A Novel Welding Method for Repairing Surface Defects of Large-Type Rotary Machinery Based on Line Structured Light Detection

Abstract Large-type rotary machinery is the core components of national major projects which is widely used aviation, electric power, metallurgy, energy and construction machinery industries. Surface defects of Large-type rotary machinery such as cracks and pits are usually processed into groove with a certain shape first, and then the processed groove is repaired by manual welding. This manual welding repair method has a low level of automation, and the repair quality of the groove is difficult to guarantee. Therefore, this paper proposes a novel welding method for repairing surface defects of Large-type rotary machinery which uses the Kollmorgen Joint Modular Robot to complete the welding repair of the processed groove. Firstly, the groove point cloud data collected by Line structured light sensor is processed by the designed algorithm to obtain the contour characteristics of the groove. Then, the arrangement of welding pass is completed based on contour characteristics of the groove. Finally, the trajectory of the welding robot is determined by the position of welding pass. The planned trajectory verification is completed on the simulation experiment platform and the result shows the accuracy and reliability of the planned trajectory which has certain theoretical and practical significance for realizing the automation of on-site maintenance of Large-type rotary machinery.