Towards automatic control of process stability for thick-wall parts in arc-directed energy deposition based on arc voltage sensing
Dashuang Chen, Hui Chen, Guangjun Zhang, Jun Xiong
Abstract
Gas tungsten arc-directed energy deposition (GTA-DED) is a highly anticipated technology to fabricate high-quality components. However, research on the automatic control of process stability in GTA-DED is very weak, especially for building thick-wall parts with multi-layer multi-pass features. This study proposes an arc voltage (AV) detection and feedback control strategy to accurately control as-built height for thick-wall parts in GTA-DED. A sensor is designed to monitor AV signals, and an ant colony optimization (ACO)-based wavelet thresholding denoising algorithm is developed to eliminate noises in raw AV signals. The novelty is that AV signal differences of six typical arc shapes under the same arc length are explored. Aiming at accurately controlling deposition height for process stability, a model-referenced adaptive controller (MRAC) is developed to control the peak AV signals by regulating the wire feed speed (WFS). A 38-layer 10-pass thick-wall component is printed to assess the MRAC's robustness. Continuous automatic printing of the thick-wall component is realized. The detected peak AV signals can well track three reference peak AV signals, and the total height difference of the thick-wall component is no >0.5 mm. This study can provide valuable guidelines and solutions for enhancing the automation level of printing thick-wall parts in GTA-DED.