Enhancing Functionalities of Metal 3D Additively Manufactured Materials Using Machine Learning

Additive manufacturing (AM) technology has attracted considerable attention because it allows for the one-piece production of metal products with complex topologies based on computer-aided design models. It is required to establish a methodology of “design for additive manufacturing (DfAM)” that efficiently designs or optimizes the complex geometries according to the high manufacturability of AM technology. Laser powder bed fusion (L-PBF), which is a representative AM process, repeats powder spreading and laser scanning on the powder bed. This process enables the rapid solidification and forms a refined non-equilibrium solidification microstructure, resulting in anomalous mechanical and functional properties. However, the optimization of process conditions is required to obtain dense products without pores and cracks. Machine learning and data science can be employed for designing complex topologies, optimizing process conditions, and quantifying the characteristic microstructures. In this article, the optimization of process conditions of the L-PBF process, the quantification of microstructures in L-PBF-processed Al-Si alloys and WC/Co cemented carbides, and the design of cellular structures for thermal management applications are reviewed.