A method is proposed for polishing fast aspherics with a lap whose shape is continuously changed under computer control as it moves over a rigid mirror blank. The required changes of radius, astigmatism and coma in a circular lap are made with edge bending levers and tensioning members with screw actuators. This method of bending has been demonstrated in the laboratory.

ESO has two telescope projects: the NTT (New Technology Telescope) with 3.5 m aperture and the VLT (Very Large Telescope) with 16 m equivalent aperture. The former is already in the engineering phase and should be completed in 1987; the latter is still in the study phase and will be described at this conference by our colleague, D. Enard.

This paper is concerned with the primary support system of the NTT. The basic principles and layout of this support have already been described in the literature 1,2,3).

The aim of this paper is to describe the support system which has been realized for the 1 m ESO active optics experiment. This experiment has been conceived in order to verify on a smaller scale the concept of the active optics, which is foreseen for the ESO NTT 3.5 m telescope. This concept has been widely discussed in references (1) (2) (3).

For the purpose of completeness we will briefly summarize here the basic concept of the experiment, while we will describe in detail the support system and the characterization of the force actuator.

ESO has investigated the technological feasibility of a metallic primary mirror for the NTT. A brief status report is given indicating that a nickel coated aluminum or aluminum-alloy mirror blank can be manufactured within 15-18 months at costs well below those for a glass or ceramic material of identical shape and dimensions.

Warping is a major concern in the use of aluminium as a material for large mirrors. The warping characteristics are influenced by:

1. * the type of material, i.e. pure aluminium or some alloy

2. * the type of fabrication, i.e. cast, rolled or forged.

So far there is little data available on the warping characteristics of aluminium mirrors. Therefore ESO has tested the warping behaviour of 18 aluminium mirrors under thermal cycling conditions with a glass ceramic mirror, which has not been cycled, as a reference mirror. The set of 500 mm mirrors selected for this experiment is described in [1].

A way to follow in order to reduce the weight of the primary mirror is to accept an higher ratio diameter-thickness and to find a remedy to the consequent greater deformability of the mirror by an active control system of it. A technique using closed loop active optics control has been proposed in references.

In these papers, a coherent scheme of active optics control for the ESO New Technology Telescope (NTT) was presented, based on analysis of the image errors in terms of an appropriate polynomial (the ESO off-line telescope test polynomial) and the production of equivalent correction terms by force modulation of the primary axial support.

The Astronomers target for the present and the next decade is the construction of large telescopes 3,5 - 4 m. primary mirror diameter and very large telescopes 8 - 9 m. primary mirror diameter, in the array or multi-mirror design, improving at the same time the performance of the instruments compared to the efficiency and precision of the same.

The construction of very large telescopes is impeded by very large costs. Much effort is going into finding ways to get below the traditional cost-scaling trend line. We will address several factors that lead to reduction in cost without reduction in performance. Clearly, a very fast primary f-ratio is necessary. The practical lower limit is set by three factors:

1) width of the Nasmyth platform

2) new optical processing techniques

and 3) maintenance of collimation.

The ESO test device is based on the proposal of Shack

Fig. 1 shows the principle of this method, which we call the SHACKHARTMANN (S-H) method as it is derived from the well-known conventional Hartmann technique. The beam from the telescope focus (6) is transmitted by a beam-splitting cube (2) of high quality to a collimating (Fabry) doublet objective (3). This images the exit pupil of the telescope to the plane (4) where a special SHACK lenticular raster is inserted. Each lens of this raster then forms a point image of that part of the parallel beam incident upon it on to a photographic plate (5) placed at the focal distance of the raster lenses from (4). The angular aberration in the telescope image is thus transferred through the system and shows as a transverse displacement of the spots.

The testing of convex secondary mirror of large telescope is known as a difficault thing, because it needs for a large mirror or lens. In USA, they used to use Hindle sphere for testing convex hyperboloids, and in West Germany as I know, they often use concave matrix probe for testing convex mirrors. In the case of Hindle method, the diameter of testing sphere always is about 2 or 3 times as the convex mirror under testing, in the later case, the matrix probe must has the same size as the convex mirror, further more, the probe itself needs another testing facilities. Some other methods had been suggested by R. Wilson, E, Gaviola and others but they have rarely been used in practice.

Every manufacturer of large mirrors is familiar with the fact that the polishing process may not only lead to large-area surface errors, such as astigmatism or spherical aberration, but also to short-period errors. In particular, this applies to aspherical surfaces because they require polishing tools of small dimensions, as compared with the mirror diameter. These errors are not entirely irregular and may therefore not be treated as statistical errors; nor are they sufficiently regular to be described in terms of amplitudes of ZERNIKE polynomials as it is now ussually done in the case of large-area aberrations (F.FRANZA e. a. 1977). Unfortunately this still is true if very high radial and tangential orders are involved, e.g.in D. ANDERSON e.a.1982. The reason is quite obvious: The description by means of ZERNIKE polynomials is based on the assumption of a two-dimensional regularity which simply does not exist in practice.

Borra (1982) has argued that it is scientifically useful and technically feasible to build very large optical telescopes (diameters over 15 meters) having as primary mirror a rotating container filled with mercury. We report here the results of various optical tests on prototype liquid mirrors. The results are very encouraging and seem to indicate that large liquid mirrors are feasible.

A 75cm alt-az telescope with a primary mirror made from the Ohara E6 glass was completed in December, 1983, at Kitakaruizawa, Japan. We describe the surface accuracy of mirror during its polishing and in a telescope cell, and also discuss some new technical points developed for this telescope.

One important large telescope design study area is the dynamic image degradation caused by wind forces upon various telescope structures and optical elements. The need to establish rapid thermal equilibration with the observatory environment further complicates the wind exposure problem. On the other hand, a telescope deeply embedded in a large enclosure is subject to dome induced seeing difficulties unless great care is used to eliminate any thermal differences among telescope structure, optics, dome interior and inside and outside air. While dome seeing is outside the scope of this paper, it should not be neglected in the consideration of enclosure design because wind protection and seeing are inexorably coupled.

Scaling laws are presented which show the dependence of the tracking error of large telescopes on structural materials, mechanical properties and dimensions of the mount, and wind speed. Based on direct measurements on the Multiple Mirror Telescope, predictions are made for future very large telescopes. It is shown that good tracking can be achieved most of the time even without a traditional dome to block the wind, and this may result in better images by eliminating “dome seeing”.

In 1969 European solar physists formed the Joint Organization for Solar Observation (JOSO). The goals of JOSO were to find an, ideal site for a large solar instrument, and to create an observatory on this site equipped with a solar telescope for high spatial resolution and polarimetric observations. This instrument is referred to by the acronym LEST which stands for Large European Solar Telescope.

Since a lot of efforts and money are being put in building larger telescopes, astronomers are becoming more concern with the effects of “atmospheric seeing” on astronomical observations. Observers need to correct their observations for image degradation and put more and more emphasis on the importance of good seeing conditions. Both observers and optical engineers try to compare the image quality between different telescopes and sites. In addition, engineers are interested in locating the origin of image degradation. What are the relative contributions of the dome and of the free atmosphere to image degradation ? As engineers become able to reduce dome seeing, the selection of a good astronomical site becomes more crucial and reliable methods for site testing become necessary.