Bayesian-updated seismic fragility of reactor containment building considering multiple damage states

The purpose of this study is to perform seismic fragility assessments of the reactor containment building (RCB) in nuclear power plants, considering multiple damage states with the aid of Bayesian framework. Fragility is a pivotal component of seismic performance that expresses the relationship between ground motion intensity and the engineering demand parameters. The accurate estimation of fragility typically requires a large number of nonlinear time history analyses that inevitably spend significant computational time. The current study employs Bayesian updating combined with Markov Chain Monte Carlo (MCMC) simulations to overcome this issue for RCB structures. The framework requires a prior belief (e.g., fragility parameters) that can be obtained from engineering judgment, experiences, previous studies, or simplified linear models. After that, a set of nonlinear time history analyses are performed to update the prior belief and then achieve posterior fragility curves. The findings show that Bayesian updating and MCMC simulation can significantly update the fragility curves of four different damage states with narrower posterior distributions and reduced posterior spread.