Reconfigurable 3D CAD feature recognition supporting confluent n-dimensional topologies and geometric filters for prismatic and curved models

摘要: Feature Recognition (FR) in Computer-aided Design (CAD) models is central for and Manufacturing. FR a problem whose computational burden intractable (NP-hard), given that its underlying task the detection of graph isomorphism. Until now, compromises have been reached by only using FACE-based geometric information prismatic CAD to prune search domain. Responding such shortcomings, this manuscript presents an interactive method more aggressively prunes space with reconfigurable tests. Unlike previous approaches, our addresses curved EDGEs FACEs. This approach allows enforcing arbitrary confluent topologic filters, thus handling expanded scope. The test sequence itself (i.e., not linear or total-order sequence). existing methods are FACE-based, present one permits combinations topologies dimensions two (SHELL FACE), (LOOP EDGE), 0 (VERTEX). system has implemented industrial environment, icon graphs rule configuration. instancing industry based customization itis faster when compared topology-based feature recognition. Future work required improving robustness conditions, treating interacting nested features, graphic input interface.