Flow physics and RANS modelling of oblique shock/turbulent boundary layer interaction

Large-eddy simulation (LES) is utilized to investigate flow physics and lower-fidelity modelling assumptions in the simulation of an oblique shock impinging on a supersonic turbulent boundary layer (OSTBLI). A database of LES solutions is presented, covering a range of shock strengths and Reynolds n...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Journal of fluid mechanics 2013-08, Vol.729, p.231-284
Hauptverfasser: Morgan, Brandon, Duraisamy, K., Nguyen, N., Kawai, S., Lele, S. K.
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Large-eddy simulation (LES) is utilized to investigate flow physics and lower-fidelity modelling assumptions in the simulation of an oblique shock impinging on a supersonic turbulent boundary layer (OSTBLI). A database of LES solutions is presented, covering a range of shock strengths and Reynolds numbers, that is utilized as a surrogate-truth model to explore three topics. First, detailed conservation budgets are extracted within the framework of parametric investigation to identify trends that might be used to mitigate statistical (aleatory) uncertainties in inflow conditions. It is found, for instance, that an increase in Reynolds number does not significantly affect length of separation. Additionally, it is found that variation in the shock-generating wedge angle has the effect of increasing the intensity of low-frequency oscillations and moving these motions towards longer time scales, even when scaled by interaction length. Next, utilizing the LES database, a detailed analysis is performed of several existing models describing the low-frequency unsteady motion of the OSTBLI system. Most significantly, it is observed that the length scale of streamwise coherent structures appears to be dependent on Reynolds number, and at the Reynolds number of the present simulations, these structures do not exist on time scales long enough to be the primary cause of low-frequency unsteadiness. Finally, modelling errors associated with turbulence closures using eddy-viscosity and stress-transport-based Reynolds-averaged Navier–Stokes (RANS) simulations are investigated. It is found that while the stress-transport models offer improved predictions, inadequacies in modelling the turbulence transport terms and the isotropic treatment of the dissipation is seen to limit their accuracy.
ISSN:0022-1120
1469-7645
DOI:10.1017/jfm.2013.301