Machine understanding of manufacturing features

Automated engineering analysis (AEA) of the solid model of a product is an essential task for the implementation of concurrent engineering. AEA has the potential to analyse several design and manufacturing activities in order to reduce the product development time and the associated costs. This especially has been the goal of many American manufacturing companies facing stiff global competition. AEA involves the use of the CAD data base of the product model for several design and manufacturing activities such as process planning, NC path planning, GT coding, assembly analysis, tolerancing, automated jig and fixture design, FEM analysis, heat transfer analysis, etc. The power of AEA can be fully realized if the output from the CAD data is in a higher-level form, i.e., features. However, CAD systems typically store the product information in a lower-level form such as points, edges, faces, etc. This paper presents a rule/graph-based approach for generating higher-level product definition in terms of manufacturing features from the B-Rep data structure. Depression, protrusion and many cases of interacting features can be recognized by the proposed methodology for their subsequent use in manufacturing planning activity. The methodology efficiently searches for 2-D features on non-convex faces and integrates this information to identify 3-D features. Some example parts are used for illustrating the methodology.