J. Am. Ceram. Soc., 89 [3] 1118–1121 (2006)
DOI: 10.1111/j.1551-2916.2005.00824.x
r 2006 The American Ceramic Society
A New Method for Precracking Beam for Fracture Toughness
Yiwang Baow,z and Yanchun Zhou
Shenyang National Laboratory for Materials Science, Institute of Metal Research Chinese Academy of Sciences,
Shenyang 110016, China
suitable for quasi-plastic ceramics or ceramic matrix composites.
Two specific reasons account for this observation: (i) an initial
indentation crack as the crack starter could not be induced on
the surface by the indentation for the quasi-plastic ceramics,10,11
and (ii) the stress resulting from the bridge load is not sufficient
to crack the quasi-plastic ceramics with Vickers indentation, because of local stress relaxation near the indentation.12,13
Reducing the brittleness of ceramics is crucial for engineering
application of ceramics.14 Precracking quasi-plastic ceramics
and in situ observation of the crack growth are two main difficulties that restrict the investigation of the fracture-resistance
behavior of ceramic materials. The purpose of this work is to
seek effective solutions to these problems. Considering the strain
dependence of crack extension15 and the increasing crack resistance in a chevron notch,16,17 a simple approach for precracking
ceramics is presented by controlling the deflection of a beam
sample with a right-angled triangular notch that makes crack
evolution observable under an optical microscope from the
smooth surface of the sample.
A simple and versatile precracking method using a triangular
notch as a crack starter in limited bending was developed, which is
suitable for both brittle ceramics and quasi-plastic materials that
are difficult to precrack by the conventional bridge-indentation
technique. Slow growth of large crack in brittle or quasi-brittle
ceramics was controlled and observed in situ in this way. The
precracking tests performed on various ceramics exhibited high
reliability and feasibility. The precracked specimens were subsequently used to measure the fracture toughness, and the resultant
data showed that the fracture toughness determined by using the
precracked specimens reflected the minimum value of the toughness measured in single edge-notched beam (SENB) tests.
I. Introduction
the fracture toughness, or resistance behavior,
and fatigue crack growth of ceramics or ceramic composites
is important to understand the reliability and safe application of
these materials.1,2 Sakai and Bradt1 pointed out in a review that
creating a starting crack with proper sharpness and length is critical for an accurate fracture toughness measurement. Because of
their low toughness and high elastic modulus, controlling crack
initiation and crack arrest while precracking ceramics is difficult,
especially for a straight-through crack. For most brittle materials,
once macrocrack initiates under a static load, it extends quickly.
Therefore, a small ceramic sample during precracking often displays either catastrophic fracture or non-cracking.
In general, cyclic fatigue on a notched sample is an effective
approach to precrack ceramic samples,3,4 but it is expensive and
time consuming. Another widely used technique to precrack fine
ceramics is the bridge-indentation method,5–7 which has been
adopted in international standard of fracture toughness tests.8
In practice, however, questions often appeared in precracking
brittle materials. Ray9 noticed that the bridge indentation technique has a very low probability of success, so he developed a
modified bridge-indentation approach that improved the success
rate of precracking in brittle ceramics. Nevertheless, the precracking technique has some disadvantages: (i) the load for precracking is always very high, so that it is difficult to perform on a
large sample; (ii) the ultimate length of the pop-in crack is not
easy to control, and it is often not straight or even; and (iii) the
evolution process from crack initiation to extension cannot be
observed and recorded in situ using conventional cameras. Our
experience indicates that the bridge indentation technique is unVALUATING
II. Experimental Procedure
The deflection-controlled precracking tests were conducted, respectively, on Ti3SiC2/SiC composite, Ti3SiC2, and Ti3AlC2 ceramics that are quasi-plastic and difficult to precrack by the
conventional indentation-bridge technique. Bar samples with a
length of 40 mm, a thickness of 4 mm, and a width of 3 mm were
machined using electrical-discharge cutting. All samples were
ground and polished to 5 mm. A triangular notch with a width of
B0.2 mm was cut at the mid-length of each sample as a crack
initiator. The shape, size, and location of the triangular notch
are schematically illustrated in Fig. 1(a). Crack extension was
monitored in situ from the non-notched surface of the sample, so
that the crack length and the extending process should be controlled artificially.
A simple device, which is capable of loading the notched
specimen horizontally under a microscope, controlling the deflection and in situ monitoring of the crack growth, was designed
and used to precrack the ceramic samples. Figure 1(b) shows the
configuration of the fixture and the specimen for the deflectioncontrolled precracking. The experimental procedure is described
as follows: (a) mount the notched sample into the fixture (the
span is 30 mm in this work; two ends of the bar sample were
stuck with a plastic film) and with the non-notched side toward
the microscope; (b) apply a compressive stress (20–50 N)
through a bolt along the direction of the sample length (to reduce the inertia of crack extension) and an initial bending load
(5–10 N) at the mid-span to fix the sample; (c) adjust the stanchion until the slip support contacts the surface of the sample;
(d) withdraw the stanchion to give an initial deflection space
(B20 mm), and then apply a bending load until the mid-span of
the beam sample is supported by the stanchion; and (e) if crack
initiation is not observed from the microscope, repeat step (d)
with a small deflection increment of about 1 mm until crack in-
G. Quinn—contributing editor
Manuscript No. 20922. Received August 25, 2005; approved October 8, 2005.
This work was supported by the National Outstanding Young Scientist Foundation
(No. 50125204 for Y. Bao and No. 59925208 for Y. Zhou), National Science Foundation of
China under Grant No. 50232040, and ‘‘The Hundred-talent plan’’ of Chinese Academy of
Sciences and ‘‘863’’ program.
Author to whom correspondence should be addressed. e-mail: [email protected]
China Building Materials Academy, Beijing, China.
March 2006
Communications of the American Ceramic Society
Fig. 1. Schematic of (a) the beam sample with a triangular notch and
the profile of the cross-section, and (b) configuration of fixture for the
strain-controlled precracking ceramic sample under a microscope and
horizontal loading.
itiates and extends to the expected length (usually half of the
sample thickness8).
The initial deflection space can be estimated from elastic
modulus E, the bending strength sb, and the geometry of the
sample, i.e., the initial deflection space should be slightly smaller
than the critical deflection Df.
Df ¼
L2 sb
6W E
where L is the span and W is the thickness of the sample. If the
strength and modulus are unknown, an initial deflection space
of B20 mm can be used. The crack growth rate is controlled by
adjusting the deflection increment slowly using the adjustable
stanchion as shown in Fig. 1(b). Thus, initiation and extension
of the crack are under direct observation and control. The
length should be controlled in the range of 0.4–0.6W, where
W is the thickness of the sample.
Figure 2 shows the crack initiation and growth process in a
Ti3SiC2/SiC sample with increasing deflection, which indicates
that stable and controllable crack growth can be realized using
this simple approach. The stable crack growth was monitored by
an optical microscope, and the final crack length was usually
controlled to be the same as the depth of the triangular notch
because unstable crack growth may occur when the crack length
is longer than the depth of the triangular notch. Therefore, the
depth of the notch should be prepared according to the requirement of the specimen size, which is usually half of the sample
The advantages of this fresh precracking approach are as follows: (i) it is feasible for various ceramics including brittle and
quasi-plastic ceramics as well as ceramic composites. Thus, those
ceramics for which the traditional bridge-indentation method is
unable to induce precracking, like layered ternary ceramics, can
be precracked easily by this method. (ii) Stable crack growth can
be monitored in situ on the sample surface by using an optical
microscope, which provides a possibility to study the resistance
behavior and the route of a macrocrack in ceramics. (iii) Com-
Fig. 2. Crack initiation and extension in Ti3SiC2/SiC composite ceramics observed in situ using the strain-controlled precracking technique; the
arrowhead indicates the crack tip. (a) Crack initiation at the root of the
triangular notch; (b) crack extension with increasing deflection; and (c)
crack reaches its length of KIC testing.
paring the conventional chevron notch, the triangle-notched
sample has the same resistance behavior and stable extension
potential, but is much easier to prepare and is convenient for in
situ crack growth observation. (iv) No material tester is required, and the small testing device is simple, inexpensive, and
the maximum load is much lower than that needed in the bridge
indentation test (over 10 000 N). Therefore, this precracking
technique has a high success rate, wide applicability, and is convenient for in situ observation.
Machinable ceramics Ti3SiC2 and Ti3AlC2 that cannot be
precracked by the traditional bridge-indentation method were
also used in the test, and the precracked samples were examined
by SEM. The crack tip in the Ti3SiC2 sample is shown in
Fig. 3(a), which indicates that the crack prepared by this approach is as sharp as natural cracks. The width of the precrack is
estimated to be B1 mm. Figure 3(b) exhibits crack deflection and
pullout of grains in Ti3AlC2, which is a typical quasi-plastic ceramic with platelet grains and a relatively weak intergrain bonding. The stable crack growth realized by this method is because
of the limited deflection adjusted by the stanchion and the compressive stress applied normal to the crack plane. Under a limited bending deflection and compression along the length
direction, the strain near the crack tip will decrease with the
crack growth toward the compressive zone, so the stress intensity factor at the crack tip does not enhance rapidly with the
crack growth. The amount of crack growth is restricted by the
decreasing strain near the crack tip. It is worth noting that the
crack prepared by using this approach is almost invisible after
unloading, especially near the crack tip. Therefore, the samples
were infiltrated with dye to determine the average length of the
precrack after fracture tests. In this work, a yellow–green ink for
nite writer pen (mark pen) was used as the dye, which was found
to be effective. For some ceramics, especially oxide ceramics, the
water in the dye may lead to an environment assisting in crack
growth. Therefore, seeking the best dye for crack identification
is necessary. In the present method, for both crack opening displacement and crack length increased with increasing deflection,
the crack size could also be measured in situ by means of a scale
Vol. 89, No. 3
Communications of the American Ceramic Society
Fig. 5. Fracture section of a precracked Ti3SiC2 sample with a triangular notch as the crack starter, after the test of toughness measurement
and dye infiltration; the long arrowheads indicate the front of the precrack and the short arrows indicate the location of the first pop-in.
Fig. 3. SEM micrographs of precracks, with arrowheads indicating the
direction of crack extension. (a) Back-scattered micrograph of the crack
tip on Ti3SiC2, and (b) the crack in Ti3AlC2 showing typical crack deflection and pulling out.
microscope. However, the measured size only represents the
crack length on the observation surface, rather than the average
depth. In general, the average value of three measured depths
distributed on the cross section is taken as the depth of the precrack.8 The crack depth can also be estimated simply by averaging the crack lengths on both sides of a sample, before the
tests of fracture toughness or resistance behavior. The different
precrack lengths on the specimen sides could be an issue for
some materials that have pronounced R-curves and bridges that
persist even with crack lengths 1 mm long or more. If the crack
lengths on two sides have great deviation and the ratio of the
difference between crack lengths on two sides to the average
length is over 0.2, the test would be considered to be invalid.
The fracture toughness of the Ti3SiC2 precracked specimens
was measured in accordance with the same procedures in the
single edge-notched beam8 (SENB) ISO standard 15732, except
for the use of the new precracking procedure. The measured
fracture toughness was 6.270.32 MPa m1/2. For comparison,
SENB and chevron notch beam (CNB) tests gave results of
6.6270.12 MPa m1/2 and 6.670.15 MPa m1/2, respectively,
using the same sample size and 0.2 mm notch width. The precrack was a sharp and straight-through crack with only a 1–2
mm gap between two crack faces. Figure 4 shows the single edge
pre-cracked beam triangular-notched specimen data in comparison with SENB results as a function of notch gap size. The resultant data indicate that the measured fracture toughness is
approximately a linear function of the notch width in a certain
range. This linear relationship between the measured toughness
and the notch width implies that the fracture toughness determined using the precracked samples represents the minimum of
toughness in SENB tests. In this precracking method, the precrack face is parallel to the face of the triangle notch, and the tip
of the straight-through precrack is approximately in a straight
line. Figure 5 shows a fracture surface of a precracked sample
after a toughness test, which displays the trace of the triangular
notch and a precrack that was not uniformly dyed; therefore, a
better dye and dye method have to be sought.
III. Conclusions
A new approach of precracking is developed, which shows a
wide applicability and a high success ratio, and is suitable to
produce, observe, and control stable growth of a macrocrack in
brittle or quasi-plastic ceramics that are difficult to precrack by
the traditional bridge-indentation technique. It allows the extension of a macrocrack in ceramics to be monitored in situ and
to be observed by using an optical microscope or SEM. In addition, this form of crack growth and arrest demonstrates the
strain dependence of the crack extension in ceramics. The fracture toughness measured by using the precracked samples reflects the fracture toughness of the natural and long crack.
The authors sincerely thank Dr. Peggy Y. Hou at Ernest Orlando Lawrence,
Berkeley National Laboratory, for helpful discussion and review of the manuscript.
Fig. 4. Comparison of measured fracture toughness in T3SiC2 ceramics,
between the data obtained from the precracked samples and samples with
various widths of straight-through notches. Each value represents an average of six specimens, and the half error bar is a standard deviation.
M. Sakai and R. C. Bradt, ‘‘Fracture-Toughness Testing of Brittle Materials,’’
Int. Mater. Rev., 38, 53–78 (1993).
March 2006
Communications of the American Ceramic Society
G. Rausch, M. Kuntz, and G. Grathwohl, ‘‘Determination of the In Situ Fiber
Strength in Ceramic–Matrix Composites from Crack-Resistance Evaluation Using
Single-Edge Notched-Beam Tests,’’ J. Am. Ceram. Soc., 83, 2762–8 (2000).
C. J. Gilbert, R. O. Ritchie, and W. L. Johnson, ‘‘Fracture Toughness and
Fatigue-Crack Propagation in a Zr–Ti–Ni–Cu–Be Bulk Metallic Glass,’’ Appl.
Phys Lett., 71, 476–8 (1997).
F. Dahmani, A. W. Schmid, and J. C. Lambropoulos, ‘‘Lifetime Prediction of
Laser-Precracked Fused Silica Subjected to Subsequent Cyclic Laser Pulses,’’ J.
Mater. Res., 15, 1182 (2000).
I. Bar-On, J. T. Beals, L. Gary, and C. M. Murray, ‘‘Fracture Toughness of
Ceramic Precracked Bend Bars,’’ J. Am. Ceram. Soc., 73 [8] 2519–22 (1990).
T. Nose and T. M. Fuji, ‘‘Evaluation Fracture Toughness for Ceramic Materials by a Single-Edge-Precraked Beam Method,’’ J. Am. Ceram. Soc., 71, 328–33
Y. Shindo, H. Murakami, K. Horiguchi, and F. Narita, ‘‘Evaluation of Electric
Fracture Properties of Piezoelectric Ceramics Using the Finite Element and SingleEdge Precracked-Beam Methods,’’ J. Am. Ceram. Soc., 85, 1243–8 (2002).
ISO/15732 (2003), Fine ceramics—Test method for fracture toughness of monolithic ceramics at room temperature by Single Edge Pre-cracked Beam (SEPB)
A. K. Ray, ‘‘A New Technique for Precracking Ceramic Specimens in Fatigue
and Fracture,’’ J. Euro. Ceram. Soc., 18, 1655–62 (1998).
I. M. Low, ‘‘Vickers Contact Damage of Micro-Layered Ti3SiC2,’’ J. Euro.
Ceram. Soc., 18, 709–13 (1998).
T. El-Raghy, A. Zavaliangos, M. W. Barsoum, and S. R. Kalidindi, ‘‘Damage
Mechanisms Around Hardness Indentations in Ti3SiC2,’’ J. Am. Ceram. Soc., 80,
513–6 (1997).
I. M. Low, S. K. Lee, B. R. Lawn, and M. W. Barsoum, ‘‘Contact Damage
Accumulation in Ti3SiC2,’’ J. Am. Ceram. Soc., 81, 225–8 (1998).
X. H. Wang and Y. C. Zhou, ‘‘Microstructure and Properties of Ti3AlC2
Prepared by the Solid–Liquid Reaction Synthesis and Simultaneous In Situ Hot
Pressing Process,’’ Acta Mater., 50, 3143–9 (2002).
B. R. Lawn, N. P. Padture, H. Cai, and F. Guiberteau, ‘‘Making Ceramics
Ductile,’’ Science, 263, 1114–6 (1994).
Y. W. Bao and R. W. Steinbrech, ‘‘Strain Criterion of Fracture in Brittle
Materials Matter,’’ Sci. Lett., 16, 1533–5 (1997).
ISO/DIS 24370, Fine ceramics–Test method for fracture toughness of monolithic ceramics at room temperature by the chevron notch beam (CNB) method
D. Munz, R. T. Bubsey, and J. L. Shannon Jr, ‘‘Fracture Toughness Determination of Al2O3 Using Four-Point-Bending Specimens with Straight Through
and Chevron-Notches,’’ J. Am. Ceram. Soc., 63, 300–5 (1980).