Computational domain and simulation setup for DNS of hypersonic turbulent boundary layers subjected to pressure gradients on smooth walls via backward facing curvature and forward facing curvature done by the OSU group, with the flow conditions and curved-wall geometries and the flow conditions of the DNS matching those measured in TAMU high-speed blown-down wind tunnel (Adapted from Nicholson et al. [5], [6]).
The approach for this INCITE project is to characterize the effects of pressure gradient and wall cooling on boundary-layer turbulence and perform a thorough evaluation of the existing turbulence models as well as the models that are currently under development.
This project aims to enable more accurate predictions for hypersonic boundary layer flows subject to the effects of pressure gradient and wall cooling. The research approach is to (i) develop a direct numerical simulation (DNS) database over parameterized wall conditions that include systematically and continuously varied surface curvature and wall temperature, and (ii) subsequently utilize this DNS database to both characterize the effects of pressure gradient and wall cooling on boundary-layer turbulence and perform a thorough evaluation of the existing turbulence models as well as the models that are currently under development.
The study will create a benchmark quality parameter-sweeping database that is essential to the training and testing of data-driven turbulence models. It will also derive flow statistics, including boundary layer profiles of mean flow, Reynolds stresses, velocity-temperature correlations, surface skin friction and heat flux, as well as various budget terms in the exact equations of turbulent kinetic energy (TKE) and Reynolds-stress transport. The evaluation of existing turbulence models will include a term-by-term comparison of TKE and Reynolds-stress budgets between DNS and the Reynolds-stress transport modeling, as well as an evaluation of the algebraic model for turbulent energy flux based on a priori and a posteriori assessments. The generated DNS datasets and turbulence statistics will be made available to other investigators to develop, improve, and validate turbulence models (both conventional and data-driven) for Reynolds-averaged Navier-Stokes equations.
