Discontinuous Galerkin Methods for inviscid low Mach number flows

Abstract

In this work we present two preconditioning techniques for inviscid low Mach number flows. The space discretization used is a high-order Discontinuous Galerkin finite element method. The time discretizations analyzed are explicit and implicit schemes. The convective physical flux is replaced by a flux difference splitting scheme. Computations were performed on triangular and quadrangular grids to analyze the influence of the spatial discretization. For the preconditioning of the explicit Euler equations we propose to apply the fully preconditioning approach: a formulation that modifies both the instationary term of the governing equations and the dissipative term of the numerical flux function. For the preconditioning of the implicit Euler equations we propose to apply the flux preconditioning approach: a formulation that modifies only the dissipative term of the numerical flux function. Both these formulations permit to overcome the stiffness of the governing equations and the loss of accuracy of the solution that arise when the Mach number tends to zero. Finally, we present a splitting technique, a proper manipulation of the flow variables that permits to minimize the cancellation error that occurs as an accumulation effect of round-off errors as the Mach number tends to zero.