During May, the SOAR 4.2-m telescope took two very important steps towards becoming a completed reality. These were the signing of contracts for the two major subsystems of the telescope: (1) the Telescope Mount System including the drives; and (2) the Active Optics System including polishing and providing the active support systems for the primary, secondary, and tertiary mirrors (M1, M2, M3). Simultaneously, the design of the enclosure and support building has moved essentially to completion, so construction on Cerro Pachon can go into high gear. The SOAR (Southern Astrophysical Research) Telescope is a joint project involving Brazil (31% of observing time), Michigan State (13%), North Carolina (16%), and NOAO (30%). Chile receives 10% of the observing time as host country. The goal is to construct and instrument a 4.2-m telescope offering the highest possible image quality over a tip-tilt corrected field of view about 7' in diameter. SOAR will be located on Cerro Pachon, 11 km from Cerro Tololo, which it will share with the Gemini South telescope. The construction cost will be $28 million. CTIO will operate SOAR for a twenty-year period as part of our contribution to the project. SOAR partners will be able to swap for time on the Blanco 4-m telescope with its wide field capabilities, so that SOAR can be specialized for programs requiring the highest image quality.
The project team is headed by Tom Sebring (Project Manager) and Gerald Cecil (Project Scientist). The approach is to have most of the design work done by the major contractors described below, with only a small in-house team whose main duty is to oversee the contracts. The lean, mean contract team consists of Victor Krabbendam (Project Engineer and active optics system), Dave Porter (mount and dome), Oliver Wiecha (electrical), German Schumacher (software), Gilberto Moretto (Optical Engineer), Eduardo Serrano (Chile Site Manager), Jeff Barr (Architect) and Christine Stone (Administrative Assistant). CTIO is very happy to have been able to second Schumacher to the project from our staff, since his return here for the SOAR operations phase will ensure continuity in the maintenance and continuing development of the telescope's software package.
The mount contract went to RSI Universal Antennas, in Richardson, Texas. This company has a long history of constructing radio and radar antennas, but also recently provided the mount for the Hobby-Eberly Telescope.
The general layout of the mount is shown in Figures 1 and 2. The telescope will be built with two Nasmyth foci and three Bent Cassegrain foci. This will support quick interchange between a variety of instruments. The large cylinders shown on Figure 1 on either side of the primary mirror area represent the volumes allocated to the two Nasmyth instrument packages. Each Nasmyth package may be subdivided between up to three permanently mounted instruments or can carry one large instrument that might be shared with Gemini. In Figure 2, the ear-like protuberances on either side of the telescope indicate the volume allocation for a Gemini instrument.
Figure 1 also shows two Bent Cassegrain instruments sticking out of the uphill side of the primary mirror area. A third such focus is located on the opposite side of the telescope.
The basic performance specifications for the mount are:
• Tracking jitter and drift <0.1" each
• Pointing accuracy and repeatability < 2.0"
• First eigenmode > 12 Hz
• Instrument payloads: 2800 kg at Nasmyth, 300 kg at Bent Cass.
The mount will use roller bearings at all locations, taking advantage of RSI's great depth of experience at building large radio antennas using this sort of drive.
The primary mirror substrate will be a meniscus, 4.25-m in diameter by 104 mm thick. This is about half the thickness of the Gemini 8-m mirrors, but the SOAR mirror actually will be stiffer than the Gemini mirrors because of its smaller diameter. An important advantage from such a thin mirror is the short thermal time constant, which will simplify the control of the mirror temperature and hence of mirror seeing. The substrate currently is being manufactured by Corning, from ULE low-expansion glass. Raytheon (Danbury, Connecticut), won the contract for polishing M1, M2 and M3 and for providing their active support systems. Their design (see Figure 3) will provide axial support for the primary mirror with 120 very stiff electro-mechanical push-pull actuators, providing high correctability of the mirror figure. In contrast to most recent support systems for meniscus mirrors, there will not be a complex lateral support system to support the mirror when the telescope is pointing at the horizon.
Rather, there will be just three tangent-bar locating links. This is because even at large zenith distances the axial support system will be able to compensate for gravitational flexure of the mirror to the very high accuracy required. This is expected to provide a very simple and easily maintainable mirror support system.
The secondary mirror (M2) will be ~ 0.6-m in diameter and will be mounted on a fully-controlled hexapod with five degrees of freedom to provide the usual sort of active optics correction for coma and focus.
All of the foci currently being implemented will have tip-tilt correction, provided by tip-tilt control of the M3 flat. This flat also will rotate to select between the various Nasmyth and Bent-Cass ports. The tip-tilt loop will be closed in most cases by on-instrument wavefront sensors, although some ports will also offer facility tip-tilt sensors. The low-bandpass active optics corrections will be calibrated using a wavefront sensor permanently mounted at one of the instrument ports. At least initially these slow corrections will then be supplied from lookup tables during actual observing.
SOAR will be an f/16 telescope, to maintain compatibility with Gemini so that future instruments can be shared. Although the optical performance is specified in terms of a structure function, in general terms the image degradation by the telescope (including dome seeing and tracking errors) will be < 0.16". SOAR's instruments will be matched to expected tip-tilt-corrected best image sizes of 0.24". It is expected that the median image size will be about two times better than is achieved by the Blanco telescope. The basic goal is to obtain essentially the same image quality as Gemini, but at half the wavelength where Gemini gets its best images, and to equip the telescope with instruments optimized for that capability.
SOAR is also pioneering new territory with its software systems. It will use PC's running a distributed PCI architecture, built around the Labview/Bridgeview software languages offered by National Instruments. The underlying operating systems probably will be LINUX for the TCS, and Windows NT for the instruments.
The SOAR partners are banding together to provide an ambitious suite of instruments for the initial instrument package. These are:
High-throughput UV Spectrometer.. This will be built by UNC and Brazil, under the leadership of Chris Clemens (UNC). It will employ volume-phase holographic gratings to get resolving powers R = 5-18,000. It will offer high UV throughput and also multi-slit spectroscopy over a 3'-5' FOV.
IFU-fed, Bench-mounted Optical Spectrometer. This is being built by Brazil under the leadership of Jacques Lépine. The IFU (Integral Field Unit) will use 1500 fibers behind a lenslet array to completely cover a field which can vary between 50"-250"2, according to the zoom setting of the foreoptics. The bench spectrograph will have a variety of modes with resolving power up to R ~ 30,000. The instrument will be used both for two-dimensional spectroscopy of extended objects, and for point sources while maintaining full spectral resolution and throughput under all seeing conditions.
Optical CCD Imager. This instrument will use a 4K × 4K mini-mosaic behind a focal reducer and atmospheric dispersion corrector producing an f/9 output beam. This will give ~ 0.08" sampling over a 5' field. The optical imager will have high UV throughput, with sol-gel coatings and UV-transmitting elements in the ADC. It is being built at CTIO under the leadership of Alistair Walker.
Infrared Imager. A 2K2 Near-IR Hg-Cd-Te imager is being built by Ed Loh at MSU. This instrument will be easily upgradable to a 4K × 4K mosaic, which with the 0.08" pixel size again will sample essentially the full isokinetic patch at the full angular resolution of the telescope. A Lyot Tunable Filter will be available, and there probably will be additional provisions for grism spectroscopy.
IR Spectrographs. Current plans are for NOAO to provide a pair of existing IR spectrographs as interim measures until a new instrument can be built. To support low-medium resolution spectroscopy of point sources, the CTIO IRS will be upgraded with a 1K × 1K HgCdTe detector. For higher resolution work, we expect to offer the R = 100,000 Phoenix spectrograph which was built in Tucson and currently is in use on Kitt Peak. It is possible that Phoenix will be shared between SOAR and Gemini-South.
The enclosure design is sketched in Figure 4. In contrast to many recent new telescopes, SOAR will use a forced ventilation system to control dome seeing. Air will be sucked in through the observing slit (which will be a small aperture defined by a shutter and a semi-porous windscreen, and which follows the telescope motion), and then will be exhausted by a ring of large fans in the non-rotating part of the enclosure. The vents for two of these fans can be seen in Figure 4. Similarly, the top surface of the primary mirror will be flushed by an active system rather than by natural ventilation. The rotating dome will be built in Brazil, and will either be a geodesic design or made of free-standing interlocking panels which do not require a supporting frame.
The construction of the fixed building will be contracted to the new AURA Observatory Support Services (formerly the CTIO Operations division), who will then subcontract out the work as needed. This approach will take maximum advantage of CTIO's long history and special knowledge of operating in Chile.
The attached support building will be kept very simple, with just a control room, one instrument workroom, a computer room, and a high-bay receiving area. Through an arrangement negotiated with Gemini, SOAR will be able to utilize space and facilities at Gemini-South telescope, 400m away. In particular, the SOAR primary mirror will be recoated in Gemini's facility.
As was noted at the start of this article, letting the major contracts for the telescope was a very important milestone and opens the way for a rapid construction schedule. The mount system and active optics systems will be tested separately in the respective factories and then shipped to Chile in late 2001 and the first quarter of 2002. First light is scheduled for mid-2002, and operational hand over will be in early 2003.
Jack Baldwin (jbaldwin@noao.edu),
for the SOAR Consortium