Research on the modification of the tool influence function for robotic bonnet polishing with stiffness modeling

•The stiffness model of the robotic bonnet polishing system is developed using Jacobi matrices and finite element simulation to include the robot joint stiffness and body stiffness.•The time term in the tool influence function (TIF) for robotic bonnet polishing is corrected to generate a time matrix of dwell points on the surface of the workpiece.•The fixed-point polishing experiments show that the maximum relative error before the TIF correction is 12.5% and the maximum relative error after the correction is 2.8%. The plane polishing experiments show that the maximum relative error before the TIF correction is 13% and the maximum relative error after the correction is 1.7%. Robotic bonnet polishing (RBP) technology is widely used in the polishing process of optical components, but industrial robots have the characteristics of a wide processing range and low body stiffness. Therefore, in the actual polishing process, there are different polishing forces due to different stiffnesses, which leads to different tool influence functions (TIF) affecting the surface quality of optical components. To achieve a uniform surface quality of the components polished by the RBP, we modeled the stiffness of the robot and modified the TIF in conjunction with the Preston equation, finally verifying the accuracy of the modified model through experiments. The results of the fixed-point polishing experiments show that the maximum relative error before the TIF correction is 12.5% and the maximum relative error after the correction is 2.8% within the robot's 600mm*600mm machining range. The results of the plane polishing experiments show that the maximum relative error before the TIF correction is 13% and the maximum relative error after the correction is 1.7%. Therefore, a method for TIF corrections from RBP based on stiffness modeling has important engineering implications.