Rapid convergence of mid-spatial-frequency errors based on a vibration-assisted polishing strategy

Silicon carbide and other high-hardness materials are widely used in lightweight space telescope mirrors and laser optical systems due to their high specific stiffness, low thermal expansion coefficient, and excellent thermal conductivity. However, their high hardness and brittle removal mechanism lead to a significant reduction in the material removal rate during polishing, making it difficult for conventional process optimization and path planning methods to substantially improve the convergence limit of mid-spatial-frequency (MSF) errors. To address this challenge, this paper presents what we believe to be a novel industrially robot-integrated, unidirectional vibration-assisted polishing tool with adjustable vibration direction. By introducing micro-scale vibrations at the tool-workpiece interface, this tool effectively expands the coverage area of abrasive particles, improves their distribution uniformity, and enhances the activation level of abrasive grains. This mechanism successfully breaks the anisotropy of the removal function in the polishing of high-hardness materials and reduces the periodic accumulation of regular errors, thereby significantly improving the convergence efficiency of the MSF error and the robustness of the process. Corresponding experimental results validate the positive contribution of the vibration-assisted polishing strategy to the convergence of mid-spatial-frequency errors.