Is a 100 meter telescope feasible?

José Ribeiro het606b 2001

When one is confronted with the idea of an optical telescope which objective has a larger surface than a soccer field, the first reaction is undoubtedly an emotive one.

In fact, during the last century the increasing rate of the diameter of the objectives of the reflecting telescopes has been of 2x per every thirty years. If in the next thirty years it will be possible to built a 100 meter telescope, the increase will be of order of 10x; a similar step to the one achieved by Galileo when he used a telescope for the first time, relatively to the naked eye observations done so far.

A lot of the research and development done in the last decade has lead to the technology which in fact may allow the building of an extremely large telescope (ELT). The optical segmentation of Keck telescope opened the possibility to the construction of large diameter mirrors. The active optics from VLT, Gemini, and Subaru proves that it is possible to build large telescopes with automatically controlled optical quality. The construction of the optics of the Hobby-Ebberly telescope, in which the segments were massively built, proved that it is possible to drastically decrease the cost of the optics manufacture [1,2]. With the development of the adaptive optics technology is possible to use large telescopes providing diffraction-limited angular resolution. These facts allow an optimism in the teams that presently study and project ELTs (OWL, XLT, CELT). From their point of view the technical limitations are not really in the diameter of the primary mirror, which “feasibility limit is estimated in the 130-150 m range” [1], but in the computer calculation capacity of the control of the active and adaptive optical systems. The costs are also a limitative relevant aspect, as until today they vary with a factor of D^2.6, which will have to be lowered in order to allow a project of this standard to be achieved. The costs must be justified by the scientific expectations an ELT may lead to.

A 100 meter telescope can achieve a visual magnitude limit of 38 in ten hours for isolated point sources [3]. This capacity of light gathering allied to diffraction limited resolution makes the difference between an ELT and any other optical system. From the present technology it would be unthinkable to put in orbit a telescope of this size. Therefore, and relatively to HST and to NGST (the actual and the future space telescopes respectively), an ELT would be a “better imager at wavelengths lower than 2.5 micrometer and a better spectrograph with resolutions of 5000 at wavelengths lower than 5 micrometer” [3], and with lower operational costs.

The optical interferometry systems, despite their good angular resolutions depending on the baseline, are limited by the capacity of light gathering of their own optics. This means that two telescopes in interferometry, with the same baseline as the diameter of an ELT, have the same angular resolution of an ELT, but gather less light than the ELT.

The science an ELT allow is placed in the study of very faint light sources and distant sources, from brown dwarfs in Magellanic Clouds to Supernovae at z<10. Due to ELT angular resolution of 2-3 miliarcseconds, it will be possible the imaging of the closest stars surfaces. It will be possible to image stars of the same mass as the Sun in galaxies of the Virgo Cluster, as well as white dwarfs in Andromeda galaxy [3,4]. The detection of very distant standard candles such as Cepheids and type Ia Supernovae will allow a better determination of the distances as well as a better determination of the Hubble constant. No doubt, an ELT has its own niche of competence in certain scientific areas for which there will not be a substitute in the near future, thus fully justifying its construction.

The design of a telescope with such dimensions presents several difficulties, from the theoretical and practical point of view. First of all, to define the requirements for the scientific objective features: collecting area, science field of view, wavelength range, Strehl ratio, angular resolution and sky coverage [6]. From these data one chooses the optical design that best fulfils them. Due to their large dimensions the primary and secondary mirrors should be built by segments. Due to the production costs, the primary mirror should be spherical, taken into account the simpler figuring segment massive manufacture. The segments are hexagonal and their dimensions are of approx. 2 to 2,3 m in order to be transported at lower cost. [5,7]. For a 100 meter diameter mirror this corresponds to about 1600 segments. In OWL’s case a 34 m plane secondary mirror is proposed, placed at a distance of 95 m from the primary mirror. The advantage of the plane mirror is to eliminate the lateral decenters [5].This mirror is formed by 210 segments. For such a telescope the manufacture of the segments will take about ten years.

Thanks to the segmentation technique it is possible today to overcome the line of the double of the diameter per generation factor. Segmentation allow a set of mirrors (segments) altogether to become a larger mirror. Therefore, it is possible to obtain extremely thin and light mirrors. The monolithic mirrors are limited by the difficulty to obtain substracts allowing stability in large diameters. Presently, it is only possible to manufacture monolithic mirrors below 10 meter diameter. However, segmentation presents several problems, such as:

“- Inhomogeneities of the Coefficient of Thermal Expansion (CTE) between segments, leading to variable focus errors from segment to segment.

- CTE inhomogeneities within individual segments, leading to higher spacial frequency errors than for a large monolithic mirror.

- Thermal gradients in the segment which require the matter to have a low ratio CTE/thermal conductivity” [4].

- Segment phasing and tilt control is a challenging task [1,7,8].

- Matching secondary and primary segment patterns is a problem to take into account.

The light from the secondary must be corrected for spherical and field aberrations. The corrector must also incorporate active optics and field stabilisation capabilities. Active optics control the segment mirrors phasing and tilt, the centering of mirrors and tracking. If top level requirements include seeing–limited performance in the technical field of view (before adaptive correction), then the corrector must perform also adaptive correction of the atmospheric turbulence [5,6]. The corrector also has the function of focusing. The operation of the telescope in the technical field of view (without adaptive optics) in seeing-limited mode is executed for telescope testing [10].

Between the technical field of view and the scientific field of view is then placed the adaptive optics system. This system aims to correct aberrations introduced by the atmospheric turbulent layers. Its functioning principle is the following one:

Sensors (Shack-Hartmann or shearing interferometer) detect the light of one or more stars in the observation field. Each time the light of these stars moves due to the alteration of the air mass between the telescope and the stars a deformable mirror (“rubber” mirror) will vary its own shape in order to correct the distortions in the wavefront. Should the observation field not contain sufficiently bright stars to activate the sensors, it will then be necessary to use artificial stars. These stars are “manufactured” by 20w power lasers tuned to the absorption and emission wavelength of Sodium. The lasers will excite the Sodium-rich layer at 90km height creating an artificial yellow-orange star bright enough for the wavefront sensor. In the case of ELTs, several stars will be needed (10 to 27), natural or laser guide stars [11] as well as 2 or 3 deformable correcting mirrors. A technique that is being recently developed is the multi-conjugate adaptive optics (MCAO) also called atmospheric tomography (a term from medical computerised axial tomography where one can obtain three-dimensional images).It calculates the three-dimensional behaviour of the atmosphere above the telescope [12]. For a 100m class, this means deformable correcting mirrors with 50000 (near-infrared) and 500000 (visible) active actuators [2]. This involves input rates of the order of 2Gpixels/s which will have to be processed by the computer, thus equivaling to a factor of 300 to 400x relatively to what today is required in adaptive optics in the class of 8 meter telescopes [4]. In the science focus a diffraction-limited resolution will be obtained, a mandatory requirement for an ELT.

Regarding the mounting of the telescope, it tends to be altazimuthal, with a rotation of 360º in azimuth and approximately 60º from Zenith for observation and 90º from Zenith for maintenance.

The mechanical structure should resist stress created by the weight of the optical systems. “The structure must provide sufficient bandpass for the motion control system as well as the required tracking accuracy and dimensional stability under varying thermal and wind loads” [13]. It should be calculated taking seismic activity into account.

An enclosure is also feasible for a telescope of the 100m size [9]. Some defend that open air circulation within the structure is enough for the system to reach thermal equilibrium giving up the need for the enclosure [1].

The choice of a suitable place for an ELT should, apart from the atmospheric aspects, must taken into account the weight and surface of the structure. The place should be as windless as possible, with small seismic activity, and with a geological feature allowing structural stability.

An Extreme Large Telescope of 100 meter diameter is feasible within this generation time. Surprisingly, the main facing obstacle is the present computational incapacity to control the active and adaptive optical systems required for such a telescope. Yet, the time it will take to build the structure and the optics (8 to 10 years) is hypothetically enough to enable the capacity of calculus to fulfil its requirements. One hopes the scientific discoveries this type of telescope achieves will have the same impact as the discoveries made by Galileo.

References

1. Progress of the OWL 100-m Telescope Conceptual Design

Philippe Dierickx and Roberto Gilmozzi

http://www.eso.org/projects/owl/

2. OWL Concept Overview

Philippe Dierickx and Roberto Gilmozzi

http://www.eso.org/projects/owl/

3. Science with 100m telescopes

Roberto Gilmozzi

http://www.eso.org/projects/owl/

4. The Future of Filled Aperture Telescopes: is a 100 feasible?

Roberto Gilmozzi et all