Professional Documents
Culture Documents
SFOREIGN
TECHNOLOGY DIVISION
THIS DOCUMENT HAS BEEN REPRODUCED FROM THE BEST COPY FURNISHED US BY THE SPONSORING AGENCY. ALTHOUGH IT IS RECOGNIZED IT
LrTICLASS IF IED
Secuwity Classilicatton,
bo"y o OVlAftf)
DATA R & D DOCUMENT kefef awII dle CONTROL ffsd n aflle n fotation msel be
onterml
when the
.o
ieel
tear#f
to
claselflecij
3*.
MgrPOnT $CUPfTY
CLASSIFIZIATION
UNCLASSIFIED
U. S. Air Force
S. NEP[IOAT TITLE
I[
OE$SC01r07v
NOT6S
(2rpe ofe
et s
Inelueive daue.)
ame)
Translation
&V T, NISI ) (Pleof name, avddle t ale . leet
Ufimtsev,
4. RaPONT
P.
Ya.
1
1b. NO.
OWN ,rs
Ga
1962 CONTiACT ON
PRO.C To
REPORT NUM9RISIr
92
F336 57-71-D-0057
L
5546
5TD-HC-23-25.9.-71
s.OTNERM RE0O0T NOES)
Wue ,pen~)
10.
OuvISuTN~l&ToN
STATEMIElNT
Wrig!1t-Patterson AFB,
Ohio
DD FN 0..147UNcSS3F
Secututv C~a~ItIs~ latln:r
FTD-HC -23-259-71l
EDITED TRANSLATION
METHOD OF EDGE WAVES IN THE PHYSICAL THEORY OF DIFFRACTION By: P. Ya. Ufintsev 223
English pages:
Source:
Metod krayevykh Voln v Fizicheskoy Teorli Difraktsli 1962, pp. 1-243 Izd-Vn Snvptskoye Radio
Translated under:
F33657-71-D-0057
UR/0000-6 2-000-000
"TMI! TRANSLATION
HAL FOREIGN TEXT WITHOUT ANY ANALYTICAL OR EDITORIAL COMMENT. STATEMENTS OR THEORIES ADVOCATED OR IMPLIEDARE THOSE OF THE SOURCE ANODO NOT NECESSARILY REFLECT THE POSITION OR OPINION OF THE FOREIGN TECHNOLOGY DIVISION.
FTD-Hc
-23-259-71
Date
797
TABLE OF CONTENTS
I :TTODUCTc
1ON
CHAPTER I.
2.
1 1
12
5 3. 4. 5 5. 6. CHAPTER II. 7. 8.
The Physical Optics Approach The Field Radiated by the Nonuniform Part
of the Current -.
18 26 32 35 3 43 48
The Physical Optics-Approach The Field from the Uniform Part of the Current-
9.
TheTotal FTle-d-Being Scattered-byv a Disk . "with Normal Irradiaton 10. The, -Pysical. Qptiv Apprbch/ " .
11. 5 12. The Field Radiated by the Nonuniform Part-6f-the Current . .
.
52
-57
CHAPTER III. DIFFPAC7ION B, YA,'F7iITB: LENGTHI.'CLINDSR. The Physical Optics Approach S 13.
14. 9 15. The Field Created by the Nonuniform Part of the Current The Total Fringing Field
73 74
80 -33
DIFFRACTION OF A PLANE WAVE INCIDENT ALONG THE SYMMETRY AXIS OF FINITE BODIES OF ROTATION The Field Created by the Nonuniform Part of the Current A Cone
90 90 95
FTD-HC-23-259-7i
"
Secondary Diffraction by a Strip (H-Polarization) Secondary Diffraction by a Strip (E-Polarization) The Scattering Characteristics of a Plane Wave by a St.ip
Secondary Diffraction by a Dlsk A Brief Review of the Literature CERTAIN PHENOMENA CONNECTED WITH THE NONUNIFORM PART OF THE SURFACE CURRENT Measurement of the Field Radiated by the Nonuniform part of the Current Reflected Wave Depolarization DIFFRACTION BY A THIN CYLINDRICAL CONDUCTOR Current Waves in an Ideally Conducting Vibrator Radiation of a Transmitting Vibrator Primary and Secondary Diffraction by a Passive Vibrator Multiple Diffraction of Edge Waves Total Fringing Field A Vibrator Which is Short in Comparison with the Wavelength (a Passive Dipole) The Results of Numerical Calculations
163 163 170 175 176 183 185 193 196 204 208 217 221
?TD-HC-23-259-71
the surface of which have discontinuities, The linear dimensions of the bodies comparison with the wavelength.
in the book.
assumed to be large in
The method
in the vicinity of the surface discontinuity and allows one to substant!ally refine the approximations of geometric and physical optics. Expressions are found for the fringing field in the distant zone. A numerical calculation is performed of the scattering characteristics, and a comparison is made with the results of rigorous theory and with experiments. The book is intended for physicists and radio engineers who are interested in diffraction phenomena, and also for students of advanced courses and aspirants who are specializing in antennas and the propagation of radio waves.
1.1
FOREWORD
First of all, one should explain the term "physical theory of dlffraction". In order to do this, let us discuss briefly the historical development of diffraction theory. If one investigates, for example, the incidence of a plane elec-
tromagnetic wave on a body which conducts well, all the dimensions of which are large in comparison with the wavelength, then the simplest solution of this problem may be obtained by means of geometric optics. It is known that in a number of cases one must add to geometric optics the laws of physical optics which are connected with the names of Huygens, Fresnel, Kirchhoff and Cotler. Physical optics uses, together with the field equations, the assumption that in the vicinity of a reflecting body geometric optics is valid. a new division of mathe-
At the start of the Twentieth Century, matical physics appeared Using it,
sphere, and infinite cylinder were obtained. Subsequently, other rigorous solutions were added; however, the total number of solutions was relatively small. For sufficiently short waves (in comparison with the dimensions of the body or other characteristic distances) these solutions, as a rule, are ineffective. Here the direct numerical methods also are unsuitable. Hence, an interest arose in approximation (asymptotic) methods which would allow one to investigate the diffraction of sufficiently short waves by various bodies, and would lead to more precise and reliable quantitative results than does geometric or physical optics. Obviously, these methods must in some way be censidered the most important results extracted from the mathematical theory of diffraction.
FD-HC-23-259-71
iv
Tn the
the
-c ut rs obtained In thtl mathematical theory of diffraction of short Here, .avo- were exactly the ones which were used and generalized. ThIs the concept of d!.ffractlon rayo advanced to the forefront. concept was exprejQ i rather a,, a physical hypothesis and was net sultable for representing the field in all of space: it was not u.;able where the formation of the diffraction field takes place (at Here it Is thie caustic, at the boundary of light and shadow, etc.). impos.;Ible to talk about rays, and one must use a wave interpretntion. clear why a large number of
What has been said above makes it works appeared in eated by other methods.
which the diffraction of short waves was InvestiAmong those applied to reflecting bodies with
abrupt surface discontinuities or with sharp edges (strip, disk, of all mention the finite cylinder or cone, etc.) one Should first These works began to appear in print in works of P. Ya. Ufimtsev. 1957, and it P. Ya. is on the basis of them that this book was written. Ufimtsev studied the scattering characteristics by such besides the currents being excited on
the surface of the body according to the laws of geometric optics (the "uniform part of the current" according to his terminology), the additional currents arising in the vicinity of the edges or borders which have the character of edge waves and rapidly attenuate with increasing distance from the edge or border (the "nonuniform part of the ceirrent"). One ma, '_Ind the radiation field created by the additional currents by comparing the edge or border with the edge of an infinite In certain cases, one is wedge or the border of a half-plane. obliged to consider the diffraction interaction of the various edges
Sthat
is, the fact that the wave cveated by one edge and propagated past another edge is diffracted by it (secondary diffraction). Such an approach to the diffraction of short waves has great physical visualizabllity and allows one to obtain rather simple approximation expressions for the field scattered by various metal bodies. This approach may be called the physical theory of diffraction.
FTD-HC-23-259-71
Thlz name is applied to many works on the diffraction of short wave. In which the mathematical difficulties are bypassed by means of physlcal considerationo. it is clear that the physical theory of diffraction is a step forward in conparison with physical optics, which In general neglects The results obtained in this book the additional (edge) currents. show that with a given wavelength the physical theory of diffraction gives a better precision than physical optics, and with a given precision the physical theory of diffraction allows one to advance Into the longer wave region and, in particular, to obtain a number of results which are of" interest for radar where the ratios of the dimensions of the bodies to the wavelength do not reach such large values as in optics. In addition, the physical theory of diffraction encompasses a number of interesting phenomena which are entirely foreign to physical optics. Thus, in a number of cases the additional currents give, not a small correction to the radiation field, but the rain contribution to this field (see especially Chapters IV and V). If a plane wave is diffracted by a thin straight wire (a passive vibrator), then the additional current falls off very slowly as one goes further from the end of the wire. Therefore, the solution Is obtained by summing the entire array of diffraction waves (secondary, tertiary, etc.) which successively arise as a consequence of the reflection of the currents from the ends of the wires. It has a resonance character. Thus, the problem of the scattering of the plane wave by a finite length wire which is a diffraction problem of a slightly unusual type is solved in Chapter VII. The resulting solution is applicable under the condition that the diameter of the wire is small in comparison with the
wavelength and length of the wire, and the ratio of the length of the
wire to the wavelength is arbitrary. The final equations which are derived 1n this book and are used for calculations are not asymptotic in the strict sense of the word.
FTD-HC-23-259)-71
vi
"
."fore, f
It
Is
natural
in
equation.; differ
mathematical
theory of diffraction?
b-fnr-hand In the
the main term of the asymptotic expansicn will not, agree with the solution obtained or the basiz of
,e-.teral case,
phi-jl
virying
con..iderations:
slowly
the decay of the fields and currents and also the diffrac-
;Z one goe, further from the edges and borders, fiAuro Inr in the main tern. influence However,
tlon Interaction of the edges and the shadowing of the edge waves will the refinement of the slowly varyis not able seen functions in the expression for the diffraction field relationships. this
to seriously from a
the quantitative
This is
obtained in
also with the results of measurements. The relationships development fraction, obtained in this book also should help the
since
L.
A.
Vaynshteyn
r;
I ,TR'O)DUCTI(AN
In recent years,
In
"J;,apt.
Such diffraction problen's with a rigorous methematical formulation reduce to an interpretation of the wave equation or Maxwell equations with consideration of the boundary conditions on the body's
surface. However, one cannot succeed !n finding solutions In the case of aotual bodies of a complicated configuration. This may be done only for bodies3 of the simplest Ceometric shape Infinitely long cylinder, a sphere, a disk, etc. It
-
such as an
the resultingr solutions permit one to effectively calculate the diffraction field only under the condition that the wavelength is lartger than, or comparable to, the finite dimensions of the body. In the "quasi-optical"case, when the wavelength is a great deal less than the dimensions of the body, practical value, and it the rigorous solutions usually lose their is necessary to add to them laborious and
complicated asymptotic studies. Here, the numerical methods for the solution of boundary value problems also become ineffective. Therefore, in the theory of diffraction the approximation methods which allow one to study the diffraction of ,ufficiently short waves by various bolies acquire great importance. The field scattered by a given body may be calculated approximately by means of geometric optics laws (the reflection equations, see,for example [1-3]), from the principles of Huygens-Fresnel an! from the equation." of Kirchhoff and Cotler [3-6].
and Cotler - the so-called physical optics approach. of' this method may be summarized as follows.
FTD-ifC-23-2')9-71
vili
ILet a plane electromagnetic wave fall on some ideally conducting body which Is, found in free space. In the physical optics approach, the ;urface cutrent density which is unilt,) equal to
P -- - In"H 0.
where c is
On the darkened side of the body the surface current is assumed to be equal to zero (J Q 0 0). Equation (A) means that on each element of the body's irradiated surface the same current is excited as on an ideally conducting surface of infinite dimensions tangent to this element. The scattered field created by the current (A) is then found by nieanv- of Maxwell's equations. It is obvious that in reality the current induced on the body's
surface will differ (as a consequence of the curve of the surface) 0 from the current J0. The precise expresslon for the surface current density has the form (B) where j is the surface density of the additional current which
results from the curve of the surface. By the curve of' the surface, we mean any of its deviations from an infinite plane (a smooth curve, a sharp bend, a bulge, a hole, etc.). If the bh Is convex and omooth and its dimensions and radii of curvature are larfre in comparlsnn with the wavwlength, mainly in then the additional current is concentrated the vicinity of the boundary between the illuminated ani
.hadowed parts of the body's surface. But if the body has an edge, bend, or point, then the additional current also arise:; near them. 0 The additional current density is comparable to the density 3 , as a rule, only at distances of the order of a wavelength from the correspodO lni7 cddge, bend, or point. Thus, if the body's dimenslens
ix
significantly
a comparatively
surface.
surface density is
called the "uniform" current j1 which is henceforth call optics approach, Therefore, it is
caused by the curve of the body's surface we will In the physical considered. gives unsatisnecessary to
only the uniform part of the current is no wonder that in a number of cases it it is
factory result..
In
this
book,
the results
of the author relating to the approxiproblems are discussed and systematized. were briefly discussed in a number of
the reader becomes accustomed to the concepts For the present, let
being used in
the surfaces of which have discontinulThe dimensions of the bodies are and their
assumed to be ideally conducting. if the edges are sufficiently far from one another, of the body's surface in to dihedral angle (a [5] 20) wedre).
Obviously,
then the current flowing on a small element the vicinity of its be the same as on a corresponding infinite In fact, in Chapter I it is shown (see also
-I
wave which rapidly decreases with the distance from the edge. Therefore, one may consider that the nonuniform part of the current By means is concentiated mainly in the vicinity of the discontinuity. of this physically obvious assumption, the field scattered by a strip (Chapter I), by a disk (Chapter II), by a finite length cylinder (Chapter III) and by certain other bGdles of rotation (Chapter IV) is calculated. For a more precise calculation, however, it is necessary to keep in mind that the actual current distribution in the vicinity of the body's edges differ's from the current distribution near the edge of the wedge. Actually, the edge wave corresponding to the nonuniform part of the current, propagated along the body's surface, reaches the adjacent edge and undergoes diffraction by it, exciting secondary edge If all the waves. The latter in turn produce new edge waves, etc. dimensions of the body are large in comparison with the wavelength, then as a rule it is sufficient to consider only the secondary diffraction. This phenomenon is siudied in Chapter V using the example of a strip and disk. In the case of a narrow cylindrical conductor of finite length, the edge waves of the current decrease very slowly with the distance from each end. Therefore, here it is impossible to limit oneself to a consideration only of secondary diffraction, and it is necessary to Chapter VII is investigate the multiple diffraction of edge waves. devoted to this problem. The uniform and nonuniform parts of the current are more than auxiliary concepts which are useful in solving diffraction problems. In Chapter VI it is shown that one is able experimentally to separate from the total fringing field that part of it which Is created by There, it is also shown that the the nonuniform part of the current. depolarization phenomenon of the reflected signal is caused only by the nonuniform part of the current.
FTD-11C-2 1-259-71
Xi
the
book. A physical representation of the nonuniform part of the current is widely used in the book, but nowhere are its explicit mathematical expressions cited. in This part of the current is generally not expressed terms of well-known functions. Obviously a direct integration of field is able to lead only Therefore, we will find
the currents when calculating the fringing to very complicated and immense equations.
the fringing field created by the nonuniform part of the current on the basis of indirect considerations without direct integration of it (see especially Chapters I - IV). The method by which the diffraction problems are solved in this
bock may be briefly summarized as follows. We will seek an approximate solution of the diffraction problem for a certain body by first having studied diffraction by its separate geometric elements.
example, for a finite cylinder such elements cylindrical surface, are: the lateral as part of an infinite plane,
For
surface
each section of the base rim as the edge of a wedge the first
ture of the rim in studied the diffraction obtain a the field which is is
by the separate elements of the body, radiated by it. that is, Then secondary, tertiary,
diffraction
stuuied -
This method appeals to physical formulating the problem but also in way differs Therefore,
considerations, its
solution process,
from the methods of the mathematical theory of diffraction. such a method may be referred to as the physical theory
of diffraction.
A whole series of other diffraction studies which appeared in the last five to ten years also are able to relate to the physical The first work which contained the idea of is evidently the paper of theory of diffraction [15] theory of diffraction. the physical Schwarzschild
FTD-HC-23-259-71
xlI
One should note that approximate solutions of diffraction problems would be impossible without the use of the results obtained in the mathematical theory of diffraction. In particular, the rigorous solution to the problem of diffraction by a wedge which is attributed to Sommerfeld [16] is widely used in this book. In Chapter I this solution is obtained by another method. The works of Fok [17, 18] served as the starting point for numerous studies on diffraction by smooth convex bodies. The rigorous solution of the problem of diffraction at the open end of a wave guide [19] revealed the mechanism for the formation of primary diffraction waves, and their shadowing by the opposite end of the wave guide. precision of the approximation theory (see Chapter V). The rigorous theory as applied to a strip and disk allows us to examine the
FTD-HC-23-259-71
xiii
CHAPTER I
DIFFRACTION BY A WEDGE
As was already said in the Introduction, the field scattered by a body may be investtigated in the form of the sum of the fields being
radiated by the uniform and nonuniform parts of the surface current. The uniform part of the current is nonuniform part generally is completely determined by the geo-
metry of the body and the magnetic field of the incident wave.
unknown. However,
The
assume that in the vicinity of the discontinuity of a convex surface It will be the same as on a corresponding wedge. Therefore, it is necessary for us to begin by studying the diffraction of a plane electromagnetic wave by a wedge. This chapter will be devoted to this problem. First we will investigate the rigorous solution of this problem ( 1 and 2). Then we will find its solution in the physical optics approach ( 3). The difference of these solutions determines the field created by the nonuniform part of the current ( 4). 1. The Rigorous Solution
The rigorous solution to the problem of diffraction of a plane wave by a wedge was first obtained by Sommerfeld by the method of branchinti wave functions [16]. Later, the diffraction of cylindrical and .-pherical waves by a wedge also was studied. A rather extensive bibliography on these problems may be found, for examplo, in the
//
[20].
rigorous
"solution to
site
he transformed it direction:
series.
integral
and is
discussed in Is
section.
detailed derivation
"hinders their
use. in free spaie (a vacuum) an ideally conto its edge 0, and . The
such a way that the z axis coincides with the wedge edge, measured from the irradiated surface. a, so that
the angle 0 is
of the source Q we will designate by r., 0" cases for the excitation of an case, it is excited by a "fila-
Let us investigate two particular electromagnetic ment of electric field. In the first
current"
r= The quantities
respectively,
and magnetic moments of the filament per unit length along the z axis,
w is
Is a
(w-Ck=
(r--
, ?--y):,)
(r -rO);Ir(?7--
Y.)
the condition
FTD-IIC-23-259-71
1~
*/
Figure 1. The excitation of a wedgeshaped rerion by a linear source. - snurce; P - the observation point;
S 5;(r-ro. v-?O)rdrdr?==l
with integration over tfie neighborhood of the point r0 , 00" Here and henceforth, we will use the absolute system of units (the Gauss system), and we will assume the dependence on time is in the form e In the first case, the "electric" vector potential Ae satisfies the equation (see, for example, [4])
AA~k'.~
~(1.03)
In
the "magnetic"
the
eq-aat ion
(1.05)
-4
natural to seek the solution of the nonhomogeneous (1.03) and (1.05) in the form
Equations
I V a.J4
(kr,)HI.(kr)sin.?osin
',?
with r~r,,
J (kr),in,,?
and '"(kr)sin v,?
J (kr) cous.p
H,"'(k,) Cos.?
(1.09)
solutions of Equations
(1.03)
and (1.05)
without (1.011)
the
and
(1.06).
(1.07)
and (1.-0S)
(0(r) enters
S when r - 0,
because It
remains
Hl(l"(kr)
is
the radiation
"I'JU'-IIC-2 *]-259t- 7 1
" ~'- dl
L
sdS,.
dS
rdrd,
(1.10)
for the contour L in the plane z = conot which is shown in Figure 1. Pere, the external normal to the, contour L is designated by the letter n. Applying Equation (1.10) to the functions Ae and Am ard performinp z z the limiting transitiono r 1 r0 and r2 r0 In It, we obtain
/[ '"-;
dI
I ,.to
.Jd?
'
I~s
(o
(1
Qrp
to Since here the integration limits are arbitrary, _. rd-O it follows from the (,oor Rr e.~u~it of he nterals that the irntegrands are equal:
I-w
,of
- a (, - Q-
(1.12)
I,,tnr
Now let us substitute Expressions (1.07) into Equality (1.11) and multiply both members of the latter by sinv,, . Then interratinC the resulting equality over 0 in the limits from 0 to a, we find
40s
In a similar way,
(1.14)
where
current
filament excites,
in
the space
E, ikA 'i:
k-P,
.5=-
k2p,
Aw
filament excites,
H, - ikA,-
with r<fo,
-- k'm,
co csp
rot H.
(1.17)
function when
kf,-..c* [21],
we have
-,.,(k,.)
et', in
H("(kro)-'".
the region r < r
0
. (1.e1 )
take the form
Then Expressions
(1.16)
and
(1.17)
~0
E, --'-i kpH") kr)X (kr) sin vpk sin v,?,
X'
8=1
e'
M,
ae X
or
.= iwk',p,HO"'(kr,) is (r. ?
-9
) --
(,.
.i, , where
o)1+ ?+ I
+r -
(1.19)
i (r
i.,e0
(,kr)
co1.0
Let us note,
furthermort!,
that in
Es,= I~k'p,"I(k,.-
21) (1.
and the field of the magnetic filament with the moment mz is determined by the relationship
H, = iksm,f"f'(kro). (1.22)
Therefore, the expressions in front of the square brackets in Equations (1.19) may be regarded as the primary field of the filament - the cylindrical wave arriving at the wedge edge. Now removing the filament of current to infinity (r 0 o),
-
plane waves E
arid
(1.23)
FTD-HC-23-259-71
t. e-
'
l-' , iv
,= 0.
(1.2
The field arisinig with the diffraction of theoe waves ty the wedge will obviouoly have the component
( 1.25 )
H.
Has [, (r,.
-- 0) +,u(r, +.
(1.26)
AiI Arco$$+..
J(kr)=-1
(1.27) (127
where the limits I - III mean that the integration contour goes from region I to region III (Figure 2). The cross-hatched sections in the plane of the complex variable 8 (8') shown in Figure 2 are regions in which Itnco%>O (Jmcos'<O) . Therefore, in the sections of the contour extendingr to infinity the integrand strives to zero, ensuring Subst-ituting Expression (1.27) into the convergence of the integral. Equation (1.20), we obtain u (r, ekfcos
t)=
[,t+
e",-e 0
After summing the infinite geometric progressions and replacint, the p) acquires the form variable 8 by $' = 8 - it, the function u (r,
,,(r. 'I) =
III,
TlH -0e
1 7e
FTD- HC-23-259-71
As a result,
-i--
(1.28)
The integration contour C is shown in Figure 3 and consists of two infinite branches. Since the integrand expression has poles at the points P.==2as--(m_--O,_ = , =2,....) , then for the values of 0 corresponding to the space outside the wedge (0<T<a') the function u (r, may be represented (with .=2,0<(p*<, ) in the following way:
ur(r, -/)--v(r, W
-,
with 2s--s<,,<22,(I.2)
where
",' O=-L
FTD-IiC-23-259-719
K
N
or
2*
+ respectively, In
Zi (30) Figures
3 and 4.
one of
(O<,,<x-%) , and
(,--r.<1<)
the plane wave "illuminates" both Let us write out in more detail the
=U(r,
+e
u(r, T+ -))
+ tihr p?0)
-
V (r,
withrwith "-a?
U(r,
?-%)-=r.?Ot)I ,(r
-- o T - ).-- (I T
and in
the case
--
<?,o<=
(Figure 6) we have
"U"(r. ?-),
- u(r, ? --? )-=--j':': (r, +with T) p
I
0<?<.|
?o.
==U(r.
V
9 +.))
p r(r, + -
,,)--=l
,,(r, -10)-av
) + e--' k-cot'('Nw
h
h
<Y." (1.32)
i'
[!
it(r, ~~~=U(r'.'T-F@)-{e-'"'''''
+- ?) -2z.
10
The direction i - 0 *0 0 corresponds to the ray reflected in a specular fashion from the first face (we will consider as the first face that face from which the angles are calculated), and the direction 7 =23-- To corresponds to the ray reflected specularly from the second face (Figure 5 and 6). The functions e-6" describe plane waves of unit amplitude: e-'k''c*describes the incident wave,
face, and
e- '*'c'+-'
-
e-'kco-u--%)
Figure 5. Diffraction of a plane wave by a wedge. The plane Figure 4. The integration contour in Equation (1.30). the wedge. incidence. is
Figure 6.
/ d.g
The plane wave Irradiates both faces. The line 2a - 0 is the boundary of the plane wave reflected from the second face
Diffraction by a wedge.
(0 =
ci).
FTD-HC-23-259-71
/
11
/I
!I
2. The integral
Asymptotic Expressions
u~r,~)=
in
-as 840`09CdO
(2.01)
in Equations (1.31) and (1.32) generally is not expressed in terms of well-known functions. However, when kr >> 1 it may be calculated approximately by the method of steepest descents [211. In integral (2.01), changing for this purpose to a new integration variable
we obtain
sin 0~ 8
do
*k.hTa
I
where n.
.(CO
(2.02)
(2.03)
It is not difficult to see that the point s = 0 is a saddle point: as one goes further from it along the imaginary axis (Re s = 0) in the plane of the complex variable s, the function e-'"* most rapidly increases, and as one goes along the real axis (Im s = 0) it decreases most rapidly. Therefore, when kr >> 1 the main contribution to the
integral (2.02) is contour in given by the integrand in the section of the the vicinity of the saddle point (s = 0).
The method of steepest descents is carried out by expanding the integrand (except for the factor e-"' ) into a Taylor series in powers of s. This series is then integrated term by term. if the integrand
FTD-HC-23-259-71
12
expansion converges only on part of the integration contour, (asymptotic). Limiting ourselves to the first term in it,
the re-
sinV
&P +
41 e- 4
!t -(r.
Co
V;-o
(2.0,4)
The remaining terms of the asymptotic series have a value on the order of I and higher.
-)r>
and describes that part of the diffraction field which has the character of cylindrical waves diverging from the wedge edge. With the incidence of the plane wave (1.23) on a wedge, the electric vector of which is parallel to the wedge edge, the cylindrical wave is determined in accordance with (1.25) B, =and (2.04) by the equation
H, = B.,.[(, [
.)V; p(.,)
where
1-I-- I
sin
-V-
so -----
y.,.
U c"o
(2.06)
9--/
When the wedge is excited by the plane wave (1.24), in which the magnetic vector is parallel to the wedge edge, the cylindrical wave has
the form
H,,g
V'2nh
(2.07)
FTD-HC-23-259-71
13
where
co
- a
(2.08)
In the vicinity of the shadow boundary ($ w + #0) and near the 1 directions of the mirror-reflected rays ( ?2a---?-) Expressions (2.04) - (2.08) are not valid, since the poles 81= V/Iel 4 sinf. = Vfe
4
sin (n S
of the integrand in
(2.02)
are close to s -
0 and,
consequently,
its
expansion in a Taylor series loses meaning. Physically, this result means that in the indicated region the diffraction wave does not reduce to plane and cylindrical waves, but has a more complicated character. An asymptotic representation of the function v (r, 4) in this region was obtained in 1938 by Pauli [22]; here we will present the derivation of the first term of the asymptotic series obtained in [22). Let us multiply and divide the integrand expression in Equation (2.02) by the quantity
(4=2cos)
(2.09)
obtain
"~~
"r.
I+cos,
a -toSo Cos
" e - ds.
J0
(2.10) FTD-HC-23-259-71 14
soI29 e
Olse
we find
ds
0jeI-'I:Oj
e'~dq(2.11)
and finally
2
0
Ia
e -'h,cia4 *
C" 4=o 1, Co
eoX e--
Cos 7 -
3
SH(2.12)
ewdq.
The next term of the asymptotic expansion for the function v (r, has a value, whose order of magnitude depends on the observation
')
direction: in the vicinity of the border of the plane waves -4- -, % its order of magnitude is , but far from it the order of magnitude is 1/kr in It form: is comparison with the term written in (2.12). in the following
con
-- o
-2
iCos-os
X
e'dq. ='=
(2.13)
FTD-IIC-23-259-71
15
Here the absolute value of the lower limit of the Fresnel integral always equals infinity, cos */2. waves Therefore, and its sign is determined by the sign of
when passing through the boundary of the plane finite discontinuity, (r, ensuring at this boundary
integral undergoes a
(2.14)
it
is
not difficult
to show that
-0)
elk,(2.15)
and consequently
c'dq:=!e, e'
S-O
ewdq---p
(with p> 1)
(2.17)
to the
determines the
in
of the direction
= 2a -
boundary of the plane wave reflected from the face 0 = a; for this purpose, it is sufficient to replace
by a -
0 and
0 by a -
0'
It (n = 2)
is
also interesting
Equation (2.13)
FTD-HC-23-259-71
16
-'-Vi
'
Q(r. ) dh =
,..e_.
x
de."
(2.19) and it may be reduced to a Fresnel integral. For this purpose, let us divide the contour Di into two parts by the point C - 0. Summing the integrals over these parts of the contour, we find that
i~r, 5(
')
e,
cos
_ ____
2 __
2
eikrcoosC Cas
I
cos
# 0
4.
s=ylie 'sin-
- and taking
we obtain
"-'4
(2.20)
Turning to Equation
we arrive at Expression (2.18) which together with relationships (1.25), (1.26) and (1.31) - give us the rigorous solution to the problem of the diffraction of plane waves by an ideally conducting half-plane.
FTD-HC-23-255-7I
17
S 3.
In the physical optics approach, the fringing field is sought as the electro-magnetic field created by the uniform part of the surface current 2a
(3.01)
Let us recall that here n designates the external normal to the body's surface, and H0 designates the magnetic vector of the incident wave. First let us investigate the case 0-?,.(a---. when the incident plane wave irradiates only one face of the wedge (Figure 5). From Equation (3.01) it follows that the density of the uniform part of the current being excited on the irradiated face by plane waves (1.23) and (1.24) has the following components, respectively 2x, and 2.,
j,=0
(3.02)
=z"=0.
(3.03)
For the purpose of calculating the field radiated by this current, we will use the following integral representation of the Hankel function (see [16], p. 866 ) I-H'," (p)==
-%-
"" 'P
--
'
-.
(
D (3 .04 )
-&+ t o ( -, 8
FTD-HC-23-259-71
18
(3.o5)
(3.03) and (3.05)
It is easy to show by means of Equations (3.02), that the vector potential A(x Y,0)A
j( d;
(3.06)
has the components A if Y,--.Eo, sin .I,, 0 and (3.08) Here, I designates (3.07)
I, =e-I1C"/H/q"(k /y'-+(.--)2ca.
(3.09)
Let us transform it
Jtu>O, d>O),
(3.04)
v = k cos t
~d--z'=P.
I,
- (k cot,,--)
( 3.1 )
FTD-HC-23-P59-71
19
0"
integral (3.11)
is
function of tyl,
and let
us change to polar coordinates according to the equations jy j = r sin w ', n ?>%. ?- w yl.=_rsinpwith ,( I
X= r Cos,?s
3 12
Furthermore,
(3.13)
1oinkCO~j+Oj
ixk
COS-jTS+ Costht
PI
t
hp
<
(3.14)
The integration contour F is shown in Figures 7a the cross-hatched areas indicate the sections in plex variable E in which |fcos(E--?)>o ; in Figure areas indicate the sections wherelJncos('A-+)>0 .
contour F into the contour G1
us change to a new integration variable with?<%, 5-(2r--?)with>ri. ) As a result, we obtain the following expressions: 0 with ?>r'-?.,
i, --
(3.15)
with y,
, co, (+2.)
(3.16)
if
0 < w and
FTD-HC-23-259-71
20
sin
II
III
II
Figure 7.
CACs Or
is*~i COfs C osV *
0 with fp<c+y.
2 e wi.th cto(9o,
++ -)
if
>
iF.
is shown in Figure 4.
(3.18)
V+(i,Y,)o7(?, p) S
E"-
wit tO<0< T
m (? y,) -- e-"k'
(<3.11)
FTD-HC-23-259-71
21
where
i/h r eosId
(?.?.) V
si
2t
k std
wl# .'-Cos G tT
(3.20)
It is not difficult to see that with =--<p.<: , when both faces of the wedge are illuminated, the field excited by the second face 0 a) will be described by the same equations if one replaces * by in them. a, and 00 by a -0 Adding the field being radiated by the uniform part of the current with the incident plane wave (1.23), we obtain the diffraction It equals field in the physical optics approach. V )
6uk '' Cos
1
_
e-.k Cos(v+v)
with O~,<
-y.
,.1
<? ,-,
with
with
:+?.<?<x-
(3.21)
if
illuminated, and
+ t
-O-c$(9
e-
cos (9 + 'e
with 01?<:--?*,
with
-- <y<"
with .
<2 -
-- ,
with
2 -
-?,<?
<,.
(3.22)
FTD-HC-23-259-71
22
If
both faces of the wedi:u (=--,<y<:) Now let us calculate the fleld ariing
H,=-O--
H 3=H=-O.
(3.23)
H,,- 1,
(3.24 )
. aith T<9.
H z=He.s,
"1with ?>'L
(3.25)
kcob
Sdw
(3.26)
along the infinite contour which passes above the pole w k cos 0" This Integral, precisely the same as integral I1, is transformed to
an integral along the contour DO. As a result, we obtain
I
where
.) ',)-e
I,
with--<Y<.
,( .... ' wlth .-+ 7.<V <2.
4(T .
(3.27)
(3.22
FTD-HC-23-259-71 23
In the case when both faces of the wedge are illuminated, the field scattered by the second face also is determined by Equations .3.27) and (3.28) in which one need only replace .0 by a and 0 by
Then adding the field radiated by the nonuniform part of the current with the incident wave (1.24), we find the diffraction field in the physical optics approach. This equals o; (',.V.) + e-f" C''t (-' e- ' o,
with Ok?<--?,.
eu (4p. %e) + e l0 (Cos.0 with '<-<+?<%
with *<
%+ -?It.
"u7(?. y;)
and
e02_'t
_.) +
with Oyt-y*,
OF --T<O)-1, with
with s--<,<,
+oTo eV'
+ , -P with
2-
(3.30)
if
FTD-rIC-23-2
5 9 -71
24
In terms of
when kr >> 1. Far from the directions 0 = w + 0 and 0 - 2a - ir - 100 the first term of the asymptotic expansion gives us the cylindrical wave diverging from the wedge edre. In the case of wedge excitation by a plane wave (1.23), these cylindrical waves are determined by the equation
-Ea__H!=E,.a.I.."
,
, = H=, 0. and with the excitation of the wedge by a plane wave (1.24) determined by the equation (3.31) they are
If,----
H,--1,=H.
The functions f0 and g
=If. -- .,.8 lA 1
have the form
sin to Cos f + cis Cos f + o
(3.32)
ye
s. e
'
(.333)
If
(O
?.
-- )
and
*
if both of its faces
Cos-
Sim .2-- YO_) coS("- V)+,-,(,-.i sin(a - f) Co osy S---1 )+Cos(i--j are illuminated.
.
'
(3.
The Index
i)
(U-.<(rO<.)
"Y" fop
(3.31)
and (j0)
FTD-11C-23-Y59-71
S 4.
In 1 and 3 we represented the rigorous and approximate expressions for the diffraction field by integrals along the same contour in the complex variable plane. By subtracting the approximate expression from the rigorous expression, we find the field created by the nonuniform part of the current. It is determined by integrals of the type
e& eC p:* dC~, ~ (4.01)
which,
S-=V2e
SW-j
are
tea,
and may be approximately descents.
q(6 ,'s)e-?aads
(4.02)
q(d,?, .. s)=q-+q&.S+q.'s'+.
(4.03)
Let us note that expansion (4.03) does not have meaning only in the particular case 9=n'-y* with %-=O; x, ) j r. with=a 9=2z- x - 0,
when the observation direction (0) coincides with the direction
(4.04)
of
propagation of the incident wave glancing along one of the wedge faces.
FTD-HC-23-259-71
26
Substituttnr serles
the field
(11.03)
Into Equatlon (4.02) and then performwe find the asymptotic expansion for
We limit omitting
radiated by the nonuniform part of the current. term of the asymptotic expansion, and higher. As a result,
o+
E,=-t1 =E4'
R,-=o11= with wedge excitation by plane wave (1.23),
I
(4.05)
and
1
H, ---E, = .,gs
',r
H,=E, =o
with wedge excitation by plane wave By calculating, (1.24). and
(4.06)
(4.06),
the vicinity
g'=g--g"
((4.07)
where in
i sin-,
g c3 s
coS,-_ .- o.CO
-COS -"
( .o8) (( (1)
Footnote appears on page 42.
FTD-HC-23-259-71
27
II
i i
and
sin +9 Cos 1. ['-Cos ff 8 -- Sint
I
(4.09)
when 0<?.<),
C
if one face of the wedge is illuminated (that is, and
go=
Cos k+os C
COSo+
sin ve
+ Cos
del
COS 0
c0s(G--1)+S(M--?*) "
(4.10)
if both faces of the wedge are illuminated (that is, when -- <0<. ). Let us recall that the functions f and g describe the cylindrical waves radiated by the total current - that is, the sum of the uniform and nonuniform parts, and the functions f0 and g0 refer to the cylindrical waves radiated only by the uniform part of the current
0). (a
01
Let us note certain properties of the functions f and g . The 1 = g function the whereas 0) is continuous, fl (a,' , function fl (al, 1, %0) undergoes a finite discontinuity when 00 = a w. The reason for this differs gated discontinuity is that the uniform part of the current (1.24) is propa-
from zero on the face along which plane wave (with 00 - a ff). In the case of radar,
the observation point coincides with the direction to the source ( - 1), both functions f1 and g1 are continuous. There is no discontinuity of the function g 1 with element does not radiate in
= 00 = a
it, because
the current
and g remain finite. In accordance with Equations they take the following values
FTD-HC-23-259-71
28
ta'ti
Cos COs
3 a
g.
i - w
%0,
and 00 <
w,
I..
-sin-
" It
Sin (a -- Co-
-- os(*- V)--os(,-Q
I si -l Cos -COS
+
ift -0
sin ),
(4.12)
ft and a
rl
s-W
. 4**I
C2 "9 C
if
(4 .1 3 )
<O < with a + i w +0 < a - w. The value - i w + %0, and < w corresponds to the angle inside the wedge, and therefore is not 2a In the direction of the mirror-reflected ray 0 of interest. 0 the functions fl and gl are determined (with a - n < 0 < by the following equation:
IN
I'==
Cos
-- - Cos -'
-_
CS+OY N;------::.
s-"
+
-.
,t _- F
2Ctg(I +(14.14)
-sinl
'~-.,+
Ce,
Cski4P.Sl
2I4.cig
(I - T.)~ ctg -.
F'TD-HC-23-259-71
29
The functions fl
and g
except
for the particular values * and *0 enumerated in Equation (O.04). The graphs of the functions fl and g (Figures 8 - 13) drawn in polar coordinates give a visual representation of the effect of the nonuniform part of the current which is concentrated near the wedge edge. they show that this effect may be usbstantial for the fringing field not only in the shadow region (a+qo<4'";,), but also in the region of light (0O<T+y.) . The continuous lines in the figures In particular, correspond to the functions fl (fl < 0). The dashed and dash-dot lines the dash lines refer to the case correspond to the functions gl gl < 0, and the dash-dot lines refer to the case g > 0.
gog ---__Fp
,',
The diagram of the Figure 8. field from the nonuniform part of the current excited by a The plane wave on a hilf-plane. function fl (or g ) corresponds to the case when the electric (or magnetic) vector of the incident wave is parallel to the wedge edge.
As is Let us turn our attention to the next important aspect. seen from 1 and 3, the nonuniform part of the current on the wedge described by a contour integral which is generally not expressed But in order to calculate the field terms of well-known functions. scattered by some convex, ideally conducting surface with discontinuiis in ties (edges), the indicated expression still Obviously, such a path is Therefore, must be integrated over able to lead only to when calculating the given surace.
henceforth,
FTD-HC-23-259-71
30
0,
-0
Figure 11.
FTD-HC-23-259-71
31
A
\
U
2 (4
Fiigre 13.
10 The same as Figure Figure 12. for the case = 00 -
will endeavour to express these integrals directly functions f1 and g which have been found.
terms of the
5.
Above,
the diffraction
us investigate
= E.eAlze09 a+ jl
P+,SS 1.o.
From the geometry of the problem, it follows that the diffraction field must have that same dependence on the z coordinate as the field of the incident wave, that is
FTD-HC-23-259-71
32
.=
(x,)elf CosT
(5.02)
rotlH=-lkE, rotE-ikH,
(5.03)
one is able to obtain the followFigure 14. Diffraction by a wedge with oblique incidence of ing expressions for the radial and a plane wave. y is the angle azimuthal components of the field: between the normal to the incident wave front and the z axis.
I I r of
Cos
Ff
H, ---
-(a,
or, ")
WY
The functions E and H in turn satisfy the wave equations
(5.04)
(5.05)
(5.06)
In 4 we found the fields (4.05) and (4.06) which satisfy the equat ions AE, +k'E,= O, AH,l+k'Hs=O
(5.07)
and which are created by the nonuniform part of the current excited on the wedge by the plane wave FTD-HC-23-259-71 33
II
E~e-
+ in the form
(5.08)
and comparing Equations (5.05) and (5.07), we easily find the field created by the nonuniforum part of the current with the irradiation of the wedge by plane wave (5.01). For this purpose, it is sufficient to replace in Equations (4.05) and (4.06) k by kl, and EOz and Hoz by
Ese"'se and Hle"' s . As a result, we obtain
1)
E-
' ('P .)
e
eICkt t
eiltco(5.10)
The angle
%0 introduced
e
here is e
c(c'.a+ Pco0
hence
tg y. coo a (5.12)
The remaining components of the field created by the nonuniform part of the current with the oblique incidence of a plane wave are found from relationships (5.04), and when kr >> 1 they equal
E,=-ctg9E, H,=-ctgTH,
tII
W'--n- 7
/ - -
(5.13)
The equiphase surfaces for these waves have the form r sin T+z co-T=const (5.14 )
FTD-HC-23-259-71
34
-x
and are conical surfaces, the generatricez of which form the angle Thus, with oblique w/2 + y with the positive direction of the z axis. irradiation of the wedge by a plane wave, the field created by the nonuniform part of the current is a set of conical waves diverging The normals to the phase surfaces of these waves from the wedge edge. form an angle y with the positive direction of the z axis and are shown in Figure 15. These waves may be represented in a more graphic for'm if one introduces Lhe components (see Figure 15): E, =E, cosT-B- sI
H=H,cos-snT. (5.15)
in
,---H, ',;T
H,.
Y--l , (5.16)
Now we are able to proceed to the application of the results which have been obtained for the solution of specific diffraction problems. infinitely The simplest of them is long strip the problem of diffraction in in the quasi-optical by an the form region
these series have a poor convergence and are not suitable for numerical calculations. of such equations Therefore, for a field the requirement arises for approximation region. The derivation will be given in equations which are useful in the quasi-optical scattered by a strin
Let us investigate diffraction by an infinitely conducting strip The orientation of the strip in space is shown in
thin,
ideally
FTD-HC-23-259-71
35
UJ
!U
Figure 15.
Figure 16. Diffraction of a plane wave by an infinitely long strip. The section of the axis y (-a < y < a) shows the transverse cross section of the strip with the plane z = 0; a is the angle of incidence.
Let the direction of propagation of this wave form an with the plane y = 0. The field of this wave is the form
E = Ee(_ cse + Yi"Z'coVl (6.01)
The uniform part of the current excited by the plane wave on the strip has the components
=0,
2xEOCOSa e
(6.02)
Substituting these values into the equation for the vector potential
FTD-HC-23-259-71
36
J1 (1) dj
_
elk
4A
(6.03)
we obtain
and taking into account relationship- (3.05) and (1.18), 2 the following expressions in the region r N> ka ; A. -0,
A,,
s - H as sin I a (s a -in
2--)J "
A, ---T ECos a
sin s: (a a
,6107 - i
2,
( 6.04)i
lH.=E 1 -=ikA,,
(6.05)
A,=A-cosy--Azsin?.
Substituting Expressions (6.04) here,
(6.06)
radiated by the uniform part of the current E,~~~~~~~I --E os=s.a fk (sin -j- si__n IF)]_ E=-H',=2E,,cos2 stiasf-sn)e yj , = H, =2WasCos
sin (kt(sins--sOny )j e sina - iny
1
(6.07)
FTD-HC-23-259-71
37
IP.,
--
II ...
- .,. ,
+ g'(2)e"(
=ss=
l..g.19
) et.
. -, ,_.
lihrl
t. Ilw
tilte r I I..' t
w; ootrld t. I o
tlls
1'' Illlw:
V Te'
pIT rtld
Crvin tedtge
Odg-e hIla
.1
I
-
(V y -:a
:I1
and TPhe
tc'lnl 1i
Ioll:-.
ee :'o'l't
en4olld t.
(y
,1111k
ly
I'n l r Id Ift It
t.he, v Il.It
-p11 I ant.
ivl'
tfi~t,
t'Il~la
"
m d"' 1.11
(6,.oo})
Itet I l: lnt t hlt' l t, Iit A:.;auIo I :gl tfith, -. t I-p II, I s .I1,11 ro x I Ilia:I vel.y
:Iu
al*I :at.otl ty t he nol nial l'o Ce1 paIll o IC t he % I Il 'fItev t l y wI t (ka , 11, ont' I thlih t,
t01illa
t.ht, vutl ,
t .t.t m.
ItI
-I
uppe'' r -! ..*-
'd ,tt'
*-ei
of
t lit' It
:i IC
and
ft'ea'r
loWe'r ,,,Io'.o
In
'-p ln .e l
-,4: ,P., .
pail
:iet'0
itallIit
rl'eoll
io nleiil
I f'oC 'll
'ronI t
t.
lew ow I It f
(t I,. y
F,
lIt IeI
r'ill o
hlie
vid' I on I
IpI(
'
(SIR *- thiV)+
flml R
.I
((2)
:-e
wli
t lit,
riellet
I 1o.;
are
d1't1t't't
I ted
Iii
I l li t' it,
lt 1 1 h1 1 '-p,1;l lt'
by I, ho
't. 1,7
fW
i() r()-l~~t)-r().6
in connection with which
COS 2 sin
1"1-12)
+.
f
-
--
c 4111
sin ,
Co, -+ 2
g (2)-
-,
(6.12)
As a result, the fringing field (the sum of the fields radiated by the uniform and nonuniform parts of the current) will equal E,H, = E,.[ (1) e'h (i
+ 1(2) e-,
= H. H.o.[g
+ g (2
i1 "+ +")
(sin ,
5)1 v.S.1n
(I)
Consequently,
the functions f and r,which determine the cylindrical wave Ini the rigorous solution (see S 2). The field is the superposItion of two such waves which diverge from the edges I (y = a) and 2 (y = -:0.
(0.'.'
-Co R0 Ia ($In:-s a
) .i.
a+9
I
E,,, 11 1co4
2
,(s,,ua4
(614
These equations are valid when r >> ka2 and j?14-. Moreover, it is assumed that ka >> 1, since only under this condition is one able to consider the nonuniform part of the current in the vicinity of the strip's edge to be approximately the same as on the corresponding half-plane. In the case of normal incidence of a plane wave (a = 0), Equations (6.14) change into expressions corresponding to the first approximation of Schwarzschild [15]. From relationships (6.03) and (6.05), it follows that the electric field is an even function, and the magnetic field an odd function, of the x coordinate measured perpendicular to the plane x = 0 (in which the current flows) E,(x)= E,(-x). H (x)=-H,(-x). (6.15)
Therefore, on the basis of Equations (6.14) and (6.15) one is able to write the expressions for the fringing field In the region x < 0 (where 2'IJ')
(Si40E, HV=:!:E"Cos(*& Sin2,4 3
E=Htcoftasfn~.:sin2
(6.16) FTD-!IC-23-259-71 40
+sin
ln
sp
Here one must select the upper sign in
*
1
< 0. < in contrast It is is to
(6.16)
> 0,
and one must select the lower sign when (6.14) and (6.16),
not diffinot
by a,
and Equation (6.16) is not changed with the replacement of a by i + -- <?<--; ) and with the replacement of a by by a - w (if and of
i-
.<w).
the indicated equations lead to a discontinuity of the tangential component Hz on the plane x
=
magnetic vector
0.
This is
conne'ted with the fact that, by considering the nonuniform part of the current in the vicinity of the strip's edge to be the same as on the corresponding half-plane, we actually assume the presence of currents on the entire plane containing the strip. In order to refine the resulting expressions, it is necessary to solve the problem of that is, diffraction of the wave travelling secondary diffraction In other words, it is from one edge of the strip to its other edge. necessary to take into account the diffraction interaction of the edges. As we see, it is also necessary to take into account strip's when the H component the secondary diffraction in the case a-:t of the fringing field must equal zero. In Chapter V, we will return to the problem of diffraction by a strip, and together with the investigation of the secondary diffraction, we will present the results of the numerical calculation based (6.07), (6.14) and (6.16). on Equations
FTD-HC- 3-259-71
4I1
FOOTNOTE Footnote (l) on page 27 The designations used here differ slightly from those used in the papers [7 - 11]. The functions f and fl there were designated by f and f, respectively.
FTD-HC-23-259-71
42
CHAPTER II
DIFFRACTION BY A DISK
The problem of diffraction by a disk has a rigorous solution [24-263; however, it is not suitable for numerical calculations in the quasi-optilcal region when the dimensions of the disk are large in comparison with the wavelength. The physical optics approach used in such cases sometimes gives erroneous results. In particular, the fringing field calculated in this approach does not satisfy the
reciprocity principle. In this Chapter a refinement of the physical optics approach is carried out. First the diffraction of a plane electromagnetic wave
by a disk with normal incidence (1 7-9) is investigated, and then 0 10-12) diffraction by a disk with oblique incidence of a plane electromagnetic wave is investigated. Normal Irradiation 1. The Physical Optics Approach
Let an ideally conductInp, infinitely thin disk of radius a (Figure 17) be irradiated by plane wave
FTD-HC-23-259-71
43
'I
\\. ,-H,=o.
1 e'/'
(7.01)
low1=
Figure 17. Diffraction by a
(7.02)
Since the diffraction fieldin the far zone (R >> ka 2 ) is interest to us, the vector potential
a
of
A(x, M. z)= }
dpJ(p, ,l)---..di,
(7.03)
="
where a is
cosfl=sinhcos(--,).
As a result, we obtain the simpler equation
.. a 2. A x.y.z) elk It eikpCos 9d,?. pdp (p, ,') eh(.ry.z)=T
(7.05)
(7.06)
H= rot A, rotH-=-ikE,
it is easy to show that in the spherical coordinate system the
(7.07)
FTD-HC-23-259-71
t 7.08 )
Asy,
(7.09)
All'=
I-..-o-
J,
.(k4sin8)
A. -a A, = 0. which result from Equations (7.02) and (7.06), let us find the field radiated by the uniform part of the current in the form
E&= f.
E, = -- H*
- ,an,. :,
The function J 1 (ka sin 0) is a first order Bessel function. using its asymptotic expression , J (kasin)Cos, (k -
a MMO
B.
... ./
a- Cos,R-
(7.13)
Xt
+eI,,e
FT,-HC-23-259-71
The resulting equations show that in the region R kal, kasinR81 the fringing field may be investigated as the sum of the spherical waves from two "luminous" points on the rim of the disk, the polar = 0 * and - r + *. It is angles of which respectively equal difficult to see that these waves satisfy the Fermi principle. Actually, *= p = a, is 0 not
of all the points on the disk's surface, the point p = a, ), and the point the closest to the observation point (RAO
-
n + 0 is
the furthest from it. describe the radiation not only from
However,
Equations (7.13)
but they determine the field radiated by region which is adjacent to the line connecting
and p - a,
= i + 4.
Let us show that the luminous region actually makes the main For this purpose, let us calcontribution to the fringing field. culate the field radiated by the currents which flow inside the sector encompassing the line 0 =
(Figure 18).
angular dimensions of the sector in such a way that its arc, which When this is done, Fresnel zone. equals 2aO0, would occupy the first the angle *0 will satisfy the equation
a (I.-- cus* sin0 4 (7.14)
-,sin
In the case being investigated by us, when the condition ka 0>1 is fulfilled, we have from Equation (7.14)
(7.15)
hence Calculation of the Figure 18. field radiated by the "luminous" region of the disk. (7.16) 0.-
--a sin
FTD-HC-23-259-71
46
flowing in
the .*ridlcated
sectcr is
d4,
* X
OP&. 0 Cos
(7.17)
A, - A,-Taking into account field the condition ka sin V.,I, one may show that the
H,
.9Ho.
(7.18)
The amplitude of the expressions which have been found Is approximatelyetimes larger than the amplitude of the first terms in Equation (7.13). Moreover, expressions (7.18) and the corresponding terms in Equation (7.13) differ slightly in their phases: the first have
the factor or is
,
the factor-*
The result
obtained of a
wave is
for example
In
0-0
more convenient
(7.19)
E-if
.-.----.
(7.20)
FTD-HIC-23-259-71
117
Consequently, in
in
the field
scattered of the
the polarizaticn
incident wave.
1 8.
Let us proceed to calculation of the field uniform part of the current with normal irradiation Since the latter edge in (p
= a),
is
corresponding to it
(80
The inner integral is calculated with ka sin#).l based on the stationary phase method (see, for example [21], p. 256), and Equation
(8.01) is transformed to the form
A (x.
g, z).
lksI
X U NI J
'e ] -i i JI ( . .) OFsn&d
8 .02)
which allows one to interpret the fringing field as the field from This line is a diameter, the polar a luminous line on the disk. angle * of the points on which equals
D,=Yandi#,s+?. (8.03)
Assuming the diameter of the disk is parison with the wavelength (ka
> 1),
sufficiently
that the nonuniform part of the current near the disk's the same as on the corresponding half-plane of 4, the field on the half-plane -oCy,a
FTD-HC-23-259-71
48
s-in"
(8.04)
and similarly the field from the current flowing on the half-plane --a'-,*400may be represented in the form 'a4') (*A., (2) . E,, 1' (2) * (h O '
E,, (2)
-" "
,-H(8.o5)
M
(8.)e
g(2)
i (&40)
a()=
A~j
-(
e (2)x
f,
(8.06)
-.
1 and the functions f and gl are determined for the right half-space (0OC -..-) by the equations
Co,-f +,An
()T.. -(1)l(2)-f(2)+ 1 1 .,
1
J
(8.08) 8.8
2 .
Sos Si., -
CisT sin
From relationships
(8.04)
(6.06),
it
follows that
FTD-HC-23-259-71
49
-.
-Las
--
,, . 1)
I
(8.09)
SA
7---
.,q elk (1 )
d in Ilia
and Figure 19. Diffraction by a disk. The half-plane L lies in the (eI
_-ikn#
=_y e
the half-plane is tangent to the cirumernc of the disk at the point y1 S a, xa = 0 (a is the radius of the disk).
The edge of
(8.1o)H.gs(1)e
e.a O -- Nkc.s* (810
In accordance with the assumption of equal currents on the disk and on the half-plane, one may consider the following equalities to be valid: i{,, ,)e-'""'fld, Ja(i)e-"b'"ndl,
~J(p,
Therefore,
Q elkeSi
0dp=
J1
1,)W
4
ikaA 6!
(8.1
"
0"'&Lr,((2' (k.
2 e_ R.
abiew
e'~asn
la
wher H;
(e
(2)
y'2Aa ~ 1g()AR
FTD-HC-23-259-71
50
..
C
)p.
(8.14)
LeH., -Cos +
H..sit?.I
(8.15)
+sin
A, =7 -L -r.4 I 1cos df
.sin
d1D,
+ldp]=o. ,,,'(p
the field from the
that is,
in
nonuniform part of the current equals zero. \ By using the Bessel functions J and J2 for the field from the nonuniform part of the current, one may write the equations
9=--H6=
+ 111
-2E
)+'
fga (0.17) IaHr.., E& = H, = -2 {fg'(2)-g'(1) JJ(kasin8)+ + lI e' (2)+ g' (ID] J (ka sin 0)) P
which with ka sin |I change to Expressions (8.12), which were nlrea.'.% found. In the directionO=0, these equations give a field which equals, in accordance with (8.16), zero, and with intermediate vwlue.:
FTD-HC-23-259-71
51
I,
V% i
ll
olp:" t
I ('. .
,', I "I w
t1, Ip to tI,
,,
"
.'
it:.k 0
wIt
Ntt'i
I: l q!I t'ii
vl'7lci1
I
H.l
, I' y'
C 1I
,I
I 1 1,
,iilll ; -.
I tit.....11
I litt
't I lie'
(r~1I1.1:0 It I t'ii'i't'it
t1 I oliltt'
I ,it
Iii
I II
t
t
i'
1,
f
1 II
llo
rll ti'w
l
.... tl wilt
a sinn:
.....
ll k
8) ] .si7
I-
*%
gi)n C
"T+ i g
1 (2
" 1(2) HJ
7
'
!'o] Ih
tilII w
'l
in-.1 1. 1
i ii tu', ' .
t7 l.l~~Jlt,,
I
t llI
1 *V": kii Nm
t1 Rll 01
it)
, l - o'l
: lv"
exproo~son:o
for theCt
e I.r. i
a z ie
at it
the ICl na
e xp reo
CosT
2
+
, k
I4
~
n theit when f W
mheoc vt~tiation.)1.
le~t C-paee(4 ha tie , hw) Cr' ruing fi'eld ts eavl ly found Iby , -;urnitill that. It..- electri k Cle Id I ;ant even t'unct lonndItmantc I'e'ie Id n odid funict.i on of tihe :;cectdi nate:
(= W,
E, (-
2).
(en:--t'quent. Iy
IIt t ho volgl on z
0 (thlat
C"
I ,
0.COS IF
RS *Is
andt
(0~.
we titt at In
which
[ -. oo
FUpivt
1I
also interesting to compare these expressions with the precise numerical results obtained by Belkina [34] by the separation of variables method in the spheroidal coordinate system. It turns out that even with ka = 5 a satisfactory agreement is mation method and the rigorous theory. observed between our approxiIn Figures 20 and 21, graphs
of the functions V"(8) and V(s)(8), are presented which allow one to calculate the fringing field on the basis of the equations
H _i , ,.
p --. .aE ,,,l
()
Cos-?.
(9.07)
The
dash-dot curve corresponds to the field from the uniform part of the current, and the dashed curve corresponds to the field calculated according to Equation (9.03) and (9.05). Oblique Irradiation 10. The Physical Optics Approach
Let us take falls on the disk at an arbitrary angle to its ixis. the spherical coordinate system in such a way that the normal to the n, would lie in the half-plane ?=r- and forrm with the z axis (Figure 22). Adherinr to the investigation procedure used in the previous sections, let us first incident wave front, an angle y (0<yT<calculate the fringing field in the physical optics approach. The uniform part of the current excited on the disk by wavw (10.01) is determined by Equation (3.01)
'z---fC,
0 .CC C
(IO.0,'
FTD-HC-23-?59-71
514
//
Figure 20.
with a normal incidence I' )pl for a disk of the wave. The various curves correspond to different approximations.
The function
The fleld radiated by it is found, as was done in 3 7, by interratrn. (with the condition R ka 2 ). In the case of E-polarization of the incident wave (E..Lyoz), this field equals E1 = 11 -/i Cos CosaO ,' -- -T sin 9, 'E,,.Cos
(10.03)
55
'TD-HC-23-259-7!
Figure 21. t The function V( "(A) for a disk with normal Incidence of a plane wave. The various curves correspond to different approximations.
il4~ cx,
(10,
are detjt--
The quantities A and u in Equations (10.03) and (10.41j) mined in the following way: - sin 0 cos p, *sin 4sin s-'inT
is--o 0.
(iy,
FT.D-HC-23-259-71
56
and==x--T
<0' ,in
let us find the field scattered by the disk in the direction toward With E-polarization of the incident wave,
. .. MaE.. Cos a
E6=':Hl-0.
and with H-polarization
(10.06)
ia H ',
ft= H6 =0. r, .
CA n-cJ,12 kasin6)**, n
(10.07)
one
able to show that when R >> ka andA,/2-Vp'pr*1 the fringing field radiated from a luminous region on the disk. In the case when the luminous region is increased and in the limit (when the entire surface of the disk starts to "shine". The Field Radiated by the 1 11. Nonuniform Part of the Current Let us calculate the field in the nonuniform part of the current,
(11.01)
VITV-0, A * - 0)
'-4)d
transformed with
(11.03)
Here
FTD-HC-23-259-71
57
9,--9, ,-Fa(11.04) ,
a are the stationary phase points and the quantity 6 is determined by the equalities
sin $=
c---
(11.05)
Figure 22. The oblique incidence of a plane wave on a disk. n is the normal to the incident wave front.
it
and
ka
-- '+
to the fringing field is by the luminous region adjacent to the line * = I 1 = *2"
the stationary phase points I' 2 physically correspond to the luminous line on the disk surface. In order to calculate the vector potential (11.03), it is necessary for us to first express the nonuniform part of the current on the half-plane in terms of its field in the far zone. For this purpose, let us introduce the auxiliary coordinate systems x1 , Yl (see Figure 23), and let us take the following designations: and x2 , Y2
al, al(a2, 82) are the angles between the normal to the incident wave front and the coordinate axes xl, Yl (x 2 Y2); -02) is the angle between the z axis and the projection 01(( 1 of the indicated normal on the plane x1 = 0; -02) is the angle between the z axis and the direction 41(01 from the coordinate origin to the point p(yl, z) which lies in the plane x1 = 0 and is the projection of the observation point P(x, y, z); ri is the distance from the origin to the point p(y,' z).
FTD-HC-23-259-71
58
,x
N/
C o)s
9f-216
.U . . .Yit"
% =;.CCO%.
sin 9,
" io 'n i
c, (.,-
,sIn'
Cos,-C_"S%i
(11.06)
f,, == e
1o
,)et
('i
-sa 9,)
omit(,?"
101.
(11.07)
If
where
.'~~ cs
(.~....)
bai3?.im,
1
I
(11.08)
,An
r (Y.., 4o_ sin
+ CsCus;
s-in ")-"=__.,_ An Y1-an
8, (T,. To ) =--
-I .---
-i
2I
It C"o,
-sn,
1,
(11.09)
1l
terms of
A (d1)
S(.)e, ,. o,+..d-5
cam
e eht cot
" +-
-,-
(12.10)
F'TD-HC-23-259-71
59
iii
,/#
IF
By means of equation
H(1(qD)-'-LS YD+, eid(.
Imq>O. D>O,
if one substitutes z = t, w =
-ip,
k = -ID,
in
it
we find that
..
~(11.12)
Taking the fact Into account that 0 >4is the nonuniform part of the current concentrated mainly in the vicinity of the half-plane edge and using the asymptotic representation of the Hankel function, we obtain , eJ (11.13)
rd"'
where
sin sin ,sin
1
) (11.14)
-- sint I cf)4,1
+(1.5)
Starting from expression (11.13), it is not difficult to show that the fringing field in the far zone is described by the followinr equations:
FTD-HC-23-259--71
60
- -. >
.....
=--ksin A, t, cos a,sin A,,Aj.i+ sin'$ *A. , "HA = sin a, cos 4,,A.. s-Lk
where
(11.16)
. 0,, . Am =
. A, - = ,(0,1
7) ( 1 1 -1
f)-5.(i
i~~
and (11.16),
connection between the nonuniform part of the current on the halfplane -- o<y&<a and its field in the far zone
a lh(#')=
(,I.?I
l"(gl 1]
g' (,p..
a completely similar way that the nonuniform part of the current excited by wave (10.01) on the half-plane One may show in
--co,a
creates,
+ iksin'ca,A,.
.20)
I,
I
FTD-HC-23-259-71.
61
//
--U
"0-,.4J
On the other hand,
5 -- m
hi
-"
(11.22)
=0 E,-S-Z*'CA,*
RJ(".%)
,g l ,,.
4)iasnij.'ig
.1
(11.23)
) ek "
S.,
.... *ti& C, ,
(.
Here
(11.24)
II
sin
In
sin
-" -sin
e,:s
(11.25)
+ Co+Y, +
+s
se
in
and (11.23),
we find
)c i
'
?o
(11.26)
Wi
si'a
In this way we established the relationship between the nonuniform part of the current on the half-plane and its field in part of the current flowing on the disk. the far zone. Now let us return to a calculation of the field from the nonuniform
FTD-HC-23-259-71
62
S.-,
I" . , .
" -:-:
Since the disk is assumed to be large in comparison with the wavelength, the nonuniform part of the current in the vicinity of its edge may be approximately considered the same as on a corresponding Consequently, the integrals in Equation (11.03) will half-plane. approximately equal the corresponding integrals from the current on the half-plane:
.(pe. e-"
' O"dp
I
(1,,
J.(p. .)e-"'
Psdp_
(p
I Pd
(11.27)
As a result,
in
,=
(11.28)
c = i[A,,sin (j, -- )- 4,c,(,--), =iA ikAE, ?)I cos #. E-ikA, =ik [A9.cos (', -- ) + A 1.sin (,-
(11
(
we find the field from the nonuniform part of the current flowing on the disk
1/ 2 -1 k
__
R1
-r
$Infl
3 CSal
-Lcos(, -- I)
cos , tg ?,J X
-
"
--
,,C.,
(I'(',. 63
W-'1'
--
(11.30)
FTD-HC- 23-259-71
63
S.../
.1L'
66a Htosi-
COS
eihR
'--"a*.~ L %inall
COS Is
sinsin'"-s
They
a.. ,=
X,/f,
HO.@ ig1 (va.
a,= ,e a-
e i
H*
2vd
Si. as
1(2stka VX +L - 'gI (?a. ln
RX .
(f,,2)eL-avr3+COS
.,
e ) -- igUS(?.,
OfI
sn,,
) I ((p. ?,) e
--0 cHp ',, OS, Xif@ if we use the identities [13 (?,-jincsh t , Toosg? sin ($-y)
Sin01COSI
+ E.,-a- o 0sin -)
t (? Ole
it '
sin (-ihr--.4
.oa VV ++
eitn
sin
xco 10
sins(,y cog 11
)e*"d COS(,,-y)
N a, COS It
Sins%
(11.33)
A = Le,*fd,?Ji (p, Y,) e'P dp is found (without direct calculation of the current) in terms of the known field of an auxiliary half-plane
FTD-HC-23-259-71
64
by a
replacement of
by the equations
'
Solution (11.32) was determined exactly in this way with kavAT'.,l, when, for auxiliary half-plane whose edge touches the rim of the
disk at the points - 6 5,
j
*
i + 6,
(11.35)
0.
Ho", sOnvl-II
""
7Ih4ir .
)
7,Mo.,3, (E 0 .,+ ca1, t,
i
5 COs
(11.36)
'?
,A,=
//
/C
(11.37)
FTD-HC-23-259-71
65
rA
. ...
I~~
-i ho
1,1.-1d k
F,
e;.'.
4 ix
F,
1'
IV...
-------------------------
'.9~~W
.7
sit
'
0.(
c.r tr
~'2ntap&
U4
*4LOA
with ?-y.i<T
.4
E* H
-
I'(2) ei2
j()
u i a wit h
ae k M Ig (
D Y
E& H
-ig' c'"' )
I(1) e
*>T
si
ith
U2
=z ?
harpne-vY'
1,
trheH
2)ad
(12.02;f.
1lel 0 ho of
.. 1 ntI . 0)
ho11,11'l
.1d 11the
-,Y<--, with
:ro
*T.iacmlreII
ki~t(,r;kn d tv ;,, xvro,. ,-U
-( Ii~~~asi
th
fuc+T;
%in
in
sin I -sin t
Oin 1'(2)--(2)+-F
- COS+
Sml-Sf
coT_.l
1(2)=--
" (12.03)
91 (2)= g (2)-I
if
Y7
and
4-y
CS
S....
sta
(2)n
o+.
2
o. -(12.04)
r
2
di
1' (2)[
9'(2)
I f
(2)-
fromy Cold
o p
te c
$In
is~~~~~~~ neiigbl
wit
nfr
It was mentioned above that when ka >> 1 in the direction O=Y (0'3<Y<,5Y).'Y= -o-the field from the nonuniform part of the current is negligibly small in Therefore, comparison with the field from the uniform
part.
current one may write, with the help of Bessel functions, with ?=4 following interpolation formulas:
iEE, H.
the
(1)
j-(2)]J(C)-
eik
E,
1
V.eA1
( 2.C, )
+ iI'
f (2)1.:~ ()+
71
/ b+i 2).;(f
IH
O= H =*
S-Hii={!
-
(I)f()--
(2)!-
(12.08)
Il (1)-+-(2)) J, ()--, 2
Ir
of W-(12.08)
0- (
)! g[ (2+ , 101--
and with
(12.00)
eI
by
E(8, ,) )(. j =
"S'-'- "i---f-I M
cor--,-I 2with
p=--i
o-It
and in
the region
+l<
Here assuming y
- 0,
Z--T),1)ere
22
t2 (9.03)
and
"
/.'In
~(9.05).
the directions #----T and 8-----(with whr th
--
scattering diagram has a principal maximum, (12.11) and (12.12) that Y)= I (Tr)=-ka C0 T
it
(12.13)
T) !r --
(-
(e).
(12.15)
FTD-HC-23-259-71
70
-.-
which corresponds to the physical optics approach [Equation (7.20)]. The functions E and Z allow one to calculate as = tag'IEl. (12.17)
which are the effective scattering surfaces with the E- and H-polarizations of the incident wave. Let us recall that, by definition, the effective scattering surface is a quantity equal to
a ==4,ER.il.
where S--
(12.18)
Re EHJ
(12.19)
which is the energy flux density averaged over one oscillation cycle (the Poynting vector) in the scattered wave, and So is a similar quantity for the incident wave. In this way, we obtained the expressions for the fringing field which approximately take into account the nonuniform part of the current. In the incident plane ?----72 , they have a form which is rather simple and convenient for calculations. It is also interesting that in this case they satisfy the reciprocity principle as distinct from expressions (10.03) and (10.04) which correspond to the uniforn part of the current. It is not difficult to prove this by verifying that Equations (12.11) are not changed with the simultaneous replacement of y by 0 and of 4 by y, and Equations (12.12) are net chanred with the replacement of 0 by - y and of y by %-0 [in the case
2
However, Equations (12.11) and (12.12) lead to a discontinuity of the magnetic field tangential component H on the plane z = 0 In which the disk lies. As in the case of diffraction by a strip, tho
FTD-HC-23-259-71
71
/-
/_r.__
-;
reason for this is that we did not consider the interaction of the edges. It is also necessary to take into account this Interaction in the case of glancing incidence of the plane wave (T1--).when the fringing field components E0 and H must be equal to zero. Let us point out once again in conclusion to this section that expressions (12.11) and (12.12) near directions 0-=T, 6= --T (with have an interpolation character, but in return they allow one to represent the fringing field in the incident plane x = 0 in a convenient (uniform) form which frequently is of greatest importance (compare 24).
/
/!
~4
FTD-HC-.23-.2
5 9 _7 1
72
. ..
"
Vr
"CHAPTER III
/
The distinctive feature of this problem is to the which is caused by the discontinuity,
that, in
addition
nonuniform part of the current on the cylinder's surface there also exists a nonuniform
part of the current arising as a consequence of the smooth curve of the surface. This part of the current has the character of waves travelling over the cylindrical surface along geodesic lines [36] that is, move along spirals on the cylinder. These waves, which as they strike the edge of the cylinder, undergo diffraction and In turn, the nonuniform part of It is clear
the current resulting from the discontinuity undergoes diffraction while being propagated over the cylindrical surface. that specific consideration of all these effects is problem. However, if all the linear dimenslons of the cylinder art, comparison with the wavelength, these effects a very complicated
sufficiently large in
may be neglected when calculating the fringing field in many caseu In particular, they may be nej-lectd which are of practical interest. when calculating the field scattered in the direction toward tfif!
FTD-11C-23-2,9-71
73
source [5,
37].
In this case it
nonuniform part of the current which is of the surface, obtained in this and we will do this
way are generalized to the case when the observation coincide with the direction of the source.
13.
on a
finite,
ideally conducting cylinder of radius a and length at the center of the cylinder, lies in and the normal n to the
1.
Let
iIt
Pigure 24. Diffraction of a plane vestigate two particular iases of wave by a finite cylinder. n is incident wave polarization: the normal to the incident wave front.
/
W()
E-polarization,
vector is
The uniform part of the current excited on the cylindrocal face by wave (13.01) has, with the E-polarizaticn, the cecrpone -t:
. ,
~FTP-HC-2 -259-71
"
"-.
-- ,
"-F.
--
L/
2aV
t==j. --
(13.02)
(13.03)
determined by
(13.05)
r::tR4-asin ,)sin ,
Ccos 0.
(13.0,q)
FD-HC-23-259-71
75
.-
A= ------
, o ,
S.,' JO(C,
q)e-'. cos
4.
(13.09)
Since the current components are described by the functions , then the problem of finding the field reduces essentially to a calculation of integrals of the type a ekz (c" cc #1' "dC- -3-! f e'P'* $114d
T-C. 1- (COS
6)
ak(Cos -CoIs) S -
(13.10)
h (Co I
2
v #I
-S
j)e'P b' d.
The integral
0S
i 1(.)e'P"
(13(1 1)
-,-g
when p > 1 is easily calculated by the stationary nhase method. The stationary phase point is determined from the condition A..-sin =O
and equals
" "(13.12)
4d;-- f(m)e-"
-2
e'w d3S
(13.13)
f ?.) e
4
='C2
PS
d=VT
FTD-HC-23-259-71
76
4'
S
-f
r
//
adwith
SA
E-polarization
.= E., tin' 1 '-- 1, A , A, 0 (13.14)
S~and
with H-polarization
As
H""
, A.=A,=O
(13.15)
where
j-(0s
1-0
-h "IC~ll-o
*
(CM1-O*
ik(cOI*T - Cos 9)
'
i 4e'"('
(13.16)
is
determined by the
= - H,= ikA,, "-l, = -H, =-ikA, sin8.f Therefore, with the E-polarization it
,=-Ho
( (13.17)
equals
E. 4=H1190 ES
nI(13.18)
ika ,
2x
29=,= o.
13.1)
The resulting equations show that the field scattered by the cylindrical surface is created mainly by a luminous band adjacent to the The radiation from thL'; cylinder's generatrix with '-=-.*---band may be represented [see Equation (13.16)] in the form of <;phorical waves diverging from its ends (points 2 and 3 in Figure 211).
Now let us write Expressions (13.18) and (13.19) in a form whcih ia most convenient for calculating the effective scatterinr area
FTD-HC-23-259-71
77
ia p , =T -E. 2
ib e1 -
-O
(8- Y
"",
Here
H. "-,-" =
AR
1(13.20)
(13.22)
--
i(13.21)
and
G-=2-/ ! 2 ak'a(slnY+Sio 8)
s 0)
~X
The index "0"
Th ne
s$in
COSiT -- Cos#t
(Cos0 -
on
onK
and
ad.
approach
current),
by a
cylindrical
determined
-%as
In
(13.23)
'
In
'?
.kt s n - 5 'i
- kal*sin 8 s
(13. ?4
,
I S -. %.
the functions
and
equal
CV
/sfinb
= -These expressions
----r-.a
sin(kic~s hI) e
*-1kin 6+i!.
4'
Cos
-"
(f13.25)
is not difficult
. 5
are valid if
ka sin 0>1.
It
FTD-HC-23-259-71
78
P
* / ."-
" ,,
:. . : ., -,. .... . . .. ..
' /
. " ... ,
:.
*,
A-
.<- ' .
_.._ , . ..
-~
,. : . _ ---.-
A ..
:,
V-.,p,
. Now let us note th-it the field scattered by a cylinder is comprised of the fields scattered by the lateral (cylindrical) surface and the base (end) of the
cylinder. disk.
In the physical optics approach, the field scattered by the end of the cylinder is equivalent to the field scattered by a But the field scattered by a disk is described by Bessel functions. Therefore, it is also advisable to express the field scattered by the cylindrical surface in terms of Bessel functions. As s result, the desired interpolation equations for the field the form scattered by the cylindrical surface may be represented in
S( s((13.26)
C = 2ka sin 8.
From this it
follows that-O q
= L,=6 -i
0 in
and with
we obtain Equations
"The field being scattered by the cylinder's end (by the disk),
in accordance with equalities (10.06) and (10.07), the physical optics approach by the equations is described in
,(-),
(13.27)
Consequently,
the field scattered by the entire surface of the cylinder will be determined in the plane S--by the equaticns:
E.
#) I(13.28)
S.....
where
E+=H,---
Hor*+(, .
<1
'/
FTD-HC-23-259-71
4.
sinb
ihJ
co b
(13.29)
2a sin 0).
These equations allow one to determine in the physical optics approach the effective scattering area of a finite cylinder. 5 14. The Field Created by the Nonuniform Part of the Current
Let us find the field from the nonuniform part of the current Figuratively speaking, the caused by the surface's discontinuity. field scattered by the cylinder is created by the "luminous" regions Mathematically this field is on its end and lateral surface. described by the sum of spherical waves from the "luminous" points 1, Obviously the field from the nonuniform 2 and 3 (see Figure 24). part of the current also will have the form of spherical waves divergIng from these same points. In the case when the length and diameter of the cylinder are sufficiently large in comparison with the wavelength, one may approximately consider that the nonuniform part of the current near The the discontinuity is the same as that on a corresponding wedge. field radiated by this part of the current in principle may be found However, such a method in the same way as in the case of the disk. We will find the desired field by a simpler is rather complicated. and more graphic method, starting from a physical analysis of the solution obtained for the disk. For this purpose, let us investigate the structure of waves (12.01) and (12.02) which are radiated by the disk. These equations include the factor
*the ia Vka (sIay +siu-.) R e
7
V -0-.R
I
(14.01)
FTD-HC-23-259-71
80
Here
Is thet
It
shows how
the field is formed with increasing distance from the disk: the diffracted wave which is cylindrical near the disk unfold: Into a spherical wave a.- the distance from it (-diny-sinO)-, Thus, in is the disk or, in only on the other words, (12.01) increases. The coefficient proportional to the width of the luminous rerion on to the width of the first Fresnel zone. the functions fI and I depend more precisely, on the character of and (12.12)
Equaticns
body's geometry -
the discontinuity. Therefore, it is entirely natural to assume that the similar waves which are being scattered by a cylinder have the same structure and differ only in the functions case to a rectangular wedge. the source, way: f1 and gl which correspond in this Consequently, in the direction. toward
in the following
-ii-,
= ua,e,-- hf,s
I)
*I--AR 1 30
(2= )=
e'? (3)
' ' -1
.,.-
+CON(l14.02)
l*
S~'
,o,
-o
251- 6
(14.03)
g' (2)
,n a.
FTD-1!C-23-259-71
81
_os
i'OS __ s5n
g' (3))
cooIN
-.
I..
2
sIms
A
(14.03)
CO2 -
where
n=
In Chapter IV,
(1..04.
that In the
the field from the nonuniform comparison with the for the field
Therefore,
e H.. isH.
R0"
1.0;
Here
(i 4
and M ,
resnectively
equal (14.07)
"Q
or
=-il (1)1-(2),
MN'1 =g'(1)-:ig'(2),
FTD-HC-23 -259-7 -
82
M'J
co-
-'
I .-""
C
COS- .
li ncos-
Cos
NI
Cos - )
RS
S-'-I
Is
_____-_--(14.08)
aI,
en,--
""I
Io-
-ca
into Equations
(14.02), and in the direction O=% they give a value equal to zero for the field. 5 15. The Total Fringing Field
Summing Expressions (13.28) and (14.05), it is not difficult to see that the total field scattered by a cylinder will equal
,
,Rf
i
ef, ( )
(15.01)
,=-H= where
H*---LO(H)
)'(0)
(15.02)
1(3) [J. e-"'*kC0, /(o) =[M J,(C) -)+ i.I,(C)J iNf. (C:)] e'*'')(0) = MJ, ()+ iN (C)|
C= 2ka sin 0
and the functions M, N and M, N are expressed only in terms of the functions f and g which correspond to the asymptotic solution for a rectangular wedge
FTD-HC-23-259-71
83
or N
A Wn
(15.03)
(15.04)
a03
~co-ICOSj--COS-
Co
The functions f(3) and g(3) in turn are determined by the equation
g()
Thus only the functions f and g are included in for the scattering characteristic
(15.05)
In the direction :=u , the functions the case of a disk, the values
as in
Z(--
(w
.kae-Ih.
(15.06)
respectively equal
+ g- M_
4
- tg
r i + n
ft
()
iJ01|
( 15 -0 7)
where
s2ka.
kZ refer to the fielei from the uniform part of the current, remaining terms refer current. 84
FTD-HC-23-259-71
-.
In accordance with (12.17), the et'flectlve scatterlng area of the cylinder to determined wvith the E-polarization of the Incident wave by the function
2t. = %aill
(15.08)
,(a'
(15.09)
Let us note that Expressions (15.02) for the scattering field may be obtained directly on the basis of an analogy with the Equations (12.06), omitting the calculation of the fields from the uniform and nonuniform parts of the current. In the same way, one may obtain the expressions
(to0+to4
+ C" A
- I !##to
(15.10)(j
S-
which are ;uitable for calculating the fringing field in the region The quantities here equal <: (,---T). 2<8 2 9-t = ka (sin
0
Xm
-s0nb.)
(15.11)
Mh~
(
'n
--'
___
Cos -W o cog --
coo W-Cos
-1T-
ft
n t-
,85
FTP-HC-?3-259)-71 8
1(3)
(3)
Expressions (15.10) Io, When #=#.,
a7
cc I--o -!-Cox -
a(S.3
that 0..
8 and
(15.02).
Equations
currents flowing only on part of the cylinder's surface: end (when z = .) and on half of the lateral surface (-r Moreover,
0).
part of the current caused by the curvature of the cylindrical surface. Therefore, they must be refined with values of f and A. which are close to .2and i. However, in the case 4 =0. - that is, in 2 the direction towards the source - these corrections may be nerlected if the parameters ka and k1 are sufficiently large. culations performed by us on the basis of Equationo that this evidently may be done already when ka The graphs of the functions and H-.-=~i4) this case in (the dashed line): urumerical cal(15.02) show and kZ = 10 r. constructed for
Figures 25 and 26 agree with the experimental curve the position of the maxima and minima basically
agree, and the number of diffraction fringes is the same. For the purpose of illustrating the effect of the ends, we constructed a iraph of the effective scattering area for those same values of ka and ki taking into account only the uniform part of the current on the A comparison of Fi7-ures 25, cylindrical surface (Figure 27). 27 shows that the effect of the ends begins to appear wihen 4 2( and 1120.
Footnote
F7TD-H{C-23-259-71
86
'
*T
'ji 4
L-ttI4
4-1
Figure 25.
Dlavrarn cf the effective .ocattering surface for a finite cylinder. The canet of
~L
E-polarization.
4-
-cheoretica2.-'x erlmnti,the
Figure 26. Diagramn of' effective scattering surface for a finite cylinder. The care of
H-polarization
FTD-1{C-23-250-71
87
0T
*4-El
I I
J-
Figure 27. The effective scatterinvr area of the lateral surface of a cylinder in the physical optics approach [see (13.26)].
FTD-HC-23-250,-71
07E.S FO 07NT 1. on pa?e 86. The experimental curves sho;:n In Figures 25 and 26 and also tho:;e 'n Flg-ures 31, 32, 65 and 71 were obtained by Ye. 11. ',Iayzels and L. S. Chugunova.
F'TD-HC-23-259-7l
89
CHAPTER 1V
D-FF.RACTY-'
OF A PLANE WAVE F
IET~~N Ph'AT
"hr ti;
='~'ae c~rcular for certain other;-c-die3of rctatlcr., wh.e We w4.11 limit cur-;elve-: tc the case wuna .:ane isczntinuitieF.
electrcmagnet-I.
Ac befcre
.".rnmrv-r
axis.
th-e
bdiles are larg-e in ccrrrparlIson with t'he wavel:n,-. curren~ts In t~he vicinity of a circular di.ca-ntlnult::
~
-I an:.
cnoe
~oinvex
surface of rotaticn may be approximatlely considered to Le t-t-e sa:Tt? , C"cnce ,Ientl as that on a ccrrespondin:7 conlca7C bc-dv.
sufficient to studv the field from. th-e no-r.unlfc-rm pnwt- If the cu-- I. whlca) Is caused by the circular disccntinuity of t',o .,,--face, u Euon- a body~ as an example. 16. Thie Field Created by the!o-nun--*,f-r:,: Part of the Current 'Iio
Let a plane elect'romaagnet.ic wave fall on n conical [cvIn, rcIive direction of the z axis (F.;7u~re 2c. rc thle rlto~
I(16.01)
H =rot A
we find the following expression.; zone : fcr the frlnginn field in the wave
(16.02)= 1.2
and 8
t -If,, = ikA,. jwith H.=ikA,
Ey
The vector potential is
(16.03)
"+ J~e'
("I(i .in4d'd
the dlst ance from the discontinuity to the observation the surface current density flowing on the irradiated and J2(a) is the current density on the shadowed Since the nonuniform part of the exponents refers to the case 4:z, and
jI(,) is
the current is concentrated mainly In the vicinity of the discontinuity, the vector potential ccrrespcnding to it may be represented in the form
A =-
i"
(16.05)
Obviously the nonuniform part of the current near the discontinuity of a conical surface may be considered to be approximately the same as on a corresponding wedge (Figure 29). In the local cyllndrical coordinate system rl, Z, the field from the nonuniform part zl'
PTD-HC-23-259-71
91
14
Wp
1! 'tVCracttonofr a ut 28. F It-re pl-ane wave by :a conl!cal body. The plane w:ave In; proparated alontr the z axio. far none (kv1
thedra I alt It Ft ure 29. I)th dI to t he d41conlt in;ot'r:'poiidin sn faco. ulty o a co l of the curr'ent a wedfe t lowinq on suvcyh In the
In4 dett-ttnined
~i t
J(:16.06;)
Wit" I-( e
S"
Hore the upper s ,its In and the lower 4I it -tnwas s;hvwn that the exponents to the ease th i
V i "
(.16.07)
ve f'er:
to the case I 1
. + In
Q1 = w.
field equ:als;
"wher'e
va i une
of t he
I nc I dent
:W t
t e"1-!'t' edre
and
are
niula'tv
/
I
(16.09)
and (16.08),
c_ (1, ,.(4)
ih2.4=
__
we find
,_ (16.10)
The components Jzl and Jol are mutually perpendicular, 0=O and ---
different orientation of the unit vector etl when 8=O and #=x Is
connected with the fact that the angle 0I is measured from the irradiated face of the wedge. In the original x, y, z coordinate system the vector J has the components J 3 =j'sin.-J-',cOs9'
I with
8=0
(16.11)
J ,
and
.cos4-4sin+
J.
J1== .
(16.12)
~ fIre'E() sin 4+
Ho, ()os.
with 0=0
(16.13)
with the current on the wedge, we find the components of vector potential (16.05)
FTD-.IC-2,3-2)59-71
93
/)
A,
,*(t)
j
(p)
(44 n eu
C%
I
=0dl6.15)
e-g'1.
4'?.With
+
Figure 30. The relative orientation of the unit vectors eland ez, in the cases 1-0Ao and
A iN
(Iin-, %'"Qin
' --
and I a.
~ Wx
iJIIIo,
+RH; d,
! with
(16.16)
Furthermore, let the plane wave be polarized In such a way that Eigrx0. Then
F, H=: 0
-;-f=flIIE, "si9
4(4.=E.~.in'. ~(4)==-.~cs#.(16.17)
Considering these relationships and substituting Expressions (16.15) and (16.16) into Equations (16.02) and (16.03), we find the fdeld
from the nonuniform part of the current which Is caused by the circular discontinuity of the conical surface
and
S
with
(16.1)
.. " Equation (16.18) Is applicable for the values Equation (16. 19) i<.< values for the and t. ind I the
and
(,'f ed
FTD---C-23- 2-50-71
disk( ,"
equal3 f
the flelft
zero on the z "-1/2 W' .
1,1,
re rl.N1ni, form
-1-
ax!,,, ,.rc,,
.,
-1/2 when-_
.
, }, )
UjInp, the rve. will calculate * I- ,for spec!f!c I ..,e:. the plane w'iv'o
!i
-it they ae
If-t. -
~*(16.20)
Itr.e:!'n.: In comparison with the wave-
A Pone
The uniform part of the current which is excited on wave (16.20). the cone's surface has the components
~~~-.= B ..
0,
%in.
'
(17.01)
and creates in
the direction
the field
S=-Ho = +E0"
tgb.
-R+ (17.02)
(i tg"-
"
tgt.)!f--eall.
F,-H,= 2 = 0.
from the
The field caused by the discontinuity of the surface at tne cone base Is a spherical wave, and Is determIned in accordance with
FTD-lHC-23-259-71
95
~/
o,
e-
1'
(17.03)
f where
-=H, =0,1
(17.04)
An asymptotic calculation of the rigorous diffraction series for a semi-infinite cone [38-40] shows that In the direction0=I one may neglect the effect of the caused by the conical point. nonuniform part of the current summing (17.02) and (17.03), Therefore,
we obtain the following expression for the fringing field: E.a= -H, =
2 ce (I -. -!
" +-h
1., e"A
(17.05 )
Let us point out the following important feature of the resulting, equation. In the problems which were investigated in the previous chapters, in the edge waves of the fringing field were expressed only f and g. But now in the equation for the in addition to the term which terms of the functions
depends on f and g [the last term in the bracket of Equation (17.05)], there is an additional term [term -/1 2 tgii"' in Equation (17.05)] which does not depend on these functions and is determined by the uniform part of the current. the functions f and g considered in [41, 44], Therefore, it is impossible to rerre.sent the resulting spherical wave from the cone's base only in terms of which characterize the total edge wave diag7ram This important fact was not as a consequence of which theIr authors (1,1(1 from the corresponding wedge edge.
FTD-HC-23-259-71
06
The effective scattering area in accordance with (12.18) is determined by the equation 2=-alfzj. where the function Z is relationship ....- ~-H=-- -E,-(-Z and equals 2
-
(17.06)
(17.07)
sin
Z= -tg'wsin kIei+-
" e
"i
(17.08)
The analogous function in the physical optics approach may be written in accordance with (17.02) in the form -- W-og.sinl e~lk-- ge'SCS0. ( 17.09 )
With the deforming of the top part of the cone into a disk Equations (17.08) and (17.09) are transformed, respectively, to the form
!=ia-,tj
t
and (17.09) that
17.10)
for larfge
Furthermore, values
It
follows
from (17.08) tg w)
2
the functions
E and Z0 may
be represented
2
ft ,
(17.'1)
cos
-Cos-, I9 97
FTD-HIC-23-259- 71
P"
.. ..
2 Z= tgw eAI'.
(17.12) > tg
2
W,
2
a
2a
(17.13)
and
@ --With this
S 2 a I
,"
g'-.(17-14
=,'(17.15)
for sufficiently short waves (or for sufficiently large dimensions of the cone) the function a is proportional to a0. The coefficient of proportionality here does not depend on the cone dimensions, but is determined only by its shape.
that is,
graphically illustrated by the curves giving the effective scattering area of a cone (w - 10025', k = w, a = 900) as a function of its length (Figure 31). Whereas our equation (the continuous line) is in satisfactory agreement with the results of measurements (the small crosses)(I) the physical optics approach (the dashed line) gives values which are smaller than the experimental values by 13-15 dB. For sharply pointed cones, the nonuniform part of the current has an especially large values. In Figure 32, a curve is constructed for the effective surface of a cone (ka
=
This result is
2.75 7, Q
90g)
Footnote
FTD-HC-23-259-71
98
ii'
"with itj
(w
-
deformation
Into a_ dlhk
900).
"optics approach
almost
here reaches
30 dB when = 20.
4j
(17.08)
S.-Expression obtained
in con-
shadowed part of the body and shows that the reflected oignal
Figure 31. The effective scattering area of a finite cone as a function of its length. The function o (the continuous line) will be larger, the closer this shape is to a funnel-shaped form
was calculated on the basis of equation (17.06) which considers the nonuniform part of
in the case w = (k
=
T)
n ")"
-
Thus,
1O0,
for example, k1 - 10
the current in the vicinity of the circular discontinuity. The function a0 (the dashed line) correspcnds to the physical optics approach.
the cone may exceed by 15 dF the value corresponding to physical optics (see Figure 33) if 0 1700.
Let us note that our Expression (17.13) is equivalent to the expression presented in the above-mentioned papers [41, 441. Hcvqever, the latter expression is applicable only for sharply pointed cone-, whereas we have, in addition to (17.13), Equatlon (17.08) which to
suitable for cones with any aperture angle . the c:n-e
The calculation method discussed may be generalized in of asymmetric irradiation, irradiation of the cone. it Is However, necessary
generally speaking,
the nonuniform part of the current caused by the point of the cone.
FTD-HC-23-25C)-71
9p
1'--
:.
,-
ITI
*
asafnto fte
/
vetxage
.
I
ts
17 i
-I lt --
0'
of
Figure 32. The effective scattering area of a finite cone as a funcftion of the vertex angle.
I in I." #' i _
z axis of the plane figure shown Figure 3)1 Integ.rating the uniform part of the current, it is
S
i,_ ifield : '
i+--= 1
is
e\
= -- t, :
Eo.L
t 2" Figure 33. The effective scatter++. ,-tg,, :g j e-' ing area of a finite cone as a function of the shape of the
+ e, IN
ebR "
shaded part.
(17.16) 100
FTD-HC-23-259-71
La
L,--J--...
Figure 34. The generatrix of the surface of rotation.
tllh
IL
must be replaced by 11 and a,, we find the field from the uniform part of the current flowing on the entire illuminated side of the body
'E
-
__i_,z -
The field radiated by the nonuniform part of the current is determined In accordance with 9 16 by the equation 2 x --r -- +tg tg-)e
where
n --(17.19)
and (17.18),
the direction
2 -g
,,ke"
(17.'0)
w 2 sin
.,
+1
e),i R
(1 7 .2 0 )
Consequently,
the effective
eaual
"r
t-'
sin..kte
n--
k-
I b
ha,- tgl
sin/dte'' j-COS
2
2*
- Cos ; a +,
+
sB
"
(17.21)
"8tga
+
e2ihI") tgu,-,] e'*-h Is When the top part of the cone is deformed into a disk(.'.T' Equations (17.21) and (17.22) take the fbrm
g'w,sin kil~eih+ (-
(17.22)
I.,-,0
,
S42a
-- ika, --
11,
Tctg
na
I-, tg ka,
, sin kie'"-+
2 sin."".
M,
.,
(17.23)
-ika,
kI~e4
(I?,2/4)sin
0,
FTD-RTC-23-2
59
-71
102
2
Sa2
-ka,
its Ctg *+
a Cos--
(72
(17.25)
.% a; ka,).
(17.26)
in
n,= 3 , n,.
Equation (17.25) was derived in 5 15, is
(17.27)
2pz
which is
1,
. E, s-n.;i
C=,
(18.01)
it
is
not difficult
to show that in
the
E,.
(,,.
--,
)2ha,
(18.02)
Here a its
is
the radius of the base of the paraboloid; 1=-*2-=*ctgao is the angle between the z axis and the tangent
=
length; a is
to the the
2pz ).
At the point z = 1,
FTD-HC-23-259-71
103
The- fleid
C'I
'
t~.tet*nn
u
.r:rp
'LPn ce ~.dtte'rmtflfo
FLue36,
oi a plane wav
1o
f el fr (-m tIhe or~rico~~.ru tC-f tile Current v'~h> CII-eed bY the Sr'o-,n cuv Of t!'e r'araVie.Pssurface (eu,,yj zero in thie case ' :;frc yritz~~ric r-d lat ion [)451J. 'i'er'e fore,
Para~l~c-Iof rtatcn.
WCfind
the frinfini--
An
0
Conseauently, t~he effective sc-itterinr area o 'dill be determined by the relation.,,&ip
f2 x
CSCos-
(18.03)
the pa!:robool
which,
Cons!)
0,-K 00
7.as
tgi
I~
k Is,
which plhysical optics ives for the effective L3c'tztterIns area, .. ,, see that they differ significantly from ono an o r'. F'irst of the oscillating character of the function aT draws our 'attenticn. FTD -!C -2 -2:rQ-0
j 1siaI
,'2, 3 ...
V,-ua
i
va:,.
fitted
aind
(:=
a maxr.im:
.~ i_ .
" ,
-. 1, 2,.3...)
1. ccrntainced in this
rerf-r.'%r.d t,,; u;i on t!.e baszl Paa-.eter. = 900, tg, althcu~h the c..c'llatinw preserved,
of .'r.uatirn
= 0.1
(1j
.0o4) shows
wIh, the
(k
= r)
t.iat, area is
the am'plitude of the oscillations is of the function a exceed the corres13 dB. and A optics approach by almost the results
e between
of our theory
physical disk
deformed into a
(Figure
ka = 3n
n, Q -
900,
900).
---
i,:#mu&r
A J!I
11!..
Figure 37. The effective scatterinr area of a finite paraboloid as a function of its length with a constant value of the angle w (tLgw 0.1). The diameter of the base varies.
FTD-4C-23-25,-71
'5
i~
.I
Ji
1-
Fiir
-'
Th
-fetv
-,,ateli
ra-
i a, "t
, II !! F "If, .37
,
lI
As~ in th ~ aeofacn,
the-ap ft he-hdoc ar
F imre 3&Q 'Phe eff'ect iwe scatterini- area f a finite parahoioci a.;, a function of .' to lenoth with a cons>tant, rdiu 15 f t., La:t.o the case of a cone , the sh}are oC" t~le shiadewcd aprt t spas out to have a suhs:taritlal influonce on the re ti]eeted s o'na]l Fr example, for a paraholoid with tihe pranot ers h~a = .J, k -- ]9 1 , t ,- 0. 1 ( k = n ). th~e reflected si onnl !'.ra vs by ~ih 12 wi fh :ii As in
I.
o - '.;-rs 3-
'
12
(r
,lis ,j 3 )
-fl':-23-Yl-7i
In concluding this section, let us dwell for a moment on the question of calculating, the
PC.&.&A.Co
effective scattering area for bodies of rotation of a complex shape, whose elements are the lateral surfaces of truncated paraboloids. The field from the
nonuniform part of the current arising in the vicinity of circular discontinuities may be determined without difficulty from Equation (17.03). The field from the uniform part of the current is found by quadratures. Thus, the field being created in the direction &== by the uniform determined
The effective scatterFigure 39. ing area of a finite paraboloid as a function of the shape of the shaded part.
part of the current which flows on the lateral surface of the truncated paraboloid r 2 = 2pz(p - a 1 tgw 1 = a 2 tgw 2 ; see Figure 40) is by the equation
E,=-
M,
J:t C")
e:N--.
(18.07)
the height of the truncated paraboloid (the distance between its Let us note that Equation (18.07) it is is a simple alrebrnte and the difference of the f0clds
bases).
SIF-tDCII
3-2 ,-71
107
1 19.
A Spherical Surface
Incident wave (16.20) excites a surface current on the surface of an .ideally conducting sphere point z The generatrix of Figure 40. the lateral surface of a truncated paraboloid of rotation.
=
(a radiu:, of
p).
f=0.
i-; E.Cos
Oe (19.01)
E.=:,,, sin 0 cos eh. The currents flowing on a spherical ring cut from the sphere's surface by the planes z = 1 and z the direction 0-=r , the field E.=--H,E..[--( +(,! tg
E, =H = 0,
create,
in
i) e2`.1,(,0 + 10
(19.02)
f
(19
.0 4 )
Here aI is
is
the rad!us
of the second cross section; w (w.) is the angle between the z nx's 1 1 + . and the tangent to the meridian at the point z = tI(z2 = Furthermore, assuming in Equation (19.02) ,, =. (v,=,const), we
obtain in the physical optics approach an expression for the fic]d scattered by the spherical segment (Firure 42)
FTD-HC-23-259-71
108
Figure 42. A ring cut from the Figure 41. surface of its sphere by the planes z - 1 and z - 1 + I 2*4 1 1
A spherical segment
f,-yz-1z
--
+ _e".
+ (19.05)
=;a,
(JU
el,= -- sin.).
are simplified if ka 1
(19.o6)
>, 1 and
and (19.05)
E'=-fV=ES' go,-!!t%,e*")e,
and the field from the spherical segment will equal "(secc-tgwe If here one assumes w s 0, then equation aE.. ,e"*
-")-.
(19.07)
(19.08)
( 19.0 (q))
The value of the will equal, in
gives us the field scattered by a hemisphere. effective scattering area corresponding to it accordance with (17.06), r7-H-2-29-.
(19.10)
,FTD-HC-23-259-71
109
/.
Now let us
considering the discontinuity of the surface; perturbation of the surface if which is the current. as a ka >> 1 [74].
Consequently,
the effective
segment
will equal I 2.
(19.12) -*
Cosa -0-Cos
In
a similar quantity is
determined
by field
-os-
(19.13)
With the deforming of the spherical surface into a disk - - f-.onst) , Equations (19.12) and (19.13) are transformed, respectively, to the form
az'ika++ I rt
0s =
(ka)$.
(19.14)
It
its
and (19.13)
-.
an oscillating function of
Numerical calculation:s (Figure showed
43) that,
In Figure a
(Q = 150),
from the nonuniform part of the current. for the effective -catterint
area of
FTD-HC-23-259-71
110
--
--
Figure 43.
Figure 44.
A comparisonof the
ing area of a spherical segment as a function of its length with a constant radius of the base. The function a (the continuous line) is calculated on the basis of Equation (19.12) which considers the nonuniform part of the current near the discontinuity fun Th tio aO thecone dashed line) is calculated from Equation (19.13), and corresponds to the physical optics approach,.hae
effective scattering area of a spherical segment (continuous line) and a finite cone (dashed line) which have the same bases.
I
spherical segment and a finite (the dashed curve) "which have the same diameter and base sae
signal depends substantially on the shape of the shaded part of the body, and increases with an increase of the concavity. However,nio since the nonuniform part of the current is concentrated mainly near the discontinuity, that part of the shaded surfaca which is several wavelengths away from the discontinuity evidently will not have a noticeable effect on the reflected sihnal apd may be an arbitraey shape.
FTD-IIC-23-259-71
Iill
It ily
is
lengths)
of the bodies,
physical optics,
contrary to the widely held opinion concerning leads to a significant discrepancy with
experiments. The method used in tAs Chapter allows one to calculate the
effective scattering area associated with the symmetric irradiation of any convex body of rotation, discontinuities. irradiation. It However, the surface of which has circular is necessary to take into may also be generalized to the case of asymmetric when doing this It caused by the point and
account the nonuniform part of the current the smooth curve of the surface.
FTD-HC-23-259-71
112
"-
-"
FOOTNOTES
1. on page 98. See footnote on page
86.
FTD-HC-23-259-71
113
CHAPTER V
SECONDARY DIFFRACTION
In the previous chapters, an approximate solution of diffraction problems was carried out which was based on the representation of the fringing field in the form of the sum of the fields from the uniform and nonuniform parts of the surface current. The first field was found by quadratures, and the second field by approximation; it was assumed that the nonuniform part of the current near the discontinuity (edge) of a surface is the same as on a corresponding wedge. However, the fields found by such a method are actually the fields from the currents flowing, not only on the flat and curved parts of the body's surface, but also to some extent on the geometric extension of these sections. The error in the expressions for the fringing field which is thus introduced is most significant with a glancing incident wave, when the edge zone occupied by the nonuniform part of the current is noticeably broadened, and also with a glancing radiation, when the direction to the observation point forms a small angle with the given section of the surface. In these cases, the results obtained earlier are in need of substantial corrections. We already talked about this briefly in 6 and 12.
FTD-HC-23-259-71
114
For the purpose of refining the solutions which were found previously, it Is necessary to assume that in actuality the currents flow only on the body's surface, and that a wave travelling from one edge to the other will undergo a perturbation at the latter. The process of forming the fringing field when this occurs may be investigated in the following way. The edge wave propagated from one of the edges is diffracted by the other edges; the waves arising with this in etc. In this chapter, we will turn are diffracted by adjacent edges,
investigate the case when the dimensions of the surface faces are so large in comparison with the wavelength that it is sufficient to limit oneself to considering the diffraction of only the primary edge waves. This phenomenon we shall call secondary diffraction. In this chapter, strip (1 20 - 23) secondary diffraction by an infinitely long
and by a circular disk ( 24) is studied. The solution of these problems may be obtained by means of the principle of duality from the solution of the diffraction problems for an infinite slit and a circular hole in a flat, ideally conducting screen. In the latter case, the physical treatment of diffraction of edge waves is significantly simpler; it is exactly for this reason, therefore, that almost all diffraction studies of edge waves are related to holes in a plane screen. However, we will not take such a path, but we shall investigate a strip and a disk directly. advantage that it is dimensional bodies. 20. Secondary Diffraction by a Strip. Formulation of the Problem. ideally conducting strip of width 2a and unlimited length be orientated in space as shown in Figure 45. A plane electromagnetic wave incident normal to the strip's edges is directed at an angle a to the plane xoz and has the following form:
E(A
HS=H.eIh...0 COS+.,
(20.01)
FTD-HC-23-259-71
115
In 6 approximation expressions were found for the fringing field in the far zone which did not consider the interaction of In the case of Ethe edges. polarization of the incident wave (E.I! Oz) , these expressions may be represented in the form == E., 1 (1 )eihC "f'l'" U+
E, ff,
x = 0, y a -a are the coordinates of the strip's edge; n is the normal to the incident plane wave front.
)e~bI~eli
H,02 ,O3
0."
1(20.03)
Let us recall that the functions f and g included here are determined (when Jai<-) by the following relationships: in the region pIp) .?1(2)= -- C_ll s -i (20.04)
g(l)=-f(2), g(2)=-10|)-
(20.05)
The first terms in Equations (20.02) and (20.03) describe cylindrical waves diverging from edge 1 (y = a), and the second terms The -a). describe the cylindrical waves diverging from edge 2 (y
FTD-HC-23-259-71
116
nonuniform part of the current on each side of the strip also has the form of waves which diverge from edges 1 and 2, and are an "analytical extension" of the corresponding terms in Equations (20.02) and (20.03). reflected from it. The current wave encountering the opposite edge is
Or else one may say that each of the cylindrical waves propagated from edge 1 or 2 undergoes diffraction by the opposite edge (secondary diffraction). If the strip's width is sufficiently large in comparison with
the wavelength,
current wave near the strip's edk will be the same as on a corresponding half-plane excited by a linear source, the moment of which is selected in a definite way. It is also obvious that the current Consequently, the waves reflected from the edge will also coincide.
problem of secondary diffraction by a strip may be reduced to the problem of the diffraction of a cylindrical wave by a half-plane.
"
to the half-plane's edge and passing through the point Q (Figure 46) may be found by means of the reciprocity principle. In the case of E-polarization, it is determined by the relationship
," ,(20.06)
and in the case of H-polarization
H, = s if.,(Q).(2
Here pz (mz)
is
the electric
(magnetic) (moz) is
current filament passing through the point P with the coordinates (0", R), and Hz (Q) or Ez (Q) is the field created by the auxiliary filament at the point Q. Now let us remove the auxiliary current filament to such a distance that the cylindrical wave arriving from it may be considered
FTD-HC-23-259-71
116a
I.
to the point Q.
I.
In this case,
in
*
Ia
accordance with 1 and 2 the field created by it at the point Q will equal
,(Q) E,. (0)fu (d. V - 1,"1) - u(d,...), +
~E #"
* I
'+')1.
(20.08)
Figure 46. Diffraction of a cylindrical wave by a half-plane. Q is the source, Q* is the mirror image of the source, and P is the observation point.
The functions u introduced here are determined (for the values OV'.<) by the equations
e-*
('',ith w"
0< NF ?
"(0
m
(d.
" 0
S.-ikdc
with
<#2x.
(20.09) and the quantities Ez (0) and Hz (0) are the values of the primary field created by the auxiliary filament at points corresponding to the half-plane's edge. In accordance with Equations (1.21) this field may be represented when kR >> 1 in the form
Ea.
2a 4R
7
and (1.22),
,0
!kpT
(e "
(20.10)
FTD-HC-23-259-71
117
Consequently,
kr)iI/"
(20.11)"
at the point P,
the field
!-) (110H,= ik'n lit (d. It is easy to see that the exponent (d.
ebi'dx-'deo(
V'V
S#
(20.12)
2 it
"
,",i
in these expressions
corresponds to the primary cyllndrical wave arriving at the observation point P, and the exponent eihI-d' corresponds to the reflected cylindrical wave. The moments mz and pz must be selected in the direction V" = w (Figure 46) such a way that in
equal to the field of the primary edge wave above an infinite, ideally conducting plane. We will conclude these calculations in the following sections, but for now let us make still one other crmment on tne formulation of the problem. In the previous chapters it was shown that the ::!attering object
may be approximated by a series cf sources "Iru;.Incuz" lines and points. Therefore, the problem of secondary diffraction may be formulated as a -roblem of searching for functions which describe the continuous change of the field of each sucli s:urce during tl.e passage through the boundary of the llzht and shadow orrcsponding to the source. 21. Seccnd-ary Diffraction by a Strip (H-Pclarization)
A current filament with the moment m, which Is positioned above an ideally conducting plane (hI 0, Figure 46) creates in space the field
F7D-HC-? 3-25!-71
11,
eio
H.- I, *k'n,i
lpr;'I IAW.& the
timary e'd't, wave
p21.02n
1 ta~kes m"
In
the d irect.ion
(itRS.+)
the va I u
(.0 12tR
who.r'e
II
lihi
rwitd or
ut
Wu
V.
(Q)
I:o Inlt
t' Ianv r
en or eC
we'C rind the
rv ev pl'Imat
In
ioal
andt
(m.'1 .
the tV I :iment
[ei~im.aly t'd1lt
Iwiimiitl
field or
to appu'exiluate
waive,
lIn
the rorm
the halI
-pI lant',~
and oovvoo'pontld I n
to otdt'o
("
~ut
.i
bl'et' iv't'svtt
itn thoic
rii
,milIietlf''i
2/
H.M () (1 + /.-
I I
I I
i,
....
Sil.
O
"X
72
a.
~~~~~
p
"H##R(1)
eL"""
(21.07)
describes the wave radiated by the source mlz. The sum of these waves equals " ' (s -! -.L) " 4
V(h-
The problem of Figure 47. secondary diffraction by a strip n . thm+and m are the sources the fields of which are used when approximating the primary
H,(I)=
"''-
"
it
-- e"'',"
'
(21.08)
The first
term in
this expression is
edge 2, and the second term ii the field radiated by the filament which is loacted above the ideally conducting plane x ='0 and has the moment
fi
4-e 4,- i.
Ho' -I "
lik~ai' A o0 .(21 -
0)
where
(.
10)
FTD-HC-23-259-71
120
iN
Summing the secondary wave which has been found with the unperturbed primary wave from edge 1; we obtain
H. (1 -2)
2VrMMieLS
Hs -;( x
c'*'dq
"(21.11)
This expression reduces to zero if one assumes 0 = - w/2; consequently, the secondary diffraction eliminates the field discontinuities which occurred in the previous approximation when 0 = - w/2. However, in the direction 0 = w/2 the field (21.11) is different from zero. Since H is an odd function of the x coordinate, the relationship
4.
components
Hz and Eo will
undergo a
S=
n/2.
discontinuity with a transition through the direction The reason for such a Jump, as before, is that in our calcu=
0 is
Again the resulting discontinuity has an order of magnitude yr= I vi However, It is clear that one may completely eliminate the field arising with multiple diffraction of the following terms in discontinuities only with consideration of multiple diffraction. the calculation of fields consideration
requires specific
order of
smallness in the expansion of the primary edge in inverse power, of yF* (see, for example, [46]). All this greatly complicates the calculations. Therefore, we, using the condition ka >> 1, will limit and in order in the plane x 0, we will proceed
ourselves to an investigation of secondary diffraction, to eliminate the discontinuities in the following way.
=
FTD-HC-23-259-71
121
to
let uz
H,(1 -2)=
w-e,(|X
e'dq
1-+("e-i'|
-+ ,,(1)
-(21.12)
will give qualitatively correct results not only when also with all other values of 0. Actually, _.4clqse to zero if yVkcos(-.+)>! and in second term remains, as must be the case.
y--,
but
may be investigated as an interpolation equation, and it may be applied with any values of . It is easy to establish that
now the fringing field does not undergo a discontinuity with the since Expression (21.12) becomes zero passage through plane x - 0, 3 when 4p_-t_.
It
is
(21.12)
automatically
Essentially,
this
substitution is
assumption that the moments of the filaments, used for approximating the primary edge waves, direction (that is,
M
in
+
i 4
H
"'* g
(1e''"
)
t
"
(21.13)
Such a determination of the moments of the auxiliary linear sources is used, for example, in the work of Millar [47]. Precisely in the same way that Equations (21.06) and (21.07) were obtained, we find (when x > 0)
FTD-HC-23-259-71
122
H+ (2)-
IS, g 12)
j
4
r-
""
"/ (21.14)
av rm cot
ik
44(osnq
((2H )
(2) "+H. eg
-e e
(21.15) ( .
These expressions give the field created by the filaments which are located above the ideally conducting half-plane -- oOiYta and have
the moments
:--nl+-
~---I~g~)eihuI-".
here
(21.16)
(21.17)
Furthermore, summing (21.14) and (21.15), we obtain
Air
H 8 (2)
H,,
g (2)
ef~dq
eIh.
sin.--sinl
4.)
FTD-HC-23-259-71
123
Ho(2-
+ H,,g (2)
.(21.19)
It is not difficult to see that the resulting expression becomes zero Consequently, the secondary diffraction if one assumes ?=-; in it. eliminates the field discontinuity which we had earlier (1 6) when I=mj, but at the same time it leads to a field discontinuity when Again the resulting field discontinuity may be eliminated T=by the above indicated method, replacing the quantity 9(2) by g(2) + depend on the obserthat is, by assuming the moments m 2 z and m vation angle *. Actually as a result of such a substitution, we obtain from (21.19) the expression
ewe"41M-11 ,,
e" -'-"
+
.--
+ H
(2'
(21.20)
. This expression may be investigated as .=-2which vanishes when an interpolation equation which describes the field created in the ,ij<-2- by the primary wave of edge 2 with consideration of region its diffraction at edge 1. Now summing (21.12) and (21.20), we obtain the following expression for the total field scattered by the strip:
(21.21)
FTD-HC-23-259-71
124
,/1
(.
,x
e-4X
eAwdq ,(21.22)
is the shading function of the primary wave travelling from edge 1, and
f; G(2, ?)=-Le
ed (21.23)
Is the shading function of the primary wave travelling from edge 2. These functions show that the primary wave from edge 1 undergoes the , and the wave from edge 2 undergreatest perturbation when '-goes the greatest perturbation when An important property of Equation (21.21) is that it becomes zero when ?=--= that is, the field discontinuity which we had earlier at the plane x - 0 is completely eliminated. In concluding this section, let us return to Expressions (21.11) and (21.19) which lead to discontinuities of the fringing field in the plane of the strip (x = 0). One may show that the sum of these expressions
A""
FTD-H-IC-23-259-71
225
/_
__
agrees, when I'kacos(-!-.+)>1 ,with the asymptotic solution obtained in the book [501 by means of integral equations. The solution found
in [50] has the greatest precision when *O,' useless if a=-or 9 22. -'-0 , and it is completely
It is known that a current filament with an electric moment z which is found at a distance h from an infinite, ideally conducting plane (see Figure 46) creates in space the field
E, =
(22.01)
trans-
si*
*R,
el
(22.02)
E, =ES() M IM
"(22.03)
where E oz(q)
*
is
the value of the incident plane wave field at the Consequently, the primary edge wave in the direction one assumes the filament moment
p,"h'A
The field, created at the point P by the current filament with a moment pz which is parallel to the half-plane's edge and passes through the point Q, is determined by Expression (20.11). Expanding
FTD-HC-23-259-71
126
the right-hand member of this expression into a series In terms of the small quantitj h(h - 0) and limiting ourselves to the first term which Is different from zero, we obtain
Esikp ;-In(.I
e
-R~d.(22.05)
By means of relatior.ships (22.04) and (22.05), one may show that current filaments with moments p-z and pz which are located on the and correspond to edge I (see Ideally conducting half-plane --a<eoo Figure 47) create In the region 1914the field
r
E1 1~7
*OS~
(i-1: )
*i4Sq+.
-sin (2:kd+
+E4/I)coSO*
(22.06)
+P which are located and correspond to
The current filp.ments with the moments Pz and above the iCeally conducting half-plane -edge 2 create In the same region the field
egy<a
j.*ib-
rcos' 1.
?
+,4) _4 1
e*dq'
+ I si (-i +
+2f4 2
0Hi-,(
r? - ()
2) .,. .-... ,
"
"EJ(2) -.
(22.07)
PTD-HC-23-259-71
127
+/
The first terms It.Express'ions (22.06) and (22.07) &re the desired secondary waves, and the last ..erms In the expressions are the fields from the current filaments located abcve the Ideally conductln- plane x a 0 and having the moments
-P
(22.08)
(22.09)
where
Suuming the secondary waves which have been round with the unperturbed primary waves, we obtain the total field scattered by the atrip
,m n=
I.op
a~R-
5
4) 0 8----.(22.10)
Now assuming, as in the case of the ',-polarization, moments plf and p depend on the angle #, by replacing
that the
7(1)
we obtain
by
M"
and T(2)
by M .(22.12)
- F(,
((
22.12)
rrD-HC-23-259-71
128
II
"where
12
F(2,
v-e "Y
*i12iha(I -sin yp)
-;
.-
edq +
+7
are the shading functions. They show that the Primary wave from edge 1 undergoes the greatest perturbation near --, and the wave from edge 2 undergoes the greatest perturbation in the vicinity of f'-. 5 23. The Scattering Characteristics of a
SPlane
(22.13)
Waveby a Stri.p-
which were obtained above the field scattered for by a strip approximately take into account interaction of the the edges and are valid when " However, are not applicable with they a glancing incidence of a plane wave on a strip (when a -:) . In order to find equations which let us proceed in the following way. for the field radiated by the strip in dence of a plane wave in the direction
E,= i-- = E,,[ F(2,
are applicable in this case, Let us write the expressions the direction a with the inci0 (Figure 45)
r)j (a ) eia(In"'-P..f+ i
,si ,c
,
g (1)
4;~st, -,,
ehI+,)
FTD-HC-23-259-71
(23.01)
129
129.
Here 1 , but * cannot approximate +w/2 Now let us note that the expressions for the fringing field must satisfy the reciprocity principle - that is, they must not change with the simultaneous replacement of a by * and * by a. Comparing Equations (21.21), (20.12) and (23.01), it is not difficult to obtain the expressions
S-'H
+F(2, 9)0(,
2)/(2)e,
,
(23.02)
which satisfy the reciprocity principle, have no discontinuities anywhere, and are suitable for making calculations with any values of a an -. I iI< ). From the second equation of (23.02), it
follows that Hz - E. Hz - E
a
0 when
that is,
-
Moreover,
a plane wave
if
e=-t.-;
that is,
polarized perpendicularly to the strip does not undergo diffraction with a glancing incidence.
The resulting Equations (23.02) may be investigated as interpolatlon equations. Actually, with Iak<-! when Ykacs(-.:.)>1 the functions F0l. 2), F(2, 2)' GO(,2) and G (2, a) are close to one, and
Equations (23.02) (22.12). But if change into the previous Expressions and VCcos( a--C 2) 1, (21.21) and then the functions
F(1, ?). F(2, p) , G (1, 0) and G (2, 0) are close to one, and EquatLone (23.02) change into Equations (23.01). Let us recall that the futi7tions F and G are determined by relationships (21.22), (21.23) and (22.13). In the direction of the principal maximum of the scattering diagram (0 - a), Equations (23.02) take the following form:
PTD-HC-23-259-7I
130
7Fl
. ) P---F .co .
. ....
.F.1.z) .. . ..
(23.03)
we have H5 = 0,
E,0
*.
.,
iv
!i d :e_.:.
...
(2 3.0o4)
It is interesting to observe that Expressions (23.02) to some extent take into account, in addition to secondary diffraction, also
tertiary diffraction. Actually, for the values I121and 1<!
we have
1) 6-osln
0iha(2-sill
4--sin
9.
, 4o
2-
-+ 2o,
the right-hand me-:'.[er of
this equation is illustrated in Figure 48 (Figure 48a corresponds to the first term; Figure 48b corresponds to the second term, etc.). Taking into account condition (6.15), one may write the equations r but l"<-for the fringing field in the left half-space ( } Thus, the functions 0(1, a), G(2, ) and in the same form as (23.02).
FTD-HC-23-259-71
131
al
h)
Figure 48. The schematic diagram of the waves corresponding to the various terms
in Equation (22.05).
as before, (22.13).
and
(1, p) jr e-)*d
- -a i 2
i *in -iI
(2,
e'4dq
(23.061
-4-r ,..
- 24. - On
F (2, ?)
K
S-
5
I.
e'dq ;z
"FD-c-3-
(23.07)
FTD-HC-23-259-71
132
toos
1.6106
also-4. W Q
2li it
I.
:itp
II
:IIwt
il .
Itui aho-
A. 1
-m
I it it11
CIt (w It
''.k1
4ett'jl r'~~ I I
t.1I 1'
11T I
th
Iho
::l it.,
Ir
aid t I l
t*(:1
h.
1' 11,1
1'i he
*e 1.1M j(P2. )
gj~"
witi'
it ,
11 o llI it C Wilvil(l
\t.' ta
0)
~-.~i-V
Calculations of the scattering characteristics were carried out based on the equations derived above. These scattering characteristics are the functions h(M, 0) and e(a, f) determining the fringing field by means of the relationships
~~~E, = E,,/ae(,
Ff H 1 ah.,)pf7e
The calculations were performed for the values In Figures 49 - 62, ka=Vs and
(23.13)
a=VIO
1) the
/
the functions
h 0 and e 0 correspond to the field from the uniform part of the current (the physical optics approach); 3) the functions h1 and e1 correspond to the field from the uniform and nonuniform parts of the current, but without consideration of the interaction of the edges; 4) the functions h 2 and e 2 correspond to the fringing field with consideration of secondary diffraction calculated on the basis of equations
(23.13),
Thus,
in accordance with 6,
h'.' '
-(23.14)
giv
where
Cosu +-
(23.15)
r.t<-,.
I'--
The results obtained show that our approximation equations arree satisfactorily with the rigorous theory already when ka.d-Y 18 although in source (? .... the given case . 2Ca ) approximately one and one-half wavelentgths In the direction toward the are fitted into the width of the strip.
FTD-HC-23-259-71
134
co 0 04-
a- 4.3
CUd
>3 CU0TA
a. z
l
4) oi
1 1
.
d-14
0CU 0) .
coo
H -1 0
03) 0 4.3
0IL 4) 1 I
FT-Hc
bd0
-T-T~ - I59toC
I~
b :
01l
a)
El
LL
V-
u -. .
. . .
r.....
Figure 51.
- -iT +
a.
when the functions e 0 and h 0 , and el and hl lead to qualitatively incorrect results, the functions e 2 and h 2 give, as in the remaining cases, fully satisfactory results. Actually, the curve 1h2 1 coincides almost everywhere with the curve lhl (Figure 49-54) within the limits of graphical precision. But the calculated values of the function le 2j differ from the corresponding values of the function lei only by hundredths of a percent (Figure 55 - 62). The better agreement with the rigorous theory associated with the Epolarization is explained by the weaker interaction of the edges in this case. A certain discrepancy of the curves jh 2j and lhi in the vicinity of the principal scattering maximum is explained by the interpolation character of our equations. As a consequence of the interpolation character of Equations (23.02), the integral scattering diameter obtained from Expressions (23.03) when a = 0 does not coincide with the integral diameter found by Clemmow [46] in the form of the first terms of an asymptotic expansion in inverse powers of k-a. However, our equations,as dl.tinct from the similar equations obtained by other authors, allow one to calculate the scattering characteristics with any Incident anrles of the plane wave. FTD-HC-23-259-71
strip (a
-1/2),
-- I
-J
--
N- L-
---
V -
IIIIF
--- t
%-b
Cfib -H4
*4
FTD-HC-23-259-71
137
*I
,,, .-
-I.
.
54.
e as
Let us note that the functions e(cx, *) and h(ci, *) for Figures were calculated on the basis of rigorous series which were 49 -52 obtained by the separation of variables in the elliptic coordinate system (compare [23)(1) 1 24. Secondary Diffraction by a Disk
Let us refine the approximate solution of the diffraction problem for a disk which was round in Chapter II. Let an infinitely thin, ideally conducting disk or radius a be found In free space. Let us orientate the spherical coordinate system in such a way that the normal n to the incident wave front would lie with the z - w/2, and form an angle Ylliptic in the half-plane s axis (Figure 63). Let us prescribe the incident plane wave field In
Ihuini ni Ee.a.if el
(24 .01)
FTD-HC-23-259-71
138
I IT
0.)
1L
0~i
Q)
CD,
0 Zz(4(D
0l w
Ei-4 74
FTD-HC-23-259-.71
139
A 41 LI% 0 9
E4 -4.)
'-r4 4)0. 35 4)4)04N)
4)-34J
0-4
4).
0 r
bob
I v
E9-4
- -IST-a
-.
-~
94
--
-q-
IA i -Tr-
FTDHC-3-247
-14
--
-- 71
00,
EU-4
InF F)
1to h
0T-C23297 14
--
--
1I
I. ..
t~ LI
/
i
Figure 62.
-v-i+ a.
H"=E - 26 e )- go J(C In accordance with 1112, the fringing field in the plane tw/2 is described (when R > ka ) by the equations
+ 1 ;
) 1(l.
J. Q)J.
(2
nandThe quantities These expressions are valid when included In them are determined by the relationships:
-. _) --- s Cos _sin -i -2+ ' 2
1(2.1) g
-,,a
u 8)-.-8), 1(2,
,a (2,B -),
(24.03)
ka(sin 0 - sin ).
(24.04)
FTD-.HC-23-259-71
1.42
>,
/,'
I with .
- 2 2
(24.05)
with
72 I (2)withcp--2~
I(2,?)= /(2) with
?-j
11() withl--
(
e(2,Zl
1 0( ) with?---3 ----'
g (2)r with
(24.06) and the functions f(l), f(2), g(l) and g(2) are determined by the Equations (12.03) and (12.04). When C >> 1, Expressions (24.02) take the form
H=.=* (2, ;)e
4
He E,4
og 2_g(2 ,
(2o7)
They show that the fringing field in this region may be investigated as the sum of spherical waves from two luminous points on the rim of the disk with the polar angle * = -it/2. The diffraction by a disk of
each of these waves may be studied as was done in but we shall proceed differently. the case of a strip,
FTD-HC-23-259-71
143
/0
Figure 63. The cross section of a disk with the plane yoz; n is the normal to the incident wave front.
Figure 64. Excitation of a halfplane by an elementary dipole which is located at the point Q.
Let us compare the shading of spherical and cylindrical waves by a half-plane. Let an ideally conducting half-plane be found in free space, and let there be an elementary dipole at the point Q (Figure 64). Let us find the field in the plane perpendicular to the half-plane's edge and passing through the point Q. In accordance with the reciprocity principle, for the electric dipole by the relationship pea and for the magnetic dipole by the relationship H. H. (Q). (24.09)
//
it
is determined
(241.08)
Here p (mz) is the electric (magnetic) dipole moment found at the point Q; poz and moz are the moments of the auxiliary dipoles which are placed at the point P; Ez(Q) and Hz(Q) are the fields created by the auxiliary dipoles at the point Q.
FTD-HC-23-259-71
144
Now let us remove the auxiliary dipoles to such a distance that the spherical wave arriving from them may be considered to be a plane wave on the section from the half-plane's edge to the point Q. In
this case, in accordance with Equation (20.08), the field created by
) - u (d. 71 + 7")].
(.10) (24.0
k'p.,
(24.n1)
(2.1
determine the fields created by the auxiliary dipoles in free space (with the absence of the half-plane) at the point 0. Consequently, the fields excited at the point P by the electric and magnetic dipoles which are found at the point Q above the halfplane equal respectively
E, = k'p, fu(d.
With the absence of the half-plane, these dipoles create at the point P the field E, =k'p, ES
,(,-,) 'Ad cs-
--et* hR
H,= k'm, 6
(24.13)
a, u (d,
FTD-HC-23-259-71
145
In the case when the current filament passes through the point
in
equations
71 -) R. - iA',n.[a (d.
4]
(2.5
With the absence of a half-plane, these sources create at the point P the field
) e'ilm"~
~" "
(24.16)
(24.16)
Comparing Equations (24.15) and (24.16), we obtain the same Expressions (24.14) for the shading functions. Consequently, a spherical wave in the direction perpendicular to an ideally conducting half-plane is shaded by it the same as a cylindrical wave. Let us note, however, that Expressions (24.14) art not equivalent to Expressions (21.22), (21.23) and (22.13), since the first represent the shading function by a half-plane of a wave from a single source, and the latter represent the shading function of an edge wave which we approximate by waves from two sources located on both sides of the corresponding half-plane. Since the shading functions of spherical and cylindrical waves are the same, the edge wave shading functions of a strip and a disk also will coincide. Therefore, the approximation expressions for a field scattered by a disk which take account of secondary diffraction may be represented in the region p=_=--. 0 (with C >> 1) in the following form:
FTD-HC-23-259-71
146
Its
-6
;')I;I, )A,
A)ei
fe
l!t, Io t: . V..I',l l. .. )
W heretz (.21. i
the" f 'w)
untll~
(ti .' u .
et
*l Io .; F andl
'r : Ioy
|
(I aifvt obt
alit
t
f'trom -4',
or o
by o.
lt I,
" 11 , of'
- f'o l ~~tilt
| owi nr+
df ' I" I
IT9
/I
AI
l Nl'
lt
I "s
w;
t A'
I he"
d vz
by.
a1 dis:k
nlt
hi
rvol'(t Iont
pi
w'~%c I: n t +1'
tlth
n lt
t v m it ! ''Incd t onf'
im t o rigth
a t
a"i anIv'.
j(I'(.,t
+II
, ""J,
l, ",'.' )g 1t.,'
)TI
Starting from Expressicns (24.18) and (24.19), it is not difficult to write interpolation equations for the fringing field which are suitable for any values of y and 8 in the interval (0; w72), but when = /2;
41, -- u
E,-- H,
S:
(10-{[ (2.0)Go(1,
a)g(,a
(24.21)
Let us note that when y - 0 these expressions will be valid for any values of the azimuth f , since then any point of space may be considered to be located in the incident plane. In the direction of the scattering diagram's principal maximum
--
that is,
when
&-I.
?---
However, these expressions have an interpolation character,and with small values of the angle y it is impossible to consider them to be more precise than the simple equations of 9 and 12. In particular, with y - 0, when the fringing field must not depend on the incident wave polarization, they give values which are different for the Epolarization and H-polarization by small quantities of the order of %-I . Therefore, in this case (when y = 0) it makes sense to use Expressions (24.20) and (24.21) only far from the z axis, switching to Equations (24.02) near the z axis.
FTD-HC-23-259-71
148
7',
Equations (24.20) and (24.21) have the following important proThey do not have discontinuities, they include the case of perties, glancing incidence of a plane wave, principle. is, From them it and they satisfy the reciprocity O when O=that
U
Y.=--!-. with any values of 8. if plane z = 0. Moreover, Eo-=H,=0 that is, a plane wave polarized perpendicular to the disk's plane does not experience diffraction with glancing irradiation of the
disk.
0) (1, G (2, 8)+ F(1. +F( =H, =-O r) F (2. G(2.,Z)f1(1, 8)]J. (C))+ 9)+ + i [G (2 . (G-O 1, &) (2, g 8)-t-
2.3
" -G(, - i) G (2, 8)(g(1, 9)] J, (C) + (1, )1] (-R + G(1,-- 0) G(2,8)g (24.23)
--airf(,8 . -
cos -
8+6
-s in( ,
the
FTD-HC-23-259-71
149
()
--
2 sinyl I -$in-(
2stia-
(24.26)
it makes sense to use Equations (24.25) only far from the z axis, changing to Expression (12.15) of the previous approximation in the vicinity of the z axis. A calculation of functions Z(--T) and Z(--T) (Figures 65 and 66) which determine the effective scattering surface [see Expressions (12.17)] was performed on the basis of these equations when ka = 5. A comparison was carried out of this calculation with the results of measurements. The two experimental diagrams (the dashed lines)(2) depicted in Figure 65 characterize the experimental precision. As distinct from the previous approximations, which lead in this case to qualitatively incorrect results [see Equations (10.06), (10.07) and (12.15)1, we observe a satisfactory agreement of theory with experiment. For verifying the results obtained, a calculation was also carried out of the functions v)() 0 ) and V02) ( ) [see Equations (9.07)] when ka = 5 (Figure 67 and 68) with normal irradiation of a disk by a plane wave. Curve 1 corresponds to the field calculated from the rigorous theory [34]; curve 2 corresponds to the field from (2)Footnote appears on page 162.
FTD-HC-23-259-71
150
,/
haS
theoretical1 -- experlmenta 1
1
r-|
j-
"
-/ -' I.
effective The diagram of a disk's Figure 65. wave's plane the when surface scattering the incimagnetic vector is perpendicular to dent plane.
FTD-HC-23-259-71
151
* ,-
,dw
Figure 66. The calculated diagram of a disk's effective scattering surface when the plane wave's electric vector is perpendicular to the incident plane.
Curve
3 corresponds to the field from the uniform and nonuniform parts of Curve 4 correthe current, but without the interaction of the edges. As sponds to the field with consideration of secondary diffraction. is seen from these graphs, consideration of the edge interaction refines the previous approximation and ensures better agreement with the rigorous theory results. The problem of secondary diffraction by a cylinder may be solved However, considering that the corrections which by a similar method.
FTD-HC-23-259-71
152
/*
A I I I I
l.,.
I:c M. Sii
i1A I i'
I'
The function "( for a
Figure 67.
disk with normal incidence of a plane wave (curve 4). Curves 1, 2 and 3 from Figure 20 are drawn for comparison.
depend on the secondary diffraction here are small (on the order of 1 dB) when ka = 7r,kl = lO, and the equations are substantially more complicated, we shall not cite them here. In the problems investigated above, the edge waves have the character of cylindrical or spherical waves -- that is, they decrease rather rapidly with the distance from the edge. Therefore, in the case when the linear dimensions of the faces are approximately two wavelengths, it is sufficient to limit ourselves to a consideration
FTD-HC-23-259-71
153
o
:Pa
ol i
sII
-1A
3~~~~~~-'
rw
o JrmFgr-1
bodis Theyseodr expressIons werepderIve weithth hnelpiat ofthysca cronsierfatdionsi which donth petgen toabes dathemaiasesloly rigoout and~~ they ar adqaefrisufficientl sor th wae.Lh iterature,
FTD-HC-23-259-71
1941
A majority of these works also are not characterized by mathematical Therefore, rigor, and they are based on certain physical assumptions. Only in one may relate them to the physical theory of diffraction. a few works (related to the simpler diffraction problems) did they succeed in obtaining specific results at a higher level of mathematical rigor - more precisely, while developing asymptotic methods of mathematical diffraction theory. We will briefly list sequence: 1. Diffraction by plane, infinitely thin plates a circular hole). (an infinite a the most important results obtained in grouping the material in a
the following
strip, a circular disk) and diffraction by auxiliary apertures in flat screen (an infinite slit, 2. cylinder,
3.
When investigating the first group of diffraction problems, it is necessary to keep in mind the principle of duality [4] which enables one to easily change from a strip to a slit, from a disk to a circular hole, etc. In the literature as a rule, they preferred to investigate apertures in an infinite flat screen, whereas in our book, diffraction by a strip and a disk was studied. This approach facilitates the transition to three-dimensional bodies beginning of this chapter). Based on the time of appearance (if we do not consider the works (see the remarks at the
of Schwarzschild [15] which we talked about in the Introduction), one should first of all mention the works of Braunbek [28 - 301 which were devoted to the diffraction of a scalar wave by a circular hole in a flat screen. Assuming that the plane wave is incident normal to
FTD-HC-23-259-71
155
the screen,
the form
of a surface integral.
incident normal to an ideally conducting screen with a circular hole The field was calculated in and also the field in the far zone and the transmission coeffi-
cient (the ratio of the energy passing through the hole to the energy falling on it) were calculated. secondary diffraction was
field intensity in the far zone agree with similar expressions following from our equations (5 9). Karp and Russek [51) studied diffraction by a slit when the incident wave's electric vector is in the case edge.
They investigated each semi-infinite p,rt of the screen as a halfplane excited by the incident wave field and a "virtual" source localized on the edge of the opposite half-plane. The moments of these sources were determined from a system of two algebraic equations which were obtained by using the asymptotic expressions resulting from the rigorous solution for the half-plane. Secondary diffraction was considered, and partially the general interaction. Special attention was allotted to calculating the transmission coefficient, but equations for the scattering characteristics which would be suitable with all directions of incident wave propagation were absent.
FTD-HC-23-259-71
156
./
The interaction of the edges was considered. was not investigated, since it turned
out to be too complicated for investigation by this The "geometric deals with diffraction amplitutude theory of diffraction" rays is
of Keller [42 -
of special interest.
each ray point on the basis of geometric considerations and the law of the conservation of energy. is The initial diffraction ray amplitude
assumed to be proportional to the incident ray amplitude at the diffraction. The unknown proportionality constant phase difference be-
point of its
tween the amplitudes and the initial from a comparison with the results tion problems. incidence found. In this way, of a plane wave ona slit the fields
s determined
of well-known solutions of diffracscattered with the normal a flat screen are and hole in
These fields
are obtained with consideration of multiple difsolutions, since Moreover, and also
fractions, their
calculation was started from approximation relations. not applicable near caustics,
In
a recently published paper of Buchal and Keller [52], problems for holes in The caustics and shadow boundaries layers,
a new a flat
Recently, particular,
the method of integral equations has been applied to problems of holes in [53, a flat screen. In 54] reduced the solution of this problem
FTD-HC-23-259-71
157
17
to an integral equation for a "shadow" current which is, In our terminology, half the nonuniform part of the current. The resulting integral equations may be solved (with any ratio between the dimensions of the hole and the wavel4ngth) by the method of successive approxiMoreover, they allow one to obtain asymptotic expressions mations. In Reference [55) Greenberg found which are suitable for short waves. an asymptotic expression for the current on a strip with ka >> 1 (2a width). Greenberg and Pimenov [56] obtained a similar is the strip's solution in hole. the case of normal incidence of a plane wave on a circular an asymptotic expression was found for the width and inner diameter of which Using the same method,
are a great deal larger than the wavelength. The above listed works [53 - 57) already relate to the mathematiterms of the asymptotic in them the first cal theory of diffraction: expansions for the current were obtained with the desire evidently to Unfortunately, the also be able to calculate the following terms. asymptotic expressions which have been found up to now refer only to currents, and one is obliged to calculate the scattering characteris-
As a consequence of the tics by means of numerical quadratures [56]. rapid oscillation of the integrands, such a method leads to rather unwieldy calculations and does not enable one to formulate a clear representation of the fringing field formation, and also does not allow one to study this field properly. Millar [58] investigated the problems of electromagnetic wave diffraction by slits in a flat screen. The system of integral equations obtained by him for the current is solved by the method of The field in the hole is calculated from successive approximations.
the currents which are found, and then on the basis of the field in the hole the field in the far zone and the transmission coefficient are calculated. All the indicated quantities are represented in the
form of an asymptotic expansion in reciprocal powers of the parameter Vka. A solution also is obtained in the case of glancing incidence of a plane wave.
FTD-HC-23-259-71
158
V ..
-.
.. .
...
.. .
._
FTD-NC-23-259-71
159
S.
.. . .
diffraction ray concept is used for calculating the scattering of scalar and electromagnetic plane waves by a finite circular cone with a flat base and also by a cone having a spherical rounding off instead of a flat base. The resulting expressions.,are not applicable in the vicinity of certain irradiation and observation directions. In 5 17 we showed that the field scattered by a cone and by certain bodies of rotation is not expressed only in terms of the functions f and g, which refer to diffraction rays diverging from a wedge edge. This result evidently attests to the impossibility, of complete calculation of the scattering characteristic with the diffraction ray concept. Diffraction problems arising in antenna theory are usually distinguished by their great complexity, since the corresponding metal bodies (mirror, horn, etc.) have a comf~icated shape. Since the dimensions of these bodies and the dimensions of the radiating apertures are considerably larger than the wavelength, the application of physical diffraction theory to antenna problems is very promising. only the first steps have been taken in this direction. Thus, Kinber [60, 61J performed a calculation of the decoupling and laterdl radiation of mirror antennas. The feature specific to mirror hntennas is
that diffraction rays arising at the mirror's edge undergo multiple
Diffraction problems relating to an antenna dipole - a thin cylindrical conductor - are investigated in Chapter VII, and references to the literature are also given there. In conclusion, let us say a few words about diffraction of short waves by smooth bodies. The basic principles relating to such problems were set forth in the fundamental works of Fok and Leontovich. These principles were established by the following methods of mathematical diffraction theory:
FTD-HC-23-259-71
160
(I
1i'ire
t:: nii
ior inizmpaI~'I n [c
,q i
auo ~'
ourmtl or)il
m Wi~
thnurrao
't'i or
Io~ l':
ioea
'~ I 4* ji y t he faiet 10
I r AO.Viid
~ r r* Ii
't, r; I It
tuo
'Ivc I r d e
t'oi
b'ody
I 701).
or an~ of ~I
Iaindvv ey
an' wa
I* P shown
cm h
In I
hi
Ito
vId gout''
Io
ci'8*i
In-
I
iat I Ion
andv~ur. v ic'v
Theuc
p.'ta
o r' c'
1.:ty
,'
'I I no t I olIn 1I
cv Itir oyl'o
I .
1dn don 1 y
FOOTNOTES Footnote (1) on page Footnote (2) 138. These calculations were performed
FTD-HC-23-259-71
CHAPTER VI
CERTAIN PHENOMENA CONNECTED WITH THE NONUNIFORM PART OF THE SURFACE CURRENT
conducted of the field radiated by the nonuniform part of the current. In this chapter we will discuss a method for measuring this field ( 26) and we will Investigate signal's depolarization ( 27). the phenomenon of the reflected
An experimental method for measuring the field from the nonuniform part of the current was first proposed for bodies of rotation in Later It w:as :-.',xfn the paper of Ye. N. Mayzel's and the author [12]. that this method has a universal character, a plane wave on any metal body [131. 26. Measurement of the Field Radiated by the Nonuniform Part of the Current in (,. and is suitable for ex(Icd by measuring the field from the nonuniform part of the current
Let an ideally conducting body of arbitrary s1!ape be fon, free space. A surface element of this body is shown in Pi .'e
The coordinate system was selected in such a way that its origin would lie near the body, and the source Q would be located in the plane
x the distance between the body and the source is a great deal larger than the body's dimensions, then the incident wave in the vicinity of the body may be investigated as a plane wav-. Let us
a
0.
If
represent it
in the form
fi----,,e
"
41=,0.
(26.01)
Here y is the angle between the normal N to the wave front and the z axis. Now let us place in front of the source, parallel to the radiated wave front, a polarizer P which transformed linear polarized radiation into a circularly polarized wave. Let the wave passing through the polarizer with an electric vector En lag in phase by 900 behind the wave with an electric vector ET (Figure 70). In this case, the polarizer achieves a clockwise rotation(l). As a result, the incident wave field at the coordinate origin will equal
,
E
-4
(26.CI')
the wave
,- E.,. H.=--E..
The field scattered by the body may be represented in zone in the following way:
I
4L
-j
i." (I 0P- R1
(26.03)
where a is a certain length characterizing the body's size and 7-0.,,) In the general case, the and Z(B?]) are unknown angular functions.
724*
field (26.03) is
an elliptically
4l 9/
P -the
polarized wave. In the direction =X--. toward the source this wave passes through the
polarizer and creates behind it field
MaE,
2
*hR
Figure 69. The problem of electromagentic wave diffraction by an arbitrary metal body. dS - is a surface element of the body. N - the normal to the incident wave front, Q - the source, where
R (26.04)
the polarizer converting linearly polarized radiation to a wave with circular polarization.
I()'
2
(26.05)
If the source radiates a wave of another polarization (H0 L yoz), then the wave reflected
by the body and passing through the polarizer is described at the point Q by similar relationships
I ON
Figure 70.
S2
&
Zj-_
"
(26.06)
the physical optics approach the the uniform part of the current exci-
diffraction of a plane linearly polarized wave by the same body. ted on the body's surface by a plane wave with E-polarization of the yoz) equals incident wave (E O
FTD-HC-23-259- !
165
=-
l
and with H-polarization (HO yOz)
j.
(26.07)
-L USX -- n. e4.
ro
(26.08)
Here Eox and Hox are the electric and magnetic field amplitudes of
the incident wave with E-polarization and H-polarization, respectively; -'.>zcos, is the incident wave phase at the point (x', y', z') on the body's surface; n., ny, n., are the components of the normal to the surface at the same point. Furthermore, calculating the vector potential in the far zone on the basis of this current and substituting its values into the equations
E =--ikAI,. }
kq
( 26.09 )
R c,
sin? -cos
. 1
""
M. ":osTsin ?I d,(26
(26.10)
-- ( lb " 7'- ,R , co 7
eo s
C s en os
(26]1,
:""-(-
-25,)-7V
!'
i k Y . _Cos, _
e l lt
e ,s . ^ S
(26.11)
--t-kecosO
In the case of radar when the observation and irradiation directions coincide (2=,. =----) , Equations (26.10) and (26.11) yield
-- H.,
n-,sin T+ (26.13)
2T2
H.14aE..
-
eI4 elk* 1- 2 R
'
(2 6 .1 4 )
(26 .15)
"INow let us represent the angular functions of fringing field (26.03) in the form
FTD-HC-23-259-71
L4
167
+ 1,
(26.16) the field radiated by Substirespectively. and (26.06) and taking the fringing field
In the case of
where the functions E0, _Z0 and El, E refer to the uniform and nonuniform part of the current, tuting these expressions into Equations (26.04) into account relationship (26.15), let us find
SE. ~~~-H
E
,
(26.17)
H O. -- , -
1-
(26.18)
The physical meaning of the result obtained is as follows. The field scattered by the body at the point Q is the sum of two waves polarized in mutually perpendicular directions. The reflected wave which is polarized the same as the primary radiation of the source is determined by the function , , and is created only by the nonuniform part of the current . The reflected wave with the perpendicular polarization is described by the function Z.-2'--Z'-' and is the field radiated by both parts of the current. Let us note that in the general case the functions E1 and 11 do not coincide, and therefore they are not balanced out in the expressions for E_. In other words, the field radiated by the uniform part of the current in this case may not be separated from the fringing field. Thus, the investigated method allows one to separate from the total field scattered by any metal body of finite dimensions *hat part of the field which is caused by a distortion of the surface (the
FTD-HC-23-259-71
16S
curvature,
note that,
etc.).
One should
of separate bodies, the separable part of the field is due not only to the surface's distortion, but also to the diffraction interaction of the bodies. It is necessary, however, to keep in mind that it is possible to
realize the indicated fringing field distribution not in an arbitrary observation direction, 10 a -_0 is fulfilled but only in a direction for which the condition in the direction towards the for example,
source. Consideration of the nonuniform part of the current also enables one to explain the reflected wave depolarization which we will investigate in the following section. Figure 71 presents the results of measurements(2)
of the effective scattering surface
and calculations
(26.19)
which is dependent upon the nonuniform part of the current excited by a plane electromagnetic wave on a disk. The disk's diameter equals 2a--(A is the wavelength). The calculations were performed with
consideration of the secondary diffraction Since it surface, is difficult on the basis of the approxi 24. flat mation equations for the functions E and F which were derived in to prepare a thin disk with a sufficiently
were performed with an obtuse cone close to the shape of a disk and having a height approximately equal to one tenth of the d.aieter.
the measurements
As is are fairly
the theoretical
and experiinental
curves
especially in
explained both by the model's conical shape and also by the (2)Footnote appears on page
174.
FTD-HC-23-259-71
169
Imeasured
theoretical
-Il
-0 al
ll
,'
Figure 71. Diagram of the radiation from the nonuniform part of the current flowing on a disk.
approximation character of the computational equations. The value y 9-0 corresponds to the direction along the disk's surface, and the value y a 00 to the direction normal to the disk. Reflected Wave Depolarization
27.
Let us again return to the problem of scatterinr of an electromagnetic wave by an arbitrary metal body. The relative position of the source Q, of a surface element of the irradiated body, and of the coordinate system is shown in Figure 69. Lot us recall that the
FTD-HC-23-259-71
17P
source Q is in the plane y0z, and radiates a linearly polarized wave. Furthermore, we shall assume that "the polarizer P which is shown in Figure 69 is now absent.
Let us designate by a the Figure 72. angle between the plane y0z and the incident wave electric vector E0 (Figure 72). The field of this wave will be represented in the form E= HOT = E.,e'k(y 91nli+z e " TOO (27.01)
E,----Esi2.
H, 8 =-E cos 2,
tg.
(27.02)
+ H..'"(Q,.81 M]---,
a ip
(.
(27.03)
Here a is E,.i(yTA) It is
a certain length characterizing the body's dimensions:, and ,.,(t. ,. ) are unknown angular functions.
-
R,
depends
FTD-HC-23-259-71
171
it
if one
investigates t?-c fringing field as the sum of the fields radiated by According to 26, the uniform and nonuniform parts of the current.
the uniform part of the current radiates the following field in the
,id -- -- 'o -
H,., 2 M(27.04)
The functions -0 and Z0 satisfy the condition r0 and are de-
let us immediately
--tg 0,
(27.05)
which means that in the physical optics approach the reflected wave Consequently, the reflected wave does not experience depolarization. depolarization is caused only by the nonuniform part of the current
or, in other words, by the surface distortion.
This is Let us derive an equation for the magnitude of angle 6. the angle by which the electric field vector of the reflected wave is turned in source. rl(2) in respect to the electric vector of the wave radiated by the For this purpose, the form let us represent the functions T (2) and
Z, (a ) ,,
-,,
I
n
(27.06)
-0
where the terms Z1 (2 ) and1-2 rl(2) 11)T correspond to the field radlatod by
the uniform part of the current, and the terms (2)
FTD-HC-23-259-71
172
v"
yo
to
/o
correspond to the field raaiated by the nonuniform part of the current. Comparing Expressions (27.04) and (27.03), we find that
-
-R
i/
V,
Theangefto the field
(E'C..
(27.08)
the
byatedequationrto the in
source (w=-ll
will equal
/, 'l CON1
Asmparn
E. .-
eoarzto (27.08)
This field's electric vector forms an angle The angle 0 is determined by the equation
tf
l-z+
ti tg -
(27.09)
As a result,
6 which characterizes
the depolarization
a=--@.(2710)
Thus, "in the field from the nonuniform part of the current, of a polarizer (5 26), separabLle
leads to depolarization
of the scattered radiation. Specific results from the depolarization calculation of waveo for example, particular, in in the works [77] scatter-
[75 -
In
Refere-ce
on the effective
4> .4
FTD-HC-23-259-71
173
1;
FOOTNOTES
164.
A system c'f metal r".ates parallel to the e vector may serve as the simplest example of such a polarizer. See the footnote on page 86.
- - -
I 71
CHAPTER VII
In almost all the works devoted to the diffraction of plane electromagnetic waves by a thin cylindrical conductor, duced in the conductor was studied, the current inHowever, obtainand then, by integrating this
current, the fringing field in the far zone was calculated. in view of the complexity of this problem, they succeeded in
ing relatively simple equations only in the particular case when the observation direction and the direction toward the source coincided, and was perpendicular to the conductor axis. In the general case when these directions did not coincide and were arbitrary, the expressions for the fringing field became very complicated and unsuitable for making calculations. Since they were obtained by integrating approximation expressions for the current, it turns out that they have still one other shortcoming they do not satisfy the principle of duality. In this chapter, explicit expressions are obtained for the fringing field which are suitable for making calculations with any direction of irradiation and observation. We shall consider' both the primary edge waves excited by the incident plane wave and also the secondary, tertiary, etc. edge waves. The total fringing fleld i1 found by summing all the diffraction waves. 175
FTD-HC-23-259-71
28.
Current Wave-
reduced to an integro-
"-<
are fulfilled, conductor and k Its where a is =-V .
and ka.(1.
(28.01)
for example,
sive approximations
82] proposed a new solution for this equation. Since we will subsequently base our work on the results of References [81, 82], let us discuss them in more detail.
Recently,
Vaynshteyn [81,
Let us assume that the vibrator's symmetry axis coincides with the z axis, and its ends have the coordinates z = zI and z = z2 (L - z2 . zl). In the case of excitation of the dipole by a concentrated external field e',=OP;) (28.02)
the current J(z) in the conductor may obviously be written in the form of the sum of the waves travelling along the conductor with a velocity
c from the excitation point z = 0 and tho ends z = z1 and z = z 2 . In
Reference [81] it was shown that the complex amplitudes of these waves are slowly varying functions of the z coordinate. These functions may be approximately expressed in terms of the function (z), so that we obtain the following expression for the current J(z):
Jlz)
FTD-HC-23-259-71
176
cc
1=.781...
(28.04)
determines the initial excitation point(. gral equation, the parameters conditions and in
value of the current wave propagated from the The function *(z) We will not list is the solution of the intealso depends on here all the propertie of satisfies the addition to the variable z it
k and a.
,( o) -- 1 .
and its
(00) _ ,
( 28 .05 )
absolute value monotonically decreases with an increase of z. which is rather slcw and does not have an exponential is due to radiation. values of the and z = z 2 , respectively,
The constants A1 and A2 determine the initial current waves originating at the points z = z and travelling in conductor. conductor ends J ,)--.J(z,)--O
(28.06)
and equal
A,,
I i,-+(z,)--4-
A,r=
where
.i ik 4,(--z,) L) e- J'ehht
(\28.07) (2".7
VOLe
(28.08)
216.
FTD-HC-23-259-71
177
geo-
metric progression,
+.
(28.09)
(L)
eML+ (z,
-- 4 (.)e'
(t "
The first term in Equation (28.10) is the primary current wave which coincides with the wave excited by a concentrated emf in an infinitely
long conductor. The second term (in all brackets) corresponds to the current resulting from the reflection of the primary current wave
from the conductor end z = Zl, and as a result of subsequent reflections from the conductor ends which arise from this wave
....... J.f--z,)e"'(z- z,)c.
all brackets)
corre-
sponds to the current resulting from the reflection of the primary wave from the end z = z2 and as a result of the subsequent reflections fro 7 the conductor ends arising from this wave It also follows from Equation (28.03) that external field (28.02)
the current
eXcites in
(28.11 )
(--
the current
(28.12)
I, (4.(It leh I)
, I )e'
d(z, -- z) e *0"-"l.
Comparing these expressions with the proper terms in (28.10), we see that the reflection of all the current waves at the end of a finite
FTD-HC-23-259-71
118
'-
(2,<z~y:).
the current
also is
represented in
(see
,_
tz,
z) e'
""-"
-+-A, ? (z
+ A ,(z. --.z)eI
where the first wave in
"I,
an infinitely
(2S.15)
The second and third terms are primary edge waves quence of the cut-off of the current terms of the functions 1P + (z) to the variable functions Se (z) and , -
k and a,
on the angle
satisfy
S(0) = ,.
',.(Z)! o
,(
(Z)
0,=O
W-.
The initial
-'
7u.e'r
Equation
(28.14)
corresponi
to sec:vdr'v, ao
etc.,
form as tlhey
-2 "3-2_-q-7117
Equation
(z8.03)].
etencs WI and
W2
1D
ar-.
and equal
; Le('
:.
-D-e IY-(L)-3
'jehu.
1(28.17)
/(4)= S {e'" -- (z
+ ..
+
f
:,) eiJ1.+Ih(E.-asJ
.10-"-
'i
+8(L)esL
Here besides the wave Seiwz and the primary edge waves,
which we
talked about in connection with Equation (28.14), the secondary, tertiary, etc. waves diverging from the ends z = z and z = z 2 are explicitly written out; they correspond to the first, second, etc. terms in the graphs. Passing to the limit in Equation (28.14) when z2 the current in the semi-infinite conductor (zl, w) J(z) -- S.lei''
-*
, we find
(3 2 .19 )
and, (-',
similarly, we find the current in the semi-infinite conductor z2 ) (Z) --s. e " . e '
FTD-I[C-2R-259-7I 1
It
is
of a Ia:'i.,
v br:t'r
the reflection of current waves at it.s' ends occurs at the end of a L;emi-irifnltc conductor.
Thas, the complex amplitudes of current waves in a thin, finite length conductor are proportional to the functionr ,(z) and ip(z) which monotonically decrease wtth an increase of z as a consequence of radiation. Let us note several properties of current waves in a vibrator. Each advancing wave in sum with the reflected wave excited by it ' gives a zero current at the conductor's end. In the case A (n=:1,2.3--.) and D % 0, a current resonance be-ins In
the vibrator. The precision of Expressions (28.03) and (2S.111) obtained by the method of slowly varying functions is different in various sections of the conductor. It is comparatively low near the conductor's er.do the vicinity of the point z = 0 of a transmittint7 vibrator) where the current waves arise, and where their complex amplitude varie-rather rapidly. As the distance from these vibrator elements increas.', the precision of these equations increases without limit. should be stated that with a more rigorous atproach the amplitudes of all the reflected waves will be determinedl It (and in
7`, 201
by dif-
ferent functions; however, the difference between them rapidly decreases with an increase of the reflection number. The functions ,(2) and *_(z) only approximately describe these current waves, but on the other hand they allow one to effectively sum them and to obtain closed
equations.
Using the variational method for the functions we obtained the approximate, (see [831)
-k
, .::-i
N(Z) = In E(2qx'l
#
( .l.i.i
FTD-HC-e23-259-71
I 1
(28.21)
and
E(Y)=Ciy+isiyJ5 1-1dt.
ci
(28.24)
(82
The integral cosine Ci y and the integral sine si y are determined by the relationships
Q -,--dt,
r'it
(28.25)
and are thoroughly tabulated functions. The equations written above for the current in a finite conductor
are distinguishable by their visualizablity, and they enable one to liken the conductor to a section of a transmission line In which, however, the attenuation of the current waves takes place, not according to an exponential law, but according to a more complicated law which is determined by Equations (28.21). In addition, the diffraction the equations. The conductor's character of the problem is reflected in
specific features as a diffraction object are included in the very As a consequence of this, it slow attenuation of the current waves. is impossible to limit oneself to considering only s-econdary and terLiary diffractions, and It is neceissary to sum all the reflected
waves.
a "resonance denominator"
D appears which takes into account the reoonance properties of a thin, finite length conductor. FTD-HC-23-259-71
29.
from a known equation by integrating the currents not advisable because, (28.03) is
different ends (z
The principle of duality gives more precise leads to the following expression
This principle in
radiation field
---= H,.-.()
2 sinOin In L, C The function
0.
(29.01)
(I
+C'
(29.02)
(29.0 2)
is
excited in
a vibrator by plane
wave
The coefficients
enables one to trace the formation of the is the radiation emf. field of an infinitely in the end z = :;, arnd (theo second
term (one)
concentrated generates
Propagating7
term).
a similar way, is
the primary
edge wave diverging from the The last two terris of .ubteB 1 and
(secondary,
The amplitudes
2
FTD-HC-23-259-71 183
-,0,
(29.04)
which means that the radiation of a finite conductor In the direction of its geometric extension must be equal to zero. These conditions, and (28.16), together with a consileration of the relationships (28.05) lead to the system of equations.
B. + B'? (L)e-i' " zj e'-"'." z,
B,'? (L)e'
" 8,
. (a,) e",,,.
(29.05)
"""(L)' , ,e!'
Ie"b'k;-" (29.06) in a
'~'
--
(L) e-j*h Ct
,
4.
-+
-
, 2RL_.iA
L)..... - e')
'.? (-z,)
e- '1' ' *1
'+
(- z.) e"*hL-'j X
(29.07)
X [(. (L) e-'*'wC#" -- , (L) :".'(L) e"'"' Co 6+ +r"1(L)e~ktji4 (L) e-rhl*o( -.-. . .. , where the secondary, second, out. tertiary, etc.
Thus, the field radiated by a transmitting vibrator arises as a result of multiple diffraction of edge waves at the vibrator's ends. Let us note in connection with this that the edge wave is diffracted by the opposite end of the vibrator in the same way as at the end. of
FTD-HC-23-259-71
184
tu
establish this
conductor excited by a
30.
Diffraction by a
Passive Vibrator Let a plane electromagnetic cylindrical For purposes of generality, electric Then, its field E wave fall z
on a thin (Figure
conductor of length L = z2 -
73).
we will consider that the incident wave's on the conductor surface will-equal
tangential component
Es = E,.-e'2.z
where
(30.01)
E, =Esin..
The current by us in 28. induced in
E=-E, cs a,
w. =
k--cos o
field
(30.02)
was investigated (28.14)
Expression
Let us also note that the fringing the principle of duality. of a passive vibrator
found by such a method does not satisfy We shall seek "he scattering characteristic
by starting
scattering picture which naturally An incident excites space. plane wave, beinrdifprimary edge waves which are Being propagated diffraction along the cmnThe latter, fringing ends. field at the opposite
follows from the previous results. fracted at the conductor ends, radiated into the surrounding ductor,
edge waves,
1url nT
sequential
FTD-HC-23-259-71
185
In 28 and 29,
w, nrti..j t.it
t:,e end.- of a fInlte ltn'it~h cciduct r thE. -.ame as from the end
.;,!m!-Infinrle c-onductcr, an] that the diffraction of these wav:z at eauch end takt.s place in the .ame way az at the end of a semi-infinite ccnductor. Therefore, the pr-,!ary edI-e waves may be found from the rr'-;Llem of scatterln. of a plane wave hy the semi-infinite conductor
:'9)
"(30.03)
where
.,g
, Cosi.
"
( 30.04 ) may be
Figure 73. The incidence of a plane wave on a thin cylindrical conductor; he is the incident
angle.
case, shin
it
Z j
(30.05)
However, hencefcrth it will not be necessary to have the rgorous expression for the function . Let us note that Equations (30.03) and (30.04) are similar to Expressions (6.13) for a strip. These latter expressions do not take into account secondary diffraction.
FTD-iHC-23-259-71
from
the enrvl Z
z2 A
; ex-
( 30.06)
from the function.; ,, of all, (z)
0
where by ,(z) we mean the functions obtained by replacing, l by 8.. it is For the purpose of calculating7 the deslred necessary for us, when applied first to find that to an infinite conductor on it.; section induced (zzc)
"secondary waves,
,
(30.06)
For this
in
an infinite
S (z
Wltn
2<
Z,
(30.07)
(IT m0 < 0). in conductor
Let us assume that w0 has a We may reg, ard the quantity accordance with equation
concentrated
emf which,
(28.03), 4_a
(30. 08)
In
I ht, f (hitd
by tite oxte'rniI
- 0'
__..
'L.0
6 .. z-
+ .
++'..
d,
in
z
in
) and
p_ 0(z
-")
2 of [82].
As- a result,
FTI-llC-2 3-259-71
187
CEo,
.l'.l'.e
l-
J(z)---- _
&h 'Ba' Ihi
ju 0(z - Z,)
--
I t 1*0cam-Y
(30.10)
Thus,
it
excites,
in
addition to the wave *p_, also the wave 0. In order to excite a "pure" i0 wave, it is necessary obviously to apply an additional external field
E:=8,(z-z.).
(30.11)
such that
(a'.-00.
(30.12)
Hence,
2ik sin,
it
Is
still
in'
21
,,i .7ka
which determines the quantity
cos, .CI2
,, I n,,,i -
( 30.14 )
FTD-HC-23-259-71
Yka co-21
I. t
h ,in'- 1
(30.16)
~L'
I0with :.zr:~
J,
<(30.17)
excites the following current in an infinite single-wire lne (with z <z 2 ) -SI.. '/'' :"'
(30.18)
Now let us study the diffraction of current waves (30.06) by the semi-infinite conductor (-a, z 2 ). For this purpose, let us use the Lorentz lemma [4i]
is the current of the auxiliary dipole with the located at point 1 with the coordinates (R, 8); H
FTD-HC-23-259-71
189
is its field on the conductor surface, where the external currents 2 are specified; E2 is the field created by these currents at point 1 (Figure 74). A The external current J 2 is determined by the well-known equation
4.=Figure 74. in
(30.20)
we have
jeE
terms of its
field
J = _ H1,4,
induced by the dipole in the conductor,
(30. 24)
we obtain from the Lorentz
:,
If field the dipole p1 is radiated by it
1(Zj dz.
(30.25)
zo, then the z0 of
FTD-HC-23-259-71
190
(-%
z2)
as a plane wave.
Then the
by the equation
"Jtz) -S
where
Ie" -e'+"'+ j
(z, - z)
("1
(30.26)
.S'-21
2k' sin' A In 'ka-sa
c,. .= --
/kcos#.
(30.27)
We will select the quantity z0 in such a way that, at a distance z2 - z 0 from the conductor end, the reflected current wave would be practically equal to zero (++(z,.-z#):O) . Substituting the function (30.26) into the right-hand member of Equality (30.25) and taking for the quantity Ee the external field (30.16), we obtain z
-a s
-
I-k
,jj(Z,-8) dz
-
XT.!.
2s= b
--
SA
ie Ie'
(30.l28)
An important feature of this relationship is that the integration is performed here not along the entire conductor (-o, z2), but only along part of it (-c, z 1 ), where the function -(zs--z) describes the current with good precision. The integrals here are calculated th,. same as in Equation (30.09). As a result, the field radiated by the
semi-infinite conductor (--, E2o # H2; z 2 ) will equal
I In
S2",,1o
.-
- 2i,
kAa son O
-#--X
eihR
FTD-IIC-23-259-71
191
c'~~~~ie y, l eI
sO
I
sin
---
, 1
U.*+r.
i,(COSC-S(')
4 (L)-In---..-.-
- +!A)11 .
iwz
(30.29)
The terms
correspond to
the desired secondary wave dlverging from the conductor's end z = z 2 . Using Equations the form this wave may be represented in
(30.30)
~r(z,) =---
4i
xsI
j 2 sn !t si,
*8 e . oil
In
~Ihcos--,
__
2 '14.4
-ya
()30.31)
necessary to investi)7ate the diffraction of primary wave (30.18) at In this case, of the semi-infinilte conductor (z,-'z-^,). the end = z the prInciple of' duality lead.- to the foelowin,- relationship
which,
(30.17)
3 )
/N
In
it
gives us 0).
lhe field
radiated by the
(zl,
2'(z. P.'
W2
IF
,(z,)
2i 2slAIn nh7sinS
(30.34)
where
4ijEe~8h i rz .- U
X -sin* sin*
(Cos#z)= +1 Co&tXt
in .(cost+ co,) In 1 ka sin 0, In . Sha sh2 (L)-
* Cos - In 2 O.2T 2
(30.35)
Otherwise, Equations
this
in
E--.
(30.30)
(30.31),
z 2 by zl,
&
by
x-0
and 8,
by
(30.34)
waves diverging from the ends of the semi-infinite conductors and (zl, -). If one excites an infinite single-wire external field
line by the
E
where
'
(z -
(31.01)
k, snAsin
then a spherical (30.30). emf wave arises which with of an infinite 0-a
(31.02)
coincides with wave line by the concentr'ated
FTD-HC-23-259-71
193
E' s=$2a(z--Z,).
where
(31.03)
(' 140
(31.04)
a wave arises which coincides with wave (30.34) when 0-a0 . It is not difficult to see that these external fields actually excite in an infinite single-wire line current waves which are equivalent to the secondary current waves in a passive vibrator [that is, equivalent to those,.waves which are expressed by the first terms in the brackets of
Equation (28.18)]. Therefore, the tertiary waves may be investigated as edge waves radiated by the semi-infinite conductors (zl, -) and (-OD, z 2 ) with their excitation by the external fields (31.01) and (31.03), respectively. From Equations (30.25) and (30.32), we find without difficulty the total field radiated with the indicated excitation by the conductor (zj, )
= E , .=
?
21 hI .i 3 # T
elk *
-R1e
-
(31.05)
(-i,
z2 )
SID jk4a1
(31.06)
As a result, we obtain for the tertiary waves diverging from the ends z and z the following expressions:
(31.07)
s,,0
- A
Y,,,
'R
e,
(31.08)
.7.1
' (z,)
In the directions
_ 2).
,"(L)e
"
(31.09)
toward the opposite end of the conductor, thue waves are equivalent to the radiation of an infinite slnle-wire line excited by the external fields
z(') S(S)(31.11)
(31.10)
Consequently, the quaternary waves again may be investigated as edge waves radiated by the semi-infinite conductors (--, z,) and (zl, -) with their excitation by the external fields (31.10) Using the reciprccity principle, we easily obtain
z,) ?() ()( 1 i'me-mZ
*i
and
(31.11).
-.
''(z,) l-,)=
2sin&In -fia-j R
jS
.
z2i
where
2s
In l On
R(1.)
fL,
. 3) (31.1
a 1 2H zin
Ec,) (z.)
C. (2 2sin 0 t
2)
In sl.
1 )
are found.
Here
FTD-HC-23-2',9-7I
19
-c~'".4
1(31-15)
Thus, the field arising with multiple diffractions (starting with the second) may be represented in the following form:
SR
e-'ae+ La(50)-(z)
=;
where
(31.17)
'.("'(,,: ;',)-m
,==
=L (z. + (L)(at,)~
n,2
tL
(31.19)
3-8
--'"
(L) e-"L
')(z,.3)
and 61
(30.35), (31.02) and (31.04). We will not write out here the rather unwieldy final expression for this field, but we will proceed with a calculation of the total field scattered by a vibrator. 1 32. Total Fringing Field
Before beginning the derivntion of the expression for the scattering characteristic, let us make the following observation.
The
FTD-HC-23-259-71
196
functions
(30.05),
satisfy relationships
our Investigation the conductor's
However,
of the successive waves arising with diffraction at ends was approximate. Expression (30.04) Therefore, it
mation expressions
* In
7ka sin -,- sin -Y
(32.01)
which were obtained by the variational method and have a precision which is sufficient for our purpose (see [833). More precisely speakIng, we will use approximation Equations (32.01) in conjunction with the rigorous Expression (30.05), and we will set
S
"("~~ ~ - cu jo
In i-Fj~" 2i In-. #- In
I"ka sin
-si
V.._
I
I
(32.02)
ik-
In " kasinX1
.E
X)
"In
ka2 sin2siu--J
(32.03)
Now summing Expressions (31.17) and (32.03), we find the total field scattered by a passive vibrator in the form
FTD-HC-23-259-71
197
":
. ( , )
( 3 2 .04 )
=
Cos In
,
21
In
iX
,0cce
-7-
csoss
0.)
s-in''--sin" -.
_t COS.COS_L
I r-"[
2
-in'
2OTO
+
T '-l-.s '
i
'
1
cos T _2a
] i,.,
,
i (Z.2C.os#+Z,
In 2 0 Coacott
(cos *.+ cos !,)I
-
0.
Ca.os
(Co I+CSO'okaco
ika ";*L ,O 0
A cosOO
o,__
20
+e-W.i 2 Cos
(32.05)
In which all the functions 0 and iP_ have the argument L. The resulting expression, despite its complexity, has a clear physical meaning. Actually, the first term in the braces corresponds to the primary edge wave radiated by the conductor's end z = zl; the second term corresponds to the primary wave radiated by the conductor's end z = z2. The terms included in the first pair of brackets refer to the secondary wave departing from the end z = zl, and the terms in the second set of brackets refer to the secondary wave departing from the end z = z 2 . The remaining terms describe the sum of all subsequent waveo arising with multiple diffraction Pnd have a resonance character. The resonance begins with L = z2 - z % n*X/2(n = 1, 2, 3...) when D % 0. FTD-HC-23-259-71
2C1 "-
is
does not
0 and
1o.
One may
also show that the vibrator does not radiate in irradiation, that is, F(O, Furthermore, istic (32.05) in )=F(%. o)-= Ft, ')F('
7)-==0.
(32.06)
using representation
(28.25)
and ,, we obtain the following expression for scattering characterthe direction of the mirror-reflected
k. 2i 21n pt $t
4-) +("!'
'
(2kL
COS'
.-2i-
4-)
+1
( it11 ++
2
iIn \,
Vt asin,,
+.(_ -Osin' $.
.oe"L
*t
(32.07) or
it
+ kkj)
a) L
""
'-*a
L1
!In L--s
i--' "I
t.,LJ,
(32.08)
which characterizes irradiation. Let us also write the expressions corresponds the reflected field magnitude with normal
Jirections coincide
(0
8o)
F(4, 0)=
"1'I
=sni
C(4o+ In ne
n "2"
c-oi-In-<i
"2 -1 a s
-, e
e- 208, Cos 0_
"(2 !.i'
In
'
sin -2.ykz
L)
"
-Cos, -, -2-1n n 2 ka cos
,+(L) e-',','
Ii
-2
c2og F#I . p(L) (L) e-" 1 0 cns iyk (L). (L) e+ ---Inj
I )"
(32.09)
O=f
Equation (32.09)
leads to Expression
The scattering characteristic (32.05) which was found above w3a obtained by summing all the waves formed with multiple diffraction.
Such a method is at the same result Ing with tertiary very graphic, but somewhat it is lengthy. One may arrive startmore quickly if assumed That the edge wave
FTD-HC-23-259-71
V Ii'r~iIop.
Ttie'cCriom
wtvh
paosvo vlhai'
ie:1"*A
MAr.V
I211a
meI~ o~'t 11,1a *n 1W %ti a Ii ka,
~wi P.
2*I'
sill" siin,
COO~
k COinn,
rsil,
a % ,Pl , In
2 **
katS2J
ykha sin
2 C40
In Y Cos
1
(L) t
, -
to,*.,.'P
Th
l .t w
l~
il.
FIltlii
.11
U1M
ov
tf:
'itli
t~tI
MIT
II
II
C,-
'kt-
'. - i n e-"'r
' 1.
.(32.14)
S 33.
A Vibrator Which iL
the Wavelength (a
J,,e
a thin cylindrical
vibrator which is discussed In this chapter is based on a number of One good aspect of this theory is the fact physical considerations.
that its precision increases with the length of the vibrator, since the current waves whose diffraction we are investigating are expressed the longer the vibrator. However, one may also show small in comparison
more clearly,
It
E# , IfI,
where the dipole moment p,
static problem. ohapt' of the vibrator'. In
(33.01 )
may be calculated by solving the electroaccordance with [92], the dipole moment of
p,--= D(1)
- ..
(]
. 2
a dmLnelns :o
on lss
ftunetion
I.
s.hown :alk-1l,
In
With
i,
one
n:ny
toh"
D(1)-T
2 In
'11
!0
If
In this
term,
then
D(I) =
The results shown in of numerical Figure calculations based on Equations we see that Z
,
(33.04)
are
the latter
equation gives a
9.
Pat
I I
I-
"
'~
A LA
A I
Figure 75.
Graph of the function D(Z) which determines the cylinder's dipole moment.
Thus,
Is
short
In
comlpari-
-n
"L
and its
scattering characteristic
form
FTD-HC-23-259-71
-24
XL _ 7
(33.06)
&ai
-Yi
two terms of the expansion of
the vibrator's scattering characteristic F in reciprocal powers of the large parameter L/a (with A - -), and we compare them with Expression (33.06). We shall limit ourselves to the case & described by the simpler Equation (32.08). and the when the function F is
) g-z-.(
em)
(33.07)
g(Z)-g(0)In and
+ eS-4
li,
(33.08)
g- d,, g(0)=In
(33.09)
corresponding terms of the aysmptotic expansion for the functions * and q,, which may be obtained from the initial determine these functions (see, for example,
and
, we have
FTD-HC-23-259-71
20l4
and
In -j.
In addition.% terms off the order (in ik-) -2are omitted In Expressions (33.10) and (33.11). Now if, we substitute these expressions into Equation (32.08), then one should omit terms of the order (n j~-~i it. Theefoethe function F'jf -may
I
-
C2LI
*L
(2I0
'
7*4tik
+i~4+..eh
~2(L)InIe. L
+ e(33.12) # M
i~~ 21
7kL
In
yk
++ + -L eihl=
W
(33.14)
-2 a
I
4,~
+iks~(~Ls
*LIn
(33-15)
IeL14
--
, it
2 24
n2\
(33.1
(33.16)
I-- 4
" + -
+(33"17) 2+0
the form
-,,,. h L'. - )
2 Vi
--
It
This result confirms the correctness of scattering characteristic (32.05) calculated by us, and shows that it is applicable for vibrators of any length. 5 34. The Results of Numerical Calculations
The function F(Oo Oo) enables one to calculate the integral scattering thickness S and the effective scattering area a of a passive vibrator. The integral scattering thickness is determined by the relationship s = ".
(34.01)
FTD-HC-23-259-71
20f
is
(34.03)
the value of the energy scattered by the vibrator into the surroundSince one may represent the fringby the equations the far zone in the direction O=-n-
F( -0,.,
.8
(34.06)
performed by us
for vibrators with the parameter Z=j1 taking the values X = -0.05 and X = -0.1, are found to be in agreement with the results of Leontovich and Levin [85]. With X = -0.1, our curve (the dotted line in Figure 76) is only slightly displaced in the direction of longer wavelengths and gives slightly higher resonance peaks. The effective scattering area a according to the definition equals
(0, 0.)= P,.4-.R p
(34.08)
where p is
and
FTD-HC-23-259-71
207
-I
wae.Cre
wer
cacltdbyIotv
au
E1
k2
3(3
09
Ii tness Slength
Figure 76.
S~and
of (with a vibrator a function ot' of' its normal as incidence a plane wave). Curves 1 were calculated by Leontovich Levin [85]. Curves 2 were calculated on the basis of Equation ' (314.07)
(3J
represents the average value of' the energy flux density scattered by
FTD-HC-23-259-7I
208
the receiving antenna operates with the same polarization as the transpitting antenna, then the corresponding value of the effectile area will equal If 7. -(-( ,# c ',. 'F(8,0')1' Ii (34.11)
In the cabe Of radar whep, the transmitting and receiving antennas are
.Equations the.basis-of
irradiation 42.)
illustrates the dependence of the function 6 on the quantity kL with a given value of 0,=21a!L-= 15, - that is, when the ratio of the
In Figure 78 vibrator's length to *itsdiameter equals L/2a = 452. of the freas a function is constructed 6 the graph of the function
quency f = c/X'10-6 (in megahertz) for the prescribed parameters Here the curves plotted by Lindroth [79] are L a 5 cm and Op = 15. drawn with a continuous line, and the curve in Figure 77 calculated by Van Vleck et al. [863 is traced by the dash-dot-curve.
1-
The curves of Lindroth and Van Vleck were calculated by integrating the current which is found as a result of the approximate However, this procedure was persolution of the integral equation. Lindroth obtained an formed in [79] and [86] in a different way. expression for the fringing field in the form of an expansion in reciprocal powers of the parameter 0 . The expression includes terms of the order of 9 -3. In [86] a different kind of approximation was used which led, as can be seen from Figure 77, to rather rough results especially in the resonance region. Our curve (the dotted area) agrees almost everywhere within the limits of graphical precisicn with A noticeable divergence is observed only in the curve of Lindroth. the magnitude of the first resonance peak.
FTD-HC-23-259-71
2oq
IM ----
II
II
41
Figure 78. The effective scattering area of a vibrator as a function of the frequency f - c/A.*lO 6 (in megahertz) with normal incidence of a plane wave. The designations are the same as those in Figure 77.
R1 In Figure 79 and 80 radar diagrams are constructed for vibrators of a length L a 0.5A and Figure 77. The effective scatter- L = 2A with the specified value ing area of a vibrator as a L/a a 900. Curves 1 were calcufunction of its length with normal incidence of a plane wave. lated by Tal using the variational Curve 1 was calculated by method [87]. Curves 3 were obLindroth [79]; curve 2 wastandbthmeodfidud calculated by Van Vleck [86] by tmdb h ehdo nue means of the method of integral ernf [86]; curves 14were obtained equations; curve 3 was calcubthaovinctemtodf lated on the basis of Equation b h bv-niae ehdo (34.12). Van Vleck. The results of calculations based on our Equations (3L4.12), (32.08) and CI2.09)-ae shown by curves 2.
FTD-HC-23-259-71
Figure 79.
A comparison of the diagrams for the effective scattering area of a half-wave vibrator calculated by various methods.
Curve 2 was calculated on the basis of Equation (34.12); Curve 3 was calculated by the method of induced emf (in the work of Van Vleck
[86]);
Curve 4 was calculated by Van Vleck [86] by the method of integral equations.
In the variational current, various approximation methods wereused. method [87] a functional was constructed for this purpose which was Then the current stationary in respect to small current variations. was sought in stants. the form of some function containing undetermined conThis method enables one to rather easily These constants were found from the condition of the func-
tional's stationarity.
FTD-HC-23-259-71
211,
III
IIIIv
hU
.. .-
U!
I'-
Figure 80. Diagrams for the effective scattering area of a vibrator calculated by various methods. The designations are the same as those in Figure 79: L = 2A
obtain the first approximation; however, its results, especially for long conductors, may depend in a substantial way on the form of the trial function. In the induced emf method [86], the current is sought in the form of a combination of trigonometric functions with unknown coefficients. These coefficients are determined by using the law of conservation of energy. This is the simplest method, but it has a number of serious defects. Thus, as a consequence of incorrectly accounting for the current component having the incident field phase,
FTD-HC-23-259-71
it leads to inaccurate results in the case of odd resonances (cspecially for long conductors), and it does not give the displacement of the resonance peaks from the values X = 2L/n (n = 1, 3, 5 .... ) in the direction of longer wavelengths. The results obtained by us are also approximate. However, our Equation (32.05) satisfies the reciprocity principle, and is applicable for any length vibrator. For very short vibrators L << X, it changes into the asymptotic expression for the scattering characteristic of a dipole (see 33). For vibrators with a length of several wavelengths (L nX, n - 1, 2, 3, 4), Equation (32.05) gives satisfactory results. Calculations performed on its basis for radar reflection with normal irradiation agree with the results of Lindroth. Good agreement is also observed with the results of Leontovich and Levin for the integral scattering characteristic. With an increase of the vibrator's length, the precision of this equation increases, and in this wiy it is favorably distinguished from the equations proposed for the scattering characteristics by other authors. Moreover, the divergence between the various approximation methods indicates the necessity of performing rather detailed calculations based on precise methods, for the purpose of evaluating the actual error of the approximation methods. Such calculations may be performed, for example, by means of the method discussed in References [88, 89] or [91].
FTD-HC-23-259-71
213
FOOTNOTE
Let us note that one may refine Equation (28.04) by multiplying its righthand member by the factor 0 (usually 00 1) calculated in Reference [84].
FTD-HC-23-259-71.
MISNC1 W,1 I ON
Pi~ii'etiO,1
lilt, f ie 1.1 or:~' :1 shavltl oiii
on
t'i:t
t lit, Ouvav
with ai ,i
been'i
p('tLri'd
!)avL a
IrraPdi ationu
t'
I e.
'
roundiL
Itar1
p1l.-I
t~:i alll
I prl0i I
ty
vi-zln~.'
i;'I,
alId
nt~ l'ot
iti:11
va 1euI 1 ltlo1,loi
Mor It'evucompete
onelitpt.
:pplictab~l) ity
t'11s eiimit'y
oipt.ic
,lvtl,
It.
vOlt' I''81
18 8
app'r'oach
onily
i' IV th k eN~
GuIo'! opn lt. Us
l,
I d i IIICn81 018
ie
Ii L'oilIpal'I i IIi~'
"I
tcivt41"
I e~gt I.
1I8 basedi
phyical
vonl8 Ih'8onl
~~~un I orin
the lit'
vvii
irill n
itv
Include part1 or the~ curr'?enit, :uiili does no1:1te ot II I'eII' parit: e I' the ourron'enl whIichI II8en, of thv bvi'o1
In Ihv~
a
L'a1eli
1 :11. 101
OI
I I the thelt,'v
i ho iiotiiii Il'opi'i
I lit
Thl'l':e l'ot'c'
Whein
1ii,
OC
dl1101 nil
part oV
on.' ari~ :e
lithe
00n:; I ul'ab 1y
waivei
J~iit
c'urrent. 0VOUlp
,
Omni!i I
Itotl
I1 I, fact.~
T1hiei'eI'ol't
onie woulld
10II
I V ui 'Ii'i''W011 110
I k1 I,0
Bput,'81''
In actuaIlt
8I~t:
:iiO ou :fl i
iothek vv on
Myehork
phyIil
o; ti
t'iri'it'i
vstI
I lt
ltttolin
l 'iitd only~
l Ilu
bodye '
or2
n diivI I'
wI v \'2
IliI
''
't't'
11'.'i'',
ph'
Iri d Ino n
par 1111
t. in'. I ii
11I 'l'lii:I.
c In n
donitt'tltei
I'vt:
lt'
i':iI "Il
titI
rr'o
ore Ikhl.-
ourv
nt
1000110 su stan
ial
Th
ve~v
y:-.caloptjo.-
quIII
the current
becomes noticeable
also in
accord-
incidence on
a cone along its axis (0 17) serves as a clear example of how important the above-indicated factors are. Although in this case the nonuniform part of the current, concentrated near the cone's vertex, has
practically no influence on the scattering, nevertheless, the physical optical approach gives values for the radar thickness which are tens of decibels smaller than the experimental values, even with large the nonuniform base
flowing in
rim of the conical surface; the nonuniform part of the current has an especially large vnlue for sharply pointed cones. Another interesting example of a of a plane wave by physical a finite similar nature is the scattering 18) where the
paraboloid of rotation (
to be a periodic function of the paraboloid length, lengths reality. The investigation of the diffraction (Chapter V) that for flat it
plates one may limit one~self to consideraif their linear dimensions are one-
tion of secondary
diffraction,
which would possess physical visualizifor mak ngr caLculations. equatiolns. which Tn
FTP-HC-23-2,39-71
21
Our purpose was not to calculate the current on the body's surface, the field in the near zone, or the integral scatterlnf thick.:ness. These questions are investigated in the physical theory of diffraction 25. In them, in particular, a number of other works based on which were already the first enumerated in
sions in
powers of A/a were obtained for the integral thickness which the total power scattered by a body. However, in these equations are missing for the scattering c.haracterand observabook and the indicated works
as a rule,
ohich are valid with any direction of irradiation Therefore, the results of this
At present,
problems have
Chapter VI an experimental method was discussed which enabled ono in a "pure form", and to measure, the field from the non-
to isolate
uniform part of the current excited by a plane wave on a metal body of any shape. is found in or, In the same chapter, it was shown that the well-known of the
phenomenon of depolarization of the wave reflected from a body which free space is in produced by the nonuniform part by the surface distortion. in Chapter VII for the problem of' current, other words,
The investigation carried out diffraction by a thin, a In natural development multiple diffraction finite
and completion of the method of considerini, the of edge waves which wa.- applied in
Chapter VII equations were derived for the scattering diaproin which lengtth with any irrailit.iton
The results
obtained in theory,
this book show the fruitfulness and enable one to arrive Such problems theory
of
problems. of diffraction
FTD-H1C-2 3-259)- 71
.1
As an example of such a problem, one may point to the first class. problem of diffraction of a plane wave by a frustum of a cone or by an inflnitely long cylinder with a polygonal transverse cross section. Those problems whose solution requires obtaining (using the methods of mathematical diffraction theory) a whole series of new results must be referred to the second class. In particular, in order to give a complete solution to the diffraction problem of a finite cone, it is necessary to have more precise knowledge on the diffraction laws of a semi-infinite cone.
Summing up, one may say that physical diffraction theory aids one in analyzing and sorting out the diffraction phenomena for complex bodies, poses problems for treatment by mathematical diffraction theory, and enables one to effectively apply the rigorous results of mathematical diffraction theory for the solution of new problems. In conclusion, this book is I express my deep thanks to L. A. Vaynshteyn for
his valuable advice and regular discussion of the questions to which devoted, and also for his attentive reading of the
manuscript and for a number of useful remarks. I also take this opportunity to express sincere thanks to M. L. Levin for his interest
in this work and his helpful remarks.
FTD-HC-23-259-71
REFERENCES
I c,k If. A. Oro r111,,u',%e aC1bI~ *~lvoy.1 Ha c.iy11n ofIIOI0350l.11"nof 1opm~u. ',.". il 1 J..1'" 0.1 .!ml "ilu. t"1 :3,,.,)X:I4t ir&, YK (Dl, NI 19A0 2.Aareiimiu Cn:?7I:0eTj'oM:4.% oc.1Ht, 1. llep.c amm.i Iiio p'. 5j. IIL (1 oai I. a. 1131.1( Is.) C(v-C7:GCe -. 1;10C,. 1930. 3 M1c iimr '1;4; P. Jaip;1;awluA it jp.icemenii paxiinsn.ui. Hep.
1958. 4. B a A. iout' .# .1 A ., 3.r ',re: mol*auu. Iba.)u -(o PHer,,Kc~ pea^1i~ 1957. 5. 11o recx ;I I( A. If l.1Hom iiaviampaKuaHm iaer.TpO(ilarlIlfTtA IX mL1031. 113.1-110 03C's-icoe j~:la;oo,198 B a %cr H. H. .Ind C o p)s'ri E. T. The isailhematic-ml theory of Mlyg.rv,~~ p1rinciple. OX!ord, 1939. 7. a p t I I. eBm 1. $1 !1p1Ii`,.vmimme4:A )ac11v.r ippaxuitim n.1o11. 1. lutm';,aiuitl Ila
Ml.ie 3i 3tare. Ac 10,) 27, .i 8, 18:0-1819, 1957. I n.. P!. [1Pih6.1 miJ4CIIt p~3cl.CT ;tioopatuitim na.10 AdXj ZX,l01fc~arjII:T:AIX 00.11 W3 ieXOTOP!Ax VICTa.1.lH14eeIX Tc~aX. it. II . Amp:t amtitsi Hla utce it xmiefl~C'mio Witimitupe. )I(To. 28. Mt 11.
8. s~t itmci
2601---2616. 1958. 10. Ytttomuci i e II. H1. Broll~umam Anitpattiimits ga lI.CKe. KTOb. 2S. N! 3. 383-591. 1938. 11. Yd) HMINIIXC ii. 51. C:I~I%.N~rpMI.II~o06.i%4CHuIie KOI#CqHU.X Ten
3;; 1?VI~. 'tIjailiortvmlug;a -i "-.1cK13.1tpu a . 6. .Y,?4. 539-567.
1961.
%WnrIMTHAN 90.111 KPytosoA nolstpmumtzI or 'wera.i.nccitax 1t'.l. qPa.a11o0Te.\IIIa it ?.e~pn a.. N.4 12. :923-1928. 1960. 13. Y ip if -! L e It 11. 51. 06 oTp)a:;c(t:,; mera.11o'wecH',H muiC.3t paaloaoxifi Kpyo03oii 110.up41113au1,il. 4pa3xIoiCXHIIca It 3.1CXrPoHIIKa,,
14. Y cp it Ntu CS m. AI. .1124I3KfIll ,l.iocxifx m.exrp)o~ta.liur~mx 30.111 l13 T0IIK0~'.IUI.11.I',IIUPI1CCKO!I rpOOi)(HII1)KC. 4P3.ZHQTCXHHX3 14 3'.1CK-
?POI;;IK.1, " ,At -2.2'0-l-6. V462) IS. S c h w a r z sch lIfIId K. Betigung find Polarisation, des Lichts. 3lathemiatisdch Annalen. 5.5,177, 1:62.
16. if) p a tifI 0. If M if 3ec P. M1Iiciepeviuma.1hmuc Hwe yj1Nulmemsfa mIat4maml'llKOA 1PH31,1(i. OHTH. 1937.
N llHTez-pa.L.
17. 0 o K B. A. Pacnp.ifeamexme TOK0as, 8036yNC.12eimx nnacitol B&II~OA Ha3 fOBePXHOC7H npoBO.1HHKIf. WSITOP. 15, Ne 12, 693--702. 194&. 18. 0 o K B. A. ITlf,)2aKlllf P32110o0.l11 ao~pyr 30MIiioll nonepXHO-
C1.M.-,I. 113.1-50 AH CCCP. 1946. 19. B a Al IM T YeA Hl Jif A. .Iit4ppaxitim 1ae)(TPO'armlitHimx it 3xyimolue Bo.11l050.a. lbtii-o aCo~ercxoe paXolmax 910.11fH3 OI'arPUTOMi
.1110. 1933. F. On tmu, diffraction anti reflection on 10. Oh b r hte t t i n gt waxes and pulses by wedges an1( corners. .Iourn. Research NBS. 61,
11.AI., 1949.
Nm.Th.1mljx
tVlijxuti.
t acrb
nep-
22. P a It I i NV. On asymiptotic series for fiimetions ill the theory o! hlff.raction of light. Ph)McaluI Rc\ kw, 54, V?~ II. 9)21--931, 1938. 23. 31a K-.1ax-.i a , if. B. Teopio, it ;i;'lmi.~emie l(1YHKUHI 3lame. It. JI.. 1953. Strenge Tle,,.rie tier Beu2. 1. liei \ it e r .1. and Anilrejews~i WV. truflg rbener cel.droi'agnetik,-hr Wellen -in der %Uolls('iflimen !etende" hn voilkournen Krri sche~l'c uid an der hicisjifrmigmi (.)fniun Ii-itenden ebjitei Schi~rn. Amiialn %der LPwsik. 7, Nes3-4, 157, 3930. 23. A id reje wsk i W. Die 13eniung ' vltromiarn-tisclser ud anmdcr k-rtis.~rmigen IWellen an der leitenden Kreissclcheib Oi'nnng iii leitendcrn cbemen Schirin Zeits. ffit aingutt.,ndte Physilt. S. V? 5. 178, 1953.
FTD-HC-2 3-259-71
M.tlathivii,.-he Funtlio. 2. eI~ i x it e. r J.. miid S c It I I I,e' F. %%' rien utid Sph~roiditunhitz':tvt. Sli inger'Verlag. 1953. 27. .1a:t ac 6e1)r I' c. Oltr~iKJ. r~i rr.l. 195 -28. Br a u n b c k W. Neur, N51:4r~m~ri:,t"3 Vi UeIir die Beugung
710.B r a uP b e ItW. Zur Regui aii d--r kreis~rinigen Offmnung Zeits. O~r Physik. 138. 1.i80 Ss. 1954. 33. 0 r a unb t k IV. Zuir fkuiquiit antOffi~tngen its nichlebenew, Schirmen. Zeits. far Phy~ik., 156. 1.65, 3959 3-2. F r a h n W. F. B.ougung e! kironwii.iietischer tWcllen in Braunbe'.schcr N5lwcrund. 1. Zoils. far I'ht%ik. IS6, 1. 78, 1959. 33. F ra hn W.E. Beiugung ck!.:romagnetischer tWellen in Braunbek-schter NMheruug. It. Zei:s. luir Pltys.. 156. Xt Z 99. 1959. 34. Sc.iuc NHa m. r. .amcppaitnita me1h rpom armHTrh1X BOau Sa incxe. B c6. c.:114paKUtim i.1exrpo.tar111fTHb1X 110.111 Hit eCKOTOPUaZ reaa31 11Paui1el'tiw11. 143a-90 'COaerCiKOe P3.11101. 1937. (crp. 148-174).
KopoTICaao.l11osaa aCI'NMTOT11I11 ANIOPSKt1NOl' a teHit IUCa.tbttO nposo~inuteru napa6n.iliwecioro uNAmNApa. aPs~noroxminua Nt ACKwpommIap. 5. Xv 3, 393-4002, 1960.
"!oro ric.13
35. H 31 m s B. It.
36. Hem tamo B. H. .IH.4paitmnw xopormxti n.oito~xt 3aexTpovarMFtHWXd. 30.?ix "a rzaaxomo timnyr-iyKm uuianupe itipH HaKJiONNom nail'. sunt. aPaamoYCxmttttt of e.1ckrOHtfi21, 5, Xt 3, i24-528, 1960. 37. raop it it Ho A. C. Armwnram'te:Koe ptctinHC 3s.2aii e AsOpaitutim n.ioc,~oA xixtrpomarmTtnnoil mooinu ma nposo~asuem uniLnJIe3. PaanaTeXrNuS u43.leK~rpottnica. 3. Xv 5, 603-614, 1938.
38. F e I en L, B. Backscattering from wide-angle and narrowangle cones Jourit. AppI. Phys. 26. Alt 3, 138-151. 195&. 39. Siegel K. M, Crispin J IV. and Schensted C. E Hiectromagrncti., and acotistical scalterlnq from a semi-infinite co" .lnurn. Afpp. Phys.. 26. Xv 3. 309-313,. 195&. 40. rap a nutta A. C. ,Iitiprax:ix n.iocxoA ,.exTpomarltHTNoI
110.71tM. PaC11pOCTp24ttfOu1ICAC
0.10.b
OCK lN0HYCZ.
ePaSAMexrMIM
AP-4. Xt 3, 312--321, 1956. 43. K c II e r J. B. Diffraction by an aperture, 1. Journal of App?. Phys. 26. NO 4, 4.16 -144., 0-57. Dif!racti-ic by an aperture. It. Jaumt AppI. Phys. 28. Y'f5. 570 -579. 1957. AP-8. .V. 2. 375- -182. 1950, Backsaltering from a finite cone-comparison of theory and experiment, Trans. IRE. AP-9. Xt 4, 413, 1961. 45. Sc hens ted C: E. Ekcrt:ninagndfir and byv t wrii.infinite hohy of revo~ution. Jnvrit. Appi.acoustic Phys., scattering 26. M~ 3 Ibfi-avt 3955
43. Siegel K. Mi., Goodrich R. F.. W~eston V. H. Comnnhents on far firlid scatterfriq I.-Pn boudies of revolution. Appl. Set. Res., Scclion- B. vo~ume 8, S. 8 1959. 12. K e I I e r J. B. Di. fract on by a convex cylinder. Trans. IRE.
Ne 5, 3957.
TrauisacCons IRE, Antennas atid Propagation. AP-4, X7*3, 28228,1956. 47. M ill a r R. F. An tapproximnate theory of the diffraction of an electromnagnetic wave by an aperliire in a plane screen. Proc. Inst. Electrical Enges., Part C. 103, Xf 3. G77. 3956. 11. MW IlIl a r R. F. The dif~ractioii fi' an electromagnetic wave h% a circular aperture. Proc. Inst. Elez.!rical Engrs., Part C. 104.
in diffraction Theory.
49. M ill1 a r R. F. The iliffiractioti of .in electromagnetic wave by a large alitrture. Proc. Inst. Efectr*.a1 Enigrs., Part C. 304. A?* 6. 1957. .50. If a it (Ilb u clIt der Ph~sik. Bd 2.5':. Kristafllptilc. Beugun. HI. Fllnl. A. W'. M~aue, K. Wtstpfahl. Theorle Berlin. Springer. 11961. 4fer tBcugung (S. 238 -5192). 5I. Ka rp S. N. and R u ss ek A. Di-!ractlon by a wide slit. Journal of Applicd Physics, 27. Ne 8, 18Q6-891, 1396. 52. Bu cIta I R. N. and K ellIe r .1. B. Boundary Layer Problems M 1, in Diffraction Theory. Conimun. Pure aid AppI. Mathem., 133. 145-114,. 396.
FTD-fHC-23-259-7122
~.ij.
-
rpi~i6eprj
i
nex~po~mvitr~ritta
oa;.uueeIe.3.144 r. A. llum~ah '111 C osroamritinfi !Ipi;%iofIIIWI111a fl.OCMOCTH 807111 4(TO. 27, Xt I I. 2.595-npo6.iem. cA po.IcTCIIIubI
54. F p H H6 e p rF. A. Metal petucemis imp~paxamoifitwl2 33aaq 0a Hy=4& n.IOCXiX 113C3.11-110 flpOBOlI$IlIX 3Kpaido3. ocHOnanuuA14 3.1111 iialOIHlblX US iKjpa14a TeHC~BIX TOKOU. ATO, 28, Xt 3. 542-4 4914W 003-468, 1958.
m.ewrpomarWH1THUZ BOAR U s53.r p itH6 ep r r. A. 0 awtopanicH nwo.1ce xoWC4iioAi oitiputi. IAH CCCP. 129. Mi 2, 295-298, 1959. 10. B. K Bolipocy 0 anelH He0o8 56. rpHii6epr F. A. Hfn
ta ea.iwiff npo~nolRtueR 1.'IocxoH 3.1CKrpo~iarHTHWuX 8.WifU (OpM1K,1H erx c iipyr.lb4m orsepCTme!Ii. Xi( T. 0. 29. Xt 10. 1206-1211, 1959. 3. A. K monp)ocy cr r. A. If N 0.1 C !l :19 0oit 57. r,)i i IfW
31. *.+TtV /1 ,1h S.M I Il a r R. F. 1'toc. Carrhbridge 3tr~p. 521, 1959. 39. K -I Ifr R. if N
1. 13- 71 . 1961. V9, Dif'ratJ.-a ov a w~tde s':-, ;ri! -:.mp:emciitary PhIa:h.ISolety. 51. 1. 179-406. 497T aif - itI y :! b. Paccevm
I~A.:t:$paxiv 3.-,0%-
TeHM C::a11l aCI4 E. 0 (IONOao 1:3.71. H i0 e p 5B. 0i. RKI *PJ-1oc:1a1T031N1 !I 1cwrp0HIIxas.V?1, 5153-58, 1961. pJC`O.1O0A0MfbI.%Il4 (1.1113KO 61. K ii :15 c :' B. E. Pa.,rm,%:t: NlCWaV 6 .\. 6, W~N.4~'Iia:a,'e;I113\II4. P321IOnTVwIa" if ~.toac~ 407-9!6, 1961. i. E. 0 1wmppahu3s %ier;)r)%armvrfHiL4 80.11 Hil 62. K it a5f
Pur:ivroft IIoBC?,vo..!it Kpp.:o:;oru uuminllpa. B.K3.6. N; s. 1273-1283. 1961. 4Pa.1Is'neCXnaxa
H 1.iex.
6. Ne 10, . riiuo6 hopvo~cp..rit 4epu. Nan1:oTextuma it mv.OIioimxa, 462---1657. 1961. spaweimi. OfI)'~~1h irij:)a~oiniaoH1 6I. 0 nK B. A. TeommI
a ilvx.fl0'tix TC.13X ami '.pUWlsro;1llrlIx4111 L60,YfaI:i Hpa11;0i:T~i:i) 11.1a.. O 41COnRCTKOC jlaatOV. 19.37. 3.1CK.o It o I A. A .11t.4.p3KUila 111.10,:KOi 65. (P o ic B. A. it 0 ca
6i6.
PO K B: A. Ioze or
C;TT~l:xa.bimnro if
a~I3.31lOlflno-
06 03lHOA1 WITOle -eMOIS 3:131 (aif "A.1.b nO ;*.x\iO%: 3CM.CKp).ltHI l.\ i! 4) pac:ipo~rpa:ivra .1'.Ila. .%If CCCP. Cep. 133. 8. 16-22, 1914. o B. A. PCI;~:011.' 132134:1 0 PaC.M. A. it 00 To s:1 68. .1 CIitr W"1C.'~il3~.1ii no MCXT;1T0%iar:mTIrIJX mmi. W30.%. IpoCTPa!Ivl!I:II Ae-o.1v[iyra;60).l::ecCloro \a~ waicin. )K3Til. 16, N?' 7. 53o7-573. 1946. .ao 8';'OIt i p'a vlxio 69. (Do B. A. Tcop,.3 pacopi~p0wrpieml IIo'4'euZa. 1139 .1II CCCP, 14. ioxi:o-5a '30 *ele .1.u11 H10n::lo.1HSr0t ia. AH CCCI', cc?. (plil3 10, ,i2, 171 -85. 1916. ,11111M C .13 11.3a 71. B a A IIITC C14 .1. A, di, e aup oa A. A. Pa:z -:e ;ii.oCehix
.1l4~)KWO'~lX .1v1%111. Xi1. 70--94. 1950. 1.1 ,0 O0K B. A_ Iloc .-MCKol 30.1Illa 46.1il31t rio3C've%!fln41~0
31--46, 1961.
* II. . ~~~~~:Wlyca.
~avliotexnlapia
it
3.T~1:3,6.
.\t 1
Na9ah114 T'V.1l1mI~le-me
.~e
:,y1T1I3T31. .0
-Pa.1,IiCTtXIIQ. If 01CKTP(H;I tl3s, 6, ";i 8, 1217-12,541. 1961. 7.3.BaAHifiITerfliH :1.A.ma.1Kciwihic'4 r. d1. 210v:1pC'IH3
.1lt4)P:!KU1'I 3. *C44"I'11:j K0CJaxOml 1o.'C P3.a;l~ya. '4. If. opalliiiIaaX. .Pa1-alewXHlloa it -e.lCT1Q~~0;.Xa-. 6, Xti 9. 1 W - i193, 14.61 . 74. 0 C.1 opos A. A. A IWmi-iutiviechoe peI:I.I.e 33:1 111 0 2W11)m023umeb ti1W a 54211,-io npo rigooA opal:Uan n.IOt-KOQ 9.10pI~omarm~mo
FTD-HC-23-,2'>)-71
75. C h yt i
~~p.Pa.11sotl:%vnNj
it ..
B Th,
iqdvpdirsation
1961. 76. Chyiyl 1D. Depoliroi~.tivii by randomly spaced scalterers. PrAr'e IRE-ZTS.V, Vj 20, Praha, 1961. 7 .7 . C. h yt(Il B. Polair'baor depoide'it scattering cross-sections. Price URE-CfSAV. X 21, Praha. 1961. M$ B r~c k in a v n P i 1-t'lvar;s,!ion of electromagnetic waves by inclined p!.1n:-. P;.ce CRE -C'iAV, .V 19, Praha, 1981. 79. 1. 1n d r o *h K. lRiI"ctioii of elemm~magnetic waves f.,om thin mdii str!ps. rran-. Roy, bisi. of Techncol., Stockholm, Sweden, .91. 1955. bO Hfall ~n r. Tabo-,tc.11 iii~.stigations Into the transmitting 30~d rvc,1.i:ig quaiitics of aJmleiIraat. Nova Acta Roy. Soc. Sci. Upsala
3. Xt 12, 1451-1 162, 1958 of elect roia gnetic waves bodwiS Price C!RE.--CSAV. .\a 17, Praha.
r~ll~
81. B aA 11 r e IN 1 .1. .%~ B,.vnu roia ii ,oaxo~, aimamastapucxiome Ilposo.'jH"Ke I. TOAc It )IN11C.'.ajic sI,'a~~eoPItipa~opa. XIT0, 29. M' 6. 673-.688, 1959. 82. B a ri itu w c fi 11 .'1. X. 80-13W& W~KSh TOIIOM WI.1IIH3;~cCKoM II;.o:oaiI:nXC. 11. Tu' it na:cwwu:,s~ ssi paru'e is m3.1ysieJ1e nepceaawOtero uzI(.paropa. )"U0. 29. X. r), 689-0D9, l~59 8.3. B a Aisus r e ii .1. A'. Boai:su roIxa a Toi5Ko~t LiI.5~jimmpi'5cxom iipo8('vifiise. Ill. B3P:137w11;s';1,-i u'ro~a it cro npistscwifeisss x TeoptmM w.jeaamhsoro ii ji-me.iamsvwo lipovoj0. )I.T40. 31, Xy 1, 29-44, 1961, St. B a Aii iusT ei,, .1. A 130~s,13 103a a TOHNOI U1I1ANAHK1p4McOM MV Bxo.iiol wini'eastc !!Atparopa it rn-nmocti. oopmiyA. 3 os2itKe. )KO1. xi 1, 45-50, 195). 83. J1 eoifro ato n 1 14 .1s11 .1 Cei 9"Is M. A.~ X~ leOmflhI!1 1101yXueisluR No-wC"amlA a isutry~aroi,%a~ iteii-s. 4AT4' 14, \'v. 9, 481-506, 2914. 8G. Van V'lec J. H. 'Blo~ch F. 'and liamermesh M. Theory o! radar WclcC~oi Irr).n v.iris rir '.hn motailic strips. Journ. Appi. Pligs, 18, Mi 3, 271---294, il4.17 S7. Ta I C. T. Clrtn.ic,0,g:,tlc b;-ch-scattcrling from cylindrical %vires.Jotirn. App!. FPh%. 23. YXi M. 903--916, 1952. 83. K an toua n1. 'a., Doox B. A is Bia PsHuwrefi Hj. A A, cssM'1erpaw14;e X0.,ca3p1iq .1
ui4euioA
sq. ro
:.1iiusub.
a)Kpus~i N4~c~lr~i~nu ~uilId7em Ilan if MIare.%far'ejC~oA *OIfHKCAV. A 4. 604--679. 1961. DO.If a I I i n E. Exact Solution of th~e Antenna Equation. Transactions of tlue Poyal linstitute ofrehooy tch~,Se den, Nr 133, 1961.TenooyStcomwe 91. D un ca n R. It. and H-lI n ch c y F. A. Cvl'ndrfcal Winenna Theory, .1.Research Nat, Bur. Standards 64D. 569, 190W. 92. B a rt Hs tu Te A it .i. .A. CTarl,'uexifse 3saaa'sp zaix no.ioro sus*
I,
maxR fllpotor
)KT'* 29i. Xt. 10, 11698Z1205. 1959. sop yi is1. If. ':.c1IcPUMMcHs I rr~epa.mbuoro ypas'sepo~ia tI.- f.lOtHoX-iI rcaca j a~urcis'e-
nnoun10,111aro
FTD-HC-23-?59-71
SYMBOL LIST
Russian
Typed
SC
Meanng
cylindrical
disk
FTD-HC-23-259-7I
2?