Implicit Massively Parallel Navier-Stokes Solver for Hybrid Structured/Unstructured Grids

最近出现的高性能大规模并行处理（MPP）计算机从理论上减少了使用计算流体动力学（CFD）设计和分析涉及流体流动的系统所需的计算机时间。然而，在实践中，今天在MPP中使用的算法是为顺序计算机开发的，因此不能利用大规模并行处理的全部能力。此外，MPP计算机没有简化结构化网格生成过程，复杂的配置可能需要数周的时间。Amtec提议开发一种新的计算流体动力学算法，以利用MPP的力量。研究将集中在MPP计算机上使用混合结构化/非结构化有限体积网格求解Navier-Stokes方程的技术。结构网格将用于解析壁面附近的边界层，而非结构网格将用于离散流场的其余部分。在第一阶段，将使用一种新的区域分解技术改进和并行化最近开发的二维混合网格隐式程序。新的求解程序将通过使用混合网格简化网格生成过程，并通过高效地使用MPP计算机减少所需的计算机时间。

The recent emergence of high performance massively parallel processing (MPP) computers has theoretically reduced the computer time required for the design and analysis of systems involving fluid flow using Computational Fluid Dynamics (CFD). However, in practice, algorithms used in MPP today were developed for sequential computers and therefore do not harness the full power of massively parallel processing. Also, MPP computers have not simplified the structured grid generation process, which may take weeks for complex configurations. Amtec proposes to develop a new algorithm for computational fluid dynamics to harness the power of MPP. The research will focus on techniques for solving the Navier-Stokes equations on MPP computers using hybrid structured/unstructured finite-volume meshes. The structured grids will be used to resolve boundary layers near walls and unstructured grids will be used to discretize the remainder of the flow field. During phase I, a recently developed implicit procedure for 2D hybrid grids will be improved and parallelized using a novel domain decomposition technique. The new solution procedure will simplify the grid generation process by using hybrid grids, and reduce the computer time required by efficiently using MPP computers.