| description abstract | This paper presents a method for automated manufacturing process selection during conceptual design. It is helpful to know which manufacturing processes can produce a design at an early stage, when the overall design can be changed for less cost. Early during new product development, geometric dimensions and tolerances may not yet be specified, but a general three-dimensional (3D) model is often under development. In this work, algorithms are presented to interrogate 3D models to calculate machining-based manufacturability metrics. These algorithms are used on a dataset of 86 computer-aided design (CAD) models classified as machined or cast-then-machined. The metrics, such as visibility, reachability, and setup orientations, seek to characterize a part's manufacturability using machining domain knowledge. These metrics serve as inputs to machine learning models, which are used to classify parts by manufacturing process with 86% accuracy. Some of the incorrectly classified parts were instances that had robust designs capable of being manufactured using machining or casting. The results of the machine learning models indicate that the machining metrics can be used to provide process selection feedback during conceptual design. | |
