A High-Order Compact Limiter Based on Spatially Weighted Projections for the Spectral Volume and the Spectral Differences Method

Abstract

This paper exposes the theoretical developments needed to design a class of spatially weighted polynomial projections used in the definition of a compact limiter dedicated to high-order methods. The spectral volume framework and its integral representation of the solution is used to introduce the degree reduction of the polynomial interpolation. The degree reduction is conducted through a linear projection onto a smaller polynomial space. A particular care is taken regarding the conservativity property and results in a parametric framework where projections can be monitored with spatial weights. These projections are used to define a simple and compact high-order limiting procedure, the SWeP limiter. Then, numerical evaluations are performed using the spectral differences method for the mono-dimensional Euler equations and demonstrate the high-order behavior of the SWeP limiter.