Deterministic microgrinding (DMG) process technology, developed by the Center for Optics Manufacturing (COM) at the University of Rochester, has been extended with the development of two new computer-controlled contouring microgrinders that can produce highly accurate aspheric and conformal optical surfaces in minutes. The Nanotech^TM 150AG Asphere Grinder was designed and built by Moore Tool Company (Bridgeport, CT) with input from the COM Machine Technical Advisory Board. Funded by DARPA, the scope of this COM-led machine development project was to create a cost- effective, high precision machine that would deterministically microgrind aspheric optical components in brittle glass materials. A critical performance requirement dictated that the ground surfaces produced by the Nanotech^TM 150AG be compatible with the COM and QED-developed magnetorheological finishing (MRF) process. With the new capabilities of this grinder and the magnetorheological finishing process, COM has demonstrated a potential 10X reduction in the cost of asphere fabrication. The Nanotech^TM 150AG is a CNC controlled, ultra-precision machining system that is capable of deterministically generating axisymmetric aspheric optical surfaces up to 100 mm in diameter in a production environment. The contouring DMG process produces aspheric surfaces within 1 micrometer of the intended surface shape with significantly lower surface roughness and less subsurface damage than conventional grinding processes. Grinding results on a variety of glasses, crystal and polycrystalline brittle materials are reported in this paper. The Nanotech^TM 500FG ultra- precision Freeform Generator is the second next-generation deterministic microgrinder. Also developed under a DARPA funded program, the Nanotech^TM 500FG was designed and built by Moore Nanotechnology Systems, LLC with input from a COM-led Machine Technical Advisory Board. The Nanotech^TM 500FG is a multi-axis, deterministic microgrinding platform capable of generating non-axisymmetric and axisymmetric optical surfaces in brittle materials. This paper reports the first results from this machine.

Conformal grinding of optical materials with CNC (Computer Numerical Control) machining equipment can be used to achieve precise control over complex part configurations. However complications can arise due to the need to fabricate complex geometrical shapes at reasonable production rates. For example high machine stiffness is essential, but the need to grind 'inside' small or highly concave surfaces may require use of tooling with less than ideal stiffness characteristics. If grinding generates loads sufficient for significant tool deflection, the programmed removal depth will not be achieved. Moreover since grinding load is a function of the volumetric removal rate the amount of load deflection can vary with location on the part, potentially producing complex figure errors. In addition to machine/tool stiffness and removal rate, load generation is a function of the process parameters. For example by reducing the feed rate of the tool into the part, both the load and resultant deflection/removal error can be decreased. However this must be balanced against the need for part through put. In this paper a simple model which permits combination of machine stiffness and process parameters into a single non-dimensional parameter is adapted for a conformal grinding geometry. Errors in removal can be minimized by maintaining this parameter above a critical value. Moreover, since the value of this parameter depends on the local part geometry, it can be used to optimize process settings during grinding. For example it may be used to guide adjustment of the feed rate as a function of location on the part to eliminate figure errors while minimizing the total grinding time required.

CNC grinding technology is increasingly used in the manufacture of precision optical components. Grinding performance is strongly influenced by the interaction of the tool and workpiece surfaces on a microscale, which in turn is influenced by the structure (topography) of the two surfaces. Unfortunately, for tool surfaces in particular, relatively little quantitative information has been available on the nature of the surfaces generated during actual grinding operations. However with the availability of advanced metrology techniques, such as optical profilometry and atomic force microscopy, it is now possible to produce detailed three-dimensional images of tool surfaces and from them to extract detailed quantitative information about the surface and its evolution. In this paper we discuss the use of optical profilometry to quantitatively characterize the surface microstructure (topography) of composite diamond tools during grinding of optical glasses. As an alternative to measuring individual diamond profiles, both the size and shape of the active diamonds on the tool surface may be evaluated by examination of the overall bearing ratio of the surface. This is quicker and has the advantage of avoiding potential bias in the selection of diamonds to be measured. Use of a micro- marker technique to precisely measure bond wear rate from a series of surface images using a micro-marker technique is also demonstrated. Grinding performance is dependent on the process conditions, but is also found to be correlated with the tool's surface structure. Moreover since the tool surface evolves during grinding, a complex relationship between the process and performance is produced. Bond wear is found to play an important role in maintaining grinding performance. Process conditions which produce a steady bond wear rate aid in establishing a quasi-equilibrium state ('self-sharpening') under which grinding performance can be maintained indefinitely.

In this paper, the chatter observed in the deterministic microgrinding of optical glasses is examined. Because chatter adversely affects ground surface quality, understanding and eliminating chatter could significantly reduce total fabrication time. First, a description of the characteristics of chatter marks observed on ground glass surfaces is presented. Next, a model for chatter generation during grinding is described. In this model, a linearized formula for chip area is derived and a parameter (the process cutting stiffness) predicting the possibility of chatter is introduced. Finally, experimental results are presented.

We review extensive data on measured subsurface damage and surface roughness resulting from lapping (loose abrasive grinding under fixed nominal pressure) and deterministic microgrinding (bound abrasive grinding under fixed nominal infeed) of commercial optical glasses with a large range of abrasive sizes. Subsurface damage is measured with the dimple method and related techniques. Surface roughness is measured with white light interferometry. Our results show that subsurface damage and its statistical scatter can both be estimated directly from the non-contact measurement of peak- valley surface roughness.

At the 1998 ASPE Meeting in Carmel, California there was much discussion of surface and subsurface damage introduced into silicon by machining operations. Many investigators have studied the problem and have defined parameters of importance. Yet there is a need for costly and time-consuming post- machining operations, such as lapping, if items of highest quality are to be produced. Significant cost reductions should be possible with machining techniques that introduce minimal damage. The combination of tool selection and treatment, speed and feed parameters, coolant choice and talent has resulted in the improved machining process. This process reduces damage without increasing machining cost. Although the McCarter Superfinish is a proprietary procedure the resulting surface condition will be compared with that of conventional machining. Data supplied by a customer is the basis for the comparison.

Deterministic grinding of large optical components (2600 - 3700 cm² area) using Electrolytic In-Process Dressing (ELID) requires strict process controls for several process parameters. In this paper we describe how the voltage and current characteristics of the ELID circuit may be used to establish viable, in-situ feedback monitoring of the grinding process. The specific approach used was to keep the ELID power supply in constant voltage mode and maintain an average power level that was optimized for each material. This was accomplished by monitoring the pulsed waveform and its frequency spectrum. By controlling the down feed rate it is possible to control the electrical characteristics of the wheel. A control loop was developed to over-ride the feed rate based upon the characteristics of the pulsed waveform. A second ELID process monitor was incorporated into the optic support scheme. To insure the part being ground was in a mechanically stable environment the optic was instrumented with eddy current gauges to detect motion during machining. Based on the data obtained from these sensors the support for the optic was optimized to minimize rigid body motion as well as bending. It has been found that creating a stable platform for machining as well as maintaining control of the ELID system is essential for a deterministic process.

The various properties of polishing pitches have different advantages and are selected according to the type of work intended. It is important to check pitch properties before a pitch lap is poured; to ensure that the final polishing lap properties will be as desired. A simple penetrometer test is utilized as a quality control tool for measuring the hardness (viscosity) of various types of pitch as received from the manufacturer. Only a small sample, 20 grams, is needed for this test. Another simple method for determining pitch quality is the measurement of the softening point. A description of this method and typical results will be described. Lastly, the 'tackiness' of pitch and its importance will be discussed.

Finish polishing of highly precise optical surfaces is one of the most promising uses of magnetic fluids. We have taken the concept of magnetorheological finishing (MRF) from the laboratory to the optical fabrication shop floor. A commercial, computer numerically controlled (CNC) MRF machine, the Q22, has recently come on-line in optics companies to produce precision flat, spherical and aspheric optical components. MRF is a sub-aperture lap process that requires no specialized tooling, because the magnetically-stiffened abrasive fluid conforms to the local curvature of any arbitrarily shaped workpiece. MRF eliminates subsurface damage, smoothes rms microroughness to less than 1 nm, and corrects p-v surface figure errors to (lambda) /20 in minutes. Here the basic details of the MRF process are reviewed. MR fluid performance for soft and hard materials, the removal of asymmetric grinding errors and diamond turning marks, and examples of batch finishing of glass aspheres are also described.

Magnetorheological Finishing (MRF) is a novel process for deterministic figure correction and polishing of optical materials that utilizes a sub-aperture lap created by moving a magnetic field-stiffened magnetorheological (MR) fluid ribbon against an optical surface. MRF has been successfully applied to a wide range of optical materials. A new research platform has been designed and built that is used to generate sub-aperture polishing profiles, i.e., polishing 'spots,' on optical flats under well-controlled conditions. This platform uses the same fluid circulation and conditioning system as the commercial computer numerically controlled MRF machine, thereby allowing fluid performance issues to be investigated. This new machine complements the capabilities of the original MRF research platform that has been in continuous use for over six years. These two machines have been used to generate polishing spots on a variety of optical materials. The spot profiles were measured to calculate material removal rates and the quality of the polished surfaces characterized by measuring the microroughness within the polishing spots. Examples are presented which illustrate how the evaluation of polishing spots was used to develop MR fluids and operating conditions for calcium fluoride, CaF₂, and potassium dihydrogen phosphate, KDP.

Magnetorheological Finishing (MRF) is a newly developed and recently commercialized method for finishing optical components. The MR fluid consists of a water based suspension of carbonyl iron (CI), nonmagnetic polishing abrasives, and small amounts of stabilizers. MRF uses both mechanics and chemistry to smooth glass surfaces to less than 10 angstrom rms. Mechanics are responsible for the microscratching of the glass surface which is initially 'softened' by hydration from water in the MR fluid. Experiments were performed to study the separate roles of mechanical abrasion and chemical softening during MRF. Chemical effects were suppressed by introducing a nonaqueous fluid in place of the water. A novel nanoindentation technique was used to determine CI particle hardness, which varied by a factor of five. The mechanics of removal were then investigated with soft and hard CI powders working against soft and hard optical glasses. Chemistry was then 'turned on' by the addition of a small amount of water to the system. Preliminary results of these experiments are presented here.

Recognizing the need to launch larger aperture telescopes to orbit with current launch capability, the Next Generation Space Telescope (NGST) program began in 1997, an aggressive technology development effort in lightweight optics. This effort has developed technologies in Silicon Carbide, Beryllium, silicon, glass and hybrid mirrors to name a few. The advancement of mirror technology to below the 15 kg/m² areal density is challenging enough without the added requirements for NGST that the mirror operate at approximately 35 Kelvin. This paper will discuss the various NGST technology developments including applications relative to bonding or joining mirror materials, subscale developments, the mirror system demonstrator program, the advanced mirror demonstrator program and development of the mirror test program. Specific attention will be given to technical issues related to lightweight mirror development and processing. Other issues to be addressed are the development of the roadmap to the NGST flight mirror development and to assess the technology readiness for NGST.

A 2-m diameter mirror is being manufactured as a demonstration for NASA's Next Generation Space Telescope (NGST). This mirror meets the challenging requirements of cryogenic operation and very low mass using an active control system. The mirror system consists of an aluminized glass membrane, 2 mm thick. This membrane is supported and controlled based on wavefront measurements with 169 remotely driven actuators. The system rigidity is provided by a lightweight carbon fiber composite structure. This entire mirror system, 2 meters across weighs less than 40 kg, and will demonstrate 20 nm surface quality in a cryogenic test facility at 35 K.

The objective of this paper is to report the recent developments in lightweight mirror technology at Composite Optics, Incorporated (COI). The developments are a result of the activities being conducted in support of the Next Generation Space Telescope (NGST) Program. The sponsors of these efforts are the NASA Marshall and Goddard Space Flight Centers. The requirements, design approach, technical challenges, hardware status, and tentative conclusions for the program are summarized. The emergence of composite materials provides exciting potential for nontraditional, accurate, lightweight, stable, stiff, and high strength mirrors. This evolving technology promises significant improvement in reducing weight, cost and cycle time for future infrared, visible, and x-ray systems. Customers currently embracing composite mirror technology for radiometric use are already reaping substantial system performance benefits. Other customers interested in LIDAR, IR, visible, and grazing incidence x-ray applications are eagerly awaiting successful completion of current technology development and demonstration efforts. 1

Optical and optomechanical structures based on silicon carbide (SiC) ceramics are becoming increasingly important for ultra- lightweight optical systems that must work in adverse environments. At IABG and Dornier Satellite Systems (DSS) in Munich, a special form of SiC ceramics carbon fiber reinforced silicon carbide (C/SiC^R) has been developed partly under ESA and NASA contracts. C/SiC^R is a light-weight, high- strength engineering material that features tunable mechanical and thermal properties. It offers exceptional design freedom due to its reduced brittleness and negligible volume shrinkage during processing in comparison to traditional, powder-based ceramics. Furthermore, its rapid fabrication process produces near-net-shape components using conventional NC machining/milling equipment and, thus, provides substantial schedule, cost, and risk savings. These characteristics allow C/SiC^R to overcome many of the problems associated with more traditional optical materials. To date, C/SiC^R has been used to produce ultra-lightweight mirrors and reflectors, antennas, optical benches, and monolithic and integrated reference structures for a variety of space and terrestrial applications. This paper describes the material properties, optical system and structural design aspects, the forming and manufacturing process including high-temperature joining technology, precision grinding and cladding techniques, and the performance results of a number of C/SiC^R optical components we have built.

This paper gives a methodological analysis of abrasive Optical Fabrication Techniques that apply an abrasive contact between tool and workpiece for the generation of refractive and reflective (a) spherical optical surfaces. Optical Fabrication Techniques comprise the following two functioning elements: the used Material Removal Process (MRP, e.g. fresh feed polishing) and the applied Shaping Method (SM, e.g. based on surface evolution calculations). Important MRP's and SM's are reviewed. Subsequently, state-of-the-art optical fabrication techniques will be discussed representing existing combinations of MRP's and SM's. Finally, it will be shown, that new optical fabrication techniques can be developed either by inventing new MRP's or SM's, or by creating new combinations of existing MRP's and SM's.

An accuracy analysis of FAUST and WAGNER (two fabrication techniques for rotary symmetric aspherical optical surfaces) is presented. Since within FAUST and WAGNER the three functioning elements of a curve generator are separated they can be used to balance each other. A procedure will be described enabling the determination of the influences of tool shape errors and misalignments on the generated surface shape, demonstrating that there is an advantageous error propagation factor between tool shape and generated surface shape. For the generation of a specific surface, this procedure can be used to determine the optimum fabrication process.

A program to simulate the production process on the FAUST machine is described and changes required to enable the simulation on the WAGNER machine are indicated. Examples of performed simulations are presented, first on a spherical surface and then on a parabolic surface. Finally, it is shown that the tool cross-section is independent of the shape of the starting surface.

Free-form surface is a kind of complex, irregular, asymmetric aspherical surface. Generally, it cannot be described with an equation, but usually by a series of digital points. Some special imaging effects can be realized by employing free-form lens. Moreover, because there is no constrain with free-form surface, the designer has more freedom of design by employing free-form lens in an optical system. However, it is quite difficult to fabricate free-form lens. This obstructs the wide application of free-form lens seriously. In this paper, a fabrication approach, named computer controlled fabrication (CCF), which includes: 1 computer numeric controlled (CNC) grinding, 2 CNC lapping, 3 profile measuring, 4 computer controlled corrective lapping, and 5 CNC polishing, is proposed. CNC grinding includes rough grinding, precision grinding, and fine grinding. Rough grinding is taken to remove the raw material rapidly. Precision grinding gets rid of the micro-crack layer and improves the profile accuracy. Fine grinding improves the figure accuracy further and generates very shallow micro-crack layer. After CNC grinding, a basic free-form surface with 'peak-valley' structure is obtained. CNC lapping is taken to get rid of the 'peak-valley' layer and micro-crack layer and to decrease the surface roughness without lowering the figure accuracy. Profile of the free-form lens is measured with a computer-controlled coordinate measuring machine. Machining error is obtained through the comparison between the measured data and theoretical profile data. Computer-controlled corrective lapping (CCCL) process provides the corrective NC code by full using the error data and unit removing profile, then corrective lapping is proceeded on a CCCL machine tool. Profile measuring and CCCL process are accomplished alternately until the required figure accuracy reached. Finally, the free-form lens is polished on a CNC polishing machine tool to lower the surface roughness. With CCF, two kinds of free-form glass lens were fabricated. Basic free-form glass lenses with figure error of plus or minus 8 micron and surface roughness about Ra0.5 micron are obtained. After CNC lapping, surface roughness decreases to less than Ra0.05 micron. Profile accuracy is improved to about plus or minus 2 micron after CCCL process, with surface roughness reaching Ra0.03 micron. After CNC polishing, free- form lens with profile accuracy of plus or minus 3 micron and surface roughness less than Ra0.01 micron is obtained.

Mirror replication includes coating, gluing and unmolding. It is a technique of producing mirror surface by molding. A coated mirror surface is replicated from a mold with a shape inverse to the required replica to a ground substrate. The distance between the mold and the substrate is filled in by the resin. By unmolding, we can get a coated replica which consists of the substrate and the resin. This paper presents the work on the 210 mm diameter flat replica, the 200 mm diameter spherical replica, and the experiment of a concave parabolic replica which we are making. The interferograms of the replicas and some parameters of the experiments are given.

The manufacture and testing of the 'core' optical substrates for the Laser Interferometer Gravitational-wave Observatory (LIGO) are described in this paper. These substrates are for use in long baseline Michelson interferometers with Fabry Perot cavities up to 4 km in length in each arm. The optical surfaces of the substrates (250 mm diameter by up to 100 mm thick) are specified either flat or curved, with radii of curvature varying between 7 and 15 km and tolerance bands on the radius equivalent to variations in the sag (over 200 mm) of twenty nanometers or so. Very strict tolerances were also placed on the astigmatism of the surfaces and the surface errors in two spatial frequency bands, one at low frequencies ('waviness') and another at high frequencies ('roughness'). In some cases the radius of the wavefront emerging from the substrate was also specified (for a collimated test beam).

The Laser Interferometer Gravitational-wave Observatory (LIGO) is a long baseline Michelson interferometer, with arms of up to 4 km in length each containing a Fabry Perot cavity. CSIRO has manufactured 32 core optical components for the LIGO interferometer consisting of five different groups of optical elements. Long radii of curvature (7 km - 15 km) and tolerances in the order of plus or minus 200 m in the radius are specified. Although the components are made of hyper pure fused silica there are some residual inhomogeneities in the material. The optics used in transmission must be figured so that the influence of these material inhomogeneities on the transmitted wave front is compensated for. This was done by correcting the surface figure on side 2 of the optics. The approach we took to manufacturing the transmission optics was to calculate the quadratic component of refractive index gradient (Delta) n of the substrate from the measurements of the transmitted wavefront and the surface profile of the two substrate surfaces, determine what shape had to be produced on side two of the substrates to compensate for this gradient and then produce this by optical polishing. The surfaces were polished on rigid solid laps of Zerodur coated with a thin layer of Teflon as the polishing matrix, a technique developed by CSIRO for super-polishing very flat surfaces.

The wavefronts reflected by and transmitted through a coated substrate will be influenced by the non-uniformities of the coatings and distortion of the substrate produced as a result of coating stress. In this paper we describe the characterization procedure and results of a coated substrate for the Laser Interferometer Gravitational wave Observatory (LIGO) project. The fused silica substrate is 250 mm in diameter, 40 mm thick and on one side a multilayer anti- reflection coating is deposited and a 50% reflectivity multilayer coating on the other. To characterize the coatings, reflected and transmitted wavefront measurements were carried out with a 300 mm aperture phase-shifting Fizeau interferometer in combination with ellipsometric measurements of the coated surfaces. The interferometric measurements allowed the deformation of the substrate by the coatings to be assessed while the ellipsometric measurements allowed the coatings' thickness variation to be measured and the resulting phase variation in the reflected and transmitted wavefronts to be estimated. The measurements revealed substrate deformation of about 45 nm with a coating relief non-uniformity of about 5 nm over a working aperture of 200 mm.

The atomic vapor laser isotope separation process uses high- average power lasers that have the commercial potential to enrich uranium for the electric power utilities. The transport of the laser beam through the laser system to the separation chambers requires high performance optical components, most of which have either fused silica or Zerodur as the substrate material. One of the requirements of the optical components is to preserve the wavefront quality of the laser beam that propagate over long distances. Full aperture tests with the high power process lasers and finite element analysis (FEA) have been performed on the transport optics. The wavefront distortions of the various sections of the transport path were measured with diagnostic Hartmann sensor packages. The FEA results were derived from an in-house thermal-structural- optical code which is linked to the commercially available CodeV program. In comparing the measured and predicted results, the bulk absorptance of fused silica was estimated to about 50 ppm/cm in the visible wavelength regime. Wavefront distortions will be reported on optics made from fused silica and Zerodur substrate materials.

Optical components needed for high-average-power lasers, such as those developed for Atomic Vapor Laser Isotope Separation (AVLIS), require high levels of performance and reliability. Over the past two decades, optical component requirements for this purpose have been optimized, and performance and reliability have been demonstrated. Many of the optical components that are exposed to the high-power laser light affect the quality of the beam as it is transported through the system. The specifications for these optics, including a few parameters not previously reported and some component manufacturing and testing experience, are described.

In most cases, conventional interferometric methods can be used to test the figure of large spherical and flat optical components. There are, however, certain types of unconventional surfaces, such as those used in grazing incidence x-ray applications, that are nearly impossible to test by conventional means. These cylindrical aspheres are usually tested by some type of scanning optical profiler. We discuss the use of a versatile slope measuring scanning interferometer system, the Long Trace Profiler (LTP), in measuring the figure error of large surfaces, particularly those that have extremely long radii of curvature in the tangential direction. Use of this instrument in different configurations has permitted measurements to be made on cylindrical asphere segments that are over one meter long, on water-cooled high heat load mirrors in ultra high vacuum synchrotron beam lines under actual operating conditions, and on complete x-ray telescope mirror shells and mandrels that are mounted in a vertical configuration to minimize gravity sag errors. Each of these different configurations has its own particular advantages and shortcomings. The ultimate performance of the LTP depends upon the thermal stability of the local environment. We illustrate the effects that temperature variations on the order of plus or minus 0.1 degrees Celsius have on the errors in the measurement of a long radius sphere.

Phase-shifting projected fringe profilometry (PSPFP) is a powerful tool in the profile inspection of a large variety of rough surfaces. In many applications, absolute PSPFP measurements, capable of compensating for the lateral distortions in the measured object shape and providing an expression of the measured shape under a predefined reference system, are highly desired. In this paper, an absolute PSPFP technique combining the lateral calibration and the phase-to- depth calibration is proposed. The principles of the proposed absolute PSPFP measurement technique will be discussed together with the calibration and measurement methods based on a particular formalism of absolute PSPFP measurements.

This paper describes a device for measuring the three dimensional surface profile using a grating projection method. A phase shifting technique without any mechanical moving is expected for profile analysis. A grating that is a key component in this technique is made using an active controlled liquid crystal (LC). This LC grating has the performances of more than 8 bits of gray levels and its grating period is 50 micro-meters per line without any colored filters. Surface profiles of some samples are measured for the demonstration of the system.

In this paper is introduced a new concept -- image point displacement (IPD), on the basis of which the principles and key tasks of light pattern projection profilometry are re- interpreted and evaluated. Two error sources that have not been addressed adequately are analyzed in detail. One is coordinate deviation, which comes from the different magnification at different depth. The other is the nonlinearity of the phase-IPD and phase-height relationship. These error sources will be more and more non-negligible as researchers make their ways to improve accuracy and increase measuring range. Some compensation approaches are also proposed, which have been verified by numerical simulations and experiments.

Measuring surface figure of large ground surfaces has been done with infrared interferometers or by means of local measurements with a spherometer to obtain a general shape of an optical surface. This paper describes a straightforward technique to obtain surface figure of plano parallel optics with an array of transducers referenced to an optical flat. The instrument utilizes 16 linear variable differential transformers (LVDT) and 16 ultrasonic transducers (UT) to measure surface figure of side 1, side 2 and the wedge in one measurement setup. The transducers are setup in a 4 X 4 array, for a total of 32 in one fixture. The data is acquired via a PC acquiring data through serial ports and an A/D card. The two 4 X 4 data sets are fit to the first ten Zernikes using the method of least squares. The data is displayed with 3D graphics to obtain a view of the optical surfaces. By using 14 bit digital LVDT's and employing the cross correlation technique for acoustic signal processing a system accuracy of plus or minus 1.0 micrometer for the LVDT array and plus or minus 2.75 micrometer for the UT array has been achieved.

The ultra-precise manufacturing of surfaces needs ultra- precise characterization. A novel optical scanning technique for testing flats, aspheres and complex surfaces is presented which offers ultra-precision for the measurement of slope and topography. The scanning technique needs no external references; it is traced back exclusively to the units of angle and length. The technique is based on a combination of two principles, namely to perform difference measurements for slopes with large shears and to use only a single pentagon prism for the difference measurements, keeping the angular position of the pentagon prism constant in space. The combination of these two principles eliminates the influences of all first- and second-order errors of the facility. Whole- body movement of the artefact under test and distortion of the facility itself do not result in errors.

For real-time 3D imaging using chirped optical pulses and a femtosecond optical Kerr shutter to measure the shape accurately, we need to develop a spectral imaging method which has a high-resolution in both space and spectrum. The spectral property of the light which is generated from the chirped pulse by the femtosecond optical Kerr shutter, the wavelength is measured as the line spectrum. Therefore, the wavelength can be determined by two detectors with different spectral properties. To obtain the line spectral image, we proposed a pair of filters, the transmissivities of which vary monotonously with the wavelength, one increases with the wavelength, the other decreases. Since the ratio between the transmissivities of the two filters changes monotonously for wavelength range of 550 - 750 nm, the wavelength of line spectrum is determined uniquely for the range with a simple function of the ratio. These filters were set in front of two monochromatic CCDs which are aligned to take an image. The resolution to determine the wavelength is tested with a monochrometer and the standard deviation for each pixel is estimated to be about 10 nm. Compared the methods of color CCD and spectrometer, this method of pair-CCDs has high spatial resolution, uniform spectral resolution, wide spectral measurement range, and simple setup.

The main methods for measuring geometric parameters used to be static or quasi-static. At present, former static measurement methods with sliding guide have met with great difficulty when measuring the outline and profile of large-scale workpiece or coordinates of objects moving in large dimension. Thus we established the laser tracking system for high precision, large range, non-contact, dynamic measurement. Laser tracking system involves in interdisciplinary knowledge and it can be made up into a new type of computer integral system with both measuring and controlling capacity. With common static coordinate measurement system, such as CMM, the probe must move along the three orthogonal guides equipped with standard rulers, while the laser tracking system has no such objective reference coordinate system. So we put forward the notion of 'Virtual Coordinate System,' that means during the measurement, the coordinate frame is not fixed. On the primary stage of our research, we established practical system for 2D moving target measurement in real-time. In order to measure the coordinate of the target, three tracking stations are used, which only measure length changes between the target and stations each, with no angles are measured. By solving non- linear least-square equations, the target and the tracking stations' coordinate can be attained respectively. We also present the optical design and the tracking controlling system based on the PID controlling principle. At the end of this paper, we will give the experiment result of the system.

Scanning Deflectometry is a powerful method to measure optical figure quality of various optical components and systems in a simple way. This principle uses detection of slope deviations rather than optical path length variations. As an example, the design of a basic deflectometer for testing flat mirrors is presented.

Development of accurate surface characterization methods is essential for testing micro-optical components, such as micro- opto-electro-mechanical systems (MOEMS), for use in complex optical systems. We consider using an array of 16 micrometer- wide micro-mirrors as programmable slits for astronomical multi-object spectroscopy, and propose a new method based upon Foucault's knife-edge test to characterize local surface deformations of individual micro-mirrors. By measuring local slopes, the surface shape of each mirror in a micro-mirror array has been reconstructed with a sub-nanometer accuracy. In addition to low-order deformation (tilt, curvature, astigmatism), each mirror is seen to be palm-tree shaped. We have checked the validity of our knife-edge test by the micro- characterization of a conventional spherical mirror.

The smooth continuous wavefront deformation function (WDF) can be expanded into Zernike polynomials. The coefficients of polynomial expansion for Fizeau or Twyman-Green interferometry can be found with ease by applying the least-square approximation. In the Hartmann test, shearing interferometry methods, or the Ritchey-Common test, coefficients can be found by using the least-square approximation as well. In these cases, the measuring data is a result from applying a linear operator to the WDF (which is the differentiation operator in the Hartmann test). By applying this operator to Zernike polynomials, new polynomials can be found for the test data. The coefficients of test data expansion and WDF expansion are equal. As in lateral shearing interferometry, the data from several test pictures can be approximated in one step of the 'truss' approximation.

On the basis of our papers, a new wedge plate shearing interferometer has been developed to evaluate and measure collimation, divergence and other parameters of laser beams and the amount of defocusing of systems of collimator and beam expander. It also can be used to test wavefront aberrations and to adjust optical systems. This is an ordinary shear plate interferometer but in order to obtain a quantitative result, it didn't apply the conventional method which is used to measure fringe width and angle, and it is impossible to give a high precision results. A reliable and high precise method is measuring the rotating angle of the plate about the center normal line of the front surface of the plate. Above method belongs to one-dimensional shearing interference. In order to test rotational asymmetric wavefront, the plate can be rotated about the optical axis of the incident laser beam. As a result any orientation shear (horizontally, vertically and at 45 degree or other angles) can be given. This instrument allows quick and easy measurement of the defocus, and third order spherical aberration, coma, and astigmatism in the system.

A novel Large Shearing Interferometer with four-plate that can adjust shearing distance freely and work in equal optical path is presented. It is used in measuring the pulse laser (from ns to ps) wavefront radius of curvature particularly. The interferometer overcome not the only shortcomings in single and double-plate shearing interferometer, but the diameter of optical plate for the interferometer is only one half or smaller than the beam splitter in three-plate cyclic shearing interferometer. Experiment results indicate that measurement error is smaller than plus or minus 5%(R: 30 to approximately 1000 meters). It also can be used in measuring the focal length and wavefront aberrations of optics.

Off axis aspheric surfaces, such as individual segments for a telescope mirror, and surfaces that do not have any optical axis are traditionally difficult to test. In addition to difficulties controlling the aspheric shape, mirror segments have tight on radius of curvature and optical axis position. This paper presents a new method of measuring these surfaces that uses a test plate with a spherical reference surface, in combination with a small computer generated hologram to compensate the aspheric departure. The example for measuring 1.8-m segments of a 10-m primary mirror is given.

In this paper we discuss the results given by the nominal design, tolerance analysis, and tests of aspheric CGH null. The optical testing configuration has been used contained a Zygo interferometer, a self-alignment CGH null and a convex aspheric surface under test with revolution symmetry. The CGH null design has been carried out while taking into consideration the parameters of the test surface such as focal ratio, axial coordinates of the test surface normal as function of radial height, as well as the diffractive technology limits. The worst case generated by Monte Carlo analysis of the tolerance has been used to determine the photolithography tolerances. Finally the CGH null testing results are compared with the nominal design and tolerance analysis predictions.

Computer generated holograms are widely used in optical testing and metrology. The primary role of the CGHs is to generate reference wavefront with any desired shape. Optical or electron-beam writers are commonly used for CGH fabrication. Limitations from the hologram fabrication processes cause errors in the reproduced wavefront. Errors in duty-cycle and etching depth have direct impact on both the amplitude and phase functions of the reproduced wavefront. A study using scalar diffraction model to simulate CGH fabrication errors and their effects on wavefront amplitude and phase functions are presented. Experimental analysis confirms the theoretical model. An example is given at the end to demonstrate one of the many applications of the wavefront sensitivity functions in CGH error budgeting for optical metrology.

The Far Ultraviolet Spectroscopic Explorer (FUSE), successfully launched in June 1999, is an astrophysics satellite designed to provide high resolution spectra ((lambda) /(Delta) (lambda) equals 24,000 - 30,000) with large effective area (20 - 70 cm²) over the interval 90.5 - 118.7 nm. The FUSE instrument consists of four co-aligned, off-axis parabolic primary mirrors which focus light into separate spectrograph channels. The mirrors are rectangular (407 X 372 mm) and fabricated from lightweighted Zerodur blanks. We describe a straightforward method for aligning these off-axis parabolas in an autocollimation setup via qualitative and quantitative analysis of static interferograms. Initial alignment is achieved rapidly by visual inspection of the interferogram as adjustments are made in vertical and horizontal alignment. Fine alignment to the limit of the optical system then proceeds with small alignment steps and fringe analysis software to find the position which minimizes wavefront error. This method was used for figure testing the FUSE primary mirrors throughout build-up and qualification of the flight mirror assemblies. The far- ultraviolet reflectivity of the FUSE mirrors is very sensitive to molecular contamination. All mirror testing thus took place in a strictly controlled class 1000 clean room environment. In addition to the challenging vibration and turbulence problems this environment presented, two of the fight mirrors were coated with lithium fluoride over aluminum. This necessitated purging the setup with dry nitrogen, as the lithium fluoride coating degrades with exposure to water vapor. We discuss the difficulties these environmental constraints presented and summarize the mitigating action.

We present the essential structures of compensating test system for high order aspherical plates and high order aspherical surfaces, of which R-C and Quasi-R-C systems with aperture of 1 meter, f/8 and FOV of 1 degree.5 are composed. All these compensating test systems reach very high compensating accuracy, the residual of most tested surfaces is less than (lambda) /100. A formula of third order spherical aberration is derived and applied to initial solution of compensator. The initial structure of the compensator is in correspondence with the optimal result very well. Finally, taking high order aspherical plate as an example, we discuss in detail the test system errors, and give some ideas of reducing test system error or loosing requirement of fabrication tolerance.

New mirror interferometers for the testing of the optical systems are discussed. Interferometers consist of the light source, very small convex mirror or lens, two small flat or spherical mirrors and optics under test. In the first new interferometer, radial shearing interferometer, a laser's light reflects from the spherical semiball mirror and goes to an optical system (or mirror) under test. The optical system (or mirror) reflects the light back and forms an image near spherical semiball mirror. Another part of beam of laser reflects from the small spherical semiball mirror, and then reflects from a small reference mirror (or reference small part of mirror under test), another small semisphere mirror, and flat mirror with central hole -- it is a reference beam. In the second new interferometer, demountable interferometer, a light passes of optical system under test, focused, passes semitransparent convex lens; the other part of the light reflects back on two small flat mirrors, reflects and both part of beam interfered. Both part of beam forms the interference pattern on a screen or CCD receiver. A modifications of the interferometers can be used also for testing ellipsoidal mirrors or another aspherical mirrors with the providing of special optical compensation and interference filters for better accuracy. An interferometers can be used at any spectral line from the far ultraviolet up to the far infrared. The interferometers are very simple, cheap and easily can be manufactured in quantity. The interferometers can be especially recommended for optics testing in space and manufacturing.

We have developed a model that predicts the effective optical path length (OPL) through a thick, refractive specimen on a reflective substrate as measured with a high numerical aperture, confocal interference microscope. Assuming an infinitesimal pinhole, only one 'magic ray' contributes to the measured OPL. It is possible to correct for the refractive errors and to unambiguously interpret the data. We present a comparison of our model predictions with experimental measurements of a fluid drop on a silicon substrate, obtained with a phase shifting laser feedback microscope.

A new approach to 3-D profilometry for the white light interferometer is presented. Recently many different methods have been used to analyze the data obtained from white light interferometric profilers. The advantage of the interferometric methods is their precision that can reach a small fraction of a wavelength. But these profilers are usually limited to relatively smooth surfaces as well as being very expensive. We detail a simple way to construct a profiler that uses a simple and efficient algorithm. It treats the data in a fast and simple manner, thus reducing both the acquisition and the analysis time. The method is based on the Focus measurement that finds a maximum variance value. The method works well with rough surfaces.

A visible wavelength interferometer has been developed for the alignment of an extreme ultraviolet lithography (EUVL) camera. Although the EUVL camera will operate at 13.4 nm, the alignment is far more conveniently done at visible wavelengths, at ambient pressure. Traditional visible interferometers are not capable of reaching the better than 1 nm accuracy required for EUVL camera alignment; so a phase shifting diffraction-limited interferometer was constructed and used to align and quantify the EUVL wavefront to an accuracy better than (lambda) visible/2000. The interferometer and alignment process are described, and camera wavefront measurements presented.

An advanced interferometer was built for measuring large mirrors at cryogenic temperatures. This instrument uses active control to compensate the effects of vibration to allow high resolution phase shift interferometry. A digital signal processor and high speed phase control from an electro-optic modulator allow phase measurements at 4000 Hz. These measurements are fed back to a real time servo in the DSP that provides a vibration-corrected phase ramp for the surface measurements taken at video rates. This instrument is planned to be integrated at NASA Marshall Space Flight Center's X-ray Calibration Facility for measuring NGST mirrors at 40 K.

A simple and robust interferometer using a laser diode subject to optical feedback is presented. Fringes phase can be locked by the optical feedback within less than 0.2(pi) (peak to valley value) even when the interferometer is placed on a wooden table where the fringe movement caused by vibration amounts to about 6(pi) (peak to valley value) in the absence of optical feedback. The locked fringe pattern with spatial carriers can be analyzed by a fringe analyzer at video rate and the measured results of a spherical mirror showed the same result and repeatability as on an optical bench. We developed an optical feedback interferometric head which is easily to set both vertically and horizontally corresponding to situation of sample.

Many algorithms for the phase-stepping (or phase-shifting) interferometry have been developed for last decade. In recent years, some algorithms are becoming more and more attractive to practical measurement because the phase steps can be calculated through fringe data itself and the calculation of phase distribution is insensitive to phase-stepping errors. A compound phase-stepping algorithm by Lissajous figures technique and iterative least-squares fitting method is proposed in this paper. The calculation of phase distribution is composed of two parts for analysis of real interferograms. First, calculate the initial phase steps by utilizing Lissajous figures and ellipse fitting. Second, calculate the exact phase distribution by solving the coupled equations obtained through the spatial least-squares fitting and the serial least-squares fitting. The iterative method with the initial phase steps obtained from the first part is applied to the second part. An optical disk has been measured to verify the performance of the new algorithm. The experimental results show that the new algorithm is insensitive to phase-stepping errors phase steps can be automatically calibrated on line.

In this paper, a He-Ne laser not only acts as a pure light source in the interferometer, but also as part of the interferometer. When an interning system may put into the cavity of the laser, the numbers of the laser oscillating modes may be changed. If the change is detected, the distance is determined appropriately. Further accuracy data is obtained by driving the mirror moving with electric-mechanical method. Because of optical gain, the output signal is very easy to further process.

We describe a high performance interferometer system which provides measurements of extremely large parts in production with a high throughput. The system has a clear aperture of 610 mm. Mechanical stability is addressed by phase shifting via wavelength tuning, which allows locking the interferometric cavity for increased stiffness. We describe in detail the special illumination and optical mounting systems and their impact on the optical configuration, phase shifting and detection systems and present results on large optical flats in both transmission and reflection.

Stitching Interferometry is rapidly emerging as an alternative to Standard Interferometry, in the measurement of Large Optics -- such as those found in Laser MegaJoule and NIF. Stitching Interferometry involves multiple overlapping sub-aperture measurements over large components, and a computer software to reconstruct the wavefronts. Obviously, the Stitching Interferometer's measurement characteristics have to be different to those of the Standard Interferometer of same nominal measurement area. Two questions emerge: (1) What metric do we choose to express these characteristics? (2) How does Stitching Interferometry compare to Standard Interferometry, using this metric? We choose to use the PSD to illustrate how Stitching Interferometry of large components compares with Standard Large-Size Interferometry, for various lateral scales. Also, we highlight some important characteristics of Stitching Interferometry, which arise from judicious use of the particular configuration of the device. Ignorance of basic propagation phenomena can lead to bad design of the Stitching Interferometer, and loss of any performance advantage over Standard Interferometry. Because many of these effects are not direct consequences of the Stitching process, we call them side effects. In this paper, we provide basic explanation, and keep the mathematics to a low profile -- indeed, it is not necessary to actually compute anything to understand the effects. However, some very basic formulas, a few numerical tables and lots of graphs are presented, in order to provide basis for discussion.

Interferometrically measuring the index of refraction variation (index homogeneity) of glass blanks requires that the blanks be made transparent to the interferometer laser. One method for achieving this is to 'sandwich' a rough ground blank between two polished flats while adding an index matching liquid at each surface interface. This is better known as oil-on-flat (OOF) or oil-on-plate testing. Another method requires polishing both surfaces and is better known as polished homogeneity (PHOM) testing or the Schwider method. Corning Inc. historically has used OOF testing to measure the index homogeneity of disk-shaped, fused silica boules over multiple 18' diameter apertures. Recently a boule polishing and PHOM testing process was developed by Corning for measuring the homogeneity over 24' diameter apertures to support fused silica production for the National Ignition Facility (NIF). Consequently, the PHOM technique has been compared to the OOF process using a number of different methods including repeatability/reproducibility studies, data stitching, and vibration analysis. The analysis performed demonstrates PHOM's advantages over OOF testing.

In a 24' phase shifting Fizeau interferometer designed and built by Veeco Metrology, coherent imaging method, which is different from one in conventional interferometers, is adopted. To decrease the some of the fixed pattern noise caused by dust on the optics within imaging system, which affects on transfer function, and to eliminate the reflections from the wedge of test glass in the measurement of homogeneity, an iris should be placed at the focal plane before final imaging lens. Its size depends both on the transfer function requirement and on the wedge angle of the test glass.

Veeco Metrology has designed and built 7 large aperture Fizeau interferometers. In these interferometers light sources need to have a selectable polarization to be either vertical or horizontal. Therefore, the method of whole light source rotation was selected. However, the assembly of whole light source needs to take time to ensure the intensity center of diode laser coincides with mechanical rotation axis. This paper describes another possible method for the laser alignment in a 610-mm large aperture Fizeau interferometer using a half-wave plate. This method is easy to align and is an option in the future. But because of time limit and cost problem we used the method of whole light source rotation in past 7 interferometers. In this paper experiments and simulated calculations were finished.

LLNL's project to construct the National Ignition Facility (NIF), a 192 beam laser system capable of generating enough light energy necessary to achieve fusion ignition, will require 26,641 small optics, many of which will be supplied in the form of cleaned, tested and aligned assemblies. These assemblies will be built to print, cleaned to specifications, and tested to performance specifications, ready to be installed in the laser system. A wide range of potential suppliers will participate in the manufacture of these sophisticated opto-mechanical systems. The injection laser system requires 7,440 precision optical components manufactured to state of the art performance specifications. In addition to 550 aspheric lenses, almost 2,000 precision spherical elements are required. Wave-fronts are specified in terms of P-V, RMS and RMS Gradient wave-front error, with strict requirements on the filtering and resolution which is required. Precision polarizers, high reflectors, leaking mirrors, high damage threshold coatings and cleanliness levels of 50 to 100 are also specified for this section of the NIF laser. The alignment and diagnostics systems for the NIF require 19,201 optics, many of which have requirements that exceed those of the injection laser system. All of these optics will be purchased using the ISO 10110 drawing notations. Other sections of the laser system will utilize commercial, off the shelf components to control cost. This paper will give an overview of the project and its objectives, with specific attention to the small optics required for the NIF.

The Injection Laser System (ILS) optical design of the National Ignition Facility (NIF) laser system is described, covering design functions, requirements and constraints, various approaches and options, and the resultant configuration. The front end compromises approximately 70 optical elements per beamline, and 8300 elements total, whose characteristic dimensions are from two to six inches. Individual beamlines span a distance of approximately 15 meters. A variety of optical element types are used: spherical and aspheric lenses, mirrors, polarizers, multi-order waveplates, Faraday rotators, and laser rods. The front end performs multiple functions, namely to: image the pupil of the NIF laser system; amplify the beam's energy with a gain of approximately 10⁴; magnify the beam size by 30x; split the beam four-fold; provide back-reflection isolation; and adjust the pupil location, along with arrival time of the pulse, on a beam-by-beam basis. Due to the high-energy nature of the beam, particular attention is paid to minimizing peak fluence throughout the system, thus reducing the likelihood of optical damage. The front end must deliver wavefront with no more than approximately a wave of P-V aberration (at 1.053 micrometer). This demanding wavefront requirement requires optics' surfaces and transmitted wavefronts to be of relatively high quality, typically 1/10 wave P-V (at 0.633 micrometer).

The NIF laser system will be capable of delivering 1.8 MJ of 351 nm energy in 192 beams. Diagnostics instruments must measure beam energy, power vs. time, wavefront quality, and beam intensity proifle to characterize laser performance. Alignment and beam diagnostics are also used to set the laser up for the high power shots and to isolate problems when performance is less than expected. Alignment and beam diagnostics are multiplexed to keep the costs under control. At the front-end the beam is aligned and diagnosed in an input sensor package. The output 1053 nm beam is sampled by collecting a 0.1% reflection from an output beam sampler and directing it to the output sensor package (OSP). The OSP also gets samples from final focus lens reflection and samples from the transport spatial filter pinhole plane. The output 351 nm energy is measured by a calorimeter collecting the signal from an off-axis diffractive beam-sampler. Detailed information on the focused beam in the high-energy target focal plane region is gathered in the precision diagnostics. This paper describes the design of the alignment and diagnostics on the NIF laser system.

The National Ignition Facility (NIF) project elected to implement ISO 10110 standard for the specifications of NIF optics drawings in 1996. More than 7,000 NIF large optics and 20,000 NIF small optics will be manufactured based on ISO 10110 indications. ISO 10110 standard meets many of the needs of the NIF optics specifications. It allows the optical engineer to quantify and clearly communicate the desired optical specifications. While no single drawing standard specifies all the requirements of high energy laser system, a combination of ISO 10110 standard with detailed notes make it possible to apply international drawing standards to the NIF laser system. This paper will briefly describe LLNL's interpretation and implementation of the ISO 10110 drawing standard, present some examples of NIF optics drawings, and discuss pros and cons of the indications from the perspective of this application. Emphasis will be given to the surface imperfection specifications, known as 5/, for the NIF optics.

In a high energy laser system such as the National Ignition Facility (NIF), the ability to focus light into as small a spot as possible at the highest possible fluence is highly dependent on the quality of the optics used in the system. Typically, surface form errors and transmitted and reflected wavefront errors are specified in terms of a peak-to-valley wavefront error (P-V), or occasionally in terms of an RMS wavefront error (RMS). It has been shown, however, that the parameter that most closely correlates with beam focusability is neither of these, but the RMS of the gradients of the wavefront error (RMS Gradient). Further, the spatial frequency of the wavefront error plays a significant role in the way that a given error effects the performance of the laser system, so careful attention must be paid to how the spatial filtering is both specified an accomplished. Since ISO 10110 has no specific provisions for a gradient specification, LLNL has developed its own notation and procedures for these critical specifications. In evaluating surface figure errors as specified by the NIF drawings, modern phase modulating interferometers (PMI) will be used. In addition to performing QA testing of the optics, LLNL intends to utilize the software capabilities of the instruments to obtain the information to model the wavefront of the 131 passes through various optical elements comprising the NIF 'front end.' A typical transmitted wavefront error call-out for an optic in the front end is: (lambda) /8 P-V, (lambda) /40 RMS, (lambda) /30/ cm RMS gradient, with these values evaluated for spatial periods greater than 2 mm only. Test will be performed and documented after coating and as installed in the specified mechanical mounts. This paper describes the evaluation of the wavefront error for NIF small optics including specifications over a given spatial period call-out, the proper low pass filtering of the data and the allowable filtering and settings that can be applied to obtain proper wavefront data. This paper also describes the origin and evolution of other NIF wavefront and roughness specifications, and gives examples. Since the wavefront requirements and hence the specifications vary for the different systems in the NIF, we will focus on one system, the injection laser system (ILS) or 'front end.' Also discussed will be the metrology and data manipulation requirements for the large aperture optics. Finally, clarification will be given to the differences between various versions of the RMS wavefront and roughness specifications allowed in ISO 10110, and how they contrast to the RMS roughness specifications used in ANSI-Y14.5.

The National Ignition Facility will require upwards of 25,000 small optical components in its various beam conditioning and diagnostic packages. A quality control program designed to ensure that the elements meet the required specifications will test these optical elements. For many of the components, damage performance is one of the critical specifications, which will require state-of-the-art performance from the industry participants. A program was initiated to understand the current performance level of such optics. The results of this study as it pertains to laser-induced damage is shown. The use of ratio reflectometry is also addressed as the method of choice for photometry measurements on these industry supplied optics.

Optical cleanliness is important to NIF because it results in beam obscuration and scatter losses which occur in the front- end (containing over 20,000 small optics) and the large- aperture portions of the laser (containing approximately equals 7,300 optics in 192 beamlines). The level of particulate cleanliness necessary for NIF, is similar to the scatter loss due to surface roughness. That is, the scatter loss should not exceed less than or equal to 2.5 X 10^-5 per surface. Establishing requirements for optical and structural surface cleanliness needs consideration of both particulate and organic thin-film cleanliness. Both forms of cleanliness may be specified using guidelines specified in Military Standard 1246C and are referred to as cleanliness Levels. This Military Standard is described briefly and displayed in tables and charts. The presence of organic thin-films on structural surfaces is of particular concern if the contaminated surface is near solgel coated optics [solgel coatings provide an antireflection (AR) quality]; or the optic is in a vacuum. In a vacuum, organic contaminant molecules have a much high probability of transporting from their source to a solgel-coated optic and thereby result in the rapid change in the transmission of the antireflection coating. Optical surface cleanliness can be rapidly degraded if a clean optic is exposed to any atmosphere containing an aerosol of small particles. The use of cleanrooms, as described in Federal Standard 209C, minimizes the settling of particulate contaminants and is described using charts and tables. These charts assist in determining the obscuration and scatter loss that can be expected when a clean surface is exposed to various Classes of cleanrooms due to particulate settling.

The most prominent physical characteristics of the 192-beam NIF laser are the 123 m length of the main laser and 400 mm aperture of each beam line. The main laser is illustrated in Figure 1, which shows half the total beam lines. Less visible are the many small optics (less than 100-mm diameter) used to align and diagnose each beam line. Commercial mounts can be used for most of the small aperture turning mirrors. This paper reviews the NIF projects effort to identify suitable commercial mirror mounts. The small mirror mounts have stability, wave front, space, and cleanliness requirements similar to the large aperture optics. While cost favors use of commercial mounts, there is little other than user experience to guide the mount qualification process. At present, there is no recognizable qualification standard with which to compare various products. In a large project like NIF, different user experience leads to different product selection. In some cases the differences are justified by application needs, but more often the selection process is somewhat random due to a lack of design standards. The result is redundant design and testing by project staff and suppliers. Identification of suitable mirror mounts for large projects like NIF would be streamlined if standards for physical and performance criteria were available, reducing cost for both the project and suppliers. Such standards could distinguish mounts for performance critical applications like NIF from laboratory applications, where ease of use and flexibility is important.

A large number of the small optics procurements for the National Ignition Facility (NIF) at Lawrence Livermore National Laboratory (LLNL) will be in the form of completely assembled, tested, and cleaned subsystems. These subsystems will be integrated into the NIF at LLNL. To accomplish this task, the procurement packages will include, optical and mechanical drawings, acceptance test and cleanliness requirements. In January 1999, the first such integrated opto- mechanical assembly was received and evaluated at LLNL. With the successful completion of this important trial procurement, we were able to establish the viability of purchasing clean, ready to install, opto-mechanical assemblies from vendors within the optics industry. Thirty-two vendors were chose from our supplier database for quote, then five were chosen to purchase from. These five vendors represented a cross section of the optics industry. From a 'value' catalog supplier (that did the whole job internally) to a partnership between three specialty companies, these vendors demonstrated they have the ingenuity and capability to deliver cost competitive, NIF- ready, opto-mechanical assemblies. This paper describes the vendor selection for this procurement, technical requirements including packaging, fabrication, coating, and cleanliness specifications, then testing and verification. It also gives real test results gathered from inspections performed at LLNL that show how our vendors scored on the various requirements.

Today the Long Trace Profiler (LTP) is widely accepted as a viable way to measure X-ray mirrors, and at some institutions is the only instrument available for measuring long, high- curvature aspheres. Although some questions of absolute accuracy over the entire LTP measurement range remain unanswered, a comparison of LTPs can still be made to assess measurement variation. Recently a round robin survey of some LTPs within the United States has been made using a single set of mirrors. These mirrors were used to characterize the performance of an LTP over its advertised range of operation. The results of this survey are presented here.

The optical manufacturing facilities, equipment and processes used to manufacture and test an 8.3 meter diameter primary mirror are described. The very successful first light of the SUBARU telescope gives the final confirmation of the manufacturing and testing procedures. This achievement resulted from the cooperative efforts of three organizations working together to produce what is currently the largest and best optical quality monolithic 8 meter class astronomical telescope. The processes outlined here have set a new standard in optical quality for all future large astronomical instruments.

The Long Trace Profiler has become the instrument of choice for mirror surface figure test and slope error measurement of synchrotron radiation and x-ray astronomy optics. In order to achieve highly accurate measurements with the LTP, systematic errors need to be reduced by precise angle calibration and accurate focal plane position adjustment. A self-scanning method is presented to adjust the focal plane position of the detector with high precision by use of a pentaprism scanning technique. The focal plane position can be set to better than 0.25 mm for a 1250 mm focal length FT lens using this technique. The use of a 0.03 arc second resolution theodolite coupled with the sensitivity of the LTP detector system can be used to calibrate angular linearity error very precisely. Some suggestions are introduced for reducing the system error. With these precision calibration techniques, accuracy in the measurement of figure and slope error on meter-long mirrors is now at a level of about 1 (mu) rad rms within whole testing range of the LTP.