The current status of the Layer-Oriented approach for the MultiConjugated Adaptive Optics techniques in the Extremely Large Telescopes case is reviewed. This includes a brief analysis of the variations of the concept as they have been conceived up to now, and which could be the developments one should expect (and hope for) in the near future, in the perspective of becoming MCAO for ELTs a reality. Then the concept of Multiple Inverse Bessel Beams, or Pseudo Infinite Guide Stars is given, along with some guidelines about what could be precursor experiments to be performed in order to validate and/or to gain experience in the field.

A handful of groups around the world are actively working on the development of the next generation of telescopes of 30 m diameter and more. Present implementations of adaptive optics will be inadequate to realize the full resolving power of these new instruments in imaging and spectroscopy. Instead, multi-conjugate adaptive optics (MCAO) systems are being contemplated. We explore here the application of MCAO using laser guide beacons, to a 30 m telescope. Using a new simulation code, we show that reliance on the expensive lasers needed to generate sodium resonance beacons can be reduced through the use of refocused Rayleigh laser guide stars at much lower cost. We show that in the geometric optics approximation, modes of the phase distortion of order less than or equal to the number of deformable mirrors in the MCAO are correctable with no isoplanatic error. A new figure of merit is derived which predicts the relative ability of a chosen beacon/deformable mirror architecture to sense and correct wavefront aberration, based solely on knowledge of the optical geometry and the statistics of the aberration to be corrected. Numerical simulation can therefore be minimized by avoiding the exploration of unpromising beacon arrangements.

For reasonable Field of View (of the order of several arcmin), the pupil overlap in the high altitude layers is marginal or non existing for 8m class telescopes and dominant for 100m class apertures. Starting from this matter of fact, we formulated a multiple resolution, multiple Field of View version of the layer oriented adaptive optics approach which looks at the same layer with field apertures and spatial resolution that are mutually compensating each other: a larger field, allowing for the collection of more photons from natural stars, is used to sample the layer with a finer spatial detail, and viceversa. In this way the small field of view, with coarse spatial sampling, gives information on the behavior of the atmosphere for a large thickness around the focused layer, at the expense of the detailed information in a thin slab centered on the layer itself. Such lacking information is obtained from the wider field channel. This requires the coaddition of the two pieces of information at the level of the Fourier pair of the reimaged layer, which can be accomplished, in a practical way, directly in the spatial domain rather than in the Fourier one. The consequences on the sky coverage estimations and on the hardware implementations are discussed. Finally some comments are given about other degrees of freedom that can be used to improve the performance of layer oriented adaptive optics.

The optical design for the proposed Euro50 extremely large telescope with integrated adaptive optics (AO) is presented. For atmospheric turbulence correction, we propose using single and dual-conjugate AO systems working with natural and laser guide stars. The corrective shape of the deformable mirrors (DMs) is derived from an analytical algorithm based on minimization of the sum of the residual power spectra of the phase fluctuations seen by guide stars after correction. Predictions for performance of the Euro50 ELT with Dual-conjugate AO are given for the K band using a seven layer atmospheric model for the atmosphere at the Observatorio del Roque de los Muchachos (ORM) on La Palma. The average Strehl ratio is used to quantify the system performance for different values of actuator pitch and DM conjugation altitudes. The influence of the outer scale and telescope pointing on the RMS stroke of the DMs is presented. It is concluded that construction of such a system is feasible and that there is a need for development of a simulation tool to verify the analytical calculations. Precise knowledge of the outer scale of the atmosphere at the ORM is needed to establish the dynamical range of the mirrors.

The design and elaboration of Extremely Large Telescopes (ELT) with primary mirror from 20m to 100m face with many challenges: mechanical, optical, computational, etc. To benefit completely of the full potentiality of such facilities, an Adaptive Optics System (AOS) have also to be designed for these telescopes. For whole field--of--view compensation and full sky coverage, the new but promising Multi--Conjugated Adaptive Optics (MCAO) technique has to be envisaged. The first step towards the design of an MCAO system is the numerical simulation. This is the first challenge we have to face. The scale of AO simulations being imposed by the ratio (D/r₀), the simulation requirements of a MCAOS for an ELT, in terms of computing power and memory available to store the data, reach and sometimes overcome the capacity of actual computers. In ESO, we have evaluated different hardware and software strategies to achieve MCAO simulations goals. Two codes have been developed to simulate MCAOS using an analytical and an end-to-end model. The goals and advantages/limitations of both approaches is shown. The hardware requirements for both methods is also given through the size of their largest matrices. And finally, results of hardware and software tests for MCAO simulations with PC--cluster and paralleled code are presented.

We describe the adaptive optical (AO) requirements, optical design, and expected performance of a near-infrared AO coronagraph for a 30 meter class giant telescope. The optical design of the instrument consists of back-to-back finite conjugate relays, each containing a collimated space between a pair of toric mirrors. The first collimated space contains the atmospheric dispersion compensator and the AO components, which are a tip/tilt mirror, a MEMS deformable mirror, and a beamsplitter for the wavefront sensing path. An occulting disk or similar focal plane mask is located at the intermediate image between the two relays, and a Lyot stop is placed at the pupil plane in the second relay. The required AO order of correction for a Strehl ratio of 0.9 at a wavelength of 2.2 microns is about 150 by 150, and the required control bandwidth is 42 Hz. The limiting magnitude at this level of performance is estimated to be 10.4.

Layer-Oriented MCAO is a promising technique that can scale up easily compared to similar competing techniques. The Multiple-Field-of-View Layer-Oriented make a further step employing more guide stars to sample the turbulence with the net effect of increasing the sky coverage enabling MCAO systems based only natural guide stars. However even a simple estimation of Layer-Oriented M-FoV MCAO performance for the Extremely Large Telescopes, in particular for OWL, has not been attempted yet. The main problem is the enormous amount of computing power (processor and memory) required to properly carry out an accurate simulation for ELTs. In this work we perform a numerical simulation using some shortcuts to simplify the problem and achieve a first cut of what the performance of Layer-Oriented M-FoV MCAO could be for a 100m-class telescope. We take a mixed analytical-numerical approach in order to significantly reduce the amount of computing power required while maintaining a good level of accuracy.

The scheme presently envisaged for the EURO50 adaptive optics is presented. The Euro50 adaptive optics will primarily work with laser guide stars (LGSs) and control of either one or two deformable mirrors (SCAO and DCAO respectively), but operation using a natural guide star (NGS) is also foreseen. The point spread function (PSF) for SCAO operation using a single NGS is evaluated. An algorithm for optimal control of the deformable mirrors (DMs) using LGSs and Shack-Hartman wavefront sensors is presented and commented upon. It is an extension of a recently developed algorithm for optimal control using NGSs and working in the spatial Fourier domain. In addition the concept of a virtual wavefront sensor is introduced to overcome the difficulty in transmitting a large number (37) of LGSs to the final DCAO focus with both adequate field and adequate aberrations. The expected performance is estimated in form of maps of the Strehl ratio versus field angle using a standard seven layer atmospheric model for the Observatorio del Roque de los Muchachos (ORM) site on la Palma for the case of the outer scale being either 20 m (nominal for ORM) or infinity (Kolmogorov - most pessimistic case).

This paper is intended to discuss the impact of noiseless CCD detectors in three significant areas of the development of large telescopes and the instruments that go on them. These are (1) CCDs that have all the characteristics that we are used to seeing in CCDs can now be made with negligible readout noise even at higher pixel rates (> 10 MHz) and will allow rather different approaches to the design of instrument is generally, (2) the technique of achieving diffraction limited imaging in ground-based telescopes known as Lucky Astronomy in which images are taken at high speed has been demonstrated to work under a variety of different conditions and (3) some suggestions as to how these methods may be applied directly to much larger diameter telescopes in order to achieve high resolution imaging and spectroscopy without the expense of laser guide stars or multi-conjugate adaptive optics.

In the troposphere water vapor plays a fundamental role in radio propagation. The refractivity of water vapor is about 20 times greater in the radio range than in near-infrared or optical regimes. As a consequence, phase fluctuations at frequencies higher than about 1 GHz are predominantly caused by fluctuations in the distribution of water vapor. On filled-aperture telescopes radio seeing shows up as an anomalous refraction (AR), i.e. an apparent displacement of a radio source from its true position. The magnitude of this effect, as a fraction of the beam width, is bigger on larger telescopes. I will thus present a model study of AR effects, obtained producing numerical simulations of two-dimensional phase screens. I will finally discuss the basic concept and requirements of a tip-tilt compensation system at millimeter wavelengths, and will also describe a proposed design based on a scanning microwave radiometer as a wave front sensing device.

Following the development of a Hartmann-Shack sensor with a variable pitch, we needed a data analysis method that does not require a priori the locations of the foci of the lenslets. Thus we developed a method for retrieving the wave front from a Hartmann-Shack sensor using a two-dimensional Fourier demodulation technique. Digital demodulation is applied twice on the pattern, along two orthogonal axes. This provides both phase gradient components in real space, smoothed over the extent of the array. If necessary, a Fourier reconstruction can be used to obtain the corresponding wave front from the gradient components. The technique is rather robust to Poisson and white noise and to relatively large deviations of the foci. The method was tested successfully on a mirror with known large aberrations.

The California Extremely Large Telescope (CELT) is a joint project of the University of California and the California Institute of Technology to build and operate a 30-meter diameter telescope for research in astronomy at visible and infrared wavelengths. The current optical design calls for a primary, secondary, and tertiary mirror with Ritchey-Chretién foci at two Nasmyth platforms. The primary mirror is a mosaic of 1080 actively stabilized hexagonal segments. This paper summarizes the recent progress on the conceptual design of this telescope.

The California Extremely Large Telescope (CELT) is a project to build a 30-meter diameter telescope for research in astronomy at visible and infrared wavelengths. The current optical design calls for a primary, secondary, and tertiary mirror with Ritchey-Chretién foci at two Nasmyth platforms. The primary mirror is a mosaic of 1080 actively-stabilized hexagonal segments. This paper summarizes a CELT report that describes a step-by-step procedure for aligning the many degrees of freedom of the CELT optics.

The current design concept for the California Extremely Large Telescope (CELT) includes 1080 segments in the primary mirror, with the out-of-plane degrees of freedom actively controlled. We construct the control matrix for this active control system, and describe its singular modes and sensor noise propagation. Data from the Keck telescopes are used to generate realistic estimates of the control system contributions to the CELT wavefront error and wavefront gradient error. Based on these estimates, control system noise will not significantly degrade either seeing-limited or diffraction-limited observations. The use of supplemental wavefront information for real-time control is therefore not necessary. We also comment briefly on control system bandwidth requirements and limitations.

Ultra-high contrast imaging with giant segmented mirror telescopes will involve light levels of order 10^-6 times that of the central diffraction spike or less. At these levels it is important to quantify accurately a variety of diffraction effects, including segmentation geometry, intersegment gaps, obscuration by the secondary mirror and its supports, and segment alignment and figure errors, among others. We describe an accurate method for performing such calculations and present preliminary results in the context of the California Extremely Large Telescope.

Considered until recently one of the best telescopes in the world, the Canada-France-Hawaii Telescope (CFHT) is now bypassed by larger telescopes. Aware of this problem, the CFHT Science Advisory Council (SAC) solicited proposals from the CFH community groups to replace the present telescope by a world-class research facility before the end of the decade. A report describing our proposal is available on our web site (www.astrsp-mrs.fr/denis/ngcfht/ngcfht.html). The motivation to design and build a new telescope is often driven by the astronomers need to observe fainter and fainter sources. The basis of the next generation CFHT (NG-CFHT) is therefore to increase the size of the primary mirror to reach fainter and more remote objects in the luminosity functions. But beyond this photon quest , the way we use the photons is also very important. The development of new technologies will permit an optimization of performances and a better image quality thanks to state-of-the-art instruments on state-of-the-art telescopes.

A design is proposed for a 20 m Canadian Very Large Optical Telescope (VLOT). This design meets the science, schedule, and availability requirements of the Canadian astronomical community. The telescope could be operational by early in the next decade to complement the science discoveries of the Next Generation Space Telescope (NGST) and Atacama Large Millimeter Array (ALMA). This design is suitable for location on the Mauna Kea summit ridge, and could replace the current 3.6 m CFHT telescope. The telescope structure provides room for two vertically oriented Nasmyth instruments, implements a very stiff monocoque mirror cell, and offers a short and direct load path to the telescope mount. A Calotte style dome structure offers many advantages over current designs including lower and more even power requirements, and a circular aperture that will better protect the telescope structure from wind buffeting. The science requirements are presented, and the telescope optical design, primary mirror pupil segmentation options, including hexagonal segments and a radial segment design with a central 8 m mirror, are considered. Point spread function plots and encircled energy calculations show that there is no significant diffraction performance difference between the options except that hexagonal segments in the 1 m point-to-point range appear to deliver poorer PSF's as compared to 2 m and larger segments. Plans for implementation of a Matlab based integrated telescope model are discussed. A summary of adaptive optics system issues for large telescopes is presented along with plans for future research in AO.

We describe a 'point design' for a 30m Giant Segmented Mirror Telescope (GSMT) aimed at meeting a set of initial science goals developed over a period of two years by working groups comprised of more than 60 astronomers. The paper summarizes these goals briefly, captures the top-level performance requirements that follow from them, and describes a plausible, first-cut technical solution developed as part of an overall systems-level analysis. The key features of the point design are: (1) a fast (f/1) primary; (2) an adaptive secondary that serves both to compensate for the effects of wind buffeting and as the first stage of three adaptive optics systems: (i) multi-conjugate AO; (ii) high-performance on-axis AO; (iii) ground-level seeing compensation; (3) a radio telescope structure; (4) multiple instrument ports (prime focus; Nasmyth foci; direct Cass); (5) an hierarchical control system comprising multiple active and adaptive elements.

The next generation 30-meter class ground-based telescopes pose an unprecedented challenge for control systems envisioned to support diffraction limited imaging. Our approach is a multi-tiered, decentralized control architecture utilizing two kinds of feedback: optical and mechanical. Based on simulations, we suggest a configuration where the optical feedback loops for the main axes, secondary mirror rigid body motions and segmented primary mirror are separated in both temporal and spatial frequency domains. The active optics system maintaining the continuity and higher order shape of the primary mirror is based on mechanical sensors while the low order shape is corrected through optical feedback. Shown are the results of numerical simulations using real, measured wind data that prove the feasibility of the suggested architecture.

To achieve diffraction-limited seeing with the next generation of giant telescopes will require a multi-tiered approach employing both active and adaptive optics. Before removing atmospheric effects via adaptive optics, the telescope structure must be actively controlled to remove structural dynamics effects from the optical wavefront. While low frequency thermal and gravitational effects may be removed by the primary mirror control system, a more difficult challenge is the higher frequency, wind-induced vibrations of the telescope structure. This paper will address the control system design for the rigid secondary mirror support structure using modern control methods. Multivariable control methods are motivated by the large number of coupled structural modes that contribute to the wavefront error at the secondary mirror. H₂ methods are applied to the secondary mirror control system in order to validate the design approach for achieving nominal performance at the system level. The approach investigated for this paper involves using wavefront information to remove wavefront error at the secondary mirror. This work will serve as a basis for demonstrating the feasibility of the overall control architecture for the GSMT point design study.

Preliminary requirements and possible technological solutions for the next generation of ground-based optical telescopes were laid down at ESO in 1998. Since then, a phase A study has been commissioned, the objective of which is to produce a conceptual design compatible, to the maximum possible extent, with proven technology, and establish realistic plans for detailed design, site selection, construction and operation for a 100-m class optical, diffraction-limited telescope. There was no doubt about how daunting such a challenge would be, but, somewhat surprisingly, it turns out to be firmly confined to adaptive optics concepts and technologies. The telescope itself appears to be feasible within the allocated budget and without reliance on exotic assumptions. Fabrication of key subsystems is fully within the reach of a properly engineered, industrialized process. A consolidated baseline is taking shape, and alternative system and subsystem solutions are being explored, strengthening the confidence that requirements could be met. Extensive development of wavefront measurement techniques enlarges the palette of solutions available for active wavefront control of a segmented, active telescope. At system level, ESO is developing enabling experiments to validate multi-conjugate adaptive optics (MAD for Multi-conjugate Adaptive optics Demonstrator) and telescope wavefront control (APE, for Active Phasing Experiment).

We consider the diffraction by segmented apertures with a very large number of segments, concentrating mostly on the effects which lead to the appearance of regular patterns of diffraction peaks, specific to extremely large telescopes (ELT). These effects are associated to gaps between segments, turned down/up segment edges and random tip-tilt errors. We discuss briefly the effect of segment piston error, which does not produce the higher-order diffraction peaks, but also affect the image quality. We deliver an analytical expression of the Point Spread Function (PSF) associated to each particular case and approximate formulae for the Strehl ratio and relative intensity of higher-order diffraction peaks.

Any future giant ground-based telescope must, at a minimum, provide foci for seeing-limited imaging over a wide field and for diffraction-limited imaging over ~1 arcminute fields corrected by adaptive optics (AO). While this is possible with a number of design concepts, our choices are constrained if we anticipate wanting to later add a second telescope for imaging with still higher resolution, and very high contrast imaging for exoplanet studies. This paper explores designs that allow for such future development. Higher resolution imaging by interferometric combination of the AO-corrected fields of two telescopes is possible without loss of point-source sensitivity or field of view, as long as the baseline can be held perpendicular to the source and can be varied in length. This requirement is made practical even for very large telescopes, provided both can move continuously on a circular track. The 20/20 telescope illustrates this concept. Telescopes so mounted can additionally be operated as Bracewell nulling interferometers with low thermal background, making possible the thermal detection of planets that would have been unresolvable by a single 20 m aperture. In practice, limits set by funding and engineering experience will likely require a single 20 or 30 m telescope be built first. This would be on a conventional alt-az mount, but it should be at a site with enough room for later addition of a companion and track. In anticipation of future motion it should be compact and stiff, with a fast primary focal ratio. We envisage the use of large, highly aspheric, off-axis segments, manufactured using the figuring methods for strong aspherics already proven for 8 m class primaries. A compact giant telescope built under these guidelines should be able to perform well on its own for a broad range of astronomical observations, with good resistance to wind buffeting and simple alignment and control of its few, large segments. We compare here configurations with adjacent hexagonal segments and close-packed circular segments. For given segment parent size and number, the largest effective aperture is achieved if the segments are left as circles, when also the sensitivity and resolution for diffraction-limited operation with AO is higher. Large round segments can also be individually apodized for high-contrast imaging of exoplanets with the entire telescope-for example 8.4 m segments will yield 10^-6 suppression 0.05 arcsec from a star at 1 μm wavelength, and at 0.25 arcsec at 5 μm.

We present a plan for making the optics of a 21 m telescope that builds on advances in mirror design and fabrication developed for the Large Binocular Telescope and other large telescopes. The 21 m telescope, with a fast f/0.7 primary mirror made of only seven large honeycomb-sandwich segments and an adaptive secondary mirror with matching segments, is much stiffer than other designs and offers simpler and more accurate wavefront control. It can be a powerful stand-alone telescope, or one of a pair that move on a circular track to achieve coherent imaging with baselines up to 120 m (the 20/20 telescope). Each segment of the 21 m primary mirror is similar to an 8.4 m LBT primary, and each segment of the 2.1 m adaptive secondary mirror is similar to an LBT secondary. The off-axis segments of both mirrors can be made with the same methods and equipment currently used at the Steward Observatory Mirror Lab, and can be polished with the same stressed-lap polishing system used for the LBT mirrors. A change in algorithm to accommodate the asymmetric surface is required, but no new hardware development is needed because the lap bending is similar to that for the LBT mirrors. Each segment can be measured interferometrically, with a combination reflective and diffractive null corrector producing an accurate template wavefront and alignment references for the segments.

The 20 20 Telescope is a 30 meter class telescope comprised of two 21.2m collector telescopes on a 100m circular track. Each collector telescope has a focal ratio of F: 0.7 and is comprised of seven 8.4 m segments. There is an instrument bridge that carries the combining instrument. The proposal for 20 20 is to have discrete combiner stations for 30,60,and 100 meter baselines. Additional focal stations are implemented for Nasmyth and bent Cassegrain. The Track has the same segmented construction and tracking motion on hydrostatic bearings as LBT. The collector telescope buildings will co-track and co-rotate on separate tracks. The 30m design has the same basic shape as a single 21 meter Collector but many aspects are different. The 30 meter telescope is a single hexagonal aperture with a primary at F: 0.5. There are 13 that are 8.74m hexagons and 6 half hexagons. The 30m telescope has primarily Nasmyth platforms behind the primary mirror. Both telescopes have a 30 meter equivalent circular aperture. Both telescopes have high structural performance, at 6.5 Hz and 5.3 Hz respectively. Both are balanced, and use similar designed components. Comparison of their characteristics and design differences can show the strengths and weaknesses of each.

Euro50 is a proposed optical telescope with an equivalent primary mirror diameter of 50 m. Partners of the collaboration are institutes in Sweden, Spain, Ireland, Finland, and the UK. The telescope will have a segmented primary mirror and an aplanatic Gregorian configuration with two elliptical mirrors. For a 50 m telescope there would be no economical advantage in going to a spherical primary. The size of the primary mirror segments (2 m) has been selected on the basis of a minimization of cost. An adaptive optics system will be integrated into the telescope. The telescope will have three operational modes: Seeing limited observations, single conjugate adaptive observations in the K-band, and dual conjugate observations also in the K-band. An upgrade to adaptive optics also in the visible down to 500 nm is foreseen. There will be an enclosure to protect the telescope against adverse weather and wind disturbances. Integrated simulation models are under development. The project time will be 10 years and the cost some 591 MEuros.

The 100-m OWL telescope being considered for open-air operation, wind is an essential disturbance affecting the tracking performance and figure of the primary and secondary mirrors. ESO has undertaken a study to build up a reliable and flexible computer model of the telescope and its environment. This model can, in a cost-effective way, be used to assess the wind loading under different conditions and configurations, before entering into more expensive wind tunnel testing. This paper presents the first preliminary results obtained with Computational Fluid Dynamic (CFD) methods about the wind action on the OWL 100-m telescope, in terms of pressure time histories and in the frequency domain. Preliminary conclusions on the effect of the wind loading on the design are also drafted.

In order for the California Extremely Large Telescope (CELT) to achieve the required optical performance, each of its 1000 primary mirror segments must be positioned relative to adjacent segments with nanometer-level accuracy. This can be accomplished using three actuators for each segment to actively control the segment in tip, tilt, and piston. The Keck telescopes utilize a segmented primary mirror similar to CELT employing a highly successful actuator design. However, because of its size and the shear number of actuators (3000 vs. 108 for Keck), CELT will require a different design. Sensitivity to wind loads and structural vibrations, the large dynamic range, low operating power, and extremely reliable operation, all achieved at an affordable unit cost, are the most demanding design requirements. This paper examines four actuator concepts and presents a trade-off between them. The concept that best met the CELT requirements is described along with an analysis of its performance. The concept is based on techniques that achieve the required accuracy while providing a substantial amount of vibration attenuation and damping. A prototype actuator has been built to validate this concept. Preliminary tests confirm predicted behavior and future tests will establish a sound baseline for final design and production.

The Giant Segmented Mirror Telescope (GSMT) is a 30-m fully-steerable ground-based optical/infra-red telescope with actively controlled segmented mirrors. This paper presents the initial point design for the telescope structure which was used to evaluate the expected structural behavior needed in developing other aspects of the telescope design, in particular, the adaptive optics system. The primary mirror consists of 618 hexagonal 1.152-m mirror segments supported on 91 rafts. A typical raft supports seven mirror segments. There are two level of actuation; each mirror segment is supported by actuators mounted on the raft structure and each raft is supported by another level of actuators mounted on the elevation structure. A radio telescope-type design is used for its structural advantages as well as for accommodation of large instruments. In particular, the location of the elevation bearings allows for an efficient support of the primary and a smaller azimuth structure. The elevation structure consists of (1) a space truss backstructure that supports the rafts, (2) a braced tripod structure that supports the prime focus instrument or secondary mirror, and (3) the transition structure and elevation wheel that connect the backstructure to the elevation bearings and provide support for the drive arc and the counterweight. The azimuth structure accommodates a large Nasmyth platform. Issues covered in this paper include: natural frequencies, response to wind buffeting, deflection limits to accommodate actuator strokes, span between elevation bearings, braced tripod to minimize blockage, and configuration to maximize space for instruments.

The drive and bearing technologies have a major impact on the static and dynamic performances of a steerable telescope. The costs related to the complexity of the design and its Reliability, Availability, Maintainability and Safety (RAMS) are not negligible. The design constraints of Extremely Large Telescopes (ELT) depart from those applicable to the current generation of 8 to 10 meter class telescopes, thus suggesting that alternative solutions should be investigated. This paper discusses the feasibility of implementing a design based on friction drives and bogies, tailored to OWL’s azimuthal and altitude degrees of freedom. The estimated static and dynamic performance of the mechanical structure, the achievable angular resolution, the optimal distribution of loads and stresses, the RAMS performance and finally its cost efficiency, make this solution particularly attractive.

The Giant Segmented Mirror Telescope (GSMT), along with other proposed Extremely Large Telescopes (ELT's) with apertures over 20-m, is likely to impose rather different site selection criteria than those for existing large telescopes. Advantageously, remote-sensing techniques allow rather more objective comparisons than was possible in the past, and the general task is aided by numerical modeling and new ground-based measurement techniques. In recognition of the difficulty of the site-selection process, co-operation between the several ELT projects is the norm. A description is given of the site survey for the GSMT, begun in late 1999, and now part of the GSMT studies and evaluation project, run by the Associated Universities for Research in Astronomy (AURA) New Initiatives Office (NIO).

During the GTC integration phase, the Commissioning Instrument (CI) will be a diagnostic tool for performance verification. The CI features four operation modes-imaging, pupil imaging, Curvature WFS, and high resolution Shack-Hartmann WFS. After the GTC Commissioning we also plan to install a Pyramid WFS. This instrument can therefore serve as a test bench for comparing co-phasing methods for ELTs on a real segmented telescope. In this paper we made a general instrument overview.

Segmented mirror telescopes take advantage of modular design to achieve large apertures at low cost. This paper describes the segment mount developed for the Southern African Large Telescope. The mount provides passive precision support for the optics, kinematic registration to the primary mirror truss, precision tip-tilt and piston adjustments, and interchangeability between segments and mounts. A trial production run of mounts is now in fabrication prior to full production of 91 units needed to populate the SALT primary.

This paper presents designs of compact 21 and 30 m aperture telescopes with primary focal of f/0.7 and f/0.56. The 20 20 telescope moves on three axes; the elevation axis (which is below the primary vertex), the azimuth axis, and a tracking axis at the center of 100 m diameter tracks. The 30 m telescope has an elevation and azimuth axis. All of the axes move on hydrostatic bearings. A primary requirement for such large telescopes is stiffness against deformation by wind gusts. The mass and stiffness needed for the structure is substantially independent of the primary mirror mass, which can therefore be set by thermal and diffraction issues. For the 21 m design, whose primary has seven 8.4 m glass segments weighing 128 tons, the total moving mass is 905 tons, and the lowest resonant frequency 6.5 Hz. For the 30 m design, whose primary has, 13 whole and 6 half, glass segments 8.7 m, across the points, weighing 256 tons, the total moving mass is 3,460 tons, and the lowest resonant frequency 5.3 Hz. These practical designs offer two versatile telescopes with high performance.

The California Extremely Large Telescope, CELT, is a proposed 30-m telescope. Choosing the best possible site for CELT is essential in order to extract the best science from the observations and to reduce the complexity of the telescope. Site selection is therefore currently one of the most critical pacing items of the CELT project. In this paper, we first present selected results from a survey of the atmospheric transparency at optical and infrared wavelengths over the southwestern USA and northern Mexico using satellite data. Results of a similar study of South America have been reported elsewhere. These studies will serve as the pre-selection criterion of the sites at which we will perform on-site testing. We then describe the current status of on-site turbulence evaluation efforts and the future plans of the CELT site testing program.

This paper presents a Navier-Stokes based numerical methodology for 3D unsteady viscous flow over a general mirror configuration. In this approach, the Reynolds averaged Navier-Stokes equations in finite volume form are solved unsteadily using Roe’s scheme on body fitted H-H type grid. A two-equation turbulence model was implemented. This method is three or five order accurate in space and first order accurate in time. The physical space modeled is a 120m×120m×120m region enclosing the primary mirror at the center. The inflow turbulence from the enclosure openings is represented by turbulence kinetic energy at the inflow boundary. The wind buffeting effect was studied by directly calculating flow field response to buffeting impinging wind. The unsteady pressure distribution on the mirror is extracted and analyzed for the amplitudes and frequencies of dynamic wind loading. The computational results are visualized to highlight the flow pattern, particularly on the mirror upper surface. Results are presented for a 30-meter aperture GSMT primary mirror. A 1:833 model of the GSMT primary mirror was tested in a water tunnel, in which the velocity distribution was measured using PIV technique. The preliminary experimental observation serves as qualitative validation of the simulation capability.

Segmented mirror technology has been successfully applied to 10m class telescopes (Keck, HET, GTC) and is widely recognized as mandatory for Extremely Large Telescopes. For optimal performance the wavefront error associated with segmentation should remain within conservative limits, typically 1/20th of a wave. Several phasing techniques and associated metrologies are under development, with a view to extrapolate such methods to the 100-m OWL telescope. We investigate a novel technique based on Mach-Zehnder interferometry, whereby the wavefront in one of the interferometer arms is spatially filtered so as to provide a reference wave, prior to having the two arms recombined to produce suitable interferograms. We introduce a theoretical description of the interferometer, as well as results of simulations, showing that with proper settings of the interferometer's parameters, the technique can be made insensitive to atmospheric turbulence and, more generally, to almost any error source not associated with the segmentation. It also appears that, in a telescope that would include more than one segmented mirror, simple processing allows to disentangle the signal associated to each of them. Finally, we outline the development still required to complete a full qualification of this approach.

Existing design rules break down as we plan for a new generation of giant optical telescopes of 20, 30, 50, even 100 meters in diameter. One might expect these telescopes to converge on the design universally ac-cepted for similarly sized radio telescopes, with their highly aspheric, ~ f/0.4, primary dish. But most of the optical design concepts now under consideration have favored spherical or relatively slow paraboloidal surfaces, leading to a much larger telescope, more subject to wind buffeting, and requiring gargantuan en-closures for protection. This paper explores issues and limitations for building and operating telescopes as the primary focal ratio is reduced to a value as small as f/0.4. Such compactness will be particularly impor-tant for mechanical stability, cost control and for large telescopes that must move continuously on a track, as in the 20/20 concept. We find that fabrication and alignment methods for telescopes using numerous small (1-m class) segments are driven to long focal ratios, while those using few large, actively controlled segments can be made as fast as f/0.5.

The diffractive performance of large segmented telescope mirrors is discussed with particular emphasis on understanding the PSF structure of arrays of hexagonal and circular mirrors. Discriminating issues needed to select mirror size and array are discussed and one solution is proposed as it has been studied for the Canada France Hawaii Telescope.

We show that the same technology developed for MMT and LBT Adaptive Secondary mirrors can be used for building segmented adaptive mirrors of essentially any size. This seems to be at the moment the most promising approach to provide the enormous number of degrees of freedom necessary for adaptive correction at visual wavelengths in giant telescopes. In this paper we recall the analytical formulation of the problem and we report recent numerical studies and initial experimental results obtained with prototype actuators for large adaptive segments.

New very large telescopes with apertures as large as 100 meters are being proposed. They will be made up of mirror segments only a meter or two in diameter and phased together. The diffraction-limited resolution of a mirror is directly proportional to its diameter, and the light-gathering-power is proportional to the square of the diameter. Near-diffraction-limited performance using adaptive optics would make such large mirrors very exciting. We have built two small prototype composite adaptive optic mirrors of graphite fiber impregnated cyanate ester resin driven by actuators spaced 4 cm apart and with a faceplate influence function of 5 cm. The second mirror assembly also makes possible a 2 cm actuator spacing. The overall figure is not yet as good as desired, but we believe that much of this problem can be corrected by mechanical adjustment of the actuator rest positions and use of low expansion mandrels. This mirror concept, when realized in primary mirror segments a meter or more in diameter, should make correction possible for atmospheric turbulence under almost any observatory seeing conditions. The composite optical faceplate in the most recent prototype had a roughness of 0.6 to 0.8 nm. Two centrifugal elutriation super-polishers, each over 1.2 meters in diameter, are in place to produce superpolished mandrels on which to form superpolished faceplates over a meter in diameter. Scattered light from such a mirror surface will be reduced by as much as a factor of ten, as compared to conventional fresh feed polishing. The name "transfer mirrors" rather than the widely recognized but poorer quality "replica mirrors" is given to such faceplates. They have an expansion coefficient comparable to ULE quartz or Zerodur, and are lightweight with 10-20, an aerial density of 17 kg/m² for the mirror with a 4 cm actuator spacing or 34 kg/m² for the mirror with 2 cm actuator spacing. In both cases the effect of gravitational sag is minimized. A 60 volt potential results in actuator displacement of 5 mm as measured inter

We propose a high-density coherentarray of eighteen 10-meter liquid-mirror telescopes. Each telescope is equipped with a reflective corrector system that enables it to point and track within an eight-degree diameter accessible field of view centered on the zenith. Equipped with adaptive optics, each telescope would deliver a diffraction-limited beam to a common focal plane. There, a dichroic optical-infrared camera will provide simultaneous imaging in six wavelength bands from 0.6 to 2.2 um. Operating coherently as a Fizeau imaging interferometer, the array would be capable of achieving 4 mas resolution at 1 um and a point source sensitivity of 40 pJy (I_AB = 34.9) in a single 30 min exposure. With incoherent beam recombination it would be capable of 25 mas resolution and a point-source detection limit of 250 pJy (I_AB = 32.9) in a single 30 min exposure. The telescope uses proven technologies and could be built in a comparatively-short time.

The trend to minimize the thickness in optical mirrors has led to several practical limits in their fabrication and operation. The design of a flexible membrane mirror segment, which overcomes most of these limitations and can be conformed to giant segmented primary mirrors, is presented. The segment consists of a lightweight multi-layer hybrid structure, which will permit precise active control of the reflecting surface by means of a continuous elastic medium interface with embedded pneumatic actuators. Conceptual designs, finite element analysis model simulations and experimental results are included.

The sitting of future giant telescopes poses a challenge to the astronomical community. Some of the top level existing observatories are already too crowded; others do not meet the specifications on particular environmental parameters such as wind or seismicity. On the other hand it has been proven that continental sites can be as good as islands, opening the way for many candidates not yet properly studied. To operate a pre-selection, it is proposed to implement a dedicated geographical information system (GIS) handling the wealth of available remote sensing data worldwide. When focusing onto particular summits, the GIS should then be complemented by mesoscale modeling of the local flow. Some examples are given to support this strategy.

Extremely large filled-aperture ground-based optical-IR telescopes, or ELTs, ranging from 20 to 100m in diameter, are now being proposed. The all-important choice of the aperture must clearly be driven by the potential science offered. We here highlight science goals from the Leiden Workshop in May 2001 suggesting that for certain critical observations the largest possible aperture - assumed to be 100m (the proposed European OverWhelmingly Large telescope (OWL) - is strongly to be desired. Examples from a long list include: COSMOLOGY: * Identifying the first sources of ionisation in the universe, out to z ≥14 * Identifying and stufdying the first generation of dusty galaxies * More speculatively, observing the formation of the laws of physics, via the evolution of the fundamental physical contants in the very early Universe, by high-resolution spectroscopy of very distant quasars. NEARER GALAXIES: *Determining detailed star-formation histories of galaxies out to the Virtgo Cluster, and hence for all major galaxy types (not just those available close to the Local Group of galaxies). THE SOLAR SYSTEM: A 100-m telescope would do the work of a flotilla of fly-by space probes for investigations ranging from the evolution of planetary sutfaces and atmospheres to detailed surface spectroscopy of Kuiper Belt Objects. (Such studies could easily occupy it full-time.) EARTHLIKE PLANETS OF NEARBY STARS: A propsect so exciting as perhaps to justify the 100-m telescope on its own, is that of the direct detection of earthlike planets of solar-type stars by imaging, out to at least 25 parsecs (80 light years) from the sun, followed by spectroscopic and photometric searches for the signature of life on the surfaces of nearer examples.

We present a concept is for a fully steerable, quasi-Cassegrain telescope in which the 36 m pri-mary and secondary mirrors are parts of concentric spheres. The 15-m secondary, supported 36 meters above the primary yields a 3°, Cassegrain focal surface, 7 meters in diameter. Though the spherical aberration is large (about 2 arcminutes), many small fields are fully cor-rected locally. Multiple small units are placed in the focal surface at regions of interest to cor-rect a field of view of a few arcseconds. These can be used for integral field spectroscopy of for direct imaging using adaptive optics. Hundreds of these units could be placed on the focal surface during the day to allow all-night exposures of the desired regions. We believe that this design offers an economical system that can be dedicated for several important types of astronomical observation.

The speckle image of a dot source is investigated in a focal plane of a large-sized telescope in a window of transparency 8-12 μ. The dot source is created with the help of a corner reflector established on geostationary satellite which is lighted with CO₂-laser. The casual realization of speckle image is generated. Contrasts of infrared (IR) speckle-image and elements speckle image and own geometrical noise of a large-sized telescope are calculated. The own geometrical noise arises because of mistakes in installation of composite mirrors of a telescope. The parameters of IR speckle image and geometrical noise are determined, at which the speckle structure disappears. The estimation of affinity of the “missed” image with diffracted one is received which determines efficiency of adaptation by the offered criterion.

If a collector is very large - square kilometers in size - it may be easier to fabricate flat optics than curved optics. If the primary objective is a flat diffraction grating, microscopic grooves would reshape the incident wave front rather than the 3D surface of a conventional parabolic reflector. I discuss diffractive optics that potentially can cover areas of square kilometers. The geometry of grazing incidence and evanescence allows for very large grating surfaces in ribbons or aggregated segmented modules. Diffraction grating primaries are attractive in that they are well suited for spectrographic astronomy, particularly the Doppler shift studies that have recently been useful in extra-solar planet detection. I propose a unique static mount for ground-based observations where the only moving part is the earth itself. For space-based installations, a new type of membrane telescope is disclosed that can be orbited as a roll of plastic and unfurled over multiple kilometer lengths. Holographic gratings are capable of very fine angular resolution, and only their extreme chromatic dispersion has inhibited their use. The bug is a feature. When taking spectra, chromatic dispersion is useful.

In the polishing process, the wear tends to be greater when the tool extends beyond the edge of the workpiece. A linear pressure distribution (between the tool and the workpiece) has been used to explain this effect, however, this model also can predict negative pressures. This could mean that material is deposited instead of being removed. We present a new pressure distribution proposal, which presents like a skin effect. This means that the pressure is significantly higher at the border points than at internal points of the glass. With this model the material removal at the border points is increased considerably since, according to Preston, the wear is proportional to the pressure. This pressure distribution model is applied to calculate the wear produced by a square tool on a glass border moving along straight lines.

We report experimental results and analysis about a new hydrodynamic radial tool (HyDra, patent pending), which expels a suspension of water and polishing particles radially on glass. With this method it is possible to locally shape optical surfaces. The depth of material removed by HyDra grows linearly with the time. The removal rate is independent of the velocity between the tool and the glass element. The HyDra has been used to fabricate successfully an optical flat and Schmidt surface.

A novel track joint was developed for the azimuth track of the 50-m diameter Large Millimeter Telescope (LMT) now under construction in Mexico at an elevation of 4,600 m. The track, which is 430 mm wide by 230 mm deep, must be flat to within ± 0.3 mm, and the material hardness at least 290 Brinell. This design uses a partial penetration narrow gap groove weld on the top surface of the track and a splice plate welded to the underside of the track. Pre-camber of the joint compensates for weld shrinkage which is small because of the use of the narrow gap groove weld. The residual deviations from flatness are reduced to the required tolerance by adjusting anchor bolts using an optimization procedure. The feasibility of the design with respect to fabrication, strength, fatigue, and alignment was demonstrated by detailed finite element analyses, trial welding and alignment of full scale joints, and testing of the mechanical properties of the joint and adjacent metal.

Future 30m telescopes provide enormous challenges for IR high resolution spectrograph design. The spectrograph collimated beam size will reach ~ 400 mm in order to reach R ~ 25,000 under 0.4 arcsec seeing-limited images. This beam size will push an IR spectrograph volume larger than that of the giant optical echelle spectrograph at 10m telescopes, e.g. the Keck HIRES is 6×6×4 m³ (Vogt et al. 1994). The cost would be enormous considering the entire instrument must be cooled to cryogenic temperatures to be feasible. Here we propose a new kind of IR spectrometer using silicon anamorphic immersion gratings as the main disperser. By operating silicon immersion gratings in an anamorphic immersion mode, the increase in spectral resolving power can be up to a factor of n² or ~12 times at Brewster’s angle (Dekker 1987). Hence, to reach the same spectral resolution, the collimated beam size is reduced to ~33mm in diameter, which makes the design of the instrument relatively easy. The recent breakthrough in silicon immersion grating technology at Penn State has allowed us to routinely fabricate high quality silicon grisms and immersion gratings with sizes of up to 2 inches, <1% integrated scattered light, and diffraction-limited performance thanks to newly developed techniques. Silicon anamorphic immersion gratings with etched dimensions of ~4 inches are being developed at Penn State. The first grating will be available for testing in late 2002. Currently, industry can supply up to 12 inch diameter silicon ingots. We plan to develop a new tool to handle this large grating size in our state-of-the-art nanofabrication facility. A silicon anamorphic grating of this size can provide a seeing-limited (0.4 arcsec) spectral resolution of R ~ 30,000 or diffraction-limited spectral resolution of R ~ 750,000 at 2.2 microns. In this paper, technical issues related to the design of an anamorphic grating spectrograph are discussed.

For future giant telescopes, control of construction and operation costs will be the key factor in their success. The best way to accomplish this cost control, while maximizing the performance of the telescope, will be through design-to-cost methods that use value engineering techniques to develop the most cost-effective design in terms of performance per dollar. This will require quantifiable measures of performance and cost, including: (1) a way of quantifying science value with scientific merit functions; (2) a way of predicting telescope performance in the presence of real-world disturbances by means of integrated modeling; and (3) a way of predicting the cost of multiple design configurations. Design-to-cost methods should be applied as early as possible in the project, since the majority of the life-cycle costs for the observatory will be locked in by choices made during the conceptual design phase. However, there is a dilemma: how can costs be accurately estimated for systems that have not yet been designed? This paper discusses cost estimating methods and describes their application to estimating the cost of ELTs, showing that the best method to use during the conceptual design phase is parametric cost estimating. Examples of parametric estimating techniques are described, based on experience gained from instrument development programs at NOAO. We then describe efforts underway to collect historical cost information and develop cost estimating relationships in preparation for the conceptual design phase of the Giant Segmented Mirror Telescope.

An enclosure design concept is proposed for a large ground-based optical telescope in the 20- to 30-metre class. The proposed configuration differs from the enclosures for existing large telescopes. Current large telescope enclosure designs have inherent inefficiencies which may be substantially magnified if these designs are scaled. Dynamic analysis studies show that motion requirements for the mechanical components of existing enclosures may be too stringent for next-generation enclosures and that these requirements should be revisited. The proposed enclosure design uses a spherical base structure with a rotating inclined cap. This design improves upon some of the mechanical, structural and operational inefficiencies of current spherical enclosures with conventional shutters. The design also offers potential advantages in the protection of the telescope from wind buffeting forces. Wind loading is expected to be one of the most significant factors governing the design of a next-generation large telescope. The enclosure design includes features which are expected to improve the air flow characteristics in and around the enclosure. Preliminary computational fluid dynamics (CFD) studies have been performed in order to analyze the effect of various enclosure details and components on the flow patterns. Future comparative and detailed CFD studies on the enclosure and telescope are proposed. A plan for practical validation of the results of CFD analysis is presented, in order to better understand the benefits of CFD in predicting the effects of wind buffeting on next-generation large telescopes.

Presented is a lightweight mirror compatible with the extremely large telescopes being proposed. The mirror exhibits an areal density of less than 12 kg/m², which is the same relative areal density as the primary mirrors on the 100-meter Robert Byrd Radio Telescope. The panels can be fabricated in 2.5-meters diameter with traceability to 4-meters. Presented are preliminary results of a set of large demonstrator mirrors, their characteristics and traceability to producing 4-meter low-cost and easily fabricated mirror segments for large telescopes.

This paper presents a science case for a wide-field camera operating at submillimetre wavelengths on an optical/infrared 100m Overwhelmingly Large Telescope (OWL). Such an instrument (dubbed "SCOWL" - a Submillimetre Camera for OWL) would offer an unprecedented simultaneous high angular resolution and imaging speed, and play a pivotal role in the future of submillimetre astronomy. Other proposed and existing facilities have either coarse resolution (current single dishes) or are relatively inefficient at carrying out large-scale survey work on a reasonable time scale (interferometers). Furthermore, if OWL is located at a dry high-altitude site, SCOWL will be able to exploit the relatively unexplored 200 and 350μm atmospheric windows with unparalleled efficiency. This is the regime where the emission from cold unevolved sources is at its peak, and is thus of unique scientific interest. By observing in the submillimetre, OWL will produce detailed, unbiased views of objects in formation, ranging from Solar-neighbourhood pre-stellar cores to proto-elliptical galaxies in the early Universe.

The success of giant telescopes is dependant not only in the optics design but also in the mechanics and electromechanics configurations. This paper addresses some of the key aspects of mechanical designs that impact this success. Areas investigated include methods for mitigating pointing errors caused by wind, thermal and gravity deflections; bearing and drive selection; and servo control issues. To reduce the cost of the telescope mount, it is hoped that conventional mechanical designs can be used to accomplish the high level of performance needed in the structure and drives area.

Curvature sensing, unlike some other wavefront sensing techniques, treats the boundary and interior of the aperture of interest in fundamentally different ways. This leads to complications for highly segmented telescopes (100 or more segments), as a significant fraction of points on the wavefront are simultaneously in the interior of the overall aperture and near a segment boundary. Although this effect may make it difficult to implement curvature sensing for adaptive optics on such telescopes, it may conversely be exploited for primary mirror segment alignment. We present numerical simulations which show that curvature methods should be an effective method for segment tip/tilt alignment on highly segmented mirrors.

In this paper I review and reflect on the contributions given at this conference and place them in a broader context. Emboldened by the recent successes of 8 to 10-meter class telescopes and by the success of adaptive optics in making these telescopes diffraction limited, astronomers and engineers are now embarking on the quest for giant telescopes. Are these plans realistic? Are we overreaching ourselves?

Performing Adaptive Optics on a 100m scale is not only a formidable enginnering goal for nowadays. It requires also to look for ideas, concepts, and ways of sorting out technicalities, that are beyond the current visions about Adaptive Optics aboard 8..10m class telescopes. Different physical regimes require to have different metrics when weighting pros and cons of one approach with respect to another. My current, and maybe very personal view, is here exposed, along with the reasons why I think one should face such a target recalling in his mind that the solution can comes together with options that can look irrespective of apparently well established configurations. Substantially, a road--map including Laser Guide Stars whose light is treated as it is the Natural Guide Stars one, is sketched.