Integrated Tool Development for Part Scale Analysis and Process Qualification of Advanced Manufactured Metal Components via Laser Powder Bed Fusion Additive Manufacturing

Abstract

The fast and advancing pace of metal additive manufacturing (AM) technology has peaked beyond its ability of supporting rapid prototyping, and in more recent years, has gained popularity in supplementing or even replacing traditional fabrication methods. Among the various metal AM methods and processes, the Laser Powder Bed Fusion (LPBF) method is popular for its ability to sufficiently produce dense components and meets the needs of fabricating low quantity parts. However, an overall lack of technical understanding or control of the process makes it difficult to ensure its repeatability and meeting qualification requirements challenging to achieve. In this dissertation, we aim to address this space through investigations of novel in-situ monitoring of thermal history, integrated modeling frameworks, and predictive thermomechanical models to better quantify and qualify the LPBF process. First, an in-situ monitoring methodology is introduced to quantify the layer-wise manufacturing process. Next, a modeling framework is presented to support the qualification of the LPBF process. Lastly, evolution of residual stresses in LPBF builds were investigated through thermomechanical finite element analysis (FEA). The simulations predicted thermal histories within reasonable errors as a result of local thermal variability in the machine, while the mechanical FEA results were able to predict local failures due to high stresses. Furthermore, the settings used in the FEA were able to be transferred across a different build, increasing reliability of the simulation model. In this work, part scale methods can be used to facilitate pre-fabrication decisions, enabling early detection of potential build risks during the design stage to improve LPBF manufacturing reliability. To conclude the dissertation, key outcomes and suggestions for future directions are provided. Overall, this dissertation establishes a comprehensive modeling and monitoring framework that can be used in a digital twin framework that advances the understanding of LPBF process behavior, supports qualification efforts, and offers predictive capabilities for improving part and process reliability in metal AM.