PWSCC Crack Growth Modeling Approaches

Flaw indications have been found in some dissimilar metal (DM) nozzle to stainless steel piping welds and reactor pressure vessel heads (RPVH) in pressurized water reactors (PWR) throughout the world. The nozzle welds usually involve welding ferritic (often A508) nozzles to 304/316 stainless steel pipe) using Alloy 182/82 weld metal. The welds may become susceptible to a form of corrosion cracking referred to as primary water stress corrosion cracking (PWSCC). It can occur if the temperature is high enough (usually >300C) and the water chemistry in the PWR is typical of operating plants. The weld residual stresses (WRS) induced by the welds are a main driver of PWSCC. Several mechanical mitigation methods to control PWSCC have been developed for use on a nozzle welds in nuclear PWR plants. These methods consist of applying a weld overlay repair (WOR), using a method called mechanical stress improvement process (MSIP), and applying an inlay to the nozzle ID. The purpose of a mitigation method is to reduce the probability that PWSCC will occur in the nozzle joint. The key to assessing the effectiveness of mitigation is to determine the crack growth time to leak with and without the mitigation. Indeed, for WOR and MSIP, the weld residual stresses are often reduced after application while for inlay they are actually increased. However, all approaches reduce crack growth rates if applied properly. Procedures for modeling PWSCC growth tend to vary between organizations performing the analyses. Currently, the prediction of PWSCC crack growth is based on the stress intensity factors at the crack tips. Several methods for evaluating the stress intensity factor for modeling the crack growth through these WRS fields are possible, including using analytical, natural crack growth using finite element methods, and using the finite element alternating method. This paper will summarize the methods used, critique the procedures, and provide some examples for crack growth with and without mitigation. Suggestions for modeling such growth will be provided.