Etude, alignement et contrôle de surfaces optiques segmentées ou discontinues. Applications en Sciences de l'Univers

摘要

Segmented or discontinuous optical surfaces have been known since the Antiquity. They generated numerous applications, the first of which probably being Archimedes's “burning mirrors” intended to concentrate solar energy and destroy hostile sea vessels by fire. This basic idea governed the construction of modern solar furnaces dedicated to fundamental researches on resistant materials placed in extreme conditions, or helio-electrical power plants. Although the manufacturing tolerances of the employed optical surfaces are typically around one millimeter, they still require the utilization of accurate alignment and control methods. Hence the principle of the “backward gazing method” developed at the IMP (Odeillo, French Pyrénées) during my PhD years can be seen as a close natural parent of the adaptive optics techniques that have now become indispensable for astronomy and astrophysics observations. But discontinuous optical surfaces can do much more than focusing light energy. The historical experiments of Fizeau and Michelson demonstrated their ability to measure core astrophysical parameters at very high angular resolution, paving the way for a new generation of astronomical observing facilities. Among those are found multi-aperture stellar interferometers whose individual telescopes may be separated by several hundred meters (such as the Very Large Telescope Interferometer in Chile), new generation Extremely Large Telescopes (ELTs) equipped with segmented primary mirrors (such as twin Keck telescopes or space borne JWST), or futuristic hyper-telescopes looking for direct images of planetary systems orbiting around nearby stars. Such complex facilities whose specifications are becoming more and more ambitious should be co-phased (i.e. behaving together as a single optical system would do) within an accuracy of one-tenth of wavelength for imaging applications, and one-thousandth in the case of nulling interferometers. It then becomes necessary to develop new techniques for modeling and controlling those extremely demanding systems. Some of these techniques are presented here in the frame of future ELTs of 30 meter diameter or greater, and of sparse apertures interferometers hunting for exoplanets such as Darwin and TPF-I. Discontinuous optical surfaces are also present in the field of modern spectroscopy: in addition to conventional diffraction gratings, they have now come into the heart of new generation spectral-imaging instruments, being able to simultaneously imaging astrophysical objects and their spectral decompositions on a single detector chip. Hence the MUSE instrument incorporating twenty-four “image slicers” composed of discontinuous arrays of mirrors will enable the VLT to observe primordial galaxies in the forthcoming years. In view of so many promising applications, it seems quite clear that the techniques of manufacturing, aligning and testing segmented or discontinuous optical surfaces should play a major role in the astrophysical science of our century. In that perspective the SIRIUS test bench developed at Observatoire de la Côte d'Azur in order to evaluate the performance of hyper-telescopes and of their co-phasing methods could represent a key step, as is explained in the conclusion.