Full text of DOI:10.1111/j.1151-2916.2003.tb03291.x

J. Am. Ceram. Soc., 86 [1] 141–48 (2003)
journal
Scratch Damage in Ceramics: Role of Microstructure
Zong-Han Xie, Mark Hoffman,* Robert J. Moon, and Paul Munroe
School of Materials Science and Engineering, University of New South Wales, NSW 2052, Australia
Yi-Bing Cheng*
School of Physics and Materials Engineering, Monash University, VIC 3800, Australia
Scratch tests were conducted using a standard pyramid in-
denter against ␣-SiAlON ceramics with different microstruc-
tures: (i) fine equiaxed grains and (ii) large elongated grains.
The formation and propagation of cracks were investigated via
focused ion-beam milling, with an emphasis on the effect of
microstructure on material removal. The fine equiaxed micro-
structure exhibited high resistance to material removal at low
loads, because of its high hardness and homogeneous struc-
ture. As the load increased, radial and lateral cracks devel-
oped, resulting in large-scale chipping. In contrast, the large
elongated microstructure showed a propensity to form micro-
cracks and microabrasion, which is characteristic of partial
grain removal, at low loads. With increasing loads, however,
the large elongated grains suppressed the propagation of
radial and lateral cracks, and, consequently, no large-scale
chipping occurred. Implications for material design in
abrasive-wear conditions have been discussed.
Many studies have shown that microstructure can have signif-
icant influence on crack formation and subsequent material remov-
al;5–8,17–22,25,26 for a fine-grained microstructure, median/radial
and lateral cracks appeared when the material was indented or
scratched with a sharp indenter3,4,13,15,17–22,24,25 (lateral cracks
have been assumed to be a major cause for material removal in
abrasive processes13,27), whereas cone cracks formed when the
material was indented or scratched with a blunt (Hertzian) indent-
er.11,28 For a coarse-grained microstructure, no median/radial and
lateral cracks were observed beyond a critical grain size when the
material was scratched with a sharp indenter; instead, subsurface
damage took the form of intergranular microcracks and intragrain
twin/slip bands.19–21 Hertzian contact studies also showed that
there is a region of maximum shear stress located beneath the
contact surface and the propensity for the formation of micro-
cracks increased as the grain size increased.5–8,11 As a result,
coarse-grained microstructures were implied to be a poor design
choice for abrasive-wear applications in which microcracking was
the controlling process for material removal.5–7,17–22
I. Introduction
These previous studies have two shortcomings. First, the devel-
opment and interaction of different crack systems and their
influence on material removal in different microstructures have not
been explored. Second, although the coarse-grained microstructure
has been shown to have a greater potential for microcracking—
and, hence, low abrasion resistance—the abrasion behavior of a
nonequiaxed microstructure has not been demonstrated.
BRASIVE wear of ceramics is a complex process. To assist in
A
elucidating the mechanisms that control material removal,
both static indentation1–11 and sliding indentation or scratch
tests12–28 have been used to simulate crack formation and abrasion
processes.
Several important crack systems in brittle materials, such as
radial and lateral cracks, have been investigated systematically by
Cook and Pharr9 using Vickers indentation. “Radial cracks”
initiate and grow during the loading segment of an elastic–plastic
contact; these cracks emanate from the corners of the indentation.
The driving force for the formation of the radial cracks results
from the localized loading that is generated by plastic deformation
during the indenter contact. “Lateral cracks” form when a material
is both loaded and unloaded, depending on the material system.
They form beneath the plastic deformation zone at an elastic–
plastic contact, are oriented parallel to the surface, and are circular
in form. The driving force for the formation of the lateral cracks
has been thought to result from the residual stress that results from
plastic deformation underneath the indenter contact.10 “Shallow
lateral cracks” are an important variation of the lateral crack.
Unlike the classic lateral crack that forms beneath the plastic zone,
the shallow lateral crack initiates at the edge of the contact
impression and propagates into the material almost parallel to the
surface; this type of crack often is bounded by the radial crack.9
In this study, we use scratch tests to elucidate the effect of
equiaxed versus nonequiaxed microstructure on abrasion behavior.
Ca ␣-SiAlON is used as a model polycrystalline system to
investigate the formation of different crack systems during scratch-
ing and the material removal that results from their interactions.
Some implications, based on this investigation, for material design
in abrasion applications are also discussed.
II. Experimental Procedure
(1) Materials and Specimen Preparation
Fine equiaxed (noted as “EQ”) and large elongated (noted
as “EL”) ␣-SiAlON samples were fabricated with the same
nominal chemical composition, defined by the formula
CaxSi12Ϫ(mϩn)AlmϩnOnN16Ϫn, where x ϭ m/2, m ϭ 2.6, and
n ϭ 1.3. The processing has been explained in greater detail in
an earlier work.29 These two microstructures are shown in Fig.
1. Some microstructural parameters and mechanical properties
are given in Table I.
Disk specimens with a diameter of 25 mm were ground, lapped,
then polished consecutively with 15, 6, and 1 ␮m diamond pastes.
Then, the polished specimens were cleaned ultrasonically in
acetone for 5 min, rinsed with alcohol and water, and dried at a
temperature of 120°C for 10 h.
N. Padture—contributing editor
(2) Scratching and Damage Evaluation
Single-pass scratch tests were conducted using a pin-on-disk-
type tribometer (CSEM Instruments, Neuchaˆtel, Switzerland). The
Manuscript No. 187110. Received March 4, 2002; approved September 20, 2002.
This investigation was supported by an Australian Research Council Large Grant
entitled “High Temperature Wear in ␣-Sialon Ceramics.”
*Member, American Ceramic Society.
141

142
Journal of the American Ceramic Society—Xie et al.
Vol. 86, No. 1
located within 10 ␮m of the surface can be captured with minimal
damage.
III. Results
Table II provides an explanation of the nomenclature used to
describe various features that are visible in the micrographs shown
in Figs. 2–8.
(1) Scratching at the Low-Load Region (1 N and 2 N)
The surface and subsurface damage that was caused by scratch-
ing on the EQ microstructure at loads of 1 and 2 N is shown in
Figs. 2(a) and (b) and 3(a) and (b), respectively. Radial cracks are
observed to propagate forward at an angle of ϳ30° in the scratch
direction, relative to the edge of the scratch track. Both the width
of the scratch grooves and the length of the radial cracks that
intersect the sample surface increased as the scratch load in-
creased. A small amount of scratch debris emerged just to the
outside edge of the scratch track at a load of 1 N. More debris
could be seen when the load increased to 2 N. The size of the
debris is of the order of the grain size and seems to result from
fine-grained detachment from the scratch surface. The correspond-
ing cross-sectional views (Figs. 2(b) and 3(b)) show that micro-
cracks formed beneath the scratch groove and are more evident at
a load of 2 N, in which a median crack can also be seen.
In contrast, Figs. 2(c) and (d) and Figs. 3(c) and (d) show that
no radial cracks appeared in the EL microstructure at scratch loads
of 1 and 2 N. Instead, there is a noticeable increase in the amount
of scratch debris at the sides of the scratch track, compared with
that observed with the EQ microstructure under similar conditions.
Further examination of the scratched surface under high magnifi-
cation revealed more details of the scratched surface, as shown in
Fig. 4. The unscratched surface microstructure, which consisted of
elongated grains and a grain-boundary glass phase, could be seen
to the right of the scratch track (see Section III(4)). In contrast, the
surface microstructure of the scratched track contained parallel
surface cracks that were perpendicular to the scratch direction. A
large amount of debris can be seen to the left-hand side of the
scratch track; here, the aggregate size is much smaller than the
length of the elongated grains, which suggests that the debris was
the product of the partial removal of individual elongated grains. In
the cross-sectional view (Figs. 2(d) and 3(d)), microcracking can
be observed underneath the scratch track; however, no median,
lateral, or radial cracks could be seen.
Fig. 1. SEM micrographs of the Ca ␣-SiAlON microstructures used in
this study ((a) fine equiaxed (EQ) and (b) large elongated (EL)). Surfaces
etched by molten NaOH.29
“pin” was a Vickers pyramid indenter (tip radius of ϳ1.5 ␮m), and
the sliding direction was parallel to the pyramid diagonal. Testing
was performed using a constant velocity of 1 mm/s, in laboratory
air with a relative humidity of 50%–60%. Normal loads of 1, 2, 5,
and 10 N were used. After scratching, the sample surface was
directly gold-coated, without any surface cleaning, thus minimiz-
ing any possible disturbance of the surface debris that may have
been produced during the testing.
After the gold coating, a focused ion beam (FIB) milling system
(Model FEI xP200, FEI Co., Hillsboro, OR) was used to generate
cross sections and examine the influence of scratch damage in the
two microstructures. The FIB milling system uses a fine (ϳ10 nm)
energetic beam of Ga ions that scans over the surface of a
specimen. At high beam currents, the beam rapidly sputters
sections through the specimen surface and allows a cross section to
be prepared. If the beam current is reduced, the secondary
electrons emitted from the specimen can be detected and used to
generate images similar to conventional scanning electron micros-
copy (SEM) images. The advantage of this sectioning technique30
is that micrometer-sized features (microcracks, pores, etc.) that are
(2) Medium-Load Region (5 N)
The scratching damage in the EQ microstructure at a load of 5
N is shown in Figs. 5(a) and (b). Radial and lateral cracking are
observed, as well as chipping, which resulted from the propagation
of these two crack systems. Chipping of the material was observed
to extend outward from the scratch track with a width of 2cЈ, which
is approximately twice the width of the scratch track (2d). The
subsurface view in Fig. 5(b) shows the lateral crack that is present
at the edge of the scratch track and extends outward into the
material. This type of lateral crack can be considered to be a
shallow lateral crack, according to the definition of Cook and
Pharr.9
In contrast, no chipping is observed in the EL microstructure
when the material is scratched under a load of 5 N. Partial grain
Table I. Properties of ␣-SiAlON EQ and EL Microstructures†
Average grain size (␮m)
Indentation test data at a load of 98 N
Aspect
ratio
Density
(g/cm3)
Vickers hardness,
Fracture toughness,
Sample identification
Diameter
Length
HV (GPa)
KIC (MPa⅐m1/2
)
Fine equiaxed (EQ)
Large equiaxed (EL)
0.35
0.70
0.39
5.04
1.1
7.2
3.19
3.21
12.8 Ϯ 0.5
12.0 Ϯ 0.2
3.7 Ϯ 0.3
7.5 Ϯ 0.3
†From Xie et al.29

January 2003
Scratch Damage in Ceramics: Role of Microstructure
143
Table II. Nomenclature Used to Describe Features Visible
in Figs. 2–8
examination on the formation of both radial and lateral cracks was
conducted, as shown in Fig. 7. Radial cracks can be seen
propagating away from the scratch track and penetrating at angles,
which can be either forward or backward to the sliding direction
(Fig. 7(b)). The lateral cracks form below the scratch track at two
different depths: Ͻ1 ␮m and Ͼ2 ␮m (Fig. 7(b)). The lateral cracks
that form within 1 ␮m of the surface are connected to the parallel
surface cracks that are oriented perpendicular to the scratch track.
The lateral crack that formed at a location Ͼ2 ␮m below the
surface is analogous to those observed in Figs. 5(b) and 6(b) and
is believed to originate below the scratch track and extend both
along the track and outward from either side of the scratch track.
Chipping was not observed for the EL microstructure; however,
severe surface damage was observed along the scratch track, as shown
in Fig. 6(c). Broken grains and rod-shaped recesses can be seen from
the enlarged view of the scratched surface that is shown in Fig. 8. The
subsurface view in Fig. 6(d) shows that discrete microcracks beneath
the scratch track seem to coalesce to form lateral cracks. These lateral
cracks do not appear to fully develop (i.e., become large and cause
chipping). Their propagation may be restricted by crack-growth-
resistance behavior that results from the elongated grains.
Symbol
Feature
d
Half-width scratch
c
Length of radial crack from middle of scratch track
Half-width of chipping
Scratch debris
cЈ
D
m
R
M
T
Microcrack
Radial crack
Median crack
Surface crack
L
Lateral crack
G
s
␣-SiAlON grain
Grain-boundary glass phase
Sliding direction of the indenter
➪
removal is still the main form of material loss, as shown in Fig.
5(c). Radial cracks can be seen propagating forward, away from
the scratch edge. No obvious increase in the amount of scratch
debris is apparent, relative to scratching under lower loads.
Meanwhile, the level of subsurface microcracking at the grain
boundaries did not change significantly (Fig. 5(d)). Propagation of
the microcracks is apparently suppressed by the elongated grains.
(4) Etching by Ion Beam
In Fig. 4, the SiAlON grains and grain-boundary glassy phase
can be clearly distinguished in the unscratched region to the right
of the figure. Such an effect becomes evident after the polished
surface has been exposed to an unfocused ion beam at a current of
70 pA for 30–50 s, depending on magnification. This is a very
useful technique for microstructure analysis, because SiAlON
materials cannot be effectively plasma-etched (because of their
crystal structure). Unlike etching with NaOH, the ion-etched
surface is essentially flat and gives good contrast between the
crystalline and glassy phases.
(3) High-Load Region (10 N)
During scratching under a load of 10 N, large-scale chipping
occurred in the EQ microstructure, as shown in Fig. 6(a). Both the
length of the radial cracks and the radius of the lateral cracks
increased, which led to the formation of large chips whose average
width was ϳ3 times that of the scratch track. The subsurface view
reveals the development of the lateral crack (Fig. 6(b)). A detailed
Fig. 2. Damage following scratching at 1 N normal load observed on (a) the surface and (b) the subsurface cross section in the EQ microstructure, and (c)
the surface and (d) the subsurface cross section in the EL microstructure.

144
Journal of the American Ceramic Society—Xie et al.
Vol. 86, No. 1
Fig. 3. Damage following scratching at 2 N normal load observed on (a) the surface and (b) the subsurface cross section in the EQ microstructure, and (c)
the surface and (d) the subsurface cross section in the EL microstructure.
Fig. 4. Detailed image of surface damage in the EL microstructure after scratching under a load of 2 N.

January 2003
Scratch Damage in Ceramics: Role of Microstructure
145
Fig. 5. Damage following scratching at 5 N normal load observed on (a) the surface and (b) the subsurface cross section in the EQ microstructure, and (c)
the surface and (d) the subsurface cross section in the EL microstructure.
IV. Discussion
is the fracture toughness for the formation of the observed radial
and lateral cracks. Kc can be expressed as a function of the applied
load (P) and the length of the crack (c) that is produced by a sliding
sharp indenter:16
Figures 2–8 demonstrate that microstructure and grain mor-
phology have significant influence on crack formation and subse-
quent material removal during scratching. Based on the observa-
tions in this present work, the scratch damage for the two distinct
microstructures will be discussed below.
P
K ϭ ␰
c
(2)
ͩ ͪ
c3/ 2
where ␰ is a constant. HS is usually given as
(1) Crack Formation and Propagation during Scratching
It is important to consider the following factors separately: (i)
the initiation of cracks that are up to 1–2 grain sizes in dimension;
(ii) the transition of these to the observed radial and lateral cracks;
and (iii) the propagation of these to large cracks, which results in
chipping. The driving force for crack initiation is expected to be
the same for both materials, because the chemistry of both EQ and
EL microstructures is identical. After crack initiation, both me-
chanical and microstructural factors become important in deter-
mining the formation of the observed radial and lateral cracks and
their propagation, as discussed below.
The major difference between static indentation and sliding
scratch tests is the presence of tangential force induced by the
sliding friction of the indenter. Veldkamp et al.12 investigated the
effect of load and speed on crack formation in brittle materials,
using a sharp diamond pyramid to scratch on different types of
ceramics. According to their work, the minimum tangential force
(Ft,min) for the formation of the observed radial and lateral cracks,
during sharp indenter scribing on brittle materials, can be de-
scribed by
4P
HS ϭ
␣␲d2
(3)
where ␣ is a geometrical constant and d is the half width of the
scratch groove. Substituting Eqs. (2) and (3) into Eq. (1), the
minimum tangential force for the formation of the observed radial
and lateral cracks can be rewritten as
6
d
Ft,min ϭ A
P
(4)
ͩ ͪ
n
c
where An is a constant that is related to the geometry of the
indenter and the crack type. Figures 2(a) and (c) and Figs. 3(a) and
(c) show that the width of the scratch groove in the EQ micro-
structure is similar to the corresponding scratch in the EL micro-
structure; thus, both materials have similar scratch hardness, based
on Eq. (3). However, although radial cracks were observed in the
EQ microstructure at loads of Ն1 N, they were not observed in the
EL microstructure until the scratch load was 5 N. These observa-
tions and Eq. (4) indicate that a greater tangential force is required
to form a radial crack of particular length c in the EL microstruc-
ture than in the EQ microstructure.
Kc4
Ft,min ϭ ␹
(1)
ͩ ͪ
HS3
where ␹ is a constant related to the geometry and orientation of the
indenter and the crack type and HS is the scratching hardness; Kc
Considering Eq. (1), therefore, it would seem that the reason
why a greater tangential force is required to extend microcracks to

146
Journal of the American Ceramic Society—Xie et al.
Vol. 86, No. 1
Fig. 6. Damage following scratching at 10 N normal load observed on (a) the surface and (b) the subsurface cross section in the EQ microstructure, and
(c) the surface and (d) the subsurface cross section in the EL microstructure.
Fig. 7. EQ microstructure scratched under a load of 10 N showing (a) the surface and (b) the subsurface views, showing the formation of radial and lateral
cracks, as well as surface cracks. Note that Fig. 7(b) corresponds to the location framed in Fig. 7(a).

January 2003
Scratch Damage in Ceramics: Role of Microstructure
(2) Material Removal during Scratching
147
For the EQ microstructure, radial cracks formed during scratch-
ing under low loads, as shown in Figs. 2(a) and 3(a), and a small
amount of material was pushed to the sides of the scratch track due
to grain dislodgement; however, no significant material removal
occurred. This phenomenon suggests that radial-crack formation
alone was not able to cause material removal at low loads. With the
increase of load, the extent of radial cracks increased and some
intersected each other in the subsurface, as shown in Fig. 7(b),
which suggests that this may be one process for chip formation.
Meanwhile, lateral cracks developed at two different depths at high
load. The lateral cracks that form within 1 ␮m of the surface
interacted with the surface cracks, which led to peeling of the
scratch surface (Fig. 7(b)). The lateral cracks that form at a depth
of Ͼ2 ␮m into the surface are believed to (i) result in partial
removal of the scratch track, as in Fig. 5(a), by extending
underneath the scratch track, and (ii) cause large-scale chipping, as
in Fig. 6(a), because of their propagation along the sides of the
scratch track and intersection with the free surface and/or radial
cracks.
Surface cracking that occurred along the scratch track has also
been observed in sliding spherical contacts on glass32 and in
pin-on-disk wear studies,33 in which the large tensile stresses
behind the sliding contact initiated surface cracks that subse-
quently propagate laterally in the near subsurface. Interestingly,
however, in this current study and in the wear study,33 the
subsurface cracks propagate in the scratching direction, whereas,
in the sliding spherical contact study, the crack propagation was in
a direction opposite to the scratching direction.
For the EL microstructure, the debris at the sides of the scratch
track demonstrates that material removal occurred; however, no
chipping appeared as the load increased. Examination of surface
and subsurface damage revealed that the material-removal mech-
anism is microcracking along the grain boundaries, which leads to
partial grain removal at low and medium load (Fig. 4) and a
mixture of partial grain removal and individual grain dislodgement
at high load (Fig. 8). According to a model proposed by Xu and
Jahanmir,20 in which the process of material removal was assumed
to be dominated by microcracking along the grain boundaries, the
rate of material removal during scratching for both microstructures
at low loads can be calculated as
Fig. 8. Detailed image of surface damage in the EL microstructure
scratched under a load of 10 N. Note that partial grain removal and
individual grain dislodgement occur on the surface.
the observed radial or lateral cracks in the EL microstructure is that
the fracture toughness for the formation of the observed radial and
lateral cracks (Kc) is greater in this material than that for the EQ
microstructure. This condition is because both materials demon-
strate similar scratch hardness.
Following the formation of the observed radial and lateral
cracks, the subsequent propagation of these cracks will be depen-
dent on the fracture toughness for crack propagation (Kc,prop).
Studies have shown that the final crack length following scratch
testing can be expressed for radial cracks as16
2/3
P
c ϭ ␨
(5)
ͩ ͪ
R
Kc,prop
and for lateral cracks as27
P5/8
E4/5
ͩ ͪ
c ϭ ␬
L
(6)
1/ 2
ͩ ͪ
1/2
c,prop
V ϭ ␥͓␺͑␤l ͒ ␴ Ϫ K ͔
P
(7)
K
T
c,s
H9/5
where both ␬ and ␨ are constants that are dependent on the scratch
shape and the elastic–plastic properties of the material. Thus, one
can conclude that the fracture toughness Kc,prop is important in
determining the magnitude of crack propagation.
where V represents the volume of material removed per unit
scratch length, ␥ is a constant independent of grain size and load,
␺ is the crack-geometry coefficient, the term ␤l is the initial crack
size (assumed to be proportional to the grain length l by a
coefficient ␤), ␴T is the tensile stress that is due to the applied load,
As mentioned previously, the driving force for crack initiation
should be the same for both materials. Figures 2(b) and (d) in fact
show the presence of subsurface microcracks in both materials,
following scratching at a load of just 1 N. However, further
increases in load reveal that these microcracks do not further
develop into radial cracks in the case of the EL microstructure until
the scratch load attains a value of 5 N, whereas in the EQ
microstructure, the formation of radial cracks is evident at a load
of 1 N. Furthermore, although lateral crack propagation is evident
at a load of 5 N in the case of the EQ microstructure, lateral cracks
are only just visible in the EL microstructure at a load of 10 N and
seem to result from microcrack coalescence rather than propaga-
tion. In the case of the EQ microstructure at a load of 10 N, radial
and lateral cracks have propagated considerably, resulting in
extensive chipping, as shown in Fig. 6(a). R-curve behavior has
been demonstrated in ␣-SiAlON.31 Vickers-indentation fracture-
toughness measurements, which incorporate the influence of
microstructure on crack propagation, revealed that the fracture
toughness in the EL microstructure was greater than that of the EQ
microstructure, as shown in Table I.29 Relative to the EQ micro-
structure, therefore, greater loads are required to cause the micro-
cracks in the EL microstructure to develop into lateral and radial
cracks and for these cracks to propagate further.
K
c,s is the short-crack toughness, E is the Young’s modulus, and H
is the hardness. In this equation, the influences of thermal-
expansion anisotropy and damage accumulation are ignored,
because the materials tested contained grain-boundary glass phase
with low solidification temperature and the test involved just one
pass. This equation shows that the microfracture-controlled pro-
cess of material removal is promoted by increasing the grain length
and load and inhibited by increasing the hardness of the material.
This prediction is consistent with the amount of apparent debris
that has been observed in this present work at low loads (Ͻ5 N), in
which a greater rate of material removal by the EL microstructure was
observed than by the EQ microstructure. It should also be recognized
that the toughness in Eq. (7), Kc,s, is not the long-crack toughness
(Kc,prop) but the short-crack toughness, which corresponds to small
initial flaws on the order of the grain size. Many studies on the
short-crack toughness of ceramics have reported that the fine-grained
microstructure exhibits greater short-crack toughness than the large/
coarse-grained microstructure34,35 and attributed this to residual
stresses within the microstructure resulting from anisotropy in thermal
expansion.36,37 In this present analysis, the large amount of glassy
phase in this ␣-SiAlON means that there would be limited residual
stress, and, hence, the short-crack fracture toughness Kc,s would be

148
Journal of the American Ceramic Society—Xie et al.
Vol. 86, No. 1
2B. R. Lawn and R. Wilshaw, “Review of Indentation Fracture: Principles and
Applications,” J. Mater. Sci., 10, 1049–81 (1975).
expected to be the same for both the EQ and EL microstructures. In
addition, similar scratch-track widths mean that the hardness H is the
same for both microstructures. For the small volumes of material
being considered under a sliding indenter, there is no reason to expect
that either of the parameters ␴T or E would be significantly different
for the two microstructures. Considering Eq. (7), therefore, differ-
ences in the rate of material removal may be attributed to the
difference in grain size only.
Some researchers predicted, according to the analysis of
short-crack/long-crack toughness effect in ceramic fracture
behavior,5–7,17–22 that the coarse-grained microstructure would
exhibit less wear resistance than the fine-grained microstructure
in abrasive-wear conditions. This conclusion is correct only
when the process of material removal is controlled by micro-
cracking along the grain boundaries.
However, as the load was increased, the advantage of the
elongated microstructure (EL) in suppressing the formation and
propagation of macrocracking, such as the radial and lateral
cracks, became apparent. As a result, large-scale chipping did not
occur in the EL microstructure, even at a load of 10 N. Observa-
tions of the subsurface structure underneath the scratched track in
Fig. 6(d) revealed that no large crack formed, which suggests that
the interlocking elongated grains can effectively impede the
propagation of the cracks and reduce the severity of surface
damage. In the case of the EQ microstructure, the fine equiaxed
grains did not lead to grain bridging, and, hence, no impediment to
radial and lateral crack propagation existed. This phenomenon
resulted in large-scale particle removal in the form of chipping.
3B. R. Lawn and D. B. Marshall, “Indentation Fracture and Strength Degradation
in Ceramics”; pp. 205–29 in Fracture Mechanics of Ceramics, Vol. 3. Edited by R. C.
Bradt, D. P. H. Hassleman, and F. F. Lange. Plenum Press, New York, 1978.
4B. R. Lawn, B. J. Hockey, and H. Richter, “Indentation Analysis: Applications in
the Strength and Wear of Brittle Materials,” J. Microscopy, 130 [3] 295–308 (1982).
5H. Cai, M. A. S. Kalceff, and B. R. Lawn, “Deformation and Fracture of
Mica-Containing Glass-Ceramics at Hertzian Contacts,” J. Mater. Res., 9 [3] 762–70
(1994).
6F. Guiberteau, N. P. Padture, and B. R. Lawn, “Effect of Grain Size on Hertzian
Contact Damage in Alumina,” J. Am. Ceram. Soc., 77 [7] 1825–31 (1994).
7H. H. K. Xu, L. Wei, N. P. Padture, B. R. Lawn, and R. L. Yeckley, “Effect of
Microstructural Coarsening on Hertzian Contact Damage in Silicon Nitride,” J.
Mater. Sci., 30, 869–78 (1995).
8A. C. Fisher-Cripps and B. R. Lawn, “Indentation Stress–Strain Curves for
‘Quasi-Ductile’ Ceramics,” Acta Mater., 44 [2] 519–27 (1996).
9R. F. Cook and G. M. Pharr, “Direct Observation and Analysis of Indentation
Cracking in Glasses and Ceramics,” J. Am. Ceram. Soc., 73 [4] 787–817 (1990).
10D. B. Marshall, B. R. Lawn, and A. G. Evans, “Elastic/Plastic Indentation
Damage in Ceramics: The Lateral Crack System,” J. Am. Ceram. Soc., 65 [11]
561–66 (1982).
11B. R. Lawn, “Indentation of Ceramics with a Sphere: A Century after Hertz,”
J. Am. Ceram. Soc., 81 [8] 1977–94 (1998).
12J. D. B. Veldkamp, N. Hattu, and V. A. C. Snijders, “Crack Formation during
Scratching of Brittle Materials”; see Ref. 3, pp. 273–301.
13M. V. Swain, “Microfracture about Scratches in Brittle Solids,” Proc. R. Soc.
London, A, 366, 575–97 (1979).
14J. C. Conway and H. P. Kirchner, “The Mechanics of Crack Initiation and
Propagation Beneath a Moving Sharp Indentor,” J. Mater. Sci., 15, 2879–83 (1980).
15Z. Bi, H. Tokura, and M. Yoshikawa, “Study on Surface Cracking of Alumina
Scratched by Single-Point Diamonds,” J. Mater. Sci., 23, 3214–24 (1988).
16S. Y. Chen, T. N. Farris, and S. Chandrasekar, “Sliding Microindentation
Fracture of Brittle Materials,” Tribol. Trans., 34 [2] 161–68 (1991).
17H. H. K. Xu and S. Jahanmir, “Scratching and Grinding of a Machinable
Glass-Ceramic with Weak Interface and Rising T-Curve,” J. Am. Ceram. Soc., 78 [2]
497–500 (1995).
(3) Implications for Material Design
18H. H. K. Xu and S. Jahanmir, “Effect of Microstructure on Abrasive Machining
of Advanced Ceramics,” Ceram. Eng. Sci. Proc., 16 [1] 295–314 (1995).
19H. H. K. Xu, S. Jahanmir, and Y. Wang, “Effect of Grain Size on Scratch
Interactions and Material Removal in Alumina,” J. Am. Ceram. Soc., 78 [4] 881–91
(1995).
Some aspects of material design for abrasive-wear applications
can be drawn from this present investigation. For applications in a
low-contact-load regime, both the EQ and the EL microstructures
exhibit low levels of material removal, with the EQ microstructure
exhibiting slightly less material removal. As the contact load in
application was increased, the EL microstructure would surpass
the EQ microstructure, because of its ability to suppress the
propagation of large cracks and subsequent large-scale chipping.
However, these implications are notwithstanding the effect of
multiple contact-load cycles.
20H. H. K. Xu and S. Jahanmir, “Microfracture and Material Removal in Scratching
of Alumina,” J. Mater. Sci., 30, 2235–47 (1995).
21H. H. K. Xu and S. Jahanmir, “Effect of Grain Size on Scratch Damage and
Hardness of Alumina,” J. Mater. Sci. Lett., 14, 736–39 (1995).
22H. H. K. Xu, N. P. Padture, and S. Jahanmir, “Effect of Microstructure on
Material-Removal Mechanisms and Damage Tolerance in Abrasive Machining of
Silicon Carbide,” J. Am. Ceram. Soc., 78 [9] 2443–48 (1995).
23Y. Ahn, T. N. Farris, and S. Chandrasekar, “Sliding Microindentation Fracture of
Brittle Materials: Role of Elastic Stress Fields,” Mech. Mater., 29, 143–52 (1998).
24O. Desa and S. Bahadur, “Material Removal and Subsurface Damage Studies in
Dry and Lubricated Single-Point Scratch Tests on Alumina and Silicon Nitride,”
Wear, 225–29, 1264–75 (1999).
V. Conclusions
The following conclusions can be drawn from the results of this
investigation:
25S. K. Lee, R. Tandon, M. J. Readey, and B. R. Lawn, “Scratch Damage in
Zirconia Ceramics,” J. Am. Ceram. Soc., 83 [6] 1428–32 (2000).
26M. G. Gee, “Low Load Multiple Scratch Tests of Ceramics and Hard Metals,”
Wear, 250, 264–81 (2001).
(1) The fine, equiaxed grain microstructure (EQ) exhibited good
resistance to material removal at low scratch loads, because of its fine
grain structure. Radial cracks formed at low loads, but no chipping
was induced. As the load was increased, radial cracks propagated.
Concurrently, lateral cracks formed and propagated, and large-scale
chipping occurred as a result of low long-crack fracture toughness.
(2) The large, elongated grain microstructure (EL) exhibited a
slightly greater material-removal behavior than the EQ microstructure
at low scratch loads, because of its larger grain size. However, at high
load, the formation and propagation of radial and lateral macrocracks
was suppressed, and no large-scale chipping occurred.
27A. G. Evans and D. B. Marshall, “Wear Mechanisms in Ceramics”; pp. 439–52
in Fundamentals of Friction and Wear of Materials. Edited by D. A. Rigney.
American Society of Metals, Metals Park, OH, 1981.
28B. R. Lawn, “Partial Cone Crack Formation in a Brittle Material Loaded with a
Sliding Spherical Indenter,” Proc. R. Soc. London, A, 299, 307–16 (1967).
29Z.-H. Xie, M. Hoffman, and Y.-B. Cheng, “Microstructural Tailoring and
Characterization of a Calcium ␣-SiAlON Composition,” J. Am. Ceram. Soc., 85 [4]
812–18 (2002).
30N. Rowlands and P. Munroe, “FIB for the Evaluation of Non-Semiconductor
Materials”; pp. 233–41 in Proceedings of the 31st Annual Technical Meeting of the
International Metallographic Society. Edited by D. O. Northwood, E. Abramovici,
M. T. Shehata, and J. Wylie. ASM International, Materials Park, OH, 1998.
31M. Zenotchkine, R. Shuba, J. S. Kim, and I-W. Chen, “R-Curve Behavior of
In-Situ Toughened ␣-SiAlON Ceramics,” J. Am. Ceram. Soc., 84 [4] 884–86 (2001).
32B. R. Lawn, S. M. Wiederhorn, and D. E. Roberts, “Effect of Sliding Friction
Forces on the Strength of Brittle Materials,” J. Mater. Sci., 19, 2561–69 (1984).
33T. E. Fisher, M. P. Anderson, and S. Jahanmir, “Influence of Fracture Toughness
on the Wear Resistance of Yttria-Doped Zirconium Oxide,” J. Am. Ceram. Soc., 72
[2] 252–57 (1989).
(3) Both microstructures developed fine subsurface micro-
cracks at low loads.
(4) The EL microstructure may be preferable to the EQ
microstructure in abrasive-wear applications This conclusion was
due to the interlocking elongated grains in the EL microstructure
that caused higher long-crack fracture toughness, which restrained
the onset of severe abrasive damage such as chipping.
34L. M. Braun, S. J. Bennison, and B. R. Lawn, “Objective Evaluation of
Short-Crack Toughness-Curves Using Indentation Flaws: Case Study on Alumina-
Based Ceramics,” J. Am. Ceram. Soc., 75 [11] 3049–57 (1992).
Acknowledgment
35N. P. Padture and B. R. Lawn, “Toughness Properties of a Silicon Carbide with
in situ-Induced Heterogeneous Grain Structure,” J. Am. Ceram. Soc., 77 [10]
2518–22 (1994).
The authors wish to thank Brian Mubaraki for technical support during the tests.
36P. Chantikul, S. J. Bennison, and B. R. Lawn, “Role of Grain Size in the Strength
and R-Curve Properties of Alumina,” J. Am. Ceram. Soc., 73 [8] 2419–27 (1990).
37R. W. Rice, R. C. Pohanka, and W. McDonough, “Effect of Stresses from
Thermal Expansion Anisotropy, Phase Transformations and Second Phases on the
References
1B. R. Lawn and M. V. Swain, “Microfracture Beneath Point Indentations in Brittle
Solids,” J. Mater. Sci., 10, 113–22 (1975).
Strength of Ceramics,” J. Am. Ceram. Soc., 63 [11] 703–10 (1980).
Ⅺ