Betegón and Hancock (1) have described the elastic-plastic behaviour of short and deeply cracked bend bars by reference to modified boundary layer formulations based on the first two terms of the Williams (2) expansion. A local cleavage criterion has been applied to these fields to indicate the effect of the loss of constraint on lower shelf toughness of shallow cracked bend bars. The work models the maximum temperature at which cleavage is possible in these geometries to show the effect of constraint and the a/W ratio of cracked bend bars on the ductile-brittle transition temperature.

Stress intensity factor solutions for surface cracks in welded structures are necessary for engineering fracture and fatigue assessments. Solutions are usually obtained by superposition of weld magnification factors derived from the analysis of two-dimensional (2D) edge cracks and planar three-dimensional (3D) surface crack solutions. This paper describes an investigation of the differences between the weld magnification factors for 2D edge cracks and realistic 3D crack shapes.

Dodds and Anderson provide a framework to quantify finite size and crack depth effects on fracture toughness when failure occurs at deformation levels where J no longer uniquely describes the state of stresses and strains in the vicinity of the crack tip. Size effects on cleavage fracture are quantified by defining a value termed JSSY: the J to which an infinite body must be loaded to achieve the same stressed volume, and thereby the same likelihood of cleavage fracture, as in a finite body. In weld metal fracture toughness testing, mismatch between weld metal and baseplate strength can alter deformation patterns, which complicates size and crack depth effects on cleavage fracture toughness. However, the virtually limitless number of weld joint geometry / crack depth combinations preclude calculation of JSSY for each individual case. This study addresses the accuracy with which JSSY for a welded single edge notch bend, SE(B), specimen can be approximated by previously published results for homogeneous specimens. The case of a crack located on the weld joint centerline is treated. The combined effects of weld groove type, degree of mismatch, and crack depth to specimen width (a/W) ratio are considered by performing plane strain elastic - plastic finite element analyses of SE(B) specimens containing a variety of common weld groove details. These results demonstrate virtually no effect of ± 20% mismatch on JSSY if the distance from the crack tip to the weld/plate interface (Lmin) exceeds 5 mm. If Lmin falls below 5 mm, there exists a deformation (applied-J) dependent value of Lmin below which reasonably accurate JSSY estimation is possible. At higher levels of overmatch (50% to 100%), it is no longer possible to parameterize departure of JSSY for a weldment from that for a homogeneous SE(B) based on Lmin alone. Weld geometry significantly influences the accuracy with which JSSY for a welded SE(B) can be approximated by JSSY for a homogeneous specimen at these extreme overmatch levels.

This paper discusses the problem of the accuracy of fracture prediction models when considering crack initiation in a weld in an engineering structure, and how that relates to crack extension in a small scale specimen. Weld metal mismatching consequences may greatly affect the crack growth resistance of such a structure, which makes integrity assessment becoming more complicated. Generally, the fracture mechanics assessment procedure assumes that the crack is located in an homogeneous material of uniform tensile and toughness properties, adopting either the weld metal properties, or the lowest tensile properties and the lowest toughness value of the base metal or the weld metal. However, this solution may lead to overconservatism in some cases and overestimates the critical size of allowable flaws. This paper presents the preliminary results of an original study which aims at determining a fracture criterion that could be applied to welds in conditions where classical fracture mechanics concepts fail. Elastic plastic finite element analyses have been performed for various specimens containing strongly overmatched cracked welds. It has been shown that overmatching can provide a substantial benefit in terms of toughness when considering short cracks in the weld metal. It is recommended also to examine cracks located in the fusion line where difference of strain carrying capacities of base and weld metal generates local concentration of shear stresses, which may affect drastically-the weld toughness properties.

This paper discusses experimental and analytical techniques for characterizing plastic strain around crack tips from deep to short crack depths. Numerically derived load deflections and plastic strain patterns for single edge notched bend bars, fabricated from ASTM A710 Grade A plate, are compared with experimental measurements. The shapes of the plastic zones are experimentally characterized employing visual techniques. Experimental trends in J and stretch zone width as crack depth is varied from the deep to short crack depth are described. Stretch zone width measurements between 0.08 a/W and 0.95 a/W have shown that there is no observable difference in stretch zone width between the deep and shallow notched specimens. This work indicates that a well defined plastic strain pattern develops around the crack, which precedes stable crack growth.

The existence and nature of surface waves initiated from a pop-in have been investigated by finite element calculations. The pop-in was simulated by a sudden release of stresses at the crack tip.

After the first experience with low CTOD values in the heat affected zone (HAZ) of low carbon micro-alloyed steels about 10 years ago, SINTEF proposed a cost effective materials characterisation programme for the Norwegian Petroleum Dirtectorate (NPD). The aim of this NPD programme was to identify the microstructures most prone to brittleness and to rank the HAZ toughness of steel qualities in a quantitative way.

The Heavy Section Steel Technology Programme (HSST) is investigating the influence of flaw depth on the fracture toughness of reactor pressure vessel (RPV) steel. To complete this investigation, techniques were developed to determine the fracture toughness from shallow-crack specimens. A total of 38 deep and shallow-crack tests have been performed on beam specimens about 100 mm deep loaded in 3-point bending. Two crack depths (a ≈ 50 and 9 mm) and three beam thicknesses (B ≈ 50, 100, and 150 mm) have been considered. Techniques were developed to estimate the toughness in terms of both the J-integral and crack-tip-opening displacement (CTOD). Analytical J-integral results were consistent with experimental J-integral results, confirming the validity of the J-estimation schemes used and the effect of flaw depth on fracture toughness. Test results indicate a significant increase in the fracture toughness associated with the shallow-flaw specimens in the lower transition region compared to the deep-crack fracture toughness. The increase in shallow-flaw toughness compared with deep-flaw results appears to be well characterised by a temperature shift of 35°C. There is, however, little or no difference in toughness on the lower shelf, where linear-elastic conditions exist for specimens with either deep or shallow flaws.

A family of self-similar fields provides the two parameters required to characterize the full range of high- and low-triaxiality crack tip fracture states. The two parameters J and Q have distinct roles: J sets the size scale over which large stresses and strains develop, while Q scales the near-tip stress distribution relative to a high triaxiality reference stress state. Careful investigations show that Q accurately describes the evolution of near-tip stress triaxiality over a wide range of loading and crack geometries. The J - Q theory provides a framework which allows the toughness locus to be measured and utilized in engineering applications. Methods for evaluating Q in fully yielded crack geometries and a scheme to interpolate for Q over the entire range of yielding are presented. An indicator of the robustness of the J - Q fields is introduced; Q as a field parameter and as a pointwise measure of stress level are discussed.

The goal of the study was to compare JR-curves from testing on relatively small conventional specimens of a pressure vessel steel with results from large slightly curved surface cracked (SCT) plates subjected to different loading histories. Six SCT experiments were carried out at 20°C and three at 60°C. Unstable crack growth occurred very shortly after crack growth initiation in four of the tests at 20°C, while some amount of stable crack growth was observed in the remaining two and in all of the small specimens. The analysis showed that initiation occurred at roughly the same value of J for all the SCT specimens. For all three SCT experiments carried out at 60°C ductile crack growth was observed. The JR-curves from the SCT tests coincided qualitatively with each other as well as with the results from CT testing. It seems that the JR-philosophy can be used if the temperature is well above the transition temperature. The initiation point is well predicted while some deviations can occur at larger crack growth increments if the loading is mainly of tensile character. This also holds for the initiation event at the lower temperature while for the further growth the situation is less clear.