Indexing and Retrieval in Metal Stamping Die Design Using Case-based Reasoning

Abstract

This paper presents a case-based reasoning (CBR) methodology for metal stamping die design, that in particular addresses the indexing and retrieval of die design cases. A feature relation graph representation of stamped metal parts are used to create a high level of geometric abstraction, which is used to index design cases quickly and accurately. Though the potential search space for case retrieval is huge, by employing a novel dual-step similarity analysis between a new stamped part and existing parts in the case library, the proposed retrieval strategy can narrow down the search space efficiently and retrieve the most similar case in a reasonable period of time. An illustrative example is included to demonstrate the operation of the proposed approach and show its effectiveness in speeding up stamping die design.