Документ взят из кэша поисковой машины. Адрес оригинального документа : http://www.apo.nmsu.edu/Telescopes/eng.papers/performance/performance1988.html
Дата изменения: Thu Nov 24 03:41:07 1994
Дата индексирования: Sun Apr 10 04:18:47 2016
Кодировка:
Performance of the 3.5m Telescope

The Performance of the Apache Point Observatory 3.5m Telescope

Edward J. Mannery, Walter Siegmund and Charles Hull

Abstract

The Astrophysical Research Consortium 3.5 m telescope facility on Apache Point (2800 m above sea level) near National Solar Observatory in southern New Mexico is nearing completion. The telescope mount has been installed and testing and fabrication of remaining subassemblies are underway. The f/1.75 lightweight honeycomb primary mirror was cast April, 1988 by the Steward Observatory Mirror Laboratory and is currently being figured.

The 3.5 meter optical telescope is an altitude over azimuth mechanical structure with Ritchey-Chretien optics. The lightweight (1800 kg) mirror leads to a mount weighing only 41000 kg; readily available rolling element bearings are used to achieve the necessary performance at low cost and without the heat dissipation of externally pressurized types. Drive torques are applied by DC servo driven capstans. These are coupled by friction to large diameter drive disks on each axis. No gears are used. Position feedback comes from low cost incremental encoders, also capstan coupled.

We have recently completed a series of measurements of the telescope mount. These measurements show that the telescope is very stiff; the lowest natural frequencies are about 7.2 Hz. Initial tracking performance is good and the mount shows high resistance to wind induced vibration. Our experience during acceptance testing suggests that routine power spectral analysis of drive motor torque and other parameters could be an important tool in the early detection of failures.

Introduction

Apache Point Observatory is a new 3.5 meter telescope facility located at an elevation of 2800 meters above sea level, in the desert of southwestern United States 18 km SE of Alamogordo, New Mexico. It is owned and operated by Astrophysical Research Consortium (ARC). Member universities are University of Chicago, University of Washington, New Mexico State University, Princeton University and Washington State University.

The telescope is intended to be a general purpose imaging and spectrographic telescope. To exploit the best seeing expected at the site, the design goal is to produce images better than 0."4 full width at half maximum under these condition. Certain types of scientific programs and modes of operation are difficult to accommodate at the national facilities. The telescope and operation environment are expected to be particularly well suited for many of these programs, for example, synoptic observations, surveys, remote observations, and flexible response (1) (2). Flexible response is needed to take advantage of transient events such as supernovae and unusual atmospheric conditions such as periods of exceptionally good image quality.

The telescope project design phase started in 1983. Major contracts were awarded in 1986 and early 1987. The telescope and buildings were essentially completed in November 1987 and are awaiting the installation of the primary mirror. The primary was cast in April, 1988 at the University of Arizona and is currently being figured. It is expected to be installed early in 1990. The project cost including instruments and staff salaries but not the primary mirror blank was $10M.

The f/1.75 primary mirror is a 3.5 m lightweight borosilicate blank with a mass of only 1800 kg (3). This is the same as a meniscus blank with a thickness of 9 cm. The primary was cast by the Steward Observatory Mirror Laboratory at the University of Arizona (details are given in an article by Angel in these proceedings).

The secondary and tertiary mirrors are lightweight borosilicate hot gas fusion blanks and were fabricated by Hextek, Inc., Tucson, Arizona. They are 780 mm in diameter and have masses of 40 kg each.

The telescope enclosure rotates with the telescope in azimuth on a track located 5.5 m above the ground. The telescope chamber co-rotating floor is at 8.5 m and gives excellent access to instruments mounted at the Nasmyth foci, 1.3 m above this floor. Openings in the rear corners of the enclosure improve air flow around the telescope (4) . The upper enclosure frame consists of weight efficient I-beams for minimum thermal mass.

An enclosed corridor, which is used as an air exhaust duct, connects the telescope enclosure and the operations building which contains the computers and electronics shop. This corridor serves as a pathway for people and equipment; this is particularly useful during the winter when up to 1 m of snow can accumulate. Cables connecting the telescope and instruments to computers in the operations building are placed in a cable tray located on one wall of this corridor.

The Telescope

The telescope mount is an altitude over azimuth design (5) . The total moving mass is 41000 kg. A preloaded pair of spherical roller bearings are used on each side of the altitude structure. The main load carrying azimuth bearing is a spherical roller bearing. This bearing is located at the apex of large steel inverted cone. At the upper end of the cone is a large diameter steel disk which serves as a bearing, drive and encoding surface. Four guide rollers, two of which are driven, contact this disk and define the azimuth rotation axis together with the bearing at the cone bottom.

To reduce wind torque on the telescope, the main telescope secondary truss tubes diameters were minimized and are only 100 mm in diameter. Even at this size, they contribute more drag than the secondary assembly. However, further reduction of truss diameters would result in unacceptable lateral stiffness. The square secondary frame is placed in compression by the secondary vanes. Vane tension is set so that the rotational mode of the secondary about the optical axis is above 10 Hz.

The mirror cell and the tube center section is a one piece weldment to reduce weight and cost. The mirror is removed from the telescope by an overhead bridge crane mounted near the ceiling of the telescope chamber.

The telescope is driven by rollers which transmit torque via friction. No gears are used. Each axis is very closely coupled to a large diameter circular disk. The altitude disk is only a portion of a full disk. Each disk is driven by a 10 cm diameter roller. This gives a reduction of about 35:1. The drive roller is part of a three stage roller speed reducer. The overall ratio from the motor to the telescope is about 1200:1.

Sony Magnesensors give absolute encoding to about 0.1 arc seconds at 15 degree intervals. A roller attached to an incremental optical encoder is driven by friction by each large disk. The encoders are made by Heidenhain, Inc. and give the mount a resolution of 0.01 arc seconds.

Dynamic Performance

To measure the natural frequencies of the telescope structure and drive system, the telescope was excited by a mechanical impulse. It was struck repeatedly at roughly random intervals of 0.5 to 2 seconds with a 3 kg wooden block. The primary sensor was a linear velocity transducer with a resolution of better than 1 micron/second made by Schaevitz, Inc. This device consists of a permanent magnet inside a pickup coil. The transducer was mounted between the telescope and the observing chamber floor. It was sampled at 100 Hz. Samples were 20 to 40 seconds long. These data were transformed to power spectra using the fast Fourier transform algorithm. Data were also obtained from the axis increment encoders.

For the azimuth structure response, the impulse was applied to one fork and directed to maximize the impulse torque about the azimuth axis. The azimuth motor shafts were prevented from rotating. The locked rotor resonance frequency was 7.8 Hz. For the altitude structure response, the impulse was applied to the edge of the steel box structure surrounding the primary mirror to maximize the impulse torque about the altitude axis. The altitude motor shaft was prevented from rotating. The locked rotor resonance frequency was 11.7 Hz.

Other frequencies of interest include the azimuth overturning frequency due to the compliance of the azimuth bearings at 7.2 Hz. The peak at 5.5 Hz has tentatively been identified as the pier rocking mode.

Separately, we have investigated the dynamic behavior of the secondary truss structure. The measured natural frequencies are all above 14 Hz. These high natural frequencies imply that the telescope is very stiff. For example, typical locked rotor resonance frequencies of all but the most recent large telescopes are below 5 Hz. High values of stiffness in the telescope and drive gives good resistance to wind induced vibration. In addition, high stiffness permits high servo loop gains and low tracking error.

Conclusions

The proposed Columbus, Magellan and National Optical Astronomy Observatory telescopes have large diameter drive, encoding and bearing surfaces coupled to the optics via very direct load paths. The ARC telescope shares many of these design features; the resulting excellent dynamic performance and low mass helps to confirm the promise of this new approach.

Our experience with dynamic power spectra suggests that this is a powerful tool for monitoring telescope performance and for early detection of failures. During acceptance testing at the factory, we discovered a strong peak in the power spectra of altitude motor torque while tracking at a constant rate. The identification of the peak with the motor shaft rotation frequency localized the fault in the motor or speed reducer, and not in the azimuth bearings. Several months later, the speed reducer failed and the component at fault was identified.

References

1. Balick, B., Loewenstein, R., Siegmund, W., and York, D.: 1988, Remote Use of the Apache Point 3.5-m Telescope, Instrumentation for Ground-Based Optical Astronomy: Present and Future, ed. L.B. Robinson, Springer-Verlag, New York, 1988
2. R. Owen, W. Siegmund and C. Hull, The Control System for the Apache Point 3.5-Meter Telescope, Instrumentation for Ground-Based Optical Astronomy, ed. L. B. Robinson, pp. 686-690, Springer-Verlag, New York, 1988
3. Siegmund, W. A., Mannery, E. J., Radochia, J., and Gillett, P. E., Design of the Apache Point Observatory 3.5 m Telescope II. Deformation analysis of the primary mirror, Proceedings of the SPIE, vol. 628, ed. L. D. Barr, p. 377, 1986
4. Siegmund, W. A. and Comfort, C., Design of the Apache Point Observatory 3.5 m Telescope I. Thermal control of the telescope enclosure, Proceedings of the SPIE, vol. 628, ed. L. D. Barr, p. 369, 1986
5. Mannery, E. J., Siegmund, W. A., and Balick, B., Design of the Apache Point Observatory 3.5-Meter Telescope: IV. optics support and azimuth structures, Proceedings of the SPIE, vol. 628, ed. L. D. Barr, pp. 397-402, 1986

Figure Captions

Fig. 1. The APO 3.5 m telescope. Visible above the floor are the forks and altitude structure (white). The main telescope secondary truss tubes are only 100 mm in diameter to minimize wind loading. The secondary counterweight will be replaced by the real secondary assembly late in 1989. The mirror cell and the tube center section is one piece to reduce weight and cost. Drive torque is transmitted to the altitude structure via the large diameter drive disk mount on the underside of the mirror cell. Openings in the rear corners of the enclosure improve air flow around the telescope.

Fig. 2. The azimuth axis power spectrum: The azimuth rocking vibrational mode due to lateral compliance of the azimuth bearings is at 7.2 Hz. The azimuth locked rotor resonance frequency due to the compliance of the azimuth drive is at 7.8 Hz. The other features seem to be vibration modes of the steel weldments except for the peak at 3 Hz which is the rocking mode of the building (the transducer was connected between the telescope and the building).

Fig. 3. The altitude axis power spectrum: The main feature is the altitude locked rotor resonance frequency due to the compliance of the altitude drive at 11.7 Hz. The other features seem to be vibration modes of the steel weldments.