Comparison of CFD simulations on two-phase Pressurized Thermal Shock scenarios

Abstract

Small Break Loss of Coolant Accident (SB LOCA) is one of the most severe transients which may lead to Pressurized Thermal Shock (PTS) on the Reactor Pressure Vessel (RPV) wall. During postulated SB LOCA Emergency Core Cooling (ECC) water is injected into the cold leg, where it mixes with the hot coolant. The mixture of cold and hot coolants flows toward the downcomer. Knowledge of transient temperature distribution in the downcomer is necessary to predict thermal gradients in the structural components of the RPV wall. For the prediction of the temperature fields and heat transfer coefficient between the fluid and wall in the cold leg and the downcomer, reliable computational fluid dynamics (CFD) simulations are needed. To validate CFD models for two-phase PTS scenarios numerical simulations of the TOPFLOW-PTS experiments were performed in the framework of the EU NURISP (NUclear Reactor Integrated Simulation Project) project. The paper presents the post-test CFD simulations of a steady-state TOPFLOW-PTS air/water experiment and the pre-test blind simulations of a steady-state TOPFLOW-PTS steam/water case with condensation. CFD simulations were performed with ANSYS FLUENT, ANSYS CFX and NEPTUNE_CFD. The simulations of the air/water test have shown that correct modeling of the ECC jet behavior is essential for the temperature prediction in the cold leg. For modeling these two-phase flows with rather smooth large free surfaces, Reynolds Averaged Navier–Stokes approach seems to be appropriate. The pre-test simulations of steam/water flow predicted a thermal stratification at the entrance of the downcomer. Finally, the simulations of the TOPFLOW-PTS experiments have depicted considerable differences between the codes and the models.