The internal representation of the solid modeller provides a description of parts which when used directly is useful for automation of the process planning function. So that the CAD model can be used toprovide the information required for manufacturing, techniques to improve machine understanding of the part as required for manufacturing are needed. This paper presents the development of the concept attributed adjacency graph (AAG) for the recognition of machined features from a 3D boundary representation of a solid. Current implementation of the feature recogniser is limited to polyhedral features such as pockets, slots, steps, bfind steps, blind slots, and polyhedral holes. Sample results that show the capabilities of the system are presented.
process planning, part description, attributed adjacency graph, feature recognition, polyhedral features
Computers are being used extensively to assist in the development of generative process planning systems. One such area is the development of a CAD interface for process planning 1 . The internal representation of the solid modeller provides a form of part description which when used directly is useful for automation of the process planning function. Part description in 3D CAD models is in a form of basic geometry and topology that is unsuitable for direct application to process planning. Process planning needs information in the form of features, which need to be extracted from this CAD model. Current generation of process planning systems require human input to interpret a part drawing or model and translate it into a set of machinable features suitable for process planning. The importance of feature recognition stems from the fact that each feature can be associated with knowledge about how the feature can be manufactured, and this information can be collectively used to generate a process plan. Hence feature extraction forms a major component of the CAD interface for process planning. Extraction of features can also be viewed as a fundamental aspect of the more general problem of machine understanding of designed parts. This understanding can also be extended to other applications besides manufacturing.
The problem of extracting geometric features involves recognising high level entities from the set of lower level entities in a geometric model. Depending on the application and the type of part representation scheme used, geometric entities such as faces, edges, vertices, primitive solids, etc. may be considered lower level entities. The definition of a