Mechanistic data-driven models of process, microstructure and mechanical properties in metal additive manufacturing

Abstract

Metal additive manufacturing (AM) offers the potential to change the manufacturing engineering landscape by producing parts with tailored performance and intricate geometries with minimal loss of material. However, limited material selection and non-optimal process parameters might lead to defects, such as porosity, roughness, and cracking, which will affect mechanical properties. The ability to accurately predict the extremely variable temperature field in detail, and relate it quantitatively to structure and properties, is a key step in predicting part performance and optimizing process design. This talk will cover the study of melt pool dynamics and microstructure evolution using mechanistic models with machine learning. Specifically, I will present a systematic modeling and experimental study on the relationship between the thermal characteristics of AM and the resultant microstructure and properties. I will also discuss how nanoparticles influence melt pool stabilities, grain sizes, and mechanical properties in metal additive manufacturing. The insights gained through the study can help better understand the mechanism of melt pool dynamics, material solidification, defect and bubble formation, and the relationship between process-structure-properties in AM.