Large-eddy simulation of laminar-turbulent transition in a swept-wing boundary layer

Abstract

The spatial development of laminar-turbulent transition in a 45 swept-wing boundary layer is investigated using the large-eddy simulation approach. Both stationary and traveling disturbances are initiated by steady and random forcings. The numerical simulations reproduce the surface shear streaks typical of the crossow instability observed in experiments. Downstream of the transition location the wall shear stress is found to turn towards the streak direction. The strength of the stationary vortices grows exponentially, initially independent of the unsteady forcing amplitude and saturates at dierent levels during the later development, depending on the magnitude of the traveling vortices. Streamwise velocity contours show the evolution from a wave-like structure to the \half-mushroom" structure for the lower-amplitude, traveling-wave case, leading to double inection points in the wallnormal velocity proles prior to transition; the nonlinear interactions involving the stationary vortices are also much stronger, as indicated by the amplication of higher harmonics of the primary mode. The frequency spectrum of the traveling waves show a high frequency secondary instability prior to transition that, however, is not prominent in the higheramplitude traveling-wave case.

Constrained large-eddy simulation of laminar-turbulent transition in channel flow

“…As pointed out by Huai et al, 34 and verified by the simulations of Schlatter et al 40 predict the transition process. Therefore, it can be inferred that the coefficients of the dynamic Smagorinsky model C DSM s can indicate the process of transition in a certain sense.…”

“…The results showed great agreement with the available data. Later, Huai et al 34 successfully applied a localized dynamic Smagorinsky model [35][36][37] to simulate a spatially developing boundary layer. (Hereafter, we denote the constant coefficient Smagorinsky model and the dynamic Smagorinsky model by SM and DSM, respectively.)…”

Assessment of the shear-improved Smagorinsky model in laminar-turbulent transitional channel flow

“…[28] The LRC correction was able to improve the transition-capturing capability of the standard SM model in the LES of transition in a flat-plate boundary layer. [28,30] In the SFIC correction, the eddy-viscosity is modified as…”

“…[28,30] When the LRC correction is applied to the standard SM model, the Smagorinsky constant C s was chosen as 0.1, and only the fluctuating strain rate was used to estimate the original SGS eddy-viscosity. It was claimed that the LRC correction will be almost inactive in the pesudo-laminar region and very close to the wall, but has some influence on the flow after transition.…”

“…[10][11][12][13][14][15][16][17][18][19] In theoretical research, a boundary layer or a plane channel was usually selected as the canonical problem. [20][21][22][23][24][25][26][27][28][29][30][31][32][33][34][35][36][37] In the present study, a K-type transition in a plane channel flow is chosen as the test case to assess the SISM model.…”