The residual stresses imparted by the welding process are a principal factor in the process of primary water stress corrosion cracking (PWSCC) of Alloy 82/182 nickel-alloy (i.e., dissimilar metal or DM) piping butt welds in PWRs. While Section XI of the ASME Code requires that residual stresses are considered in crack growth calculations, there is little guidance or requirement on how to calculate them. Analytical models are frequently used to simulate the welding process in order to predict the residual stress distribution in the weld and base material as an input to crack growth calculations. The crack growth calculations, in turn, have demonstrated a high sensitivity to the welding residual stress distribution inputs.
While significant progress has been made in understanding and reducing the variability in calculated residual stress among modelers as well as the variability in measured residual stress among different techniques, there remains some uncertainty regarding any given measured or calculated distribution. A feasible alternative to calculating through-wall stress distributions with analytical models on a case-by-case basis is to develop a set of standardized through-wall stress distributions that are applicable to DM welds. Examples of standardized through-wall distributions for residual stress are found in numerous consensus code and standards. The benefit of established through-wall stress distributions is that evaluations for flaws in welds would start from a uniform basis on one of the key inputs to the crack growth calculation, reducing the time required to perform and review flaw evaluations.
This paper presents and describes the technical basis for a set of through-wall distributions for common DM welds found in the US nuclear industry. The basis of the distributions include the results of analytical models, including uncertainty, as well as measured data for through-wall stress in DM welds.