Frequently encountered difficulties in large optics fabrication projects are discussed with emphasis on four generic types of problem areas. Scaling, innovative techniques, optical uniqueness and specifications are covered. Insights into how problems arise in these areas and how they affect fabrication, project management and final project results are given. The basic conclusions of the paper are that you will probably: (1) underestimate the difficulty of the scaling problem, (2) place too much emphasis on an innovative new technique which probably won't work, (3) take longer to complete the project than esti-mated and (4) become heavily mired down in setting, interpreting or meeting specifications. You will do all of these things even if you read this paper, but hopefully, to a lesser degree.

The successful development of space-based infrared surveillance systems and space-based lasers is predicated primarily upon the availability of large, high precision optics. Current technology and existing facilities are insufficient to produce, in a timely manner, either the quality or the quantity of high performance mirrors required for envisioned systems. In addition, the performance requirements associated with space based, diffraction limited optical systems demand ever increasing figure quality. In response to this challenge, the government has sponsored the Rapid Optics Fabrication Technology (ROFT) Program.

Future large optics systems currently under consideration require the availability of mirrors much larger than have previously been demonstrated. Taking a 4-meter diameter as a representative size for large monolithic mirrors, we will consider three material classes and their limitations to fabrication in volume. The materials considered are ceramics, metals, and advanced composites. After surveying currently demonstrated capabilities, we will project two years into the future to determine the potential production rate for the various materials, as well as identifying developmental limitations which must be addressed. Our purpose in writing this paper is to identify process and facility improvements required for each mirror type, along with alternate fabrication processes. That way, while planning future systems requirements, these long lead items can be kept firmly in mind.

Coatings with lower emissivity than the traditional aluminum coating are required for large telescope mirrors. Multilayer coatings consisting of silver with ultraviolet boosting overlayers and copper underlayers for increased environmental resistance have been proposed. These multilayers are examined briefly, and some of their implications for the coating of large astronomical telescope mirrors are considered. A major problem is that they demand much greater uniformity than simple metallizing. A possible coating plant layout is suggested.

In late 1982 the Optical Sciences Center decided to purchase a numerical control optical generator. The purpose was to support the work of Roger Angel's mirror blank fabrication work at Steward Observatory'-4 and to generate the mold segments for use in fabricating carbon fiber-reinforced plastic panels for the Sub Millimeter Telescope, a joint project of the University of Arizona and the Max Planck Institut fur Radioastronomie in Bonn, West Germany.

The Hubble Space Telescope uses a primary mirror of area 5 m². We consider the mirrors needed for two possible future telescopes, of 50 m² and 500 m² in area. The former would be an 8m diameter monolithic mirror of "zero" expansion glass, and would be diffraction limited at optical wavelengths (0.015 arcseconds images). The 500 m² mirror, the area proposed for the Large Deployable Reflector, would be diffraction limited down to 30μ wavelength (0.3 arcsecond images). Both types of mirror could be of honeycomb sandwich construction made by a new air inflation method. The larger reflector would consist of ultra lightweight glass panels.

Telescopes had tended to follow established designs until costs made it essential to look toward significant departures from conventional designs. Attempts to introduce some new concepts were without success until the MMT broke the established pattern. Now several new possibilities are being engineered for the very large telescopes of the future.

The University of Texas 7.6m telescope design has evolved through several stages, each bringing added strength. and efficiency to a well-planned overall approach. The current design is based on an f/1.8 primary focal ratio with an f/13.5 Nasmyth beam directed through the elevation bearings. Recent studies include consideration of a class of correctors defined for an fill.° internal (Epps) focus by Charles F. W. Harmer, optical consultant. A standard Surrurier truss arrangement for the optical support structure assures equal gravity flexure. The Nasmyth flat and corrector housing are supported by a rigid elevation box structure. In the area of primary mirror design and support, research toward a superlative design is still in progress, primarily to take advantage of the rapidly emerging techniques of high efficiency borosilicate spin casting developed at the University of Arizona. There are some significant tradeoffs in terms of rib structure, span of unribbed sections, total weight, casting difficulty, and design of a tunable mirror support. A ribbed borosilicate mirror of the 7.6m class is very nearly passively supportable, provided the image specifications are not too stringent, and its figure stability can be made considerably better than any mirror extant with a modest amount of force variation in the axial support mechanism. Mirror support consists of a combination of pneumatic and counterweighted flotation systems with a low bandwidth tuning system added for figure improvement. Large classes of external loads, to be expected in a realistic observatory environment, may be obviated by such a system.

To help gain an engineering overview of the efficacy of a low bandwidth figure control or "tuning" system, the Aerospace Corporation, during a review of the 7.6m telescope design, produced a short 16mm film of a simulation designed by Dr. R. Hefner of Aerospace. In this design, the mirror is derived from a blank proposed by the Corning Glass Works, only 5 inches thick, of ULE material. An oscillating point load is placed at a point near the edge of the mirror, half-way between two of the collimating points. Two stages of control are depicted, in quasi-real time, with a clock display inculded. The first is rate damping at the 14 controlled points. The second is based on direct output feed-back of position information derived from the controlled points. A point spread function displays the energy concentration improvement gained by utilizing such figure control methods. Although rate damping can give a measure of improvement, it is clear from this single point case that to obviate the effects of all potentially degrading external forces, direct output feedback is the more desirable approach.

The University of California and the California Institute of Technology are collaborating to build and operate the Keck Ten Meter Telescope and Observatory. The telescope will be a ten-meter diameter, ground-based, astronomical telescope for observations at visible and infrared wavelengths. The telescope primary mirror is a mosaic of thirty-six hexagonal segments, each 1.8m in diameter. Piston and tilt in two directions of each segment are actively con-trolled by computer. The fabrication of the segments is a major challenge. The material must be stable under gravitational and ther-mal perturbations. Each segment has a surface shape that is the off-axis section of the parent hyperboloid and this surface shape is difficult to achieve with conventional polishing methods. In this paper we describe in detail the proposed segments and we describe the theoretical basis of the method proposed for achieving the segment figure, Stressed Mirror Polishing.

A 2-meter diameter off-axis parabola has been produced using the stressed mirror polishing technique. Surface accuracy obtained was better than 0.25μm (rms) or within about 1% of the intended goal. Further improvement is limited by mechanical factors. Results are presented. No further work is planned.

After several years of technology development and concept evaluation, the "four-barrel" Multiple Mirror Telescope concept has been selected for the 15-M National New Technology Telescope. Performance goals are presented. The principal optical configurations are described with emphasis on trade-offs under current study. Discussion of the thermal environment needed for the primary mirror is presented.