论文部分内容阅读
径向基函数广泛应用于网格变形、气动外形优化设计、网格优化等领域。近年来,基于径向基函数的动网格技术得到了深入的研究和广泛的应用。本文结合计算流体力学和高性能计算的应用背景,从径向基函数对网格的变形质量和变形效率进行了总结和进一步研究:在网格变形方面,重点对比了不同基函数对同一网格运动变形能力和同一基函数对不同网格运动的适应能力;在网格变形效率方面,分析了算法在计算和存储的瓶颈所在,考虑了OpenMP和GPU这两种共享内存的加速方式,得到较好加速比。最后,分析了当网格规模增大时,动网格在分布式计算和存储模型(MPI)下的处理方法。
Radial basis functions are widely used in the field of grid deformation, aerodynamic shape optimization, grid optimization and so on. In recent years, the dynamic grid technique based on radial basis function has been deeply studied and widely used. In this paper, based on the application of computational fluid dynamics (CFD) and high-performance computing, the quality and deformation efficiency of the grid are summarized and further studied from the radial basis function. In terms of grid deformation, Motion deforming ability and adaptability of the same basis function to different grid motions. In terms of grid deformation efficiency, the bottleneck of the algorithm in computing and storage is analyzed, and the acceleration methods of two shared memories, OpenMP and GPU, are considered. Good speed ratio. Finally, the method to deal with moving grid under distributed computing and storage model (MPI) is analyzed when the size of grid increases.