Treatment of iimviscid fluxes in compressible turbulent flow computations

The present study presents a comparative investigation between two methodologies for treating inviscid fluxes, namely, an artificial dissipation-based finite-volume scheme and a second-order upwind difference scheme. Computations made of the turbulent, supersonic flow past an axisymmetric afterbody have revealed certain disadvantages in using the artificial dissipation-based approach to the discretization of inviscid fluxes. Noticeably, the artificial dissipation-based approach does not maintain the integrity of the flow physics and reveals upstream influences of the rapid expansion at the base corner. Accuracy of the results obtained of flow fields with complex features, such as a streamline curvature or strong shocks, depends significantly on the treatment of the inviscid fluxes, and from the comparison presented here, the flux-vector splitting scheme offers better results for the supersonic flow problem involving sharp comers.