In chapter one we mentioned that the development of large reflector antennas has been spurned strongly by the needs of radio astronomers. Interestingly, this is not entirely the case for the spectral range, called millimeter-wave radio astronomy, which we define as the frequency range from 30 - 300 GHz (wavelength 10 to 1 mm). The first reflectors, sufficiently accurate to operate at 3 mm wavelength, were the 4.9 m diameter antenna at the University of Texas (Tolbert et al.,1965) and the 4.6 m diameter dish of the Aerospace Corporation in El Segundo, California (Jacobs and King, 1965). Both instruments were not built primarily for radio astronomy, but were however quickly "taken over" by radio astronomers. Eugene Epstein became "corporate astronomer" at the Aerospace Corporation and made a career out of making millimeter wavelength observations near the beach of the Pacific Ocean. The Texas antenna was moved to the McDonald Observatory in 1967, where it operated until 2000. It has now been relocated to the 4600 m high site of the LMT in Mexico (see Sec.7.7). In 1964 the National Radio Astronomy Observatory (NRAO) proposed to develop a large millimeter telescope and operate it from a high and dry site to avoid the strong absorption from atmospheric water vapour at these wavelengths. Peter Mezger (1964) of NRAO summarized the prospects of astronomy at 3 mm wavelength in a report with the conclusion that a limited number of objects might be observable: "planets, compact HII-regions, a few quasars, perhaps recombination lines of ionised hydrogen". Today, this would not have resulted in funding, but the NSF provided 1 million dollars to NRAO to pursue the project. The result was the 36-ft telescope on Kitt Peak, which started operation in 1968. With the discovery of the spectral line of Carbon-Monoxide (CO) at 115 GHz with this telescope (Wilson et al,1970), millimeter wavelength astronomy established itself as a major branch of radio astronomy. The story of this highly productive instrument has been told vividly by Gordon (2005) in his book "Recollections of 'Tucson Operations'". Short descriptions of these and other mm-antennas are presented in a "Special Issue on Millimeter Wave Antennas and Propagation" of the IEEE Transactions on Antennas and Propagation, July 1970.
In his position as director at the Max-Planck-Institut für Radioastronomie (MPIfR), Peter Mezger concluded in 1972 that observing capabilities needed to be augmented with a large telescope for millimeter wavelengths. Preliminary studies led to a proposal for a reflector of 30 m diameter with a surface accuracy of 0.1 mm, which would marginally enable observations at the 20 spectral line of CO at 230 GHz, where the beamwidth would be about 10 arcsecs. A critical aspect of the proposal was the placement of the telescope at a site suitable for observations at 1 mm wavelength. At about the same time French radio astronomers were making plans for an interferometric array for mm-observations, while plans were being developed also in the United Kingdom for a dedicated telescope for wavelengths even shorter than 1 mm - the submillimeter range of the spectrum. In a collaborative effort, the three groups proposed the creation of a joint observatory for these telescopes on a suitable site. The proposal met with support from the respective governments and resulted in the end of 1975 in negotiations between the Max-Planck-Gesellschaft and the Centre National de Recherche Scientifique about the establishment of a joint mm-observa-tory, later named IRAM. The British proceeded on their own, which resulted in the JCMT on Hawaii in a collaboration with the Netherlands and later with Canada.
The MPIfR obtained main funding for the Millimeter Radio Telescope (MRT) from the Volkswagen Foundation and started serious design effort in the second half of 1975 with the same joint venture of companies, which built the Effelsberg telescope. At this time the author joined the MPIfR and became Project Manager of the MRT jointly with Ben Hooghoudt.
The specifications of the MRT formed a formidable challenge for the designers, who however had the Effelsberg telescope as a good point of departure. A straightforward downscaling of the 100-m antenna could be expected to show the required small surface error of not more than 0.1 mm rms. There were however serious complications with this. First, we needed to account for the severe weather conditions, which could be expected at the high mountain site, planned for the telescope. These could include icing storms and very high wind speeds. Second, we wanted to arrange for a spacious and easily accessible room for the receivers to ease their operation and maintenance and to allow a quick change of system depending on atmospheric conditions. This precluded prime focus operation and a Cassegrain room in a scaled Effelsberg antenna would be both difficult to access and too small. Thus it was clear that the highly symmetrical "umbrella" support of the Effelsberg telescope had to be abandoned. Nevertheless, a highly homologous reflector structure would be required to fulfill the surface specification.
The pointing specification of 1 arcsec also was well beyond what had been realised in antennas of similar size, and moreover, this accuracy had to be maintained in a wind speed of 10 m/s. Experience with other telescopes had taught us that there must be a match between surface accuracy and pointing for a telescope to be effective. Older instruments often showed a better surface than expected, but this could normally not be exploited because of lack of pointing accuracy and stability. Thus the designers were told that pointing and surface were of equal importance and should remain "matched" under all operational conditions. Next to the wind, we determined that variations and gradients of the temperature throughout the structure could easily cause deformations large enough to jeopardise the telescope performance. Controlling the influence of these thus became a major requirement for the design.
The requirements for the MRT originated in the experience with other telescopes, notably the NRAO 140-ft and 36-ft telescopes, as well as early operation of the Effelsberg telescope. For the first time this was all bundled into one comprehensive specification, more in terms of operational performance than of structural tolerances. Consequently, a very close collaboration between the industrial designers and the astronomers/engineers of the MPIfR was established from the early design phase throughout the commissioning of the instrument.
The partners in the industrial joint venture, Krupp and MAN, started with presenting each three conceptual designs. Of those, the one shown in the cross-section of Fig. 7.10 was chosen. It is a compact, "turnstile", alt-azimuth antenna, well suited for operation exposed to the hostile climate of high mountains. All drive and receiver systems are accommodated in a two-storey cabin. The entire structure, apart from the reflector surface, of course, is covered by heat-insulating panels, which themselves can be heated to avoid excessive deposit of ice during icing storms (Fig. 7.11). By 1975 advanced finite element analysis programs, like NASTRAN and STRUDL, were available along with sufficient computing power, so that complete structures
Fig. 7.10. Cross section of the MRT. A concrete pedestal (1) supports a 5 m diameter azimuth bearing (2). The 2-storey cabin (3) is placed between the elevation bearings. A yoke and cone-section (4), supported at the elevation bearings carries the reflector (5). Thermal insulation (6) covers the entire outside of the antenna; it is also present between reflector panels and their support structure.
like the reflector could be analysed. This made the detailed analysis of structures with limited symmetry possible. In the MRT the reflector backup structure (BUS) is composed of a space frame of 20 sections. It is a homologous structure and in order to exploit that, it must be supported by a mount structure which shows equal stiffness at the attachment points. This was realised by a box-like yoke structure, supported at the elevation bearings, with a cone shaped extension terminating in a round, flat plate of 14 m diameter. The BUS is attached to this plate at 40 points on the outer circumference. The entire structure will now behave in a homologous way, if the plate remains flat and round with varying elevation angle. (Brandt and Gatzlaff, 1981, Eschenauer et al., 1980). This structure does not reach a homologous behaviour as well as the Effelsberg design. To obtain the required small deformations a considerable amount of stiffness, hence steel, was put into the structure. Actually, to survive the specified survival loads, 200 km/h wind with 30 cm of ice on the structure, more steel was needed than required for stiffness. This is a general feature of exposed antennas. There was another reason not to concentrate solely on achieving maximum homology. As pointed out above, achieving excellent pointing stability under operational conditions was of equal importance. This led to the idea to incorporate the quadripod support of the subreflector into the overall structural design. Often the
quadripod has been supported by the mount, independent from the reflector. Here it was made part of the BUS structure and the deformations of the entire structure were optimised to obtain small pointing and reflector errors, both as function of elevation angle and under wind influence. Thus in the MRT the pointing errors caused by the bending of the reflector and the quadripod (see Sec. 5.5.1) compensate each other to a large degree. This is shown in Table 7.2, where the static pointing errors under gravity for the primary and secondary focus are given as function of elevation angle. The pointing jitter under wind of 12 m/s was calculated to be less than one arcsecs.
Table 7.2 Static pointing error due to gravity
Was this article helpful?