Towards multiprocessor feature recognition

The availability of low-cost computational power is enabling the development of increasingly sophisticated CAD software. Automation of design and manufacturing activities poses many difficult computational problems-significant among them is how to develop interactive systems that enable designers to explore and experiment with alternative ideas. As more downstream manufacturing activities are considered during the design phase, computational costs become problematic. Creating working software-based solutions requires a sophisticated allocation of computational resources in order to perform realistic design analyses and generate feedback. This paper presents our initial efforts to employ multi-processor algorithms to recognize machining features from solid models of parts with large numbers of features and many geometric and topological entities. Our goal is to outline how improvements in computation time can be obtained by migrating existing software tools to multiprocessor architec-tures. An implementation of our approach is discussed. The availability of low-cost computational power is enabling the development of increasingly sophisticated CAD software. Software tools designed to reduce time-consuming build-test-redesign iterations are becoming essential for increasing engineering quality and productivity. Examples include tools for finite element analysis, mechanism analysis, simulation, and rapid prototyping. Such tools have become crucial components for research in collaborative engineering and engineering design. Automation of the design process and construction of such tools, however, pose many difficult computational problems. To realize the advantages of collaborative engineering, more downstream engineering activities are considered during the design phase. As design is an interactive process, developing techniques to manage computational costs better is critical in systems that enable designers to explore and experiment with alternative ideas during the design stage. Achieving reasonable levels of interactivity between design and-37 downstream activities (such as analysis, process planning , and simulation) requires an increasingly sophisticated allocation of computational resources in order to perform design analyses and generate feedback. It is becoming increasingly evident that one necessary component of an automated design analysis tool is a subsystem for recognizing manufacturing features directly from a CAD or solid model. This problem has been the focus of extensive research aver the last decade. Feature recognition is used for a variety of applications, including the generation of process plansss,t4st5, translation between design and manufacturing features, and generation of redesign suggestions5,6. What has also become evident is that feature recognition, for realistic classes of parts with multiple and interacting feature interpretations, requires extensive geometric reasoning and is computationally expensive. Hence, generating the features from a part is …