国家数值风洞(NNW)工程中的黏性自适应笛卡尔网格方法研究进展

相比传统的结构和非结构网格，笛卡尔网格具有自动化和高质量兼顾的优势，是未来网格技术发展的重要方向。面向笛卡尔网格生成软件开发需求，依托国家数值风洞（NNW）工程，介绍了黏性自适应笛卡尔网格技术相关课题的研究工作进展，重点从网格生成技术、自适应方法、黏性物面边界处理方法等方面展开。网格生成技术方面，基于全线程树数据结构的思想，开展了叉树数据结构的优化；针对物面离散单元信息的快速检索，构建了稳定、平衡、高效的叉树式物面网格数据结构；对于网格单元类型的判断，发展了染色方法和包围盒方法以提高效率；针对大规模网格量，发展了高性能并行计算技术。自适应技术方面，发展了基于几何特征和流场解特征的自适应方法，建立了流场结构捕捉判据，并开展了三维复杂构型的应用；为了降低全局网格规模，发展了各向异性自适应方法，构建了各向异性特征识别的综合判据；发展了边界层法向射线加密技术，以减少边界层内的网格数目。物面边界处理方面，发展了重叠笛卡尔网格方法，构建了重叠网格尺度匹配技术和数据交互方法；开展了壁面函数方法在笛卡尔网格框架下的适应性应用，以放宽壁面网格尺度限制、降低边界层网格规模，并通过典型算例验证了方法的可靠性。

Advances in viscous adaptive Cartesian grid methodology of NNW Project

Compared with traditional structured and unstructured grids, the Cartesian grid has the advantages of automation and high quality, and therefore is an important development direction of grid technology in the future. Relying on the National Numerical Windtunnel (NNW) Project basic research system, this paper studies viscous adaptive Cartesian grid methodology, focusing on the grid generation technology, adaptation method, and viscous wall boundary treatment method, for the development of Cartesian mesh generation software. For the grid generation technology, optimization of the quadtree or octree data structure is conducted, starting from optimization of the Cartesian grid data structure and based on the idea of the fully thread tree data structure; a more stable K-dimensional tree method compared to the traditional one is constructed aiming at the information search of facets. For the determination of grid types, the painting method and axis aligned bounding box method are developed to improve the efficiency, and a high-performance parallel computing technology based on message passing interface is developed for large-scale grids. In the aspect of adaptation technology, an adaptation method based on geometric features and flow field solution features is developed, the flow field structure capture criteria established, and the application of three-dimensional complex configuration carried out; to reduce the scale of global mesh, the anisotropic adaptive method is developed, and the comprehensive criteria for anisotropic feature recognition constructed; the normal ray refinement technology is developed to reduce the number of grids in the boundary layer. In the aspect of wall boundary processing, the overset Cartesian grid method is developed, and the overset grid scale matching technology and data interaction method constructed. Meanwhile, the adaptive application of the wall function method in the Cartesian grid framework is carried out to relax the size limitation of the wall grid and reduce the number of boundary layer grids, and finally the reliability of the method is verified by typical examples.