Toolpath smoothing and motion planning for the diamond turning of microlens arrays with improved uniformity

Abstract

Microlens arrays fabricated by ultra-precision machining commonly suffer from surface nonuniformity, which results in the inconsistency and degradation of their optical performances. One major possible cause is the uneven distribution of toolpath kinematic characteristics, especially the excessive local acceleration and jerk values concentrated at lens edges. To address the issue, this study proposes an optimization strategy comprising toolpath smoothing and motion planning. The toolpath directly generated using equal-angle discretization is first processed by spline interpolation-based smoothing to eliminate the abrupt turnings at lens edges. The smoothed toolpath is then optimized by a replanning process which adaptively adjusts the cutter location point (CLP) density according to local curvature. Lastly, the replanned toolpath and CLPs are further finely tuned by genetic algorithm in which acceleration and jerk feedbacks are integrated using penalty terms. Theoretical analysis demonstrates greatly reduced acceleration and jerk peak values by over 90 % compared to equal-angle discretization. Significantly improved surface uniformity, both among differently located lenses and within a single lens, is verified by experimental investigation. These findings provide a potential solution to the nonuniformity of machined microstructured surfaces, enabling the advanced and consistent performances of optical systems.