Proceedings of the 2012 9th International Pipeline Conference
IPC2012
September 2428, 2012, Calgary, Alberta, Canada
MULTITIER TENSILE STRAIN MODELS FOR STRAINBASED DESIGN PART 2 – DEVELOPMENT AND FORMULATION OF TENSILE STRAIN CAPACITY MODELS
Ming Liu, YongYi Wang, and Yaxin Song
Center for Reliable Energy Systems Dublin, OH, USA
David Horsley
BP Calgary, Alberta, Canada
Steve Nanney
PHMSA, US DOT Houston, TX, USA
ABSTRACT
This is the second paper in a threepaper series related to
the development of tensile strain models. The fundamental basis of the models [1] and evaluation of the models against experiment data [2] are presented in two companion papers. This paper presents the structure and formulation of the models. The philosophy and development of the multitier tensile strain models are described. The tensile strain models are applicable for linepipe grades from X65 to X100 and two welding processes, i.e., mechanized GMAW and FCAW/SMAW. The tensile strain capacity (TSC) is given as a function of key material properties and weld and flaw geometric parameters, including pipe wall thickness, girth weld highlow misalignment, pipe strain hardening (Y/T ratio), weld strength mismatch, girth weld flaw size, toughness, and internal pressure. Two essential parts of the tensile strain models are the crack driving force and material’s toughness. This paper covers principally the crack driving force. The significance and
determination of material’s toughness are covered in the
companion papers [1,2].
KEYWORDS
Strainbased design, tensile strain capacity, tensile strain models, crack driving force, crack tip opening displacement (CTOD)
INTRODUCTION
The applications of strainbased design under a variety of ground movement hazard conditions are described by Wang, et al. in a companion paper [1]. Tensile strain models at varying level of maturity are briefly reviewed. The TSC of a pipeline is controlled by its girth welds, due to the possible existence of
weld defects and often deteriorative metallurgical and/or mechanical property changes from welding thermal cycles. Consequently, TSC is an integral part of girth weld procedure specification and field welding flaw acceptance criteria.
The most recognized procedures for TSC design are DNV OS F101 and DNV RP F108 for offshore and CSA Z662 Annex C for onshore applications. In the Annex C of CSA Z662 2007 Edition [3], TSC is given in a set of parametric equations which incorporate the influence of material properties and girth flaw size. The initiation toughness, alternatively termed apparent toughness is used as to represent the material’s toughness under low cracktip constraint conditions [4,5,6,7,8,9].
Some factors which were later found to have a strong impact on TSC were not explicitly considered in the Annex C of CSA Z662. These factors include internal pressure [10,11,12], girth weld highlow misalignment [13,14,15], HAZ softening [16], and weld strength mismatch. The effects of internal pressure effect were implicitly considered by setting a hard limit on the maximum allowable apparent toughness (0.3 mm) to prevent nonconservative TSC prediction. Girth weld strength overmatch was required to use the parametric equations of Annex C.
A joint US government (Pipeline and Hazardous Materials Safety Administration, or PHMSA of US DOT) and industry (PRCI) project was initiated in 2006 to develop refined tensile strain models. The work was jointly carried out by the Center for Reliable Energy Systems (CRES), CFER Technologies, and Microalloying International. The description of this multi year multimillion dollar effort can be found in a set of freely available publications on US DOT PHMSA web site [17,18]. The TSC model development effort led by CRES is the focus of
1 Copyright © 2012 by ASME
This work is in part a work of the U.S. Government. ASME disclaims all interest in the U.S. Government’s contributions.
this paper. Some parts of the newly developed models are a continuation of the CSA Z662 Annex C approach, albeit with significantly more refinement.
Parallel to the US DOT and PRCI efforts, independent work has been carried out by other organizations, notably SINTEF [19,20,21,22], ExxonMobil [23,24,25,26,27] and Osaka/JFE [28,29,30].
STRUCTURE OF THE TSC MODELS Limit States
Two limit states, initiationcontrol and ductileinstability control, are incorporated in the tensile strain models.
Limit State Based on Initiation Control
The initiationbased tensile limit state is defined as CTOD _{F} = CTOD _{A} , where CTOD _{F} is the crackdriving force and CTOD _{A} (or δ _{A} ) is the apparent toughness. The apparent toughness is the toughness corresponding to the onset of stable tearing. Both CTOD _{F} and CTOD _{A} are represented by the crack tip opening displacement, i.e., CTOD. The initiationbased limit state is schematically illustrated in Figure 1.
The apparent toughness represents the resistance of the material to the applied strain in the presence of weld flaws.
The term “apparent” is adopted in recognition that the
traditional singleparameter based fracture mechanics is not strictly valid in the presence of large cracktip plasticity, which
is almost always the case for strainbased design.
Figure 1 Definition of initiationcontrol based limit state
Limit Stated Based on Ductile Instability
The ductileinstability based limit state is illustrated in Figure 2. The fracture toughness is expressed as a function of flaw growth as shown in Figure 2, which is usually termed as the tearing resistance curve (i.e., CTOD _{R} curve). The limit state is defined as the tangent point of the crack driving force curve and the fracture toughness curve.
TSC Model Structures
The tensile strain models presented here were established with the following philosophical considerations:
(1)
A flexible framework is necessary for the adoption of
current technology and the incorporation of future
development. Tensile strain design is still an evolving engineering discipline. A number of potentially valid approaches are being developed. A flexible framework allows for the incorporation of those approaches as they become mature. One example of such flexibility is the use of both CTOD and Jintegral as fracture mechanics parameters. Although CTOD is used as the fracture mechanics parameter in the current models, the approach established here can be easily converted to use J as the fracture mechanics parameter. The offshore industry has been moving towards Jbased assessment methodology in recent years.
Figure 2 Definition of ductileinstability based limit state
(2) The most appropriate approach for the tensile strain design of a particular project depends on the scale of the
project and many design and maintenance considerations. No single approach may be appropriate for all projects. There can be tradeoffs between extensive material property testing and qualification versus simply “overdesigning” the system. For instance, lower grades and thicker walls may be used to achieve high strain capacity with a relatively simplified welding procedure qualification. (3) The overall framework should be flexible enough to use various forms of material toughness representation. Depending on the vintage of a pipeline or if the methodology is being applied to new constructions or an inservice pipeline, the availability of girth weld toughness varies greatly. For instance, Charpy impact energy may not be considered as a rigorous fracture mechanics parameter. For many inservice pipelines, the Charpy impact toughness could be the closest toughness representation. With those philosophical considerations as a guide, a multitier structure, represented by four levels, is established for the TSC models. As the level progresses, the assessment methodology becomes more sophisticated; thus more material property data are required.
2
Copyright © 2012 by ASME
Level 1 Models – Initial Screening and Feasibility Studies
The Level 1 models are intended for quick estimations of likely tensile strain capacity. The TSC is tabulated for selected pipe dimensions, material properties, and flaw sizes [18]. The apparent toughness is estimated from upper shelf Charpy impact energy.
Level 2 Models – Nominal Assessment with Standard Toughness Data
The Level 2 models are given in a library of parametric equations as described in the following sections. The apparent toughness can be estimated from either upper shelf Charpy impact energy or the upper shelf standard CTOD toughness.
Level 3 Models – Advanced Assessment with Low Constraint Toughness
The Level 3 models have two options. Level 3a uses an initiation control limit state. Level 3b uses a ductile instability control limit state. In Level 3a, the TSC is given by the same library of parametric equations as in Level 2. The apparent toughness may be obtained by a number of low constraint test options, including shallownotched SENB, SENT, and CWP. In Level 3b, the crack driving force, CTOD _{F} , is expressed by a group of isostrain curves constructed from the same library of parametric equations. In this application, various levels of strains are given as a function of flaw depth derived from the parametric equations, see Figure 2.
Level 4 Models – Advanced Analysis with Direct FEA Calculation
The Level 4 models are structured in two options representing two limit states. In contrast to Level 3 models, where the driving force relations are expressed in parametric equations, the driving force relations are directly obtained from finite element analysis (FEA) in this level. The toughness options are the same as in Level 3 models. This level allows for special cases when the specific weld geometry and material property conditions do not allow the use of the first three levels. The Level 4 models should only be exercised by seasoned experts. As stated in a prior section, the two key components of tensile strain models are crack driving force and the material’s toughness. The development of the crack driving force relations, as shown schematically in Figure 1 (CTOD _{F} ), is described in details below.
DEVELOPMENT OF CTOD _{F} RELATIONS Material Property Representation
Background
The crack driving force (CTOD _{F} ) is influenced by the shape of a material's stressstrain curve. The stressstrain curve
is often represented by three key material parameters: yield strength (YS), ultimate tensile strength (UTS), and uniform strain (uEL). In the development of TSC models, material's full stressstrain curves are used in FEA. These stressstrain curves are fully described by some mathematic equations. Depending on the selection of the mathematical equations, different stress strain curves can be formed even with the same values of the three key parameters, YS, UTS (or Y/T), and uEL. Therefore, proper mathematic equations are needed to represent the shape of the stressstrain curves of actual pipe and weld materials. The mathematical equations should be able to (1) uniquely determine a full stressstrain curve using the three material parameters, i.e., YS, UTS (or Y/T), and uEL and (2) statistically give a reasonable representation of the stressstrain curves of actual materials.
Two widely used stressstrain curve equations, i.e., the RambergOsgood (RO) equation and the CSA Z662 equation were examined. The CSA Z662 equation was found to give a better representation of pipe stressstrain curves than the RO equation [18]. The CSA Z662 equation was selected to represent and generate full stressstrain curves. The CSA Z662 equation defines the relationship between the engineering stress () and engineering strain () as,
y
y
,
(1)
where E is the Young's modulus, _{y} is the YS at 0.5% strain, and n is the strain hardening exponent of the CSA equation.
For
a given
set
of YS
( _{y} ), UTS ( _{u} ), and uEL ( _{u} ), a
unique n can be determined by Eq. (2). Therefore, the CSA equation can uniquely determine a full stressstrain curve which satisfies the YS, UTS (or Y/T), and uEL exactly.
u
y
(2)
Representation of Pipe Properties
The relationship between the strain hardening exponent n and the Y/T ratio is shown in Figure 3 for a collection of 76 pipe longitudinal stressstrain curves. The range of the material properties is summarized in Table 1. The YS (longitudinal direction) ranges from 56 ksi to 114 ksi, and therefore roughly covers pipe grades from X65 to X100. The uEL varies from 5% to 17%. The key material parameters, i.e., YS, UTS, and uEL, were directly measured from the stressstrain curves and used to calculate the Y/T and the strain hardening exponent n of the CSA equation (i.e., Eq. (2)). A clear correlation between n and Y/T can be observed in Figure 3 which indicates the key linepipe parameters, i.e., YS, UTS (or Y/T), and uEL are interrelated to a certain degree. In the following analysis, the
3
Copyright © 2012 by ASME
Eng. Stress (MPa)
n (CSA)
Calculated uEL (mm/mm)
uEL was treated as a dependent material parameter. Its value is uniquely related YS and Y/T.
Table 1 Summary of the material database (pipe)
80
60
40
20
0
YS: 56 ksi  114 ksi
No. of StressStrain Curves: 76
uEL: 5%  17%
0.70
0.75
0.80
0.85
0.90
0.95
1.00
Y/T
Figure 3 Correlation between n and Y/T (pipe)
A representative pipe stressstrain curve is shown in Figure
4. The curve is of roundhouse shape but becomes very flat
near UTS. The flat curve near UTS is typical for modern linepipe steels which can create certain challenges in determining uEL. In addition, due to the flat part of the curve, the representation of the stressstrain curve with a single mathematical equation can be difficult. Therefore, the stress train curve was divided into two parts as shown in Figure 4 and represented by two different equations, respectively:
= _{u}
y
y
n
( ≤ _{0} ) ,
( _{0} < ≤ _{u} ) ,
(3)
(4)
where n, _{0} , and _{u} are calculated from the following equations,
650
600
550
500
450
400
350
300
^{} 0
0.00
0.02
0.04
0.06
0.08
0.10
Eng. Strain (mm/mm)
Figure 4 A sample pipe stressstrain curve
0.30
0.25
0.20
0.15
0.10
0.05
0.00
YS: 56 ksi  114 ksi
No. of StressStrain Curves: 76
0.00
0.05
0.10
0.15
Measured uEL(mm/mm)
0.20
Figure 5 Calculated uEL v.s. measured uEL
Representation of Weld Properties
Two sample weld stressstrain curves are shown in Figure 6, one for X65 pipe and the other for X100 pipe. In contrast to the roundhouse type stressstrain curves of linepipes, the weld metal stressstrain curves show a discontinuous yielding point.
The stressstrain relationship is linear prior to reaching the yield
point. Furthermore, it has been frequently observed that the weld metal stressstrain curves often contain a Lüder’s strain as high as 2%. Some of the recent data suggest that the existence
of the Lüder’s strain is closely related to specimen locations.
4
Copyright © 2012 by ASME
n (CSA)
YS (ksi)
Eng. Stress (MPa)
(8)
(9)
y
y
n
(10)
where a 1% Lüder’s strain was used.
The strain hardening exponent n can be calculated from the following correlation equation:
.
(11)
Equation (11) was derived from a collection of 64 weld stressstrain curves with YS ranging from 66 ksi to 132 ksi as shown in Figure 7. The range of the material properties is summarized in Table 2. Specimens of different types and extracted at different locations are all plotted together.
To further reduce the number of independent parameters for generating a weld stressstrain curve, a correlation equation for the YS and UTS of the weld was established for the same set of weld stressstrain curves. As shown in Figure 8, the correlation equation is given as:
_{y} = _{u}  13.2,
(12)
where the unit of YS ( _{y} ) and UTS ( _{u} ) is ksi. By the use of Eqs. (8)  (12), a unique weld metal stressstrain curve can be generated for a given UTS.
Table 2 Summary of the material database (weld)
0.96 96.9
0.89 34.2
0.81 15.9
n
(CSA)
1000
800
600
400
200
X65
X100
0.00
0.05
0.10
0.15
0.20
Eng. Strain (mm/mm)
Figure 6 Sample weld metal stressstrain curves
120
100
80
60
40
20
0
n= 2.58/(1Y/T) ^{1}^{.}^{1}^{7}
Y/T: 0.81  0.96
uEL: 4%  13%
Experiment Data
Fitted Curve
No. of StressStrain Curves: 64
YS: 66 ksi  132 ksi
0.80
0.85
0.90
0.95
1.00
Y/T
Figure 7 Correlation between n and Y/T (weld)
160
140
120
100
80
60
No. of StressStrain Curves: 64
60
80
100
120
140
160
UTS (ksi)
Figure 8 Correlation between YS and UTS (weld)
Crack Driving Force Development
FE Models
Commercial finite element (FE) software ABAQUS® was
used to calculate crack driving forces. A typical finite element
model is shown in Figure 9, where due to symmetry boundary
conditions in the circumference direction, only half of the pipe was modeled. The length of the model was 6 times of pipe OD to eliminate endeffects on crack opening and to obtain a finite zone of uniform strain in pipe longitudinal direction. Three dimensional linear solid/brick elements were used where non linear geometric (largedeformation) effect was enabled. The selected element type is proven to be effective on large plasticity analysis, especially associated with cracks.
The flaw was of a semielliptical shape and located on the ID surface of the pipe and at the fusion line (i.e., HAZ flaws). Due to weld metal overmatching, which is generally required in
strainbased design, the HAZ flaws usually see larger crack driving force than the weld metal flaws. The girth weld high low misalignment was modeled as a relative shift of the two pipes on either side of the welds. The flaw depth was defined as the distance of the tip to the pipe ID surface, which is
5
Copyright © 2012 by ASME
adjacent to the flaw, as shown in Figure 9.
The flaw size was
kept constant during simulation and no flaw growth was
Figure 9 Finite element model
The FEA was conducted in two steps. In the first step, an internal pressure was applied to the ID surface where the two pipe ends were free of constraint and pressure. In the second step, a uniaxial tension was applied to the pipe while maintaining the internal pressure.
The key output of the FEA is the crack driving force relations, i.e., the relationship between the crack tip opening displacement (CTOD _{F} ) and the remote strain. The CTOD _{F} is calculated from the deformed crack tip profile following the traditional 45ºline method as shown in Figure 10. The remote strain was calculated in the uniform strain zone using a gage length of half OD, as shown in Figure 9.
Figure 10 Schematic illustration of CTOD _{F} calculation Weld Profiles
Weld and HAZ Profile
The profile of actual welds varies. To characterize the size and shape of a weld, the simplified parameters such as bevel angle, weld root width, cap height, and HAZ width are often used. In general, the total volume of the weld, determined by
the root width and bevel angle, is the dominant parameter for the CTOD _{F} . However, the HAZ width (with softened HAZ), the shape and height of the cap (with large bevel angles), and
the alignment of the bevel angle/HAZ with plastic strain shear
band (with softened HAZ) can also have impact on the CTOD _{F} .
The weld geometries are mainly determined by the welding
procedures. Although the exact weld shape can be affected by
many other conditions, the major parameters such as the bevel angle, root width, and HAZ width are repeatable within certain
ranges. The shape and height of the weld cap, on the other
hand, can vary significantly, even for the same weld. Mechanized GMAW and FCAW (or SMAW) are normally considered to be two of the major welding processes for pipeline welding. The mechanized GMAW welds normally have a small bevel angle and weld volume. The FCAW (or SMAW) welds have larger bevel angles and weld volume. Due to its high heat input, the FCAW (or SMAW) welds usually have a wider HAZ than mechanized GMAW welds.
The weld profiles of those two welding processes (i.e., GMAW and FCAW (or SMAW)) were selected in the analysis matrix (see Figure 11). The weld geometric parameters were determined from representative welds for each welding process. Conservative values were used for parameters that are highly variable such as the weld cap height.
2.5 mm
Figure 11 Weld geometry profiles and parameters
HAZ Softening
HAZ 
softening 
has been 
frequently 
found 
in modern 

linepipe welds, even for the pipe grades as low as X65. 
The 

hardness 
map 
in the weld area usually indicates that 
the 
6
Copyright © 2012 by ASME
strength of HAZ materials varies with location. It should be noted that the width of the softened zone can be twice as wide as the visible HAZ. By controlling the welding procedure, the level of HAZ softening can be controlled. For welds of reasonable quality, the maximum softening should be expected to be below 15% and likely around 10%. Previous research has demonstrated that a 10%15% softened HAZ of 2mm wide can increase the crack driving force by 20%.
To
obtain
reasonably
conservative
solutions,
a
10%
softening was used in all analyses. The strength of the HAZ materials was assumed to vary with the location in the HAZ and follow a 2 ^{n}^{d} order function of a bathtub shape. The maximum softening was reached at the midwidth of the HAZ and the strength of the HAZ material gradually changed to the
strength of the pipe and weld materials, respectively at the corresponding interface.
FE Matrix
Based on the degree of interaction between the parameters, two types of analysis matrix were developed for different key parameters, namely full and partial permutation analysis matrix.
Partial Permutation Matrix
The effect of wall thickness and internal pressure on CTOD _{F} was analyzed with a partial permutation matrix as shown in Table 3 and Table 4, respectively. In those analyses, the complete permutation was implemented only on selected parameters.
Full Permutation Matrix
The key parameters such as flaw depth, flaw length, weld highlow misalignment, pipe Y/T, and weld OM are highly interacting. Those parameters were named full permutation parameters in the following matrix. To understand and model their coupling effect, full permutation of all those parameters needs to be implemented. The corresponding FE matrix is called the full permutation analysis matrix (see Table 5). The full permutation parameters are highlighted in bold. A total of 1152 cases are included in this matrix for two different welds (GMAW and FCAW). The Y/T ranges from 0.75 to 0.94; and the weld overmatch (OM) varies from 0% to 30%. The flaw depth changes from 2 mm to 8 mm; and the flaw length changes from 25 mm to 100 mm. The weld highlow misalignment varies from 0.0 mm to 3.2 mm.
MULTITIER TENSILE STRAIN MODELS BY THEIR LEVELS Levels 1, 2 and 3a Models  Initiation Control
The Levels, 1, 2, and 3a models use initiation control. The TSC is determined by equating the crack driving force with the apparent toughness. The Levels 2 and 3a models are expressed
in parametric equations. The Level 1 models are effectively a tabular format of the Level 2 models.
Table 3 FE matrix for wall thickness effect




Values 


Parameters 
Thickness I 
Thickness II 
Thickness III 

YS (ksi) 
80 
80 
80 

Pipe 
Y/T 
0.81, 0.87 
0.81, 0.87 
0.81, 0.87, 0.94 

Material 
Lüder's Strain 
1.0% 
1.0% 
1.0% 

Weld 
OM 
0%, 15% 
0%, 15% 
0%, 15% 

HAZ 
Softening 
10% 
10% 
10% 

OD (inch) 
42 
42 
42 

Pipe 
WT (mm) 
12.70 
19.05 
25.40 

Depth (mm ) 
2, 4 
2, 4 
3, 6 

Geometry 
Flaw 
Length (mm) 
25, 50 
25, 50 
25, 50 

Highlow (mm ) 
0.0, 1.6, 3.2 
0.0, 1.6, 3.2 
0.0, 1.6, 3.2, 4.8 

Weld 
Type 
GMAW 
GMAW 
GMAW 

Load 
Pressure 
Design Factor 
0.72 
0.72 
0.72 

Total Cases 
192 



Table 4 FE matrix for pressure effect 

Values 

Parameters 
Nominal 
Y/T 
OM 
Depth 
Length 
Misalignment 

YS (ksi) 
80 
80 
80 
80 
80 
80 

Pipe 
Y/T 
0.87 
0.75, 0.94 
0.87 
0.87 
0.87 
0.87 

Material 
Lüder's Strain 
1.0% 
1.0% 
1.0% 
1.0% 
1.0% 
1.0% 

Weld 
OM 
15% 
15% 
0%, 30% 
15% 
15% 
15% 

HAZ 
Softening 
10% 
10% 
10% 
10% 
10% 
10% 

OD (inch) 
42 
42 
42 
42 
42 
42 

Pipe 
WT (mm) 
15.875 
15.875 
15.875 
15.875 
15.875 
15.875 

Depth (mm ) 
6 
6 
6 
2, 8 
6 
6 

Geometry 
Flaw 
Length (mm) 
75 
75 
75 
75 
25, 100 
75 

Highlow (mm ) 
1.6 
1.6 
1.6 
1.6 
1.6 
0.0, 3.2 

Weld 
Type 
GMAW 
GMAW 
GMAW 
GMAW 
GMAW 
GMAW 

Load 
Pressure 
Design Factor 
0.0, 0.2, 0.3, 0.4, 0.5, 0.6, 0.72, 0.8 
0.0, 0.4, 0.6, 0.72, 0.8 

Total Cases 
64 

Table 5 Full permutation analysis matrix
YS (ksi) 
80 
1 
Y/T 
0.75, 0.81, 0.87, 0.94 
4 
Lüder's Strain 
1% 
1 
OM 
0%, 15%, 30% 
3 
Softening 
10% 
1 
OD (inch) 
42 
1 
WT (in) 
5/8 
1 
Depth (mm) 
2, 4, 6, 8 
4 
Length (mm) 
25, 50, 75, 100 
4 
Highlow (mm) 
0.0, 1.6, 3.2 
3 
Type 
GMAW, FCAW 
2 
Design Factor 
0.72 
1 
Pressure
Weld
Flaw
Parameters
Pipe
Geometry
Pipe
Material
Weld
HAZ
Load
Development Process
The development of the Levels 2 and 3a models follows the following process:
7
Copyright © 2012 by ASME
(1) 
The crack driving force databases of the full analysis 
matrix in Table 5 were fitted with two sets of parametric 

equations, one for GMAW and one for FCAW/SMAW. 

(2) 
A wall thickness correlation relation was developed 
using a curve fitting procedure based on the FEA database of pressurized pipes shown in Table 3. 

(3) 
The effects of internal pressure on TSC were analyzed 
by comparing the TSC of pressurized pipes with the TSC of nonpressurized pipes (Table 4). A scaling factor for the TSC of nonpressurized pipes was developed by comparing their
TSC with the TSC of pressurized pipes at a pressure factor of
0.72.
(4) 
An interpolation function was developed between zero 

pressure and the 72% pressure based on the cases in Table 4. 

The 
outcome 
of 
the 
above 
process 
is 
a 
set 
of TSC 
equations for a full range of internal pressure, a range of wall 

thickness, and two weld profiles. 

Input Parameters 
pressure factor, ratio of applied hoop stress to pipe yield strength.
The equations are in the general format of
(13)
where _{u} represents pipe uniform strain (uEL) and the functions G(t) and P(f _{p} ) characterize the effect of wall thickness and internal pressure, respectively. The function G(t) is given as,
0.80961 1.503
1.229
,
(14)
and the function P(f _{p} ) is in the form of,
p
max
p
max
p
.
p
(15)
A conservative fixed value of 1.5 was initially given to P _{m}_{a}_{x} in [18]. A more refined relation was developed later as,
The TSC models include the following parameters: 
P 
2.252a 
(16) 

(1) 
Geometric parameters: 
max 
t 

t 
pipe wall thickness, mm, pipe yield strength, MPa, 
The function TSC _{p} in Eq. (13) is given as 

a 
flaw height, mm, 
F 


2c 
flaw length, mm, and 
TSCp 
A 
A1 F 

A 
(17) 

h 
girth weld highlow misalignment, mm. 
where the unction F( _{A} ) is 

(2) _{y} 
D 

Material parameters: 
FACABA. (18) A, B, C and D in Equations (1) and (3) are fitted functions of 

_{} _{U} 
pipe ultimate tensile strength, MPa, 
normalized geometry and material properties. 

_{U} ^{W} 
weld metal tensile strength, MPa, and 
The fitted functions of A, B, C and D for GMAW are 

A(3) 
girth weld apparent CTOD toughness, mm. Loading parameter: 
A 
a e1 
a 
2/ e3ea 4 /B 
1a5ab 1b2 
6b 3 
a 7/b45b( 
a81a9bb67) 
10aa11b8 
b9 
a 1213aa, 

,(19) (20) 

pressure factor, ratio of applied hoop stress to pipe 
c2 

fpTSC Equations 
yield strength. 
c 
1 
/e(1D 
c3 )1c4c5 d3 dd 12 1 d4 
c e c e67c1d5d6 1d 7 
8c9d8 
c10 c11dd 910 . 
c12 
,(21) (22) 
The TSC equations are given as a function of the following normalized geometric and material parameters, apparent CTOD toughness and pressure factor:
= a/t = 2c/t = h/t
normalized flaw depth, normalized flaw length, normalized weld highlow misalignment,
= _{y} / _{U} base metal Y/T ratio, = _{U} ^{W} / _{U} weld metal strength mismatch ratio
girth weld apparent CTOD toughness, mm,
The coefficients are listed in Table 6. The fitted functions of A, B, C and D for FCAW/SMAW are
,(23)
,
(24)
2
1
3
5
10
12
,
.
(25)
(26)
The coefficients are listed in Table 7.
8
Copyright © 2012 by ASME
Applicable Range
The applicable ranges of the input parameters of TSC equations are:
= a/t = 2c/t = h/t
=
_{y} / _{U}
= _{U} ^{W} / _{U}
0.05 – 0.5, 1.0 – 20.0, 0.0 – 0.2, 0.75 – 0.94, 1.0 – 1.3, 0.2 – 2.5 mm, and
0.0 – 0.8.
The applicable range of pipe wall thickness is 12.7 mm  25.4 mm (i.e., 0.5 inch  1.0 inch). The applicable range of pipe OD is 304 mm – 1,219 mm (i.e., 12 inch – 48 inch). The applicable range of pipe yield strength is 386 MPa  690 MPa (i.e., 56 ksi  100 ksi).
Table 6 Coefficients of GMAW equation
Table 7 Coefficient of FCAW equation
2.023E+00
a6
9.434E01
1.057E+00
2.603E01
d5
c5
b5
1.609E+00
c1
Level 3b  Ductile Instability Control
The TSC from the initial control limit state is determined by equating the crack driving force with the apparent toughness. The parametric equations described in the previous sections are effectively the crack driving force relation, if the apparent toughness in those equations is replaced with the crack driving force CTOD _{F} . These parametric equations can therefore be used to develop the isostrain curves of Figure 2.
Alternatively, the TSC equations of Eqs. (13)  (26) can be used to calculate the TSC directly for the ductile instability
limit state. A relationship between the apparent toughness ( _{A} ) and the flaw depth (a) can be generated from Eq. (17). By incrementing the flaw depth, the _{} _{A} vs. Da (where Da =aa _{0} ) relation is similar to a resistance curve. The parameter a _{0} is the initial flaw depth. The strain value, i.e., TSC in Eq. (17), corresponding to varying amount of flaw growth (Da) can be calculated by using a = a _{0} + Da and _{A} = CTOD _{R} (Da). Here the CTOD _{R} (Da) is the experimentally measured material’s resistance. This process is equivalent to forcing the flaw to obey the material’s resistance curve. The computed strain values from this process first increase and then decrease with an incrementally larger Da. The maximum strain value is the tensile strain capacity (TSC) at the point of ductile instability. The process is fundamentally the same as determining the instability point using the isostrain curves. With the parametric TSC equations, the ductile instability limit state can be exercised without generating the isostrain curves.
Level 4 Models
The Level 4 models are effectively similar to Level 3 models. The major difference is that, at Level 4, the crack driving force is obtained on a casebycase basis. At Levels 2 and, the crack driving force is given by a set of parametric equations. Expanded weld features can be incorporated in the Level 4 analysis. It should be emphasized that developing crack driving force in a consistent and robust manner requires extensive experience and validation. Only seasoned experts in the tensile strain design should attempt the Level 4 approach.
CONCLUDING REMARKS
Multitier tensile strain models are presented in this paper. Newly identified issues which can affect TSC, such as internal pressure, girth weld highlow misalignment, HAZ softening, and weld strength mismatch, have been systematically investigated and addressed. The tensile strain models are applicable to linepipe grades X65X100 and two welding processes, i.e., GMAW and FCAW/SMAW.
ACKNOWLEDGMENTS
The financial support of DOT/PHMSA and PRCI is gratefully acknowledged. The authors wish to recognize the invaluable contributions of Drs. Xiaoxia Wu and Honggang Zhou to the work presented here.
9
Copyright © 2012 by ASME
REFERENCES
1 Wang, Y.Y., Liu, M., Zhang, F., Horsley, D., and Nanney, S., “MultiTier Tensile Strain Models for StrainBased Design Part I  Fundamental Basis,” Proceedings of the 2012 9 ^{t}^{h} International Pipeline Conference, Paper No. IPC201260690, Calgary, Alberta, Canada, September 24 28, 2012.
2 Liu, M., Wang, Y.Y., Horsley, D., and Nanney, S., “Multi Tier Tensile Strain Models for StrainBased Design Part III  Model Evaluation against Experimental Data,” Proceedings of the 2012 9 ^{t}^{h} International Pipeline Conference, Paper No. IPC201260660, Calgary, Alberta, Canada, September 2428, 2012.
3 CSA Z662, “Oil and Gas Pipeline Systems,” Canadian Standards Association, 2007.
4 Wang, Y.Y., Rudland, D., Denys, R., and Horsley, D. J., “A Preliminary StrainBased Design Criterion for Pipeline Girth Welds,” Proceedings of the International Pipeline Conference 2002, Calgary, Alberta, Canada, September 29 October 3, 2002.
5 Wang, Y.Y., Cheng, W., McLamb, M., Horsley, D., Zhou, J., and Glover, A., “Tensile Strain Limits of Girth Welds with SurfaceBreaking Defects Part I an Analytical Framework,” in Pipeline Technology, Proceedings of the 4th International Conference on Pipeline Technology, Edited by Rudi Denys, Ostend, Belgium, May 913, 2004.
6 Wang, Y.Y., Horsley, D., Cheng, W., Glover, A., McLamb, M., and Zhou, J., “Tensile Strain Limits of Girth Welds with SurfaceBreaking Defects Part II Experimental Correlation and Validation,” in Pipeline Technology, Proceedings of the 4th International Conference on Pipeline Technology, Edited by Rudi Denys, Ostend, Belgium, May 913, 2004.
7 Horsley, D. and Wang, Y.Y., “Weld Mismatch Effects on the Strain Limits of X100 Girth Welds,” in Pipeline Technology, Proceedings of the 4th International Conference on Pipeline Technology, Edited by Rudi Denys, Oostende, Belgium, May 912, 2004.
8 Wang, Y.Y., Cheng, W., and Horsley, D., “Tensile Strain Limits of Buried Defects in Pipeline Girth Welds,” Proceedings of the International Pipeline Conference 2004, Calgary, Alberta, Canada, October 48, 2004.
9 Wang, Y.Y., Liu, M.,
Horsley,
D.,
and
Zhou,
J.,
“A
Quantitative Approach to Tensile Strain Capacity of Pipelines,” 6 ^{t}^{h} International Pipeline Conference, Paper No.
IPC200610474, September 2529, 2006, Calgary, Alberta, Canada.
10 Mohr. W. C., “StrainBased Design of Pipelines,” final report to MMS, October 2003.
11 Mohr, W., “Weld Area Mismatch and Pressure Effects in StrainBased Design,” Proceedings of the 4th International Conference on Pipeline Technology, Edited by Rudi Denys, Ostend, Belgium, May 913, 2004, pp. 279290.
12 Liu, M. and Wang, Y.Y., “Significance of Biaxial Stress on the Strain Concentration and Crack Driving Force in Pipeline Girth Welds with Softened HAZ,” Proceedings of the 26 ^{t}^{h} International Conference on Offshore Mechanics and Arctic Engineering (OMAS 2007), San Diego, California, USA, June 1015, 2007.
13 Wang, Y.Y. and Liu, M., “The Role of Anisotropy, Toughness Transferability, and Weld Misalignment in the Strain Based Design of Pipelines,” Proceedings of the 17 ^{t}^{h} International Offshore and Polar Engineering Conference (ISOPE 2007), Lisbon, Portugal, July 16, 2007.
14 Wang, Y.Y. and Liu, M., “Weld Integrity from the Perspective of StrainBased Design of Large Diameter and High Strength Pipelines,” Special Issue of the Proceedings of “Pipeline Technology Now and Then” at the 8 ^{t}^{h} International Welding Symposium, 8WS, November 1618, 2008, Kyoto, Japan.
15 Kibey, S. A., Minnaar, K., Issa, J. A., and Gioielli, P. C., “Effect of Misalignment on the Tensile Strain Capacity of Welded Pipelines,” Proceedings of the 18 ^{t}^{h} International Offshore and Polar Engineering Conference (ISOPE 2008), Vancouver, Canada, July 611, 2008.
16 Liu, M., Wang, Y.Y., and Horsley, D., “Significance of HAZ Softening on Strain Concentration and Crack Driving Force in Pipeline Girth Welds,” Proceeding of the 24 ^{t}^{h} International Conference on Offshore Mechanics and Arctic Engineering , 2005, June 1317, Kalkidiki, Greece.
17 Wang, Y.Y., Liu, M., Long, X., Stephens, M., Petersen, R., and Gordon, R., 2011, “Validation and Documentation of Tensile Strain Limit Design Models for Pipelines,” US DOT
Contract No. DTPH5606T000014, final report to US DOT and PRCI, August 2, 2011,
18 Wang, Y.Y., Liu, M., and Song, Y., “Second generation models for strainbased design,” US DOT Contract No. DTPH5606T000014, final report to US DOT and PRCI, August 30, 2011, http://primis.phmsa.dot.gov/matrix/PrjHome.rdm?prj=201.
10
Copyright © 2012 by ASME
19 Østby, E, “New strainbased fracture mechanics equations including the effects of biaxial loading, mismatch and misalignment”, Proceeding of the 24 ^{t}^{h} International Conference on Offshore Mechanics and Arctic Engineering, Paper No. OMAE200567518, June 1217, 2005, Halkidiki, Greece.
20 Sandivk, A., Østby, E, Naess A., Sigurdsson G., and Thaulow, C., “Fracture control – offshore pipelines: probabilistic fracture assessment of surface cracked ductile pipelines using analytical equations”, Proceeding of the 24 ^{t}^{h} International Conference on Offshore Mechanics and Arctic Engineering, Paper No. OMAE200567517, June 1217, 2005, Halkidiki, Greece.
21 Østby, E., Torselletti, E., and Levold, E., "A strainbased approach to fracture assessment of pipelines". Proceedings of FITNET 2006, International Conference on Fitnessfor service, May 1719, Amsterdam, Netherlands.
22 Østby, E., “Proposal for a strainbased fracture assessment procedure for offshore pipelines”. Proceedings of the 17 ^{t}^{h} International Offshore and Polar Engineering Conference, July, 16, 2007, Lisbon, Portugal.
23 Minnaar, K. Gioielli, P.C., Macia, M.L., Bardi, F., Biery, N.E., and Kan, W.C., “Predictive FEA modeling of pressurized fullscale tests”, Proceedings of the 17 ^{t}^{h} International Offshore and Polar Engineering Conference, July, 16, 2007, Lisbon, Portugal, P3114.
24 Kibey, S.A., Minnaar, K., Cheng, W., and Wang, X., “Development of a physicsbased approach for the predictions of strain capacity of welded pipelines”, Proceedings of the 19 ^{t}^{h} International Offshore and Polar Engineering Conference, June, 2126, 2009, Osaka, Japan, P132.
25 Kibey, S., Issa, J.A., Wang, X., and Minnaar, K., “A simplified, parametric equation for prediction of tensile strain capacity of welded pipelines”, Pipeline Technology Conference, October 1214, 2009, Paper No. Ostend2009 039, Ostend, Belgium.
26 Kibey, S.A., Wang, X., Minnaar, K., Macia, M.L., Fairchild D.P., Kan, W.C., Ford, S.J., and Newbury, B., “Tensile strain capacity equations for strainbased design of weld pipelines”, Proceedings of the 8th International Pipeline Conference, September 27  October 1, 2010, Calgary, Alberta, Canada, IPC201031661.
27 Fairchild D.P., Macia, M.L., Kibey, S.A.,
Wang, X.,
Krishnan, V.R., Bardi, F., Tang, H., and Cheng, W, “A
multitiered procedure for engineering critical assessment of strainbased pipelines”, Proceedings of the 21 ^{s}^{t}
International Offshore and Polar Engineering Conference, June, 1924, 2011, Maui, Hawaii, USA, P698.
28 Otsuka, A., et al., “Effect of Stress Triaxiality on Ductile Fracture Initiation of Low Strength Steel,” J. of the Society of Material Science, Japan, Vol. 29, No. 322, pp. 717723,
1980.
29 Toyoda, M., et al., “Criterion for Ductile Cracking for the Evaluation of Steel Structure under Large Scale Cyclic Loading,” Proc. Of 20 ^{t}^{h} International Conference on Offshore Mechanics and Arctic Engineering, OMAE 2001 MAT3103, 2001.
30 Igi, S. and Suzuki, N., “Tensile Strain Limits of X80 High Strain Pipelines,” Proceedings of the 17 ^{t}^{h} International Offshore and Polar Engineering Conference, Lisbon, Portugal, July 16, 2007.
11
Copyright © 2012 by ASME
Гораздо больше, чем просто документы.
Откройте для себя все, что может предложить Scribd, включая книги и аудиокниги от крупных издательств.
Отменить можно в любой момент.