Numerical Prediction of Turbulent Thermal Mixing in the T-Junction Using ANSYS-CFX Code

Abstract The main objective of this study is to characterize thermal mixing downstream of a T-junction through which two fluids flow at different temperatures inducing cyclical thermal stresses. The generated turbulent mixing is frequently observed in many industrial applications, particularly in nuclear field. The inevitable temperature variations in fluid circuits can induce cyclic thermal stresses in the walls, particularly, at the points where the flows at different temperatures meet and in areas of thermal stratification. This study aims to estimate the predictive performance of the CFD (computational fluid dynamics) code used to model the turbulent mixing phenomenon, with a view to contributing to its validation. The use of numerical simulation makes it possible to analyze the physical mechanisms involved in detail and to extrapolate the results beyond the experimental configurations available. For this purpose, ANSYS-CFX code was used with the LES (large eddy simulation) model. The numerical predictions were compared with reference experimental data from the Vattenfall installation. A good agreement between the two results was found. This work contributes to the understanding of the thermal mixing phenomenon in T-junctions and shows the predictive capability of the CFD codes to predict thermal hydraulic parameters and the responses of the plant performances.