Weld quality measurement of T-joints based on three-dimensional scanner
Jiadong He,Haibo Wang,Yajun Zhang
DOI: https://doi.org/10.1007/s00170-024-13847-7
IF: 3.563
2024-07-13
The International Journal of Advanced Manufacturing Technology
Abstract:We propose a weld quality measurement system specifically designed for T-joints using a surface-structured light stereo scanner. The T-joint, which consists of a rib plate and a base plate, undergoes assessment based on the welding angle, height, and width. The system begins by employing a scanner to capture point cloud data of the rib plate and base plate. To simplify this data, pass-through filtering and voxel meshing methods are incorporated. Following this, the T-joint point cloud is segmented using a region-growing algorithm, and the normal vectors of the rib plate and base plate are calculated using the random sample consensus (RANSAC) algorithm. To extract the weld lines from the T-joint point cloud, the fast point feature histogram (FPFH) algorithm is utilized. The welding angle is subsequently determined from the normal vectors of the two plates, while the weld height and width are extracted using the weld line point cloud. To evaluate the effectiveness of the system, it is applied to a T-joint in a Ti-6Al-4 V alloy–stiffened plate. The results demonstrate that the system can accurately detect the welding angle, height, and width, thus providing valuable technical support for the timely monitoring of weld quality in T-joints throughout the welding process.
engineering, manufacturing,automation & control systems