Computational study of a complex three-dimensional shock boundary-layer interaction

Shock boundary–layer interactions occur in many high-speed aerodynamic flows and they can have a notable impact on design considerations due to the aerodynamic and heat transfer effects. Consequently there is a notable interest in understanding the ability of computational tools to calculate the complex flow fields that can arise in a range of engineering applications. Three-dimensional complex shock boundary layer interaction studies are expensive in both time and computational resources. Although recent studies have begun to focus on the use of more complex computational methods such as large eddy simulations, the aim of this research is to assess the ability of steady Reynolds averaged Navier Stokes turbulence models to simulate the interaction of a planar shock impinging on a cylindrical body under supersonic conditions and to determine if these models have a role to play in engineering design applications. The performance of both eddy viscosity and Reynolds stress models are evaluated relative to an established experimental test case. The impact of Reynolds number and impinging shock strength are also considered. Of the eddy viscosity models it was shown that the Spalart-Allmaras model is unsuitable for this complex interaction and that the k- and Reynolds stress methods both gave notably better agreement with the measured surface static pressures. Overall it was considered that the Reynolds stress method was the best model as it also provided better agreement with the measured surface flow topology. It was concluded that, although a steady Reynolds averaged Navier Stokes approach has known limitations for this type of complex interaction, within an engineering context it can also provide useful results when applied appropriately.