The technology and components required to implement discrete adaptive optics systems capable of compensating wavefront errors caused by atmospheric turbulence in ground-based astronomical telescopes are reviewed. Characteristics of the major types of deformable mirrors, wavefront sensors and wavefront reconstructors are described. The effects of device limitations such as the size of the compensation subapertures and the signal to noise ratio of the wavefront sensor detector on the overall performance of adaptive optics systems are discussed. This review indicates that the technology exists to enable conventional adaptive optics systems to perform close to their inherent performance limits, the major impediment being the high cost of the components required. However, a larger problem exists in that the usefulness of adaptive optics for ground-based astronomy is severely limited by external factors such as the small size of the isoplanatic patch and the small photon flux available from most astronomical objects. The conclusion is that new system concepts are needed to overcome these external limitations and to make adaptive optics a useful technique for ground-based astronomy. Among the new approaches that have already been proposed are laser guide stars and multiple wavefront correctors.

Correlation functions of the Zernike-polynomial expansion coefficients of turbulence induced aberrated wavefronts are presented and experimentally confirmed using laboratory-generated Kolmogorov turbulence. These functions are convenient for characterizing the anisoplanatic error for adaptive optics. The isoplanatic patch-size of such systems is evaluable according to their specific working parameters. We find, for example, with a 4 m telescope working at 3.5 μm and sensing the wavefront phase fluctuation on 20 Zernike polynomials, an isoplanatic patch about 22 arcseconds wide.

A kind of intermittent turbulent flows which was observed using a hot wire anemometer and a platinum wire thermometer on tower in Beijing at night has been introduced . This turbulence flow produced by wind shear in the stable stratified flow is consist of some individual or a crowdof cold eddies .The temperature spectral power law of this turbulence lies between -2 and -5/3 so that the C_m value measured by various methods which correlate with scale size of eddies would be different.The mathematical representation of intermittent turbulent flows in the atmosphere was developed. The abrupt variance of temperature at the surface of cold eddies may cause a reflection of radio wave or sound wave , therefore the value of C_m measured by various methods such as radar,scintillometer,thermometer, would be different.

The characterization of seeing degradation induced by near-ground microturbulence is useful not only for selecting the optimum astronomical observatory site, but also for determining the suitable height above the ground for telescope installation. The near-ground contribution to seeing can be established by measurements of stellar image motion, and microthermal activity as a function of altitude. These techniques are applied to several superior sites known to have seeing < 1 arcsecond.

Adaptive optical systems and their applications in astronomy have been discussed for over a decade. Meanwhile the main components like deformable mirrors, wavefront sensors etc. for these real time correction systems of atmospheric turbulence effects, are commercially available. The principles of of this technology, its predicted performance and the current programs underway to implement adaptive optics for astronomical purposes are summarized.

One way to overcome the turbulence perturbations problem on ground based telescopes is to use adaptive optics. The main components of an adaptive optics system developed for several years at ONERA are reviewed here. The different structures of 21 actuators monolithic piezoelectric and electrostrictive deformable mirrors are described. The Hartmann-Shack wavefront sensing technique is presented. The wavefront reconstruction and the mirror control algorithms and processors are mentioned. An adaptive optics test bench associating these components is detailed. Some simulation results dealing with aberrations correction are given.

The paper presents the status of the COME-ON (C GE Observatoire de Meudon ESO ONERA) experiment. This instrument, developed and tested by several European laboratories, is an adaptive optical system with a 19 actuators deformable mirror and a Hartmann Shack type wavefront sensor. The wavefront sensing is performing at visible wavelengths; a special computer drives the deformable mirror which should achieve diffraction limited infrared imagery with large optical telescope. The different components and their individual characteristics are described. The results of the tests of some components are given: 19 actuators deformable mirror, tip-tilt mirror. The expected performances are summarized and possible applications of the instrument to astronomical sources are presented. The isoplanaticity aspect, the required temporal bandwidth and reference source brightness is discussed. The conclusions of the experiment will be used for the design of adaptive optics for the ESO Very Large Telescope.

A number of experiments have been made with two adaptive optical systems. A 19 element deformable mirror with serial hilt-climbing control system is used to correct static wavefront errors. A system with 21-element deformable mirror and real-time wavefront sensing with shearing interferometer is used to correct dynamic errors. With this system two kinds of experiments have been made, in room correction of simulated disturbance and atmospheric compensation over a horizontal path of 340 meters. In these experiments reduction of the speckle pattern to the Airy diffraction disc is demonstrated.

We describe an adaptive optics system designed to obtain diffraction limited imaging in the near infrared (~2 microns) with the Mayall 4-meter telescope at Kitt Peak National Observatory. The basic system consists of a 55 element adaptive mirror, a 37 element Hartmann-Shack wavefront sensor, and a dedicated microprocessor for analysis of sensor data and active mirror control. This polychromatic adaptive optics system utilizes the visible radiation from a semi-stellar source, or nearby star, for wavefront sensing while imaging the near-infrared radiation on an InSb array detector. Recent laboratory tests of the working system are summarized.

In wavefront compensation systems, a compensation cycle includes three basic procedures, namely wavefront detection, generation of correction command and response of the corrector. On considering that the coherence time of the atmosphere is limited, and it should usually take some limited time to complete each of these procedures one after another successively when digital computation is involved, a step-by-step model should be suitable to study the performance of active systems. Based on that model, the statistical optical transfer function (SOTF) of the system is evaluated. SOTF is first associated with the structure function of the compensated wavefront. The latter is then associated with its detection error, the corrector setting error, the spatial response function of atmosphere, under some assumptions. The results are checked for severl special cases. It is found that they are either identical with those published by other authors or quite reasonable and easily explainable.

The concept of curvature sensing and compensation in adaptive optics is investigated by means of a computer simulation. Satisfactory correction is observed when the signal from a 13 element sensor is directly applied to a simulated membrane mirror.

The Johns Hopkins University has recently started a a program to develop adaptive optics instrumentation for astronomy. We have taken an evolutionary approach where the simplest element, wavefront tilt correction, is being developed first, along with a stellar coronagraph designed to fully utilize the higher resolution images, and the more complex elements, to correct for higher order wavefront distortions, will be phased in as they are developed. A wavefront tilt correcting coronograph is now complete and is being commissioned at the Swope 40 inch telescope of the Las Campanas Observatory in Chile. In this paper we describe this system and the initial results of its laboratory evaluation.

Most of the adaptive optics experimented for astronomy uses a Shack-Hartmann wavefront sensor. By their principle these sensors require a large amount of real time calculations. The time spent to perform these calculations may become a limiting factor of the bandwidth. It should be probably the case for the mirrors with high numbers of actuators which will be built for the very large telescope projects. Recently, F. Roddier has proposed an analog curvature sensing method which could theoritically avoid any calculations. We will demonstrate that optical filterings which transform wavefront distorsions in intensity modulations can be used with a reasonable amount of calculations to perform the control of an adaptive optical system. Some of these filtering which permits differential measurements offers the possibility to eliminate the inexpected speckle intensity modulations.

The "Standardized Electodisplacive Transducer" or the SELECT actuator combines a novel fabrication methodology with state-of-the-art lead magnesium niobate (PMN) electrostrictive materials to produce a new generation precision displacement transducer. Historically, PMN actuators have undergone a plethora of material and process modifications to make a more reliable and reproducible device. Actuator designs were also custom for each individual application. The SELECT actuator concept provides a standardized actuator configuration with fixed geometries, fixed active layers number, and fixed process parameters consistent with large scale production. Technical improvements include increased sensitivity, improved uniformity, and larger displacements. In addition SELECT provides significant fiscal advantages. When fully integrated into the production facility, actuator costs are potentially reduced to less than $100 per channel with no long lead materials and a significantly reduced schedule.

Modular adaptive optics provides the means to produce a scalable, state-of-the-art wavefront correction device with both high spatial and temporal frequency. A modular methodology, traceable back to the revolutionary monolithic piezoelectric mirror (MPM), is being incorporated into a new generation of wavefront corrector. The modular approach combines established conventional wavefront corrector technology with technology developed for the electronic microcircuit industry. Electrostrictive lead-magnesium-niobate (PMN), a material which operates at low voltages, dissipates little power, and exhibits negligible hysteresis, is the material of choice for the advanced microactuator development. In addition to meeting the technical performance issues, additional emphasis is placed on providing a design which is easily maintained, scalable, and cost effective.

Using microelectronics, a 37-element hex pattern has been deposited on a thin PZT bimorph consisting of two wafers which produce changes in curvature with voltage applied to individual facets. The fabrication and optical performance of these devices are described, together with utilization schemes for atmospheric wavefront restoration based on rapid curvature sensing and correction.

There are currently a number of groups developing adaptive optics instruments for diffraction limited imaging from ground-based optical telescopes. With few exceptions these programs are intended initially for large (3 - 5 metre) telescopes and, ultimately, the next generation of very large (8 - 10 metre) telescopes. In view of the technical difficulties associated with the large number of elements required for adaptive correction at optical wavelengths most systems will operate at 2 - 5 μm, and employ IR imaging cameras4 as the system detectors.

Accurate measurement of the wavefront perturbations caused by atmospheric turbulence is vital to the operation of an adaptive optics system. Current systems typically use a Hartmann-Shack wavefront sensor to perform this measurement. This type of sensor uses a lens to translate the problem of phase tilt detection to one of spot centroid detection at the lens focal plane. The accuracy with which a CCD array in the focal plane measures the centroid of this spot is analyzed, taking into account the finite pixel dimension, pixel quantity and dead space between pixels. The optimum design of the wavefront sensor is discussed and the required flux from the reference source is computed. If we require the corrected telescope to have a Strehl ratio of 0.85, and assume typical values for the detector parameters, then the required return flux per seeing cell of size r_o is approximately 115 photocounts per measurement.

We discuss that one needs to create an artificial reference source to operate adaptive optics devices at visible wavelengths, since the probability is too low to find a star bright enough in the isoplanatic patch surrounding a given direction in the sky. That artificial source could be the light backscattered from a laser beam emitted through the telescope aperture, e.g. by Sodium atoms in the mesosphere. We describe the ATLAS experiment which is being built to check the validity of the concept of the laser probe technique; we use the 300mJ doubled Nd:YAG laser of the CERGA Lunar Laser Ranging Station and specially developped Shack-Hartmann sensors. The goal is to compare the wavefronts restored from almost simultaneous observations of a real bright star and of a laser spot. The finite distance of the laser source compared to that of the remote star may induce systematic errors in the phase corrections determined from the laser spot and applied to the stellar wavefront. The magnitude of that "cone effect" depends on the telescope diameter, the isoplanatic patch and the wavelength. We propose to overcome it by measuring the 3D map of the ray deflexions in the turbulent layers in the atmosphere. From SCIDAR measurements, the refractive index structure function shows few well peaked features, so that the layers are assumed discrete. The correction of the cone effect requires to observe a small number of laser spots, typically 4 with an 8m telescope. It is a definite advantage to spread the spots outside the projection of the pupil onto the sky, so that the wavefront can be restored for directions largely beyond the isoplanatic patch. This method to correct the cone effect and to widen the isoplanatic domain could be tested with ATLAS and with real multiple stars.

Laser guide stars can be produced by Rayleigh scattering of light emitted by a pulsed excimer laser. The necessary laser output power is calculated to be 78 watts for an adaptively corrected 2 meter telescope working in conditions where r_o ≈18.5 cm. Because an array of laser guide stars must be used in the system, the required power increases as the square of the telescope aperture.

Atmospheric turbulence severely limits the resolution of ground-based astronomical telescopes. In good seeing conditions at the best observatory sites, resolution at visible wavelengths is typically limited to ~1 sec of arc. During the past 15 years adaptive optical systems using electrically deformable mirrors have been developed to compensate for turbulence effects. Unfortunately, these systems require bright reference sources adjacent to the object of interest and can only be used to observe the brightest stars. Artificial guide stars suitable for controlling an adaptive imaging system can be created in the upper atmosphere by using a laser to excite either Rayleigh backscattering or the mesospheric sodium layer. The design requirements for an adaptive telescope utilizing Na laser guide stars are discussed in this paper and those for the Rayleigh guide stars in a companion paper. The brightness of the Na guide star depends upon many factors including Na abundance, laser bandwidth, laser pulse energy, pulse length, zenith angle and seeing cell diameter. Based upon our initial experiments at Mauna Kea Observatory and our detailed theoretical calculations, it should be possible to obtain near diffraction limited images for zenith angles down to 30° from ground-based telescopes as large as 10m in diameter with a laser power of less than 30 watts.

Saturation effects may severely limit the brightness of guide stars created in the mesospheric sodium (Na) layer. Saturation arises when the laser energy density is large enough to significantly alter the population densities of the atomic states within the layer. These altered state populations lead to nonlinear absorption of the laser energy thereby producing in a reduced rate of fluorescence and an increased rate of stimulated emission. The level of saturation is determined by the laser's pulse length, pulse energy, beamwidth and linewidth. We have quantified the saturation effects in terms of these laser parameters and have formulated design equations which will allow us to design a laser capable of achieving a specified guide star brightness while at the same time minimizing the power and pulse length requirements. Our calculations show that a Na laser having a pulse energy and pulse length on the order of 106 mJ and 69 μs respectively and an linewidth of 600 MHz will produce a guide star at zenith which is bright enough to drive an adaptive optics system with a seeing cell size of r_o = 18.5 cm. This energy corresponds to a laser power of 21 watts at 200 pulses per second.

By the simultaneous observation of the atmospheric wavefront of a number of adjacent objects in the sky it appears possible to determine the atmospheric wavefront distortion in 3 dimensions using tomography techniques. Once obtained it becomes then possible to correct the atmospheric wavefront distortions in detail using adaptive optics at the conjugates of a number of atmospheric layers.

Long baseline stellar interferometers are very large optical systems that have extremely tight mechanical tolerances. In the Mark III interferometer, there were two major active systems that maintained the alignment and optical path control needed to make astronomical measurements. One system, called the star tracker system, kept the two interfering wavefronts parallel. The other system, the fringe tracker/delay line kept the optical paths equal in the two arms of the interferometer. This paper reviews the operation of the Mark III active systems and introduces systems that will be built into the next generation six element imaging array being designed. The key difference between the two systems is that with a two element interferometer, light from the star must be split for use in the star tracker and fringe tracker while in interferometers with more than two elements, it is possible to use the same photons to run both the star tracker and fringe tracker.

We present the ASSI (Active Stabilization in Stellar Interferometry) interferometric experiment presently under realisation for use with the 2-Telescope interferometer of CERGA (I2T). A detailed presentation of the instrument project was given in ref.¹, and progress on the realization were indicated in ref.². We review the principles and expected performances of our system and, in particular, the real time, flux optimized, fringe tracking system that we developped using synchronous detection technique on cooled silicium diodes in the near Infrared.

The Meudon Observatory NIR 4.2-m light collector, which incorporated a primitive form of active optics, was con-ceived, built and successfully tested between 1967 and 1974; then, the entire project was scrapped, and lies wholly forgot-ten today. Since a complete description has been given (together with a picture of the finished system, on the July 77 cover of Applied Optics), there is no need to present here more than a brief summary. We mostly want to draw a few les-sons from an undertaking that has obviously turned out to be a technical success, but a scientific failure.

In a Cassegrain telescope with a segmented primary mirror, wavefront aberrations are introduced by segment figure errors, segment positioning errors, secondary figure errors, and secondary positioning errors. We give here an analytic description of these errors. The various sources contribute different aberrations to the wavefront. These differences suggest an ordering to the aberration measurement and correction, and allow options in alignment procedures and tolerances. The segmentation and control of the primary mirror segments allow the partial correction of figure aberrations in the secondary mirror and aberrations caused by secondary misalignment. We use the W. M. Keck Ten Meter Telescope as an example for these considerations. The ten-meter diameter actively-controlled primary mirror is a mosaic of thirty-six hexagonal segments and is currently under construction.

The Keck Observatory primary mirror is comprised of 36 segments from which 216 glass temperatures and 168 measures of relative displacement are continuously read. Mirror figure adjustments are under the active control of 108 adjustable-length actuators, three per segment. We describe our collection method for mirror segment displacement data and glass temperature data and the supplying of individual control signals to the actuator controllers and sensor electronics, both located near each mirror segment, and discuss the mirror's computerized control system. Much of the sense and control electronics is close to the rear surface of the primary mirror. From an historical perspective, we describe the impact of the need to minimize heat sources and cabling, on the communications method used and on its implementation. We conclude with a brief structural overview of the hardware in use to actively maintain primary mirror figure.

Each of the 108 hydraulic actuators on the Keck ten meter primary mirror lengthens or shortens under electronic control in steps of about four nanometers. The starts and stops of longer moves are profiled to help prevent the stimulation of mirror system mechanical resonances by abrupt actuator length changes. At the move endpoint, the actuator is kept still to prevent the pumping of periodic energy into these resonances. Endpoint stiffness is maintained by electronic feedback. The combined power dissipation of the actuator and its local control system is constrained by the need to minimize heat generation in the vicinity of the mirror. In this paper, we discuss observed stictional and other mechanical phenomena associated with very short bi-directional actuator length changes and trace the impact of the various constraints on actuator controller design and implementation. We present some results of tests with an actuator controller driving an actuator which.in turn drives a displacement sensor. The units under test are production devices to be used in the Keck primary mirror active figure control.

Improvements in state-of-the-art focal plane sensors mean that the next level of performance improvement for astronomical instruments must come from an increase in the aperture. While both monolithic and segmented mirrors are being considered for ground-based systems, weight and volume constraints make segmented reflectors the only practical approach for space systems. Segmented reflectors require an active segment-alignment control system in order to make the reflecting surface have the optical performance of a monolith. This paper describes an experimental apparatus developed at the Lockheed Palo Alto Research Laboratory called the Advanced Structures/Controls Integrated Experiment (ASCIE). The ASCIE consists of a Cassegrain optical configuration with a 2-m seven-segment actively controlled primary mirror supported by a light, flexible truss structure. The ASCIE is a response to the need for experiments that can simulate the complex dynamic behavior of a large structure and which address the myriad problems associated with precision control of optical surfaces. This paper describes the ASCIE test bed, presents details of the control and optical measurement systems, and reports on preliminary performance results.

Wind buffet deformation of large flexible mirrors is, from its frequency bandpass and other characteristics, an overlap area between normal low bandpass "Active Optics" and high bandpass "Adaptive Optics". The fundamental problem posed is that the time frequencies involved do not permit the elimination of external seeing effects by appropriate time integrations with the image analyser, as in normal low bandpass closed-loop operation. Confusion between the wind deformation and atmospheric functions results. For wind buffet correction in these circumstances, we analyse the possibilities and limitations of the closed-loop principle using one or more image analysers in a number of configurations. Some important parameters of the atmospheric function apear to be insufficiently known and should be measured experimentally if rapid practical progress is to be made. Cross-checks betweens closed-loop and open-loop methods in a two-pronged approach seem highly desirable.

The idea of a "New Technology Telescope" was born in 1979. At this time, however, the only significant new technology feature envisaged was an Alt-Az mount. By 1982 the design concept had evolved much further to include not only the major new technology feature of active optics but also a new building concept derived from the experience of the MMT building. In this year, Switzerland and Italy entered ESO enabling the financing of the NTT project with DM 24 million. It was thereby defined as a telescope of 3.5 m aperture, exploiting the most advanced technology to achieve at the same time not only much reduced costs compared with the conventional ESO 3,6 m telescope but also an optical performance pushed to such limits that even the best conceivable seeing could always be fully exploited.

We start off with the 1 meter experiment. The principles of active optics and the results of the 1 meter experiment are described in detail in three papers 1, 2, 3. The support geometry is identical to the one of the NTT primary mirror. The 78 actuators, 75 of them astatic levers with movable counterweights, are located on four rings with three fixed points on the third ring. The mirror is a spherical mirror of 1 meter diameter and a thickness of 19 mm.

A concept of active structural control will be established for large optical telescopes to improve their mechanical behavior. The description of the concept starts with the mechanical design which is based on new construction principles. The mechanical performances achieved by passive means alone, will be presented first. The next step will introduce active components to perform the tight tolerances in tracking and aligning the optical components. A structured control scheme of independent loops is elaborated and discussed in some details.

Optical telescopes of unprecedented size, high resolution optical arrays having large number of telescopes and more performant auxiliary instrumentations will be built by strongly applying the concept of active optics. Taking apart developments of adaptative optics relative to seeing compensation, the meaning of "active optics" must be considered as having at least three different senses i) maintaining, in situ, the best figure of the telescope optics during astronomical observations, ii) figuring highly aspheric optics and/or unaxisymmetric optics, and iii) modifying, in situ, the optical figure parameters as, for instance, the asphericity of a Cassegrain mirror for a better field optimization with respect to various focal instruments. This paper discusses some thoughts on these questions and presents some results and astronomical applications.

The control system for the Multipurpose Multiple Telescope Testbed (MMTT) is described. The phased array consists of four 20 cm telescopes, and has a 30 arcminute field of view. The maintenance of lateral pupil geometry to within one micron of the ideal is the key feature of the control system. Two mirrors behind each afocal telescope combine four beams into one, and are actuated to control both tilt and piston. Together the two beam combiner mirrors control lateral pupil geometry. The optical sensor system slowly scans the telescope field of view while monitoring the piston error verses field angle. A discrete-time-varying Kalman filter then processes these measurements to estimate lateral pupil geometry errors. The controller integrates the Kalman filter estimates, along with the x and y tilt measurements which are treated deterministically. Multiplication by an axis separator matrix converts the control input commands into appropriate piezoelectric transducer actuator commands. The entire control algorithm is implemented on a high speed array processor board. Preliminary test results show that the control system is accurately controlling lateral pupil geometry for small field angles. For larger field angles, the system was unstable due to erroneous tilt error measurements. The source of this error is discussed and a solution is proposed.

The phased array telescopes have been an active area of research at the US Air Force Weapons Laboratory over the years. Its function is to synthesize a single aperture from separate optical systems combined and properly controlled to form a synthetic aperture. When the aperture is synthesised electronic control is needed to position optical mirrors at closed loop frequency of greater than 1kHz. The system considered is a 14th order and includes an optical path difference adjuster (OPDA), a high voltage amplifier (HVA), a compensator and a digital position monitor (DPM) which itself includes an 8th order digital filter. The system disturbances such as wind and deterioration of components would often lead to changes in the plant parameters which would not normally by tractable by a constant gain controller. Therefore, a new control philosophy which can cope with plant's unmodeled dynamics is employed for continuous maintenance of the telescope's performance. For this purpose, adaptive control has been applied to continuously change the gains of a controller. This control philosophy is known as model-referenced adaptive control which is based on an "ideal" behavior of the system called "reference" model. Then through an iterative mechanism, called adaptation algorithm, the gains of the controller is adjusted. This paper describes a software and hardware implementation of a model-reference adaptive control for the phased-array telescope. Simulation results are presented using MATRIXx package on a VAX 11/780 computer. Theoretical development and experimental results of this new control approach for a telescope systems will be given here.

A concept for an extremely lightweight primary mirror for a large space telescope is proposed. The mirror uses a sandwich structure composed of aluminum faceplates with an aluminum foam core. Modal analysis indicates that a fundamental frequency of 100 Hz can be obtained for a 6 meter diameter mirror with a tapered back and having a total mass of about 3500 kg. An array of thermal actuators embedded in the mirror corrects for figure errors. Such an active thermal figure control system does not have a quick response, but is well suited to space applications where disturbances have relatively long time constants. The advantage of thermal actuators is very high reliability, a primary concern in any space system requiring a large number of actuators.

The Precision Segmented Reflectors (PSR) project is described in this paper. The project, which involves the close cooperation of two NASA Centers, represents a first step toward developing the technology base needed to support future advanced astrophysics missions. The focus of this project is to develop a lightweight, low cost option for building large reflecting telescopes that can be assembled in space. Four major technology areas, i.e., reflector panels, structures, controls, and system integration are being developed. A principal driver for PSR technology development is the proposed Large Deployable Reflector (LDR) telescope. The technologies being developed under the PSR program are, however, generic and can be applied to many future missions such as optical communications, optical interferometers , and other missions requiring large reflectors.

To better understand the relationship between active, segmented precision reflectors and manufacturing and alignment errors, a FORTRAN routine has been written for the Precision Segmented Reflector (PSR) and Large Deployable Reflector (LDR) programs that simulates these errors and provides optical path difference (OPD ) input for CODE V ^TM through the 'ADD' option in the pupil map and point spread function features. Diffraction analysis,including point spread function and encircled energy can then be performed. The manufacturing errors are input for each segment in terms of 37 Zernike terms whose origin is the segment's center. Alignment errors are input for each segment in terms of the six degrees of freedom, dx, dy,dz, α,β, γ. Angles are defined with respect to the segments center. The program works in two modes. The first mode simply looks at the effects that the errors have on image quality without any compensating parameters. The second mode uses three degrees of freedom to help compensate for the effects of the other errors. In this mode, dz, α and β are used to minimize the sag difference between the real and ideal parabolic surface to insure proper phasing of all mirror segments. This paper discusses the methodology and results of this endeavor.

The 20-m Large Deployable Reflector (LDR) will have a segmented primary mirror. Achieving diffraction-limited performance at 50 μm requires correction for the errors of tilt and piston of the primary mirror. This correction can be obtained in two ways, use of an active primary or correction at a de-magnified pupil of the primary. A critical requirement is the means for measurement the wavefront error and maintaining phasing during observation of objects that may be too faint for determining the error. Absolute phasing can only be determined using a cooperative source. Maintenance of phasing can be done with an on-board source. A number of options are being explored as discussed below. The many issues concerning the assessment and control of an active segmented mirror will be addressed with early construction of the Precision Segmented Reflector (PSR) Testbed.

Light weight, high precision, low cost structural composite mirrors have tremendous potential for enabling affordable space telescope systems. The Large Deployable Reflector (LDR) is an example of such a system. It is a 20 meter diameter, earth orbiting submillimeter telescope. Its technology requirements are for panels that are from 1 to 2 meters in size with areal densities of 5 to 10 Kg/m² and surface figure precision of a few microns. JPL and the Hexcel Corp. have entered into a joint technology activity, sponsored by the NASA Precision Segmented Reflector (PSR) Program, for the development of such mirrors. Highly specialized manufacturing and materials processing techniques have been developed by Hexcel for the production of high precision, light weight and low cost composite mirrors. JPL has developed an analytical simulation capability for composite mirrors that characterizes their mechanipal and thermal performance in terms of the materials properties and configurations. This capability is the basis of detail panel designs for thermal stability, test simulation, test/analysis correlation and projection of performance for specific applications. This combination of capabilities from both organizations has resulted in the development of graphite/epoxy mirrors up to 1.0 meter in size with surface precision of a few microns rms while weighing only 6 Kg/m². This paper describes that development program. The PSR Panel Program, over a four year period is for mirrors up to 1.5 meters with surface precision and LDR orbital thermal stabilities on the order of one micron.

An increasing number of large aperture telescopes and antennas are being designed to meet a wide range of scientific objectives in both terrestrial and orbital environments. Because of the large scale of these instruments, studies have shown that materials and structures drive the design of these systems. Mass, surface figure, and overall dimensional stability are critical for most subsystems in these applications. Accordingly, light weight, low cost, structurally stiff materials are the most attractive candidates. To address these reflector technology challenges JPL has initiated directed research that has led to major technology programs whose objectives are to evaluate, advance and apply structural composite materials technologies to high precision, thermally stable, lightweight mirrors. JPL has taken a balanced, iterative approach that combines analytical simulation, hardware fabrication, and mirror performance testing. This paper will focus on the computer models, mirror performance evaluation and their correlation.

Progress in the area of precision segmented reflectors (PSR) for space applications has been driven by panel development activities. A number of small to medium size panels have been fabricated to demonstrate feasibility. The primary emphasis to date has been on making the panels lightweight and with high precision surfaces. Because of this emphasis, composite materials, and in particular, graphite fiber reinforced epoxy materials, were an obvious choice for construction of the panels. Indeed, many of the panels in existence have been fabricated from materials of this type. In terms of space applications, however, where stability and durability are of concern, it appears that these materials may not possess the balance of properties required for long term missions. Testing of these materials in simulated use environments has shown that they are deficient in either their thermomechanical properties, their stability (dimensional and environmental) or their fabricability. Recognizing that the development of new or modified materials for panel construction will be imperative if PSR technology is to be utilized in long term space missions, a program has been initiated to achieve this end. An initial set of material requirements has been developed based on a variety of mission scenarios. It is clear that no one material will be applicable to all such missions. Some potential candidate materials or generic classes of materials have been identified. Because of the basic requirements of light weight and stiffness, attention has continued to focus on advance composites. In the near term, it is likely that modified fiber reinforced organic matrix composites have good applicability. In the longer term materials such as carbon/carbon composites, graphite/glass composites and metal matrix composites may ultimately provide the best mix of properties. In this paper, information on PSR panel material requirements, as well as, test data on state-of-the-art materials will be presented. In addition, potential candidate alternate materials and progress toward their development will be discussed.

Precision Segmented Reflector (PSR) technology is currently being developed for a range of future applications such as the Large Deployable Reflector (LDR). This paper outlines the structures activities at NASA Langley Research Center in support of the PSR program. Design concepts are explored for erectable and deployable support structures which are envisioned to be the backbone of these precision reflectors. Important functional requirements for the support trusses related to stiffness, mass, and surface accuracy are reviewed. Proposed geometries for these structures and factors motivating the erectable and deployable designs are discussed. Analytical results related to stiffness, dynamic behavior, and surface accuracy are presented and considered in light of the functional requirements. Results are included for both a 4-meter-diameter prototype support truss which is currently being designed as the Test Bed for the PSR technology development program, and for two 20-meter-diameter support structures. For the most likely ground support conditions, the maximum gravity-induced deflection of the Test Bed support truss (with 10 kg/m² panels) was determined to be approximately 50 μm, and the rms surface error was 12 μm. For the same support conditions, the Test Bed fundamental frequencies were between 30 Hz and 40 Hz. It is shown that if the secondary optical system is supported by a simple tripod design, the first six vibration modes are likely to be dominated by the secondary system. The 20-meter-diameter support trusses were found to be quite stiff for structures of such large size, having maximum deflections on the order of 0.35 mm in a 1-g environment. When considered as part of a reflector system, these support trusses had maximum deflections of 6-11µm under slewing loads, and free-free fundamental frequencies of 6-8 Hz.

The Precision Segmented Reflector (PSR) represents an initial step in the development of enabling technologies for large, lightweight segmented optical telescopes. It is a technology development program leading to a ground-based demonstration of the design, fabrication, and assembly or deployment of a 5-m class segmented optical telescope, intended to achieve an initial alignment of the segments comprising the primary reflector surface, and maintain that surface figure in the presence of environmental disturbances (principally thermal). To achieve these objectives, the structure which supports the primary optical surface in the PSR system must meet stringent mass, stiffness, thermal stability, and fabrication precision requirements. The structure which is designed must be fabricated to an initial precision on the order of ±100 microns across its 3.8 meter surface, and must maintain that precision in the presence of a nominal temperature gradient representative of orbital operating conditions. It must additionally provide three stable points of support to each of 18 hexagonal composite reflector panels, and have removable members to accommodate the introduction of vibration suppression elements. The derivation of the design requirements for the PSR structures is discussed, as well as the approaches being used to analyze these systems, and plans for ground-based testing.

Ground-based testing of large, precision segmented optical systems poses many unique challenges. Among the factors complicating this process is the extreme sensitivity of the system to environmental disturbances typical of the test laboratory. Floor to ceiling thermal gradients, micro-seismic vibrational disturbances, and acoustic noise become increasingly significant influences in the testing of these systems as the dimensions of the optical elements extend into the 10 meter range. An analytical model of the Precision Segmented Reflector (PSR) Test Bed has been used to investigate the influence and significance of known or anticipated test environmental conditions. Results of these analyses and their implications for ground-based testing of large, precision segmented optical systems are presented and discussed.

The development of a lightweight one meter composite mirror that can be controlled and adjusted on-orbit, is presented. The data in this paper show that long wave distortion errors can be corrected by using embedded piezoelectric ceramic actuators. The proposed concepts were verified by both mathematical simulations and laboratory experiments.

This paper describes an ongoing effort at the Jet Propulsion Laboratory on the vibration suppression for the Precision Segmented Reflector backup structure. The effort is centered on the vibration damping augmentation through a system consisting of active and passive damping members. An active member is a structural member with built-in piezoelectric actuator and sensors. A bridge feedback technique developed in the communication engineering is applied locally to the active member for active damping augmentation. An efficient method was developed for optimal placement of active and passive damping members in the truss-type backup structure. A simple synergistic model between the active and passive damping was proposed based on a weighted energy dissipation criterion. A baseline passive member design with constrained viscoelastic material treatment was used as the source of passive damping.

This paper describes an advanced technology figure control system for a generic class of large space based segmented reflector telescopes. A concept definition study was recently carried out at the Jet Propulsion Laboratory to develop the sensing and control mechanization for the actively controlled primary reflector. The reflector segments are made up of hexagonal composite panels individually pre-assembled to a panel control module that contains the sensing, actuation and control electronics hardware for each segment module or "unit cell." Unique features of this integrated segment control module include a completely determined kinematic mounting to the underlying truss structure, inherent attenuation of truss vibrations, prevention of load transfer to mirror panels and reaction moments to the truss nodes, three degree-of-freedom (piston and tilts) active position control, and passively restrained in-plane lateral and rotational motion. The unit cell physical architecture also provides a reference station for interferometric sensing of relative motion between adjacent mirror segments, a mounting for linear electrodynamic actuators, and forms a robust, readily transportable unit which can be attached to the truss during space assembly without risk of damage to the figure control mechanisms and mirror segment. Performance requirements are derived in part from the Large Deployable Reflector, which is a representative mission, and error allocations are made which consider mirror panel surface errors, position measurement and figure estimation, and position control of both quasi-static and dynamic disturbances. Major technology and design motivations for selection of sensing, actuation, and mechanism approaches result from the high precision and very low mass and power goals for the reflector system. Finally, analysis of performance, mass, and power for the unit cell mechanization are shown to be compatible with these objectives.

This paper presents an overview of the control analysis activity related to the development of figure control technologies for large space telescopes with precision segmented, actively controlled, primary reflectors. The issues addressed here involve the development of geometric and dynamic models, characterization of figure estimation errors and optimal sensor placement, and the development of quasi-static and dynamic control concepts. The most significant simulation results obtained during an extensive performance evaluation are also presented.