Unsteady RANS simulation of OECD-TAMU cold-leg mixing benchmark

Abstract Unsteady Reynolds-Average Navier-Stocks (URANS) simulation was conducted, in which the mixing of two immiscible fluids with different densities were measured using the multi-component material model for the OECD-TAMU cold-leg mixing benchmark. The turbulence models (i.e. realizable κ-e, standard κ-ω and κ-ω SST) were validated against the particle image velocimetry (PIV) data and sensitivity analysis of uncertain input parameters were carried out. Turbulence models for the prediction of mean velocity components and root mean square velocity components, the mixing phenomena in the cold-leg were found in good agreement with the PIV data. However, the turbulence models unable to predict the close agreement with PIV data in the downcomer. Numerical uncertainty of two meshes was conducted for cold-leg and downcomer. It is observed that the numerical uncertainty band is within the range of experimental measurement data. Furthermore, The sensitivity of five uncertain input parameters (i.e. Schmidt number, density difference, heavy fluid viscosity, light fluid viscosity) was analyzed and their influence on the output parameters was noticed. Therefore, the turbulent Schmidt number and density difference, revealing significant effect on the out parameters. As a consequence, the major contribution of turbulent Schmidt number and density difference will be observed in the propagation of input uncertainty.