A coordinated constraint-based modeling and design advisory system for mechanical components and assemblies
Product modeling, in which the product information is managed throughout the design and manufacturing activities, is the key technology for integrating CAD/CAM in a concurrent engineering system. Researchers in this area are still investigating and trying to improve the product design during the design process. An intelligent CAD system should be able to assist a designer during the product design process by detecting design problems, providing alternative suggestions and correcting design violations. This research concentrates on developing a collaborative design advisory system based on a constraint-based product modeling environment.
We classify engineering constraints during the product design process as component design constraints, manufacturing constraints and assembly constraints. Engineering constraints are modeled and managed in close relation to assembly and component models. We develop an enhanced assembly modeling scheme based on spatial relationships and introduce mating bonds in the assembly for easy and systematic integration with the design advisory system. The representation and management of constraints are elaborated at the component level and assembly level using a graph-based approach.
We develop a model for a design advisory system which is embedded in the constraint modeling scheme. This system considers product design from three different aspects: component dimensioning, manufacturing and assembly, and each corresponding sub-model is elaborated. The interactions among these three sub-models is reasoned from the constraint graphs. The assembly design advisory system plays an important role from the viewpoint of the whole product design, so it is amplified in detail. The assembly design advisory system composes a conflict-based communication with the other advisory systems to negotiate the design modification of the product such that an optimal product design can be achieved. The product design optimization is formulated as a bipartite graph matching problem and resolved by integer programming techniques.
0548: Mechanical engineering