A wire deflection detection method based on image processing in wire + arc additive manufacturing

In wire and arc additive manufacture (WAAM), the twist of wire during a robot’s movement can result in the sudden changes of the wire-feeding position and thus cause deposition defects and dimensional errors. In the worst case, it may cause wire jamming and damage of the wire-feeding system. Therefore, online monitoring and correction of the wire deflection are very important for WAAM. In this paper, a vision-based measuring method is proposed for detecting the deviations of the wire-feeding position of a plasma welding-based WAAM process. It uses adaptive threshold and Hough transform to extract the wire edges, judges and merges the coincident lines, and applies Radon transform to measure the wire deflection. Software to automatically detect the wire deviation was developed based on the proposed method. The method and the software were verified with experiments.