Effect of Crack Front Curvature on CMOD

Compliance and Crack Length Evaluation for

Singleedge Bend Specimens

Z. Yan and W. Zhou

Department of Civil and Environmental Engineering
The University of Western Ontario, Canada

presented by

Zijian Yan
[email protected]

Introduction
Objective
Methodology
Results and Discussion
Conclusion
Outline
2
Fracture mechanics is the theory that describes the behavior of
bodies containing cracks and sharp notches, and is one of the most
important development in the theory of mechanics.
Linear Elastic Fracture Mechanics and Elastic-Plastic Fracture
Mechanics
A path-independent contour integral proposed by Rice (1968)
known as J-integral or J
Introduction
3
i
i
u
J wdy T ds
x
I
c
| |
=
|
c
\ .
}
Introduction
4
For brittle materials, when J reach the critical value, J
c
, the fracture
behavior results in the development of a rapid and unstable crack
extension.
For more ductile (tougher) materials, the ductile fracture behavior
can lead to a slow and stable crack extension.
Fracture toughness (resistance) can typically increase as the crack
length grows. The relationship between J and the crack extension,
called J -Resistance or J -R curve, is used to characterize the
fracture toughness property of ductile material.
Introduction
5
J -Resistance or J -R curve is an important input in the integrity
assessment and strain-based design of energy pipelines.

Introduction
6
J -Resistance or J -R curve is typically measured on small scale
specimens, e.g. SE(B), C(T) and SE(T) specimens.
Introduction
7
Single-edge bend or SE(B) specimen
Crack Mouth Opening Displacement (CMOD)
Introduction
8
Why the specimen crack front can be curved ?
All the machine notched crack need to be fatigue pre-cracked
to simulated the natural narrow crack, as specified in test
standards, e.g. ASTM E1820, BS 7448 and ISO 12135.
Real crack after fatigue pre-cracking
Introduction
9
Why the specimen crack front can be curved ?
All the machine notched crack need to be fatigue pre-cracked
to simulated the natural narrow crack, as specified in test
standards, e.g. ASTM E1820, BS 7448 and ISO 12135.
What does the real crack front look like ?
Introduction
10
How is the crack front curvature considered in test standards ?
Test standards usually specify the allowable deviation of a curved crack front
from a straight front based on the so-called nine-point measurement method
instead of evaluating the impact of the curvature on the J-R curve.
a
i
(i = 1, 2, , 9)

=
1
8

1
+
9
2
+

=8
=2

Introduction
11
ASTM E1820 requires that none of the nine physical measurements of the
initial (final) crack size differ by more than 0.05 times specimen thickness
from the average initial (final) crack length a
ave
obtained from the nine
measurements.
Test specimens that do not meet this criterion are deemed unacceptable and
therefore rejected.
= max(a
max9
- a
ave
, a
ave
- a
min9
)/B 0.05
a
i
(i = 1, 2, , 9)

=
1
8

1
+
9
2
+

=8
=2

Introduction
12
Single specimen unloading compliance method (Clarke et al., 1976)
J - R curve
0
400
800
1200
1600
2000
2400
2800
0 1 2 3 4
J

(
k
J
/
m
2
)
Aa (mm)
P - CMOD curve
=

=

2
1
2

Intermediate crack length, a =
f (elastic unloading compliance, C)
Introduction
13
The relationship between crack length, a and CMOD compliance,
C, for SE(B) specimens (Wu (1984) and Joyce (1992)):
= 0.999748 3.950 + 2.9821
2
3.21408
3
+51.5156
4
113.031
5

= (1.01878 4.5367 +9.0101
2
27.333
3
+74.400
4
71.489
5
)
for deeply-cracked SE(B) specimens (0.45 a/W < 1)
for shallow-cracked SE(B) specimens (0.05 a/W < 0.45)
where =
1
4

/
1/2
:1

These equations are all derived based on 2D analysis, and the
crack front curvature has never been considered.
E

- plane strain
E - plane stress
E
e
3D (Wang et al. 2013)
Objective
14
Investigate the effect of crack front curvature on :
CMOD compliance, C
predicted crack length, a from the CMOD compliance
experimentally determined J using the formulas derived
from 2D analysis (ASTM E1820)
Yan, Z. and Zhou, W. (2014). Effect of Crack Front Curvature on
Experimental Evaluation of J-integral for Single-edge Bend Specimens.
Journal of Testing and Evaluation, accepted and in press.
Methodology
15
Characteristics of Curved Crack Front
Symmetric
power-law expression
0 1 -1
a/W
2x/B
() = (0)

2

a(0) is the crack length at the mid-plane;
a(B/2) is the crack length at the free surfaces;
= a(0)/W - a(B/2)/W
Methodology
16
Nikishkov et al. (1999) - 110 specimens -
a fixed value of p = 3.0
Our experimental study - 3 specimens -
Wang et al. (2013)
CANMAT MTL - 14 specimens - Park et
al. (2011)
p = 3.0 (& 2.5)
() = (0)

2

() = (0)

2

Methodology
17
= a(0)/W - a(B/2)/W

=
1
8

1
+
9
2
+

=8
=2

= 0 5 0.25

2

+ 0.25

2

1
8
5

=8
=1


2

= max(a
max9
- a
ave
, a
ave
- a
min9
)/B 0.05
=
1

0.25

2

4

1
8
5

=8
=1

p = 3.0
a
ave
/W =0.3, 0.5, 0.7
=0.01 to 0.10
= 3.0
= 0.3, 0.5, 0.7
= 0.01 to 0.10
Methodology
18
Commercial Finite Element Program: ADINA 8.7.4
Linear Elastic FEA, E = 200GPa and v = 0.3
20-node 3D isoparametric brick elements with full integration
Three specimen thickness-to-width ratios: B/W = 1, 0.5 and 0.25
consistent with the range specified in ASTM E1820
Standard span-to-width ratio S/W = 4
Three different relative average crack lengths: a
ave
/W = 0.3, 0.5,
0.7
Curved crack fronts with = 0.01 to 0.10 with an increment of
0.01
Finite Element Analysis
Methodology
19
A quarter with 10 layers
Typically contains 11,000 elements
Finite Element Analysis
Results and Discussion
20
0.93
0.94
0.95
0.96
0.97
0.98
0.99
1
1.01
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1
SE(B) B/W = 0.25
a/W = 0.3
a/W = 0.5
a/W = 0.7
0.93
0.94
0.95
0.96
0.97
0.98
0.99
1
1.01
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1
SE(B) B/W = 0.5
a/W = 0.3
a/W = 0.5
a/W = 0.7
0.93
0.94
0.95
0.96
0.97
0.98
0.99
1
1.01
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1
SE(B) B/W= 1
a/W = 0.3
a/W = 0.5
a/W = 0.7
ASTM E1820 Criteria
Results and Discussion
21
Effect of crack front curvature on the predicted crack length, a
p
Error =(a
p
- a
ave
) / a
ave
Relative crack length a
ave
/W = 0.3, 0.5 and 0.7
Specimen thickness-to-width ratio B/W = 1, 0.5 and 0.25
a
p
predicted based on

E

- plane strain
E - plane stress
E
e
3D (Wang et al. 2013)
Results and Discussion
22
-3.0%
-2.0%
-1.0%
0.0%
1.0%
2.0%
3.0%
4.0%
5.0%
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1
E - plane stress
E' - plane strain
Effective modulus
/W = 0.3, B/W = 1, p = 3
ASTME1820-11 Limit
-3.0%
-2.0%
-1.0%
0.0%
1.0%
2.0%
3.0%
4.0%
5.0%
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1
E - plane stress
E' - plane strain
Effective modulus
/W= 0.5, B/W = 1, p = 3
ASTME1820-11 Limit
-3.0%
-2.0%
-1.0%
0.0%
1.0%
2.0%
3.0%
4.0%
5.0%
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1
E - plane stress
E' - plane strain
Effective modulus
/W= 0.7, B/W = 1, p = 3
ASTME1820-11 Limit
-1.0%
0.0%
1.0%
2.0%
3.0%
4.0%
5.0%
6.0%
7.0%
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1
E - plane stress, p = 3
E - plane stress, p = 2.5
E' - plane strain, p = 3
E' - plane strain, p = 2.5
Effective modulus, p = 3
Effective modulus, p = 2.5
/W = 0.3, B/W = 0.5, p = 3 & 2.5
ASTME1820-11 Limit
-3.0%
-2.0%
-1.0%
0.0%
1.0%
2.0%
3.0%
4.0%
5.0%
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1
E - plane stress, p = 3 E - plane stress, p = 2.5
E' - plane strain, p = 3 E' - plane strain, p = 2.5
Effective modulus, p = 3 Effective modulus, p = 2.5
/W = 0.5, B/W = 0.5, p = 3 & 2.5
ASTME1820-11 Limit
-3.0%
-2.0%
-1.0%
0.0%
1.0%
2.0%
3.0%
4.0%
5.0%
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1
E - plane stress, p = 3
E - plane stress, p = 2.5
E' - plane strain, p = 3
E' - plane strain, p = 2.5
Effective modulus, p = 3
Effective modulus, p = 2.5
/W = 0.7, B/W = 0.5, p = 3 & 2.5
ASTME1820-11 Limit
-1.0%
0.0%
1.0%
2.0%
3.0%
4.0%
5.0%
6.0%
7.0%
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1
E - plane stress
E' - plane strain
Effective modulus
/W = 0.3, B/W = 0.25, p = 3
ASTME1820-11 Limit
-3.0%
-2.0%
-1.0%
0.0%
1.0%
2.0%
3.0%
4.0%
5.0%
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1
E - plane stress
E' - plane strain
Effective modulus
/W = 0.5, B/W = 0.25, p = 3
ASTME1820-11 Limit
-3.0%
-2.0%
-1.0%
0.0%
1.0%
2.0%
3.0%
4.0%
5.0%
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1
E - plane stress
E' - plane strain
Effective modulus
/W = 0.7, B/W = 0.25, p = 3
ASTME1820-11 Limit
-3.0%
-2.0%
-1.0%
0.0%
1.0%
2.0%
3.0%
4.0%
5.0%
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1
E - plane stress
E' - plane strain
Effective modulus
/W = 0.3, B/W = 1, p = 3
ASTME1820-11 Limit
E
e
E

E

Conclusion
23
SE(B) specimen with a curved crack front satisfying the straightness
criterion as specified in ASTM E1820, i.e. 0.05, the corresponding
CMOD compliance differs by less than 1% compared with the
compliance of the specimen with a straight crack front.
Crack front curvature has a negligible impact on the predicted crack
length for all the considered values regardless of a
ave
/W and B/W ratios.
The use of E
e
can lead to more accurate prediction than E and E.
In J-R curve tests, the specimen crack front straightness criterion is
governed by the evaluation process of J-integral instead of crack
extension, as discussed in Yan and Zhou (2014).
Acknowledgements
24
Natural Sciences and Engineering Research Council of
Canada (NSERC)
TransCanada Pipelines Limited
Dr. G. Shen from CANMAT MTL, Hamilton, Ontario
25
Thanks!

[email protected]