Guowu Shen

Fracture Toughness Evaluation of High Strength Steel Pipe

Stress fields and constraint parameters (Q and A2 ) of circumferentially-cracked high strength pipe in displacement-controlled tension are compared with those of small-scale single-edge notched samples tested in tension (SE(T)) and bending (SE(B)). The factors affecting transferability of fracture toughness (J-resistance) data from small-scale laboratory tests to cracked high strength pipe are discussed. The crack-tip stress field is of similar form for a circumferential crack in a pipe and a SE(T) test specimen, while for a SE(B) specimen there is a significant gradient in the crack-tip stress field. Hence, the fracture toughness can be characterized by only two parameters (J and Q or J and A2 ) for tension-loaded pipe and SE(T) tests, but for SE(B) tests one more parameter is needed to describe the bending term. It is concluded that the constraint in a SE(T) test with ratio of span between load points to width H/W = 10 provides a reasonable match to that for a circumferential crack in a pipe subjected to tensile loading.© 2008 ASME

Evaluation of Fracture Toughness of X100 Pipe Steel Using SE(B) and Clamped SE(T) Single Specimens

J-resistance testing using a single-specimen unloading compliance technique has been performed on single-edge-notched tension (SE(T)) specimens of X100 pipe steel base material at room temperature and at −20°C, using a procedure developed at CANMET. J-resistance testing using single-edge-notched bend (SE(B)) specimens according to ASTM E1820 was also conducted for comparison. The specimens included two nominal through-thickness pre-crack aspect ratios (a/W = 0.25 and 0.5). The results show that shallow-cracked (a/W∼0.25) bend and tension specimens produce higher resistance curves than deeply-cracked (a/W∼0.5) specimens; ductile propagation was observed at both temperatures. Resistance curves are slightly higher at −20°C than at room temperature for both bending and tension, especially for shallow-cracked specimens. Crack length predicted from unloading compliance of crack mouth opening displacement for the SE(T) specimens was validated by optical measurement of initial crack length (ao ) and final crack extension (Δa>1.0 mm) after heat-tinting, as per ASTM E1820. Predicted crack growths show acceptable agreement with measured values in all cases. The effect of side-groove depth on the resistance curve and straightness of the crack front was briefly investigated. For both bending and tension, resistance curves for 10% (total) side-grooved specimens were close to those from plain-sided specimens when other testing conditions, such as precrack and testing temperature, were the same, whereas 20% (total) side-grooved specimens showed lower toughness. It was occasionally observed that the crack grew faster at the side for 20% side-grooved bend and tension specimens, resulting in a crack front of concave curvature. For 10% side-grooved specimens a rather straight crack front or slightly faster crack growth in the middle of the specimen (convex curvature) was observed.Copyright

Effect of Side Grooves on Compliance, J-Integral and Constraint of a Clamped SE(T) Specimen

The effect of side grooves on crack mouth opening displacement (CMOD) compliance, distribution of J-integral and crack-tip constraint parameters Q and A2 along the thickness of a clamped single-edge-notched tension (SE(T)) specimen were studied by finite element analysis (FEA). Focus was on the effect of depth of side grooves on J-integral and constraint parameters Q and A2 for shallow and deep cracks. The 3-D results were compared with those of SE(T) specimens in plane strain. The results show that the effective thickness equation used in ASTM E 1820 to evaluate compliance of side-grooved SE(B) and C(T) specimens can be used for clamped SE(T) specimens with reasonable accuracy. The results also suggest that the depth of the side grooves affects the distribution of the J-integral: the highest J-integral is at the center of the thickness for a SE(T) specimen with side grooves equal to or less than 10% of total thickness, and near the root of the side grooves for side grooves greater than 10% for a deeply-cracked specimen when the applied load P≥PY . The FEA results also show that the depth of side grooves affects the distribution of the constraint parameters: the crack-tip constraint is highest at the center of the thickness for a specimen with 0% side grooves (plain-sided), and near the root of the side grooves for side grooves equal to or greater than 10%. It was also found from FEA that the crack-tip constraint of a SE(T) specimen with 20% side grooves with shallow (a/W = 0.2) or deep (a/W = 0.5) crack is higher than that of a SE(T) specimen with the same crack depth in plane strain. As a result, the J-resistance of a SE(T) specimen with 20% side grooves may be lower than that of the same specimen in plane strain.Copyright

Fracture Toughness Testing of Pipeline Girth Welds

The guidelines and recommendations for fracture toughness testing of pipeline girth welds outlined in CSA Z662-03, Annex K are reviewed in this work. In Annex K of CSA Z662-03, the specimen type and notch location have been grouped into four categories and the CTOD tests are to be carried out in accordance with either BSI Standard 7448 or ASTM Standard E 1290. In the present study, CTOD tests have been conducted on a manual shielded-metal-arc weld (SMAW) that was prepared in a high strength X80 pipeline steel. The experimental results obtained by applying the two testing standards are compared. The focus was to identify the differences between these two standards that may significantly affect the test results, such as the requirements for straightness of the fatigue crack, and the equations and parameters used for evaluation of CTOD. Some additional factors affecting the testing, such as selection of test specimen location and procedures for targeting specific weldment microstructures as well as the application of local compression, are also discussed. The variation of strength and toughness with clock position around the circumference of the girth welds has also been studied.Copyright

CMOD Compliance of B×B Single Edge Bend Specimens

In ASTM standard E1820, the single edge bend (SE(B)) geometry is one of those recommended for fracture toughness testing. The width to thickness (W/B) ratio recommended in E1820 for this specimen is 2. However, in certain cases, it is desirable to use specimens having alternative W/B ratios; the range of W/B suggested in E1820 is 1 to 4. In E1820, the crack size a may be evaluated during J-integral or CTOD resistance testing using the crack mouth opening displacement (CMOD) elastic unloading compliance C. The equation given to relate a to C using a dimensionless compliance BCE incorporates Young’s modulus E. For the three-dimensional (3-D) SE(B) specimens that are in neither plane stress nor plane strain condition, this parameter E may be considered as a normalizing parameter varying between extremes E (plane stress) and E/(1−ν2) (plane strain) depending on crack depth (a/W) and specimen W/B ratio.In the present study, 3-D finite element analysis (FEA) was used to evaluate the CMOD compliance of B×B SE(B) specimens with shallow and deep cracks and compared with that from Tada’s plane stress equation. Crack sizes evaluated using plane stress and plane strain assumptions with the 3-D CMOD compliance obtained from FEA were compared with the actual crack size of the specimens used in FEA. It was found that the errors in crack size using plane strain or plane stress assumptions can be larger than 5%, especially for shallow-cracked specimens. In the present study, an effective modulus with values between plane stress and plane strain is proposed and evaluated by FEA for the 3-D B×B SE(B) specimens. The values were fitted to a polynomial equation as a function of u = 1/(√(BCE)+1) for use in estimating the dimensionless compliance for crack size evaluation for B×B SE(B) specimens. It is shown that the errors in crack size evaluation can be significantly reduced using this effective modulus.Copyright

Mechanical Properties and Microstructure of Weld Metal and HAZ Regions in X100 Single and Dual Torch Girth Welds

The main objectives of the current study were to further develop tensile and toughness testing protocols and to provide a better understanding of the factors that control both weld metal and HAZ microstructure and properties in pipeline girth welds. In this investigation, two series of rolled (1G) girth welds were made in X100 pipe of 36 in. diameter and 0.750 in. wall thickness using two pulsed-gas metal arc welding process variants: single and dual torch. The small-scale testing program included evaluations of all-weld-metal tensile strength, Charpy impact and standard fracture toughness measured by single-edge bend SE(B) tests, along with preliminary fracture toughness results using a single-edge tension SE(T) test developed at CANMET. Additional information was obtained from detailed microstructural characterizations of weld metal and HAZ regions along with microhardness testing. All-weld-metal tensile tests using round and strip tensile specimens showed variations with through-thickness location and in some case with clock position. Full stress-strain curves were generated, and 0.2% offset yield strength, flow stress, ultimate tensile strength, and uniform strain were measured and compared with pipe properties using calculated weld strength mismatch factors based on these properties. Charpy V-notch transition curves were generated for both weld metal and HAZ (notched within 0.5 mm of the fusion line). Fracture toughness of both weld metal and HAZ regions of single torch welds was assessed using standard SE(B) testing procedures with Bx2B preferred specimens notched through–thickness at the weld centerline and in the HAZ (within 0.5 mm of the fusion line). Full J-resistance curves were measured using SE(T) tests of surface-notched WM and HAZ specimens; the SE(T) test was designed to match the constraint of full-size pipeline girth welds.Copyright

Fatigue Crack Driving Force for Axial Surface Cracks in Pipes

Fatigue life assessment procedures require knowledge of the fatigue crack driving force, such as stress intensity factor range (ΔK) and cyclic J-integral (ΔJ), for the flaw geometry detected during inspection. Because three-dimensional closed-form crack driving force solutions are not available for typical flaws in pipelines, it is common practice to obtain these solutions from finite element analysis (FEA) or to adopt a closed-form crack driving force solution for the equivalent flawed plate and include a correction factor to take account of the pipe bulging effect. In the present study, pipes and plates with an axial rectangular crack with filleted corners under fatigue loading are simulated by FEA. The initial results show that the stress intensity factor range (ΔK) for a thin-walled pipe with a shallow crack (a/t < 0.5) is given reasonably well by the bulging factors given in BS 7910 combined with the stress intensity factor equation given by Newman and Raju for a plate with a semi-elliptical crack. However, the stress intensity factor is significantly over-estimated for a long and deep crack using this procedure. Different parameters for elastic-plastic fatigue are calculated and are proposed to be correlated with the rate of crack growth for thin-walled pipes with an axial rectangular crack with filleted corners. It is intended to use the results presented here in combination with full scale experimental fatigue data to obtain pipeline fatigue crack growth formulations, to accurately predict the rate of crack growth within a pipeline due to fluctuating internal pressure.Copyright

Evaluation of J/CTOD for Clamped SE(T) Specimens With Welds

In BS 7448, Part 2, the stress intensity factor, J-integral and crack tip opening displacement (CTOD) equations developed for evaluation of fracture toughness of a homogeneous material using experimentally measured quantities, such as load-load line displacement, are applied to SE(B) specimens with yield-strength-mismatched welds. The accuracy of this procedure was studied by Gordon and Wang using finite element analysis (FEA). Recently, the so-called “η factor” method for J-integral evaluation of SE(T) specimens with weld-center-line-cracked and yield-strength-mismatched welds was studied by Ruggieri using detailed FEA calculations and the load separation method proposed by Paris et al. For application to strain-based design of pipelines, CANMET has developed equations to evaluate J-integral and CTOD resistance curves for clamped SE(T) specimens of homogeneous materials using experimentally measured load and crack-mouth-opening displacement (CMOD) in a single-specimen procedure similar to that in ASTM E1820. In the present study, the accuracy of using these equations for J-integral evaluation of clamped SE(T) specimens with weld-center-line-cracked and strength-mismatched welds was studied. It was found that the errors in J and CTOD using the equations developed for SE(T) specimens of homogenous materials for these strength-mismatched welds are similar to those for SE(B) specimens with the same weld geometry and mismatch level as reported by Gordon and Wang. It was also found that using the higher of the strength of base and weld metals σY (= (σYS +σTS )/2), (i.e. (σY )w for overmatching and (σY )B for undermatching) in converting J to CTOD gives reasonable and conservative CTOD evaluations for specimens with weld-center-line-cracked and yield-strength-mismatched welds.Copyright

Plastic Collapse Load Determination Using Laboratory-Scale SE(T) Samples Under Fixed-Grip Loading

A high-toughness steel was tested in tension under fixed-grip loading using Charpy-size SE(T) samples. The objective was to determine plastic collapse loads using laboratory-scale samples under conditions similar to those experienced by girth weld defects in pipelines. The selection of the high-toughness steel was to inhibit significant crack extension from notches at maximum loads and this was largely achieved based on fractographic observations and ductile tearing analysis. This paper includes detailed experimental measurements of maximum loads, fractographic observations, and comparison of the experimental values with FEA results in 2-D plane stress and plane strain and in 3-D. The plastic collapse limit in BS 7910 was found to be conservative for plane-sided samples by 17 to 30% (average 23%) for the steel and geometries used in this work.© 2006 ASME