You are on page 1of 12

Materials and Design 91 (2016) 294–305

Contents lists available at ScienceDirect

Materials and Design

journal homepage:

Design and impact resistant analysis of functionally graded Al2O3–ZrO2

ceramic composite
Chin-Yu Huang a, Yu-Liang Chen b,⁎
School of Defense Science, Chung Cheng Institute of Technology, National Defense University, No. 75, Shiyuan Rd., Daxi Dist., Taoyuan City 33551, Taiwan
Dept. of Power Vehicle and Systems Engineering, Chung Cheng Institute of Technology, National Defense University, No. 75, Shiyuan Rd., Daxi Dist., Taoyuan City 33551, Taiwan

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

Article history: Gradient concept is one of the solutions in advanced material design and structure control. Functionally graded
Received 14 August 2015 materials (FGMs) are materials which change their composition, constitution or structure continuously along
Received in revised form 23 November 2015 the direction of thickness; the property and function of those materials appear as changeable gradient character-
Accepted 24 November 2015
istics. In this research, powder metallurgy method was used to make the functionally graded Al2O3–ZrO2 com-
Available online 2 December 2015
posite, which was backed with 6061-T6 aluminum alloy plate. For comparing purpose, we also make
multilayer Al2O3–ZrO2 composite and pure Al2O3 composite under the same areal density. Based on NIJ IV stan-
FGM dard, ballistic test was executed on the three kinds of samples by using 0.30″ armor piercing projectile (AP). Finite
Ballistic test element software ANSYS/LS-DYNA was also used to simulate stress distribution and fracture of these composites
ANSYS/LS-DYNA under impact. It was shown from ballistic test that FGM had the best impact resistant performance. Normally,
Impact resistance multilayer ceramic structure will be fractured after impact because of the tensile wave caused by the interlayer.
Microstructure Through SEM observation, delamination was not found in the interlayer of FGM. This phenomenon reduced the
tensile wave occurred in the interface between layers and delayed the crack propagation, resulting in a higher
ballistic resistance capability structure.
© 2015 Elsevier Ltd. All rights reserved.

1. Introduction inflicted on the ceramic and that the rigid adhesive reduced the fractur-
ing more than the soft adhesive did.
Ceramics have been used in protective equipment for more than 50 Arias et al. [5] developed a novel adhesive by using ceramic particles
years. In the Vietnam War, ceramics were widely used for the first time and a polymeric matrix. They applied the adhesive between ceramic
in bulletproof vests and helicopter seats [1]. Ceramics have received and metal plates and conducted ballistic experiments to test the
considerable international attention in recent years and have been sub- bullet-resistant abilities of the plates.
jected to extensive research. The increasing use of small-caliber armor- Lopez-Puente et al. [6] investigated the influence of the thickness of
piercing bullets on battlefields has necessitated the development of an adhesive on the impact resistance of ceramic–metal composite
novel light-weight armor [2]. Numerous recent studies have investigat- plates. The results showed that reducing the thickness of the adhesive
ed the wave impedance performance of ceramics and rear plate combi- layer reduced the shear-stress-induced fracturing on the rear plate,
nations, adhesives used in between materials, sandwich layer structural and the simulation software program they used indicated an optimal
material combinations, and multilayer material combinations. adhesive layer thickness of 0.3 mm.
Grouch et al. [3] used DRA-DYNA-2D to simulate the impact resis- Chen et al. [7,8] combined two layers of aluminum with aluminum
tance of aluminum multilayer plates having various adhesive layer foam in a sandwich structure and conducted ballistic tests as well as
thickness levels. They also used SHPB to test the dynamic behavior of an LS-DYNA numerical simulation to analyze the effect of the projectile
epoxy resin at varying strain rates and temperatures and modified the shape on the multilayer structure. Furthermore, they derived mathe-
Cowper–Symonds equation. matical models for penetration depth, ballistic limit, and residual termi-
Zaera et al. [4] used two types of adhesives, namely polyurethane nal velocity.
(soft adhesive) and epoxy resin (rigid adhesive), to conduct impact re- The variations in the elastic modulus, hardness, and density between
sistance experiments on ceramic composite materials. For each type of a ceramic front plate and a tough rear plate lead to a mismatch of inter-
adhesive, three thicknesses were considered: 0.5, 1.0 and 1.5 mm. The layer wave impedance and localization of stress. The tensile wave in-
results showed that the 1.5 mm adhesive reduced the fracturing duced by a bullet penetration causes severe fracturing to the ceramic
front plate; however, when mismatched adhesives are used, the impact
⁎ Corresponding author. wave can easily cause interlayer delamination, considerably reducing
E-mail address: (Y.-L. Chen). the ability of the ceramic armor to resist multiple impacts. Regarding
0264-1275/© 2015 Elsevier Ltd. All rights reserved.
C.-Y. Huang, Y.-L. Chen / Materials and Design 91 (2016) 294–305 295


C wave velocity
E Young's modulus
F reflection coefficient
M material parameter
T transmission coefficient
V velocity
X percentage by volume
ΔE absorptive energy
σ strength
ρ density Fig. 1. Radical weight reduction for future ground vehicles.

ν Poisson's ratio

the fabricated material. Subsequently, they used bullets measuring

7.62 mm to compare the ballistic resistance. The results revealed that
a initial velocity
the FGM produced using the SPS method exhibited excellent ballistic re-
b residual velocity
sistance; small grain growth and denser materials can be attained by
c material (Al2O3)
using the SPS method and short-duration low-temperature sintering.
d material (ZrO2)
Sun. et al. [12] discovered that ceramic sintering produces residual
i incident wave
stress, mainly because of a dissimilar sintering temperature and thermal
t transmission wave
expansion coefficient. They designed an optimal shrinkage and
r reflection wave
sintering curve, producing a three-layer Al2O3/ZrO2 FGM.
0 tensile
Zhang et al. [13] used a DOP method to analyze the ballistic re-
1 front plate
sistance of materials containing 95% alumina ceramic and 10%
2 rear plate
zirconia-toughened alumina (ZTA) ceramic when impacted by

the establishment of a mathematical model for sandwich structures or

multilayer structures, the condition of adhesive layers has been
disregarded in most assumptions because the differences in the wave
impedance vary considerably with the adhesive layer, and the produced
reflection stress wave is relatively complex. Functionally graded mate-
rials (FGMs) can help reduce the post-impact delamination between ce-
ramic and metal plates and alleviate the effect of the reflection waves;
thus, the twenty-first century scholars have proposed new graded
Impact resistance design emphasizes light weight, low cost, and sim-
ple manufacturing processes; therefore, “gradient” is a crucial novel
technology in material design and structure control. FGMs are novel
materials with continuously changing constituent elements (composi-
tion and structure), for example, continuous changes in thickness.
Such changes lead to gradient variations in the properties and functions
of materials. FGMs were first designed in the field of aeronautics and as-
tronautics. Once the concept of such design emerged, it received consid-
erable attention among international scholars and it has already been
applied in electronics, nuclear energy, optics, biomedicine, and the mil-
itary. Weapon attacks and armor protection have been improved be-
cause of the technological advancement. When considering protection
and mobility, increasing impact resistance capability and reducing
weight is always an objective in armor design. Since 2001, the U.S. De-
partment of Defense has allocated a large yearly budget to FGM research
to develop lighter and stronger protective equipment that can with-
stand multiple impacts as well as to conduct research on novel material
manufacturing processes. The Department has also gradually published
numerous research reports. Therefore, the design of protective battle-
field materials is trending toward lightweight materials (Fig. 1) [9].
Guo et al. [10] used hydroxyapatite (HA) ceramics and yttria-
stabilized tetragonal zirconia (Y-TZP) composite powders and adopted
the spark plasma sintering (SPS) method to manufacture HA/Y-TZP
FGM. A bending test revealed that the strength of the produced FGM
was 200 MPa, which was twice that of pure HA ceramics.
Pettersson et al. [11] employed SPS technology to produce a TiB2–Ti
graded material and applied a hot isostatic pressing sintering method to Fig. 2. Specimens for ballistic test.
296 C.-Y. Huang, Y.-L. Chen / Materials and Design 91 (2016) 294–305

Fig. 3. Profile of 0.30″ AP bullet and target.

long rod tungsten projectiles. They found that the ZTA ceramic en-
hanced ballistic resistance.
Ewais et al. [14] designed a tailored zirconia–mullite/(0–100 vol%)
alumina FGM. They used reaction sintering to produce an 11-layer
FGM; a mechanical property test indicated that the FGM possessed
high strength and mechanical and thermal properties.
Gupta et al. [2] used SPS sintering to produce TiB2/(10, 20 wt.%)Ti
FGM. They used small-caliber armor-piercing bullets to test the ballistic
resistance of the fabricated material. The results showed that the ballis-
tic efficiency (η) of the TiB2/Ti FGM, TiB2/(10 wt.%)Ti, TiB2/(20 wt.%)Ti,
and TiB2 (monolithic) composite materials were in the range 5.1–5.9.
In addition, the ballistic efficiency decreased when the thickness of the
FGM was increased from 5 to 7.8 mm.
Übeyli et al. [15] used powder metallurgy to produce SiC/(80, 90,
100 wt.%) Al 7075 and SiC/(60, 80, 100 wt.%) Al 7075 FGMs. They
used armor-piercing bullets to test impact resistance and analyzed the
fracture behavior in the microstructure of the ceramic by using scanning
electron microscopy (SEM) to irradiate fragments. The results revealed
that FGM (up to a thickness of 25 mm) did not exhibit improved ballistic
Lin and Xue et al. [16,17] implanted a TiNi Shape Memory Alloy
(SMA) into a fiber composite material and then used 9 mm full metal
jacket bullets to test ballistic resistance. They altered the arrangement,
quantity, and thickness of the SMA and adopted the Taguchi method
and ANOVA analysis method to analyze absorptive energy. They con-
Fig. 5. Model diagram of numerical simulation.
cluded that the optimal combination resulted in a 19% increase in the
absorptive energy. Furthermore, they used the laws of thermodynamics
to calculate a phenomenological constitutive model for FGM-SMA. This element model, to simulate a uniaxial compression test on FGM-SMA
model was then combined with another mathematical model, a finite cylinders. The simulation and test results revealed the mean stress
and strain relation, cross-sectional stress distribution, and absorptive
energy properties of the FGM-SMA cylinders.
The preceding literature review indicates that many scholars
have already conducted experiments and simulations on ceramic-
toughened composite materials and FGM design and evaluated the
impact resistance of the materials. However, few studies have been
published on the processes of fabricating materials and evaluating
the mechanical properties of ceramic–ceramic FGMs, and the impact
resistance of the resulting materials has not been evaluated.

Table 1
MAT_PLASTIC_KINEMATIC model parameters [27].

Materials ρ0 (g/cm3) E (Mbar) ν σy Et

(Mbar) (Mbar)

Steel 7.89 2.02 0.3 0.01 0.02

Epoxy resin 1.15 0.0114 0.49 0.00043 0.00036
Fig. 4. Schematic diagram of ballistic test.
C.-Y. Huang, Y.-L. Chen / Materials and Design 91 (2016) 294–305 297

Table 2


ρ0 (g/cm3) G (Mbar) σy (Mbar) Eh C0 (cm/μs) S1 γ0

Copper 8.93 0.477 0.0012 0.00288 0.394 1.489 2.02

6061-T6 2.7 0.26 0.0027 0.00235 0.524 1.4 1.97

Therefore, in the current study, Al2O3 and ZrO2 were used to produce
an FGM and its impact resistance properties were analyzed.

2. Experimental setup and procedure

2.1. Materials selection and processing

In selecting FGMs, using similar thermal expansion coefficients and Fig. 6. Initial velocity against residual velocity of different specimens.
sintering conditions is crucial because doing so can prevent stress
concentration and spallation. Therefore, the ceramics Al2O3 and
3 mol% Y2O3-stabilized zirconia (3YSZ) were used in the present calculated using the law of energy conservation, as following
study; the thermal expansion coefficient of Al2O3 is 8.6 × 10−6/°C and
1 1
that of 3YSZ is 10.1 × 10−6/°C. The powder was spray granulated, and ΔE ¼ mV 2a − mV 2b ð1Þ
the Al2O3 particle size d50 was 44 μm and the 3YSZ particle size was 2 2
17.9 μm. Powder metallurgy sintering was used to create a four-layer
(100, 90, 80, 70 vol.%) Al2O3/(0, 10, 20, 30 vol.%) ZrO2 FGM at a sintering
3. Stress wave theory
temperature of 1550 °C, and the relative density of each ceramic layer
was 99.4%–99.8%.
The transmission of stress waves is similar to that of light waves.
When a stress wave is transmitted from one medium into another neigh-
2.2. Ballistic test boring medium, the dissimilar wave impedance of the two media creates
a reflection and transmission phenomenon of the stress wave on the in-
The impact-resistant ceramic specimens were regular hexagons terface between the two media. If both media have the same cross section
with a side length of 60 mm. The dimension of 6061-T6 Al backing area, the following equations are obtained by stress wave theory [18].
plate was 200 mm (length) × 200 mm (width) × 1 mm (thickness).
Epoxy resin was used as the adhesive between the ceramics and the ðρ2 C 2 Þ−ðρ1 C 1 Þ
F¼ ð2Þ
rear plate. The rear plate dimensions remained unchanged and only ðρ2 C 2 Þ þ ðρ1 C 1 Þ
those of the ceramics were varied. The design factors include areal den-
sity of ceramics 4.64 g/cm2 and the thickness 11 mm (without the adhe- 2ðρ2 C 2 Þ
T¼ ð3Þ
sive layer and rear plate). Under these conditions, the impact-resistant ðρ2 C 2 Þ þ ðρ1 C 1 Þ
ceramic specimens designed for the experiment comprised FGM, multi-
layer ceramic composite materials (MCMs), and purity ceramic com- σ r ¼ Fσ i ð4Þ
posite materials (PCMs) (Fig. 2). The projectiles used for the ballistic
σ t ¼ Tσ i ð5Þ
experiment were 0.30″ armor-piercing bullet, the profile of 0.30″ AP
bullet and target is shown in Fig. 3. The test specifications were in accor-
dance with the NIJ IV standard and the initial velocity was 868 ± 15 m/s.
The initial and terminal velocities were obtained using optical shutter (1) when ρ2C2 N ρ1C1, F N 0, and T N 1. The reflection and transmis-
velocimetry (Fig. 4), and the energy absorbed by the target plates was sion waves have the same properties as the incident wave; thus,

Table 3
Johnsone–Holmquist model parameters [28].

Materials ρ0 (g/cm3) G (Mbar) A B C M N

100% Al2O3 3.93 1.56 0.88 0.28 0.007 0.6 0.64

90% Al2O3/ 4.14 1.53 1 0.6 0.006 0.6 0.62
10% ZrO2
80% Al2O3/ 4.35 1.38 1 0.92 0.005 0.6 0.6
20% ZrO2
70% Al2O3/ 4.56 1.34 1 0.92 0.004 0.6 0.58
30% ZrO2

Materials T (Mbar) Hel (Mbar) Phel (Mbar) D1 D2

100% Al2O3 0.00259 0.0591 0.0185 0.01 0.7 2.5

90% Al2O3/ 0.00397 0.0978 0.0305 0.02 0.83 2.6
10% ZrO2
80% Al2O3/ 0.00429 0.1057 0.0033 0.03 0.96 2.34
20% ZrO2
70% Al2O3/ 0.00671 0.1657 0.0517 0.04 1 2.28
30% ZrO2
298 C.-Y. Huang, Y.-L. Chen / Materials and Design 91 (2016) 294–305

Table 4 A 1/4 model was used for establishing the ceramic composite plate
Experimental results. model (Fig. 5). Solid 164, a 3D modeling program, was used to establish
Average a model, and the model's specifications were the same as those of the
Areal density Thickness Average residual
(g/cm2) (mm) velocity (m/s)
absorptive energy experimental specimens. The adhesive used in this study is epoxy,
(J) which is usually used by ceramic manufacturing factory and the thick-
FGM 4.64 11 259 3349 ness of the adhesive layer is 0.5 mm, a reasonable thickness compared
MCM 4.64 11 573 2047 with that of ceramic layer (3 mm). To simplify the complexity of the
PCM 4.64 11.8 288 3295
projectile simulation, the lead was replaced with copper. The mesh
3.93 11 359 3108
size was approximately set as 0.5 mm [22,23] and the minimum ele-
ment size of the projectile was approximately 0.4 × 0.6 × 0.6 mm3
the compression wave (tensile wave) is reflected as the compres- and that of the target was 0.5 × 0.8 × 0.8 mm3 in the present study. In
sion wave (tensile wave); the three models, the mesh number of the FGM was 61,710, that of
(2) when ρ2C2 b ρ1C1, F b 0, and 0 b T b 1. The reflection and inci- the MCM was 75,898 and the PCM was 74,510. The mesh number of
dent waves have opposite properties, the transmission and inci- the projectile was 15,498 and adhesive was 14,400. Two materials,
dent waves have the same properties. Thus, the compression namely steel and copper, were used for creating the 0.30″ armor-
wave (tensile wave) is reflected as the tensile wave (compression piercing bullets. In the LS-DYNA simulation, MAT_PLASTIC_KINEMATIC
wave); was used for the steel part and MAT_ELASTIC_PLASTIC_ HYDRO was
(3) when ρ2C2 = ρ1C1, the two media share the same wave imped- used for the copper part. These materials were modeled using the
ance. Therefore, F = 0 and T = 1. When the wave is transmitted Mie–Gruneisen equation of state (EOS_GRUNEISEN). Epoxy resin, ce-
to the interface between the two media, the transmission is com- ramic, and aluminum were used as the target plates. Furthermore,
plete and no reflection wave is generated. Although ρ and C differ MAT_PLASTIC_KINEMATIC was used for the epoxy resin and
for both media, no reflection is generated when the stress wave MAT_ELASTIC_PLASTIC_HYDRO was used for the 6061-T6 aluminum
reaches the interface as long as the wave impedance is the same. plate. The material parameters used in the model were shown in
Tables 1–2 [24]. MAT_JOHNSON_HOLMQUIST_CERAMICS was used
for the ceramic. The material parameters used were shown in
Table 3, in which the density (ρ 0), shear modulus (G), tensile
4. ANSYS/LS-DYNA numerical simulation strength (T) and bulk modulus (K1 ) were obtained by tensile test
and ultrasonic non-destructive test. Hel and Phel were obtained by
In this study, ANSYS/LS-DYNA was used for conducting the simula- Eq. (6) and Eq. (7) [24,25]. Other parameters (i.e. A, B, C, M, N, D1
tion. LS-DYNA is an explicit nonlinear dynamic finite element analysis and D 2 ) cited the data listed in [25] or by mixture law M =
software program that can solve the dynamic response of various 2D McXc + MdXd. These materials were modeled using EOS_GRUNEISEN.
and 3D non-elastic structures during large deformation induced by The contact mode between the projectiles and target plates was de-
high-velocity impact and explosions [19–21]. ANSYS/LS-DYNA is a com- fined as CONTACT_ERODING_SURFACE_TO_SURFACE [26].
bination of the explicit finite element analysis program LS-DYNA with
the powerful pre- and post-processing functions of the ANSYS program.
The explicit LS-DYNA algorithm can rapidly solve transient large defor- 1−υ
Hel ¼ 8 σ0 ð6Þ
mation dynamics, large deformations, and multiple nonlinear quasi- ð1−2υÞ2
static problems as well as complex contact-impact problems. When
this combined program is employed, ANSYS can be used to establish a
model and LS-DYNA can be used to derive explicit solutions. LS-DYNA 1 1þυ
Phel ¼ Hel ð7Þ
can then be used for post-processing to examine the results. This 3 ð1−υÞ
study investigated the fracturing and deformation phenomena of ce-
ramic composite plates subjected to a high-velocity impact. Further-
more, to reduce the number of meshes and the computation time, a
Lagrange algorithm was used.
5. Results and discussion

5.1. Results of experiment and simulation

As shown in Fig. 6 and Table 4, the experimental results showed that

the FGM possessed optimal ballistic resistance under the conditions in-
volving both equal areal density (4.64 g/cm2) and equal thickness
(11 mm). The results of simulation showed good agreement with
those of ballistic test. Fig. 7 indicates that the error between the simula-
tion results and the experimental results is within 5%, confirming the ac-
curacy of the finite element simulation.
The initial and terminal velocities obtained in the experiment and
simulation can be used to calculate the target's absorptive energy ac-
cording to the energy conservation equations (Eq. (1)). Table 4 showed
that the absorptive energy of the FGM increased only by 1.64% when
compared with PCM under equal areal density (4.64 g/cm2) condition,
but the thickness of FGM was less than that of PCM. Under equal thick-
ness (11 mm) condition, the FGM was heavier than PCM by 18.1%.
However, FGM could absorb more energy than that of PCM by 7.8%.
The MCM showed poor capability in energy absorption because of the
Fig. 7. Absorptive energy of experiment and simulation. adhesive effect.
C.-Y. Huang, Y.-L. Chen / Materials and Design 91 (2016) 294–305 299

Table 5
Parameters of materials.

Materials Density ρ (kg/m3) Young's modulus E (GPa) Wave velocity C (m/s) Impedance coefficient ρC (106)

100% Al2O3 3930 393 10881 42.76

90% Al2O3/10% ZrO2 4147 370 10397 43.11
80% Al2O3/20% ZrO2 4355 357 9962 43.38
70% Al2O3/30% ZrO2 4560 344 9557 43.58
Epoxy resin 1150 1.14 996 1.14
6061-T6 Al 2700 67.6 5003 13.51

5.2. Stress wave analysis substantially decreased. Similar results were found by Tasdemirci et al.
[26]. The results from their work showed that the interlayer could
When a projectile comes into contact with a target, an impact wave have a strong effect on the fragmentation behavior of the ceramic,
perpendicular to the target is generated. This impact wave is the com- caused mainly by the reflection of compressive waves at the interlayer
pression wave. Because material layers of the target have different due to the impedance mismatch. Besides, PCM was made of pure
wave impedance, the compression wave is reflected and transmitted Al2O3, therefore no layer effect on it. According to the experimental re-
at the interface between each layer. Dissimilar material properties also sults, the ballistic resistance of the FGM and PCM were superior to that
generate diverse reflection coefficients. When the reflection coefficient of the MCM plate.
is positive, the reflected wave is the compression wave; when the re-
flection coefficient is negative, the reflected wave is the tensile wave. 5.3. Microstructural analysis
After the parameters were calculated (Table 5), the results indicated
that the compression wave was reflected as a tensile wave when the re- SEM was used to observe the microstructure of the ceramic plates
flection coefficient between the ceramic and adhesive was F b 0 (the after they were subjected to the high-velocity impact. The structural de-
minus sign indicates the direction) and the transmission coefficient sign of the FGM does not include an adhesive layer structure. Therefore,
was 0 b T b 1. In addition, a higher absolute value of the reflection coef- the reflection tensile stress can be reduced and the bonding of the inter-
ficient indicates a stronger reflection wave. Therefore, increased reflec- layer interfaces preserved, substantially reducing delamination. Fig. 9
tion tensile stress causes increased tensile stress in ceramics. As shown indicates that each interface bond remains intact post-impact with no
in Table 6(a) (columns I, III and V) and (b) and in Eq. (4), the reflection severe cracking or delamination. Because of the adhesive layer effect,
wave produced by the MCM exceeded that exhibited by the FGM. Be- the compression wave produced after the high-velocity impact was
cause the tensile strength of the ceramic material is considerably reflected as a tensile wave, causing the ceramic to rapidly crack progres-
lower than its compressive strength, the ceramic material easily sively. As depicted in Fig. 10, obvious cracks were observed between the
fractures and cracks. The transmission intensity increased with the ceramic and adhesive layers. Delamination was also noticed between
transmission coefficient, and because the stress wave velocity was con- the ceramic and adhesive layers (Fig. 10(a)). In addition, Table 6(a)-IV
siderably greater than that of the projectile, the projectile did not reach and Fig. 10 shows that when the transmission coefficient and transmis-
the next ceramic layer. However, the transmission wave already frac- sion intensity increase, the transmission wave penetrates the adhesive
tured the next ceramic layer, thus reducing the ceramic's ability to resist layer and produces a relatively large range of cracking. Thus, the ballistic
the projectile's penetration [29]. Table 6(a) (columns II, IV and VI) and resistance of the FGM was superior to that of the MCM.
Eq. (5) show that the transmission intensity of the MCM is greater
than that of the FGM. The adhesive layer effect in the MCM also resulted 5.4. Numerical simulation analysis
in an increase in reflection intensity and a dramatic increase or decline
in the transmission intensity. By contrast, the reflection and transmis- In LS-DYNA software, the internal energy is the total strain energy
sion intensities of the FGM decreased. The experimental results and is calculated by adding the stress and strain produced in six direc-
(Fig. 8) showed that the level of cracking was less severe in the FGM tions of every element. The simulation results (Fig. 11) showed that
than it was in the MCM, and the ceramic cone area formed in the FGM the maximal internal energy in the MCM occurred at 16 μs, 4 μs earlier
by the impact was integral, whereas the cone area of MCM exhibited ir- than that in the FGM. This confirms that the adhesive layer caused the
regularities. The reason for this outcome is that the FGM exhibited no reflection coefficient to increase, which in turn increased the reflected
adhesive layer effect, therefore the reflection coefficient was extremely energy that induced the 100% Al2O3 ceramic layer of the MCM to
small, and then the tensile stress-induced fracturing in the ceramic reach maximal strain energy, causing the ceramic to fracture early. In

Table 6
The parameters of reflection coefficient and transmission coefficient.

(a) MCM

Epoxy resin—90% 90% Al2O3/ Epoxy resin—80% 80% Al2O3/ Epoxy resin—70%
100% Al2O3—epoxy resin
Interface Al2O3/10% ZrO2 10% ZrO2—epoxy resin Al2O3/20% ZrO2 20% ZrO2—epoxy resin Al2O3/ 30% ZrO2


Reflection coefficient (F) −0.94806 0.94848 −0.94848 0.94879 −0.94879 0.94902

Transmission coefficient (T) 0.05193 1.94848 0.05152 1.94879 0.05120 1.94902

(b) FGM

100% Al2O3–
90% Al2O3/10% ZrO2 80% Al2O3/20% ZrO2
Interface 90% Al2O3/
–80% Al2O3/20% ZrO2 –70% Al2O3/30% ZrO2
10% ZrO2

Reflection coefficient (F) 0.00409 0.00314 0.00225

Transmission coefficient (T) 1.00409 1.00314 1.00225
300 C.-Y. Huang, Y.-L. Chen / Materials and Design 91 (2016) 294–305

Fig. 8. The results of ballistic test of MCM and FGM.

Fig. 9. The microstructure of FGM after impact (dark: Al2O3, white: ZrO2).

addition, the compression wave was reflected as a tensile wave, expe- withstand fracturing induced by the transmission wave before the pro-
diting the ceramic cracking and lowering the ballistic resistance of the jectile reaches the second layer. The result is that the ballistic resistance
MCM plate. Fig. 12 indicates that the second layer plate (90% Al2O3/ in the MCM diminishes. Fig. 14 indicated that FGM reaches a maximal
10% ZrO2) of the MCM reaches a maximal internal energy at 30 μs, internal energy at 30 μs, 8 μs later than that of the FGM. The results im-
8 μs earlier than that of the FGM. Table 3(a)-II shows that the amount plied that the additives ZrO2 can increase the toughness of FGM, thus re-
of transmitted energy increases with the transmission coefficient, in- ducing crack propagation and increasing the capability to withstand
ducing the second ceramic layer plate of the MCM to undergo an early bullet impact. The maximum internal energy of FGM was lower than
fracture. When the projectile was still moving in the first ceramic that of PCM (Fig. 14), however a maximum internal energy of AP bullet
layer (100% Al2O3) at approximately 14 μs, the distribution of the stress was obtained when target was FGM (Fig. 15). The reason is that the
exerted on the second ceramic layer of the MCM was wider than that structure of FGM strengthens eroding effect on bullet, showing a supe-
exerted on the FGM (red area) (Fig. 13). The primary reason for this is rior ballistic resistance capability.
that the transmitted energy increases with the MCM transmission The penetration mechanism of the projectile can be described in
coefficient. This phenomenon requires the second ceramic plate to three phases [30]:

Fig. 10. The microstructure of MCM after impact.

C.-Y. Huang, Y.-L. Chen / Materials and Design 91 (2016) 294–305 301

Fig. 11. Internal energy of the 100% Al2O3 layer. Fig. 14. Total internal energy of the ceramic layers.

Fig. 15. Internal energy of AP bullet.

Fig. 12. Internal energy of the 90% Al2O3/10% ZrO2 layer. trajectory and the diameter and thickness of the bottom area of
the ceramic cone decrease progressively. The compression
simultaneously affects the rear plate, which begins to deform.
(1) The impact: After the projectile impacts the ceramic surface, the (3) The projectile and remainder of the ceramic cone jointly impact
effect of the incident and reflection waves causes interlacing ra- the rear plate: The projectile and remainder of the ceramic
dial and circumferential cracks in the ceramic front plate, cone travel at the same velocity, the ceramic cone ceases to
forming cone-shaped fracture areas on the front plate. The rear exert abrasion on the projectile, and the plastic deformation of
plate is not yet affected by the compression effect of the ceramic the rear plate increases. When the deformation reaches a certain
cone at this time. level, the rear plate fractures or is perforated.
(2) The projectile compresses the ceramic cone: As the projectile
compresses the ceramic cone, each fragment of the material is
crushed further at an accelerating speed. Part of the ceramic frag- The results obtained from the simulation of the FGM structure
ments starts to flow in an opposite direction to the ballistic (Fig. 16(a)) showed that because of the small difference in wave

Fig. 13. Distributive stress of a quarter model in LS-DYNA.

302 C.-Y. Huang, Y.-L. Chen / Materials and Design 91 (2016) 294–305

Fig. 16. Distributive stress of ceramic composite materials in LS-DYNA.

impedance, the reflection coefficient decreased considerably and exhib- strongly increasing or decreasing transmission coefficient. Each ceramic
ited a systematic descent with the transmission coefficient. This caused plate layer must endure a complex interaction effect of reflection and
the impact energy to be transferred in a graded state and the stress to transmission intensity. According to the simulation results depicted in
fracture the ceramic, forming the ceramic cone. Therefore, the FGM Fig. 16(b), the stress fracture distribution of the ceramic layers did not
structure can promote the formation of a complete ceramic cone after exhibit a graded transfer. The impact stress caused every layer to frac-
the ceramic is penetrated by the projectile, thus stopping the projectile ture, hindering the formation of a complete ceramic cone after the
from advancing and enhancing ballistic resistance. Regarding the MCM, MCM is penetrated by the projectile and consequently diminishing the
the adhesive layer induces an extremely high reflection coefficient and a MCM's ballistic resistance.

Fig. 17. The ceramic cone of PCM, FGM and MCM after impact.
C.-Y. Huang, Y.-L. Chen / Materials and Design 91 (2016) 294–305 303

Fig. 18. Fracture of PCM, FGM and MCM after impact.

Fig. 19. Comparison of ceramic cone semi-angle of PCM, FGM and MCM.

The stress distribution (Fig. 16(c) and (d)) showed that the ceramic PCM, resulting in a smaller ceramic cone angle. Because the graded
cone was formed earlier in the PCM than it was in the FGM. The stress structure of the FGM delayed the ceramic cone formation after impact,
fracture distribution in the PCM was also greater than that in the FGM, the development of the radial and circumferential cracks was also
indicating that the ceramic cracking was more severe in the PCM than reduced, which may increase the abrasion effect between the ce-
it was in the FGM. After 36 μs, the stress fracturing began to decrease ramic and the projectile and increase the ceramic cone's ability to
in the PCM, indicating that the ceramic cone had already entered the impede the projectile's advance. The deformation area of alumi-
second phase because the projectile was continuously moving and in- num backing plate is proportional to the ceramic cone angle. Refer-
ducing the ceramic cone to crack early. At 44 μs (Fig. 16), the abrasion ring to Fig. 21, the deformation area of backing plate is in the
exerted on the projectile by the PCM was considerably lower than that following sequence: PCM N FGM N MCM. The stress distribution of
exerted by the FGM. The semi-angle variation of ceramic cone is about aluminum backing plates is compatible with deformation areas
63°–68° [31–33]. Through measuring top and bottom diameters of ce- (Fig. 22). Overall, the ballistic resistance of the FGM was superior
ramic cone of the three specimens (Fig. 17 and Fig. 18), the semi- to that of PCM and MCM.
angle of those ceramic cones was obtained after calculation, i.e. 66° for
PCM, 61.5° for FGM and 40° for MCM (Fig. 19). The values of PCM and
FGM close to those shown in references. The stress distribution in sim- 6. Conclusion
ulation is matching with the ceramic cone angles (Fig. 20). Due to adhe-
sive effect of MCM, the forming ceramic cone angle is small. The In this paper, we used experiment and simulation to study the
experimental results (Fig. 18) signified that the ceramic cracking was impact resistance capability of Al 2 O 3 /ZrO 2 FGM and compared
more severe in the PCM (average sizes 16 mm) than it was in the with that of MCM and PCM. The results confirmed that the FGM
FGM (average sizes 23.7 mm). The small fragments of PCM reduce its re- demonstrated optimal energy absorption experiencing impact.
sistant capability under impact. Because the ZrO2 can increase tough- Through investigation and theoretical explanation, several com-
ness of FGM, the average fragments size of FGM is larger than that of ments can be given in following.

Fig. 20. Stress distribution of PCM, FGM and MCM.

304 C.-Y. Huang, Y.-L. Chen / Materials and Design 91 (2016) 294–305

Fig. 21. Deformation area of backing plates of PCM, FGM and MCM.

Fig. 22. Stress distribution of backing plates of PCM, FGM and MCM.

(1) The FGM structure can avoid delamination and increase the abra- [4] R. Zaera, S. Sánchez-Sáez, J.L. Pérez-Castellanos, C. Navarro, Modelling of the adhe-
sive layer in mixed ceramic/metal armours subjected to impact, Compos. A: Appl.
sive effect of each ceramic layer on the projectile, improving im- Sci. Manuf. 31 (2000) 823–833.
pact resistance. [5] A. Arias, R. Zaera, J. López-Puente, C. Navarro, Numerical modeling of the impact
(2) The FGM structure can delay ceramic cone formation and reduce behavior of new particulate-loaded composite materials, Comput. Struct. 61
(2003) 151–159.
circumferential and radial crack propagation, maintaining the in- [6] J. López-Puente, A. Arias, R. Zaera, C. Navarro, The effect of the thickness of the ad-
tegrity of the ceramic plate. hesive layer on the ballistic limit of ceramic/metal armours. An experimental and
(3) The FGM structure can enable all layers to exhibit similar wave numerical study, Int. J. Impact Eng. 32 (2005) 321–336.
[7] H.C. Chen, Y.L. Chen, B.C. Shen, Ballistic resistance analysis of double-layered com-
impedance, reducing the interaction effect of the reflection posite material structures, Theor. Appl. Fract. Mech. 62 (2012) 15–25.
wave and the transmission wave. [8] Y.L. Chen, H.C. Chen, Penetration depth of closed-cell aluminum foam sandwich
(4) Under same thickness or areal density, FGM showed optimal structures under low velocity impact, Trans. Jpn. Soc. Aeronaut. Space Sci. Space
Technol. Jpn. 10 (2012) 51–58.
ballistic resistance capability when compared with MCM and
[9] E.S.C. Chin, Army focused research team on functionally graded armor composites,
PCM. Mater. Sci. Eng. A 259 (1999) 155–161.
(5) Ceramic–metal composite is widely used to resist ballistic [10] H. Guo, K.A. Khor, Y.C. Boey, X. Miao, Laminated and functionally graded hydroxyap-
impact. Furthermore, present study confirmed that a ceram- atite/yttria stabilized tetragonal zirconia composites fabricated by spark plasma
sintering, Biomaterials 24 (2003) 667–675.
ic–ceramic FGM design can improve impact resistance. There- [11] A. Pettersson, P. Magnusson, P. Lundberg, M. Nygren, Titanium–titanium diboride
fore, mixture of high-hardness and high-toughness ceramics composites as part of a gradient armour material, Int. J. Impact Eng. 32 (2006)
can be used in future FGM design and manufacture. 387–399.
[12] L. Sun, A. Sneller, P. Kwon, Fabrication of alumina/zirconia functionally graded ma-
terial: from optimization of processing parameters to phenomenological constitu-
tive models, Mater. Sci. Eng. A 488 (2008) 31–38.
[13] X.F. Zhang, Y.C. Li, On the comparison of the ballistic performance of 10% zirconia
Acknowledgments toughened alumina and 95% alumina ceramic target, Mater. Des. 31 (2010)
[14] E.M.M. Ewais, D.H.A. Besisa, Z.I. Zaki, A.E.H.T. Kandil, Tailoring of functionally graded
We thank all the graduate students in the ballistic laboratory for zirconia–mullite/alumina ceramics, J. Eur. Ceram. Soc. 32 (2012) 1561–1573.
their wholehearted support in the experimental tests and material [15] M. Ubeyli, E. Balci, B. Sarikan, M.K. Oztas, N. Camuscu, R.O. Yildirim, et al., The ballis-
tic performance of SiC-AA7075 functionally graded composite produced by powder
manufacturing process.
metallurgy, Mater. Des. 56 (2014) 31–36.
[16] Y.C. Lin, Y.L. Chen, H.W. Chen, Low velocity impact performance analysis of fiber
References composite structure embedded with shape memory alloy, Trans. Jpn. Soc. Aeronaut.
Space Sci. Space Technol. Jpn. 12 (2014) 35–41.
[1] J.W. McCauley, G. D'Andrea, K. Cho, M.S. Burkins, R.J. Dowding, W.A. Gooch Jr., Status [17] L. Xue, G. Dui, B. Liu, L. Xin, A phenomenological constitutive model for functionally
report on SPS TiB2–TiB–Ti functionally graded materials (FGMs) for armor, Report graded porous shape memory alloy, Int. J. Eng. Sci. 78 (2014) 103–113.
Documentation of US Army Research Laboratory 2006, pp. 1–26. [18] W. Johnson, Impact Strength of Materials, Edward Arnold, London, 1972.
[2] N. Gupta, V.V. Bhanu Prasad, V. Madhu, B. Basu, Ballistic studies on TiB2–Ti function- [19] H. Kurtaran, M. Buyuk, A. Eskandarian, Ballistic impact simulation of GT model vehi-
ally graded armor ceramics, Def. Sci. J. 62 (2012) 382–389. cle door using finite element method, Theor. Appl. Fract. Mech. 40 (2003) 113–121.
[3] I.G. Grouch, L.J. Greaves, C. Ruiz, J. Harding, Dynamic compression of toughened [20] T. Børvik, O.S. Hopperstad, T. Berstad, M. Langseth, A computational model of
epoxy interlayers in adhesively bonded aluminium laminates, J. Phys. IV: JP 4 viscoplasticity and ductile damage for impact and penetration, Eur. J. Mech. A. Solids
(1994) C8-201-C8-6. 20 (2001) 685–712.
C.-Y. Huang, Y.-L. Chen / Materials and Design 91 (2016) 294–305 305

[21] T. Børvik, O.S. Hopperstad, T. Berstad, M. Langseth, Perforation of 12 mm thick steel [27] D.J. Steinberg, Equation of state and strength properties of selected materials, UCRL-
plates by 20 mm diameter projectiles with flat, hemispherical and conical noses: MA-106439, Lawrence Livermore Natioanal Laboratory, USA, 1996.
part II: numerical simulations, Int. J. Impact Eng. 27 (2002) 37–64. [28] H.L. Ren, W. Chen, T.T. Guo, Numerical simulation on the anti-penetration properties
[22] R. Bobbili, A. Paman, V. Madhu, A.K. Gogia, The effect of impact velocity and target of ceramic target, Trans. Beijing Inst. Technol. 33 (2013) 111–115.
thickness on ballistic performance of layered plates using Taguchi method, Mater. [29] H. He, M. Zhang, S.Y. Zeng, Effect of target inosculation condition on the anti-penetra-
Des. 53 (2014) 719–726. tion property of ceramic composite armour, Adv. Manuf. Manag. 25 (2006) 35–37.
[23] M. Grujicic, B. Pandurangan, B. d'Entremont, The role of adhesive in the ballistic/ [30] D.P. Gonçalves, F.C.L. de Melo, A.N. Klein, H.A. Al-Qureshi, Analysis and investigation
structural performance of ceramic/polymer-matrix composite hybrid armor, of ballistic impact on ceramic/metal composite armour, Int. J. Mach. Tools Manuf. 44
Mater. Des. 41 (2012) 380–393. (2004) 307–316.
[24] Z. Rosenberg, On the relation between the Hugoniot elastic limit and the yield [31] G.H. Liaghat, H. Shanazari, M. Tahmasebi, A. Aboutorabi, H. Hadavinia, A modified
strength of brittle materials, J. Appl. Phys. 74 (1993) 752–753. analytical model for analysis of perforation of projectile into ceramic composite tar-
[25] N.K. Bourne, J.C.F. Millett, M. Chen, J.W. McCauley, D.P. Dandekar, On the Hugoniot gets, Int. J. Compos. Mater. 3 (2013) 17–22.
elastic limit in polycrystalline alumina, J. Appl. Phys. 102:073514 (2007) 1–9. [32] R.L. Woodward, A simple one-dimensional approach to modelling ceramic compos-
[26] A. Tasdemirci, G. Tunusoglu, M. Güden, The effect of the interlayer on the ballistic ite armour defeat, Int. J. Impact Eng. 9 (1990) 455–474.
performance of ceramic/composite armors: experimental and numerical study, [33] A.L. Florence, T.J. Ahrens, Interaction of projectiles and composite armour, Final
Int. J. Impact Eng. 44 (2012) 1–9. Report, US Army 1967, p. 1.