K. Abburi Venkata

Round robin prediction of residual stresses in the edge-welded beam R6 validation benchmark problem

A simple austenitic steel beam specimen with a single autogenous weld bead laid along its top edge is an ideal training problem for novice weld modellers. This geometry may be analysed using either 3D or 2D FE models and employing either block-dumped or moving heat source techniques, with a variety of material constitutive models. This paper describes a residual stress simulation round robin performed on the validation benchmark problem of this type recently added to the R6 defect assessment procedure. A variety of solution strategies of increasing complexity are applied to the problem and the results compared with accuracy targets specified in the validation benchmark. It is found that the greatest solution accuracy is obtained using a 3D moving heat source, and mixed isotropic-kinematic material hardening behaviour.Copyright

Assessment of the effect of residual stresses in elastic-plastic fracture of dissimilar welded components

Abstract Residual stresses in welds pose a significant threat to the structural integrity of a component, especially in the presence of defects and are required to be accounted for in assessing component safety. Although the R6 assessment procedure suggests various approximate methods for incorporating these effects in defect assessment, most of them are overly conservative and not very cost-effective. A more reliable approach is to characterise the weld residual stresses around a defect and study how they interact with primary load. The current paper analyses the effects of weld residual stresses on the fracture of a dissimilar weld in the presence of defect. The weld is made between modified 9Cr–1Mo steel and 316LN stainless steel using autogenous electron beam welding. A C(T) specimen was extracted from the centre of the weld and a crack introduced in the fusion zone using electro-discharge machining. The residual stresses around the crack were measured on a grid of measurement points at mid-thickness of the C(T) specimen using neutron diffraction on the strain diffractometer SALSA at ILL, Grenoble. The measured residual stresses around the crack-tip were incorporated into a finite element model and the interaction of these with applied load was predicted under fracture.