Wall-Modeled Large-Eddy Simulation of a High Reynolds Number Separating and Reattaching Flow.

Area of Science:

• Computational fluid dynamics
• Turbulence modeling

Background:

• Accurate simulation of high Reynolds number flows with separation is crucial.
• Traditional large-eddy simulation with no-slip walls requires very fine grids.
• Wall modeling offers a potential solution for coarse grid simulations.

Purpose of the Study:

• To compare the performance of two Reynolds-averaged Navier-Stokes wall models.
• To evaluate their effectiveness in simulating high Reynolds number separating and reattaching flow.
• To identify the strengths and weaknesses of equilibrium and nonequilibrium wall models.

Main Methods:

• Large-eddy simulation (LES) of flow over the NASA wall-mounted hump.
• Implementation and comparison of two wall models: equilibrium and nonequilibrium.
• Validation against experimental data for low-order statistics and fluctuations.

Main Results:

• Wall modeling significantly improves flow prediction on coarse grids.
• Low-order statistics from wall-modeled LES agree well with experiments.
• Wall-pressure fluctuations match experimental data, but wall shear-stress fluctuations are underpredicted.
• The nonequilibrium wall model outperforms the equilibrium model in the separation and recovery regions.

Conclusions:

• Wall modeling is essential for accurate LES of separated flows on coarse grids.
• The nonequilibrium wall model is more suitable for complex separated and reattaching flows.
• Assumptions of equilibrium wall models are invalid in separated flow regions.