DentMaterll:ill-112,
March,1995
Fracture mechanics principles
John J. Mecholsky,
Department of Materials Science and Engineering,
ABSTRACT
The principles of linear elastic fracture mechanics (LEFM) were devel-
oped in the 1950s by George Irwin (1957). This work was based on
previous investigations of Griffith (1920) and Orowan (1944). Irwin
(1957) demonstrated that a crack shape in a particular location with
respect to the loading geometry had a stress intensity associated with
it. He also demonstrated the equivalence between the stress intensity
concept and the familiar Griffith criterion of failure. More importantly, he
described the systematic and controlled evaluation of the toughness of
a material. Toughness is defined as the resistance of a material to rapid
crack propagation and can be characterized by one parameter, K,,. In
contrast, the strength of a material is dependent on the size of the
initiating crack present in that particular sample or component. The
fracture toughness of a material is generally independent of the size of
the initiating crack. The strength of any product is limited by the size of
the cracks or defects during processing, production and handling. Thus,
the application of fracture mechanics principles to dental biomaterials is
invaluable in new material development, production control and failure
analysis. This paper describes the most useful equations of fracture
mechanics to be used in the failure analysis of dental biomaterials.
THEORETICAL BACKGROUND
Complex loading on pre-existing cracks can be resolved
into three modes. If the loading is tensile, it is referred to
as Mode I; if it is in-plane shear, Mode II; and out-of-plane
shear (or torsion), is referred to as Mode III (Jayatilaka,
1979). Most brittle materials fail in Mode I. If two materi-
als arejoined, but still fail primarily in a tensile mode, then
the original crack exhibits a mixed mode condition com-
bining Modes I & II. Most of the cases with which we will
be concerned will be Mode I. The Mode I stress intensity,
K,, of a crack in a material can be determined, in general,
using the relationship: _ Q
I.IJlt
K, =[-]o&
@ local plastic deformation at the crack tip. An estimate of
K, =[Y(O)]o& (lb)
where Y(O) is a geometrical factor which accounts for the
location and geometry of the crack and loading (Randall,
19671 and 0 is the angle from the surface to a point
anywhere along the crack periphery, @ is a flaw shape
parameter,0 is the applied stress, a is the semi-minor axis,
and b the semi-major axis of an elliptical crack. K1in this
case is a measure of the effect of loading on a crack of a
given geometry. As the load increases, the stress intensity
increases. Irwin (1957) defined failure at the point when
K1reaches a critical value, i.e., fracture occurs when:
K, 2 K,C = [Y(O)] o&i (2)
where K,, is in many cases a material constant and is
Jr.
University of Florida, Gainesville, Florida, USA
considered a measure of the toughness of the material.
Irwin (1957) further recommended that standardized tests
beperformed which can measure the critical stress inten-
sity factor for different materials. From this work, and
through the work of many other investigators (ASTM,
1984a; Benchmark Committee, 1989), a compilation of
several standardized tests for determining K,, values for
metals are available from ASTM (198413).Toughness can
be measured by either the critical stress intensity factor,
Krc, fracture energy, y,, or critical strain energy release
rate, G, (G, = 2~). Note that K,, is measured in units of
stress-(distance)ln, e.g., MPa rnln, and yc and G, are
??
measured in energy/unit area, e.g., J/m”.
For the purposes of this paper in the application of
fracture mechanics, we will only consider brittle materi-
als, such as ceramics, glass, glass ceramics, and brittle
polymers, and materials with limited ductility, i.e., local
plastic deformation at or near the crack tip, such as metals
and some polymers. Global yielding is governed by plastic
deformation and is limited by the yield stress.
For small elliptical surface cracks in brittle materials,
i.e., cracks whose depths are less than -0.25 times the
thickness and with no local residual stress, Eqs. 1 and 2 can
be reduced to:
K,, =1.240&i (3)
where @ = 1.57 (from Eq. 1) for a semi-circular crack and
c is the radius of an equivalent semi-circular crack for a
semi-elliptical crack of semi-minor axis a and semi-major
axis b. For most applications offailure in brittle materials,
this is an extremely useful approximation.
For materials with limited ductility, Eqs. 1 and 2 have
to be modified: 11 I-Z
K,f = .!+!!&&
where Q =@’ - 0.212 [o/oy I2and Keflisthe value ofK,c with
the effect of local plasticity can be obtained from an
evaluation of the second term in Q with respect to (D*.The
value of @varies from 1 to 1.57 for cracks with aspect ratios
(a/b) ranging from 0 to 1, respectively. Obviously, the yield
stress for the material, oy, has to be known or determined
for this analysis.
Toughening mechanisms that increase the resistance
to crack growth with crack size do not necessarily invali-
date the test, but rather provide a multi-valued property
(Broek, 1986). These latter materials have been termed “R-
curve” or “T-curve”materials (Cook et al., 1985). It appears
that this type of behavior is common in tougher materials.
The name “R-curve” does not identify the toughening
mechanism. Thus, if brittle or ductile materials exhibit
Dental Materials/March 1995 111
R-curve behavior, then the mechanism for this behavior
should be identified to take advantage of any further
toughening.
TESTING PROCEDURES
Strength alone will not provide sufficient information to
decide whether or not a treatment process has enhanced
the resistance to fast fracture. Since strength is depen-
dent on crack size, it can vary with the handling proce-
dure, finishing procedure or with random processing
flaws. Various strength tests are used in the literature.
These include the diametral compression test (Shetty et
al., 1987), the flexural beam (Rudnichet al., 1968) and the
tensile specimen (Hayden et al., 1965). In most tests, the
crack or flaw size at the fracture origin is not controlled or
measured. The results of these tests are subjected to
statistical scatter due to the distribution of crack sizes.
Toughness tests, on the other hand, are generally
independent of crack size, except for the case of R-curve
materials. The measurement of toughness in brittle mate-
rials has been discussed extensively (Freiman, 1979). All
of the tests used for determination of toughness are based
on Eq. 1 combined with Eqs. 2 or 4. In the case of R-curve
materials, many investigators list the large crackvalue as
the toughness of the material. However, in reality, the
entire curve is a measure of the fracture resistance. The
crack size observed in components is the guide to deter-
mine the appropriate value of toughness to use in design
calculations. Experimental techniques used in the deter-
mination of toughness can be divided into large and small
crack techniques. The large crack techniques can be
classified as plate or beam type tests. For example, many
large crack techniques use the double cantilever beam
(DCB) analysis for determination of fracture toughness.
DCB type tests include compact tension (Anderson, 19911,
constant moment DCB (Freiman et al., 19731, and the
chevron notch short bar tests (Barker, 1983). Large crack
beam tests include the single edge or double edge notched
beams loaded in either flexure or tension and the chevron
notch flexure beam test. Small crack techniques are
usually based on flexure tests on beams either with
controlled cracks introduced by indentation (Chantikul et
al., 1981) or by fracture surface analysis techniques ap-
plied to “natural” cracks (Mecholsky, 1993).
Application of Linear Elastic Fracture Mechanics. De-
termination of the failure origin is essential in the appli-
cation of linear elastic fracture mechanics (LEFM). The
techniques for determining the initial point of failure are
the subject of the area of fracture surface analysis and are
discussed in the following paper (Mecholsky, 1995). The
principles of LEFM discussed above can be used in the
calculation of strengths for failed dental materials if the
appropriate material constants, i.e., KIc or G,, have been
previously determined. If the stress at failure can be
determined, then the value of the effective fracture tough-
ness can be determined through the use of Eqs. 3 or 4. The
value obtained can then be compared to published or
independently obtained values for the material so that a
determination can be made as to the proper fabrication of
the part. The size of the crack relative to the microstruc-
tural features, i.e., grain size, glassy phase, pores, etc., is
critical to the effective value of the toughness. The local
112 Mecholsky/Fracture mechanics principles
behavior controls the apparent global behavior of the
material. Thus, it is critical to material development that
the fracture toughness be properly obtained. To character-
ize fracture toughness completely, both large and small
crack techniques should be used. The test that matches
the conditions of the particular application will prove to be
the most reliable in predicting behavior.
Presented at the 1994 ADM annual meeting, Bermuda, November 1994.
See following paper for corresponding author’s address.
REFERENCES
Anderson TL (1991). Fracture mechanics. 2nd ed. Orlando:
CRC Press, 709-718.
ASTM (1984a). Plane strain fracture toughness of metallic
materials. ASTM Standards, 519-553.
ASTM (198413).Fracture testing. ASTM Standards, 674.
Barker LM (1983). Compliance calibration of a family of short
rod and short bar fracture toughness specimens. Eng
Fract Mech 17:289-312.
Benchmark Editorial Corn. of the SESA Fracture Committee
(1989). A critical evaluation of numerical solutions to the
benchmark surface flaw problem. Exp Mech: 253-264.
Broek D (1986). Elementary Engineering Fracture Mechan-
ics. 4th ed. Boston: Martinus-Nijhoff, 130-140.
Chantikul P, Anstis GR, Lawn BR, Marshall DB (1981). A
critical evaluation of indentation techniques for measur-
ingfracture toughness: II. Strength method. JAm Cer Sot
64:539-43.
Cook RF, Lawn BR, Fairbanks CJ (1985). Microstructures-
strength properties in ceramics: I. Effect of crack size on
toughness. J Am Cer Sot 68:604-15.
Freiman SW (1979). Fracture mechanics applied to brittle
materials. In: ASTM STP 678. Phila: Am Sot Test Mater.
Freiman SW, Mulville DR, Mast PW (1973). Crack propaga-
tion studies in brittle materials. JMuter Sci 8:1527-1533.
Griffith AA (1920). The phenomena of rupture and flow in
solids. Phil Trans Series A 221:163-198.
Hayden HW, Moffat SG, Wulff J (1965). The Structure and
Properties of Materials III. New York: J. Wiley, l-9.
Irwin GR (1957)Analysis of stresses and strains near the end
of a crack transversing a plate. J Appl Mech 24:361-64.
Jayatilaka AS (1979). Fracture of engineering brittle mate-
rials. Applied Sciences Pub., 80-113.
Mecholsky JJ (1993). Quantitative fracture surface analysis
of glass materials. In: Simmons CJ, El-Bayoumi 0, Eds.
Experimental Techniques of Glass Science. Westerville:
Am Cer Sot, 483-520.
Mecholsky JJ (1995). Fractography: Determining the sites of
fracture initiation. Dent Mater 11:113-116.
Orowan E (1944). The fatigue of glass under stress. Nature
154:341-343.
Randall PN (1967). Plain strain crack toughness testing of
high strength metallic materials. In: ASTM STP 410.
Phila: Am Sot Test Mater, 88-126.
Rudnich A, Marshall CW, Duckworth WH, Emerick BR
(1968). The evaluation and interpretation of mechanical
properties of brittle materials. Defense Information Cen-
ter Rpt. 68-3 AF’ML-TR-67-361.
Shetty DK, Rosenfeld AR, Duckworth WH (1987). Mixed-
mode fracture in biaxial stress state: Application of the
diametral compression test. Eng Fract Mech 26:825-40.