In-Situ Monitoring of Additive Manufacturing Process Based on Vibration Data

Abstract

With the increasing application of additive manufacturing in various fields, it is particularly important to monitor the manufacturing process, product quality, and printer status aiming at obtaining high-quality manufactured products. This paper introduces a kind of additive manufacturing process monitoring based on the vibration data of the fused deposition modeling(FDM). The random forest model has been applied to identify the normal state and filament jam state of the printer, which could achieve an accuracy rate of 94 %. The experimental results demonstrate that the proposed method can accurately identify the mechanical faults in the process of additive manufacturing, and effectively monitor the machine health and ensure the fabricated parts quality.