For many years Schott has been supplying the Zerodur glass ceramic, a very low expansion material, for a variety of applications, in particular, mirror substrates. With the requirements shifting towards larger diameters, a decrease in the weight of the mirror substrate becomes increasingly important. This goal can be achieved by applying a supporting rib structure to a thin face sheet thus providing the structural rigidity or by actually actively supporting a thin meniscus shaped blank. Considerable progress has been made in the forming of Zerodur, in particular, by applying the fusion, slumping and spincasting techniques. Several samples have been fused with diameters up to .5 m and aerial densities around 60 kg/m2. The slumping technique has been demonstrated for diameters up to 1 m with a thickness range between 4 and 100 mm. The spincasting technique has been successfully developed and demonstrated by the manufacture of several mirror blanks up to a diameter of 4.1 m at thicknesses of down to 57 mm. The production of spincast Zerodur mirror substrates is now deemed feasible for diameters larger than 8 m. The parameters for the selection between the slumping and spincasting processes are geometry, flexibility and economy.

A wide variety of metal oxides with high index of refraction can be prepared by Metal-Organic Chemical Vapor Deposition. We present some recent optical and laser damage results on oxide films prepared by MOCVD which could be used in a multilayer structure for highly reflecting (HR) dielectric mirror applications. The method of preparation affects both optical properties and laser damage threshold.

Thermal lensing of optics used in high power CO2 laser cavities and beam delivery systems is a constant problem facing the designer and the end user. CO2 laser mirrors, used as total reflectors in the laser cavity and beam turning optics in the delivery system, play a key role in the performance of the laser system. Silicon and copper are the two most popular substrate materials used for high power CO2 laser mirrors today. The amount of thermal lensing in these mirrors depends on the amount of absorption in the mirror coating and the ability of the mirror substrate to dissipate the heat energy absorbed by the coating. Since the coating applied to silicon and copper mirrors is the same, the decision as to which substrate material to use can be based on the mechanical and thermal characteristics of the respective materials. This paper will present current data on mirror coating absorptivities, a comparison of the mechanical and thermal properties of the two substrate materials, and finally, a figure of merit analysis based on data from a finite element analysis program.

By sandwiching a single layer of wire screen mesh mechanically between two surfaces, a simple heat transfer enhance-ment mechanism has been demonstrated recently (Fleishman and Yuen, 1988) to be effective for general high heat flux applications. Average heat transfer coefficients of up to 9.5 W/cm2K were measured for a smooth plane surface cooled by water. Based on a first-order mathematical model, the enhancement factor was shown to be a function of mesh wire diameter, thermal conductivity, screen pitch, surface heat transfer coefficient, and a non-dimensional thermal contact resistance. In this work, additional data are generated to further understand quantitatively the performance characteristics of the mesh-enhanced heat transfer mechanism. Specifically, average heat transfer coefficients ranged between 0.7 to 6.2 W/cm2K are measured for three different screen materials (brass, stainless steel, and copper) in a 26 cm2 fixture cooled by city water under line pressure. Heating rates range from 200 to 350 kW/m2 and flow rates from 0.004 to 0.086 kg Is . Effect of macroscopic screen parameters such as thermal conductivity, the weaving pattern of the wire and the "flatness" of the screen are investigated. The mathematical model is further developed to predict parametrically the optimized conditions for maximum heat transfer. The applicability of the present mesh-enhanced heat transfer mechanism for the cooling of high power density mirrors is assessed.

This paper concerns the problem of describing and evaluating thermal lensing phenomena that occur as a result of the absorption of laser light in solid windows. The aberration-function expansion method is applied for deriving the two optical distortion coefficients X+ and x_ that characterize the degradation in light intensity at the Gaussian focus of an initially diffraction-limited laser beam passing through a weakly absorbing stress-birefringent window. In a pulsed mode of operation, the concept of an effective optical distortion coefficient Xeff, which properly combines the coefficients X+ and X_ in terms of potential impact on focal irradiances, then leads to the definition of a figure of merit for distortion. The theory and the calculations presented in this and earlier papers provide simple analytical tools for predicting the optical performance of a window-material candidate in a specific system's environment.

A new process for producing large, single, oriented crystals of calcium fluoride (CaF2) has been developed which overcomes the limitations of current growing methods. This process has been reduced to practice and has yielded oriented crystals 17.5 3 x 17.5 x 5 cm3. Currently nearing completion is a system for producing 35 x 35 x 7.5 cm3 single crystals. A scale up to one-meter-square is considered feasible. This crystal growing process makes possible the fabrication of very large CaF2 windows. Suitability for very high power lasers, however, requires attention to properties beyond mere size. A process to generate higher purity growth stock (starting material) was also developed. The additional purification of the growth stock contributes to lower bulk absorption, the absence of color centers and increased radiation hardness. Also identified were several specific impurities which correlate with radiation hardness. A correlation was found between color centers induced by laser radiation and ionizing radiation. The ability of the new process to produce oriented single crystals permits entire laser windows to be fabricated with the <111> direction parallel to the direction of propagation. This orientation is known to reduce thermo-optic distortion and stress induced birefringence. Other CaF2 crystal properties such as tensile strength, absorption and laser damage thresholds were studied and are discussed.

The direction of this paper is to set out the analytical expressions which may be used to approximate the maximum flux density, or fluence, which will be required to cause thermal shock catastrophic damage to various and sundry transmissive optical components. The equations will describe the temperature, and subsequent stress gradients that are engendered, to cause the transmissive optic to reach the Modulus of Rupture and, self destruct in tension. Because of the nature of absorption on the surface, vis-a-vis, the bulk absorption, the surfaces of transmissive optics are very susceptible to substantial thermal gradients. The absorption on the surfaces of optics has been shown to be several times to hundreds of times greater than the bulk. If the optics are of the production line type, the absorption can be expected to be substantially more than those optics which are prepared for use in the laboratory environment Assembly line manufactured optics appear to be more greatly affected by defects, scratches, and residues which lead to greater surface absorption and are, subsequently, more susceptible to thermal shock catastrophic damage. We will look at the impact of Continuous Wave, Single Pulse, and Repetitive Pulsed laser environments on an optic which is transparent at some wavelengths, and not transparent at others. We will look at the equations that describe the temperature gradients developed in real time for both the non-transparent and transparent case. Damage flux density, and fluence thresholds are provided for a number of common materials used for windows and domes on optical systems that have no forced or free convection and those that are used in high velocity flow, forced convection, environments.

A spherical vacuum - interface window and an axial - flow impingement cooling system for high - energy laser applications are described. The vacuum - interface window is a 65 - cm diameter, 0.5 - cm thick, fused silica dome with a 10° edge angle. Both window optical surfaces are coated with anti - reflective coatings. Finite element analysis techniques were used to determine window thermal, structural, and optical performance characteristics for both uniform and Gaussian, 50 - cm diameter laser beam profiles. The axial - flow impingement cooling system provides convective cooling on a window surface. Helium gas is drawn into a throat duct, flows axially towards the window, impinges the window, and then flows radially outwards through a diffuser. Thermal, fluid, and optical performance testing was performed on a sub - scale apparatus to optimize convective cooling, minimize pumping power, and determine optical path distortion. A 40% scale factor was selected based on similitude. For the configurations tested, a diffuser angle of 30 degrees and a diffuser gap width to duct radius ratio of 0.3 maximized heat transfer with a Nusselt number of 610 at a flow Reynolds number of 740,000. This Nusselt number corresponds to a 30 - second thermal response time constant for the window. A gap width ratio of 0.4 minimized pumping power but at a 20% reduction in heat transfer. Use of turbulence promoters in the throat duct can increase convective cooling by 50%.

Feasibility of a 3.5 m diameter, segmented, solid, exit window for a high-power laser beam director was investigated. Effects of laser heating, gravity, and differential pressure loads were analyzed. A design was found for which structural deflections and stresses were within allowable limits. The time-varying optical aberrations introduced by the window were less than 0.01 1.tm rms after removal of bias, tilt, and focus aberrations for two window materials. Static aberrations caused by differences in optical thickness of the finished window panels were found to be relatively large and potentially uncorrectable if common optical finishing tolerances were assumed. Three potential solutions for accommodating the optical thickness variations are discussed.

A comprehensive capability for testing optics exists at the Michelson and Lauritsen Laboratories, Naval Weapons Center. We can evaluate coated or uncoated mirrors up to 4 m in diameter or with focal lengths of over 25 m. Measured parameters include optical figure, surface defects and surface roughness, total and angular dependence of scattered light, thermal distortion at temperatures from -190 to +100°C, optical anisotropy, coating uniformity, vibration characteristics of the mount, and optical properties (such as phase change on reflection and absolute reflectance). An excellent diamond-turning facility and an active optics shop are on site, as is an optics repository for the temporary storage of optics under clean conditions. A description of these facilities is given.

The uncontrolled thermal boundary layers associated with optical surfaces exposed to high power laser radiation can significantly degrade optical performance. The thermal boundary layers associated with the centrally obscured primary mirror or exit window of an on axis beam expander can be controlled by placing a radial array of gas jets in the central obscuration. The resulting momentum and thermal boundary layers were experimentally investigated, and the results are used to predict both time-averaged and time-varying optical path differences through the boundary layer of a 3.5 m diameter heated optical surface.

A previously patented design is re-examined as to its advantages over conventional stable and unstable resonators for large bore lasers. Two design options are identified which make the resonator more practical to employ. The equations for specifying the resonator parameters are presented along with a discussion of the unique advantages of this design.

The aberrations have been measured for the finished mirrors that are part of the Burst Mode ring resonator of the Free Electron Laser (FEL) being constructed at the Boeing Aerospace Company in Seattle, Washington. This paper presents analysis of these measurements using the GLAD code, a diffraction ray-tracing code. The diffraction losses within the resonator due to the aberrations are presented. The analysis was conducted in two different modes, a paraxial approximation and a full 3-D calculation, and good agreement between the two approaches is shown. Finally, a proposed solution to the problems caused by the aberrations is presented and analyzed.

A chemical vapor deposition process has been used to replicate shapes, patterns or highly reflective surfaces in infrared transmissive optical materials (ZnS, ZnSe) and mirror materials (Si, SiC) for a variety of applications, such as ZnS domes or meniscus lenses, ZnSe lens arrays for adaptive optics and low f-number Si/SiC mirrors for space optics. Conditions for achieving a high degree of replication are specified and replication results on several different substrate materials are presented. Techniques to obtain replication on those substrates which either are attacked in the CVD environment or whose thermal expansion coefficients are considerably different from that of the deposit are also discussed.

The ion beam mirror coating via a broad beam ion source has produced dielectric mirrors of extremely low optical loss and excellent damage resistance. The coating speed and the throughput should improve drastically in the near future.

We have observed damage to 80-cm-diam fused-silica disks and lenses subjected to high-fluence pulses (up to 2.3 J/cm2) from an upgraded Nova laser beamline (wavelength 351 nm; pulse duration 2.35 ns; beam diameter 70 cm; energy up to 8 kJ). Damage occurred in the center of each element, where a 6-cm-wide obscuration prevented direct illumination. We believe that light strongly scattered by transverse stimulated Brillouin scattering (SBS) interacts with the surface and with the bulk of the substrate, producing two kinds of acoustic waves that propagate to its center, where they become strong enough to do damage. In the surface interaction, scattered light is absorbed by an 0-ring near the perimeter of the optic, creating a Rayleigh wave that propagates along the surface to the center of the optic. The resulting damage takes the form of crater-shaped fractures about 8 mm in diameter and 4 mm deep. In the bulk interaction, transverse SBS strongly compresses the optic in large regions transverse to the direction of beam polarization at the perimeter of the beam. The compression may result from electrostriction: the SBS intensity is several times that of the incident beam. Compressive waves resulting from the relaxation of these regions propagate to the perimeter of the optic, where they are reflected as bulk tensile waves. The focusing of these tensile waves in the center of the optic results in cracks along the direction of polarization. Up to 25 percent of the incident beam energy is lost to SBS at these high fluences. Frequency chirping of the laser beam by 45 GHz strongly suppresses the SBS, and reduces the amplitude of the stress waves by about an order of magnitude; no energy loss, cratering, or cracking occurs under these conditions. We propose design rules for avoiding acoustic damage in large optics and compare observed thresholds for transverse SBS with predictions in the literature.

Ellipsometric data for several diamond-turned samples were measured at a wavelength of 5µm and at multiple angles of incidence. These data were reduced by the least-square-fit programs for different models to obtain the corresponding best-fit parameters. Calculations from the three-dimensional anisotropic model for diamond-turned surfaces show that the ellipsometric parameters are not sensitive to the sample orientation for parallel depolarization factors smaller than 0.03. The best-fit results for different models are compared and discussed. A special model for two-dimensional symmetric rough layers can give rms errors as low as δψ ≤ 0.005° and δΔ ≤ 0.02°; its best-fit parameters agree with the profilometric rms roughness and rms slope.

A holographic technique is described that can be used to evaluate the susceptibility to thermally induced distortion in mirror blanks prior to optical finishing. Polished blanks also can be evaluated using this technique. Primarily, this technique has been used for measuring the distortion at cryogenic temperatures; however, with minor modifications, it also can be used to evaluate mirror performance above room temperature.

Recent advances in High Average Power (HAP) solid state lasers and the development of new concept lasers with the potential of ultra-high average power output have put increasing demands on the transparency of optical window materials. To gain a better understanding of the current status of window materials and to direct research towards more nearly transparent materials, we have constructed an optical characterization facility with the purpose of making quantitative optical loss measurements in the sensitivity range of 10-3 to 10-6 cm-1. The cornerstone of this facility is a scanning optical lossmeter in which loss is determined by comparing the decay time of an optical cavity with and without a transparent solid present. The lossmeter has been successfully applied to measurements of the optical loss of witness samples of highly transparent fused silica. A description of the lossmeter and a compilation of preliminary loss measurements are presented here.

Optical component specifications for high average power laser systems must address performance, efficiency, maintainability, and reliability. The requirements for high performance optics goes beyond conventional methods of specification. Methods of specification for this application are discussed in addition to several areas that need attention within the optics community. The specifications must be based upon the available methods of testing. The ability to test the optical components is crucial to both successful manufacturing and system operation. This means that both the manufacturer and user must have testing capability. For large system applications, the test facility must be designed to meet the demand of a high volume of optics as well as measurement precision. The facility for testing these parts is described for an application requiring thousands of optical components. This test capability is also an important part of system design and development when the performance of high performance optics can be defined and incorporated during planning for a large facility.

For several years, the Optical Component Evaluation Laboratory (OCEL) and three sister laboratories have been building and operating optical measurement equipment. This contractual work is being done at Kirtland Air Force Base, New Mexico. Optical measurement capabilities include reflectance, transmittance, scatter, and laser absorption at one or more wavelengths including 351, 442, 633, 514, 1064, and 1318 nm.

Analyses of optical components consisting of a metal reflecting layer, a beryllium energy-sharing layer, and a fused silica substrate were done using LXRT, Livermore's version of the XRT (X-Ray Transport) computer code. These are compared with calculations for components lacking the beryllium layer. Our analyses show that the addition of the beryllium significantly hardens the system against x-rays. The mechanisms for this increased hardness are apparent from time-dependent tem-perature and enthalpy plots presented here.

The major physical reason for laser induced damage of real dielectric materials, their surfaces, and thin films is laser heating of the absorbing inclusions /1/. The role of the inclusions in the process of damage initiation is well experimentally established for both bulk damage /2-5/, and surface damage of transparent dielectrics /6-8/ and thin films /9/. The theoretical models of laser damage associated with the absorbing inclusions have been considered in /6,10-15/. The model of thermal explosion most consistently considered in /14-15/ for the case of bulk damage is evidently most adequate for the description of laser damage physics. Despite the general nature of thermal explosion of absorbing inclusions located in the dielectric bulk, surface layer, or thin film, their characteristics (threshold temperature Te and threshold intensity Ie) may differ. To investigate these differences one has to take into account numerous and, as a rule, purely controllable factors. For instance, it is well known that physico-chemical properties of the dielectric surface layer strongly differ from its properties in the bulk, the degree of difference depending on the surface treatment techniques. For this reason, a detailed investigation of the differences in the inclusion thermal explosion in the bulk and dielectric surface layer should be performed parallel with control of physico-chemical properties of this layer. However, in order to clear up ultimate resistance of the dielectric material surface, one has to study thermal explosion of the inclusion located in a perfect dielectric surface layer, whose physico-chemical properties coincide with those of the bulk. In the present paper we determine a relationship of the thermal explosion characteristics for the inclusions in the bulk and on a perfect surface, analyze thermal explosion kinetics, study the dielectric surface laser damage statistics with due regerd to the thermal explosion kinetics, and determine the dependences of the surface damage threshold Isd, upon laser pulse width end spot size. The results of the work are compared with experimental data.