ICONE19-43592 Improvement of MARS Code Reflood Model

Abstract

A specifically designed heat transfer model for the reflood process which normally occurs at low flow and low pressure was originally incorporated in the MARS code. The model is essentially identical to that of the RELAP5/MOD3.3 code. The model, however, is known to have under-estimated the peak cladding temperature (PCT) with earlier turn-over. In this study, the original MARS code reflood model is improved. Based on the extensive sensitivity studies for both hydraulic and wall heat transfer models, it is found that the dispersed flow film boiling (DFFB) wall heat transfer is the most influential process determining the PCT, whereas the interfacial drag model most affects the quenching time through the liquid carryover phenomenon. The model proposed by Bajorek and Young is incorporated for the DFFB wall heat transfer. Both space grid and droplet enhancement models are incorporated. Inverted annular film boiling (IAFB) is modeled by using the original PSI model of the code. The flow transition between the DFFB and IABF, is modeled using the TRACE code interpolation. A gas velocity threshold is also added to limit the top-down quenching effect. Assessment calculations are performed for the original and modified MARS codes for the Flecht-Seaset test and RBHT test. Improvements are observed in terms of the PCT and quenching time predictions in the Flecht-Seaset assessment. In case of the RBHT assessment, the improvement over the original MARS code is found marginal. A space grid effect, however, is clearly seen from the modified version of the MARS code