| dc.description.abstract | A thin-shell CAD model can typically be divided into inner and outer regions, on which various structural and functional features are resided. The recognition of various types of faces on a thin-shell CAD model is an essential step in feature recognition that can be used for automatic generation of better type of meshes. The essence of this paper is to propose a method for detaching face types, including inner, auxiliary and outer faces, on thin-shell parts. Firstly, a thin-shell part is split into two types of regions: observable and unobservable, based on the mold opening direction, called inner and outer faces, respectively. Secondly in auxiliary faces recognition, the inner faces are then divided into several types, including: transition, wall, next-layer wall, base, ridge and inner protrusion faces, where the first five types are termed auxiliary faces. Finally, in outer faces recognition, the outer faces are divided into the following face types: flange, shell, ground and outer protrusion faces. Also, the thin-shell CAD models are divided into two types: concave-shell and non-concave shell cases, for facilitating the development of outer-faces recognition algorithm. A detailed description of the algorithms for inner-and-outer faces splitting, auxiliary faces recognition, and outer faces recognition is provided in this thesis. Also, realistic thin-shell CAD models are tested to demonstrate the feasibility of the proposed method. With these types of faces recognized, appropriate meshing algorithms that can yield a better type of solid meshes could be developed for the meshing of thin-shell CAD models. | en_US |