Parallel Adaptive Methods for Modeling Nonequilibrium Fluids and Plasma's

9318181 Gombosi这项高度跨学科的工作是在并行超级计算机上使用自适应精细网格开发非平衡流体和等离子体建模工具。并行自适应网格方法的发展，包括负载平衡和高效利用高性能微处理器的有效战略，将对解决流体和等离子体物理所有领域的问题产生重大影响。具体的例子包括地球空间环境建模、高空空气动力学、大气科学、一般非线性系统和先进推进系统。该模型开发的第一个应用将是开发一个全球磁层模型，该模型将提供给磁层物理界。该模型将使用玻尔兹曼方程的高阶矩，并在3D自适应改进的非结构化网格上提供解决方案，这将允许多尺度分析，包括亚网格大小的动力学效应。＊＊＊

9318181 Gombosi This highly interdisciplinary effort is to develop tools for modeling nonequilibrium fluids and plasmas using adaptively refined grids on parallel supercomputers. The development of parallel adaptive-grid methodology, including effective strategies for load balancing and efficient utilization of high-performance microprocessors, will have a major impact on the solution of problems in all areas of fluid and plasma physics. Specific examples include geospace environment modeling, high-altitude aerodynamics, atmospheric science, nonlinear systems generally and advanced propulsion systems. The first application of this model development will be the development of a global magnetospheric model which will be made available to the magnetospheric physics community. This model will use high-order moments of the Boltzman equation with solutions on a 3D adaptively-refined, unstructured mesh which will allow multiple scale analyses, including sub-grid- size kinetic effects. ***