Assessment of the MARS Code Using the Two-Phase Natural Circulation Experiments at a Core Catcher Test Facility

A core catcher has been developed to maintain the integrity of nuclear reactor containment from molten corium during a severe accident. It uses a two-phase natural circulation for cooling molten corium. Flow in a typical core catcher is unique because (i) it has an inclined cooling channel with downwards-facing heating surface, of which flow processes are not fully exploited, (ii) it is usually exposed to a low-pressure condition, where phase change causes dramatic changes in the flow, and (iii) the effects of a multidimensional flow are very large in the upper part of the core catcher. These features make computational analysis more difficult. In this study, the MARS code is assessed using the two-phase natural circulation experiments that had been conducted at the CE-PECS facility to verify the cooling performance of a core catcher. The code is a system-scale thermal-hydraulic (TH) code and has a multidimensional TH component. The facility was modeled by using both one-and three-dimensional components. Six experiments at the facility were selected to investigate the parametric effects of heat flux, pressure, and form loss. The results show that MARS can predict the two-phase flow at the facility reasonably well. However, some limitations are obviously revealed.