Th design mcthods of nonimaging optics for illumination by light sourccs which arc of• finite extcnt arc well cstablishcd1. Two classcs of algorithms havc b2cn found which work at or ncar th2 s2cond—law of thermodynamics limit of prformancc. Th2 first of thsc is the "string" or "cdgc—ray" mcthod. It may bc succinctly characterized as a gcncralization of Fcrmat's principle. While all of imaging optics can in principle be derived from: _1 ndl = constant along a ray, the string solutions can be similarly derived from: I ndl = constant along a string. The second class of algorithms places reflectors along the line of flow of a radiation field set up by a radiating source. In cases of high symmetry such as a sphere or disc, we obtain ideal solutions in both two and three dimensions.

Concentrators based on geometrical optics increase the irradiance by increasing the projected solid angle, but conserve the radiance of radiation. The general principle for increasing the radiance, and thereby concentrating even diffuse radiation, resembles a light trap. Light, which enters the trap through a selective filter, is shifted in photon energy, for example, by a Stokes luminescent process. It is subsequently trapped because it is reflected by the filter. Concentration is limited, in the ideal case, by the reverse (anti-Stokes) process, which reaches equilibrium when incoming and concentrated radiation reach equal chemical potential. The laser is discussed as an example for a concentration not limited by thermodynamics. The limits imposed by quantum mechanics are derived. Real systems, with various losses, are discussed.

Analytic solutions have been derived for the optical characteristics of V-trough concentrators, both for accepted and rejected light radiation. The optical efficiency, for beam and diffuse radiation and exchange factors for radiation heat exchange, are expressed in terms of the solutions obtained. Sample results are presented.

The energetic and economic attractiveness of linear solar concentrators can be significantly improved by the use of properly secondary non-imaging (CPC-type) concentrators. Two specific illustrative cases are analyzed. One is the optical re-design of a commercial two-stage solar concentrator which generates process steam at 150 degree(s)C. The primary is a linear Fresnel reflector with one-axis horizontal tracking. The receiver is a stationary, non- evacuated, glazed tubular receiver with secondary CPC. We have re-designed the initial, manufacturer-designed secondary so as to noticeably improve collector thermal output. Details of secondary design and system performance are presented. The other is a new concept in secondary CPC-type concentrators for parabolic trough collectors with tubular receivers and large rim angles (typically 80 degree(s)-120 degree(s)). It had been though that such large-rim- angle concentrators could not benefit from secondary concentrators, since the second-stage concentration boost goes as 1/sin(rim angle). However, by introducing multiple asymmetric CPC-type devices, we can increase the geometric concentration of a 90 degree(s) rim angle parabolic trough by roughly a factor of 3. Furthermore, certain secondary designs can be accommodated within the annulus of currently-manufactured evacuated receiver tubes, and still offer a flux concentration improvement of about a factor of 2.5. Examples of the new secondary designs, and achievable concentration gains, are presented.

A computer simulation of the relative performance of certain truncated symmetrical and asymmetrical fixed reflector designs for solar energy collection was performed. The major results were as follows: (1) Annual solar fractions in excess of 90% seems to be feasible with a fixed load matching collector, in a climate where 70% of hot water requirements is the norm from flat plate collectors. Consumer interaction could either improve or lower this figure, depending upon circumstances. (2) Symmetrical CPC reflectors always gave the best annual output performance per unit of mirror area, and allowed the lowest receiver area for situations of constant annual load. (3) Asymmetrical fixed concentrators are most cost-effective for seasonally asymmetrical load patterns. (4) Fixed parabolic systems were not competitive. (5) The concentration levels utilizable in fixed systems are much higher than previously supposed, with approximately 3.1:1 in an asymmetrical reflector being optimal for the domestic load pattern used. (6) With seasonal load matching, the storage required to achieve solar fractions above 90% appears to be approximately one day of typical load.

Using anidolic or nonimaging optics principles, we present a study of the role of light confinement, with special application to photovoltaics. In particular, we study how and to what extent the reflectance of an absorbing surface can be reduced. Also, we study how and to what extent the absorption of light in a volume can be increased. In both cases the use of light confining cavities is the key for these improvements. Cavities different from the Helmholtz sphere, but rather based in an angular-spatial limitation of the escaping beam, are presented as a key tool for practical designs.

A compact non-imaging lens is described and analyzed: the Totally Internally Reflecting (TIR) lens. It constitutes a major class of optical devices distinct from reflectors and Fresnel lenses. It is a transmissive device that redirects light passing through it via the action of a multiplicity of prismatic facets basically acting as annular Harting-Dove prisms that rotationally `wash-out' image structure. They can achieve much larger bend angles (well over 90 degree(s)) than those of the refraction-only facets of Fresnel lenses. As a consequence, TIR lenses are extremely compact, typically having a thickness about one fifth their diameter. This paper discusses their applications as collimators for small light sources (LEDs and HID lamps), injectors for fiber- optic illumination systems, and solar concentrators, and how close their performance comes to the ideal thermodynamic limit.

We have demonstrated the feasibility of a high temperature cool-wall optical furnace that harnesses the unique power of concentrated solar heating for advanced materials processing and testing. Out small-scale test furnace achieved temperatures as high as 2400 C within a 10 mm X 0.44 mm cylindrical hot-zone. Optimum performance and efficiency resulted from an innovative two-stage optical design using a long-focal length, point-focus, conventional primary concentrator and a non-imaging secondary concentrator specifically designed for the cylindrical geometry of the target fiber. A scale-up analysis suggests that even higher temperatures can be achieved over hot zones large enough for practical commercial fiber post- processing and testing.

Application of non-imaging optical concentrators to infrared light detection is discussed. It is shown that dielectric non-imaging concentrators can enhance the sensitivity of IR detection systems with minimal reduction in the field of view. Potential laser warning applications are discussed.

Two types of non-imaging systems have been modeled, a lighthouse and a crossed compound parabolic concentrator (CPC). Design tradeoffs have been made to control the radiant intensity of each system. An asymmetric design for the lighthouse system is evaluated and optimization of the crossed CPC design at the field edge is discussed. A tolerance study showing the effect of the surface reflectivity and source misalignment is presented. A comparison of theoretical predictions and system measurements agree within 5 percent.

A new method of design of nonimaging concentrators is presented and two new types of concentrators are developed. The first one is an aspheric lens, and the second one is a lens- mirror combination. A ray tracing of 3-D concentrators (with rotational symmetry) has also been done, showing that the lens-mirror combination has a total transmission as high as that of the full compound parabolic concentrators, while their depth is much smaller than the classical parabolic mirror-nonimaging concentrator combinations (8 times smaller or more). Another important feature of this concentrator is that the optical surfaces are not in contact with the receiver, as occurs in other nonimaging concentrators.

A fundamental problem in illumination optics, with important applications for lighting and infrared heating, is to design reflectors that can concurrently fulfill two conditions: (1) maximum efficiency (all rays from the source reach the target); (2) uniform flux density on the target plane. The problem is difficult when the source is extended (rather than a point or a line) and when the reflector must be small compared to target distance and target size, i.e., when the field of view subtends a large angle. In general, an exact solution is impossible with a finite number of optical elements. To find practical solutions that approach the goal, we take the compound parabolic concentrator (CPC) of non-imaging optics as a starting point because it achieves the first condition by its very design. However, its flux density distribution falls off like cos3 ((theta) ) where (theta) is the angle from the normal of the aperture. To gain an extra degree of freedom for the design, we modify the CPC by introducing a gap between source and reflector. We present results for symmetrical configurations in two dimensions (troughlike reflectors) for flat and for tubular sources. For fields of view of practical interest (half angle in the range of 40 to 60 degree(s)), these devices can achieve minimum-to-maximum intensity ratios of around 0.7, while remaining compact and incurring low reflective losses.

Laser diode arrays are currently replacing flashlamps as optical pumping sources for high power solid state laser systems. With this advanced technology comes the challenge of transferring the energy from these sources to the laser material. Coupling the energy emitted by these arrays is important as simultaneously it determines the laser's efficiency and beam quality. Nonimaging reflectors offer several advantages over coupling schemes using imaging devices for this application. Various tradeoffs affecting the laser's output are discussed, while model-predicted and experimental deposition profiles will be presented.

Diode-pumped slab lasers require concentrators for high-average power operation. We detail the properties of diode lasers and slab lasers which set the concentration requirements and the concentrator design methodologies that are used, and describe some concentrator designs used in high-average power slab lasers at Lincoln Laboratory.

Many optical characteristics of the CPC (compound parabolic concentrator) can be analyzed with remarkable simplicity and elegance if one employs the reciprocity relations of radiative heat transfer. While the characteristics that are of interest for solar energy have been studied extensively, those for lighting and infrared heaters have remained unexplored. The design considerations are quite different from solar applications because the reflector is placed close to the source rather than close to the target. In this note we show that the flux density produced by an untruncated CPC at a point P of a distant target plane is essentially the radiation shape factor from P to the aperture of the CPC--a simple trigonometric expression obtained without detailed ray tracing. Absorption losses can be analyzed with equal simplicity in terms of the average number of reflections. In additions to their usefulness for the reflector configurations in question, the resulting closed form solutions, being exact, can serve as test of the accuracy of ray trace programs.

Angular momentum conservation and phase space are used for an exact calculation of bend loss in rectangular light pipes without rayiracing. It is found that the loss fraction ofrays in the bend plane is greater than that of the full 3-D distribution for any bend radius. There is excellent agreement between the calculated and measured loss.

Nonimaging optics can be used to reduce systematic measurement errors associated with diffuse reflectance measurement instrumentation. An analysis of measurement errors related to a sample's scattering characteristics, a detector's field-of-view and the reflectometer's geometry is presented. Recent designs for hemi-ellipsoid and integrating sphere reflectometers, incorporating nonimaging optics, are reviewed.