Optomechanics of the Large Binocular Telescope

J. M. Hill

University of Arizona, Large Binocular Telescope Project
Steward Observatory,Tucson, AZ 85721
jhill@as.arizona.edu

P. Salinari
Osservatorio Astrofisico di Arcetri,
Large Binocular Telescope Project
Largo Enrico Fermi 5, 50125 Firenze, ITALY
salinari@arcetri.astro.it

http://medusa.as.arizona.edu/lbtwww/tech/optomech.htm

Proceedings of SPIE conference on Advanced Technology Optical Telescopes V, 2199, p. 64 (1994)

Abstract

1. Introduction

2. Optics

3. Telescope Structure

4. Summary

5. Acknowledgements

6. References

Abstract

We describe the optical layout of the Large Binocular Telescope. Recent changes in the baseline optical configuration include moving the wide field foci above the primaries to allow a one degree field at F/4. The infrared F/15 secondaries are now a Gregorian design to allow maximum flexibility for adaptive optics. The F/15 secondaries are undersized to provide a low thermal background focal plane which is unvignetted over a 4 arcminute diameter field-of-view. The interferometric focus combining the light from the two 8.4 meter primaries will reimage two folded Gregorian focal planes to a central location. The telescope elevation structure accommodates swing arms which allow rapid interchange of the various secondary and tertiary mirrors. Using swing arms has allowed us to remove most of the superstructure supporting the spiders in an earlier version of the design. Maximum stiffness and minimal thermal disturbance continue to be important drivers for the detailed design of the telescope. By concentrating the structural mass between the two elevation C-rings, we are able to achieve a 2 Hz increase in the lowest eigenfrequency without increasing the mass of the elevation structure. The telescope structure accommodates installation of a vacuum bell jar for aluminizing the primary mirrors in-situ on the telescope. The detailed design of the telescope structure will be completed during 1994.

1. Introduction

The Large Binocular Telescope (LBT) Project has evolved from concepts first proposed in 1985. There have been many changes in the LBT over the past two years. Perhaps most obvious is the change of name --- the former ``Columbus Project'' has changed its name to the politically correct ``Large Binocular Telescope''. Other changes include the configuration of the secondary mirrors and the design of the optical support structure of the telescope. The technical aspects of these changes are discussed below.

The partners involved in the design and construction of this 8.4 meter binocular telescope are the University of Arizona, Italy represented by the Osservatorio Astronomico di Arcetri and the Research Corporation based in Tucson. These three partners have committed sufficient funds to build the enclosure and the telescope populated with a single 8.4 meter optical train --- approximately 40 million dollars (1989). Based on this commitment, design and construction activities are now moving forward. Additional partners are being sought to provide funding to complete the second optical train of the binocular and to build facility instruments --- approximately 20 million dollars (1989). We have continued to use 1989 dollars as the cost basis to reassure our administrations that the scope of the project has not changed. A reassessment of the costs in current dollars suggests that the overall project can be accomplished for approximately 70 million dollars (1994). Just over 8 million dollars have been spent through the end of 1993.

The telescope configurations shown in Figures 1 and 2 provide a diffraction baseline of 22.8 meters and a collecting area equivalent to an 11.8 meter circular aperture. The telescope structure is now in the phase of final design drawings. The final design of the telescope is being done by a consortium of ADS Italia (Lecco) and European Industrial Engineering (Venezia). A contract for architectural services will be placed shortly to produce the final enclosure design. The first 8.4 meter honeycomb mirror is next in the queue of the Steward Observatory Mirror Lab with casting projected for Fall 1995. Polishing of this mirror should be completed in the Spring of 1998.

The decision to locate the telescope on Mt. Graham was reached at a relatively early stage. Mt. Graham is over 3200 meters in altitude and lies in the same Sonoran desert climatic zone as other Arizona sites such as Kitt Peak and Mt. Hopkins. Seeing measurements have shown that several sites within the 150 acre Emerald Peak astrophysical area deliver image quality comparable to the best recorded at any site. Measurements combined from a Polaris monitor and microthermal sensors indicate that the median site seeing should be 0.6 arcsec FWHM if the telescope is mounted above the local boundary layer at treetop level. In December 1993, with Forest Service approval, one acre of trees were cut on the LBT site on Emerald Peak 200 m East of the other two telescopes on Mt. Graham. Already built are the 10 m submillimeter telescope (SMT) and the 1.8 m Lennon telescope (VATT) --- both begin science operation in 1994.

2. Optics

2.1 Optical Requirements

The main LBT design drivers from the beginning of the project have been:

The error budget specifies that the telescope and its optics will produce images to match an r 0 = 45 cm atmosphere. Hill$^1$ reviews the error budget and detailed performance specifications. This specification produces an image which is 0.22 arcsec FWHM in the visible and produces a diffraction-limited image at wavelengths longer than 2 microns. As the design process moves along, it appears that we have selected an achievable goal.

The observational priorities which have guided the telescope design include:

A representative set of instruments which might be built shortly after first light include: an optical direct imager, an optical interferometric imager, a faint object optical spectrograph, a high resolution optical spectrograph, a near infrared camera, a medium resolution infrared spectrograph, a thermal infrared imager and spectrograph. This list is not intended to show any priority order, nor is it complete, but rather to illustrate the diversity of observational interests among the community of LBT astronomers. Provisions for adaptive optics feature prominently in the design of the telescope.

2.2 Primary Mirrors

From the outset, it was agreed to exploit the borosilicate honeycomb primary mirror technology because of its relatively simple thermal control and support characteristics and its prospective lower overall implementation costs. This technology has now been demonstrated at the 6.5 m size as described by Olbert et. al. 2. A great deal of our technical effort over the last four years has gone into understanding in detail the best practical way to support the 6.5 and 8.4 meter honeycombs. These blanks are stiffer than other types of mirrors, but they are still very susceptible to small force errors. Finite element analysis of the support of these mirrors is described by Parodi, Hill and Salinari3. We will be supporting the mirrors with active pneumatic force actuators. A brief description of the mirror support system is given by Gray et. al. 4. Some results from a similar system implemented on a 3.5 m mirror are discussed by Martin et. al. 5.

The relevant dimensions of the primary mirrors are:

The 8.4 meter mirrors have a focal ratio of 1.14. This allows the telescope structure to be very compact, and thus very stiff. The compact focal ratio also allows us to build a smaller enclosure. The fast focal ratio requires the mirrors to be polished to a very aspheric shape. The LBT primaries and secondaries will be among the most aspheric primaries and secondaries ever made. Burge et. al. 6 describe the null correctors used to test these mirrors. Anderson et. al. 7 summarize the polishing of a 3.5 m F/1.5 and two 3.5 m F/1.75 mirrors all to a surface accuracy of 16 -- 20 nm rms.

2.3. F/15 Infrared

LBT is not unique among large telescope projects in wanting to achieve good performance in the thermal infrared. The infrared optical design was recently changed to be an F/15 Gregorian focus with only 0.89 meter obscuration on the 8.4 meter primary. This is the smallest fractional obscuration of any large telescope. The change to Gregorian optics was principally motivated to allow us to gain maximum benefit from an adaptive secondary which can be tested independently from the primary. All of the wide field optical configuration is arranged to swing out of the beam when not in use to avoid contaminating the thermal infrared beam. The secondary is made sufficiently undersize to account for a 4 arcminute field-of-view, for ±0.7 mm of centration error and diffraction from the edge of the secondary at thermal infrared wavelengths. The undersized Gregorian secondary which gives an F/15 beam from an F/14.7 parent is 87 cm in diameter.

2.4. Interferometry

The binocular mounting of the two 8.4 meter LBT primaries gives us an interferometric baseline for diffraction of 22.8 meters. Because the two telescopes track on a common mount, we are able to bring a field-of-view several arcminutes in diameter to the combined focus. The experience of the MMT (6 mirrors) has taught us that phasing a large multiple mirror telescope is possible. The challenge for LBT has been preserving a low-vibration, low-emissivity, phased field while allowing (and not compromising) other observing modes in the telescope. A major activity in recent months has been designing the appropriate relay optics to bring together the two images at a combined focal plane. This exercise involves designing an instrument and a telescope simultaneously. We anticipate that a stable, phased field combined with developments in adaptive optics will yield significant observational benefits. Additional discussion of the interferometric optics is provided by Hill8 and by Byard and Bonaccini9.

2.5. Wide Field F/4

Inspired by the previous work of Epps, Bonaccini has designed a 3-element refractive corrector which provides excellent images over a 1.0 degree field-of-view. The F/4 corrected focal plane will be used mainly for multiobject spectroscopy, but the inner 30 arcminutes are optimized for imaging as well. F/4 provides a good match to fiber optics and to the size of CCD pixels. We have recently switched to the trapped focus configuration to increase the field-of-view while decreasing the focal ratio and avoiding conflict with the infrared focus over the hole in the primary. The wide field Cassegrain secondaries are 1.3 m in diameter.

3. Telescope Structure

3.1 Structural Requirements

The prime requirement on the structure is that of accommodating on the same mount the two primary mirrors, the baseline optical configurations and the corresponding focal stations described above. These focal stations include: a phased combined focus, two direct Gregorian foci, two bent Gregorian foci and two trapped wide field foci. The provisions for a possible future implementation of a quasi-coudè focus and two prime foci are included in the design.

A second requirement that had a large impact on the telescope design is that of providing the possibility of switching among all foci and configurations during the night in a short time. The solution that was recently adopted, based on swing arms to exchange secondary mirrors, brought a significant evolution of the telescope design.

A third important requirement is that of handling mirrors and cells to install them in the telescope. This affected the cell design and also the telescope geometry because of the need to provide access for large equipment either to remove and replace the mirrors or for removal and replacement of the coating. We have chosen to strip and aluminize the mirrors on-board the telescope since that appears more practical than removing the mirrors to a remote location.

Telescope locked rotor frequency (goal, both axes) > 8 Hz
Optical path difference due to vibrations (>8Hz) < 0.025 µ m
Maximum angular speed (both axes) 1.5 deg sec -1
Maximum angular acceleration 0.3deg sec-2
Short term tracking specification 0.03 arcsec rms
Whole sky pointing specification 0.3 arcsec rms
Wind speed for pointing and tracking specification 24 km hour-1
Maximum operating wind speed 80 km hour-1

Table 1: Telescope Performance Specifications.

Further specifications of more quantitative nature are reported in Table 1. The resonant frequency specification is of fundamental importance for the mechanical design of the structure and for the selection of the type of drives. Setting the locked rotor frequency at 8 Hz requires that the very large structure loaded with all the accessory parts has all vibration modes (that can be excited by the drives) above 8 Hz and that all the elements transmitting the motion are rigid enough not to degrade this performance below 8 Hz. The vibration specification is a reminder of the interferometric use of the telescope and therefore of the care that must be taken at all levels in the design to avoid excitation of local or global vibrations, while all the other specifications affect primarily the dimensioning of drives and encoders.

3.2. Function of the Structure

The present telescope design is a descendant of that proposed by Davison 10. The mount has undergone a great deal of optimization and tuning over the years; see for example Del~Vecchio et al.11. This design has shown to provide better global performance than a number of alternative structural concepts that have been considered. The 325 ton (including optics and instruments) elevation structure is supported by two large rolling sectors of 14 m diameter, which rotate on four radial hydraulic supports and are laterally constrained by two pairs of hydraulic supports acting on the edges of each sector. The 100 ton azimuth platform is a simple frame that connects the elevation supports to the four vertical azimuth supports acting on a rail with 14 m external diameter. The azimuth platform is radially constrained by a central bearing. The main mechanical advantages of this geometry are in the direct transmission of the loads from the mirror cells to the elevation platform to the azimuth rail and in the large diameters available for the drives and for the encoders. The primary mirror cells and the swing arms supporting secondary and tertiary mirrors are directly connected to the vertical rolling sectors that are the stiffest parts of the elevation structure. The motivation behind swing arm spiders supported from only one end is discussed in more detail by Del Vecchio, Miglietta and Davison 12. The swing arm spiders combined with the central concentration of mass in the elevation structure provide a 2 Hz increase in resonant frequency compared to the previous superstructure supporting the spiders from both ends. Compared with previous ``closed'' structures, the open elevation structure also reduces the cross section for the wind and the thermal mass of the telescope.

View Figure 6 here

The results of a number of finite element studies of the telescope structure are described by Del~Vecchio, Miglietta and Davison 12 . The simple summary is that we are able to achieve locked rotor resonant frequencies above 9 Hz. The pinion and gear drives, with four pinions directly driven by torque motors for both azimuth and elevation axis, were selected for the high rigidity, for reproducibility and for the simple implementation of the safety equipment. Because the drive acts at a 7 meter radius and has a simple and stiff transmission it reduces by only about one Hertz the resonant frequencies of a telescope structure with ~10 Hz lowest frequency. Commercially available strip encoders can provide the required resolution (0.01 arcsec) on 7 m radius.

3.3. Thermal Aspects

The whole surface of the telescope must track very closely the temperature of ambient air to minimize local seeing. This requirement dictated a number of measures in the present design of the telescope. We have adopted four basic strategies: thin structural sections, natural air circulation, forced heat removal using ambient air and emissivity control.

All structural elements, with the exception of the vertical rolling sectors and of most of the azimuth platform, have wall thickness of less than 10 mm if exposed to ambient air only on one face (20 mm if both faces are exposed). This maintains the thermal time constant shorter than about one hour even in very low wind conditions. The elevation sectors and the azimuth platform, that need substantially thicker elements, will be shielded from contact with ambient air to avoid thermal pollution. Rather than using simple thermal insulation, that would bring a large temperature difference between the massive and insulated parts and the rest of the structure, we adopted a ``stealth'' scheme based on the combination of insulation and forced ventilation. A lightweight perforated wall surrounds the massive parts of the telescope. Ambient air will be sucked through the perforated wall and will come in thermal contact with the structural parts of the telescope. The perforated surface will only be subject to a modest heat load (a few tens of W m -2) and thermalized by an air flow of typically 100 l s -1 m -2 . This results in a maximum temperature difference between air and the perforated skin of about 0.2 C °. The corresponding relatively large air flow, about 50 m 3 s -1 , is used to thermalize the massive parts of the structure, about 200 tons of steel. We expect these to track ambient temperature with a thermal time constant of about one hour and that a typical temperature difference of less than one degree could be maintained in most circumstances. The modest temperature difference between parts of the structure with different thermal time constant reduces the thermal deformations, that are difficult to model, and therefore increases pathlength stability and the pointing and tracking performance obtainable with open-loop modeling. The thermally contaminated air will be dumped outside away from the telescope. Other measures foreseen for the thermal control of the telescope are: removal of the heat that is produced locally by motors and electronics, pre-cooling of the hydraulic fluid of the telescope supports and reduction of the radiative losses of the structure by covering it with low emissivity aluminum tape. Similar precautions are taken for the enclosure as described by Salinari and Hill 13 .

3.4. Aluminization

In order to avoid the risk and expense of annually removing the primary mirrors from the telescope for recoating, we have elected to strip and recoat the LBT mirrors in their cells on the telescope. This requires us to make the primary mirror cells compatible with the rough vacuum behind the mirrors. An external bell jar is put in place to create the clean (1 microTorr) vacuum needed in front of the mirrors for evaporation of aluminum. Our coating strategy for obtaining maximum scientific results includes designing a facility with the capability to produce coatings with: high reflectivity, low emissivity and uniform thickness. To further enhance performance in the telescope we will implement a plan of regular cleaning with CO 2 snow, careful maintenance and protection from the weather. Our design studies in this area have not found any special problems associated with aluminizing the mirrors on the telescope or with handling the bell jar. Rather, the challenges come from applying a high reflectivity coating on such a large surface and from applying a uniform coating using a relatively compact bell jar. Atwood and Sabol14 describe the results of some experimental studies.

4. Summary

5. Acknowledgements

 Drawings of the telescope structure shown here were supplied by ADS Italia, of Lecco, Italy. We thank Warren Davison, Roger Angel and Walter Gallieni for many years of valuable discussions about the telescope design.

6. References

  1. Hill, J. M. 1990,
    ``Optical Design, Error Budget and Specifications for the Columbus Project Telescope'',
    S.P.I.E, 1236, pp. 86-107.

  2. Olbert, B., Angel, J. R. P., Hill, J. M. and Hinman, S. F. 1994,
    ``Casting 6.5 meter mirrors for the MMT Conversion and Magellan'',
    S.P.I.E., 2199 (These proceedings).

  3. Parodi G., Hill J. M. and Salinari P. 1992, ``Supporting the 8.4m honeycomb mirrors of Columbus'', Proceedings of the ESO Conference on Progress in Telescope and Instrumentation Technologies, ed. M.-H. Ulrich, (Garching:ESO), pp. 301-306.

  4. Gray, P. M., Hill, J. M., Davison, W. B., Callahan, S. P. and Williams, J. T. 1994, ``Support of large borosilicate honeycomb mirrors'', S.P.I.E., 2199 (These proceedings).

  5. Martin, H. M., Davison, W. B., DeRigne, S. T., Hill, J. M.,
    Hille, B. B. and Trebisky, T. J. 1994, ``Active supports and force optimization for a 3.5-m honeycomb sandwich mirror'',
    S.P.I.E., 2199 (These proceedings).

  6. Burge, J. H., Anderson, D. S., Ketelsen, D. A. and West S. C. 1994, ``Null test optics for the MMT and Magellan 6.5-m f/1.25 primary mirrors'', S.P.I.E., 2199 (These proceedings).

  7. Anderson, D., Burge, J., Ketelsen, D., Martin, B., West, S.,
    Poczulp, G., Richardson, J. and Wong, W. 1993,
    ``Fabrication and testing of the 3.5 m, f/1.75 WIYN primary mirror'',
    S.P.I.E., 1994, pp. 193-207.

  8. Hill, J. M. 1994,
    ``Strategy for interferometry with the Large Binocular Telescope'',
    S.P.I.E., 2200 (Companion proceedings).

  9. Byard, P. and Bonaccini, D. 1994, ``Optical design for interferometry with the Large Binocular Telescope'', S.P.I.E., 2200 (Companion proceedings).
  10. Davison, W. B. 1987,
    ``Structural innovations in the Columbus Project:
    an 11.3 meter optical telescope'',
    S.P.I.E., 748, pp. 31-37.

  11. Del Vecchio, C., Davison, W., Hill, J. M. and Gatti, R. 1992,
    ``Finite element analysis of the Columbus Telescope Project elevation
    structure'', Proceedings of the ESO Conference on Progress in
    Telescope and Instrumentation Technologies    , ed. M.-H. Ulrich,
    (Garching:ESO), pp. 79-82.

  12. Del Vecchio, C., Miglietta, L. and Davison, W. B. 1994,
    ``Mechanical Design of the Large Binocular Telescope'',
    S.P.I.E., 2199 (These proceedings).

  13. Salinari, P. and Hill, J. M. 1994,
    ``Enclosure of the Large Binocular Telescope'',
    S.P.I.E., 2199 (These proceedings).

  14. Atwood, B. and Sabol, B. A. 1992,
    ``Studies of some aspects of aluminizing large astronomical mirrors'',
    Proceedings of the ESO Conference on Progress in
    Telescope and Instrumentation Technologies    , ed. M.-H. Ulrich,
    (Garching:ESO), pp. 265-268.