Collaborative Research: Towards Intelligent Scan Sequence Generation to Reduce Local Overheating, Distortion, and Residual Stress in LPBF Additive Manufacturing

Laser powder bed fusion is increasingly being used to produce metallic parts in a variety of high-value industries, like aerospace, biomedical, and automotive. However, parts manufactured are prone to shape distortion and excessive heat or stress build up due to uneven temperatures across the part during the printing process, leading to cracks or other defects. Prior research has shown that scan sequence (i.e., the order in which geometric features on the part are scanned by the laser) can help in homogenizing temperature distribution across a part, thus reducing distortion, overheating and excessive stress. However, scan sequence is currently determined based on trial-and-error or heuristics, leading to inconsistent and suboptimal results. This project supports a scientific investigation into an approach for optimally determining scan sequence using models of the printing process. The knowledge created through this investigation will enable 3D printing of complex metallic parts with fewer failed or defective prints, thus improving the economic viability of laser powder bed fusion. The research will enrich an outreach program to excite middle school students in Detroit and inspire them to pursue careers in STEM fields. The objective of the project is to mathematically, numerically, and experimentally uncover the relationships between optimal scan sequences, temperature distribution, distortion, and residual stress in laser powder bed fusion using physics-based and data-driven thermal or thermomechanical models. The impacts of optimal scan sequences on microstructure and other part quality metrics will also be investigated. This objective will be achieved by: (1) incorporating advanced thermal effects into the determination of optimal scan sequences using data-driven models; (2) numerically investigating when optimal scan sequences generated using only thermal models do not adequately reduce distortion or residual stress; and (3) introducing thermomechanical effects into the determination of optimal scan sequences in cases where thermal models alone are deficient. The methods will be validated experimentally. Translation of knowledge from this project to application may accelerate the adoption of additive manufacturing in broader industries. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.