Improving the manufacturing yield of investment cast turbine blades through robust design

Abstract

The manufacturing of turbine blades is often outsourced to investment casting foundries by aerospace companies that design and build jet engines. Aerospace companies have found that casting defects are an important cost driver in the price that they pay the foundries for the turbine blades. Defect types include porosity, stress, grain, fill, and mold-related defects. In order to address the defect problem, aerospace companies have adopted a design for manufacture approach to drive the cost of the turbine blades down. The principal research objective of this thesis was to discover how the critical part features on the turbine blade drive the number of manufacturing defects seen in the casting process. This problem was addressed by first selecting and evaluating a casting simulation software package. Secondly, a robust design of experiments was performed by using the simulation software. In the experiment, the dimensions of the critical part features were varied in order to quantify how the critical part features relate to manufacturing defects.