International Journal of Solids and Structures

Similar documents
Treatment of Constraint in Non-Linear Fracture Mechanics

Engineering Fracture Mechanics

Elastic-Plastic Fracture Mechanics. Professor S. Suresh

Efficient 2-parameter fracture assessments of cracked shell structures

Thermal load-induced notch stress intensity factors derived from averaged strain energy density

FRACTURE MECHANICS FOR MEMBRANES

Influence of impact velocity on transition time for V-notched Charpy specimen*

Critical applied stresses for a crack initiation from a sharp V-notch

DEVELOPMENT OF TEST GUIDANCE FOR COMPACT TENSION FRACTURE TOUGHNESS SPECIMENS CONTAINING NOTCHES INSTEAD OF FATIGUE PRE-CRACKS

Crack Tip Plastic Zone under Mode I Loading and the Non-singular T zz -stress

Constraint effects on crack-tip fields in elasticperfectly

Weld Fracture. How Residual Stresses Affect Prediction of Brittle Fracture. Outline. Residual Stress in Thick Welds

FAILURE ASSESSMENT DIAGRAM ASSESSMENTS OF LARGE-SCALE CRACKED STRAIGHT PIPES AND ELBOWS

Characterization of crack-tip field and constraint for bending specimens under large-scale yielding

On the Path-Dependence of the J-Integral Near a Stationary Crack in an Elastic-Plastic Material

Topics in Ship Structures

Cracks Jacques Besson

INFLUENCE OF A WELDED PIPE WHIP RESTRAINT ON THE CRITICAL CRACK SIZE IN A 90 BEND

A fracture parameter for welded structures with residual stresses

Materials and Structures

TOUGHNESS OF PLASTICALLY-DEFORMING ASYMMETRIC JOINTS. Ford Research Laboratory, Ford Motor Company, Dearborn, MI 48121, U.S.A. 1.

Weibull stress solutions for 2-D cracks in elastic and elastic-plastic materials

IMECE CRACK TUNNELING: EFFECT OF STRESS CONSTRAINT

Fracture mechanics fundamentals. Stress at a notch Stress at a crack Stress intensity factors Fracture mechanics based design

Non-linear fracture mechanics in LS-DYNA and LS-PrePost

Volume 2 Fatigue Theory Reference Manual

NOTCH FRACTURE OF MEMS SENSORS MADE OF SINGLE CRYSTAL SILICON

A modified quarter point element for fracture analysis of cracks

MECHANICS OF MATERIALS

Lecture #7: Basic Notions of Fracture Mechanics Ductile Fracture

EFFECT OF THERMAL AND WELD- INDUCED RESIDUAL STRESSES ON THE J-INTEGRAL AND CTOD IN ELASTIC-PLASTIC FRACTURE ANALYSES SINTAP/SAQ/03

Size effect in the strength of concrete structures

G1RT-CT A. BASIC CONCEPTS F. GUTIÉRREZ-SOLANA S. CICERO J.A. ALVAREZ R. LACALLE W P 6: TRAINING & EDUCATION

J-T AND J-Q CHARACTERIZATION OF SURFACE CRACK TIP FIELDS IN METALLIC LINERS UNDER LARGE-SCALE YIELDING

Stress intensity factors under combined tension and torsion loadings

A Simplified Eigenstrain Approach for Determination of Sub-surface Triaxial Residual Stress in Welds

Evolution of Tenacity in Mixed Mode Fracture Volumetric Approach

Modelling the nonlinear shear stress-strain response of glass fibrereinforced composites. Part II: Model development and finite element simulations

Shock wave speed and stress-strain relation of aluminium honeycombs under dynamic compression

Basic studies of ductile failure processes and implications for fracture prediction

Elastic and Elastic-Plastic Behaviour of a Crack in a Residual Stress Field

5 ADVANCED FRACTURE MODELS

Multiscale modeling of failure in ABS materials

436 A. Barani and G.H. Rahimi assessment models have been employed to investigate the LBB of cracked pipes that are not for combined load [8]. Yun-Jae

Fracture Mechanics, Damage and Fatigue Non Linear Fracture Mechanics: J-Integral

The plastic behaviour of silicon subjected to micro-indentation

Examination in Damage Mechanics and Life Analysis (TMHL61) LiTH Part 1

UNIT I SIMPLE STRESSES AND STRAINS

Module 5: Theories of Failure

Chapter 7. Highlights:

Finite Element Investigation on the Stress State at Crack Tip by Using EPFM Parameters

Tensile behaviour of anti-symmetric CFRP composite

Methodology for the evaluation of yield strength and hardening behavior of metallic materials by indentation with spherical tip

INFLUENCE OF TEMPERATURE ON BEHAVIOR OF THE INTERFACIAL CRACK BETWEEN THE TWO LAYERS

FCP Short Course. Ductile and Brittle Fracture. Stephen D. Downing. Mechanical Science and Engineering

Computers and Geotechnics

Stress concentration factor in plates with transverse butt-weld misalignment

Mechanical Properties of Polymer Rubber Materials Based on a New Constitutive Model

Finite-Element Analysis of Stress Concentration in ASTM D 638 Tension Specimens

Distributed: Wednesday, March 17, 2004

The waves of damage in elastic plastic lattices with waiting links: Design and simulation

FEA A Guide to Good Practice. What to expect when you re expecting FEA A guide to good practice

Massachusetts Institute of Technology Department of Mechanical Engineering Cambridge, MA 02139

J-Integral Evaluation of Surface Cracks in Round Bar under Mode III Loadings

5. STRESS CONCENTRATIONS. and strains in shafts apply only to solid and hollow circular shafts while they are in the

NORMAL STRESS. The simplest form of stress is normal stress/direct stress, which is the stress perpendicular to the surface on which it acts.

NON-LINEAR FRACTURE BEHAVIOR OF DOUBLE CANTILEVER BEAM

A new computational method for threaded connection stiffness

G1RT-CT D. EXAMPLES F. GUTIÉRREZ-SOLANA S. CICERO J.A. ALVAREZ R. LACALLE W P 6: TRAINING & EDUCATION

Plasticity R. Chandramouli Associate Dean-Research SASTRA University, Thanjavur

The effect of a crack-tip radius on the validity of the singular solution

Apparent thermal conductivity of periodic two-dimensional composites

Available online at ScienceDirect. 20th European Conference on Fracture (ECF20) Yu.G. Matvienko*

Determination of Dynamic Fracture Toughness Using Strain Measurement

CRITICAL CONDITIONS OF PRESSURIZED PIPES

University of Sheffield The development of finite elements for 3D structural analysis in fire

A novel approach to predict the growth rate of short cracks under multiaxial loadings

Prediction of Elastic Constants on 3D Four-directional Braided

PROPAGATION OF CURVED CRACKS IN HOMOGENEOUS AND GRADED MATERIALS

SIZE EFFECTS IN THE COMPRESSIVE CRUSHING OF HONEYCOMBS

CHAPTER THREE SYMMETRIC BENDING OF CIRCLE PLATES

Proceedings of the ASME th International Conference on Ocean, Offshore and Arctic Engineering OMAE2016 June 19-24, 2016, Busan, South Korea

Lecture #8: Ductile Fracture (Theory & Experiments)

Optimization of blank dimensions to reduce springback in the flexforming process

Outline. Tensile-Test Specimen and Machine. Stress-Strain Curve. Review of Mechanical Properties. Mechanical Behaviour

MMJ1133 FATIGUE AND FRACTURE MECHANICS E ENGINEERING FRACTURE MECHANICS

On characterising fracture resistance in mode-i delamination

Stresses Analysis of Petroleum Pipe Finite Element under Internal Pressure

Siping Road 1239, , Shanghai, P.R. China

Supplementary Figures

Burst pressure estimation of reworked nozzle weld on spherical domes

On the study of elastic wave scattering and Rayleigh wave velocity measurement of concrete with steel bar

J. Sladek, V. Sladek & M. Hrina Institute of Construction and Architecture, Slovak Academy of Sciences, Bratislava, Slovakia

Prediction of Elastic-Plastic Behaviour of Structures at Notches

CRACK INITIATION CRITERIA FOR SINGULAR STRESS CONCENTRATIONS Part I: A Universal Assessment of Singular Stress Concentrations

Laboratory 4 Bending Test of Materials

Pressure Vessels Stresses Under Combined Loads Yield Criteria for Ductile Materials and Fracture Criteria for Brittle Materials

Determination of Poisson s Ratio of Rock Material by Changing Axial Stress and Unloading Lateral Stress Test

V Predicted Weldment Fatigue Behavior AM 11/03 1

Lecture 7, Foams, 3.054

Transcription:

International Journal of Solids and Structures 46 (2009) 2629 2641 Contents lists available at ScienceDirect International Journal of Solids and Structures journal homepage: www.elsevier.com/locate/solstr Effect of residual stresses on the crack-tip constraint in a modified boundary layer model X.B. Ren a, Z.L. Zhang a, *, B. Nyhus b a Department of Structural Engineering, Norwegian University of Science and Technology (NTNU), N-7491 Trondheim, Norway b SINTEF Materials and Chemistry, N-7465 Trondheim, Norway article info abstract Article history: Received 26 May 2008 Received in revised form 26 January 2009 Available online 26 February 2009 Keywords: Residual stresses Crack-tip constraint Cleavage fracture Failure assessment Residual stresses play a crucial role in structural integrity assessment. In this study, a large cracked cylinder with a weld in the center is applied to investigate the effect of residual stresses on the cracktip constraint. A modified boundary layer model with a remote displacement-controlled elastic K-field and T-stress under small scale yielding has been used to simulate the problem. A two-dimensional tensile residual stress field due to the weld is introduced into the model by the so-called eigenstrain method. It has been shown that the residual stresses can significantly elevate the crack-tip constraint and thus increase the probability for cleavage fracture. The constraint parameter R introduced by the authors can be used to rank the crack-tip constraint induced by the bi-axial residual stresses. The R value decreases with the increase in the applied J-integral. The residual stress-induced constraint is also coupled with the T-stress. The R value becomes smaller with larger T-stress. Ó 2009 Elsevier Ltd. All rights reserved. 1. Introduction * Corresponding author. Tel.: +47 73 59 25 30; fax: +47 73 59 47 01. E-mail address: zhiliang.zhang@ntnu.no (Z.L. Zhang). Process-induced residual stresses are inevitable for most mechanical or thermal operations used in processing engineering materials. Residual stresses have a significant effect on structural integrity assessment of welded structures. The proper treatment of the welded residual stresses in integrity assessment is becoming an increasing important safety factor when high strength steels are utilized more widely in the offshore industry. In practice, it has been demonstrated that the residual stresses are either overestimated in most cases or underestimated in others (Dong and Brust, 2000). Significant progress has been made in the prediction of the welding residual stresses, and several commercial finite element codes are available on the market (WeldsimS, SYSWELD). The study of the effect of residual stresses on the crack-tip driving force, constraint, failure mechanisms and integrity assessment has also received attention recently. The studies carried out by Panontin and Hill (1996) and Hill and Panontin (1998) confirm that the residual stresses contribute to both the crack driving force and the crack-tip constraint. Xu and Burdekin (1998) investigated the effect of residual stresses on the crack-tip constraint and found that the tensile residual stresses parallel to the crack flank increase the constraint at the crack tip while compressive residual stresses in this direction have the opposite effect, but a biaxial nonlinear residual stress state may also increase the crack-tip constraint despite the residual stress component parallel to the crack flank being compressive. Lei et al. (2000) observed that the standard Rice s J-integral (Rice, 1968) becomes path dependent in the presence of a residual stress field. They suggested a modified J-integral that is path-independent for both a pure residual stress field and a combination of residual stress field and the applied load. Residual stresses were also shown to affect fracture mechanisms, such as brittle fracture, plastic collapse, fatigue, creep, stress corrosion cracking (Withers, 2007) as well as hydrogen embrittlement (Toribio and Elices, 1991). Understanding how residual stresses along with the presence of known or worst case defects affect the limit life without unnecessary conservation is crucial for the industry. Most recently, the residual stress-induced crack-tip constraint has been investigated by Liu et al. (2008) by using single edge notched bending specimens and a one-dimensional residual stress field. A new parameter R was proposed. By combining the CTOD (Crack Tip Opening Displacement) and Q-stress (O Dowd and Shih, 1991, 1992), a so-called CTOD-Q-R three-parameter formulation is proposed to describe the near-tip stress field. In this study, a modified boundary layer (MBL) model has been used to investigate the effect of a two-dimensional residual stress field on the crack-tip constraint. An ideal problem, a large round cylinder with a spot weld in the center, was studied, see Fig. 1. A rigid analytical surface was placed on the symmetrical line to model the contact of the crack surfaces. The cylinder was simulated by an MBL model with a remote boundary controlled by the elastic K-field and T-stress, and with a crack located at the weld 0020-7683/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.solstr.2009.02.009

2630 X.B. Ren et al. / International Journal of Solids and Structures 46 (2009) 2629 2641 Fig. 1. Illustration of the problem. (a) The round cylinder; (b) spot welding with radius of c in the center and one sharp crack is introduced; (c) applied load. metal. Material property mismatch between the weld metal and the base metal has not been taken into consideration. The overall aim is to study the structure of the near-tip stress field in the presence of two-dimensional residual stresses and the relationship of this field to the applied J-integral and T-stress. The paper is organized as follows: the studies of crack-tip constraints including the geometry constraint, mismatch constraint, prestrain history-induced constraint are reviewed first. The numerical procedure for this study is briefly discussed in Section 3. The findings and results are discussed in detail in Section 4. Finally, the paper ends with the concluding remarks. 2. Crack-tip constraint In a weldment there are basically four factors which influence the level of a crack-tip constraint. Constraint in fracture mechanics is a term that is widely used but vaguely defined or understood. In the present context we prefer to understand the level of constraint as an indicator of the near-tip stress state, and the constraint is regarded as the factors or conditions which influence the transferability and invalidate the one-to-one relation between the crack driving force and near-tip stress field. The geometry constraint is caused by crack size, specimen dimensions and loading mode; inhomogeneous material properties can induce the mismatch constraint at the crack tip (Zhang et al., 1996; Betegón and Peñuelas, 2006; Burstow et al., 1998); both the prestrain history (Eikrem et al., 2007) and the welding residual stresses influence the crack-tip constraint as well. In recent decades, a series of studies has been carried out to characterize the different crack-tip constraints. According to William s solution, the first two terms of smallstrain linear elastic expansion of the crack-tip stress field possess the following form (Williams, 1957), K I r ¼ pffiffiffiffiffiffiffiffi f ðhþþtd 1i d 1j ð1þ 2pr where K I is the mode I elastic stress intensity factor and T is a stress parallel to the crack. Larsson and Carlsson (1973) demonstrated that the second term in the series was important to modify the boundary solution to fit the real crack problem, and the T-stress has a significant effect on the plastic zone size and shape. Du and Hancock (1991) studied the effect of T-stress on the small scale yielding field of a crack in plain strain conditions and found that a positive T-stress causes plasticity to envelop the crack tip and exhibits a Prandtl field. This corresponds to the limit solution of the HRR field (Hutchinson, 1968; Rice and Rosengren, 1968) for a non-hardening material, while a compressive T-stress reduces the stress triaxiality state at the crack tip. Betegón and Hancock (1991) suggested a two-parameter framework J-T to characterize the effect of the constraint induced by the geometry. But, T-stress is only valid in an elastic regime. O Dowd and Shih (1991, 1992) developed the J Q two-parameter theory and gave a precise meaning to the term constraint caused by the geometry and loading mode. They showed that the full range of high- and low-triaxiality fields within the J Q annulus are members of a family of solutions parameterized by Q when distances are normalized by J=r 0, where r 0 is the yield stress. The near-tip stress field can be expressed by two-term expansion: r ¼ r HRR þ Qr 0 r J=r 0 where r HRR ¼ q ^r ðh; nþ ð2þ 1 J nþ1 r a 0 r 0 ~r ðh; nþ 0 I n r is the J-controlled HRR stress field, r and h are polar coordinates centered at the crack tip; n is the hardening component; 0 is the yield strain ( 0 ¼ r 0 =E), and a is a material constant. Their study showed that jqj 1; and when jhj < p=2, 1 < r=ðj=r 0 Þ < 5, the stress components ^r rr ^r hh constant and j^r rh jj^r hh j. Thus, Q is a hydrostatic stress parameter. In this two-parameter formulation, J sets the size scale over which large stress and strains develop, and Q characterizes the crack-tip stress distribution and the stress triaxiality achieved ahead of the crack. Q is therefore a quantitative measure of the crack-tip constraint caused by geometry. It must be noted that the J Q theory fails to characterize the crack-tip fields and quantify the constraint level in a bending-dominated large deformation regime. Zhu and Leis (2006) proposed a bending modified J Q theory, by which the crack-tip stress fields for bending specimens at all deformation levels can be characterized. In welded components, the crack located at the interface between the weld metal and the heat affected zone is generally the most critical one. Because of the nature of welding, there is often a mismatch between the weld metal and the base metal. By considering the interface crack as a bi-material system, Zhang et al. (1996) carried out a numerical investigation on the near-tip stress field and found that the near-tip field in the forward sector can be separated into two parts. The first is characterized by the J-integral for a reference material; the second part which influences the absolute levels of stresses at the crack tip and measures the deviation of the field from the first part can be described by a mismatch constraint parameter, M (Zhang et al., 1997b): r r Ref ðjþþmr 0Ref f M ðh þ 12bÞ ð4þ where b ¼ 0 for overmatch and b ¼ 1 for undermatch, r 0Ref is the yield stress of reference material and f M represents the angular function of the difference fields caused by mismatch, which only depends on the reference material. The study also showed that radial dependence of M-field is weak. Similar studies have been car- ð3þ

X.B. Ren et al. / International Journal of Solids and Structures 46 (2009) 2629 2641 2631 ried out for the crack in the middle of weld (Betegón and Peñuelas, 2006; Burstow et al., 1998). Zhang et al. (1997a) further studied the effect of T-stress on the crack-tip stress field of an elastic plastic interface crack. They found that the T-stress can shift the near-tip stress level up and down without significantly affecting the mismatch constraint parameter M, which indicates that the constraints caused by geometry and mismatch are independent of each other. A so-called J Q M formulation to describe the near-tip stress field in the presence of both geometry and material mismatch constraints was then proposed: r r M¼0;T¼0 Q þ Qr 0Ref ^f ðhþþmr M 0 Ref ^f ðh þ 12bÞ ð5þ Here, the Q parameter describes the geometry constraint. Similarly, the M value is used to rank the material mismatch effect on the crack-tip constraint. Plastic prestrain history common in reeled pipes has also been found to influence on the crack driving force and crack-tip stress field (Eikrem et al., 2007). By considering single prestrain cycles, the authors developed a new parameter to quantify the prestrain induced crack-tip constraint: P ¼ Mr hh r hh ¼ r hh ; h ¼ 0 ð6þ r 0 x=d¼4 r 0 r 0 ¼0 x=d¼4 where r hh h 0 implies the case with prestrain history and r hh h 0 denotes the monotonic loading case. Thus, P value represents the ¼0 amplitude of the difference stress field caused by the prestrain history and can be used to rank its constraint. A three parameter formulation CTOD Q P is suggested to describe the near-tip stress field of a cracked specimen with prestrain history. The formulation has the following form: r hhðxþ ¼r¼0 hh ðx; d; r 0; nþþq þ P at h ¼ 0 ð7þ It has been shown that the residual stresses also have a significant effect on the crack diving force and crack-tip constraint. Liu et al. (2008) studied a one-dimensional residual stress field perpendicular to the crack plane in single edge notched tension and bending specimens. They showed that residual stress can enhance the crack-tip constraint and defined a parameter R to characterize the effect. Following the same approach for investigating crack-tip constraint (O Dowd and Shih, 1991, 1992; Zhang et al., 1996; Eikrem et al., 2007) the structure and behavior of the near-tip stress field under the combined load of a two-dimensional residual stresses field and external load in a well defined MBL model is studied, and the parameter R used to quantify the constraint induced by residual stresses is further investigated. R represents the difference between the full near-tip stress field and reference field. 3. Numerical procedure 3.1. Finite element model The modified boundary layer (MBL) model used for this study consists of a weld metal region located in the center of the model and an outer base metal region. The problem considered is a plane strain formulation with an assumed crack in the center of weld metal region. The load was applied to the remote edges of the model through a displacement field (u, v) controlled by the elastic asymptotic stress field of a plane strain mode I crack 1 þ m uðr; hþ ¼K I E 1 þ m vðr; hþ ¼K I E rffiffiffiffiffiffi r cos 1 ð3 2p 2 h 4m cos hþþt rffiffiffiffiffiffi r sin 1 2 h ð3 4m cos hþ T 2p 1 m2 r cos h E mð1 þ mþ r sin h E ð8þ p where K I ¼ ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi EJ=ð1 m 2 Þ under plane strain condition; E is Young s modulus, m is Poisson s ratio; r and h are polar coordinates centered at the crack tip with h ¼ 0 corresponding to the crack tip. ABAQUS/CAE was used for the analyses. Only the upper-half plane was modeled because of symmetry. The crack is assumed to be a sharp crack without initial radius and the radius of the MBL model was taken as 1000 mm to ensure that the small scale yielding condition is fulfilled. The model was meshed by standard eight-node elements with reduced integration, CPE8R, with a finer mesh in the crack-tip region and the interface between the weld metal region and the base metal region. The size of smallest elements near the crack tip is 0.1 mm. The finite element model has 1408 elements and the meshes are shown in Fig. 2. 3.2. Material properties The weld metal and base metal were assumed to have the same elastic properties (E ¼ 2 10 5 MPa, m ¼ 0:3) and plastic properties. The rate-independent power law strain hardening materials were assumed to have the form of r f ¼ r 0 n ð9þ 1 þ p 0 where r f is the flow stress; p is the equivalent plastic strain, r 0 ¼ 400 MPa the yield stress, 0 ¼ r 0 =E the yield strain and n is the plastic strain hardening exponent. Different thermal expansion coefficients for the base metal (a b ) and weld metal (a w ) are assumed to introduce the residual stresses into the model by so-called eigenstrain method. It should be noted that the thermal expansion coefficients here are not physical thermal coefficients, but are just used to introduce the residual stress into the computational model. The eigenstrain method was also called inherent strain method when first introduced by Ueda et al. (1975). The basic idea of the eigenstrain method is that the source of residual stress is an incompatible strain field caused by plastic deformation, thermal strains and phase-transformation etc. (Hill and Nelson, 1995). Thus, if the distribution of the eigenstrain is known, the distribution of residual stresses can be obtained through linear elastic calculation by using the finite element method. In this study, the isotropic distribution of eigenstrain in both the base metal and the weld metal regions was assumed; the eigenstrain values were set to be equal to the thermal expansion coefficients of two regions respectively. The residual stresses were then introduced Fig. 2. Modified boundary layer model, (a) global mesh; (b) crack-tip mesh.

2632 X.B. Ren et al. / International Journal of Solids and Structures 46 (2009) 2629 2641 by loading the model with a unit temperature change. Similar approaches were also applied in the literature (Hill and Nelson, 1996; Mochizuki et al., 1999), and their results showed that the distributions of residual stresses obtained by such approaches agreed well with the experimental results. 3.3. Residual stress field introduced by eigenstrain method In this study, a spot weld with a round shape located in the center of the model was assumed. The eigenstrain for the base metal was assumed to be zero, and isotropic non-zero value for the weld was assigned. The size of the weld region was described by radius c, and three different sizes were investigated. Fig. 3 shows the redistributed residual stress after the crack was introduced for the case with a w ¼ 0:003 and a b ¼ 0. The stress components were normalized by the yield stress, and the distance from the crack tip was normalized by c. It can be seen that the residual stresses along both the parallel and opening directions have a sharp turning point at the interface between the base metal and weld metal. The reason for this sharp turning point is that the assumption of eigenstrain distribution is not continuous between the two regions. It can also be observed that the normalized residual stress fields collapse into one curve for different radius of weld region. Both the parallel and opening residual stress components in the weld metal are tensile and the peak values are approximately 1000 and 1520 MPa, respectively. In the base metal region, the residual stress parallel to the crack plane is also tensile in a large range while the opening residual stress component is compressive to counter balance the tensile stress in the weld. The effect of the biaxial residual stress on the (a) (b) Fig. 3. The redistribution of the residual stress fields after the crack was introduced. (a)r R 11 ; (b)rr 22. E=r0 ¼ 500, m ¼ 0:3, n = 0.1; aw ¼ 0:003 and a b ¼ 0.

X.B. Ren et al. / International Journal of Solids and Structures 46 (2009) 2629 2641 2633 crack-tip constraint will be investigated in the following by using the residual stress field with weld size c ¼ 20 mm. 4. Results and discussion The J-integral is adopted by the majority of the integrity assessment procedures currently used as the elastic plastic fracture parameter. But for a crack in a residual stress field or the combination of mechanical and residual stresses a general path-independent J-integral appears to be path-dependent (Lei et al., 2000). In our study, the computed J-integral by ABAQUS has been investigated and compared with the applied J-integral (external loading) to the MBL model. Residual stresses as an additional stress field induce an initial J-integral, which is about 0.04% of the final J-integral caused by the combination of residual stress and external load. The residual stress was found to have significant effect on the path independence at the early stage of loading, while the path dependence becomes less severe with the increase of external loading. We also found that the J-integral in both the cases with and without residual stress loses path independence in the finite strain region, beyond which the J-integral are practically path-independent. The difference between the computed J-integral and applied J-integral is 0:2% 2:78% when a residual stress is present with applied J-integral from 200 to 600 N/mm, and it is 1:1% 3:25% for the case without residual stresses. The calculated J-integral has been used in the calculation of the stress field, J in the following means the computed J-integral. 4.1. Reference solution and the Q-field The reference solution is important for studying the crack-tip constraint. The stress field distribution according to the HRR singularity or the small scale yielding solution (SSY) from MBL analysis is generally considered as the reference field. The difference between the HRR singularity and SSY solution was found to be very small. Dodds et al. (1993) showed that the choice of HRR field or SSY solution as reference field does not result in any significant difference. But, applying SSY solution as the reference field can extend the applicability of the approach to a much broader range of materials, because the HRR singularity is limited to deformation plasticity. In this study, we used a homogeneous SSY solution without residual stresses and the T-stress as the reference field. Small scale yielding conditions are enforced by not allowing the plastic zone size r p to exceed 0.2 times the radius of the MBL model. The stress distribution obtained from the small strain analysis for T = 0 under different loading levels is presented in Fig. 4. Here, opening stresses are normalized by the yield stress r 0 ; r is the radial distance of the material in the undeformed state measured from the crack tip and normalized by J=r 0. As shown in Fig. 4, the opening stresses for different external loadings collapsed into a single curve. In other words, the reference field is independent of applied load. In this study, solution with J applied ¼ 200 N=mm was taken as the reference one. In order to better understand the effect of residual stresses on the crack-tip constraint, the Q-field is revisited. The Q value in the J Q theory (O Dowd and Shih, 1991, 1992) represents the crack-tip constraint induced by specimen geometry, crack size or loading mode. The small scale yielding solution was used as the reference solution to measure the Q value. Fields of different crack-tip constraint levels were induced by applying different combinations of K and T. Bilby et al. (1986) showed that the near-tip stress distribution depends on T, but is independent of K. Therefore, the K-field was fixed in this section, but the T-stress varied in the range of 1 < T=r 0 < 1. The effect of loading path will be further investigated in Section 4.5. The opening stresses of the case with different T are shown in Fig. 5. It can be seen that the finite-strain effect is significant in the range r=ðj=r 0 Þ < 1, beyond which the stress field shows similarity. The stress distribution of T=r 0 ¼ 0 corresponds to the small scale yielding solution and the stress distribution for T=r 0 ¼ 0:5 and 1 are almost identical. The reason for this is that the crack-tip field will approach full plasticity and a further increase of T-stress does not change the crack-tip field anymore when T=r 0 is greater than certain value, as was shown by Du and Hancock (1991). However, negative T=r 0 values cause a significant downward shift of the stress field. O Dowd and Shih (1992) have demonstrated that there was a one-to-one correspondence between T and Q under the conditions that the remote stress field is given by the first two terms of the small-displacement-gradient linear elastic solution (Eq. (1)), in which the applied load and geometry affect Q only through T. i.e. Q ¼ FðT=r 0 ; nþ ð10þ Fig. 4. Small scale yielding solution without residual stresses and the T-stress. E=r 0 ¼ 500, n = 0.1, m ¼ 0:3.

2634 X.B. Ren et al. / International Journal of Solids and Structures 46 (2009) 2629 2641 Fig. 5. Opening stresses at different T-stress. E=r 0 ¼ 500, n = 0.1, m ¼ 0:3. Here, Q also depends on the material hardening component n, but the dependence is weak. The relationship between Q and T in this study is shown in Fig. 6. 4.2. Stress field with residual stress and the definition of R As mentioned in Section 2, residual stresses affect the crack-tip constraint. However, the issue of how to quantify the effect remains open. In order to investigate the effect of residual stress fields on crack-tip constraint, different biaxial residual stress fields (see Fig. 3) were introduced by varying the eigenstrain values under the same external loading (J applied ¼ 200 N=mm) controlled by the K-field. The stress distributions including residual stresses are compared with the reference SSY field in Fig. 7. Here, the stress components along the crack line (h ¼ 0) were shown in range 0 < r=ðj=r 0 Þ < 5. It can be seen from Fig. 7 that the presence of residual stresses elevates the stress level compared with reference solution, and the elevation of the stress level increases with the increase of the eigenstrain level. It can be observed that the finite-strain effects are significant in the range r=ðj=r 0 Þ < 1:5, beyond which the stress distributions seem to be parallel to each other. It should be noted that the magnitudes of normalized opening stress are greater than that of the stress component parallel to the crack flank. Due to the symmetrical condition, the shear stress component is zero. A difference stress field has been calculated between the full stress field with residual stresses and the reference solution (Mr ¼ r with r ref stresses and r ref, where r with is the stress field with residual is reference SSY solution). The difference fields for the three eigenstrain levels a w ¼ 0:002, 0.0025 and 0.003 are shown in Fig. 8. In our earlier work, Liu et al. (2008) showed that the residual stress-induced difference field could be approximated by a hydrostatic stress with both principle components almost identical and shear component zero. However, as shown in Fig. 8, Mr 11 and Mr 22 are different for the same eigenstrain level. With the increase in the eigenstrain level, the difference between Mr 11 Fig. 6. Relationship between Q and T. The Q value was taken out at r=ðj=r 0Þ¼2. E=r 0 ¼ 500, n = 0.1, m ¼ 0:3.

X.B. Ren et al. / International Journal of Solids and Structures 46 (2009) 2629 2641 2635 (a) (b) Fig. 7. Comparison of the reference field and the stress field including residual stresses along h ¼ 0, J applied ¼ 200 N/mm. n = 0.1, E=r 0 ¼ 500, m ¼ 0:3. (a) r 11; (b) r 22. and Mr 22 increases. It should be noted that uniaxial residual stresses perpendicular to the crack flank were used in Liu et al. (2008) while biaxial residual stress fields were introduced in this study. The difference between the present results and the results in Liu et al. (2008) may be explained by the different residual stress components. Biaxial residual stresses tend to have more significant effects on the crack-tip constraint than uniaxial ones. Similar features have also been reported by Xu and Burdekin (1998). It is known that the cleavage fracture is controlled by the critical levels of the opening stress acting over a microstructurally significant distance ahead of the crack tip (Dodds et al., 1993). In order to quantify the effect of residual stresses on the crack-tip constraint, a parameter R can be defined based on the difference in the opening stresses. The same reference stress used in the previous section was used here. The definition of R is: R ¼ r 22 ðr 22 Þ SSY;T¼0 r 0 at r ¼ 2J=r 0 ð11þ The distance r=ðj=r 0 Þ¼2 is chosen so that R is evaluated outside the finite-strain region. It can also be seen that the difference between the finite strain solution with T = 0 and reference small scale yielding solution is negligible when the distance is greater than r=ðj=r 0 Þ¼2. 4.3. Effect of external loading on R Welded structures with residual stresses are subject to various service loading conditions. It is interesting to investigate the effects of external loading on the crack-tip constraint induced by residual stresses. A residual stress field with eigenstrain value a b ¼ 0; a w ¼

2636 X.B. Ren et al. / International Journal of Solids and Structures 46 (2009) 2629 2641 Fig. 8. Difference stress between the full stress field and reference field along h ¼ 0 with T =0.n = 0.1, E=r 0 ¼ 500, m ¼ 0:3. 0:003 was introduced into the MBL model, and the crack-tip constraint was investigated under five external loading levels ðj applied ¼ 200; 300; 400; 500 and 600 N=mmÞ. The opening stresses of combined external loading and residual stresses are shown in Fig. 9 together with the reference solution. Fig. 9 shows that in the presence of residual stresses the cracktip opening stresses exceed the reference solution for all the loading levels considered. However, the opening stress decreases with the increase of external loading. The difference between the opening stresses and reference solution are plotted in Fig. 10(a). The R values quantifying the residual stress-induced crack-tip constraint have been calculated using Eq. (11) and are plotted as a function of the external load in Fig. 10(b). It can be seen from Fig. 10(a) that different stresses under the various external loading levels are parallel to each other to a large extent when r=ðj=r 0 Þ > 1:5. The residual stress-induced constraint R decreases with the increase in the external loading, as shown in Fig. 10(b). Liu et al. (2008) observed a similar trend in their studies. The behavior is in agreement with common knowledge that the external loading and plasticity can reduce the effects of residual stresses. It is also interesting to note that R is different to the mismatch-induced constraint parameter M that depends on the material properties but is independent of external loading and geometry constraint (Zhang et al., 1996, 1997a,b). The above results indicate that the residual stress as an additional stress field has interaction with the applied stress fields and depends strongly on the residual stress field itself. 4.4. Interaction of R and Q In Section 4.2 it has been noted that the specimen geometry, crack size and loading mode influence the crack-tip constraint and the geometry constraint can be characterized by the Q parameter (O Dowd and Shih, 1991, 1992). It is interesting to study how Fig. 9. Comparison of reference field with stress field combining external loading and residual stresses along h ¼ 0 with T =0.E=r 0 ¼ 500, m ¼ 0:3.

X.B. Ren et al. / International Journal of Solids and Structures 46 (2009) 2629 2641 2637 (a) (b) Fig. 10. Effect of external loading. (a) different opening stress fields along h ¼ 0; (b) R as a function of external loading. T =0,n = 0.1, E=r 0 ¼ 500, m ¼ 0:3. the geometry constraint interacts with the residual stress-induced constraint. In a boundary layer formulation, the normalized cracktip stress fields depend on the remote T-stress but are supposed to be independent of the K-field. By changing the T-stress, different geometry constraint levels can be obtained. In the following, the near-tip stress field in the presence of both a residual stress field and T-stress has been investigated. The residual stress is fixed with eigenstrain value a b ¼ 0; a w ¼ 0:003, and seven T-stress values have been considered. The normalized opening stresses at J applied ¼ 200 N=mm are shown in Fig. 11. Fig. 11 shows that the stress field was shifted up and down by different T-stresses compared with the reference solution. It can be seen that the compressive T-stress affects the stress field more significantly than the tensile T-stress. In order to investigate the interaction between R and Q, R was calculated according to Eq. (11) for different T-stresses and compared with Q. Here, it should be noted that R defined in Eq. (11) represents the cracktip constraint induced purely by the residual stress. However, the R value here (designated as R Q ) includes both the residual stress-induced constraint and geometry constraint, i.e. R Q ¼ R þ Q. The comparison of R Q and Q with different T is plotted in Fig. 12. The difference between R Q and Q is the constraint induced by the residual stress, i.e. R showed by the shaded area in Fig. 12. It can be seen that the difference between R Q and Q decreases with the increase of T=r 0, which indicates that the higher the geometry constraint, the weaker the residual stress effect on the crack-tip constraint. Liu et al. (2008) also showed that residual stress induced constraint in the tensile specimen is in general higher than that in the bending specimen. It is known that for the same geometry and crack size, the bending specimen displays a higher geometry constraint than the tensile specimen.

2638 X.B. Ren et al. / International Journal of Solids and Structures 46 (2009) 2629 2641 Fig. 11. Comparison of reference solution with stress field including residual stress and geometry effects along h ¼ 0. n = 0.1,E=r 0 ¼ 500, m ¼ 0:3. The T-stress was applied as a uniaxial tension or compression parallel to the crack flank to change the crack-tip stress triaxiality in the boundary layer formulation. In this study, the T-stress which generates different Q stresses was designated as T Q. The biaxial residual stress component parallel to the crack flank can also be considered as a T-stress component, designated as T R. Thus, the interaction between R and Q can be explained as the interaction of T R and T Q. As shown in Fig. 3, the residual stress component parallel to the crack flank is mainly tensile. Therefore, the superposition of T Q and T R enhances the total T-stress that results in a higher crack-tip constraint. A positive T R can reduce the constraint loss significantly when T Q is compressive while it has a weak effect on the crack-tip constraint when T Q is positive. However, when the crack-tip achieved the full plasticity, a further increase in tensile T- stress does not have any significant effect. 4.5. Effect of loading path on Q and R For the same external displacement field applied at the outer boundary of the MBL model, different loading path may induce a different crack-tip constraint, which was generally neglected by most of the work in the literature. There are generally two loading paths to apply to remote displacements: the proportional loading path controlled by fixing the ratio of K/T (Path I); and the sequent loading path by applying the T-field first and then the K-field (Path II). In this study, the effect of the loading path on the crack-tip constraint was investigated both with and without residual stress cases. T=r 0 ¼ 0:5 and K-field with J applied ¼ 200; 300; 400; 500 and 600 N/mm were studied and the same reference field was used. Fig. 13(a) and (b) show the effect of the loading path on Q and R, respectively. Q R + Q R + Q Fig. 12. Comparison of R and Q. n = 0.1, E=r 0 ¼ 500, m ¼ 0:3, a b ¼ 0; a w ¼ 0:003.

X.B. Ren et al. / International Journal of Solids and Structures 46 (2009) 2629 2641 2639 Fig. 13. Effect of the loading path on the crack-tip constraint with T ¼ 0:5, n = 0.1, E=r 0 ¼ 500, m ¼ 0:3. (a) effect on Q; (b) effect on R. a b ¼ 0; a w ¼ 0:003. It can be seen that proportional loading path can generate a higher crack-tip constraint than the sequent loading path both with and without the residual stress cases. It should be noted that the compressive T-stress represents the loss of crack-tip constraint. Thus, the lower crack-tip constraint induced by loading path II indicates that the T-stress applied as an additional field affects the crack-tip constraint. We also observed that the effect of the loading path on R is stronger than the effect on Q for the same external loading. This indicates that the effect of residual stresses on the crack-tip constraint can be regarded as the superposition of components parallel to the crack flank and the additional T-field. It was also found that under the same geometry constraint level (T=r 0 ¼ 0:5), the effect of residual stresses decreases with the increase in external loading, as shown in Fig. 13(b). 4.6. Effect of hardening component on R Finally, it is interesting to investigate the influence of material properties on the residual stress-induced constraint. The same residual stress field generated by eigenstrain values a b ¼ 0 and a w ¼ 0:003 was introduced and the same external loading J applied ¼ 200 N/mm was applied for three hardening components. Fig. 14 shows the difference in the opening stress between the case with residual stresses and the corresponding reference solution. The values of R for different n were marked by circles. Fig. 14 shows that the difference in the opening stresses increases with the increase of the strain hardening component. R is also higher for the case with stronger strain hardening. For the materials with weak hardening, the crack tip can easily develop

2640 X.B. Ren et al. / International Journal of Solids and Structures 46 (2009) 2629 2641 Fig. 14. Difference opening stress for the stress fields with residual stresses and the corresponding reference solution with n = 0.05, 0.1, and 0.2 along h ¼ 0. E=r 0 ¼ 500, m ¼ 0:3. full plasticity. Thus, for the same residual stress field, its effect on the crack-tip constraint is smaller for weaker hardening materials. It should also be noted that the finite strain effect becomes more significant for a material with weaker strain hardening. 5. Concluding remarks Welding residual stresses are unavoidable and play a crucial role in the integrity assessment procedure. Residual stresses affect both the crack driving force and the crack-tip constraint. This study has focused on the latter effect by using a new parameter to quantify its effect. The modified boundary layer model with a remote displacement field controlled by the K-field and T-stress was used. A two-dimensional residual stress field was introduced into the model by the eigenstrain method. A small scale yielding solution without residual stress and geometry constraint (T=r 0 ¼ 0) was taken as the reference field. It has been shown that the difference in the stress fields between the full stress field with residual stresses and the reference solution show similarity. Unlike previous findings, we found that the stress components of the difference fields parallel and perpendicular to the crack flank are not equal. Thus, parameter R is not a hydrostatic term for the cases examined. Since the cleavage fracture is more sensitive to the opening stress, a parameter R was defined based on the opening stress difference to quantify the welding residual stress-induced constraint. The results showed that external loading can remedy the residual stress-induced constraint R that decreases with the increase in the external loading. R is different to the mismatch-induced constraint parameter M which is independent of the external loading. The results also indicate that the geometry constraint interacts with the constraint induced by the residual stresses. For a higher geometry constraint, the effect of the residual stresses becomes weaker. This could be explained by the fact that the residual stress components parallel to the crack flank interact with the remote T-stress. The study also indicates that a sequential loading path with the T- field taken as an additional field will result in lower crack-tip constraint. The loading path effect is stronger for the cases with residual stresses. The residual stress-induced constraint depends on the material hardening component as well. R increases with the increase of the hardening, in which the near-tip plasticity plays an important role. Full plasticity is easier to cover the crack tip for the weaker hardening material and then the residual stresses have a smaller effect on the crack-tip constraint. The present work is concerned with the residual stress-induced crack-tip constraint. The parameter R was defined to quantify the residual stress-induced crack-tip constraint. Further work will be addressed to implement the residual stress-induced constraint into the failure assessment. Acknowledgement The funding from the Research Council of Norway through the STORFORSK Project No.167397/V30 is greatly acknowledged. References Betegón, C., Hancock, J.W., 1991. Two-parameter characterized of elastic plastic crack-tip fields. J. Appl. Mech. 58, 104 110. Betegón, C., Peñuelas, I., 2006. A constraint based parameter for quantifying the crack tip stress fields in welded joints. Eng. Fract. Mech. 73 (13), 1865 1877. Bilby, B.A., Cardew, G.E., Goldthorpe, M.R., Howard, I.C., 1986. A finite element investigation of the effect of specimen geometry on the fields of stress and strain at the tips of stationary cracks. Size Effects in Fracture. Institution of Mechanical Engineers, London, England, 36 46. Burstow, M.C., Howard, I.C., Ainsworth, R.A., 1998. Influence of constraint on crack tip stress fields in strength mismatched welded joints. J. Mech. Phys. Solids 46 (5), 845 872. Dodds Jr., R.H., Shin, C.F., Anderson, T.L., 1993. Continuum and micromechanics treatment of constraint in fracture. Int. J. Fract. 64, 101 133. Dong, P., Brust, F.W., 2000. Welding residual stresses and effects on fracture in pressure vessel and piping components: A millennium review and beyond. J. Press. Vess. Technol. 122 (3), 329 338. Du, Z.Z., Hancock, J.W., 1991. The effect of non-singular stresses on crack-tip constraint. J. Mech. Phys. Solids 39 (1), 17 26. Eikrem, P.A., Zhang, Z.L., Nyhus, B., 2007. Effect of plastic prestrain on the crack tip constraint of pipeline steels. Int. J. Press. Vess. Pip. 84 (12), 708 715. Hill, M.R., Nelson, D.V., 1995. The inherent strain method for residual stress determination and its application to a long welded joint. ASME Press. Vess. Pip. 318, 343 352. Hill, M.R., Nelson, D.V., 1996. Determining residual stress through the thickness of a welded plate. ASME Press. Vess. Pip. Div. 327, 29 36. Hill, M.R., Panontin, T.L., 1998. Effect of residual stress on brittle fracture testing. In: Fatigue and Fracture Mechanics, vol. 29. ASTM STP, p.1332. Hutchinson, J.W., 1968. Singular behaviour at the end of a tensile crack in a hardening material. J. Mech. Phys. Solids 16 (1), 13 31. Larsson, S.G., Carlsson, A.J., 1973. Influence of non-singular stress terms and specimen geometry on small-scale yielding at crack tips in elastic plastic materials. J. Mech. Phys. Solids 21, 263 277.

X.B. Ren et al. / International Journal of Solids and Structures 46 (2009) 2629 2641 2641 Lei, Y., O Dowd, N.P., Webster, G.A., 2000. Fracture mechanics analysis of a crack in a residual stress field. Int. J. Fract. 106, 195 216. Liu, J., Zhang, Z., Nyhus, B., 2008. Residual stress induced crack tip constraint. Eng. Fract. Mech. 75 (14), 4151 4166. Mochizuki, M., Hayashi, M., Hattori, T., 1999. Residual stress analysis by simplified inherent strain at welded pipe junctures in a pressure vessel. J. Press. Vess. Technol. (Trans. ASME) (USA) 121 (4), 353 357. O Dowd, N.P., Shih, C.F., 1991. Family of crack-tip fields characterized by a trixiality paramater. Part I: structure of fields. J. Mech. Phys. Solids 39 (8), 989 1015. O Dowd, N.P., Shih, C.F., 1992. Family of crack-tip fields characterized by a trixiality paramater. Part II: application. J. Mech. Phys. Solids 40 (5), 939 963. Panontin, T.L., Hill, M.R., 1996. The effect of residual stresses on brittle and ductile fracture initiation predicted by micromechanical models. Int. J. Fract. 82 (4), 317 333. Rice, J.R., 1968. A path independent integral and the approximate analysis of strain concentration by notches and cracks. J. Appl. Mech. 35, 379 386. Rice, J.R., Rosengren, G.F., 1968. Plane strain deformation near a crack tip in a power law hardening material. J. Mech. Phys. Solids 16 (1), 1 12. Toribio, J., Elices, M., 1991. Influence of residual stresses on hydrogen embrittlement susceptibility of prestressing steels. Int. J. Solid Struct. 28 (6), 791 803. Ueda, Y., Fukuda, K., Nakacho, K., Endo, S., 1975. A new measuring method of residual stresses with the aid of finite element method and reliability of estimated values. Trans. Jpn. Weld. Res. Inst. 4 (2), 123 131. Williams, M.L., 1957. On the stress distribution at the base of a stationary crack. J. Appl. Mech. 24, 109 114. Withers, P.J., 2007. Residual stress and its role in failure. Rep. Prog. Phys. 70 (12), 2211 2264. Xu, W.G., Burdekin, F.M., 1998. Effects of residual stresses on constraint and fracture behaviour of wide plates. Roy. Soc. A (454), 2505 2528. Zhang, Z., Hauge, M., Thaulow, C., 1997a. The effect of T-stress on the near tip stress field on an elastic plastic interface crack. In: Advances in Fracture Research, Proceedings of the 9th International Conference on Fracture, vol. 5. Pergamon, Oxford, Sydney, Australia, pp. 2463 2650. Zhang, Z.L., Hauge, M., Thaulow, C., 1996. Two-parameter characterization of the near-tip stress fields for a bi-material elastic plastic interface crack. Int. J. Fract. 79 (1), 65 83. Zhang, Z.L., Thaulow, C., Hauge, M., 1997b. Effects of crack size and weld metal mismatch on the HAZ cleavage toughness of wide plates. Eng. Fract. Mech. 57 (6), 653 664. Zhu, X.-K., Leis, B.N., 2006. Bending modified J Q theory and crack-tip constraint quantification. Int. J. Fract. 141, 115 134.