This action might not be possible to undo. Are you sure you want to continue?
The differential equation is derived that describes the reflector of an ideal two-dimensional radiation concross centrator with an absorber of arbitrary convex shape. For the special case of an absorber with circular section, the equation can be solved in closed form if suitable coordinates are used. The effect of absorption on at the reflector is considered, and formulas are presented for determining the attenuation of radiation its passage from aperture to absorber.
onto the smallest possible absorber area consistent with 1 2 For example, if the second law of thermodynamics. radiation has a restricted range of incidence angles, IOij < Oa, a two-dimensional ideal concentrator can con-
In an ideal concentrator, radiation is concentrated
sorber by polar coordinates (r,O)and to characterize any point B on the reflector by its distance p = BC from the shown in Fig. 3.
I1. Reflector Shape for an Arbitrary Absorber
point C at which the tangent CB touches the absorber. The angle 0 is measured from the negative y axis as
centrate it by a factor
Cideal, 2dim = 1/sinOa.
ments6 ' 7 :
The reflector slope is fixed by the followingrequire-
In this note we deal exclusively with two-dimensional sorber areas. The compound parabolic concentrators
concentrators, also called cylindrical or troughlike. The concentration is, of course, the ratio of aperture to ab4 is an example of an ideal concentrator; it is designed for a flat absorber which is illuminated on one side only. During the course of recent solar energy research, sev-
+ (r/2) any ray emitted tangentially • Oa 01 (1) For 1 from a point C = (r,O)of the absorber toward the reflector must be reflected back onto itself; hence, the corresponding portion of the reflector is a convolute (2) For Oa+ (ir/2) < IO < (3r/2) - Oa any ray emitted tangentially from a point C = (r,0) of the absorber toward the reflector must be reflected so as to make an
angle Oawith the y axis.
eral new types of ideal concentrators have been dis6 covered.5 In particular, Winston and Hinterberger
have shown that in two dimensions, radiation with IOij by a factor 1/sinOa onto an < Oa can be concentrated absorber of arbitrary (convex) cross section. However, plicitly. In fact, until now the reflector shape has been known only for absorbers consisting entirely of straight sections as in the examples in Fig. 1, where the reflector is formed by appropriate parabolic and circular sections (for details see Ref. 2). this equation in general and to solve it for the important special case of a round absorber (i.e., a tube) as shown
they did not calculate the required reflector shape ex-
These conditions are equivalent to first-order differential equations which determine the reflector uniquely. The requirements that the curve start at point A(0 = 0) and that it be continuous at point F[O= Oa+ (r/2)] provide the boundary conditions. The equatiorn of the convolute is well known: the distance p = BC is equal to the arclength AC along the absorber circumference 2 (-) 2]1/ (2) 'i dOforI0110 + L= f.30 [r2 (thus allowing the convolute to be drawn by unwinding a string from the absorber). For the Oa + (7r/2) < 0 < (37r/2) - Oa portion of the reflector, requirement (2) imposes the following relations among the angles a, , 'y, 0, Oa, and 5 in Fig. 3:
Oa+ 3+ 2a = r, a + Oa= (3) (4)
The reflector is determined by a first-order differential equation. The purpose of this note is to derive in Figs. 2 and 3. As usual, with such equations, tre-
mendous simplification can result from a suitable choice
of variables. We find it convenient to describe the ab-
The author is with Argonne National Laboratory, Argonne, Illinois 60439. Received 8 December 1975.
-) (r - 0)+ (r - d) + (7r
July 1976 / Vol. 15, No. 7 / APPLIEDOPTICS 1871
7r)/2]. but the notation is appropriate for arbitrary absorber shapes. 1(c) is illuminated on both sides and must be considered to consist of three straight sections: AB. 1. (c) concentrator for horizontal fin. p. Section AD of the reflector is a convolute (circle with center at D'E is parabolic with focus at C. r' = dr/dO. Coordinates and parameters used for describing absorber and reflector. Differentiating the equations x = r sin0 coso (8) and y = -r cos . the A CIRCLE Fig. and = 7r/2. A subtle complication arises when the absorber contains straight segments. AXISOF. the curved portion between reduces to a point. For example. Therefore. and section 111. dp/dO = r + [(dr/dO)+ pI tan[( -Oa - -)/2]. With the substitution For the special case of a circular absorber (tube).. (b) concentrator for vertical fin. (7) and (10) and solve for p' to obtain the equation for the reflector in the range a + (7r/2)< I < (37r/2) Oa. (6) E / Fig. Equation (11) is still applicable if r and r' are (14) with f(0) =_ constant] is parabolic with focus at that point. r vanishes. 15. section DD' is parabolic with focus at B.AXISOF CPC a Ad APERTURE determined entirely by the curved portions of the absorber. No. and 0. 7 / July 1976 Fig. Solving for the angle y and inserting it into dy/dx = tany we can write the reflector slope as dy/dx = tan[(0 + 0° + a . Ideal concentrator for tube. \ a a. Since all rays emitted tangentially from points on a straight absorber segment are where r = r(O).) . 2. (7) concentration = 1/(sina) = 2 (ratio of aperture width to absorber diameter = 270. = tan'1[r/(dr/d0)]. Finally we combine Eqs.PARABOLA a>| \ i both set equal to zero. and = tan-l(r/r') are of identical. Examples of ideal concentrators with flat absorbers: (a) compound parabolic concentrator (CPC). If the absorber contains adjacent straight segments. BC. the absorber in Fig. The example shown has Let us reexpress dy/dx in terms of r.p sin0 (9) with respect to 0 leads to dy (r-p') tanO-r'-p dx (r' + p) tano + r-p" (10) where r' = dr/dO and p' = dp/dO. (11) course given by the absorber shape. Special Case of a Circular Cylindrical Absorber a) \ l. and CA. The corresponding solution [Eq. 3. <PARABOLA h FOCUSOF PARA BOLA 'ABSORBER radius r is constant r = a. (The example shown happens to have a round absorber. the reflector is 1872 APPLIEDOPTICS/ Vol. B). they strike only one point of the reflector (a point that has already been determined by the curved portion of the absorber).
= Oa Therefore the complete solution for the reflector shape is p = a for I• Oa+ (/2) (17) a Heat Pipe" (1976). On its passage from surface BCDE to the absorber tube ACD itself. Heat Mass Transfer. 5. Patent by F. 207. J. Opt. 7 / APPLIEDOPTICS 1873 . presented at National Science Foundation Solar Thermal Review (March 1974).1094 (1966). Meinel. 2. A. (19) 9. (14) which will indeed solve Eq. Two stage arrangements (e. R. A. For simplicity. K. A. Ideal concentrators with restricted < 02." Argonne National Laboratory Report SOL 75-02. Sov. a = (18) where a is the tube radius.. 2. 37. Trombe.5 higher function of concentration.S.e. 33. K. "Comparison of Solar Concentrators. published Solar Energy 18. In Ref. 255 (1975). 2 and 3 is about the same as for the configthan Fig. Soc. 2." Argonne National Laboratory Report SOL 75-01 (1975). 3. see Ref. In most practical centration. 2. (18) has also been derived by Ugur Ortabasi of Corning Glass. "Evacuated Tubular Collector Utilizing where the constant must be set equal to a [Oa+ (r/2)] in order to insure continuity of the reflector curve at 0 + (/2). Hinterberger and R. a large portion of the reflector at the aperture end can be cut off with little decrease in conDue to absorption at the reflector surface only a fraction r of the radiation incident on the aperture will be transmitted to the absorber. A.5. Numerically (n) for the configuration in Figs. 15. V.. 1976). No. Melnikov. Meinel et al. H. Rabl. This suggests the ansatz p(o) = [fl)]/(cos20). 4. (13) In particular. An example with 30° (concentration = 2) is shown in Fig. July 1976 / Vol. this equation could be integrated directly to yield a cos-20 dependence for p (parabola)." Argonne National Laboratory Report SOL 75-03 (May 1975). Winston. 2. aperture width = tube circumference) corresponds to Oa = 7r/2. Energy Research and Development Administration. which in turn is about 0. I was informed that Eq. B. (15) The solution to this is straightforward: f( ) = a ( + . (12) the differential Eq. 6. Rabl. Applied Solar Energy: An Introduction (Addison-Wesley. Rev. this radiation undergoes 2 = [Oa+ (ir/2)1 /47r (n)convolute Rabl and R. 1976).I Oout to ten) it ranges from about 1 to about 1. A. 7. 8. Instrum. 408 (1966). J. to be reflections at the convolute reflector section EAB and is attenuated by a factor published in Int.. Winston. Technol. a cusp with unit concentration8 (i.concentration C = sin0 2/sinO). Hinterberger. Opt. Since in concentrators of this type the number of reflections varies both with angle and with point of incidence. Am. 2 '4 applications. Meinel and M. 1(b). (n) can be calculated either analytically of T is difficult. an exact calculation But in most cases a good approximation is provided by the formula. J. (13) provided 2 df/db = 2a cos o. r = p(fn) wherep is the reflectivity of the reflector and ( n) is the average number 9 of reflections. 93(1976).hence (n) = 7r/4. Mass. "Optical and Thermal Properties of Compound Parabolic or by ray tracing. see Ref. Ideal concentrators whose concentration varies with angle of incidence. Argonne National Laboratory Report SOL 76-01 (Jan. R. we assume symmetry about the optical axis. 4. 2 is < Oa. Solar Energy 17. (n) has been plotted as a For low concentrations (two Concentrators. it will be uniformly distributed over all angles IOind totally diffuse when it reaches the surface BCDE. and 0+ Oa+ (7r/2) . Winston and H.. 245 (1970).sin2o) + constant. Reading. and the transmission factor is pr/4 If a were equal to zero. A. References 1. exit angles (IOin• 01. for Oa+2 | <0-2 0 a. (11) becomes dp/d k = 2a + 2 p tano.asymptotically it increases logarithmically with concentration. P. uration in Fig. "Radiation Transfer Through Specular Passages. Sci. Winston. Rabl. to be published in Solar Energy.cos(O 1+ sin(0-a) Oa) This work was supported by the U. the formula for the average number of reflections is particularly simple. For the convolute portion of Fig. 60. Baranov and G.(/2)]. conventional concentrator such as a Fresnel lens in tandem with an ideal second stage concentrator).IT = p(Oa+ir/2) 2 /4ir 2 [°Oa O - . 1(a). Asymmetric concentrators can be constructed in a similar manner. If the radiation incident on the aperture FG of Fig. p. (16) After completion of this paper.g.
4 11 -- ) ' I .Optical Society >2 -~ '7.
This action might not be possible to undo. Are you sure you want to continue?
We've moved you to where you read on your other device.
Get the full title to continue reading from where you left off, or restart the preview.