Data-Driven Process Analysis for Iron Foundries with Automatic Sand Molding Process

Abstract This paper proposes a methodological framework to develop data-driven process control using pure industrial production data from cast iron foundry, despite the limitation of complete casting traceability. The aim is help sand foundries produce good castings. A reference which produces mainly automotive and oven parts with automatic molding pouring machines, was selected. Past data, where only castings were produced, extracted database determine parameter limits (upper lower limits) aid statistical approach. To identify critical parameters associated defects, zero high scrap batches systematically compared. method clearly identified unstable without exact synchronization between inline part quality data. Molding moisture, temperature compactability, liquidus melt, phosphorus content, carbon equivalent found be stabilized. Finally, regression model for predicting controlling moisture melt created. determined boundaries models helpful foundry in assisting ongoing correction inputs achieve target .