流场自适应叉树网格数值模拟三维复杂超声速流场

建立了基于分区结构网格的三维贴体叉树形网格的数据结构,并阐述了在此基础上进行自适应分裂/合并判别方法.为节省网格量和保证流场结构捕捉质量,提出对自适应程度进行区域性控制,以及对流场结构进行"保护"性预加密的优化方式.通过应用该网格对三维复杂超声速流场算例的计算,证明该方法对网格加密控制方便,对流场结构分辨率高.

Numerical Investigation of Complex 3-D Supersonic Flows with a Self-Adaptive Tree Mesh

Based on an initial structural mesh, a data structure for a 3-D body-fitted tree mesh was built. Simplified algorithms were developed for anisotropic self-adaptive split and mergence. In order to reduce the number of mesh ceils we set restrictive self-adaptive levels for different zones divided in advance. A method of protective mesh split for structures in flow field was developed. Numerical simulations with different complex supersonic flow patterns were shown. It is proved that the anisotropic self-adaptive tree mesh is effective with less mesh cells and convenient for mesh refinement. High resolution for Now field is obtained.