Extending Graph-Based Feature Recognition to Functional Features for Enhanced Process Planning

Feature recognition (FR) is a crucial step in bridging the gap between computer-aided design (CAD) and other computer-aided systems like computer-aided process planning (CAPP) or computer-aided manufacturing (CAM). This work introduces functional features like keyways and relief grooves to enable manufacturing process selection (MPS) for multi-process manufacturing chains. It also enables automatic association of required tolerances and surface finish specifications with relevant surfaces of the part, facilitating an early-stage comparison between manufacturing requirements and available manufacturing processes and equipment capabilities. A graph-based FR-approach is used, leveraging the Attributed Adjacency Graph (AAG). The methodology extracts geometrical and topological information from the B-Rep of a CAD component, which is then used to construct the AAG, where nodes correspond to faces and arcs correspond to edges. The AAG is divided into sub-graphs using graph-decomposition. The sub-graphs are classified based on defined geometric, topological, and other criteria. The proposed methodology has been validated on representative CAD models containing both general and functional features for machining processes, with a complex shaft as an application example. The results demonstrate the system’s robustness and potential for automating a MPS system using only the CAD-model of a component as an input.