Research on Reconstruction of Weld Pool Surface Based on Shape from Shading During Robot Aluminum Alloy Pulse GTAW

Abstract

In the interest of controlling the weld bead formation quality during robot aluminum alloy gas tungsten arc welding (GTAW) process, it’s necessary to monitor the shape and surface height information of the weld pool in real-time, which can indicate the penetration and formation. This paper presents an algorithm of shape from shading (SFS) based on single image for weld pool surface reconstruction. Two classic SFS algorithms are analyzed with the results of reconstruction from the synthetic half sphere image. Based on the Zheng and Chellapa algorithm, this paper proposes some measures for the improvement, such as smoothing process, strengthening the boundary constraints, using the known characteristics of the object surface and weighting error functions. And the improved algorithm is used to reconstruct the aluminum alloy weld pool surface during the robot pulse GTAW process. The result shows that the algorithm can successfully reconstruct the weld pool surface with effectiveness and accuracy and the computation of the surface height can be applied in real-time.