Source-containing wall-attached phase layers and angular heat-flux redistribution in an internally heated conjugate molten-salt pool: a controlled computational case study

Internally heated liquid pools undergoing wall solidification occur in molten-salt systems, latent-heat thermal storage, metallurgical processing, and severe-accident in-vessel retention. Their behaviour is governed by the coupled interaction of natural convection, phase change, porous-layer mobility, turbulent mixing, internal heat generation, conjugate wall conduction, and competing heat-removal paths. While these mechanisms are individually well established, their relative influence on phase structure, thermal stratification, and angular wall loading is difficult to separate within a single phase-changing calculation. The present study addresses this problem through a verified assessment of porous-phase transport, turbulence treatment, boundary-condition sensitivity, and numerical uncertainty within a common computational framework. A two-dimensional conjugate molten-salt/steel semicircular pool is analyzed using a multi-region OpenFOAM enthalpy–porosity formulation. The study systematically varies mushy-zone resistance, turbulence treatment, turbulent Prandtl number, liquid-fraction formulation, phase-dependent turbulence attenuation, and boundary heat-loss parameters. Numerical credibility is established through independent grid and time-step studies, comparison across two mesh families, conservation and interface checks, and external comparisons with RASPLAV-Salt and the finite-freezing-interval LIVE-L4/L5L experiments. The analysis shows that the wall-attached solid and mushy regions behave as active thermo-hydraulic layers rather than purely passive conductive resistances: internal heat generation, phase state, and hydraulic accessibility jointly determine their interaction with the convecting pool and the cooled wall. Porous resistance primarily influences basal mush development, crust–mush partitioning, local mobility, and bottom-wall heat transfer, while numerical verification distinguishes these robust responses from mesh-dependent upper-wall effects. The LIVE benchmark provides complementary model-selection evidence: the phase-gated SST formulation strongly mixes the stably stratified lower pool, reducing melt superheat and altering crust development, whereas the corresponding calculation without an eddy-viscosity model reproduces substantially more of the measured stratification, melt temperature, downward wall flux, and wall-power partition. Boundary-condition variations significantly affect the global heat-path distribution but have a weaker influence on the principal basal phase-layer response. Relative to earlier internally heated enthalpy–porosity studies, including Xi et al., the present work extends the analysis by examining how porous resistance, turbulence treatment, boundary effects, and discretization propagate through the same conjugate phase-changing system. The results indicate that total solid inventory or a single permeability parameter is insufficient to characterize local wall loading; phase partition, local mobility, thermal stratification, turbulence treatment, and angular heat-flux redistribution must be considered together. These findings provide a physics-based basis for interpreting model behaviour and for improving reduced-order descriptions of internally heated solidifying pools in molten-salt, thermal-storage, and nuclear-safety applications.