The key requirements for an optical mirror material include low density, high Young's modulus, low coefficient of thermal expansion, high thermal conductivity, and high diffusivity. Not included among these are fracture toughness and stress corrosion constant, which control slow crack growth and long-term reliability under static or dynamic loads during manufacturing and in-service. The reliability requirement becomes crucial as the mirror size increases and/or its mission takes on strategic importance. This paper compares the critical properties of three ultralow expansion materials, namely ULE^TM, Zerodur and Astrositall^TM. It demonstrates how these properties affect the bending rigidity and safe allowable stress for the mirror subjected to different types of loading, namely: (i) its own weight and (ii) external load. An analysis of bending rigidity, bending stress, and safe allowable stress shows that mirror blanks of two different materials can be designed to be equivalent in terms of their rigidity without any weight penalty. The lower modulus and lightweight material like ULE glass requires about 10 percent higher thickness which reduces the bending stresses 20 percent compared to those in Zerodur or Astrositall mirrors of identical size. The lower stress, according to Power law fatigue model, is highly beneficial in that it improves the mechanical reliability of ULE mirror during manufacturing, transportation, installation and in-service by two orders of magnitude over that of Zerodur and Astrositall mirrors. The fatigue and fracture data for the three materials are used to estimate the safe allowable stress for facilitating mirror design from mechanical reliability point of view.

In FTIR film thickness measurement system, traditional process for determining film thickness is optical interference method or absorption method. In practice, both interference effect and absorption effect make a partial contribution to spectrum respectively. When both of the tow effects have influence on measurement, neither interference method nor absorption method can be individually used to determine the film thickness accurately. A new mathematical correction method is described for FTIR film thickness measurement in this case. Employing this new mathematical correction method, the effects of interference and absorption can be apart form each other effectively. The film thickness measurement accuracy of (lambda) /100 has been achieved.

Ultra low expansion (ULE) material of Corning Glass Work was chosen for the Gemini primary mirrors. The ULE mirror blank becomes monolithic by a fusion process which seals together 55 piece parts from a total of 44 hexagonal segments (hexes). As a consequence of this fusion process, an optical surface distortion due to inhomogeneity in the coefficient of thermal expansion (CTE) is induced. The precise location of the individual hexes in the blank was determined by a detailed analysis in the optimization process. This analysis accommodates two thermal environments, thermal soak of -25' C and thermal gradient of 3' C from the top to bottom surfaces. A parametric design study was conducted to determine an optimized pattern of the hex placements for the Gemini primary mirrors. Active optics corrections were performed to determine the optimum hex patterns. The results indicated that the optical surface distortion due to the CTE deviations was minimized based on the optimized location of the individual hexes. The thermal surface distortion and the optical image quality as well as the plate scale error of the primary mirrors met the design and the scientific requirements. The effect of random errors of the CTE measurements was within the tolerance error budget.

The photoacoustic tow-beam phase lag method was described for studying the thermal diffusivity of metal and superconductor materials in this paper. A beam of laser modulated from an argon ion laser is divided into the two beams, which illuminate respectively front and rear surface of the sample. The phase photoacoustic signal difference (phase lag) can be measured by lock-in amplifier. The phase lag is related to thermal diffusivity of sample. We measured the metal samples such as lead, copper, aluminium and stainless-steel in room temperature, and the results are very agreeable with the other method. We also have studied the temperature dependence of the thermal diffusivity of the high-temperature superconductor Bi_1.5Pb_0.5Sr₂Ca₂Cu₃O. The experimental results are presented.

This paper describes the analyses performed to estimate the on-orbit distortion of the Mars Orbiter Laser Altimeter (MOLA II) primary mirror. MOLA II is one of five scientific instruments that will be flown on the Mars Global Surveyor. The MOLA II instrument will map the surface profile of Mars for a full Marian year to a resolution of 2 meters vertical and 160 meters horizontal. The MOLA II telescope is an f/6 Cassegrain telescope with a 0.85 milliradian (mrad) field of view. The telescope is made entirely of Brush Wellman S200F vacuum hot pressed beryllium. The primary mirror diameter is 508 mm with a base radius of curvature of 711.2 mm. This mirror is plated first with electroless nickel and then with electrolytic gold. The purpose of these analyses was (1) to estimate the on-orbit distortion of the large primary mirror due to thermal loading, interface stresses, and gravity release and (2) to calculate the expected damage to the mirror surface due to micrometeroid impacts. A detailed NASA structural analysis program finite element model was used as a tool for evaluating the mirror performance. The results of the analyses indicate that a stability error of 2.4 microns peak-to-valley and 0.6 microns root mean square is expected for the on-orbit distortion of the primary mirror surface. The estimated surface damage due to micrometeoroids is 0.03 cm², which is 0.002 percent of the total surface area. Both of these results are within mission acceptance parameters.

Improvements made in the ultrasonic thermal expansion measurement methods used at Corning Incorporated have significantly increased the precision of absolute CTE measurements on ULE^TM glass. Precise CTE data were required for modeling the performance of 8-meter mirror blanks. Several repeatability and reproducibility studies for many of the measurements performed throughout the 8- meter process were performed. Measurement equipment and procedures were then improved to obtain the required precision. A volume of absolute CTE data collected during the manufacture of three 8-meter mirror blanks is summarized, showing excellent results. Comparison with previously demonstrated precision is also given.

During the decade of the 1980's, silicon carbide was funded primarily as the water cooled mirror material for the future and secondarily as a lightweight tactical alternative to beryllium and glass. With the perceived deployment of Star Wars, the payoff for the silicon carbide investment was imminent. Wrong assumption. The emphasis shifted from cooled optics to lightweight, uncooled optics and structures during the early 1990's. CERAFORM SiC became more attractive as a mirror material as the forming process produced lighter, closed back mirrors and a polishing process was developed to finish the bare material to 10 angstroms rms. Cost became the major limitation to penetrating commercial markets and with the defense cut-backs in 1993 UTOS ceases operations. The facilities and intellectual property associated with CERAFORM was at the mercy of bean counters. In March 1995 Xintics officially purchased form the United Technologies Corporation all intellectual property including patents, processes, proposals, engineering notebook, and trademarks pertaining to CERAFORM SiC. In a subsequent deal, part of the furnace facility was also obtained.

Silicon carbide mirrors have been made ultra lightweight with an areal density below 10 kg/mm² and have been made in sizes as large as 1.2 meters in diameter. The CERAFORM SiC process provides a cost-effective means to make lightweight substrates in either the open back or closed back form. Optical finishes below 10 angstrom rms have been achieved on both the chemical vapor deposited beta phase and the silicon infiltrated alpha phase. COmplex structures with triangular, square, or hexagonal core geometry has been produced with web thicknesses as thin as 0.015 inches and depth to diameter aspect ratios as large as 50:1. By designing to specific sectional stiffness, SiC offers performance which exceeds that of beryllium and glass, especially in extreme thermal environments. By polishing bare CERAFORM SiC to better than 10 angstrom rms, the single greatest impediment to SiC being used as an optical material was resolved.

Recent advances in polishing the bare CERAFORM SiC surface to finishes as smooth as 1OA rms has enhanced the viability of SiC as a mirror material those applications requiring thermal stability over a broad temperature range. In addition PVD silicon claddings have been developed to provide a low cost polishing option for more environments which are less severe. With the ability to make complex shapes in sizes up to 1 meter, CERAFORM SiC provides a cost-effective alternative to beryllium and glass.

Large sapphire components are required to meet the challenging needs of commercial and military optical applications. However, the desirable properties of sapphire also make it difficult to grind and polish. Fabrication costs can represent 50 percent of the price of large sapphire components. Cutting and grinding studies on sapphire were carried out with four types of diamond tools to correlate tool characteristics and process parameters with the grinding mechanism. The Twyman effect was also investigated to relate to crystallographic properties of sapphire with fabrication concerns. If grinding and microgrinding techniques can be optimized, costs associated with fabrication of sapphire and other optical materials will ultimately be reduced.

We present here the study of the aging under thermal neutrons irradiation of NiC/Ti multilayers used in supermirror neutron guides. After irradiation, it is shown that titanium undergoes a change of structure. Under specific conditions, this change of structure is accompanied by the formation of an hydride. Moreover, stresses in nickel layers have been determined to show that they increase with the irradiation rate.

REOSC has been selected for the design, manufacturing and integration of the four ESO very large telescope (VLT) secondary mirrors. The VLT secondary mirrors are 1.12 m lightweight convex hyperbolic mirrors made of beryllium. Despite the VLT active optics correction capabilities, the use of a metal for the mirror structure implies specific manufacturing processes and associated design rules in order to ensure its dimensional stability during the telescope required life time. This paper describes how the fabrication process of the VLT secondary mirror has been optimized in order to maximize the dimensional stability of its structure. The beryllium properties are analyzed in parallel with the mirror requirements, the choices for the manufacturing, at all levels, are presented. A short work progress is presented, with the achieved mirror properties.