Developing a rule engine for automated feature recognition from CAD Models

The detailed development phase in modern engineering project lifecycles is characterised by the iterative use of a number of engineering software tools. Inefficient integration between these tools often results in a high volume of manual data manipulation, for example the derivation of analysis and manufacturing models from detailed design models. The automatic recognition of engineering features from product geometry has potential to improve integration efficiency and reduce time and costs of downstream processes of Computer Aided Design (CAD) system based design. This paper introduces a methodology for developing and executing rules to identify engineering features from geometric data. The methodology has been implemented in an Automated Feature Recognition (AFR) system that identifies and extracts analysis features to feed stress analysis algorithms.