Investigating the dynamic behaviour of high performance fibres

Faculdade de Engenharia da Universidade do Porto Investigating the Dynamic Behaviour of High Performance Fibres Luis Pedro Correia da Costa MSC THESIS Mestrado Integrado em Engenharia Mecânica Supervisor FEUP: Prof. Dr. Pedro Ponces Camanho Supervisor ICL: Prof. Dr. Lorenzo Iannucci Dr. Lucio Raimondo October, 2014 Investigating the Dynamic Behaviour of High Performance Fibres Luis Pedro Correia da Costa Mestrado Integrado em Engenharia Mecânica October, 2014 Investigating the dynamic behaviour of high performance fibres ii Investigating the dynamic behaviour of high performance fibres iii If I have seen further than others, it is by standing upon the shoulders of giants. Isaac Newton Investigating the dynamic behaviour of high performance fibres iv Abstract The main goal of this thesis is to obtain the dynamic characterization of high performance fibres using a miniaturized Hopkinson bar design in Imperial College London. Beyond the dynamic characterization it was performed quasi-static tests to determine the different behaviours between high and low strain rates. To test the filaments under low strain rates, it was used a Linkam TST350 Tensile Tester and an Instron 5969 while the fibres tested under high strain rate were tested using a miniaturized Hopkinson bar. In order to measure the strain and the strength under high strain rates it was used a high speed camera and a piezoelectric load cell, respectively. The quasi-static tests performed show that there is a limit of strength for S2-Glass® and Vectran® fibres and that the Weibull analysis cannot be used to predict Vectran® strength. On the other hand Dyneema® SK76 fibres slipped through the glue and for that reason they were not tested dynamically. The dynamic strength of Vectran® and S2-Glass® fibres tested is lower than the quasi-static strength which means there is a strain rate dependence. Investigating the dynamic behaviour of high performance fibres v Investigating the dynamic behaviour of high performance fibres vi Resumo A presente tese tem como principal objetivo caracterização dinâmica de fibras de alta performance utilizando uma versão miniatura da barra de Hopkinson, desenhada no Imperial College London. Para além da caracterização dinâmica, foram realizados testes quási-estáticos para determinar as diferenças de comportamento entre elevadas e baixas taxas de deformação. As fibras testadas a baixas taxas de deformação foram testadas usando duas máquinas de teste, uma Linkam TST350 e uma Instron 5969, e os testes a elevadas taxas de deformação foram testadas usando uma miniatura da barra de Hopkinson. Para a medição da deformação e da tensão a elevadas taxas de deformação foram usadas, respetivamente, uma câmara de alta velocidade e uma célula de carga, mais propriamente um piezoelétrico. Os testes quási-estáticos executados demonstram um limite para a tensão das fibras de S2-Glass® e Vectran® e que a análise de Weibull não pode ser usada para prever a tensão de Vectran®. Por sua vez, as fibras de Dyneema® SK76 escorregaram através da cola e por essa razão não foram testadas dinamicamente. A tensão dinâmica das fibras de Vectran® e S2-Glass® testadas é inferior à tensão quási-estática, o que significa que a tensão depende da velocidade da taxa de deformação. Investigating the dynamic behaviour of high performance fibres vii Acknowledgements I would like to express my gratitude to Dr. Lucio Raimondo for guidance and help through the project. I couldn’t wish a better supervisor. He was always there for me and he was very patient man, because I was very annoying. He was more than a supervisor. This experience was fantastic because he was fantastic. I want to express my gratitude to Dr. David Anthony. He helped me with the fibre preparation, otherwise I definitely wouldn’t finish this project. If I had any question he took some of his time to help me. He was very important for guidance through the project. I would like to express my gratitude to Prof. Lorenzo Iannucci for the amazing opportunity he gave me and for the guidance and help through the project. I really enjoyed this project. I also would like to express my gratitude to Dr. Pedro Camanho who accepted to be my supervisor. I would like to thank Franco Giammaria, Jonathan Cole, Gary Senior, Keith Wolstenholme, Joseph Meggyesi, Stefano Del Rosso, Roland Hutchins and Martin Boyle for all the help they gave through this project and for the companionship: they made me feel at home. I want to express my gratitude to my girlfriend, Daniela Santos. If it wasn’t her I would never have the courage to do an experience like this, do a master thesis in Imperial College London. But I can’t forget all the support she gave me and the patience she had with me. I want to express my gratitude to my friends. In their one way, they helped me achieving my goals. They never let me down and I will be forever thankful to them. In moments of choices they don’t allow me to do bad ones. Every time that I need something they were there to help me. Lastly, I would like to thank my mother, Maria Correia. I am what I am because she taught me this way, at her own image. She never gave up on me and she gave me everything she could. She told that if I give always my best I can be succeed. I would like to thank my father, Fernando Costa. Usually we think that we can’t make an omelette without breaking eggs, but he taught exactly the opposite: there is always a way to do something. It may look impossible but if we work hard it will be the easiest thing ever. Finally, I would like to thank my entire family. Investigating the dynamic behaviour of high performance fibres 1 Investigating the dynamic behaviour of high performance fibres 2 1.1 Introduction Most of the mechanical properties of materials are obtained under quasi-static loading conditions, but not always the structures are exposed to this type loading. The dynamic characterization of materials should be studied because the behaviour of materials may be different from quasi-static loading conditions. In this thesis it will be studied the dependence of the single fibres on strain rate, testing Vectran®, Dyneema® and S2-Glass® fibres under quasi-static and high strain rate. Some materials present strain rate dependence which means they don’t exhibit the same behaviour under quasi-static and high strain rate. This behaviour becomes important if the materials are exposed to a different loading conditions than the mechanical properties were obtained. Composite materials are being used in aircraft structures due to their low weight but their use requires a perfect understanding of their behaviour. The mechanical properties of single fibres under high strain rates are difficult to achieve due to the small size of the specimens. Before the failure of the fibre there must be a state of equilibrium and ensure that the filament don’t slip during the procedure. In order to avoid the slipping of the fibre, it can be used glue or a clamping system but this way there will be stress concentrations. 1.2 Split Hopkinson Pressure Bar The Split Hopkinson Pressure Bar (SHPB), also known as the Kolsky bar, is used to determine the mechanical properties of materials under high strain rates and Hopkinson bar experiments can reach strain rates between 102 to 104 𝑠−1. The Hopkinson bar consists of two bars, the incident (or input) bar and the transmission (or output) bar, in which between them there is a sample of the material to be studied. A striker usually accelerated by a gas gun hits the incident bar causing an elastic wave pulse. When this wave reaches the end of the incident bar, a part of it will be reflected and the rest will pass through the sample. This reflection is due to the difference of impedance between the sample and the input bar. 1 Literature review Investigating the dynamic behaviour of high performance fibres 3 Frequently, to measure the strain, it is used strain gauges attached to both bars to measure the reflected and the transmitted wave although it can be used high speed cameras triggered manually or by the signal from the strain gauges. [3, 4] The reliability of the mechanical properties obtained using the Hopkinson bar test is achieved if, during the failure of the specimen, the specimen is in a state of dynamic equilibrium and the strain rate is constant. 1.3 Miniaturized Hopkinson Bar Single fibre experiments require a small load (less than 2N), due to the sample size. Since the load is small, it leads to a new version of the Hopkinson bar. Usually the diameter of the bars are smaller (the lower setup is due to the need of smaller load) but there are some other variations, such as replacing the transmission bar with a piezoelectric load cell. This last version uses the piezoelectric load cell to measure the stress and will be the approach used on this thesis for dynamic experiments. Lim et al. [8] designed a miniaturized Hopkinson bar to test PPTA single fibres. The authors replaced the output bar with a piezoelectric load cell. To produce the wave, they used a striker, launched by a spring system into a flange where this flange was part of the incident bar. The striker was separated from the incident bar by a brass tube, minimizing noise during the reading. In order to obtain a constant amplitude pulse and a constant strain rate during the experiment, the authors used a pulse shaper placed in the flange. The sample was glued between two small plates at the end of the input bar and the end of the load cell. [7, 8, 9] Figure 1 - Miniaturized Hopkinson bar with a piezoelectric load cell. [8] Investigating the dynamic behaviour of high performance fibres 4 1.4 High performance fibres The fibres used in the experiments made by Huang et al. [25] were produced by DSM and tested in a bar-bar tensile impact apparatus. This setup consists in a short metal bar connected to a block and to an input bar. When the hammer hits the block, the block will deform and break the short bar producing a stress impulse as in the Hopkinson bar experiment. These experiments were conducted at two different temperatures and the table shows the dependence of the mechanical properties on the temperature. Benloulo et al. [26] characterized the dynamic properties of woven fabric of polyethylene and unidirectional composite of polyethylene. The specimens produced from the previous materials were glued to the bars (the authors used the Hopkinson bar experiment) to avoid the use of a clamping system. However, the specimens slipped, and the authors had to design a clamping system without changing the wave propagation. Languerand et al. [32] analysed the tensile behaviour and fracture mechanisms. The difference between the two samples used during the experiments, beyond the crystalline order, was the difference between the number of filaments in each specimen: HPME-A fibre bundles had 120 filaments (38 μm of diameter for single fibre) and of HPME-B fibre bundles had 240 filaments (26 μm of diameter for single fibre). The authors used a laser detector to measure the initial fibre bundle length and the fibre bundle elongation. Justo [19] characterized Dyneema® SK66, testing the samples using the Hopkinson bar. For the dynamic tests, Justo used four different specimens: type 3 specimen with 5 layers, type 3 specimen with 2 layers, type 4 specimen with 5 layers and type 4 specimen with 2 layers. In both cases, the type 3 and type 4 specimens had the same gauge length, 20 mm. However, the width was different: type 3 had 12 mm and type 4 had 10 mm. The 5 layers specimen was tested at 200 s-1 strain rate and under quasi-static loading conditions: the ultimate strength increase by 37% and the strain failure decrease by 29%. The 2 layers specimen was tested at 135 s-1 and under quasi-static loading conditions: the ultimate strength increase by 13%. Adrian [31] studied the dynamic behaviour of Spectra® 900 and Spectra Shield® LCR. The author developed a clamping system to hold the specimens. The specimens of Spectra® 900, before the tests were heated for 30 minutes in a chamber placed at the end of the input bar and the beginning of the output bar. The Spectra Shield® LCR consists of two plies of unidirectional Spectra® 1000 extendedchain laid perpendicular to each other, and sandwiched between two thermoplastic films. Koh et al. [34] developed an algorithm to obtain correct results, due to the grips that introduced an impedance mismatch with the input/output bars. Investigating the dynamic behaviour of high performance fibres 5 Error! Reference source not found. shows, the strain rate affects the mechanical properties of the aramid fibres. All the fibres present a significant change comparing quasi-static and dynamic experiments, except the experiments made by Dooraki et al [33]. For the same fibre, Kevlar® 49, there are two different behaviours: the tests performed by Wang et al. [27] shows an increasing ultimate strength with increasing strain rate, while the tests performed by Languerand et al. [32] the opposite behaviour occurred. GPa % (vs QS) - % (vs QS) GPa % (vs QS) 0.001 0.405 - 0.102 - - - 1 0.46 13.6 0.064 -37.3 - - 1000 0.631 55.8 0.025 -75.5 - - 0.001 - - - - - - 1 0.372 - 0.053 - - - 1000 0.654 75.8 0.016 -69.8 - - 0.001 0.27 - 0.05 - - - 1 0.32 18.5 0.05 0.0 - - 1000 0.411 52.2 0.011 -78.0 - - 0.001 0.359 - 0.059 - - - 1 0.445 24.0 0.06 1.7 - - 1000 0.51 42.1 0.029 -50.8 - - 300 (25 ℃) 2.55 - 0.0652 - 80 - 300 (70 ℃) 2.47 - 0.0754 - 61 - 700 (25℃) 2.55 - 0.0626 - 82 - 700 (70℃) 2.48 - 0.0657 - 68 - Quasi-static 2.40 - 0.0410 - 66 - 340 0.96 -60.0 0.1400 241.5 25 -62.1 530 0.80 -66.7 0.1240 202.4 40 -39.4 800 1.07 -55.4 0.1940 373.2 19 -71.2 Quasi-static 3.25 - 0.0290 - 113 - 540 1.82 -44.0 0.1930 565.5 33 -70.8 670 2.03 -37.5 0.1940 569.0 36 -68.1 0.001 (20 1.641 - 0.052 - - - 433 (20℃) 2.519 53.5 0.029 -44.2 - - 0.001 (40 1.574 - 0.067 - - - 477 (40℃) 2.46 56.3 0.04 -40.3 - - 0.001 (60 1.334 - 0.076 - - - 510 (60℃) 2.463 84.6 0.038 -50.0 - - 0.002 (20 0.2458 - 0.048 - - - 465 (20℃) 0.7492 204.8 0.031 -35.4 - - 0.001 1.5 - 0.05 - - - 365-433 2.5 66.7 0.029 -42.0 - - Strain Rate (1/s) Reference Gauge Length (mm) Ultimate Strength Failure Strain Young's Fiber Spectra Shield Spectra Shield Spectra 900 SK66 UD66 UHMWPE fibres PET1 PET2 30 [26] HPME-A HPME-B 25 [34] 3.416 to 4.539 [32] [25] 8 [31] 25 3360 1333 Fibre Bundles Fibre Bundles Yarns Yarns Yarns Specimen Linear density (dtex) N.A. N.A. N.A. N.A. N.A. Table 1 - Mechanical properties of Polyethylene fibres Investigating the dynamic behaviour of high performance fibres 6 Lim et al. [8] used a modified Hopkinson bar to performed their experiments. The authors used a miniaturized Hopkinson bar, replacing the transmission bar with a piezoelectric load cell. The incident bar had 6.35 mm of diameter and 1651 mm length, made of aluminium. They tested five different gauges length, 2.5, 5.5, 10, 50, 100 and 250 mm, and removed the Kevlar fibres from woven fabric, in warp and weft directions, and from a yarn that not suffered a weaving process. The high strain rates experiments were performed at 1500 s-1, and, to achieve a constant-amplitude incident pulse, GPa % (vs QS) - % (vs QS) GPa % (vs QS) Quasi-static 2.85 - - - - - Dynamic 2.9 1.8 - - - - Quasi-static 2.87 - - - - - Dynamic 3.08 7.3 - - - - Quasi-static 2.75 - - - - - Dynamic 2.78 1.1 - - - - Quasi-static 2.66 - - - - - Dynamic 3.27 22.9 - - - - 0.0001 2.34 - 0.0329 - 97 - 0.01 2.47 5.6 0.0333 1.2 100 3.1 140 2.94 25.6 0.0354 7.6 112 15.5 440 3.02 27.5 0.0364 10.5 119 22.0 1350 3.08 25.2 0.0386 16.1 125 25.0 Quasi-static 3.4 - 0.0310 - 77 - 500 1.57 -53.8 0.0500 61.3 59 -23.4 850 1.89 -44.4 0.0950 206.5 48 -37.7 Quasi-static 3.0 - 0.0370 - 63 - 560 2.9 -3.3 0.1000 170.3 41 -34.9 580 2.4 -20.0 0.1060 186.5 41 -34.9 Quasi-static 3.1 - 0.0240 - 120 - 500 3.8 22.6 0.1000 316.7 99 -17.5 540 3 -3.2 0.1000 316.7 105 -12.5 8 Fiber PPTA-B (Kevlar 29) Twaron (1350 fibres) Strain Rate (1/s) Ultimate Strength Failure Strain Young's Gauge Length (mm) Reference [33] 3.416 to 4.539 [32] Kevlar 49 8 [27] Kevlar 129 (1154 fibres) Kevlar KM2 (1166 fibres) Kevlar LT (1080 fibres) PPTA-A (Kevlar 49) 1270 PPTA-C 1100 440 950 940 fibre bundles Specimen Fibre Bundles Fibre Bundles Linear density (dtex) N.A. Table 2 - Mechanical properties of Aramid fibres Investigating the dynamic behaviour of high performance fibres 7 the authors used a pulse shaper. From the results, the research team concluded that the fibres studied did not show a significant strain rate or gauge length dependence. GPa % (vs QS) - % (vs QS) GPa % (vs QS) Quasi-static 4.12 - - - - - Dynamic 6.55 59.0 - - - - 0.001 0.503 - 0.068 - - - 1 0.533 6.0 0.069 1.5 - - 1000 0.575 14.3 0.039 -42.6 - - Young's Gauge Length (mm) Reference Fiber Strain Rate (1/s) Ultimate Strength Failure Strain Zylon (1268 fibres) PP [26] 30 [33] 8 3360 560 Yarns Fibre Bundles Specimen Linear density (dtex) Table 3 - Mechanical properties of other fibres Investigating the dynamic behaviour of high performance fibres 8 Dooraki [33] performed high strain rates tests on aramid fibres in a miniaturized Hopkinson bar. He used a high speed camera to measure the deformation of the specimens during the experiments. The average strain rate was calculated by averaging the displacement rate obtained from each frame. In order to measure the effect of specimen size on failure stress, Dooraki [33] tested single fibres and multi-fibre specimens of Kevlar® 129 with various gauge lengths( 5, 16, 25, 50 and 100 mm, for the single fibre experiments, and 24, 100 and 170 mm for the multi-fibre experiments), but the author only tested the specimens under quasi-static loading conditions. However, it can be seen a significant dependence of Failure Stress with increasing gauge length: the higher gauge length, the lower failure stress was found, although this effect is more evident for multi-fibre. Another effect was perceptible: for the same gauge length, there were two different specimens, one with 1154 fibres and other with 2308 fibres. The specimen with more fibres had a lower failure stress, and this effect can be explain due to the friction between the fibres. Hill and Okoroafor [36] performed tests on fibre bundles to obtained tensile properties of fibres, and they used fibre bundles lubricated and fibre bundles without any lubrication. The authors concluded that the lubrication does not affect the Young’s Modulus, however, the ultimate strength and the failure strain are significantly reduced in the dry bundle tests. 1.5 Problems characterizing fibre bundles In order to measure the mechanical properties of fibres, in this thesis, it was used single fibres, although it could be used fibre bundles. Measuring mechanical properties using single fibres can introduce errors due to fibre damage during the preparation of the samples, but on the other hand using fibre bundles can minimize this problem. It is expected that different methods leads to different results. Hill and Okoroafor [36] performed tests using lubricated and non-lubricated fibre bundles at low strain rates, and tried to explain the difference between the results obtained. One of the main reasons can be explained due to interfibre interaction/friction during the experiment. The authors concluded that the lubrication does not affect the Young’s Modulus of the fibres, however the Ultimate Strength and the Failure Strain are significantly affected by lubrication. The lubricated fibre bundles had higher Ultimate Strength and Failure Strain values. Investigating the dynamic behaviour of high performance fibres 9 2.1 Fibre preparation In order to perform the quasi-static tests, it was need to prepare the samples to be tested. The single fibres were placed in a frame template printed out on card (if the paper is not strong enough, the fibre can be damage during the preparation and handling). In the template there is double-sided tape to align and hold the fibres before glue them with an appropriate glue. The gauge length is set by the window cut in the template. For the experiments the gauge lengths tested on the Linkam TST350 were 15, 20 and 25 mm (the Tensile Stress Tester only allow at least 14 mm) and on the Instron 5969, the gauge lengths tested were 4, 6 and 8 mm. To test the single filaments using the Instron 5969, it was need to develop a clamping system. The clamping system (Error! Reference source not found.) consists in two aluminium parts and two screws to tight the template and avoid any slippage. On the other end of the template there is a pneumatic grip. 2 Quasi-Static Experiments Figure 2 - Clamping system used on the Instron 5969 Tensile Tester Investigating the dynamic behaviour of high performance fibres 10 The glue used to glue the fibres was Araldite Rapid, an epoxy glue. It should not be used super glue unless a standard explicitly says to use it. Although the ultimate strength is the same if it is used super glue or epoxy glue, the Young’s Modulus will be affected by the super glue. A complete cure of the epoxy glue used should last at least 3 nights. After the cure, the double-tape attached to the template is cut off and the samples are ready to be tested. Figure 3 shows the frame template after the preparation. 2.2 Quasi-static tests The quasi-static tests were performed using a Linkam TST350 Tensile Stress Tester, with a 20N load cell and an Instron 5969 with a 10N load cell. The frame template is clamped and aligned before the test start. After clamping the frame, the card that keep the fibre stretched (Figure 3) should be cut, otherwise the fibre won’t be properly tested. A previous pre-cut before glue the fibre to the frame template will allow an easier cur and avoid any damage on the fibre. Before perform the test the force and displacement were set to zero (both software, Linkam and Instron, had an option to set these values to zero). The strain rate was set to 0.001/s, the lowest strain rate that the Linkam Tensile Stress Tester could achieve, even though the Instron could achieve lower strain rates. Figure 3 - Frame Template Figure 4 - Linkam TST350 Tensile Tester Investigating the dynamic behaviour of high performance fibres 17 Figure 12 - Typical Stress vs Strain curves (before system compliance correction) Investigating the dynamic behaviour of high performance fibres 18 Figure 13 - Corrected Strain at failure of S2-Glass® single fibres (Linkam Tensile Tester) Figure 14 - Tensile strength of S2-Glass® single fibres (Linkam Tensile Tester) Investigating the dynamic behaviour of high performance fibres 19 Figure 15 - Corrected Young's Modulus of S2-Glass® single fibres (Linkam Tensile Tester) Figure 16 - Tensile strength of S2-Glass® single fibres (Instron Tensile Tester Investigating the dynamic behaviour of high performance fibres 20 Figure 17 - Corrected Strain at failure of S2-Glass® single fibres (Instron Tensile Tester) Figure 18 - Corrected Young's Modulus of S2-Glass® single fibres (Instron Tensile Tester) Investigating the dynamic behaviour of high performance fibres 21 S2-Glass® fibres tested using the Linkam Tensile Tester exhibit a significant gauge length dependence. The strength and the strain at failure increased when the gauge length decreased. Although the strength of the fibres tested on the Instron is not the same for the different gauge length it is not correct to say that there is a gauge length dependence. Since the strength practically remains the same, there must be a limit of strength for those fibres and that limit is between 8 and 15 mm gauge length which means that no matter the gauge length under that limit, the strength won’t be too far for the results obtained in Error! Reference source not found.. Weibull Analysis Experimental 44021 3550 63770 3906 83601 3884 Gauge length [mm] Strength [MPa] Figure 19 - Weibull analysis for S2-Glass® Table 6 - Experimental and prediction of S2-Glass® fibres strength Investigating the dynamic behaviour of high performance fibres 22 Figure 20 - Typical Stress vs Strain curves (before system compliance correction) Investigating the dynamic behaviour of high performance fibres 23 Figure 21 - Tensile strength of Dyneema® SK®76 single fibres (Linkam Tensile Tester) Figure 22 - Corrected Strain at failure of Dyneema® SK76 single fibres (Linkam Tensile Tester) Investigating the dynamic behaviour of high performance fibres 24 Dyneema® SK76 fibres were only tested using the Linkam Tensile Tester. However, from the results obtained, the strength is not affected by the gauge length. However, it is not possible to compare the prediction from the Weibull analysis and the experimental results because the fibres were not tested using the Instron 5969 Tensile Tester. Figure 24 - Weibull analysis for Dyneema® SK76 Figure 23 - Corrected Young's Modulus of Dyneema® SK76 single fibres (Linkam Tensile Tester) Investigating the dynamic behaviour of high performance fibres 25 Figure 25 - Typical Stress vs Strain curves (before system compliance correction) Investigating the dynamic behaviour of high performance fibres 26 3.1 Miniaturized Hopkinson Bar The miniaturized Hopkinson Bar was developed by Dr. Lucio Raimondo and Prof. Lorenzo Iannucci in which they used the previous split Hopkinson pressure bar and adapted to the new version. This new version, the miniaturized Hopkinson Bar, consists of two titanium bars, with different lengths and diameters, by a striker and a piezoelectric load cell. The input bar consists in a 2.1 m length titanium bar with 12.7 mm of diameter and in a 0.85 m length titanium bar with 4 mm of diameter. The smaller diameter bar is screwed into the other and on the other end there is a silver steel pin where the specimen is glued. The piezoelectric load cell has another silver steel pin where the other end of the specimen will be glued. Both pins are screwed into the titanium and the piezoelectric load cell so that they can be removed after the test and screw two new pins. 3 Dynamic Tests Figure 26 - Flange at the end of the input bar Investigating the dynamic behaviour of high performance fibres 33 Figure 33 - FIbre misaligned Investigating the dynamic behaviour of high performance fibres 34 4.1 Conclusions In this thesis, it was performed single fibres tests under high strain rates and under low strain rates. In order to investigate the dynamic behaviour of high performance fibres, it was performed single fibres tests at quasi-static conditions using an Instron 5969 and a Linkam TST350 to compare both behaviours. From the quasi-static tests it was possible to realize that the Weibull analysis cannot be used to predict the strength of Vectran® fibres. The strength of Vectran® is not affected by the gauge length and it seems there is a limit of strength for these fibres which is under 4GPa. S2-Glass® fibres exhibit a gauge length dependence, although this dependence is more significant between 15 and 35mm. Under 15mm, the strength does not change, which may indicate that the limit of strength of these fibres has been reached. Dyneema® SK76 fibres weren’t tested under high strain rates because the fibres slipped through the glue while being tested using the Linkam TST350 Tensile Tester. The dynamic strength of Vectran® and S2-Glass® is lower than the quasi-static strength: S2-Glass® strength diminished 5% while Vectran® strength diminished 25%. Although it was tried to align the fibres, and at naked eye they look aligned, using the high speed camera it was possible to realize that some of the fibres tested weren’t in fact aligned. This misalignment affected the strength. 4.1 Future Work As previously discussed, during the project it was found some problems that with more time would certainly be resolved. One of them was that sometimes after fire the striker, due to its small diameter, the striker was touching the 12.7 mm bar and then the bar moves before the striker hits the flange. This affect the results because the fibre was being stretched before the wave runs through it increasing the time that the test lasts and decreasing the strain rate. Another problem faced was the glue used to glue Dyneema® fibres. If it was possible to test Dyneema® SK76 fibres under quasi-static strain rate using a low viscosity cyanoacrylate, under high strain rate the fibre slipped using the same glue. Although the results obtained using the miniaturized Hopkinson bar are reasonable, due to the glue used, the time to test one fibre was too long (glue the fibres plus test it dynamically). Araldite 2021 was more than enough to hold the fibres during the test but for a test as fast as 100 microseconds 4 Conclusions and Future work Investigating the dynamic behaviour of high performance fibres 35 (the worst scenario), 20 minutes is too much. To solve this problem it could be used a different glue, although this was the best found, or develop a clamping system. Investigating the dynamic behaviour of high performance fibres 36 [1] – Cwik, Tomasz Krzysztif. 2013. “ Highly Instrumented Static, Dynamic, and Impact Testing of High Performance Materials”. PhD, Imperial College London. Access at 10 of February of 2014. [2] – HBM (Hottinger Baldwin Messtechnik). “Split Hopkinson bar”. Access at 13 of February of 2014. http://www.hbm.com [3] – Chen, Weinong and Song, Bo. 2011. Split Hopkinson (Kolsky) Bar. Springer. [4] – Pinto, José Manuel Grilo Taveira. 2009. “ Avaliação do Comportamento Mecânico de Blindagens Balísticas”. MSc, Instituto Superior Técnico, Universidade Ténica de Lisboa. Access at 13 of February of 2014. [5] – Restivo, Maria Teresa; Almeida, Fernando Gomes; Chouzal, Maria de Fátima; Mendes, Joaquim Gabriel and Lopes, António Mendes. Laboratórios de instrumentação para medição. Access at 20 of February of 2014 [6] – Vaz, Mário A.P. and Gomes, J.F. Silva. Análise Experimental de Tensões, Access at 20 of February of 2014 [7] – Siviour, Clive R. and Jordan, Jennifer L.. 2004. A Miniaturized Split Hopkinson Pressure Bar for Very High Strain Rate Testing. Access at 22 of February of 2014 [8] – Lim, Jaeyoung; Zheng, James Q.; Masters, Karl; and Chen, Weinong W., 2011. “ Effects of gage length, loading rates, and damage on the strength of PPTA fibres”. US Army Research. Paper 134. Access at 24 of February of 2014 http://digitalcommons.unl.edu/usarmyresearch/134 [9] – Ramesh, K.T. 2008. High Rates and Impact Experiments. In Handbook of Experimental Solid Mechanics. 929-960. Springer [10] – Casem, Daniel T.. 2009. “A Small Diameter Kolsky Bar for High Rate Compression”. US Army Research. Access at 25 of February of 2014 [11] – Whitenton, Eric. The NIST Koslky Bar Data Processing System. National Institute of Standards and Technology (NIST). Access at 25 of February of 2014 [12] – Leblans, P. J. R.; Bastiaansen, C. W. M. and Govaert, L. E. . Viscoelastic Properties of UHMW-PE Fibres in Simple Elongation. DSM Research BV and T.U. Eindhoven. Access at 26 of February of 2014 References Investigating the dynamic behaviour of high performance fibres 37 [13] – Bouwmeester, J.G.H.; Marissen, R. and Bergsma, O. K. .2008. “Carbon/Dyneema® Intralaminar Hybrids: New Strategy to Increase Impact Resistance or Decrease Mass of Carbon Fibre Composites”. 26th International Congress of the Aeronautical Sciences. [14] – Jalili, I. ; Nouri, Z. H. ; Aabady, A.Z. and Akbari, K. . 2012. “ Strengthening the composite protective shield of light-weight ship against ballistic impacts: Analytical and Experimental”. Latin American Journal of Solids and Structures. Access at 26 of February of 2014 [15] – Chocron, Sidney; Pintor, A.; Cendon, A.; Rosello, C. and Sanchez-Galvez, V. . 2002. Characterization of Fraglight Non-Woven Felt and Simulation of FSP’s Impact in it. Access at 26 of February of 2014 [16] – Chocron, S.; Rodríguez, J. and Sánchez-Gálvez, V. . 1997. “A Simple ANalytical Model for Ballistic Impact in Composites”. Supplément au Journal de Physique III d’août 1997. Access at 26 of February of 2014 [17] – Technical brochure: Dyneema® in marine and industrial applications. Access at 26 of February of 2014 http://www.dekkerwatersport.nl/pdf/d12.pdf [18] - Utomo, B.D. Heru and Ernst, L.J. 2008. “Detailed Modeling of Projectile Impact on Dyneema® Composite Using Dynamic Properties”. Journal of Solid Mechanics and Materials Engineering. Access at 26 of February of 2014 [19] – Justo, J. M. C. F.. 2005. “Estudo do Comportamento ao Impacto de Alta Velocidade de Estruturas em Materiais Compósitos”, PhD, Faculdade de Engenharia, Universidade do Porto. Access at 26 of February of 2014 [20] – Song, Bo and Chen, Weinong. 2005. “Split Hopkinson Pressure Bar Techniques for Characterising Soft Materials”. Latin American Journal of Solids and Structures”. Access at 27 of February of 2014 [21] - Hoogsteen, W. ; Hooft, R. J. van der; Postema, A. R.; Brinke, G. ten and Pennings, A. J. . 1988. “Gel-spun Polyethylene fibres”. Journal of Materials Science”. Access at 27 of February of 2014 [22] – Lim, Jaeyoung; Chen, Weinong W. and Zheng, James Q.. 2010. “Dynamic small strain measurements of Kevlar® 129 single fibres with a miniaturized Kolsky bar”. Elsevier. Access at 1 of March of 2014 [23] – Xia, Yuanming; Yuan, Jianming and Yang, Baochang. 1995. “A Statistical Model and Experimental study of the strain-rate dependence of the strength of fibres”. Composites Science and Technology. Access at 3 of March of 2014 [24] – Zhu, D.; Mobasher, B.; Erni, J.; Bansal, S. and Rajan, S. D.. 2012. “Strain rate and Gage length effects on tensile behaviour of Kevlar® 49 single yarn”. Elsevier. Access at 3 of March of 2014 [25] – Huang, Wen; Wang, Yang and Xia, Yuanming. 2004. “Statistical dynamic tensile strength of UHMWPE-fibres”. Elsevier. Access at 3 of March of 2014 [26] – Benloulo, I.S. Chocron; Rodríguez, J.; Martínez, M.A. and Gálzez, V. Sánchez. 1996. “Dynamic tensile testing of aramid and polyethylene fibre composites”. Access at 4 of March of 2014 [27] – Wang, Yang and Xia, Yuanming. 1998. “ The effects of strain rate on mechanical behaviour of Kevlar® fibre bundles: an experimental and theoretical study”. Elsevier. Access at 4 of March of 2014 Investigating the dynamic behaviour of high performance fibres 38 [28] – Hearle, J. W. S. . 2001. High-perfomance Fibres. USA. Woodhead Publishing Ltd. Access at 07 of March of 2014 [29] – Kevlar Technical Guide. Access at 7 of March of 2014 [30] – Shukla, Arun; Ravichandran, Guruswami and Rajapakse, Yapa D.S.. 2009. Dynamic Failure of Materials and Structures. Springer. Access at 12 of March of 2014 [31] – Adrian, Koh Chien-Ping. 2009. “Response of yarn systems to impact loading”. PhD, National University of Singapore. Access at 13 of March of 2014 [32] – Languerand, D.L.; Zhang, H.; Murthy, N.S.; Ramesh, K.T. and Sansoz, F. . 2008. “Inelastic behaviour and fracture of high modulus polymeric fiber bundles under high strain rates”. Materials Sience and Engineering A, 216-224. Elsevier. Access at 13 of March of 2014 [33] – Dooraki, Babak Farsi. 2006. “Study of Parameters Affecting the Strength of Yarns”. MSc. McGill University. Access at 15 of March of 2014 [34] – Koh, A.C.P.; Shim, V.P.W. and Tan V.B.C.. 2010. “Dynamic behaviour of UHMWPE yarns and addressing impedance mismatch effects of specimen clamps”. International Journal of Impact Engineering, 324-332. Elsevier. Access at 16 of March of 2014 [35] – Wang, Zhen. 1995. “ Experimental Evaluation of the Strength Distribution of E-Glass Fibres under high strain Rates”. Applied Composite Materials, 257-264. Kluwer Academic Publishers. Access at 16 of March of 2014 [36] – Hill, R. and Okoroafor, E. U.. 1994. “Weibull statistics of bibre bunfle failure using mechanical and acoustic emission testing: the influence of interfibre friction”. Compostites 26, 699705. Elesevier. Access at 23 of March of 2014 [37] - Anthony, David Benbow; PhD thesis: Improved Synthesis of Carbon Nanotube Grafted Carbon Fibre: Towards Continuous Production, Chemical Engineering, Imperial College London, 233 pages (2013)