Grid Study on Blunt Bodies with the Carbuncle Phenomenon

Abstract

the capturing of shock waves when low-dissipative, upwind schemes are used to numerically analyze high speed flows. The carbuncle phenomenon is best illustrated by the distorted bow shock predicted upstream of a blunt body in a supersonic flow. Various cures have been proposed for the carbuncle problem. These cures generally involve adding dissipation to the numerical routine in order to eliminate the carbuncle. This work will show results of a detailed study of how the structured grid aects the carbuncle phenomenon and how well it captures a strong shock. The CFD (computational fluid dynamics) code used to perform this study is AVUS (Air Vehicles Unstructured Solver). During this work it was found that heat transfer profiles in the stagnation region of a hypersonic blunt body are sensitive to perturbations upstream in the flow field. It is believed that the upstream perturbations are errors generated in the shock which are convected downstream to the surface of the blunt body. These upstream perturbations are amplified in the calculated heat flux profiles on the body surface. These errors arise from the Riemann solver which depends on grid quality in the region of the shock. It appears that the grid quality in the region of the shock is a major factor contributing to the inability of some Riemann solvers to accurately predict the flow field in this region. The grid study proposed here will provide recommendations on what types of structured grids should be used to accurately capture strong shocks and accurately predict the heat transfer profile on a body surface.