Die-Casting Feature Recognition for Automated Parting Direction and Parting Line Determination

Abstract

Determining parting direction and parting line for die-cast parts is a nontrivial task that not only depends upon shape and topology of the part, but also on many process related factors. Normally, a die-casting expert decides parting direction and parting line, intuitively taking into account a large number of factors, and this process can be time consuming and cumbersome in many cases. This study addresses automated determination of parting direction and parting line for a die-cast part from part CAD model. The proposed methodology takes STEP file of the part as input for extracting die-casting features, which consists of protrusion or depression regions of the part. These features are classified into those with single, double, or multiple withdrawal directions. Geometric reasoning is used for feature recognition, which includes nested and interacting features. Global visibility instead of local visibility is used for planning withdrawal direction, which makes the decision arrived by present system closer to industrial practice. Parting line is determined based on selected candidate parting direction considering process constraints and priorities. The contribution of this paper is in terms of development of an automated parting direction and parting line determination system, which is more comprehensive and overcomes limitations of the previous work. Results of this system have been validated with those arrived at by experts from the die-casting industry.