Polymer Testing 31 (2012) 261–266

Contents lists available at SciVerse ScienceDirect

Polymer Testing
journal homepage: www.elsevier.com/locate/polytest

Material behaviour

Stress wave attenuation in composites during ballistic impact

Kedar S. Pandya, Lokesh Dharmane, Jayaram R. Pothnis, G. Ravikumar, N.K. Naik*
Aerospace Engineering Department, Indian Institute of Technology Bombay, Powai, Mumbai 400 076, India

a r t i c l e i n f o a b s t r a c t

Article history: Experimental studies are presented on stress wave attenuation during ballistic impact for
Received 15 October 2011 four types of polymer matrix composites. The materials considered are plain weave E-
Accepted 26 November 2011 glass/epoxy, 8H satin weave T300 carbon/epoxy and two types of hybrid composite made
using plain weave E-glass fabric and 8H satin weave T300 carbon fabric with epoxy resin.
Keywords: Strain profiles were obtained during ballistic impact event at certain distances from the
Stress wave attenuation
point of impact. There is stress wave attenuation leading to reduction in peak strains
Strain profile
obtained as the stress wave propagates away from the point of impact. Further, it is
Polymer matrix composite
Hybrid composite observed that ballistic limit velocity, V50, can be increased compared to carbon only
Ballistic impact composites by adding E-glass layers to T300 carbon layers.
Ó 2011 Elsevier Ltd. All rights reserved.

1. Introduction thickness, lay-up sequence, geometry and boundary

Polymer matrix composites are used in high perfor-
mance structural applications as well as in day-to-day
applications because of their excellent mechanical prop-
erties such as high specific stiffness, high specific strength,
high resistance to corrosion and increased fatigue life.
Structural components undergo different loading condi-
tions during their service life and impact loading is one of
the critical conditions. Hence, resistance to penetration/
perforation is one of the critical requirements of structural
components. Polymer matrix composites are generally
susceptible to impact loading and, for their effective use in
high performance applications, their behaviour under such
loading conditions should be fully understood.
During an impact event, different types of waves prop-
agate in the impacted bodies depending on how the motion
of the particles of the body is related to the direction of
propagation of the waves and the boundary conditions. The
most common types of waves are given in [1].
Many factors affect the response of composites to
impact loading, including type of fiber and matrix, laminate
* Corresponding author. Tel.: þ91 22 2576 7114; fax: þ91 22 2572 2602.
E-mail address: nknaik@aero.iitb.ac.in (N.K. Naik).
conditions. Size, shape and kinetic energy of the projectile
also influence the ballistic impact response. Experimental
studies are available in literature on the ballistic impact
behaviour of polymer matrix composites covering various
aspects, such as effect of target thickness [2,3], form of
reinforcement of target [4–9], damage due to impact
loading [10–15], post-impact mechanical behaviour [16,17]
and penetration/perforation [18–26]. Since the test condi-
tions and the target materials are different in those studies,
the results cannot be compared.
The studies available are generally on penetration/
perforation, effect of target material and geometry, damage
due to impact loading, post-impact mechanical behaviour
and determination of ballistic limit velocity. It would be
interesting to note stress wave attenuation during ballistic
impact which can be represented in terms of strain profiles.
Hence, in the current work, experimental investigations
were carried out to obtain strain profiles as a function of
time and distance from the point of impact. The materials
used for the studies were plain weave E-glass/epoxy, 8H
satin weave T300 carbon/epoxy and two types of hybrid
composite made using plain weave E-glass fabric, 8H satin
weave T-300 carbon fabric and epoxy resin. Strain profiles
are provided at certain distances from the point of impact.

0142-9418/$ – see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.polymertesting.2011.11.006
262 K.S. Pandya et al. / Polymer Testing 31 (2012) 261–266
2. Experimental studies
2.1. Experimental facility
Experimental studies were carried out on a single stage
gas gun operated ballistic impact test apparatus. A photo-
graph of the ballistic impact test apparatus used is shown in
Fig. 1. The apparatus consists of a projectile propelling
mechanism, chronograph for velocity measurement,
support stand for holding the specimens, containment
chamber, safety devices and strain measuring facility.
A single stage compressed gas gun was used with air as
the propelling gas. Compressed gas cylinder, barrel for
propelling and guiding the projectile, quick release valve
and a nitrogen gas based solenoid valve to operate the
quick release valve are the components of the propelling
mechanism. The barrel was 1.5 m long with its inner
diameter governed by the diameter of the projectile used.
Projectile speed was varied up to 200 m/s by changing the
air pressure in the cylinder.
A hollow shield was used from the tip of the barrel up to
the impact chamber in order to enclose the projectile after
propulsion. This was used for protecting the user of the
ballistic impact test apparatus from accidental stray firings.
The center lines of the barrel, hollow shield, chronograph
and the center of the specimen after clamping in the holder
were aligned.
Experimental studies were carried out on flat speci-
mens of 125 mm  125 mm with thickness of 3 mm. The
specimen dimensions are governed by the specimen
holding device which is part of the ballistic impact test
apparatus. The mass of the hardened steel projectile was
6.42 g and its diameter was 6.36 mm. Experimental studies
were carried out on at least 3 specimens for each impact
condition.
Fig. 1. Photograph of a typical ballistic impact test apparatus.
2.2. Planning for experiments
The study of stress wave attenuation during ballistic
impact of polymer matrix composites is a relatively unex-
plored field of research as evident from the available liter-
ature. Hence, experiments were planned in order to study
the stress wave attenuation of polymer matrix composites
in terms of strain profiles as a function of time and distance
from the point of impact.
2.3. Specimen details
Two types of hybrid composites, H4 and H5 were made
using 8H satin weave T300 carbon fabric and plain weave E-
glass fabric with epoxy resin. Specifications of tows/
strands, fabrics, resin and composites for plain weave E-
glass/epoxy and 8H satin weave T300 carbon/epoxy are
presented in the Appendix.
The lay-up sequences used were: For hybrid H4, [C2G2]S
For hybrid H5, [G2C2]S
Here, G represents plain weave E-glass fabric whereas C
represents 8H satin weave T300 carbon fabric.
For comparison, 8H satin weave T300 carbon/epoxy and
plain weave E-glass/epoxy composites were also studied.
 Fiber volume fractions: For hybrids H4 and H5, Vf ¼ 0.53
For plain weave E-glass/epoxy, Vf ¼ 0.51.
For 8H satin weave T300 carbon/epoxy, Vf ¼ 0.56
 Composite volume fraction
Composite volume fraction is defined as the ratio of
volume of one of the constituents of the hybrid composite
to the volume of the hybrid composite.
For hybrids H4 and H5: T300 carbon composite volume
fraction, Vc ¼ 0.55.
E-glass composite volume fraction, Vg ¼ 0.45
 Volume fraction of a particular type of reinforcement as
a part of volume of total reinforcement
This is defined as the ratio of volume of one type of
reinforcement to the volume of total reinforcement of the
hybrid composite.
For hybrids H4 and H5: T300 carbon fiber volume
fraction, Vcf ¼ 0.57.
E-glass fiber volume fraction, Vgf ¼ 0.43.
Fig. 2 presents the schematic arrangement of a test
specimen and locations of centres of strain gauges. Fig. 3
presents a photograph of a test specimen with strain
gauges mounted on it.
2.4. Test procedure, data acquisition and analysis
Strain gauges of type FLA-6-11-3LT with gauge factor of
2.11, resistance 120 U and gauge length of 6 mm were used
together with a Wheatstone circuit in quarter bridge
configuration. Data from the strain gauges was routed
through DC-96A dynamic strain meters and a TDS 1002
 K.S. Pandya et al. / Polymer Testing 31 (2012) 261–266 263

S3 S1 S2
125

15 30

50

125

Fig. 2. Schematic arrangement of a specimen tested, distances in mm; S1,

S2, S3 indicate locations of centers of strain gauges.
Fig. 3. Photograph of a typical specimen tested.

dual channel oscilloscope to a PC for further processing.

The strain gauges and strain meters were supplied by M/s The strain profile at location S1 is represented by the
Tokyo Sokki Kenkyujo Co., Ltd., Japan. The oscilloscope was curve a1b1 during the ballistic impact event (Fig. 4b). As
procured from M/s Tektronix Inc., USA. the time progresses, the strain increases from a1 to b1.
Strain gauges were mounted on the specimen at
distances of 15 mm (strain gauge S1), 30 mm (strain gauge
S2) and 50 mm (strain gauge S3) away from the point of
impact and the corresponding strains were obtained.
Strain profiles were obtained as a function of time and
distance from the point of impact using the strain
measuring facility.

3. Results and discussion

Strain profiles obtained on the oscilloscope during

ballistic impact for plain weave E-glass/epoxy are pre-
sented in Fig. 4. For this case, V50 ¼ 98 m/s. Strain as
a function of time at d ¼ 15 mm is presented in Fig. 4a. In
this figure, the curve AB indicates the strain profile at strain
gauge S1. The point B shows the peak strain during pene-
tration/perforation. It is observed that the value of strain
decreases beyond point B. This is due to energy dissipation
as a result of multiple wave reflections and transmissions
along the path of wave propagation. The decrease in strain
sensed by the strain gauge would continue beyond point B
until perforation of the target by the projectile takes place.
The strain gauge would also sense the strain in the target
after perforation, but this would represent post-impact
behaviour of the target.
Strain profiles at ballistic limit velocity at distances
15 mm, 30 mm and 50 mm are presented in Fig. 4b. In this
figure, oa1 represents time taken for the stress wave to
travel from the point of impact to the centre of strain gauge
S1. In a similar way, oa2 and oa3 represent the time taken
Fig. 4. Strain profiles obtained on oscilloscope, plain weave E-glass/epoxy,
for the stress wave to travel from the point of impact to the h ¼ 3 mm, dp ¼ 6.36 mm, mp ¼ 6.42 gm, V50 ¼ 98 m/s: (a) strain as
centres of strain gauges S2 and S3, respectively. It can be a function of time, d ¼ 15 mm, (b) at ballistic limit velocity at different
observed that oa1 < oa2 < oa3. locations.
264 K.S. Pandya et al. / Polymer Testing 31 (2012) 261–266
Fig. 5. Target peak strain profiles during ballistic impact event at ballistic
limit velocity, along warp, h ¼ 3 mm, dp ¼ 6.36 mm, mp ¼ 6.42 gm.
Here, b1 represents the peak strain reached at location S1.
In a similar way, peak strains reached at locations S2 and S3
are represented by b2 and b3, respectively. It can be noted
that the peak strain reached at S1, represented by point b1,
is higher than that at S3, represented by point b3. The peak
strain reached at S2, represented by point b2, is in between.
This is because of stress wave attenuation along the path of
propagation. The hierarchy of peak strains reached at the
three strain gauge locations during ballistic impact event at
ballistic limit velocity is as follows:
Peak strain at S1 > Peak strain at S2 > Peak strain at S3
It should be noted that the peak strain is not obtained
instantaneously at location S1. Due to the propagation of
the stress wave away from the point of impact, the
measured strain is distributed over the area covered by the
strain gauge at every time instant. As a result, the strain
gauge S1 yields an average value of the strain distribution
with respect to time. Since the time taken by the stress
wave to propagate over the length of the strain gauge S1 is
finite, the peak strain is not reached instantaneously. This
can be seen from the strain profile represented by a1b1
(Fig. 4b).
Fig. 5 presents target peak strain profiles during ballistic
impact at ballistic limit velocity for the four types of poly-
mer matrix composites tested. As the distance increases
Fig. 6. Strain profile obtained on oscilloscope, 8H satin weave T300 carbon/
epoxy, h ¼ 3 mm, dp ¼ 6.36 mm, mp ¼ 6.42 gm, V50 ¼ 81 m/s, at ballistic
limit velocity at different locations.
Fig. 7. Strain profile obtained on oscilloscope, hybrid H4, h ¼ 3 mm,
dp ¼ 6.36 mm, mp ¼ 6.42 gm, V50 ¼ 84 m/s, at ballistic limit velocity at
different locations.
from the point of impact, the peak strain reached decreases
for all the four materials. This is due to stress wave atten-
uation caused by multiple wave reflections and trans-
missions at the various interfaces along the path of
propagation.
The hierarchy of peak strains obtained for the different
materials studied is as follows:
½Plain weave E  glass=epoxy > ½Hybrid composite H5
> ½Hybrid composite H4
> ½8H satin weave
T300 carbon=epoxy
It should be noted that the hierarchy of ballistic limit
velocities, V50, for the different materials tested is as
follows:
½Plain weave E  glass=epoxy > ½Hybrid composite H5
> ½Hybrid composite H4
> ½8H satin weave
T300 carbon=epoxy
The values of V50 for hybrid composite H5, hybrid
composite H4 and 8H satin weave T300 carbon/epoxy are
87 m/s, 84 m/s and 81 m/s, respectively.
Fig. 8. Strain profile obtained on oscilloscope, hybrid H5, h ¼ 3 mm,
dp ¼ 6.36 mm, mp ¼ 6.42 gm, V50 ¼ 87 m/s, at ballistic limit velocity at
different locations.
 K.S. Pandya et al. / Polymer Testing 31 (2012) 261–266 265

The peak strain readings for the different materials Appendix

studied are taken at their respective ballistic limit velocities
Specifications of tows/strands, fabrics, resin and composites.
with the same target thickness. Since the peak strain values
obtained depend on projectile impact energy and target Property T300 carbon/epoxy E-glass/epoxy
material properties, it can be inferred that the hierarchy for Reinforcement T300 carbon E-glass
ballistic limit velocity, V50 and peak strains obtained Filament diameter (mm) 7 20
would be the same, as observed experimentally. It can be Filament density (gm/cc) 1.76 2.62
observed that the peak strain obtained for plain weave Filaments per tow/strand 3000 182
Tow/strand tex (gm/Km) 203 150
E-glass/epoxy is higher than that for 8H satin weave T300
Type of weave 8H satin weave Plain weave
carbon/epoxy. Since the ballistic limit velocity for plain No. of counts (per cm) 7.16y 12.1y
weave E-glass/epoxy is higher than that for 8H satin weave Crimp (%) 0.1y 0.9y
T300 carbon/epoxy with the same target thickness, Fabric thickness (mm) 0.34 0.28
Fabric areal weight (gm/m2) 312 388
the resulting ballistic impact force is higher in the former
Fiber volume fraction 0.56 0.51
case. Further, since the elastic modulus of plain weave E- Void content (%) 0.5 0.75
glass/epoxy is lower than that of 8H satin weave T300 Matrix Epoxy LY556 with hardener HY951
carbon/epoxy, the peak strain obtained in case of plain Process used Matched die molding
weave E-glass/epoxy is comparatively higher. y
Along both warp and fill.
Carbon layers are in the exterior and glass layers are in
the interior for hybrid composite H4, whereas glass layers
are in the exterior and carbon layers are in the interior for References
hybrid composite H5. It can be inferred that placing E-glass
layers in the exterior and carbon layers in the interior [1] M.A. Meyers, Dynamic Behavior of Materials, Wiley Publications,
provides higher ballistic limit velocity than placing carbon New York, 1994.
[2] E.P. Gellert, S.J. Cimpoeru, R.L. Woodward, A study of the effect of
layers in the exterior and E-glass layers in the interior. target thickness on the ballistic perforation of glass-fibre-reinforced
It can be seen that the ballistic limit velocity, V50, can be plastic composites, International Journal of Impact Engineering 24
increased by adding E-glass layers to T300 carbon layers (5) (2000) 445.
[3] G. Caprino, V. Lopresto, D. Santoro, Ballistic impact behaviour of
compared with only carbon composites. stitched graphite/epoxy laminates, Composites Science and Tech-
Figs. 6,7 and 8 present strain profiles obtained during nology 67 (3–4) (2007) 325.
ballistic impact at ballistic limit velocity for 8H satin weave [4] F. Larsson, Damage tolerance of a stitched carbon/epoxy laminate,
Composites: Part A 28 (11) (1997) 923.
T300 carbon/epoxy, hybrid composite H4 and hybrid
[5] R.A.W. Mines, A.M. Roach, N. Jones, High velocity perforation
composite H5, respectively. The qualitative behaviour as behaviour of polymer composite laminates, International Journal of
given in these figures is similar to the qualitative behaviour Impact Engineering 22 (6) (1999) 561.
for plain weave E-glass/epoxy as given in Fig. 4. [6] A.P. Mouritz, Ballistic impact and explosive blast resistance of
stitched composites, Composites: Part B 32 (5) (2001) 431.
[7] N. Tarim, F. Findik, H. Uzun, Ballistic impact performance of
4. Conclusions composite structures, Composite Structures 56 (1) (2002) 13.
[8] M.V. Hosur, U.K. Vaidya, C. Ulven, S. Jeelani, Performance of stitched/
unstitched woven carbon/epoxy composites under high velocity
Experimental studies have been conducted on stress impact loading, Composite Structures 64 (3–4) (2004) 455.
wave attenuation during ballistic impact for four types of [9] C.R. Cork, P.W. Foster, The ballistic performance of narrow fabrics,
polymer matrix composites. The specific observations are: International Journal of Impact Engineering 34 (3) (2007) 495.
[10] W.J. Cantwell, J. Morton, The influence of varying projectile mass on
the impact response of CFRP, Composite Structures 13 (2) (1989) 101.
 There is stress wave attenuation leading to reduction in [11] W.J. Cantwell, J. Morton, Comparison of low and high velocity
peak strains obtained as the stress wave propagates impact response of CFRP, Composites 20 (6) (1989) 545.
[12] S.W.R. Lee, C.T. Sun, Dynamic penetration of graphite/epoxy lami-
away from the point of impact. nates impacted by a blunt-ended projectile, Composites Science and
 The hierarchy of peak strains obtained for different Technology 49 (4) (1993) 369.
materials studied is as follows: [13] W.L. Cheng, S. Langlie, S. Itoh, High velocity impact of thick compos-
ites, International Journal of Impact Engineering 29 (1–10) (2003) 167.
[14] L.M. Nunes, S. Paciornik, J.R.M. d’Almeida, Evaluation of the
damaged area of glass-fiber-reinforced epoxy-matrix composite
materials submitted to ballistic impacts, Composites Science and
½Plain weave E  glass=epoxy > ½Hybrid composite H5 Technology 64 (1) (2004) 945.
[15] A.R. Boccaccini, S. Atiq, D.N. Boccaccini, I. Dlouhy, C. Kaya, Fracture
> ½Hybrid composite H4 behaviour of mullite fibre reinforced–mullite matrix composites
under quasi-static and ballistic impact loading, Composites Science
> ½8H satin weave
and Technology 65 (2) (2005) 325.
T300 carbon=epoxy [16] E. DeLuca, J. Prifti, W. Betheney, S.C. Chou, Ballistic impact damage
of S2-glass-reinforced plastic structural armor, Composites Science
and Technology 58 (9) (1998) 1453.
[17] J.E.L. da Silva Jr., S. Paciornik, J.R.M. d’Almeida, Determination of the
post-ballistic impact mechanical behavior of a 45 glass–fabric
Acknowledgements composite, Polymer Testing 23 (5) (2004) 599.
[18] M. Colakoglu, O. Soykasap, T. Ozek, Experimental and numerical inves-
This work was partially supported by the Structures tigations on the ballistic performance of polymer matrix composites
used in armor design, Applied Composite Materials 14 (1) (2007) 47.
Panel, Aeronautics Research & Development Board, [19] H.L. Gower, D.S. Cronin, A. Plumtree, Ballistic impact response of
Ministry of Defense, Government of India, Grant No. DARO/ laminated composite panels, International Journal of Impact Engi-
08/1051302/M/I. neering 35 (9) (2008) 1000.
266 K.S. Pandya et al. / Polymer Testing 31 (2012) 261–266
[20] J. Lopez-Puente, R. Zaera, C. Navarro, Experimental and
numerical analysis of normal and oblique ballistic impacts on
thin carbon/epoxy woven laminates, Composites: Part A 39 (2)
(2008) 374.
[21] A.R. Sabet, M.H. Beheshty, H. Rahimi, Experimental study of sharp-
tipped projectile perforation of GFRP plates containing sand filler
under high velocity impact and quasi-static loadings, Polymer
Composites 30 (10) (2009) 1497.
[22] E. Sevkat, B. Liaw, F. Delale, B.B. Raju, A combined experimental and
numerical approach to study ballistic impact response of S2-glass
fiber/toughened epoxy composite beams, Composites Science and
Technology 69 (7–8) (2009) 965.
[23] A.H. Sheikh, P.H. Bull, J.A. Kepler, Behaviour of multiple composite
plates subjected to ballistic impact, Composites Science and Tech-
nology 69 (6) (2009) 704.
[24] G. Zhao, C. Cho, S. Lu, Z. Wang, Experimental study on impact
resistance properties of T300/epoxy composite laminates, Journal of
Composite Materials 44 (7) (2010) 857.
[25] Y. Hirai, H. Hamada, J.K. Kim, Impact response of woven glass-fabric
composites. I. Effect of fibre surface treatment, Composites Science
and Technology 58 (1) (1998) 91.
[26] Y. Hirai, H. Hamada, J.K. Kim, Impact response of woven glass-fabric
composites.II. Effect of temperature, Composites Science and Tech-
nology 58 (1) (1998) 119.