Subaperture stitching surface errors due to noise

Abstract

Subaperture stitching is a popular method for extending small, subaperture interferometer measurements to cover largeaperture optics. The method is simple in that there are only two steps: 1) make multiple measurements across the surface and 2) use well-established software techniques to merge the individual measurements into one surface estimate. Because parts of the system must move between measurements, small misalignments between subapertures are unavoidable, but easily accommodated within the software. Unfortunately this process has the potential to introduce errors. In this work, we show how random noise in a circular ring of subapertures creates artifacts in low-order surface shape estimates. The magnitude of these errors depends on setup parameters such as the number of subapertures and their overlap, as well as the measurement noise within a single subaperture. Understanding the relationships between subaperture stitching configuration and surface artifacts is important when designing high-accuracy metrology systems which rely on subaperture stitching. This work will help metrology system designers incorporate subaperture stitching into error budgets and tolerances.