太阳物理研究组知识库

流体动力学(Hydrodynamics,HD)与磁流体动力学(Magnetohydrodynamics,MHD)高效数值计算方法被广泛应用于天体物理领域，用来研究和解释天体系统中有关磁场和流场的演化以及能量转化过程，如磁重联、恒星演化及形成、吸积盘等问题。通过对三个开源程序ZEUS,ATHENA和NIRVANA的探讨，详细阐述经典网格方法的起源、发展及适用于求解MHD方程的数值方法。同时介绍了基于此类程序已展开的科研工作和获得的成果，旨在通过对一些具体实例的计算的介绍，分析相应算法的优缺点，探索有关算法和程序在天体物理领域中的应用。希望提供给读者关于MHD数值优化方面一些参考与借鉴。

Hydrodynamics (HD) and Magnetohydrodynamics (MHD) simulation technics have been wildly developed, in the last decade, in order to describe the plasma behavior in a magnetic field related to some kinds of astrophysical system, such as magnetic reconnection in solar eruptions or the interstellar medium, star formation, and accretion flows. As we know, MHD approximation considers the plasmas as a collisionless neutral fluid and focuses on the macro description of dynamic phenomenons. Basically, we have two ways to describe fluids. One of them is Lagrangian method, while the other is Eulerian method. And MHD simulations usually belong to Eulerian class. In this work, three Eulerian grid-based codes:ZEUS, ATHENA and NIRVANA (2D or 3D) are introduced to solve MHD problems. We discuss the category of grid generation (Structured mesh, Unstructured mesh and Cartesian mesh) as well. Indeed, the main challenge for computational MHD in astrophysics is the huge discrepancy in space scaling and complex physical process coupling (shock-turbulence).