Multi-Objective Surrogate Optimization of Process Parameters for Additive Manufacturing With Applications in Laser Powder Bed Fusion

Abstract Widespread adoption of additive manufacturing (AM) is hindered by challenges in achieving part quality using metrics such as geometry accuracy and porosity. These metrics are affected by the microstructure and mechanical properties of fabricated parts which can be controlled by the AM’s process parameters. Fine-tuning these parameters can enable control over the part quality. In this study, an optimization-based approach for selecting the AM process parameters is proposed for achieving part quality. The proposed approach integrates design of experiments, AM process simulation, surrogate modeling, and multi-objective optimization. While the proposed approach is general and applicable to any AM process, the applicability of the proposed approach is demonstrated through a laser powder bed fusion (LPBF) process. Three LPBF process parameters, namely layer thickness, laser power, and scanning speed are considered for obtaining optimized part quality considering geometric accuracy and porosity. A cylinder and a heat exchanger example are used to demonstrate the effectiveness of the proposed approach with the LPBF process. For these examples, it is shown that with the optimized process parameters, the part has about 17% better geometric accuracy when compared to the unoptimized part while satisfying a porosity requirement. The results also reveal that laser power is the most influential process parameter affecting both the geometric accuracy and porosity.