A multi-order candidates weighting framework for discontinuity capturing of hyperbolic conservation laws
Yankai Ma,Meiliang Mao,Zhen-Guo Yan,Jinwei Bai,Huajun Zhu,Xiaogang Deng
DOI: https://doi.org/10.1016/j.jcp.2024.113064
IF: 4.645
2024-05-05
Journal of Computational Physics
Abstract:Weighted compact nonlinear schemes (WCNSs) represent a popular family of high-order methods used for simulating compressible flows. Within these methods, the nonlinear interpolation process is crucial for the scheme's ability to capture shocks. In this study, we introduce a generalized weighting framework that utilizes candidates of varying orders, known as the Multi-order candidates (MOC) weighting, to design nonlinear schemes capable of capturing discontinuities. Analyzing within the MOC framework reveals that the Z nonlinear weights are equivalent to the JS nonlinear weights, supplemented by an additional fifth-order candidate. The convergence accuracy of the MOC weighting is examined, indicating that higher-order candidates in the MOC weighting require smoothness indicators with higher-order leading terms for better accuracy. Based on this convergence analysis, consistently high-order smoothness indicators (CHOSI) are proposed for the MOC weighting, which are essential for achieving the designed high-order accuracy and high resolution. It is found that many existing nonlinear weighting methods can be integrated into the MOC framework. Three of these methods are analyzed and further improved within the MOC framework. Finally, the newly proposed schemes are tested across a range of benchmark problems to showcase their enhanced performance, also illustrating the MOC framework's capability in developing innovative schemes.
physics, mathematical,computer science, interdisciplinary applications