high-temperature deformation of al2o3/y-tzp particulate laminates
TRANSCRIPT
Acta Materialia 52 (2004) 4685–4693
www.actamat-journals.com
High-temperature deformation of Al2O3/Y-TZP particulate laminates
Jue Wang a, Eric M. Taleff a,b, Desiderio Kovar a,b,*
a Materials Science and Engineering Program, The University of Texas at Austin, Austin, TX 78712, USAb Department of Mechanical Engineering, The University of Texas at Austin, 1 University Station C2200, Austin, TX 78712, USA
Received 23 April 2004; received in revised form 18 June 2004; accepted 21 June 2004
Available online 22 July 2004
Abstract
Al2O3/Y-TZP particulate laminates with varying compositions and ratios of layer thickness were fabricated by tapecasting, lam-
ination, and sintering. The resulting particulate laminates were tested in compression at a temperature of 1350 �C over strain rates
from 1.00·10�5 to 3.16·10�4 s�1. Microstructural changes during testing were observed to be minor. Stress exponents were meas-
ured to be approximately two and are consistent with previous data for particulate composites. Using parameters determined from
particulate composites, the behaviors of the particulate laminate composites are accurately predicted using a constrained isostrain
model without additional fitting parameters.
� 2004 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.
Keywords: High-temperature deformation; Compression test; Laminates
1. Introduction
In the past decade, laminated ceramic composites
have emerged as promising candidates for use in struc-
tural applications [1–3]. By controlling microstructure,
many properties of laminated ceramic composites can
be tailored to be superior to their monolithic counter-
parts [4,5]. For example, greatly enhanced fracture
toughness has been observed in Al2O3/Ce-TZP lami-
nates compared to either monolithic Al2O3 or Ce-TZP [6]. Although, the deformation behaviors of some
laminated metal composites have been explored [7–9],
there have been only a few studies of the creep of
laminated ceramic composites [10–12] and, as yet, de-
formation behavior of laminates is not well under-
stood. In contrast, the high-temperature behavior of
Al2O3/Y-TZP particulate composites has been exten-
1359-6454/$30.00 � 2004 Acta Materialia Inc. Published by Elsevier Ltd. A
doi:10.1016/j.actamat.2004.06.034
* Corresponding author. Tel.: +1-512-471-6271; fax: +1-512-471-
7681.
E-mail address: [email protected] (D. Kovar).
sively studied [13–17], because they exhibit remarkably
high tensile elongations of up to 625% [17].In the present work, Al2O3/Y-TZP particulate com-
posites with different compositions were used as starting
materials to build up a novel layered architecture, i.e.
particulate laminates. The objective of this work is to
study the high-temperature behavior of Al2O3/Y-TZP
particulate laminates in compression as a function of
composition, specimen orientation, and layer thickness
ratios. The data obtained for particulate laminates arecompared to existing theoretical models and to data
for particulate composites.
2. Experimental procedure
2.1. Processing
Fine-grained yttria-stabilized tetragonal zirconia pow-
der (3Y-TZP, Tosoh, Tokyo, Japan) and high-purity
ll rights reserved.
4686 J. Wang et al. / Acta Materialia 52 (2004) 4685–4693
alumina powder (AKP-50, 99.99% purity, Sumitomo
Chemical Co. Ltd., Tokyo, Japan) were used as raw
materials. Ceramic powders were ball-milled in solvents
with a dispersant, a plasticizer, and a polymer binder.
Tapecasting was performed using a doctor blade on a
glass substrate. After drying, tapes with a length of 47mm, a width of 23 mm, and a thickness of 90 lm were
stacked and then laminated at 120 �C within a metallic
die to form a billet. The polymeric binder within the
billet was subsequently pyrolyzed by heating slowly in
flowing air, and the billets were then pressureless
sintered at 1450 �C for 1 h. Details of the tapecasting,
laminating and sintering processes can be found else-
where [15].Tapecasting slurries were prepared with solids con-
tents of 20, 40, 50, 60, and 80 vol% Al2O3, with the bal-
ance Y-TZP. Tapes of two different compositions were
stacked in alternating layers to form symmetric particu-
late laminates. Compositions in this study are desig-
nated by the volume fraction of Al2O3; for example,
20A contained 20 vol% Al2O3 and 80 vol% Y-TZP.
For laminates, the layer containing more Y-TZP is des-ignated as the soft layer; the layer containing more
Al2O3 is designated as the hard layer. The ratios of soft
layer to hard layer thicknesses varied from 1:4 to 4:1.
For 1:1, both layers were 450 lm thick prior to sintering;
for 1:2 (or 2:1) and 1:4 (or 4:1), the thinner layer was 450
lm and the thicker layer was twice and four times as
thick, respectively, prior to sintering. The total thickness
of each green billet was about 8 mm. After sintering, theshrinkage of the layer thickness was approximately 22%
for all the compositions. The particulate laminates are
designated by their constituent layers and ratios of layer
thicknesses, e.g. 20A/40A (1:1) consisted of 20A and
40A layers with a layer-thicknesses ratio of 1:1. A sum-
mary of all the particulate laminates fabricated for this
study is given in Table 1. Some compositions, indicated
with an asterisk, were fabricated but cracked duringprocessing due to differential sintering rates and thermal
expansion mismatch between layers [18] and, therefore,
were not tested.
Table 1
Compositions of the particulate laminates used in this study are shown
1:1 1:2 2:1 1:4 4:1
Particulate laminates
20A/40A 20A/40A 20A/40A
20A/50A 20A/50A 20A/50A
20A/60A 20A/60A 20A/60A 20A/60A 20A/60A
20A/80A 20A/80A 20A/80A* 20A/80A* 20A/80A*
40A/80A 40A/80A 40A/80A
50A/80A 50A/80A 50A/80A
60A/80A 60A/80A 60A/80A
40A/60A 40A/60A
Asterisks indicate compositions that fractured during processing.
2.2. Microstructure analysis and mechanical testing
The density of each billet was measured using the Ar-
chimedes method, according to ASTM C373-88 [19],
with water as the immersion medium. Microstructures
of the particulate laminates were characterized using ascanning electron microscope (SEM). Specimens for
SEM observation were polished to a final grit size of 1
lm using diamond abrasives and were then thermally
etched at 1370 �C for 20 min to reveal grain structure
[15]. Mean grain sizes (�d) of both Al2O3 and Y-TZP
were calculated from measurements of equivalent circu-
lar diameter (D), using the following expression [20]:
�d ¼ 1:27�PN
i¼1DN
; ð1Þ
where N is the number of measured grains. The mor-
phology of particulate laminates before and after testing
was observed using an optical microscope.
The sintered laminates were cut and ground into rec-tangular bars 6 mm·4 mm·4 mm with the largest di-
mension either perpendicular or parallel to the layer
interfaces. Compression strain-rate-change (SRC) tests
[15] were performed at 1350 �C in air using a split-tube
furnace with MoSi2 heating elements. The applied uni-
axial, compressive stress was either perpendicular or
parallel to the layer interfaces. Fig. 1 illustrates these
testing configurations, which are referred to as the iso-stress orientation, Fig. 1(b), and the isostrain orienta-
tion, Fig. 1(c) [21].
The compression SRC tests consisted of nine steps in
engineering strain rate, defined as the ratio of displace-
ment rate to the sample length at the beginning of each
step. A prestrain of approximately 2% was initially ap-
plied at a strain rate of 1.00·10�4 s�1 in order to ensure
mating between the sample and compression platens andto stabilize the sample microstructures. Following the
prestrain step, seven steps with strain rates from
1.00·10�5 to 3.16·10�4 s�1 were applied. The strain
Fig. 1. Illustrations of testing orientation for: (a) particulate compos-
ites, (b) particulate laminates, isostress orientation and (c) particulate
laminates, isostrain orientation are shown.
J. Wang et al. / Acta Materialia 52 (2004) 4685–4693 4687
rate of 1.00·10�4 s�1 was repeated in the final step of
the series to check for repeatability of the measurements.
Engineering strain and stress were obtained from load–
displacement curves by assuming that the change in
displacement of the crosshead corresponded to the
reduction in the height of the specimen, after compensa-tion for elastic deflection. True stress, true strain, and
true strain-rate were derived from engineering stress,
strain, and strain rate under the assumptions of uniform
deformation and volume conservation. Due to the small
strain within each step, the true strain-rate was within
2% of the engineering strain-rate for each step.
Fig. 2. Scanning electron micrographs of particulate laminates at
different magnifications are shown for: (a) 60A before testing; (b) 20A/
60A (1:1) after testing at 1350 �C. An optical micrograph is shown for
(c) 20A/60A (1:1) before testing. The arrows indicate the orientation of
applied stress in (b).
3. Results
3.1. Microstructure
An example of the microstructure of a typical Al2O3/
Y-TZP particulate composite, prior to testing, is shown
in Fig. 2(a). The grains with the lighter shading are
Y-TZP and the grains with the darker shading areAl2O3. Both Al2O3 and Y-TZP phases are generally
well-dispersed and have equiaxed grain shapes. In a pre-
vious study of particulate composites [15], it was found
that the mean grain size of Al2O3 ranged from 0.35 to
0.46 lm, and the mean grain size of Y-TZP varied from
0.32 to 0.22 lm, with the Al2O3 grain size increasing as
the volume fraction of Al2O3 in the Al2O3/Y-TZP par-
ticulate composites increased [15].A SEM micrograph of a representative 20A/60A (1:1)
particulate laminate after compression SRC testing is
shown in Fig. 2(b), where the darker layers are 60A.
An optical micrograph of the 20A/60A (1:1) laminate
before testing is shown in Fig. 2(c), where the layers with
the lighter shading are 60A. From these micrographs, it
can be seen that the interfaces between the layers of par-
ticulate composites with different compositions arestraight and remain bonded during testing. Previously,
we have shown that, within the test temperature and
range of strain rates used in the current study, no statis-
tically significant changes in grain size or grain shape, as
measured by SEM microscopy, occurred during com-
pression SRC testing of particulate composites contain-
ing from 20 to 80 vol% Al2O3 [15]. Observations of the
microstructures in the particulate laminates tested underthe same conditions indicate that they also have no
significant change in grain size or shape during SRC
testing.
The relative densities of selected laminates are shown
in Table 2 before and after testing. The densities prior to
testing range from 95% to 99%. The highest densities are
obtained in materials for which the composition differ-
ence between layers is small (e.g. 20A/40A and 40A/60A). These results are consistent with previous studies
on similar Al2O3/Y-TZP laminates, which found that
large composition differences result in differential sinter-
ing rates between the layers and low density [18]. Com-
paring the densities before and after testing (Table 2), it
is observed that only a slight (1%) increase in density oc-
curred during compression testing.
3.2. Deformation behavior
The dimensions of all the (1:1), (1:2), and (2:1) speci-
mens were measured after SRC testing. The macroscopic
Table 2
Relative densities of Al2O3/Y-TZP particulate laminates before and after testing (in the isostrain orientation) are shown
Laminates Relative densities (%)
2:1 1:1 1:2 1:4
Before After Before After Before After Before After
20A/40A 98.9 99.1 99.1 99.1 98.4 99.2
20A/50A 97.6 98.5 97.9 98.9 99.1 99.4
20A/60A 97.3 98.0 97.3 98.3 98.6 99.0 97.3 97.8
20A/80A 95.8 96.4 95.9 96.6
40A/60A 98.4 99.2 97.1 98.1
4688 J. Wang et al. / Acta Materialia 52 (2004) 4685–4693
strains calculated from the final specimen dimensions
in the directions perpendicular to the loading axis were
equal, indicating that the macroscopic, in-plain plastic
strains were isotropic. Measurements of strains after
SRC testing revealed no significant differences between
the strains in the two directions perpendicular to the ap-
plied stress for specimens tested in either the isostress or
isostrain orientations, indicating that the in-plane defor-mations are approximately isotropic. Fig. 3 shows data
from a representative SRC test for the 20A/40A (1:2)
material tested in the isostress orientation. The total true
strain for this test is e=0.115. After a brief transient at
every rate change, a reasonably steady-state stress is
achieved. It is also apparent that for the repeated steps
at a strain rate of 1.00·10�4 s�1, the steady-state flow
stress at a given rate is nearly constant over the rangeof strains imposed in any single SRC test. Behaviors
similar to those shown in Fig. 3 were observed for all
the particulate laminates tested.
Fig. 4 shows data accumulated from SRC tests at
1350 �C for all of the (1:1) laminates as plots of true
strain-rate against true stress on log–log scales. Figs.
4(a) and (c) contain data for specimens tested in the iso-
stress orientation and Figs. 4(b) and (d) contain data forspecimens tested in the isostrain orientation. From these
plots, it is apparent that the flow stresses for a given ma-
terial are very similar in both testing orientations. The
Fig. 3. A plot of true stress versus true strain is shown with SRC test
data from a 20A/40A (1:2) specimen in the isostress orientation. The
dashed line indicates the average flow stress at a strain rate of
1.00·10�4 s�1.
slope in Fig. 4 is equal to the stress exponent, n, from
the phenomenological equation for creep, which can
be written at a constant temperature as [22]
r ¼ K _e1=n; ð2Þwhere K is a constant for a given material. For all these
materials, the stress exponents are approximately equal
to 2 over the range of measured strain rates. A slight
negative curvature is apparent in all of the data, suggest-
ing that a transition occurs to lower values of stress ex-
ponent as strain rate increases which is consistent with
previous studies [23,24]. Similar behavior is observed
for all the laminates tested and has been observed previ-ously in Al2O3/Y-TZP particulate composites [15].
The influence of layer composition on the high-tem-
perature deformation of particulate laminates can be as-
sessed from Fig. 4. In Figs. 4(a) and (b), the composition
of the hard layer is fixed at 80A and the composition
of the soft layer is varied. As the fraction of Al2O3 in
the soft layer is increased, the resistance to deforma-
tion in the particulate laminates increases slightly. InFigs. 4(c) and (d), the composition of the soft layer is
fixed at 20A and the composition of the hard layer is
varied. In this case, the resistance to deformation in
the particulate laminates increases significantly as the
volume fraction of Al2O3 in the hard layer is increased.
Values of strain rate at a stress of 50MPa, interpolated
from the data in Figs. 4(a)–(d), are shown in Fig. 5, as
a function of total Al2O3 volume fraction for all of the(1:1) laminates in both the isostress and isostrain orien-
tations. Also shown in Fig. 5 are values of strain rate for
Al2O3/Y-TZP particulate composites with similar grain
sizes at a stress of 50 MPa, interpolated from previous
work (see Fig. 4 in [15]). The solid line and dashed line
are trend lines for the particulate composites and partic-
ulate laminates, respectively. The values of strain rate in
Fig. 5 indicate that the creep rates in the Al2O3/Y-TZPparticulate laminates decrease with increasing Al2O3
content, consistent with observations of the Al2O3/Y-
TZP particulate composites [15]. Moreover, comparing
particulate laminates and particulate composites with
the same volume fraction of Al2O3 indicates that the
particulate laminates creep more slowly than do the par-
ticulate composites for Al2O3 volume fractions up to
Fig. 4. The influence of layer composition on deformation response is shown for: (a) isostress and (b) isostrain orientations for which the
composition of the hard layer is held constant and the composition of the soft layer is varied; and (c) isostress and (d) isostrain orientations for which
the composition of the soft layer is held constant and the composition of the hard layer is varied.
Fig. 5. The dependence of strain rate on volume fraction of Al2O3
is shown. The solid line indicates the trend for particulate compos-
ites [15] and the dashed line indicates the trend for particulate
laminates.
J. Wang et al. / Acta Materialia 52 (2004) 4685–4693 4689
0.6, i.e. the particulate laminates appear to be stronger
than the particulate composites.
Fig. 6 shows SRC data for 20A/60A particulate lam-
inates with different layer thickness ratios as plots of thetrue strain-rate against true stress on log–log scales.
Figs. 6(a) and (b) are for specimens tested in the isostress
and isostrain orientations, respectively. Data for 20A
and 60A particulate composites are shown in Fig. 6
for comparison [15]. As expected, the flow stresses for
the 20A/60A laminates lie between the flow stress of
20A and 60A particulate composites at any given strain
rate. There is little difference between the data for spec-imens tested in the isostress and isostrain orientations.
4. Discussion
4.1. Microstructure
Prior to testing, the particulate laminates used in thecurrent study exhibited lower densities than the particu-
late composites tested previously [15]. As a result, many
of the specimens exhibited a slight increase (61%) in
density during compression testing. To evaluate the in-
fluence of changes in density that occurred during the
deformation, the strain attributed to densification was
calculated and found to be less than 1%. Compared to
the total engineering strain (�11%) during testing, thestrain from densification is relatively small and was,
therefore, neglected. The flow stress at a given strain rate
is nearly constant over the strains used in testing (see
Fig. 3), further confirming that densification and other
Fig. 6. The influence of thickness ratios on deformation response of 20A/60A particulate laminates is shown for: (a) isostress orientation, (b)
isostrain orientation. Data from particulate composites are shown for comparison [15].
4690 J. Wang et al. / Acta Materialia 52 (2004) 4685–4693
changes in microstructure do not strongly influence the
deformation behavior of these particulate laminates.
4.2. Stress exponents
Previous studies of Al2O3/Y-TZP particulate compos-
ites suggest that stress exponents are approximately 2
for composites with high Al2O3 contents and high strain
rates and approximately 3 at low Al2O3 contents and
low strain rates [15]. In addition, Jimenez-Melendo et
al. [11] have reported stress exponents of approximately
2 over similar stress and temperature ranges for Al2O3/
Y-TZP hybrid laminates consisting of alternating layersof Y-TZP and Al2O3/Y-TZP. The present results for
particulate laminates are similar, with a stress exponent
of approximately 2 at high strain rates and a transition
to higher values as strain rate decreases. This suggests
that there is likely a common deformation mechanism
between Al2O3/Y-TZP particulate composites and par-
ticulate laminates.
4.3. Models for deformation behavior
French et al. [21] proposed that the creep of particu-
late composites could be analyzed using either an iso-
stress or an isostrain model. The isostress model
assumes that the average stress in the composite, rc, isequal to that in each component of the composite, r1and r2 for a two-component composite, i.e. rc=r1=r2.The average strain rate in the composite, _ec, is then given
by
_ec ¼ V 1 _e1 þ V 2 _e2; ð3Þwhere V1 and _e1 are the volume fraction and strain rate
of component one and V2 and _e2 are the volume fraction
and strain rate of component two. The isostrain model,on the other hand, requires that the strain in each com-
ponent of the composite equal the average composite
strain, i.e. ec= e1= e2. Thus, the strain rates must also
be equal, i.e. _ec ¼ _e1 ¼ _e2. The average stress in the com-
posite can be represented for the isostrain case as
rc ¼ V 1r1 þ V 2r2: ð4ÞAssuming that each component obeys Eq. (2), the iso-
strain model predicts the composite flow stress to be
rc ¼ V 1K1 _e1=n1c þ V 2K2 _e
1=n2c ; ð5Þ
where Ki ¼ EiðAiÞ1=niðdi=biÞpi=ni expðQci=niRT Þ for eachcomponent i, where E is the dynamic, unrelaxed
Young�s modulus, A is a material constant, d is the grain
size, b is the magnitude of the Burgers vector, p is the
grain-size exponent, n is the stress exponent, Qc is the ac-
tivation energy for creep, R is the universal gas constant,
and T is the absolute temperature [22].
To determine whether existing models could be used
to describe the behavior of particulate laminates, the be-havior of particulate composites with compositions
from 20A to 80A were first examined. Previously, it
was shown that the isostress model did not fit the exper-
imental data well for Al2O3/Y-TZP particulate compos-
ites. Although the isostrain model yielded better fits to
the data, physically unrealistic values of the fitting pa-
rameters were observed [15]. The deformation behaviors
of these particulate composites were, however, accurate-ly modeled using a constrained isostrain model for
which the stress exponent was constrained to a single
value, i.e. n1=n2, and differences in grain size were ac-
counted for. This model predicts
rc ¼ ðV 1K1 þ V 2K2Þ_e1=2c ; ð6Þwhere the stress exponent for Al2O3 and Y-TZP were
both taken to be 2, i.e. n1=n2=2, [15].
To model the behavior of Al2O3/Y-TZP particulatelaminates, the laminates are considered to consist of lay-
ers with two compositions, each of which are particulate
composites. The particulate composites, which all exhibit
n�2 within compositions ranging from 20 to 80 vol%
Al2O3, have been shown to closely obey Eq. (6) [15]. Be-
Table 3
The properties of Al2O3/Y-TZP particulate composites are shown
(T=1350 �C, r=50 MPa, n=2) from [15]
Particulate composites _e ðs�1Þ K (s�2 MPa)
20A 2.73·10�4 3026.1
40A 1.66·10�4 3880.8
50A 1.00·10�4 5000.0
60A 5.51·10�5 6735.9
80A 2.77·10�5 9500.1
J. Wang et al. / Acta Materialia 52 (2004) 4685–4693 4691
cause the particulate laminate composites also exhibit
n�2, it is hypothesized that their behaviors may also
obey Eq. (6), when each particulate composite is consid-
ered a component of the laminate composite, i.e. one
particulate composite layer provides V1 and K1 andthe other provides V2 and K2. Because the values of
K1 and K2 were previously determined [15] for particu-
late composites within the range of compositions used
in the laminate composites, Eq. (6) can be used to directly
predict the behavior of each particulate laminate com-
posite with no additional fitting parameters required.
The isostress model, Eq. (3), can be evaluated in a sim-
Fig. 7. The dependence of strain rate on the volume fraction of the hard layer
laminates. The solid symbols represent the data for the 20A and 60A particu
particulate laminates.
ilar manner for comparisons with predictions from the
constrained isostrain model, Eq. (6). The data for partic-
ulate composites used in these models are summarized in
Table 3 [15].
Predictions of the isostress and the constrained iso-
strain models are shown in Figs. 7(a)–(d) as dashedand solid lines, respectively. It is apparent that when
the composition differences between the layers are small
(e.g. 20A/40A and 60A/80A), the differences between the
predictions of the isostress and the constrained isostrain
models are small, as shown in Figs. 7(a) and (d). When
the composition differences between the layers are larger
(e.g. 20A/60A and 40A/80A), the differences between the
predictions are large, as seen in Figs. 7(b) and (c). InFigs. 7(a) and (d) the differences between the two models
are within the range of experimental scatter. However,
in Figs. 7(b) and (c) it is clear that the constrained iso-
strain model fits the experimental data best both when
specimens are tested in the isostress orientation and
when tested in the isostrain orientation.
To further examine the applicability of the con-
strained isostrain model to Al2O3/Y-TZP particulate
is shown for: (a) 20A/40A; (b) 20A/60A; (c) 40A/80A; and (d) 60A/80A
late composites [15], and the open symbols correspond to the data for
Fig. 8. The flow stresses at three different strain rates are shown for
20A/60A particulate laminates. The solid lines are predictions from the
constrained isostrain model. The solid symbols are data for 20A and
60A particulate composites [15], and the open symbols are data for
particulate laminates.
4692 J. Wang et al. / Acta Materialia 52 (2004) 4685–4693
laminates, flow stresses are predicted at three different
strain rates using Eq. (6) and compared to experimental
data. The values of K1 and K2 for particulate composites
at a flow stress of 50 MPa, as shown in Table 3, were
used to predict flow stresses of the particulate laminates,
and these predictions are shown as solid lines in Fig. 8.
The experimentally determined values of flow stress arealso shown in Fig. 8. Predictions and data are for strain
rates of 1.00·10�5, 1.00·10�4, and 3.16·10�4 s�1. The
solid symbols correspond to the data for particulate
composites from previous work (see Fig. 4 in [15]). Note
that there are no fitting parameters used in the predictive
equations since all the parameters in Eq. (6) are deter-
mined independently of the data in Fig. 8. Excellent
agreement is observed between data in both the isostressand isostrain orientations and the constrained isostrain
model. As shown in Fig. 8, the strength of the particu-
late laminates increases with increasing volume fraction
of the hard layer (i.e. 60A), especially at high strain rate,
e.g. 3.16·10�4 s�1.
These results also explain why the particulate lami-
nates appear to be stronger than particulate composites
with the same overall composition, as shown in Fig. 5.For a given stress, the strain rate does not vary linearly
with composition for materials such as Al2O3/Y-TZP
particulate composites that obey the constrained iso-
strain model [15]. Since the particulate laminates con-
sisting of layers A and B were shown to follow a
constrained isostrain model, and since the constituent
layers A and B themselves obey the constrained isostrain
model, the deviation of strain rate from a rule-of-mixtures with composition is further increased for the
particulate laminates.
As discussed by Jimenez-Melendo et al. [11], the dis-
crepancy between the isostress model and experimental
data for samples tested in the isostress orientation re-
sults from inter-layer constraint imposed by the hard
layer on the soft layer, which is not accounted for in
the isostress model. Because volume is conserved during
plastic deformation and because the layers remain
bonded without slipping at the interface, the strains in
each layer must be equal, except near free surfaces, where
this constraint is relaxed away from the layer interfaces.
Note that since the volume fraction of material near sur-
faces is relatively small for the test specimens used in thisstudy and the strains during SRC testing were also
small, the loss of constraint at free surfaces is a minor
effect and the measured deformation behavior is domi-
nated by material that is constrained. As a result, the be-
havior of specimens tested in both orientations should
be similar, as was observed. The slightly higher strength
of specimens tested in the isostrain orientation results
from this orientation having fewer free surfaces whereinter-layer constraint can be relaxed (2 for the isostrain
versus 4 for the isostress orientation). Jimenez-Melendo
et al. [11] also found that the isostrain model could be
used to predict the behavior Y-TZP/Al2O3+Y-TZP
hybrid laminates tested in the isostrain orientation.
Although they did not attempt to predict the behavior
of specimens tested in the isostress orientation, it is ap-
parent from visual inspection that the isostrain modelcould also be used to predict the behavior of their spec-
imens tested in the isostress orientation.
5. Conclusions
The high-temperature deformation behaviors of
Al2O3/Y-TZP particulate laminates with varying com-positions and ratios of layer thickness, fabricated by
tapecasting, lamination, and sintering, have been investi-
gated in compression at 1350 �C with a stable, fine-
grained microstructure. Experimental results show that
stress exponents of all the Al2O3/Y-TZP particulate lam-
inates are approximately 2, which is similar to that meas-
ured on particulate composites within the same range of
composition. Experimental measurements indicate thatlaminates in the isostrain orientation are very slightly
stronger than laminates in the isostress orientation. This
is a result of good layer bonding, which requires the
strains in each layer to be identical, except near free sur-
faces where constraint is relaxed; the difference in free sur-
face area where inter-layer constraint is relaxed between
the isostress and isostrain orientations results in their
slight difference in strength. The constrained isostrainmodel, with n1=n2=2, provides a good prediction of
the flow behavior in particulate laminates tested in both
the isostrain and isostress orientations.
Acknowledgement
This work has been supported by the Texas Ad-vanced Research Program under project #003658-
0426-1999.
J. Wang et al. / Acta Materialia 52 (2004) 4685–4693 4693
References
[1] Marshall DB. Design of high-toughness laminar zirconia com-
posites. Am Ceram Soc Bull 1992;71:969–73.
[2] Padture NP, Pender DC, Wuttiphan S, Lawn BR. In situ
processing of silicon–carbide layer structures. J Am Ceram Soc
1995;78:3160–2.
[3] Kuo DH, Kriven WM. A strong and damage-tolerant oxide
laminate. J Am Ceram Soc 1997;80:2421–4.
[4] Moya JS. Layered ceramics. Adv Mater 1995;7:185–9.
[5] Chan HM. Layered ceramics: processing and mechanical behav-
ior. Annu Rev Mater Sci 1997;27:249–82.
[6] Marshall DB, Ratto JJ, Lange FF. Enhanced fracture-toughness
in layered microcomposites of Ce–ZrO2 and Al2O3. J Am Ceram
Soc 1991;74:2979–87.
[7] Wang JG, Hsiung LM, Nieh TG. Formation of deformation
twins in a crept lamellar TiAl alloy. Scr Mater 1998;39:
957–962.
[8] Snyder BC, Wadsworth J, Sherby OD. Superplastic behavior in
ferrous laminated composites. Acta Metall 1984;32:919–32.
[9] Lesuer DR, Syn CK, Sherby OD, Wadsworth J, Lewandowski JJ,
Hunt WHJ. Mechanical behavior of laminated metal composites.
Int Mater Rev 1996;41:169–97.
[10] Jimenez-Melendo M, Clauss C, Domı´ nguez Rodrıguez A, San-
chez Herencia AJ, Moya JS. Microstructure and high-tempera-
ture mechanical behavior of alumina/alumina–yttria-stabilized
tetragonal zirconia multilayer composites. J Am Ceram Soc
1997;80:2126–30.
[11] Jimenez-Melendo M, Clauss C, Domı´ nguez-Rodrıguez A, De
Portu G, Roncari E, Pinasco P. High temperature plastic
deformation of multilayered YTZP/ZTA composites obtained
by tape casting. Acta Mater 1998;46:3995–4004.
[12] Jimenez-Melendo M, Gutierrez-Mora F, Domı´ nguez-Rodrıguez
A. Effect of layer interfaces on the high-temperature mechanical
properties of alumina/zirconia laminate composites. Acta Mater
2000;48:4715–20.
[13] Wakai F, Kodama Y, Sakaguchi S, Murayama N, Kato H,
Nagano T. Superplastic deformation of ZrO2/Al2O3 duplex
composites. In: Superplasticity, MRS Conference Proceedings,
International Meeting on Advanced Materials, 1989;7:259–66.
[14] Wakai F, Kato H. Superplasticity of TZP/Al2O3 composite. Adv
Ceram Mater 1988;3:71–6.
[15] Wang J, Kovar D, Taleff EM. High-temperature deformation of
Al2O3/Y-TZPparticulate composites.ActaMater 2003;51:3571–83.
[16] Nieh TG, McNally CM, Wadsworth J. Superplastic behavior of a
20% Al2O3/YTZ ceramic composite. Scr Metall 1989;23:457–60.
[17] Nieh TG, Wadsworth J. Superplasticity in fine-grained 20%
Al2O3/YTZ composite. Acta Metall 1991;39:3037–45.
[18] Cai PZ, Green DJ, Messing GL. Constrained densification of
alumina/zirconia hybrid laminates, I: experimental observations
of processing defects. J Am Ceram Soc 1997;80:1929–39.
[19] ASTM C 373-88, American Society for Testing and Materials.
[20] Exner HE. Analysis of grain- and particle-size distributions in
metallic materials. Int Metall Rev 1972;17:25–42.
[21] French JD, Zhao JH, Harmer MP, Chan HM, Miller GA. Creep
of duplex microstructures. J Am Ceram Soc 1994;77:2857–65.
[22] Sherby OD, Burke PM. Mechanical behavior of crystalline solids
at elevated temperature. Prog Mater Sci 1968;13:325–90.
[23] Cannon RM, Rhodes WH, Heuer AH. Plastic deformation of
fine-grained alumina (Al2O3): I, interface-controlled diffusional
creep. J Am Ceram Soc 1980;63:46–53.
[24] Wakai F, Nagono T. The role of interface-controlled diffusion
creep on superplasticity of yttria-stabilized tetragonal ZrO2
polycrystals. J Mater Sci Lett 1988;7:607–9.