Research On Topological Changes In Feature Modeling

Semantic feature modeling is one of the most important technologies in the integrated system of CAD or CAM, which is the key problem in product model design. At the same time, the parametric design technique is applied to feature-based design system, where the feature can be adjusted, which improves the efficiency of designing process. So, the parametric design technique based on feature has become very important in most CAID systems.Current feature models do not explicitly represent the relation between the parameters and the topology of the model. Therefore, the topology of model may change when user fine-tuning a model unsuitably by modifying parameter of the model. Thus the semantic of feature changed undesirable, and the model departure the intention of designer.Aim at the problem described above and the deficiency of current solution, this paper presents a new method to detect topological changes of a model. The considered feature model consists of a system of geometric constraints, relating parameters to feature geometry, and a cellular model. The cellular model partitions Euclidean space into quasi-disjoint cells, determined by the intersections of the feature geometry. Our method creates a new system of geometric constraints to relate the parameters of the model to topological entities in the cellular model by carriers. And then decompose the geometric constraints system into different clusters according to relative freedom degree, determine entities that dependent on variant parameter. To every entity that dependents on variant parameter, this paper presents a method to determine it's degenerate constraint system by de-compositing the geometric constraints and compare equivalent constraints, then use a normalized expression based method to translate the geometric constraints to an algebraic constraint problem. By solving the problem we can obtain a critical value. Repeating this process we can obtain all critical values and determine the parameter range. Thus we can determine weather the parameter input by user will change the topology of the model before the modify excuted, reduce modifying times and improve editing efficiency.