Adaptive Unstructured Surface Mesh Generation Method for Arbitrary Geometry with Proximity Features

Automatic simulation is an urgent challenge to be solved. For complex geometries, 73 percent of the total available CPU time is spent on mesh generation, which is still a bottleneck in the preprocessing module. Therefore, an unstructured mesh generation algorithm for arbitrary geometry with proximity features is presented. The refinement field is constructed according to mesh size, boundary curvature, surface curvatures and other geometry features. The geometry fitting techniques and the nearest distance methods are adopted to approximate the solid boundary of the mesh nodes. In order to effectively construct the core jagged mesh, a fast intersection algorithm is established to solve the intersection of mesh topology elements with the solid model boundary, which improves the efficiency of intersection calculation. For complex entities with proximity features, such as the models used in analysis of fatigue crack growth, a new type of proximity feature template is proposed, which successfully solves the problem of entity boundary fitting of the traditional out-side-in mesh method. Numerical examples show that the proposed method can perform quadrilateral-based meshing for solid models with proximity features, and the generated mesh has the characteristics of high cell quality and accurate expression of key features in the region. And the research findings have been seamlessly integrated into the domestically-produced and highly-autonomous 5aCAE software.