Application of data-driven methods for laser powder bed fusion of Ni-based superalloys: A review

Ni-based superalloys, with their high mechanical integrity at high temperature, have become critical engineering materials in modern advanced mechanical systems. However, this characteristic also makes them difficult to process. The laser powder bed fusion (LPBF) technique, with its high processing temperature, high design flexibility, and no tooling requirements, presents a promising way for manufacturing high-performance Ni-based superalloy components. However, the LPBF process also presents additional challenges, such as controlling precipitates, suppressing micro-defects, and achieving the desired microstructure. To address these challenges, it is crucial to enhance the intelligence of LPBF technology through advanced engineering software and computer techniques. Therefore, in this review, we introduce the concept of the “ready-to-use” mode of LPBF technology, that is, real-time control. We survey recently reported studies related to data-driven LPBF of Ni-based superalloys, focusing on the relevant achievements in the processing stage. Simulation and machine learning are two main techniques that have received increasing attention. The review begins with a description of the complex microstructural features of Ni-based superalloys, providing a basis for discussing the data-driven optimization of alloy microstructure in the following sections. The process parameters involved in LPBF and their complex effects on microstructure are subsequently reviewed. The main section reviews the specific applications of simulation and machine learning in the LPBF of Ni-based superalloys, such as microstructure-related, manufacturing process-related, correlation analysis-related, and process monitoring-related. This review provides insights for researchers studying the LPBF of Ni-based superalloys or other difficult-to-machine materials to achieve intelligent manufacturing through data-driven approaches.