Scale adaptive turbulence modeling for in-vessel sodium thermal hydraulics

Abstract

Computational fluid dynamics is a powerful tool for the simulation of nuclear reactor coolant flows, such as in sodium fast reactors. In these reactors, the phenomenon of thermal striping -- characterized by oscillatory turbulent mixing of non-isothermal coolant flows -- has the potential to damage the structural integrity of reactor instrumentation and structural materials. At present, large eddy simulation is the only turbulence modeling approach which can sufficiently resolve and predict the mixing behavior of thermal striping, including temperature fluctuation and fluctuation frequencies. The extreme computational cost requirements of large eddy simulation application preclude the use of CFD for large engineering applications. In this work, the performance of the newly developed STRUCT hybrid turbulence model (Lenci, 2016) is evaluated on three representative test cases in comparison to traditional unsteady Reynolds-Averaged Navier-Stokes (URANS) and large eddy simulation (LES) models. Results indicate excellent potential for application of the STRUCT approach to sodium thermal striping flows. Best practice guidelines are developed and discussed.