Plane wave diffraction by two oppositely placed,

parallel two-part planes

’I.H. Tayyar and A. Buy
. ukaksoy
.

Abstract: The problem of plane wave diffraction by two oppositely placed, parallel two-part
planes, consisting of the junction of perfectly conducing and resistive half-planes, is solved with the
Fourier transform technique. The problem is formulated as two simultaneous Wiener–Hopf
equations that are uncoupled by the analytical properties of the functions that occur. The solution
involves an infinite number of expansion coefficients satisfying an infinite system of linear algebraic
equations. The expansion coefficients are determined numerically, and some graphical results
showing the influence of the surface resistivity and the distances between the two-part planes on the
diffracted and transmitted waves are presented.

1 Introduction A time factor e
iot, with o being the angular frequency,

is assumed and suppressed throughout the paper.
In the present work, a Wiener–Hopf matrix equation
connected with a new canonical diffraction problem is 2 Analysis
solved explicitly. We consider the diffraction of Ez-polarised
electromagnetic plane waves by two oppositely placed, Let the incident wave be an Ez-polarised plane wave, the
parallel, perfectly conducting half-planes with semi-infinite electric field of which is given by
resistive sheet extensions. This configuration can be used for
the purpose of electromagnetic shielding and of optimum Ezi ¼ ui ðx; yÞ ¼ e
ik½x cos f0 þy sin f0  ð1Þ
coupling between incident and transmitted waves. The where k denotes the wave number of the space, and f0
special case where the resistive sheet extensions are absent denotes the angle of incidence. The problem consists of
was treated in [1, 2, 4–6]. studying the scattering of this incident wave by two
The aim of this work is to reveal the influence of the oppositely placed, parallel two-part planes defined by
resistive sheet extensions on the transmitted fields. The S 1 ¼ S 11 [ S 12 , with S 11 ¼ fðx; y; zÞ; x 2 ð
1; 0Þ; y ¼
representation of the solution to the boundary-value d; z 2 ð
1; 1Þg; S 12 ¼ fðx; y; zÞ; x 2 ð
1; 0Þ; y ¼ 0;
problem in terms of Fourier integrals leads to a Wiener– z 2ð
1; 1Þg; and S 2 ¼ S 21 [ S 22 , with S 21 ¼ fðx; y; zÞ;
Hopf matrix equation, the solution of which requires a x 2 ð0; 1Þ; y ¼ d; z 2 ð
1; 1Þg; and S 22 ¼fðx; y; zÞ; x 2
square matrix to be split into a product of two matrices with ð0; 1Þ; y ¼ 0; z 2 ð
1; 1Þg, respectively, as depicted in
non-vanishing determinants, such that the entries of these Fig. 1.
matrices as well as their inverses are regular and of algebraic We suppose that the half-planes S11 and S22 are perfect
growth in certain overlapping halves of the complex plane. electric conductors, and the half-planes S12 and S21 are
Owing to the non-commutativity of the matrix multi- partially transparent thin dielectric sheets characterised by
plication, no general method exists to achieve the Wiener– resistive-type boundary conditions. A thin dielectric sheet
Hopf factorisation of an arbitrary square matrix. The kernel with thickness kt  1, dielectric permittivity e and con-
matrix encountered in this work is in a form to which the ductivity s is modelled by [7]
‘weak factorisation method’ introduced by Abrahams [1]
½~ ~þ¼0
n  E ð2aÞ
and Idemen [3] is applicable. This method reduces the

solution of the related Wiener–Hopf matrix equation to the
determination of an infinite number of expansion coeffi- ½~ ~ þ ¼
1~
n  H n  ð~ ~
n  EÞ ð2bÞ


cients satisfying infinite systems of linear algebraic equa- R
tions. The effects of the spacing between the two parallel,
two-part planes and the surface resistivities on the
diffraction phenomenon are shown graphically.

r The Institution of Engineering and Technology 2006

IEE Proceedings online no. 20050094
doi:10.1049/ip-smt:20050094
Paper first received 29th November 2005 and in final revised form 20th March
2006
’
I.H. Tayyar is with the Department of Electronics Engineering, Gebze Institute
of Technology, 41400, Gebze, Kocaeli, Turkey
A. B.uy.ukaksoy is with the Department of Mathematics, Gebze Institute of
Technology, 41400, Gebze, Kocaeli, Turkey
E-mail: [email protected] Fig. 1 Geometry of the problem

168 IEE Proc.-Sci. Meas. Technol., Vol. 153, No. 4, July 2006

Authorized licensed use limited to: ULAKBIM UASL - GEBZE INST OF TECH LIBRARY. Downloaded on March 23, 2009 at 05:49 from IEEE Xplore. Restrictions apply.
 with R being the surface resistivity, given by account the following radiation and edge conditions
i [8, 9]:
h  
R¼ si ð2cÞ
pffiffiffi @u pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
2ot ðe
e0 Þ þ i r
iku ! 0 r ¼ x2 þ y 2 ! 1 ð5gÞ
o @r
þ
In the above relationships, ½ 
denotes the discontinuity
across the transmissive sheet, and ~ n is the unit normal u ¼ Oðr1=2 Þ r ! 0 ð5hÞ
vector directed to the side indicated by the + sign. The By inserting (4a–c) into (5a–f ) and then inverting the
resistivities of the half-planes S12 and S21, are denoted by
pffiffiffiffiffiffiffiffiffiffiffi resulting integral equations, we obtain
z1;2 ¼ R1;2 =Z0 , with Z0 ¼ m0 =e0 being the characteristic
impedance of the free space. AðaÞ ¼ BðaÞeiKd þ CðaÞe
iKd ð6aÞ
For analysis purposes, it is convenient to express the total
field as follows: AðaÞ ¼
þ
1 ðaÞ ð6bÞ
8
< ui þ ur þ u1 y4d
uT ðx; yÞ ¼ u2 y 2 ð0; dÞ ð3aÞ iAðaÞ½KðaÞ
k=z1 
iKðaÞ½BðaÞeiKd
CðaÞe
iKd 
:
u3 yo0
2ik sin f0 e
ikd sin f0
Here, ur denotes the field that would be reflected if the ¼
þ


1 ðaÞ ð6cÞ
2piða
k cos f0 Þ
whole plane y ¼ d was perfectly conducting
ur ðx; yÞ ¼
expf
ik½x cos f0
ðy
2dÞ sin f0 g ð3bÞ BðaÞ þ CðaÞ ¼ DðaÞ ð6dÞ
In (2a), un ðx; yÞ ðn ¼ 1; 2; 3Þ satisfy the Helmholtz equation
Dun þ k 2 un ¼ 0 n ¼ 1; 2; 3 DðaÞ ¼


2 ðaÞ ð6eÞ
in the regions where they are defined. For the sake of and
mathematical convenience, we assume that k has a small
positive imaginary part. The results corresponding to media DðaÞ½KðaÞ
k=z2  þ KðaÞ½BðaÞ
CðaÞe
iKd 
with negligibly small conductivity can be obtained by 2ik sin f0 e
ikd sin f0
making Imk ! þ0 at the end of the analysis. ¼
þ


1 ðaÞ ð6f Þ
2piða
k cos f0 Þ
It is appropriate to use the following Fourier integral
representations: with
Z Z 1
u1 ¼ AðaÞeiKðaÞðy
dÞ
iax da ð4aÞ 1

þ
1 ðaÞ ¼ u1 ðx; dÞeiax dx ð6gÞ
L 2p 0
Z
Z  
u2 ¼ ½BðaÞeiKðaÞy þ CðaÞe
iKðaÞy e
iax da ð4bÞ 1 0 @ @ ik
L

1 ðaÞ ¼ u1 ðx; dÞ
u2 ðx; dÞ
u1 ðx; dÞ eiax dx
2p
1 @y @y z1
and
Z ð6hÞ
u3 ¼ DðaÞe
iKðaÞy
iax da ð4cÞ Z
L 0
1
Here, K(a) denotes the square-root function


2 ðaÞ ¼ u3 ðx; 0Þeiax dx ð6iÞ
2p
1
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
KðaÞ ¼ k 2
a2 ð4dÞ and
which is defined in the complex a-plane, cut along a ¼ k to Z 1 
1 @ @ ik
a ¼ k þ i1 and a ¼
k to a ¼
k
i1, such that
þ
2 ðaÞ ¼ u2 ðx; 0Þ
u3 ðx; 0Þ
u3 ðx; 0Þ eiax dx
Kð0Þ ¼ k. The integration line L is a straight line lying in 2p 0 @y @y z2
the strip Imð
kÞoImðaÞoImðk cos f0 Þ. ð6jÞ
The spectral coefficients AðaÞ, BðaÞ, CðaÞ and DðaÞ are to In the region x o 0, where the sources are absent, the field
be determined through the following boundary and u1 behaves like a cylindrical wave that emanates from the
continuity relationships: origin and propagates towards infinity with the wave
u1 ðx; dÞ ¼ u2 ðx; dÞ x 2 ð
1; 1Þ ð5aÞ number k (radiation condition):

ikx 
u1 ðx; dÞ ¼ 0 xo0 ð5bÞ e
u1 ðx; yÞ ¼ O pffiffiffiffiffiffi x !
1 ð7aÞ
@ @ @
x
ui ðx; dÞ þ ur ðx; dÞ þ u1 ðx; dÞ In the region x 4 0, where the sources are present, relying
@y @y @y
upon the boundary condition (5c), it is an easy matter to
@ ik

u2 ðx; dÞ ¼ u1 ðx; dÞ x40 ð5cÞ show that the asymptotic behaviour of u1 is the same as that
@y z1 of the incident field, namely
u2 ðx; 0Þ ¼ u3 ðx; 0Þ; x 2 ð
1; 1Þ ð5dÞ
u1 ðx; yÞ ¼ Oðe
ikx cos f0 Þ x ! 1 ð7bÞ
u3 ðx; 0Þ ¼ 0 x40 ð5eÞ If we consider (7a) and (7b), we observe that

þ

1;2 ðaÞ and
1;2 ðaÞ are regular functions of a in the
@ @ ik
u2 ðx; 0Þ
u3 ðx; 0Þ ¼ u3 ðx; 0Þ xo0 ð5f Þ half-planes ImðaÞ4Imðk cos f0 Þ and ImðaÞoImðkÞ,
@y @y z2 respectively.
To obtain a unique solution to the mixed boundary value The elimination of A(a), B(a), C(a) and D(a) between
problem stated by (3) and (5a–f ), we have to take into (6a) and (6f ) yields the following Wiener–Hopf matrix

IEE Proc.-Sci. Meas. Technol., Vol. 153, No. 4, July 2006 169

Authorized licensed use limited to: ULAKBIM UASL - GEBZE INST OF TECH LIBRARY. Downloaded on March 23, 2009 at 05:49 from IEEE Xplore. Restrictions apply.
 equations: with
2 3 2 3
1 1 8 9
0 " þ # 1

 > k >
6 MðaÞN1 ðaÞ 7
1 ðaÞ 6 MðaÞ 7 i
1 ðaÞ >
> i$ d
2
sin ð$dÞ2 >
>
6 7
6 7 >
>d z1;2 >
>
4 1 5 þ 4 1 5 i

>
> 2 3>>

1

2 ðaÞ
0 2 ðaÞ >
>
2 >
>
Z 1>
>p k >
>
MðaÞ MðaÞN2 ðaÞ < 2
6 $
2 sin ð$dÞ
2
7=
  q1;2 ðaÞ ¼ PV 6 z1;2 7
k sin f0 e
ikd sin f0 1 0 >
> ip6
6
7>
7>
¼ ð8aÞ >
>
> 4 þ i2$ k sinð$dÞ cosð$dÞ 5> >
>
ipða
k cos f0 Þ 0 >
> 2 >
>
>
> z1;2 >
>
with >
> >
>
: ;
sin½KðaÞd
MðaÞ ¼ ð8bÞ  
KðaÞ Kð$Þ þ a
 ln d$ ð10bÞ
and Kð$Þ
a
KðaÞ As to the explicit expression for M þ ðaÞ appearing in (9b), it
N1;2 ðaÞ ¼ is given in [11] as
½k=z1;2
KðaÞ sin½KðaÞd
iKðaÞ cos½KðaÞd
ð8cÞ       
sinðkdÞ 1=2 iad 2p p
M þ ðaÞ ¼ exp 1
C
ln þi
k p kd 2
Y1    
2.1 Solution of simultaneous Wiener–Hopf a iab
 1þ exp ð11aÞ
equations m¼1
a m mp
As is well known, the first and most important step in
solving a matrix functional equation of the Wiener–Hopf and
type is the factorisation of the kernel matrix. To this end, let
us multiply (8a) from the left by the following matrix: M
ðaÞ ¼ M þ ð
aÞ

N1;2 þ
ðaÞ ¼ N1;2 ð
aÞ ð11bÞ


L1 ðaÞ 0 In (11a), C is the Euler constant given by C ¼ 0:57721 . . ..
ð9aÞ
Lþ After the Wiener–Hopf decomposition have been
0 2 ðaÞ
performed, the matrix equation given by (9b) can be rear-
to obtain ranged as
" #" #
1=Lþ
1 ðaÞ 0
þ
1 ðaÞ
þ
1 ðaÞ L
ðk cos f0 Þk sin f0 e
ikd sin f0 L
ðaÞ

þ
1 ¼
i 1
ðaÞ
Lþ
2 ðaÞ=MðaÞ
Lþ
2 ðaÞ
þ
2 ðaÞ L1 ðaÞ ipða
k cos f0 Þ MðaÞ 2
" #" #
L


1 ðaÞ
L1 ðaÞ=MðaÞ i


1 ðaÞ k sin f0 e
ikd sin f0

þ iL


1 ðaÞ
1 ðaÞ þ
1=L
i

ipða
k cos f0 Þ
0 2 ðaÞ 2 ðaÞ
 
k sin f0 e
ikd sin f0
1  ½L


1 ðaÞ
L1 ðk cos f0 Þ ð12aÞ
¼ L ðaÞ ð9bÞ
ipða
k cos f0 Þ 1 0 and
Here, Lþ
1;2 ðaÞ
denotes certain functions that regular and



2 ðaÞ iLþ ðaÞ þ
free of zeros in the half-planes ImðaÞ4Imð
kÞ and ¼
2
ðaÞ þ iLþ þ
2 ðaÞ
2 ðaÞ ð12bÞ
ImðaÞoImðkÞ, respectively, resulting from the Wiener–


L2 ðaÞ MðaÞ 1
Hopf factorisation of The regularity of the right-hand side of (12a) in the lower
half-plane can be violated by the simple poles occurring at
L1;2 ðaÞ ¼ MðaÞN1;2 ðaÞ ð9cÞ the zeros of MðaÞ ¼ sin½KðaÞd=KðaÞ lying in the lower
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
as half-plane, namely at a ¼
am ¼
k 1
ðmp=kdÞ2 ,
m ¼ 1; 2; . . . , and the regularity of the right-hand side of
L1;2 ðaÞ ¼ Lþ

1;2 ðaÞL1;2 ðaÞ (12b) in the upper half-plane can be violated by the simple
þ
¼ N1;2 ðaÞM þ ðaÞN1;2


ðaÞM
ðaÞ ð9dÞ poles occurring at the zeros ofqMðaÞ lying in the upper half-
ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
The explicit expression of N1;2
ðaÞ can be obtained by plane, namely at a ¼ am ¼ k 1
ðmp=kdÞ2 , m ¼ 1; 2; . . ..
P P1
following the procedure outlined in [10] If the infinite series of poles 1m¼1 am =ða þ am Þ and m¼1
bm =ða
am Þ are subtracted from both sides of (12a) and
" #1=2 (12b), respectively, we obtain
þ KðaÞ
N1;2 ðaÞ¼  
k=z1;2
KðaÞ sin½KðaÞd
iKðaÞ cos½KðaÞd
þ L

ikd sin f0 X1
1 ðaÞ 1 ðk cos f0 Þk sin f0 e am
 1=4    þ
2
kþa KðaÞd a þ iK ðaÞ L1 ðaÞ ipða
k cos f0 Þ m¼1
a þ am
 exp ln
k
a p k L
X1
1 ðaÞ
am
¼
i
2 ðaÞ þ iL


1 ðaÞ
1 ðaÞ

   MðaÞ a þ am
iad 2a m¼1

0:5iKðaÞd þ 0:5q1;2 ðaÞ exp ln
p k k sin f0 e
ikd sin f0

þ ½L ðaÞ
L

1 ðk cos f0 Þ ð13aÞ
ð10aÞ ipða
k cos f0 Þ 1

170 IEE Proc.-Sci. Meas. Technol., Vol. 153, No. 4, July 2006

Authorized licensed use limited to: ULAKBIM UASL - GEBZE INST OF TECH LIBRARY. Downloaded on March 23, 2009 at 05:49 from IEEE Xplore. Restrictions apply.
 and a ¼
k cos t, x ¼ r cos f and y
d ¼ r sin f, with
f 2 ð0; pÞ. The saddle point occurs at t ¼ f, whose


X1
iLþ contribution is
2 ðaÞ bm þ þ 2 ðaÞ þ 

¼ iL ðaÞ
ðaÞ

1 ðaÞ
L
ðaÞ a
a m
2 2
MðaÞ eip=4 2k sin f0 sin fe
ikd sin f0 L


1 ðk cos fÞL1 ðk cos f0 Þ
2 m¼1 u1 ðr; fÞ¼ pffiffiffiffiffiffi
X
1 2p ðcos f þ cos f0 Þ
bm X1

ð13bÞ am eikr
a
am þ 2pik sin fL

1 ðk cos fÞ pffiffiffiffiffi
m¼1
m¼1 m
a
k cos f kr
To cancel out the unwanted poles, the expansion coeffi- ð16Þ
cients am and bm should satisfy the following equations:
Similarly, if we use the change of variables a ¼
k cos t,
x ¼ r cos f and y ¼ r sin f, with f 2 ð0; pÞ, the saddle
Lþ

1 ðam Þ
2 ð
am Þ
am ¼
i m ¼ 1; 2; . . . ð13cÞ point of the integrand occurs at t ¼ 2p
f, with
_
Mð
a mÞ f 2 ð0; pÞ, and the asymptotic evaluation of the transmitted
and field (4c) gives
pffiffiffiffiffiffi
u3 ðr; fÞ ¼
2pe
ip=4 k sin fLþ2 ðk cos fÞ
Lþ þ
2 ðam Þ
1 ðam Þ X1
bm ¼
i m ¼ 1; 2; . . . ð13dÞ bm eikr
_ mÞ
Mða  pffiffiffiffiffi ð17Þ
m¼1 m
a þ k cos f kr
In (13b) and (13c), the dot (  ) stands for the derivation with
respect to a. The application of the analytical continuation
principle together with the Liouville theorem to the
equations (13a) and (13b) produces 4 Computational results

X1 In this Section, some numerical results concerning the

þ
1 ðaÞ L

1 ðk cos f0 Þk sin f0 e

ikd sin f0
am
¼ þ variation of the amplitude
pffiffiffiffiffi of the diffracted and transmitted
Lþ
1 ðaÞ ipða
k cos f 0 Þ m¼1
a þ am fields ð20 log ju1;3  krjÞ against the observation angle f
ð14aÞ are presented.
All the numerical results were derived by truncation of
the infinite series and the infinite systems of linear algebraic
L
1 ðaÞ
k sin f0 e
ikd sin f0 equations after the first N terms. Figure 2 shows the
iL


1 ðaÞ
1 ðaÞ
i
2 ðaÞ ¼
variation of the modulus of the diffracted field against
MðaÞ ipða
k cos f0 Þ
X1 the truncation number N. It is seen that the amplitude of
am
 ½L


1 ðaÞ
L1 ðk cos f0 Þ þ ð14bÞ the diffracted field becomes insensitive to the increase of the
m¼1
a þ am truncation number after N ¼ 3.
In Figs. 3 and 4, the variation of the diffracted and


X
1
2 ðaÞ bm transmitted fields for different values of surface resistance z1

¼ ð14cÞ
L2 ðaÞ m¼1 a
am and z2 are presented. It is shown that when the surface
resistances increase up to certain critical values, and then
and they start to diminish. It is also observed that the curves
iLþ X1
2 ðaÞ þ bm corresponding to the transmitted field amplitude are shifted
iLþ þ
2 ðaÞ
2 ðaÞ

1 ðaÞ ¼ ð14dÞ to the left as the separation distance increases.
MðaÞ m¼1
a
am
As an example of the influence of the geometrical
The solution of the simultaneous Wiener–Hopf equations parameter on the diffraction phenomenon, Figs. 5 and 6
has now been obtained in terms of infinite series of illustrate the variation of the diffracted and transmitted
constants am and bm. By using (13c), (13d), (14a) and (14c)
we can show that these constants are to be solved through
the solution of the following two infinite sets of linear
algebraic equations:
Lþ ðam ÞLþ X1
2 ðam Þ bn
am ¼ i 1 ð15aÞ
_
Mð
a Þ a þ am
m n¼1 n

and
Lþ ðam ÞLþ
2 ðam Þ
bm ¼
i 1
_
Mðam Þ
" #
L1 ðk cos f0 Þk sin f0 e
ikd sin f0 X

1
an
 þ
ipðam
k cos f0 Þ a þ am
n¼1 n
ð15bÞ

3 Analysis of the fields

Let us consider the integral representation of the diffracted

field u1 in (4a). To perform the asymptotic evaluation of this
integral, let us make the following change of variables Fig. 2 Truncation number N against the diffracted field amplitude

IEE Proc.-Sci. Meas. Technol., Vol. 153, No. 4, July 2006 171

Authorized licensed use limited to: ULAKBIM UASL - GEBZE INST OF TECH LIBRARY. Downloaded on March 23, 2009 at 05:49 from IEEE Xplore. Restrictions apply.
 Fig. 5 Diffracted field amplitude against observation angle for
Fig. 3 Diffracted field amplitude against z1 for different values of d different values of distance d between two-part planes

Fig. 4 Transmitted field amplitude against z2 for different Fig. 6 Transmitted field amplitude against observation angle for
values of d different values of distance d between two-part planes

Wiener–Hopf equations and then solved rigorously with the

fields with the separation distance d between the two-part Wiener–Hopf technique.
planes. The transmitted field amplitude increases with For the special case, where z1 ¼ z2 ¼ 1, L1 ðaÞ and
increasing values of the separation distance d. L2 ðaÞ reduce to
It can be checked easily that the effects of z1 and z2 on the
sin KðaÞd iKðaÞd
transmitted and diffracted fields, respectively, are very L1 ðaÞ ¼ L2 ðaÞ ¼ i e ¼ iLðaÞ ð18Þ
weak, as expected. KðaÞ
and the diffracted and transmitted fields in the regions y4d
and yo0 read
5 Concluding remarks
 i3p=4
e 2k sin f0 sin fe
ikd sin f0 L
ðk cos fÞL
ðk cos f0 Þ
In the present work, we consider the diffraction of high- u1 ðr; fÞ¼ pffiffiffiffiffiffi
2p ðcos f þ cos f0 Þ
frequency Ez-polarised plane waves by two, oppositely
placed, parallel two-part planes, one half of which is pffiffiffiffiffiffi X
1
am eikr
perfectly conducting, whereas the other part is partially þ 2pe
ip=4 k sin fL
ðk cos fÞ pffiffiffiffiffi
a
k cos f
m¼1 m
kr
transmissive. To obtain the explicit expression of the
diffracted field, the problem is first reduced to two coupled ð19Þ

172 IEE Proc.-Sci. Meas. Technol., Vol. 153, No. 4, July 2006

Authorized licensed use limited to: ULAKBIM UASL - GEBZE INST OF TECH LIBRARY. Downloaded on March 23, 2009 at 05:49 from IEEE Xplore. Restrictions apply.
 In this case, the two infinite sets of linear algebraic
equations in (15a) and (15b) are reduced to

½Lþ ðam Þ X
2 1
bn
am ¼
ð23aÞ
Mð
am Þ n¼1 an þ am

and
" #
½Lþ ðam Þ2 L

ikd sin f0 X1
1 ðk cos f0 Þk sin f0 e an
bm ¼ þ
Mðam Þ pðam
k cos f0 Þ a þ am
n¼1 n

ð23bÞ

Finally, Fig. 7 depicts the transmitted field amplitude for

the special case, where the resistive sheet extensions are
absent ðz1 ¼ z2 ¼ 1Þ.

6 References
1 Abrahams, I.D.: ‘Scattering of sound by two parallel semi-infinite
screens’, Wave Motion, 1987, 9, pp. 289–300
Fig. 7 Transmitted field amplitude against observation angle for 2 Abrahams, I.D., and Wickham, W.G.: ‘The scattering of water waves
different values of d by two semi-infinite opposed vertical walls’, Wave Motion, 1991, 14,
pp. 145–168
3 Idemen, M.: ‘A new method to obtain exact solutions of vector
Wiener–Hopf equations’, Z. Angew. Math. Mech. (ZAMM), 1979,
59, pp. 656–658
and 4 B.uy.ukaksoy, A., Topsakal, E., and Idemen, M.: ‘Plane wave
pffiffiffiffiffiffi
ip=4 diffraction by a pair of parallel soft and hard overlapped half-planes’,
u3 ðr; fÞ ¼
2pe k sin fLþ ðk cos fÞ Wave Motion, 1994, 20, pp. 273–282
5 B.uy.ukaksoy, A., and Serbest, A.H.: ‘Matrix Wiener-Hopf factoriza-
X1
bm eikr tion methods and applications to some diffraction problems’, in
 pffiffiffiffiffi ð20Þ Hashimoto, Idemen, Tretyakov (Eds.): ‘Analytical and numerical
m¼1 m
a þ k cos f kr techniques in electromagnetic wave theory’ (Science House, Tokyo,
Japan, 1993, chap. 6)
respectively. The explicit expression of L ðaÞ is given 6 B.uy.ukaksoy, A., Uzg.oren, G., and Birbir, F.: ‘The scattering of a
plane wave by two parallel semi-infinite over-lapping screens with
in [11] as dielectric loading’, Wave Motion, 2001, 34, pp. 375–389
pffiffiffi Kd aþiK iad 2p p
 7 Senior, T.B.A., and Volakis, J.L.: ‘Sheet simulation of a thin dielectric
Lþ ðaÞ ¼ d e½ p lnð k Þ e p 1
Cþln kd þi2 layer’, Radio Sci., 1987, 22, pp. 1261–1272
2sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 3 8 Idemen, M.: ‘Confluent edge conditions for the electromagnetic wave
Y1  2 at the edge of a wedge bounded by material sheets’, Wave Motion,
4 1
kd ad ad 2000, 32, pp. 37–55

i 5ei np ð21Þ 9 Braver, I.M., Fridberg, P.SH., Garb, K.L., and Yakover, I.M.: ‘The
n¼1
np np behavior of the electromagnetic field near the edge of a resistive half-
plane’, IEEE Trans. Antennas Propagat., 1988, AP-36, pp. 1760–1768
10 Mittra, R., and Lee, S.W.: ‘Analytical techniques in the theory of
L
ðaÞ ¼ Lþ ð
aÞ ð22Þ guided waves’ (Macmillan, New York, 1971)
11 Noble, B.: ‘Methods based on the Wiener–Hopf technique’ (Pergamon
with C ¼ 0:577215 being the Euler constant. Press, New York, 1958)

IEE Proc.-Sci. Meas. Technol., Vol. 153, No. 4, July 2006 173

Authorized licensed use limited to: ULAKBIM UASL - GEBZE INST OF TECH LIBRARY. Downloaded on March 23, 2009 at 05:49 from IEEE Xplore. Restrictions apply.