大样本恒星演化研究组知识库

自适应网格(Adaptive Mesh Refi nement,AMR)是处理时空多尺度问题的有力工具。本文针对无力场磁流体动力学,将前期工作中设计的基于WENO格式的高阶交错Godunov类型方法推广到自适应网格。其中网格自适应基于并行块结构自适应网格管理软件包PARAMESH实现。此外我们针对磁场的无散条件设计了加细比为2的块结构自适应网格上的无散提升算法。本文提出的自适应算法和无散提升算法的正确性和有效性通过在磁主导等离子体撕裂模不稳定性问题中的应用得到了验证。

Adaptive Mesh Refinement (AMR) is a powerful technique for solving multi-spatio-temporal scale problems. In this work, we applied AMR to a force-free magnetohydrodynamics problem. A high-order WENO-based staggered Godunov-type with constrained transport proposed in our previous work is extended to AMR grids. The AMR is realized using parallel block structured grids management software package PARAMESH. In addition, we designed a divergence-free prolongation scheme for block structured AMR grids with refi nement ratio of two.The proposed schemes are tested on a tearing instability problem in magnetically dominated plasma. Numerical results demonstrated the effectiveness of the AMR and the divergence-free prolongation scheme.