Telescope structures are described, beginning with Galileo's hand-held telescope, and continuing the evolution to the modern Multi-Mirror Telescope, to the Keck Telescope now being built, and to the Large Deployable Telescope of tomorrow.

The six Cassegrain telescopes which comprise the MMT are held in place by a steel space frame called the optics support structure (OSS). The design and development of this structure, from concept to final implementation, is presented in the context of the history of the MMT project.

The Multiple Mirror Telescope is a 150 ton optical instrument composed of an array of six 1.8 m diameter folded Cassegrain telescopes. A mount for the array uses conventional ball bearings and gear trains to permit the telescope to track in two axes, azimuth and altitude. A computer-controlled servo system uses a 26-bit absolute encoder and associated electronics to achieve an all-sky pointing accuracy approaching 1.0 arcsecond RMS, and a tracking accuracy approaching 0.1 arcsecond RMS.

The Multiple Mirror Telescope is an array of six 1.8 m diameter folded Cassegrain telescopes used to achieve the collecting area equivalent to that of a telescope with a single 4.5 m primary mirror. For most applications, the six images are co-aligned in such a way that overlapping images appear to result from a single telescope. Another application requires that the wavefronts from all six telescopes be co-phased within 1 micron, concurrent with co-alignment. This paper describes the successful operation of the co-alignment and co-phasing capabilities, the control system and devices used in the applications, and test results to date.

The Columbus Project's goal is to develop and construct an 11.3 meter telescope by the 500th anniversary of the discovery. of American, 1992. The consortium that was formed to complete this task with this optimistic time schedule includes the University. of Arizona, Ohio State University, Arcetri Observatory (representing the Italian Telescope Group), and the University of Chicago. The mirror work being done by Roger Angel at the University of Arizona is the major defining parameter. The telescope is foreseen as being constructed of two 8-meter diameter, F:1 primaries with a 14 meter center separation.

Using computer-aided engineering (CAE) tools and techniques for the design of a space optical system, from preliminary design through testing of hardware, provided an excellent means for compressing project schedules. The three principal steps in the process were (1) solids modeling, (2) structural analysis, and (3) design test. Geometric solid modeling requires that the components of the system be described expli-citly in the CAE database. In this case the components consist of a number of electronic chassis and their associated circuit boards, the optical head, the gimballed mirror and the optical bench. The CAE database was used to support: (a) checks of layouts and parts interference, (b) development and modification of designs, (c) fabrication drawing preparation, (d) computation of properties for further computer analysis, (e) project team communication, and (f) generation of presentation aids for customer communication.

The WF/PC is designed for use on the Hubble Space Telescope (HST). It has complex optics with tight alignment and design requirements. The camera is subjected to various static, dynamic and thermal environments during its ten plus years life. To ensure the camera meets its design objectives and functions properly on orbit, an extensive structural analysis was performed using MSC/NASTRAN. This paper describes the camera, its analysis, and shows the versatility of the finite element method applied to an optical system.

A variety of approaches have been proposed in recent years for construction of Solid Optics. After a brief review of the goals and history, we describe a new, systematic approach. The new approach involves the total removal of air gaps from the system. This requires "solid air" for light propagation. Solid air, in turn, behaves differently from the more familiar air. In particular, it has a larger index of refraction and more dispersion. Modular, highly manufacturable systems approaches are described.

Light power attenuation occurs in deformed fibre. We use this measure pressures applied into one or many fibres without any additional mechanical. Different configurations could be considered : - an elementary loop - a surface sensor woven with a single fibre - a surfacic sensor woven multiple fibres. So we obtain a flexible sensor insensitive to bending with and sensitive to pressure like a skin. The different configurations emerge on different applications - binary working (presence detector reconnaissance shape) - continuous working (pressure may- strain sensor).

Future electro-optical space systems require low ambient vibration response levels in the presence of numerous broad and narrow-band disturbances. A structural design methodology which addresses this vibration suppression problem is devised. Dynamic response is described in terms of random vibration theory. Input power spectral density is represented compactly in terms of its poles and zeroes. Output power spectral density is integrated in closed form using residue theory to obtain root mean square re-sponse. An optical performance index based on this response is minimized subject to a constraint on structural weight. That structural design which provides the best combination of stiffening, tuning and mode shape changes is thus employed. Large gains in optical performance with little or no weight penalty are demonstrated.

The optical pathlength differences for complex variations of index or geometry due to temperature, moisture, or stress effects can be approximated for axial beam optics using existing structural finite element computer codes. Existing pre/post-processors for finite element codes can be used to conveniently display and interpret the results.

A method has been developed to least-mean square fit Zernike polynomials to optical component deformations during finite element structural analysis. This procedure uses a pre-processor to MSC/NASTRAN and allows the polynomial fit to be calculated during dynamic response as well as static loading.

With the proliferation of digital computer capabilities the industry has developed many different ways to analyze optomechanical systems. These methods can be roughly grouped into three categories; 1) classical closed form analysis, 2) data base management analysis and 3) optical analog analysis. This paper describes the three groups, discusses their applications and compares their performance on the bases of cost, schedule and accuracy. The evaluations are based upon the author's experience on four recent projects which are used as benchmarks for the comparisons.

Lockheed's approach to studying Large Space Optical System (LSOS) vihration suppression is to assemhle an integrated optics/structures/controls simulation and test the performance of the primary control system, high authority control/low authority control (HAC/LAC) against a defined suppression requirement. This simulation featured all of the key elements of an integrated suppression system analysis including: (1) partially validated finite element model; (2) requisite optical performance algorithms, and (3) closed loop HAC/LAC model. The integrated system analysis demonstrates that HAC/LAC meets specified requirements of -40 dB and supports the choice of HAC/LAC as the primary control system. Further, data from this analysis was used to show that selected actuators are of minimal weight (0.5 lb), will require eight locations, and do not cause significant changes in either modal frequencies or shape. The analysis showed: (1) eight actuators give jitter attenuation of 143 and 253 in x, y directions; (2) largest force level required ~1.0 lb; (3) HAC used 25 of 44 modes, and (4) LAC was not required for stability.

The control problem examined is to stablize a large, spaceborne Cassegrain telescope. Modal gain factors and known characteristics of disturbances are used to determine which structural modes affect line-of-sight (LOS) the most and are candidates for active control. The approach is then to: (0 actively control/maintain alignment of optical components; (ii) place structural control actuators for optimum impact on the selected modes for active vibration control; (iii) feed back the best available estimate of LOS error for direct LOS control. Local analog loops are used for high BW control; multi-variable digital control for lower BW control. The control law is synthesized in the frequency domain using the charac-teristic gain approach. Robustness is measured by employing conicity, which is an outgrowth of the positivity approach to robust feedback system design.

A design for the active structural control of a two-element laser beam expander is described. The lengths of selected struts of a trussed, quadrapod graphite/epoxy structure are monitored and controlled by sensor/actuators using a modal control technique with rate-damping enhancement. The details of structural design, optical arrangement, strut sensors and actuators, and control algorithms are discussed.

This paper will discuss control laws for a large space antenna. The discussion will focus on analysis of the multivariable, closed loop system with respect to nominal performance, robust stability, and robust performance. The analysis employs singular values and structured singular values of multivariable frequency responses. Consistency between the frequency response analysis and time simulations is also presented. A simplified model of a large flexible space antenna was used for the analysis.

This paper reviews an optics/vibration experiment involving the Harris Deployable Multi-Hex Prototype - a reflector structure comprising a seven-panel array. Design considerations needed to ensure the emulation of vibration pathologies characteristic of SDI systems are discussed. The key aspects of dynamic complexity, deployability and proper combination of generically distinct vibration suppression methods are emphasized. We describe the experimental setup which follows as a consequence of these considerations. The projected test plan emphasizes combined orchestration of active and passive methods.

One of the key parameters usually considered in a light weighted mirror design is its flexural rigidity. In part I of this paper, the relative merits of symmetric sandwich and open back light-weighted designs in the 10 to 30 kg/m² range (of mirror weight per unit area) are examined for ULE, Heraeus fused silica, Zerodur and beryllium, subject to some representative state-of-the-art fabrication constraints. Part II of the paper addresses the same issue independently of any specific material. This is made possible by normalization of several parameters by the weight equivalent solid thickness of the mirror. Also included in the study are nonsymmetric sandwich cross sections with unequal front and rear facesheet thicknesses. The study shows that, for given values of mirror weight per unit area and the core (or rib) solidity ratio, there exists a transitional value of the normalized front facesheet thickness above which an open back configuration is structurally superior to a sandwich configuration. Moreover, for the increasingly demanding light-weighting goals of moderns optical systems, the optimum geometrical configurations may be beyond the fabrication capability of the current state of the art.

The design of high-performance quality lightweight mirrors necessitated by payload requirements must be shown to be capable of resisting environmental-load-induced distortion. Such loading can include thermal gradients in the presence of flux loading, or extreme thermal soak in the cryogenic environments demanded by IR systems. Additionally, for aircraft systems, the optics may be subject to a changing gravitational vector, causing performance error. For orbital systems, gravitational error is a major concern as well, as it is necessary to perform meaningful ground tests prior to the zero-g release condition. These mirrors must exhibit excellent stiffness and thermal expansion characteristics, particularly in a passive system, and often in an active system as the mirror size increases but the aerial density requirement does not. To meet the stringent requirements, analyses for mirrors of various sizes, both solid and lightweight, are presented to show the effects of material properties and inhomogeneities on performance characteristics in the presence of a thermal and gravitational environment. Included is the effect of kinematic mount location and coefficient of thermal expansion uncertainties. Passive and active focus performance is compared, and design points are indicated for actively controlled deformable mirror requirements.

Classical precision mirrors in optical systems are right circular glass cylinders with a 6-to-1 diameter-to-thickness ratio. The typical weight of a mirror of this type is W = 255 x D2.92 where: W is the weight in kg, D is the diameter in meters.

This paper presents a closed form solution of E. Reissner's shallow spherical shell equations for the thermal bowing of a curved circular mirror, which is kinematically supported and subjected to a generalized nonuniform thermal load distribution. A variety of nonsymmetric functional forms of thermal moment and thermal force distributions can be constructed from the assumed form of the thermal load distribution. Several test cases of the closed form solution are compared with independent, Nastran based, finite element solutions. The comparison shows excellent correlation between the two methods in all cases. However, the closed form approach is significantly more efficient, convenient and cost-effective. The closed form solution is applied to the development of normalized thermal deflection curves and related RMS figure errors in curved circular mirrors. The solution and the data are also applicable for the thermal bowing component of the thermal distortion of light-weighted curved circular mirrors.

The support of mirrors often requires bonding inserts into holes in the mirror material. Temperature change and shrinkage of the bonding material and temperature change of the insert produce stresses that deform the optical surface. Using both classical analytical methods and finite element techniques the deformation of plates and shells due to temperature and shrinkage effects are computed. The depth of the hole can be chosen to minimize the deflection of the optical surface. A variety of geometries which include hole diameters both large and small compared to the plate thickness are studied. In addition to the general formulation of the problem, some quantitative results for the Keck Observatory Ten Meter Telescope mirror segments are presented.

The National New Technology Telescope program at NOAO is developing technology to be used in the design of a 15-meter multiple-mirror telescope. One configuration being considered for the primary mirrors is a lightweight honeycomb sandwich structure cast from borosilicate glass. This material has a relatively large coefficient of thermal expansion (about 3 parts per million per degree C) which can produce optically-significant distortions when the mirror is exposed to a typical observatory thermal environment. To study this problem we have tested a prototype borosilicate glass mirror, 1.8 meters in diameter, in various thermal environments. Finite-element modeling was used to predict the thermal distortions of the optical surface. This led to an analysis in which temperature patterns were described by a least-squares fit to a polynomial expression, and the polynomial was then used to predict nodal temperatures of the model. The individual terms of the polynomial describe patterns of temperature such as linear-diametral gradients, radial gradients, etc. The optical distortions resulting from these cases are described in this paper. An analytical solution for a 3-dimensional linear temperature gradient was developed to check the finite-element results, and is also presented.

The cyclic fatigue behaviour of a set of light guiding glass rods is studied and the results are compared to data of soda lime glass rods.The rods were excited uniaxially in a high frequency (ultrasonic) system. A statistical staircase method is described and was used to compare the strengths of the soda lime and light guiding rods.Residual stresses in the light guides were determined from fractographic measurements and their possible influence on the observed lower strenght of the light guides is discussed.