Date of Award
Industrial and Manufacturing Systems Engineering
CC BY-NC-ND 4.0
Currently, parts designed via CAD are defined and stored as geometric and topologic solids. The descriptions of these parts exist in terms of low-level information such as faces, edges and vertices or primitive volumes that are unrelated to manufacturing. Consequently, CAM is not able to make full use of the CAD-generated part description. To link between CAD and CAM, automatic interpretation of part description is very important. It involves a re-interpretation of the part description to extract manufacturing-specific semantic knowledge about the part. This knowledge can then enable the process planning and other CAM operations to proceed without human intervention. An algorithm is developed in this thesis to extract semantic knowledge useful to manufacturing in the form of part features from CAD database. It is capable of recognizing features from a polyhedral part given all its surfaces are perpendicular to each other. The algorithm consists of iteratively creating AAG (Attributed Adjacency Graph), recognizing features, and removing features. In each iteration, an individual feature is identified. The identified feature is then removed, creating a simpler solid model. Finally, a block is formed. In this thesis, the neural network method is used to recognize intersecting features. The output of the algorithm can be used as part feature input for CAM systems.Dept. of Industrial and Manufacturing Systems Engineering. Paper copy at Leddy Library: Theses & Major Papers - Basement, West Bldg. / Call Number: Thesis1993 .L585. Source: Masters Abstracts International, Volume: 32-02, page: 0697. Adviser: N. Singh. Thesis (M.A.Sc.)--University of Windsor (Canada), 1993.
Li, Yuan., "A feature recognition algorithm for polyhedral parts." (1993). Electronic Theses and Dissertations. 604.