Process stability control of corner structures in robotic gas tungsten arc additive manufacturing via arc sensing

Aiming at a higher deposition rate, robotic gas tungsten arc additive manufacturing (GTA AM) always uses the front feed mode in which the wire is fed in front of the arc along the torch travel direction. However, fabricating corner structures in this mode is still a challenging issue in robotic GTA AM, since the serious hump height at corner points during torch rotation will hinder the process stability and automation of corner structures. The novelty of this study is to solve this issue by developing an in-situ monitoring and control strategy. An arc voltage sensor is designed to monitor the arc voltage that can characterize the deposition height. The arc voltage decreases first and then keeps stable at the corner points during torch rotation. It increases rapidly as the torch runs away from the corner points. An empirical mode decomposition (EMD) algorithm filters noises in sampled arc voltages. The peak arc voltage is selected to characterize the arc length since it presents fewer fluctuations than the base arc voltage. A Fuzzy-PID controller is developed for the peak arc voltage control by regulating the wire feed speed (WFS). Open-loop and closed-loop control tests of corner structures confirm that the deposition process with the fixed WFS is forced to terminate due to the serious hump height, and that with the controlled WFS can lower the hump height markedly at the corner points. This study demonstrates that the developed arc voltage sensing and control system can remove metal humps at the corner points and enhance the process automation of corner structures in robotic GTA AM.