Materials and Design 58 (2014) 305315

Contents lists available at ScienceDirect

Materials and Design

journal homepage: www.elsevier.com/locate/matdes

The ballistic performance of metal plates subjected to impact

by projectiles of different strength
Deng Yunfei a,1, Zhang Wei b,, Yang Yonggang a, Wei Gang b
a
College of aeronautical Engineering, Civil Aviation University of China, Tianjin, 300300, PR China
b
Hypervelocity Impact Research Center, Harbin Institute of Technology, P.O. Box 3020, Science Park, No. 2 Yikuang Street, Nan Gang District, Harbin 150080, PR China

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

Article history: In this paper, the ballistic performance of monolithic, double- and three-layered steel plates impacted by
Received 2 November 2013 projectiles of different strength is experimentally investigated by a gas gun. The ballistic limit velocity for
Accepted 29 December 2013 each conguration target is obtained and compared based on the investigation of the effect of the number
Available online 10 January 2014
of layers and the strength of projectiles on the ballistic resistance. The results showed that monolithic
plates had higher ballistic limit velocities than multi-layered plates for projectiles of low strength regard-
Keywords: less their nose shape, and also the ballistic limit velocities of plates decreased with the increase of the
Projectile
number of layers. Moreover, monolithic plates showed greater ballistic limit velocities than multi-layered
Impact
Failure model
plates for ogival-nosed projectiles of high strength, and also the ballistic limit velocities of plates
Layered plate decreased with the increase of the number of layers. However, monolithic plates had lower ballistic limit
Target velocities than multi-layered plates for blunt-nosed projectiles of high strength, and also the ballistic
limit velocities of plates increased with the increase of the number of layers. The differences in the bal-
listic limit velocities between various impact conditions can be related to the transitions of perforation
mechanisms and failure models of plates and projectiles.
2014 Elsevier Ltd. All rights reserved.

1. Introduction residual velocities of the projectiles for double-layered plates were

When a single plate is replaced by several layered plates, the
number, thicknesses and order of layers, the air gap between lay-
ers, the nose shape and strength of projectiles affect the failure
models of plates that lead to the differences of ballistic perfor-
mance between different impact conditions. A lot of investigations
have been conducted on the study of the ballistic resistance of
monolithic and multi-layered plates experimentally, numerically,
and theoretically.
Zhang and Deng et al. [13] carried out an extensive experi-
ments on the monolithic and multi-layered steel plates struck by
blunt-, ogvial- and hemispherical-nosed projectiles of high
strength in order to investigate the effect of the air gap between
layers, the number, order and thickness of layers on the on the bal-
listic performance of plates.
Gupta et al. [4] investigated the ballistic performance of multi-
layered aluminum plates under the impact of at-, ogival- and
hemispherical-nosed steel projectiles. It was observed that the
Corresponding author. Tel.: +86 451 86417978 13; fax: +86 451 864020 55.
E-mail addresses: dengyunfeihit@gmail.com (D. Yunfei), zhdawei@hit.edu.cn
(Z. Wei).
1
Tel.: +86 451 86417978 18; fax: +86 451 864020 55.
comparable to those of the single plates of equivalent thicknesses.
However, the single plates offered more resistance against perfora-
tion when the number of layers was increased. Ogival-nosed pro-
jectiles were found to be the most efcient penetrator, but
hemispherical-nosed projectiles required maximum energy for
perforation. Moreover, results of the nite element analysis were
compared with experiments and a good agreement was found.
Radin and Goldsmith [5] compared the ballistic resistance of
monolithic and multi-layered aluminum plates that impacted by
blunt- and conical-nosed projectiles. It was found that the ballistic
resistance of in-contact multi-layered plates was inferior to that of
monolithic plates of equal thickness, and also spaced multi-layered
plates were less effective than in-contact multi-layered plates.
Almohandes et al. [6] conducted an extensive experiments on mild
steel plates struck by standard bullet to investigate the effect of
target conguration on the ballistic performance. They concluded
that monolithic plates were more effective than multi-layered
plates of equal thickness. Moreover, the ballistic resistance of
multi-layered plates decreased with the increase of the number
of the layers. The ballistic performance of double-layered plates
can be enhanced by using a thicker back plate, and also in-contact
multi-layered plates had higher ballistic resistance than spaced
multi-layered plates. Nurick and Walter [7] studied the ballistic

http://dx.doi.org/10.1016/j.matdes.2013.12.073
0261-3069/ 2014 Elsevier Ltd. All rights reserved.
306 D. Yunfei et al. / Materials and Design 58 (2014) 305315

resistance of multi-layered steel plates using conical- and at- interpreted in relation to mass, radius, nose shape of projectile,
nosed projectiles. It was revealed that the ballistic limit velocities plate thickness and number of layers but not in relation to the duc-
of monolithic plates were 48% higher than those of the in-contact tility of the projectile material. In other words, the investigation on
multi-layered plates of equivalent thickness. Woodward and the projectile deformation and fracture behavior in the impact test
Cimpoeru [8] carried out experimental studies on multi-layered is limited and patchy. Especially, reports regarding the effect of
aluminum alloy plates of the same total thickness that perforated projectile strength on the failure and ballistic limit velocities of
by blunt- and conical-nosed projectiles. It revealed that multi-layered plates are few. In fact, the failure models of projec-
multi-layered plates of two equal plates presented the best perfor- tiles greatly affect the perforation efciency of projectiles and the
mance followed by multi-layered plates of a thicker front plate and ballistic performance of plates. So, further detailed studies, espe-
a thinner back plate, multi-layered plates of a thinner front plate cially experimental investigation is required. Thereby, this paper
and a thicker back plate. The monolithic plates were more effective will concentrate on the ballistic capacity of monolithic and
than multi-layered plates, and also these conclusions of blunt- multi-layered plates, where the varying factors will be the number
nosed projectiles were the same to those of conical projectiles. of layers in relation to the strength and nose shape of the
Marom and Bodner [9] carried out experimental and theoretical projectiles. To do this, the ballistic resistance of monolithic and
studies on the perforation behavior of multi-layered thin alumi- multi-layered plates of the same equivalent thickness against
num beams under impact by spherical-nosed bullet projectiles. blunt- and ogival-nosed projectiles impact is studied.
They concluded that the general order of the ballistic resistance
of the beam targets, starting with maximum, was multi-layered
at beams in contact, an equivalent weight uniform beam, and sep- 2. Experimental details
arated at beams of equal weight. Corran et al. [10] carried out a
series of experiments on the performance of multi-layered mild 2.1. Experimental setup
steel plates under impact of at projectiles. They found that the
in-contact layered plates were superior to the monolithic plates The impact tests were conducted in a one-stage gas gun at
whose energy absorption dominated by membrane stretching. Hypervelocity Impact Research Center in Harbin Institute of
Teng and Dey et al. [1113] recently reported comprehensive Technology, where the impact velocity was controlled directly by
experimental and numerical studies on the perforation resistance the gas pressure, as depicted in Fig. 1. Square targets with a side
of double-layered steel armor plates. They found that the ballistic of 230 mm were xed to a thick armor plate by means of eight
limit velocity of a double-layered shield was 30% higher than that bolts arranged on a 170 mm diameter pitch circle in the target
of the monolithic case in the case of a blunt-nosed projectile. How- chamber, as shown in Fig. 2.
ever, these advantages seemed to disappear when ogival-nosed A Photron FASTCAM SA5 high speed camera was used to record
projectiles were used. the process of projectile penetrating target and the projectile mo-
It appears from the above literature review that there have been tion. From the digital images, traveled distance, impact angle,
many papers which examined the deformation and failure behav- velocity and acceleration as a function of penetration time for
ior of plates that most focus on the rigid projectiles impact tests. the projectiles during penetration were obtained. The selected
However, the effect of deformation and failure behavior of projec- frame rate was 50,000 per second, so that a frame was taken every
tiles to the ballistic performance of steel plates is limited. It is
poses an interesting question: Under what kind of projectile im-
pact would a multi-layered shield be superior in the ballistic limit
velocity than a monolithic plate of the same total thickness?
Besides the studies of rigid projectiles, the deformable projec-
tiles were also used for the ballistic tests of monolithic plates.
Dongquan and Stronge [14] examined the effect of various param-
eters on the extent of global deformation and localized failure
mechanisms of the target plate that as a function of the relative
size of plate and projectiles, ow stress and fracture strain of the
plate, projectiles mass and deformability at impact velocities close
to the ballistic limit velocities. Chen et al. [15] constructed a rigid-
plastic model to assess the effect of a soft nose on the perforation
of metallic plates struck by stubby projectiles. Effects of transverse Fig. 1. Schematic of one-stage gas gun experimental system.
shear, bending and membrane deformations on the perforation
process were included in a rigid-plastic analysis. Chen et al. [16]
described experimental and numerical studies concerning the im-
pact of steel blunt-nosed projectiles against harder steel plates at
impact velocities between 200 and 800 m/s. Tang [17] investigated
the inuence of soft nose of projectiles on the perforation of steel
plates by experiments. The initiated ahead structural response of
target plate was benecial to decrease the relative impact velocity
between projectile and target and to increase the time duration of
perforation. Similarly, Kim et al. [18] carried out the numerical
simulations and experiments that indicated that the soft nose acts
as an energy absorber to be benecial for the integrity of projectile
structure in the perforation.
It is evident from the literature review, the assumption of non-
deformable (rigid) projectiles is usually employed in the analysis of
the ballistic resistance of monolithic and multi-layered plates, and
also the experimental data for the ballistic limit velocity has been Fig. 2. Target supports.
 D. Yunfei et al. / Materials and Design 58 (2014) 305315
20 ls and the resolution was 512  128 pixels. A photocell detec-
tor system having two identical light-barriers with laser sources
and light-sensing photodiodes was used to correct aberrations that
may to appear in the high speed camera system. The velocities cal-
culated using the high-speed camera images were in close agree-
ment (within 12%) with those obtained by the photocell system
[1,2].
2.2. Sample specications
To examine the effects of various parameters on the ballistic
resistance, two groups of tests were conducted. The rst group of
tests consisted of ring hardened steel projectiles to plates. The
projectiles are made of 38CrSi. After machining, they are oil hard-
ened to a maximum Rockwell C value of 58 in order to prevent the
plastic deformation during impact process for projectiles of high
strength. The hardened steel projectiles experienced negligible
inelastic deformation during the impact process and consequently
can be considered as rigid projectiles. The high strength cylinder
projectiles have nominal average strength (HRC 53.1), diameter
(12.7 mm) and mass (34.5 g). A second group of tests were con-
ducted to examine the effect of deformability of ductile projectiles
on the perforation behavior and the ballistic limit velocities of
plates. To achieve an extensive range of projectile deformation
characteristics, the tests in the second group were performed with
projectiles made of original 38CrSi, while nominal average strength
about HRC 19.1 for projectiles of low strength [19]. The geometry
of blunt-nosed and ogival-nosed projectiles can be found in
Fig. 3(a and b).
The target plates are made of 45 steel as mild carbon alloy steel,
whose yield stress and strength are about 714 MPa and 29 HRC
after heat treatment. A series of tensile tests were performed on
smooth plates to determine strength of 45 steel based on Chinese
national standard of GB/T 228-2002 [20] that update from ISO
6892:1998, as shown in Fig. 4. These tests provided typical true
stressstrain curves for the steel alloy to failure, as shown in
Fig. 5. The result is obtained from a standard quasi-static tensile
test on a smooth axial-symmetric specimen at room temperature.
The thickness of monolithic plate is 6 mm and the thicknesses
of layers of multi-layered plates are 2 and 3 mm, respectively.
For the investigation of the effect of the layering, ten congurations
are considered. In paper, the target code T3T3 indicates a in-con-
tact double-layered target whose layers thickness are all 3 mm
and T2T2T2 represents a three-layered target whose layers thick-
ness are all 2 mm, and also T6 denotes a monolithic target whose
thickness is 6 mm.
3. Experimental results
With various combinations of the metal plates, a series of im-
pact tests were conducted on the gas gun. The ballistic resistance
of targets is evaluated by comparing their ballistic limit velocities.
Fig. 3. Geometry and dimensions for projectiles (unit: mm).
307
Fig. 4. Geometry of the smooth plate (unit: mm).
900
Engineering stress (MPa)
800
700
600
500
400
300 Test 1
200 Test 2
Test 3
100
0
0 0.1 0.2 0.3
Engineering strain
Fig. 5. Engineering stressstrain curves.
The experimental results of the response of targets impacted by
ogival-nosed projectiles are presented in Tables 1 and 2. The im-
pact velocity vi, residual velocity vr, velocity drop vd = vivr, initial
kinetic energy Ei and residual kinetic energy Er of projectiles, and
also energy absorbed by the targets Ea = EiEr are shown in the
Table 1. Moreover, the residual length lr, nal maximum diameter
dmax and residual mass mr of projectiles of low strength after im-
pact are given in Table 2.
The change in energy absorption and failure models of the pro-
jectiles can induce the differences in ballistic limit velocities of tar-
gets. Comparing with the original, all the residual projectiles lose a
little amount mass and their lengths shorten, and also the head of
projectiles obviously become blunter and bigger, as indicated in
Table 2. With increasing the impact velocity of projectiles, the loss
of length and mass of projectile become larger, and also the head of
projectiles gradually become bigger. Consequently, as it is harder
for the projectiles of low strength to perforate the plates due to
their plastic deformation, and also more energy is consumed for
projectiles of low strength compared to projectiles of high strength
due to the change in deformation and failure model, which means
that there is increment in the ballistic limit velocities when projec-
tiles of low strength compared to projectiles of high strength. In
addition, the plastic deformation and mass loss of projectiles of
low strength that perforate multi-layered plates are smaller than
perforate monolithic plates. Thereby, the decrease of the plastic
deformation for projectile of low strength seems to be one major
reason for the decrease in the ballistic limit velocity when layering
308 D. Yunfei et al. / Materials and Design 58 (2014) 305315

Table 1
Experimental results of ogival-nosed projectiles.

Impact type Target code vi (m/s) vr (m/s) vd (m/s) Ei (J) Er (J) Ea (J)
Projectiles of high strength T6 316.99 0 316.99 1733.33 0 1733.33
341.18 93.48 247.7 2007.97 150.74 1857.23
347.22 116.55 230.67 2079.69 234.32 1845.37
365.47 153.03 212.44 2304.05 403.96 1900.09
380.39 166.42 213.97 2496.02 477.75 2018.27
394.477 192.31 202.167 2684.31 637.96 2046.35
441.35 237.42 203.93 3360.12 972.35 2387.77
T3T3 291.38 0 291.38 1464.56 0 1464.57
328 128.21 199.79 1855.82 283.55 1572.27
356.13 170.94 185.19 2187.79 504.053 1683.74
420.43 253.59 166.84 3049.13 1109.31 1939.82
432.69 266.27 166.42 3229.56 1223.02 2006.54
480.77 320.51 160.26 3987.16 1772.04 2215.13
T2T2T2 261.31 0 261.31 1177.88 0 1177.88
293.69 140.97 152.72 1487.88 342.80 1145.08
300.36 110.13 190.23 1556.23 209.22 1347.01
372.76 238.62 134.14 2396.89 982.21 1414.68
416.06 293.69 122.37 2986.08 1487.88 1498.20
445.71 330.4 115.31 3426.84 1883.08 1543.76
474.42 330.4 144.02 3882.53 1883.08 1999.45
560.67 468.06 92.61 5422.55 3779.13 1643.42
Projectiles of low strength T6 434.78 0 434.78 3260.83 0 3260.83
487.45 53.11 434.34 4098.73 48.66 4050.07
507.25 163.04 344.21 4438.47 458.54 3979.93
513.83 144.93 368.9 4554.37 362.33 4192.04
531.4 163.04 368.36 4871.16 458.54 4412.62
543.48 204.6 338.88 5095.14 722.11 4373.04
553.36 217.39 335.97 5282.08 815.21 4466.87
559.01 217.39 341.62 5390.49 815.21 4575.28
615.94 273.63 342.31 6544.34 1291.57 5252.78
T3T3 347.83 0 347.83 2087.00 0 2087.00
362.32 40.92 321.4 2264.51 28.88 2235.623
376.81 81.06 295.75 2449.26 113.35 2335.91
383.63 60.66 322.97 2538.72 63.47 2475.24
395.26 122.63 272.63 2694.98 259.41 2435.57
440 155.28 284.72 3339.6 415.93 2923.67
465.84 207.94 257.9 3743.37 745.87 2997.50
507.24 281.33 225.9164 4438.41 1365.28 3073.13

Table 2 v r av pi  v pbl 1=p 1

Experimental results of ogival-niosed projectiles of low strength.

Impact type vi (m/s) vr (m/s) lr (m/s) dmax (mm) mr (g) Here, the ballistic limit velocities vbl are always taken directly from
T6 434.78 0 33.51 14.96 33.57 the experimental data. In the original model a = mp/(mp + mpl),
487.45 53.11 33.46 15.19 32.09 where mp and mpl are the masses of the projectiles and plugs, both
507.25 163.04 33.56 15.56 32.45 a and p are tted to the experimental results by using the method of
521.74 144.93 32.52 15.48 32 least squares. The values of a, p and vbl obtained from the experi-
531.4 163.04 33.17 16 32.3
543.48 204.6 33.83 16.16 32.3
mental data by Eq. (1) are given in Table 3.
553.36 217.29 32.56 16.34 31.88 Based on the result of impact tests, the initial velocity vs. resid-
559.01 217.39 32.77 16.21 31.75 ual velocity curves of the projectiles are constructed and com-
615.94 273.63 32.45 17.05 31.78 pared, as shown in Figs. 6 and 7, where the solid lines through
T3T3 347.83 0 39.56 12.86 33.61
the data points are tted to Eq. (1) It should be noticed that the
362.32 40.92 39.33 12.91 33.55
376.81 81.06 39.12 12.95 33.48 experimental data of blunt-nosed projectiles have been presented
383.63 60.66 38.83 13.02 33.3 before [23], and also is therefore mainly given for comparison. It is
395.26 122.63 38.29 13.28 33.6 seen from the initial-residual velocity curves that the impact veloc-
440 155.28 38.61 13.16 33.58 ity, nose shape and yield strength of projectiles signicantly affect
465.84 207.94 38.88 13.34 33.43
507.2464 281.33 38.64 13.45 34.28
the ballistic resistance of plates.
From Figs. 6 and 7 and Table 3, it can be concluded that the
number of layers and the strength of projectiles signicantly affect
the ballistic resistance of targets. The observations illustrate that
the plates. Moreover, the mass loss of blunt-nosed projectiles is the initial-residual velocity curves for projectiles of high strength
bigger than that of ogival-nosed projectiles of low strength under seem to be rather close and parallel as the impact velocities be-
the same impact condition. come high compared to the respective ballistic limit velocities,
The ballistic resistance of plates is evaluated by comparing their especially for blunt-nosed projectiles of high strength. Namely,
ballistic limit velocities. The ballistic limit velocity is calculated as the inuence of target conguration on the ballistic resistance de-
the average between the highest impact velocity not giving perfo- creases with the increase of the impact velocity. For projectiles of
ration and the lowest impact velocity giving complete perforation low strength, on the other hand, the distance between the curves
of the target [21].The ballistic limit velocity is tted to an analytical seems to bigger than those of projectiles of high strength, this is
model originally proposed by Recht and Ipson [22]. also to say that the effect of layering is less distinct for projectiles
 D. Yunfei et al. / Materials and Design 58 (2014) 305315 309

Table 3
Ballistic limit velocities and model constants of targets (vbl: m/s).

Blunt-nosed projectiles Ogival-nosed projectiles

High strength Low strength High strength Low strength
T6 T3T3 T2T2T2 T6 T3T3 T6 T3T3 T2T2T2 T6 T3T3
a 0.82 0.76 0.72 0.8 0.71 0.7 0.84 0.96 0.6 0.76
p 2.1 2.95 2.95 2.45 2.4 2.45 2.05 1.98 2.61 2
vbl 215 237.5 252 485.4 424 325 294.6 276.5 482.5 358.5

350 400 Experimental data of T6

Experimental data of T3T3
Residual velocity (m/s)

Residual velocity (m/s)

300 350 Experimental data of T2T2T2
250 300
250
200
200
150
Experimental data of T6 150
100 Experimental data of T3T3 100
50 Experimental data of T2T2T2 50
0 0
250 300 350 400 450 500 550 600 150 200 250 300 350 400 450 500 550
Initial velocity (m/s) Initial velocity (m/s)
(a) Ogival-nosed projectiles; (b) Blunt-nosed projectiles
Fig. 6. Residual velocities vs. initial velocities for projectiles of high strength.

400 Experimenatal data of T6

Experimenatal data of T3T3 Experimental data of T6
Residual velocity (m/s)

350 400
Residual velocity (m/s)

Experimental data of T3T3

300 350
250 300
250
200
200
150
150
100 100
50 50
0 0
300 350 400 450 500 550 600 650 400 450 500 550 600 650 700
Initial velocity (m/s) Initial velocity (m/s)
(a) Ogival-nosed projectiles; (b) Blunt-nosed projectiles
Fig. 7. Residual velocities vs. initial velocities for projectiles of low strength.

of high strength than projectiles of low strength. The obvious rea-
son for this is that the projectiles of low strength yield plastic
deformation during impact as discussed above, while the plastic
deformation can be neglected in projectiles of high strength after
perforation. It should be noticed that the residual velocity curve
for T6 exceeds that of T3T3 at the highest impact velocities when
impacted by blunt-nosed projectiles of low strength, and also it
indicates that even though the ballistic resistance is higher for T6
than for T3T3 at impact velocities close to the ballistic limit veloc-
ities, the situation is opposite at higher velocities. A similar behav-
ior is seen between T6 and T3T3 for blunt-nosed projectiles of high
strength, so the initial impact velocities of projectiles has greatly
inuence on the ballistic resistance of plates.
For blunt-nosed projectiles of high strength, the ballistic limit
velocities of plates increase with the increase of the number of lay-
ers, so the ballistic limit velocity is enhanced when using multi-
layered plates instead of monolithic plates. For example, the ballis-
tic limit velocity of T6 is 215 m/s and that of T3T3 is 237.5 m/s, and
also that of T2T2T2 is 252 m/s, so the increment in the ballistic lim-
it velocity of 10.5% and 17.2% are obtained when T3T3 and T2T2T2
compared with T6. However, the ballistic limit velocity of T6 is
485.4 m/s and that of T3T3 is 424 m/s for blunt-nosed projectiles
of low strength, so the increment in the ballistic limit velocity of
14.5% is found when T6 compared to T3T3, and also the ballistic
limit velocity is decreased when using multi-layered plates instead
of monolithic plates. From above analyses, it indicates that the
change in the ballistic limit velocity is less distinct for blunt-nosed
projectiles of high strength than those of low strength when using
multi-layered plates instead of monolithic plates.
The results clearly show that the ballistic limit velocities of
plates increase with the increase of the number of layers for
blunt-nosed projectiles of high strength, while the opposite trend
is seen for ogival-nosed projectiles of high strength. For ogival-
nosed projectiles of high strength, the ballistic limit velocities of
plates decrease with the increase of the number of layers. For
example, the ballistic limit velocity of T6 is 325 m/s and that of
T3T3 is 294.6 m/s, and also that of T2T2T2 is 276.5 m/s, so the
increment in the ballistic limit velocity of 10.3% and 17.5% are ob-
tained when T6 compared with T3T3 and T2T2T2. Therefore, it can
be concluded that the ballistic limit velocities are decreased when
310 D. Yunfei et al. / Materials and Design 58 (2014) 305315

using multi-layered plates instead of monolithic plates. And also

Ballistic limit velocity (m/s)

for ogival-nosed projectiles of low strength, the ballistic limit
velocities are decreased when using multi-layered plates instead
of monolithic plates. For example, the ballistic limit velocity is
482.5 m/s for T6 and 358.5 m/s for T3T3. Thus, an increment in
the ballistic limit velocity of 34.5% is found when T6 compared
to T3T3. From above analyses, it indicates that the change in ballis-
tic limit velocity is less distinct for ogival-nosed projectiles of high
strength than those of low strength when using multi-layered
plates instead of monolithic plates.
The strength of projectiles greatly inuences the ballistic limit
velocities of plates, and also the effect of strength is in related to
the nose shape of projectiles, as shown in Table 3. The experimen-
tal results indicate that the projectiles of high strength are further
stronger than projectiles of low strength. For blunt-nosed projec-
tiles, an increment in the ballistic limit velocity of 104.4% is found
for T6 when projectiles of low strength compared to projectiles of
high strength, and also an increment of 78.5% is obtained for T3T3
when projectiles of low strength compared to projectiles of high
strength. Moreover, for ogival-nosed projectiles, an increment in
the ballistic limit velocity of 48.5% for T6 is found when projectiles
of low strength compared to projectiles of high strength, and also
the increment of 21.7% is obtained for T3T3 when projectiles of
low strength compared to projectiles of high strength. From above
analyses, it reveals that the ballistic limit velocities of blunt-nosed
projectiles are more sensitive to their strength than those of ogival-
nosed projectiles.
The nose shape of projectiles greatly inuences the ballistic lim-
it velocities of plates, and also the effect of nose shape is in related
to the strength of projectiles. For example, the increment in the
ballistic limit velocity of 51.2%, 24.1% and 9.7% are obtained for
T6, T3T3 and T2T2T2 when ogival-nosed projectiles compared with
blunt-nosed projectiles of the same high strength, so the ballistic
limit velocities of blunt-nosed projectiles are smaller than those
of ogival-nosed projectiles, especially for the monolithic plates.
However, the increment in the ballistic limit velocity of 0.6% and
18.3% are obtained for T6 and T3T3 when blunt-nosed projectiles
compared with ogival-nosed projectiles of the same low strength,
so the ballistic limit velocities of blunt-nosed projectiles are bigger
than those of ogival-nosed projectiles, especially for the multi-
layered plates, as depicted in Table 3.
From above analyses, it reveals that the layering effect of plates
for the projectiles of low strength is signicantly bigger than that
of projectiles of high strength, especially for ogival-nosed projec-
tiles, as shown in Fig. 8. This is because that the mass loss and
deformation of blunt-nosed projectiles is bigger than those of
550 Projectiles of high strength
Projectiles of low strength
500
450
400
350
300
250
200
150
100
T6-B T3T3-B T6-O T3T3-O
Target configuration
Fig. 8. Comparisons of the ballistic limit velocities of T6 and T3T3.
ogival-nosed projectiles of low strength under the same impact
condition as analyzed above.
The inuence of the impact velocity on velocity and kinetic en-
ergy variable is depicted in Fig. 9.
Fig. 10, the solid lines through the data points have been calcu-
lated using the initial vs. residual velocity curves obtained from the
RechtIpson model [21], where the T6-O and T6-B represent tests
of T6 impacted by ogival-nosed and bunt-nosed projectiles respec-
tively, and also the same to T3T3-O and T3T3-B. The kinetic energy
absorbed by the plates increase with the increase of the initial pro-
jectile velocity, as shown in Fig. 9(a).
Fig. 10(a), the decrease of the impact velocity of projectiles (Vi)
results in the signicant increase of velocity degradation, as shown
in Fig. 9(b).
Fig. 10(b), the velocity variable vs. initial velocity curves exhibit
almost discontinuous jump close to the ballistic limit velocities.
4. Failure mechanisms
The different failure models may be developed in a target by
changing the impact condition. A failure model with higher energy
absorption can signicantly improve the ballistic resistance of a
shield. The difference in target capacity between different congu-
rations can be related to the transition in deformation mechanism
and failure models. In general, the effect of projectile on the target
plates varies with various parameters such as strength, impact
velocity and nose angle or nose radius of the projectile. Moreover,
the interaction of the perforation process with overall structural
deformation and the localized mechanism of perforation is an
important aspect in analyzing the multi-layered plate problem.
Fig. 11 shows some high-speed camera images of the
perforation process for plates struck ogival-nosed projectiles of

3000 T6-B T6-B

T3T3-B 400 T3T3-B
Kinetic energy variable (J)

T2T2T2-B T2T2T2-B
Velocity variable (m/s)

T6-O T6-O
2500
T3T3-O T3T3-O
T2T2T2-O 300 T2T2T2-O
T6-H T6-H
2000 T3T3-H T3T3-H
T2T2T2-H T2T2T2-H
1500 200

1000
100
200 300 400 500 200 300 400 500
Initial velocity (m/s) Initial velocity (m/s)
(a) Kinetic energy variable vs. initial (b) Velocity variable vs. initial velocity
velocity
Fig. 9. Velocity and kinetic energy variable of projectiles of high strength vs. initial velocity.
 D. Yunfei et al. / Materials and Design 58 (2014) 305315
T6-O
Kinetic energy variable (J)
6000 T3T3-O
T6-B
T3T3-B
5000
4000
3000
2000
400 500 600
Initial velocity (m/s)
(a) Kinetic energy variable vs. initial
velocity
Fig. 10. Velocity and kinetic energy variable of projectiles of low strength vs. initial velocity.
high strength. Note that the given times refer to the image taken
closest to the assumed time of impact. Ogival-nosed projectiles
seem to penetrate the plate mainly by ductile hole enlargement,
pushing the material in front of the projectile aside. No plug is seen
in any test for ogival-nosed projectiles, but petals are formed on
rear side of the indentation.
Photographs of some typical target plates impacted by ogival-
nosed projectiles of high strength are given in Fig. 12. For mono-
lithic plates, the plates failed by hole enlargement that the domi-
nant failure models are signicant localized bulging and petaling
which involve bending and stretch tearing.
While for multi-layered plates, the failure models of layers de-
pend on their set up. For example, the failure mechanism in the
rst layer of T3T3 is severely localized bulging with small cracks
on the tip, whereas in the second layer is localized bulging and
petaling, and also the petals of the second layer is greater than that
of the rst layer, but the number of petals of the second layer is sig-
nicantly fewer than that of the rst layer. Moreover, the perfora-
tion mechanism and failure models of plates of T2T2T2 are the
same to those of T3T3, namely, the plate deformation are domi-
nated by localized bulging and petaling with small global response.
Fig. 11. Selection of high-speed camera images showing perforation of plates impacted by ogival-nosed projectiles of high strength. (a) T6,
(b) T3T3, vi = 432.7 m/s, vr = 266.3 m/s; (b) T2T2T2, vi = 474.4 m/s, vr = 330.4 m/s.
311
500 T6-O
T3T3-O
Velocity variable (m/s)
T6-B
T3T3-B
400
300
700 400 500 600 700
Initial velocity (m/s)
(b) Velocity variable vs. initial
velocity
In addition, the petals of the rst layer are smaller but more than
the second layer which is in turn than the next layer. The second
layer deforms more than the rst layer, the thinning of material
on the bulge is found except the last layer which due to the com-
pression between the projectile and the next layer. Furthermore,
the global deformation of plates increases with the increase of
the number of layers for a given total thickness, as shown in
Fig. 12(b and c). From above analyses, it can be seen that the tran-
sitions of failure models seen for plates impacted by blunt-nosed
projectiles do not take place when impacted by ogival-nosed pro-
jectiles [22], this is also to say that the dominant response does
not change when using multi-layered plates instead of monolithic
plates that the plates response are dominated by bulging and
petaling, while the different trend is seen for tests of blunt-nosed
projectiles of high strength.
In the sub-ordnance impact speed range, previous experiments
focused on effects of plate material properties and the relative size
of projectiles and plate. The effect of projectile deformability was
neglected by using a very hard projectile so there was little plastic
deformation generated in the projectile during the impact process.
In this present investigation, monolithic and multi-layered plates
vi = 341.2 m/s, vr = 93.5 m/s;
312 D. Yunfei et al. / Materials and Design 58 (2014) 305315

Fig. 12. Deformed plates after impacted by ogival-nosed projectiles of high strength (a) T6, vi = 365.47 m/s, vr = 153.03 m/s; (b) T3T3, vi = 420.43 m/s, vr = 253.59 m/s;
(c) T2T2T2, vi = 372.76 m/s, vr = 238.62 m/s.

impacted by projectiles of low strength are also considered in order
to investigate the inuence of yield strength of projectiles on the
failure models and ballistic resistance of plates. Some typical
high-speed camera images from the impact tests when plates im-
pacted by ogival-nosed projectiles of low strength are given in
Fig. 13. The projectiles of low strength yield obvious deformation
after impact the plates, and also the nose shape of projectiles
changes from ogive to hemisphere that somewhat like a cap. In
addition, a visible plug is detected for the monolithic plate, as
shown in Fig. 13(a). Moreover, there are some fragments are found
for the double-layered plates, as shown in Fig. 13(b).
Photographs of monolithic and double-layered plates after im-
pacted by projectiles of low strength are given in Figs. 14 and 15,
and also signicant differences in the failure models during perfo-
ration are detected. When monolithic plates subjected to the im-
pact of ogival-nosed projectiles of low strength, the plates are
caused failure by plugged, and also the plugs are seen to be both
smooth and cylindrical that like caps. It is noticed that the plugs
have a signicant bulge on the distal side, and also a stretch zone
height is observed on the surface of the plug, as shown in
Fig. 14(b). In addition, comparing with the original, all the residual
projectiles lose a little amount mass and their lengths shorten. The
head of projectile obviously become blunter and look like a mush-
room cylinder with a smooth spherical cap, this is because of the
strength of projectile smaller than that of plate that the projectile
yields deformation during impacting, and also the deformation in-
creases the area of the contact surface between projectile and plate
that increases the resistance force to projectile. The dominant fail-
ure models of plates are bulging, where the failure is caused by
plugging due to intense stretch stress. It should be noticed that
the frontal side of indentation is smooth while the rear side of
indentation is coarse, as depicted in second picture in

Fig. 13. Selection of high-speed camera images showing perforation of plates impacted by ogival-nosed projectiles of low strength (a) T6: vi = 559.01 m/s; vr = 217.39 m/s; (b)
T3T3: vi = 496.89 m/s; vr = 289.86 m/s.
 D. Yunfei et al. / Materials and Design 58 (2014) 305315
Fig. 14. Deformed specimens of monolithic plates after impacted by low strength ogival-nosed projectiles (a) vi = 535.12 m/s; vr = 31.06 m/s; (b) vi = 543.48 m/s; vr = 204.6 m/s;
(c) vi = 553.36 m/s; vr = 217.29 m/s.
Fig. 15. Deformed specimens of double-layered plates after impacted by ogival-nosed projectiles of low strength (a) vi = 347.83 m/s; vr = 0 m/s; (b) vi = 502 m/s; vr = 281 m/s;
(c) vi = 362 m/s; vr = 79 m/s.
Fig. 14(b and c), where the impact velocities are high. The photo-
graphs reveal that small cracks start to grow in front of projectile
from the initial impact, then the cracks propagate toward the rear
side of the target plate and a plug is formed. However, the visible
cracks are irregular when the impact velocity is relative low, as
shown in Fig. 14(a).
Relating to the double-layered plates against the impact of pro-
jectiles of low strength, the plates are also caused failure by plug-
ging that two plugs with reduced thickness and diameter are
detected, and also the global deformation is larger than that of
monolithic plates, as shown in Fig. 15. The projectiles cause failure
by plugging due to stretch and bend the material in contact with
projectiles that resulting in circular indentations in the plates that
cracks are detected near the indentations, and also thinning before
fracture can be clearly observed in the rst plates. The second plates
deform much more than the rst plates. Furthermore, the projec-
tiles become shorter and blunter that the nose shape changes from
ogive to somewhat like hemisphere. Moreover, the gures also
show that the dominant failure models of plates are bulging. From
above analyses, it reveals that there is not transition in dominant
failure models can be observed when using multi-layered plates
instead of monolithic plates for ogival-nosed projectiles of low
strength that the dominant failure models are bulging.
The difference of the ballistic limit velocities between targets
with various numbers of layers can also be related to the change
in bend stiffness between layers. The bending stiffness of plate
K = ET3/12(1v2) [24], where E is the elastic modulus, T, the plate
thickness, and v, Poissons ratio. Since bending stiffness follows
313
plate thickness to the third power, for example, the thickness of
T6 is three times thicker than T2, so the stiffness of T6 is 27 times
the size of T2. When the T6 and T2T2T2 bear the same impact
force, the T2T2T2 is much easier to yield deformation than T6, so
the ballistic resistance of T6 is signicantly bigger than that of
T2T2T2, which is possibly the main reason for the increment in bal-
listic limit velocity when T6 compared with T2T2T2, and also the
plastic deformation and mass loss of projectiles of low strength
that perforate multi-layered plates are smaller than perforate
monolithic plates. Thereby, the number of layers has great inu-
ence on the ballistic resistance of plates if the total thicknesses
are equal.
5. Discussion
In this paper, the ballistic tests are conducted to investigate the
inuence of nose shape and strength of projectile on the ballistic
performance of monolithic and multi-layered plates. There are dif-
ferences in the ballistic resistance between monolithic and multi-
layered plates of the same total thickness under various impact
conditions. The results indicate that monolithic plates have higher
ballistic limit velocities than multi-layered plates for ogival-nosed
projectiles regardless their strength, so layering plates decrease the
ballistic limit velocities of plates, this conclusion is in agreement
with the dominant previous work [2,46,8,12,23,25] where the
projectiles with high strength. However, the monolithic plates
have lower ballistic limit velocities than multi-layered plates of
the same total thickness for blunt-nosed projectile of high strength
314 D. Yunfei et al. / Materials and Design 58 (2014) 305315
is in agreement with Refs. [12,1], but it is opposite to Refs.
[1,4,5,7,8,10] that the total thicknesses are small. In addition, the
different results are also found for blunt-nosed projectiles of low
strength that monolithic plates have higher ballistic limit velocities
than multi-layered plates of the same total thickness. From above
analyses, it is also reveals that the monolithic plates have higher
ballistic limit velocities than multi-layered plates for projectiles
of low strength regardless their nose shape. Furthermore, the dif-
ference of the ballistic limit velocities between plates with various
numbers of layers can be related to the change in bend stiffness be-
tween layers and the transitions in deformation mechanism and
failure models of plates.
The contrary nding of layering effect is an interesting phenom-
enon due to the different transitions of the dominant failure mod-
els and energy dissipation manners between monolithic plates and
multi-layered plates, which is in relation to the nose shape and
strength of projectiles. In other words, the difference in target
capacity with different impact conditions can be related to the
change in deformation and failure models. A failure model with
higher energy absorption can signicantly improve the ballistic
resistance of a target. The dominant failure models change from
shearing to tensile stretching and bending which absorb more
kinetic energy of projectile when using multi-layered plates in-
stead of monolithic plates for blunt-nosed projectiles of high
strength, namely, the bending and stretching action can be en-
hanced and thus the multi-layered plates may undergo consider-
able deformation before fracture, so the ballistic limit velocities
of monolithic plates are smaller than those of multi-layered plates.
However, there is no change in energy dissipation models when
moving from monolithic plate to a multi-layered plate for blunt-
nosed projectiles of low strength and ogival-nosed projectiles
regardless of their strength. Furthermore, the bending stiffness of
multi-layered plates is signicantly smaller that of monolithic
plates that result in the decrease of ballistic resistance to projec-
tiles, so the ballistic limit velocities of monolithic plates are bigger
than those of multi-layered plates, and also the ballistic limit
velocities of plates decrease with the increase of the number of
layers.
The strength of projectiles has great inuence on the ballistic lim-
it velocities of targets. It can be concluded that the ballistic limit
velocities of blunt-nosed projectiles are smaller than those of
ogival-nosed projectiles of the same high strength, especially for
the monolithic plates, which is in agreement in nding in Refs.
[1113,2628], but it is opposite to the results in Refs.
[4,5,7,8,29,30] when the total thickness when the total thickness is
small. However, the ballistic limit velocities of blunt-nosed projec-
tiles are bigger than those of ogival-nosed projectiles of the same
low strength, especially for the multi-layered plates.
6. Conclusions
The perforation behavior of monolithic and multi-layered steel
plates impacted by projectiles are studied based on the investiga-
tion of the inuence of the numbers of layers and the strength of
projectiles on the ballistic resistance of plates. Based on the exper-
imental observations, the following conclusions are drawn:
(1) The effect of layering depends on the strength and nose
shape of projectiles. On one hand, the monolithic plates have
lower ballistic limit velocities than multi-layered plates of
the same total thicknesses under impact by blunt-nosed pro-
jectiles of high strength, and also the ballistic limit velocities
increase with the increase of the number of layers. Con-
versely, the effect of layering decreases the ballistic limit
velocities for ogival-nosed projectiles of high strength. On
the other hand, the monolithic plates give greater ballistic
limit velocities than multi-layered plates for projectiles of
low strength regardless their nose shape, and also the ballis-
tic limit velocities decrease with the increase of the number
of layers. The differences in the ballistic limit velocities
between various impact conditions can be related to the
transitions of perforation mechanism and failure models of
plates and projectiles. Moreover, the change in the ballistic
limit velocities is less distinct for high strength than for
low strength projectiles when using the multi-layered plates
instead of monolithic plates, so the layering effect of plates
for the projectiles of low strength is signicantly bigger than
projectiles of high strength, especially for ogival-nosed
projectiles.
(2) The strength of projectiles greatly inuences the ballistic
limit velocities of plates, and also the effect of strength is
in related to the nose shape of projectiles. The projectiles
of high strength are further stronger than projectiles of
low strength, and also the ballistic limit velocities of blunt-
nosed projectiles are more sensitive to their strength than
those of ogival-nosed projectile
(3) The nose shape of projectiles greatly inuences the ballistic
limit velocities of plates, and also the effect of nose shape is
in related to the strength of projectiles. The ballistic limit
velocities of blunt-nosed projectiles are smaller than those
of ogival-nosed projectiles of the same high strength, espe-
cially for the monolithic plates. However, the ballistic limit
velocities of blunt-nosed projectiles are bigger than those
of ogival-nosed projectiles of the same low strength, espe-
cially for the multi-layered plates.
(4) The inuence of plate congurations on the ballistic resis-
tance decreases with the increase of the impact velocity.
The observations illustrate that the initial-residual velocity
curves for projectiles of high strength seem to be rather
close and parallel as the impact velocity becomes high com-
pared to the respective ballistic limit velocities, especially
for blunt-nosed projectiles of high strength.
Acknowledgement
The authors would like to thank the National Natural Science
Foundation of China (No. 11072072) for supporting this present
work.
Appendix A. Supplementary material
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.matdes.2013.12.
073.
References
[1] Zhang Wei, Deng Yunfei, Cao Zongsheng, Wei Gang. Experimental
investigations on the ballistic performance of monolithic and layered metal
plates subjected to impact by blunt rigid projectiles. Int J Impact Eng
2012;49:11529.
[2] Deng Yunfei, Zhang Wei, Cao Zongsheng. Experimental investigation on the
ballistic resistance of monolithic and multi-layered plates against ogival-nosed
rigid projectiles impact. Mater Des 2013;44:22839.
[3] Deng Yunfei, Zhang Wei, Cao Zongsheng. Experimental investigation on the
ballistic resistance of monolithic and multi-layered plates against
hemispherical-nosed projectiles impact. Mater Des 2012;41:26681.
[4] Gupta NK, Iqbal MA, Sekhon GS. Effect of projectile nose shape, impact velocity
and target thickness on the deformation behavior of layered plates. Int J Impact
Eng 2008;35:3760.
[5] Radin J, Goldsmith W. Normal projectile penetration and perforation of layered
targets. Int J Impact Eng 1988;7(2):22959.
 D. Yunfei et al. / Materials and Design 58 (2014) 305315
[6] Almohandes AA, Abdel-Kader MS, Eleiche AM. Experimental investigation of
the ballistic of steelberglass reinforced polyester laminated plates. Compos
Part B: Eng 1996;27:44758.
[7] Nurick GN, Walters CE. The ballistic penetration of multiple thin plates
separated by an air gap. In: Proceedings of SEM conference on experimental
mechanics. Springer; 1990. pp. 631637.
[8] Woodward RL, Cimpoeru SJ. A study perforation of aluminium laminate
targets. Int J Impact Eng 1998;21(3):11731.
[9] Marom I, Bodner SR. Projectile perforation of multilayered beams. Int J Mech
Sci 1979;21(8):489504.
[10] Corran RSJ, Shadbolt PJ, Ruiz C. Impact loading of plates-an experimental
investigation. Int J Impact Eng 1983;1:322.
[11] Teng X, Dey S, Brvik T, Wierzbicki T. Protection performance of double-
layered metal shields against projectile impact. J Mech Mater Struct
2007;2(7):130930.
[12] Dey S, Bovik T, Teng T, Wierzbicki T, Hopperstad OS. On the ballistic resistance
of double-layered steel plates: an experimental and numerical investigation.
Int J Solids Struct 2007;44:670123.
[13] Teng X, Wierzbicki T, Huang M. Ballistic resistance of double-layered armor
plates. Int J Impact Eng 2008;35:87084.
[14] Li Dongquan, Stronge WJ. Ballistic limit of metal plates struck by blunt
deformable missiles: experiments. Int J Solids Struct 2000;37:140323.
[15] Chen XW, Yang YB, Lu ZH, Chen YZ. Perforation of metallic plates struck by a
blunt projectile with a soft nose. Int J Impact Eng 2008;35:54958.
[16] Chen XW, Chen G, Zhang FJ. Deformation and failure modes of soft steel
projectiles impacting harder steel targets at increasing velocity. Exp Mech
2008;48:33554.
[17] Tang P. Experimental investigation on the effect of soft nose of semi-piercing
warhead. Explo Shock Waves 1995(Suppl):292 [in Chinese].
[18] Kim HS, Yeom KS, Kim SS, Sotsky L. Numerical simulation for the front section
effect of missile on the target perforation. In: Proc 22nd Int Symp Ballistics,
Vancouver, Canada, Nov 1418, 2005. p. 1094101.
315
[19] Xiao Xinke, Zhang Wei, Wei Gang, Zhongcheng Mu. Effect of projectile
hardness on deformation and fracture behavior in the Taylor impact test.
Mater Des 2010;31:491320.
[20] GB/T 228-2002. Metallic materials-tensile testing at ambient temperature.
China Standard Press, 2002.
[21] Borvik T, Hopperstad OS, Langseth M, Malo KA. Effect of target thickness in
blunt projectile penetration of Weldox 460 E steel plates. Int J Impact Eng
2003;28:41364.
[22] Recht RF, Ipson TW. Ballistic perforation dynamics. Int J Appl Mech (Trans
ASME) 1963;30:38490.
[23] Yunfei Deng, Wei Zhang, Guanghui Qing Guang, Gang Wei, Yonggang Yang,
Peng Hao. The ballistic performance of metal plates subjected to impacted by
blunt-nosed projectiles of different strength. Mater Des 2014;5:105667.
[24] Zukas JA, Schfer DR. Impact effects in multilayered plates. Int J Solids Struct
2001;38:33218.
[25] Flores-Johnson EA, Saleh M, Edwards L. Ballistic performance of multi-layered
metallic plates impacted by a 7.62-mm APM2 projectile. Int J Impact Eng
2011;38(12):102232.
[26] Borvik T, Langseth M, Hopperstad OS, Malo KA. Perforation of 12 mm thick
steel plates by 20 mm diameter projectiles with at, hemispherical and conical
noses Part I: experimental study. Int J Impact Eng 2002;27:1935.
[27] Dey S, Borvik T, Hopperstad OS, Leinumc JR, Langseth M. The effect of target
strength on the perforation of steel plates using three different projectile nose
shapes. Int J Impact Eng 2004;30:100538.
[28] Wingrove AL. The inuence of projectile geometry on adiabatic shear and
target failure. Metall Trans A 1973;4:182933.
[29] Wilkins ML. Mechanics of penetration and perforation. Int J Impact Eng
1978;16:793807.
[30] Gupta NK, Iqbal MA, Sekhon GS. Effect of projectile nose shape, impact velocity
and target thickness on deformation behavior of aluminum plates. Int J Solids
Struct 2007;44:341139.