An efficient search algorithm for calculating minimum wall distance of unstructured mesh

The wall distance is a parameter that must be specified to solve the Reynolds-averaged NS equations with turbulence models. The efficiency for computing the wall distance of unstructured mesh is low due to the arbitrary topological relationship between vertices and cells. To enhance the computational efficiency, we propose a layer-by-layer search algorithm that calculates only the minimum wall distance from any grid cell in a flow field to the target wall. The core of the algorithm consists of: (1) supposing that the scattered unstructured grids have the layer by layer arrangement, we start from a solid surface unit and use the relations among the adjacent grids to diffusely search for the spatial grid cell which is the nearest to the surface unit layer by layer; (2) the search range of each spatial grid cell is effectively restricted to the precise wall unit and several other wall units adjacent to them, thereby enhancing the computational efficiency greatly. We select three meshes of complex solid wall surface as calculation examples to verify the effectiveness of our search algorithm by comparing its computation time and computation results with those of the enumeration method. The computation results, given in Figs. 5, 8 and 11 and Tables 1 through 6, and their analysis show preliminarily that: (1) the computation accuracy of our search algorithm has no significant difference from that of the enumeration method; (2) our search algorithm can raise the computational efficiency of small and medium-sized mesh by two orders of magnitude, while the computational efficiency of the larger mesh that has more surface units can be more than two orders of magnitude higher.