Ritchey-Common test for low uncertainty metrology of ELT pre-focal station M6N mirrors

The pre-focal stations of the Extremely Large Telescope contain the final optical components before the light collected by the 39m primary mirror is focused into the various scientific instruments of the telescope. The deployable fold mirrors must, therefore, be of very high quality to maintain the accuracy of the wavefront that reaches them. The two primary elliptical mirrors of each pre-focal station were polished by Glyndˆwr Innovations Ltd, to a required flatness specification of less than 30nm RMS for each mirror. To achieve this precision on the larger, 1460mm wide, M6N mirrors, a Ritchey-Common test was used to measure the surface error of the mirrors after each polishing run. This type of test utilises a spherical reference mirror to refocus the light from the interferometer, creating a spherical test where the flat mirror is in the path. The surface error of the flat mirror must be disambiguated from all the other sources of error and uncertainty in the test. Since the uncertainty of the test had to be exceptionally low, many factors were considered in depth, including mirror support stresses, integration variations, test alignment, and reference mirror shape fluctuations. Without careful measurement and calibration, any one of these factors could easily cause an uncertainty higher than the required flatness specification. The final Ritchey-Common test uncertainties were quantified to be less than 14nm RMS for the full aperture of the M6N mirror. The method by which this uncertainty was achieved is detailed through the design and verification of the test.