An optimal procedure for fragility analysis of nuclear containment structures under internal pressure

This study aims to develop a novel efficient procedure for determining the fragility function of nuclear containment structures under internal pressure. This procedure can substantially reduce the computational cost by optimizing the required number of nonlinear structural analyses, which is initially unclear in the pressure fragility analysis. The core principle of the proposed procedure is to monitor the changes of the mean ( P m ) and standard deviation ( β S ) of ultimate pressure at failure while gradually increasing the number of structural analyses. The process is terminated when the changes in P m and β S between two consecutive steps drop below a prescribed threshold. The proposed procedure was applied to the pressure fragility analysis of an existing type of containment structure. Since the proposed procedure involves the random selection of additional value sets of the uncertainty parameters at each step, it was repeated 10 times to ensure a fair evaluation. The fragility curve obtained from as small as 40 analyses was nearly identical to that from 100 analyses. On average, over the 10 repeated cases, the computation time was reduced by approximately 47 % compared to the case of 100 analyses. The results confirm that the proposed procedure not only significantly reduced the computational demand but also ensured the reliable generation of the pressure fragility function.