Study on wavenumber adaptive simulation scaling behavior for channel flow and the simulation of a spatially developing mixing layer

Wavenumber Adaptive Simulation (WAS) is a turbulence model closure based on partial-averaging closure for hybrid turbulence modeling. The turbulent viscosity is computed using k−ω Shear Stress Transport (SST) closure and scaled directly and locally by the unresolved-to-total turbulent kinetic energy ratio, f_k. The scaling function fk is computed via partial integration of the turbulent kinetic energy spectrum, dependent on local cell cut-off wavenumber. This work analyzes the scaling function behavior for the WAS model using Direct Numerical Simulation (DNS) data of a channel flow and the simulation of a spatially developing mixing layer for model validation.

Previous analysis on scaling function behavior was performed using DNS data of forced homogeneous isotropic turbulence with results showing f_k estimated by the wavenumber adaptive model and that computed by filtering DNS data closely matched. This analysis was extended to DNS data of a channel flow. DNS data was filtered, and one-dimensional spectra were calculated and compared to modeled spectrum predictions.

Lastly, a spatially developing mixing layer (r = 0.6) was simulated using WAS. Computational results were compared to experiments for model validation. Additional simulations were performed using the k − ω SST unsteady Reynolds-Averaged Navier-Stokes (URANS) model. Results highlight the improved ability of WAS to resolve unsteadiness and mean profiles compared to URANS models.