Autonomous Optimization in Casting Process Simulation

Computer processing speed has changed dramatically in the last 10 years. Traditional casting process modeling is based on "what-if" scenarios, requiring the user to make a decision, implement changes and start a simulation. To gain the most advantage out of computer advances, casting process modeling is now combined with genetic autonomous optimization tools, which produce a simulation based on a range of parameters rather than specific design points. Besides the optimization of designs, process parameters, as well as mechanical properties can be optimized too. The production of high quality castings depends on many factors. As state-of-the-art simulation tools consider many of these factors, multi-objective autonomous optimization opens a whole new level of accuracy. Input parameters and thermophysical properties can be optimized to characterize the specific melt of a foundry or production line. This paper will show the background of genetic optimization tools and examples of optimized castings utilizing this technology.