A novel weld-pool-length monitoring method based on pixel analysis in plasma arc additive manufacturing

Abstract

The real-time monitoring of the weld pool during deposition is important for automatic control in plasma arc additive manufacturing. To obtain a high deposition accuracy, it is essential to maintain a stable weld pool size. In this study, a novel passive visual method is proposed to measure the weld pool length. Using the proposed method, the image quality was improved by designing a special visual system that employed an endoscope and a camera. It also includes pixel brightness-based and gradient-based algorithms that can adaptively detect feature points at the boundary when the weld pool geometry changes. This algorithm can also be applied to materials with different solidification characteristics. Calibration was performed to measure the real weld pool length in world coordinates, and outlier rejection was performed to increase the accuracy of the algorithm. Additionally, tests were carried out on the intersection component, and the results showed that the proposed method performed well in tracking the changing weld pool length and was applicable to the real-time monitoring of different types of materials.