A profile measurement method of large aspheric optical surface based on optimal stitching planning

• An optimal stitching planning for two-dimensional profile measurement of large aspheric surface by applying small range profiler as the measuring machine is proposed.
• A data matching algorithm is developed to match the adjacent segments due to the invariance of curvature radius of overlap region.
• The accumulative stitching errors of three different curves are predicted within sub-micron level through simulations.
• The optimal stitching method is verified by experiments of two groups of 110-mm aspheric curves. The PV (Peak-to-valley) values of the stitching errors of the two curves are in the range of ±1 μm and the standard deviations are less than 0.2 μm.

This paper proposes an optimal stitching planning for two-dimensional profile measurement of large aspheric surface by applying small range profiler as the measuring machine. The originality of this paper reflects on two aspects: (1) According to the applied range of probe, the amount of overlap region and the geometric characteristic of large aspheric surface, a segmentation planning algorithm is presented to determine the segment parameters during the measuring process. Considering the measurement efficiency and stitching accuracy synthetically, the applied range of probe with ±4 mm and the amount of overlap region with 20% of segment length are regarded as the optimal parameters. (2) Based on the invariance of curvature radius of overlap region, a data matching algorithm is developed to match the adjacent segments. Through the simulation analysis, the deviation of adjacent segments is reduced to 0.03 mm with the sampling interval of 0.01 mm. Then on the basis of the simulation settings with the translation error of 1 μm and the rotation error of 30′, the accumulative stitching errors of three different curves are predicted within sub-micron level. Finally the optimal stitching method is verified by experiments of two groups of 110-mm aspheric curves. The PV (Peak-to-valley) values of the stitching errors of the two curves are in the range of ±1 μm and the standard deviations are less than 0.2 μm.