Sensitivity analysis of wall-modeled large-eddy simulation for separated turbulent flow

While there have been some valuable investigations to assess the impact of grid resolution and subgrid-scale (SGS) models on large-eddy simulation (LES) accuracy and convergence, there remains a lack of systematic studies analyzing the sensitivities of wall-modeled LES, particularly for separated turbulent flows. In this study, we conduct a parametric analysis to evaluate the sensitivities of wall-modeled LES with respect to SGS models, grid resolution, and grid aspect ratio, with a specific focus on separated turbulent flow. The evaluations are performed using simulations of flow over a two-dimensional Gaussian-shaped bump at moderately high Reynolds number, which involves smooth-body separation of a turbulent boundary layer under pressure-gradient and surface-curvature effects. The no-slip condition at the wall is replaced by three different forms of boundary condition based on the mean stress-balance equations and the mean wall-shear stress from direct numerical simulation to avoid the additional complexity of modeling the wall-shear stress. Various statistics, including the mean separation bubble size, mean velocity profile, and eddy viscosity from the SGS model, are compared and analyzed. The results reveal that capturing separation bubble strongly depends on the choice of the SGS model. Furthermore, while grid convergence is attainable in DNS-like resolutions, above this limit, the LES predictions exhibit intricate sensitivities to grid size and aspect ratio. Additionally, the boundary condition has a noticeable impact on the flow predictions with coarse grid resolutions.