The carbuncle phenomenon is a numerical instability that manifests itself in problems with strong shock waves in the form of sawtooth perturbations of the shock-wave front, a tumor-like growth on the shock wave, or its nonphysical bending. The aim of this study is to compare methods of suppressing carbuncle phenomenon based on various Riemann problem solvers and to identify the most effective one among them. To achieve this goal, the main methods of combating carbuncle are considered, a methodology for comparing methods of suppressing a carbuncle is outlined, and the selected methods are applied to a test problem. The problem of interaction between a shock wave and a vortex is chosen as the test problem. Two types of errors are identified: startup-errors and the carbuncle. To eliminate the first type of error, it is recommended to use the Godunov, Roe, HLLE, and HLLEC methods (and their modifications). The HLLEs (a dissipative solver) and Rotated Riemann Solver (RRS) approaches are used to eliminate the second type of error. When comparing these approaches, the best result in terms of calculation accuracy is shown by the RRS method using the HLLEsC solver for the flow normal to the physical gap and the HLLEs solver for the flow along the gap.
A. V. Kazantsev (Sun,) studied this question.