Abstract
The linear programming (LP) model for optimizing the average efficiency of molten alloy is a comprehensive yet simple tool for scheduling a die casting process in real foundries. In this paper, the authors present a heuristic algorithm for a post-optimization scheme to further enhance the efficiency of a scheduling result obtained by LP. The proposed scheme is motivated by the unique property of casting operations that raw material or alloy is provided as a set of solid-state ingots with a discrete size, and scrap is cut off from a solidified casting and recycled with a constant rate. This scheme adjusts the number of ingots to be melted by calculating a priori the amount of castings per melting furnace and by recycling scrap left in the previous casting shifts. A case study using practical data from a real foundry demonstrates the applicability and confirms the enhanced efficiency of the proposed scheduling mechanism and adjustment scheme.