LAUSR

Abstract Reactor pressure vessel (RPV) failure analysis is a thermo-fluid-structure coupled problem involving the melt pool heat transfer and RPV structural failure. It is of great importance to the qualification of severe accident mitigation strategies as well as the general assessment of severe accident risk in light water reactors. Two coupling approaches, i.e. volume loads mapping (VLM) and surface loads mapping (SLM), are widely employed to resolve this problem. The present study was intended to compare the performance of the two coupling approaches by simulating vessel failure problems of the FOREVER-EC2 experiment and a reference boiling water reactor (BWR) during a postulated severe accident. Simulations of the experiment using the two different approaches showed good agreements with experimental data in terms of total deformation in vertical direction of bottom point and wall thickness changes. The spatial distributions of creep strain and total deformation were also similar between the two approaches. For the reference BWR case, good agreements were achieved between the two approaches in predicting the maximum creep strain and the total deformation in vertical direction of bottom point. The deformation rate of the vessel wall was slow at the beginning, and then increased with time, and finally accelerated to a drastic deformation prior to vessel failure. Similar spatial distributions of creep strain and total deformation were also predicted by the two approaches. Although the SLM approach predicted the deformation acceleration began slightly earlier than the VLM, the difference between the predicted failure times was negligible if considering the very long vessel failure time (~ 10 4 s) and various uncertainties. Generally speaking, the VLM approach was computationally more efficient than the SLM, while both approaches had similar performances in terms of their predictability of experiment and applicability to reactor case.