Advanced Turbulence-Heat Flux Sub-Models for Hypersonic CFD Applications

Current RANS-based heat-flux closures assume fixed turbulent Prandtl numbers and gradient diffusion—assumptions that break down in hypersonic, cold-wall, and shock-dominated regimes. We propose a physics-informed turbulence-heat-flux sub-model formulated within the Favre-averaged, reacting Navier–Stokes framework. The model introduces (i) variable Prandtl number linked to local turbulent Reynolds and Peclet numbers, (ii) anisotropic-flux corrections that decouple momentum and energy transport, and (iii) compressibility and wall-temperature dependencies derived from first-principles scaling. All terms depend only on locally available RANS variables, ensuring turbulence-model agnosticism and seamless drop-in compatibility with Navy CREATE-AV/Kestrel and other compressible CFD solvers. Phase I develops the algebraic forms, implements them in a reference FANS solver, and conducts a priori and a posteriori verification/validation against DNS and experimental datasets for wavy-plate, cold-wall and shock-impingement boundary layers. Feasibility is judged by =10 % accuracy improvement in surface heat-flux prediction, numerical robustness across Mach and wall-temperature ratios, and negligible runtime penalty relative to baseline models. The effort culminates in deliverables that include model source, verification cases, and a transition plan jointly authored with the HPCMP CREATE team. Follow-on work will calibrate constants with expanded data, embed the model natively in Kestrel, and execute full-geometry V&V on Navy glide-body and fin configurations. Successful transition offers the Navy a higher-fidelity, modular tool for thermal-protection sizing and trajectory trades, while commercial CFD vendors gain a solver-agnostic plug-in that reduces design conservatism in propulsion, turbomachinery, EV thermal systems, and space-access platforms.