On the Vickers Indentation Fracture Toughness Test

George D. Quinn
w
National Institute of Standards and Technology, Gaithersburg, Maryland 20899-8529
Richard C. Bradt
Department of Metal. & Mater. Science, College of Engineering, University of Alabama, Tuscaloosa, Albama 35487-0202
The Vickers indentation fracture toughness test, or VIF, is ad-
dressed by considering its origins and the numerous equations
that have been applied along with the technique to estimate the
fracture resistance, or the K
Ic
of ceramics. Initiation and prop-
agation of cracks during the VIF test are described and contrast-
ed with the pre-cracking and crack growth for internationally
standardized fracture toughness tests. It is concluded that the
VIF test technique is fundamentally different than standard frac-
ture toughness tests. The VIF test has a complex three-dimen-
sional crack system with substantial deformation residual stresses
and damage around the cracks. The VIF test relates to an ill-
dened crack arrest condition as opposed to the rapid crack
propagation of the standardized fracture toughness tests.
Previously published fracture toughness results employing the
VIF technique are reviewed. These reveal serious discrepancies
in reported VIF fracture toughness values. Finally, recent frac-
ture resistance measurements by the VIF technique for the
Standard Reference Material SRM 2100 are presented. These
are compared with standardized test results for the same mate-
rial. It is concluded that the VIF technique is not reliable as a
fracture toughness test for ceramics or for other brittle materi-
als. What the VIF actually measures in terms of fracture re-
sistance cannot be readily dened. It is recommended that the
VIF technique no longer be acceptable for the fracture toughness
testing of ceramic materials.
I. Introduction
T
HE Vickers indentation fracture (VIF) test has been a pop-
ular experimental technique for the estimate of the fracture
resistance of brittle ceramics for the past three decades. The
concept of monitoring cracks created by hardness indentations
originated with Palmqvist
1
in the 1950s, half a century ago, but
the concept was rst advocated as a fracture toughness test by
Evans and Charles
2
in the mid-1970s. Evans and Charles applied
the VIF to materials from single crystal oxides to cemented car-
bides, all of which seemed amenable to the test. The technique
rapidly achieved popularity because of its expediency. It seemed
to be a convenient way to determine K
Ic
. It requires only a small
volume of material and the actual measurements can be quickly
completed with only a short sample preparation time and at a
minimum nancial cost. In nearly all respects the VIF method
seemed to be practically ideal, although it is admittedly a rather
non-conventional approach for determining the fracture tough-
ness.
The VIF technique has been thoroughly described in two in-
dependent, comprehensive reviews of the fracture toughness
testing of brittle materials.
3,4
Recently, a third review, that by
Morrell
5
categorizes the VIF as not meeting fracture mechanics
criteria, a stunning indictment. The present authors believe a
fair assessment of the VIF is that most currently active research-
ers do not consider the VIF to be an accurate, nor a reliable
method for determining K
Ic
or any other fracture resistance pa-
rameter. These concerns were amplied when the VIF test ex-
perienced difculties in three international round robin
exercises.
6,7
Unfortunately, the VIF method is being considered
in some quarters as a possible method for evaluating fracture
toughness for materials specications. These combined issues
have prompted this re-examination of the validity of the VIF
technique for the fracture resistance testing of ceramics.
This paper rst reviews the origin of the VIF technique and
then considers the crack initiation and crack propagation during
a VIF measurement. The VIF multiple cracks are then com-
pared with the single crack of approved standardized fracture
toughness tests. This paper considers the numerous equations
that are applied in the VIF method. The fundamental origins
and the accuracies of these equations are questioned. Next, the
results are addressed from the numerous equations which have
been applied to the VIF technique as they have been summar-
ized in several comprehensive archival papers. Finally, recent
VIF data were obtained and compared with similar measure-
ments by approved international standard fracture toughness
tests for a NIST Standard Reference Material developed espe-
cially for standardized fracture toughness testing of ceramics. In
combination, these considerations have lead to the conclusion
that the VIF technique should be discontinued for the measure-
ment of the fracture toughness of brittle materials.
Feature
D. Greencontributing editor
This work was supported by the U. S. Department of Energy, Heavy Vehicle Materials
Program with Oak Ridge National Laboratory.
w
Author to whom correspondence should be addressed. e-mail: geoq@nist.gov
Manuscript No. 22201. Received September 1, 2006; approved November 22, 2006.
J
ournal
J. Am. Ceram. Soc., 90 [3] 673680 (2007)
DOI: 10.1111/j.1551-2916.2006.01482.x
r 2007 The American Ceramic Society
No claim to original US government works
II. Origin of the VIF Technique
The Vickers indentation technique was introduced to the ce-
ramic fracture community by Evans and Charles
2
in a brief
communication note approximately 30 years ago. It appeared
during a period when considerable attention was focused on the
nature of deformation and cracking around Vickers indenta-
tions,
813
a eld that became known as indentation mechanics.
z
The Evans and Charles note was not very extensive by any
measure, but it undeniably suggested that one might utilize the
cracks which emanate from the corners of a high testing load
Vickers microhardness indentation to determine the fracture
toughness of brittle materials. Evans and Charles presented a
generalized equation and a normalized calibration curve that
appeared to apply to many different materials, some with
Palmqvist cracks and some with median cracks. It appeared
that this technique might be a universal fracture toughness test.
It certainly was a very convenient one.
Subsequently, numerous other authors have used similar
curve tting methods and dimensional analyses to relate the ex-
tending crack lengths from a Vickers indentation impression to
the fracture toughness for both mixtures of Palmqvist and me-
dian cracks and for each type of crack independently. Many of
the researchers have been puzzled with their results from the
VIF measurements, which has lead to a proliferation of numer-
ous equation modications to satisfy those researchers in their
quest for a satisfactory value of the fracture toughness. At last
count, more than 30 equations have been presented in the liter-
ature. Not surprising, only rarely do any two of them produce
the same result for the fracture toughness for the same test con-
ditions.
Fracture toughness determination by the VIF technique pro-
ceeds by rst preparing a high-quality smoothly polished test
specimen surface. The specimen is not precracked. The polished
specimen surface is then indented by a Vickers pyramidal mi-
crohardness indenter. Usually a conventional hardness testing
machine is used to gradually apply load to the indenter to peak
load, hold the load constant for a set dwell period, and then
unload the indenter. The sample is indented at a high testing
load to create a deformed region beneath and in the vicinity of
the impression, which leads to the generation of four cracks
emanating from the corners of the square Vickers diamond im-
pression. Figure 1 illustrates a Vickers indentation with four
such cracks radiating from the impression corners. Not all in-
dentations and their cracks are as ideal as those shown in Fig. 1,
as Fig. 2 illustrates. The lengths of the cracks, along with the
indentation load, the impression size, the hardness and elastic
modulus of the material, and an empirical calibration constant
are used to compute the fracture toughness of the material.
Considerable care must be used to measure the length of the
cracks. Variations in microscopic techniques and equipment as
well as observer skill and subjectivity can cause considerable
uncertainty. Although the result is not a valid fracture tough-
ness, most papers call the resulting value K
Ic
.
During Vickers indentation, with increasing indentation
loads, one observes a transition during the development of the
cracks about the impression. For low test loads, the indentation
begins as a crack-free pyramidal microhardness impression. It
then evolves or transforms to an impression with individual
cracks emanating from the four corners of the impression at the
surface, but also containing an extensive crack pattern beneath
the surface. Finally, it experiences extensive lateral cracking and
often considerable spalling about the impression for the higher
indentation test loads as shown in Fig. 2. These crack-type in-
dentation transitions occur at different applied test load levels
for different ceramic materials. The material surrounding the
impression is left with a complex residual stress state dependent
on the indentation load and the multiple cracking.
There are many questions as to the exact sequence of crack
growth from the Vickers indentation corners for different ce-
ramic materials.
1416
For example, it may be that the cracks rst
form as Palmqvist cracks that later extend to form median
cracks, or median cracks may form directly from the deform-
ation beneath the indenter. The cracks may form either during
indenter loading or unloading. The environment may also play a
role. It is not easy to ascertain the sequence from the end result
that is visible only on the sample surface, but in all cases the
cracks extending from the four corners of the pyramidal Vickers
indentation are measured, as is the size of the indentation itself.
The latter point, the size of the indentation, is often a quite
challenging measurement for the higher test loads especially
where there has been considerable spalling about the indentation
impression such as shown in Fig. 2.
III. Equations for Fracture Resistance from the VIF Technique
As previously noted, numerous VIF equations by different au-
thors have been summarized in various forms in different jour-
nal and proceedings publications. It is not sufcient to simply
state that none of them are based on exact theory. Many equa-
tions are the products of dimensional analyses modied with
experimentally derived calibration factors with occasional vague
Fig. 1. A 98 N Vickers indentation with cracks in a NIST Standard
Reference Material 2100 test piece.
Fig. 2. A 98 N Vickers indentation in a boron carbide. Note the severe
spalling.
z
Lawn and Fuller
9
suggested as early as 1975 that: Indentation fracture presents itself
as a useful addition to the mechanical testing repertoire of materials scientists.
674 Journal of the American Ceramic SocietyQuinn and Bradt Vol. 90, No. 3
allusions to a theoretical basis. Many are just the manipulation
of previous equations with new calibration constants in an at-
tempt to achieve reasonably correct values of the fracture tough-
ness for the material under investigation. Most, if not all, of
these equations and their empirical derivations are highly ques-
tionable for none of them are the product of an accurate stress
intensity factor solution. The complex crack network and re-
sidual stress damage zone around indentations are not amenable
to a straightforward analysis as in most traditional fracture me-
chanics test congurations. Yet, the equations for the VIF have
been attractive for they sometimes yield, fortuitously, what ap-
pear to be reasonable values for the fracture toughness. This has
encouraged the continued use of the VIF technique for the de-
termination of the fracture toughness of brittle materials in nu-
merous academic and industrial laboratories. This continued use
has persisted in spite of the suspect reputation of the VIF tech-
nique and a general skepticism and mistrust by the traditional
fracture mechanics community.
Following the original Evans and Charles note, Marshall and
Evans
17
rened the original equation and published a simplied
formula. This was followed by two papers by Anstis et al.
18
Chantikul et al.
19
that presented further renements to the
proposed equations. Two additional papers by Niihara
20,21
followed, the latter specifically addressing Palmqvist cracks.
An equation by Miyoshi
22
for median cracks that is an adap-
tation of Anstis et al.s
18
has also been used. It is the equations
from these papers that seem to be the most frequently
quoted and the most often applied. However, a plethora of ad-
ditional equations in other publications has followed. A list of
equations will not be presented here. Many of them have been
tabulated in the series of four papers by Li et al.,
23
Ponton and
Rawlings,
24,25
and Ghosh et al.
26
which are discussed in some
detail later.
Many derivations make assumptions about the plastically
deformed damage zone underneath the indentation. The dam-
age zone is commonly modeled by an expanding cavity in the
solid. The stress distribution from the expanding cavity is in turn
converted to an equivalent force that acts to open the median or
Palmqvist cracks. Other models have the plastic zone acting as a
wedge or a spring. There are numerous assumptions about de-
formation and fracture built into these models. It appears that
many of the equations are manipulations of previous equations
with new calibration constants in an attempt to achieve reason-
ably correct values of the fracture toughness for the material
under investigation. Ponton and Rawlings
24,25
have summarized
many of the equations along with the assumptions and the in-
consistencies in their 1989 review papers.
It is important to note a couple of features of many of these
equations, features which cast considerable doubt on their fun-
damental nature and, ultimately, their value. All of the equa-
tions contain the indentation test load, the crack length, the
indentation size, the elastic modulus, and the hardness of the
material in one form, or another. Often these parameters are
raised to peculiar non-integer exponents that consist of fractions
or decimals, such as 0.4, 0.6, and 0.7. In other applicable equa-
tions, the same parameters are raised to 0.5, 1.5, or 2.0 powers.
These aspects of the multitude of equations do not inspire con-
dence in their utilization. The peculiar power quantities are not
surprising when it is considered that the equations are neither
the result of a closed-form analytical solution, nor a formal
stress intensity analysis of the cracks that result, but rather are
just sort of an informal adjustment of previous equations. In
addition to, or perhaps because of the above inconsistencies,
various geometrical effects and other factors are often collected
into a dimensionless calibration factor at the beginning of the
equation. In most instances, it is a formidable, if not impossible,
challenge for any scholar to obtain that calibration factor from
rst principles. This factor has often been adjusted or tuned to
t expected results for a particular material. To simply state that
these equations are highly suspect would be an understatement
for there is really little, if any, defensible basis for their justi-
cation from a fundamental perspective.
IV. Concept of Fracture Toughness
Almost all of the publications employing the VIF technique
refer to the resulting measurement as the fracture toughness, or
K
Ic
. For that reason, it is important to consider the definition of
the concept of fracture toughness before further considering the
crack initiation and growth situation in the VIF technique and
contrast it with that of a single crack in the accepted standard-
ized tests. For a basic definition of K
Ic
, a quote from the recent
text of Munz and Fett
27
sufces: If a component or a test
specimen with a crack is loaded, K
I
increases with increasing
load until unstable crack propagation occurs at a critical value
of K
I
. This critical value is the fracture toughness, K
Ic
. Perhaps
it is the specicity of this precise definition that led Morrell
5
to
classify the VIF as one of the Test methods not meeting frac-
ture mechanics criteria. Although undetected slow crack
growth and R-curves have created some problems with fracture
toughness testing of ceramics in the past, it is now generally ac-
cepted that many brittle materials actually do have a definite
value of the fracture toughness in the classical sense. These in-
clude many ne-grained polycrystalline ceramics as well as
coarse-grained ceramics that fracture transgranularly without
bridging. For example, sintered a silicon carbide with small
amounts of boron and carbon sintering aids and no boundary
phase has a specific fracture toughness, K
Ic
.
23,26,28
The concept of K
I
is derived from linear elastic fracture me-
chanics, LEFM. It is a measure of the magnitude of the stress
eld in the crack tip region. When the Roman numeral I is af-
xed to K as a subscript, it refers to the tensile or opening mode
condition of failure, as opposed to a shear mode such as speci-
ed by a II or a III. Thus the quantity known as the fracture
toughness, K
Ic
, refers to the critical value of the stress intensity
factor in the opening mode when fracture initiates and unstable
crack propagation occurs. A part of that definition is the sub-
script c for it is considered to refer to critical or catastrophic
crack propagation. As such it is usually referred to the particular
condition of crack growth that proceeds rapidly and completely
through the test specimen. In fact, several standardized fracture
toughness tests experience just that: a catastrophic fracture with
the single pre-crack propagating completely through the test
specimen once K
Ic
is reached at the crack tip during loading.
One exception is the chevron notch in bending method, a stand-
ardized test method in which the crack is propagated stably
through a triangular cross sectional area.
35,y
Because of the complexity of interacting stress elds from
multiple cracks, the standardized tests for the measurement of
the fracture toughness or K
Ic
consist of only a single well-dened
crack in a geometrically simple test specimen that is subjected to
a well-dened loading conguration. These points are reected
in the makeup of the internationally accepted standardized frac-
ture toughness tests for K
Ic
, such as those that have been ap-
proved for ceramics and other brittle materials including the
chevron-notched beam, (CNB), single-edge pre-cracked beam
(SEPB), and the surface crack in exure (SCF) methods.
2935
Although fracture toughness is believed to be a material con-
stant, like all other engineering parameters, experimental K
Ic
measurements are subject to statistical scatter. In principle, all
valid techniques for fracture toughness measurement should
yield the same value for a material within the normal statistic-
al uncertainties of the test itself. To achieve such accuracy, sev-
eral criteria for a standard K
Ic
test have been established by the
international fracture testing standards community. The test
must consist of a well-dened simple specimen geometry for
which the state of stress of the test piece is known during load-
ing. For simplicity, there must be a sharp single pre-crack in the
specimen for which the alteration of the uncracked stress eld is
known, so that K
I
is well-dened with an accurate stress inten-
sity solution for the particular test specimen. There must not be
y
There are other congurations, such as the double cleavage-drilled compression method
and Obreimoffs wedge-loaded mica ake arrangement, that rely on special geometries to
promote stable crack extension. Congurations that promote stable extension may be valu-
able for studying R-curve behavior. Nevertheless, most K
Ic
methods are intended to measure
the critical value of the stress intensity by initiating fracture with unstable crack extension.
March 2007 On the Vickers Indentation Fracture Toughness Test 675
any ambiguity relative to the value of K
Ic
at the crack tip during
loading when the test specimen achieves the critical condition
and fails catastrophically. These criteria demand a definite
mathematical expression for the fracture toughness, an expres-
sion for K
Ic
that involves the specimen geometry, the descriptive
pre-crack parameters and the maximum load at failure during
the test. All of the accepted standardized tests for ceramic ma-
terials meet these stringent criteria.
2934
This is the reason why
the standardized tests have been demonstrated to yield the same
value of the fracture toughness for a NIST Standard Reference
Material as will be discussed later.
V. VIF in Contrast to Standardized Fracture Toughness Tests
Having described both the non-conventional VIF technique and
the standard fracture toughness tests, it is necessary to compare
the two on the critical points. First is the specimen geometry.
For the standard tests, it is a simple, well-dened geometry for
which the loading conditions and the state of stress are well-
known during the loading process. The geometry is usually one
for which a closed-form solution of the stress state is available in
many standard mechanics texts. For the VIF technique, any
shape or chunk of material is adequate so long as it can be
mounted and polished. The standard tests employ pre-cracked
specimens with a single well-dened crack for which there exists
a formal stress intensity solution. There is no pre-crack in the
VIF specimen: multiple cracks are generated during the actual
test when the indenter is forced into the polished specimen sur-
face. Despite allusions to the contrary, the VIF cracks do not
have an applicable stress intensity solution.
In the standard tests, loading is applied in a universal me-
chanical testing machine, but in the VIF technique the specimen
is loaded through a hardness indenter. During loading in the
standard tests, the pre-crack extends in an accelerating cata-
strophic manner through the entire test specimen. In the VIF
method, deformation occurs below and around the indentation,
cracks form and propagate beneath the indenter and then inter-
sect the surface. The cracks decelerate to an arrest condition
away from the indentation. The VIF leaves a multiple crack
condition about the indentation. At the end of a standardized
fracture toughness test, the single crack has halved the specimen,
but in the VIF method multiple cracks have arrested in a post-
test conguration and material environment that consists of
multiple cracking and residual stresses of considerable complex-
ity. Often major spalling occurs about the indentation, especially
for the higher test loads in the VIF technique. The multiple
cracks in a VIF test do not remotely ascribe to the definition of
fracture toughness in the text of Munz and Fett,
27
nor the many
other similar definitions for fracture toughness that appear in
various other textbooks.
VI. Published Summary Results for the VIF Technique
There are numerous publications involving the use of the VIF
method, but four papers with similar thrusts highlight the basic
problems of the VIF. They extensively summarize the VIF
method by evaluating the results of the technique using many
of the equations which have been suggested for the VIF.
2326
Although these publications are now more than a decade old,
when the four are considered together, they are quite revealing
and demonstrate the failings of the VIF technique. The papers
of Li et al.,
23
Ponton and Rawlings,
24,25
and Ghosh et al.
26
summarize a large number of VIF results using the numerous
different equations for the same materials. Their compilations
alone are impressive. The sintered alpha silicon carbide used by
Li et al.
23
and Ghosh et al.
26
was a candidate for a NIST K
Ic
Standard Reference Material. The papers listings of fracture
toughness numbers will not be duplicated here, only the most
important results will be summarized. The three most critical
conclusions of these four publications relative to the credibility
of the VIF method are as follows:
(1) When the many different proposed equations are ap-
plied to calculate the fracture resistance at the same, or at very
nearly the same crack lengths in a ceramic material, then widely
ranging
z
values of K
Ic
result.
(2) None of the equations are able to generate accurate re-
sults for different materials. An equation may, by chance or by
force tting the calibration factors, produce plausible results for
one specific material, but usually gives poor results for others
even within a single material class.
(3) When the different equations are applied to estimate the
fracture resistance as it varies with the crack length from the
indentations, as determined for different indentation test loads,
then sometimes the fracture resistance appears to increase and
sometimes it decreases with crack length for the same material.
The R-curve like results are frequently contradictory for the
same material when determined by the different fracture resist-
ance equations for the VIF.
These conclusions from four different, but closely related pa-
pers clearly illustrate the basic deciencies of the VIF method.
The rst is that the VIF method is not consistent. The technique
just does not produce a unique value for K
Ic
, which is not sur-
prising. This suggests that the VIF technique does not and
probably cannot measure the value of K
Ic
for brittle materials
such as ceramics. The range of fracture resistance values cer-
tainly points out the unreliability of the VIF method. A funda-
mental problem with the VIF method is that brittle materials do
not deform and fracture underneath an indentation in a self-
similar manner as many have assumed. Covalently-bonded hard
ceramics (e.g., silicon carbide) deform and fracture very differ-
ently than ionically-bonded cubic ceramics (e.g., magnesium ox-
ide) or glasses. Fine-grained and coarse-grained ceramics often
do not deform and fracture the same way. Furthermore, the
cracks often are not the idealized semicircular median or Palmq-
vist cracks assumed in the models. The VIF crack system is ac-
tually a three-dimensional network of intersecting orthogonal
and interacting cracks. Hence, it should not be surprising that
the quest for the universal indentation fracture equation or cal-
ibration constant has been in vain.
The observation that fracture resistance appears to vary with
different length VIF cracks is equally disturbing when one con-
siders the contradictory nature of the results in light of the other
reported R-curve trends for brittle ceramics. R-curves are either
at, as for glass and some ne-grain size ceramics, or they have
an increasing resistance trend with increasing crack length as for
coarse-grain size polycrystalline ceramic materials and ceramic
composites. R-curves should not be decreasing with increasing
crack length. This makes the VIF results even more confusing
and problematical if one wishes to consider that the VIF tech-
nique actually measures some form of crack arrest phenomenon,
or a value for K
Ia
somewhere along the R-curve of the test ma-
terial.
The R-curve, K
Ia
interpretation of the crack arrest situation in
the VIF actually has considerable intrinsic appeal, for the cracks
from the indentation corners obviously become arrested in the
VIF test. Unfortunately, the VIF condition of crack arrest is
virtually impossible to specify with any degree of certainty be-
cause of the auxiliary multiple cracking and spalling about the
indentations at higher test loads. In addition there are substantial
residual stresses from the indentation deformation and the pile
up of owed material about the impression.
36,37
This extremely
complicated situation no doubt results in crack opening displace-
ments (COD) for the VIF cracks that are very different from the
COD values that occur in the typical double cantilever beam or
chevron V-notched specimens which are commonly utilized for
ceramic R-curve determinations. Furthermore, the VIF cracks
arrest in a multiple crack environment, where the multitude of
cracks varies, dependent on the auxiliary cracking and spalling.
In spite of its R-curve like appeal, the contradictory results
z
Ponton and Rawlings
25
showed that for identical indentation data the computed frac-
ture toughness varied by as much as a factor of 7 for the 19 different VIF equations they
tabulated. Variations of almost a factor of 2 were obtained with some of the commonly used
VIF equations.
676 Journal of the American Ceramic SocietyQuinn and Bradt Vol. 90, No. 3
from the different VIF equations indicate that this is probably
not a valid interpretation. The many equations for the VIF do
not consistently predict the expected trends of increasing crack
growth resistance with increasing crack length and they most
certainly do not all predict the same R-curve trends.
The collective results in the four quoted papers
2326
clearly
signaled a serious problem with the VIF technique over a decade
ago. Unfortunately, many researchers have continued to use the
VIF method in spite of these red ag warnings in the archival
literature. It is puzzling to say the least. It is no doubt the con-
sequence of the expediency of the VIF technique to obtain a
result, even though the result is probably not a correct one. Of
course, it is also possible that a VIF toughness result may for-
tuitously appear to be reasonable, another point which has lead
to its continued application.
VII. New VIF Results for the NIST SRM 2100 and Sintered a
Silicon Carbide
The general state of fracture toughness testing has improved
considerably in the past 30 years since the appearance of the VIF
method. In the 1970s only a few academic laboratories and even
fewer industrial and government laboratories were active in the
measurement of fracture resistance. Today, standards organiza-
tions in Europe, Japan, and the US have rened and optimized
fracture testing techniques.
2931
The International Organization
for Standards (ISO) approved standard test methods are now
available for the SEPB,
32
the CNB,
34
and SCF
33
tests. In the late
1990s, SRM 2100, was prepared by NIST.
3840
It is a commer-
cial hot-pressed ne-grain size silicon nitride that has been dem-
onstrated to yield the same K
Ic
value irrespective of the test
method. The certied value for the fracture toughness, K
Ic
, of
SRM 2100 is 4.57 MPa m
1/2
with an uncertainty of 0.11
MPa m
1/2
for the average of ve test outcomes at the 95%
condence level. Thus, if one makes ve measurements of the
fracture toughness, 95% of the time the result should be between
4.46 and 4.68 MPa m
1/2
. The uncertainty is remarkably small
and attests to the consistency of the SRM. Almost identical re-
sults were obtained by SEPB, CNB, and SCF methods despite
the dramatic difference in the precrack types and sizes as shown
in Fig. 3.
For VIF measurements of SRM 2100, test bar halves (3
mm4 mm23 mm) which had previously been used to es-
tablish the certied value for SRM 2100 were mounted and pol-
ished. Indentations were put into surfaces with different
orientations in order to account for possible preferred orienta-
tion effects in this hot-pressed material. Indentations were made
with a diamond Vickers indenter in a conventional hardness
tester that was checked with a NIST SRM 2831 reference disk.
41
Dwell time was 15 s. Ten indentations were made at each of four
indentation loads: 9.8, 19.6, 49, and 98 N. Indentations at 196 N
were made in one specimen. One interesting observation made
during all these experiments was that the crack lengths for a
single indentation often differed by 2%11%. Some indenta-
tions had cracks with similar sizes. The crack length differences
for a single indentation were greater in the rst test piece (Fig. 4a
and Table I). The longer cracks were in the hot pressing plane,
which is expected to have lower fracture toughness. The second
specimen with the cracks running perpendicular to the hot
pressing plane had much less variability in crack lengths. An
average crack length size was used in all of our calculations.
Wilantewicz
16
has observed that one median or radial crack
system often forms before the perpendicular crack system and
could interfere with the latter. Further details of these tests are
described elsewhere.
7
Values of the fracture toughness were calculated using three
different popular equations: Anstis et al.s equation,
18
Niiharas
equation
20,21
as used in a ceramic ball bearing specication,
42
Fig. 3. Fracture surfaces of two silicon nitride Standard Reference Ma-
terial 2100 specimens. The single-edge precracked beam test piece on the
left had a 2 mm deep precrack (arrows), whereas the surface crack in
exure test piece on the right had a 50 mm deep semielliptical precrack
(arrow). Virtually identical fracture toughnesses were obtained despite
the dramatic difference in precrack sizes and shapes.
Fig. 4. Representation of the Vickers indentation fracture results for two Standard Reference Material 2100 test pieces with different orientations.
Miyoshis equation seems to fortuitously match in one case.
March 2007 On the Vickers Indentation Fracture Toughness Test 677
and Miyoshis equation
22
as used in Japanese Industrial
Standard (JIS) R 1607.
30
These equations have been discussed
in detail by Quinn.
7
The latter two equations were deve-
loped and calibration constants adjusted with silicon nitride
specifically in mind. The results of the VIF measurements
are summarized in Tables I and II and are presented in
Figs. 4 and 5 along with the certied value for K
Ic
. There are
small differences in the VIF results for the two test pieces at low
indentation loads. These diminish to less than 5% at 49 and
98 N.
It is evident from the results in the Tables and Fig. 4 that none
of these three popular equations consistently yields the correct,
certied value of fracture toughness of 4.57 MPa m
1/2
for SRM
2100. The VIF results from Anstis et al.s and Niiharas equa-
tions are incorrect at all loads. In fact, the standard value does
not even fall within their error bars. Results from Miyoshis
equation do not match when the indentations were made in the 4
mm25 mm specimen surface (Table I and Fig. 4(a)) and the
data do not fall within the error bars. The data do match for the
other case (Table II and Fig. 4(b)). All three equations suggest a
slightly rising R-curve with about a 10% increase of the average
fracture resistance over the range of indentation loads that were
applied in the testing.
Miyoshis equation produces a match in one instance, but the
agreement should not come as a surprise. Miyoshis VIF equa-
tion calibration factor was empirically force tted to match valid
data from other methods for a ne-grained equiaxed silicon ni-
tride.
22
How well does Miyoshis equation fare with sintered a silicon
carbide? This homogenous ne-grained material was the back
up material for the NIST fracture toughness SRM. It has a at
R-curve and no sensitivity to slow crack growth at room tem-
perature. Ghosh et al.s
26
original data were re-analyzed using
Miyoshis equation. The results are shown in Fig. 5. Analyses
and testing
28,29,35
of the sintered a silicon carbide have rened
the K
Ic
values to 2.6 MPa m
1/2
to 2.8 MPa m
1/2
depending
upon plate density. Figure 5 shows that Miyoshis equation
substantially overestimates the actual fracture toughness
of the silicon carbide by nearly 20%. The standard test K
Ic
values do not even fall within the VIF error bars. The other two
equations do not t the standard K
Ic
values very well either at
2535 N.
VIII. Discussion
One primary conclusion that must be drawn from all of the
earlier studies
2326
is that even if one equation does match ex-
pected fracture toughness results for a particular material, it
often does not produce reliable results for another material. The
outcomes in the present study conrm this beyond any doubt.
Ponton and Rawlings analyses
24,25
for 16 materials devoted
much energy and space seeking the one equation that gave the
best t to the available data, but their search was fruitless. None
of the numerous equations were satisfactory even for simple
ranking of materials between classes. It should be clear that
there never will be a universal VIF equation that will be reliable
for all brittle materials.
Is it surprising that the VIF technique is incapable of con-
sistently producing results that yield the correct results? Not
really, for the previous technical discussions and the aforemen-
tioned publications
2326
indicate that the VIF method is dis-
tinctly different than a fracture toughness test. The VIF does not
produce a reliable K
Ic
value. The crack initiation and propaga-
tion in the VIF test is not the same as the sequence of crack
processes in the standardized fracture toughness tests. Why
should the VIF technique be expected to yield the correct frac-
ture toughness value?
The toughness versus indentation test load curves presented
in Fig. 4 closely resemble similar curves in the four earlier pub-
lications
2326
of over 10 years ago that were discussed previous-
ly. These recent results simply reconrm the conclusions of those
archival papers regarding the VIF technique for fracture tough-
ness measurements. Those conclusions are that the VIF is not a
suitable fracture toughness measurement method. The VIF
should not be used for fracture toughness testing, even if just
for a comparative basis. The VIF does not provide legitimate
toughness values.
IX. A Word of Caution about Nanoindentation Fracture
Toughnesses
Recently there has been an escalating interest in applying some
of the VIF technique concepts to the nanomeasurement of the
fracture toughness of materials by measuring the crack lengths
emanating from the three corners of sharp nanoindenters.
43,44
Table II. Vickers Indentation Fracture Resistance Results for a Standard Reference Material (SRM) 2100 Silicon Nitride Test Piece
with Indentations in a 3 mm25 mm Face
Indentation
load (N)
Average diagonal
half length a (mm)
Average crack half
length c (mm)
SRM 2100
w
certied K
Ic
(MPa m
1/2
)
Anstis
z
K
c
(MPa m
1/2
)
Niihara
y
K
c
(MPa m
1/2
)
Miyoshi
z
K
c
(MPa m
1/2
)
9.8 16.770.3 30.970.5 4.57 3.8970.13 5.7870.18 4.5570.15
19.6 23.770.3 49.670.9 4.57 3.8470.10 5.7070.14 4.4970.12
49 37.970.4 91.771.4 4.57 3.8670.10 5.7270.15 4.5170.12
98 53.670.4 145.673.7 4.57 3.8770.15 5.7370.22 4.5270.17
196 76.570.7 222.076.4 4.57 4.1570.18 6.1370.27 4.8570.21
w
The uncertainty of SRM 2100 is 0.11 MPa
.
m
1/2
for the average of ve test outcomes at the 95% condence level. At a 1 standard deviation condence level, the
uncertainty would be 70.06 MPaOm.
z
Anstis et al. equation.
18 y
Niiharas median crack equation,
20,21
with a constant value of 10.4 as per ref.
42 z
Miyoshis equation
22
as
used in JIS R 1607.
30
Uncertainties are 1 standard deviation unless otherwise noted. The notes are the same as for Table I.
Table I. Vickers Indentation Fracture Resistance Results for a Standard Reference Material (SRM) 2100 Silicon Nitride Test Piece
with Indentations in a 4 mm25 mm Face
Indentation
load (N)
Average diagonal
half length a (mm)
Average crack
half length c (mm)
SRM 2100
w
certied
K
Ic
(MPa m
1/2
)
Anstis
z
K
c
(MPa m
1/2
)
Niihara
y
K
c
(MPa m
1/2
)
Miyoshi
z
K
c
(MPa m
1/2
)
9.8 16.270.3 33.071.4 4.57 3.4470.20 5.1470.30 4.0270.23
19.6 23.570.1 52.071.0 4.57 3.5670.11 5.2970.16 4.1670.13
49 37.570.2 94.071.5 4.57 3.6970.09 5.4770.13 4.3170.10
98 53.370.3 148.971.9 4.57 3.7270.08 5.5170.11 4.3470.09
w
The uncertainty of SRM 2100 is 0.11 MPa
.
m
1/2
for the average of 5 test outcomes at the 95% condence level. At a 1 standard deviation condence level, the uncertainty
would be 70.06 MPa
.
m
1/2
.
z
Anstis et al. equation.
18 y
Niiharas median crack equation,
20,21
with a constant value of 10.4 as per ref.
42 z
Miyoshis equation
22
as used in JIS R
1607.
30
Uncertainties are 1 standard deviation unless otherwise noted. K
c
designates the apparent fracture toughness using the equation from the listed source.
678 Journal of the American Ceramic SocietyQuinn and Bradt Vol. 90, No. 3
Interest has been particularly high for characterizing thin
lms.
45
It can be called the NIF test technique. The triangular
Berkovich nanoindenter geometry and that of cube corner in-
denters is even more sharply pointed than the Vickers diamond
pyramid and they readily produce cracks. These sharp three-
sided indenters may be expected to create extensive impression
corner cracking at even lower comparable indentation test loads
than a Vickers indenter. They will also create complex multiple
crack congurations and residual stresses.
In view of the above collective observations and conclusions
about the VIF technique for fracture toughness testing, one
must consider the nanoindentation fracture toughnesses, or the
NIF technique, with similar reservations as have been expressed
for the VIF determined toughness values. In other words, so
long as there is an obvious lack of a definite formal solution for
the stress intensity factor at the crack tip and the applicable
equations are no more than dimensional analyses with calibra-
tion factors, the validity of any results of nanoindentation frac-
ture toughness measurements must be considered highly suspect.
Nanoindentation or NIF toughnesses should be seriously ques-
tioned and probably should never be used, even with extreme
caution. Similar to the VIF toughnesses, NIF toughnesses are
unlikely to be accurate or valid, except fortuitously. There clear-
ly is a need to develop fracture toughness test methods at the
very small scale that have the same accuracy and precision as is
possible for micro- and macroscale tests.
X. Summary and Conclusions
The VIF for the determination of fracture resistance has been
reviewed. Its history and origin have been discussed along with
the numerous equations which different authors have proposed
to apply to the VIF test. The cracking processes have been de-
scribed and were contrasted with those of standardized fracture
toughness tests with reference to the basic definition of fracture
toughness. Extensive compilations of VIF test results previously
summarized in the literature were also reviewed and considered.
Finally, additional VIF experiments were performed on a Stand-
ard Reference Material to establish whether the VIF test had the
ability to determine a valid K
Ic
. One equation produced results
that came close, but only because it had an empirical calibration
constant force tted to a similar silicon nitride. The same equa-
tion did not procedure accurate results for sintered a silicon
carbide, another well-characterized ceramic. It undoubtedly fails
for many others as well.
What then does the VIF test actually do or measure? It does
not appear to correctly measure any crack propagation param-
eter. The VIF does appear to represent some form of a complex
crack arrest phenomenon. That crack arrest process occurs in
the presence of a multiple-cracked material environment and in
a highly complex residual stress condition, often accompanied
by extensive spalling of material from around the indentation.
These features create an incredibly complex crack tip arrest en-
vironment, one that is impossible to describe. The crack arrest
situation of the VIF dees meaningful description.
It must be concluded that the VIF technique is not suitable
for the measurement of the fracture toughness, K
Ic
, or any other
form of the fracture resistance of ceramics or other brittle ma-
terials. Any agreement with accepted fracture toughness, K
Ic
data is either fortuitous or due to force tting the VIF calibra-
tion constants. It is recommended that use of the VIF be dis-
continued for all ceramic materials. It is not a reliable fracture
test method. It should not be applied to or be acceptable for any
basic fracture resistance measurements of ceramics or any other
materials. It should never be used for determination of the
property K
Ic
in materials specications.
Acknowledgments
The rst author (G. D. Q.) thanks his colleagues who have so painstakingly
helped rene fracture toughness test methods and help prepare the NIST SRM
2100 including R. Gettings, K. Xu, Dr. J. Swab, Dr. J. Salem, Prof. M. Jenkins,
and Prof. I. Bar-On. We also thank M. Sakai, Z. Li, A. Kobayashi, A. Ghosh, C.
Henager, and K. W. White for consultations and advice.
References
1
S. Palmqvist, Method att Bestamma Segheten hos Spread Material, Sarskilt
Hardmetaller, Jernkortorets Ann., 141, 3007 (1957).
2
A. G. Evans and E. A. Charles, Fracture Toughness Determinations by In-
dentation, J. Am. Ceram. Soc., 59 [7-8] 3712 (1976).
3
T. Fujii and T. Nose, Evaluation of Fracture Toughness for Ceramic Mate-
rials, IJIS Int., 29 [9] 71725 (1989).
4
M. Sakai and R. C. Bradt, Fracture Toughness Testing of Brittle Materials,
Int. Mater. Rev., 38 [2] 5378 (1993).
5
R. Morrell, Fracture Toughness Testing for Advanced Technical Ceramics:
Internationally Agreed Good Practice, Adv. Appl. Ceram., 105 [2] 111 (2006).
6
G. D. Quinn, The Fracture Toughness Round Robins in VAMAS: What We
Have Learned; pp. 10726 in Fracture Resistance Testing of Monolithic and Com-
posite Brittle Materials, ASTM STP 1409, Edited by J. A. Salem, G. D. Quinn,
and M. G. Jenkins. ASTM, West Conshohocken, PA, 2002.
7
G. D. Quinn, Fracture Toughness of Ceramics by the Vickers Indentation
Crack Length Method: A Critical Review, Ceram. Eng. Sci. Proc., 27 [3] (2006).
8
B. R. Lawn and M. V. Swain, Microfracture Beneath Point Indentations in
Brittle Solids, J. Mater. Sci., 10, 11322 (1975).
9
B. R. Lawn and E. R. Fuller, Equilibrium Penny-like Cracks in Indentation
Fracture, J. Mater. Sci., 10, 201624 (1975).
10
A. G. Evans and T. R. Wilshaw, Quasi-Static Solid Particle Damage in Brit-
tle Solids, I. Observations, Analysis and Implications, Acta Met., 24, 93956
(1976).
11
B. R. Lawn and A. G. Evans, A Model for Crack Initiation in Elastic/Plastic
Indentation Fields, J. Mater. Sci., 12, 21959 (1977).
12
B. R. Lawn and D. B. Marshall, Hardness, Toughness, and Brittleness: An
Indentation Analysis, J. Am. Ceram. Soc., 62 [78] 34750 (1979).
13
B. R. Lawn, A. G. Evans, and D. Marshall, Elastic/Plastic Indentation
Damage in Ceramics: The Median/Radial Crack System, J. Am. Ceram. Soc., 63
[9-10] 57481 (1980).
14
J. T. Hagan and M. V. Swain, The Origin of Median and Lateral Cracks
Around Plastic Indents in Brittle Materials, J. Phys. D., 11, 2091102 (1978).
15
R. F. Cook and G. M. Pharr, Direct Observation and Analysis of Indenta-
tion Cracking in Glasses and Ceramics, J. Am. Ceram. Soc., 73 [4] 787817
(1990).
16
T. E. Wilantewicz, Vickers Indentation Fracture in Optical Glass Composi-
tion, to be publ., in Fractography of Glasses and Ceramics, Vol. 5, Edited by G. D.
Quinn, J. R. Varner, and M. Wightman. American Ceramic Society, Westerville,
OH, 2007.
17
D. B. Marshall and A. G. Evans, Reply to Comment on Elastic/Plastic In-
dentation Damage in Ceramics: The Median/Radial Crack System, Comm. Am.
Cer. Soc., 64 [12 C] 1823 (1981).
Fig. 5. Fracture toughness of sintered alpha silicon carbide using the
data from Ghosh et al.
26
The average K
Ic
from all standard test methods
is very consistent and falls between 2.6 and 2.8 MPa m
1/2
as shown by
the band. Note that the Vickers indentation fracture data with Miyoshis
equation falls well outside the correct range.
March 2007 On the Vickers Indentation Fracture Toughness Test 679
18
G. R. Anstis, P. Chantikul, B. R. Lawn, and D. B. Marshall, A Critical
Evaluation of Indentation Techniques for Measuring Fracture Toughness: I, Dir-
ect Crack Measurements, J. Am. Ceram. Soc., 64 [9] 5338 (1981).
19
P. Chantikul, G. R. Anstis, B. R. Lawn, and D. B. Marshall, A Critical
Evaluation of Indentation Techniques for Measuring Fracture Toughness: II,
Strength Methods, J. Am. Ceram. Soc., 64 [9] 53943 (1981).
20
K. Niihara, Indentation Fracture Toughness of Brittle Materials for Palmq-
vist Cracks; pp. 97105 in Fracture Mechanics of Ceramics, Vol. 5, Edited by
R. C. Bradt., et al. Plenum Publishing Co., NewYork, 1983.
21
K. Niihara, A Fracture Mechanics Analysis of Indentation-Induced Palmq-
vist Cracks in Ceramics, J. Mater. Sci. Lett., 2, 2213 (1983).
22
T. Miyoshi, A Study on Evaluation of K
Ic
for Structural Ceramics, Trans.
Jap. Soc. Mech. Eng., Ser. A., 51A, 248997 (1985).
23
Z. Li, A. Ghosh, A. S. Kobayashi, and R. C. Bradt, Indentation Fracture
Toughness of Sintered Silicon Carbide in the Palmqvist Crack Regime, J. Am.
Ceram. Soc., 72 [6] 90411 (1989).
24
C. B. Ponton and R. D. Rawlings, Vickers Indentation Fracture Test. Part 1:
Review of Literature and Formulation of Standardised Indentation Toughness
Equations, Mater. Sci. Tech., 5 [9] 86572 (1989).
25
C. B. Ponton and R. D. Rawlings, Indentation Fracture Toughness Test.
Part 2, Application and Critical Evaluation of Standardised Indentation Tough-
ness Equations, Mater. Sci. Tech., 5 [10] 96176 (1989).
26
A. Ghosh, Z. Li, C. H. Henager, A. S. Kobayashi, and R. C. Bradt, Vickers
Microindentation Toughness of a Sintered SiC in the Median Crack Regime; pp.
21932 in Fracture Mechanics of Ceramics, Vol. 12, Edited by R. C. Bradt., et al.
Plenum Publishing Co., NewYork, 1996.
27
Munz D. and T. Fett, Ceramics: Mechanical Properties, Failure Behavior, and
Materials Selection, pp. 20. Springer Verlag, NewYork, 1999.
28
G. D. Quinn and J. A. Salem, Effect of Lateral Cracks Upon Fracture
Toughness Determined by the Surface Crack in Flexure Method, J. Am. Ceram.
Soc., 85 [4] 87380 (2002).
29
ASTM C 1421-99 Standard Test Methods for the Determination of Fracture
Toughness of Advanced Ceramics at Ambient Temperature. Annual Book Standards,
Vol. 15.01. ASTM Int., West Conshohocken, PA, 1999.
30
JIS R 1607 Testing Methods for Fracture Toughness of High Performance Cer-
amics. Japanese Standards Association, Tokyo, 1990.
31
CEN TS 14425 Technical Specication, Advanced Technical CeramicsTest
Methods for Determination of Fracture Toughness of Monolithic Ceramics, Parts 1
5. European Committee for Standardization, Brussels, 2003.
32
ISO 15732 Fine Ceramics (Advanced Ceramics, Advanced Technical Ceram-
ics)Test Method for Fracture Toughness at Room Temperature by Single Edge
Precracked Beam (SEPB) Method. International Organization for Standards,
Geneva, 2003.
33
ISO 18756 Fine Ceramics (Advanced Ceramics, Advanced Technical Ceram-
ics)Determination of Fracture Toughness of Monolithic Ceramics at Room Tem-
perature by Surface Crack in Flexure (SCF) Method. International Organization
for Standards, Geneva, 2003.
34
ISO 24370 Fine Ceramics (Advanced Ceramics, Advanced Technical Ceram-
ics)Test Method for Fracture Toughness of Monolithic Ceramics at Room Tem-
perature by Chevron Notched Beam (CNB) Method. International Organization
for Standards, Geneva, 2005.
35
M. Jenkins, G. D. Quinn, J. A. Salem, and I. Bar-On, Development, Ver-
ication and Implementation of a National Full-Consensus Fracture Toughness
Test Method Standard for Advanced Ceramics; pp. 4975 in Fracture Resistance
Testing of Monolithic and Composite Brittle Materials, ASTM STP 1409, Edited
by J. A. Salem, M. G. Jenkins, and G. D. Quinn. ASTM, West Conshohocken,
PA, 2002.
36
D. Bleise and R. W. Steinbrech, Flat R-Curve from Stable Propagation of
Indentation Cracks in Coarse-Grained Alumina, J. Am. Ceram. Soc., 77 [2] 315
22 (1994).
37
T. A. Michalske and J. M. Collins, Fractographic Determination of Crack-
Tip Stress Intensity, pp. 22939 in Fractography of Glasses and Ceramics, eds.
J. R. Varner and V. D. Frechette. American Ceramic Society, Westerville, OH
(1988).
38
SRM 2100 Fracture Toughness of Ceramics, Standard Reference Materials
Ofce. NIST, Gaithersburg, MD, 1999.
39
G. D. Quinn, K. Xu, J. A. Salem, and J. J. Swab, SRM 2100: The Worlds
First Fracture Toughness Reference Material; pp. 499530 in Fracture Mechanics
of Glasses and Ceramics, Vol. 14, Edited by R. C. Bradt, D. Munz, M. Sakai, and
K. W. White. Klewer/Plenum, NewYork, 2005.
40
G. D. Quinn, K. Xu, R. J. Gettings, J. A. Salem, and J. J. Swab, Does Any-
one Know the Real Fracture Toughness? SRM 2100: The Worlds First Ceramic
Fracture Toughness Reference Material; pp. 7693 in Fracture Resistance Testing
of Monolithic and Composite Brittle Materials, ASTM STP 1409, Edited by J. A.
Salem, G. D. Quinn, and M. G. Jenkins. ASTM, West Conshohocken, PA, 2002.
41
SRM 2831 Vickers Hardness of Ceramics and Hardmetals. Standard Reference
Materials Ofce, NIST, Gaithersburg, MD, 2001.
42
ASTM F 2094-01 Standard Specication for Silicon Nitride Bearing Balls; in
Annual Book of Standards., Vol. 1.08. ASTM Int., West Conshohocken, PA, 2003.
43
D. J. Morris, S. G. Myers, and R. F. Cook, Sharp Probes of Varying Acuity;
Instrumented Nanoindentation and Fracture Behavior, J. MaterRes., 19 [1] 165
75 (2004).
44
J. M. Jungk, B. L. Boyce, T. E. Buchheit, T. A. Friedmann, D. Yang, and W.
W. Gerberich, Indentation Fracture Toughness and Acoustic Energy Release in
Tetrahedral Amorphous Carbon Diamond-like Thin Films, Acta Mater., 54 [5]
404352 (2006).
45
M. D. Michel, L. V. Muhlen, C. A. Achete, and C. M. Lepienski, Fracture
Toughness, Hardness and Elastic Modulus of Hydrogenated Amorphous Carbon
Films Deposited by Chemical Vapor Deposition, Thin Solid Films, 496, 4818
(2006). &
Richard Dick Bradt is a profes-
sor of materials engineering at the
University of Alabama, Tusca-
loosa, AL, USA, where he is the
Alton N. Scott Professor of En-
gineering. Dick is a graduate of
MIT and RPI and has previously
held faculty positions at Penn
State University, the University
of Nevada and the University of
Washington, the latter where
he held the endowed Kyocera
Professorship. Prof. Bradt has directed over 100 graduate
theses and coauthored nearly 400 papers with students and
colleagues. More than 100 of these manuscripts have appeared
in the Journal and Bulletin of the American Ceramic Society.
Dr. Bradt has served as coeditor of over 20 proceedings of
international conferences that primarily address fracture and
the mechanical properties of ceramic and glass materials. Prof.
Bradt is a Fellow of the American Ceramic Society and also
ASM-International, as well as Distinguished Life Member of
UNITECR, the international refractories organization.
George Quinn is a Ceramic En-
gineer in the Ceramics Division at
the National Institute of Stan-
dards and Technology. A gradu-
ate of Northeastern University in
Boston, Mr. Quinn worked for 21
years at the U.S. Army Materials
Research Laboratory in Water-
town, MA, before coming to
NIST in 1990. He spent 1 year in
1987 as an exchange scientist at
the German Aerospace Research
Establishment in Cologne. He works on mechanical testing,
reliability analysis, and fractography, with emphases on re-
nements to and standardization of test methods. He is a
master fractographer and has analyzed many commercial,
military, and research fracture problems. He has led many
national and international standardization programs includ-
ing six successful international round robin projects on ex-
ural strength, fracture toughness, hardness, and fractography.
He has authored or coauthored 18 standards and prepared
three standard reference materials. He is a fellow of the
American Ceramic Society and also the American Society
for Testing and Materials.
680 Journal of the American Ceramic SocietyQuinn and Bradt Vol. 90, No. 3