Abstract
A methodology for evaluating the probability of baffle-former bolt cracking was developed for applicability to PWRs. The methodology is based upon observed IASCC test results for the stainless steels most commonly used for baffle-former bolts (e.g., solution annealed UNS S30400 or S34700 or cold worked S31600) and predictions of the representative stress patterns in those bolts that were developed as inputs to the model. The predictive methodology for IASCC develops a single parameter that incorporates dose and stress ratio (applied stress divided by yield strength, where the yield strength includes irradiation hardening), then defines the statistical distribution of that parameter using a Weibull distribution.
The model was benchmarked against field experience by taking the calculated bolt-by-bolt stresses in a typical PWR, and the accumulated fluence or dose level; then the model was used to predict the probabilities (or numbers) of baffle-former bolt failures due to IASCC over time (i.e., at various Effective Full Power Years). The model provides a probabilistic estimate of the number of cracked bolts that might be expected to be found during any future refueling outage with inspections of the baffle-former bolts. Such a priori knowledge is important because the plan for inspection of the baffle-former bolts may require additional contingencies depending on the likely outcome.
