Original Article

Journal of Thermoplastic Composite

Materials
Investigating the behavior 1–22
ª The Author(s) 2018
of silicon-coated Kevlar Reprints and permissions:
sagepub.co.uk/journalsPermissions.nav

fabric under low-velocity DOI: 10.1177/0892705718772875

journals.sagepub.com/home/jtc

impact: An experimental
and numerical study
A Abbaszadeh1, M Yazdani1,2 ,
F Abbasi3,4 and A Rashed1

Abstract
This study reports experimental and numerical behaviors of both dry and silicon-coated
twill-weave Kevlar fabrics under low-velocity impact. Initially, the fabrics are augmented
in silicon aqueous suspension with various particle concentrations, and then, increase in
the weight and friction coefficient are studied. The low-velocity impact test results show
that the best particle concentration to meet the mentioned requirements is about 10
wt%. The experiments indicate high-impact resistance of the target by increase in the
number of fabric plies. It is found that silicon-coated fabrics under drop-weight test show
more time duration of impact and better performance than dry fabrics. Furthermore, the
tests show that in the dry fabrics, broader region stretches due to impact, while in
silicon-coated fabrics, the damage is limited to the impact point. The numerical simu-
lation is performed for the coated fabric, and the effect of fabric augmentation with
silicon is introduced as yarn friction. The numerical results are in good agreement with
the experimental results.

Keywords
Twill weave, silicon-coated Kevlar fabric, inter-yarn friction, drop-weight impact,
LS-DYNA

1
Department of Mechanical Engineering, Sahand University of Technology, Tabriz, Islamic Republic of Iran
2
Dynamic Behavior of Materials Research Laboratory, Sahand University of Technology, Tabriz, Islamic
Republic of Iran
3
Institute of Polymeric Materials, Sahand University of Technology, Tabriz, Islamic Republic of Iran
4
Department of Polymer Engineering, Sahand University of Technology, Tabriz, Islamic Republic of Iran

Corresponding author:
M Yazdani, Department of Mechanical Engineering, Sahand University of Technology, PO Box 51335-1996,
Sahand New Town, Tabriz, Islamic Republic of Iran.
Email: [email protected]
2 Journal of Thermoplastic Composite Materials XX(X)

Introduction
Kevlar fabrics with high-strength and high-modulus fibers were developed in 1960s. The
use of these materials improved and enhanced body armors to protect against various
harms and damages.1 Kevlar fibers have exceptional engineering properties, such as high
modulus, high strength, low weight, high chemical resistance, and thermal stability.2
Nowadays, Kevlar fabrics are widely applied in armors, aeronautics, and automobile
industries due to higher impact performance.3 Many researchers studied the impact
performance of these materials experimentally and numerically, and they devised ana-
lytical formulations.4
Various factors influence the impact performance of the fabric. These factors
include studying the role of friction,5 fabric clamping conditions,6 velocity and geo-
metry of the projectile,7 fabric target shape and size,8 number of plies,9 and yarn
material properties.10
Inter-yarn coefficient of friction is one of the most effective factors in enhancing the
impact performance and energy absorption capacity of Kevlar fabrics. Inter-yarn friction
causes effective yarn interlocking, and more energy is required for yarn pullout. In fact,
warp–weft close interlocking causes delay in fabric yarn breakage, and this enhances
energy-absorbing potential of the fabric.
By working experimentally on Kevlar fabrics, Briscoe et al.11 mentioned that
energy absorption capacity of fabrics against projectile impact decreased by reduction
in the inter-yarn coefficient of friction. Duan et al.5 performed numerical simulation
for the fabric with zylon yarns, and they showed that by increasing the coefficient of
friction between the zylon yarns, impact performance of the fabric increased against
the projectile impact, that is, it resulted in lower residual velocity of the projectile after
the impact.
Nowadays, most research studies concerning improvement in the impact performance
of the fabrics are associated with increasing the coefficient of friction of yarn. These
studies include fabric augmentation with various materials and estimation of impact
performance. The materials with lubricious properties, that is, greasy, reduce the inter-
yarn friction and, consequently, the impact performance of the fabric, or vice versa.
Bazhenov12 studied the ballistic performance of wet fabric against spherical bullet
impact. He concluded that water reduced the impact performance of the fabric due to
reduction in friction between the bullet and the yarns. He showed that dry laminates
resisted against the projectile impact, while wet laminates were perforated by that
impact. Lee et al.13 showed that shear thickening fluid (STF)-impregnated fabric had
higher impact performance than the neat fabric, and the results exhibited that four layers
of Kevlar fabric impregnated with 8-mm-thick STF absorbed the same amount of impact
energy as 14 layers of neat Kevlar fabric did. Dischler et al.14 studied the energy
absorption capacity of yarns coated with various powders. For example, two layers of
fabric coated by potassium chloride displayed 23% improvement in energy absorption,
due to friction enhancement, with 5% of weight increase. LaBarre et al.15 studied the
ballistic performance of carbon nanotube (CNT)-augmented Kevlar fabric. The authors
concluded that the ballistic resistance was doubled due to fabric augmentation. They also
Abbaszadeh et al. 3

Table 1. Product data of Kevlar fabrics.

Weave Weight Thickness Warp count Yarn size

style (g/m2) (mm) (yarns/cm) (Denier) (g/9 km)

Twill woven 400 0.38 6 3020

performed numerical simulation for CNT-augmented Kevlar and introduced the effect of
CNT as inter-yarn coefficient of friction, and finally, they found good agreement
between the numerical and the experimental results.

Problem description
In this study, low-velocity impact performance of twill-weave Kevlar fabric has been
reported to provide protection against damage. This study also compares the impact
performance of neat Kevlar fabric with silicon suspension-augmented fabric experi-
mentally. The test is performed with a cylindrical flat-end–shaped drop-weight impactor.
The test specimens are made of single-, two-, and four-ply fabrics. The criteria for
comparing the impact performance of the results include time history diagram of force
and acceleration for various targets, with time duration of impact as quantitative para-
meters and yarn crimp and yarn breakage of the fabrics as qualitative indicators. Fur-
thermore, the practical clamping condition is presented so as to achieve the most
efficient clamping and to avoid test result deviations.
Numerical simulation is performed with shell elements for modeling cross section of
fabric yarns. A yarn level of resolution is used in this model. The linear stress–strain
behavior of the Kevlar fabric is modeled according to the material model given by
Khodadadi et al.16 Finally, the coated fabric yarns are introduced in the form of inter-
yarn coefficient of friction in LS-DYNA.
The final goal of this research is to present a low-weight and high-performance fabric
to be used against impact damage.

Materials and experiments

Materials
Kevlar fabrics used for the impact test are made by Kevlan Company, Australia. The
as-received Kevlar fabric is twill-woven. The weight per unit area is 400 g/m2, and the
thickness is 0.38 mm. The yarns are 1.6 mm wide and 0.19 mm thick. Complete detailed
properties and product data of Kevlar fabrics are given in Table 1.
The fabrics are cut into 15  15 cm2 pieces, and the weight is measured up to three
decimals accuracy before augmentation (Figure 1).
The silicon used for coating purposes is made in a German company “Wacker
Chemical” with the “ELASTOSIL® LR 3003/40” brand. The silicon belongs to the
group of polydimethylsiloxane with (C2H6OSi)n chemical formula with two types A and
4 Journal of Thermoplastic Composite Materials XX(X)

Figure 1. Cutting the Kevlar fabric to prepare for augmentation with silicon suspension.

B, that is, it contains platinum catalyst and a cross-linker, respectively, with equal
density of 1.13 g/cm3. To prepare silicon suspension with an arbitrary particle con-
centration, 50% of each type A and B is solved with the presence of N-hexane solvent at
room temperature. Hexane (C6H14) is a colorless, transparent, and volatile liquid with a
density of 0.65 g/ml, and it is widely used in various products such as adhesives, paints,
plastics, and rubber.
The coefficient of friction for silicon-coated fabric is measured according to the
model proposed by Dong and Sun.2 Initially, the fabrics are augmented with silicon
suspension with 5%, 10%, and 20 wt% concentration. By comparing the coefficient of
friction enhancement relative to increase in weight, the optimum suspension concen-
tration is chosen to augment the remaining Kevlar fabrics.
To prepare the desired silicon concentration, an arbitrary amount of solvent is poured
into the graduated cylinder, then the mass is measured, and finally the required mass of
silicon to reach 5, 10, and 20 wt% concentration is calculated. Half of the required silicon
mass, types A and B with equal mass, that is, 50% of each, is added to the solvent.
Finally, the suspension is completely mixed up with the mechanical disperser machine.
The dip coating method is a common method to coat fabrics with liquids, suspensions,
and so on. This method may be applied using a dip coater machine with an arbitrary
infinitesimal coating thickness. Figure 2 displays the common procedure of dip coating
method.
According to Figure 2, at the first stage, the sample is submerged into the suspension,
and it is augmented with silicon. In this experimental procedure, the speed of the dip
coating machine is 6 mm/min. At the second stage, the sample is pulled out of suspension
with the same speed. At the third stage, a thin layer of coating is visible on the sample,
and the layer thickness depends on various factors such as dip coating speed, density, and
viscosity of the suspension. At the final stage, the sample is placed in a convection oven
for the extra solvent to evaporate.
Abbaszadeh et al. 5

Figure 2. Schematic procedure of dip coating method.

Figure 3. Schematic experimental setup for measuring the cross-yarn friction.2

Inter-fiber friction test

The model proposed by Dong and Sun2 is presented in this study to measure the coef-
ficient of friction of yarn. The drum is made by attaching two small plastic gears to a
solid cylinder. For a fabric under consideration, the Kevlar yarn is removed from the
fabric first and then wrapped onto the drum. The yarns are wrapped along the drum axial
direction and the Kevlar yarns are held in position by gear teeth (Figure 3).
During the test, the drum is fixed by fixing drum’s shaft. A weight with a definite
mass M1 is kept constant during the test. The second weight M2 is a container with sand.
Its weight is increased by adding sand to the container until it exceeds the balance and
starts to move downward.
6 Journal of Thermoplastic Composite Materials XX(X)

The coefficient of friction is calculated using the following equation:

 
1 M2
 ¼ ln ð1Þ
 M1

According to Dong and Sun2, two gears with a similar diameter of 77.7 mm and
60 gear teeth are used to perform the friction test. The yarns are separately wrapped
between the gear teeth along drum axis, and they are fastened to the shafts. As half of
the gear teeth are subjected to perform yarn sliding on another yarn, 30 yarns are
wrapped between 30 teeth. The last yarn is wrapped around perpendicular to previous
yarns with similar weights hung over both sides. Finally, the coefficient of friction of
yarns is measured.

Impact test
Impact test experiment is performed with drop-weight impact test machine. The 4-m
high pneumatic impact device is composed of support fixture, impactor, latch
mechanism, crosshead, electromotor, impactor accelerometer, two guide rails, two 4-m
high screws, two springs above the end of the guide rails to increase kinetic energy of the
impactor, and finally four springs to stop the impactor. The device is connected to a
computer; furthermore, it can be controlled manually.
The target is supported by 30  30 cm2 support plates. The plates can support a
cylindrical-shaped target with a diameter of 15 cm, and the plates have holes centered
with a diameter of 8 cm, the region that exposes the target to impact load (Figure 4(a)
and (c)).
In this research, targets are cut into blocks with 10  10 cm2 dimensions, and they
are put between the support plates that are clamped by screws around plates inside the
fixture. The impactor used for the test is given in Figure 4(b). The impactor assembly
has a mass of 5.9 kg with a flat-end striker. The impact test machine has an accel-
erometer located on the impactor, and it measures the acceleration of the impactor from
the beginning of the release of the latch mechanism, that is, free falling of the impactor,
until the end of the impact event. The output is the time history of acceleration of the
impactor.
According to the fundamentals of physics,17 the initiation time of impact event during
free falling of the impactor is calculated from the following equation:
sffiffiffiffiffiffiffiffiffiffiffi
2h
t ¼ ð2Þ
g

In equation (2), t is the time during the impactor drop, g is acceleration due to
gravity, and h is the impactor drop height. Accordingly, drop-weight impact test
machine approximates the initial contact time of the impactor with the target, and it
filters the initial response of acceleration. In the next step, the data are smoothed and
curve-fitted by machine.
Abbaszadeh et al. 7

Figure 4. The impact test machine setup: (a) impactor assembly, (b) striker, (c) fixture, and
support plates.

Yarn pullout from the clamps is one of the major problems occurring through the
experiments. The reason is that the impactor is too heavy. Different methods are per-
formed to avoid this problem. One of the methods applied is the use of a file to make a
rough surface finish around the fabric for strict clamping of the target (Figure 5).
This is not so much helpful. Another method applied is the fastening of yarns with
stitches inside the clamp, but this method fails too (Figure 6). Other procedure used to
stop yarn pullout is to apply various adhesives, but they are not supportive.
The final solution for suitable clamping is to apply resin epoxy. A resin epoxy with
proper proportionality of hardener is applied carefully around the fabric as it prevents the
flow of the resin inside the impact region (Figure 17), and no yarn pullout occurs through
the impact event, and the impactor penetration results in yarn breakage.

Numerical simulation
In this study, Ansys software is used to model and mesh yarns, fabric, and the impactor.
For analyzing purposes, LS-DYNA software is used to simulate the impact event.
8 Journal of Thermoplastic Composite Materials XX(X)

Figure 5. Yarn pullout through the clamps under drop-weight impact due to improper clamping in
the (a) front side (b) back side of the fabric.

Figure 6. Yarn pullout of the fabric stitched to the clamps under drop-weight impact.

In this research, fabric yarns are modeled explicitly, and inter-yarn friction, impactor–
yarn interaction, and impactor penetration through warp/fill tows are considered in the
modeling procedure. The fabric experimentally tested in this research is a twill-woven
fabric. Plain weave is a base style. Thus, the fabric is initially modeled in plain style, and
then twill weave is created from the base style.18
In this research, fabric yarns and the impactor are modeled with shell 163 elements for
lower time processing purposes in Ansys software. A yarn level of resolution is used in
this model. Each yarn in contact with other yarns takes elliptically shaped cross-section
throughout the fabric. To model the curved shape of the yarn cross-section, six elements
are used across the yarn width as mentioned by Nilakantan et al.19 Shell elements have
nonuniform nodal thicknesses as given in Figure 7.
Abbaszadeh et al. 9

Figure 7. Yarn cross section using shell element with non-uniform nodal thicknesses.19

Table 2. Thickness across the yarn width.

Node Thickness (mm)

t1 0.07080
t2 0.14161
t3 0.17913
t4 0.19000

Figure 8. Yarn schematic geometry.20

The thickness across the yarn width is given in Table 2.

The tows in fabric are undulated, and they take sinusoidal form, so the curve modeling
of yarn undulation for plain-weave fabric may be as follows20:
 t  z
y¼ cos ð3Þ
2 s
In the above equation, t is thickness of the yarn along Y direction, s is yarn span, and
the yarn length is along z direction (Figure 8).
The number of elements should be chosen very carefully so that the undulation is
maintained with no sharp step creation along the elements. Twill-weave fabric yarns are
composed of undulating and straight parts. To simplify the modeling, equation (3) is
used for modeling sinusoidal part, so the repetitive unit cell of the twill-weave fabric is
generated according to Figure 9. Numerical analysis of single-ply fabric is performed
with 92,628 elements.
10 Journal of Thermoplastic Composite Materials XX(X)

Figure 9. Repetitive unit cell of twill-weave fabric.

Figure 10. Model of (a) fabric with circularly clamped boundary condition (b) impactor.

The drop-weight impact test for the fabric is performed by clamping support plates.
As mentioned before, the plates support a cylindrically shaped target with a diameter of
15 cm and a hole centered inside with a diameter of 8 cm. In other words, just the circular
region of the fabric with 8 cm diameter is exposed to the impact event. To perform the
similar numerical analysis, the outside of the circular boundary condition with 8 cm
diameter is clamped as shown in Figure 10(a).
The mechanical properties of the yarn is according to the reported values by Nila-
kantan et al.21
The impactor used in experiments is made up of two parts: a striker and crosshead-
mounted impactor. The striker is made up of three geometrical shapes as shown in Figure
10(b): a cylindrical tip with diameter of 8 mm and length of 30 mm, cylindrical extension
with diameter of 10 mm and length of 20 mm, and finally, partial conic tail with the base
diameter of 40 mm. The crosshead has a mass of 5.2 kg, and it is simulated as a
cylindrical mass with a rectified density to maintain 5.2 kg mass. This crosshead geo-
metry has no effect on impact event results, and just the mass should be accounted to
Abbaszadeh et al. 11

maintain the initial kinetic energy. The striker has the mass of 0.7 kg, so the total mass of
the weight is 5.9 kg. The element used in the impactor modeling is shell element.
The impact event consists of free falling of the weight with gravitational acceleration.
This phenomenon is simplified as a weight with initial impact velocity in the model. The
impact velocity is calculated according to the principle of energy conservation. Before
fall, the impactor has potential energy (equation (4)). After falling of the impactor and
getting contact with the target, the potential energy is converted to kinetic energy
(equation (5)). The impact velocity of the impactor (equation (6)) is obtained from the
equality of equations (4) and (5).
U ¼ mgh ð4Þ
1
K ¼ mv2 ð5Þ
2 ffiffiffiffiffiffiffiffi
p
v ¼ 2gh ð6Þ
In the above equations, m is the mass of the impactor, h is free falling height, U is the
potential energy, v is the impactor velocity, and K is the kinetic energy.17 For each test, to
simulate the impact event, the impactor height is measured, and then from equation (6),
the impact velocity is calculated and introduced in LS-DYNA.
In this research, the impactor and Kevlar target models are according to Khodadadi
et al.16 They considered that Kevlar has a linear stress–strain behavior.
The contacts used in yarn-level interaction and impactor–fabric interaction
for single-ply twill-weave fabrics are both CASS and CAG, that is, Con-
tact_Automatic_Surface_To_Surface and Contact_Automatic_General, algorithms,
otherwise the initial penetration of elements is inevitable. For multiple-ply twill-weave
fabrics, CAG contact algorithm is used for yarn-level interaction, fabric plies, and
impactor–fabric interaction. For shell elements with CASS contact algorithm, badly
shaped elements make the solution unstable, so CAG contact algorithm along with
CONTROL-CONTACT best suits for multiple-ply fabric modeling. Using CONTROL-
CONTACT, contact stiffness parameters can be increased, and it results in premature
failure of elements under the projectile, while the stress quickly reaches the ultimate
strength. The yarn-level and projectile-fabric contact are introduced by defining coeffi-
cient of friction obtained from the current experimental data for both dry and silicon-
coated fabrics.

Results and discussion

The coefficient of friction obtained for dry fabric according to the experiments given by
Dong and Sun2 is 0.228.
The weight increment and coefficient of friction enhancement due to augmentation of
yarns are given in Table 3.
Due to the reasonable relation between friction coefficient enhancement and weight
increment in 10% augmented fabric, the fabrics suit this level of augmentation con-
centration. Despite the low-weight increment for 5% concentration coating of fabric, the
friction coefficient is low, and it is not suitable for the fabric coating. On the other hand,
12 Journal of Thermoplastic Composite Materials XX(X)

Table 3. Weight increment and friction coefficient enhancement of augmented fabrics relative to
dry ones.

Suspension Weight Friction coefficient

Material type concentration wt% increment % enhancement %

Augmented fabric 5 11 7.1

Augmented fabric 10 47.1 18.7
Augmented fabric 20 >200 29.9

20% concentration coating of fabric increases its weight about 210%, and it is not
suitable for low-weight armors.

Impact test on dry fabrics

For the experimental impact test, single-, two-, and four-ply fabrics are used with three
samples from each. An extra three-ply sample for making comparison with four-ply
fabric is tested.
Three samples of single-ply fabrics are tested with drop-weight impact machine, and
the acceleration data are recorded. One of the samples experienced yarn pullout with
unusual acceleration data, so the tests including samples no yarn pullout or just yarn
breakage at the impact region were accepted as the right impact test.
Figure 11 displays the time history diagram of force for two samples of single-ply dry
fabrics with 30 J initial kinetic energy of the impactor. The diagram of the third sample is
not displayed here due to yarn pullout of the fabric. As mentioned before, the accel-
erometer on the impactor measures the time history of acceleration for drop-weight
impactor, and the force exerted by is calculated from Newton’s second law:
F ¼ Ma ð7Þ
In the above equation, M is the mass of the impactor, a is the measured acceleration,
and F is the force exerted by impactor.17
Figure 11 shows that the time history diagram of force for similar targets is in
agreement with one another. The impact event includes loading condition from 0 to 6 kN
and unloading condition from peak force to zero within 4 ms. This long period of the
impact event is due to flexibility of the fabric target, and the test results in perforation of
dry fabrics.
In addition to the impactor drop height, the time during the impactor drop depends
on some other parameters. One of the parameters is time delay in latch mechanism
release. This mechanism is pneumatic, and it may differ infinitesimally in the time that
it releases the piston and consequently the crosshead. In addition, the crosshead is
mounted on two guide rails, and the friction between rails and crosshead bearings may
raise the time of free falling condition. Therefore, time during the impactor drop may
differ for two similar drop heights. To achieve better comparison of the results, toe
compensation of the data is required. In other words, the initial time of impact is
Abbaszadeh et al. 13

Figure 11. Time history diagram of force for two samples of single-ply dry fabrics at 30 J impact
energy.

equalized for all tests, and the lowest time of start of impact is set as the base data for
all tests; Therefore, time during the impact condition, that is, final time of impact event
subtracted from the initial time, is the main parameter to be compared with neglect of
the start time of the impact event.
Figure 12 displays the time history diagram of force for three samples of two-ply
dry fabric targets at 30 J, 40 J, and 70 J impact conditions. Unlike the single-ply
fabric, the two-ply fabric is not perforated at 30 J impact energy, and no yarn breakage
is visible after the impact event. The 40-J impact energy breaks upper yarns of the
fabric with no perforation occurrence. Finally, the impact energy of 70 J perforates
both layers of the fabric.
As given in Figure 12, for 30-J impact energy, the fabric stops the impactor with 3800
N force exerted on, while for the 40-J impact energy, the fabric exerts 4500 N force on
the projectile, and the curve peak data shifts left. The increase in the impact energy to 70
J makes perforation in the fabric target with similar peak force duration compared with
40 J impact energy. A similar trend is shown in the time history diagram of force for 40 J
and 70 J impact energies, and the fact that the 40-J impact energy breaks some of the
upper yarns of the fabric reveals that the threshold impact energy is 40 J for the two-ply
dry fabric. Thus, by increasing the impact force energy over 40 J, it makes no change in
the peak force duration exerted by the fabric.
As it is shown in Figures 11 and 12, the peak force obtained for 30 J impact energy in
single- and two-ply dry fabrics equal to 6000 N and 3800 N, respectively. This decrease
in force value with respect to increase in the number of plies is due to greater damping
14 Journal of Thermoplastic Composite Materials XX(X)

Figure 12. Time history diagram of force for two-ply dry fabrics.

behavior of the two-ply fabric. As damping rises, the loss of impact energy increases, and
the impact performance of the fabric is improved. It is concluded that increase in the
number of plies of the fabric will better protect against damage.
As shown in Figure 12, the initial trend of the curve is ascending to some extent, then
the curve becomes horizontal, and finally it starts to ascend again. This is because of the
resistance of two layers of the fabric one after another during the impact event. Initially,
the upper layer of the fabric resists against the impactor, and it stretches through the
impact event. The resistance is apparent as the force increases in the curve. In the second
stage, while stretching, the first layer comes in contact with the second layer, and the
second layers gets involved in resisting against the impactor energy. Due to flexibility of
the Kevlar fabric, the second layer does not resist immediately after contact. This layer
stretches to some extent, and then it gets involved in resisting against the impactor.
The stage between the first and the second layer in which the layers resist includes
sliding of layers on each other. While the impactor pushes the first layer along the impact
direction, the second layer contacts with the first layer so they slide on each other until
they become interlocked, and then they resist together against the impact. The stage in
which sliding occurs explains the horizontal trend of the force history curve. As impact
energy increases, the sliding duration of layers decreases, and the interlocking occurs
quickly, so the horizontal trend of the curve shortens quickly. Furthermore, while the
impact energy increases from 30 J to 40 J, the interlocking of the two layers occurs
quickly, and the peak force duration shifts to the left in the diagram.
In four-ply dry fabric samples, a sample is tested at 100 J impact energy with no
perforation occurrence. Two other samples are tested against the 140-J impact energy.
Figure 13 displays the time history of force for two samples. Both targets are tested by an
Abbaszadeh et al. 15

Figure 13. Time history diagram of force for two samples of four-ply dry fabrics at 140 J impact
energy.

impact velocity of 6.89 m/s, that is, 140 J energy. In the first sample shown with dashed
line, the impactor penetrates the fabric, but in the second sample shown with solid line,
no penetration occurs. The time history of force for the two samples is similar, and the
peak force is nearly the same. The only difference between the two diagrams is the time
duration of the impact event. This time duration is larger for nonperforated sample than
the other because the sample stretches much to resist against the impactor. Although no
perforation occurs in the second sample, there are signs of yarn breakage at the impact
region, and it can be mentioned that 140 J impact energy is the energy threshold to
perforate four-ply dry fabric.

Impact test on coated fabrics

To perform the impact tests on coated fabrics, 15 fabric samples are prepared and coated
with silicon. The samples include three single-ply fabrics, two double-ply fabrics, and
two four-ply fabrics. Due to yarn pullout, resin epoxy adhesive is applied on all fabric
edges as described before. Among the three samples of single-ply fabric, just one of the
samples passed the impact test with no yarn pullout condition. Figure 14 displays the
force history diagram of dry and coated single-ply fabrics. The impact energy exerted
equals 30 J, and both targets are perforated. The coated fabric displays more time during
the impact than the dry fabric. As shown in Figure 14, time during the impact for coated
and dry fabric are 4.5 and 3.6 ms, respectively, and the peak forces exerted to coated and
dry fabrics are 5335 N and 5933 N, respectively.
In two-ply fabrics, yarn pullout results in windowing mode at 70 J impact energy.
This phenomenon occurs due to partial yarn pullout, and consequently, the impactor
16 Journal of Thermoplastic Composite Materials XX(X)

Figure 14. Comparison of force history diagram for dry and coated single-ply fabrics at 30 J
impact energy.

Figure 15. Windowing phenomenon for two-ply coated fabric at 70 J impact energy (back side).

passes through the yarns. Figure 15 displays the windowing phenomenon for two-ply
silicon-coated fabric at 70 J impact energy. This phenomenon reduces the impact per-
formance of the fabric.
Among four-ply fabrics, windowing occurs in one of the two samples. Four-ply
silicon-coated fabrics at 140 J impact energy are compared with dry fabrics. The
coated fabric with no yarn pullout resists against the impactor and stops it. Figure 16
displays time history diagram of force for dry and coated four-ply fabrics. In both targets,
the impactor is stopped with no perforation occurrence.
According to Figure 16, the peak force history diagram of the coated fabric is higher
than the dry one. Similar to single-ply fabric (Figure 14), the time history diagram of
Abbaszadeh et al. 17

Figure 16. Comparison of force history diagram of dry and coated four-ply fabrics at 140 J impact
energy.

force for coated four-ply fabric has more gradual slope and more time duration of impact
than the one for dry fabric.
Figure 17 displays the final status of dry and coated four-ply fabric after the 140-J
impact energy. Despite the fact that both dry and coated fabrics resist against impact
energy, that is, no perforation occurrence, yarn crimp is visible in the mid-side of the
edges of dry fabric, and yarn breakage happens at the impact point, but in the coated
fabric, neither yarn breakage at the impact point nor yarn crimp take place.

Numerical results
In numerical simulations, the initial impact energy is defined as impact velocity.
Figure 18 displays the numerical and experimental time history of acceleration for
single-ply dry fabric at 30 J impact energy. The toe compensation is made to have similar
starting time of impact event for better comparison of the results.
Figure 18 shows that the peak acceleration for experimental test equals 1050 m/s2
with 3.6 ms duration of impact. In finite element simulation, the peak acceleration is 884
m/s2 with 3.1 ms duration of impact. Although the numerically simulated and experi-
mental time history curve have some differences, the time duration of impact and peak
acceleration values have little deviations that equal 13.8% and 16%, respectively. The
curve shape differences may be due to the use of shell elements for warp/fill yarns and
the impactor in 2D simulation.
Figure 19 displays numerical simulation results of acceleration history for single-ply
silicon-coated fabric compared with experimental results at 30 J impact energy. For
numerical simulation purposes, the effect of coating in the fabric is introduced as the
coefficient of friction. So, for 10 wt% suspension concentration, the coefficient of
friction introduced equals 0.271. Figure 19 shows that time during the impact for
18 Journal of Thermoplastic Composite Materials XX(X)

Figure 17. Comparison of fabric target after impact: (a) front side of dry four-ply fabric, (b) back
side of dry four-ply fabric, (c) front side of coated four-ply fabric, and (d) back side of coated four-
ply fabric.

Figure 18. Finite element simulation and experimental time history of acceleration for single-ply
dry fabric at 30 J impact energy.
Abbaszadeh et al. 19

Figure 19. Numerical and experimental time history diagram of acceleration for single-ply silicon-
coated fabric at 30 J impact energy.

numerical simulation is less than the experimental one. It is shown that numerical and
experimental time during the impact are 3.3 and 4.5 ms, respectively, and the simulated
projectile perforates the target after the mentioned time with 26% deviation from the
experimental results. The peak accelerations for numerical and experimental results are
909 m/s2 and 904 m/s2, respectively, with good agreement.
Figure 20 shows the numerical simulation of the single-ply dry fabric and deforma-
tion mechanism of the yarns and perforation of the fabric.
As the fabric is perforated, or it resists against the impact, the boundary condition with
the weakest clamping condition, due to circular clamping, affects the yarn crimp in the
connecting path to the impact point. This phenomenon is visible in both experimental
tests and numerical simulation as shown in Figure 21.
Table 4 lists the numerical results of single-ply dry and coated fabric at 30 J impact
energy. Comparison of the exit velocity for both targets reveals that the silicon-coated
fabric has higher impact performance than the dry fabric due to 3.6% lower residual
velocity of the impactor.
In silicon-coated fabric with 10 wt% concentration, the coefficient of friction is
0.271, 18.7% higher than the dry fabric, and the increase in weight is 3.9%, but the
coated fabric reduces the projectile velocity just 3.6% more than the dry fabric. This
shows little performance improvement of silicon-coated fabric. The fact that the
friction coefficient of the coated fabric is high, and the impact performance is not
significantly improved can be supported by the research performed by Zeng et al.22
They concluded that the friction coefficient above the 0.2 value might increase stress
concentration in fabric yarns at the impact point, thus it reduced the impact perfor-
mance of the target; therefore, the coated fabric with 0.271 friction coefficient does not
have enhanced impact performance.
20 Journal of Thermoplastic Composite Materials XX(X)

Figure 20. Deformation mechanism of numerically simulated single-ply fabric at 30 J impact

energy (a) weight at the impact point with fabric, (b) stretching of yarns, and (c) perforation of
the fabric and failure of the yarns.

Figure 21. Yarn crimp in middle sides and fabric perforation in both (a) numerical simulation and
(b) experimental test of single-ply fabric at 30 J impact energy.

Summary and conclusions

In this article, the low-velocity impact performance of dry and silicon-coated twill-
weave fabric is studied experimentally and numerically. To find out a proper method
to coat the fabrics, they are augmented in various suspension concentrations, and the
Abbaszadeh et al. 21

Table 4. Numerical results of single-ply dry and coated fabric at 30 J impact energy.

Impactor exit Impactor exit Peak acceleration

Fabric type velocity (m/s) kinetic energy (J) (m/s2)

Coated fabric 2.14 13.5 909

Dry fabric 2.22 14.4 884

weight increase and friction coefficient enhancement are measured. The best coating
concentration by considering lower weight increase relative to friction enhancement is
chosen. In the next step, the impact test is performed for the selected fabric. In this study,
10 wt% suspension concentration is found to be the best choice with 18.7% friction
enhancement, 47.1% fabric yarn weight increase compared with the dry fabric. The
impact test performed for both coated and dry fabrics shows impact performance
enhancement by increasing the number of plies. As the number of plies increases, the
damping of the fabric rises, and it leads to higher impact performance of the fabric and
higher impact energy loss. The silicon-coated fabrics have more time duration of impact
and better impact performance compared with dry fabrics. Unlike the broader region of
dry fabric subjected to yarn stretch, the damage in silicon-coated fabric is limited to the
impact point. The impact test is numerically simulated in LS-DYNA software, and the
effect of coating is introduced as the coefficient of friction. The numerical results for
single-ply dry and coated fabrics at 30 J impact energy reveal that the coated fabric
reduces the residual velocity of the impactor 3.6% more, resulting in higher impact
performance of the coated fabric.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research,
authorship, and/or publication of this article.

Funding
The author(s) received no financial support for the research, authorship, and/or publi-
cation of this article.

ORCID iD
M Yazdani http://orcid.org/0000-0003-4321-4263

References
1. Cheeseman BA and Bogetti TA. Ballistic impact into fabric and compliant composite lami-
nates. Compos Struct 2003; 61(1): 161–173.
2. Dong Z and Sun CT. Testing and modeling of yarn pull-out in plain woven Kevlar fabrics.
Compos A 2009; 40(12): 1863–1869.
3. Ha-Minh C, Kanit T, Boussu F, et al. Numerical multi-scale modeling for textile woven fabric
against ballistic impact. Comput Mater Sci 2011; 50(7): 2172–2184.
22 Journal of Thermoplastic Composite Materials XX(X)

4. Tabiei A and Nilakantan G. Ballistic impact of dry woven fabric composites: a review. Appl
Mech Rev 2008; 61(1): 010801.
5. Duan Y, Keefe M, Bogetti TA, et al. Modeling the role of friction during ballistic impact of a
high-strength plain-weave fabric. Compos Struct 2005; 68(3): 331–337.
6. Nilakantan G and Gillespie JW. Ballistic impact modeling of woven fabrics considering yarn
strength, friction, projectile impact location, and fabric boundary condition effects. Compos
Struct 2012; 94(12): 3624–3634.
7. Nilakantan G, Wetzel ED, Bogetti TA, et al. Finite element analysis of projectile size and
shape effects on the probabilistic penetration response of high strength fabrics. Compos Struct
2012; 94(5): 1846–1854.
8. Nilakantan G and Nutt S. Effects of fabric target shape and size on the V50 ballistic impact
response of soft body armor. Compos Struct 2014; 116: 661–669.
9. Mamivand M and Liaghat G. A model for ballistic impact on multi-layer fabric targets. Int J
Impact Eng 2010; 37(7): 806–812.
10. Rao MP, Duan Y, Keefe M, et al. Modeling the effects of yarn material properties and friction
on the ballistic impact of a plain-weave fabric. Compos Struct 2009; 89(4): 556–566.
11. Briscoe BJ and Motamedi F. The ballistic impact characteristics of aramid fabrics: the influ-
ence of interface friction. Wear 1992; 158(1): 229–247.
12. Bazhenov S. Dissipation of energy by bulletproof aramid fabric. J Mater Sci 1997; 32(15):
4167–4173.
13. Lee YS, Wetzel ED and Wagner NJ. The ballistic impact characteristics of Kevlar® woven
fabrics impregnated with a colloidal shear thickening fluid. J Mater Sci 2003; 38(13):
2825–2833.
14. Dischler L, Moyer TT and Henson JB. Dilatant powder coated fabric and containment articles
formed therefrom. USA: Milliken Research corporation, 1998.
15. LaBarre ED, Shanaman MT, Tiffany JE, et al. Multi-scale testing techniques for carbon
nanotube augmented kevlar. In: Dynamic Behavior of Materials. Proceedings of the 2013
Annual Conference on Experimental and Applied Mechanics, Conference Proceedings of the
Society for Experimental Mechanics Series, The Society for Experimental Mechanics Inc.,
2014, pp. 59–68.
16. Khodadadi A, Liaghat G, Akbari M, et al. Numerical and experimental analysis of penetration
into Kevlar fabrics and investigation of the effective factors on the ballistic performance. Mod
Mech Eng (In Persian) 2013; 13(12): 124–133.
17. Resnick R, Halliday D and and Walker J. Fundamentals of physics. John Wiley, 1988.
18. Bacarreza O, Abe D, Aliabadi MH, et al. Micromechanical modeling of advanced composites.
J Mult Model 2012; 4(02): 1250005.
19. Nilakantan G, Keefe M, Bogetti TA, et al. Multiscale modeling of the impact of textile fabrics
based on hybrid element analysis. Int J Impact Eng 2010; 37(10): 1056–1071.
20. Nilakantan G, Keefe M, Gillespie JW, et al. Novel multi-scale modeling of woven fabric
composites for use in impact studies. In: 10th International LSDYNA users Conference.
Michigan: Dearborn, 2008, pp. 19–38.
21. Nilakantan G, Keefe M, Bogetti TA, et al. On the finite element analysis of woven fabric
impact using multiscale modeling techniques. Int J Solid Struct 2010; 47(17): 2300–2315.
22. Zeng XS, Tan VBC and Shim VPW. Modelling inter-yarn friction in woven fabric armour. Int
J Num Method Eng 2006; 66(8): 1309–1330.