The typical experimental procedure for testing stress corrosion cracking initiation involves an interval-censored reliability test. Based on these test results, the parameters of a Weibull distribution, which is a widely accepted crack initiation model, can be estimated using maximum likelihood estimation or median rank regression. However, it is difficult to determine the appropriate number of test specimens and censoring intervals required to obtain sufficiently accurate Weibull estimators. In this study, we compare maximum likelihood estimation and median rank regression using a Monte Carlo simulation to examine the effects of the total number of specimens, test duration, censoring interval, and shape parameters of the true Weibull distribution on the estimator uncertainty. Finally, we provide the quantitative uncertainties of both Weibull estimators, compare them with the true Weibull parameters, and suggest proper experimental conditions for developing a probabilistic crack initiation model through crack initiation tests.
It is extremely difficult to predict the initiation time of cracking due to a large time spread in most cracking experiments. Thus, probabilistic models, such as the Weibull distribution, are usually employed to model the initiation time of cracking. Therefore, the parameters of the Weibull distribution are estimated from data collected from a cracking test. However, although the development of a reliable cracking model under ideal experimental conditions (e.g., a large number of specimens and narrow censoring intervals) could be achieved in principle, it is not straightforward to quantitatively assess the effects of the ideal experimental conditions on model estimation uncertainty. The present study investigated the effects of key experimental conditions, including the time-dependent effect of the censoring interval length, on the estimation uncertainties of the Weibull parameters through Monte Carlo simulations. The simulation results provided quantified estimation uncertainties of Weibull parameters in various cracking test conditions. Hence, it is expected that the results of this study can offer some insight for experimenters developing a probabilistic crack initiation model by performing experiments.
In general, the fatigue life of a safety critical pressure component is estimated using best-fit fatigue life curves (S-N curves). These curves are estimated based on underlying in-air condition fatigue test data. The best-fitting approach requires a large safety factor to accommodate the uncertainty associated with large scatter in fatigue test data. In addition to this safety factor, reactor component fatigue life prognostics requires an additional correction factor that in general is also estimated deterministically. This additional factor known as the environmental correction factor Fen is to cater the effect of the harsh coolant environment that severely reduces the life of these components. The deterministic Fen factor may also lead to further conservative estimation of fatigue life leading to unnecessary early retirement of costly reactor components. To address the above-mentioned issues, we propose a data-analytics framework which uses Weibull and Bootstrap probabilistic modeling techniques for explicitly quantifying the uncertainty/scatter associated with fatigue life rather than estimating the lives based on a best-fit based deterministic approach. We assume the proposed probabilistic approach would provide the first hand information for assessing the maximum and minimum effects of pressurized water reactor water on the reactor component. In the discussed approach, in addition to the probabilistic fatigue curves, we suggest using a probabilistic environment correction factor Fen. We assume the probabilistic fatigue curve and Fen would capture the S-N data scatter associated with the bulk effect of material grades, surface finish, strain rate, etc. on the material/component fatigue life.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.