The frequency response characteristics of an angle-resolved optical scatterometer and a mechanical profilometer are examined. The transfer function for the scatterometer is found to be constant within the spatial frequency bandwidth of the system. The mechanical profilometer is found to have a nonlinear response. The effect of the nonlinearity on the frequency response of the mechanical profilometer is examined.

A needed capability for making optical scatter measurements led the Lockheed company to design, build, and use an initial scatter measurement system. Experience with this early system revealed a number of unanticipated shortcomings. A new and improved system, designated as the BOSM Facility, was designed and built to correct the shortcomings and to improve functional capabilities. This facility will be described in this paper. The facility consists of three principle modules. First, there is a collimated light source with a variety of beam control features based on a fixed beam propagation direction. Second, there is a sample holder with three gimbal motions and a variety of controls. It is configured to facilitate and to optimize BRDF/BTDF measurements. Third, there is a PMT based photometer, with special electronic processing, mounted on a boom structure that pivots about the vertical center of the test sample. The gimbal and boom angle combinations provide the required bi-directionality. A facility computer with specialized software automatically controls illumination spot size and intensity, gimbal and boom angles, photometer measurements, and processing of raw measurement data into BRDF/BTDF output data. The software organizes these results into descriptive, tabular, and graphical reports. Other features, believed novel, are included in the facility design.

The comparison of BRDF (Bi-directional Reflectance Distribution Function) measurements made on the same system but at different times or days is a method of determining potential sample or system change. If the measurements lie within the acceptable error region of the system, confidence is gained in the results after repeated testing. However, when a system is replaced by a new, improved facility, both the time delay associated with the change and the different character of the systems may preclude easy, direct comparison. When the test pieces are carefully stored to minimize contamination and when both systems use a common standard, similar results can be anticipated in a comparison test. Such comparisons are needed and can be made. Two mirrors were tested, using an in-house designed and fabricated BRDF measuring system. The mirrors were immediately replaced in their protective boxes and stored. A comparison was made between the mirror test results and tests from similar mirrors. No abnormal results were found. The initial system was subsequently replaced by a new system, the LMSC BOSM (Bi-directional Optical Scattering Measurement) Facility. After installation and evaluation, the two reserved mirrors were retested on the new system and comparison between old and new data is shown.

An in-vacuum BR/TDF measurement apparatus has just been assembled as part of a larger space environmental effects facility. The sample stage and detector assemblies are mounted within and the requisite optics train and light sources external to a 48-in. diameter vacuum chamber. The chopped and field stopped f/60 incident beam enters the chamber through a window of appropriate material. Argon ion, helium neon, and CO2 lasers, as well as a xenon arc lamp, are used as light sources. Beam size on the sample is about 1-cm diameter with the detectors mounted on a 60-cm arm. The detectors can be rotated from within about 2° of the incident beam to directly facing the incident beam and can measure reflectance of a normally incident beam to about 83° and transmittance from 150° to 180°.

Mirrors designed to survive exposure to damaging radiation are being irradiated and then measured to determine the mechanisms of failure and to improve the ability of analysis codes to predict an exposure damage threshold. The differences between survival and catastrophic failure are easily recognized and recorded by macro photography. However, the goal of this project is to quantify the onset of mirror degradation utilizing non contact methods that have good measurement sensitivity to small changes in reflectivity (material properties) and light scatter (roughness). A new angular resolved scatterometer is described that has an extended dynamic range and integrated analysis capable of displaying the surface power spectral density (PSD) over large bandwidths of surface spatial frequencies. Graphical displays of the scattered light power before and after exposure to the radiation are compared and integrated over equivalent spatial bandwidths of sensitivity for other instruments to compare calculated RMS roughness values.

An antireflection overcoat for opaque baffle coatings in the FIR/submillimeter region has been made from a simple Teflon spray-on lubricant. The Teflon overcoat reduced the specular reflectance of four different opaque coatings by nearly a factor of two. Analysis, based on the interference term of a reflecting-layer model, indicates that in the submillimeter region the reduced reflectance depends primarily on the refractive index of the overcoat and very little on its thickness.

Deposition of 0.1 to 1.0 μM thick metal films on a CaF2 coated Si wafer causes a reduction in optical scatter and related microroughness. This effect is not observed when the coated surfaces are examined using a surface profilometer (Talystep). Scanning electron micrographs indicate that coated surfaces are smoother than uncoated surfaces.

Bidirectional scattering functions of numerous optical baffle materials and of spacecraft thermal control coatings and surfaces are presented. Measurements were made at 254 nanometers and at 633 nanometers. The coatings and surfaces include high-reflectance white paints, low-reflectance optical blacks, thermal control blankets and various conversion coatings on aluminum.

Light-scattering profiles were measured for small circular fibers. In particular, two fused silica fibers of different diameter were measured at both 10.6 and 0.6328-μm wavelengths, with a range in size parameter (circumference to wavelength ratio) from 1/2 to 20. Data were obtained for incidence normal to the fiber axis and for almost 360 degrees of rotation in the plane of incidence. The measurements were taken with an instrument designed for the optical surface evaluation of the bidirectional reflectance distribution function (BRDF) In addition, the measured scattering profiles were compared, with good agreement, to the Mie theory prediction for scattering from an infinite cylinder.

The stray radiation analysis tool presented here is based on a general ray tracing program which is integrated with the "Euclid" 3-D solid modeling CAD system. It is an analysis/design tool whose main features are its highly accurate representation of mechanical structures, its interactivity, its visualization capability and the understanding it gives the user about the paths by which stray radiation propagates through the opto-mechanical system.

The new Off-Axis Rejection Design Analysis Software (OARDAS) aids in modeling complex baffle configurations for generally asymmetric and folded optical systems in order to estimate their background suppression performance. This interactive program uses real raytracing and a combination of Monte Carlo and deterministic methods to aid the user in finding stray radiation paths and calculating stray backgrounds. User friendliness and flexibility were two primary goals in the program design. With its extensive graphics capability, OARDAS is a powerful tool for diagnosing and understanding complex radiation paths. This paper first gives an overview of OARDAS philosophy and capabilities and then shows how OARDAS is used to analyze a complex baffle design.

STRAY has been developed to quickly but roughly model the amount of stray radiation reaching the focal plane of a well-baffled telescope. This radiation arises from the scattering and diffraction of the light of bright off-axis sources, such as the Sun, Earth and Moon, as well as the thermal emission of the telescope itself. A series of specular reflections, scatters and diffractions propagates this energy to the focal plane. The sum of the various paths yields the focal plane irradiance. The telescope may be of Cassegrain, Gregorian or Herschelian design, as well as prime focus. A two-section conical aperture shade is included, which may be sliced off at an angle to the optical axis. Data on Class 100, 300 and 500 mirror contaminations, as well as a number of clean mirrors, are available for BRDF calculations. The stray radiation signal accumulated after a series of telescope nods and/or chops against a time-varying background may also be evaluated. The user enters the nodding and chopping frequencies, the chopping amplitude and duty cycle, and the time rate-of-change of the off-axis angle or temperature.

The Arizona's Program for the Analysis of Radiation Transfer/Program for the Analysis of Diffracted Energy (APART/PADE) stray light analysis program has been under continuous development and maintenance for about 15 years, and represents over 25 man years of development. APART/PADE is a deterministic computer program that can be used to calculate the propagation of stray radiation via scatter, diffraction, or thermal emission, and it can mix and match the modes in any manner. It has undergone dramatic changes in its structure, code, capabilities, and even its name, in the past seven years. The later change in name has been necessary in order to keep it properly in step with the program's capabilities. The latest enhancement allows the program to perform the deterministic calculations using real rays, thereby removing one of the long-time limitations of APART/PADE. Greatly enhanced graphics have added a new dimension by allowing the user to clearly view the modeled sensor from various positions in order to see that the system is correctly modeled. Internal and user-definable macro and help files, and even user-definable Fortran BRDF subroutines can open up virgin territory to meet the needs of the users. A simple, single baffle command allows the user to direct the program to make a first cut baffle design for eccentric pupil designs, from the entrance aperture to the detector if desired. Many of the other new factors are described in the following text.

In well-baffled optical systems it is expected that improved stray light suppression can be obtained by using a specular black coating for baffle surfaces as compared to a diffuse black coating. In this context, a well-baffled optical system consists of an all-reflective, narrow field-of-view, unobscured reimaging optical design with a field stop and Lyot stop and baffle vanes for trapping stray light. There is a concern that a specular baffle coating will cause problematic focusing of stray light and result in excessive or spatially varying stray irradiance at the focal plane. A computer simulation was undertaken to study these issues using the General Unwanted Energy Rejection Analysis Program (GUERAP). Three different baffle configurations were studied. They were obtained by optimizing for three different baffle coatings. The coatings are a perfectly specular coating, an ideal Lambertian coating, and Cat-a-lac black coating as measured at wavelength 10.6 μm. Each of the three baffle designs was simulated using each of the three coating models. The point source transmittance (PST) was calculated and the specular baffle coating was shown to give superior stray light suppression, particularly at large off-axis source angles. Specular baffle coatings are not recommended for obscured or nonreimaging systems due to the possibility of specular glints propagating to the focal plane.

The main goal of the Hipparcos mission is to measure the position of 100 000 stars, down to magnitude 9, within an accuracy of 2 mas. Achievement of these performances is only possible by reducing sufficiently observation dead times due to straylight originated by earth or moon. After a brief description of the Hipparcos mission principles, this paper will describe how straylight protection requirements have been taken into account in the design of the Hipparcos telescope and in the development of the corresponding hardware. The logic and methods used to demonstrate compliance with straylight protection requirements will be presented and the tests methods and test hardware will be described. First test results will be briefly given.

Various types and sources of grating stray light are described. A format for stray light measurements is presented so that stray light measurements made by different investigators can be directly compared. Finally an investigation is made to determine an order of magnitude Bidirectional Reflectance Distribution Function for input into the stray light programs APART and GUERAP for the narrow field of interference (holographically recorded) gratings.

A stray light analysis of the Cryogenic Limb Array Etalon Spectrometer (CLAES), which is to be flown on NASA's Upper Atmospheric Research Satellite (UARS), was performed utilizing the Arizona's Paraxial Analysis of Radiation Transfer/Paraxial Analysis of Diffracted Energy (APART/PADE) computer code. Developed by the Electro-optics Laboratory of the R&D division of Lockheed Missiles and Space Co., Inc., CLAES is a Gregorian-Mersene optical system in a z-configuration especially designed to minimize the amount of earth infrared radiation scattered to the detector. Diffraction phenomena are controlled by field and aperture stops, thus becoming a second order effect. A high level of stray light rejection is crucial for CLAES in order to detect the signatures of trace molecular species in the stratosphere. Nominally, the earth's limb will be imaged 0.20° off the edge of the detector array. The dominant source of scattered radiation in this system is produced by the laam and primary mirrors, therefore, the polishing, coating and cleaning of these surfaces is vital to the ultimate overall system performance. Contamination control has thus become an important issue for the CLAES project. In order to help evaluate the accuracy of the CLAES APART/PADE model and to validate diffraction and scattered light control, a stray radiation analysis of the CLAES Brassboard Telescope was undertaken. The resultant Point Source Transmittance (PST) predictions are compared to actual off-axis response measurements. BRDF measurements of the Brassboard mirrors were used as a means to obtain the optimal CLAES APART/PADE model. Both the analytical results of the CLAES instrument and the Brassboard are presented in this paper.

A systems level analysis technique has been developed for predicting the direct solar scatter in the focal plane of an optical sensor due to particle deposition on critical surfaces. The technique consists of: 1) developing a simplified scattering model for predicting the baseline sensor performance, 2) developing predictions of the surface cleanliness levels of critical sensor optical surfaces based on the expected exposure to a specified contamination environment, 3) developing the particle scattering functions for the critical sensor surfaces using a geometrical scattering model, and 4) combining the particle scattering functions with the system model and predicting the contaminated system performance. This paper describes this process using a typical Cassegrain sensor as a sample demonstration case. The particle scattering model is developed and compared to measured particle scattering functions. The system scattering model is developed and compared to system scattering test data. Surface cleanliness levels are predicted based on facility particle fallout test data for a standard sensor assembly and test timeline. The facility fallout data are calculated at sensitive surface locations using a full-scale mockup of the sensor. The overall sensor performance is predicted for the clean and contaminated system based on the total scattered power incident on the sensor detector.

The origin of scattered light, particularly near-angle scattered light, and its effect on optical resolution have been analyzed. Also, an instrument is described that can measure near-angle scatter. An application of Rayleigh and Mie scattering theories shows that a few large defects cause a high level of near-angle scatter, which leads to a loss of resolution. The degradation in resolution is severe when trying to resolve a bright object near a faint object. A relatively simple apparatus has been built that can measure near-angle scatter at angles as small as 0.03 degrees and at levels of 10^-6 peak intensity.

A technique for monitoring the degradation in performance of an infrared instrument caused by contamination of the primary mirror is needed. A scatterometer, measuring the light from an off-axis source transmitted through the instrument, is proposed. The design and analysis of this device is described.

The following is an edited transcript of the panel discussion held at the SPIE Conference called "Stray Radiation V" at SPIE's 30th Annual International Technical Symposium. The Panel Discussion was on Stray Light Problems, Concepts, and Tools.

It has been proposed to use interferometry to suppress scattered radiation in optical systems. We have measured gains in signal-to-noise ratio using a model interferometric design with monochromatic light and an extended scatter source. The available gains are about an order of magnitude.

Stray light from optical components in BRDF/BTDF measuring instruments is investigated regarding small angle limitations. Angular limits and associated features in the instrument function have been identified. The angular characteristics of the focused laser beam which is specularly reflected/transmitted by the sample are derived in terms of the Gaussian distribution. Measurements of BRDF and instrument function are consistent with the analysis.

Reflective baffles are proposed to reject off-axis light entering a telescope. Toroidal surfaces and adjacent cones are positioned so that off-axis rays make multiple reflections between these two surfaces. Meridional rays are reflected approximately parallel to the entering direction. Skew rays are reflected obliquely, but leave the telescope aperture. The range of incident angles for which these reflections are obtained is approximately 45°. A system is described that is designed specifically for the Space Shuttle Infrared Telescope Facility (SIRTF). Because of its reflective properties, the proposed baffle system rejects about 90° of the heat load from the SIRTF sunshade that would be absorbed in systems of conventional black baffles.

This paper presents the results of a full three-dimensional analysis of a new type of reflecting baffle vane cavity proposed by Dr. William Linlor of NASA's Ames Research Center. The analysis is divided into two separate parts. The first employs advanced geometrical ray tracing techniques (the ASAP/RABET computer program) to the analyze the specular reflecting properties of the cavities. The second part uses deterministic radiation transfer methods (a later version of the APART/PADE stray radiation analysis program) to analyze the diffuse scattering properties of the reflecting cavities relative to absorbing ones. In both cases, an operating wavelength of 10.6 microns is initially used. These results are then used to estimate the performance at 2.0 microns.

A helium-neon portable scatterometer that was built and described previously has been modified. The intent was to obtain better data at angles very close to specular. We incorporated a twelve-element, linear array of silicon detectors in place of the first detector near the specular reflection direction, and removed the collimator to reduce the beam diameter. Attention was paid to scattered light from the instrument. A new scheme to orient the instrument to the sample surface using the array and the specular beam is described. Design details of the detection system capable of detecting signals ranging over five orders of magnitude and utilizing lock-in detection software are given. Results showing that the modification improved the scatterometer performance so that a Bidirectional Reflectance Distribution Function (BRDF) value of 1 sr^-1 at one degree from specular can be measured with reasonable accuracy for a specular sample are discussed.

Using effective medium theory, the surface roughness and the void fraction of very rough surfaces can be measured nondestructively using infrared ellipsometry. The results correlate well with stylus measurements in cases where the latter can be made. The ellipsometric technique may be useful in evaluating the roughness of optical baffle materials, which are usually too rough to be conveniently measured using other techniques.

Spurious response in a scanning optical system was traced to various pathways for stray radiation. The term "spurious response" refers to any false pulses appearing in the signal-train of an electro-optical system. Such responses can result from a variety of stray light paths. A description is given of a particular scanning optical system which exhibited several modes of spurious response.

Low scatter optical surfaces degrade as a function of time even under controlled clean room conditions. A technique to determine the clean room cleanliness level and period of exposure to achieve a given change in thebidirectional reflectance distribution function is described. Class 10 clean rooms will protect low scatter optics during assembly and test.

An experiment using low energy oxygen ions to clean organic films from bare aluminum mirrors was performed. Film removal was determined by measuring the reflectance of the mirrors in the ultraviolet region of the spectrum. The results of this study show that complete removal of hydrocarbon films is obtainable. This method may not be fully effective in removing silicon-based polymers. The removal rate for a hydrocarbon oil contami nant was determined to be 2.1 X 10^-14 Å/ion.

Three models have been developed to describe light scattering from a sphere positioned on a mirror. The actual system is modeled as a two-sphere system consisting of a real and an imaginary sphere. Light scattering from each sphere is assumed to follow Mie theory. No further interaction is assumed between the sphere and the mirror. The models differ in their consideration of the interaction of scattered fields from the spheres. The results of the models are compared with experimental values with the size parameter 4.88 and 7.00 with s- and p-polarization. The comparison indicates that a best fit can be obtained by assuming the two spheres are coherent light sources with a phase difference ofπ.