Research On Online Detection And Motion Tracking For Friction Stir Welding

Though Friction Stir Welding(FSW)becomes the preferred welding method for high-speed trains,it will cause some welding defects.Therefore,online detection and tracking in the process of FSW has very important engineering significance.The current method can't meet the requirements of FSW which requires weld seam gap to be less than 0.2mm.In order to solve these problems,a new research is necessary to be done.Technology of online detection and tracking in the process of FSW based on the principle of laser monocular vision is researched.By the way of images capture and sensor calibration,the 2D point cloud can be converted into 3D point cloud.Then methods of calculating weld seam features based on the measured point cloud data are researched.Besides,series of experiments were conducted to verify the methods.We have done some experiments to verify the accuracy and repeatability of the measurement system.The results show that our device can reach good results.