arrow
返回

A parallel explicit/implicit time stepping scheme on block-adaptive grids

delete2006-09-01
delete68
PRE
AI
DOI:10.1016/j.jcp.2006.01.029delete
delete原文链接
delete原文求助
delete分享
delete收藏
摘要

摘要

En 中文
We present a parallel explicit/implicit time integration scheme well suited for block-adaptive grids. The basic idea of the algorithm is that the time stepping scheme can differ in the blocks of the grid for a given time step: an explicit scheme is used in the blocks where the local stability requirement is not violated and an implicit scheme is used in the blocks where the explicit scheme would be unstable. The implicit scheme is second order in time. The non-linear system of equations is linearized with Newton linearization. The linear system is solved with a preconditioned Krylov subspace iterative scheme. The Schwarz type preconditioning is also based on the block structure of the grid. We discuss load balancing for parallel execution and the optimal choice of the time step for speed and robustness. The parallel efficiency of the scheme is demonstrated for the equations of magnetohydrodynamics with a geophysics application in three dimensions. The control of the numerical divergence of the magnetic field in combination with the explicit/implicit time stepping scheme is also discussed. (c) 2006 Elsevier Inc. All rights reserved.
Keyword:
models
numerical methods
numerical approximation
time dependent initial-boundary value problems
parallel computation
compressible fluids and gas dynamics
magnetohydrodynamics
AI总结

AI总结

对已上传原文的论文进行重点信息的提取,主要内容包括:简要概述、研究摘要、背景介绍、关键亮点、图文解析、展望与总结。

期刊

Journal of Computational Physics 封面图
Journal of Computational Physics
IF:
3.8
论文数:
1.6W
被引数:
7.4W

机构

暂无机构信息
引用论文

引用论文

err分享
err收藏
A solution-adaptive upwind scheme for ideal magnetohydrodynamics
err1999-09-01
err1.3K
PREAI
errPowell, KG; Roe, PL; Linde, TJ; Gombosi, TI; De Zeeuw, DL
err分享
err收藏
Neuromodulation in epilepsy: state-of-the-art approved therapies癫痫的神经调节: 最先进的批准疗法
err2021-12-01
err0
PREAI
errPhilippe Ryvlin; Sylvain Rheims; Lawrence J Hirsch; Arseny Sokolov; Lara Jehi
err分享
err收藏
学者 查看更多内容