Designing the Least Expensive Charge Mix Using Data Analytics and Optimization for Gray Cast Iron (Grade FG 220)

In a foundry, optimizing the charge mix is critical to achieving consistent quality, cost-efficiency, and desired qualities in the final metal or alloy product. This paper describes a data analytics-driven strategy for optimizing the charge mix by lowering the cost of the scrap used to prepare the molten metal while maintaining the required chemical composition, tensile strength, and hardness required by the foundry for manufacturing gray cast iron products (Grade FG 220). The linear programming approach is used for this purpose where all the constraints are strictly met. Three categories of constraints are used for this purpose, i.e., composition constraint, foundry constraint, and material grade constraint. In the linear programming approach, the feasible region is considered as an ellipsoidal region and the developed convex optimization problem is iteratively solved. The result showed potential cost savings could be obtained, accompanied by the needed alloy chemical composition and quality.