Project R-7168


A high-order discretization method for low-Mach flows (Research)


In applications from computational fluid dynamics (CFD), flow problems that are nearly incompressible - i.e., with a small Mach number - pose severe challenges to the numerical discretization. To overcome some of these difficulties, a novel splitting technique, that splits the equation into stiff and non-stiff parts, has been introduced recently; it was named RS-IMEX. In the context of ODEs and low-order methods, it has shown great potential regarding stability and accuracy. The goal of this proposal is to extend the splitting idea to the setting of a high-order discretization in both time and space. Recent state-of-the art methods as the hybridized discontinuous Galerkin method and uniformly convergent IMEX ODE integrators will serve as building blocks. Ultimately, this will yield an efficient method for the solution of a variety of low Mach number problems in a stable and accurate manner.

Period of project

20 June 2016 - 31 December 2019