The Fracture Mechanics Method (da/dN- K)

June 2nd – 6th, 2014,

Aalto University,
Espoo, Finland

Prof. Grzegorz Glinka

University of Waterloo, Canada

© 2010 Grzegorz Glinka. All rights reserved. 1

 Information path for fatigue life estimation based on
the da/dN- K method

LOADING GEOMETRY, Kt MATERIAL

F
E
PSO

0
t

Stress-Strain
Analysis

MATERIAL
da n
dN
Damage Analysis

DKth DK

Fatigue Life

© 2008 Grzegorz Glinka. All rights reserved. 2

Steps in Fatigue Life Prediction Procedure Based on the
da/dN- K Approach (cont’d)
i)

af
f)
Stress intensity factor, K
(indirect method)

Crack depth, a
Weight function, m(x,y)
(x, y)
K x, y m x, y dxdy Fatigue Life
A

K
Y
n a
ai
Stress intensity factor, K
(direct method) Number of cycles , N

g)
KI yFE 2 xFE h) Integration of Paris’ equation
a
or
m
K E
dU
EG
ai C Ki Ni
da
N

K af a0 ai
Y
n a i 1

N Ni

© 2008 Grzegorz Glinka. All rights reserved. 3

 The Similitude Concept in the da/dN – K Method
10-6
a) Structure

Crack Growth Rate, m/cycle

10-7
H

F 10-8
Q

10-9
b) Weld detail
a
10-10

10-11

10-12
K 1 10 100

K , MPa m
a c) Specimen

The Similitude Concept states that if the stress

P intensity K for a crack in the actual component
and in the test specimen are the same, then the
fatigue crack growth response in the component
and in the specimen will also be the same and
can be described by the material fatigue crack
P growth curve da/dN - K.

© 2008 Grzegorz Glinka. All rights reserved. 4

Crack tip stress dependence on the stress intensity factor K
y
S S

K1 < K2 < K3
K3
yy
2 x Stress components, ij, at the
K2 crack tip depend on the stress
yy
2 x intensity factors KI which is
influenced by:
uy - the load, S
x
-crack dimensions, a
-geometry, Y
r
yy
a
The stress field , ij, around
the crack tip can be described
xx
by one universal function
xy
valid for all cracks of Mode I,
i.e. for =0

S S K
yy
2 x
© 2010 Grzegorz Glinka. All rights reserved. 5
 G. Irwin’s fundamental Fracture Mechanics principles:
1. The near crack tip stress field expressions above are universal, i.e. the stress
distributions in the vicinity of the crack tip have the same general mathematical
form regardless of the crack geometry, loading and geometrical shape of the body.

2. The strain energy release rate GI is related to the stress intensity factor KI and
therefore it is justified (and easier) to calculate the strain energy release rate (and
the critical stress) from the purely elastic local (near the crack tip) stress
distribution (i.e. from the Stress Intensity Factor).

2
S aY K I2
GI plane stress
E E
2
S aY K I2
2 2
GI 1 1 plane strain
E E
© 2010 Grzegorz Glinka. All rights reserved. 6
 A crack becomes unstable (fracture) when the stress intensity factor,
KI , exceeds the critical, for given material, stress intensity factor KIc!

KI > KIc
Stress Intensity
Force [MN] Stress [MPa] Factor [MPa m]

Strength
parameters P KI Fracture
Mechanics
traditionally parameters
used for the used for the
sttrength
analysis of
PY Y Kc strength
analysis of
engineering engineering
components
and structures: Pcr ut s K Ic components
and structures:

Leonardo da Vinci Euler, Cauchy Irwin

17th-century 19th -century 20th-century

© 2010 Grzegorz Glinka. All rights reserved. 7

 General Stress Intensity Factor Expressions for Cracks in Mode I
The stress intensity factor is defined as:

KI S a Y
in which S is the stress (usually the nominal) away from the crack. The geometry
factor, Y , accounts for the effect of geometry of the crack and the body, the
boundary conditions and the type of loading.
Determining stress intensity factor means in essence the derivation of the function
describing the geometrical factor Y. One of the confusing issues while determining
stress intensity factors is that the remotely applied stress S and the geometry factor
Y are inter-related. The value of parameter Y depends on the definition of the remote
(termed often as nominal) stress S. In cases where the nominal or hot spot stress is
well defined there is no problem in the definition of the remote stress S. However, if
the stress distribution is non-uniform it may not be clear which stress should be
used in the expression for the stress intensity factor. Theoretically, any reference
stress S can be chosen for the determination of the geometrical factor Y, as far as
the stress varies proportionally with the applied load. However, the user of given
expression for K has to use the same definition of the reference stress while
carrying out fatigue and fracture analyses. Nominal or the maximum stress in the
case of non-uniform stress distributions is most often used in stress intensity factor
expressions. Therefore, it is a good professional practice to define the reference
stress S when quoting the geometry factor Y.
© 2010 Grzegorz Glinka. All rights reserved. 8
 Center Crack Plate under Uniform Tension

2a

W
t
Reference: C.F. Federsen (1), H. Tada (2)
2a
KI a FI ( ), , Y FI ( ) Method: Empirical formula based on Isida’s results
W
(1) Accuracy: +0.3% for 2a/W 0.7 and 1.0% for 2a/W=0.8
FI ( ) sec
2
or

(2) 2 4 Accuracy: Better than 0.2% for any value of

KI a F ( ) 1 0.025 0.06

(Y. Murakami et. al)

© 2010 Grzegorz Glinka. All rights reserved. 9

 The SIF geometry correction factor Y; KI = S Y; (central crack)
Geometry correction factor, FI( ) = Y
2a/W 0.000 0.001 0.002 0.003 0.004 0.005 0.006 0.007 0.008 0.009 0.010
0.200 1.025 1.026 1.026 1.026 1.026 1.027 1.027 1.027 1.028 1.028 1.058
0.210 1.028 1.028 1.029 1.029 1.029 1.029 1.030 1.030 1.030 1.031 1.031
0.220 1.031 1.031 1.032 1.032 1.032 1.032 1.033 1.033 1.033 1.034 1.034
0.230 1.034 1.034 1.035 1.035 1.035 1.035 1.036 1.036 1.036 1.037 1.037
0.240 1.037 1.037 1.038 1.038 1.038 1.039 1.039 1.039 1.040 1.040 1.040
0.250 1.040 1.041 1.041 1.041 1.042 1.042 1.042 1.043 1.043 1.043 1.044
0.260 1.044 1.044 1.045 1.045 1.045 1.046 1.046 1.046 1.047 1.047 1.047
0.270 1.047 1.048 1.048 1.049 1.049 1.049 1.050 1.050 1.051 1.051 1.051
0.280 1.051 1.052 1.052 1.052 1.053 1.053 1.054 1.054 1.054 1.055 1.055
0.290 1.055 1.056 1.056 1.056 1.057 1.057 1.058 1.058 1.059 1.059 1.059
0.300 1.059 1.060 1.060 1.061 1.062 1.062 1.062 1.062 1.063 1.063 1.064
0.310 1.064 1.064 1.064 1.067 1.066 1.066 1.066 1.067 1.067 1.068 1.068
0.320 1.068 1.069 1.069 1.070 1.070 1.071 1.071 1.072 1.072 1.072 1.073
0.330 1.073 1.073 1.073 1.074 1.075 1.075 1.076 1.076 1.077 1.077 1.078
0.340 1.078 1.078 1.078 1.079 1.080 1.080 1.081 1.081 1.082 1.082 1.083
0.350 1.083 1.083 1.083 1.085 1.085 1.086 1.086 1.087 1.087 1.088 1.088
0.360 1.088 1.089 1.089 1.090 1.090 1.091 1.092 1.092 1.093 1.093 1.094
0.370 1.094 1.094 1.094 1.096 1.096 1.097 1.097 1.098 1.098 1.099 1.100
0.380 1.100 1.100 1.100 1.101 1.102 1.103 1.103 1.104 1.104 1.105 1.106
0.390 1.106 1.106 1.107 1.107 1.108 1.109 1.109 1.110 1.111 1.111 1.112
0.400 1.112 1.112 1.113 1.114 1.114 1.115 1.116 1.116 1.117 1.118 1.118
0.410 1.118 1.119 1.120 1.120 1.121 1.122 1.122 1.123 1.124 1.124 1.125
0.420 1.125 1.126 1.126 1.127 1.128 1.128 1.129 1.130 1.131 1.131 1.132
0.430 1.132 1.133 1.133 1.134 1.135 1.136 1.136 1.137 1.138 1.138 1.139
0.440 1.139 1.140 1.141 1.141 1.142 1.143 1.144 1.144 1.145 1.146 1.147
0.450 1.147 1.148 1.148 1.149 1.150 1.151 1.151 1.152 1.153 1.154 1.155
0.460 1.155 1.155 1.156 1.157 1.158 1.159 1.159 1.160 1.161 1.162 1.163
0.470 1.163 1.164 1.164 1.165 1.166 1.167 1.168 1.169 1.170 1.170 1.171
0.480 1.171 1.172 1.173 1.174 1.175 1.176 1.176 1.177 1.178 1.179 1.180
0.490 1.180 1.181 1.182 1.183 1.184 1.185 1.186 1.186 1.187 1.188 1.189
0.580 1.277 1.279 1.280 1.281 1.283 1.284 1.285 1.287 1.288 1.289 1.291
0.590 1.291 1.292 1.293 1.295 1.296 1.297 1.299 1.300 1.302 1.303 1.304
0.600 1.304 1.306 1.307 1.309 1.310 1.311 1.313 1.314 1.316 1.317 1.319
0.610 1.319 1.200 1.322 1.323 1.325 1.326 1.328 1.329 1.331 1.332 1.334
0.620 1.334 1.335 1.337 1.338 1.340 1.342 1.343 1.345 1.346 1.348 1.350
0.630 1.350 1.351 1.353 1.354 1.356 1.358 1.359 1.361 1.363 1.364 1.366
0.640 1.366 1.368 1.370 1.371 1.373 1.375 1.376 1.378 1.380 1.382 1.383
0.650 1.383 1.385 1.387 1.389 1.391 1.392 1.394 1.396 1.398 1.400 1.402
© 2010 Grzegorz Glinka. All rights reserved. 10
 Example
A thick center-cracked plate of a high strength aluminum alloy is 200 mm wide and contains a crack
of length 80 mm. If it fails at an applied stresses of 100 MPa, what is the fracture toughness of the
alloy? What value of applied stress would produce fracture for the same length of crack in:
a) an infinite plate
b) a 120 mm wide plate?

2a

2W

© 2010 Grzegorz Glinka. All rights reserved. 11

a ) Finite width plate c) Plate 120 mm wide
2W 200 mm, 2a 80 mm, 100 MPa 2a a 40
0.6666
a W handbook W example 60
K a Y; Y ; f
W 2a
Y 0.6666 1.413
See notation in the Handbook : W
2a 2a a Kc K; 39.42 0.04 1.413
W 2W W
handbook example example
39.42
2a a 40
78.7 MPa
0.4 0.04 1.413
W handbook W example 100
2a
Y 0.4 1.112
W
K a Y 100 0.04 1.112 39.42 MPa m
K Kc 39.42 MPa m

b) Ininitely wide plate

K a Y; and Y 1
Kc K; 39.42 0.04
39.42
111.20 MPa
0.04

© 2010 Grzegorz Glinka. All rights reserved. 12

 Geometry Effects on the Stress Intensity Factor
Stress Intensity factors for cracks in a butt weldment and flat plate of the same thickness

Y S S

(a)
KI a (x)=S

S a a
t t

0 a/t 0.5

S y S
Y
a
(b)
KI a (x)
x
S a 0
t t
(a)
S
0 a/t 0.5
© 2010 Grzegorz Glinka. All rights reserved. 13
The Weight Function method for calculating Stress Intensity Factors
1 3
1
2 P1 x1 2 x1 x1 2
K AP1 m x1 , a, P1 1 M1 1 M2 1 M3 1
2 a x1 a a a
1 3
1
S 2 P2 x2 2 x2 x2 2

y K AP2 m x2 , a, P2 1 M1 1 M2 1 M3 1
2 a x2 a a a

x2

(x)
x1
x
K AP1 P2
m x1 , a , P1 m x2 , a , P2
A
P1 P2
a
a
t x
KA x m x , a dx
0

S
© 2010 Grzegorz Glinka. All rights reserved. 14
Geometrical parameters and notation for weight functions
y y

P
x P
x x
A x x
A

a 2a
W 2W

1 3
1
2 x 2 x x 2
K AP1 m x, a, P 1 M1 1 M2 1 M3 1
2 a x a a a

© 2010 Grzegorz Glinka. All rights reserved. 15

Central through crack in a finite width plate subjected to symmetric loading
2 3 4 5 6 7
a a a a a a a
M1 0.06987 0.40117 5.5407 50.0886 200.699 395.552 377.939 140.218
w w w w w w w
2 3 4 5 6
a a a a a a
M2 0.09049 2.14886 22.5325 89.6553 210.599 239.445 111.128
w w w w w w
2 3 4 5 6 7
a a a a a a a
M3 0.427216 2.56001 29.6349 138.40 347.255 457.128 295.882 68.1575
w w w w w w w

Edge crack in a finite width plate

2 3 4 5 6 7
a a a a a a a
M1 0.0719768 1.51346 61.1001 1554.95 14583.8 71590.7 205384 356469
w w w w w w w
8 9 10
a a a
368270 208233 49544
w w w
2 3 4 5 6 7
a a a a a a a
M2 0.246984 6.47543 176.457 4058.76 37303.8 181755 520551 904370
w w w w w w w
8 9 10
a a a
936863 531940 127291
w w w
2 3 4 5 6 7
a a a a a a a
M3 0.529659 22.3235 532.074 5479.53 28592.2 81388.6 128746 106246
w w w w w w w
8
a
35780.7
w

© 2010 Grzegorz Glinka. All rights reserved. 16

 The Weight Function method for calculating
Stress Intensity Factors
y S y S y

(x) (x)

x x x

a a a
t t t

S S
a) b) c)

Ka = Kb
The Stress Intensity Factor for any loading case is equal to the stress intensity
factor obtained by applying to the crack faces the stresses that used to be there
when there was no crack.
© 2010 Grzegorz Glinka. All rights reserved. 17
 Stepwise Procedure for the Stress Intensity
Calculation using the Weight Function Method
1. Calculate stress distribution (x) in the prospective crack plane in the
absence of the crack (un-cracked body, linear elastic analysis).

x f 0 ,x
2. Apply the stress distribution (x) to the crack surface as tractions.

3. Choose appropriate weight function, i.e. parameters M 1, M 2 and M3.

1/ 2 1 3/ 2
2 x x x
m ( x, a) 1 M1 1 M2 1 M3 1
2 a x a a a

4. Integrate the product of the stress distribution (x) and the weight function m(x,a).

K x m ( x , a ) dx
a
© 2010 Grzegorz Glinka. All rights reserved. 18
 The superposition principle for calculation of stress
a) intensity factors using the weight function approach;
T2

a) stress distribution in the prospective crack plane in the un-cracked body;

b) the “un-cracked stress field” applied to the crack surfaces of identical
r body with a crack;

t Prospective crack plane (y)

S

y o
b) a
T2
a
KI m y, y dy
t
crack 0
x

t (y)

y
© 2010 Grzegorz Glinka. All rights reserved. 19
Through the plate thickness stress distributions in a T-butt weldment
obtained for r/t = 1/25, = 45o (in the weld toe cross section)

FEM

n
(y) /

y/T
© 2010 Grzegorz Glinka. All rights reserved. 20
 Geometrical Stress Intensity Correction Factor “Y” for
an Edge Crack Emanating from the Weld Toe
(Comparison of WF and FEM data)

T-butt welded joint; Tension loading T-butt welded joint; Bending load

© 2010 Grzegorz Glinka. All rights reserved. 21

 Calculation of SIF for cracks
at notches using the weight
(x) (x) functions for edge and
through cracks

d b b b

(x) (x)
(x)

b b b
a 2a
t 2t

© 2010 Grzegorz Glinka. All rights reserved. 22

 Fracture Mechanics Approach to Fatigue Crack
Growth Analysis

•Fatigue crack growth equations

•Integration of fatigue crack growth expressions

•The effect of the initial crack size

•The effect of the weld geometry

•Residual stress effect

•Example

© 2010 Grzegorz Glinka. All rights reserved. 23

Fatigue Crack Growth Micro-Mechanism
A sharp crack in a tension stress field causes a high stress
concentration at the its tip resulting in slip and plastic deformation in
the crack tip vicinity. The material above and below the crack tip may
a
slip along a favorable slip plane in the direction of maximum shear
stress.

(c)

Stress
b
(d)
(b)

0 (e)
(a) Time
c

d
(R.Pelloux, ASTM,
STP 415, 1967)
(C. Laird, ASTM,
STP 415, 1967)
a
e
© 2010 Grzegorz Glinka. All rights reserved. 24
Experimental data for the determination of the fatigue crack growth curve

S1 > S2 > S3 af , Nf
Crack length, a

a
S S2 S N
1 3 Applied nominal stress history

da/dN= a/ N
a0

Stress, S
0 Number of applied cycles, N

S
S - Experimental test data
0 Time, t
Crack length, a

The ‘a vs. N’ data is obtained in

practice by periodic measurement of
ai, Ni the crack length, a, together with the
S
da/dN number of cycles, N. The raw data is
usually given in the form of series of
a0
points as shown in the figure.
0 Number of applied cycles, N
© 2010 Grzegorz Glinka. All rights reserved. 25
 The Fracture Mechanics approach to fatigue or the da/dN - K method is a technique based on the
analysis of fatigue crack growth. The combination of load/stress and geometry parameters, necessary
for the quantification of damage due to crack growth, is represented by the stress intensity factor, K,
in the case of monotonic load and by the range of the stress intensity factor, K, in the case of cyclic
loading.
The fatigue material properties are characterized by the threshold stress intensity range, Kth, the
fatigue crack growth rate relationship, da/dN vs. K, and the critical stress intensity factor, Kc, to be
often the same as the fracture toughness, KIc. The crack growth rate is then described by an
expression being function of the stress intensity range:

da
f K
dN ,

The stress intensity range associated with a stress cycle is calculated as:

K K max K min S max a Y S min a Y

where – a is the crack size, Smax and S min is the maximum and minimum nominal (or reference) stress
respectively, characterizing a stress cycle, and Y is the geometry correction factor. The aim of the
final analysis of the da/dN- K data is to determine necessary constants and parameters appearing in
expression f( K).
It should be noted that the ‘da/dN - K’ curve in fracture mechanics represents the material fatigue
resistance similarly to the S-N curve in the nominal stress approach or the ‘ - N’ relationship in the
local strain-life methodology.
As soon as the crack growth curve for the material of interest is known the fatigue life of the structural
component can be determined as shown in the figure below.
© 2010 Grzegorz Glinka. All rights reserved. 26
 The notation for the cyclic stress history parameters and the steps necessary for the determination of
the da/dN - K relationship are explained later in the following sections of the notes.

The fatigue life in terms of the number of cycles necessary to propagate the crack from its initial size,
a0, to the final or critical crack size, af, is determined by integrating the crack growth equation.

af af

da da
N
f K f S a Y
a0 a0

The determination of the integral above needs a numerical treatment because the geometry
correction factor, Y, becomes frequently a complex function of the crack size, a.

Subsequent stages of the fatigue life prediction method based on the crack growth analysis are
shown graphically in the Figure.

© 2010 Grzegorz Glinka. All rights reserved. 27

 Constant Amplitude Cyclic Load - Notation

Smax

Sa
Stress, S

S
Sm

Smin
1 cycle

0 Time, t

Smin - minimum stress

Smax - maximum stress
S = Smax- Smin - stress range
Sa = S /2 = (Smax- Smin )/2 - mean stress
R = Smin/ Smax - stress ratio

© 2010 Grzegorz Glinka. All rights reserved. 28

 Fatigue Crack Growth Rate vs. Stress Intensity Factor

S
K S a Y
and

K S a Y

© 2010 Grzegorz Glinka. All rights reserved. 29

 Scatter of fatigue crack
growth data; Low alloy steel
18G2VA

4 mm thick plate 4 mm thick welded plate

© 2010 Grzegorz Glinka. All rights reserved. 30

 The Fatigue Crack Growth Expression – The Paris equation
The first mathematical relationship relating fatigue crack growth rate and the stress intensity range was
proposed by Paris and Erdogan. This relationship is up to date the most popular mathematical expression used
if various fatigue/fracture mechanics analyses. It was obtained by fitting power law curve into the experimental
data.
da m
C K
dN
Where: da/dN - fatigue crack growth rate [in/cycle or m/cycle]
C- Paris’ equation parameter (valid for given R)
m- Paris’ equation exponent
K- stress intensity range

K K max K min for K min 0

K K max for K min 0
Where:
a - crack length/depth
max K max S a Y
Smax - maximum stress in a stress cycle
Smin - minimum stress in a stress cycle
min K min S a Y
Kmax - maximum stress intensity factor
Kmin - minimum stress intensity factor
Y - geometry correction factor in the stress intensity factor expression

© 2010 Grzegorz Glinka. All rights reserved. 31

 Complete Fatigue Crack Growth Rate Curve, da/dN - K
Soon after the Paris equation gained wide acceptance as a tool for fatigue crack growth analysis, it was
found that the simple expression proposed by Paris and Erdogan had some limitations. As the Figure
below illustrates the complete log-log plot of da/dN vs. K is sigmoidal rather then linear and limited by the
threshold stress intensity range, Kth, and the critical stress intensity factor Kc.
At low growth rates, the da/dN vs. K curve becomes steep and appears to approach a vertical asymptote
denoted Kth, which is called the fatigue threshold stress intensity range or fatigue crack growth threshold.
This quantity is interpreted as a lower limiting value of the stress intensity factor range K below which
fatigue crack growth does not ordinarily occur. The fatigue crack growth threshold is analogous to the
fatigue limit in the S-N approach.
At high growth rates, the da/dN vs. K curve may again become steep. This is due to rapid unstable crack
growth just prior to final fracture when Kmax Kc. The increase of the fatigue crack rate near the critical
stress intensity factor Kc is due to mixture of static (monotonic -fracture) and fatigue mechanisms driving
the crack growth.
Also, the fatigue crack growth rate exhibits a dependence on the stress ratio ‘R’. The stress ratio R affects
the fatigue crack growth rate in a manner analogous to the effects observed in the S-N and -N methods,
i.e. for a given K, increasing R-ratio increases the fatigue crack growth rate, and vice-versa.
The effect of the R -ratio (or mean stress) on Fatigue Crack Growth is most often explained using the
phenomenon discovered by Elber. By measuring the compliance of specimens with fatigue cracks he
noticed that the crack tip got closed during the descending part of the stress cycle in spite of the fact that
the applied stress/load remained tensile (see Figure). Elber postulated that crack closure decreases the
fatigue crack growth rate by reducing the effective stress intensity range.

© 2010 Grzegorz Glinka. All rights reserved. 32

 K, Stress Intensity Range, [ksi in]
5 10 20 50 100 3
10
Fatigue crack growth
2
10
A533B-1 steel, rates for a ductile
u=627 MPa

10 4
pressure vessel steel
Crack Growth Rate, da/dN [mm/cycle]

(the Paris equation)

da/dN. Crack Growth Rate, da/dN [inches/cycle]

(c)
3
10

10 5
da m
10 4
(b)
C K
K th
dN
6
da m
10
C K
10 5 dN The da/dN- K curve is the
fatigue material curve
7

R=0.1
10 independent of the geometry,
10 6 i.e. the same curve for all
(a) geometrical crack-body
10 8 configurations!
7
10
5 10 20 50 100 Source: N. Dowling, ref(2)
K, Stress Intensity Range, MPa m
© 2010 Grzegorz Glinka. All rights reserved. 33
 For simplicity reasons the
complete fatigue crack
growth rate is usually
approximated by three
linear pieces with the two
of them being vertical
limiting asymptotes.

© 2010 Grzegorz Glinka. All rights reserved. 34

 Paris’ equation constants for steel materials at R = 0

Ferritic-Pearlitic Steel:
ys
252
10-6 273 da 12 3.0
392 6.9 10 K
Crack Growth Rate, m/cycle

415
dN
Martensitic Steel:
da 10 2.25
10-7 1.4 10 K
dN
Austenitic Stainless Steel:
da 12 3.25
5.6 10 K
10-8 dN
5 10 100
K, MPa m for: da/dN in [m/cycle] and K in [MPa m]

J. Barsom, “Fatigue Crack Propagation in Steels of Various Yield Strengths”

Journal of Engineering for Industry, Trans. ASME, Series B, Vol. 93, No. 4, 1971, 1190-1196

© 2010 Grzegorz Glinka. All rights reserved. 35

 Estimation of the Fatigue Crack Propagation Life
Basic Steps:
1. Estimate the initial crack size and shape, ao;
- non-destructive testing - ao
- proof load - ao
2. Estimate the critical crack size ac based on the fracture toughness KIC, i.e. the crack size
that the component will tolerate when the applied stress reaches its maximum Smax.

2
1 K IC
K IC S max ac Yc ac
Smax Yc
3. Using the same expression for the stress intensity factor calculate the stress intensity
range K.

K S aY for R 0
K Smax aY for R 0 (if r 0!!)

© 2010 Grzegorz Glinka. All rights reserved. 36

 4. Substitute K into fatigue crack growth equation (Paris or Forman)

da
C ( K )m C ( S a )m Y m
dN
5. Integrate the equation above from a = ao to a =ac and determine the number of
cycles, N, necessary to grow the crack from the initial crack size of ao to the critical
size of ac. This is the estimated fatigue crack propagation life of given component!

da
dN
C ( K )m
ac ac
da da
N
C ( K )m C( S a Y )
ao ao

Note! In most practical cases the integration requires numerical solution due to the
complexity of the geometric factor Y.

© 2010 Grzegorz Glinka. All rights reserved. 37

 Integrated Paris’ Equation for a Constant Geometric
Factor, Y = const.
da m
C K C( S aY ) m
dN

for m 2
2 1 1
N ;
( m 2)C ( SY ) m m/2
ao ( m 2) / 2
ac ( m 2) / 2

for m 2
1 ac
N 2 2
ln ;
C S Y ao
© 2010 Grzegorz Glinka. All rights reserved. 38
 Numerical Integration of the Paris Equation
If the Y factor is not constant a numerical technique has to be applied. The most
often used is the cycle by cycle technique based on the calculation of crack
increments ai corresponding to each load cycle. In this case, the infinitesimal
increments da and dN are replaced by finite differences a and N= 1.
N
ai m m
C ( Ki ) C ( Si ai 1 Yi 1 ) ; ai ao ai ; ai C ( Si ai 1Yi ) m Ni
Ni i 1

N0 0 a0 0 a0 a0 ;
m
N1 1; N1 1; a1 C S1 a0 Y0 ; a1 a0 a1 ;
m
N 2 1; N 2 1; a2 C S2 a1 Y1 ; a2 a1 a2 ; Calculations have to be
m
carried out for each cycle !!
N 3 1; N 3 1; a3 C S3 a2 Y2 ; a3 a2 a3 ;
m
N 4 1; N 4 1; a4 C S4 a3 Y3 ; a4 a3 a4 ;
...........................................................................................
m
Ni 1; Ni 1; ai C Si ai 1 Yi 1 ; ai ai 1 ai ;
until ai ac

© 2010 Grzegorz Glinka. All rights reserved. 39

 Subsequent stages of fatigue life prediction method based on the crack
growth analysis
Analysis of external forces acting on the structure and the component in question (a),
Analysis of internal loads in chosen cross section of a component ( b),
Selection of individual welded joints in the structure (c),
Identification of appropriate nominal or reference stress history (d),
Extraction of stress cycles (rainflow counting) or reversals from the stress history
(Fig.e),
Determination of the stress intensity factor (i.e. the factor Y) for postulated or existing
crack,
- indirect method (Fig.f):
analyze un-cracked weldment and determine the stress field, (x,y), in the
prospective crack plane; normalize the calculated stress distribution with respect to
the nominal or any other reference stress, i.e. (x,y)/ n,
choose appropriate weight function,
calculate stress intensity factor
determine the stress or displacement field near the crack, or the strain energy
release rate,
calculate stress intensity factor using.
Determination of crack increments for each stress cycle (Fig. h),
Determination of the number of cycles, N, necessary to grow the crack from its initial
size, a0, up to the final size, af.
A summary of necessary input data and procedures used in the, da/dN - K, approach
to fatigue life estimation is also presented in the Figure.
© 2010 Grzegorz Glinka. All rights reserved. 40
Example: A very wide SAE 1020 cold-rolled thin plate is subjected to constant amplitude uni-axial cyclic
loads that produce nominal stresses varying from Smax=200MPa (29ksi) to Smin=-50 MPa (-7.3ksi). The
monotonic properties for this steel are Y=630 MPa (91 ksi), uts=670 MPa (97 ksi), E=207000 MPa
(30000 ksi), Kc =104 MPa m (95 ksi in). What fatigue life would be attained if an initial through-thickness
edge crack existed and was 1 mm (0.04 in) in depth?
S
The fatigue crack growth data are:
Kth(r=0)=6 MPa m, and Paris’ equation
parameters C=6.9 10-12 and m=3.
Smax

Stress
time
0
Smin

W >> a

© 2010 Grzegorz Glinka. All rights reserved. 41

 A. What is the stress intensity factor expression?
Semi-infinite plate with an edge crack.

K max S max a Y S max a 1.12

B. Is Linear Elastic Fracture Mechanics

(LEFM) applicable?

Nominal stress level :

S max 0.8 Y 0.8 630 504 MPa YES !

Plastic zone size :

K max S max a Y 200 0.001 1.12 12.6 MPa m
2 2
1 K max 1 12.6
ry 0.0000635 m 0.0635 mm
2 Y 630
ry 0.0635 1
0.125 YES !
a 1 8
© 2010 Grzegorz Glinka. All rights reserved. 42
 C. The effective stress range

S Smax Smin for Smin 0

S Smax for S min 0
Smax 200 MPa and Smin 50 MPa
thus
S Smax 200 MPa
D. Is the Paris equation applicable?

Paris equation is valid for K K th !

Smallest K K 0 occurs for a a0 0.001m.
K0 S a0 Y 200 0.001 1.12 12.6 MPa m
K 0 12.6 K th 6 MPa m YES , Paris equation is applicable !

© 2010 Grzegorz Glinka. All rights reserved. 43

 E. What is the critical/final crack size?
Kc K final S max ac Y
2 2
1 Kc 1 104
ac 0.068 m 68 mm
E. Integration of the Paris equation Smax Y 200 1.12
Analytical integration is possible because Y= const.

da m
C K C( S aY ) m
dN
ac ac ac
da da 1 da
N m
C K C( S aY ) m C Sm m2
am 2 Y m
a0 a0 a0

for m 2 and Y const

2 1 1
N ;
( m 2) C ( S Y ) m m/ 2
a0 ( m 2) / 2
ac ( m 2) / 2

2 1 1
N
(3 2) 6.9 10 12 (200 1.12)3 3/ 2
0.001(3 2) / 2
0.068(3 2) / 2

1 1 1 1
4631 ( m 2) / 2 ( m 2) / 2
4631 4631 31.645 3.834 128792 cycles
a0 ac 0.0316 0.2608
© 2010 Grzegorz Glinka. All rights reserved. 44
 Fatigue Crack Growth under Variable Amplitude Loading:
the retardation effect

da m
C K
dN
da
f K ,K max
dN

© 2007 Grzegorz Glinka. All rights reserved. Page 45

46
 Evolution of the crack tip plastic zone ahead of
a fatigue crack & crack tip closure
A
Stress, S

Stress, S

Stress, S

Stress, S
B
D
C
0 0 0 Time, t 0
Time, t Time, t Time, t

Crack Crack

Crack tip Crack tip Extended crack Crack tip

open starts to close tip closure starts to open

Crack tip
plastic zone

Envelope of plastically
deformed material in the
wake of a growing crack
© 2010 Grzegorz Glinka. All rights reserved. 47
 The stress-strain evolution and the monotonic
and plastic zone ahead of a fatigue crack tip
2
A Kmax 1 K max
rp
at: A ys

Stress
K

ys

K Crack
x
B Kmin

0 Time rp Monotonic plastic zone

2
at: B
A 1 K

Stress
rc
Stress,

4 ys

Crack x

0 Strain,
ys

B rc Cyclic plastic zone

© 2010 Grzegorz Glinka. All rights reserved. 48

 The effect of the crack tip closure

Kmax K min
R
K max
Stress intensity factor, K

K opn

Keff
?
K max

K
Kth K eff
U
Kopn K
Kth,eff

Kth
or
1
U
Kmin 1 R

0 Time, t

© 2010 Grzegorz Glinka. All rights reserved. 49

 Fatigue Growth of Corner Cracks in a Lug Subjected to
a VA Loading History
F

2W

(x) (x)

r r
x x

d a+r

Experimental data from: Jong-Ho Kim, Soon-Bok Lee, Seong-Gu Hong, Int. Journal of Fatigue, vol. 40, 2003

© 2010 Grzegorz Glinka. All rights reserved. 50

 The VA Load/Stress History

The Lug Loading Histories, Pmax=21 kN

100% Clipping
20
90% Clipping
18
80% Clipping
16

14
Load/Force [kN]

12

10

0
0 500 1000 1500 2000
No. of Reversals

© 2010 Grzegorz Glinka. All rights reserved. 51

 Calculations vs. Experiment
Fatigue life for 80% and 100% Clipped Loading History+Load Shedding: Al7050 T7

25
80% Clipped Experimental

80% Corner+EdgeRestrained+Shedding_SP=A045B023q065
20
100% Clipped Experimental

100% Corner+EdgeRestrained+Shedding_SP=A043B0182q065
Crack length [mm]

15

10

0
0 200 400 600 800 1000 1200 1400 1600 1800 2000
Number of blocks

Experimental data from: Jong-Ho Kim, Soon-Bok Lee, Seong-Gu Hong, Int. Journal of Fatigue, vol. 40, 2003
© 2010 Grzegorz Glinka. All rights reserved. 52
 Crack Shape Evolution; quarter circular initial crack

Predicted vs. Measured Crack Shape Evolution in the Lug,

Predicted vs. Mesured Crack Shape Evolution in the Lug,
100% Clipping
80% Clipping;
10 10

9 9

8 8

7 7

Depth, b, [mm]
6 6
Depth, b [mm]

5 5

4 4

3 3

2 2

1 1

0 0
0 1 2 3 4 5 6 7 8 9 10 0 1 2 3 4 5 6 7 8 9 10
Surface, a [mm] Surface, a, [mm]

Experimental data from: Jong-Ho Kim, Soon-Bok Lee, Seong-Gu Hong, Int. Journal of Fatigue, vol. 40, 2003

© 2010 Grzegorz Glinka. All rights reserved. 53

Weight Function for Arbitrary Planar Cracks
KA m Ai ( x , y; P )
i

P 2 c b
1 2
i Pj Ai c

bi

2 Ai Ai
PjAi

n
Pj(x,y)
n i
Pj(x,y) 1

ci A1
c-inverted contour of
the crack front;
2
b-inverted contour of
the external boundary;

3 A2 - distance between
PjAi
the point load P and point
A on the crack front
3
A3

© 2010 Grzegorz Glinka. All rights reserved. 54

 Location of maximum
shear stress

D=18.5 mm, d=4 mm

Fig. 10. Geometry and dimensions of the spring

© 2010 Grzegorz Glinka. All rights reserved. 55

 Relative dimensions of the inclusion (d=20-30 m) and the final
crack size (2af= 700 m)

© 2010 Grzegorz Glinka. All rights reserved. 56

 2-D Stress Field in the Spring Critical Cross Section and the Location
of the Initial Crack (non-metallic inclusion)
D
16 F
16T 2 8 FD 8F
ref torque 4.625
d3 d 3
d3 d2

C 1.55 ref ref

1.0
B ref 0.4 ref

B 0 time, t
C

0
0.14 D 1.0 ref
0 ref

A 0.85 ref A D

© 2010 Grzegorz Glinka. All rights reserved. 57

 0.25
Edge of the Boundary
cross section Start
0.2 1000000
2000000
3000000
4000000
0.15 5000000
5100000
5200000
5250000
0.1
Nf = 5,296,900 cycles 5270000
5280000
5290000
0.05 5295000
5296000
5296900

0
-0.25 -0.2 -0.15 -0.1 -0.05 0 0.05 0.1 0.15 0.2 0.25

-0.05
0.5

0
-3 -2 -1 0 1 2 3

-0.5 -0.1
-1

-1.5

-0.15
-2

-2.5

-3 -0.2
-3.5

-4

-0.25

Fatigue crack growth; d=0.03x0.02 mm, depth 0.25 mm, C, max= 1030 MPa, C, min= 390 MPa
© 2010 Grzegorz Glinka. All rights reserved. 58
 Main steps in fatigue design – flow chart

Start
1 2 3

Selecting criteria Simulation

Collecting data

3
5
The
4 End !

Data analysis Designing Stress relive!

© 2010 Grzegorz Glinka. All rights reserved. 59