Dynamic Strain Hardening Recovery and Its Influence on Weld Residual Stress Simulation of Stainless Steel Welds

In the past decade or so, stress corrosion cracking (SCC) in the dissimilar metal welds of reactor pressure vessel nozzles has become one of the critical problems faced by the nuclear industry and the regulatory bodies. The existence of high tensile stresses in the weld region is a main driving force for the SCC crack initiation and growth. In a recent international round-robin on weld residual stress validation organized jointly by NRC and EPRI, there exists a significant scattering in the predicted results submitted from 15 organizations around the world. One of the key factors affecting the prediction of the weld residual stress is the selection of strain hardening laws of stainless steels and nickel based super alloys, as these alloys exhibit very strong strain hardening – the ratio of ultimate tensile strength to yield strength typically exceeding 2. It is well known that the deformation at elevated temperatures of these alloys involves dislocation recovery and/or recrystallization, thereby far more complex than the rate-independent elastic-plastic constitutive relations commonly used in computational models for welding residual stress simulations. In this work, the dynamic strain hardening behavior of stainless steels is systematically studied using the Gleeble thermal-physical simulator. Specially designed Gleeble samples and testing procedures are utilized to achieve acceptable level of spatial-uniformity in thermal-mechanical loading conditions during testing which are essential to obtain meaningful temperature and time dependent stress-strain relationship. Gleeble specimens are rapidly heated and cooled during the test that mimics the temperature transients typical of multipass welding of dissimilar metal welds in reactor vessel nozzles. The effects of deformation at elevated temperatures on the room temperature strain hardening are quantified and discussed. The implications of such effects in weld residual stress modeling are evaluated.