Hybrid composites under high strain rate compressive loading

N.K. Naik , Veerraju Ch, Venkateswara Rao Kavala

Aerospace Engineering Department, Indian Institute of Technology Bombay, Powai, Mumbai 400076, India

a b s t r a c t
Hybrid composites consist of two or more types of bres and/or matrices in a composite. By combining two or more types of bres, it is possible to club advantages of both the bres while simultaneously mitigating their less desirable qualities. Investigations on high strain-rate behaviour of a typical hybrid composite under compressive loading are presented. The hybrid composite is made using satin weave carbon and plain weave E-glass with epoxy resin. Studies were also carried out on satin weave carbon/epoxy and plain weave E-glass/epoxy. Compressive split Hopkinson pressure bar (SHPB) apparatus was used for the studies. Compressive properties were evaluated along all the three principal directions in the strain-rate range of 5461503 s1 . During SHPB testing of the specimens, it was observed that the peak force obtained from the strain gauge mounted on the transmitter bar is lower than the peak force obtained from the strain gauge mounted on the incident bar. The explanation for this is provided based on stress wave attenuation studies.

Keywords: Hybrid composite Woven fabric High strain rate Compressive loading

1. Introduction Polymer matrix composites are nding increasing uses during the last four decades in high technology as well as conventional applications. A variety of bre and resin systems are available to the designers for both high technology and conventional applications. The designers are looking for high specic stiffness, high specic strength, enhanced dimensional stability, energy absorption, corrosive resistance as well as reduced cost while selecting a material system for typical applications. Composites made of a single reinforcing material system may not be suitable if it undergoes different loading conditions during the service life. Hybrid composites may be the best solution for such applications. Hybrid composites consist of two or more types of bres and/or matrices in a composite. By combining two or more types of bres, it is possible to club the advantages of both the bres while simultaneously mitigating their less desirable qualities. Normally, one of the bres in a hybrid composite is a high-modulus and high-cost bre such as carbon, boron and the other is usually a low-modulus bre such as E-glass, Kevlar. The high-modulus bre provides the stiffness and load bearing qualities, whereas the low-modulus bre makes the composite more damage tolerant and keeps the material cost low. Hybrid composites offer an effective way of increasing mechanical properties and reducing the cost of advanced composite

materials. The mechanical properties of a hybrid composite can be varied by changing volume ratio and stacking sequence of different plies. High-modulus bres such as carbon, boron are widely used in many aerospace applications because of their high specic modulus. However, the impact strength of composites made of such high-modulus bres is generally lower than conventional steel alloys or glass reinforced composites. An effective method of improving the impact properties of high-modulus bre composites is to add some percentage of low-modulus bres like E-glass. Depending on the geometrical pattern of bre/yarn/ply, hybrid composites can be classied as interply (interlaminated) and intraply (intermingled). Interply hybrid composites consist of different bre plies bonded together in a matrix. In intraply hybrid composites, each fabric consists of two or more kinds of bres. There are many studies on hybrid composites under different loading conditions: tensile [19], compressive [10,11], low velocity impact [1217], exure [12,14,18,19] and fatigue [19,20]. It is generally observed that the failure strain increases signicantly for hybrid composites compared with those of high-modulus and low-strain component of the hybrid composites. The strength also increases for hybrid composites compared with rule of mixtures predictions. The impact properties are also enhanced for hybrid composites compared with those of high-modulus and low-strain component of the hybrid composites. High strain rate loading is one of the possibilities in many structural applications. It has always been a case for concern that the mechanical properties of composite materials may be different at high strain rate loading compared with those at quasi-static load-

88

Fig. 1. Compressive split Hopkinson pressure bar (SHPB) apparatus, (a) schematic arrangement and (b) photograph.

ing. This necessitated to carry out investigations on how mechanical properties of composites would change with strain rate. The widely used technique for the determination of behaviour of composites under high strain rate loading is the split Hopkinson pressure bar (SHPB) apparatus. The working of this apparatus is based on onedimensional wave propagation theory in elastic bars. Various studies were carried out on specimen design [21], stress uniformity [22] and stress wave reverberation [23,24]. It was observed that the loading pulse with larger rise time generates better stress uniformity than the loading pulse with lesser rise time [22,23]. Rise time of the incident pulse can be increased by placing a pulse shaper such as a small strip of soft material at impact end of the incident bar. Pulse shaper technique was used in SHPB experiments to obtain accurate stressstrain data at high strain rates for metals [25,26], composites [2729] and elastic-plastic materials [30]. Various studies were carried out on high strain-rate behaviour of unidirectional and cross-ply glass composites [3134] and woven fabric glass composites [28,3437] under compressive loading. Generally, it was observed that the compressive strength and modulus increase compared with those at quasi-static loading. It was observed that there is a signicant scatter in the case of failure strain. Various studies were also carried out on high strain-rate behaviour of unidirectional and cross-ply carbon composites [3843] and woven fabric carbon composites [4447] under compressive loading. In this case also, it was observed that the compressive strength and modulus increase compared with those at quasi-static loading. It was observed that there is a signicant scatter in the case of failure strain. Typical studies are well documented in the form of review articles [4850]. Even though there are typical studies on the compressive behaviour of composites under high strain rate loading, studies on

hybrid composites are not available. Further, possible stress wave attenuation effect, especially for woven fabric composites, is not discussed in the previous studies. The objective of the present study is to determine the behaviour of a typical satin weave carbonplain weave E-glass and epoxy hybrid composite under high strain rate compressive loading. Compressive strength, modulus and ultimate strain were evaluated and are presented along warp, ll and thickness directions. Studies were also carried out on satin weave carbon/epoxy and plain weave E-glass/epoxy composites for comparison. Force versus time plots based on strain gauge signals obtained from incident bar and transmitter bar are derived and compared. During testing of specimens, it was observed that the peak force obtained from the strain gauge mounted on the transmitter bar is lower than the peak force obtained from the strain gauge mounted on the incident bar. The explanation for this is provided based on stress wave attenuation studies. 2. Experimental set up The standard compressive SHPB apparatus as shown in Fig. 1 was used for the studies [51,52]. The main parts of the compressive SHPB apparatus are: propelling mechanism, striker bar, incident bar, transmitter bar and support stand. The diameter of the incident and transmitter bars is 12 mm and the length was varied in the range of 10001400 mm. The length of the striker bar was varied in the range of 150430 mm. The bars are made of SUS440C martensite stainless steel with Youngs modulus of 203 GPa and density of 7667 kg/m3 . The entire strain/deformation history within the specimen can be obtained by taking measurements along the incident and transmitter bars from strain gauges with the assistance of amplier

89

and oscilloscope. From these signals and using one-dimensional wave propagation theory, strain rate versus time, strain versus time, stress versus time and stress versus strain history in the specimen can be determined. The techniques used for controlling pulse duration and amplitude and strain rate are presented in Appendices A and B, respectively. 2.1. Specimen design and dimensions Specimen design is one of the most important considerations in SHPB testing. From the literature it is observed that cylindrical specimens with L/D ratio of 0.52.0 are suitable for compressive SHPB testing of polymer matrix composites [21]. Here, L and D represent length and diameter of the specimen. The specimen geometry and L/D ratio are important to achieve dynamic stress equilibrium within the specimen. In all the experiments, cylindrical specimens with L/D ratio of 0.75 were used. The diameter of the specimen used was 8 mm. This ensures the impact on full cross-section of the specimen and also permits the specimen to expand along radial direction within the cross-sectional area of the bars after the compressive load is applied. 2.2. Pulse shaper technique To increase the rise time, smoothen the pulse and to modify the shape of the pulse, a pulse shaper was used. Three different materials, namely copper, brass and aluminum were used for making the pulse shapers. The diameter of pulse shaper is 12.5 mm whereas the thickness was varied in the range of 0.53 mm. 3. Hybrid composites Hybrid composites were made using satin weave carbon and plain weave E-glass with epoxy resin. The lay-up used was: [G/C]9s . Here, G represents plain weave E-glass fabric whereas C represents satin weave carbon fabric. Satin weave carbon/epoxy, plain weave carbon/epoxy and plain weave E-glass/epoxy composites were also used for the study. Fibre volume fractions: Hybrid composite, Vf = 0.52 Plain weave E-glass/epoxy, Vf = 0.51. Satin weave carbon/epoxy, Vf = 0.56. Plain weave carbon/epoxy, Vf = 0.51. Composite volume fraction Composite volume fraction is dened as the ratio of volume of one of the constituents of the hybrid composite to the volume of hybrid composite. Carbon composite volume fraction, Vc = 0.55. E-glass composite volume fraction, Vg = 0.45. Volume fraction of a particular type of reinforcement as a part of volume of total reinforcement This is dened as the ratio of volume of one type of reinforcement to the volume of total reinforcement of the hybrid composite. Carbon bre volume fraction, Vfc = 0.57. g E-glass bre volume fraction, Vf = 0.43. 4. Theory The design of SHPB is based on one-dimensional wave propagation in elastic bars which deals with the motion of particles in longitudinal direction. One-dimensional system can ideally be considered to be of innite length and negligible diameter. Since it is
Fig. 2. Compressive SHPB test results obtained during calibration, (a) strain gauge signals on oscilloscope and (b) comparison of force vs. time behaviour, derived from strain gauge signals.

not possible in practice, the theory is adopted with certain approximations. The analytical relations to calculate strain rate, strain and stress as a function of time in the specimen in SHPB testing are: The strain rate, S (t) = 2CO lS R (t) 2CO lS
t

(1) R (t) dt
O

The average strain, The stress,

S (t)

S (t) = AB T (t) AS

(2) (3)

= E

where CO is elastic wave velocity in the bars, lS is specimen gauge length, AB is cross-sectional area of the bars, AS is cross-sectional area of the specimen, E is Youngs modulus of the bars, R is reected strain pulse, T is transmitted strain pulse and t is time duration. 5. Calibration of compressive SHPB apparatus For commissioning and assessing the accuracy of SHPB apparatus, calibration was carried out rst. During calibration, the two elastic bars were wrung together without a specimen sandwiched between them. Lubrication was applied between the bars to minimize friction. With this the incident and transmitter bars can be treated as a single bar. Strain gauge signals on the oscilloscope during calibration are presented in Fig. 2(a). Channel 1 indicates the output of the strain gauge mounted on the incident bar whereas channel 2 indicates the output of the strain gauge mounted on the

90 Table 1 Strain rate effect on compressive properties of satin weave carbonplain weave E-glass and epoxy hybrid composite Along Warp Strain rate, (s1 ) 592 (318) [1235] 1184 (820) [1685] 1321 (501) [1503] 546 (410) [1002] 865 (412) [1593] 1139 (501) [1457] 1275 (+136) [1594] 1503 (+410) [1503] Strength,
ult

(MPa)

Ultimate strain, ult (%) 1.48 (0.05, +0.08) 1.95 (0.24, +0.10) 2.09 (0.10, +0.07) 1.09 (0.01, +0.13) 1.21 (0.07, +0.06) 1.82 (0.61, +0.05) 1.67 (0.11, +0.13) 2.06 (0.06, +0.14)

Modulus, E (GPa) 21.2 (1.3, +0.7) 22.5 (1.1, +2.6) 21.9 (0.6, +1.7) 21.8 (0.9, +0.5) 23.7 (1.6, +0.3) 22.3 (2.1, +5.0) 12.7 (0.5, +1.3) 13.6 (1.0, +0.8)

292 (19, +03) 334 (07, +16) 358 (04, +11) 282 (14, +06) 341 (10, +11) 360 (12, +15) 465 (09, +16) 680 (15, +08)

Fill

Thickness

Overall bre volume fraction, Vf = 0.52. Carbon composite volume fraction, Vc = 0.55. E-glass composite volume fraction, Vg = 0.45.

Table 2 Strain rate effect on compressive properties of satin weave carbon/epoxy Along Warp Strain rate, (s1 ) 774 (455) [1503] 1139 (911) [1776] 1230 (865) [1787] 728 (637) [1731] 1002 (820) [1822] 1184 (774) [1640] 1503 (+273) [1503] Strength,
ult

(MPa)

Ultimate strain, ult (%) 1.36 (0.02, +0.10) 1.70 (0.35, +0.12) 1.64 (0.05, +0.13) 1.35 (0.03, +0.08) 1.50 (0.14, +0.09) 1.70 (0.10, +0.08) 2.71 (1.20, +2.10)

Modulus, E (GPa) 20.3 (0.9, +1.8) 25.0 (2.3, +1.5) 23.9 (1.0, +1.7) 22.2 (1.6, +0.8) 25.7 (2.5, +1.6) 24.4 (0.7, +1.1) 11.3 (0.1, +2.5)

296 (07, +19) 347 (17, +08) 323 (10, +18) 294 (05, +12) 367 (20, +08) 334 (11, +16) 621 (10, +38)

Fill Thickness

Overall bre volume fraction, Vf = 0.56.

transmitter bar. Here, I is the incident pulse with pulse duration equal to a1 a2 whereas T is the transmitted pulse with pulse duration equal to b1 b2 . During calibration, reected pulse (R) is not present. The amplitude and duration of incident and transmitted pulses are nearly the same. Force versus time plots are obtained from the strain gauge signals and are presented in Fig. 2(b). The force history obtained based on the strain gauge mounted on the incident bar is indicated by F1 whereas the force history obtained based on the strain gauge mounted on the transmitter bar is indicated as F2 . It may be noted that the forces obtained F1 and F2 match very well. This indicates that the stress states within the incident bar and transmitter bar are exactly the same. This ensures that the SHPB apparatus is perfectly aligned and friction free. The apparatus is ready for further investigations. 6. Experimental studies Experimental studies were carried out on high strain-rate behaviour of satin weave carbonplain weave E-glass with epoxy
Table 3 Compressive properties of various composite materials at high strain rate loading Along Material Carbon plain weave Carbon satin weave Hybrid E-glass plain weave Carbon plain weave Carbon satin weave Hybrid E-glass plain weave Carbon plain weave Carbon satin weave Hybrid E-glass plain weave Strain rate, (s1 ) 1000 1000 1000 1000 1000 1000 1000 1000 1503 1503 1503 1503

hybrid composite and satin weave carbon/epoxy under compressive loading using SHPB apparatus along warp, ll and thickness directions. Studies are also presented for plain weave E-glass/epoxy and plain weave carbon/epoxy. Studies were carried out in the strain-rate range of 5461503 s1 . Strain rates used, compressive properties and overall bre volume fraction are given in Tables 13 and Figs. 321. 6.1. Satin weave carbonplain weave E-glass with epoxy hybrid composite Strain gauge signals obtained on oscilloscope during testing along warp are presented in Fig. 3(a). The durations of incident and reected signals are represented by a1 a2 and a3 a4 , respectively. Here, P represents a point on the signal at the end of rise time. In the present case, a1 a2 = 200 s. It may be noted that the durations of incident and reected signals are nearly the same. Force versus time plots are obtained from the strain gauge signals and are presented in Fig. 3(b). Force history on the incident bar is plotted based on strain gauge signals I + R, whereas force history on

Strength, 301 317 325 279 313 339 347 278 532 621 680 593

ult

(MPa)

Ultimate strain, ult (%) 1.45 1.57 1.80 1.72 1.56 1.49 1.51 1.61 4.71 2.71 2.06 4.91

Modulus, E (GPa) 22.4 23.2 22.0 21.4 23.3 25.6 23.0 19.8 12.0 11.3 13.6 8.9

Warp

Fill

Thickness

91

Fig. 3. Compressive SHPB test results for satin weave carbonplain weave E-glass and epoxy hybrid composite, along warp, ds = 8 mm, ls = 6 mm, (a) strain gauge signals and (b) comparison of force vs. time behaviour, derived from strain gauge signals.

the transmitter bar is plotted based on strain gauge signal T. The force history obtained based on signals I + R is referred to as F1 and the force history obtained based on signal T is referred to as F2 for further discussion. Force F1 would be acting on the interface between the incident bar and the specimen whereas force F2 would be acting on the interface between the transmitter bar and the specimen. Force versus time plots can be subdivided into two regions: one, until the peak force is reached; two, after the peak force is reached. The region two indicates the behaviour after the specimen has failed. Obviously, in this region, F1 and F2 would be different. The forces F1 and F2 are not matching in region one also. The magnitude of peak force F2 is about 13% less than the magnitude of peak force F1 . This is because of stress wave attenuation within the woven fabric composites. Details about stress wave attenuation in woven fabric composites are presented in Appendix C. It may be noted that the peak force is attained at time duration of 27 s in the case of F1 . During this period 16 number of transits take place. One transit equals to time required for a pulse to travel from one end of the specimen to another end of the specimen during testing. Since F1 and F2 are not equal, the specimen would not be under uniform stress during the loading process. The compressive properties reported further are based on F2 . This is to obtain conservative estimates of compressive properties. Time versus strain rate, strain and stress plots along warp are given in Fig. 4. These plots are obtained based on strain gauge signals and Eqs. (1)(3). As indicated earlier, point P indicates the end of the rise time. Point A indicates rst peak strain rate whereas point

Fig. 4. High strain rate compressive test results for satin weave carbonplain weave E-glass and epoxy hybrid composite, along warp, ds = 8 mm, ls = 6 mm, (a) time vs. strain rate plot, (b) time vs. strain plot and (c) time vs. stress plot.

B indicates peak stress. Point C indicates overall peak strain rate. It may be noted that, the specimen has failed at a time duration corresponding to point B. Strain rate effects on the compressive properties of the hybrid composite are presented in Table 1 and Fig. 21. The strain rates indicated here are with respect to point A. The values given in round bracket are with respect to point B whereas the values indicated in square bracket are with respect to point C. The negative sign for strain rates in the tables indicates that the strain is compressive whereas the positive sign for strain rates

Fig. 5. Stress vs. strain plot from high strain rate compressive test on SHPB, along warp, ds = 8 mm, ls = 6 mm, satin weave carbonplain weave E-glass and epoxy hybrid composite, = 1184 s1 .

92

Fig. 6. Compressive SHPB test results for satin weave carbonplain weave E-glass and epoxy hybrid composite, along thickness, ds = 8 mm, ls = 6 mm, (a) strain-gauge signals and (b) comparison of force vs. time behaviour, derived from strain-gauge signals. Fig. 7. High strain rate compressive test results for satin weave carbonplain weave E-glass and epoxy hybrid composite, along thickness, ds = 8 mm, ls = 6 mm, (a) time vs. strain-rate plot, (b) time vs. strain plot and (c) time vs. stress plot.

indicates that the strain is tensile. In the present case, the strain rate at point A is higher than at point B. The strain rate at point A is taken as reference strain rate. A typical stressstrain plot along warp is presented in Fig. 5. The stressstrain plot can be subdivided into two regions. Region one represents the behaviour of the material until the compressive strength is reached. Region two represents post-failure behaviour of the material. Figs. 3(a) and 4(a) show the rise time of 17 s. During this period, which is the initial stage of loading, the strain rate is not constant. Hence, the Youngs modulus obtained based on the strain gauge data during this period would not be exact. It can be seen, from Fig. 3(a), that the strain rate variation is relatively less beyond the rise time. The stressstrain behaviour obtained beyond the rise time of the pulse can be considered to be the actual behaviour of the material. The end of rise time is represented by point P. The stressstrain plot, as given in Fig. 5, can be considered to represent the actual behaviour of the material beyond the point P. As a rst approximation, by joining point P to the origin, Youngs modulus can be determined. This would indicate the lower bound of Youngs modulus. Youngs modulus can also be found by extrapolating the stressstrain curve in region one from point P to the origin by a smooth curve. From Table 1 and Fig. 21, it can be observed that the compressive strength is enhanced with increasing strain rate. Experimental studies were carried out along ll direction also (Table 1 and Fig. 21). The compressive properties are nearly the same as that along warp. This is because the fabrics used were nearly balanced. Experimental studies were also carried out along thickness direction. Strain gauge signals obtained on oscilloscope during test-

ing along thickness are presented in Fig. 6(a). In the present case, a1 a2 = 175 s. It may be noted that the duration of incident and reected signals are nearly the same. Force versus time plots are presented in Fig. 6(b). The forces F1 and F2 are not matching. It may be noted that the peak force is attained at a time duration of 27 s in the case of F1 . During this period 11 number of transits take place. Since F1 and F2 are not equal, the specimen would not be under uniform stress during the loading process. The compressive properties reported further are based on F2 . Time versus strain rate, strain and stress plots along thickness are given in Fig. 7. It is interesting to note that the strain rate changes from negative to positive during loading. This is possibly because

Fig. 8. Stress vs. strain plot from high strain rate compressive test on SHPB, along thickness, ds = 8 mm, ls = 6 mm, satin weave carbonplain weave E-glass and epoxy hybrid composite, = 1503 s1 .

93

Fig. 9. Compressive SHPB test results for satin weave carbon/epoxy, along warp, ds = 8 mm, ls = 6 mm, (a) strain gauge signals and (b) comparison of force vs. time behaviour, derived from strain gauge signals.

Fig. 10. High strain rate compressive test results for satin weave carbon/epoxy, along warp, ds = 8 mm, ls = 6 mm, (a) time vs. strain-rate plot, (b) time vs. strain plot and (c) time vs. stress plot.

6.2. Satin weave carbon/epoxy the specimen has not fully failed during loading and some strain energy is stored within the specimen during high strain rate loading along thickness direction. This phenomenon is explained later with the help of Fig. 22. As indicated earlier, point P indicates the end of the rise time. Point A indicates rst peak negative strain rate whereas point B indicates peak compressive stress. Point D indicates change of strain rate from negative to positive whereas point E indicates peak positive strain rate. Point C indicates overall peak strain rate. It may be noted that, the specimen has failed at a time duration corresponding to point B. Strain rate effects on the compressive properties of the hybrid composite along thickness are presented in Table 1 and Fig. 21. The strain rates indicated here are with respect to point A. The values given in round bracket are with respect to point B whereas the values indicated in square bracket are with respect to point C. A typical stressstrain plot along thickness is presented in Fig. 8. This behaviour is explained later with the help of Fig. 22. The energy input would be absorbed by the specimen partly in the form of fracture energy and partly in the form elastic strain energy stored within the specimen. During the later part of compressive loading, the elastic strain energy stored within the specimen would be released. As a result of this, the extent of compressive strain within the specimen would be decreased. During this phase the failure of the specimen would take place. Figs. 6(a) and 7(a) show the rise time of 17 s. During this period, which is the initial stage of loading, the strain rate is not constant. From Table 1, it can be observed that the compressive strength is enhanced with increasing strain rate. The experimental studies were carried out on satin weave carbon/epoxy along warp, ll and thickness directions. The results are presented in Table 2 and Figs. 914 and 21. The qualitative behaviour for satin weave carbon/epoxy and satin weave carbonplain weave E-glass with epoxy hybrid composite is nearly the same. The Figs. 911 are for loading along warp direction. In this case, a1 a2 = 190 s (Fig. 9a). The durations of incident and reected signals are nearly the same. Force versus time plots are obtained from the strain gauge signals and are presented in Fig. 9(b). The peak force is attained at time duration of 22 s in the case of F1 . During this period 13 number of transits take place. The compressive properties reported further are based on F2 .

Fig. 11. Stress vs. strain plot from high strain rate compressive test on SHPB, along warp, ds = 8 mm, ls = 6 mm, satin weave carbon/epoxy, = 1139 s1 .

94

Fig. 12. Compressive SHPB test results for satin weave carbon/epoxy, along thickness, ds = 8 mm, ls = 6 mm, (a) strain gauge signals and (b) comparison of force vs. time behaviour, derived from strain gauge signals.

Time versus strain rate, strain and stress plots along warp are given in Fig. 10. Strain rate effects on the compressive properties of the satin weave carbon/epoxy are presented in Table 2 and Fig. 21. The strain rates indicated here are with respect to point A. A typical stressstrain plot along warp is presented in Fig. 11. Figs. 9(a) and 10(a) show the rise time of 18 s. From Table 2, it can be observed that the compressive strength is enhanced with increasing strain rate initially, and then, it decreases marginally. Experimental studies were carried out along ll direction also (Table 2 and Fig. 21). The qualitative behaviour of compressive properties is nearly the same as that along warp. This is because the fabrics used were nearly balanced. Experimental studies were also carried out along thickness direction. In this case, a1 a2 = 185 s (Fig. 12a). The durations of incident and reected signals are nearly the same. Force versus time plots are presented in Fig. 12(b). The forces F1 and F2 are not matching. The peak force is attained at time duration of 19 s in the case of F1 . During this period 8 number of transits take place. The compressive properties reported further are based on F2 . Time versus strain rate, strain and stress plots along warp are given in Fig. 13. The behaviour is similar to that of hybrid composite. A typical stressstrain plot along thickness is presented in Fig. 14. The behaviour is similar to that of hybrid composite. Figs. 12(a) and 13(a) show the rise time of 18 s. During this period, which is the initial stage of loading, the strain rate is not constant. 6.3. Plain weave E-glass/epoxy The experimental studies were carried out on plain weave Eglass/epoxy along warp, ll and thickness directions. The results are presented in Figs. 1521.

Fig. 13. High strain rate compressive test results for satin weave carbon/epoxy, along thickness, ds = 8 mm, ls = 6 mm, (a) time vs. strain-rate plot, (b) time vs. strain plot and (c) time vs. stress plot.

The Figs. 1517 are for loading along warp direction. In this case, a1 a2 = 182 s (Fig. 15a). The durations of incident and reected signals are nearly the same. Force versus time plots are obtained from the strain gauge signals and are presented in Fig. 15(b). The peak force is attained at time duration of 27 s in the case of F1 . During this period 13 number of transits take place. The compressive properties reported further are based on F2 . Time versus strain rate, strain and stress plots along warp are given in Fig. 16. Strain rate effects on the compressive properties of the plain weave E-glass/epoxy are presented in Fig. 21. The strain rates indicated here are with respect to point A. A typical stressstrain plot along warp is presented in Fig. 17. Figs. 15(a) and 16(a) show the rise time of 16 s.

Fig. 14. Stress vs. strain plot from high strain rate compressive test on SHPB, along thickness, ds = 8 mm, ls = 6 mm, satin weave carbon/epoxy, = 1503 s1 .

95

Fig. 15. Compressive SHPB test results for plain weave E-glass/epoxy, along warp, ds = 8 mm, ls = 6 mm, (a) strain gauge signals and (b) comparison of force vs. time behaviour, derived from strain gauge signals. Fig. 16. High strain rate compressive test results for plain weave E-glass/epoxy, along warp, ds = 8 mm, ls = 6 mm, (a) time vs. strain-rate plot, (b) time vs. strain plot and (c) time vs. stress plot.

From Fig. 21, it can be observed that the compressive strength is enhanced with increasing strain rate. Experimental studies were carried out along ll direction also (Fig. 21). The compressive properties are nearly the same as that along warp. This is because the fabrics used were nearly balanced. Experimental studies were also carried out along thickness direction. In this case, a1 a2 = 185 s (Fig. 18a). The durations of incident and reected signals are nearly the same. Force versus time plots are presented in Fig. 18(b). The forces F1 and F2 are not matching. The peak force is attained at time duration of 15 s in the case of F1 . During this period 8 number of transits take place. The compressive properties reported further are based on F2 . Time versus strain rate, strain and stress plots along thickness are given in Fig. 19. It may be noted that, in this case, strain rate does not change sign during loading. A typical stressstrain plot along thickness is presented in Fig. 20. In this case, the input energy is primarily absorber as fracture energy by the specimen. Figs. 18(a) and 19(a) show the rise time of 26 s. During this period, which is the initial stage of loading, the strain rate is not constant. Further details about high strain-rate behaviour of plain weave E-glass/epoxy are available in Ref. [53]. 6.4. Comparison of test results Schematic of strain state in the specimen during high strain rate compressive loading along thickness direction is presented in Fig. 22. As the striker bar hits the incident bar, energy input

into the system is absorbed in the form of fracture energy during the failure process of the specimen. The schematic of strain state for plain weave E-glass/epoxy is shown in Fig. 22(a). This can be seen in Figs. 19 and 20. The strain rate is compressive throughout. For the cases of satin weave carbonplain weave E-glass with epoxy hybrid composite and satin weave carbon/epoxy, the specimens did not fail fully. In other words, some elastic energy was also stored in the specimen during the loading process. This behaviour

Fig. 17. Stress vs. strain plot from high strain rate compressive test on SHPB, along warp, ds = 8 mm, ls = 6 mm, plain weave E-glass/epoxy, = 880 s1 .

96

Fig. 18. Compressive SHPB test results for plain weave E-glass/epoxy, along thickness, ds = 8 mm, ls = 6 mm, (a) strain gauge signals and (b) comparison of force vs. time behaviour, derived from strain gauge signals.

is typical of compressive loading along thickness direction. As the specimen is not fully failing during loading, the stored elastic strain energy in the specimen would be released during the later part of the loading process. During this stage, the strain rate would be positive. The schematic of strain state for satin weave carbonplain weave E-glass and epoxy hybrid composite and satin weave carbon/epoxy is shown in Fig. 22(b). As the stored elastic strain energy in the specimen is released, the specimen would expand and exert pressure on both the incident and transmitter bars. Under such a condition the specimen can be considered as a second striker bar. The pulses generated due to this would be superimposed on the earlier transmitted and reected pulses. Since the specimen has nearly failed, the resulting strain is compressive as can be seen from Figs. 7, 8, 13 and 14. As the specimens had nearly failed, in this case, the peak stresses generated are taken as compressive strengths. The compressive properties under high strain rate loading are presented in Tables 13 and Fig. 21 in a consolidated form. For comparison, results for plain weave carbon/epoxy are also presented. It is generally observed that the compressive strength increases with increasing strain rate at high strain rate loading initially, and then decreases marginally. The ultimate strain also increases with increasing strain rate initially, and then decreases marginally. There is no signicant change in compressive modulus. The compressive strength under high strain rate loading is higher for satin weave carbonplain weave E-glass with epoxy hybrid composite compared with satin weave carbon/epoxy. For hybrid composite, the ultimate strain increases considerably compared with that for satin weave carbon/epoxy. It is generally

Fig. 19. High strain rate compressive test results for plain weave E-glass/epoxy, along thickness, ds = 8 mm, ls = 6 mm, (a) time vs. strain-rate plot, (b) time vs. strain plot and (c) time vs. stress plot.

observed that the high strain rate compressive strength of: (satin weave carbon-plain weave E-glass with epoxy hybrid composite) > (satin weave carbon/epoxy) > (plain weave carbon/epoxy) > (plain weave E-glass/epoxy). For satin weave carbonplain weave E-glass with epoxy hybrid composite, the ultimate strain increases considerably compared

Fig. 20. Stress vs. strain plot from high strain rate compressive test on SHPB, along thickness, ds = 8 mm, ls = 6 mm, plain weave E-glass/epoxy, = 1300 s1 .

97

(i) Generally, the compressive strength increases with increasing strain rate at high strain rate loading. (ii) Generally, the ultimate strain also increases with increasing strain rate. There is no signicant change in compressive modulus. (iii) The compressive strength under high strain rate loading is higher for satin weave carbonplain weave E-glass with epoxy hybrid composite than for satin weave carbon/epoxy. Further, It is generally observed that the high strain rate compressive strength of: (satin weave carbon-plain weave E-glass with epoxy hybrid composite) > (satin weave carbon/epoxy) > (plain weave carbon/epoxy) > (plain weave E-glass/epoxy). (iv) For satin weave carbonplain weave E-glass with epoxy hybrid composite, the ultimate strain increases considerably compared with that for the other cases. (v) The compressive strength is signicantly higher along thickness direction compared with that at along warp or ll.
Fig. 21. Strain rate vs. compressive strength plots for various composite materials, (a) along warp and (b) along ll.

Acknowledgement This work was partially supported by the Structures Panel, Aeronautics Research & Development Board, Ministry of Defense, Government of India, Grant No. DARO/08/1051302/M/I. Appendix A. Controlling pulse duration and shape during SHPB testing The pulse duration is directly proportional to the length of the striker bar. It is the time required for the pulse to travel twice the length of the striker bar. The duration of the pulse, t= 2L1 c0 (A.1)

Fig. 22. Schematic of strain state in the specimen during high strain rate compressive test on SHPB, along thickness, (a) specimen failed and (b) specimen not fully failed.

with the other cases. The compressive strength is signicantly higher along thickness direction compared with that at along warp or ll.

where t is pulse duration, L1 is length of the striker bar and c0 is elastic wave velocity in the striker bar. Normally, the diameter of incident, transmitter and striker bars is the same. During controlling the pulse duration, precaution has to be taken on the maximum limit of the length of the striker bar. Generally, the strain gauges are placed at the centers of the incident and transmitter bars. The length of the striker bar must be less than half the length of the incident/transmitter bar. This ensures that the overlap of the pulses does not take place. To increase the rise time, smoothen the pulse and to modify the pulse shape, a pulse shaper with different dimensions is used. In this study, copper, brass and aluminum were used for making the pulse shapers. The diameter of the pulse shaper was marginally more than the diameter of the elastic bars whereas the thickness was varied in the range of 0.53 mm. Appendix B. Controlling strain rate during SHPB testing

7. Conclusions The compressive properties of a typical satin weave carbon plain weave E-glass with epoxy hybrid composite, satin weave carbon/epoxy, plain weave carbon/epoxy and plain weave Eglass/epoxy are presented under high strain rate loading. The properties are presented with respect to all the three principal directions, i.e., along warp, ll and thickness. The strain rate in the specimen is obtained using Eq. (1) as given below: S (t) = 2CO lS R (t)

The strain rate depends on: elastic wave velocity in the bars, length of the specimen and reected wave pulse.

98

Generally, the conguration of SHPB and the materials for different components, especially for incident and transmitter bars, are worked out based on the range of applications visualized. On a specic SHPB apparatus, in order to achieve higher strain rates, either larger amplitude input pulses or shorter specimens can be used. The amplitude of input pulses is limited by the yield strength of the incident bar or by the amount of force that can be generated by the particular experimental apparatus. The length of the specimen is limited by L/D ratio suggested for SHPB testing. The possibility of controlling strain rate by varying the gauge length of the specimen is marginal. The best way to control the strain rate is by governing the amount of force applied on the incident bar. As it can be seen form Eq. (1), the strain rate is directly proportional to the amplitude of the reected strain pulse. It may be noted that the reected strain pulse is related to the amplitude of the incident strain pulse, which in turn, is related to the amount of force applied on the incident bar. By varying the force applied on the incident bar, the amplitude of the reected strain pulse, and hence, the strain rate on the specimen can be controlled. The force exerted by the striker bar is controlled by varying the velocity of the strike bar. Caution has to be taken during testing at lower strain rates. At lower strain rates the specimen needs more time to reach to higher strains. Higher strains can be obtained by increasing the duration of the pulse. The procedure for controlling the duration of the pulse during testing is given Appendix A. Appendix C. Stress wave attenuation studies in woven fabric composites A woven fabric is produced by the process of weaving in which the fabric is formed by interlacing warp and ll stands/yarns. In the case of plain weave, warp and ll stands/yarns are interlaced in a regular sequence of one under and one over. When the striker bar hits the incident bar, an elastic stress wave pulse is generated. It travels trough the incident bar, specimen and the transmitter bar. As the stress wave pulse passes through the specimen, it would encounter many warp-ll, warp-resin and ll-resin boundaries. As it encounters the boundaries, the incident wave would partly transmit and partly reect. This phenomenon would take place at every boundary leading to stress wave attenuation as the wave travels from the interface between the incident bar and the specimen to the interface between the transmitter bar and the specimen. In a similar way, stress wave attenuation would take place when stress wave travels along thickness direction.

An independent analytical study was carried out to investigate stress wave attenuation behaviour of plain weave fabric composites under transverse ballistic impact. The transverse ballistic impact was at the center of the plate. Based on the ballistic impact analysis, intensity of in-plane radial and through the thickness stress waves was determined. Fig. 23 presents stress wave attenuation in a typical plain weave fabric lamina under ballistic impact unit step loading for plate dimensions of 400 mm 400 mm at time interval of 40 s. At this time interval, the wave has traveled up to a distance of 110 mm on either side of point of impact. At this distance the magnitude of stress wave is nearly zero. In other words, the stress wave has attenuated and its intensity is nearly zero at a distance of 110 mm and beyond, up to the boundary of the plate. Experimental studies were also carried out to determine strain magnitudes along in-plane radial direction from the point of impact. The strain was the maximum near the point of impact and was decreasing along the radial direction while moving away from the point of impact. At a distance of 110 mm and beyond, the strain was nearly zero. As presented in Section 6, the magnitude of force F2 was less than the magnitude of force F1 . This can be explained based on stress wave attenuation in plain weave fabric composites. References
[1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] A.R. Bunsell, B. Harris, Composites 5 (1974) 157164. J. Aveston, J.M. Sillwood, Journal of Materials Science 11 (1976) 18771883. P.W. Manders, M.G. Bader, Journal of Materials Science 16 (1981) 22462256. H. Fukuda, Journal of Materials Science 19 (1983) 974982. S.J. Fariborz, D.G. Harlow, Journal of Composite Materials 21 (1987) 856 875. K. Saka, J. Harding, Composites 21 (1990) 439447. Q.D. Zeng, Composites Science and Technology 52 (1994) 481487. S.C. Khatri, M.J. Koczak, Composites Science and Technology 56 (1996) 181192. Z. Qingdum, L. Xuehui, Acta Mechanica Solida Sinica 16 (2003) 3340. M.R. Piggott, B. Harris, Journal of Materials Science 16 (1981) 687693. S.F. Hwang, C.P. Mao, Composites Science and Technology 61 (2001) 15131527. D.F. Adams, A.K. Miller, Journal of Materials Science 11 (1976) 16971710. D.W. Woods, P.J. Hine, I.M. Ward, Composites Science and Technology 52 (1994) 397405. R. Park, J. Jang, Composites Science and Technology 58 (1998) 16211628. S.S. Cheon, T.S. Lim, D.G. Lee, Composite Structures 46 (1999) 267278. N.K. Naik, R. Ramasimha, H. Arya, S.V. Prabhu, N. ShamaRao, Composites: Part B 32 (2001) 565574. C. Thanomsilp, P.J. Hogg, Composites Science and Technology 63 (2003) 467482. S.C. Khatri, M.J. Koczak, Composites Science and Technology 56 (1996) 473482. H. Harel, J. Aronhime, K. Schulte, K. Friedrich, G. Marom, Journal of Materials Science 25 (1990) 13131317. G. Fernando, R.F. Dickson, H. Reiter, B. Harris, Journal of Materials Science 23 (1988) 37323743. E. Woldesenbet, J.R. Vinson, AIAA Journal 37 (1999) 11021106. W.E. Jahsman, Journal of Applied Mechanics (1971) 7582. L.M. Yang, V.P.M. Shim, International Journal of Impact Engineering 31 (2005) 129150. E.D.H. Davies, S.C. Hunter, Journal of the Mechanics and Physics of Solids 11 (1963) 155184. B. Song, W. Chen, Experimental Mechanics 44 (2004) 622627. W. Chen, F. Lu, B. Zhou, Experimental Mechanics 40 (2000) 16. L. Ninan, J. Tsai, C.T. Sun, International Journal of Impact Engineering 25 (2001) 291313. B. Song, W. Chen, T. Weerasooriya, Journal of Composite Materials 37 (2003) 17231743. A. Jadhav, E. Woldesenbet, S.S. Pang, Composites Part B 34 (2003) 339346. D.J. Frew, M.J. Forrestal, W. Chen, Experimental Mechanics 45 (2005) 186195. S.T. Jenq, S.L. Sheu, Composite Structures 25 (1993) 427438. K. Oguni, G. Ravichandran, Journal of Materials Science 36 (2001) 831838. M. Vural, G. Ravichandran, Journal of Composite Materials 38 (2004) 609623. A. Haque, M. Ali, Journal of Composite Materials 39 (2005) 423450. J. Harding, Composites 24 (1993) 323332. B.A. Gama, J.W. Gillespie, H. Mahfuz, R.P. Raines, A. Haque, S. Jeelani, T.A. Bogetti, B. Fink, Journal of Composite Materials 35 (2001) 12011228. M.V. Hosur, A. Abraham, S. Jeelani, U.K. Vaidya, Journal of Composite Materials 35 (2001) 11111133. H.M. Hsiao, I.M. Daniel, Composites Part B 29B (1998) 521533. H.M. Hsiao, I.M. Daniel, R.D. Cordes, Experimental Mechanics 38 (1998) 172180.

Fig. 23. Stress wave attenuation in plain weave fabric lamina under high velocity impact unit step loading for plate dimensions of 400 mm 400 mm, at time interval of 40 s.

[38] [39]

99 [40] H.M. Hsiao, I.M. Daniel, R.D. Cordes, Journal of Composite Materials 33 (1999) 16201642. [41] Z. Li, J. Lambros, Composites Science and Technology 59 (1999) 10971107. [42] J.R. Vinson, E. Woldesenbet, Journal of Composite Materials 35 (2001) 509521. [43] M.V. Hosur, J. Alexander, U.K. Vaidya, S. Jeelani, Composite Structures 52 (2001) 405417. [44] J.P. Hou, C. Ruiz, Composites Science and Technology 60 (2000) 2829 2834. [45] M.V. Hosur, M. Adya, U.K. Vaidya, A. Mayer, S. Jeelani, Composite Structures 59 (2003) 507523. [46] M.V. Hosur, M. Adya, S. Jeelani, U.K. Vaidya, P.K. Dutta, Journal of Reinforced Plastics and Composites 23 (2004) 491514. [47] M.V. Hosur, J. Alexander, U.K. Vaidya, S. Jeelani, A. Mayer, Composite Structures 63 (2004) 7585. [48] R.L. Sierakowski, Journal of Applied Mechanics 50 (1997) 741761. [49] B.A. Gama, S.L. Lopatnikov, J.W. Gillespie Jr., Applied Mechanics Review 57 (2004) 223249. [50] G.C. Jacob, J.M. Starbuck, J.F. Fellers, S. Simunovic, R.G. Boeman, Journal of Applied Ploymer Science 94 (2004) 296301. [51] R.L. Sierakowski, S.K. Chaturvedi, Dynamic Loading and Characterization of Fiber Reinforced Composites, A Wiley-Interscience Publication, New York, 1997. [52] H. Kuhn, D. Medlin (Eds.), Mechanical Testing and Evaluation, 8, ASM International, Materials Park, 2000, pp. 427529. [53] N.K. Naik, Venkateswara Rao Kavala, Materials Science and Engineering A 474 (2008) 301311.