Validation of Wall Friction Model in Multidimensional Component of Mars with Two-Phase Flow Experiments Describing ECC Behavior in Downcomer

Nowadays, high-precision analysis of multidimensional phenomena is one of the issues in nuclear engineering to predict thermal–hydraulic phenomena in a nuclear reactor more precisely under accident conditions. For this reason, the nuclear reactor safety analysis codes have adopted 3D component which is for better prediction of the multidimensional phenomena. However, current 3D components have a fundamental flaw in their applicability since most of the constitutive models in the components were ones developed for the one-dimensional pipe flow. Nevertheless, the constitutive models have been validated rarely based on local measurement database that is from multidimensional phenomena. This insufficient validation of constitutive models led to the present study to simulate multidimensional experiments and validate associated physical models. This study focused on the two-phase flow occurred in the upper downcomer region of a PWR (Pressurized Water Reactor) which adopts the direct vessel injection system during the reflood phase of LBLOCA (Large Break Loss of Coolant Accident). In this study, two sets of experiment describing ECC (Emergency Core Coolant) behavior in the upper downcomer were selected for providing benchmark data to assess the wall friction model in 3D component of MARS-KS (MARS-MultiD). One is the 2D film flow experiment and the other is the MIDAS (Multidimensional Investigation in Downcomer Annulus Simulation) test. Both experiments were conducted at KAERI (Korea Atomic Energy Research Institute). In the former experiment, liquid film velocity and liquid film thickness were measured and these measured data were compared with the simulation results of MARS-MultiD in order to assess the wall friction model in the code. The comparison result was that wall friction model in MARS-MultiD, H.T.F.S. (Heat Transfer and Fluid Flow Service) correlation underestimates the magnitude of the wall friction force. Then the code was modified by introducing Wallis model instead of H.T.F.S. correlation as a wall friction model in MARS-MultiD and the improved results could be obtained with this modified code. After that, the second experiment, MIDAS test, was simulated with the modified code. In this simulation, ECC bypass fraction was calculated and compared with the experimental data. The result showed that the modified MARS-MultiD predicts the trend of bypass fraction more accurately than the default MARS-MultiD. From these assessment results with two experiments, the necessity of modifying the default wall friction model in MARS-MultiD was confirmed to predict the ECC behavior in downcomer precisely.