The Multiple Mirror Telescope (MMT) incorporates many new technological features because its designers wanted to construct as large and as good as possible a telescope at as low as possible an expense. They did succeed at this and as a result the MMT may be viewed as the first of the new generation of Advanced Technology Telescopes. A substantial effort is given by the MMT Observatory staff and by the staff of the two sponsoring institutions (Smithsonian Astrophysical Observatory and University of Arizona Observatories) to optimize the performance of the lurr especially in those areas where it serves as a test bed or prototype of the telescopes of the future. Notable features in this respect are (i) the MMT altitude-over-azimuth. mount, (ii) the corotating MMT building, (iii) the light weight. honeycomb primary mirrors, (iv) the multiple mirror optics configuration, (v) the coalignment system for these mirrors, (vi) the seeing associated with the unconventional telescope and building and (vii) the capability to do interferometry using two or more of the MMT telescopes. This introductory paper will summarize the properties and performance of the MMT in these and other respects. It will set the stage for the in-depth description by the other papers in this sequence.

There are several innovative and unusual features incorporated into the MMT alt-azimuth mount and optical support structure (OSS). Recent work has, and still is being done, to assess and optimize these features. Measurements and analysis of the azimuth drive system have allowed us to increase the azimuth resonant frequency from 2.2 Hz to 3.0 Hz by just repositioning one motor. By adding two additional drives we can increase the frequency again to about 4.1. Friction in the azimuth ball bearing and drives were measured and the data is very surprising. The intricate truss structure of the OSS has several variable area members (tuning members) that can be adjusted to partially compensate for gravitational flexure. The adjustment of these useful members has revealed a flaw of conservatism in the structural model. These adjustments coupled with thermal radiation shielding have allowed the six telescopes to be aligned passively to 20 arc seconds. Although details will vary, the knowledge of the success and limitations of these unusual features might prove very useful to advanced technology optical telescope designers and builders.

Image quality rivals telescope aperture in importance for astronomical observing. For many astronomical experiments the telescope diameter to image diameter ratio defines the merit of the telescope system, so that decreasing the image size by, say, 10% is as important as increasing the telescope diameter by 10%. Since the former can be done often at a fraction of the cost of the latter we have made considerable effort to improve the MMT seeing. By using largely passive means for controlling the thermal environment of the MMT we have eliminated most of the seeing effects originating in and around the telescope and building. On excellent nights the effects of external seeing associated with the atmosphere above the site are undetectable so that on those nights the telescope optics defines the 0.5 arc sec FWHM image size.

Operation of the MMT is heavily dependent on the use of computers for a variety of tasks. An altitude azimuth mount requires the ongoing conversion of celestial coordinates into local terrestrial coordinates. A sixteen bit minicomputer performs the conversion, compares the resulting commanded position with the current position readout by a digital encoder and generates an error signal to servo the telescope drives. The telescope co-alignment system (TCS) uses a computer to analyze a digitized television frame and control the tilt and focus of each of the six telescope secondary mirrors. All of the data at the MMT is collected by digitizing signals from photoelectric detectors in the astronomical instrumentation. (The MMT has no photographic darkroom facilities.) A computer with several peripherals to aid in display and analysis of data is dedicated to the data collection task. Our approach is to use three independent computers which can communicate over digital links. Each computer is dedicated to a specific, clearly identified task: tracking, coaligning, and data collection. It is clear that computers are an essential component of the MMT. The MMT has even been described as a large peripheral device Software for the systems is written in a sophisticated dialect of FORTH which allows multitasking.

The Multiple Mirror Telescope (MMT) has an absolute pointing accuracy of about one arc second RMS, which is considerably better than that of other telescopes. This paper discusses the characteristics of the telescope and pointing control system which may be of interest to telescope designers. Examination of known problems indicates that further improvements are possible. The MMT is exposed to large thermal radiation and wind loads because of the exposure allowed by the compact enclosure design. A theoretical analysis of the mount tracking servo system has been used to optimize the resistance to wind buffeting. The combination of mechanical and electronic servo system improvements and small, fixed wind baffles covering the yoke arms has obviated the need for a conventional wind-screen.

A crucial feature of large telescopes using multiple objectives is the ability to align the optics so that the different star images coincide to within a fraction of the image size. Initially this was intended to be done in the MMT with the "laser active optics" system in which the six telescopes were coaligned by means of artificial laser-star images generated by a seventh guide-alignment telescope. That system was found to be less than satisfactory mostly because of internal seeing and scattered light effects. Instead we have constructed the Telescope Coalignment System (or TCS) in which the telescopes are co-aligned on stellar images. The TCS is an autoguider which works simultaneously on six telescopes and which therefore automatically coaligns the telescopes and tracks them on a field star. When a field star is not present in the small 4 arc minute field of view of the MMT the TCS system coaligns the MMT on a nearby star and then tracks and maintains co-alignment of the six telescopes under computer control using separate flexure corrections for each one.

Observations of image size and motion have been made at the MMT and 2.3m telescopes. At the 2.3m, simultaneous observations were made of image size and position at 0.8μ and various IR wavelengths. The MMT observations were all at 0.6μ, and were used to measure the isoplanicity of the atmosphere for image motion, the time development of image position, and the telescope to telescope correlation of image motion. At both telescopes the image size was larger than expected for the image motion. The MMT results showed that although there is a finite outer scale of turbulence for the atmosphere, it did not appreciably reduce image motion. Therefore the excess image size is caused by excess small scale turbulence (presumably inside the dome), together with various telescope imperfections. It is inferred that sub-arc second seeing is frequently available at both Mt. Hopkins and Kitt Peak, but that the image is substantially deteriorated in the observations. While correction for image motion is a necessary nart of a program to shrink telescopic images, there are other high priority cures also needed.

Large gains in spatial resolution have been achieved at 0.5 and 5.0 µm by operating the MMT as a two-element interferometer with a maximum baseline of 6.9 m. Measurements of the resulting Michelson fringes at 5.0 µm have determined pathlength errors within the MMT and characterized pathlength stability versus elevation angle, temperature, and perturbations of the optical elements. At 0.5 μm, a "coherent beam-combiner" has successfully reconstructed a phased entrance pupil achieving a field >5 arcsec in diameter with 15 milli-arcsec resolution. Scientific measurements have fully resolved the 0.024 arcsec diameter disk α Tau (Aldebaran); resolved the 0.1 arcsec binary, β Tau; and provided new measurements of the spectroscopic binary, α Aur (Capella). Instrument designs employing only two additional reflections in each beam have been developed to phase the entire MMT for simultaneous applications at optical and infrared wavelengths. When configured as a phased array of four or six elements, the MMT responds to the full range of spatial frequencies present in a filled 6.9 m aperture with only a small degree of redundancy. A fully phased MMT possesses significant advantages for low background infrared photometry.

The astronomical instruments currently in use at the MT are briefly described and highlights of some astronomical results obtained with them are summarized.

The Multiple Mirror Telescope (MMT) has been used for the first time to make astronomical observations as a coherent phased array at sub-millimeter wavelengths. The success of the initial experiment indicates that the precipitable water vapor over Mt. Hopkins is sufficiently low to permit routine operation at 870 μm wavelength and frequent operation at 450 μm and 350 μm wavelength. The six telescopes of the MMT were initially phased to an accuracy of λ/30 at λ = 870 μm and remained phased to that accuracy over a four day period and over a wide range of elevation angle. The measured beam pattern of the telescope matches diffraction theory, and the width of the main lobe is 26 arc seconds. A coherent super-heterodyne receiver and a novel beamcombiner were used to make spectral line maps of the J = 3-2 line of CO. The combination of large collecting area, high angular resolution, and good atmospheric transmission makes the MMT superior to other existing telescopes for both continuum and spectral line astronomy at sub-millimeter wavelengths.

With the benefit of experience gained from operation of the MMT, we are defining the characteristics of a much larger instrument. Scaling the MMT to 15m divides into two main aspects, one directly related to how one performs scientific observations, the other related to how one makes and keeps a telescope operating near its theoretical limits. In the observational area we discuss phasing for both interferometry and long wavelength imaging, matching images to solid state detectors, and multiple fiber and single object spectroscopy. In the more direct telescope area we discuss the mechanical characteristics of the scaled up structure, optimizing the diffraction pattern, optical and mechanical implementation of different foci, matching atmospheric demands on image quality, thermal control of the mirrors and structure, and mirror coating losses.

The ESO NTT project is still in the conceptual phase but it is hoped early in 1982 to enter the engineering phase. With an aperture of 3.5 m it is seen also as a basic development aspect of our Very Large Telescope (VLT) project. The basic characteristics of the NTT are briefly reviewed with particular reference to the nature of the primary and its support system, which is strongly influenced by the active optics approach. Some relevant results of image analysis of existing conventional telescopes as well as some dome seeing and imaging monitoring results at the present ESO 3.6 m telescope are given. Brief comments are made on the concept of the ESO VLT.

For over two years the McDonald Observatory of The University of Texas at Austin has been actively designing a very large optical telescope. The design is largely constrained by the two requirements of completion by the late 1980's and cost within the realm of private philanthropy. Within these constraints, we have concluded that the largest possible telescope is one which incorporates the following properties: (1) a monolithic primary mirror of diameter near 7.6 meters, (2) an ultra-thin, meniscus, fused silica primary of thickness 10-15 centimeters or a "hondycomb" borosilicate primary of near classical thickness, (3) a fast (f/2) Ritchey-Chretien primary relayed to two f/13.5 Ilasmyth foci, (4) an alt-azimuth mount, (5) an MMT-style building instead of a classical dome, and (6) a sepa-rate building to house the aluminizing chamber.

The University of California is planning to build a segmented, ten-meter effective diameter optical and infrared telescope for use by its astronomers. The anticipated site for the telescope is Mauna Kea, Hawaii. The details of the design and the current activities on prototypes as well as a technical demonstration will be described. The design employs 36 actively controlled hexagonal mirror segments that form a mosaic primary mirror. As the primary will be parabolic, the segments will be off-axis sections of a paraboloid, each segment to be 7.5 cm thick and 1.8 m in diameter. The segments will be polished to their desired shape by stressed mirror polishing, a technique for making non-axisymmetric surfaces. The active control system employs displacement sensors at the edges of the segments that allow the positions of all the mirror segments to be determined; the segments' positions are adjusted by three displacement actuators per segment. The basic characteristics of the control system and its critical components will be described. The overall optical design of the telescope will be a Pitchey-Cretien f/1.75 - f/15 system. The goals and the design of the surrounding building and dome will be described. The dome is extremely compact, and rotates on a stationary building in the manner of conventional telescope domes.

We have considered the performance of the University of California Ten Meter Telescope (TMT) in the thermal infrared. At around 10 μm the thermal background seen by a detector in the focal plane is dominated by emission from the telescope itself since the atmospheric emission is often very low. Several aspects of telescope design are of crucial importance for optimum performance to be achieved in the infrared, the main goals being to minimize the background radiation and to keep fluctuations in that background to an acceptably low level. We have considered the effects of telescope design parameters in two classes: 1. The usual set of mechanical and optical constraints on design that are encountered in optimizing the performance of a conventional monolithic IR telescope. These include minimization of background radiation from telescope mirrors and support structure, chopping secondary design, etc. 2. Potential additional constraints arising from the segmented design of the primary mirror. These include the effects of the "cracks" between segments and consideration of the active control mechanism. We presently see no problems in the telescope design which will adversely affect IR performance and we consider that the actual performance of the TMT will be determined solely by the quality of its instrumentation and the properties of the atmosphere.

A number of designs for future astronomical telescopes call for large primary mirrors that are mosaics of smaller mirrors. We describe here a study of some characteristics of the images expected from a telescope with a primary mirror composed of 36 hexagonal segments. Various effects caused by the segmentation and imperfections in the segment fabrication and control have been analyzed using physical optics. The diffraction-limited image distribution from the segmentation geometry of the primary is derived, and the diffraction spikes are shown to be similar to those caused by secondary support struts in existing telescopes. A general relation between surface quality and image quality is given, and the implications for surface quality tolerances are discussed. The optical effects of segment phasing errors are derived, showing that at least for visible light observations the phasing is unimportant. For observations at 10 μm near diffraction-limited perfor-mance can be achieved with a 10 meter aperture requiring that the segments be phased correctly at this wavelength.

Very good imaging performance can be achieved in telescopes with fast parabolic primaries, with the aid of two-mirror correctors. We give details of a Paul-Baker type corrector for an f/l primary, relaying to an f/2 final focus. Images with 80% of the energy into 0.2 arc second diameter, or better, are obtained over a field of 10 diameter. The mechanical structure and enclosure of a large telescope built with these fast optics would be considerably smaller and less expensive than those for conventional optics. We explore an MMT design in which six off-axis primaries in a circle together form a Parabolic f/l surface. Other applications are in ultraviolet astronomy, where the high resolution with no refractive elements to give chromatic aberration is an advantage, and in infrared astronomy, where an off-axis part of the f/l parent gives a wide field telescope of very low emissivity. We also consider the use of the correctors for telescopes in which the aperture is a long thin rectangle, of fast focal ratio in its long dimension. This configuration has special advantages for diffraction limited imaging and interferometrv. Such telescopes can be as efficient for spectroscopy and photometry as circular mirrors of the same area, but have much higher angular resolution, and only weak side lobes. For space use the parts of such a deployable telescope would take advantage of the shape of the shuttle payload bay. On the ground we show that the rectangular array can provide a convenient alt-azimuth telescope when it is fed by a long thin flat mirror whose long axis can be rotated. We consider configurations appropriate for an optical telescone, a sub-mm telescope and a transit telescope. The manufacturing techniques for these fast, and sometimes off-axis, mirrors needs to be proved before these designs can become contenders for the construction of actual telescopes.

An important part of a balanced future program in space astronomy is a fast, wide-angle telescope (a Space Schmidt) capable of imaging in the ground-inaccessible ultraviolet wave-length range (1100-3000 Å) and capable of reaching substantially fainter background light levels in the visible and near-infrared than are groundbased telescopes (limited by airglow sky background). In 1978, NASA appointed a working group to assess the scientific need and objectives for such a telescope, and to study its feasibility of implementation. In the latter task, the working group was assisted by the Goddard Space Flight Center and by the Perkin Elmer Optical Technology Division (under contract to GSFC). Among the principal observational objectives which were outlined are: Detection of hidden hot objects UV morphology of galaxies Determine presence of dust in galactic fields and extragalactic objects Detection and study of faint extended objects Detection and study of emission line objects Observation of solar system objects (particularly diffuse ones). In addition to the scientific objectives, other concerns such as technical feasibility, complexity, and cost resulted in the following guidelines for the engineering aspects of the feasibility study: Confine space to one orbiter pallet. Use one of the standard pointing systems. Minimize cost. 5 degree field of view 0.75 m diameter aperture, minimum 1.0 arc sec image resolution Fast focal ratio (~ f/3.0) 170 mm diameter detector format (~ 10 μm pixels). The feasibility study determined that all of these specifications could be met with reason-able assurance, and baseline configurations were derived for the telescope optics, structure, and electrographic detector. Supporting studies of these and various other aspects of the telescope system are continuing. Since completion of the study, prospects for a long-duration space platform (as an alternative to extended-duration Spacelab missions) have come to the forefront, resulting in significant changes in the program plan. By far the most important of these is that the baseline scientific objective can now be upgraded (from the observation of a few dozen selected fields of special interest) to a complete full-sky survey in the far-ultraviolet to a limiting V magnitude (for an unreddened BO star) of mv = +27.0. An international consortium of astronomical institutions has now joined together in proposing to carry out this program from instrument development to the production and distribution of the final far-ultraviolet all-sky survey atlas.

Continuing plans for an orbiting submillimeter and infrared observatory are discussed. The Large Deployable Reflector (LDR) is seen as a dedicated observatory with a 10-year life-time, which will be useful for a wide range of astronomical observations in a largely unexplored portion of the electromagnetic spectrum. The present concept is for a 10- to 30-m-diameter clear aperture telescope operating at wavelengths from 1000 μm to a diffraction-limited 30 μm. The primary reflector will be composed of a number of close packed hexagonal segments of glass or lightweight composite material. The individual reflector segments will be attached to a truss integrating structure through position actuators providing three degrees of freedom for each segment. Other technical issues discussed are optical design, surface measurement systems, deployment and detectors.

A concept is presented for a phase-coherent optical telescope array which may be deployed in orbit by the Space Shuttle in the 1990's. The system would start out as a four-element linear array with a 12 m baseline. The initial module is a minimum redundant array with a photon-collecting area three times larger than Space Telescope and a one-dimensional resolution of better than 0.01 arc seconds in the visible range. Thermal structural requirements for the optical bench are assessed, and major subsystem concepts are identified.

We are developing numerical methods of image reconstruction which can be used to produce very high angular resolution images at optical wavelengths of astronomical objects from an orbiting array of telescopes. The engineering design concept for COSMIC (coherent optical system of nodular imaging collectors) is currently being developed at Marshall S.F.C., and includes four to six telescope modules arranged in a linear array. Each telescope has a 1.8 meter aperture, and the total length of the array is about 14 meters. This configuration, when controlled to fractional wavelength tolerances, will yield a diffraction pattern with an elongated central lobe about 4 milli-arc-sec wide and 34 milli-arc-sec long, at a wavelength of 0.3 microns, and correspondingly larger at longer wavelengths. The goal of image reconstruction is to combine many images taken at various aspect angles in such a way as to reconstruct the field of view with 4 milli-arc-sec angular resolution in all directions. We are developing a Fourier transform method for extracting from each individual image the maximum amount of information, and then combining these results in an appropriately weighted fashion to yield an optimum estimate of the original scene. The mathematical model is discussed, and the results of preliminary numerical simulations of data are presented.

A design definition study has been made of the image properties of the Nasmyth and prime foci for an f/2.0 primary mirror. Spherical and aspheric corrector subsystems have been examined with the goal of 0.5 arcsec images over a field of 30 arcmin and with final f-ratios from f/1.5 to f/14. Departure of the Ritchey-Chretien hyperbolic primary to a larger conic constant yields a remarkably good field at the prime focus with the addition of two aspheric reflectors as the field corrector. Intercomparison has been made of a number of alternates for performance and practicality aspects.

"Local seeing" effects degrade image quality with conventional impaired performance of mirror alignment systems. Essential to the design of advanced technology telescopes such as the National New Technology Telescope (NNTT) is a knowledge of these dome-induced image motion effects. Characterization of the turbulent, small scale thermal disturbances had been achieved with differential microthermal sensors able to detect millidegree centigrade thermal gradients with a bandwidth of 100 Hertz. Image degradation associated with dome microthermal activity is shown for sub-arcsecond images. Thermal spectra prepared with Fourier transforms from 0.02 to 65 Hertz show increases in thermal activity as the dome is opened. Relative motion of star pairs as a function of field separation is studied using a CID camera system. Light intensity profiles and centroids are recorded along an axis passing through the two separated stars. The relative motion for the binary star γ Andromeda is presented as image motion correlation spectra from .03 to 15 Hertz.

Site seeing at good locations has r_ο (5000 Å) ≈15 cm, and the 5 percentile seeing is about three times better. There are seeing effects at particular locations, but no known geographic trends. The dominant observatory problem is mirror seeing, and ventilated honeycomb mirrors are the current best solution. The detailed effects of phasing precision are discussed. It is assumed that adaptive optics can control separate 2m sized portions of the entrance pupil. Finally the efficiency of use of a telescope is estimated as a function of the image size produced by misfigure, misfocus, misalignment and local seeing etc.

The next generation of large telescones currently in the planning stage will be exposed to higher wind loading due to the cost cutting reduction of dome size in relation to aperture. Thin actively figured mirrors and tighter pointing requirements increase their sus-ceptibility to the general wind loading including higher frguency componants. The figure control and pointing servo loops will require wider bandwidth and higher gain to maintain gool optical performance. To design these servo loops requires knowledge of the wind power spectra to assess loop stability and correction performance. Wind data was collected from several different observing sites. The measurements were made with a Pitot tube anemometer mounted on a tower at the leading edge of a site and four differential pressure sensors mounted across the mirror plane. Anemometer measurements were made at two different elevations to provide site wind vs. elevation extrapolation data. Hundred second measurements, with the data sampled at 50 Hz, were made. Fourier Transforms of the data cover a frequency range of 0.02 to 25 Hz.

A finite element model using an assemblage of plate elements with membrane stress for a 7.6 meter diameter meniscus mirror was extensively explored, for aspect ratios in the range of 50 to 75, and for various materials. Analyses of the support forces necessary for the meniscus shape and a possible design of the mirror cell revealed the necessity of laterally suspending the mirror from a set of points in its midsurface. A map of the heat distribution through the thickness of the meniscus for a typical ambient temperature change and an assumed air conditioning capacity is presented, derived from the solution of a one-dimensional thermal analysis. An equivalent thermal gradient is derived. A thermal edge effect was found, mostly due to the extreme aspect ratios involved, both for fused silica and for pyrex. Edge control becomes particularly necessary with a higher expansion coefficient, and the analysis shows to what extent.

We describe the deflections of uniform-thickness plates supported by discrete points and by continuous rings. The calculations are based on the theory of deflections of thin plates. In some cases the effect of shear on the deflections is also included. The optimum locations of the support points for a wide variety of simple geometries are given. The deflections and methods for estimating the deflections for the limiting case of a large number of support points are also described. Since the slope errors induced in a mirror by its support may be relevant to optical image quality, we describe a method of relating the surface deflections to the surface slopes, and hence the geometric optics image blur that results from a support system. These results are particularly useful for the support of very thin mirrors, where the optimum support is needed.

Traditional telescope control uses rows of switches with individual wires carrying 110 VAC to relays and contactors that operate the various telescope functions. The operation of the control panels is seldom straight-forward and the design is customized for the application. Computer control is often added as an after thought with access provided by means of intimidating and cryptic commands from a typewriter keyboard. This paper reports on Kitt Peak National Observatory's experience with new console technology: using the computer to operate the telescope more reliably and efficiently and to keep track of telescope status for the observer, using solid state and serial techniques to minimize operator hazard and cabling expense while maximizing reliability, replacing the keyboard and customized control panel with interactive CRT video computer terminals to minimize capital expense and maintenance while making the human interface simpler and less error-prone.

For the past year Kitt Peak National Observatory has been conducting a pilot program of remote observing experiments. Observers who would ordinarily have travelled to Kitt Peak have instead observed from their home institutions. Three voice grade telephone lines are used. One line is used for voice communications with the telescope operator, another line is used to interface with the telescope computer which controls the telescope and instrument, and the third line is used to transmit a slow scan (34 second update time) television picture from the telescope field acquisition camera. Observers have found this system to be quite effective. Problems with telephone line noise have been encountered and several schemes for alleviating these problems are presently being implemented. Plans for more rapid video refresh rates over voice grade lines, the use of greater bandwidth communication links and the multiplexing of voice, data and video over single channel are presently being considered. The goal of the experiments is to investigate the effectiveness of this technique with respect to observing efficiency and cost savings.

Communication technology makes it possible to extend the link between the telescope and control room from tens of feet to thousands of miles. Reasons for doing so include: 1) avoiding the health risks and observing inefficiencies caused by hypoxia at high-altitude sites; 2) facilitation of new telescope scheduling schemes; 3) saving travel time and money; and 4) providing troubleshooting backup by the headquarters' engineers and astronomers. The required data rate is estimated by assuming that: 1) the data from a mosaic of nine 1000 x 1000 CCDs will be transmitted every ten (10) minutes; 2) troubleshooting will be supported by transmitting television pictures at a few frames per second. With these assumptions a 500-Kbs data rate is needed to accommodate peak data rates and to have adequate catch-up capability. A two-step implementation of remote observing at Mauna Kea is considered in detail. The first step is installation of a microwave link or glass fiber land line between Mauna Kea's summit and Waimea. The second step is to connect the island headquarters at Waimea to a mainland headquarters, or each U.C. campus, with a satellite link. Cost estimates are given for each step.

An analysis is made of the application of active optics of various degrees of complexity to large ground-based telescopes, using field stars as reference sources. The performance of active compensation systems is evaluated as a function of the number and size of the active zones, reference star magnitude, turbulence strength, and isoplanatic patch size. The results show that for nighttime observations, the average field star distribution allows real-time compensation not only for quasi-static wavefront errors due to optical misalignment and mirror figure, but also for image motion, dome seeing, and some atmospheric turbulence effects. Such compensation is especially valuable under good seeing conditions, when residual errors become a significant factor. It is suggested that all astronomical telescopes could benefit from the use of compensation systems with even a small number of active zones. In large segmented-mirror telescopes, the segments themselves can be used to compensate for random wavefront errors occurring in the entire optical path. In fixed-primary telescopes, the same function may be performed with an auxiliary deformable mirror.

The light propagation through the atmosphere limits the high angular resolution in astronomical imaging. Active optics is a method to overcome this problem. It allows a real-time optimization of the resolving power. An active mirror was developed which consists of, an electrostatically deformable membrane with 5 cm in diameter. The 0.5 micron thick aluminized polymer foil is elongated by an electrode array with 63 hexagonal elements arranged in a ring structure. The sensitivity is in the region of 0,05 microns per volt. It works up to 4 kHz without resonances. The maximum local tilt of the membrane is 3 microns per 5 millimeter. For an atmospheric tilt compensation of the wavefront the mirror housing is in a gimbal-mount. Piezo-electric actuators provide a total mirror angular movement up to 20 Hz and angular sensitivity of 77 arcsecs per kV. With this active mirror device and a multiprocessor-microprocessor control unit the stabilisation of the star-speckle pattern positions and the deconvolution of the speckle patterns are possible. Two control methods are in development and analized in comparison. One of them is working with a modified shearing interferometer as a wave-front sensor and a feedback with cross-talk compensation. The other is sensing the optical information in the image plane by a diode-array and estimating the wave-front by trial and error or different more sophisticated algorithms. The previous system is designed for the 0,75 m RC-telescope with alt-az mount at the Landessternwarte in Heidelberg, FRG.

In adaptive optics, the compensating phase distribution can be generated by an expansion of the turbulence phase distortions into modes of a set of basis functions, i.e. Zernike-polynomials and Karhunen-Loeve-functions. By applying a modal control concept to a practical adaptive optical system, the electrodes of a membrane mirror can be controlled in parallel with greatly reduced cross-talk. The coefficients of the approximating functions, each of which corresponds to a certain mode of the mirror surface, are fed back to the actuators by a modal control matrix. The atmospheric turbulences are taken into account by extending the local differential operator in an "observer"-like structure.

A theoretical study was conducted on the performance of the four-wave mixer as an adaptive optical element in an atmospheric optical communication system. As an active element, the mixer introduced its own quantum noise, and therefore, this noise was considered in the system performance. This analysis is directly applicable to atmospheric compensation for large telescopes in that the four-wave mixer phase conjugation property will restore planar phase fronts to incoming aberrated waves. Results indicated that the average received noise power generated in the four-wave mixer was 60 dB below the average received signal power.

Speckle interferometry techniques are especially attractive with large future telescopes. For example, a 10m telescope can yield an angular resolution of 0.01 arc sec in spite of image degradation by the atmosphere. The limiting magnitude of speckle interferometry is about 20^m. We intend to discuss the following aspects of speckle methods: (a) The influence of telescope aberrations on speckle interferometry measurements. We show laboratory simulations of speckle interferometry with severe stationary telescope aberrations. (b) Roll deconvolution: a method for reconstructing high-resolution images from Space Teles-cope data. In the case of the 2.4m Space Telescope a resolution of about 0.01 arc sec can be achieved. (c) The reconstruction of direct images (instead of autocorrelations) from speckle interfero-grams. Applications on data recorded with large telescopes are discussed. Reconstruction of high-resolution objective prism spectra with the speckle spectroscopy method.

The figuring of the 3.5 m primary mirror of the large telescope is described. The execution of final tests was carried out in a vertical test tunnel by means of a laser interferometer, using a compensating system. A special sequence of interferograms was applied aiming to suppress the thermal, accidental, and environmental influences. The wave front errors were evaluated accordingly. The results show that the standard deviation due to the mirror is 24 nm, out of total standard deviation of 49 nm taken as an avarage of 24 interferograms.

We have begun a program to develop very large lightweight mirrors needed for ground and space based astronomy. A rigid structure of honeycomb sandwiched between faceplates is formed of borosilicate glass by melting it into a complex mold. Thermal gradients in the glass that degrade the figure of thick borosilicate mirrors during use can be largely eliminated in a honeycomb structure, by internal ventilation (in air) or careful control of the radiation environment (in space). We anticipate that, when the seeing is good, ground based telescopes with honeycomb mirrors will give better images than those with solid mirrors. Even slight temperature differences in the air near a slowly eauilibrating solid mirror can significantly degrade the wavefront at a turbulent boundary. We discuss in this paper materials and techniques and the experience that has been gained making trial mirrors and test castings. Our future program calls for making a 1.8m blank that will be figured and tested in the MMT in 1983. After further experience with 3m blanks, we plan to build a large furnace to make mirror blanks up to 8m or even 10m diameter.

The historic market for large mirror blanks can be characterized as a "custom" market, which now appears to be evolving toward higher volumes, multiple modules, and lower mass for application in both land-based telescopes and in space systems. Corning Glass Works is currently evaluating alternatives that would evolve the manufacturing processes to keep pace with the emerging market. An overview of some of the considerations, including alternatives as to types of glass, initial glass form, glass utilization, equipment and mechanization, frit versus fusion bonding, etc. is presented.

The theoretical design and experimental characteristics of multilayer coatings for astronomical telescope mirrors are described. The results show that performance superior to that of a single-layer aluminum coating should be possible.

A method of precise generation of optical surfaces is described. The work is turned about one axis, while a grinding head or cutting tool is gradually moved across by an arm turning about a second axis. This geometry can be used to directly generate spherical surfaces or aspherical surfaces that we call hulahoids. We analyze the properties of these surfaces to determine how well large paraboloids and their off-axis segments can be generated. Very close approximations are possible, for example, 1.5-m segments of a 10-m paraboloid differ by only 0.6 μm peak to peak from the best fit hulahoid when the focal ratio is f/2, and 60 μm when it is f/0.4. For accurate generation the cutting tool may need numerically controlled travel of 1 mm or less, with the position set as a function of the swirlg arm angle.

A new kind of surface measuring machine has been developed under Government contract at Itek Optical Systems, a Division of Itek Corporation, to assist in the fabrication of large, highly aspheric optical elements. The machine uses four steerable distance-measuring interferometers at the corners of a tetrahedron to measure the positions of a retroreflective target placed at various locations against the surface being measured. Using four interferometers gives redundant information so that, from a set of measurement data, the dimensions of the machine as well as the coordinates of the measurement points can be determined. The machine is, therefore, self-calibrating and does not require a structure made to high accuracy. A wood-structured prototype of this machine was made whose key components are a simple form of air bearing steering mirror, a wide-angle cat's eye retroreflector used as the movable target, and tracking sensors and servos to provide automatic tracking of the cat's eye by the four laser beams. The data is taken and analyzed by computer. The output is given in terms of error relative to an equation of the desired surface. In tests of this machine, measurements of a 0.7-m-diameter mirror blank have been made with an accuracy on the order of 0.2 μm rms.

Mirror surface ripple acts as a phase grating to diffract light out of the central maximum of the point spread function (PSF), reducing image quality. The effects of ripple on image quality are examined with the aid of computer simulations using rotationally symmetric wavefront error models, and through interferometric measurements of a mirror known to have significant surface ripple. Image quality is evaluated in terms related to the performance requirements of large orbital astronomical telescopes that must perform in both ultra-violet and visible light. Techniques for measuring and specifying ripple are discussed.

The construction of very large optical telescopes can be prohibitively expensive due to the massive size and weight of the required primary mirrors. One method of reducing the cost of such telescopes is the use of a segmented primary mirror. The computer controlled polisher (CCP) approach is ideally suited to the fabrication of telescope mirror segments. Since the process achieves controlled material removal by regulating the tool motion, there is no need for any surface symetry. Recent experiments have demonstrated the ability of the CCP to figure mirror segments.

The objective of the "Stressed Mirror Polishing" experiment is to investigate the practicality and economical feasibility of manufacturing precisely specified, large diameter, off-axis paraboloidal mirror segments, derived from a long radius of curvature parent paraboloid. Toward this end, the KPNO Large Optical Component Manufacturing Team is attempting to produce two nearly identical, 2-meter diameter segments, of 4.3-meter off-axis center distance, derived from a 40.5-meter radius of curvature parent paraboloid. Optical fabrication methods, optical alignment techniques, optical testing procedures, mechanical properties of the back support, and the bending fixture are discussed from the perspective of their use during manufacturing and testing.

This paper describes the results of recent research undertaken to examine the feasibility of employing laser interferametry to precisely measure absolute distance over extended ranges. Results are presented which show a resolution of 0.03 μm (RMS) for measurements over distances up to 10 meters. The technology developed for achieving these results is based on two-color, synthetic Michelson interferometry employing a new CO2 laser source. Indeed, the new laser is the key element in this process: it was specifically designed to sequentially switch between four sets of stable R- and P-line pairs and thereby provide a basis for forming simultaneous equations which were employed to greatly reduce the half wavelength ambiguity typical of single wavelength interferometers. Potential applications to future optical telescopes - particularly the large, multipanel telescopes under consideration for 10-15 years hence - their initial alignment and control, are suggested.

Shortwave adaptive optical systems are required for ground-to-space transfer of laser power. These systems require high temporal bandwidth wavefront correction elements with between 1,000 and 10,000 degrees of freedom. We have developed a prototype segmented mirror whose wavefront correction capability can match these requirements. Details of the design and suggestions of how this technology could impact the NGT design are given.

The University of California's Ten Meter Telescope project is currently working toward a hardware demonstration of the proposed segmented mirror control system. A prototype with a full size morror is being fabricated to prove the hardware and software concepts while providing hands-on experience with full scale equipment fabrication, testing, installation and operation. A brief derivation of the response function of the system is given, utilizing the calculated mechanical behavior of the mirror support with measured wind power spectra as the forcing function. Design parameter values are then established for the mirror control loop. The dynamic measurement and control computer system is described.

The central problem in the design of a phased array beam expander is devising an alignment technique for pointing and phasing the subaperture telescopes. The advantages and disadvantages of the standard alignment methods are discussed briefly. The most promising candidate technique, heterodyne interferometry, is developed into an optical configuration suitable for breadboard test. The optical configuration is discussed on the basis of reduction of misalignment modes, athermal design, coherence length and polarization effects. Fabrication methods are considered. It is concluded that the optical configuration presented has the capability to meet the severe alignment requirements of a multiple telescope laser beam expander.

Achieving optimal performance from a multiple element telescope is critically dependent on an element position sensing and control system. We first discuss the design of a novel absolute distance measuring interferometer with performance suitable for such a sensing system, and secondly describe how it can be simply integrated into the telescope. Features of the interferometer include an extended ambiguity range, high temporal bandwidth, and very high accuracy. The concept for the overall optical alignment system is expected to compensate in part for local atmospheric distortions as well as mechanical vibrations while not interfering with the optical performance of the telescope.

This paper summarizes the salient features of a telescope for the sub-mm wavelength region, which we have begun to design. Operating from a high and dry mountain the telescope will be used for astronomical observations in the atmospheric transmission windows between 350 μm and 3 mm wavelength. The surface accuracy will be ≤ 15 μm, the reflector diameter ≈ 10 m and the pointing accuracy 1". Extensive use of graphite epoxy material shall eliminate thermal deformations.

There are several possibilities for an optical telescope design to have light gathering power equivalent to that of a 15 metre diameter monolithic mirror telescope. These include a segmented mirror telescope, using as many as ninety hexagonal mirrors; a multi-mirror telescope with six mirrors of over 6 metres diameter; or an array of twenty-five 3 metre telescopes with a common focus. It is the purpose of this paper to show the advantages of an array telescope, these being in summary that: 1) the individual telescopes, using thin meniscus mirrors of short focal length are certainly quite practicable and of low cost; 2) the array is flexible and can be added to without difficulty; 3) the array can be coherent over path lengths of a least 70 metres with existing technology and can have a finite field of at least 5 minutes of arc with good definition; 4) very high resolution interferometry will be possible with such an array, equivalent optically to the VLA in radio wavelengths, as well as the use for photometry, polarimetry, spectroscopy and direct photography or finite field imaging with solid-state detectors.

An optical test facility has been built for testing candidate mirror materials for the Shuttle Infrared Telescope Facility (SIRTF). Mirrors as large as 66 cm in diameter can be tested at temperatures down to about 10K for changes in optical figure of a fraction of a wave from their room temperature figure. Tests of two fused silica mirrors, 50 cm in diameter, are underway. The test mirror is heat sunk to the helium reservoir with copper straps whose connection to the mirror is accomplished by soldering individual strands of copper to small silver spots diffused into the unfigured side of the mirror. This permits relatively fast, conductive cooling of the mirror. In the first test, cooling from 300 to 80K took 4 days; cooling from 80 to 12.5K took 24 hours. Optical access to the cold mirror is through a small (5 cm diameter) glass port in the vacuum chamber placed a few cm short of the radius of curvature of the mirror. A Shack interferometer is used to examine the mirror figure throughout the cool-down. Interferograms are photographed and the fringe patterns are digitized. Contour plots of mirror figure are then calculated using the University of Arizona's FRINGE program on our CDC 7600 computer. Preliminary analysis of interferograms of one of the mirrors shows very little change in figure between 293K and 10.5K (change in rms OPD=0.027 waves).

Two types of monolithic lightweight mirrors with arched backs are discussed: the center-supported single arch and the ring-supported double arch. The theoretical deformations of a 20-in.-diameter double arch mirror are compared with the actual deformations. A mirror of this size weighs about 50% less than an equivalent conventional mirror. The double arch design may be scaled up to 144 in. where the mirror weighs less than 40% of the eauivalent conventional mirror. Further weight savings are possible due to the reduced size and simplicity of the support required by the double arch mirror design.

Two 20-in, diameter, f/4 fused silica mirrors were lightweighted by contouring the rear sides to near parabolic cross sections. One mirror was lightweighted into a single-arch, thick-hub design. The other mirror took the form of a cantilevered double arch. Details of the fabrication are outlined. Following fabrication, these mirrors were interferometrically tested at their centers of curvature in both a face-up and face-down support mode. By subtracting the wavefront errors in these two support modes, the gravity deflections were determined in spite of some residual figure error. Details of the data analysis and the resulting gravity-induced deflections are reported. These empirical results are compared with finite element analyses with favorable results.

The use of a telescope with a "one dimensional" aperture (10 x 800 cm for example) is suggested for high angular speckle interferometry. In the image plane, the speckles are elongated in a direction perpendicular to the entrance pupil. For a given amount of atmospherical turbulence, a one-dimensional pupil gives a contrast gain in the high frequencies region compared to the circular objective having the same resolution limit. Associated with a spectroscope, a one-dimensional telescope allows the representation of the spatial-spectral plane with no loss of photons. This is obtained because the image of the telescope aperture is the entrance slit of the spectroscope. It permits precise, high resolution astronomical measurements as a function of wavelength as well as velocity speckle-interferometry at high level of photons with CORAVEL type experiments. The study of a domeless, altazimuthal mount, low-cost prototype of 4 cm x80 cm, intended for solar observations, is developed at the Nice Observatory.

This paper describes the development of a metrology mount system for large spaceborne optical elements operating at diffraction limit. Verification tests of the mount, together with confirmation of the finite element model of the mount and mirror system, demonstrate that it is capable of simulating a zero-g environment within an uncertainty band of λ/300 rms for a mirror whose characteristic gravity deformation is 12λ. Using a highly detailed finite element model of the sandwich mirror, the effects of stiffness and mass nonuniformity due to faceplate thickness variations and calibration uncertainties in each of the support points on gravity-release errors were assessed and/or compensated for. The mathematics of optimizing these forces to yield a minimum rms figure error and the method and results of the faceplate thickness mapping to determine the spatial weight and stiff-ness variations are also described. Finally, verification of the finite element model is discussed where predicted, and full aperture interferometrically measured figure changes due to discrete changes in the applied force field are compared.