Minimizing the surface roughness in L-PBF additive manufacturing process using a combined feedforward plus feedback control system

The poor surface quality during the laser powder bed fusion (L-PBF) process adversely impacts the mechanical properties of the final product and even can result in failure of the process. This study aims at minimizing the top surface roughness of the parts manufactured by the L-PBF process by deploying a feedforward plus feedback control system. The most common factors affecting the surface quality, namely, balling, lack of inter-track overlap, overlapping curvature of laser scan tracks, and spatters, were investigated through a monitoring system consisting of a high-speed camera, a zooming lens, and a short-pass filter. The desired melt pool width and the critical value for the level of spatters were determined using the imaging system and subsequent image processing. An experimental model was developed, and the control system was designed accordingly. The performance of the control system was evaluated by simulations and experiments. In all cases, the control system showed an excellent transient performance to reach the desired melt pool width only after printing a few layers. The results obtained from this study showed that the average arithmetic mean surface roughness value (Sa) reduced from 10.48 to 5.91 mu m and from 9.61 to 6.64 mu m at 500 mm/s and 400 mm/s scanning speed, respectively. In addition, evaluating the controller on a bridge geometry showed that controlling the geometry of the melt pool can mitigate significant defects occurring during the process and minimize the top surface roughness.

Automated summary

This study focuses on the poor surface quality in the laser powder bed fusion (L-PBF) process and aims to minimize the surface roughness of the manufactured parts by using a feedforward plus feedback control system. The factors affecting surface quality were investigated through a monitoring system, and the desired melt pool width and critical spatter level were determined. An experimental model and control system were developed, and the performance was evaluated through simulations and experiments. The results showed a significant reduction in surface roughness and the ability to control the geometry of the melt pool to mitigate defects.