Ocean Engineering 128 (2016) 81–93

Contents lists available at ScienceDirect

Ocean Engineering
journal homepage: www.elsevier.com/locate/oceaneng

Progressive inelastic deformation of a girth-welded stainless steel pipe

under internal pressure and cyclic bending
crossmark
⁎
Wan-Gon Baea, Kyong-Ho Changb, Chin-Hyung Leec,
a
Department of Civil Engineering, Chung-Ang University, 84, Huksuk-ro, Dongjak-ku, Seoul 06974, Republic of Korea
b
Department of Civil and Environmental & Plant Engineering, Chung-Ang University, 84, Huksuk-ro, Dongjak-ku, Seoul 06974, Republic of Korea
c
The Graduate School of Construction Engineering, Chung-Ang University, 84, Huksuk-ro, Dongjak-ku, Seoul 06974, Republic of Korea

A R T I C L E I N F O A BS T RAC T

Keywords: This study aims to characterize numerically the ratcheting behavior of a girth-welded straight stainless steel
Pressurized girth-welded stainless steel pipe pipe in combined action of internal pressure and cyclic bending loading. Finite element (FE) thermal simulation
Weld-induced residual stresses of the girth butt welding process is first performed to identify weld-induced residual stresses. Three-
Cyclic bending dimensional (3-D) elastic-plastic FE analyses incorporated with the cyclic plasticity constitutive model capable
Ratcheting
of describing the cyclic plastic performance are next conducted to scrutinize the local (circumferential strain)
Cyclic plasticity constitutive model
and global (cross-section diameter change) ratcheting responses of the girth-welded stainless steel pipe under
Finite element analysis
internal pressure and cyclic bending, which take the residual stresses and plastic strains obtained from the
preceding thermal simulation as the initial condition. The analytical results demonstrate that welding residual
stresses in combination with the internal pressure have significant effects on the hoop strain rate and the in-
plane and out-of-plane diameter changes, and the degree and shape of the ovalization which occurs during the
multiaxial ratcheting are dependent on the applied loads.

1. Introduction
Pressurized steel pipes which are the most basic elements in
offshore pipelines are subject to cyclic loading induced by extreme
weather, earthquake and wave, etc. One of the major concerns on these
pipes is ratcheting. In a component under a primary load with
secondary cyclic stressing that exceeds the elastic limit of the material,
progressive accumulation of plastic strain occurs, which is called
ratcheting even though the origin of the cyclic permanent strain
accumulation remains not completely known (Taleb and Cailletaud,
2011). The inelastic strain accumulation combined with the fatigue
damage during cyclic loading may result in reduction of the fatigue
crack initiation life and thus the fatigue life of the component (Rahman
et al., 2008). Ratcheting leads to larger in-grain misorientation in
microstructure compared to fatigue, which is attributed to the larger
accumulated plastic strain during ratcheting (Paul et al., 2015).
Ratcheting is therefore one of the most critical structural problems to
be investigated in a pressurized piping component subjected to cyclic
loading. In most piping networks, connection of the pipes is mainly
implemented by girth welding. In a girth-welded steel pipe, the
presence of unavoidable welding residual stresses is well known
(Karlsson and Josefson, 1990; Brickstad and Josefson, 1998; Yaghi
et al., 2006; Deng and Murakawa, 2006; Lee and Chang, 2011a, 2014;
Lee et al., 2013a), which are induced as a result of plastic strains
caused by solidification, phase transformation and circumferential
shrinkage during welding. Weld-induced residual stresses increase
the vulnerability to stress corrosion cracking, fatigue damage and
brittle fracture (Withers, 2007). It is also well recognized that
mechanical behavior of a girth-welded pipe under monotonic loading
is significantly affected by welding residual stresses (Lee and Chang,
2011b, 2013a, 2013b; Lee et al., 2013b, 2014a), i.e. the girth weld-
induced residual stresses cause premature yielding and loss of stiffness
and eventually lead to deterioration of the load-carrying capacity.
However, the residual stress effect on the ratcheting response of a
girth-welded steel pipe subjected to internal pressure and cyclic loading
remains unclear.
In last three decades, a large number of research works have been
dedicated to understand the cyclic plastic behavior. Along with the
experimental studies (Hassan and Kyriakides, 1992, 1994a, 1994b;
Hassan et al., 1992; Portier et al., 2000; Taleb and Hauet, 2009; Paul
et al., 2012; Song et al., 2014; Taleb et al., 2014) which have provided
useful information for the uniaxial/multiaxial ratcheting characteris-
tics, various constitutive models capable of simulating the cyclic plastic
response as accurate as possible have been developed (Armstrong and

⁎
Corresponding author.
E-mail address: ifinder11@gmail.com (C.-H. Lee).

http://dx.doi.org/10.1016/j.oceaneng.2016.10.027
Received 17 March 2016; Received in revised form 29 July 2016; Accepted 13 October 2016
Available online 20 October 2016
0029-8018/ © 2016 Elsevier Ltd. All rights reserved.
W.-G. Bae et al.
Nomenclature
b material constant
ck material constant
f yield function
k* initial size of elastic domain
m material constant
n unit normal vector to yield surface at current stress point
p accumulated plastic strain
ṗ accumulated plastic strain rate
H plastic modulus
Q saturated constant value of isotropic hardening
Frederick, 1966; Chaboche and Rousselier, 1983; Nouailhas et al.,
1985; Burlet and Cailletaud, 1986; Chaboche, 1986, 1991, 2008; Ohno
and Wang, 1993; Corona et al., 1996; Jiang and Sehitoglu, 1996;
Abdel-karim and Ohno, 2000; Bari and Hassan, 2000, 2002; Kang
et al., 2002; Chen and Jiao, 2004; Kang and Kan, 2007; Kang and Liu,
2008; Abdel-Karim, 2009; Feigenbaum et al., 2012; Lee et al., 2014b;
Do et al., 2015), most of which have generally been limited to the
materials level. It has been demonstrated that the ratcheting response
of a material has a significant dependence on the stress history, which
in a pipe relies on the external load and the geometry (Hassan et al.,
1998). Considering all these parameters experimentally is cost prohi-
bitive and impractical. Thus, finite element (FE) simulation based on
cyclic plasticity constitutive model should be employed to analyze the
global (diameter change) and local (circumferential strain) ratcheting
responses of a pressurized pipe for different cyclic loading patterns and
configurations. For the structural ratcheting behavior of piping com-
ponents such as straight pipes, elbows and branch pipes, many studies
have been conducted (Chen et al., 2013). Nevertheless, a small number
of research works so far have dealt with simulation of the plastic
response of a straight piping component under cyclic bending and
internal pressure. Hassan et al. Hassan et al., (1998) incorporated an
improved constitutive model, which consists of the Armstrong-
Frederick kinematic hardening rule (Armstrong and Frederick, 1966)
and the Drucker-Palgen plastic modulus equation (Drucker and Palgen,
1981), into ANSYS FE package and validated the modified code by
replicating the four-point displacement controlled cyclic bending tests
of straight stainless steel pipes at constant internal pressure and
correlating the analytical results with the experimental measurements.
Gao et al., (2006) performed ratcheting simulations of pressurized
straight low carbon steel pipes under cyclic quasi-three point bending,
which involved bending moment gradients and shear loads through the
pipe length, by using ANSYS program into which the Ohno-Wang
model and the modified Ohno-Wang models were incorporated. The
effects of bending load, internal pressure and loading history on the
multiaxial strain accumulation were discussed. But, only hoop ratchet-
ing strains were examined at various positions. Rahman et al., (2008)
evaluated the performance of several constitutive models in simulating
the structural ratcheting responses of straight pipes of alloy steel 4130
under combined curvature-symmetric cyclic bending and internal
pressure against the measured local and global ratcheting responses.
They claimed that the model parameters should be refined to improve
the cyclic plasticity modeling. However, they focused on the perfor-
mance evaluation of the constitutive models and hence limited load
and structural parameters were employed in their investigation. Zakavi
et al., (2010) used the Armstrong-Frederick model with isotropic/
kinematic hardening rule to anticipate the circumferential ratcheting
strains of pressurized straight carbon and stainless steel pipes sub-
jected to simulated seismic bending moment. The FE results were
compared with those obtained from the experiments and the capability
of the combined hardening model was evaluated. From the literature
review, it seems that ratcheting responses for internally pressurized
82
Ocean Engineering 128 (2016) 81–93
R isotropic hardening variable
α back stress tensor
α′ deviatoric back stress tensor
α̇′ deviatoric back stress increment tensor
α′k asymptotic value of the kth back stress
ε̇p plastic strain increment tensor
σ stress tensor
σ′ deviatoric stress tensor
σ̇′ deviatoric stress increment tensor
γk material constant
φ material constant
λ̇ plastic multiplier
piping components under cyclic bending have not been fully investi-
gated. Moreover, as for the ratcheting analysis of a girth-welded steel
pipe subjected to internal pressure and cyclic bending, very limited
works have been reported to date due to the truly complex analysis
procedure involved in welding and subsequent cyclic loading problems
and therefore deserves special attention. Actually, Lee et al., (2004)
evaluated the thermal ratcheting of welded cylindrical structure using
inelastic analysis that employed the Chaboche-Rousselier nonlinear
combined hardening model (Chaboche and Rousselier, 1983). They
assumed the residual stress distribution at the welded joint according
to the UK R6 procedure using axisymmetric FE model. However, the
residual stresses are by no means axisymmetric, i.e. they vary spatially
along the circumference due to the moving arc and welding start/stop
effects (Lee et al., 2013a; Lee and Chang, 2014). Thus, the exact
assessment of the residual stress effect on the ratcheting behavior could
not be achieved.
This study attempts to scrutinize the ratcheting responses of a
girth-welded straight stainless steel pipe under cyclic bending and
internal pressure through the numerical simulation. The following
approaches are taken in the present investigation: first, thermal
simulation of the girth butt welding process is first conducted to obtain
weld-induced residual stresses by using a sequentially coupled three-
dimensional (3-D) thermo-mechanical FE analysis model developed by
the authors; second, a cyclic plasticity constitutive model which can
simulate the multiaxial ratcheting behavior is presented and verified
against the test data, and third, parametric comparative studies in
which the ratcheting responses of the girth-welded stainless steel pipe
exposed to internal pressure and cyclic bending are explored taking the
residual stresses and plastic strains as the initial condition are carried
out by using a 3-D elastic-plastic FE analysis method which incorpo-
rates the cyclic plasticity model as the material constitutive equation.
The ratcheting behavior of the pressurized steel pipe in cyclic bending
without considering the residual stresses is also investigated in order to
clarify the relevance of weld-induced residual stresses to the cyclic
responses.
2. FE thermal simulation of the girth butt welding process
FE thermal simulation of the girth butt welding process should first
be performed to attain weld-induced residual stresses and plastic
strains, which are required input to the mechanical model for analyzing
the effects that the residual stresses have on the ratcheting responses of
the pressurized girth-welded stainless steel pipe under cyclic bending.
Welding process is essentially a coupled thermo-mechanical process.
The thermal histories strongly affect the stress fields, whilst the
mechanical fields have a weak influence on the temperature profiles.
Therefore, in this work, a sequentially coupled 3-D thermal-mechanical
FE analysis model able to precisely capture the 3-D feature of welding
residual stress distribution in the girth-welded stainless steel pipe,
which was developed by the authors (Lee and Chang, 2014) based on
the in-house FE-code (Lee, 2005) and its accuracy was confirmed
W.-G. Bae et al.
against the experimental measurements, was employed to gain weld-
induced residual stresses. The model comprises two parts: a transient
heat transfer model followed by a transient stress model. The thermal
model identifies the temperature and phase evolution as a function of
time, and the structural model utilizes the previous results to calculate
displacements at nodes and stresses at integration points. The forth-
coming section briefly describes the thermo-mechanical FE analysis
model and the specific details can be found in (Lee and Chang, 2014).
2.1. Thermal and mechanical models
The thermal model is based on the heat conduction, the convective
and radiative boundary conditions with the moving heat source. The
combined heat source model is adopted to simulate the heat of the
welding arc and the melt droplets (Pardo and Weckman, 1989), i.e. the
heat of the welding arc is modeled by a surface heat source with a
Gaussian distribution, and that of the melt droplets is reproduced by a
volumetric heat source with uniform density. In order to take into
account the heat transfer due to fluid flow in the weld pool, an
artificially increased thermal conductivity is assumed for temperatures
above the melting point. The liquid-to-solid phase transformation
effects of the weld pool are modeled by considering the latent heat of
fusion. The combined heat transfer coefficient for the convection and
the radiation (Lee and Chang, 2014) is applied on all surfaces.
The subsequent mechanical model entails the use of the tempera-
ture histories predicted by the preceding heat transfer model for each
time increment as thermal loads for calculating transient thermal
stresses and deformations. The temperature values obtained at every
time step when the weld filler is completely deposited are employed in
the structural model. Each temperature step is divided into optimized
sub-steps considering the state of the welding process and the
convergence. For the stress-strain relations, rate-independent elastic-
plastic constitutive equation taking the Von Mises yield criterion,
temperature-dependent mechanical properties and linear isotropic
hardening rule into account is incorporated into the structural model.
The solid-state phase transformation effect is neglected here because
the metallurgical phase change does not occur during welding of the
austenitic stainless steel used in the present study. Since the thermal
elastic-plastic analysis is a nonlinear problem, a full Newton-Raphson
iterative solution technique (Bathe, 1996) is used for solving the
problem. In the thermal and mechanical models, the process of
sequential weld metal deposition is controlled by using a consistent
filler activation/deactivation scheme (Lindgren, 2001).
2.2. Model geometry
Fig. 1(a) schematically shows the configuration and dimensions of
the analysis model. The welding arc travel direction and the welding
start/end position (θ =0°) are also illustrated in the figure. A single pass
girth butt welding of two pipes with a single V-groove was considered.
The welding conditions and process parameters chosen for the present
model were as follows: welding method, gas tungsten arc (GTA)
welding process; welding current, 200 A; welding voltage, 12.5 V; and
welding speed, 3 mm/s, respectively. The welding procedures used
were typical of industrial practice (Malik et al., 2008). The 3-D FE
mesh model with eight-noded isoparametric solid elements is shown in
Fig. 1(b). Due to the symmetry conditions with respect to the weld
centerline, only one half of the pipe geometry needed to be built, and
two layers were adopted through the thickness since the residual stress
variation through the inside surface or the outside surface is insignif-
icant for a girth-welded thin-walled steel pipe (Lee and Chang, 2013a).
The adopted mesh configuration was a result of the considerations
regarding the girth weld-induced residual stresses and the ratcheting
behavior of the girth-welded steel pipe in internal pressure and cyclic
bending. In the weld region and its vicinity, the mesh is suitably refined
to accurately capture the anticipated high temperature and stress
83
Ocean Engineering 128 (2016) 81–93
gradients there. A few preliminary mesh convergence studies leaded
to using the present FE mesh with the smallest element size of 0.9 mm
(axial) ×1.5 mm (thickness) ×12.8 mm (circumference). Specific de-
scriptions on the FE mesh scheme for examining the ratcheting
responses are made next (Section 4.1). The same FE mesh refinement
scheme between the thermal and mechanical models was employed
except for the element type and the applied boundary conditions in
order to facilitate nodal data mapping between them. The FE element
type adopted in the heat transfer model is the one which has single
degree of freedom, temperature, on its each node, while for the
structural model the element type is the other with three degrees of
freedom at each node: translations in x, y, z directions. The mechanical
boundary conditions were prescribed for hindering rigid body motion
of the weld piece.
2.3. Material model
The base material selected for this investigation is SUS304 auste-
nitic stainless steel. The physical constants such as thermal conductiv-
ity, specific heat and density as well as the mechanical properties
including Young's modulus, Poisson's ratio, thermal expansion coeffi-
cient, yield strength and strain hardening rate vary with temperature.
Thus, temperature-dependent thermo-physical and thermal-structural
properties were taken into account in the FE analysis and are shown in
Figs. 2 and 3 (Lee and Chang, 2014). Note that in Fig. 3(a), both the
yield stress and the elastic modulus are reduced to 5.0 MPa and
5.0 GPa, respectively, at the melting temperature to simulate low
strength at high temperatures (Barsoum, 2008). The same material
properties were postulated for the base material and the weld metal.
The work hardening induced by the thermal cycles during welding in
the weld area and its neighborhood was also considered. In this study,
temperature-dependent strain hardening rule was used (Lee and
Chang, 2014).
2.4. Results and discussion
In order to clarify how the axial and hoop residual stresses change
along the circumferential weld, four sections which have different
circumferential angles from the weld start/end position, θ are selected
to represent welding residual stress distributions. Fig. 4(a) and (b)
depict the axial residual stress distributions at the four locations on the
inside surface and the outside surface, respectively. Note that the stress
profiles are reported only in the weld region and its neighborhood for
clarity. Actually, beyond the region given in the figure, the stresses
converge to zero. Bending axial stress distribution through the thick-
ness in and around the girth weld can be observed, i.e. the axial
residual stresses are tensile on the inside surface and compressive on
the outside surface along the perimeter. This is because the circumfer-
ential shrinkage during the girth welding process results in a local
inward deformation near the welded zone, producing tensile axial
residual stresses on the inside surface balanced by compressive axial
residual stresses on the outside surface. Compressive axial residual
stresses are formed on the inside surface away from the weld center-
line, and tensile axial residual stresses on the outside surface due to
self-equilibration. It is also worth noting that the axial residual stresses
are sensitive to the circumference. This is ascribed to the fact that the
sequential weld metal deposition as the welding torch travels along the
girth makes the internal restraint change spatially. The biggest varia-
tion occurs at the overlapping region (θ =0°). Fig. 5(a) and (b) portray
the hoop residual stresses at the four positions on the inside surface
and the outside surface, respectively. Turning to the hoop residual
stresses, their magnitude is affected by the axial residual stresses. This
explains why the hoop residual stresses on the inside surface, which is
experiencing axial tension, are more tensile at the weld area and its
vicinity compared to those on the outside surface. Furthermore, careful
observation of the results unveils that spatial variation of the stress
W.-G. Bae et al. Ocean Engineering 128 (2016) 81–93

(a)

Y X

(b)
Fig. 1. Model for analysis: (a) Configuration and dimensions of the analysis model and the welding arc direction and (b) 3-D FE mesh model.

profile is present along the circumference due to the moving arc and
the welding start/end effects.
3. Constitutive model
3.1. Cyclic plasticity constitutive model
The constitutive model is based on the isotropic linear elasticity
with the generalized Hooke's law, a von Mises yield criterion, the
associated plastic flow rule and the hardening rule. The rate-indepen-
dent cyclic plasticity constitutive model considered in this study, which
assumes cyclic hardening material behavior and J2 -type plasticity, has
the following three fundamental aspects:
(i) von Mises yield criterion: the yield criterion prescribes the
condition for onset of plastic strain. The yield surface is repre-
sented by
⎡3 ⎤1/2
f (σ − α ) = ⎢ (σ ′ − α′)⋅(σ ′ − α′) ⎥ − R − k* = 0
⎣2 ⎦ (1)
The dot “⋅ ” implies the inner product at a⋅b = aij bij .
(ii) Flow rule: the total strain increment is additively decomposed into
the elastic and the plastic strain increment. The incremental

84
W.-G. Bae et al. Ocean Engineering 128 (2016) 81–93

1000 Axial_90º
Axial_180º
800 Axial_270º
Axial_0º

Residual stress (MPa)

600

400

200

-200
0 20 40 60 80 100 120
Distance from the weld centerline (mm)

(a)
Fig. 2. Temperature-dependent thermo-physical constants of the base material.
400 Axial_90º
Axial_180º
Axial_270º
200 Axial_0º

Residual stress (MPa)

0

-200

-400

-600
0 20 40 60 80 100 120
Distance from the weld centerline (mm)

(b)

Fig. 4. Axial residual stresses at locations with different circumferential angle θ : (a)
inside surface and (b) outside surface.

incorporated into the cyclic plasticity model, which is expressed as

l l ⎧ ⎛ ′ ⎞m
αij̇ ′ = ∑ (αij̇ ′ )k = ∑ γk ⎨ 2 ck εij̇ p − ⎜ αk ⎟[φ (αij′)k + (1 − φ)((αrs′ )k ⋅nrs ) nij ]
k =1 k =1 ⎩3 ⎝ ck ⎠
⎫
p⎬
̇
⎭ (3)

For the isotropic hardening rule, the following equation is adopted

(Lee et al., 2014b; Do et al., 2015).

Fig. 3. Temperature-dependent thermal-structural properties of the base material.
stress-plastic strain relation is given by
∂f 1 2 ∂f
ε ̇p = λ ̇ = σ ′̇ ⋅n n , n =
∂σ H 3 ∂σ
where denotes the MacCauley bracket
(iii) Hardening rule: the hardening rule dictates changes that occur in
the yield surface during plastic deformation. In general, the yield
surface evolves through size change (the yield surface expansion
by isotropic hardening), shape change (the yield surface distor-
tion) and/or variation in center location (the yield surface
translation by kinematic hardening) due to plastic deformation.
In this work, the evolution of the yield surface with plastic loading
increments is accommodated by considering the isotropic and the
kinematic hardening rule. For the kinematic hardening rule, the
decomposed nonlinear kinematic hardening equation proposed by
the authors (Lee et al., 2014b; Do et al., 2015) is employed and
R = Q (1 − e−bp ) (4)
The plastic strain path dependence is not taken into account here
(Zakavi et al., 2010).
The cyclic constitutive model has three decomposed kinematic
(2)
hardening rules (n = 3), i.e. the number of the back stress components
is three (Bari and Hassan, 2000; Koo and Lee, 2007). Thus, the
implementation of the cyclic constitutive model needs a total number
of eleven material parameters, i.e. the material parameter k *, the
material constants of c1, c2 , c3, γ1, γ2 , γ3, m , φ for computing the kinematic
hardening rate and those of Q , b for calculating the isotropic hardening
rule are required to simulate the cyclic plastic behavior. The parameter
k* is obtained from the linear part of the stable uniaxial stress-strain
hysteresis curve. The cyclic model parameters for kinematic hardening
c1, c2 , c3 and γ1, γ2 , γ3 are determined by using only the stabilized uniaxial
stress-strain hysteresis loop with pure kinematic hardening in accor-
dance with the procedures specified in (Bari and Hassan, 2000). The
material constants Q and b can be acquired from the isotropic
hardening evolution curve with respect to accumulated plastic strain

85
W.-G. Bae et al. Ocean Engineering 128 (2016) 81–93

1000 Hoop_90º
Hoop_180º
800 Hoop_270º
Hoop_0º
600
Residual stress (MPa)

400

200

-200

-400
0 20 40 60 80 100 120
Distance from the weld centerline (mm)

(a)

600 Hoop_90º
Hoop_180º
Hoop_270º
400 Hoop_0º
Residual stress (MPa)

200

-200

-400
0 20 40 60 80 100 120
Distance from the weld centerline (mm)

(b)

Fig. 5. Hoop residual stresses at locations with different circumferential angle θ : (a)
inside surface and (b) outside surface.

by assuming that the cyclic hardening is only induced by isotropic

hardening, which is determined by the strain-controlled symmetric
cyclic loading test (Do, 2013). The additional kinematic hardening
parameters m and φ are identified from the ratcheting experiments by
the trial and error method (Do, 2013).
3.2. Verification
In order to confirm the accuracy of the cyclic constitutive model, the
experimental works by Kang et al., (2002) in which the stress-strain
response of SUS 304 stainless steel under fully reversed strain-
controlled cyclic loading and the uniaxial ratcheting of the stainless
steel under the asymmetric stress cycling were measured were simu-
lated using the FE analysis technique for direct comparisons of the
analytical predictions calculated by the cyclic constitutive model with
the test data. Specific details on the experiments are presented in (Kang
et al., 2002). Fig. 6 compares the simulated and the measured
hysteresis loops for the strain-controlled symmetric cycling with a
strain amplitude of 0.6%. The experimental data can be available only
in the form of figure and hence it is difficult to compare the analytical
and experimental results in single figure. It can be found that the
overall locus of the predicted stress-strain response curve is in a good
agreement with that of the experiment and the cyclic constitutive
model precisely mimics the cyclic hardening behavior. Comparison is
next made in Fig. 7 between the experimental measurement on the
uniaxial ratcheting response (Kang et al., 2002) and the analytical
prediction computed by the cyclic constitutive model. The inserted
graph in Fig. 7(b) shows the ratcheting progress in last cycles. All the
material parameters for the cyclic constitutive model are gained from
the experiments and are tabulated in Table 1. There is a high
correlation between the predicted and the experimental results.
Fig. 6. Comparison of the predicted and the experimental hysteresis loops for the
symmetric strain cycling: (a) experiment and (b) simulation.
Comparison is finally made between the test data on the multiaxial
ratcheting of a thin-walled pipe specimen of 1026 carbon steel
subjected to axial strain-controlled symmetric cycling in the presence
of a constant internal pressure (Corona et al., 1996) and the analytical
prediction from the corresponding FE simulation for the verification
purpose. Details on the material, the preparation of the specimen and
the loading conditions are presented in (Corona et al., 1996). The
material parameters needed to implement the cyclic plasticity consti-
tutive model were obtained from the experimental measurements
found in (Corona et al., 1996). The axial strain versus circumferential
strain curve predicted by the constitutive model is shown and
compared with the experimental relations in Fig. 8. Like the preceding
uniaxial ratcheting experiment, the test data can be accessible only in
the form of figure. It can be seen that the overall trends between the
simulation and the test match reasonably well. Therefore, the rate-
independent cyclic plasticity constitutive model is believed to be
suitable for multiaxial ratcheting simulation as well as for describing
uniaxial ratcheting behavior.
4. Ratcheting behavior of the girth-welded stainless steel
pipe under cyclic bending and internal pressure
4.1. 3-D FE model for the analysis
After obtaining weld-induced residual stresses by the FE simulation
of the girth butt welding process, 3-D elastic-plastic FE analyses to
explore the local (circumferential strain) and global (diameter change)

86
W.-G. Bae et al. Ocean Engineering 128 (2016) 81–93

Fig. 7. Ratcheting response of the stainless steel under the asymmetric uniaxial stress
cycles: (a) experiment and (b) simulation.

Table 1
Model parameters for the material.

m φ c1 (MPa) c2 (MPa) c3 (MPa) γ1

0.5 1.0 30.11 82.71 70.8 3.466

γ2 γ3 Q (MPa) b k* (MPa)
350.0 53.19 50.0 12.5 225.0

ratcheting responses of the girth-welded stainless steel pipe submitted

to steady internal pressure and cyclic bending were carried out through
the in-house FE-code (Do, 2013) by incorporating the cyclic plasticity
model into the material constitutive equation. In this study, the same
material parameters as for the uniaxial ratcheting simulation were
utilized to replicate the ratcheting responses of a pressurized girth-
welded stainless steel pipe under cyclic bending due to the scarcity of
Fig. 8. Comparison of the analytical and the experimental axial strain versus circumfer-
the measured data, even though the additional material parameters
ential strain response. (a) analytical prediction and (b) experimental measurement.
calibrated from the multiaxial ratcheting test could yield more accurate
results. Actually, for many cases, the same material constants can
Table 2
duplicate fairly well both uniaxial and multiaxial ratcheting experi-
Loading conditions for the analyzed cases.
ments (Abdel-Karim, 2009). Rahman et al. (2008) also demonstrated
that the cyclic plasticity constitutive model in which the material Case Displacement amplitude (mm) Internal pressure (MPa)
parameters were determined from the uniaxial ratcheting rate only
1 6.16 5.6
produced similar ratcheting responses in the multiaxial ratcheting
2 8.22 5.6
simulations at the structural level as well as at the material level to 3 10.3 5.6
those predicted by the cyclic plasticity model whose material para- 4 8.22 4.2
meters were obtained through both the uniaxial and multiaxial 5 8.22 7.0
ratcheting rates. Furthermore, this hypothesis seems to be sensible

87
W.-G. Bae et al.
Fig. 9. Schematics of the pressurized girth-welded steel pipe in four-point cyclic bending
loading (a) loading condition and (b) diameter change of the pipe cross-section under the
cyclic bending.
Fig. 10. Cyclic mid-span moment-displacement response curves of the pipe model for
the Case-2: (a) simulation with consideration of welding residual stresses and (b)
simulation without considering welding residual stresses.
and justified for parametric comparative studies because material
behavior contributes equally in the results of all cases and discrepan-
cies in structural response can be attributed to the different conditions.
88
Ocean Engineering 128 (2016) 81–93
Fig. 11. Comparison of the in-plane diameter change-displacement curves: (a) predic-
tion with consideration of welding residual stresses and (b) prediction without
considering welding residual stresses.
Nevertheless, it should be mentioned that the cyclic plasticity consti-
tutive model which employs the material parameters listed in Table 1 is
not able to describe the non-proportional hardening effect, i.e. the
additional hardening of the stainless steel pipe due to non-proportional
loading (Halama et al., 2012). The radial return method introduced by
Simo and Taylor (1985) was used to numerically solve the equations in
the cyclic constitutive model. The multiaxial ratcheting simulations
were conducted by reproducing four-point displacement controlled
cyclic bending at constant internal pressure. In the FE analysis,
symmetry conditions with respect to the pipe mid-span (weld center-
line) can be applied and the maximum hoop strain and diameter
change occur near the pipe mid-span. Thus, the same FE mesh density
as the thermal stress model, with different loading and boundary
conditions, was utilized to facilitate data mapping between the two
structural FE models. A uniform mesh configuration away from the
mid-span was carefully determined by carrying out a mesh convergence
study to achieve accurate results whilst minimizing computational
requirements, e.g., the suitable mesh size was found to be a uniform
mesh of 15.0 mm (axial) ×12.8 mm (circumference). 3-D eight-noded
isoparametric brick elements with three translational degrees of free-
dom at each node were used to discretize the computation domain. The
axial and hoop residual stresses and plastic strains of magnitudes and
distributions as given by the FE thermal simulation were introduced as
the initial condition into the FE model. In this work, the ratcheting
behavior of the girth-welded steel pipe under cyclic bending combined
with internal pressure was analyzed for five cases with various loading
conditions listed in Table 2. The displacement amplitude and the
W.-G. Bae et al.
Fig. 12. Comparison of the out-of-plane diameter change-displacement curves: (a)
prediction with consideration of welding residual stresses and (b) prediction without
considering welding residual stresses.
magnitude of internal pressure were adopted as the main loading
parameters. Three different levels of internal pressure-induced hoop
stress (σh = pDi /2t , p is the magnitude of internal pressure, Di is the
internal diameter and t is the thickness) to yield stress ratios, σh / σ0 of
0.3, 0.4 and 0.5 were considered (Østby et al., 2005). The actual
magnitudes of the internal pressure selected in this study correspond-
ing to the values of the hoop stresses above are 4.2, 5.6 and 7.2 MPa,
respectively as seen in Table 2.
As stated earlier, the multiaxial ratcheting simulations involved
four-point displacement controlled cyclic bending of the girth-welded
pipe in steady internal pressure. At first, the internal pressure was
applied in one step at the inside surface as a distributed load assuming
the open-ends condition (i.e. the axial force at the ends due to internal
pressure load is zero: only radial component (force) of pressure load is
applied), and then the cyclic symmetric displacements were imposed
gradually at the quarter points to the specified magnitude while the
internal pressure was held at the same value as shown in Fig. 9(a). The
effect of ring collars located at the loading points and the end supports,
which are typically used in four-point bending test so as to avoid stress
concentrations and to restrain the radial displacement of the specimen,
was taken into consideration by ensuring that the respective cross-
sections remained undeformed at these locations using kinematic
coupling (Theofanous et al., 2009). The boundary conditions were
chosen such that the girth-welded steel pipe is simply supported. As
indicated before, autogenous weldment was assumed during the girth
welding. Therefore, the cyclic stress-strain relations for the base
89
Ocean Engineering 128 (2016) 81–93
Fig. 13. Comparison of the hoop strain-displacement responses: (a) prediction with
consideration of welding residual stresses and (b) prediction without considering welding
residual stresses.
material can be used to capture the multiaxial ratcheting responses
of the girth-welded stainless steel pipe. The mechanical properties at
room temperature presented in Fig. 3 were used for the FE analyses.
The ratcheting responses of the pressurized pipe model under bending
without considering weld-induced residual stresses were also simulated
for the comparison by which the contribution degree of the residual
stresses to the multiaxial ratcheting behavior was identified.
4.2. Results and discussion
Steel pipes subjected to symmetric cyclic pure bending in the
presence of internal pressure experience progressive accumulation of
ovalization of the pipe cross-section (Fig. 9(b)) and hoop strain
ratcheting (Rahman et al., 2008; Gao et al., 2006). In the present
investigation, the diameter changes in both the plane of bending and its
vertical plane as well as the cumulative hoop strains were reported at
the mid-span. Fig. 10(a) depicts the simulated cyclic mid-span
moment-displacement loop for the Case-2, compared with the non-
linear prediction of the cyclic mid-span moment-displacement re-
sponse curve excluding welding residual stresses in Fig. 10(b). The
sign convention is established such that the moment generated by the
positive displacement is positive. It is immediately obvious that the
cyclic behavior is significantly affected by the residual stresses, i.e. the
cyclic moment response of the pipe model with no residual stresses
becomes saturated after slight cyclic softening in the initial few cycles,
whilst the cyclic moment-displacement curve of the pipe model with
W.-G. Bae et al. Ocean Engineering 128 (2016) 81–93

20 30
Case-1 Case-1
Case-2 Case-2
Case-3 25 Case-3
16

20

Hoop strain (%)

Moment (kN-m)

12

15
8

10

4
5

0
0 10 20 30 40 50 60 70 80 0
0 10 20 30 40 50 60 70 80
Number of cycles
Number of cycles
(a)
(a)
20
Case-4 18
Case-2 Case-1
Case-5 Case-2
16 15
Case-3

In-plane diameter change (mm)

12
Moment (kN-m)

12

8
6

4 3

0
0
0 10 20 30 40 50 60 70 80
-3
Number of cycles 0 10 20 30 40 50 60 70 80
(b) Number of cycles

Fig. 14. Cyclic moment responses of the girth-welded steel pipe under various loading (b)
conditions: (a) mid-span moments at the positive displacement peaks versus number of 15
cycles under different bending loads with steady internal pressure and (b) mid-span Case-1
moments at the positive displacement peaks versus number of cycles under various Case-2
Out-of-plane diameter change (mm)

internal pressures with constant bending loading. Case-3

12

the girth weld is stabilized after considerable cyclic deterioration of the

load-carrying capacity along with the premature yielding of the girth- 9
welded steel pipe. In general, under cyclic loading, welding residual
stresses tend to relax to some extent after a certain number of cycles
and become stabilized. The magnitude of the residual stress relaxation 6
depends on the loading direction as well as the amplitude of applied
stress (Lee et al., 2015). In this study, the residual stresses are not fully
3
released during the mechanical cyclic loading. The initial residual
stresses cause premature yielding of the pipe cross-section and the
relaxed residual stresses result in the deterioration of the load-carrying 0
capacity. Moreover, the presence of constant internal pressure pro- 0 10 20 30 40 50 60 70 80
motes the cyclic softening. Figs. 11 and 12 represent the in-plane and Number of cycles
out-of-plane diameter change responses for the Case-2, respectively
(c)
with and without considering welding residual stresses after fifty four
cycles after which the analysis program which took the residual stresses Fig. 15. Local and global ratcheting responses of the girth-welded steel pipe under
different bending loads with steady internal pressure: (a) hoop strains at the negative
into account was terminated. The diameter change profiles follow
peak displacements versus number of cycles, (b) in-plane diameter changes at the
similar trends observed in the experiment (Rahman et al., 2008), i.e.
positive peak displacements versus number of cycles and (c) out-of-plane diameter
the in-plane diameter change-displacement curve develops into convex changes at the positive peak displacements versus number of cycles.
shape, whilst the out-of-plane diameter change-displacement curve
concave form. The results also unveil that the diameter of the pipe those of the pressurized pipe which has no girth weld, which is incurred
cross-section gradually increases in both directions with progress in from the residual stresses combined with the internal pressure. The
cycles owing to the steady internal pressure. Note that the in-plane hoop strain ratcheting responses for the Case-2 at the top surface with
diameter change occurs at a higher rate than the out-of-plane diameter and without considering weld-induced residual stresses are portrayed
change, indicating that some ovalization takes place. Moreover, it in Fig. 13. Referring to the results, the hoop strain ratchets with
should be recognized that both the in-plane and out-of-plane diameters progressive cycles and the cyclic strain rate of the pressurized pipe with
of the pressurized girth-welded pipe expand remarkably compared to the residual stresses is much higher than that of the pipe model with no

90
W.-G. Bae et al.
Fig. 16. Local and global ratcheting responses of the girth-welded steel pipe under
various internal pressures with constant bending load: (a) hoop strains at the negative
peak displacements with respect to the number of cycles, (b) in-plane diameter changes
at the positive peak displacements with respect to the number of cycles and (c) out-of-
plane diameter changes at the positive peak displacements with respect to the number of
cycles.
girth weld.
Results are next given for investigating influences of the internal
pressure and the displacement amplitude on the ratcheting responses.
The cyclic moment responses of the girth-welded pipe under different
bending loads with constant internal pressure and vice versa are
compared in Fig. 14(a) and (b), respectively by plotting the mid-span
moment at the positive displacement peak in each cycle versus the
91
Ocean Engineering 128 (2016) 81–93
number of cycles. These figures demonstrate that the cyclic moment
responses from the loading conditions are almost stable after marked
cyclic softening in a number of cycles and that the larger the moment or
the internal pressure becomes, the higher the rate of weakening
becomes. The evolution of hoop strain calculated at the negative
displacement peak of every cycle with respect to the number of cycles
of the girth-welded pipe subjected to different bending loads with
steady internal pressure is shown in Fig. 15(a). Furthermore, the
variations of in-plane and out-of-plane diameter changes computed at
the positive displacement peak of each cycle versus number of cycles
are depicted in Fig. 15(b) and (c), respectively. The increases in the
local (circumferential strain) and global (cross-section diameter
change) ratcheting responses with progress in the displacement
amplitudes can be seen. The hoop strain increment also advances with
increasing the displacement amplitude and the ovalization becomes
more notable as the cycling proceeds. Fig. 16(a), (b) and (c) represent
in turn the simulated hoop strains at the negative peak displacements,
the in-plane and out-of-plane diameter changes at the positive peak
displacements of the girth-welded pipe exposed to various internal
pressures with constant bending loading. It is apparent that in addition
to the hoop strain and its rate, the diameter changes increases as the
internal pressure increases. Careful observation of the results indicates
that the ovalization degree and shape are dependent on the loading
conditions.
5. Conclusions
In this study, efforts were made towards charactering the ratcheting
behavior of a girth-welded straight stainless steel pipe under cyclic
bending and steady internal pressure through the numerical simula-
tion. Thermal simulation of the girth butt welding process was first
performed to identify weld-induced residual stresses by using a
sequentially coupled 3-D thermo-mechanical FE analysis method. A
rate-independent cyclic plasticity constitutive model capable of simu-
lating the multiaxial ratcheting responses was next presented and
incorporated into the cyclic constitutive equations in the 3-D elastic-
plastic FE analyses to scrutinize the local (circumferential strain) and
global (cross-section diameter change) ratcheting responses of the
girth-welded stainless steel pipe subjected to various internal pressures
and constant cyclic bending loading and vice versa. The residual
stresses and plastic strains obtained from the preceding thermal
simulation were taken as the pre-stress condition for the analyses.
The ratcheting behavior of the pressurized steel pipe in cyclic bending
without considering the residual stresses is also investigated in order to
elucidate the effects that weld-induced residual stresses have on the
cyclic performance. Based on the results in this work, the following
conclusions can be drawn.
a) Weld-induced residual stresses in combination with the internal
pressure significantly affect the plastic behavior of the pressurized
girth-welded steel pipe in cyclic bending by enhancing considerably
the hoop strain rate and expanding notably the in-plane and out-of-
plane diameters, as compared to those of the pressurized steel pipe
with no girth weld. Hence, weld-induced residual stresses should be
taken into account in assessing the cyclic plastic responses.
b) The hoop strain and its rate, and the in-plane and out-of-plane
diameter changes increase with increasing the bending load or the
internal pressure.
c) The in-plane and out-of-plane diameters of the pipe cross-section
become larger as the cycling progresses with the diameter change
rates of the two orthogonal planes different each other, indicating
that some ovalization takes place. The ovalization degree increases
with advance in cycles.
d) The ovalization degree and shape are governed by the magnitudes
of applied loads such as the internal pressure and the cyclic
bending.
W.-G. Bae et al.
e) Accurate evaluation of the local and global ratcheting responses of a
girth-welded steel pipe in combined action of internal pressure and
cyclic bending loading by the FE analysis method can be useful to
understand the ratcheting-fatigue failure mechanism of a pressur-
ized welded piping component under cyclic bending.
f) The ratcheting creep tests may be needed to identify the origin of
the cyclic inelastic strain accumulation during the uniaxial asym-
metric stress cycling and a time-dependent cyclic plasticity con-
stitutive model could be developed accordingly to accurately reflect
the physical meaning of the progressive plastic deformation, at
which the future works are aimed.
Acknowledgments
This research was supported by Mid-career Researcher Program
through the National Research Foundation of Korea (NRF) funded by
the Ministry of Science, ICT & Future Planning (NRF-
2015R1A2A2A01006390).
References
Abdel-karim, M., Ohno, N., 2000. Kinematic hardening model suitable for ratcheting
with steady-state. Int. J. Plast. 16, 225–240.
Abdel-Karim, M., 2009. Modified kinematic hardening rules for simulations of
ratchetting. Int. J. Plast. 25, 1560–1587.
Armstrong P.J., Frederick C.O., 1966. A mathematical representation of the multiaxial
bauschinger effect. CEGB Report RD/B/N731. Berkely Nuclear Laboratories,
Berkely, UK.
Bari, S., Hassan, T., 2000. Anatomy of coupled constitutive models for ratcheting
simulation. Int. J. Plast. 16, 381–409.
Bari, S., Hassan, T., 2002. An advancement in cyclic plasticity modeling for multiaxial
ratcheting simulation. Int. J. Plast. 18, 873–894.
Barsoum, Z., 2008. Residual stress analysis and fatigue of multi-pass welded tubular
structures. Eng. Fail Anal. 15, 863–874.
Bathe, K.J., 1996. Finite Element Procedures. Prentice Hall, New Jersey.
Brickstad, B., Josefson, B.L., 1998. A parametric study of residual stresses in multi-pass
butt-welded stainless steel pipes. Int. J. Press. Vessels Pip. 75, 11–25.
Burlet, H., Cailletaud, G., 1986. Numerical techniques for cyclic plasticity at variable
temperature. Eng. Comput. 3, 143–153.
Chaboche, J.L., 1986. Time-independent constitutive theories for cyclic plasticity. Int. J.
Plast. 2, 149–188.
Chaboche, J.L., 1991. On some modifications of kinematic hardening to improve the
description of ratcheting effects. Int. J. Plast. 7, 661–678.
Chaboche, J.L., 2008. A review of some plasticity and viscoplasticity constitutive theories.
Int. J. Plast. 24, 1642–1693.
Chaboche, J.L., Rousselier, G., 1983. On the plastic and viscoplastic constitutive
equations-part I: rules developed with internal variable concept. J. Press Vessel
Technol. 105, 153–158.
Chen, X., Jiao, R., 2004. Modified kinematic hardening rule for multiaxial ratcheting
prediction. Int. J. Plast. 20, 871–898.
Chen, X., Chen, X., Yu, D., Gao, B., 2013. Recent progresses in experimental investigation
and finite element analysis of ratcheting in pressurized piping. Int. J. Press. Vessels
Pip. 101, 113–142.
Corona, E., Hassan, T., Kyriakides, S., 1996. On the performance of kinematic hardening
rules in predicting a class of biaxial ratcheting histories. Int. J. Plast. 12, 117–145.
Deng, D., Murakawa, H., 2006. Numerical simulation of temperature field and residual
stress in multi-pass welds in stainless steel pipe and comparison with experimental
measurements. Comput. Mater. Sci. 37, 269–277.
Do, N.V.V., 2013. Finite element modeling of fatigue damage and its evolution in steel
structures (Ph.D. thesis). Chung-Ang University, Korea.
Do, N.V.V., Lee, C.H., Chang, K.H., 2015. A constitutive model for uniaxial/multiaxial
ratcheting behavior of a duplex stainless steel. Mater. Des. 65, 1161–1171.
Drucker, D.C., Palgen, L., 1981. On stress-strain relations suitable for cyclic and other
loading. J. Appl. Mech. 48, 479.
Feigenbaum, H.P., Dugdale, J., Dafalias, Y.F., Kourousis, K.I., Plesek, J., 2012. Multiaxial
ratcheting with advanced kinematic and directional distortional hardening rules. Int.
J. Solids Struct. 49, 3063–3076.
Gao, B., Chen, X., Chen, G., 2006. Ratchetting and ratchetting boundary study of
pressurized straight low carbon steel pipe under reversed bending. Int. J. Press.
Vessels Pip. 83, 96–106.
Halama, R., Šofer, M., Fojtík, F., 2012. Choice and calibration of cyclic plasticity model
with regard to subsequent fatigue analysis. Eng. Mech. 19 (2/3), 87–97.
Hassan, T., Kyriakides, S., 1992. Ratcheting in cyclic plasticity, part I: uniaxial behavior.
Int. J. Plast. 8, 91–116.
Hassan, T., Kyriakides, S., 1994a. Ratcheting of cyclically hardening and softening
materials, part I: uniaxial behavior. Int. J. Plast. 10, 149–184.
Hassan, T., Kyriakides, S., 1994b. Ratcheting of cyclically hardening and softening
materials, part II: multiaxial behavior. Int. J. Plast. 10, 185–212.
Hassan, T., Corona, E., Kyriakides, S., 1992. Ratcheting in cyclic plasticity, part II:
92
Ocean Engineering 128 (2016) 81–93
multiaxial behavior. Int. J. Plast. 8, 117–146.
Hassan, T., Zhu, Y., Matzen, V.C., 1998. Improved ratcheting analysis of piping
components. Int. J. Press. Vessels Pip. 75, 643–652.
Jiang, Y., Sehitoglu, H., 1996. Modeling of cyclic ratchetting plasticity, part I:
development of constitutive relations. J. Appl Mech. 63, 720–725.
Kang, G.Z., Kan, Q., 2007. Constitutive modeling for uniaxial time-dependent ratcheting
of SS304 stainless steel. Mech. Mater. 39, 488–499.
Kang, G.Z., Liu, Y., 2008. Uniaxial ratchetting and low-cycle fatigue failure of the steel
with cyclic stabilizing or softening feature. Mater. Sci. Eng. A 472, 258–268.
Kang, G.Z., Gao, Q., Yang, X., 2002. A visco-plastic constitutive model incorporated with
cyclic hardening for uniaxial/multiaxial ratcheting of SS304 stainless steel at room
temperature. Mech. Mater. 34, 521–531.
Karlsson, R.I., Josefson, B.L., 1990. Three-dimensional finite element analysis of
temperatures and stresses in a single-pass butt-welded pipe. J. Press Vessel Technol.
112, 76–84.
Koo, G.H., Lee, J.H., 2007. Investigation of ratcheting characteristics of modified 9Cr-
1Mo steel by using the Chaboche constitutive model. Int. J. Press. Vessels Pip. 84,
284–292.
Lee, C.H., 2005. A Study on the Mechanical Characteristics of High Strength Steel for the
Application to the Steel Bridge (Ph.D. thesis). Chung-Ang University, Korea.
Lee, C.H., Chang, K.H., 2011a. Prediction of residual stresses in high strength carbon
steel pipe weld considering solid-state phase transformation effects. Comput. Struct.
89, 256–265.
Lee, C.H., Chang, K.H., 2011b. Numerical investigation on the behavior of
circumferentially butt-welded steel circular hollow section flexural members. J.
Struct. Eng. 137, 1395–1404.
Lee, C.H., Chang, K.H., 2013a. Failure pressure of a pressurized girth-welded super
duplex stainless steel pipe in reverse osmosis desalination plants. Energy 61,
565–574.
Lee, C.H., Chang, K.H., 2013b. Influence of the residual stresses and distortions on the
structural behavior of girth-welded cylindrical steel members. Constr. Build Mat. 41,
766–776.
Lee, C.H., Chang, K.H., 2014. Comparative study on girth weld-induced residual stresses
between austenitic and duplex stainless steel pipe welds. Appl. Therm. Eng. 63,
140–150.
Lee, C.H., Chang, K.H., Park, J.U., 2013a. Three-dimensional finite element analysis of
residual stresses in dissimilar steel pipe welds. Nucl. Eng. Des. 256, 160–168.
Lee, C.H., Do, N.V.V., Chang, K.H., 2014b. Analysis of uniaxial ratcheting behavior and
cyclic mean stress relaxation of a duplex stainless steel. Int. J. Plast. 62, 17–33.
Lee, C.H., Chang, K.H., Do, N.V.V., 2015. Finite element modeling of residual stress
relaxation in steel butt welds under cyclic loading. Eng. Struct. 103, 63–71.
Lee, C.H., Chang, K.H., Park, K.T., Shin, H.S., Kim, T., 2013b. Bending resistance of
girth-welded stainless steel circular hollow sections. Thin-Walled Struct. 73,
174–184.
Lee, C.H., Chang, K.H., Park, K.T., Shin, H.S., Lee, M., 2014a. Compressive strength of
girth-welded stainless steel circular hollow section members: stub columns. J.
Constr. Steel Res. 92, 15–24.
Lee, H.Y., Kim, J.B., Lee, J.H., 2004. Evaluation of progressive inelastic deformation
induced by a moving axial temperature front for a welded structure. Int. J. Press.
Vessels Pip. 81, 433–441.
Lindgren, L.-E., 2001. Finite element modelling and simulation of welding, Part 2
Improved material modeling. J. Therm. Stress. 24, 195–231.
Malik, A.M., Qureshi, E.M., Dar, N.U., Khan, I., 2008. Analysis of circumferentially arc
welded thin-walled cylinders to investigate the residual stress fields. Thin-Walled
Struct. 46, 1391–1401.
Nouailhas, D., Chaboche, J.L., Savalle, S., Gailletaud, G., 1985. On the constitutive
equations for cyclic plasticity under nonproportional loading. Int. J. Plast. 1,
317–330.
Ohno, N., Wang, J.D., 1993. Kinematic hardening rules with critical state of dynamic
recovery, part I: formulations and basic features for ratcheting behavior. Int. J. Plast.
9, 375–390.
Østby, E., Jayadevan, K.R., Thaulow, C., 2005. Fracture response of pipelines subject to
large plastic deformation under bending. Int. J. Press. Vessels Pip. 82, 201–215.
Pardo, E., Weckman, D.C., 1989. Prediction of weld pool and reinforcement dimensions
of GMA welds using a finite element model. Met. Mater. Trans. B 20, 937–947.
Paul, S.K., Sivaprasad, S., Dhar, S., Tarafder, S., 2012. True stress-controlled ratcheting
behavior of 304 LN stainless steel. J. Mater. Sci. 47, 4660–4672.
Paul, S.K., Stanford, N., Taylor, A., Hilditch, T., 2015. The effect of low cycle fatigue,
ratcheting and mean stress relaxation on stress-strain response and microstructural
development in a dual phase steel. Int. J. Fatigue 80, 341–348.
Portier, L., Calloch, S., Marquis, D., Geyer, P., 2000. Ratchetting under tension-torsion
loadings: experiments and modeling. Int. J. Plast. 16, 303–335.
Rahman, S.M., Hassan, T., Corona, E., 2008. Evaluation of cyclic plasticity models in
ratcheting simulation of straight pipes under cyclic bending and steady internal
pressure. Int. J. Plast. 24, 1756–1791.
Simo, J.C., Taylor, R.L., 1985. Consistent tangent operator for a rate-independent
elastoplasticity. Comp. Method Appl. Mech. Eng. 48, 101–118.
Song, D., Kang, G., Kan, Q., Yu, C., Zhang, C., 2014. Non-proportional multiaxial
transformation ratcheting of super-elastic NiTi shape memory alloy: experimental
observations. Mech. Mater. 70, 94–105.
Taleb, L., Hauet, A., 2009. Multiscale experimental investigations about the cyclic
behavior of the 304 L SS. Int. J. Plast. 25, 1359–1385.
Taleb, L., Cailletaud, G., 2011. Cyclic accumulation of the inelastic strain in the 304 L SS
under stress control at room temperature: ratcheting or creep? Int. J. Plast. 27,
1936–1958.
Taleb, L., Cailletaud, G., Sai, K., 2014. Experimental and numerical analysis about the
W.-G. Bae et al. Ocean Engineering 128 (2016) 81–93

cyclic behavior of the 304L and 316 L stainless steels at 350 °C. Int. J. Plast. 61,
32–48.
Theofanous, M., Chan, T.M., Gardner, L., 2009. Flexural behavior of stainless steel oval
hollow sections. Thin-Walled Struct. 47, 776–787.
Withers, P.J., 2007. Residual stress and its role in failure. Rep. Prog. Phys. 70,
2211–2264.
Yaghi, A., Hyde, T.H., Becker, A.A., Sun, W., Williams, J.A., 2006. Residual stress
simulation in thin and thick-walled stainless steel pipe welds including pipe diameter
effects. Int. J. Press. Vessels Pip. 83, 864–874.
Zakavi, S.J., Zehsaz, M., Eslami, M.R., 2010. The ratchetting behavior of pressurized
plain pipework subjected to cyclic bending moment with the combined hardening
model. Nucl. Eng. Des. 240, 726–737.

93