Data-Driven Modeling of Directed Energy Deposition (DED) for Defect-Free Builds using Real-Time Monitoring Approaches

Additive Manufacturing (AM) will play a large role in the development for NASArsquo;s next generation of space flight systems. However, without a deeper understanding of various additive processes available, it is difficult to use the technologies reliably. In order to predict the results for dimensional accuracy, microstructure formation, and defect initiation points, FormAlloy and New Mexico State University proposes to develop a data-driven model of the directed energy deposition (DED) process. The models will be derived from data generated by in-situ monitoring with acoustic emissions (AE), thermal, and vision sensing modalities during deposition process and mechanical, metallurgical, and nonlinear ultrasound evaluations post-build.The sensing modalities will be installed within one of FormAlloyrsquo;s award-winning DED systems and data will be collected during the deposition of two materials of interest for NASA, Inconel 625 and GRCop-42. Several parameter sets will be utilized to generate samples for evaluation with a wide range of microstructures and mechanical properties. The approach presented here shall potentially set the framework other material systems within DED and other AM processes of interest, such as laser powder bed fusion (L-PBF). Once the models are validated, FormAlloy and NMSU would proceed with a Phase II to develop a software package to deliver accelerated development cycles and improved part quality for additive manufacturing processes.