Numerical analysis of retained residual stresses in C(T) specimen extracted from a multi-pass austenitic weld and their effect on crack growth

Abstract

Small scale fracture mechanics test specimens of austenitic stainless steel weld and heat affected zone material are often extracted from non-heat-treated weldments, which contain significant weld residual stresses. Although these stresses are substantially relaxed by the process of specimen extraction, they may still reach levels that can affect subsequent testing if the applied loads are low and deformation is elastic. Long-term creep crack growth testing is one such case, where failure to take account of retained residual stresses could result in unrealistically high measurements of creep crack growth at applied load levels equivalent to those in operating plant. This paper describes a research programme to predict the start-of-creep-test levels of retained residual stress and residual stress intensity factor in compact tension C(T) specimen blanks extracted from non-post heat-treated AISI 316 weldments. The simulations were validated using neutron diffraction and slitting residual stress measurements and stress intensity factor measurement. A pass-by-pass finite element simulation of the original weldment is performed first, and followed by extraction of the C(T) specimen blank. The predicted retained residual stresses in the specimen are compared with residual stress measurements made on similar blank using neutron diffraction, and slitting techniques. The elastic stress intensity factor due to residual stress is then evaluated on the crack plane of the C(T) specimen and compared with experimental measurements made using the slitting method. Good agreement is achieved between measurement and simulation, providing validated basis for future modelling of long term creep crack growth tests.