Large-eddy simulation of helium–xenon binary mixture in a turbulent channel flow with significant heat transfer

This study employs wall-resolved large-eddy simulations (LES) with real thermophysical property models to investigate the turbulent flow and heat transfer characteristics of helium–xenon (He–Xe) binary mixtures in a channel with significant heat transfer while neglecting buoyancy effects. The simulations capture the effects of temperature-dependent density, viscosity, thermal conductivity, and He–Xe stratification. Despite the strong property variations across the channel, the nondimensionalized Favre-averaged velocity and temperature profiles can be largely collapsed into incompressible results through appropriate semi-local, wall-property, or Van Driest transformations, demonstrating the quasi-similarity of near-wall momentum transport. The distributions of Reynolds stress and turbulent heat flux deviate from incompressible behavior, with higher near-wall viscous and turbulent contributions, yet the strong correlation between stress and heat flux implies that the Reynolds analogy remains approximately valid. Although the different thermal diffusion coefficients of the He–Xe mixture lead to helium enrichment in the high-temperature near-wall region, this species stratification exerts a limited impact on the turbulent transport and can be eliminated by the semi-local scaling. The resulting property variations are effectively accounted for by the semi-local scaling, which preserves the collapse of the normalized profiles. Based on the LES data, new semi-local correlations for near-wall root-mean-square velocity, temperature fluctuations, and a modified turbulent Prandtl number (Prt) model for y+ < 20 are proposed. The integrated Prt model significantly improves the agreement with experimental data, bridging the gap between He–Xe mixture physics and Reynolds-averaged Navier–Stokes modeling for advanced gas-cooled and space reactor applications.