Dormant deficiency: a novel concept to direct cause-effect CAD model analysis

Fully constrained features and sound associativity are prerequisites for robustness and alterability of parametric feature-based CAD models. However, errors in associativity are very difcult to detect with traditional static analysis approaches, due to efects that remain hidden until parameter changes and model re-creation take place. Currently, studies on associativity-related CAD model defciency have not advanced to the point of being a part of model analysis. In this paper, the novel concept of dormant defciency, together with a three-level classifcation, a graph-based knowledge network, a human readable visualization of cause and efect relationships, and a software tool are presented in a newly developed approach to dynamic CAD model analysis. Within this approach, dormant defciencies are triggered to facilitate a methodic knowledge-driven method of detecting errors in associativity. This is achieved through systematic analysis of defciency generating efects and their related symptoms, followed by systematic backtracking to their root causes. A selection of representative examples used for testing and evaluation of the approach is included within the empirical results from practice.