The Euclid STM VIS Focal Plane Assembly metrology: description of the method to measure the CCDs position and the flatness of the full camera

摘要

In the frame work of the ESA Euclid mission to be launched in 2021, the Euclid Consortium is developing an extremely large and stable focal plane for the VIS instrument. After an extensive phase of definition and study over 4 years made at CEA on the thermo-mechanical architecture of that Focal Plane, the first model (Structural and Thermal Model) has been assembled qualified and delivered to MSSL in June 2017. The VIS Focal Plane Assembly integrates 36 CCDs (operated at 153K) connected to their front end electronics (operated at 280K). This Focal Plane will be the largest focal plane (～0.6 billion pixels) ever built for space application after the GAIA one. The CCDs are CCD-273 type specially designed and provided by the Teledyne e2v company under ESA contract. The Structural and Thermal Model is fully representative of the Flight Model regarding the thermo-mechanical architecture. The STM FPA thus integrates 36 CCDs representative of the flight model except for the detection function. We have implemented specific equipment in order to perform the metrology of the full FPA. It consists of the measurement of the flatness of the full camera as well as the determination of the position of its 36 CCDs. The purpose is to measure the dimension of the sensitive area and to localize each CCDs' image area with an uncertainty of +/-50 μm in X- and Y-directions. These positions are then given at room temperature in the reference frame of the main FPA structure that is interfaced with the Euclid telescope. The metrology also implies the verification of the flatness of the focal plane in the range of +/-60 μm with an uncertainty of +/-10 μm. Indeed, we must check that the design and the assembly of the FPA meet this requirement that consists of considering that the full photosensitive area is included in a volume of 120 μm high. Based on a marble with a flatness of 10 um and two motorized linear stages, the test bench combines a CCD camera and a confocal sensor. The camera allows localizing the four fiducial crosses of each CCD-273 and thus to define a grid of 9 equally spaced points on the image area. We can then measure thanks to the confocal sensor the flatness of the full sensitive area in 324 points across the FPA. In this paper, we describe the test bench and the method that we have validated for the STM program. We thus report the results for the STM FPA5 with an estimation of the uncertainty of +/-10μm for the flatness measurement and around +/-24 μm (including a coverage factor of 2 for a level of confidence of 95%) for the relative position of the CCDs, which corresponds to twice the pixel size of the CCDs. We finally indicate the improvement that we plan to implement to better estimate the CCDs' position in the FPA coordinates.