Feature extraction from solid model for manufacturability assessment
journal contributionposted on 01.01.1989, 00:00 authored by Friedrich B. Prinz, Young Choi, Jyh-Huei Chern, Carnegie Mellon University.Engineering Design Research Center.
Abstract: "Assessment of manufacturability of a design with minimum human intervention is in increasing demand. It is, however, very difficult to investigate manufacturability problems with analytical tools in most of the manufacturing processes. In a real manufacturing situation, manufacturability problems are considered only by designers based on the standard design practice and their past experience. In general, manufacturing knowledge is hardly reflected at initial design stage. An expert system equipped with knowledge of manufacturability can be usefully applied to solve these problems. An expert system that can evaluate designs for their manufacturability must have knowledge of manufacturing requirements as they relate to the design of the part to be produced. However, this type of knowledge is generally in the form of relationships or rules that refer to geometric features of designs.Therefore, an appropriate form of input data for the expert system should be generated from the part design data base. This article describes a method to extract feature information from the data base of a 3D surface boundary representation for the injection molded part design. The feature information is then passed into a knowledge-based system to perform manufacturability assessment."