Euler Equations Of Three-dimensional Adaptive Cartesian Grid-based Applied Research On The Complex Shape

In this thesis, a method of adaptive Cartesian grid based on the omni-tree and finite volume method of the Euler equations is presented. By applying such approaches, the grid is generated. Firstly, the initial mesh is generated after the size of the flowfield is determined. Secondly, the initial mesh is divided according to geometric and the body curvature. Thirdly, the smoothness of mesh should be checked and the cell that is not smooth enough has to be divided. And then, the boundary information is calculated and the solid mesh is deleted. Finally, the mini volume cell of the boundary mesh should be found and merged with its large neighbor.In this thesis, the omni-tree (also 2~N tree) data structure is applied to divide the mesh. Compared with octree data structure, the omni-tree data structure could reduce the total number of the mesh and get better mesh quality. The cell-centered finite volume spatial discretization, four-stage Runge-Kutta time-stepping scheme, some standard convergence acceleration techniques such as local time stepping and enthalpy damping, and solution-adaptive refinement of mesh are also applied to resolve high-gradient regions of the flow.Extentive calculations are carried out for pratical flow fields of M6 wing and DLR-F4 wing-body configuration. From the comparison between the computation result and the experiment data, it is showed that the agreement is good while CPU time is reduced.