Journal of Applied Mathematics and Physics, 2013, 1, 54-56
Published Online November 2013 (http://www.scirp.org/journal/jamp)
http://dx.doi.org/10.4236/jamp.2013.16011
Open Access JAMP
Effect Analysis of Gurson Model Parameters on Crack
Extension of Pipeline
Tieping Li, Xinlu Tian
Nuclear and Radiation Safety Center of MEP, Beijing, China
Email: everlasting_cat@sina.com, xinlu724@163.com
Received August 2013
ABSTRACT
Leakage-before-break technique is widely used in high energy pipeline of nuclear plant, for which crack stability of
pipeline under complex loading condition is a key issue, and crack growth resistance curve of pipeline material is the
important foundation for crack stability analysis. In this paper, ferritic steel A533B is studied, Gurson damage model is
used to simulate crack process of contact tension specimen under uni-tension, and effect of Gurson mod el parameter on
simulation result is discu ssed. The following results are found dur ing simulation: initial porosity factor
0
f
is the main
parameter, when it increases gradually, unstable crack extension will be observed; however, only initial J toughness is
affected by critical po rosity factor
N
f
; the minor parameter is loa d step control
α
, when it increases, stable and con-
vergent result is obtained. All results in this paper can be used to determine parameters in Gurson mode, which will be
foundation for cra c k extension anal y s i s of pi pe line.
Keywords: Leakage-Before-Break; Crack Growth Resistance Curve; Gurson Damage Model; Crack Extension
1. Introduction
In present, leakage-before-break (LBB) technique is
widely used for collant pipeline in most operating and
abuilding nuclear plants. In LBB technique, leakage can
be detected in pipeline before unstable crack extension
occurring in pipeline, and protective hardware and whip
restraints can be removed for convenience of inspection
operability, from which economic efficiency can be
guaranteed. Since austenitic steel with high toughness is
used for nuclear pipeline, elastic-plastic fracture is the
main failure mode. Elastic-plastic fracture is a complex
process, and fracture toughness
IC
J
cannot be adapted
directly to fracture analysis in complex structure [1].
Combination of fracture test and material model is pro-
posed by some researchers since fracture mechanism is
required by fracture process simulation [2], material
model parameters can be deduced from combination re-
sult [3], which can be used to simulate fracture process in
any complex structure.
Gurson model is proved to simulate fracture process
well for metallic material from a lot of tests and theoretic
studies [4]. Contact tension specimen made of ferritic
steel A533B is studied in this paper, Gurson model is
combined to describe micro fracture process in A533B
material, and effect of Gurson model parameter on simu-
lation results is analyzed by finite element method.
2. Gurson Model
Metallic material is divided into two parts in Gurson
model: basement material and porosity. In itial porosity is
assumed as
0
f
, when material is loaded, bigger poro sity
is formed by connection of smaller porosity, and new
crack surface is obtained at the same time. Material fail-
ure can be judged by the following equation:
( )
22
2
13
3
2 cosh10
2
em
q
g qfqf
σσ
σσ
 
=+−+ =
 
 
(1)
In Equation (1),
e
σ
is regarded as equivalent stress of
the whole material, m
σ
is hydrostatic pressure,
σ
is
equivalent stress of basement material,
f
is real-time
porosity rate of the whole material. When
f
is larger
than
N
f
, which is defined as the critic porosity, crack
extension will be found. In above equation, q1, q2, q3, is
aspect ratio of porosity, which can be deduced by
stress-strain curve of uni-tension test [5].
3. Simulation Analysis
3.1. Material Property
The stress-strain curve of ferritic steel A533B can be
described as the following relationship:
T. P. LI, X. L. TIAN
Open Access JAMP
55
1
0
0
N
E
σσ
εσ

=

(2)
In Equation (2),
0
σ
is initial yielding stress of ma-
terial, N is strain hardening index, E is young’s modulus,
and for material of A533B, parameters defined in Equa-
tion (2) are fixed a s
.
3.2. Finite Element Model
According to ASME fracture toughness test standard
E1820 [6], the geometric character of contact tension
specimen can be plotted as Figure 1: w is 50.8 mm as
width of specimen, B is 127 mm as depth of specimen,
and size of the remanent part can be deduced by the pro-
portion in Figure 1. Geometric size in Figure 1 is used
to build the finite element model.
Software Warp3d is used here for model creation and
element C3D8 is selected. Since plastic deformation is
included in simulation, crack tip singularity is not consi-
dered in this paper. Plot in Figure 2 is the half model of
contact tension specimen, one-layer element with Gurson
model property is assigned to the crack face, and the
elastic-plastic material relationsh ip in Equation (2) is still
suitable for the remanent part. In this paper, some para-
meters such as aspect ratio and Gurson element height
are fixed in simulation as
12 3
1.46, 0.931, 2.131qq q= ==
and
200Dm
µ
=
, where
D
is the height of element
with Gurson model property.
The boundary condition is: all node freedom in depth
direction is fixed; node freedom in vertical direction for
symmetric surface except crack face is fixed; in order to
remove rigid displacement of the whole model, one of
the nodes far from loading position and crack surface is
Figure 1. Schematic plot of contact tension specimen.
Figure 2. Finite element model of contact tension specimen.
fixed in horizontal direction. Displacement loading me-
thod is used.
3.3. Calculation Results
Software Warp3d is used to simulate fracture process of
contact tension specimen. Effects of initial porosity
0
f
,
critic porosity
N
f
and load control parameter
α
em-
bedded in Warp3d on fracture process are studied in si-
mulation.
1) Effect of load control parameter
α
on
J
value
For the simulation of this part, initial crack length ratio
is
/ 0.3aw=
, initial porosity rate is
0
0.005f=
, critic
porosity ratio is
0.2
N
f=
and only load control para-
meter is changeable. It is shown in Figure 3 that, curve
of
J
value vs. displacement loading is affected little by
load control parameter, and the specimen is perfect now
with no crack extension. As displacement
increasing,
curves are distinct with different
α
: smaller
α
value
is related with stable crack extension, such as there is a
platform in curve corresponding to
0.005
α
=
; curves
with larger
α
are related with unstable crack extension
since there is a downtrend in these curves.
2) Effect of critic porosity
For the simulation of this part, initial crack length ratio
is defined as
/ 0.5aw=
, load control parameter is de-
fined as
0.005
α
=
, initial porosity is defined as
0
0.003f=
, and only critic porosity
N
f
is changeable.
It is shown in Figure 4 that almost no effect of
N
f
on
result can be observed.
3) Effect of initial porosity
For the simulation of this partinitial crack length ratio
is defined as
/ 0.5aw=
, load control parameter is de-
fined as
0.005
α
=
, critic porosity is defined as
0.2
N
f=
, and only initial porosity
0
f
is changeable. It
is shown in Figure 5 that curves of
J
value vs. dis-
placement loading are coincident with smaller loading.
As loading
increases, yielding is found around crack
tip since smaller
0
f
(
0
0.001f=
) corresponding to
more perfect material, and there is an uptrend for
J
value because yielding zone is expanding. As
0
f
in-
creases, there is a downtrend in curve, since material
with larger porosity rate is fragile, and relaxative stress
can be observed around crack tip with crack surface ex-
tension.
T. P. LI, X. L. TIAN
Open Access JAMP
56
012345
0
50
100
150
200
α=0.005
α=0.01
α=0.02
α=0.05
J(N/m)
(mm)
Figure 3. Effect of load-control parameter on calculation
results.
010 20 30 40 50
0
50
100
150
200
250
J(N/m)
fN=0. 15
f N=0. 20
f N=0. 25
P(KN)
Figure 4. Effect of critic porosity on calculation.
012345
0
50
100
150
200
250
300
350
(mm)
J(N/m)
f 0=0.001
f 0=0.003
f 0=0.004
f 0=0.005
Figure 5. Effect of initial porosity on calculation results.
4. Conclusions
The fracture process of contact tension specimen made of
ferritic steel A533B is simulated by finite element me-
thod combining with Gurson damage model, and effect
of parameters in Gurson model on simulation result is
studied, from which the following conclusions can be
obtained:
1) The initial porosity rate is the leading role for simu-
lation result, when
0
f
increases, there is an unstable
trend for crack extension, and only
J
value in initial
loading period is affected by critic porosity rate
N
f
.
2) With decreasing load control parameter of
α
, sta-
ble and convergent crack extension can be observed.
5. Acknowledgements
This work was financially supported by Natural Key
Projects of China (Grant No. 2011ZX06002-010).
REFERENCES
[1] Z. R. Huang and Y. K. You, Analysis of the Load-Dis-
placement Relationship to Determine the JR Resistance
Curve of Carbon Steel Piping,” Journal of Jiangsu Insti-
tute Petrochemical Technology, Vol. 12, No. 3, 2000, pp.
1-3.
[2] T. L. Anderson, “Fracture Mechanics: Fundamentals and
Applications,” CRC Press, Boca Raton, 1994.
[3] F. Jonas, X. S. Gao and C. F. Shih, “Cell Model for Non-
linear Fracture Analysis: I. Micromechanics Calibration,”
International Journal of Fracture, Vol. 89, No. 4, 1998,
pp. 355-373.
http://dx.doi.org/10.1023/A:1007421420901
[4] X. Lin and C. F. Shih, “Ductile Crack Growth I: A Nu-
merical Study Using Computational Cells with Micro-
structurally-Based Length Scales,” Journal of the Me-
chanics and Physics of Solids, Vol. 43, No. 2, 1995 pp.
233-259.
http://dx.doi.org/10.1016/0022-5096(94)00064-C
[5] “WARP3D-Release 15.9,” University of Illinois, 2008.
[6] ASTM E1820-05a Standard Test Method for Measure-
ment of Fracture Toughness, ASTM International, 2008.