International Journal of Impact Engineering 86 (2015) 157e166

Contents lists available at ScienceDirect

International Journal of Impact Engineering

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

An experimental study on the projectile defeat mechanism of hard

steel projectile against boron carbide tiles
S.G. Savio *, Ponguru Senthil, Vajinder Singh, P. Ghoshal, V. Madhu, A.K. Gogia
Defence Metallurgical Research Laboratory, Kanchanbagh, Hyderabad 500058, India

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

Article history: Failure mechanism of 7.62 mm AP projectile impacting boron carbide tiles was studied by post ballistic
Received 21 June 2014 examination of the failed projectiles. It is found that the projectile failure originates simultaneously from
Received in revised form at least two locations in the projectile during ballistic impact, one originating from the target interaction
13 July 2015
front of the projectile and another from a location near the tail end of the projectile. The failure origi-
Accepted 15 July 2015
nation from the target interaction front of the projectile is due to severe deformation at the interaction
Available online 4 August 2015
front. The second failure origination from locations near tail end of the projectile is due to stress wave
generated micro-cracks. At the target interaction front of the projectile a temperature rise of above
Keywords:
Ballistic impact
1160
C has been estimated. No remarkable change in failure mechanism of the hard steel projectile with
Projectile failure respect to tile thickness (5e9.2 mm) or projectile velocity (600e820 m/s) has been observed. The vari-
Projectile dwell ation in remnant projectile weight with respect to tile thickness and projectile velocity can be attributed
Projectile ceramic interaction to the change in duration of dwell or total projectile interaction time with ceramic target.
Armor-piercing 2015 Elsevier Ltd. All rights reserved.

1. Introduction the dwellepenetration transition velocity with tungsten long

Historically, ceramic composite armor systems were designed to
defeat armor piercing (AP) kinetic energy projectiles mainly in
small arms and heavy machine gun category [1,2]. The AP pro-
jectiles used in small arms are generally made with hard steel (HRc
60e64) core, but some are even made with harder tungsten carbide
(WC) cores [3]. The projectile core is covered with a thin ductile
metal jacket for interior ballistics or aerodynamic considerations
[3]. The penetration performance of the bullet however is
controlled only by the core properties [3]. Small arm projectiles
usually have a length to diameter (L/D) ratio in the range of 3e5
with muzzle velocities of less than 1 km/s [3].
The projectile penetration and its failure mechanism on impact
with brittle materials have been studied by many authors during
the past two decades [4e12]. Michael et al. [4] used ash X-rays
to capture the penetration event of 7.62 mm AP projectile into
boron carbide target. They observed that while at 6 ms the nose
of the projectile core eroded, the projectile did not penetrate into
the ceramic. The projectile core started penetration into the
ceramic material only between 16 and 25 ms. Leavy et al. [5] studied
* Corresponding author. Armour Design and Development Division, Defence
Metallurgical Research Laboratory, Hyderabad 500058, India. Tel.: 91 40
24588010; Fax: 91 40 24588504.
E-mail address: geasin@dmrl.drdo.in (S.G. Savio).
rods into silicon carbide ceramic targets and found that the
dwellepenetration transition velocity was primarily determined by
the material properties and not by the thickness of the target.
However, it was observed that the penetration-rate was inuenced
by the ceramic target thickness. In studies by Anderson et Al. [6,7] it
has been shown that tile thickness has inuence on duration of
dwell, but duration of dwell is relatively independent on projectile
velocity. As per Vlasov et al. [8], the penetration of hard projectile
(L/D 10) into ceramic target can be considered as a two stage
process. During the rst stage the penetration velocity increases as
the ceramic is progressively damaged and during the second stage
the penetration velocity reaches a quasi-steady state value where
projectile penetrates into the failed ceramic. Sinani et al. [9] studied
the penetration process of hard steel (hardness e 8 GPa) and WC
(hardness e 17 GPa) projectiles in alumina and sapphire targets.
They found that alumina and sapphire performed similarly against
steel projectile. Against a WC cored projectile, the alumina target
defeated the shot by fragmenting the bullet into pieces, whereas in
sapphire the projectile perforated the target without being
deformed. The inferior ballistic performance of sapphire against
WC core was attributed to the decrease in sapphire hardness below
WC hardness with increase in load.
It is important to understand that the projectile impact event is
a high strain rate phenomenon, and the fracture patterns in high

http://dx.doi.org/10.1016/j.ijimpeng.2015.07.011
0734-743X/ 2015 Elsevier Ltd. All rights reserved.
158 S.G. Savio et al. / International Journal of Impact Engineering 86 (2015) 157e166
strain rate events are likely to be different from quasi-static fracture
events because during high strain rate event, fracture is indepen-
dently nucleated at many sites [13,14]. Further, in quasi-static
conditions the surface/boundary defects (aws) play a major role
since the crack initiation takes place from those areas, whereas in
high strain rate phenomena the spall fracture occurs at the interior
of the body unaffected by the surface defects [14]. Therefore, the
fracture morphologies observed in dynamic fracture process are
expected to be different from quasi-static one.
Even though extensive literature on projectile penetration is
available, not many studies pertaining to post ballistic examination
on failed projectile have been published. In the present study,
projectile failure mechanism of 7.62 mm AP projectiles when
impacted against boron carbide tiles having different thicknesses
and varying projectile velocities has been investigated. Detailed
observations on the fracture surfaces as well as interior regions of
the failed projectiles have been employed to identify the failure
mechanisms.
2. Experimental
2.1. Projectile and target
The 7.62 mm AP projectiles used in this study consisted of a hard
steel core covered with a copper sheath. The steel core had a
diameter of 6.1 mm and a length of 28.4 mm with a mass of 5.3 g.
The nominal weight of the projectile including the sheath was
10.4 g. Photographs of typical projectile with jacket along with hard
steel core are shown in Fig.1. The hardness of hard steel projectile is
found to be 765 7 VHN.
Tiles of boron carbide used in these experiments were of 40 mm
diameter. Four different tile thicknesses i.e. 5 mm, 6 mm, 7 mm, and
9 mm were used to study the effect of target thickness for which a
projectile velocity of about 820 m/s was used. The 6 mm thick boron
carbide tiles were used to study the projectile velocity effect with
600 m/s and 700 m/s velocities. Aluminium alloy (6063-T6) blocks
with 50 mm thickness were used as backing material for all the
tests. The properties of boron carbide and backing material can be
referred in our previous publication [15].
Fig. 1. Photograph of the 7.62 mm AP projectile (a) covered with copper jacket and (b)
hard steel core without jacket.
Steel
confinement
Al-alloy backing
Ceramic tile
Projectile
Penetration
Fig. 2. Schematic of the hard steel projectile penetration process into the ceramic and
Al-alloy target in DOP test conguration.
2.2. Ballistic test conguration
The DOP tests were conducted as per the conguration shown in
Fig. 2. The detailed experimental setup has been described else-
where [15]. The boron carbide tiles were tightly tted in a steel
connement using brass shim and these conned tiles were placed
over the aluminium alloy backing material without application of
any bonding material. The 7.62 mm AP projectiles were red at
these targets from a ried gun. The debris produced from the
projectile and ceramic tile during ballistic impact event was
collected using a front fabric covered steel box. The remnant broken
shots present along with the collected boron carbide debris were
separated using magnetic separation for further examination.
3. Results and discussion
3.1. Studies on the target interaction front of the projectile
A detailed post ballistic study on the target interaction front of
the failed projectiles, after interaction with boron carbide and
backing aluminium alloy, was carried out. A typical failed projectile
after ballistic impact is shown in Fig. 3. The projectile is intact along
its longitudinal axis and failure is observed only at the target
Fig. 3. Failed 7.62 mm AP projectile after ballistic impact with 5 mm thick boron
carbide tile backed by Al6063 alloy.
 S.G. Savio et al. / International Journal of Impact Engineering 86 (2015) 157e166
Fig. 4. Typical fracture surface observed at the target interaction front of a 7.62 mm AP
projectile failed due to ballistic impact with boron carbide tile backed by Al6063 alloy.
interaction front. A close-up view of the target interaction front of
the failed projectile is shown in Fig. 4. This fracture surface has two
dominant features, a comparatively larger region which has smooth
surface and remaining smaller region with rough surface. Similar
features were observed in many failed projectiles especially when
the remnant projectile had longer length. The chemical composi-
tion of the smooth surface region in Fig. 4 was analysed with energy
dispersive X-ray spectroscopy (EDS) technique and the results are
given in Fig. 5. It was found that the smooth surface region was
enriched with aluminium. It suggests that the aluminium on the
target interaction front of the failed projectiles should have come
from the backing aluminium alloy during penetration into it after
defeating the ceramic target.
At higher magnications, the smooth regions show at least two
surface morphologies. The rst surface feature is a solidied molten
metal ow pattern as shown in Fig. 6a and the second surface
feature is a ow pattern with tear drops (Fig. 6b). Since the impact
velocity of the projectile in this study is relatively lower in com-
parison with long rods, material strength effect [6,16] of the pro-
jectile and target should be considered and it is unlikely for the
hydrodynamic ow to occur. Therefore, the presence of molten
aluminium at the target interaction front of the remnant projectile
indicates that the temperature at the target interaction front of the
projectile should have reached above the melting temperature of
aluminium alloy 6063 which is in between 616 and 654
C [17].
Moreover, many of the remnant projectiles with shorter length
(like Fig.10a) did not exhibit smooth failed surface and had only
Fig. 5. EDS analysis on the smooth surface region of the target interaction front of the projectile (see Fig. 4) shows the presence of aluminium.
159
rough fracture surface regions. It suggests that the recovered
remnant projectile should have separated from the front end of the
projectile before or during penetration into the aluminium alloy
backing.
The magnied image of the target interaction front of the failed
projectile at location E on Fig.11b is shown in Fig. 7. The SEM
micrograph (Fig. 7a) shows three distinct regions with varying
contrasts. The chemical analysis using EDS on the inner layer (L)
shows only Fe, while the middle layer (M) is iron rich with a
composition of Fe e 58.5 at.% and Al e 41.5 at.% . An analysis on
the dendritic region of the outside layer (R) shows a composition of
Fe e 26.0 at.% and Al e 74.0 at.% as shown in Fig. 7b and c. From the
FeeAl phase diagram [18] it is found that the composition of the
dendritic region is very close to the non-stoichiometric interme-
tallic compounds such as Fe2Al5 and FeAl3. Therefore, the phases
were identied by using electron backscatter diffraction (EBSD)
which obtains crystallographic information at the microscopic level
by kikuchi diffraction technique. Crystallographic information such
as lattice parameters, space group, symmetry elements, atomic
position, occupancy of expected phases from FeeAl phase diagram
were fed into the EBSD software. The EBSD software database was
matched with experimental pattern to identify the phases in den-
dritic region. The EBSD (Fig. 8) analysis clearly showed the pre-
dominant presence of FeAl3 phase in the dendritic region. Since,
dendrites of non-stoichiometric intermetallic compound FeAl3, a
phase which forms above 1160
C, has been observed at the target
interaction front of the projectile, the temperature rise at this re-
gion could have reached above 1160
C [18]. Further, the shear
deformation observed at the target interaction front of the hard
steel projectile is shown in Fig. 9. This shear deformation occurs due
to rise in temperature and large stresses generated at the target
interaction front of the projectile on interaction with ceramic
target.
3.2. Studies on the failure mechanism of the projectile
A study on the mechanism of failure of hard steel projectiles
during ballistic impact with ceramic targets was carried out by
examining different fracture patterns present in the failed pro-
jectiles. While examining various failed projectiles, it was observed
that some of the projectiles have failed along its longitudinal axis
and others have not as shown in Fig.10. In order to understand this
difference in failure patterns of projectiles, the projectiles which
have not failed along the longitudinal axis were cut along its lon-
gitudinal axis for further studies. The longitudinal sections of two
projectiles, which have interacted with 5 mm and 6 mm thick boron
carbide tiles respectively, are shown in Fig.11. Close examination on
160 S.G. Savio et al. / International Journal of Impact Engineering 86 (2015) 157e166

Fig. 6. SEM microstructure of the interaction front of 7.62AP projectile after interaction with 5 mm thick boron carbide tile. (a) Solidied molten metal ow pattern of aluminium
alloy is seen; (b) ow pattern with tear drop pattern is seen.

Fig. 7. (a) Molten aluminium region found at the target interaction front of the projectile on the longitudinal cross-section of the failed projectile with three distinct regions L, M
and R; (b) dendrite formation found at the region R; (c) EDS analysis of L, M and R regions.
 S.G. Savio et al. / International Journal of Impact Engineering 86 (2015) 157e166
Fig. 8. EBSD analysis on the dendrites of molten aluminium region on the target interaction front of the projectile along the longitudinal cross-section of the failed projectile. (a and
b) Secondary Electron image. (c) EBSD phase mapping shows the presence of FeAl3 phase.
the longitudinal cross section of the failed projectiles revealed the
presence of two independent failure origins located at two different
locations in the projectile. The rst common type of failure origin is
marked as A in Fig.11a and b which is found at the target inter-
action front of the projectile. The rst type of failure initiation is
mainly due to impact induced stresses generated at the target
interaction front of the projectile on ballistic impact with high
hardness ceramic target. The decrease in projectile length due to
ballistic impact occurs mainly due to failure initiation from rst
type of failure origin.
The second type of failure origin is found at a place which is far
away from the target interaction front of the projectile and closer to
projectile tail end. One such type of failure is marked as B in
Fig.11a. The stress conditions at the location of second failure origin
are not clearly known. But the initiation of second type of failure
may not be due to spall since the occurrence of spall fracture re-
quires interaction of two strong plane tensile waves under uniaxial
conditions [14]. Further, the spall fracture occurs in a characteristic
Fig. 9. Shear band observed at the target interaction front of the 7.62 mm AP projectile
due to interaction with hard ceramic target.
161
plane which lies perpendicular to the direction of wave propaga-
tion and in this case the characteristic plane is expected to lie
perpendicular to the longitudinal axis of the projectile. Notwith-
standing the reasons for failure initiation, a conclusion can be
drawn that during projectile impact if the rst failure origin dom-
inates then the length of projectile decreases gradually from the
target interaction front of the projectile and if the second failure
origin dominates then projectile failure occurs along the longitu-
dinal axis of the projectile.
Further, examination along the longitudinal cross-section of the
failed hard steel projectile revealed a number of dynamic crack
nucleation sites as shown in Fig.12a. These dynamic micro-cracks
are uniformly distributed throughout the longitudinal cross-
section of the projectile. It is also seen from Fig.12a and b that
many of the dynamic crack nucleation sites have their origination at
the triple points and grain boundaries. This type of dynamic crack
nucleation is generally caused by tensile stress waves which are
produced due to reection of compressive stress waves from free
surfaces and these failure features are very common in high strain
rate failure of materials [13,14,19]. Also the tensile stress amplitude
required for dynamic crack nucleation process is equal to its static
value for crack nucleation [14]. Therefore these micro voids should
have been produced at lower intensity stress waves. Also from
Fig.12c, which is observed near B region of Fig.11a, it is seen that
Fig. 10. Failure of 7.62 mm AP projectile after impacting with boron carbide tiles. (a)
Projectile which has failed along its transverse axis. (b) Projectile which has failed
along its longitudinal axis.
162 S.G. Savio et al. / International Journal of Impact Engineering 86 (2015) 157e166

Fig. 11. The longitudinal cross-section of the 7.62 mm AP projectiles which have impacted with (a) 5 mm thick boron carbide tile and (b) 6 mm thick boron carbide tile. (A) Target
interaction front; (B) projectile tail region; (C and D) stress wave generated cracks; (E) formation of thin layer of FeAl3 crystals.

Fig. 12. SEM microstructure of the polished and etched surface of failed 7.62 mm AP projectile along the longitudinal cross-section shows (a) the presence of uniformly distributed
stress wave generated crack nucleation sites; (b) the presence of crack nucleation sites at the inclusion boundaries; (c) crack formation through coalescing together of micro-cracks;
(d) smaller cracks joining together to form larger crack.
 S.G. Savio et al. / International Journal of Impact Engineering 86 (2015) 157e166 163

Fig. 13. Magnied images of different locations in Fig. 10(a) shows (a) the crack propagation along stress wave generated crack nucleation sites observed in region C and (b) zigzag
inter-granular crack propagation mode (shown with arrow marks) observed at the leading edge of the propagating crack in both B and C regions.

Fig. 14. SEM image of the fracture surface of failed 7.62 mm AP projectile due to ballistic impact.

nearby micro-cracks coalesce together to produce crack extension.
These extended small cracks join together to form larger cracks as
shown in Fig.12d. Therefore, fragmentation of projectile along the
axis occurs due to extension of these cracks during projec-
tileeceramic impact process.
The magnied images of the region C (see Fig.11a) on the
longitudinal cross-section of the failed hard steel projectile are
shown in Fig.13. Similar to the previous observations, numerous
dynamic crack nucleation sites visible adjacent to the crack paths in
Fig.13a clearly reveal the dynamic crack propagation mechanism
through bridging of micro-cracks within the projectile. Further, the
zigzag crack path present at the leading edge of the propagating
crack is shown in Fig.13b. This feature which was observed at the
leading edge of the propagating crack in both B and C regions
(Fig.11a) of the projectile is nothing but the inter-granular crack
propagation within the projectile. This inter-granular crack prop-
agation is conrmed from the fracture surface of the failed hard
steel projectile as shown in Fig.14. The fracture surface exhibits

Fig. 15. Polished and etched surface of the longitudinal cross-section of the failed projectile at the target interaction front of the projectile shows the presence of shear bands
(marked with arrows) at the interface as well as nearby subsurface.
164 S.G. Savio et al. / International Journal of Impact Engineering 86 (2015) 157e166

Fig. 16. (a) Crack propagation through shear band and (b) interaction among shear band cracks produce ne fragmentation at the target interaction front of the projectile.

Fig. 17. Fracturing of projectile into ner particles due to interaction among stress wave generated cracks (magnied image of region D in Fig. 10(b)).

inter-granular fracture features and dimples with localised ductile
deformation. The dimples present in the fracture surface are due to
pulling out of carbide particles during localised plastic deformation.
The zigzag crack propagates in both trans-granular (the interface
between carbide particles and the matrix) and inter-granular (grain
boundaries) fashion resulting in a mixed fracture mode.
The presence of adiabatic shear bands at the target interaction
front of the failed projectile is shown in Fig.15. These shear bands
are only observed in areas which are very close to the target
interaction front of the projectile. Crack propagation through shear
bands as well as interaction among shear band generated cracks is
shown in Fig.16. It is suggested that the failure of projectile due to
shear bands leads to ne fragments. The ne fragmentation of
projectile is also produced by interaction among stress-wave
generated cracks (see D region in Fig.11b) as shown in Fig.17,
but observed only at areas closer to target interaction front of the
projectile. Interaction of shear band generated cracks with stress
wave generated cracks is shown in Fig.18. These observations
clearly indicate that the ne fragmentation produced near the
target interaction front of the projectile is due to interaction of at
least two fracture modes like shear band cracks and stress wave
generated cracks.
3.3. Effect of tile thickness and projectile velocity on projectile
defeat
The results on the effect of tile thickness and projectile velocity
on the size of failed projectile have been published in our earlier
publication [15]. It is observed that as the ceramic tile thickness
increases the average recovered projectile weight decreases. In
other words, larger quantity of projectile is consumed. Though
the remnant projectile weight is found to decrease with increasing
tile thickness, the study of fracture surfaces revealed similar kind
of fracture features (see Fig.14) in all the failed projectiles

Fig. 18. Interaction of shear bands (marked with black arrow) with the stress wave generated cracks (marked with white arrow) near projectileeceramic interface which produce
ne fragmentation of the projectile.
 S.G. Savio et al. / International Journal of Impact Engineering 86 (2015) 157e166
Fig. 19. The 7.62 mm AP projectile fracture surface due to ballistic impact with boron carbide tiles with projectile velocities (a) 600 m/s, (b) 700 m/s, and (c) 820m/s shows similar
inter-granular fracture with carbide particle pull out features.
independent of tile thickness. It is also known that the duration of
projectile dwell on ceramic tile is related to the delay in failure of
ceramic material during ballistic impact. Hence, the dwell of pro-
jectile on ceramic target depends on the ceramic tile thickness and
backing material properties [6]. Since the failure types observed in
failed projectiles with respect to different tile thicknesses are
similar, the decrease in the weight of recovered remnant failed
projectile with respect to increase in tile thickness can be attributed
to the increase in duration of dwell of the projectile with ceramic
target.
The effect of projectile velocity on the projectile defeat mecha-
nism was also studied in a similar manner like that of tile thickness.
As the projectile velocity increases from 600 m/s to 820 m/s, a
marginal increase in average recovered remnant projectile weight is
observed with respect to projectile velocity [15]. The fracture sur-
faces of 7.62 mm AP projectiles with different velocities, which had
interacted with boron carbide tiles of same thickness, are shown in
Fig.19. The fracture surfaces of all three projectiles show similar
features and no difference in failure types between different pro-
jectile velocities is observed within this studied velocity range. It is
also expected that the duration of dwell time does not change within
variation of few hundred meters per second in projectile velocity [6].
Therefore, the marginal increase in recovered projectile weight of
the failed projectile with respect to projectile velocity can be solely
attributed to the reduction in total interaction time of projectile with
the ceramic target as the projectile velocity increases.
4. Conclusions
The fracture modes and failure mechanism of 7.62 mm AP pro-
jectiles, on impacting boron carbide tiles backed by aluminium
165
alloy 6063 target, were studied through microstructural examina-
tion of the failed projectiles. From this study the following con-
clusions are derived:
(1) At the target interaction front of the projectile, features like
solidied molten aluminium as well as new dendritic crys-
tals of FeAl3 were observed. This indicates a temperature rise
of over 1160
C at the target interaction front of the projectile.
(2) Disintegration of projectiles into large pieces is primarily due
to interaction of stress wave generated micro-cracks with
impact induced inter-granular cracks. Fine fragmentation
occurs only at the target interaction front of the projectile
and is mainly due to interaction among shear band generated
cracks and stress wave generated cracks.
(3) While the macroscopic failure of projectile along its radial
cross-section is mainly due to extension of cracks originating
from the target interaction front of the projectile, the failure
along its longitudinal cross-section is due to joining of cracks
originating from near tail end region of the projectile with
cracks originating from target interaction front of the
projectile.
(4) Tile thickness and projectile velocity do not appear to have
any signicant effect on fracture features or failure mecha-
nism. Reduction in weight of remnant projectile at higher
thickness or lower velocities is due to increase in dwell or
total interaction time of projectile with boron carbide tile.
Acknowledgements
The authors thank DRDO for funding this research work and
Director, DMRL, for granting permission to publish this work. The
166 S.G. Savio et al. / International Journal of Impact Engineering 86 (2015) 157e166
support of all staff members of Armour Division of DMRL on con-
ducting the ballistic tests is gratefully acknowledged. Further, help
provided by Mr. P Kamalakar Rao and other groups of DMRL is also
acknowledged.
References
[1] Viechnicki DJ, Slavin MJ, Kliman MI. Development and current status of ar-
mour ceramics. American Ceramic Society Bulletin 1991;70(6):1035e9.
[2] Medvedovski E. Alumina ceramics for ballistic protection e Part 1. Am Ceram
Soc Bull 2002;81(3):27e32.
[3] Gooch WA. An overview of ceramic armor applications, ceramic armour
materials by design. Ceram Trans 2002;134:3e21.
[4] Normandia MJ, LaSalvia JC, Gooch Jr WA, McCauley JW, Rajendran AM. Pro-
tecting the future: ceramic research leads to improved armour performance.
The AMPTIAC Quarterly 2004;8(4):21e7.
[5] Leavy B, Krautauser C, Houskamp J, LaSalvial J. Fundamental investigation of
high-velocity impact of ductile projectiles on conned ceramic target. In:
Parmentola JA, Rajendran AM, editors. Orlando, GL: Tech Science Press; 2006.
Proceedings of the 25th Army science conference.
[6] Anderson Jr CE, Walker JD. An analytical model for dwell and interface defeat.
Int J Impact Eng 2005;31(7):1119e32.
[7] Anderson Jr CE, Orphal DL, Johnson GR, Grosch DJ, Walker JD, Beissel SR. An
investigation to develop ultralight weight armour. SwRI Report 18e1147,
Southwest Research Institute, San Antonio, TX, September. 2000.
[8] Vlasov AS, Emelyanov YA, Zilberband EL, Kozhushko AA, Kozachuk AI,
Pugachev GS, et al. Kinetics of high velocity penetration into high-hardness
brittle media. Phys Solid State 1999;41(10):1638e40.
[9] Sinani AB, Kozhushko AA, Zilberbrand EL. Resistance of brittle solids to high-
velocity projectile penetration at the initial stage of impact. Tech Phys Lett
2008;34(2):103e5.
[10] Hazell PJ, Roberson CJ. On the Critical Thickness of Ceramic to Shatter WC-Co
Bullet Cores. In: Flis WJ, Scott BR, editors. Canada: DESTech Publications;
2005. Proceedings of the 22nd International Symposium on Ballistics, Van-
couver, 14e18 November.
[11] Horsfall I, Edwards MR, Hallas MJ. Ballistic and physical properties of highly
fractured alumina. Adv Appl Ceram 2010;109(8):498e503.
[12] Hazell PJ, Appleby-Thomas GJ. More on penetration of ceramic based targets
by non-deforming projectiles. Adv Appl Ceram 2012;111(3):171e3.
[13] Meyers MA. Dynamic behavior of materials. New Jersey, USA: John Wiley &
Sons, Inc; 1994. p. 488e90.
[14] Antoun T, Curran DR, Razorenov SV, Seaman L, Kanel GI, Utkin AV.
Spall fracture. In: Davison L, Horie Y, Graham RA, editors. High-pressure
shock compression of condensed matter. New York: Springer-Verlag; 2003.
p. 1e33.
[15] Savio SG, Ramanjaneyulu K, Madhu V, Balakrishna Bhat T. An experimental
study on ballistic performance of boron carbide tiles. Int J Impact Eng
2011;38(7):535e41.
[16] Anderson Jr CE, Orphal DL. Analysis of the terminal phase of penetration. Int J
Impact Eng 2003;29:69e80.
[17] Al 6063-T6 data sheet, ASM Aerospace Specication Metals, Inc [online],
http://asm.matweb.com/search/SpecicMaterial.asp?bassnumMA6063T6.
accessed 12.12.14.
[18] Kottcamp Jr EH, Langer EL. ASM handbook. In: Baker H, editor. Alloy Phase
Diagrams, vol. 3. Materials Park, Ohio: ASM International; 1992. p. 44.
[19] Erzara B, Buzaud E. Shockless spalling damage of alumina ceramic. Eur Phys J
Spec Top 2012;206:71e7.