Rule-based automatic part feature extraction and recognition from CAD data

Abstract

Automatic extraction and recognition of part features directly from a CAD (Computer-Aided Design) database can form a vital link between CAD and CAPP (Computer-Aided Process Planning). This paper reports the effort of developing such a system for rotational part features using the expert system approachs. Part geometric and topological data in IGES (Initial Graphics Exchange Specifications) format are read by a feature extraction data-compactor (a pre-processor of the system). Then a geomtry-to-feature translator captures manufacturing features in its decision logic expressed as production rules written in LISP. Successful functioning of the system is demonstrated by two illustrative examples. Future research to include a generic feature representation scheme for direct interface with CAPP is given by a framework for this integration.