Increasing stability in robotic GTA-based additive manufacturing through optical measurement and feedback control

Abstract

Abstract Additive manufacturing employing Gas Tungsten Arc (GTA) as the heat source is capable of fabricating fully dense metal components layer upon layer. However, this process is sensitive to various disturbances and needs on-line detection and adjustments. In this work, a visual sensor, comprising a camera and composite filters, is developed for automatically real-time sensing of the fabrication process. The aim is to keep stable manufacture, and the deviations of the deposited height are compensated by designing an integral separation PID controller to adjust the wire feed speed in the next layer. The optical measurement technique and the controller are estimated via building multi-layer single-pass walls. The results show that the process stability in GTA-based additive manufacturing is well controlled when the designed visual sensor and the proposed closed-loop controller are applied.