Full-mold casting has the advantage of reducing the production costs of complex shapes, but also the drawback of generating residual defects. One effective countermeasure against residual defects is the design of efficient traps. Moreover, in recent years, numerical fluid dynamics (CFD) has been utilized for the design of trap shapes. However, trap design remains a challenge for complex product shapes with significant changes in combustion cross-sectional areas using conventional design methods.
In this study, to prevent residual defects occurring in complex product shapes with large variations in combustion cross-sectional area during full-mold casting, we aimed to clarify how such product shapes influence the molten metal flow, the generation of residuals, and their behavior. Based on these findings, we developed an automated residual trap design method that considers the effects of product shape on molten metal flow. Specifically, we developed a velocity model that considers the impact of product shape, head pressure, and temperature variations in the molten metal on the molten metal flow, allowing the identification of defect-prone areas. We also utilized artificial intelligence to automate the design of traps for effectively capturing residuals.
As a result of the analysis, the predicted positions of residuals using the developed model matched the actual residual positions in cast products, confirming the validity of this approach. Additionally, experimental verification results demonstrated that the designed traps effectively prevent residual defects. This method contributes to improving the accuracy and reliability of lost foam casting (LFC), and provides a promising approach to enhancing casting outcomes and reducing defects in complex castings.
