MNL41-EB/Nov.

1999

Stress Analysis for Members

with Cracks--K

2.1 Introduction
FRACTURE MECHANICSis a method of characterizing the fracture and fatigue be-
havior of sharply notched structural members (cracked or flawed) in terms that
can be used by the engineer, namely, stress (o-) and flaw size (a).
This chapter describes the stress analysis procedures for structural members
with cracks. The difference between stress-concentration factors (used to analyze
stress at a point in the vicinity of well-defined notches) and stress-intensity factors
(used to analyze the stress field ahead of a sharp crack) is first described. Then,
the stress analysis of members with sharp cracks is introduced. The complete
stress analysis involves complex variables and other forms of higher mathematics
and may be found in various references [1-7]. Emphasis in this book is on the
applications and use of fracture mechanics.
Stress-intensity factors (Ki) are presented for most of the cracks commonly
found in structures. Stress-intensity factors for other crack geometries can be
found in various handbooks [8-10]. In fatigue, where the driving force is &K =
KMAX -- KMIN, the same KI relations described in this chapter also are used.
The stress-intensity factor is a mathematical calculation relating the applied
load and crack size for a particular geometry. The calculation of K~ is analogous
to the calculation of applied stress, or, in an unflawed member. To prevent yield-
ing, the engineer keeps the applied stress, o-, below the material yield strength,
Crys. In an unflawed member, o- is the "driving force," and r (the yield strength)
is the "resistance force.'" The driving force is a calculated quantity while the
resistance force is a measured value. In the same sense, KI is a calculated "driving
force" and Kc (described in Chapter 3) is a measured fracture toughness value
and represents the "resistance force" to crack extension. To prevent brittle frac-
ture, the engineer keeps the calculated applied stress intensity factor, K~, below
28
Copyright9 1999 by ASTM International www.astm.org
 Stress Analysis for Members with Cracks--K~ 29

the measured fracture toughness value, K c, in the same manner that o- is kept
below O'ys to prevent yielding.
This chapter describes the calculation of K, for various crack geometries.
Chapter 3 describes the measurement of fracture toughness values, Kc, under
various conditions of loading and constraint.

2.2 Stress-Concentration Factormkt

Most structural members have discontinuities of some type, for example, holes,
fillets, notches, etc. If these discontinuities have well-defined geometries, it is
usually possible to determine a stress-concentration factor, k t for these geometries
[11]. Then, the engineer can account for the local elevation of stress using the
well-known relation between the local maximum stress and the applied nominal
stress:
(rma x = k t 9 O'nom (2.1)
In fact, for most structures, the engineer usually relies on the ductility of the
material to redistribute the load around a mild stress concentration. The effects
of mild stress concentrations (holes, smooth fillets, etc.) in ductile materials are
addressed in various codes and standards. However, if the stress concentration
is severe, for example, approaching a sharp crack in which the radius of the
crack tip approaches zero, the use of fracture mechanics becomes necessary to
analyze structural performance.
To illustrate this point, let us analyze the stress concentration at the edge of
an ellipse as shown in Figure 2.1. This is given as:

"~ om (2.2)

Rearranging Equation 2.2 gives o-.... for an ellipse,

O'max = O'nom (1 q- 2 b ) (2.3)

For very sharp cracks,

As shown in Figure 2.1, kt becomes very large as a / b becomes large. The radius
at the end of the major axis can be approximated by p = b2/a. For sharp cracks,

a >> 19, O'max = O-nom " 2 ~ (2.5)

and
30 FRACTURE AND FATIGUE CONTROL IN STRUCTURES

O"

llttl t t

= Kto-

o-

I5 I I 1 1 I I

_JZ
7-
O__I--

n.'F-

TO
5-
0
1
o.I 0.5 1.0 2.0:5.04.0 6.0
(a/b)
FIG. 2.1 Stress-concentration factor for an elliptical hole,

O'ma x
kt - = 2/~ (2.6)
O'no m ~P
For sharp cracks, p ---* 0 and k t ~ ~ . Thus, the use of the stress-concentration
approach becomes meaningless. Consequently, an analytical method different
from the stress-concentration approach is needed to analyze the behavior of
structural or machine components that contain cracks or sharp imperfections.
The first analysis of fracture behavior for components containing cracks was
developed by Griffith [1] as described in the appendix to this chapter. Presently,
the fracture behavior for such components can be analyzed best by using frac-
ture-mechanics technology.

2.3 Stress-Intensity Factor, K~

Linear-elastic fracture-mechanics technology is based on an analytical procedure.

This procedure relates the stress-field magnitude and distribution in the vicinity
 Stress Analysis for Members with Cracks--K~ 31

of a crack tip, the nominal stress applied to a test specimen or a structural m e m -

ber, and the size, shape, and orientation of a crack or crack-like discontinuity.
The f u n d a m e n t a l principle of fracture mechanics is that the stress field ahead
of a sharp crack in a test specimen or a structural m e m b e r can be characterized
as a single parameter, K, w h i c h is the stress-intensity factor. This parameter, K,
is related to b o t h the nominal stress level (o-) in the m e m b e r a n d the size of the
crack (a), and has units of ksiX/~n. (MPaX/mm). Thus, all structural m e m b e r s or
test specimens that have flaws can be l o a d e d to various levels of K. This is
analogous to the situation w h e r e unflawed structural or mechanical m e m b e r s
can be l o a d e d to various levels of stress, ~r.
To establish m e t h o d s of stress analysis for cracks in elastic solids, it is con-
venient to define three types of relative m o v e m e n t s of two crack surfaces [5].
These displacement m o d e s (Figure 2.2) represent the local deformation ahead of
a crack. The opening mode, M o d e I, is characterized b y local displacements that

X
MODE I

i MODE

MODE lIT

FIG. 2.2 The three basic modes of crack surface

displacements (Ref. 5).
32 FRACTURE AND FATIGUE CONTROL IN STRUCTURES

are symmetric with respect to the x-y and x-z planes. The two fracture surfaces
are displaced perpendicular to each other in opposite directions. Local displace-
ments in the sliding or shear mode, Mode II, are symmetric with respect to the
x-y plane, and s k e w symmetric with respect to the x-z plane. The two fracture
surfaces slide over each other in a direction perpendicular to the line of the crack
tip. M o d e III, the tearing mode, is associated with local displacements that are
s k e w symmetric with respect to both x-y and x-z planes. The two fracture surfaces
slide over each other in a direction parallel to the line of the crack front. Each of
these modes of deformation corresponds to a basic type of stress field in the
vicinity of crack tips. In any analysis, the deformations at the crack tip can be
treated as one or a combination of these local displacement modes. Moreover,
the stress field at the crack tip can be treated as one or a combination of the three
basic types of stress fields.
Most practical design situations and failures correspond to M o d e I displace-
ments. Accordingly, M o d e I is emphasized in this book.
By using a m e t h o d d e v e l o p e d by Westergaard [6], Irwin [7] found that the
stress and displacement fields in the vicinity of crack tips subjected to the three
m o d e s of deformation are given by:
Mode I
KI
cos~-
o[ 1 - o
sin~-sin
o-x -- -(2wr)1/2
-

KI
o-y = -(21Tr)1/2
-cos~- 1+ sin-~sin

0 0 30
"rxy - (21Tr)1/~ sin ~- cos ~ cos - f -

ez = v(r + %), ~x~ = -ry~ = 0 (2.7)

KI[F]1/2cos~-0[ 1 -
u = ~- L~-~j 2v + s i n 2
~]
K~ [ r ] 1/2 O [ _02]
v = - ~ [~-~j sin-~ 2 - 2v- cos 2

w=O
Mode II

KII Sill ~ 2 + COS COS

{Yx = (2Trr)l/2 ~-

KII | | 30
~ry -= 2~rrl/2 sin ~ cos ~- cos -~- (2.8)

%y (2~rr)~/------
~ cos ~- 1 - sin ~- sin
 Stress Analysis for Members with Cracks--K I 33

= + %), = = 0
KII[ r ] 1/2 ~ [ ~]
1 sin 2-2 ,+cos 2

KII[r] 1/2 0 [ _ 1 + 2,o + sin2 ~ ]

V= LG/ cos?-

W=0

Mode III

KIII {~
"rx~ - (2,rrr)l/2sin ~-

- - - - C O S - -

TYz (2a-rr)1/2 2

(Yx = (Yy = (Yz = Txy ~" 0 (2.9)

w = 2 sin

u=v=0

where the stress components and the coordinates r and 0 are shown in Figure
2.3; u, v, and w are the displacements in the x, y, and z directions, respectively;
v is Poisson's ratio, and G is the shear modulus of elasticity.
Equations (2.7) and (2.8) represent the case of plane strain (w = 0) and ne-
glect higher-order terms in r. Because higher-order terms in r are neglected, these
equations are exact in the limit as r approaches zero and are a good approxi-
mation in the region where r is small compared with other x-y planar dimensions.
These field equations show that the distribution of the elastic-stress fields, and
the deformation fields in the vicinity of the crack tip, are invariant in all com-
ponents subjected to a given mode of deformation.
The magnitude of the elastic-stress field can be described by single-term
parameters, KI, Kip and Km, that correspond to Modes I, II, and III, respectively.
Consequently, the applied stress, the crack shape, size, and the structural config-
uration associated with structural components subjected to a given mode of de-
formation affect the value of the stress-intensity factor but do not alter the stress-
field distribution.
Dimensional analysis of Equations (2.7), (2.8), and (2.9) indicates that the
stress-intensity factor must be linearly related to stress and must be related to
the square root of a characteristic length. Based on Griffith's original analysis of
glass components with cracks and the subsequent extension of that work to more
ductile materials, the characteristic length in a structural member is the crack
length. Consequently, the magnitude of the stress-intensity factor must be related
directly to the magnitude of the applied nominal stress, (~. . . . and the square root
34 FRACTURE A N D FATIGUE C O N T R O L IN STRUCTURES

o- v
Y ~ Txy

FIG. 2.3 Coordinate system and stress components ahead of a crack

tip.

of the crack length, a. In all cases, the general form of the stress-intensity factor
is given by:
K = ffnom~ a "f(g) (2.10)
wheref(g) is a parameter that depends on the geometry of the particular member
and the crack geometry. Fortunately, a large number of relationships between the
stress-intensity factor and various b o d y configurations, crack sizes, orientations,
shapes, and loading conditions have been published [5,8-10]. The more common
ones are presented in this chapter. One key aspect of the stress-intensity factor,
K~, is that it relates the local stress field ahead of a sharp crack in a structural
member to the global (or nominal) stress applied to that structural member away
from the crack. Specifically, Figure 2.3 shows the stresses just ahead of a sharp
crack. Most fractures occur under conditions of Mode I loading (Figure 2.2).
Accordingly, the stress of primary interest in Figure 2.3 and in most practical
applications is cry. For o-v to be a maximum in Equation (2.7), let 0 = 0,

O'y - - K~ (2.11)

Rearranging this expression shows that:

KI = o'yX/2"rrr (2.12)
At locations farther and farther from the crack tip (increasing r), the stress,
 Stress Analysis for Members with Cracks--KI 35

o-y, decreases. However, KI remains constant and describes the intensity of the
stress field ahead of a sharp crack. This same stress-intensity factor is also related
to the global stress by Equation (2.10) for various crack geometries as described
in this chapter. Hence, KI describes the stress field intensity ahead of a sharp
crack in any structural member (plates, beams, airplane wings, pressure vessels,
etc.) as long as the correct geometrical parameter, fig), can be determined. Ex-
pressions for different crack geometries in various structural members are pre-
sented in the next section.

2.4 Stress-Intensity-Factor Equations

Stress-intensity-factor equations for the more common basic specimen geometries
that are subjected to Mode I deformations are presented in this section. Extensive
stress-intensity-factor equations for various geometries and loading conditions
are available in the literature in tabular form [5,8,9].

2.4.1 Through-Thickness Crack

The stress-intensity factor for an infinite plate subjected to uniform tensile
stress, o-, that contains a through-thickness crack of length 2a is:
KI = o-X/~a (2.13)
Note that from the Griffith analysis given in the appendix to this chapter there
is a direct relationship between Griffith's analysis and Equation (2.13).
A tangent-correlation factor having the form:
2b "wa~1/2
~aa tan - ~ / (2.14)

is used to approximate the/(i values for a through-thickness crack in plates of

finite width, 2b (Figure 2.4). Thus, the stress-intensity factor for a plate of finite
width 2b subjected to uniform tensile stress, (r, and that contains a through-
thickness crack of length 2a (Figure 2.4) is:
2b ~a~ 1/2
KI = o-X/-~a ~a tan ~-~] (2.15)

The values for the tangent correction factor for various ratios of crack length to
plate widths are given in Table 2.1. Equation 2.14 is accurate within 7% for a/b
-< 0.5.

2.4.2 Single-Edge Notch

The stress-intensity-factor equation for a single-edge-notch in an infinite
width specimen (Figure 2.5) is
KI = 1.12o-V~ (2.16)
The 1.12 factor is referred to as the free surface correction factor. Thus, when
36 F R A C T U R E A N D FATIGUE C O N T R O L I N S T R U C T U R E S

o-
r t t

J
2b D

1 1
o"
1
FIG. 2.4 Finite-width
plate containing a
through-thickness
crack.

the through-thickness crack of Figure 2.4 is " c u t " in half to form an edge crack
in a plate, the KI value is increased b y about 12%.
For single-edge-notch specimens h a v i n g finite width, an additional correc-
tion factor is necessary to account for b e n d i n g stresses caused b y lack of sym-
m e t r y in the single-edge-notched specimen. Equation (2.17) incorporates the var-
ious correction factors for a single-edge-notch in a finite w i d t h plate, Figure 2.5.

K1 = 1 . 1 2 ~ r ~ (b) (2.17)

The values of the function k ( a / b ) are tabulated in Table 2.2 for various values

TABLE2.1. Correction Factors for a Finite-width

Plate Containing a Through-thickness
Crack (Ref. 5).
a/b [2b/~a 9tan ~a/2b] 1/2
0.074 1.00
0.207 1.02
0.275 1.03
0.337 1.05
0.410 1.08
0.466 1.11
0.535 1.15
0.592 1.20
 Stress Analysis for Members with Cracks--K I 37

o-
r t

9
2b

l l
cr
l
FIG. 2.5 Single-edge-
notched plate of
finite width.

of the ratio of crack length to specimen width. Note that for a/b = 1.0, the notch
depth is one-half the width of the plate, and the correction factor is quite large.

TABLE 2.2. Correction Factors for

a Single-edge Notched
Plate (Ref. 5).
a/b k(a/b)
0.00 1.00
0.10 1.03
O.2O 1.07
0.30 1.15
0.40 1.22
0.50 1.35
0.60 1.50
0.70 1.69
0.80 1.91
0.90 2.20
1.00 2.55

2.4.3 Embedded Elliptical or Circular Crack in Infinite Plate

The stress-intensity factor at any point along the perimeter of elliptical or
circular cracks embedded in an infinite body subjected to uniform tensile stress
(Figure 2.6) is given by [12]:
O.N/-~ ( a2 )1/4
KI - Q sin 2 ~ + ~ cos2 ~ (2.18)
38 FRACTURE AND FATIGUE CONTROL IN STRUCTURES

O"

FIG. 2.6 Elliptical crack in an infinite body subjected

to uniform tension (Ref. 5).

In Equation (2.18), KI corresponds to the value of the stress-intensity factor

for a point on the perimeter of the crack whose location is defined by the angle
f~. Q is a flaw-shape parameter that depends on cr/O'y~ and a/2c, as shown in
Figure 2.7, where O-y~is the yield strength of the material. This figure is the
"Boeing Design Curve," [5], which is based on the solution of an elliptical inte-
gral, ~b, that is equal to X/Q as described in various handbooks [8-10]. This curve
is widely used to solve for KI factors for cracks with various a/2c ratios.
The stress-intensity factor for an embedded elliptical crack reaches a maxi-
m u m at [~ = -rr/2 and is given by the equation:

KI=~ ~a Q (2.19)

Values of the ratio of nominal applied stress, r to yield stress, O-ys,are in-
tended to account for the effects of plastic deformation in the vicinity of the crack
tip on the stress-intensity-factor value and are incorporated into the value of Q
which can be obtained from Figure 2.7.
For a circular crack where a = c and Q -~ 2.4, Equation (2.19) becomes:
K~ = 0.65 o-N/~ = 1.15 crV~a (2.20)
The exact expression for an embedded circular crack [5] is:
 Stress Analysis for Members with Cracks--K I 39

0.50

0.45 _ Surface Flaw K1 = 1'12c "/~~--C~" MK

0.40

0.35

0.30

~ l a 0.25 -
0.01.0----~j~-~- ,~'eroa, Raw _

~176
0.20

0.15 -

0.10

0.05

Q
FIG. 2.7 Effect of a / 2 c ratio a n d ~r/~ry s ratio on flaw-shape p a r a m e t e r O .

2
KI = ~ o-Va = 1.13 o'Va (2.21)

Note that the two values agree within 2% of each other.

2.4.4 S u r f a c e C r a c k
The stress-intensity factor for a part-through "thumbnail" crack in a plate
subjected to uniform tensile stress (Figure 2.8) can be calculated by using Equa-
tion (2.19) and a free-surface-correction factor equal to 1.12. Remember that this
1.12 factor occurs any time a crack originates at a free surface. The stress-intensity
factor for 13 = ~r/2 (which is the location of maximum stress intensity) is given
by:

K~ = 1.12 ~r~/'rr ~a " M k (2.22)

This equation is identical to Equation (2.19) except for the 1.12 multiplier
that corresponds to the front free-surface correction factor and M K, which cor-
responds to a back free-surface correction factor. M K is approximately 1.0 as long
as the crack depth, a, is less than one-half the wall thickness, t. As a approaches
t, M K approaches approximately 1.6, and a useful approximation is:

/
for values of a / t >- 0.5.
40 FRACTURE AND FATIGUE CONTROL IN STRUCTURES

--A-

r
E

FIG. 2.8 Plate

containing a semi-
elliptical surface
crack.

2.4.5 Cracks Growing from Round Holes

The stress-intensity factor for cracks growing from a circular hole (e.g., a bolt
hole) in an infinite plate (Figure 2.9) is given by:

where r = radius of hole, and

a = crack length from the side of the hole.
Note that as a / r approaches zero, f(a/r) for either one or two cracks ap-
proach a value of about 3. This is the value of the stress-concentration factor, kt,
at a round hole. Thus, the KI factor can be thought of as being equal to KI
kto-X/~a~afor very short cracks near a stress concentration where the local stress is
elevated by the stress concentration.
As the crack grows and (a/r) becomes large, the effect of the stress concen-
tration (in this case the round hole) decreases, and the expression approaches
that of a through-thickness crack in an infinite plate.

2.4.6 S i n g l e C r a c k in B e a m in B e n d i n g
The stress-intensity factor for a beam in bending that contains an edge crack
(Figure 2.10) is:

K~ = B ( W 6M
- a)3/2 " g ( W ) (2.25)

where M is the moment, and the values of g ( a / W ) are presented in Table 2.3 for
various ratios of crack length, a, to beam depth, W.
 Stress Analysis for Members with Cracks--K~ 41

O-

)
}-*--2r "1

@ )
)
K I = f(o/r) 9 o-~/-~-o

O"

o/r f(a/r), ONE CRACK f(a/r), TWO CRACKS

0.00 3.39 3.39
O. 10 2.73 2.73
0.20 2.30 2.41
0.30 2.04 2.15
0.40 1.86 1.96
0.50 1.73 1.83
0.60 1.64 1.71
0.80 1.47 1.58
1.0 1.37 1.45
I .5 I . 18 1.29
2.0 1.06 1.21
3.O 0.94 1.14
5.O 0.81 1.07
10.0 0.75 1.03
oo 0.707 1.00
FIG. 2.9 Cracks growing from a round hole (Ref. 5).

2.4.7 Holes or Cracks Subjected to P o i n t or Pressure Loading

The relation for KI for a crack growing from a hole subjected to centrally
applied point forces, P, such as a bolt, Figure 2.11 is:

P
KI - ~ (2.26)

w h e r e P = force/thickness of plate.
42 FRACTURE AND FATIGUE CONTROL IN STRUCTURES

~M
C M

FIG.2.10Edge-notched
beaminbending.
TABLE 2.3. Stress-intensity-factorCoefficients for Notched Beams (Ref. 5).
a/W 0.05 0.1 0.2 0.3 0.4 0.5 0.6 (and larger)
g(a/W) 0.36 0.49 0.60 0.66 0.69 0.72 0.73

Note that for a given P, KI decreases as the crack size, a, increases. Thus,
u n d e r the influence of a concentrated load, P, the driving force, KI decreases as
the crack length increases and the crack m a y be arrested.
The stress-intensity factor for a crack subjected to an internal pressure p, in
psi, as s h o w n in Figure 2.12 is given b y the equation:
K~ = 1.12 pV~-~a (2.27)
This expression can be used to calculate the additional effect of a v e r y high
pressure in a thick-walled pressure vessel with an internal surface crack as:

KI = 1.12 p (2.28)

2.4.8 Estimation o f Other K I Factors

The preceding stress intensity factors have been sufficient to analyze most
of the design cases and structural failures familiar to the authors. For v e r y un-
usual cases, the reader is referred to the various h a n d b o o k s of stress-intensity
factors available in the literature [8-10]. However, there are several common-
sense guidelines that can be applied to estimate stress-intensity factors for crack
geometries not described above.
These guidelines are as follows:
1. Free-Surface Correction
For e v e r y free-surface at which a crack might originate, increase KI b y 1.12.
For example, the KI for an e m b e d d e d circular crack (Figure 2.6) was given
b y Equation 2.21 as:
2
KI = ~ ~V~a (2.21)

If this were to occur at the corner of a plate, Figure 2.13, the KI expression
 Stress Analysis for Members with Cracks--Kx 43

/ /

FIG. 2.11 Crack subjected to point loads.

4 2a

FIG. 2.12 Crack subjected to internal pressure.

44 FRACTURE AND FATIGUE CONTROL IN STRUCTURES

FIG. 2.13 Circular corner crack.

would be:
2
K~ = (1.12)(1.12) ~ r (2.29)

because there are two (1.12) free surface flaw corrections.

2. Inclined Cracks
Experience has shown that cracks tend to grow perpendicular to the
direction of applied stress, particularly in fatigue or repeated loading.
Accordingly, inclined cracks of length L, such as is shown in Figure 2.14,
can be treated as a projected crack length of 2a. If 0 approaches 90 ~ the
crack is parallel to the direction of applied stress and generally is harmless,
i.e., KI ~ 0. This fact will be particularly important in Chapter 10, Fatigue
and Fracture Behavior of Welded Components, where some discontinuities
in welded structures are parallel to the direction of the applied stress and
thus are essentially harmless.
3. Bounding Using Existing Solutions
In many circumstances, the particular crack of interest has an irregular
shape. For most of these cases, the stress-intensity factor can be estimated
 Stress Analysis for Members with Cracks--K I 45

T~

l FIG. 2.14 Inclined crack.

from equations for simple and familiar shapes such as tunnel, circular, or
elliptical cracks.
Estimating stress-intensity factors for irregularly shaped cracks can be made
easier if one remembers some basic relationships and trends. A tunnel crack is a
special case of the equation for an embedded elliptical crack (Equation 2.18) with
the crack-shape parameter, Q, equal to 1.0. Furthermore, the equation for an
embedded circular crack is also a special case of Equation (2.18) for a = c (Q
2.4). Thus, as the crack shape changes from a circular crack front to a straight
crack front, the value of Q changes from 2.4 to 1.0 and the magnitude of the
stress-intensity factor, Kl, increases b y a factor of about 1.5. A corollary to these
observations is that, for a tunnel crack that is curved, KI is higher for the convex
front of the crack than for the concave front.
A study of the crack-shape parameter, Q (Figure 2.7), shows that when
a/2c is less than about 0.15 (i.e., when the crack length on the surface, 2c, is
larger than about six times its depth, a), the crack-shape parameter approaches
1.0 and the crack behavior approaches that for a long tunnel crack where a/2c
--* 0. An approximate linear relationship between the change in crack geometry,
a/2c, and the change in the crack-shape parameter, Q, between the extreme of a
circular crack and a long crack is as follows:
2.4(circular crack, a/2c = 0.5) > Q > 1.1(long crack, a/2c --~ 0.15) (2.30)
Thus, for every 0.1 decrease in the magnitude of a/2c, the value of the crack-
46 FRACTURE AND FATIGUE CONTROL IN STRUCTURES

shape parameter, Q, decreases by about 0.35. This observation is derived from

Figure 2.7, where the relationship between a / 2 c and Q is shown to be approxi-
mately linear in this range.
The application of the preceding information to estimate the stress-intensity
factor along the perimeter of an irregularly shaped crack can be demonstrated
by analyzing the embedded crack shown in Figure 2.15. The crack has an irreg-
ular shape and is subjected to a uniform tensile stress, o-, that is perpendicular
to the crack plane.
The K] for the crack in region 1 (where the crack is denoted as a 1 in Figure
2.15) can be represented by a circular crack having a radius al because the crack
is essentially circular in shape. The Kn value for this perimeter, which opens into
a long tunnel crack, should be slightly higher than the K~ for a circular crack of
radius a I and should be less than the KI value for a long tunnel crack having a
width equal to 2ap Therefore,

O'~-~ < KI1 < < o'~/-~al (2.31)

where Q = 2.4 or:
0.65 o ' X / ~ 1 < Kn < < o-X/-~al (2.32)
Thus, a reasonable estimate for K~ is closer to a circular crack, or:
KI1 ~ 0.75 o~V/~l (2.33)
is a reasonable estimate for that portion of the crack perimeter.
The crack front in region 2 corresponds closely to a long tunnel crack and,
therefore,

CONCAVE
KI2 ,~..//

_Z---. K,1

ti... +
2o 3 WITH G/2c = 0.2

FIG. 2.15 Estimating stress-intensity factors for an embedded crack having

an irregular shape.
 Stress Analysis for Members with Cracks--K I 47

/(i2 ~ (rX/--~a2 (2.34)

The elliptical crack front in region 3 can be represented b y a circular crack
of radius a3, or a tunnel crack of width 2a3. Because of the general elliptical shape,
the Kx for this region is better described for a tunnel crack than for a circular
crack. Therefore,
0.65 0-~/~3 < < KI3 ( c r V ~ a3 (2.35)
A reasonable estimate for K~3 is closer to a tunnel crack, or:
KI3 ~'~ 0.9 o - V ~ a3 (2.36)
The crack front in region 3 can be approximated also by an elliptical crack
with a minor axis equal to a3, and for example a crack geometry, a/2c, of about
0.2. The decrease in the value of a/2c from 0.5 for a circular crack to a value of
0.2 for the ellipse is equal to 0.3. Thus, the decrease in the value of Q from 2.4
for a circular crack is 1.05 (i.e., 0.3 • 0.35/0.1 = 1.05). Consequently, the value
of Q for the ellipse is approximately 1.35 (i.e., 2.4 - 1.05 = 1.35), and

KI3-~- o - / ~ = 0.86 o - V ~ a3 (2.37)

This value is a reasonable estimate of KI3, especially for the lower portion of
the crack front, which fits the ellipse more closely than the upper portion. Be-
cause the upper portion of the perimeter has less curvature than the ellipse, the
value for K~3 in that region should be increased slightly. Consequently,
KI3 ~ 0.9 o-V~a~ (2.38)
should be a good estimate for this region.

2.4.9 S u p e r p o s i t i o n o f S t r e s s - I n t e n s i t y F a c t o r s
Components that contain cracks may be subjected to one or more different
types of Mode I loads such as uniform tensile loads, concentrated tensile loads,
or bending loads. The stress-field distributions in the vicinity of the crack tip
subjected to these loads are identical and are represented by Equation (2.7). Con-
sequently, the total stress-intensity factor can be obtained by algebraically adding
the stress-intensity factors that correspond to each load.
For example, Figure 2.16 shows a plate with an edge crack subjected to
tension plus bending loads. The combined stress intensity factor for that case is
KI TOTAL= KI (EDGE) q- KI (BENDING) (2.39)

KI TOTAL = 1.12 ~X/-~a 9 k ( b ) + B ( W 6M

- a) 3/2 " g ( W )

A fairly common situation is a plate loaded in tension at one end and this
force being withstood by a bolt force P. The solution, KI, is a superposition of KI
48 FRACTURE AND FATIGUE CONTROL IN STRUCTURES

Ill
op- o~- o~-
> + >

2b

FIG. 2.16 Superposition solution for tensile and bending stresses applied to a single-
edge notched plate.

solutions for a through-thickness crack (Equation 2.13), and a crack subjected to

point loads (Equation 2.26), as shown in Figure 2.17.
Figure 2.18 shows a portion of a thin-walled pressure vessel subjected to a
hoop stress, O'HOOP, and the crack surface subjected to the internal pressure, p.
K~ = KI (HOOP) + K~ (PRESSURE)

KI = 1.12 ~HooP + p (2.40)

/<i = 1.12 pD + P

Note that as D / t becomes large, the effect on K~ of the pressure stress on the
crack becomes 10% or less and thus generally is neglected, particularly for shal-
low flaws.
Some components may be subjected to loads that correspond to various
modes of deformation. Because the stress-field distributions in the vicinity of a
crack, Equations (2.7), (2.8), and (2.9) are different for different modes of defor-
mation, the stress-intensity factors for different modes of deformation cannot be
added. Under these loading conditions, the total energy-release rate, G, described
in the appendix, rather than stress-intensity factors, K, can be calculated b y al-
gebraically adding the energy-release rate for the various modes of deformation.
 Stress Analysis for Members with Cracks--K~ 49

cr cr

2b--
f t t
p = cr(2b)

t +
~P
~P

o" o"

KII = KI2 + KI3 - KI4

o" o"

-, I I
P = o-(2b)
~P
+
~P

1 2 3

o- o"

KI4 + KI1 = KI2 + Kz3

k ~r J

2 KI1 = KI2 + KI3

1
KI1 = ~ (K[2 + KI3]

K~ = ~~ [ o-,/-~ + ~~ j

FIG. 2.17 S u p e r p o s i t i o n solution for a c r a c k e m a n a t i n g f r o m a hole.

2.5 C r a c k - T i p D e f o r m a t i o n and Plastic Zone Size

The stress-field equations, Equations (2.7), (2.8), and (2.9), show that the elastic
stresses in the vicinity of a crack tip where r < < a can be very large. In reality,
such high stress magnitudes do not occur because the material in this region
undergoes plastic deformation, thus creating a plastic zone that surrounds the
crack tip. Figure 2.19 is a schematic presentation of the change in the distribution
of the y component of the stress caused by the localized plastic deformation in
the vicinity of the crack tip.
50 FRACTURE AND FATIGUE CONTROL IN STRUCTURES

7 t
7
Intema
Z
2c- )

/ / /
~
FIG. 2.18 Superposition of K~ factors for hoop stress and
internal pressure.

ELASTIC STRESS DISTRIBUTION

STRESS DISTRIBUTION
, ', ~_ /AFTER LOCAL YIELDING
',, !~/
,, ! ~
cr

G . . . . . . . . . . . .

I
I
i
I
:
1I
CRACK 2ry !,,I
TIP
-','-PLASTIC ZONE-'~-

FIG. 2.19 Distribution of the Cry stress c o m p o n e n t in the crack-tip region.

 Stress Analysis for Members with Cracks--K~ 51

The size of the plastic zone, ry, can be estimated from the stress-field equa-
tions by setting the y component of stress, o-y, equal to the yield strength, o-ys,
which results in:

ry = G1 (~ys)2 (2.41)

Irwin [13] suggested that the plastic-zone size under plane-strain* conditions
can be obtained b y considering the increase in the tensile stress for plastic yield-
ing caused by plane-strain elastic constraint. Under these conditions, the yield
strength is estimated to increase b y a factor of V3. Consequently, the plane-strain
plastic-zone size becomes:

ry = ~1 (~sys) 2 (2.42)

The plastic zone along the crack front in a thick specimen is subjected to
plane-strain conditions in the center portion of the crack front where w = 0 and
to plane-stress conditions near the surface of the specimen where o-z = 0. Con-
sequently, Equations (2.41) and (2.42) indicate that the plastic zone in the center
of a thick specimen is smaller t h a n at the surface of the specimen. A schematic
representation of the variation of the plastic-zone size along the front of a crack
in a thick specimen is shown in Figure 2.20. Irwin [13] suggested that the effect
of small plastic zones corresponds to an apparent increase of the elastic crack
length by an increment equal to ry. This plastic-zone correction factor is valid for
small plastic-zone sizes and as described in the next section can be used to es-
timate K~eft (effective) factors for loadings that result in moderate plastic zone
sizes.

2.6 Effective K I Factor for Large Plastic Z o n e S i z e

Figure 2.21 shows the effective crack length, 2a + 2ry, for a wide plate loaded in
tension. As O-app,ed is increased a n d the plane stress plastic zone size increases,
use ae~~ = a 0 + ry to establish KI:

Kie ff ~ O ' a p p l i e d ~ (2.43)

K~e~f= O'applied "IT a + ~ \ o-y~! ! (2.44)

*Plane strain and plane stress are conditions of maximum and variable constraint, respectively,
as discussed in Chapters 3 and 4.
52 FRACTURE A N D FATIGUE CONTROL IN STRUCTURES

Midsection~/~ !

A-Overall View B-Edge View

--ry (PlaneStrain)
Machine Notch
Fatigue Crack----~ 3_
Plastic Zone----~,- ry (PlaneStress)
7~t"~ Pla~ne T
Strain
IRegi~ I

SpecimenCross-section
FIG. 2.20 Schematic representation of plastic zone ahead of crack
tip.

t t t t t t t t t tO'Applied

(~Applied

FIG. 2.21Effective crack length,

2ae. = 2a + 2ry.
 Stress Analysis for Members with Cracks--K~ 53

Solving Equation (2.44) for K~ as K;ff because of the larger plastic zone size, we
obtain:

( KIef______L/2= ~a + 1 (KI~f.___2~
O'app/ 2" \ O'ys//
2

[ .2
(Kieff)2 1 1 - (~aPE1 2 (2.45)
2~ 2s ~ = ~ra- O'app
O'app~a

For (Yapp -~- 3/40"y s (a reasonable upper limit to the applied nominal stress in a
structure):
O-app~/~--a~a O'app'~/~
K~,,, = [ 1 (: ~)211/2 ( 1 - .28) 1/2 (2.46)
1 -- ~ O'Ys

KI~ - .85 1.18 O-appV~

Note that this is only an 18% increase in K I compared with KI for ry - 0,

and thus should give a reasonable approximation to the driving force for fairly
large stress levels.
For this same case, and a crack length 2a = 2", the absolute value of rp is
still fairly small:

1 (KI~ 2 1 (.75~ (2.47)

rp = - ~ \~ s,/ ~ G \ O'ys //
rp = .28 in.
Thus, calculating KI as a driving force for cr = 3/40-y s is not an unrealistic
extension of linear-elastic fracture mechanics into the beginning of the elastic-
plastic regime. In fact, for (Yapp = O'ys, as might be the case for weldments where
the residual stress m a y be equal to trys, the approximation of K;ff is:
O-yf~--a~a
KI-~ 1 1 - ~ 1 ( ( 1 ) 2 ) ] 1/2 (2.48)

KI --- 1.40ysV~
If the structural component is heavily constrained, the linear portion of the
54 FRACTUREAND FATIGUE CONTROL IN STRUCTURES

cr-s curve would be elevated such that the yield stress, O-y~,m a y well be elevated
40%, and the behavior w o u l d still be close to linear-elastic.
In summary, as a first approximation, use of KI = KI~,~= flO-app, a q- J'y) appears
to be a reasonable engineering approach, even for stresses approaching the yield
strength.

2.7 JI a n d 51 Driving Forces

Two other fracture mechanics parameters that can be used to calculate driving
forces are the path-independent integral, Ji, and the crack-tip-opening displace-
ment (CTOD), 8I. These two parameters are used primarily in the elastic-plastic
regime, whereas the stress intensity parameter, Kx, is used primarily in the linear-
elastic regime.

2.7.1 J Integral
The path-independent J-integral proposed by Rice [14] is a method of char-
acterizing the stress-strain field at the tip of a crack by an integral path taken
sufficiently far from the crack tip to be analyzed elastically, and then substituted
for the inelastic region close to the crack-tip region. Thus, even though consid-
erable yielding (elastic-plastic behavior) may occur in the vicinity of the crack
tip, if the region away from the crack tip can be analyzed elastically, behavior of
the crack-tip region can be inferred. Elastic stress strain behavior is assumed,
even though the stress-strain curve may be nonlinear. This technique can be used
to estimate the fracture characteristics of materials exhibiting elastic-plastic be-
havior and is a means of extending fracture-mechanics concepts from linear-
elastic (Kk) behavior to elastic-plastic behavior.
For linear-elasfic behavior, the J-integral is identical to G, the energy release
rate per unit crack extension (see Appendix). Therefore,
(1 - v2)/~r
J1 = G1 = E (2.49)

There are procedures available to calculate J1 driving forces and these are
introduced in the appendix to this chapter. The most widely used procedure is
to use 2D or 3D finite element analysis programs to evaluate the J-integral ahead
of a crack in the geometry of interest. It requires that a stress-strain curve be
available and was first used in the pressure vessel industry, primarily for nuclear
pressure vessels. However, in most cases, the K~ driving force is used, recognizing
that:

KI = ~.1 JIE
~ v2
for plane strain conditions and,
 stress Analysis for Members with Cracks--K~ 55

KI = ~ (2.50)
for plane stress conditions.
As will be discussed in Chapter 3, it is common in most structural situations
to measure the critical resistance fracture toughness using the J-integral proce-
dure (Jc or Jir convert these values to Kc or KIc, and then compare the driving
and resistance forces in terms of KI and K~. For complex structures, such as nu-
clear pressure vessels, where elastic-plastic analyses can be justified, J~ is deter-
mined using finite element analyses. J~ is measured as described in Chapter 3.
Fracture control then consists of keeping Ji < Jc.

2.7.2 C T O D (8~)
In 1961, Wells [15] proposed that the fracture behavior in the vicinity of a
sharp crack could be characterized by the opening of the notch faces, namely,
the crack-tip opening displacement, CTOD, as shown in Figure 2.22. Further-
more, he showed that the concept of crack-opening displacement was analogous
to the concept of critical crack extension force (Gc as described in the appendix
to this chapter), and thus the CTOD values could be related to the plane-strain
fracture toughness, K~. Because CTOD measurements can be made even when
there is considerable plastic flow ahead of a crack, such as would be expected
for elastic-plastic or fully plastic behavior, this technique can be used to establish
critical design stresses or crack sizes in a quantitative manner similar to that of
linear-elastic fracture mechanics.
The CTOD relationship for a center crack in a wide plate is developed ha
the appendix and is as follows:
2
8~ - "~appa (2.51)
Eo-ys
Because,
K I = O'appV~ (2.52)

81 -
Eo'y s
Thus, in the same manner as the design use of the J-integral, it is common
to measure ~ (see Chapter 3) and then convert 8c values to Kc values. These K~
resistance values obtained from CTOD test results are then compared to the driv-
ing force in terms of KI.

2.8 S u m m a r y

One of the underlying principles of fracture mechanics is that unstable fracture

occurs when the stress intensity factor at the crack tip, KI, reaches a critical value,
Kc. KI represents the stress intensity ahead of a sharp crack in any material and
56 FRACTURE A N D FATIGUE CONTROL IN STRUCTURES

CLIP GAGE DISPLACEMENT

(a) KIc/CTOD TEST SETUP

"DUCTILE" MATERIAL
...... ~ U L A T E 8c)
LOAD,
P ,L "BRITTLE" 1
/ Ki~\iAILc~.U
).L/~ITE
MATERIAL K j

t I I
O CTOD CTOD
AT FRACTURE AT MAX. LOAD
CL1P GAGE DISPLACEMENT, inches

(b) P = CTOD RELATIONSHIP

FIG. 2.22 Relation between Kjc and C T O D test behavior, Kjc-
C T O D test setup, and load displacement records from K~c-
C T O D tests, as described in Chapter 3.

is a mathematical calculation based on the crack size and the geometry of the
member, as well as the applied load. K~ relations for various crack geometries
have been presented in this chapter. Kc represents the fracture toughness of a
particular material at a given temperature, loading rate, and constraint level, and
must be measured. Procedures used to measure Kr are presented in Chapter 3.
The stress intensity factor, KI, although based on linear-elastic concepts, is
quite useful for calculating driving forces in structures even if the plastic zone
size ahead of the crack becomes moderately large. However, most structures are
loaded in the elastic range at stresses less than yield. Therefore, as described in
or
this chapter, the use of either KI Kieff as a driving force appears to be a realistic
engineering approach.
Elastic-plastic methods, such as the J-integral and the CTOD (crack tip open-
ing displacement, 8) parameter, also can be used to calculate driving forces. How-
 Stress Analysis for Members with Cracks--K I 57

ever, most engineers will m e a s u r e the fracture toughness in terms of ]c or 8c and

convert those values to a Kc. Then the comparison of driving force and resistance
force can be made by using the stress intensity factors KI and K~. The advantage
of this approach is that K~, the driving force, is expressed as a function of stress
and flaw size, terms familiar to engineers. Also, there are m a n y K~relations avail-
able for various crack geometries and structural configurations.
As mentioned previously, Chapter 3 describes the procedures used to mea-
sure Kc under various conditions of loading rate and constraint. Chapter 4 then
will describe the effect of temperature, loading rate, and constraint on the resis-
tance force, K~.
As will be described in Part IV, KI should be kept below Kc at all times to
prevent fracture of members with flaws in the same manner that the applied
stress, r is kept below the yield strength, Cys, to prevent yielding in the design
of members that do not have flaws.

2.9 References
[1] Griffith, A. A., "The Phenomena of Rupture and Flaw in Solids," Transactions, Royal Society of
London, Vol. A-221, 1920.
[2] Inglis, C. E., "Stresses in a Plate due to the Presence of Cracks and Sharp Comers,"
Proceedings, Institute of Naval Architects, Vol. 60, 1913.
[3] Irwin, G. R., "Fracture Dynamics," in Fracturing of Metals, American Society of Metals,
Cleveland, 1948.
[4] Orowan, E., "Fracture Strength of Solids," in Report of Progress in Physics, Vol. 12, Physical
Society of London, 1949.
[5] Paris, C. P. and Sih, G. C., "Stress Analysis of Cracks," in Fracture Toughness Testing and Its
Applications, ASTM STP 381, American Society for Testing and Materials, Philadelphia, 1965.
[6] Westergaard, H. M., "Bearing Pressures and Cracks," Transactions, ASME, Journal of Applied
Mechanics, 1939.
[7] Irwin, G. R., "Analysis of Stresses and Strains Near the End of a Crack Transversing a Plate,"
Transactions, ASME, Journal of Applied Mechanics, Vol. 24, 1957.
[8] Tada, H., Paris, P. C., and Irwin, G. R., Eds., Stress Analysis of Cracks Handbook, Del Research
Corporation, St. Louis, MO, 1985.
[9] Sih, G. C., Handbook of Stress-Intensity Factorsfor Researchers and Engineers, Institute of Fracture
and Solid Mechanics, Lehigh University, Bethlehem, PA, 1973.
[10] Rooke, D. P. and Cartwright, D. J., "Compendium of Stress Intensity Factors," Her Majesty's
Stationary Office, London, Hillingdon Press, 1976.
[11] Pilkey, W. D., "Petersons Stress Concentration Factors," Second Edition, John Wiley & Sons
Inc., 1997.
[12] Irwin, G. R., "The Crack Extension Force for a Part Through Crack in a Plate," Transactions,
ASME, Journal of Applied Mechanics, Vol. 29, No. 4, 1962.
[13] Irwin, G. R., "Plastic Zone Near a Crack and Fracture Toughness," 1960 Sagamore Ordnance
Materials Conference, Syracuse University, 1961.
[14] Rice, J. R., "A Path Independent Integral and the Approximate Analysis of Strain
Concentration by Notches and Cracks," Journal of Applied Mechanics, Transactions ASME, Vol.
35, June 1968.
[15] Wells, A. A., "Unstable Crack Propagation in Metals--Cleavage and Fast Fracture: Cranfield
Crack Propagation Symposium, Vol. 1, September 1961, p. 210.
[16] Dugdale, D. S., "Yielding of Steel Containing Slits," Journal of Mechanics and Physics of Solids,
Wiley Interscience, New York, 1963, p. 103.
58 FRACTURE A N D FATIGUE C O N T R O L I N S T R U C T U R E S

[17] Sorem, W. A., Dodds, R. H., and Rolfe, S. T., "An Analytical Comparison of Short Crack and
Deep Crack CTOD Fracture Specimens of an A36 Steel," Fracture Mechanics: 21st Symposium,
ASTM STP 1074, J. P. Gudas, J. A. Joyce, and E. Hackett, Eds., American Society for Testing
and Materials, Philadelphia, 1990, pp. 3-23.
[18] Wilson, A. D. and Donald, K., "Evaluating Steel Toughness Using Various Elastic-Plastic
Fracture Toughness Parameters," Nonlinear Fracture Mechanics: Volume Ih Elastic-Plastic Fracture,
ASTM STP 995, J. D. Landes, A. Saxena, and J. G. Merkle, Eds., American Society for Testing
and Materials, Philadelphia, 1989, pp. 144-168.
[19] Dodds, R. H. and Read, D. T., "Elastic-Plastic Response of Highly Deformed Tensile Panels
Containing Short Cracks," ASME Special Publication, Computational Fracture Mechanics--
Nonlinear and 3-D Problems, Vol. 85, PVPD, June 1984, pp. 25-34.
[20] Dodds, R. H., Read, D. T., and Wellman, G. W., "Finite-Element and Experimental Evaluation
of the J-Integral for Short Cracks," Fracture Mechanics, Vol. I: Theory and Analysis, ASTM STP
791, American Society for Testing and Materials, Philadelphia, 1983, pp. 520-542.
[21] API 579--Recommended Practice for Fitness--For Service, Appendix B: Stress Analysis
Overview for a FFS Assessment, Draft in Preparation, 1999.

Appendix

2.10 GRIFFITH, CTOD AND J-INTEGRAL THEORIES

2.10.1 The GrifJith Theory
The first analysis of fracture behavior of components that contain sharp dis-
continuities was developed by Griffith [1]. The analysis was based on the as-
sumption that incipient fractures in ideally brittle materials occur when the mag-
nitude of the elastic energy supplied at the crack tip during an incremental
increase in crack length is equal to or greater than the magnitude of the elastic
energy at the crack tip during an incremental increase in crack length. This en-
ergy approach can be presented best by considering the following example.
Consider an infinite plate of unit thickness that contains a through-thickness
crack of length 2a (Figure 2.4) and that is subjected to uniform tensile stress, o-,
applied at infinity. The total potential energy of the system, U, may be written
as:

u : Uo - G + G (A-l)

where Uo = initial elastic energy of the uncracked plate,

Ua = decrease in the elastic energy caused by introducing the crack in
the plate, and
 Stress Analysis for Members with Cracks--Ky 59

Uv = increase in the elastic-surface energy caused by the formation of

the crack surfaces.
Griffith used a stress analysis that was developed by Inglis [2] to show that:
gr~2a 2
U a- E (A-2)

the elastic-surface energy, Uy, is equal to the product of the elastic-surface energy
of the material, "Ye,and the new surface area of the crack:
U~ = 2(2a'Ye) (A-3)
Consequently, the total elastic energy of the system, U, is:
,rro-2a2
U = Uo - ~ + 4a~te (A-4)

The equilibrium condition for crack extension is obtained by setting the first
derivative of U with respect to crack length, a, equal to zero. The resulting equa-
tion can be written as:
dU 2a'rro-2
da - 0 E + 4ye = 0 (A-5)

2acro-2 = 4~leE (A-6)

o-'Va = (2~--~E)1/2 (A-7)

which indicates that crack extension in ideally brittle materials is governed by

the product of the applied nomh~al stress and the square root of the crack length
as well as by material properties. Because E and ~/~ are material properties, the
right-hand side of Equation (A-7) is equal to a constant value characteristic of a
given ideally brittle material. Consequently, Equation (A-7) indicates that crack
extension in such materials occurs when the product cX/a attains a critical value.
This critical value can be determined experimentally by measuring the fracture
stress for a large plate that contains a through-thickness crack of known length
and that is subjected to a remotely applied uniform tensile stress. This value can
also be measured by using other specimen geometries, which is what makes this
approach to fracture analysis so powerful.
Equation (A-7) can be rearranged in the form:
,rro-2a
E - 2V~ (A-8)

The left-hand side has been designated the energy-release rate, G, and rep-
resents the elastic energy per unit crack surface area available for infinitesimal
60 FRACTURE AND FATIGUE CONTROL IN STRUCTURES

crack extension. The right-hand side of Equation (A-8) represents the material's
resistance to crack extension, R.
In 1948 Irwin suggested that the Griffith fracture criterion for ideally brittle
materials could be modified and applied to brittle materials and to metals that
exhibit plastic deformation. A similar modification was proposed by Orowan at
about the same time. The modification recognized that a material's resistance to
crack extension is equal to the sum of the elastic-surface energy and the plastic-
strain work, 7p, accompanying crack extension. Consequently, Equation (A-8) was
modified to:
,rr(y2a
G - E = 2('Ye q- Vp) (A-9)

Because the left-hand side is the energy-release rate, G, and because (rV'~wa
represents the intensity, KI, of the stress field at the tip of a through-thickness
crack of length 2a, the following relation exists between G and K~ for plane-strain
conditions,

"a'cr2a~-- G = --~-K~(1
I - 122) (A-10)

The energy-balance approach to crack extension defines the conditions re-

quired for instability of an ideally sharp crack. This approach is not applicable
to analysis of stable crack extension such as occurs under cyclic-load fluctuation
or under stress-corrosion-cracking conditions. However, the stress-intensity pa-
rameter, K, is applicable to stable crack extension, and therefore development of
linear-elastic fracture-mechanics theory has assisted greatly in improving our un-
derstanding of subcritical crack extension and crack instability.

2.10.2 Crack-Tip Opening Displacement (CTOD) and

the Dugdale Model
The tensile stresses applied to a b o d y that contains a crack tend to open the
crack and to displace its surfaces in a direction normal to its plane. For small
crack-tip displacements and small plastic deformation at the crack tip, the stress
and strain fields in the vicinity of the crack tip can be described by linear-elastic
analyses. Under these loading conditions, the fracture instability can be predicted
by using the critical plane-strain stress-intensity factor, Kic.
As the size of the plastic zone and of the displacements at the crack tip
increase, the stress and strain distributions in that neighborhood can be charac-
terized better by using elastic-plastic analyses rather than linear-elastic analyses.
Wells [15] argued that the opening displacement at the crack tip reflects the strain
distribution in that region. He also proposed that fracture would initiate when
the strains in the crack-tip region reach a critical value, which can be character-
ized by a critical crack-tip opening displacement.
 Stress Analysis for Members with Cracks--K I 61

Using a crack-tip plasticity model proposed by Dugdale [16], referred to as

the strip-yield-model analysis, it is possible to relate the CTOD to the applied
stress and crack length. The strip yield model consists of a through-thickness
crack in an infinite plate that is subjected to a tensile stress normal to the plane
of the crack (Figure A.1). The crack is considered to have a length equal to 2a +
2ry. At each end of the crack there is a length, ry, that is subjected to yield-point
stresses that tend to close the crack or, in reality, to prevent it from opening.
Another w a y of looking at the behavior of this model is to assume that yield
zones of length ry spread out from the tip of the real crack, a, as the loading is
increased. Thus, the displacement at the original crack tip, 8, which is the CTOD,
increases as the real crack length increases or as the applied loading increases.
The basic relationship developed by Dugdale [16] is:

o"

tttttt

2a + 2ry

111111
o-

ry ~1 _ J 2a "y t
I

9 -t --L -

FIG. A.1 Dugdale strip-yield model.

62 FRACTURE AND FATIGUE CONTROL IN STRUCTURES

8 = 8 --ffff in sec -~ (A-11)

where O'ys = yield strength of the material, ksi,

a = 1,4 real crack length, in.,
o- = nominal stress, ksi, a n d
E = m o d u l u s of elasticity of the material, ksi.
Using a series expansion for In sec [(~/2)(o-/O-y~)], this expression becomes:

8 = 8~y;a [~ (2 ~ys)2+ 1-~ (2 ~ys)4 q- 4-~ (2 ~s) 6 ""] (A-12)

For nominal stress values less than Cry~, a reasonable approximation for 8,
using only the first term of this series, is:
,a-ff2a
8 - (A-13)
Eo'ys
In this chapter, it was shown that, for a through-thickness crack of length 2a,
KI = ~rX/~a (A-14)
Thus BEerys -- K~I,and since E = O-yJSy~, the following relation exists:

'~ys \O'ys//
- - = - - (A-15)

Also, the strain energy release rate, G, is:

~o-2a
G- E (A-16)

Therefore,
G = 8 9 O-ys (A-17)
At the onset of crack instability under plane-strain conditions, where KI
reaches K~c and CTOD reaches a critical value, 8c,
8c (Klcl 2
'Sys \O-ys//
- - = - - (A-18)

Because (K~c/Sys)2 can be related to the critical crack size in a particular struc-
ture, it is reasonable to assume that the parameter 8c/Sys can likewise be related
to the critical crack size in a particular structure. The advantage of the CTOD
approach is that the CTOD values can be measured throughout the entire plane-
strain, elastic-plastic, and fully plastic behavior regions, whereas K~c values can
be measured only in the elastic plane-strain region or approximated in the early
portions of the elastic-plastic region. These regions are described in Chapter 3.
As with the KI analysis, the application of the CTOD approach to engineering
structures requires the measurement of a fracture-toughness parameter, 8c, which
 Stress Analysis for Members with Cracks--K I 63

is a material p r o p e r t y that is a function of temperature, loading rate, specimen

thickness, and specimen geometry. The CTOD test m e t h o d o l o g y is described in
Chapter 3.

2.10.3 J-Integral
For linear-elastic behavior, the J-integral is identical to G, the e n e r g y release
rate per unit crack extension, described previously in this appendix. Therefore a
/-failure criterion for the linear-elastic case is equivalent to the Krc failure criterion.
For linear-elastic plane-strain conditions,

(1 - 1)2)G (A-19)
lic = Gic - E

The CTOD parameter, 8, also is related to J (and K) as follows:

G K2
8 - = - - (A-20)
m(Yys m~ysE
because J = G,
j ~ rr/O'ys8 (A-21)
where l -< m <- 2.
Recent studies [17,18] of the relations b e t w e e n J and 8 indicate that o-now =
(Crys + %1t)/2 should be used in Equation (A-21). Also, m = 1.7 fits most exper-
imental data. Thus,
J = 1.7Cr~ow8 (A-22)
is a preferred relation b e t w e e n J and 8.
The l-integral, a mathematical expression, is a line or surface integral that
encloses the crack front from one crack surface to the other as s h o w n in Figure
A.2. It is used to characterize the local stress-strain field a r o u n d the crack front
for either elastic or elastic-plastic behavior.
The line integral is defined as follows:

I = W dy - T ~ ds (A-23)

where F = any contour s u r r o u n d i n g the crack tip as s h o w n in Figure

A.2 (note that the integral is evaluated in a counterclock-
wise m a n n e r starting from the lower flat notch surface
and continuing along an arbitrary p a t h F to the u p p e r
flat surface),
W loading w o r k per unit v o l u m e or, for elastic bodies, the
strain e n e r g y density = J'0~crdr
T = the traction vector at ds defined according to the o u t w a r d
n o r m a l n along F, T / = o-/jnj,
displacement vector at ds,
64 FRACTURE AND FATIGUE CONTROL IN STRUCTURES

T
ds

FIG. A.2 Crack-Up coordinate system and typical line

integral contour.

ds = arc length along contour F,

T -~x ds = the rate of work input from the stress field into the area
enclosed by F.
For any linear-elastic or elastic-plastic material treated by deformation the-
ory of plasticity, Rice [14] has proven path independence of the ]-integral. Dodds
et al. [19,20] have described experimental and analytical methods for direct eval-
uation of the J-integral.
The most widely used analyses methods to determine the driving force, ]i,
are either 2D or 3D finite element analyses, depending on the structure under
consideration. API 579--Recommended Practice for Fitness-For-Service [21] de-
scribes this process in detail. Briefly, the steps involved include:
Step 1 Develop a finite element model of the structural component including
all relevant geometry and flaw characteristics.
Step 2 Define all relevant loading conditions and apply them to the model of
the structural component.
Step 3 An accurate stress-strain curve of the material used in the structural
component should be included in the model.
Step 4 Perform the analysis using an evaluation of the J-integral defined in
Equation A-23.
Step 5 Measure Jc as described in Chapter 3.
Step 6 As stated previously for fracture control, keep ]~ < Jc-