Euclid. VII. Structural-thermal-optical performance

Context. The performance of the Euclid system is defined in terms of image quality metrics tuned to the weak gravitational lensing cosmological probe. The weak lensing measurement induces stringent requirements on the shape and stability of the VIS instrument system point spread function (PSF). The PSF is affected by error contributions from the telescope, the focal plane, and image motion, and it is controlled by a global error budget, with error allocations for each contributor.Aims. During development of the spacecraft, we verified through a structural-thermal-optical performance (STOP) analysis that the built and verified telescope with its spacecraft interface meets the in-orbit steady-state and transient image quality requirements under temperature-induced loads in all permitted spacecraft attitudes after all permitted attitude transitions. Based on data from its first year in orbit, we compared the performance we expected with the actual performance.Methods. For the purposes of the STOP analysis, we set up a detailed finite-element mathematical model and defined a standard set of test cases, both steady-state and transient, comprising combinations of worst-case boundary conditions. Iterations of the analysis were performed in conjunction with the major reviews of the spacecraft verification cycle. After launch, we applied the model in sensitivity analyses using realistic boundary conditions.Results. The STOP analysis addressed the interaction of all spacecraft components in transmitting temperature-induced loads that lead to optical train deformation. The results of the prelaunch analysis demonstrated that temperature-induced optical perturbations would be well below the allowable limits for all permitted observing conditions. We used the STOP analysis predictions to help interpret the measured performance of the spacecraft as a function of environmental variables during its first year in orbit. We discovered unpredicted disturbances (heat pulses from instrument operation propagating into the telescope), and unexpected sensitivities (e.g. a high dependence of the telescope baseplate temperature on the solar aspect angle; a nearly absent dependence on the azimuth angle after the attitude domain was redefined for stray light avoidance). In-orbit temperature variations are small (<300 mK) and so are their effects on the telescope structure (displacements < 1 mu m, rotations < 1 mu rad), but they are detected in the time histories of the image quality metrics and are a non-negligible factor in the point spread function stability budget demanded by the weak lensing science (Delta e < 2x10(-3) over 11 000 s). Taking everything into account, our analysis confirms the overall excellent performance of the telescope.