Numerical study of transonic buffet on supercritical airfoil with different boundary layer states

Accurate numerical simulations of flow over airfoils play an increasingly important role in the design of aircraft major components such as wings and turbo- machinery blades. These lifting devices often operate in demanding aerodynamic conditions for optimum performances, and may experience the presence of shock waves in operating conditions. Shocks may become unsteady under specific conditions, undergoing a large-scale, low-frequency periodic motion, which affects the entire flow-field. This unsteady phenomenon, named transonic buffet, is the subject of the present numerical investigation, with an oscillating shock over the suction side of the airfoil. In this study, a range of transonic Mach numbers and angles of incidence are considered, but the bulk of the analysis is carried out for flow conditions at free- stream Mach number M∞ = 0.7 and angle of incidence α = 7°, which show well established buffet. Large-eddy simulations (LES) with natural and forced transition carried out at chord Reynolds number Re = 3000000 clearly highlight the effects of the incoming boundary-layer state on the shock oscillations. While a laminar upstream boundary layer yields weak oscillations of the shock, a turbulent incoming boundary layer yields significant buffet. The LES database has been used to establish veracity (or not) of suggested buffet pathways, mainly based on the alleged existence of an acoustic feedback loop. This mechanism is actually found to consist of two separate patterns: coherent pressure disturbances convected from the shock to the trailing edge, and acoustic waves scattered at the trailing edge, feeding the shock motion. Additional exploration of the pressure side role in the unsteadiness reveals that is has but marginal effect on the phenomenon. Direct numerical simulations (DNS) at lower Reynolds number (Re = 300000) suggest a reversal in the previously observed trend. In this case, a laminar incoming boundary layer yields stronger buffet as compared to its turbulent counterpart, highlighting strong dependence of the buffet phenomenon on the Reynolds number when natural transition is considered. In order to passively control buffet, we consider devices whose design is similar to large-eddy break-up devices (LEBU), consisting of a thin circular-arc airfoil placed between shock and trailing edge, with the main goal of: i) breaking the eddies originating at the shock, responsible for the acoustic scattering at the trailing edge; ii) manipulating the acoustic field in the aft part of the airfoil. RANS simulations show potential for this kind of device for complete stabilization of buffet. On the other hand, DNS shows that the device is able to curtail the buffet, but not to eliminate it. Additional tests are needed in order to assess the effectiveness of the control device, whose practical impact might be very larg