In-situ optical monitoring and analysis of weld pool based on machine vision for wire and arc additive manufacturing

In the additive manufacturing process, the geometric characteristics of the weld pool directly reflect the shape accuracy of the weld bead. Therefore, it is particularly important to extract the characteristic dimensions of the weld pool and establish a mathematical model between the welding process parameters and the weld pool characteristics to guide the online control of welding. However, there are strong arc light and splash interference during the forming process, which hinders the observation of the forming area and precludes the acquisition of real-time forming information. For this reason, an in-situ optical monitoring system for arc additive weld pool is established in this paper, and a filter component is designed to obtain a clear weld pool image. A weld pool contour extraction algorithm is proposed, and the weld pool contour size is calculated according to the camera calibration parameters. Subsequently, based on the extracted geometric features of the weld pool, a three-factor three-level orthogonal experiment is designed to investigate the effects of welding process parameters on the weld pool characteristics. The influence trends and degrees of different parameters on the weld pool characteristics are compared through response surface analysis. Furthermore, a mathematical model is established between welding process parameters and weld pool contour dimensions based on the regression analysis method, which is used for online prediction of weld pool morphology. Eventually, the verification experiment shows that the error between the predicted value and the actual value of the model is less than 0.55%, which can be used for the geometric characteristics of the weld pool under different process parameters.