Overview of HiFi -- implicit spectral element code framework for multi-fluid plasma applications
Overview of HiFi -- implicit spectral element code framework for multi-fluid plasma applications
复制标题
DOI：
--
复制
发表时间：
2016-08
期刊：
影响因子：
--
通讯作者：
V. Lukin;A. Glasser;W. Lowrie;E. Meier
中科院分区：
文献类型：
--
作者：
V. Lukin;A. Glasser;W. Lowrie;E. Meier
概述的算法和采样的等离子体应用的隐式，自适应高阶有限（谱）元素建模框架，HiFi，提出。HiFi代码的区分能力包括具有灵活几何形状的自适应谱元素空间表示，高度并行化的隐式时间推进，以及可以在其框架中实现的偏微分方程和边界条件的一般通量源形式。早期的算法开发和二维版本的代码的广泛验证研究，称为SEL，先前已经描述[A.H. Glasser & X.Z. Tang，Comp. Phys. Comm.，164（2004）; V.S. Lukin博士Princeton University（2008）]。在这里，大量的算法改进和扩展的HiFi框架的二维和三维应用的例子一起提出。其中包括一个笛卡尔二维不可压缩磁流体动力学模拟低耗散磁重联在一个大系统中，一个二维轴对称模拟自相似压缩的磁等离子体约束配置，和一个三维霍尔磁流体动力学模拟spheromak倾斜和松弛。一些计划的努力，以进一步改善和扩大的HiFi建模框架的能力进行了讨论。
An overview of the algorithm and a sampling of plasma applications of the implicit, adaptive high order finite (spectral) element modeling framework, HiFi, is presented. The distinguishing capabilities of the HiFi code include adaptive spectral element spatial representation with flexible geometry, highly parallelizable implicit time advance, and general flux-source form of the partial differential equations and boundary conditions that can be implemented in its framework. Early algorithm development and extensive verification studies of the two-dimensional version of the code, known as SEL, have been previously described [A.H. Glasser & X.Z. Tang, Comp. Phys. Comm., 164 (2004); V.S. Lukin, Ph.D. thesis, Princeton University (2008)]. Here, substantial algorithmic improvements and extensions are presented together with examples of two- and three- dimensional applications of the HiFi framework. These include a Cartesian two-dimensional incompressible magnetohydrodynamic simulation of low dissipation magnetic reconnection in a large system, a two-dimensional axisymmetric simulation of self-similar compression of a magnetic plasma confinement configuration, and a three-dimensional Hall MHD simulation of spheromak tilting and relaxation. Some planned efforts to further improve and expand the capabilities of the HiFi modeling framework are discussed.