Analysis of Unsteady Behavior in Shock/Turbulent Boundary Layer Interactions with Large-Eddy Simulations

Abstract

Various numerical and physical aspects of shock/turbulent boundary layer interactions (STBLI) are discussed in the context of Large-Eddy Simulations (LES). A spatially developing incoming equilibrium turbulent boundary layer is obtained with a force based tripping mechanism. Previous results are augmented with a detailed analysis of this procedure. In particular, it is shown that relatively coarser meshes may be employed if the strength of the trip is increased, together with an adjustment of the wall thermal condition for a short distance downstream of the trip. This approach yields both adiabatic as well as constant wall temperature results. Additionally, it is shown that precursor events provide estimates for the size of the trip with only a small portion of the domain in the vicinity of the trip, thus yielding an efficient technique. A simulation of a STBLI using this spatially developing incoming turbulent boundary layer confirms the low frequency signal observed in the STBLI by other approaches which use recycling techniques, and reproduces the particular unsteadiness shown in a reference experiment. A detailed analysis is performed of the evolution of the Reynolds stresses through the interaction. The amplification of turbulence energy is quantified, and the evolution of two point correlations is explored. The resultsindicate that the spanwise decorrelation distance first decreases and then increases through the interaction. A preliminary assessment of the effect of pulsed arc filament based actuators is provided. The results indicate a reduction in the separation length when pulsed at a St corresponding to the peak frequency near the reattachment point.

Progress in shock wave/boundary layer interactions

“…Frame (b) shows power spectral densities and highlights the large amplitudes at frequencies much lower than those associated with turbulence. In a recent study by Mullenix and Gaitonde [42], a detailed comparison of LES results was made with new data from Webb et al [43]. The normalized and weighted power spectral density of the pressure signal on the surface is shown in Fig.…”

“…[174]. Mullenix and Gaitonde [42] examined the problem computationally by placing the actuators at the same locations as those of the experiments i.e., upstream of the interaction (x=δ ¼ 52:6) (see Fig. 16(a)), but by exciting them at St¼0.5, representing the peak frequency inside the separated region, rather than the local peak St $ 0:03.…”

“…Elements of unsteadiness in impinging shock SBLI (a) pressure trace versus time[41]; (b) premultiplied power spectral density[41]; and (c) normalized weighted PSD at different points in interaction, compared with experiments[42].…”

Progress in Shock Wave/Boundary Layer Interactions

“…6 taken from Ref. 55 where a Mach 2.3 flow is considered with an impinging shock generated by a 9 o wedge to reproduce the experiments of Refs. 56,57.…”

“…Current efforts are focused on making the process more efficient by determining the required force field for any Mach and Reynolds numbers with small computations focused on signatures in the recovering boundary layer near trip region. 55 Unsteadiness in genuinely 3-D configurations, such as the sharp or double fin, has been predominantly the domain of experiments, mostly performed until the late 1990s. Dolling and colleagues have examined numerous such configurations including swept corners, sharp fins, blunt fins and doublefins.…”