Analysis and prediction of thin-airfoil stall phenomena with hybrid turbulence methodology

Analysis and prediction of thin-airfoil stall phenomena of a NACA 64A006 airfoil are numerically investigated using large-eddy simulation (LES)/Reynolds-averaged Navier-Stokes (RANS) hybrid methodology with a high-order compact differencing scheme. Subsonic flow of M-infinity = 0.17 with the high Reynolds number of Re = 5.8 x 10(6) is considered, and the angle of attack is varied from 4.0 to 11.0 deg. The results illustrate the possibility of the present LES/RANS hybrid methodology for the prediction of the massively separated high Reynolds number flows with laminar separation and turbulent reattachment within more practical computational cost than that of a pure LES approach. Thin-airfoil stalling aerodynamic characteristics are successfully predicted using the LES/RANS hybrid methodology with a high-order compact difference scheme that is less costly than pure LES approaches. For the prediction of thin-airfoil stall phenomena, it is necessary to resolve properly the laminar small bubble near the leading edge at relatively low angles of attack and the growth of the bubble where suction pressure peak collapses with increasing angles of attack. From the instantaneous and time-averaged flows, it is confirmed that the laminar small bubble near the leading edge is a phenomenon that appears when the unsteady small vortices shedding from the leading edge are averaged for a certain length of time. High-order compact differencing scheme provides extremely high-fidelity results for the complicated and separated flowfields associated with a NACA 64A006 airfoil near stall, even under the reasonable number of grid points. Thin-airfoil stall characteristics are well predicted with the numerical transition method implemented in the original Baldwin and Lomax turbulence model that detects the transition point automatically from the computation. The LES/RANS hybrid methodology with the simple transition method is considered to be an effective prediction tool for flows where the RANS can predict the transition reasonably well.