High Order Large Eddy Simulation For Shock Boundary Layer Interaction Control By A Micro Ramp Vortex Generator PDF Download

Shock wave/boundary layer interactions (SBLIs) are important issues for high-speed vehicles. SBLIs reduce the performance of aerodynamic surfaces and engine inlets, amongst a number of adverse effects. Micro-vortex generators (MVGs) are a flow control method with strong potential to mitigate the effects of SBLI by energizing the boundary layer through momentum transfers from the freestream. They have been implemented in actual configurations at low speeds. The present research includes a combined experimental and theoretical analysis of the evolution of the perturbation downstream the MVG, the formation of vortices, and their interaction with the shock front. Experiments were performed with a baseline MVG configuration of 90° trailing edge on flat plate, ramp alone, and also with MVG mounted ahead of a 20° ramp. The surface flow visualization and particle image velocimetry (PIV) results are presented; the surface flow visualization shows a substantial suppression of SBLIs. A new method to quantify the effectiveness of the MVG on the shock recompression is presented. Moreover, the PIV results were used as the initial input values for the simulation work. A theoretical analysis of the interaction of the MVG perturbation with the boundary layer is performed by assuming linear dynamics of the perturbation. The major assumption is that the interaction between MVG perturbation and the shear flow is affected by transient growth as a result of the non-orthogonality of the linearized Navier-Stokes equations. A new method to perform the projection of the measured perturbation on the continuous modes of the boundary layer is presented. The method takes advantage of the biorthogonality of the direct and adjoint modes. The implementation of such a method using both the Chebyshev polynomials and a shooting algorithm is discussed. The results of the theoretical analysis are encouraging and display a similar trend as the experiments. Both experimental and theoretical results yield perturbation decay not far downstream from the MVG: about 72 mm for experiments, and about 95 mm for simulation. The experiments display two distinguishable growths downstream of MVG, while the simulation predicted one small growth at the very beginning. Both works show trends that agree well with each other.