Boss recognition algorithm for application to finite element analysis

In finite element analysis (FEA), computer-aided design (CAD) models must be converted into solid meshes so that the solver can perform the desired analysis and simulation. Hexahedral and prism meshes are better than tetrahedral meshes, but are inherently more complex and difficult to generate. The purpose of this study was to propose an approach based on feature recognition for generating better quality solid meshes for FEA applications. Particularly, this study focused on the development of a boss recognition algorithm, the output of boss data for meshing, and the development of a process for automatic boss meshing. The proposed boss recognition method contains three parts: the preliminary functions, the data base, and the boss recognition. The first two parts provide a framework that can be used for other types of feature recognition. The core of the boss recognition is the tube recognition, which involves five main steps of validation. The output data of a boss includes the rib, tube, and hole data, which record not only feature data, but also the meshing data that can be used for automatic meshing. The meshing of a boss is divided into rib meshing and tube meshing. Both hexahedral meshes and prism meshes may be used depending on the regularity and orthogonality of the shape. This paper provides a detailed description of the proposed algorithm and presents several examples to illustrate its feasibility.