ONE STEP BEYOND: FROM SIMULATION TO OPTIMIZATION

摘要

Computer simulation of the casting process is being used increasingly in commercial investment foundries to improve quality and decrease cost and lead time. One drawback of this technology is that it depends on the system user to analyze the results and make changes that will improve the process; trial-and-error on the shop floor is replaced with trial-and-error on the computer. This paper describes case studies moving one step beyond solidification modeling to multi-variable optimization. Optimization technology enables the computer to search for an optimum solution without human intervention; this technology has previously been used successfully to redesign automotive components for reduced weight and improved stress resistance and is now being successfully applied to the production of metal castings, including those produced by the investment casting process. Application of multi-variable optimization to solidification modeling requires an initial process design; selection of design variables and ranges within which they may vary; selection of constraints; and identification of an objective function which can be used to determine when an optimum design has been reached. The simulation program is then controlled by the optimization module, with design changes progressively specified by the optimization algorithm as it searches a response surface. The end result should be an optimum process design for producing a given casting, produced entirely by the computer software. This paper presents a series of case studies of using simulation and optimization to cure foundry problems, rig new jobs and optimize the casting process. Details of the optimization technique and application are also given.