UBC Faculty Research and Publications

2.7-m liquid-mirror telescope. Hickson, Paul 2011

You don't seem to have a PDF reader installed, try download the pdf

Item Metadata


Hickson_SPIE_2199_922.pdf [ 918.81kB ]
JSON: 1.0107612.json
JSON-LD: 1.0107612+ld.json
RDF/XML (Pretty): 1.0107612.xml
RDF/JSON: 1.0107612+rdf.json
Turtle: 1.0107612+rdf-turtle.txt
N-Triples: 1.0107612+rdf-ntriples.txt

Full Text

A 2.7-rn Liquid-Mirror Telescope Paul Hickson, Gordon A. H. Walker University of British Columbia, Department of Geophysics and Astronomy 2219 Main Mall, Vancouver, British Columbia V6T 1Z4, Canada Ermanno F. Borra and Rémi Cabanac Université Laval, Département de Physique Québec, Québec G1K 7P4, Canada ABSTRACT An astronomical telescope employing a 2.7-meter diameter rotating liquid mercury mirror has recently begun oper- ation at a site near Vancouver. The telescope achieves seeing-limited performance, and can detect galaxies as faint as 21st magnitude. Equipped with a 2048 x 2048 pixel low-noise CCD detector, the telescope is now surveying a 20 arcmin wide strip of sky centered at +49°declination. The CCD is operated in TDI mode, providing continuous imaging with a resolution of O.6"/pixel and an integration time of 129 seconds. The primary scientific program of this instrument is to obtain spectral energy distributions of all objects in the survey area, by means of imaging through a series of 40 interference filters spanning the wavelength range 0.4 - 1.0 urn. These data will then be used to identify and estimate redshifts for of order x iO galaxies and x i03 quasars. 1. INTRODUCTION Over the past decade, research into the technology of rotating liquid mirrors (LMs)1'2'3 resulted in the development of mirrors several meters in diameter. Laboratory optical testing has demonstrated that they can achieve diffraction- limited performance2 . Because of this it is now feasible to construct large astronomical telescopes using liquid primary mirrors. The theoretical optical performance of liquid mirrors is well understood. Rotation, at angular frequency w,of a liquid in a uniform gravitational field g produces a paraboloid with focal length f = g/2w2. The rotation of. the earth and the nonuniformity of its gravitational field produce very minor and easily correctable effects on the image4. Because the liquid mirror must rotate about a vertical axis, liquid-mirror telescopes (LMTs) are presently restricted to observing an area of sky within about a degree of the zenith. In future, it may be possible to extend the observable area either by means of adaptive correctors5'6 or by introducing magnetic particles into the mercury and deforming the shape of the mirror by magnetic fields7. Over the past several years, a 2.7-meter LMT has been designed and built at the University of British Columbia ( UBC). Regular observations were begun February 1, 1994. This paper summarizes the main features of the telescope and provides a preliminary assessment of its performance. A more complete description of the telescope and its observing program will be published elsewhere8. 2. THE UBC/LAVAL 2.TM LMT In 1987 a collaborative project was initiated between UBC and Laval University to develop a large LMT and to use it for astronomical research. A new design, developed for large mirrors, was employed for the 2.7-meter mirror9. This design has formed the basis for subsequent large LMs at the University of Western Ontario, Lava!, and NASA. These mirrors employ a skin of Keviar surrounding a foam and aluminum core. This forms a lightweight, and very stiff, structure. The upper surface of the core of the mirror was machined to a parabolic shape, before applying 922 / SPIE Vol. 2199 Q-8194-1494-8/94/$6.OO Downloaded from SPIE Digital Library on 20 Sep 2011 to Terms of Use:  http://spiedl.org/terms the Keviar, and then covered with a spun-cast layer of polyester which supports 150 kg of mercury. The mirror is supported by an air bearing, and turned by a synchronous motor at an angular frequency of 1 radian/sec, giving a 5-meter focal length (Figure 1). After several hours exposure to air, the mercury develops a thin coating which both stabilizes the surface and virtually eliminates mercury evaporation. The final surface is a highly accurate paraboloid with a reflectivity of about 80%. The mercury layer is quite thin, typically 2 mm. The reason for this is both to reduce weight, and to provide viscous damping of gravity waves. Because surface tension tends to prevent mercury from wetting the surface, the layer is established by a process which involves rapidly changing the rotation speed until the surface of the mirror is completely covered. The mirror and air bearing are protected by a set of six rollers located 0.7 mm below the mirror rim. In normal operation these rollers do not touch the mirror. However, if the mercury surface breaks asymmetrically causing an unbalanced load, the rollers prevent the mirror from tipping, or transmitting high lateral torques to the air bearing. The rollers are located on composite steel-nylon posts which compensate for thermal expansion. These posts also support electromechanical brakes, which deploy in the event of a power failure. The mirror is cleaned periodically by stopping the rotation and scraping the surface of the mirror with a rubber hose to collect dust and other contaminents. Contaminated mercury is then collected by a vacuum system and the mirror is restarted. The entire process takes less than an hour, and involves the removal and occasional replacement of only SPIE Vol.2199/923 Figure 1. The 2.7-meter liquid mirror. The surface Figure 2. Schematic drawing of the liquid mirror consists of a 2mm-thick layer of mercury. telescope. motor Downloaded from SPIE Digital Library on 20 Sep 2011 to Terms of Use:  http://spiedl.org/terms a few ml of mercury. The telescope is equipped with a Loral 2048 x 2048 -pixel CCD camera housed in a UBC cryogenic dewar9. This detector is aligned with columns oriented east-west, and is clocked continuously at a rate of one row every 63 ms. Because of this, charge in the CCD is transferred at the same rate as the image motion produced by the Earth's rotation. This technique (TDI) produces sharp images of all objects with an integration time of 127 s, the time taken for an object to traverse the CCD. (A small distortion occurs due to the curvature of the star trails and the discrete motion of the charge, but these effects are negligible in our application'0.) The image scale on the CCD is 0.6 arcsec per pixel, which gives an image 21 arcmin wide. Off-axis aberrations are removed by a 5-element corrector, designed by Dr. C. Morbey of the Dominion Astrophysical Observatory, providing sub-arcsecond image quality over the entire 1/2 degree diameter field of the telescope. 10-cm diameter optical filters are inserted in the beam to define the observing band. We use a series of 40 interference filters. These cover the wavelength range 0.4 — 1.0 um with uniform logarithmic spacing. The filter bandwidths are1 log ii = 0.02 FWHM, which corresponds to a width of 23 nm at at wavelength of 0.5 um. The detector and corrector assembly is supported above the primary mirror by a tripod (Figure 2). A set of three screws, driven by a stepping motor, provides vertical positioning with a resolution of 1 .5 urn. Thermal expansion of the tripod will be compensated by focus motion based on readings of temperature sensors attached to the telescope. The telescope is housed in a 4.5-meter diameter cylindrical wood frame building 6.5 meters tall. The roof is fitted with a retraction mechanism which uncovers a central 2.9 meter diameter opening. Access to the prime focus is provided by a mezzannine and retractable platform. The floor of the observatory is covered with a plastic spill container, which prevents any accidental escape of mercury. The building is ventilated by a small fan, to equalize air temperatures inside and outside the building. Because of the experimental nature of this project, it was decided to locate the observatory near Vancouver in order to reduce costs and provide easy access from UBC. The chosen site is approximately 40 km south-east of the city, in a rural area with reasonably dark skies. The observatory is on the north slope of a gentle hill, about 150 m above sea level, and is sheltered from the wind by a tall stand of trees. Data from the CCD are recorded continuously throughout the night on 8mm helical-scan tape cartridges. Each record corresponds to one row of CCD data. At a data rate of about 32000 pixels per sec, this corresponds to over 2 GBytes of data per night. Because of the very large quantities of data that the telescope will produce, data analysis must be automated. Analysis will be done using algorithms developed by Hickson for the automatic detection and measurement of faint objects. By combining data taken on different nights through the various filters, spectrophotometry can be obtained for every detected object. In this way we plan to produce a catalog of an estimated x i0 objects. The spectrophotometric data will allow faint stars and galaxies to be distinguished on the basis of their differing spectra. Moreover, extensive computer simulations have shown that it should be possible to determine both the Hubble type and redshift for galaxies as faint as 20th magnitude using the multi-filter technique". 3. PERFORMANCE Initial optical tests with the telescope commenced in the fall of 1992, using a small commercial CCD video camera and no corrector. Star images were recorded with FWHM ranging from 1.5 - 2.0 arcsec. Engineering tests continued using an intensified CCD camera, and a small 3-element corrector. In January 1994, the controller and TDI data acquisition system for the large CCD camera were completed and installed on the telescope. An extended period of poor weather precluded any useful observations until the first week of March. A sample image is shown in Figure 3. The field is at high galactic latitude, and shows many faint stars and galaxies. No filter was used for this observation; the effective bandwidth was about 500 nm. As a result, the sky was very 924 ISPIE Vol. 2199 Downloaded from SPIE Digital Library on 20 Sep 2011 to Terms of Use:  http://spiedl.org/terms bright. Comparison with Palomar Observatory Sky Survey prints reveals that the limiting magnitude is R 21. The FWHM of star images on this frame is 2.8 arcsec. Trailed images obtained by momentarily stopping the CCD scan indicate that distortion in the small corrector (which was not designed for such a large CCD) contributes significantly to the FWHM. We expect that the images will be further improved with the installation of our new wide-field corrector in late March. TABLE 1 Telescope Specifications mirror diameter (m) 2.65 effective mirror area (m2) 5.18 focal length (m) 5.00 focal ratio 1.89 image scale (arcsec/mm) 41.25 pixel size (urn) 15 x 15 pixel size (arcsec) 0.619 detector size (pixels) 2048 x 2048 detector width (arcrnin) 21.1 detector quantum efficiency (0.55 um) 0.30 detector read noise (e ) 7 observatory latitude 49° 03'41" scan rate (arcsec/sec) 9.815 readout rate (pixels/sec) 32471 integration time (sec) 128.9 available sky area (sq deg) 83.0 typical seeing FWHM (arcsec) 2 estimated sky brightness (V mag/sq arcsec) 20 One can readily calculate the expected limiting magnitudes for various observing conditions, given some reasonable assumptions about the seeing and sky brightness. A summary of the telescope specifications is given in Table 1. Performance estimates are given in Table 2. Our initial results, while still quite limited, are very encouraging, and indicate that the telescope will achieve its expected performance. TABLE 2 Expected Performance (single night, 25 nm bandpass at 550 nm wavelength) area surveyed in 8 hr night (sq deg) 27.7 sky counts per pixel (e—) 1075 sky S/N per pixel 33 limiting magnitude (V mag, S/N = 3) 22.5 limiting surface brightness (V mag/sq arcsec, S/N = 3) 23.1 SP!E Vol. 2199 / 925 Downloaded from SPIE Digital Library on 20 Sep 2011 to Terms of Use:  http://spiedl.org/terms 4. ACKNOWLEDGEMENTS We are pleased to acknowledge financial support from the Natural Sciences and Engineering Research Council of Canada. 5. REFERENCES 1. E. F. Borra, R. Content, L. Girard, S. Szapiel, L. M. Tremblay, and E. Boily, "Liquid mirrors: optical shop tests and contributions to the technology," Asirophys. J., 393, 829-847 (1992). 2. E. F. Borra, R. Content, L. Girard, "Optical shop tests of a f/1.2 2.5-meter diameter liquid mirror," Asirophys. .1., 418, 943-946 (1993). 3. P. Hickson, B. K. Gibson, and D. W. Hogg, "Large astronomical liquid mirrors," Pub. Asir. Soc. Pacific, 105, 501-508 (1993). 4. B. K. Gibson, and P. Hickson, "Liquid mirror surface aberrations. I. Wavefront analysis," Asirophys. J., 391, 409-417 (1992). 5. E. F. Borra, "On the correction of the aberrations of a liquid mirror observing at large zenith distances," As1r. Asfrophys., 278, 665-668 (1993). 6. M. Wang, G. Moretto, E.F. Borra and G. Lemaitre, "A simple corrector design for liquid mirror telescopes observing at large zenith angles," Asir. Asirophys., in press (1994). 7. W. L. H. Shuter and L. A. Whitehead, "A wide sky coverage ferrofluid mercury telescope," Asrophys. J. LeU., in press (1994). 8. P. Hickson, E. F. Borra, G. A. H. Walker, B. K. Gibson, R. Cabanac and R. Content, "A 2.7-metre astronomical liquid mirror telescope," in preparation (1994). 9. B. Campbell, G. A. H. Walker, R. Johnson, T. Lester, S. Yang, and J. Auman, "Precision radial velocities and residual problems with Reticon arrays," Proc. Soc. Phot. Opt. Inst. Eng., 290, 215-218 (1981). 10. B. K. Gibson and P. Hickson, "Time-delay integration CCD read-out technique: image deformation," Mon. Not. R. Astr. Soc., 258, 543-551 (1992). 11. P. Hickson, B. K. Gibson and K. A. S. Callaghan, "Multi-narrowband imaging: a new technique for multi-object spectrophotometry," Mon. Not. Roy. Astr. Soc., in press (1994). 926 ISPIE Vol. 2199 Downloaded from SPIE Digital Library on 20 Sep 2011 to Terms of Use:  http://spiedl.org/terms Figure 3: A 2048 x 2048 pixel CCD image from the LMT. This is an image of a high-galactic latitude field centred at 11h34m+490i01l(epoch 2000). It was taken in white light and has an integration time of 129 sec. Many stars and galaxies are visible. The faintest objects have a red magnitude of 21. SPIEVo!. 2199/927 Downloaded from SPIE Digital Library on 20 Sep 2011 to Terms of Use:  http://spiedl.org/terms


Citation Scheme:


Usage Statistics

Country Views Downloads
Japan 3 0
City Views Downloads
Tokyo 3 0

{[{ mDataHeader[type] }]} {[{ month[type] }]} {[{ tData[type] }]}


Share to:


Related Items