Note: Descriptions are shown in the official language in which they were submitted.
~8~S9~
1 This is a divisional application of patent application
serial number 272,986 filed on March 2, 1977.
Background of the Invention
This invention relates to the useful recovery of heat.
In particular, it relates to devices and techniques e~fective to
return radiant energy emitted from a source opening back to that
source (e.g., a work~iece within an industrial furnace).
The problem of preventing heat loss (i.e., conserving
energy) in a wide variety of industrial situations has been
common for many years and has recently become increasingly im-
portant as the world's energy supplies dwindle and the cost of
generating heat increases. Although more, and be-tter, insulation
surrounding a unit to be main~ained at an elevated temperature
(e.g., an industrial furnace) may be valuable, it is not a
complete answer, since, in many circumstances, there must be ~ ;
continuous access to the heated region ~e.g., an open door to the
furnace). Such openings are common, for example, where items to
be treated in a heated environment are processed on a continuous,
rather than a batch, basis. ;~ ;
While convection currents through such an opening can
be inhibited in various ways, heretofore there has been no way
to prevent the radiation of substantial amounts of infrared
energy through the opening. Although the energy losses resulting
from such radiation have been substantial, there have remained
as an unsolved problem because there have been no suitable
devices or techniques to deal with them.
In view of the foregoing discussion, it is a principle
object of the present invention to provide devices and techniques
for recouperating heat in the form of radiant energy. It is a
further ob~ect to provide such devices which are relatively
inexpensive to manufacture and relatively easy to install and
maintain.
,:,, ~. .
-1-
;~` , "'' - :
,
.
-
~86~9~
1 Summary of the Invention
Briefly, in one aspect oE the present invention
features structures and methods for reflecting an incident beam
of radia-tion, the beam being incident from any direction within
a large solid angle. The recouperator reflects the beam, in at
least one reference plane, through an angle of substantially
1~0 to return it to the radiative source. The recouperator
comprises a plurality of reflecting cells, each comprising a
substrate which defines a plurality of planar surfaces inter- -
secting each other in angles of approximately 90 and each
coated on its surfaces facing the radiation with a material
which reflects a major fraction of incident radiation in the
wavelength band of about one micron to about 20 microns~ In
order to reduce scattering losses, each of the planar surfaces
is in the form of a geometric figure having at least ~four edges.
In another aspect, whatever the shape of the reflective
cells of the array, one or more secondary arrays of cells may
be provided and oriented with respect to the primary array such
that the radiant energy scattered from the primary array ~i.e~
not returned to the original source of heat) is intercepted by
a secondary array and returned to the primary array for subse- ;
quent reflection, by the primary array, to the original source
of heat. If the incident radiation has direction cosines of ~ -
a, b, c, in a coordinate system oriented with the intersections
of the planar surfaces of the reflective cells, preferably any
secondary array is oriented with respect to the primary array in a
plane approximately perpendicular to the direction defined by
direction cosines chosen from the follo~ing group:
-2-
~086S9~ -
1 -a, b, c,
-a, -b, c,
a, -b, c,
a, -b, -c,
a, b, -c,
-a, b, -c.
~.' ;' '.
In yet another aspect of the invention, dispersion
losses from the heat recouperator defined by two or more inter-
sectiny surfaces are controlled by orienting those surfaces to ~-
intersect at angles of t~2 - S) and by providing each cell of
the array with an aperture dimension of ~. Preferably, such a
heat recouperator is used with a source that is disposed a dis-
tance from the array of cells that is less than ~and preferably
exactly one half)the "focal distance" of the array of cells
having the facet angles and aperture dimensions as defined
To this end, in one of its aspects, the invention
provides a heat recouperator for returning an incident beam of
radiative energy to its source, said beam incident from any
direction within a predetermined solid angle, said recouperator
comprising a main array of radiant heat reflecting ceIls each
comprising a substrate defining at least three planar surfaces
intersecting each other in angles of approximately 90, each of
.. : . .
: said planar surfaces being in the form of a geometric figure ~ :
having at least four edges, front substrate surfaces facing
said solid angle being reflective of a major fraction of incident
electromagnetic radiation in the wavelength band of about 1
micron to about 20 microns, said incident beam of radiation
having direction cosines (a,b,c) in the coordinate system
oriented with the intersections of said planar surfaces, the ~:
recouperator further including at least one secondary array of
said reflecting cells, said secondary array positioned with
; 30 respect to said main array at a direction defined by a set
of direction cosines chosen from the following group: ~ :
-3-
" ' ' ' ' , ~. .. : -
.
1 -a, b, c;
-a, b, c;
a, -b, c;
a, -b, -c;
a, b, -c;
-a, b, -c.
Brief Descr-i-ption_of the Drawings
Other objects, features, and advantages of the inven-
tion will appear from the following descrip-tion of various
aspects thereof, taken together with the accompanying illustrative
drawingsO In the drawings:
Fig. 1 is a schematic illustration of one embodiment
of a heat recouperator constructed in accordance with the
present invention as used to return radiant energy to an opening
which is the source of the radiant beami
Fig. 2 is a plan view of the arrangement depicted in
Fig. l; `~
Fig. 3 is a schematic illustration, similar to Fig. 2
showing a modification;
~ Fig. 4 is a schematic illustration`of a trihedral
- - .-
(i.e., three faceted) reflecting cell;
Fig. 5 is a schematic illustration of an array of
trihedral reflecting cellssuch as that shown in Fig. ~
-
` 30 3a-
.
,
.
108~iS~
1 Fig. 6 is a schematic illustration of the effects of
scattering at a reflecting cell;
Fig, 7 is a plan view of a single trihedral reflective
cell indicating the regions o~ the cell that are primarily
responsible for scattering losses;
Fig. 8 is a view similar to Fig. 7 illustrating one
embodiment of the modified reflective cell accordiny to the ~;
present invention that avoids large portions of the scattering
losses inherent in the previous design;
Fig. 9 is a view of another embodiment of a modified
reflective cell;
Fig. 10 is an illustration of yet another embodiment
of a modified reflective cell;
Fig. 11 is a plan view of an array of reflective cells
as illustrated in Fig. 10; ;
Fig. 12 is a schematic illustration of the use of
primary and secondary arrays of cells to reduce scattering losses;
Fig. 13 schematically illustrates dispersion effects
associated with a right angle dihedral reflective cell; ~ '
Figs.14A and 14s schematically illustrate dispersion
as a function ofsource location for a dihedral cell having a
facet angle of (~r/2 - ~) radius; ~;
Fig. 15 is an illustration, similar to Fig. 14A showing
additional dispersion effects; and
Fig. 16 schematically illustrates the modification,
for the case of trihedral cells, of parameters identified in
Figs. 14A-15.
Description of Preferred Embodiments :
General
The devices and techniques to be described are
essentially reflectors of infrared radiation intended for use in
--4--
.. ~ .
. -. .
5~
1 indu~trial si-tuatlons (e.g., to return to a furnace heat radiated
from a furnace port which must either remain open/ or must be
opening frequently to permit processing of items or materials
into and/or out oE the furnace). Given this function and
environment of operation, devices to achieve the reflection oE
infrared radiation in industrial situations ta) require high
specular reflectance in the infrared, lb) should be easily
cleanable, (c) should not require exacting alignment, (d) should
not be overly fragile, ~e) should minimize to the degree possible
scattering losses, and ~f) should minimize to the degree possible
the effects of radiation dispersion.
According to the present invention it has been found ~;
possible to achieve each of these characteristics by providing
a heat recouperator formed as an array of reflecting units or
cells, each of which presents multiple reflecting surfaces for
returning an incident beam of radiation back to the source sub~
,.
stantially independently of the angle of incidence. Each re~
flecting cell comprises a reflective coating (e.g. gold) deposited
upon surfaces of a substrate to provide the reflecting surfaces.
An overlying protective layer (e.g., TiO, A1203, etc.) covers
the, typically delicate, reflective coating to permit easy
cleaning of the recouperator. The multiple reflections in the
cells eliminate critical alignment problems common in most
optical systems. As is shown in Fig. 1, when a ray 10 of
infrared radiation is reflected from surfaces 12, 14 of reflector
16 (which meet at the angle ~) the ray will be turned through
the angle 2~. The reflector 16 is one of an array formed in a
substrate 36. The reflectors 16 face a radiative opening 18 in
a furnace wall 20. (Relative dimensions have been substantially
altered from the most typical situations in order to simplify
--5--
. - ' . .; . .. ~,.....
; ~
:
~l08~S~
1 the explanations. In particular, the aperture/~ , of the
reflector 16 would usually be a few centimeters while the dis-
tance D, between the reflectur and the opening 18 would typically
be a few me-ters).
A dihedral recouperator 30 (i.e., two surfaces --12
and 14-- per cell 16) such as shown in Fig. 1 is particularly
suitable for use in recouperating infrared radiation radiated
frorn long openings (e.g. slots in furnace walls). In construct-
ing such a recouperator 30, a stacked array of reflecting cells
16 is supported such that the region of the beam in front of
the radiating opening 18 is substantially filled with the
reflecting facet~ 12, 1~ of the stacked reflecting cells. The
reflecting facets or surfaces of each cell 16 intersect in a
line of intersection 32 and adjacent reflecting cells 16 abut
in a line of abutment 34. The lines 32 and 34 should be quite
sharp (i.e., a small radius of curvature), since radiation
incident thereupon will be scattered to all angles and not
returned uniformly to the opening 18. The reflecting surfaces
12, 14 can be provided as surfaces of an integral substrate 36
(e.g. r glass). Other possible arrangements include the provision
of separate plates (e.g., glass) defining each of the facets 12,
14 and supported on a framework. Whatever arrangement is
employed to provide the surfaces 12, 1~, all such surfaces would
be coated to provide high reflectance in the infrared, (e.g., a
thin layer of gold) and then, preferably provided with a pro
tective overlying layer (e.g., TiO~) to prevent degradation of
the reflective coating (e~g., as by oxidation, ~oiling,
scratching, etc.).
As is evident from Fig. 1, radiation striking the
dihedral recouperator is turned through an angle of abouc 180
659~
1 relative to a vertical plane (i.e., the plane of Fiy. 1). Fig.
2 is a schematic illustration (again, with dimensions exaggerated)
of the same arranyement of Fig. 1, but viewed from the top.
Looking down at the recouperator 30, of course, the various
facets of the reflecting surfaces are obscured. Seen in Fig. 2,
however, is the length of the recouperator 30 and of the opening
18 (i.e., D is approximately 2.5 meters). As is evident from
Fig. 5, any ray of radiation 38 tllat is emitted from the opening
18 in a plane which is not normal to the plane of the opening
18 (indicated by the reference line~o) will not be reflected
through an angle of 180 in a horizontal plane. Fig. 3 is a view
similar to Fig. 2 showing a dihedral reflector 30 which has been ~;
modified by the inclusion of reflecting plates 42 which project
toward the opening 1~ adjacent the front of the recouperator 30.
Plat~s ~2 are supported in any conventional fashion and have
their surfaces coated with both a reflective and a protective
coating in a manner discussed above. An obliquely radiated ray
38 will strike one of the plates ~2 after its reflection from the
dihedral faces of the recouperator 30 and will be turned in a
horizontal plane and d;-rected back to part of the opening 18
where it initiated. While there may be some scattering with such ~;
an arrangement, an appropriate choice of the spacing between the
reflective plates 42 can reduce th~ scattering to a minimum for
any given application of the recouperator.
For clarity of explanation, the dimensions of the
reflective cell are greatly enlarged compared to the typical
opening 18 in an industrial furnace. For example, in the typical
industrial situation, the opening 18 may be over two meters long, -
while each reflective cell according to the present invention
would preferably have an aperture,~ , of between 1 cmD and 15cm. across.
7- !~
sg~, .
1 Referring to Fig. 4, a trihedral reflector cell 54 may
be formed by three intersection orthogonal surfaces 56, 58, S0.
Viewed from the front, such a reflector looks like an equilateral
triangle. Any incident ray 64 is returned, after multiple
reflections in the reflector cell S~, to its source as a
reflected ray 66. ~ group 62 of six such reflectors is hexagonal,
as shown in Fig. 5. Hexagons, of course, can fill a plane and
thus are a desirable shape as building blocks for a large array
of reflecting units.
Even with a large array of cells positioned to in-ter-
cept all radiation from opening 18, scattering losses, which
depend upon the angle of incidence of the incident rays, can be
expected. Such losses can be explained with reEerence to Fig.
6, which shows, for simplicity, a dihedral (two surfaced) re-
flector. At any angle of incidence ~ an incoming ray 68 which
lies in the band A~A of an incident beam will make two reflections
on cell facets and, therefore, will be reflected, as desired,
to form a parallel reflected ray 70. An incoming ray 72, however,
which lies on the band B-B will make only one reflection from
one facet 12 and will therefore be scattered as the associated
reflected ray 74~ ; ;
- In particular, I have discovered that when collimated
radiation impinges on such a cell, with the direction of the
radiation lying parallel to the normal to the frontal plane of
the reflective trihedral cell, one third of the radiation fails
to make the threQ reflections required for recouperation, and
is thus lost through scattering of the type just discussed for
the simplified dihedral case. The radiation which is scattered
is that which strikes the peripheral cell areas 76 of FigO 7
lFig. 7 refers specifically to scattering from a cell in which
''. ' ' ' ~, ' ' .
~01!ii659~
1 the three ~acets are exac-tly orthogonal to each other. Scatter-
ing from facets between which the angle is slightly less than
a right angle.
To eliminate the losses from scattering there is shown
in Fig. 8, a reflective cell having a generally hexagonal frontal
configuration and eliminating the cell surfaces previously
responsible for the ma~or portion of scattering is formed by
cutting off the corners of the triangular cell of Fig. 7 (which
corners define the areas 76) to provide the generally hexagonal
cell 78 formed by three pentagonal facets 80.
In Fig. 9, the modified reflective cell 82 is generally
triangular in shape and is formed by three facets 84, each of
which is somewhat irregular in shape (i.e., not a parallelogram).
In Fig. 10, the portions 76 that contribute to the
bulk of scattering are eliminated by cutting off the cell 86
facets to provide square facets 88. Note that with the square-
faceted cells of Fig. 10, surface discontinuities between cells
do not occur, as is illustrated in Fig. 11, where the vertex of
each cell 86 is inaicated at 90.
Furthermore, in the case of reflective cells with
triangular facets, a pair of two cells must be considered in -~
order to obtain the complete response of the reflective array of
cells to variations in the polar angle of incidence of the
radiation. With the use of square facets such as illustrated
in Figs. 10 and 11, however, a single cell will have the same
response as the entire reflecting array with respect to ~ariations
in the polar angle of incidence. This fact has implications `~
regarding the size of facets one ma~ use to construct reflecting
arrays; implications which are reflected in manufacturing costs.
For example, in constructing a reflecting array with square
.
.
. .
~156S9~ -
; facets, one could use ~acets that could be approximately twice
as large as triangular facets, that would produce the same
reflecting properties over the same area. Natùrally, cells
with larger facets are easier, and less expensive, to manufacture.
I have further discovered -that the radiation which
i5 scattered from an array of trihedral reflecting cells falls
into fairly sharply defined beams. If the incident radiation
has direction cosines (a, b, c,) in a coordinate system oriented
with the intersections of the cell facets, the scattered radia-
tion occurs with orientations defined by the followin~ sets ofdirection cosines:
-a, b, c,
-a, -b, c,
a, -b, c,
a, -b, -c,
a, b, -c, or
-a, b, -c.
Thus, placement of ~econdary reflecting arrays 92
(see Fig. 12), of the same basic design as the primary array g4,
at proper orientations to the primary array, would return the
scattered radiation to the primary array. If the facts of the
reflecting cells of the secondary arrays were e~actly orthogonal,
s~ that reflected radiation were exactly parallel to the incident
radiation, and if the frontal dimension o the reflecting cells
of the secondary arrays were small compared with the size of
those on the primary array, then the scattered radiation incident
on the secondary arrays would be returned not only to the primary
array, but to facets of the same orientation as those from which
it made its final reflection prior to travelling to the secondary
array. The scattered radiation would then be returned from the
primary array directly back to its source.
--10--
,
~8~553~
1 In this process -the scattered radiation would have to
undergo A total of either 5 or 6 reflections. The total radia-
tion returned to the source, of the part which is normally lost
through scatteriny and which undergoes the process described
above, will then be either the fourth power of the reflec-tance
or the fi~th power of the reflectance. For example, with a
reflectance of 0.90 the radiation returned to the source will be
either (o.90)6 = 0.531, or (0.90)5 = 0.590, of the fraction which
was originally scattered. For a reflectance of 0.98 the
corresponding fractions would be 0.886 or 0.904.
The practical implications of this are as follows.
Suppose the angle of incidence of radiation on a primary reflect-
ing panel were 15~. Further, suppose that the reflectance of
the individual facets were somewhere between 0.98 and 0.90, say
0.95. Then the scattering losses from the primary panel would
represent approximately 38~ to 36% of the total energy incident
upon the primary panel. By employing secondary panels of
appropriate size and orientation, and with facets of~the same ;~
reflectance, one could reduce this loss so that 0.308 to 0.326
of the radiation that would normally be lost via scattering
would be returned to the high temperature source. This would
raise the efficiency of the recouperator system from approximately
0.657 to approximately 0.965.
Another physical mechanism by which energy may fail to
be returned to ~he source is dispersion. ~his phenomemon can be
explained with reference to Fig. 13 which is a schematic illus-
tration of a single dihedral reflector 96 which has been greatly
enlarged, for clarity, relative to the distance from the plane
98 of source 100 along the axis 10~ of the reflector cell. With
the angle 104 between the facets of the cell 96 set precisely
.. . . . . . .
~6S9~
1 at a right anyle, the extrerne rays 106 ~hich will be doublP
reflected (i.e., not scattered) undergo a first reflection on
one face a~ a location 108 spaced inwardly from the extreme
edge 110 of the respective face and receive a second reflection
at the extrem~a edge 110 of the other face, thereby being turned
into a reflected ray 112. The extreme reflected ray 112 defines
a region on the plane 98 of source 100 that has twice the width
as the width of cell 96 (i.e., the distance across the open end
of the cell 96 between the extreme edges 110 of the cell facets)~
This return of a wider beam than that radiated is what is meant
by "dispersion"
The effect of providing an angle 10~ between the cell
facets that is slightly less than a right angle is illustrated
in Fig. 14a. As is known, a property of such an angle is that
an incoming ray is returned as the reflected ray making an angle
of 2~ where the angle 10~ is t~/2 -~) radians. As is evident
from Fig. 14A, the result is still a substantial dispersion
producing a reflected image on the plane 98 ~hat is substantially
twice the cell aperture~ . In this illustration, however, the
distance, D, from the source 100 to the frontal plane of the
cell 96 is D a~ /2S. As is illustrated in Fig. 14B, by locating
the source 100 at a distance of D/2 from the frontal plane of
the reflecting cell 96, the width ~f the image on the plane 98
of the source 100 is substantially~, rather than 2~ .
Analysis of the paths or both the inner and outer rays
radiated from the source 62 leads to a further discovery, as
illustrated in Fig. 15. With the source 100 placed at the full
distance D from the frontal plane of the reflecting cell 96,
it is found that the innermost rays adjacent the axis 102 of the
cell 96 define the outer boundaries of the image that has a width
- ~ -
1~65~
1 of 2~at the plane 98 of source 100. The outermost rays are
reflected to inte~sect at the source having an angle of 2 ~
between them. The result is a reflected radiation pattern behind
the source 100 (i.e., to the right of the source 100 as viewed
in Fig. 15) in the ~orm of a diverging cone having a central dark
cone therewithin and an illuminated annulus of radius ~.
The envelope of this dispersion pattern is indicated
at 114. Between the reference plane 116 located at a distance
D/2 from the frontal plane of the reflected cell 96 and that
cell itself, the dispersion envelope 114 is defined by the outer
ray emitted from the source. seyond the plane 116, however, the
outer envelope 114 is defined by the inner ray radiated from the
source which is dispersed by the cell 58 in a manner to cross
the outer ray at the location of plane 116. seyond the plane
98 located at a distance D from the frontal plane of the
reflective cell 96, the inner ray still defines the outer
envelope 114 of the dispersion pattern and the conical interior ~-
umbra having an apex angle of 21 appears.
The preceding discussion of dispersion effects has, ~ `
~ for simplicity, considered the case of a dihedral reflective cell
96. While essentially the same analysis applies, in a somewhat
more complicated geometrical form, to a trihedral reflective
cell, there is one changed relationship between the parameters
involved. Referring to Fig. 16, there is illustrated the rela-
tionship that the "focal distance" (i.e., the distance D of
Figs. 14A, 14B) for a trihedral cell is ~3/2)1/2 times as large
as that of a dihedral cell having the same values of ~ and ~ .
The value of ~ , the cell aperture, is preferably in
the range of about 1 cm. to about 15 cm. ~lthough small ~ 's
result in reduced dispersion losses, very small ~'s lead to botn
::.:;: : -:
high manufacturing costsand undesirable diffraction effects. -
,: ,
-13-
... , . , . :
,: , - ,. ~ ;
~)8~ii9~
1 AlthoucJh -the disclosure describes and illustrates a
pre~erred embodiment of the invention, it is to be understood
the invention is not restricted to this particular embodiment.
..
. .
- 14 -
'
