Development of geometry-driven quantitative prediction for shrinkage porosity in T-junction of steel sand castings

Abstract

Shrinkage porosity poses a significant challenge in metal casting processes, impacting both productivity and energy efficiency, especially when dealing with components that are not accepted or reprocessed. Addressing this issue requires proactive measures, and predictive techniques play a crucial role in minimizing its occurrence. Among these methods, the Criterion Function stands out as a valuable empirical model extensively explored in the literature. By intricately linking solidification processes to the development of shrinkage porosity, the Criterion Function leverages key process parameters, including thermal gradient, molten metal velocity during solidification, and cooling rate, to offer predictive insights into the location and presence of porosity. However, a criterion function is needed that also considers the effect of geometric variations as well as the size of the defect (shrinkage porosity). In this study, a casting with three T-joints was taken as a benchmark shape to develop a geometry-based quantitative prediction model for plain carbon steel castings. Real experimental results were combined with solidification simulation results to produce reliable data, which were then used to extrapolate the results. The developed quantitative prediction model, which includes the effect of geometric changes, has been validated and proven effective in predicting shrinkage porosity.