Optimized Turbine Blade Cooling and Defect Mitigation for Additive Manufacturing

Cornerstone Research Group proposes to develop a computational framework for optimization of turbine engine blades with cooling flow paths enabled by additive manufacturing. This framework will be driven by mechanical, thermal, and fluid responses produced using coupled finite element models as well as a machine learning based defect prediction model. The defect prediction model will be trained using features extracted from the blade geometry, the AM process parameters (orientation, scan variables, powder properties), and projected thermal history from solidification modeling to incorporate cooling rate. Phase I will established feasibility of the defect prediction algorithm using a limited amount of NDT data from experimental fabrication efforts as well as the manufacturing capability to produce the type of geometries anticipated during follow on phases with suitable properties. Phase II will enhance the defection prediction and optimization methodology to guide turbine blade design for subsequent additive production and experimental cooling performance evaluation.