Effect of Crack Orientation on Toughness of Contemporary EW Pipe Seam Welds

Abstract

Abstract Toughness values of pipe steel and welds are key mechanical properties for material specification and integrity assurance. Toughness is influenced by crack size (deep vs. shallow), loading mode (bending vs. tension), crack orientation and sampled microstructure (notch location in weldments). To the best of the authors’ knowledge, no data exist in the open literature for crack orientation effects on Electric Welded (EW) welds. Charpy notch toughness and single-edge-notched bend (SE(B)) fracture toughness tests using surface-notched specimens of three modern (EW) pipe seam welds were performed at −20°C and −45°C, and the results are compared to those for through-thickness-notched (TTN) specimens. Charpy results indicated that the tests were in lower-transition or lower-shelf regions. Charpy impact V-notch absorbed energies (CVNs) of EW welds displayed large scatter; the effect of crack orientation was small. The initiation J-integral values of SE(B) surface-notched specimens were higher than those of TTN specimens, although when brittle fracture initiation occurred the effect of orientation on toughness was negligible. Note that the specimen geometry likely affects the results; surfaced-notched specimens were B×B and through-thickness-notched specimens were B×2B, although they are all standard SE(B) specimens in ASTM E1820. A limited study of geometry effects on crack-tip stress was done in the present work by Finite Element Analysis (FEA). It was found that these geometries all have similar maximum crack-tip opening stresses at the same J value although the stress gradient through the ligament and hence the constraint parameter Q (measured at a distance 2J/σy from the crack tip) and the J-Q curves differ. Thus, differences in fracture toughness values for brittle fracture would be expected to be small because cleavage is stress controlled, although ductile fracture toughness may reflect the differences in J-Q curves. There was considerable scatter in both notch toughness and fracture toughness, and much of this may be attributed to slight variations in notch or crack location with respect to the weld bond line because the toughness can vary markedly across an EW weld. The variation in location is superimposed on the variation in toughness inherent in each weld microstructure. The present data shows that the conventional use of TTN specimens is appropriate for EW weld toughness specification because toughness in this orientation is either equivalent to or lower than the toughness of surface-notched specimens (i.e. it is conservative).