Specific objective functions and algorithms are presented, which by means of dedicated finite element simulation software, calculate optimized control variables for an industrial Bridgman casting furnace. A gradient method and an evolution strategy have been integrated into an efficient optimization tool in order to minimize an objective function which characterizes the quality of directional solidified (DS) or single crystal (SC) turbine blades. The goal is to minimize the manufacturing costs for a desired quality. For a dummy turbine blade geometry, the macrograin structures produced by the optimized withdrawal velocity profile and an equivalent constant velocity, which had the same total process time, are compared. This comparison shows that the developed optimization techniques are applicable to the casting of DS/SC turbine blades.

No abstract

No abstract