We establish a fundamental bound on the field of view over which strictly uniform far-field irradiance can be achieved in symmetric 2D (trough-like) and 3D (cone-like) illumination systems. Earlier results derived for particular 2D devices are shown to be special cases of the general formula. For a rotationally-symmetric 3D luminaire with a lambertian disc light source and a prescribed uniform cone region half-angle (theta) _c, no more than tan²((theta) _c) can be projected within a uniform core region. Hence the efficiency with which such illuminators can produce uniform flux is severely limited for many problems of practical interest. Being guided by the Tailored Edge-ray Device formalism for the design of 2D luminaires, we develop a 3D reflector that produces extremely uniform far-field illuminance.

We develop a range of practical nonimaging devices for optical fiber applications where rays emerging from a fiber over a restricted angular range (small numerical aperture) must illuminate a small near-field detector at maximum radiative efficiency. These designs range from pure reflector (all-mirror), to pure dielectric (refractive and based on total internal reflection), to lens-mirror combinations. Sample designs are presented for a specific infrared fiberoptic irradiation problem of practical interest. Optical performance is checked with computer 3D raytracing. Compared to conventional imaging solutions, our new nonimaging units offer considerable practical advantages in compactness and ease of alignment, as well as noticeably superior radiative efficiency.

The design of non-sequential mirrors within the Simultaneous Multiple Surface design method of nonimaging concentrators is presented. The formulation of the edge ray theorem for this problem is stated, which defines the rays to be considered in the design. As a result three new nonimaging concentrators are developed, which work efficiently and close to the thermodynamic limit of concentration in two dimensions.

Theoretical upper limits on measures of flux-transfer performance due to skewness conservation in rotationally symmetric nonimaging optical systems have recently been discovered and quantified. These limits can have an adverse impact on the performance of projection or coupling optics which collect light from 3D sources. In this contribution we show that these limits can be exceeded by employing nonrotationally symmetric configurations. We consider the problems of maximizing flux transfer from both a homogeneous spherical source and a homogeneous cylindrical source to a homogeneous disk-shaped target of equal etendue. We find that the performance limits due to skewness conservation for these problems can be overcome by numerically optimized reflectors possessing a nonrotationally symmetric star-like cross-section.

In this paper we propose a particular experimental setup for measuring the instrument function, the Weyl transform of the non-negative definite Hermitian operator representing the effect of the instrument in a measurement. Once the instrument function is known, the results of various measurements involving wavefields of any state of coherence can be calculated. For simplicity, we will concentrate our discussion on the case of one transverse spatial dimension.

In this paper, the phase-space formulation is applied to the evaluation of nonimaging optics sub-systems. Brightness (luminance) efficiency is introduced as a Figure of Merit for system performance maximization procedures that can be applied, for example, to plasma diagnostics (by utilizing coherent fiber imaging).

The conditions of the radiance increase are analyzed in detail on the base of the equation of radiance transfer and requirements of thermodynamics. It is shown, that the most important of these conditions are the nonreversability of the beam propagation process and inappropriateness of the beam approach.

A 2D-error model for nonimaging concentrators composed by multiple optical surfaces is presented. The concentrator surfaces can be either dioptrics, sequential mirrors (as a conventional parabola) or non-sequential mirrors (i.e. CPC- like mirrors). Under the hypothesis of the model, the slope errors of all the surfaces can be transferred to entry aperture and combined there, and the effect of errors can be studied with a single probability density function as in the case of one-mirror concentrators. A single number, the concentrator error multiplier, is defined to characterize the concentrator tolerance to errors. This number and the concentrator acceptance angle are the key to analyze the error sensitivity of concentrators. Finally the model is used to quantify the maximum tolerance on the concentrators surfaces to guarantee a specified quality.

The problem of generating a uniform irradiance distribution on a target plane using extended sources has been addressed by several authors in recent arti1 The techniques described in these papers are generally applicable to any source to target distance. However, most effort has been put into uniform irradiance on a distant target ( one whose distance is large compared to the size ofthe illumination source and its uminair'24 ' The techniques ofnon imaging optics have generally been applied to design these luniinaires. The tailored edge-my method is used to adjust the irradiance on target while maintaining collection efficiency at or near the theoretical limit. The present effort was stimulated by the desire to heat a substrate with a high degree of uniformity. The sources are tungsten-halogen filament lamps and the luminaires surround them. The luininaires in this work were consirained to fit into existing hardware and replace existing conventional luininaires. Section 2 lays out the theory ofluminaire design and the methods used to generate the luminaire shape. Section 3 provides the method of tracing rays through the luininaire to test the uniformity of irradiance. Section 4 provides the results of the effort to get uniform irradiance and Section 5 summarizes these results and draws some conclusions.

We present a novel condenser system based on a nonimaging TIR lens and an associated apodizing aspheric lens. This system provides uniform collimated illumination of the image plane of a projector such as an overhead viewgraph or LCD projector. The TIR lens enables the system to be very compact. The aspheric lens lies between the light source and the TIR lens, so that the TIR lens output illumination is spatially uniform.

We present two new applications for light emitting diodes of the Total Internal Reflection (TIR) lens, a non-imaging optical device presented at previous SPIE conferences on nonimaging optics. The first is a flat circularly symmetric lens that efficiently forms a highly collimated beam from the light output of Hewlett-Packard's Super Flux LEDs. The second is a linear TIR lens with die-on-board LEDs of several wavelengths positioned along its focal line. HP's Super-Flux LED package has an output half angle of 55 degree(s). Only the TIR lens can accept such a wide range for beamforming, and do it with high efficiency. We have designed and prototyped 1' models with half-power half angles of only 1.5 degree(s), utilizing a hyperbolic central section in place of the usual Fresnel lens. There are numerous applications for arrays of these lenses, since they emit more lumens per electrical watt than filtered incandescent lamps with parabolic mirrors. Moreover, they are more compact than conventional lamps, and LED lifetimes are much longer. The TIR lens in its linear form has been applied successfully to fluorescent downlighting products with much narrower transverse illumination angles than previously available with trough mirrors. More recently, light emitting diodes (LEDs) have been placed on the focal line of a linear lens. In this paper, we describe the optical properties and biomedical applications of the linear TIR lens when the LEDs have several different emission wavelengths. This single device can uniformly illuminate an extended target with several wavelengths either simultaneously, sequentially, or in complex programmed combinations. It can replace the complex systems of dichroic mirrors used with conventional white-light sources.

Beam-shape transformers are used in high-efficiency projectors to match round source beams to rectangular LCD targets. Two common alternative approaches use (1) light- pipes and (2) lenslet arrays. Both approaches outperform the simple overfilling used in simpler systems. We review the two approaches conceptually and develop simple theoretical models which elucidate each approach's trade-offs between uniformity, efficiency, etendue, and compactness. For the light-pipe approach, we develop a detailed analytical theory for idealized, uniform sources, and we compare the predictions to ray-trace results. Finally, we compare the two approaches and find that the light-pipe approach offers similar performance with higher compactness. The lens array approach, however, may be favored in complex systems where multiple elements already result in long optical trains, or in systems where etendue conservation is not a priority.

For the last ten years, the company OPTIS has worked on non imaging system design software. The software is needed to simulate a complete photometric path that could take into account all the major needs of optronic system designers. Different ways of simulation the system photometry will be described, showing the improvement which has come through experience. The analysis and simulation include different parts of an optronic system. The main input data includes: characterization of sources in the near and far fields, thermal emission information, geometrical data, and the surface and volume properties of materials. The emission and propagation of radiation through the system with the material interactions (refraction, reflection and scatter) are simulated for coherent and incoherent light. The simulation uses a Monte Carlo method which has been improved for use in stray light analysis. The main research emphasis has been on developing and testing algorithms to reduce the calculation time without sacrificing precision. This is a fundamental point we describe for IR countermeasures applications and for cases of very low levels of signal flux.

The heliostat field plays a crucial role in defining the achievable limits for central receiver system efficiency and cost. Increasing system efficiency, thus reducing the reflective area and system cost, can be achieved by increasing the concentration and the receiver temperature. The concentration achievable in central receiver plants, however, is constrained by current heliostat technology and design practices. The factors affecting field performance are surface and tracking errors, astigmatism, shadowing, blocking and dilution. These are geometric factors that can be systematically treated and reduced. We present improvements in collection optics and technology that may boost concentration (up to 20,000 peak), achievable temperature (2,000 K), and efficiency in solar central receiver plants. The increased performance may significantly reduce the cost of solar energy in existing applications, and enable solar access to new ultra-high-temperature applications, such as: future gas turbines approaching 60% combined cycle efficiency; high-temperature thermo-chemical processes; and gas-dynamic processes.

A new approach towards large scale solar systems is described. Solar Reflective Tower systems for thermal applications that can achieve peak solar concentration of 10000 are the first systems that can be used for combined cycle gas turbines. Using a combination of high solar concentration and spectrum splitting it is possible to operate simultaneously solar thermal, solar lasers and concentrated photo-voltaic systems. After demonstration in laboratory scale two industrial projects were initiated to develop the technology to industrial scale. Within these projects, close cooperation between several industries, universities and a research institute take place. The goal of these projects is to industrialize the technologies towards the year 2000.

The wireless international communication these days faces a practical bandwidth limitation since the frequency band offered by satellites is almost exhausted. In addition, the energy source of a satellite is limited, hence, broadening the frequency domain in a wireless system increases the cost of operation. Free-space optical communication, using direct sun to laser conversion, can contribute to overcome these problems. A lab simulation for outer space solar laser communication was performed. A combination of imaging and non-imaging concentrators for pumping solar laser (oscillator and amplifier) were built and analyzed. An approximated 3D-CPC was designed, assembled and tested. The laser crystal characteristics were measured under the system's pumping conditions. An oscillator-amplifier system was designed built and analyzed. Direct and indirect gain measurements were done. Gain up to 1.4 was measured at a low sun flux of 630 w/m². Communication experiments at a boud rate of 1.5 MHz were conducted using intensity modulation outside the cavity.

Two new types of two-mirror solar concentrator for tubular receiver, the snail concentrator and the helmet concentrator , are presented. The main feature of these concentrators is that they have a sizable gap between the secondary mirror and the absorber, and they still achieve concentrations close to the thermodynamic limit with high collection efficiencies. This characteristic makes them unique and, on the contrary to the present two-stage designs, allows for the location of the secondary outside the evacuated tube. One of the differences between the snail and the helmet concentrators is that the last is symmetric (as the conventional parabolic trough) but the first is not. For an acceptance angle of (alpha) equals +/- 0.73 degs and a collection efficiency of 96.8% (i.e. 3.2% of the rays incident on the primary mirror within the acceptance angle are rejected), the snail concentrator and the helmet concentrator achieve an average flux concentration of 91.1% and 72.8% of the thermodynamic limit, respectively. The gap between the absorber and the secondary mirror is 6.8 and 12.1 times the absorber radius for each concentrator. Moreover, both concentrators have also high rim angles of the primary mirror: +/- 86.2 degs (helmet) and 3.1 - 98.8 degs (snail). This is of interest for a good mechanical stability of the collector.

Flexible optical fibers and fiber bundles can be used to transfer solar energy to a desirable place, where it could be used either to pump a laser crystal or to carry out other useful mechanical, chemical or thermal processings. Two flexible fiber-optic bundles were built. Each bundle consists of 19 optical fibers of 1.5 mm diameter each. The input section of each single fiber is polished to form a hexagonal column. When the input columns were joined together, two compact fiber-optic bundles were formed, leaving no dead space between the fibers and hence, the concentrated solar energy was transmitted without extra loss. Two off-axis parabolic mirrors with hexagonal form were held onto a solar tracker which continuously tracks the Sun. With an incident intensity of 650 W/cm², each primary mirror captured 143 W solar energy and concentrated it into a light spot of hexagonal form, which matches well with the input area of the fiber-optic bundle. Solar energy of 100 W was successfully delivered by each bundle, with transmission efficiency of 70%. The two fiber bundles were also combined to form a large bundle for 200 W solar energy delivery.

Science Applications International Corporation has used a unique nonimaging-optical global optimization computer code, NICOS, to design an innovative secondary concentrator for the National Renewable Energy Laboratory (NREL). NICOS allows for the optimal design of such devices to achieve a variety of irradiance distributions on a desired target. The case of interest to NREL called for a uniform irradiance of concentrated sunlight over a relatively large area and at a reasonable working distance from the exit of the device. Because the irradiance at the nominal focal point of NREL's High-Flux Solar Furnace (HFSF) was reshaped from a near- Gaussian distribution to a nearly uniform one, the designs generated have been called irradiance redistribution guides (IRG). A design featuring reentrant optics was selected for fabrication and testing. This IRG has been fabricated and tested at the HFSF to compare predicted and measured performance. The IRG's performance is close to the theoretical predictions. Much of the performance difference can be explained by discrepancies between the actual HFSF performance relative to that assumed in the NICOS predictions. This IRG will be useful for applications in which uniform solar concentration at moderate flux is required. In general, the design methodology and resulting devices can provide a new way to satisfy diverse flux tailoring needs.

The Analex Corporation, under contract to the NASA Lewis Research Center (LeRC), Cleveland, Ohio, recently evaluated the feasibility of utilizing refractive secondary concentrators for solar heat receivers operating at temperatures up to 2500 K. The feasibility study pointed out a number of significant advantages provided by solid single crystal refractive devices over the more conventional hollow reflective compound parabolic concentrators. In addition to the advantages of higher concentration ratio and efficiency, the refractive concentrator, when combined with a flux extractor rod, provides for flux tailoring within the heat receiver cavity. This is a highly desirable, almost mandatory, feature for solar thermal propulsion engine designs presently being considered for NASA and Air Force solar thermal applications. Following the feasibility evaluation, the NASA-LeRC, NASA-Marshall Space Flight Center, and Analex Corporation teamed to design, fabricate, and test a refractive secondary concentrator/flux extractor system for potential use in the NASA-MSFC `Shooting Star' flight experiment. This paper describes the advantages and technical challenges associated with the development of a refractive secondary concentrator/flux extractor system for this application. In addition it describes the design methodologies developed and utilized and the material and fabrication limitations encountered.

In this paper, we discuss the features of different types secondary concentrators used in solar energy for dish- thermal and high flux applications. We include a preliminary comparison of a new type of nonimaging concentrator with the more traditional ideal concentrators.

Rectangular core multimode fibers with core dimensions of (10 - 22) micrometers X (170 - 200) micrometers and numerical aperture of 0.15 have been designed for high power diode lasers coupling with maximum preservation of laser radiation power density and laser radiation linear state of polarization.

An upper limit on concentration for any optical device has previously been derived from the conservation of etendue. In this contribution we derive more stringent upper limits for the efficiency and the concentration of rotationally symmetric optical devices which are a consequence of skewness conservation. If the desired source and target have different skewness distributions, then losses or dilution or both will limit the performance of the optical system. For all skewness values, for which the source contains more radiation than the target, the difference is lost. Conversely, for all skewness values, for which the target contains a large etendue than the source, the difference remains empty and results in dilution. We calculate the limiting curve of optical transfer efficiency versus concentration relative to the maximum concentration possible and provide a design example that is practically at this limit. We also provide another design example that addresses the challenge posed at the last SPIE meeting, namely to transfer the maximum radiation from a Lambertian spherical source to a disk target of equal etendue under a reflector- to-source minimum distance constraint. We conjecture that even rotationally symmetric problems may benefit from asymmetric optical systems.

We generalize a previous derivation of theoretical upper limits on measures of flux-transfer performance of nonimaging optical systems to include the case of sources and targets having inhomogeneous distributions of radiance and importance weighting. This generalization is of practical importance in understanding the limitations of optimally designed projectors which collect light from complex 3D sources such as filament or HID sources. The performance limits are derived from the conservation of etendue and--for the case of rotationally symmetric optics-- from skewness conservation. The limits on performance are calculated for examples involving the use of rotationally symmetric optics to transfer flux from homogeneous and inhomogeneous spherical sources to a homogeneous disk shaped target having a phase-space volume equal to that of the source. It is shown that an optical system which is optimal for the case of a homogeneous source and target will not necessarily provide the best achievable performance when used in conjunction with an inhomogeneous source and/or target occupying the same regions of phase space as the homogeneous source and target. The theoretical upper limits are shown to be consistent with the performance of three numerically optimized reflector designs.

In solar tower plants, where a rotationally symmetric field of heliostats surrounds the tower, an axisymmetric secondary concentrator such as a Compound Parabolic Concentrator (CPC) or a tailored concentrator or a cone is the obvious choice. For locations at higher latitudes, however, the reflecting area of the heliostats may be used more efficiently if the field of heliostats is located opposite to the sun as seen from the tower. Then the field is asymmetric with regard to the tower. In the case of an asymmetric field, an axisymmetric concentrator necessarily has a concentration significantly lower than the upper limit. Furthermore, the area on the ground from which a tilted axisymmetric concentrator accepts radiation is an ellipse, including also heliostats very distant to the tower producing a large image of the sun. Therefore we investigate asymmetric secondaries. From the shape of the edge ray reflectors constructed for rays in the central south-north plane we conclude that a skew cone reflector might be appropriate for the field and optimized its free parameters by means of raytracing. Asymmetric concentrators may increase the concentration by up to 25% at the same efficiency compared to optimized axisymmetric CPC or cone reflectors.

The amount of light that can be coupled from a noncoherent lamp into an optical fiber of small size and limited numerical aperture is restricted by fundamental principles. High power lamps have the additional problem that there is too much light resulting in burned fiber or wasted light. This paper describes a new high efficiency coupling approach that is particularly applicable to high power lamps such as the 1000 Watt electrodeless Sulfer lamp manufactured by Fusion LIghting Inc. A semi spherical array of lenses is used to focus light onto individual large core fibers or fiber bundles. The design constraints are the size and shape of the optical source, and the size and numerical aperture of the fibers. The variables are the number of lenses in the array (which determines the number of fibers) and the lens power. There are enough degrees of freedom to design a system that can theoretically couple nearly 100% of the light into the fibers.

The recent development of large core diameter (5 to 25 mm) flexible illumination fiber raises some interesting questions about the properties and theory of geometric illumination beams. Examples are shown of the peculiar beam properties of light from large diameter, large numerical aperture fibers focused by a lens. It is demonstrated that the axial position for the tightest beam focus does not occur at the location predicted by the conventional geometric imaging equation. This focus shift is very similar to that found with coherent laser beams and described by the Gaussian Beam Propagation Equations. In fact in the geometric illumination beam case it is found that nearly identical beam focus equations apply. Furthermore the illumination beam equations can be derived entirely from geometric optics, independent from wave optics, and are simply the refraction equations for a paraxial skew ray. These theoretical considerations are shown to have a significant impact on the design of fiber coupling and other beam handling optics for use with the new large core fiber. Examples, such as 90 degree bend optics and other optical `plumbing' are given.

A lab simulation which demonstrates the power beaming concept, based on solar pumped laser and photovoltaic cell, was performed. The simulation included a parabolic dish, a 3D CPC, a 2D CPC as a laser head for transmission and a photovoltaic cell for converting the laser light into electricity. A waveguide was used in order to obtain a uniform illumination upon the photovoltaic cell. A Nd:YAG laser rod was solar pumped using imaging and nonimaging systems producing 52 Watt laser at sun flux of 830 Watt/m². In successive experiments the solar cell was exposed to a laser light using Nd:YAG and Alexandrite lasers. The efficiencies achieved were 33% laser to electricity efficiency for the Nd:YAG laser and approximately 40% for the Alexandrite.