Bachelor of Engineering
School of Mechanical, Materials & Mechatronic Engineering
Wire and arc additive manufacturing technology has the unique characteristics of high deposition rate, open architecture and high energy efficiency during the manufacturing of metal products. This thesis focuses on controlling the quality of deposited beads in the single layers of a multilayered wall using gas tungsten arc welding-based additive manufacturing. The quality of a deposited bead is strongly associated with its geometry, hence the width of weld bead should be monitored and controlled. A passive vision sensor system has been combined with an adequate composition of filters to filter out the arc light. An adaptive exposure-time control method was developed to obtain clear weld bead images under various welding conditions. The significance of welding parameters on weld bead geometry was thoroughly analyzed. The welding parameters with the strongest effect on weld bead geometry were isolated using the ANOVA statistical analysis technique. A closed-loop controller was developed to regulate the weld bead width. This approach was based on the features of deposited bead images extracted through an improved edge detection algorithm and image enhancement techniques. The results indicate that the proposed system is capable of capturing clear weld bead images during the welding process and controlling the width of deposited beads.
Ma, Zihan, Vision-sensing and bead width control of the first layer of multi-layered part in GTAW based Additive Manufacturing, Bachelor of Engineering thesis, School of Mechanical, Materials & Mechatronic Engineering, University of Wollongong, 2017. https://ro.uow.edu.au/theses1/55