Accuracy analysis of RANS turbulent models on predicting thermal stratification in the SUPERCAVNA facility

The accuracy in simulating thermal stratification, a critical issue in thermohydraulic research of liquid-metal-cooled nuclear reactors, has garnered significant attention. In this study, a symmetric 3D-CFD model has been established based on the SUPERCAVNA facility to analyze the performance of four RANS turbulent models. The results indicate that the SKE is the most accurate turbulent model in predicting the temperature distribution in the SUPERCAVNA with an average relative temperature error of 2.96 % and a standard deviation of 2.7 %. A parametric study has been conducted using SKE to discuss the effect of turbulent Prandtl number, the value of which is recommended to be between 1.5 and 2.0 from the results. Moreover, the Boussinesq approximation can be regarded as an alternative method for setting the temperature-dependent variables to conduct calculations involving gravity-induced buoyancy. These results are valuable for further investigations of thermal stratification and thermohydraulic research in liquid-metal-cooled nuclear reactors.

Automated summary

In this study, a 3D-CFD model based on the SUPERCAVNA facility was established to analyze thermal stratification in liquid-metal-cooled nuclear reactors. The results show that the SKE model is the most accurate in predicting temperature distribution, and further parametric studies were conducted.