This indicates that the effect of impact repetitions plays a significant role in the variation of peak axial stress. at E630 (Figure C. (1997b) as the main reason why square section tubes are generally less effective at absorption energy than circular ones. however. and E420.99999 whilst 0. found to be less influential on the variation of peak stress at the corner of the tube as their values are similar for all corresponding number of impacts.99996 for the 20th impact. This can be evidenced by the figure whereby the peak stress intersects at one point located at the corner of the tube. increases when the location is approaching to its corner. The peak axial stress induced on the tube is relatively lower at the middle.15 also shows that the variation of the peak axial stress with the location is more pronounced for higher number of impacts.. Their effects. the peak axial stress at the middle (i.15 shows that the peak axial stress varies its distance from the middle node of the tube for E630. reported that the square section tubes have 0. the normalised stressed value at 57th (for E480) and 95th impact (for E420) is 0. Behaviour of glass FRP composite tubes under repeated impact for piling application 15 .

99986 1st impact 30 57 0.16 that the affected distance from the top of the tube increases when the number of impact increases.99986 1st impact 40 80 0.99998 0. The values in the x-axis were obtained by dividing the degradation at each location by the value at the extreme top edge. and E420.99994 0.3 Variation of peak axial strength degradation with the number of impact Figure C.99990 0. at E480.99990 1st impact 5 10 15 20 0. and 80th impacts for E630. It can also be observed from Figure C.99998 0.99978 0 10 10 50 95 20 20 60 30 30 70 40 50 Distance from the middle node (mm) d) E420 Figure C. However when the tube starts to fail or when it is impacted by 57 Behaviour of glass FRP composite tubes under repeated impact for piling application 16 .00002 0.99994 0.99982 0. respectively. This trend can be witnessed to up to the 15th.99982 0.99998 0.EJ Guades Peak axial stress (normalised) Appendix C Middle node 0 mm 25 50 1.16 that the strength degradation is much higher at the extreme top edge as compared to the bottom of the tube regardless of the level of incident energies and number of impacts. As an example.16 illustrates the peak strength degradation with its axial location on the tube for the three incident energies.99994 0.00002 Peak axial stress (normalised) Peak axial stress (normnalised) a) Distance reference 0.00002 0.99978 0 10 20 30 40 50 Distance from the middle node (mm) b) E630 1. It can be observed from Figure C.99982 10 40 20 50 0. the degradation is concentrated on the upper portion of the tube. 50th.15 Variation of peak axial stress in transverse direction C.99990 0.99986 0. The degradation of the tube is calculated as the difference of the initial peak strength (or stress) and the peak strength at corresponding number of impacts.3. the degradation extends up to the bottom of the tube when subjected by 50th repeated impact loading. When the tube started to fail. however.99978 0 10 20 30 40 Distance from the middle node (mm) c) E480 50 1. E480.

2 0. 20th. the affected distance from the top of the tube increases with decreasing incident energy.17 shows the comparison of the absolute peak axial strength degradation of the collapsed (failed) and non-collapsed tubes. E480.8 Peak axial stress degradation (normalised value) c) E480 1 10 20 30 40 50 60 70 80 0 0 1st impact 0 0. This indicates that after the occurrence of failure.2 a) Distance reference 0. A detailed discussion to this effect in terms of damage caused by the increase of incident energy and number of impacts is presented in Section C..8 1 b) E630 400 400 Distance from bottom (mm) Distance from the bottom (mm) 0.6 0.4 Peak axial stress degradation (normalised value) 300 200 100 1st impact 10 20 30 40 50 57 0 300 200 100 0. It should be noted that the two other simulated incident energies also follow the same trend. and E420.3.4 0. 375 Distance from the bottom (mm) 400 300 200 100 300 200 1st impact 5 100 10 15 20 0 0 0 mm 0. 57th.16 Variation of peak axial strength degradation with number of impacts C. and 95th impact for E630. the damage caused by the succeeding impacts is concentrated on the top of the tube only.6 95 0.2 0. At the start of collapse (i.5. One interesting result that can be obtained from Figure C.4 0.6 0. a distance of approximately 60 mm from its bottom found not to be affected.3. It should be noted that the number of Behaviour of glass FRP composite tubes under repeated impact for piling application 17 .4 Comparison between the absolute peak axial strength degradation of collapsed and non-collapsed tubes Figure C.8 1 Peak axial stress degradation (normalised value) d) E420 Figure C.e.Appendix C EJ Guades impact repetitions.16 is on the comparison of the distance affected by strength degradation. respectively).

00 5.50 Peak axial stress degradation (%) c) Non-collapsed tubes 5. To compare the magnitude of strength degradation of the impacted tubes.17b). on the other hand.20 Peak axial stress degradation (%) d) Average values Figure C.80 5.00 4.Appendix C EJ Guades impacts adopted to characterise the non-collapsed tubes are 15. it can be observed that the strength degradation is imminent at the upper portion with not much on the lower side of the tube.17 Absolute peak axial strength degradation at failure Behaviour of glass FRP composite tubes under repeated impact for piling application 18 .00 3.20 Peak axial stress degradation (%) a) Distance reference b) Collapsed tubes 300 200 E630 at 15th impact 100 E480 at 52th impact E420 at 90th impact 0 3. the degradation varies from 5% at the extreme top and 4.13b and C. and 90 for E630.20 4. They are selected since these numbers of impacts are relatively near on the occurrence of failure and therefore can give a fairly higher axial strength degradation values.60 4. Figure C. For collapsed tubes.00 4. 52.60 4.40 4.20 4.17d illustrates that the strength degradation of the non-collapsed tubes ranges from 4. Figure C. respectively.3% at its extreme top and 3.00 5.40 4.80 5. These numbers of impacts are 5 impacts prior to the occurrence of failure for the three incident energies. E480 and E420.7% at the bottom.17c shows that strength of the entire axial length of the non-collapsed tubes found to degrade relative to the extreme top edge. the average value obtained from Figures C.6% at a distance of approximately 120 mm from the bottom of the tube. 375 Distance from the bottom (mm) 400 300 200 100 E630 at 20th impact E480 at 57th impact 300 E420 at 95th impact 200 100 0 0 mm 4. For collapsed tubes (Figure C.60 3.00 Distance from the bottom (mm) Distance from the bottom (mm) 400 Average (Non-collapsed tubes) 400 Average (Collapsed tubes) 300 200 100 0 3.50 4. On the other hand.13c was plotted and is shown in Figure 17d.80 4.

It should be noted that these lengths are specifically referenced from the corner since it was found out that the maximum damage occurred in this location as a result of stress concentration. Figure C.19 in which the maximum length of damage for the collapsed tubes occurred at its corner. The length of the damage on the tube was obtained by setting a limit in the element result display of the Strand7 output.19. In this modelling.18 Comparison of the damaged length at the start of failure Behaviour of glass FRP composite tubes under repeated impact for piling application 19 . This was substantiated by Figure C. the strain value at the extreme top edge of the tube during the start of collapse (20th. This result points out that the damage sustained in the area near the impact point of the composite tube is higher for a tube failed in several impacts with less energy per impact.18 shows that the damaged length of E630 is relatively lower compared to E480 and E420 at the start of collapse. used laminates and not composite tubes as adopted in the present investigation.Appendix C EJ Guades C.18 demonstrates the comparison of the simulated damaged length of the tube at the start of collapse for the three incident energies. and 95th impact for E630. Their study. respectively) was used as limit. This outcome was also reported by Wyrick and Adams (1998) whereby the specimens with holes resulting from a single perforating impact exhibited a much lesser damage compared to those perforations resulted from a number of impacts.5 Simulated damaged length at the corner of the tube Figure C. These lengths are measured from the corner of the tube as schematically shown in Figure C. and E420. however. The damaged length of E480 and E420 is approximately 2 and 3 times to that of E630. 57th. respectively.3. Damaged length from the top of the tube (mm) 12 E630 at 20th impact E480 at 57th impact E420 at 95th impact 8 4 0 Figure C. E480.

19 Damaged length simulation using FE analysis Behaviour of glass FRP composite tubes under repeated impact for piling application 20 .Appendix C EJ Guades Damaged length a) E630 Damaged length b) E480 Damaged length c) E420 Figure C.

As a result. the damaged length near the contact point is higher for a tube failed by a lesser incident energy with several impacts. On the contrary.3% for non-collapsed tubes and 4. Its effect. however. The degradation ranges from 3. Finite element analysis using Strand 7 software was carried out to investigate the variation of peak axial stress and strength degradation of the square composite tubes subjected by repeated impact loading. The strength degradation is higher at the extreme top of the tube as compared to the bottom edge. the degradation increases when the number of impact increases. Therefore it is recommended that a rigorous analysis is needed using sophisticated FE software package to accurately predict the impact behaviour of glass fibre-reinforced polymer (FRP) composite tubes. Apart from the effort of the author in attempting to model the repeated impact behaviour of FRP tube. The result shows that the difference of the acceleration values at the mid-height and at the top most portion of the tube is less than 5% indicating that the former can be used in the analysis. The number of impacts significantly affects the stress variation along the axial length of the composite tube. Stress concentration along the corner of the tube is likely to happen during repeated impact loading. Moreover. the damage during impact loading generally initiates along the corners leading to the formation of splitting cracks.7% to 4.6% to 5% for collapsed/failed tubes. the FE study presented here is still considered a preliminary work and as this does not provide other important information such as the simulated failure mode caused by the impact event. For tubes in which failure is not achieved within the impact duration. the degradation caused by the succeeding impacts is concentrated only on the top of the tube. Behaviour of glass FRP composite tubes under repeated impact for piling application 21 .4 Conclusions Analytical modelling was performed to verify the accuracy of using the acceleration at the mid-height point of the tube as its response.Appendix C EJ Guades C. when failure is achieved. found to be less influential on the stress variation along the corner of the tube. The result of the FE analysis showed that the applied peak axial stress on the tube is concentrated on the impact point and attenuates when the location moves away from this point.

40 375.1 Top portion (see the location on Figure 5.50 5.43 375.03 2 12.50 5.24 114.27 0.30 E630-10 100.68 100.24 5.15 Length (mm) 114.31 114.15 E480-40 100.38 Length (mm) 375.58 100.22 114.00 5.21 114.50 Thickness (mm) 5.1 Summary of the dimension of the tubes Tube ID E0-0 Depth (mm) 100.2 Summary of results of coupon compressive test D.68 3 12.60 100.05 21.45 Width (mm) 100.50 5.38 Standard deviation 0.00 5.60 5.2.43 374.24 E480-80 100.16 5.28 E480-10 100.24 430.00 5.16 0.07 0.96 4 12.28 Average 100.19 E160-80 100.24 114.15 5.30 Length (mm) 114.12 440.22 5.54 100.42 374.02 17.30 450.23 Peak stress (MPa) 444.69 100.12 5.18 431.29 100.08 1.42 100.12 419.16 398.03 3 12.85 5.38 Standard deviation 0.43 377.00 5.2 Coupon dimension and compressive test result for E160-80 (Top) 1 Width (mm) 12.06 D.14 0.1 Summary of the details of the tubes Table D.14 425.82 Specimen no Table D.22 Thickness (mm) 5.00 5. D.32 113.38 375.58 373.10 E320-80 100.27 E740-10 100.99 Average 12.51 100.25 E630-30 100.03 0.Appendix D EJ Guades Appendix D – Summary of specimen dimension and results in residual properties testing The details of the specimen and the results discussed in Chapter 5 are presented here.82 2 12.21 114.24 114.48 4 12.3 Coupon dimension and compressive test result for E320-80 (Top) 1 Width (mm) 12.79 5.23 Standard deviation 0.16 5.39 Specimen no Behaviour of glass FRP composite tubes under repeated impact for piling application 1 .18 0.30 376.27 395.55 376.32 100.83 Thickness (mm) 5.2) Table D.25 Average 12.18 Peak stress (MPa) 441.

169 Specimen no Behaviour of glass FRP composite tubes under repeated impact for piling application 2 .37 49.07 0.00 5.13 5.51 0.10 0.4 Coupon dimension and compressive test result for E480-10 (Top) 1 Width (mm) 12.71 5.07 434.45 46.149 Standard deviation 0.32 4 12.45 Length (mm) 112.03 0.51 5.23 457.20 114.2) Table D.02 - Average 12.33 - Average 12.11 Peak stress (MPa) 437.20 113.17 415.65 Specimen no Table D.39 48.64 5.49 113.85 1.76 5.379 2 12.95 5.22 436.03 16.29 Modulus (MPa) 53.42 5.28 5.6 22.04 5.29 Length (mm) 114.39 3 12.12 0.15 5.802 Standard deviation 0.36 398.05 440.55 5.25 1143.37 113.61 3.24 425.79 2 12.33 Thickness (mm) 5.61 5.27 0.04 5.64 Thickness (mm) 5.30 114.46 112.02 4 12.7 Coupon dimension and compressive test result for E320-80 (Middle) 1 Width (mm) 12.66 Average 12.20 5.24 Peak stress (MPa) 446.52 Length (mm) 113.90 Standard deviation 0.318 2 12.92 Thickness (mm) 5.34 114.2.27 114.88 50.226 3 12.32 452.44 - 4 12.01 425.Appendix D EJ Guades Table D.12 0.979 3 12.577 Specimen no Table D.74 - 4 12.29 114.29 114.14 Standard deviation 0.6 Coupon dimension and compressive test result for E160-80 (Middle) 1 Width (mm) 12.19 5.10 114.96 3 12.12 434.31 434.16 Length (mm) 113.47 114.36 Modulus (MPa) 51.04 5.27 431.63 Specimen no D.51 Thickness (mm) 5.14 Peak stress (MPa) 426.43 2 12.45 112.25 Average 12.22 415.08 21.5 Coupon dimension and compressive test result for E630-10 (Top) 1 Width (mm) 12.2 Middle portion (see the location on Figure 5.39 444.26 450.46 112.39 113.30 Peak stress (MPa) 412.35 467.19 0.

93 - 4 12.694 2 12.21 Thickness (mm) 5.20 467.35 1.87 1.40 113.64 50.88 - 4 12.357 3 12.38 0.51 48.26 Peak stress (MPa) 438.76 Thickness (mm) 5.88 51.58 Modulus (MPa) 49.13 0.692 3 12.30 377.09 425.44 115.8 Coupon dimension and compressive test result for E480-10 (Middle) 1 Width (mm) 12.944 Standard deviation 0.9 Coupon dimension and compressive test result for E630-10 (Middle) 1 Width (mm) 12.84 50.37 115.47 - Average 12.54 - Average 12.04 0.360 Standard deviation 0.11 49.37 113.20 443.84 5.041 2 12.39 Modulus (MPa) 49.10 Coupon dimension and compressive test result for E480-40 (Middle) 1 Width (mm) 12.196 2 12.68 5.27 5.43 114.32 - Average 12.58 5.08 19.38 5.58 Modulus (MPa) 52.22 506.06 22.01 5.81 50.22 56.03 5.47 114.61 5.31 Length (mm) 113.73 Thickness (mm) 5.01 5.18 5.33 Peak stress (MPa) 424.38 5.07 49.72 5.Appendix D EJ Guades Table D.679 3 12.43 115.43 114.37 443.43 114.45 114.78 Modulus (MPa) 43.11 Coupon dimension and compressive test result for E480-80 (Middle) 1 Width (mm) 12.06 12.01 0.11 435.319 Specimen no Behaviour of glass FRP composite tubes under repeated impact for piling application 3 .40 113.25 407.24 Peak stress (MPa) 431.28 425.30 Length (mm) 114.36 426.27 441.46 113.32 5.026 Standard deviation 0.08 332 Specimen no Table D.025 2 12.748 Specimen no Table D.279 Standard deviation 0.37 113.10 415.39 113.03 467.746 Specimen no Table D.05 0.44 113.21 395.82 6.43 0.43 114.534 3 12.43 Length (mm) 113.06 0.37 Peak stress (MPa) 415.39 - 4 12.01 - Average 12.80 - 4 12.29 432.70 50.48 Length (mm) 115.35 453.50 5.15 0.94 5.44 Thickness (mm) 5.71 5.

25 50.38 469.48 5.Appendix D EJ Guades Table D.436 3 12.29 444.64 5.95 Average 12.21 Peak stress (MPa) 433.032 Standard deviation 0.03 0.12 Coupon dimension and compressive test result for E630-30 (Middle) 1 Width (mm) 12.37 Modulus (MPa) 48.28 410.35 113.45 Length (mm) 113.40 113.35 113.37 407.11 5.44 113.30 416.02 480.862 2 12.32 113.32 Peak stress (MPa) 411.33 2 12.13 Peak stress (MPa) 435.14 Coupon dimension and compressive test result for E160-80 (Bottom) 1 Width (mm) 12.15 Coupon dimension and compressive test result for E320-80 (Bottom) 1 Width (mm) 12.649 Standard deviation 0.21 0.39 445.24 Standard deviation 0.2.61 5.20 5.04 403.056 2 12.07 442.42 0.87 3 12.75 Thickness (mm) 5.92 2 12.38 - 4 12.25 5.57 5.008 3 12.27 411.74 Thickness (mm) 5.06 7.13 0.62 49.30 113510 5.33 Length (mm) 114.7 23.57 Specimen no Table D.79 5.16 5.44 Thickness (mm) 5.33 Length (mm) 114.35 414.95 50.25 419.05 27.66 Modulus (MPa) 49.34 114.07 Specimen no Behaviour of glass FRP composite tubes under repeated impact for piling application 4 .24 5.70 3 12.34 Length (mm) 112.71 51.18 Average 12.06 10.08 0.41 5.21 114.01 976 Specimen no Table D.86 5.28 114.40 113.21 Peak stress (MPa) 405.13 Coupon dimension and compressive test result for E740-10 (Middle) 1 Width (mm) 12.74 4 12.39 Standard deviation 0.71 - 4 12.2) Table D.83 Thickness (mm) 5.33 400.30 113.96 5.24 114.19 413.35 113.62 114.31 441.06 0.10 448.84 4 12.26 - Average 12.60 5.50 5.28 114.46 0.27 0.44 - Average 12.3 Bottom portion (see the location on Figure 5.01 787 Specimen no D.37 113.

39 462.24 114.84 Standard deviation 0.70 5.13 0.11 479.39 114.38 0.36 Length (mm) 113.79 2 12.38 Length (mm) 114.35 Length (mm) 113.17 Coupon dimension and compressive test result for E630-10 (Bottom) 1 Width (mm) 12.46 0.07 0.19 Coupon dimension and compressive test result for E480-80 (Bottom) 1 Width (mm) 12.40 4 12.02 26.30 114.70 5.87 5.84 2 12.00 Thickness (mm) 5.18 5.09 448.32 454.76 Thickness (mm) 5.30 444.66 3 12.99 5.34 Peak stress (MPa) 455.44 4 12.74 Specimen no Table D.08 0.40 112.Appendix D EJ Guades Table D.02 Average 12.32 114.30 469.47 5.28 2 12.36 498.41 114.06 0.37 419.28 5.06 12.86 5.69 3 12.30 5.28 Peak stress (MPa) 408.39 113.45 114.08 450.58 4 12.21 5.30 441.06 14.09 Thickness (mm) 5.13 Thickness (mm) 5.26 114.48 Specimen no Table D.38 115.18 Coupon dimension and compressive test result for E480-40 (Bottom) 1 Width (mm) 12.72 5.55 Standard deviation 0.30 Average 12.05 Peak stress (MPa) 408.25 5.47 2 12.40 114.22 0.23 Peak stress (MPa) 437.03 0.20 434.61 5.27 456.25 112.92 3 12.18 Length (mm) 115.32 5.06 465.35 112.37 450.34 114.82 Average 12.39 5.32 114.07 34.90 3 12.82 4 12.33 Specimen no Table D.16 Coupon dimension and compressive test result for E480-10 (Bottom) 1 Width (mm) 12.22 445.36 422.87 Specimen no Behaviour of glass FRP composite tubes under repeated impact for piling application 5 .37 Standard deviation 0.38 114.35 Standard deviation 0.38 Average 12.15 5.

21 Coupon dimension and compressive test result for E740-10 (Bottom) 1 Width (mm) 12.32 0.96 Modulus (MPa) 44.44 113.390 Standard deviation 0.42 - 4 25.09 0.22 0.3 Summary of results of coupon tensile test Table D.50 5.29 444.20 463.98 230.67 41.00 5.94 5.50 5.66 5.642 Specimen no Behaviour of glass FRP composite tubes under repeated impact for piling application 6 .29 Peak stress (MPa) 428.28 442.64 4 12.31 113.16 606.57 5.63 5.70 5.17 Length (mm) 113.35 0.23 Coupon dimension and tensile test result for E320-80 1 Width (mm) 25.39 1.41 0.38 452.19 604.60 38.25 618.33 435.07 8.30 113.57 3 12.99 231.76 5.56 5.17 571.71 - Average 25.30 454.19 697.Appendix D EJ Guades Table D.99 230.30 444.00 Thickness (mm) 5.032 2 25.82 2 12.29 Modulus (MPa) 42.6 52.32 Length (mm) 113.92 Thickness (mm) 5.60 Thickness (mm) 5.86 231.748 3 25.98 40.587 2 25.81 Standard deviation 0.22 Coupon dimension and tensile test result for E160-80 Width Length Thickness Peak stress Specimen no (mm) (mm) (mm) (MPa) 1 25.20 Coupon dimension and compressive test result for E630-30 (Bottom) 1 Width (mm) 12.51 4 12.75 Average 12.13 627.01 Specimen no D.01 0.13 5.04 20.98 Standard deviation 0.707 Standard deviation 0.89 230.32 5.94 230.32 114.98 230.05 Peak stress (MPa) 571.00 5.25 Specimen no Table D.23 - 4 25.24 113.34 Peak stress (MPa) 461.00 5.56 38.07 5.20 573.21 2 12.32 113.62 3 12.00 5.04 0.80 - Average 25.02 Average 12.880 Table D.29 113.24 470.03 16.19 600.05 Length (mm) 231.12 0.00 5.21 585.95 230.44 113.76 230.77 2.828 3 25.

08 40.15 Peak stress (MPa) 594.53 41.00 Thickness (mm) 5.16 0.50 41.21 601.04 552.00 5.00 5.11 Peak stress (MPa) 625.995 3 26.21 Peak stress (MPa) 623.479 Specimen no Table D.50 5.05 598.14 611.94 231.953 2 25.50 5.56 Modulus (MPa) 38.03 231.01 231.41 0.20 562.51 - 4 26.87 Length (mm) 231.03 614.32 - Average 26.05 0.93 233.41 0.18 647.261 3 26.87 0.93 231.78 2.88 230.72 41.81 42.439 3 26.24 Peak stress (MPa) 634.04 37.Appendix D EJ Guades Table D.89 4.293 2 25.00 Thickness (mm) 5.14 Length (mm) 231.253 Standard deviation 0.08 36.95 778 Specimen no Behaviour of glass FRP composite tubes under repeated impact for piling application 7 .45 - 4 26.24 Coupon dimension and tensile test result for E480-10 1 Width (mm) 26.00 Thickness (mm) 5.08 230.00 5.07 21.09 610.22 613.82 231.213 3 25.03 - 4 25.00 5.32 - Average 25.25 627.20 642.93 231.00 5.16 603.00 5.54 1.22 639.63 5.25 Coupon dimension and tensile test result for E630-10 1 Width (mm) 26.38 5.253 Standard deviation 0.039 Standard deviation 0.86 41.00 5.89 231.26 Coupon dimension and tensile test result for E480-40 1 Width (mm) 25.69 231.00 5.23 543.03 231.27 - Average 25.00 5.23 36.08 23.30 Modulus (MPa) 45.47 Modulus (MPa) 40.186 Specimen no Table D.43 0.44 - 4 25.90 Length (mm) 230.15 44.00 5.75 5.067 2 26.27 Coupon dimension and tensile test result for E480-80 1 Width (mm) 25.473 Standard deviation 0.04 0.50 5.06 230.11 231.00 Thickness (mm) 5.01 231.12 Length (mm) 232.07 0.960 Specimen no Table D.18 584.817 2 25.82 Modulus (MPa) 41.65 - Average 25.09 607.90 230.

991 Specimen no Table D.803 Standard deviation 0.00 Thickness (mm) 5.88 5.04 38.18 230.613 3 26.15 Length (mm) 150.075.27 153.31 152.83 5.95 2.05 - 4 26.15 581.16 Thickness (mm) 5.17 975.18 150.02 Modulus (MPa) 42.46 - Average 25.79 40.24 2.49 - 4 25.57 2.90 230.89 5.72 231.13 230.32 0.07 0.00 5.721 Specimen no Behaviour of glass FRP composite tubes under repeated impact for piling application 8 .50 5.38 41.242 2 26.02 901.217 3 15.619 4 15.07 38.21 152.13 231.92 230.19 39.28 Coupon dimension and tensile test result for E630-30 1 Width (mm) 25.75 5.06 0.10 869.31 943.28 5.604 Standard deviation 0.72 5.93 5.41 0.Appendix D EJ Guades Table D.058 Average 15.05 36.25 152.23 150.99 Length (mm) 231.92 - Average 26.767 3 15.51 5.13 601.15 602.24 941.365 3 26.63 5.439 Standard deviation 0.937 Standard deviation 0.14 0.19 Peak stress (MPa) 575.195 4 15.21 152.994 Average 15.17 1.93 5.25 881.50 5.4 Summary of results of coupon flexural test Table D.14 33.95 39.29 Coupon dimension and tensile test result for E740-10 1 Width (mm) 25.147 Table D.06 79.00 5.00 Thickness (mm) 5.11 955.40 37.90 1.41 40.31 Coupon dimension and flexural test result for E320-80 1 Width (mm) 15.04 0.65 37.17 974.56 0.438 Specimen no D.24 Peak stress (MPa) 965.04 587.24 150.30 Coupon dimension and flexural test result for E160-80 Width Length Thickness Peak stress Specimen no (mm) (mm) (mm) (MPa) 1 15.85 Length (mm) 230.02 552.07 20.91 5.737 2 15.54 Modulus (MPa) 44.09 150.00 5.25 Peak stress (MPa) 607.44 0.11 628.50 5.14 37.00 40.10 230.595 2 25.918 2 15.81 Modulus (MPa) 34.10 231.16 635.20 649.30 37.69 Modulus (MPa) 37.

17 913.06 0.33 Coupon dimension and flexural test result for E630-10 1 Width (mm) 15.50 Thickness (mm) 5.171 3 15.24 5.24 39.32 Coupon dimension and flexural test result for E480-10 1 Width (mm) 15.633 Standard deviation 0.76 0.16 152.00 5.00 5.031.231 2 15.34 1.22 902.02 0.54 0.130 Standard deviation 0.29 905.35 Coupon dimension and flexural test result for E480-80 1 Width (mm) 15.493 Specimen no Behaviour of glass FRP composite tubes under repeated impact for piling application 9 .25 151.93 Modulus (MPa) 38.67 Thickness (mm) 5.21 39.18 Length (mm) 151.94 5.30 38.28 37.26 902.06 0.29 5.850 Average 15.74 38.22 151.27 151.24 152.64 5.849 Average 15.01 5.30 152.27 Length (mm) 151.23 895.22 Peak stress (MPa) 858.75 0.03 0.362 3 15.32 152.37 Modulus (MPa) 35.88 35.087 4 15.50 5.25 5.13 941.06 19.54 Thickness (mm) 5.22 918.13 897.32 151.06 37.26 Peak stress (MPa) 917.15 922.00 5.Appendix D EJ Guades Table D.15 972.03 71.662 4 15.27 151.167 2 15.08 850.04 21.571 4 15.25 Length (mm) 150.40 Length (mm) 151.140 4 15.19 37.26 0.50 5.26 151.71 Modulus (MPa) 33.466 Specimen no Table D.24 150.22 153.424 Average 15.01 42.23 943.00 Thickness (mm) 5.24 944.25 151.656 Specimen no Table D.79 37.42 1.30 153.71 37.00 5.804 Specimen no Table D.71 5.96 41.23 151.17 Peak stress (MPa) 1.748 3 15.58 39.30 3.808 Standard deviation 0.462 Standard deviation 0.55 2.13 984.07 34.433 Average 15.34 Coupon dimension and flexural test result for E480-40 1 Width (mm) 15.05 23.33 37.26 Peak stress (MPa) 954.13 5.26 893.13 5.70 5.33 946.27 153.01 38.208 2 15.255 3 15.974 2 15.58 Modulus (MPa) 35.

24 151.24 151.19 878.173 Specimen no Table D.721 Average 15.24 899.21 152.75 5.54 Modulus (MPa) 39.00 5.06 0.636 2 15.64 2.34 152.03 60.00 5.26 Length (mm) 150.50 Thickness (mm) 5.24 151.056 2 15.23 900.00 Thickness (mm) 5.07 43.31 948.25 964.37 Coupon dimension and flexural test result for E740-10 1 Width (mm) 15.23 150.43 41.26 Peak stress (MPa) 945.00 5.12 Peak stress (MPa) 938.50 38.00 5.03 0.69 37.43 0.21 0.29 861.993 Standard deviation 0.17 39.378 4 15.009 3 15.18 151.662 Specimen no Behaviour of glass FRP composite tubes under repeated impact for piling application 10 .00 5.27 809.36 Coupon dimension and flexural test result for E630-30 1 Width (mm) 15.00 5.893 Average 15.23 40.91 36.435 4 15.Appendix D EJ Guades Table D.24 151.91 2.73 33.50 37.049 Standard deviation 0.20 851.88 5.043 3 15.20 Length (mm) 152.65 Modulus (MPa) 41.