Machined Feature Recognition Based on Feature Loop and Convex Polyhedron

摘要

This study deals with a machined feature recognition algorithm. This algorithm is used for recognizing features such as a hole or a slot in the surface form from a boundary representation based solid model of a machined part. The algorithm is composed of two main processes. One is the feature extraction process and the other is the feature recognition process. Through the first process, faces organizing the feature are obtained from a depression or a protrusion area in the machined part. These areas are surrounded by closed edge loops (hereafter called feature loops). Through the second process, a convex polyhedron (CVP) corresponding to a feature is generated from all organizing faces of the feature, and the other faces adjacent to the same feature loop. The generated CVP is used to recognize simple features of the part. Since the generation of CVP for a complex feature is not straightforward, a decomposition method is used to recognize the feature as a set of simple features. Finally, the effectiveness of the proposed recognition method is clarified through case studies.