Large-eddy simulation study of unsteady wake dynamics and geometric sensitivity on a skewed bump

Wall-modeled large-eddy simulation of non-equilibrium turbulent boundary layers

“…Here, this separation point is not well predicted from the cases B1 and B2, whereas that from the cases B3 and B4 is similar to the LES result. In addition to the large separation bubble, the thin secondary separation bubble was also observed at the leeside of the bump in the previous LES study by Ching & Eaton (2019). These secondary separation bubbles are captured in Figure 11 in the range of 0.1 < x/H < 0.3 for cases B2-B4.…”

“…The bump angle in the present study is at θ = 10 • and the simulation results are compared with those of previous experimental and LES studies (Ching et al 2018a;Ching & Eaton 2019). The coordinate origin is located on the bottom wall under the center of the bump and the Reynolds number based on the bulk velocity (0.83 m/s), and the bump height is 16,000.…”

“…Figure 11 shows a zoomed-in view of the streamwise velocity contours at z/H = 0 from cases B1-B4 of the present WMLES. In the reference LES study (Ching & Eaton 2019), the separation point on the centerplane was shown to be located around x/H = 0. Here, this separation point is not well predicted from the cases B1 and B2, whereas that from the cases B3 and B4 is similar to the LES result.…”

“…Figures 8, 9, and 10 show the mean streamwise velocity profiles at various streamwise locations from the present WMLES along with the reference experiment and LES data Here, the profile at x/H = −2 is shifted by 2 on the horizontal axis. Lines, present WMLES (cases B1,B2,B3,B4); dashed lines, no-slip LES on the case B3 mesh (without wall model); dash-dot lines, LES (Ching & Eaton 2019). (Ching et al 2018a;Ching & Eaton 2019).…”

“…Lines, present WMLES (cases B1,B2,B3,B4); dashed lines, no-slip LES on the case B3 mesh (without wall model); dash-dot lines, LES (Ching & Eaton 2019). (Ching et al 2018a;Ching & Eaton 2019). Before the flow separates around the top of the bump (i.e., at the upstream section of the bump (Figure 8) and the windward side of the bump (Figure 9(a))), the present WMLES predicts reasonable velocity profiles regardless of the mesh distribution.…”