A probabilistic environmentally assisted cracking (PEAC) model was developed to describe the propagation of primary water stress corrosion cracking for Alloy 600 in roll-transition region of steam generator (SG), a severe environmentally assisted cracking problem in pressurized water reactors (PWRs). In the PEAC model, crack growth rate (CGR) and probability of failure (POF) were obtained by adopting a Bayesian inference that decreases the uncertainties of unknown parameters and their distributions in theoretical equations. The CGR is mainly dependent on three factors: probability of detection (POD), initial crack size distribution, and stress distribution. The POD, which is a logistic link was updated with Bayesian inference based on SG inspection data. The crack size distribution, which is relative to initiation time expressed by a Weibull function, was also updated with Bayesian inference using POD. The stress distribution caused by mechanical rolling is considered to be a major contributing factor along the SG tube. It based on finite element analysis is deterministic model unlike POD and initial crack distribution. According to this model, the uncertainty of hyperparameters in the CGR which are parameters of a prior distribution was reduced, and the appropriate level of confidence was achieved by utilizing the available data. Moreover, a benchmark study for the SG tube was performed to evaluate reliability of Alloy 600 SG components in nuclear power plants. The POF was estimated from the developed PEAC model and failure criteria by taking into account the effects of inspection and repair of defective tubes. The results from this study are applied to demonstrate risk reduction in PWRs by adopting risk-informed in-service inspection.
Large datasets of single carbon and glass fibre mechanical properties obtained with automated testing equipment
