Surface Grinding of Borosilicate Crown Glass Optics via a Robotic Approach Based on Superposed Trajectories

The production of large-sized optical components with complex shapes requires several phases, including surface finishing. Currently, mainly skilled workers can correctly perform this operation, divided into the successive steps of grinding and polishing, leading to long production times, poor reproducibility of results, and exposure to human error. For this reason, the industry is trying to move towards automation involving, for example, high-precision machine tools and machining centers. However, these solutions require high investment costs and long setup times. Using robotic cells helps to reduce these expenses, manufacture larger components, and increase the flexibility in the production chain. In this research, we present an unconventional approach to the robot-assisted grinding of optical samples made of borosilicate crown glass. The samples were guided by a six-degree-of-freedom industrial robot on a rotating grinding disc while imposing to them different trajectories with complex geometry. We avoided regular grinding patterns, which are easily recognizable by human eyes and affect the quality assessment, by superposing multiple relative movements between the machined surface and the abrasive grains. The ground surfaces of the samples were characterized based on average roughness values, profile error data, and surface topography images. Finally, we selected the best robotic grinding procedure matching the trajectory and strategy with optimal surface quality, processing time, and productivity. The suggested methodology not only shortens the manufacturing sequence by eliminating manual methods but also provides components with optical properties within the required specifications for subsequent polishing steps.