Many papers have been written on the design of gradient-index optical systems. The purpose of this paper is to review current technology in gradient-index materials research and to review the types of materials that are currently available. These include glass, materials made by ion exchange, chemical vapor deposition, sol-gel, and phase-separated techniques, gradient-index germanium, zinc-selenide/zinc sulfide, and gradients in plastics.

Optical designers were among the first to use the computer as an engineering tool. Powerful programs have been written to do ray-trace analysis, third-order layout, and optimization. However, newer computing techniques such as database management and expert systems have not been adopted by the optical design community. For the purpose of this discussion we will define a relational database system as a database which allows the user to specify his requirements using logical relations. For example, to search for all lenses in a lens database with a F/number less than two, and a half field of view near 28 degrees, you might enter the following: FNO < 2.0 and FOV of 28 degrees ± 5% Again for the purpose of this discussion, we will define an expert system as a program which contains expert knowledge, can ask intelligent questions, and can form conclusions based on the answers given and the knowledge which it contains. Most expert systems store this knowledge in the form of rules-of-thumb, which are written in an English-like language, and which are easily modified by the user. An example rule is: IF require microscope objective in air and require NA > 0.9 THEN suggest the use of an oil immersion objective The heart of the expert system is the rule interpreter, sometimes called an inference engine, which reads the rules and forms conclusions based on them. The use of a relational database system containing lens prototypes seems to be a viable prospect. However, it is not clear that expert systems have a place in optical design. In domains such as medical diagnosis and petrology, expert systems are flourishing. These domains are quite different from optical design, however, because optical design is a creative process, and the rules are difficult to write down. We do think that an expert system is feasible in the area of first order layout, which is sufficiently diagnostic in nature to permit useful rules to be written. This first-order expert would emulate an expert designer as he interacted with a customer for the first time: asking the right questions, forming conclusions, and making suggestions. With these objectives in mind, we have developed the Optics Toolbox. Optics Toolbox is actually two programs in one: it is a powerful relational database system with twenty-one search parameters, four search modes, and multi-database support, as well as a first-order optical design expert system with a rule interpreter which has full access to the relational database. The system schematic is shown in Figure 1.

It is demonstrated that realistic unobscured mirror systems may be fabricated using only ordinary (rotationally symmetric) surfaces. The low-order aberrations may be balanced by careful selection of the curvatures and tilt angles. The remaining higher-order aberrations are tolerable for slow systems and may be corrected with slight asymmetric surface departures for faster systems.

A null-lens design approach for centrally obscured components is described. This opti-mization method, which explicitly takes advantage of the obscuration, can achieve improved wavefront quality of the null test for a given level of lens complexity than a design approach that does not take advantage of the obscuration. The fabricated aspheric part, while likely to be different in shape from that predicted by the null test by a constant phase factor, will still allow satisfactory performance of the system.

Excellent results for optimization of one all-reflective system are gained by choosing a limited number of variables as indicated by vector aberration theory. An aberration analysis shown here can be used to extend the method to other systems.

A satisfactory bifocal contact lens has not yet reached the marketplace because of deficiencies in the present designs. The two-zone concentric design allows both the near and far images to be present simultaneously, however the image quality is not very good. Recently several new bifocal contact lenses with multiple zones have been suggested. This paper presents computer calculations of the image quality of several of these new designs. The zone-plate design, with an intermediate blaze angle, gives the best image quality of the lenses that were examined.

A monocentric element, that is, one whose surfaces are concentric, can be valuable in an optical design. Such elements are, however, very difficult to fabricate. A departure from true monocentricity will make the fabrication problem more tractible and can usually be achieved without loss of design performance. A relationship is derived which will indicate the minimum departure from monocentricity necessary to make fabrication feasible.

I have been asked to speak on the flight of optical manufacturing to the far east and because my wife has long ago taught me to do as I am told, I will do it, but it would be much more fun and perhaps productive, to discuss with you ways to overcome such flight. Don't tell Warren Smith, but I intend to use some of my alloted time to discuss this with you.

Optical performance results are reported for four different molded-glass lenses. All four lenses are intended for use as focus objectives in optical memory applications, including Compact Disc, Laser Disc, and optical storage pickups. All lenses have measured performance below 0.04 waves RMS on-axis at 0.780 micron. The lenses differ in the different performance features they emphasize. These features include excellent off-axis performance, longer working distance, a finite conjugate, and a higher numerical aperture.

Hybrid mounting of optical components, combined perhaps with integrated optical waveguides and lenses on a large area silicon, wafer-scale integrated (WSI) electronic circuit provides one potential approach to combine advanced electronic and photonic functions. The desire to achieve a high degree of parallelism in multi-wafer WSI-based architectures has stimulated study of three-dimensional interconnect structures obtained by stacking wafer circuit boards and. providing interconnections vertically between wafers over the entire wafer area in addition to planar connections. While presently it is difficult for optical interconnects to compete with electrical interconnects in the wafer plane, it is appropriate to look at vertical optical interconnections between wafer planes since the corresponding conductive structures would be large in area and may impede system repairability. The ability to pass information optically between circuit planes without mechanical electrical contacts offers potential advantages for multi-wafer WSI or other dense three-dimensional architectures. However, while optical waveguides are readily fabricated in the plane of the wafer, waveguiding vertically through the wafer is difficult. If additional processing is required for waveguides or lenses, it should be compatible with standard VLSI processing. This paper presents one straightforward method of meeting this criterion. Using optical device technology operating at wavelengths beyond the ≈1.1μm Si absorption cutoff, low loss, through-wafer propagation between WSI circuit boards can be achieved over the distances of interest (≈1mm) with the interstitial Si wafers as part of the interconnect "free-space" transmission medium. The thickness of existing VLSI layers can be readily adjusted in featureless regions of the wafer to provide antireflection windows such that the transmittance can be raised to ≈77% for n-type and to ≈97% for p-type silicon. Optical interconnect source-receiver coupling efficiency can be increased and crosstalk decreased using VLSI process compatible apertures and phase reversal zone plate lenses.

In this paper, the measuring accuracy about the angstrommeter is searched, and a special measuring method about the thickness of the optical coating is developed. A "fully assembling optical coating block" is used successfully. It consists of three kinds of optical coating from which 7 sorts of thickness can be formed. Properly choosing the arrangement model, we can composs 19 coudition equations. According to the theory of matrix square-error, we deduced the measuring accuracy of the coating thickness and the measuring accuracy of the interferometer itself. The method is also useful for some other films.

A Technique For Monitoring Optical Thin Film Thickness By Direct Evaporation Onto The End Of An Optical Fiber Is Described. The Optical Fiber Is Suspended Inside A Coating Chamber So That The Fiber Distal End Faces The Evaporation Source And Is Illuminated With A Chopped Light Source. The Light Which Is Reflected From The Fiber End Is Returned Through The Fiber And Is Collectyed By A 45° Mirror With A Hole In It. The Signal Is Detected And Displayed On Led Displays And A Chart Recorder. Multiple Fibers Are Used To Monitor Individual Layers In A Stack, Or A Single Fiber Is Used After Cleaving The End With A Tool To Produce A Clean, Flat, Uncoated Surface Without Breaking The Vacuum.

A review of the application of optical interferometry to aerodynamics measurements and the determination of optical degradation due to the aircraft flow field is presented. Early work on transonic aerodynamics demonstrated that holographic interferometry could provide accurate flow visualization and quantitative results for general wind tunnel testing. Two-dimensional transonic flow fields investigated were used to obtain quantitative measurements which were compared to other flow measurements including surface pressure and laser Doppler velocimeter data. Even in large-scale facilities, the results were in good agreement. Investigations on the related problem of optical transmission through turbulent compressible flow fields were also discussed. These results provided information on the optical distortion produced by the flow field in addition to information on the flow field itself.

The advantages of solid state laser scanning are explained and contrasted with the conventional software-intensive approach to machine vision. A new laser gauge design is described and applications are explored.

When defects in optical elements scatter light they change its direction, intensity, and polarization. We discuss defect scattering in the context of the Stokes vectors and Mueller matrices that characterize the interaction.

An imaging system model, based upon linear systems theory, has been developed for a microcomputer with less than 32K of free random access memory (RAM). The model includes diffraction effects of the optics, aberrations in the optics, and atmospheric propagation transfer functions. Variables include pupil geometry, magnitude and character of the aberrations, and strength of atmospheric turbulence ("seeing"). Both coherent and incoherent image formation can be evaluated. The techniques employed for crowding the model into a very small computer will be discussed in detail. Simplifying assumptions for the diffraction and aberration phenomena will be shown along with practical considerations in modeling the optical system. Particular emphasis is placed on avoiding inaccuracies in modeling the pupil and the associated optical transfer function knowing limits on spatial frequency content and resolution. Memory and runtime constraints are analyzed stressing the efficient use of assembly language Fourier transform routines, disk input/output, and graphic displays. The compromises between computer time, limited RAM, and scientific accuracy will be given with techniques for balancing these parameters for individual needs.

This paper shows how Soramerfeld's multiple-valued function method can be extended to solve Fresnel three-dimensional diffraction problems for a point radiation source and perfect conductors in two different configurations: a semi-infinite plane and a circular disc. Characteristically, the solutions are in closed form, can be directly calculated and are without any restriction on the relative sizes of radiation wavelength and scatterer.

Sommerfeld's multiple-valued solution of Maxwell's equations is applied to the diffraction of a plane wave by a flat conducting strip and a circular conducting channel. Solutions are derived valid for any given wavelength and in closed form. The method of applying Sommerfeld's procedure indicates how Sommerfeld's solution can be generalized.

The propagation of a generally astigmatic Gaussian beam through an arbitrary optical system can be expressed solely in terms of coordinate independent transverse ray vectors. With this formulation, it is relatively easy to take standard ray trace data and deduce the equivalent Gaussian beam shape. In addition, by a simple rearrangement of the equations, the beam shape can also be expressed as the convolution of two geometrically propagated Gaussians.

The author has developed a computational algorithm for the calculation of Kirchoff-Fresnel diffraction in optical systems and radio antennas [1]. It can be used for investigating the complex amplitude of a wavefront in most regions of interest in imaging optical systems. The method employs the two-dimensional Fast Fourier Transform (FFT), and thus is computable in time proportional to N log N, where N is the number of picture elements in the wavefronts to be analyzed. The paper consists of two parts: Part I briefly describes the algorithm, discusses computational errors, and gives some comparisons to other calculations. Part II describes several applications in which the algorithm has been used, including the design of a lens for the improved efficiency of radio antennas, and the study of aberrations in optical systems.

Methods are presented for calculating the radiation intensity in the focal region resulting from the focusing of a laser beam by a spherical surface lens. An approach based upon Fourier optics is employed to include the effects of aberrations as well as diffraction and interference, and digital signal processing techniques are used to speed the calculations. Calculated results are displayed on a digital image processor to aid in visual interpretation. Sample calculations are presented for intensity fields in the transverse and meridional planes of the focal region for circular, gaussian and annular laser beams incident on the lens perpendicularly and at small oblique angles.

In this paper the applications of multiplex holograms are discussed for optical processing of wideband spectral signals. The comparative analysis of different multiplex systems can be performed by using Fourier space formalism, particulary in determining the angular acceptance and dispersion of holographic structures. The limitations of transmission and reflection holograms are analyzed, including cross-coupling effects, chromatic dispersion and bandwidth control.

Experimental observations and measurements are essential sources of information for the correct development of mathematical models of real materials. Moire interferometry offers high sensitivity in full-field measurements of the in-plane displacements on the surface of a specimen. Although it is a powerful method in experimental stress analysis, it has some shortcomings. One is that existing systems require highly coherent light. The only sufficient source of light for this application is a long cavity laser, which is relatively expensive and at best quite cumbersome. Another shortcoming is that measurements must be performed in a vibration-free environment, such as that found on a holographic table. These requirements limit the use of existing moire interferometers to a holographic laboratory. In this publication a modified concept of compensation is developed that permits the use of a chromatic source of light in a compact moire system. The compensator provides order in the angles of incident light for every separate wavelength, so that the virtual reference gratings created by each wavelength in a continuous spectrum are identical in frequency and spatial position. The result is a virtual reference grating that behaves exactly like one created in coherent light. With this development the use of a laser diode, which is a noncoherent light source of tiny dimensions, becomes practical. The special configuration of the optics that create the virtual grating allows its synchronization with the specimen grating and leads to the design of an interferometer that is relatively insensitive to the vibrations found in a mechanical testing laboratory. Sensitivity to relative motion is analyzed theoretically. This development provides the opportunity to apply moire interferometry to solid mechanics problems that cannot be studied in an optics laboratory. Experimental verification of the optical concepts is provided.

The structure and design of diffractive optical diffusers for display applications is discussed and a new technique for designing complex relief structures with new, improved diffuser behaviour is described. Such diffusers can be fabricated by laser beam writing techniques. A specific example is given of a diffractive diffuser for improving the subjective quality of color CRT data displays by acting as a low-pass spatial-frequency filter to reduce the noise generated by the matrix dot structure.

A simple but accurate method is described for calibrating detector responsivity The method involves creating a far-field intensity pattern via diffraction from a circular aperture, and then computing the integrated irradiance over the detector area using well-known relations. No experimental verification has been performed, but several error sources are discussed and are expected to be small.

Diffraction limited use of optics, especially in Ultra Violet, requires correction of residual wavefront errors ranging from slight decentring to rough ripples. A new and very sensitive method using coherent UV light (257 nm) and a diverging Michelson Twymann interferometric set up, allows to generate by phase conjugation a perfect replica of residual wavefront errors directly in a photoresist thin film, in a one step process. Used in conjunction with the recording mirror or objective, this treated (coated) photoresist compensates any residual wavefront deviations down to possibly x/40 peak to peak. A 10 cm spherical mirror of phase defects amplitude x/14 peak to peak on the mirror surface (x/7 on the wavefront) was used to generate such a phase conjugated corrector. Interferometric testing in the UV (257nm) of the mirror plus corrector assembly yield a very good figure : x/24 peak to peak on the wavefront. Such a system is already diffraction limited down to 200 nm and even further in UV since the correlation length of the residual defects is large. This method could also be applied to large mirrors correction (the interferometer is actively stabilized using synchrone detection) which might significatively reduce their polishing cost.