Point-driven toolpath curve and orientation smoothing in robotic belt grinding for turbine blade

The presence of noise or other abnormal points in the measured point clouds of the turbine blade can result in local discontinuities of the tool path curves, fitted by the machining path points generated by slicing the measured point cloud. In addition, the fluctuation exists in the tool orientation vectors corresponding to the cutter-contact (CC) points in the toolpath curves. These issues can lead to poor smoothing of robotic motion during the grinding process, thereby affecting the quality of the blade grinding. To overcome the above problems, a novel toolpath smoothing method for the measured point cloud model of the turbine blade is proposed. In this method, the initial path points are firstly generated by slicing point clouds of the turbine blade. Next spline segments replace the abnormal points in the initial path points and obtain the smooth toolpath curves. Then, for the initial tool orientation vectors distributed at the smoothed toolpath curves, an objective function by considering the directional deviation between tool orientation vectors and slicing planes is established to reduce the macro fluctuation of these vectors. Based on this, another objective function is established to filter out micro fluctuation among these vectors by considering the energy model of the motion surface of the grinding tool. The surface contour smoothness of the blade with the proposed method is improved by over 20 % compared to other toolpath planning methods.