Escolar Documentos
Profissional Documentos
Cultura Documentos
Department of Electrical & Computer Engineering Rutgers University, Piscataway, NJ 08854 orfanidi@ece.rutgers.edu www.ece.rutgers.edu/~orfanidi/ewa
Department of Applied Electronics, University "Roma Tre" via della Vasca Navale, 84 00146, Rome, Italy davide.ramaccia@gmail.com alessandro.toscano@gmail.com
Exercises Chapter 1
Table of Contents
Chapter1 Maxwell's Equations ................................................................................... 1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 Exercise.......................................................................................................... 1 Exercise.......................................................................................................... 6 Exercise........................................................................................................ 12 Exercise........................................................................................................ 16 Exercise........................................................................................................ 18 Exercise........................................................................................................ 20 Exercise........................................................................................................ 25 Exercise........................................................................................................ 29 Exercise........................................................................................................ 30
D.RamacciaandA.Toscano
Exercises Chapter 1
(1.1.1)
x U V = Ux Vx
y Uy Vy
z Uz = Vz
(1.1.2)
= x U y Vz U z Vy y ( U x Vz U z Vx ) + z U x Vy U y Vx
Now we can prove the algebra identities with simply mathematical substitutions:
)
)) =
(1.1.3)
A ( B C ) = A x B y C z Bz C y y ( B x C z Bz C x ) + z B x C y B y C x =
) ( ( ) y ( A x ( B x C y B y C x ) A z ( B y C z Bz C y ) ) + z ( A x ( B x C z Bz C x ) A y ( B y C z Bz C y ) )
x A y B x C y B y C x + A z ( B x C z Bz C x )
((
( ) + y ( A x B y C x A x B x C y + A z B y C z A z Bz C y ) + z ( A x Bz C x A x B x C z A y B y C z + A y B z C y )
+ x A y B x C y A y B y C x + A z B x C z A z Bz C x
A ( B C ) =
(1.1.4)
Let us write eq. (1.1.4) in matrix form, separating the terms with the minus sign and the terms with the plus sign:
D.RamacciaandA.Toscano
Pag.1
Exercises Chapter 1
C yA x Bx 0 C y A z Bz Cz A x Bx C z A y B y (1.1.5) 0
Note that the elements of the diagonal of each matrix are zero. Each term can be filled with the product of the three component with the same subscript ( a ii = A i B i C i ) :
Bx AxCx ByAxCx BzAxCx Cx Ax Bx CyAx Bx CzAx Bx A ( BC) = Bx AyCy ByAyCy BzAyCy Cx AyBy CyAyBy CzAyBy = Bx AzCz ByAzCz BzAzCz Cx AzBz CyAzBz CzAzBz
( ) ( ) ( ) Ax Bx ( Cx x + Cyy + Czz ) AyBy ( Cx x + Cyy + Czz ) AzBz ( Cx x + Cyy + Czz ) = (1.1.6) = B ( AxCx + AyCy + AzCz ) C( Ax Bx + AyBy + AzBz ) =
= +AxCx Bx x + Byy + Bzz + AyCy Bx x + Byy + Bzz + AzCz Bx x + Byy + Bzz = B ( A C) C( A B)
( A x B y C z A x Bz C y ) ( A y B x C z A y Bz C x ) + ( A z B x C y A z B y C x ) = ( A x B y C z + A y Bz C x + A z B x C y ) ( A x Bz C y + A y B x C z + A z B y C x )
B (C A) = B x C y A z C z A y y ( C x A z C z A x ) + z A x C y A y C x
((
)) =
(1.1.7)
( B x C y A z B x C z A y ) ( B y C x A z B y C z A x ) + ( Bz A x C y Bz A y C x ) = ( Bx C y A z + B y C z A x + Bz C x A y ) ( Bx C z A y + B y C x A z + Bz C y A x ) = ( A x B y C z + A y Bz C x + A z B x C y ) ( A x Bz C y + A y B x C z + A z B y C x )
C (A B) = C x A y Bz A z B y y ( A x Bz A z B x ) + z A x B y A y B x
((
)) =
(1.1.8)
order them
( C x A y Bz C x A z B y ) ( C y A x Bz C y A z B x ) + ( C z A x B y C z A y B x ) = ( C x A y Bz + C y A z B x + C z A x B y ) ( C x A z B y + C y A x Bz + C z A y B x ) = ( A x B y C z + A y Bz C x + A z B x C y ) ( A x Bz C y + A y B x C z + A z B y C x )
((
)) =
(1.1.9)
order them
If we compare the last row of each expression, we note that they are identical so the algebra identity is verified. D.RamacciaandA.Toscano Pag.2
Exercises Chapter 1
c)
AB + A B
= A
2 2 A B + A B = x A y Bz A z B y y ( A x Bz A z B x ) + z A x B y A y B x
+ A x B x + A y B y + A z Bz
( A y Bz A z B y )
2
+ ( A x Bz A z B x ) + A x B y A y B x
2
+ A x B x + A y B y + A z Bz
( A y Bz A z B y )
+ ( A x Bz A z B x ) + A x B y A y B x
2
) + ( A x B x + A y B y + A z Bz )
2
A 2 B2 + A 2 B2 2A y Bz A z B y + A 2 B2 + A 2 B2 2A x Bz A z B x + z y x z z x y z A 2 B2 + A 2 B2 2A x B y A y Bx + A x Bx + A y B y + A z Bz x y y x
2 2 2 2 A 2 Bz + A z B2 2A y Bz A z B y + A 2 Bz + A z B2 2A x Bz A z B x + y y x x
A 2 B2 + A 2 B2 2A x B y A y Bx + A 2 B2 + A 2 B2 + A 2 B2 + x y y x x x y y z z 2A x B y A y B x + 2A x Bz A z Bx + 2A x B y A y Bx
cancel the opposites
2 2 2 2 2 2 A 2 Bz + A z B2 + A 2 Bz + A z B2 + A 2 B2 + A 2 B2 + A 2 B2 + A 2 B2 + A z B z = y y x x x y y x x x y y
n = n x x + n y y + n z z 2 2 2 n = nx + n y + nz = 1
(1.1.10)
D.RamacciaandA.Toscano
Pag.3
Exercises Chapter 1
( n A ) n = x ( n y Az n z A y ) y ( n x Az n z A x ) + z ( n x A y n y A x ) n =
( ) y ( n y A z n z A y ) n z ( n x A y n y A x ) n x + + z ( n y A z n z A y ) n y ( n z A x n x A z ) n x =
+ x ( n z A x n x A z ) n z n x A y n y A x n y + z +x n 2 Ax n x n z Az n x n y A y + n 2 A x + y
2 y n y n z Az n z A y n 2 A y + n y n x A x + x y +z n 2 Az n z n y A y n z n x Ax + n 2 Az x
(1.1.11)
( ) y n x ( A x n y A y n x ) n z ( A y n z A z n y ) + +z n x ( Az n x A x n z ) n y ( A y n z Az n y ) =
+x Ax n 2 A y n y n x Az n z n x + Ax n 2 + y z
2 y A x n x n y A y n 2 A y n z + A z n z n y + x 2 2 +z Az n x Ax n x n z A y n y n z + Az n y
+x n y A x n y A y n x n z ( A z n x A x n z ) +
(1.1.12)
It is very easy to show that ( n A ) n = n ( A n ) . The second term of the identity can be written as:
D.RamacciaandA.Toscano
Pag.4
Exercises Chapter 1
)( n x x + n y y + n z z ) =
( ) + ( ) + + z n z ( n x A x + n y A y + n z A z ) =
x +x n 2 A x + n x n y A y + n x n z A z + +y n y n x A x + n 2 A y + n y n z Az + y
2 +z n z n x A x + n z n y A y + n z Az
(1.1.13)
+x A x n 2 A y n y n x Az n z n x + A x n 2 + n 2 A x + n x n y A y + n x n z Az + y z x 2 2 2 +y A y n x + A y n z A x n x n y Az n z n y + n y n x Ax + n y A y + n y n z Az + +z Az n 2 A x n x n z A y n y n z + Az n 2 + n z n x A x + n z n y A y + n 2 Az = x y z
2 + xA x n 2 + n z + n 2 + + yA y n 2 + n 2 + n 2 + + zA z n 2 + n 2 + n 2 = A x z y y z y x x
(1.1.14)
D.RamacciaandA.Toscano
Pag.5
Exercises Chapter 1
1.2 Exercise
Prove the vector analysis identities: 1. ( ) = 0 2. ( ) = 2 + 3. ( ) = 2 2 4. ( A ) = ( ) A + A 5. ( A ) = ( ) A + A 6. ( A ) = 0 7. A B = B ( A ) A ( B ) 8. ( A ) = ( A ) 2 A First of all we have to express the operator in general orthogonal coordinates in four common applications. All vector components are presented with respect to the normalized base ( e1 , e 2 , e 3 ) : (Green's first identity) (Green's second identity)
e1 e2 e3 = h q + h q + h q 1 1 2 2 3 3 1 h 2 h3 h1h3 h1h 2 2 = + + h1h 2 h3 q1 h1 q1 q 2 h 2 q 2 q3 h3 q3 1 F = ( F1h 2 h3 ) + ( F2 h1h3 ) + ( F3h1h 2 ) h1h 2 h3 q1 q2 q3 h1e1 h 2e2 h3e3 F = 1 = h1h 2 h3 q1 q 2 q3 h1F h 2 F2 h3F3 1 e e + 1 ( h3F3 ) ( h 2 F2 ) + 2 ( h1F1 ) ( h3F3 ) + q3 q1 h 2 h3 q 2 h1h3 q3 e + 3 ( h 2 F2 ) ( h1F1 ) q2 h1h 2 q1 (1.2.1)
where ( h1 , h 2 , h 3 ) are the metric coefficients. For common geometries they are defined as follow:
D.RamacciaandA.Toscano
Pag.6
Exercises Chapter 1
(rectangular coordinates) (cylindrical coordinates) (spherical coordinates)
(1.2.2)
For simplicity, the proves are done using rectangular coordinates ( h1 = 1, h 2 = 1, h 3 = 1) : Identity n 1
e1 ( ) = q1
e2 q 2
e3 q3
q1 q 2 q3 e1 e2 + q1 q3 q3 q1 q 2 q3 q3 q 2 = =0 +e3 q1 q 2 q 2 q1
For the property of linearity of the derivate operator parentheses vanishes and also the result. Identity n 2
( ) = e1 + e2 + e3 = q1 q 2 q 3 = = + + q1 q1 q 2 q 2 q 3 q 3
2 2 2 = + + + + + = q1 q1 q1 q 2 q 2 q 2 q 3 q 3 q 3 2 2 2 = + + + + = 2 + + q1 q 2 q 3 q1 q1 q 2 q 2 q 3 q 3
D.RamacciaandA.Toscano
Pag.7
Exercises Chapter 1
( ) = e1
Now we can apply the dot product:
+ e2 + e3 q1 q 2 q3 q1 q 2 q3
Identity n4
( A ) = ( A1e1 + A 2 e2 + A3e3 ) = ( A1 ) + ( A 2 ) + ( A3 ) = q 2 q3 q1 A A 2 A3 = 1 + A1 + A2 + A3 = + + q1 q 2 q 2 q3 q3 q1 A1 A 2 A3 = A1 + A2 + A3 + + = ( ) A + A + q 2 q3 q1 q 2 q3 q1
Identity n 5
e1 ( A ) = q1 A1 e2 q 2 A 2 e3 = q 3 A 3
D.RamacciaandA.Toscano
Pag.8
Exercises Chapter 1
+e1 ( A3 ) ( A 2 ) + e2 ( A1 ) ( A3 ) + e3 ( A 2 ) ( A1 ) = q3 q1 q 2 q1 q 2 q3 A3 A 2 =e1 A3 + A + q3 2 q3 q 2 q 2 A3 A + e 2 + A1 + 1 A3 + q3 q1 q1 q3
A 2 A + e3 A2 + A1 + 1 = q1 q 2 q 2 q1 =e1 A3 A 2 + e2 A1 A 3 + e3 A2 A1 + q3 q1 q 2 q1 q 2 q3 A A A A A 2 A + e 2 1 3 + e3 2 1 = ( ) A + A +e1 3 q3 q1 q 2 q1 q 2 q3
Identity n 6
A A 3 A1 A 2 A1 A 2 ( A ) = e1 3 e2 + e3 = q 2 q 3 q1 q 2 q1 q 3 A 3 A 2 = q1 q 2 q 3 A 3 A1 A 2 A1 + = q 2 q1 q 3 q 3 q1 q 2
A 3 A 2 A 3 A1 A 2 A1 = + =0 q1 q 2 q1 q 3 q 2 q1 q 2 q 3 q 3 q1 q 3 q 2
divergence of vector A B . This choise is obligated by the fact that if first we calculated the divergence of the vector A , the results would be a scalar. Cross product with the vector B would be impossible. So we have:
D.RamacciaandA.Toscano
Pag.9
Exercises Chapter 1
A 2 B A3 B B A = B3 + A 2 3 B2 + A 3 2 B3 1 + A1 3 + q1 q1 q1 q1 q 2 q 2 A3 B A1 B B A 2 + B1 + A 3 1 + B2 + A1 2 A 2 1 + B1 = q 2 q 3 q3 q 3 q3 q 2 A A1 A3 A 2 A1 A 2 =B1 3 + B2 + B3 + q1 q 2 q 2 q 3 q3 q1 B B B B B B + A1 2 3 + A 2 3 1 + A 3 1 2 = q 2 q1 q3 q 2 q1 q3 =B ( A ) A ( B )
Identity n 8
A A 3 A1 A 2 A1 A 2 ( A ) = e1 3 e2 + e3 = q1 q 2 q 2 q 3 q1 q 3 A 2 A1 A1 A 3 =e1 + q q 2 q 3 q 3 q1 2 q1 A 2 A1 A 3 A 2 e2 + q q q 2 q 3 q 2 q 3 1 1 A1 A 3 A 3 A 2 + e3 = q q q1 q 2 q 2 q 3 1 3 A 2 2 A1 2 A1 A 3 + =e1 + q 2 q1 q 2 q 3 q 3 q1 2A2 A1 A 3 2 A 2 e2 + + q1 q1 q 2 q 3 q 2 q 3 A1 2 A 3 2 A 3 A 2 + e3 + = q1 q 3 q1 q 2 q 2 q 3
D.RamacciaandA.Toscano
Pag.10
Exercises Chapter 1
A 2 A1 A1 A 3 A 3 A 2 + + =e1 e2 + e3 + q 2 q1 q 3 q1 q1 q 2 q 3 q 2 q1 q 3 q 2 q 3 2 A1 2 A1 2 A2 2 A2 1 2 + +e e q 2 q1 q 3 q 3 2 A3 2 A3 3 = +e q1 q 2
A 2 A 3 A1 A 3 A1 A 2 + + + =e1 + e2 + e3 q 3 q1 q 2 q1 q 2 q 3 q 2 q1 q 3 2 A1 2 A1 2 A2 2 A2 1 2 e + + e q 2 q 3 q 3 q1
2 A3 2 A3 3 + + = ( A ) 2 A e q1 q 2
D.RamacciaandA.Toscano
Pag.11
Exercises Chapter 1
1.3 Exercise
Consider the infinitesimal volume element xyz shown below, such that its upper half lies in medium 1 and its lower half in medium 2 . The axes are oriented such that
n = z
Fig. 1.3.1: Infinitesimal volume element between two medium. 1. Applying the integrated form of Ampre's law to the infinitesimal face abcd, show that
H 2 y H 1y = J x z + D x z t
(1.3.1)
2. In the limit
z 0
whereas the first term will be nonzero and may be set equal to the surface current density, that is, J sx lim z 0 ( J x z ) . Show that this leads to the boundary condition
H 1y H 2 y = J sx . Similarly, shows that H 1x H 2 x = J sy , and that these two boundary
(1.3.2)
3. Apply the integrated form of Gauss's law to the same volume element and show the boundary condition D1z D 2 z = s = lim z 0 ( z ) .
Solution
Question n 1 In its historically original form, Ampre's circuital law relates the magnetic field to its electric current source. The law can be written in two forms, the integral form and the differential form. The forms are equivalent, and related by the KelvinStokes theorem. The identity demonstrated by Stokes is the follow:
( F ) dS =
S c(S)
Fd
(1.3.3)
So applying (1.3.3) to the second Maxwell's equation, we obtain the Ampere's law in integral form with few simply steps: D.RamacciaandA.Toscano Pag.12
Exercises Chapter 1
H = J +
Integrate terms of the identity over an opened surface S:
D t
D n dS t
( H ) ndS = J ndS +
S S S
c(S)
H d = J ndS +
S S
D ndS t
(1.3.4)
where
is the infinitesimal vector, tangent to the curved line c that bounds the surface S.
Now we can consider the infinitesimal face abcd, that has area S = zy and perimeter
p = 2z + 2y . The lefthand side of (1.3.4) can be decomposed into a sum of four integral
expression, one for each infinitesimal side of the rectangular abcd, and we have to define the sense
of integration. Choose an counterclockwise path so that, using the righthand rule, the normal is x.
Note that the zparallel sides have the first half in the medium 1 and the second in medium 2. So the integral on that part of the path needs to be decomposed into two integral with different arguments. For simplicity, denote the points of contact between mediums along the segments ab and cd with O1 and O 2 , respectively. On the contrary, to solve the rightside of (1.3.4) we have to identify the correct component of J and D that flows through the face abcd, i.e. the component Jx and D x . So we obtain:
H1 dz
a O1
O1
H 2 d z + H 2 dy +
b
O2
H 2 dz +
O2
H1 dz H1 dy =
d
= J x z y +
D x z y t
H 1 and H 2 are constant inside each medium, so the line integrals can be written as:
H1z
D x z z z z H1y y = J x zy + zy H 2z + H 2y y + H 2z + H1z t 2 2 2 2
D x z y t
i.e.
H 2 y y H 1y y = J x z y + H 2 y H 1y = J x z +
D x z t
H 1x H 2 x = J sy . In order to obtain eq. (1.3.2), we can subtract vectorially these two boundary
conditions:
y ( H1x H 2x ) x H1y H 2y = Jsx x + Jsy y
n ( H1 H 2 ) = J s
where n = z .
Question n 3 Gauss's law relates the electric field to its electric charge sources. Like Ampre's circuital law, it can be written in two forms, the integral form and the differential form. The forms are equivalent, and related by the divergence theorem:
( F ) dV =
V S(V)
F ndS
(1.3.5)
So applying (1.3.5) to the third Maxwell's equation, we obtain the Gauss's law in integral form with few simply steps:
D =
Integrate terms of the identity over a volume V:
( D ) dV = dV
V V
S(V)
D ndS =
dV = Qin
V
(1.3.6)
where n is the outgoing unit vector normal to the closed surface S that bounds the volume V.
Now consider the volume V = xyz . The lefthand side of (1.3.6) can be decomposed into two integrals with arguments D1 and D2 , respectively in the medium 1 and medium 2. The righthand side of (1.3.6) is a simple volume integral. So we have:
( D1 n ) dS1 + ( D2 n ) dS2 = dV = xyz
S1 S2 V
(1.3.7)
where S1 and S2 are portions of S in the medium 1 and medium 2, respectively and is considered constant inside the volume V. The terms on the righthand side of eq. (1.3.7) can be decomposed into several surface integrals, one for each side of parallelepiped xyz : D.RamacciaandA.Toscano Pag.14
Exercises Chapter 1
D1z xy + D1y
D 2z xy + D 2y
i.e.
D1z D2z = z
In the limit z 0 , the amount z collapses in s which is the surface electric charge density.
D.RamacciaandA.Toscano
Pag.15
Exercises Chapter 1
1.4 Exercise
Show that time average of the product of two harmonic quantities (t) = Re Ae j t and
A(t) B (t) =
1 T
T 0
A(t) B (t)dt = 2 Re AB
(1.4.1)
where T = 2 is one period. Then show that the timeaveraged values of the cross and dot products of two timeharmonic vector quantities A (t) = Re A e j t and B (t) = Re Be j t can be
(1.4.2) (1.4.3)
Solution
First of all, we express the harmonic quantities A(t) and B(t) in their extended form:
A(t) = A cos ( t + 1 ) B (t) = B cos ( t + 2 )
(1.4.4)
A(t) B (t)dt =
1 AB cos ( t + 1 ) cos ( t + 2 ) dt T
0
(1.4.5)
e jx + e jx cos x = 2 jx jx sin x = e e 2j
(1.4.6)
D.RamacciaandA.Toscano
Pag.16
Exercises Chapter 1
dt =
A(t) B(t) =
j t 1 T j t 1 e e j t e2 + e j t e2 AB e e + e = dt = 2T 2 0 T + + AB e2 j t e( 1 2 ) + e( 1 2 ) + e ( 1 2 ) + e2 j t e ( 1 2 ) dt = = 2 2T
)(
e j t e2 + e j t e2 2
0 T
Operating in similar way, we can demonstrate the timeaveraged values of the cross and dot products of two timeharmonic vector quantities. Cross Product
A (t) B (t) = 1 T
T 0
( A(t) B (t) ) dt =
A(t) B(t) =
Dot Product
A(t) B (t) =
= 1 T
T 0
( A(t) B (t) ) dt =
D.RamacciaandA.Toscano
Pag.17
Exercises Chapter 1
1.5 Exercise
Assuming that B = H : 1. Show that Maxwell's equations
E = j B H = J + j D D = B = 0
(1.5.1)
where J tot = J + j D . 2. Extracting the realparts of both sides of eq. (1.5.1) and integrating over a volume V bounded by closed surface, show the timeaveraged form of energy conservation:
S(V)
(1.5.2)
which states that the net timeaveraged power floating into a volume is dissipated into heat. 3. For a lossless dielectric, show that the integrals in (1.5.2) are zero and provide an interpretation.
Solution
Question n 1 Using the identity ( E H ) = H ( E ) E ( H ) and Maxwell's equations, we have:
E H = H ( E ) E H = H ( j B ) E J * j D* = = j H H E J tot
Question n 2 Integrate over a volume V the righthand side of eq. (1.5.1) and apply the divergence's theorem:
( ( E H
V
)) dV =
S(V)
( E H ) ndS
Exercises Chapter 1
(1.5.3)
( (
) )
(1.5.4)
The real part of j H H is zero because the product H H = H 2 is real and so the quantity
j H H is imaginary. Only the term associated with the heat survives and we can write:
S(V)
1 1 Re E H ndS = Re E J dV tot 2 2
V
(1.5.5)
The minus sign is been associated with the lefthand side because it identifies the quantity of energy that goes in the volume V while the Poynting's vector is defined outgoing from V and the righthand side represents the energy dissipated as heat. Question n 3 Inside a lossless dielectric, the current density J is zero while the displacement current D is simply equal to E . So:
2 Re E ( j E
V
) dV = 0
(1.5.6)
being the real part of j E E zero. Moreover zero for the righthand side of (1.5.5), that represents the quantity of energy ingoing the volume bounded by the surface S, implies that not all the energy remains inside the volume. Exactly in steady state the quantity of energy ingoing is equal to the outgoing one. It is correct because electromagnetic wave pass through the dielectric.
D.RamacciaandA.Toscano
Pag.19
Exercises Chapter 1
1.6 Exercise
Assuming that D = E and B = H , 1. Show that Maxwell's equations imply the following relationships:
Ex + D = Ex E x E2 t x 2 B 1
(1.6.1)
( J B )x +
D 1 B = Hx H x H2 2 t x
(1.6.2)
where the subscript x means the xcomponent. 2. From eq. (1.6.1) and (1.6.2), derive the following relationship that represent momentum conservation:
fx + G x = Tx t
(1.6.3)
Tx = Ex E + Hx H x
1 E 2 + H2 2
3. Write similar equations of y, z components. The quantity G x is interpreted as the field momentum (in the xdirection) per unit of volume, that is, the momentum density.
Solution
Question n 1 Let us begin with eq. (1.6.2) because it is easy to note from the lefthand side that it is the cross product of the second Maxwell's equation (i.e. H = J +
and then we
extract the xcomponent. So we have to demonstrate the righthand side of eq. (1.6.2). We can write:
( H ) B = x H z H y H z H x y z z x y x y H y H x + z B = y x z
H y H H z H x H y H x = z z z x y x y Bx By Bz
(1.6.4)
D.RamacciaandA.Toscano
Pag.20
S.J. Orfanidis Electromagnetic Waves and Antennas Now we consider only the xcomponent, writing Bi = Hi :
Exercises Chapter 1
( H ) B x component = H z
H y H z H x H x = H z + Hz Hy + Hy x z x y
H z H x z x
H y H x H y x y =
(1.6.5)
From the forth Maxwell's equation (i.e. B = 0 ) and the constitutive relation B = H , we can add to eq. (1.6.5) the term H x ( H ) and the couple of terms H x H x , because they're both zero:
x
( H ) B x component
H y H z H x H x H x + Hz H y + Hy + Hx + H z x z x y x = = H y H z H x H x +Hx + Hx + Hx H x y z x x
(1.6.6)
H y H y H z H x H z H x 1 Hz + Hy + Hx + 2H y + 2H x = 2H z = x x x x x x 2 2 1 H 2 H y H 2 1 x z + + = x H 2 x x 2 x 2
and now consider the remaining terms:
H y H x H x H x H z H x + Hy + Hx + Hx + Hx + Hx = z y x y z x order them =
Hz
H y H x H x H x H z = 2H x + Hx + Hy + Hx + Hz x y y z z
H 2 H y H x (Hx Hz ) = H x H x x + H y y + H z z = x + + y z x
( (
) ) = ( H x H )
1 ( H ) B xcomponent = ( Hx H ) x H2 = 2 1 = H x H x H2 2
(1.6.7)
D.RamacciaandA.Toscano
Pag.21
Exercises Chapter 1
B Eq. (1.6.1) is obtained in similar way. In the lefthand side, there is the term D that t x
and then we extract the xcomponent. So we have to demonstrate the righthand side of eq. (1.6.1). We can apply the cross product to the first Maxwell's equation:
D ( E) = D
B t
From the properties of the cross product, it's possible to invert the order of the terms in the left hand side and change the sign in the righthandside:
( E) D = D
Now consider the term ( E ) D :
( E ) D = x E z E y E z E x y z z x y x y
B t
E y E x + z D = y x z
E y E E z E x E y E x = z z z x y x y Dx Dy Dz
(1.6.8)
(1.6.9)
As for eq. (1.6.5), we can add to eq. (1.6.9) the third Maxwell's equation (i.e. D = 0 ), but in this case there is the term and it's correct for the results that we want to obtain. In fact, multiplying it with E x , the term Ex completes the lefthand side of eq. (1.6.1), changing its sign. With these considerations, we can add to eq. (1.6.9) the term E x D and the couple of terms
Ex E x : x
D.RamacciaandA.Toscano
Pag.22
Exercises Chapter 1
( E ) D x component
E y E z E E E x + E z x E y + Ey x + Ex + E z x z x y x = E y E z E x E x +E x + Ex + Ex E x y z x x
(1.6.10)
E z
E y E y E z E E E 1 Ey E x x = 2E z z + 2E y + 2E x x = x x x 2 x x x 2 1 E 2 E y E 2 1 z = + + x = x E 2 x x 2 x 2
E y E x E x E x E z E x + Ey + Ex + Ex + Ex + Ex = z y x y z x order them
Ez
E y E x E x E x E z = 2E x + Ey + Ex + Ex = + Ez x y y z z E 2 E x E y ( Ex Ez ) = E x E x x + E y y + E z z = x + + y z x
( (
) ) = ( E x E )
( E) D xcomponent
1 = ( E x E ) x E2 = 2 1 = E x E x E2 2
(1.6.11)
Question n 2 The identity (1.6.3) is obtained adding eq.(1.6.1) and (1.6.2) as follow:
B 1 1 D B = E x E + Hx H x E2 x H2 E x + ( J B )x + D + t x t 2 2 x 1 1 E x + ( J B )x + ( D B ) = E x E + H x H x E2 x H2 t 2 2 x
It is easy to note the presence of f x , G x and Tx as defined in the text of the exercise.
t
D.RamacciaandA.Toscano
Pag.23
Exercises Chapter 1
Operating in the similar way to question n1, it is possible to demonstrate that the relationships (1.6.1) and (1.6.2) can be written for the y and zcomponent as follow:
B 1 2 D t E y = y 2 E E y E y D 1 ( J B ) y + B = y H2 Hy H t y 2
B 1 2 D t + E z = E z E z 2 E z 1 D B = Hz H z H2 ( J B )z + 2 t z
(1.6.12)
(1.6.13)
From eq. (1.6.12) and (1.6.13) as in question n2, we can derive the relationship that represents momentum conservation for y and zcomponent:
fy +
fz +
G y t
= Ty
(1.6.14) (1.6.15)
G z = Tz t
where f y = ( J B ) y E y G y = ( D B ) t y 1 Ty = y E 2 + H 2 E y E H y H 2 f z = ( J B ) z + E z G z = ( D B ) t z 1 2 2 Tz = E z E + H z H z E H 2
D.RamacciaandA.Toscano
Pag.24
Exercises Chapter 1
1.7 Exercise
Consider the permittivity of a dispersive material
( ) = 0 +
2 0p 2 0 2 + j
(1.7.1)
Ne2 p = 0m
and measures the rate of collisions per unit of time.
(1.7.2)
Show that the casual and stable timedomain dielectric response of eq. (1.7.1) is given as follows:
( t ) = 0 ( t ) + 0 ( t ) 2 p t/2 e sin (0 t ) u ( t ) (t) = 0
2 where u ( t ) is the unitstep function and 0 = 0 2 4 , and we must assume that < 20 , as
typically the case in practice. Discuss the solution for the case 2 > 0 .
Solution
For the linearity of Fourier transform, we have
{ ( )} =
2 0 p 0 + 2 0 2 + j
1 2 = 1 { 0 } + 0p 1 2 2 0 + j
(1.7.3)
where 1 denotes the inverse Fourier transform operator. The first term of eq. (1.7.3) is constant, so it's easy to calculate:
1 { 0 } = 0 (t)
(1.7.4)
(t)e
j t
dt = e j t
t =0
=1
D.RamacciaandA.Toscano
Pag.25
Exercises Chapter 1
The second term of eq. (1.7.3) is more complicate and it is necessary to simplify the argument. First of all, we can reduce the denominator in the product of two polynomials of first degree. So we
1,2 =
Assuming that
2 j 2 + 40
(1.7.5)
1,2 = j 0
2
(1.7.6)
where it's important to note that has been substituted by because of the minus sign of the denominator. Now we can write:
1 1 2 2 0 + j 1 A B 1 1 + = = ( 1 )( 2 ) ( 1 ) ( 2 )
where A and B are two constant that we calculate applying the method of weighted residuals:
A = lim
1 1 1 1 = = = 20 1 ( 2 ) (1 2 ) j 0 j 0 2 2
1 1 1 1 B = lim = = = 2 ( 1 ) (2 1 ) 20 j + 0 j + 0 2 2 so:
1
1 1 20 20 1 1 + 2 = 2 0 + j ( 1 ) ( 2 )
(1.7.7)
Now the problem is only to transform the trivial expression 1 ( i ) and then to apply the result to eq. (1.7.7). To solve this problem, consider:
je ji t u(t) ,
D.RamacciaandA.Toscano
Pag.26
Exercises Chapter 1
+
j i ) t
dt =
= j = =
1 e j( i )t = 0 j ( i )
0 +
(1.7.8)
1 j( i )t j i ) t e ( e = ( i ) t + t 0 1 1 [ 0 1] = ( i ) ( i )
1 j t 1 = je i u ( t ) i
Using eq. (1.7.9) in eq. (1.7.7), we have:
(1.7.9)
(1.7.10)
j2 t
j j t j j + t e 2 0 e 2 0 u ( t ) = u t = j ( ) 20
t t + j t j0 t = e 2e 0 u (t) = j e 2 e 20
= = =
1 20 1 2 0 1
je
t 2
e j0t e+ j0t u ( t ) =
t 2
(1.7.11)
je
2 jSin (0 t ) u ( t ) =
t 2 Sin
(0 t ) u ( t )
=
(1.7.12)
(0 t ) u ( t ) = 0 ( t ) + 0 ( t )
Pag.27
D.RamacciaandA.Toscano
Exercises Chapter 1
D.RamacciaandA.Toscano
Pag.28
Exercises Chapter 1
1.8 Exercise
Show that the plasma frequency for electrons can be expressed in the simple numerical form:
fp = 9 N
(1.8.1)
where fp is in Hz and N is the electron density in electrons/m3. What is fp for the ionosphere if
N = 1012 ?
Solution
Plasma frequency id defined as
fp = 1 2 Ne 2 0m
(1.8.2)
where e is the electron charge, 0 is the permittivity of vacuum and m is the mass of electron. So we have to demonstrate the follow identity:
1 2 e2 =9 0m
(1.8.3)
The charge of an electron is 1, 602 10 19 C and its mass is about 9,10 10 31 Kg . The electric permittivity is about 8, 85 10 12 F m . So:
1 8, 988925.... 2 3,14 9,10 1031 8,85 1012
D.RamacciaandA.Toscano
Pag.29
Exercises Chapter 1
1.9 Exercise
Show that the relaxation equation
2 (r , t) + (r , t) + p (r , t) = 0
(1.9.1)
where is the charge density in the conductor, is the measurement of collisions per unit of time and p is the plasma frequency, can be written in the following form in term of dcconductivity
2 = 0 p = Ne 2 m :
(r , t) + (r , t) +
(r , t) = 0 0
(1.9.2)
+ =0 t
in the limit = 1 0 . Show also that in this limit, Ohm's law
J (r , t) =
2 p
(1.9.3)
(1.9.4)
takes the instantaneous form J = E , from which the naive relaxation constant relax = 0 was derived.
Solution
Eq. (1.9.2) is obtained dividing eq. (1.9.1) by :
1
2 p
(r, t) + (r, t) +
(r, t) = 0
2 where p = 0 . It's easy to note that if = 1 0 , then the term (r, t) 0 and the eq.
(1.9.2) is reduced to eq. (1.9.3). The Ohm's law (1.9.4) can be written, highlighting we have to solve only the integral of an exponential:
2 J (r , t) = p 0 E (r , t)
e (t t ') dt '
So
D.RamacciaandA.Toscano
Pag.30
Exercises Chapter 1
1 (t t ') t 1 1 = 1 0] = e [
dt ' =
e (t t ') d[ (t t ')] =
E(r, t) = E(r, t)
(1.9.5)
D.RamacciaandA.Toscano
Pag.31
Exercises Chapter 1
1.10 Exercise
Conductors and plasmas exhibit anisotropic and birefringent behavior when they are in the presence of an external magnetic field. The equation of motion of conduction electrons in a constant external magnetic field is
mv = e(E + v B) m v
that E = x E x + y E y and v = x v x + y v y . 1. Show that in component form, the above equations of motion read:
e E x + B v y v x m e v y = E y B v x v y m vx =
(1.10.1)
with the collisional term included. Assume the magnetic field is in the zdirection, B = z B , and
(1.10.2)
where B = eB m is the cyclotron frequency. What is the cyclotron frequency in Hz for electrons in the Earth' magnetic field
B = 0.4 gauss = 0.4 10 4 T esla ?
2. To solve this system, work with the combinations v x jv y . Assuming harmonic time dependence, show that the solution is:
e E x jE y v x jv y = m + j ( B )
(1.10.3)
(1.10.4)
where ( ) = conductivity.
0 + j ( B )
(1.10.5)
where ( t ) = 0 e t e j B t u ( t ) is the inverse Fourier transform of ( ) and u(t) is the unitstep function. 5. Rewrite eq. (1.10.5) in component form:
D.RamacciaandA.Toscano Pag.32
Exercises Chapter 1
J x (t) = xx (t t ')E x (t ') + xy (t t ')E y (t ') dt ' J y (t) = yx (t t ')E x (t ') + yy (t t ')E y (t ') dt '
0 0 t
(1.10.6)
and identify the quantities xx (t) , xy (t) , yx (t) , yy (t) . 6. Evaluate eq. (1.10.6) in the special case E x (t) = E x u(t) and E y (t) = E y u (t) , where E x and E y are constants, and show that after a long time the steadystate version of eq. (1.10.6) will be:
Jx = 0 Jy = 0 E x + bE y 1 + b2 E y bE x 1 + b2
(1.10.7)
Fig. 1.10.1: Conductor with finite extent in ydirection. where b = B . If the conductor has finite extent in the ydirection, as show in Fig. 1.10.1, then no steady current can flow in this direction, J y = 0 . This implies that if an electric field is applied in the xdirection, an electric field will develop across the yends of the conductor, E y = bE x . The conduction charges will tend to accumulate either on the right or the left side of the conductor, depending on the sign of b, which depends on the sign of the electric charge e. This is the Hall effect and is used to determinate the sign of the conduction charges in semiconductors, e.g. positive holes for ptype, or negative electrons for ntype. What is the numerical value of b for electrons in copper if B is 1 gauss? 7. For a collisionless plasma ( = 0 ), show that its dielectric behavior is determined from
D x jD y = ( ) E x jE y , where
2 p 1 ( ) = 0 ( B )
(1.10.8)
D.RamacciaandA.Toscano
Pag.33
Exercises Chapter 1
Solution
Question n1 First of all, divide eq. (1.10.1) by m:
v=
and expand the terms
e (E + v B) v m
vx x + v y y =
e Ex x + E y y + v y Bx vx By vx x + vy y m
) (
e vx = E x + v y B vx m v = e E v B v x y y m y
e eB vx = Ex + v y vx m m v = e E eB v v y y m y m x
where is easy to note the cyclotron frequency B . The cyclotron frequency for electrons ( e = 1, 602 10 19 C , m = 9,10 10 31 Kg ) in the Earth's magnetic field is:
fB = 1, 602 1019 0, 4 104 2 3,14 9,10 10
31
so we have:
e E x + B v y v x m e j v y = E y B v x v y m j v x =
(1.10.9)
e j E y B v x v y m
that is
j vx jvy =
D.RamacciaandA.Toscano
e Ex jE y + B vy jvx vx jvy m
) (
(1.10.10)
Pag.34
Exercises Chapter 1
In Eq. (1.10.10) there is the term B v y jv x that we have to express in the form C vx jvy , where the constant C is to be found. If we take out of the parentheses j , we obtain
jB v x jv y and the constant C = jB . So eq. (1.10.10) becomes:
j vx jvy =
and we obtain:
(
(
)
)
e Ex jE y jB vx jv y vx jvy m
) (
(
)
)
j v x jv y jB v x jv y + v x jv y = e ( v x jv y ) ( + j ( B ) ) = m ( E x jE y )
e E x jE y = m + j ( B )
) (
e E x jE y m
(
Question n3
v x jv y
(1.10.11)
Substituting eq. (1.10.11) in the expression for the induced currents ( J = Nev ), we have:
J x jJ y = Ne v x jv y
where we can identify ( ) as
Ne2 m E jE y = + j ( B ) x
(1.10.12)
Ne2 0 m ( ) = = + j ( B ) + j ( B )
where 0 =
Ne 2 is the dc value of the conductivity. m
(1.10.13)
(1.10.14)
0 1
1 1 1 1 1 = j 0 = j 0 (1.10.15) j + ( j B ) + j ( j B ) +
D.RamacciaandA.Toscano
Pag.35
Exercises Chapter 1
where = j ( j B ) . It's easy to note that the inverse Fourier transform in (1.10.15) is already calculated in exercise 1.7 (eq. (1.7.9)). So we have:
( t ) = j 0 1
1 j t = j 0 je u ( t ) = +
(1.10.16)
= 0e j t u ( t ) = 0e t e jB t u ( t )
where u ( t ) is the unit step function. Now we can write eq. (1.10.12) in the timedomain version:
J x (t) jJ y (t) = (t t ') E x (t ') jE y (t ') dt '
0 t
(1.10.17)
Question n5 It's possible to decompose eq. (1.10.17) in its two component as follow:
J x (t) + jJ y (t) = + (t t ') E x (t ') + jE y (t ') dt ' J x (t) jJ y (t) = (t t ') E x (t ') jE y (t ') dt '
0 0 t t
(1.10.18)
(t t ') (t t ') j + E y (t ') dt ' (1.10.21) 2 (t t ') + (t t ') j + E y (t ') dt ' (1.10.22) 2j
D.RamacciaandA.Toscano
Pag.36
Exercises Chapter 1
(1.10.23)
I1 and I2 can be solved separately. Let's start with I1 substituting E x (t) = E x u(t) and the definition of xx (t) . So we obtain:
Now we solve separately I11 and I12, substituting the definitions of + (t) and (t) respectively:
I11 = + (t t ')E x u(t ')dt ' = 0e j+ (t t ') u ( t t ') E x u(t ')dt ' =
0 0 t j+ (t t ') t t
= 0 E x e
0
e = 0 E x j +
j+ (t t ') t
1 e j+ t = 0 E x 0 j+
1 0 I11 = 0Ex Ex = j+ ( + jB )
In the same way we can solve I12 and obtain:
(1.10.24)
1 0 I12 = 0Ex Ex = j ( jB )
D.RamacciaandA.Toscano
(1.10.25)
Pag.37
Exercises Chapter 1
Now combining eq. (1.10.24) and (1.10.25), we have the solution of integral I1:
0E x 1 1 0E x + I1 = (I11 + I12 ) = = 2 2 ( + jB ) ( jB )
= =
0E x
2
0E x jB + + jB 1 1 + = 2 ( + jB ) ( jB ) 2 2 + B
= (1.10.26)
0E x
2
2 2 2 = 2 0 2 Ex + 2 + B B
The integrals I21 and I22 have the same structure of integrals I11 and I12. So the results are known:
I 21 = I 22
0 Ey ( + jB ) 0 = Ey ( jB )
(1.10.27)
j 0 E y j 0 E y jB jB 1 1 = (1.10.28) = 2 2 ( + jB ) ( jB ) 2 2 + B j 0 E y 2 jB 2 = 2 B 02 E y 2 2 + B + B
Now it is possible to write Jx (t) in steady state when a constant electric field is applied:
J x (t) =
2 E x + B E y 2 0 E x + 2 B 02 E y = 0 = 2 2 2 + B + B 2 + B 2 E x + B E y 2 2 2 + B 2
= 0 E x + bE y 1 + b2
= 0
where b = B .
D.RamacciaandA.Toscano Pag.38
Exercises Chapter 1
Consider now the expression of J y (t) in eq. (1.10.22) and divide it in two integrals:
(t t ') (t t ') + (t t ') + (t t ') J y (t) = + E x (t ')dt ' + E y (t ')dt ' = 2j 2 0 0 = I3 + I 4
t t
I3 and I4 can be solved separately. Let's start with I3 substituting E x (t) = E x u(t) and the definition of yx (t) . So we obtain:
Now we solve separately I31 and I32, substituting the definitions of + (t) and (t) respectively:
I31 = + (t t ')E x u(t ')dt ' = 0e j+ (t t ') u ( t t ' ) E x u(t ')dt ' =
0 0 t j+ (t t ') t t
= 0E x e
0
e = 0E x j +
j+ (t t ') t
1 e j+ t = 0E x 0 j+
1 0 I31 = 0E x Ex = j+ ( + jB )
In the same way we can solve I32 and obtain:
(1.10.29)
1 0 I32 = 0E x Ex = j ( jB )
Now combining eq. (1.10.29)and (1.10.30), we have the solution of integral I3:
I3 =
(1.10.30)
Exercises Chapter 1
The integrals I41 and I42 have the same structure of integrals I31 and I32. So the results are known:
I 41 = I 42
0 Ey ( + jB ) 0 = Ey ( jB )
(1.10.32)
0E y
2
0 E y jB + + jB 1 1 + = 2 2 2 + B ( + jB ) ( jB )
0E y 2 2 0 E = = 2 2 y 2 2 + B 2 + B
Now it is possible to write J y (t) in steady state when a constant electric field is applied:
2 E y B E x B 0 2 0 J y (t) = 2 E + 2 E = 0 = 2 x 2 y 2 + B + B 2 + B 2 E y B E x
= 0
2 2 2 + B 2
= 0
E y + bE x 1 + b2
where b = B . Question n7 To solve this point it is necessary to obtain the expression of ( ) for a collisionless plasma because it is note the relationship:
( ) = 0 +
( ) j
(1.10.34)
The definition of the conductibility ( ) has just obtained in this exercise, i.e. eq. (1.10.13). We have only to set = 0 for indicating that there is no collision between the electrons and the medium structure:
D.RamacciaandA.Toscano Pag.40
Exercises Chapter 1
(1.10.35)
Ne2 2 p m 1 = ( ) = 0 ( B ) 0 ( B )
2 where 0p =
(1.10.36)
Ne2 . m
The numerical value of b for electrons in copper can be find out using:
b = B 1 = eB m
The result is different from the one in the text which is 43.
D.RamacciaandA.Toscano
Pag.41
Exercises Chapter 1
1.11 Exercise
This problem deals with various properties of the KramersKronig dispersion relations for the electric susceptibility, given by:
(1.11.1)
Hermitian symmetry property ( ) = ( ) , which is equivalent to the even symmetry of its real part, r ( ) = r ( ) , and the odd symmetry of its imaginary part, i ( ) = i ( ) . 1. Using the symmetry properties, show that eq. (1.11.1) can be written in the folded form:
' ( ' ) 2 r ( ) = 2 i 2 d ' 0 '
i ( ) = 2 d ' 0 ' 2
2. Using the definition of the principalvalue integrals, show the following integral:
r ( ' )
(1.11.2)
d '
2 2 0 '
=0
(1.11.3)
dx 1 a+x
a 2 x 2 = 2a ln a x
3. Using eq. (1.11.3), show that the relations (1.11.2) may be rewritten as ordinary integrals (without the instruction) as follows:
r ( ) =
d '
i ( ) =
r ( ' ) r ( ) '2 2
(1.11.4)
d '
Hint: You will need to argue that the integrands have no singularity at ' = .
4. For a simple oscillator model of dielectric polarization, the susceptibility is given by:
D.RamacciaandA.Toscano Pag.42
Exercises Chapter 1
( ) = r ( ) j i ( ) =
=
2 2 p 0 2
2 0
2 2
(02 2 )
2 p 2
(1.11.5)
+ 2 2
Show that for this model the quantities r ( ) and i ( ) satisfy the modified Kramers Kronig relationships (1.11.4).
Hint: You may use the following definite integrals, for which you may assume that
0 < < 20 :
2
dx
(02 x 2 )
= + 2x 2
1
2 0
x 2 dx
(02 x 2 )
+ 2x 2
Indeed, show that these integrals may be reduced to the following ones, which can be found in standard tables of integrals:
2
0 0
1 2 y 2 C os + y 4
dy
1 2 y 2 C os + y 4
y 2 dy
1 2 (1 C os )
where S in ( 2 ) = ( 2 0 ) . 5. Consider the limit of Eq. (1.11.5) as 0. Show that in this case the functions r , i are given as follows, and that they still satisfy the KramersKronig relations:
r ( ) =
2 p
2 p
0 +
i ( ) =
2 p
( 0 ) ( + 0 ) . 20
Solution
The Cauchy's principal value is defined for the integration of a function f ( x ) with limited value in the interval [ a, b ] except for the x = x 0 .This is a special case of Reimann's principal value:
f ( x ) dx = lim
a b x 0 1
1 0+
f ( x ) dx + lim
2 0 + x + 0 2
f ( x ) dx
where 1 and 2 are independent of each other, whereas in the Cauchy's principal value 1 = 2 , so:
f ( x ) dx = lim
a b x 0 a
0+
f ( x ) dx + f ( x ) dx
x 0 +
Exercises Chapter 1
Simmetry property of i
0 +
+ 0
+ ( '+ ) i ( ') + ( ' ) i ( ') 1 = d ' = 0 '2 2 + ' ( ' ) + ( ' ) + ' ( ' ) ( ' ) 1 i i i i = 2 2 0 '
d ' =
+ 0
d '
D.RamacciaandA.Toscano
Pag.44
Exercises Chapter 1
Simmetry property of r
0 +
+ 0
d ' =
Question n2 Using the definition of the principalvalue integrals and the hint, we write:
d '
(1.11.6)
= =
2 1 2 + lim ln ln 1 + ln 1 ln = 2 2 1 2 + lim ln ln =0 2
Question n3 The integrands of equations (1.11.2) have a singularity in ' = and we have to use the principalvalue for solving the integral. But if we also introduce a singularity at the numerator in
' = , we will have the integrand that will not diverge. Besides subtracting eq. (1.11.3), that is
D.RamacciaandA.Toscano
Pag.45
Exercises Chapter 1
Being equal to zero, the second term can be multiplied by a constant and we choose 2 ( ) . So we have:
r ( ) =
(1.11.7)
Because now the integrands don't present a singularity in ' = , we can cancel instruction and obtain:
2 ' i ( ' ) i ( ) d ' '2 2 0
r ( ) =
i ( ) =
Question n4
(1.11.8)
Substitute the quantity i ( ) expressed in eq. (1.11.5) inside the modified KramersKronig relationships (1.11.4) for r ( ) :
r ( ) =
x i ( x ) i ( ) x2 2
dx,
where
i ( ) =
2 p 2 0 2
+ 2 2
where we have substituted ' x to distinguish better the arguments of the integral. Let us denote the denominator of i ( ) as Den[] . So we have that:
r ( ) =
= =
2 2 xp p x dx = x 2 2 Den[x] Den[ ] 0
2 2p
2 2p
D.RamacciaandA.Toscano
Pag.46
Exercises Chapter 1
Now we can expand the numerator x 2 D en[ ] 2 D en[x ] inside the integral using the extended form for Den[x] and Den[] :
2 2 2 2 x 2Den[ ] 2Den[x] = x 2 0 2 + 2 2 2 0 x 2 + 2 x 2 = 4 2 4 2 = x 2 0 + 4 20 2 + 2 2 2 0 + x 4 20 x 2 + 2 x 2 = 4 2 4 2 = x 20 + x 2 4 2x 20 2 + x 2 2 2 20 2 x 4 +2 20 x 2 2 2 x 2 = 4 4 4 = x 20 + x 2 4 20 2 x 4 = 0 x 2 2 2 x 2 x 2 2 = 4 = x 2 2 0 2 x 2
)(
)(
Now it is possible to use the note results suggested by the text of the exercise:
2 2 2 2 p 0 p 4 1 2 1 0 = = r ( ) = 2 Den[ ] Den[ ] 0
2 2 p 0 2
(02 2 )
+ 2 2
that is exactly the expression of r ( ) for a simple oscillator model of dielectric polarization in eq. (1.11.5). Now we have to demonstrate the dual expression for i ( ) :
i ( ) =
2 r ( x ) r ( )
x2 2
dx
where
r ( ) =
2 2 p 0 2
(02 2 )
+ 2 2
D.RamacciaandA.Toscano
Pag.47
Exercises Chapter 1
i ( ) =
2 2 2 2 x2 p 0 2 p 0 Den[x] Den[ ] x2 2
) dx =
2 2p
2 where Den[x] = 0 x 2
2 + 2 x 2 and Den[ ] = 0 2
+ 2 2 .
2 2 Now we can expand the numerator 0 x 2 Den[ ] 0 2 Den[x] inside the integral using
(02 x 2 ) Den[ ] (02 2 ) Den[x] = (02 x 2 ) (02 2 ) + 2 2 (02 2 ) (02 x 2 ) + 2x 2 = (02 x 2 ) 04 + 4 202 2 + 2 2 (02 2 ) 04 + x 4 202x 2 + 2x 2 =
2 2 2 4 2 4 2 6 +0 + 40 20 2 + 2 20 x 20 x 2 4 +2x 20 2 2 x 2 2 + 6 2 4 2 4 2 0 x 40 + 20 x 2 2 x 20 + 20 + x 4 2 20 x 2 2 + 2 x 2 2 =
( ) ( ) ( ) ( ) ( ) 2 4 2 4 0 ( x 2 2 )( x 2 + 2 ) + 20 ( x 2 2 ) 20 ( x 2 2 ) 0 ( x 2 2 ) + x 2 2 ( x 2 2 ) = ( x 2 2 ) 02 ( x 2 + 2 ) 202 + 04 + x 2 2
2 4 2 4 0 x 4 4 + 20 x 2 2 20 x 2 2 0 x 2 2 + x 2 2 x 2 2 =
D.RamacciaandA.Toscano
Pag.48
Exercises Chapter 1
i ( ) =
2 2p
x 2 2 2 x 2 + 2 2 2 + 4 + x 2 2 0 0 0 dx = Den[x]Den[ ] x2 2 1
2 2 2 2 p 0 0 2 2 0 4 2 + 02 + = 2 2 Den[ ] 0 0 0 2 2 p p = = Den[ ] 2 2 2 + 2 2 0
2 2 2 p 0 2 2 0 2 + + = Den[ ]
that is exactly the expression of i ( ) for a simple oscillator model of dielectric polarization in eq. (1.11.5). Question n 5 Let us compare the first integrals in x and y respectively:
2
0
dx
(02 x 2 )
+ 2x 2
1 2y 2 cos + y 4
dy
They differ only in the denominator, so we can compare them to find the condition of equality, substituting y x :
(02 x 2 )
that is:
+ 2 x 2 = 1 2x 2 C os + x 4
4 2 0 + x 4 20 x 2 + 2 x 2 = 1 2x 2C os + x 4
4 2 0 2x 2 0
4 2 4 + x = 1 2x C os + x 2
(1.11.9)
Using the suggested relation Sin ( 2 ) = ( 2 0 ) and the relation C os = 1 2Sin ( 2 ) , we have:
2 2 C os = 1 2Sin 2 = 1 2 2 = 1 2 2 40 20
Now it is possible substitute eq. (1.11.10) in eq. (1.11.9):
4 2 0 2 0 x 2 C os + x 4 = 1 2 x 2 C o s + x 4
(1.11.10)
(1.11.11)
Pag.49
D.RamacciaandA.Toscano
Exercises Chapter 1
In order to match the right and leftside of eq. (1.11.11), it is necessary that 0 1 and it is correct because the integrals in y are a reduced form of the integrals in x. Indeed the result of integrals in y can be written as follow:
1 2 (1 C os ) = 1 2 2 1 1 2 40 = 1
2 2 20
0 1
that is exactly the result of integrals in x when 0 1 . Question n 6 Starting from the expression of ( ) as in eq. (1.11.5) and applying the limit as 0 we get:
0 ( ) = lim ( ) = lim
0
2 p
2 0 0 2 + j 2
(1.11.12)
0,
respectively.
(1.11.13)
2 2 p p 1 = 2 = 0 + 0 2 20 0 2 p
2 For what concerns i0 ( ) = p lim
2 0
2 2
( x ) = lim
1
2
0+
x +2
(1.11.14)
So, we first manipulate i0 ( ) in order to apply eq. (1.11.14). Dividing and multiplying numerator and denominator by 2 and respectively, we have:
D.RamacciaandA.Toscano
Pag.50
Exercises Chapter 1
2 i0 ( ) = p lim
2
2 0 2
= +
=
2 p
(1.11.15)
2 0 2 0 2
lim
2 +
The limit 0 of eq. (1.11.15) has the same form of eq. (1.11.14), assuming and
x
2 0 2 . So it is possible to write:
i0 ( ) =
2 p
2 0 2
(1.11.16)
In order to write (1.11.16) in a simpler form, we apply the following two properties of the Dirac delta function: 1. consider a function f ( x ) with n zeros:
f ( x ) = 0 f ' ( x ) 0 in x i = x1 , x 2 , ... , x n
then
(f ( x )) =
( x xi )
f ' ( xi )
(1.11.17)
i =1
x2 a2 =
d 2 x a2 dx
(x a)
) x =a
d 2 x a2 dx
(x + a)
) x=a
1 1 1 (x a) + (x + a) = ( x a ) + ( x + a ) 2a 2 a 2a
2. Consider a function g ( x ) :
g ( x ) ( x x 0 ) = g ( x 0 ) ( x x 0 )
(1.11.18)
In order to apply (1.11.17) to (1.11.16), first of all we have to evaluate the zeros of the function
f ( ) =
2 0 2 :
f ( ) = 0
= {+ 0 , 0 }
(1.11.19)
Pag.51
D.RamacciaandA.Toscano
S.J. Orfanidis Electromagnetic Waves and Antennas and its first derivate respect to
Exercises Chapter 1
2 2 d 0 2 d d 0 2 2 2 d (1 ) 1 + = f ' ( ) = f ( ) = = 0 d d d d (1.11.20) 2 2 2 2 1 = 0 2 2 = 2 0 2
i0
2 2 p 0 2 = ( ) =
2 p
( i ) = i=1 f ' (i )
2 p ( 0 ) ( + 0 ) + = 2 2 2 2 2 0 2 0 2 0 2 0 0 0 =
2 p
(1.11.21)
( 0 ) + ( + 0 ) 2
2 p 1 1 ( 0 ) + ( + 0 ) = 2 2 p 1 1 ( + 0 ) = ( 0 ) 0 2 0 2 p
Let us, then, apply the second property of the Dirac delta function (1.11.18):
i0 ( ) =
= =
(1.11.22)
( 0 ) ( + 0 ) 20
This expression still satisfies the KramersKronig relations. Substitute the quantity i0 ( ) expressed in eq. (1.11.22) inside the modified KramersKronig relationship (1.11.4) for r ( ) :
r ( ) =
x 0 ( x ) 0 ( )
i i
x2 2
dx
where we have substituted ' x to distinguish better the arguments of the integral. So:
D.RamacciaandA.Toscano
Pag.52
Exercises Chapter 1
r ( ) =
2 2 p 1 x 2 2 x ( ( x 0 ) ( x + 0 ) ) ( ( 0 ) ( + 0 ) ) dx = 2 0 0
x x 2 ( x 0 ) dx 2 ( x + 0 ) dx 2 2 2 p p 0 x 2 0 x = [ I1 + I2 + I3 + I4 ] = 0 0 ( 0 ) dx + 2 ( + 0 ) dx + 2 x 2 x 2 0 0
The integrals I3 and I4 vanishe as demonstrated in (1.11.6). Also the integral I2 vanishes because it is over the realpositive values of x while the Dirac delta function is always zero on this interval (it is not zero only for x = 0 , which is real negative). So we can write:
r ( ) =
2 p
0 02 2
2 p
02 2
= 0 ( )
r
(1.11.23)
0 The KramersKronig relationship (1.11.4) for i ( ) with r ( ) = r ( ) can be written using 0 the definition of r ( ) as in eq. (1.11.12):
r0
So:
( ) = lim
2 2 p 0 2
(02 2 )
(
+ 2 2
i0
( ) =
0
2 2 2 2 2 p 0 2 p 0 x lim dx 2 2 2 2 2 x 2 2 0 2 x 2 2 + 2 x 2 0 + 0
Now it is easy to note that the argument of the limit has already been solved in the solution of question n4 of this exercise resulting in i ( ) . Then, the limit 0 to i ( ) should be applied, which is given by eq. (1.11.14).
D.RamacciaandA.Toscano
Pag.53
Exercises Chapter 1
1.12 Exercise
Derive the KramersKronig relationship:
( ) =
+ ( ' ) d ' j '
(1.12.1)
by starting with the causality condition ( t ) u ( t ) = 0 and translating it to the frequency domain, that is, expressing it as the convolution of the Fourier transforms of ( t ) and u ( t ) .
Solution
The convolution is a mathematical operation on two functions f and g, producing a third function that is typically viewed as a modified version of one of the original functions. In signal theory, it represents the transformation obtained when a signal passes through a blackbox system with a known impulse response. In similar way, in frequency domain the output of the system is the product of the Fourier transformations of the input signal and the impulse response. So the convolution in time domain is also the corresponding operation of the product in frequency domain and vice versa. It is defined as:
+ +
( f g )( t )
f ( ) g ( t ) d =
g ( ) f ( t ) d
( ) = { ( t )} = { ( t ) u ( t )} = { ( t )} {u ( t )} =
1 ( ) U ( ) 2
where ( ) and U ( ) are the Fourier transforms of ( t ) and u ( t ) , respectively. The Fourier transformation of Heaviside step function is:
U ( ) = 1 + ( ) j
So we have:
1 ( ) = ( ' ) U ( ' ) d ' 2 = = 1 1 1 ( ' ) d ' + 2 j ( ') 2
+ + + +
D.RamacciaandA.Toscano
Pag.54
Exercises Chapter 1
Rearranging terms and canceling a factor of 1 2 , we obtain the KramersKronig relation in its complexvalue form:
( ) =
+ ( ' ) d ' . j ( ' )
D.RamacciaandA.Toscano
Pag.55
Exercises Chapter 1
1.13 Exercise
An isotropic homogeneous lossless dielectric medium is moving with uniform velocity
v
with
respect to a fixed coordinate frame S. In the frame S' moving with dielectric, the constitutive relations are assumed to be the usual ones, that is, D ' = E ' and B' = H' . Using the Lorentz transformations:
E' = ( E + c B ) H ' = ( H c D )
1 B ' = B E c
E'/ / B '/ /
1 D' = D + H c
H '/ / D'/ /
(1.13.1)
= E/ / = B/ /
1
2
= H// = D/ /
D = E + av ( H v E ) B = H av ( E + v H )
(1.13.2) (1.13.3)
where a =
0 0 . 1 v 2
Solution
It is possible to express the constitutive relations D ' = E ' and B' = H' as follow:
(1.13.4)
vector v . Considering the first equation of set (1.13.4), we can substitute to all of component with the superscript with the correspondent definition given by the Lorentz transformation:
D.RamacciaandA.Toscano
Pag.56
Exercises Chapter 1
)
(1.13.5)
D + H + D / / = ( E + c B ) + E / / c D + H + D / / = ( E + c B ) + E / /
c c
D + H + D / / = ( E + c B ) + E / /
Now it is possible to substitute c = v and c = v c 2 and separate the component parallel and perpendicular at the velocity vector v : 1 D + 2 v H = E + v B c D = E // // ( component) ( // component) (1.13.6)
The relation between the parallel components of vector D and vector E is the same in the two frame S and S'. On the contrary, the perpendicular component depends on both electric and magnetic field. In the fixed frame S the constitutive relation B = H is not valid, so we have to evaluate it using the set (1.13.1) and the constitutive relation B' = H' in the frame S' where it is valid. Considering only the perpendicular component, we have:
B ' = H '
B E = ( H c D )
1 B E = H c D c 1 B = H + E c D c
1 c
(1.13.7)
D.RamacciaandA.Toscano
Pag.57
Exercises Chapter 1
The double cross product v v D can be evaluated using the BAC-CAB rule:
v v D = v ( v D ) D ( v v ) = 0 v 2 D = v 2 D
so 1 D v2D = E + 2 v H + 2 v v E c c
(1 v2 ) D = E + ( 00 ) v H + c2 v v E
D =
(1.13.8)
(1 v )
2
E +
( 00 ) v H
(1 v )
2
c 1 v2
2
v v E
(1 v )
2
E + av H +
c 1 v
2
v v E
(1.13.9)
Comparing eq. (1.13.9) and eq. (1.13.2), we can note that the coefficient of E and v v E are different. It is possible to think that probably we have to add and subtract a unknown quantity. To find it, we will compare the our expression and the expression suggested in eq. (1.13.2), that is:
E + v v E 2 2 1 v c 1 v 2 E 0 0 v v E 1 v 2
(1 v )
2
0 0
=a.
Now we can identify with X the unknown quantity and we can impose that:
E + XE = E 2 1 v 0 0 v v E + X E = v v E 2 2 c 1 v 1 v 2
D.RamacciaandA.Toscano
Pag.58
Exercises Chapter 1
+X = 1 v 2 0 0 v2 + X = + v 2 c 2 1 v 2 1 v 2
(1.13.10)
where in the second equation is present the term v 2 because of v v E = v 2 E . From the first of (1.13.10), we obtain that:
X =
1 v 2
+1 v 2 1
1 v 2
v 2
1 v 2
(1.13.11)
v2
2 2
) (1 v2 ) (1 v2 )
0 0 + 0 0 =
v2
0 0
v2
1 2 + 0 0 = c
2
(1 v )
(1 v2 )
(1.13.12)
It is easy to note that X has the same module, but opposite sign. This confirms our argument of finding a quantity to add and subtract at eq. (1.13.9). So finally we can write:
D = =
1 v2
E + av H +
c2 1 v2
v v E + XE XE =
(1 v )
2
E + av H +
c 1 v
2
v v E +
(1 v )
2
v2
(1 v )
2
v2
(1.13.13)
1 v
v 2
2
1 v
E =
(1 v )
2
(1 v 2 ) E = E
E = v v E =
(1.13.14)
c2 1 v2 c 1 v
2
( (
) )
v v E + v v E
v 2
1 v 2
2
(1 v )
(1 v2 )
=
1 2 v v E c
(1.13.15)
0 0
1 v 2
v v E = av v E
D.RamacciaandA.Toscano
Pag.59
S.J. Orfanidis Electromagnetic Waves and Antennas Substituting (1.13.14) and (1.13.15) in eq. (1.13.13), we obtain:
D = E + av H av v E
Exercises Chapter 1
(1.13.16)
Now using the second equation of (1.13.6) and eq. (1.13.16), it is possible write the expression of vector D in the fixed frame S:
D = D + D/ / = E + a v H av v E + E / / = = ( E + E / / ) + av H av v E + av H / / a v v E / / = (1.13.17)
null null
= E + av H av v E = E + a v ( H v E )
It is easy to note that eq. (1.13.17) and eq. (1.13.2) are identical. In the same way, we can demonstrate eq. (1.13.3), but we kwon that the relations are dual and we can obtain eq. (1.13.3) just operating a changing of variables as follow:
D B E H v H v E
So:
D = E + av ( H v E ) B = H av ( E + v H )
(1.13.18)
D.RamacciaandA.Toscano
Pag.60
Exercise book
Chapter 2: Uniform Plane Waves
Sophocles J. Orfanidis1 Davide Ramaccia2 Alessandro Toscano2
Department of Electrical & Computer Engineering Rutgers University, Piscataway, NJ 08854 orfanidi@ece.rutgers.edu www.ece.rutgers.edu/~orfanidi/ewa
Department of Applied Electronics, University "Roma Tre" via della Vasca Navale, 84 00146, Rome, Italy davide.ramaccia@gmail.com alessandro.toscano@gmail.com
Exercises Chapter 2
Table of Contents
Chapter 2 Uniform Plane WavesEquation Chapter 2 Section 1 ............... 1
2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8 2.9 Exercise ........................................................................................................ 1 ExerciseEquation Section (Next) ................................................................. 3 ExerciseEquation Section (Next) ................................................................. 4 ExerciseEquation Section (Next) ................................................................. 9 ExerciseEquation Section (Next) ............................................................... 11 ExerciseEquation Section (Next) ............................................................... 11 ExerciseEquation Section (Next) ............................................................... 12 ExerciseEquation Section (Next) ............................................................... 14 ExerciseEquation Section (Next) ............................................................... 18
2.10 ExerciseEquation Section (Next) ............................................................... 20 2.11 ExerciseEquation Section (Next) ............................................................... 21 2.12 ExerciseEquation Section (Next) ............................................................... 23 2.13 ExerciseEquation Section (Next) ............................................................... 24 2.14 ExerciseEquation Section (Next) ............................................................... 26 2.15 ExerciseEquation Section (Next) ............................................................... 28 2.16 ExerciseEquation Section (Next) ............................................................... 37 2.17 ExerciseEquation Section (Next) ............................................................... 39 2.18 ExerciseEquation Section (Next) ............................................................... 40 2.19 ExerciseEquation Section (Next) ............................................................... 42
Exercises Chapter 2
2.20 ExerciseEquation Section (Next) ............................................................... 44 2.21 ExerciseEquation Section (Next) ............................................................... 46 2.22 ExerciseEquation Section (Next) ............................................................... 48 2.23 ExerciseEquation Section (Next) ............................................................... 49 2.24 ExerciseEquation Section (Next) ............................................................... 53 2.25 ExerciseEquation Section (Next) ............................................................... 55 2.26 ExerciseEquation Section (Next) ............................................................... 58 2.27 ExerciseEquation Section (Next) ............................................................... 62 2.28 ExerciseEquation Section (Next) ............................................................... 66 2.29 ExerciseEquation Section (Next) ............................................................... 69
Exercises Chapter 2
(2.1.1)
(2.1.2)
(2.1.3)
Thus,
Solution
First of all, we have to evaluate the derivates:
D.RamacciaandA.Toscano
Pag.1
Exercises Chapter 2
(a) (b)
(2.1.4)
(c) (d)
so, multiplying eq. (a) and (c) of (2.1.4), we have 2 z = and, multiplying eq. (b) and (d) of (2.1.4), we have c2t = . Now we can substitute them inside eq. (2.1.1) to obtain: 2 2 + that is:
2 2 E ( , ) = 0
E ( , ) = 0
Using the relationships (a) and (b) of (2.1.4), we can rewrite the forward equation in (2.1.2) as follow:
E + E = +
(2.1.5)
In the same way, using the relationships (c) and (d) of (2.1.4), the backward equation in (2.1.2) becomes:
E E =
and, consequently
E =0
(2.1.6)
D.RamacciaandA.Toscano
Pag.2
Exercises Chapter 2
pulse is launched towards the positive zdirection. Determine expressions for E ( z, t ) and H ( z, t ) and sketch them versus z at any given t.
Solution
For a forwardmoving wave, we have E ( z, t ) = F ( z ct ) = F ( 0 c ( t z c ) ) , which implies that
(2.2.1)
Fig. 2.2.1: Expanding wavefont at time t and t + t . Because of the unit step, the nonzero values of the fields are restricted to t z c 0 , or, z ct , that is, at the time t the wavefront has propagated only up to the position z = ct . Fig. 2.2.1 shows the expanding wavefronts at time t and t + t .
D.RamacciaandA.Toscano
Pag.3
Exercises Chapter 2
E 0 = z H z j
j E 0 H z
(2.3.1)
where the matrix is meant to apply individually to the x, y components of the vector entries. Show that the following similarity transformation diagonalizes the transition matrix, and discuss its role in decoupling and solving the above system in terms of forward and backward waves:
1 Z0 0 1 Z0 j j 1 Z0 0 1 Z0
1
jk = 0
0 jk
(2.3.2)
where k = c , c = 1
, and Z0 = .
Solution
For a singlefrequency wave, we can assume a timedependence as e j t . So the electric and magnetic field can be expressed as E ( x, y, z, t ) = E ( x, y, z ) e j t , H ( x, y, z, t ) = H ( x, y, z ) e j t , respectively. The Maxwell's equations can be written in the form: H( x,y,z) ejt jt E( x,y,z) e = t E( x,y,z) ejt jt H( x,y,z) e = t
) )
(2.3.3)
Evaluating the derivate in the righthand side of both equations, the term e jt can be simplified:
E ( x, y, z ) e j t = j H ( x, y, z ) e j t H ( x, y, z ) e j t = j E ( x, y, z ) e j t
The curl of the electric (or magnetic) field can be written as determinant of the following matrix:
x E = x y y z z
where i with i = x, y, z are the partial derivates. But we are in presence of an uniform plane waves propagating along zdirection, so the electric filed vector lies on the xy plane, i.e. the z D.RamacciaandA.Toscano Pag.4
Exercises Chapter 2
component of the electric field is null( Ez = 0 ), and also in each plane the vector has constant amplitude, i.e. the derivates along x and y are null. The curl becomes: x y z E = 0 0 z E x (x, y, z) E y (x, y, z) 0 The only applicable derivate is z , so (2.3.4) is similar to:
x E = 0 E x (x, y, z) z y 0 E y (x, y, z) z z 1 0
(2.3.4)
(2.3.5)
E ( x, y, z ) : z
E ( x, y,z ) z H ( x, y,z ) z
= j H ( x, y,z ) = + j E ( x, y,z )
(2.3.6) (2.3.7)
Consider eq. (2.3.6) and apply the crossproduct with z to both of side:
E z z = j H z z
(2.3.8)
where we used the condition zEz = 0 for a plane wave. So eq.(2.3.8) can be written as follow:
E = j H z z
(2.3.9)
On the contrary, eq. (2.3.7) needs only to invert the cross product at the lefthand side and to change the sign at the righthand side:
H z = j E z
(2.3.10)
Now it is easy to write in matrix form eq. (2.3.9) and (2.3.10) to obtain (2.3.1).
D.RamacciaandA.Toscano
Pag.5
Exercises Chapter 2
The transition matrix has to be diagonalized and we need of eigenvectors to create the matrix for the base change. The eigenvectors are found from the eigenvalues that are the roots of the characteristic polynomial:
Det ( A I ) = 0
0 where A = j
So we have:
which gives:
1 = j = jk 2 = + j = + jk
It has two separate eigenvalues, so it is diagonalizable. The diagonal matrix is simply:
1 0 jk 0 0 = 0 + jk 2
(2.3.11)
that is the righthand side of (2.3.2). The lefthand side is composed by the product of the matrix
A and two matrixes for the base change. These matrixes are made putting in row the eigenvectors
1 and 2 calculated as follow: A i = i i
with i = 1, 2 and i = i1 i2
Expanding it two different systems of equations, one for each eigenvalue, we have:
0 j 0 j j 11 11 = 1 0 12 12 j 21 21 = 2 0 22 22
or equivalently,
j12 = 1 11 j11 = 1 12 j 22 = 2 21 j 21 = 2 22
(2.3.12)
(2.3.13)
Solving (2.3.12) and (2.3.13), we find that the eigenvectors are given by: D.RamacciaandA.Toscano Pag.6
Exercises Chapter 2
(2.3.14)
Now we can write the matrix for the base change as:
T 1 11 12 1 Z0 = = 1 Z0 T 2 21 22
(2.3.15)
where the superscript T indicate the vector transpose. In order to verify that eq. (2.3.2) is correct, we have to calculate the inverse of the matrix (2.3.15):
P
1 1
1 Z0 = 1 Z0
where Cij (P) is the matrix of cofactors of P. The cofactor in position ( i, j) is defined as follow:
Cij (P) = ( 1)
i+ j
where Minor ( P,i, j) represents the matrix obtained by P cancelling the ith row and jth column. It is easy to evaluate it:
Z Cij (P ) = 0 Z0 1 1
(2.3.16)
1 Z0 2Z0 1
12 Z0 1 2 = 1 ( 2Z ) 1 ( 2Z ) 1 0 0
So: 1 Z0 0 PAP 1 = 1 Z0 j j Z0 = j Z0 12 j 1 2 1 ( 2Z ) 1 ( 2Z ) = 0 0 0
12 j 1 2 = j 1 ( 2Z0 ) 1 ( 2Z0 )
(2.3.17)
D.RamacciaandA.Toscano
Pag.7
S.J. Orfanidis Electromagnetic Waves and Antennas where k = and (2.3.17) is equivalent to (2.3.2). The diagonal matrix given in (2.3.17) can be substituted in (2.3.1) as follow: E jk 0 E = z H z 0 + jk H z that is z E = jkE H z = + jkH z z case vacuum, as follow:
Exercises Chapter 2
(2.3.18)
(2.3.19)
The electric and magnetic field are related by the characteristic impedance of the medium, in this
E = Z0 H k = Z0 H z
where k = z , being the electromagnetic wave propagating along zdirection.
(2.3.20)
(2.3.21)
It is possible to note that the diagonalization (2.3.2) allow us to decouple the electric and the magnetic field as in (2.3.19) and, using the relationship (2.3.20), we are able to express the electric field in term of forward and backward wave.
D.RamacciaandA.Toscano
Pag.8
Exercises Chapter 2
Solution
The wavelength is the distance by which the phase of the sinusoidal wave changes by 2
radians. Since the propagation factor e jkz accumulates a phase of k radians per meter, we have by
definition that k = 2 . The wavelength can be expressed via the frequency of the wave in
Hertz, f = 2 ,asfollows:
2 2 c 2 c c = = = k 2 f f
(2.4.1)
Using the relation (2.4.1), we can calculate the frequency range for the electromagnetic visible spectrum:
c 380 109
>f >
c 780 109
(2.4.2)
D.RamacciaandA.Toscano
Pag.9
Exercises Chapter 2
D.RamacciaandA.Toscano
Pag.10
Exercises Chapter 2
wavelength of 20 m?
Solution
Using eq. (2.4.1), it is easy to obtain the result:
f= c 3 108 20 10
6
= 15 1012 = 15 THz
Solution
Using eq. (2.4.1), we have:
c 3 108 = = 30000 m = 30 Km f 10 103 c 3 108 = = 30 m f 10 106 c 3 108 = = 0.030 m = 30 mm f 10 1012
3 108 21 102
The frequency in GHz of the 21cm hydrogen line observed in the cosmos is:
f= c
= 1.43 GHz
The wavelength in cm of the typical microwave oven frequency of 2.45 GHz is:
D.RamacciaandA.Toscano
Pag.11
Exercises Chapter 2
and (you may assume they are both positive.) By inserting E ( z, t ) into Maxwell's equations,
determine the k relationship as a consequence of these equations. Determine also the magnetic
field H ( z, t ) and verify that all Maxwell's equations are satisfied. Repeat the problem if
E ( z, t ) = xE 0e j t + jkz
and
E ( z, t ) = yE 0e j t jkz
Solution
Consider the sourcefree Maxwell's equations:
B E = t H = D t
and now substitute the expression of E ( z, t ) in the first Maxwell's equation. Assuming valid the constitutive relation B = H :
xE 0e j t jkz = H t
(2.7.1)
The crossproduct in the lefthand side of eq. (2.7.1) can be expanded as follow:
yjkE 0e j t jkz = H t
It is easy to note that H (z, t) has to depend by z and t in the same way as E ( z, t ) because the
variables t and z are presented only in the exponential. So H(z, t) = yH0e j t jkz :
D.RamacciaandA.Toscano
Pag.12
Exercises Chapter 2
kE0 = H0
(2.7.2)
Assuming valid the constitutive relation D = E , substitute the expression of E ( z, t ) and H (z, t) in the second Maxwell's equation:
yH0e j t jkz = xE 0 e j t jkz t
(2.7.3)
The crossproduct in the lefthand side of eq. (2.7.3) and the derivate in the righthand side can be expanded as follow:
x H 0e j t jkz + z H 0e j t jkz = j xE 0e j t jkz z x
which gives:
kH0 = E0
(2.7.4)
Thanks to the relationships (2.7.2) and (2.7.4) we can find the k relation:
kE 0 = H 0 kH 0 = E 0
k 2 E 0 H 0 = 2 E 0 H 0
and, consequently:
k =
(2.7.5)
D.RamacciaandA.Toscano
Pag.13
Exercises Chapter 2
3x y e jkz
E = E 0 ( jx y ) e jkz E = E 0 ( x + 2 jy ) e jkz
(2.8.1)
Solution
The polarization of a plane wave is defined to be the direction of the timevarying realvalued field E ( z, t ) = e E ( z ) e j t where E ( z ) = E0e jkz . At any fixed point z, the vector E ( z, t ) may be along a fixed linear direction or it may be rotating as a function of t, tracing a circle or an ellipse. Consider the following expression for the electric field:
E ( z, t ) = xA x e jx + yA ye
jy
)e
j t jkz
= xA x e (
j t jkz+x )
+ yA y e
j t jkz+y
) (2.8.2)
where Ax and Ay are realpositive quantities. Extracting the real part for each component, we find the corresponding realvalued x,y components:
E x ( z, t ) = A x cos ( t kz + x ) E y ( z, t ) = A y cos t kz + y
(2.8.3)
The sign of kz is defined by the direction of propagation of the wave: forwardmoving fields have the minus sign, e.g. kz , backwardmoving fields the plus sign, e.g. +kz . In order to determine the polarization type of the waves, we consider the timedependence of these fields at some fixed point along zaxis. For convenience we choose z = 0 :
E x ( z, t ) = A x cos ( t + x ) E y ( z, t ) = A y cos t + y
(2.8.4)
The parameters Ax, Ay, x , y allow us to determine the type of polarization: Linear polarization ( a = b = 0 or a = 0 , b = ): the two components Ex , Ey are in phase and the electric field vector oscillates along a straight line. It is of interest the direction,
D.RamacciaandA.Toscano
Pag.14
Exercises Chapter 2
respect the xaxis, along which the electric field oscillates with angular frequency . It is directly related to the amplitudes of the components Ex , Ey :
= arctan
Ey slope Ay/Ax
Ay Ax
(2.8.5)
Ex Ex Ey slope -Ay/Ax
Fig. 2.8.1:Directions along the electric field oscillates in linear polarization. Elliptical polarization ( x y = 2 ): Ex and Ey have different amplitudes and are in quadrature phase because one is always 90 out of phase respect to other. Circular polarization ( A x = A y and x y = 2 ): this is a particular case of elliptical polarization when the amplitudes of the components are equal. The sign of the relative phase = x y suggests the sense of rotation: counterclockwise ( = 2 ) and clockwise ( = + 2 ) and consequently, according to the direction of propagation, left or right elliptical polarization (or circular in particular cases).
D.RamacciaandA.Toscano
Pag.15
Exercises Chapter 2
Ex(t) t Ey(t)
Fig. 2.8.3: Clockwise rotation of the electric field vector. To decide whether this represents a right or left polarization, we use the IEEE convention. Curl the fingers of your left and right hands into a fist and point both thumbs towards the direction of propagation. If the fingers of your right (left) hand are curling in the direction of rotation of the electric field, then the polarization is right (left) polarized. y E rightpolarized x z E y leftpolarized x z
y E -z
leftpolarized E -z
rightpolarized
the field in its realvalued form in z = 0 , in order to obtain an expression similar to eq. (2.8.4): E ( t, z ) = E 0e ( x + y ) e ( so
j t kz )
= E 0 x cos ( t kz ) + E 0 y cos ( t kz )
E x ( t, z ) = E0 cos ( t kz ) E y ( t, z ) = E0 cos ( t kz )
D.RamacciaandA.Toscano Pag.16
Exercises Chapter 2
It is easy to note that it is a forwardmoving wave, because of the term kz , and linear polarized, being x = y = 0 . Using eq. (2.8.5), the direction of electric field vector is 45.
On the contrary, let us solve the exercise for the case (c): E = E 0 ( jx + y ) e jkz . Its realvalued
form in z = 0 is:
j t kz ) = E 0e xe j 2 + y e j( t kz ) = E ( t, z ) = E 0e ( jx + y ) e ( = E 0 x cos ( t kz + 2 ) + E 0 y cos ( t kz )
so
E x ( t, z ) = E0 cos ( t kz + 2 ) E y ( t, z ) = E0 cos ( t kz )
The wave is still forwardmoving, but the relative phase = x y = 2 , so it is in general elliptical polarized. In this case E x = E y , i.e. A x = A y , then it is circular polarized. According to Fig. 2.8.2 and Fig. 2.8.3, the sense of rotation is counterclockwise. Now we apply the IEEE convention and the find that the field is rightcircular polarized. Table 2.8.1 contains the results of the exercise for each given electric field:
# a b c d e f g h Expression Polarization Type Linear Linear Circular Elliptical Linear Linear Circular Elliptical Direction/ Sense of Rotation 45 -60 Counterclockwise Counterclockwise -45 -30 Clockwise Clockwise
E = E 0 ( x + y ) e jkz
E = E 0 x 3y e jkz
3x y e jkz
E = E 0 ( jx y ) e jkz E = E 0 ( x + 2 jy ) e jkz
D.RamacciaandA.Toscano
Pag.17
Exercises Chapter 2
E ( z, t ) = 2xCos ( t kz ) + 4ySin ( t kz )
(2.9.1)
1. Determine the vector phasor representing E ( z, t ) in the complex form E = E0e j t jkz . 2. Determine the polarization of this electric field (linear, circular, elliptic, lefthanded, righthanded). 3. Determine the magnetic field H ( z, t ) in its realvalued form.
Solution
Question n1 First of all, we need to manipulate (2.9.1) in order to obtain an expression with components similar to:
E x ( z, t ) = A x cos ( t kz + x ) E y ( z, t ) = A y cos t kz + y
(2.9.2)
So we can write:
E x ( z, t ) = 2cos ( t kz ) E y ( z, t ) = 4cos ( t kz 2 )
from which, we can obtain the complexvalued electric field:
E ( z, t ) = x 2e (
j t jkz )
(2.9.3)
+ y 4e (
j t jkz 2 )
= 2x + 4ye j
) e jt jkz =
= ( 2x j4y ) e (
j t kz )
Question n2
In (2.9.3) it is easy to note that the relative phase = x y = 2 , so according to Fig. 2.8.2 and Fig. 2.8.3, the sense of rotation is counterclockwise. The field is forwardmoving and, using the IEEE convention, the field is rightelliptical polarized, being E x E y . Question n3 Using the relation:
E = Z0 ( H z )
where Z0 is the characteristic impedance of vacuum, we find:
D.RamacciaandA.Toscano
Pag.18
Exercises Chapter 2
1 1 z E ( z, t ) = z E x ( z, t ) x + E y ( z, t ) y = Z0 Z0 1 E x ( z, t ) y E y ( z, t ) x = Z0 1 ( 2 cos ( t kz ) y 4 cos ( t kz 2 ) x ) Z0
D.RamacciaandA.Toscano
Pag.19
Exercises Chapter 2
E x = cos ( t kz ) ,
E y = 2sin ( t kz )
(2.10.1)
1. Its phasor form has the form E = ( Ax + By ) e jkz . Determine the numerical values of the
complexvalued coefficients A, B and the correct sign of the exponent. 2. Determine the polarization of this wave (left, right, linear, etc.). Explain your reasoning.
Solution
Question n1 First of all, we need to manipulate (2.10.1) in order to obtain an expression with components similar to (2.9.2):
E x ( z, t ) = cos ( t kz ) E y ( z, t ) = 2cos ( t kz 2 )
from which, we can obtain the complexvalued electric field:
E ( z, t ) = xe (
j t jkz )
+ y 2e (
j t kz )
j t jkz 2 )
= x + 2ye j
) e jt jkz =
= ( x j2y ) e (
Question n2
The polarization of the wave is elliptical because the module of x, y components are different. It is also right polarized because the relative phase = x y = 2 .
D.RamacciaandA.Toscano
Pag.20
Exercises Chapter 2
For both cases, determine the polarization of the wave (linear, circular, left, right, etc.) and the
Solution
Case (a) First of all, we rewrite the first field of (2.11.1) as follow:
E x ( z, t ) = cos ( t + kz 2 ) E y ( z, t ) = 2cos ( t + kz )
from which, we can obtain the complexvalued electric field:
E ( z, t ) = xe (
j t + jkz 2 )
(2.11.2)
+ y 2e (
j t + jkz )
= xe j
+ 2y e j t + jkz =
= ( 2 y jx ) e (
j t + kz )
In (2.11.2) it is easy to note that the relative phase = x y = 2 , so according to Fig. 2.8.2 and Fig. 2.8.3, the sense of rotation is clockwise. The field is backwardmoving and, using the IEEE convention, the field is rightelliptical polarized ( E x E y ). Case (b)
= e ( (1 + j) x (1 j) y ) ( cos ( t kz ) + jsin ( t kz ) ) = = ( x y ) cos ( t kz ) ( x + y ) sin ( t kz ) = = ( x y ) cos ( t kz ) ( x + y ) cos ( t kz 2 ) = = x ( cos ( t kz ) cos ( t kz 2 ) ) y ( cos ( t kz ) + cos ( t kz 2 ) )
(2.11.3)
D.RamacciaandA.Toscano
Pag.21
Exercises Chapter 2
+
2
cos
(2.11.4)
(2.11.5)
and it is easy to note that the relative phase = x y = 2 , so according to Fig. 2.8.2 and Fig. 2.8.3, the sense of rotation is clockwise. The field is forwardmoving and, using the IEEE convention, the field is leftcircular polarized ( E x = E y ). In similar way, the exercise can be solved writing the complex amplitude of each component in the form:
2 2 a + jb = Me j where M = a + b = arctan b a
(2.11.6)
and, consequently
E ( z, t ) = e
j t kz ) = 2e j 4 x + 2e j3 4 y e (
j t kz ) = = 2 e e j 4 x + e j3 4 y e ( = 2xcos ( t kz + 4 ) + 2y cos ( t kz + 3 4 )
D.RamacciaandA.Toscano
Pag.22
Exercises Chapter 2
E ( t, z ) = x cos ( t kz 4 ) + y sin ( t kz + 4 )
1. Determine the complexphasor form of this electric field. 2. Determine the corresponding magnetic field H ( t, z ) given in its realvalued form. 3. Determine the polarization type (left, right, linear, etc.) of this wave.
(2.12.1)
Solution
Question n1 First of all, we rewrite (2.12.1) as follow:
E x ( z, t ) = cos ( t kz 4) E y ( z, t ) = cos ( t kz 4)
from which, we can obtain the complex form as eq. (2.8.2) j t kz 4 ) + ye j( t kz 4 ) = xe j E ( z, t ) = xe ( = Question n2 1 j t kz ) (1 j) ( x + y ) e ( 2
(2.12.2)
+ ye j
) e jt jkz =
E = Z0 ( H z )
where Z0 is the characteristic impedance of vacuum, we find:
H ( t, z ) = = Question n3
In (2.12.2) it is easy to note that the relative phase = x y = 0 , so the wave is linear polarized, tilted by 45 with respect to the xaxis.
D.RamacciaandA.Toscano
Pag.23
Exercises Chapter 2
E ( z ) = 1 + j 3 x + 2y e jkz
E ( z ) = (1 + j) x (1 j) y e jkz
jk x + z E ( z ) = x z + j 2y e ( )
2
Solution
Case (a) We have to writing the complex amplitude of each component Ex , Ey in the form Ae j , using (2.11.6):
(1 + j 3 ) =
so:
(1)2 + (
3 e
2 jarctan
31
) = 2e j 3
E ( z ) = 2 xe j
+ y e jkz
The relative phase = x y = 3 , so it is inside the interval [ 0, 2] and, according to Fig. 2.8.2 and Fig. 2.8.3, the sense of rotation is counterclockwise. The field is backwardmoving and, using the IEEE convention, the field is leftcircular polarized, being E x = E y . Case (b) Case (c)
See case (b) of exercise n 2.11. In this case the electric field doesn't propagate along the z direction, but it is tilted with respect to the z axis and lies on the z-x plane. So we have to identify a new coordinate system in order to apply the wellknown steps to solve the problem. Express the electric field in the realvalued form:
jk x + z 2 E ( t, z ) = e x z + j 2y e j t e ( ) = = e x z + j 2y cos t k ( x + z ) 2 + jsin t k ( x + z )
( (
) )( (
= ( x z ) cos t k ( x + z )
D.RamacciaandA.Toscano
2 2y sin t k ( x + z )
2 =
))
Pag.24
Exercises Chapter 2
This wave does not propagate along the zdirection, and consequently the plane with constant phase are not identified for any constant value of z. So we have to apply a change of coordinate system (x, y, z) (x ', y, z ') where the y axis is the same because E ( t, z ) has constant phase for any y. z z 45
x Fig. 2.13.1: Rotation of the coordinate system. The expression of x' and z' are given by:
x + z 2 = z' x z = x' 2 z ' x ' z = 2 x = x '+ z ' 2
and we can rewrite the electric field in the new coordinate system:
E ( t, z ) = 2x 'cos ( t kz ') 2y sin ( t kz ' ) = = 2x 'cos ( t kz ' ) 2y cos ( t kz ' 2 ) = = 2x 'cos ( t kz ' ) + 2y cos ( t kz '+ 2 ) =
The relative phase = x y = 2 , so according to Fig. 2.8.2 and Fig. 2.8.3, the sense of rotation is clockwise. The field is forwardmoving and , using the IEEE convention, the field is leftcircular polarized , being E x = E y .
D.RamacciaandA.Toscano
Pag.25
Exercises Chapter 2
E ( t, z ) = Ax cos ( t kz + a ) + By cos ( t kz + b )
Show that:
(2.14.1)
(2.14.2)
Solution
The cross product of two vectors A = A x , A y , A z determinant of the following matrix:
x A B = Ax Bx y Ay By z Az Bz
and B = B x , B y , B z
is defined as the
(2.14.3)
x E ( t + t, z + z ) E ( t, z ) = A cos a A cos a
y Bcos b Bcos b
z 0 = Det ( M ) 0
Det ( M ) = z ABcos b cos a ABcos a cos b = = zAB cos b cos a cos a cos b cos ( ) + cos ( + ) 2
(2.14.4)
The expression inside the brackets can be simplified using the producttosum identity for cosine:
cos cos =
and, consequently, (2.14.4) can be written as:
(2.14.5)
D.RamacciaandA.Toscano
Pag.26
Exercises Chapter 2
Det ( M ) = zAB cos b cos a cos a cos b = 1 + 2 cos (t kz + b a ) + cos ( 2 t 2kz + t kz + b + a ) + = zAB = 1 cos (t kz + ) + cos ( 2 t 2kz + t kz + + ) a b b a 2 1 = z AB cos (t kz + b a ) cos (t kz + a b ) 2
( (
) )
The expression inside the brackets can be still simplified using now the productto sum identity for sine:
sin sin =
and, consequently, we have:
cos ( ) cos ( + ) 2
(2.14.6)
D.RamacciaandA.Toscano
Pag.27
Exercises Chapter 2
E2 y B
2
ExEy AB
cos = sin 2
(2.15.1)
(2.15.2)
Fig. 2.15.1: General polarization ellipse. one must determine the tilt angle such that the following matrix condition is satisfied:
cos sin 1 sin A 2 cos cos AB cos AB cos 1 sin 2 B 1 2 sin A = sin 2 cos 0 0 (2.15.3) 1 B2
tan 2 =
2AB A 2 B2
cos
(2.15.4)
( A2 B2 2 )( B2 A2 2 ) = A2B2 sin2 2 (1 2 )
Using this property, show that the semiaxes A', B' are given by the equations:
A = 2 A 2 B2 2 1 2 , B = 2 B2 A 2 2 1 2
(2.15.5)
(2.15.6)
Exercises Chapter 2
2
A = B =
1 2 s A + B2 + 2 2 1 2 s A + B2 2 2
( A 2 + B2 ) ( A 2 + B2 )
+ 4A 2 B2 cos 2
(2.15.7)
+ 4A 2 B2 cos 2
where s = sign(A B) .Finally, show that A', B' satisfy the relationships:
A2 + B2 = A 2 + B2 , A B = AB sin
(2.15.8)
Solution
The polarization ellipse in eq. (2.15.1) can be written in matrix form as follow:
1 2 Ex E y A cos AB cos AB E x = sin 2 1 E y B2
(2.15.9)
and, from the matrix form of (2.15.2), it is possible to obtain its inverse:
E cos x E = y sin
(2.15.10) into (2.15.9):
cos Ex Ey sin
sin E x cos E y
E x cos E = y sin
Noting that the first vector in (2.15.9) is the transposed of the one in (2.15.10), we can substitute
1 sin A 2 cos cos AB cos AB cos 1 sin 2 B
(2.15.11)
Eq. (2.15.11) represents the tilted ellipse shown in Fig. 2.15.1. The ellipse is not rotated with respect to the axes E , Ey and it is possible to define new values of the minor and major axis in x this rotated coordinate system. As suggested from Fig. 2.15.1 the minor axis is B and the major axis is A and the equation of the tilted ellipse can be rewritten as:
1 2 Ex Ey A 0 0 E x =1 1 Ey B2
(2.15.12)
D.RamacciaandA.Toscano
Pag.29
Exercises Chapter 2
0 E x = sin 2 1 Ey B2
1 2 sin = sin 2 A cos 0
(2.15.13)
and comparing eq. (2.15.11) and eq. (2.15.13), we note they are equal if and only if:
cos sin cos AB cos 1 sin 2 B 0 (2.15.14) 1 B2
The relationship (2.15.14) is fundamental to solve the whole exercise. Firstly, it can be manipulated in order to demonstrate eq. (2.15.4). Left multiplying both side by:
1
cos sin
we obtain: cos sin sin cos
1
sin cos
cos sin
1 1 2 sin A cos 0
cos = sin
sin cos
1 sin A2 cos 0 0 1 B2
(2.15.16)
Let us now divide both side of (2.15.16) by cos and, defining = tan , we get:
cos 1 1 1 A2 AB = sin 2 1 1 1 0 B2 0 1 B2
(2.15.17)
S.J. Orfanidis Electromagnetic Waves and Antennas 1 cos sin 2 = 2 AB A 2 A 1 cos sin 2 + = AB B2 B2 sin 2 cos + 2 = AB B A 2 cos sin 2 2 = B2 AB A
1 cos sin 2 = 2 AB A A 2 1 cos sin 2 + = B2 AB B2 cos + = 1 cos 2 AB B2 AB A cos cos 1 AB 2 = 2 + AB A B
cos cos 1 + 2 = 2 AB B AB A cos cos + 2 = 2 2 AB B AB A cos cos 2 = 2 2 AB A AB B cos cos 1 2 = 2 + AB A AB B
Exercises Chapter 2
(2.15.19)
(c) (d)
It is easy to show that (2.15.19) (c) and (2.15.19) (d) are equivalent. In fact:
(c) (d)
)
Pag.31
D.RamacciaandA.Toscano
Exercises Chapter 2
(2.15.20)
It is known that: tan 2 = 2 so (2.15.20) can be written as tan 2 = which is the same of eq. (2.15.4). Eq. (2.15.19)(c), or (d), can be viewed as a quadratic equation in : A 2 B2 cos cos + 2 2 =0 A B AB AB 2 1
2
tan 1 tan
2
=2
1 2
2AB A 2 B2
cos
(2.15.21)
2
with its solution given by:
(2.15.22)
1,2 =
A 2 B2 2 2 A B
A 2 B2 cos 2 +4 2 2 2 2 A B A B = cos 2 AB
A 2 B2 1 2 2 2 2 A B A B =
( A 2 B2 )
cos 2 AB
2
+ 4A 2 B2 cos 2 = (2.15.23)
= where:
B2 A 2
( A 2 B2 )
+ 4A 2 B2 cos 2
2ABcos
B2 A 2 + sD = s 2ABcos
s = 1 ;
D=
( A 2 B2 )
+ 4A 2 B2 cos 2 =
( A 2 + B2 )
4A 2 B2 sin 2 ;
The choice between s = 1 and s = 1 does not matter because both are solution of eq. (2.15.22). So, for simplicity, we can set the value of s as:
s = sign ( A B)
(2.15.24)
which is the same value set in the text of the exercise. This choice is according to the fact that in general one defines the major axis with A and the minor axis with B and we have that s = 1 .
D.RamacciaandA.Toscano Pag.32
Exercises Chapter 2
The values s , s satisfy the identities (2.15.19) (c) and (2.15.19) (d) and so, if we substitute them inside (2.15.19) (a) or (b), we are able to eliminate in the definition of A , B and A, B. We want to remark that the direct substitution is not a good choice because it leads to an expression difficult to be managed. So it is better to follow a longer way to solve the problem, but easier to be understood. Let us now stop to consider the linear system (2.15.19) and let us diagonalize the matrix:
1 2 A cos AB cos AB 1 B2 1 2 A 0 0 1 B2
diagonalize
The eigenvalues are the roots of the characteristic polynomial: 1 2 Det A cos AB that is
1 1 cos 2 2 2 2 = 0 2 A B A B =0 B2 A 2 B2 A 2 + B2 1 2 2 2 + 2 2 1 cos 2 = 0 A B A B A 2 B2 A2 1
cos AB 1 0 = 0 1 0 1 B2
+ 2
cos 2
( ) 2 ( A 2 B2 ) ( A 2 + B2 ) + (1 cos 2 ) = 0 2 ( A 2 B2 ) ( A 2 + B2 ) + sin 2 = 0
Finally we get:
(2.15.25)
( =
A 2 + B2 + s
) (
A 2 + B2
4A 2 B2 sin 2
2A 2 B2
A 2 + B2 + sD 2A 2 B2
with s = 1 (2.15.26)
s are the eigenvalues, useful to simplify the expressions when substituting s in eq. (2.15.18)
(a) and (b). In fact:
D.RamacciaandA.Toscano
Pag.33
Exercises Chapter 2
(2.15.27)
(2.15.28)
Comparing (2.15.28) with (2.15.19) (a) and (2.15.19) (b), we can write:
sin 2 = s A 2 2 sin 2 = s B
(2.15.29)
and consequently 2 sin 2 sin 2 = A = s s s s 2 2 2 sin sin B = = s s s s The product s s is given by:
(2.15.30)
s s =
( A 2 + B2 ) = ( =
A 2 =
A 2 + B2
A 2 + B2 + sD A 2 + B2 sD = 2A 2 B2 2A 2 B2
2
D2
4A 4 B4
(2.15.31)
) (
2
A 2 + B2
+ 4 A 2 B2 sin 2
4 A 4 B4
sin 2 A 2 B2
s s
sin 2
s =
sin 2 sin A 2 B2
2
1 s = A 2 B2s = A 2 + B2 + sD 2
B2 =
s s
s =
sin 2 sin A 2 B2
2
s = A 2 B2s
1 = A 2 + B2 sD 2
(2.15.32)
D.RamacciaandA.Toscano
Pag.34
Exercises Chapter 2
giving the expressions for A and B as required by the exercise [see eq. (2.15.7)]. In the same way, starting from (2.15.32): A2 = A 2 B2s B2 = A 2 B2s we can substitute the other value of s as in eq.(2.15.28), that is:
1 cos s = 2 s AB A = 1 + cos s B2 s AB
and we have: cos 1 = A 2 + s ABcos A 2 = A 2 B2 2 + s AB B cos 1 2 B2 = A 2 B2 2 s = B s ABcos AB A From eq. (2.15.21) we have also: 2 s = 2AB A 2 B2 cos
(2.15.33)
2 1 s
( A 2 B2 ) s ABcos =
2 1s
(2.15.34)
( A2 B2 ) s = A2 B2 s2 (A
2 1s 2
B2 = B2 s ABcos = B2 s AB
B s
2 1s
(2.15.35)
2 1 s
2 A 2 s 2 1 s
Eq. (2.15.35) gives the expression of A and B as required by the exercise [see eq.(2.15.6)]. Adding A and B , we obtain:
A2 + B2 =
2 A 2 B2 s 2 1 s 2 B2 A 2 s 2 1 s
2 B2 s
+B
2 1s
2 A 2 s
( A2 + B2 ) (1 s2 ) = A2 + B2 =
2 1 s
(2.15.36)
s s = A 2 B2
sin 2 A B
2 2
= A 2 B2 sin 2
AB = AB sin
(2.15.37)
D.RamacciaandA.Toscano
Pag.35
Exercises Chapter 2
It is easy to note that eq. (2.15.36) and (2.15.37) are equal to eq. (2.15.8) in the text of the exercise.
D.RamacciaandA.Toscano
Pag.36
Exercises Chapter 2
E ( 0 ) E ( t ) = zABsin sin t
where = a b . Then prove the more general relationship:
(2.16.1)
E ( t1 ) E ( t 2 ) = zABsin sin ( ( t 2 t1 ) )
Discuss how linear polarization can be explained with the help of this result.
(2.16.2)
Solution
Using (2.14.3), we can write:
x E ( 0) E ( t ) = y z A cos a B cos b 0 A cos ( t + a ) B cos ( t + b ) 0
(2.16.3)
and we have:
E ( 0 ) E ( t ) = z ABcos a cos ( t + b ) ABcos b cos ( t + a ) = = zAB cos a cos ( t + b ) cos b cos ( t + a )
(2.16.4)
The expression inside the brackets can be simplified using the producttosum identity for cosine:
cos cos =
and, consequently, (2.16.4) can be written as:
cos ( ) + cos ( + ) 2
(2.16.5)
E ( 0 ) E ( t ) = zAB cos a cos ( t + b ) cos b cos ( t + a ) 1 + 2 cos (a t b ) + cos (a + t + b ) + = zAB = 1 cos ( t ) + cos ( + t + ) b a b a 2 AB cos (a t b ) cos (b t a ) = =z 2 AB =z cos ( t + ) cos ( t ) 2
( (
) )
D.RamacciaandA.Toscano
Pag.37
Exercises Chapter 2
The expression inside the brackets can be still simplified using now the productto sum identity for sine:
sin sin =
and, consequently, we have:
cos ( ) cos ( + ) 2
(2.16.6)
E ( 0) E ( t ) = z
(2.16.7)
The more general relationship (2.16.2) can be proven in the same way of eq. (2.16.1):
x E ( t1 ) E ( t 2 ) = Det A cos ( t1 + a ) A cos ( t 2 + a ) y z B cos ( t1 + b ) 0 = B cos ( t 2 + b ) 0
= zAB cos ( t1 + a ) cos ( t 2 + b ) cos ( t1 + b ) cos ( t 2 + a ) = AB cos ( t1 + a t 2 b ) + cos ( t1 + a + t 2 + b ) + =z = 2 cos ( t + t ) cos ( t + + t + ) 1 b 2 a 1 b 2 a AB cos ( ( t1 t 2 ) + ) cos ( ( t1 t 2 ) ) = =z 2 AB cos ( ( t 2 t1 ) ) cos ( ( t 2 t1 ) + ) = =z 2 = zABsin sin ( ( t 2 t1 ) )
(2.16.8)
When the electric field is linear polarized, the electric field vector, sampled in any t , is always along a fixed direction, so the angle between the two vectors E ( t1 ) and E ( t 2 ) , represented by the relative phase = a b , is always zero and the crossproduct is null at any t = t 2 t1 .
D.RamacciaandA.Toscano
Pag.38
Exercises Chapter 2
c =
(2.17.1)
where c = j is the complex value of permittivity of a lossy media and k c = j , show the following relationships:
e c 1 = = 2
(2.17.2)
Solution
From the first property, let us express the characteristic impedance as:
c =
and invert it:
c 1 =
kc kc
(2.17.3)
(2.17.4)
Now it is possible to substitute k c j in eq. (2.17.4) and extract the real part: j e c 1 = e =
D.RamacciaandA.Toscano
Pag.39
Exercises Chapter 2
= tan =
+ d = d
(2.18.1)
(2.18.2)
Solution
Using the definition of k c in (2.17.1) and the relationship (2.18.1), we can write:
k c = c = ( j ) = = (1 jtan ) = (1 jtan )
The complexvalued permittivity c is also defined as
12
(2.18.3)
c = j = d j d +
(2.18.4)
where d = d j d is the permittivity of dielectric and its conductivity. So in (2.18.3) we can substitute d and tan sin cos to obtain:
sin k c = d 1 j cos
12
=
12
=
12
12
1 cos
(2.18.5)
12
( cos )1 2 =
In the same way it is possible to express c as in (2.18.2), starting from its definition in (2.17.1):
D.RamacciaandA.Toscano
Pag.40
Exercises Chapter 2
c =
= = = c j j tan
12
1 = d 1 jtan =
12
(2.18.6)
j e d
1 2
( cos )1 2 =
D.RamacciaandA.Toscano
Pag.41
Exercises Chapter 2
Solution
First of all we have to express the complexvalues of the permittivity as follow:
2 2 a + jb = Me j where M = a + b = arctan b a
(2.19.1)
so, using the superscripts 1 and 2 to indicate the permittivity of potatoes and carrots, respectively, we have:
1 c = ( 65 j25 ) 0 = 652 + 252 e jArc tan( 25 65) 2 c = ( 75 j25 ) 0 = 752 + 252 e jArc tan( 25 75)
Using
k c = 0 c = 0 0 ( c 0 ) = 0 0 c 0 = k 0 c 0 ,
we
calculate
the
wavenumbers:
k1 c = j = 51.31 65 j25 428.18e
2 kc j0.1835
51.31 69.64e
0.367 2
=
1
(2.19.2)
0.322 2
=
1
(2.19.3)
1 =1 1 = 12.8 cm 2 =1 2 = 13.7 cm
D.RamacciaandA.Toscano
Pag.42
Exercises Chapter 2
This heating method is effective because the penetration depths are bigger than the dimension of mesh potatoes and carrots. The energy in the electromagnetic waves reheat successfully the foods. The attenuation of the electric field (in dB and absolute units) at a depth of 1 cm from the surface of the food is:
A1 ( z = 1 cm ) = 8.686 z 1 = 8.686 12.8 = 0.68 dB dB
2 AdB ( z = 1 cm ) = 8.686 z 2 = 8.686 13.7 = 0.63 dB A1 dB A1 = 10 20
(2.19.4)
= 0.925 = 0.930
A1 = 10
2 AdB
(2.19.5)
20
Thus, the fields at a depth of 1 cm are 92.5% and 93% of their values at the surface.
D.RamacciaandA.Toscano
Pag.43
Exercises Chapter 2
Solution
The parameters , and in a good conductor are:
= =
=
1 =
2
2
= f
= 1 f
(2.20.1) (2.20.2)
The conductibility of the copper is 5.8 107 Siemens/m, so the skin depth at frequency f is:
1 1 = f 1 2 = 0.0661 f 1 2 7 7 f 4 10 5.8 10
(2.20.3)
A dB ( z ) = 8.686 z
(2.20.4) with the assumption that AdB ( z ) 50dB , we have:
A dB ( z ) = 8.686 z 50dB z
(2.20.4)
and its minimum value over the frequency range of interest is at least 50 dB. So inverting the eq.
The inequality (2.20.5) can be plotted as function of frequency in the range 10 kHz1 GHz:
D.RamacciaandA.Toscano
Pag.44
Exercises Chapter 2
Fig. 2.20.1: Thickness of copper shield in mm for 50dB of attenuation. The high frequency interference is attenuated of 50dB using a copper shield with thickness very low, exactly, at 1 GHz, 0.012 mm of copper are sufficient. On the contrary at low frequencies the thickness is more, exactly, at 10 kHz, z =3.8 mm, that represents the minimum thickness of the shield in order to satisfy the attenuation limit.
D.RamacciaandA.Toscano
Pag.45
Exercises Chapter 2
Solution
First of all we have to evaluate the skin depth as function of the frequency f:
=
The attenuation in dB is:
1 1 = f 1 2 7 7 f 4 10 3.5 10
0.0851 f 1 2
(2.21.1)
A dB ( z ) = 8.686 z
(2.21.2) with the assumption that AdB ( z ) 50dB , we have:
A dB ( z ) = 8.686 z 50dB z
(2.21.2)
and its minimum value over the frequency range of interest is at least 50 dB. So inverting the eq.
The inequality (2.21.3) can be plotted as function of frequency in the range 10 kHz1 GHz:
D.RamacciaandA.Toscano
Pag.46
Exercises Chapter 2
A typical thickness of aluminium is about 1 mm, and this shield reduces by 50 dB only electric fields with frequency greater than 150 kHz.
D.RamacciaandA.Toscano
Pag.47
Exercises Chapter 2
(2.22.1)
Solution
For a forwardmoving wave, we have E ( z, t ) = F ( z ct ) = F ( 0 c ( t z c ) ) , which implies that
j1( t + z c )
+ E 2e
j2 ( t + z c )
)=
)
(2.22.2)
H ( z, t ) =
where Hi = Ei Z0 with i = 1, 2 .
(2.22.3)
D.RamacciaandA.Toscano
Pag.48
Exercises Chapter 2
(2.23.1)
1) Verify that these expressions satisfy all of Maxwell's equations. 2) Show that the timeaveraged energy flux in the zdirection is independent of z and is given by:
1 1 2 2 Pz = e E ( z ) H ( z ) = E+ + E 2 2
(2.23.2)
3) Assuming = 0 and = n 2 0 , so that n is the refractive index of the dielectric, show that the fields at two different zlocations, say at z = z1 and z = z2 are related by the matrix equation:
E ( z1 ) cos k = 0 H ( z1 ) j sin k
where
j 1 sin k E ( z 2 ) cos k 0 H ( z 2 )
(2.23.3)
location z. Show the relationships at the location z1 and z2 : Z ( z1 ) = Z ( z 2 ) + j 1 tan k , 1 + j Z ( z 2 ) tan k Y ( z1 ) = 1 + j 1Y ( z 2 ) tan k Y ( z 2 ) + j tan k (2.23.4)
What would be these relationships if had we normalized to the medium impedance, that is,
Z ( z) = E ( z) H ( z) ?
Solution
Question n 1 Assuming an harmonic time dependence e jt , the Maxwell's equations can be written as follow:
D.RamacciaandA.Toscano Pag.49
Exercises Chapter 2
(2.23.5)
In a source less, linear, isotropic and uniform medium, the quantities and J are zero and the constitutive relations B = H and D = E are valid. So the set (2.23.5) becomes: E = j H H = j E E = 0 H = 0 Now it is possible to verify the first Maxwell's equation in set (2.23.6): x xE ( z ) = x E (z) y y 0 z z = y z E ( z ) z y E ( z ) = 0
(2.23.6)
(2.23.7)
From exercise 2.7 we know the k relationship and as consequence also the expression of the
(2.23.8)
2 y E ( z ) = j y E (z) =
E (z)
(2.23.9)
= j y
= j H ( z )
It is very simple to verify the second Maxwell's equation in the same way. The third and fourth equation are the divergence of the electric and magnetic field respectively:
E = xE ( z ) = x E ( z ) = 0
H = yH ( z ) = y H ( z ) = 0 Question n 2
(2.23.10) (2.23.11)
D.RamacciaandA.Toscano
Pag.50
Exercises Chapter 2
The energy flux can be evaluated as dot product of the Poynting vector and the unit vector along the zdirection:
1 Pz = P z = e E H z 2
(2.23.12)
)(
(2.23.13)
Question n 3
Consider the expression for E ( z1 ) and multiply it by the neutral term e jkz2 e jkz2 :
E ( z1 ) = E + e jkz1 e jkz 2 e jkz 2 + E e jkz1 e jkz 2 e jkz 2
consequently,
E ( z1 ) = E + e jkz 2 e jk + E e jkz 2 e jk
(2.23.14)
where
= z2 z1 . Using the Euler's formula e jx = cos x + jsin x , eq. (2.23.14) can be written as:
E ( z1 ) = E + e jkz2 ( cos k + jsin k ) + E e jkz2 ( cos k jsin k
)=
(2.23.15)
E ( z1 ) = E ( z 2 ) cos k + j
0H ( z 2 )
n
sin k
(2.23.16)
(2.23.17)
Pag.51
S.J. Orfanidis Electromagnetic Waves and Antennas The normalized wave impedance at location z1 is:
Z ( z1 ) =
Exercises Chapter 2
0 H ( z1 )
E ( z1 )
and we can substitute the electric and magnetic field with their respective relationships (2.23.16) and (2.23.17). So: sin k n Z ( z1 ) = jnE ( z 2 ) sin k + 0 H ( z 2 ) cos k Dividing the numerator and denominator by 0 H ( z 2 ) cos k , we have: E ( z 2 ) cos k sin k + jn 1 0 H ( z 2 ) cos k cos k Z ( z 2 ) + jn 1 tan k = Z ( z1 ) = E ( z 2 ) sin k 1 + jnZ ( z 2 ) tan k 1 + jn 0 H ( z 2 ) cos k The admittance Y ( z1 ) is the inverse of Z ( z1 ) : E ( z 2 ) cos k + j
0 H ( z 2 )
(2.23.18)
(2.23.19)
Y ( z1 ) =
n 1 1+ j tan k Z ( z2 )
1 + jn 1Y ( z 2 ) tan k
If we had normalized Z ( z ) and Y ( z ) to the medium impedance, simply we have to cancel the term n of refractive index inside eq. (2.23.19) and (2.23.20): Z ( z 2 ) + jtan k Z ( z1 ) = 1 + jZ ( z 2 ) tan k Y ( z ) = Y ( z 2 ) + jtan k 1 1 + jY ( z 2 ) tan k
D.RamacciaandA.Toscano
Pag.52
Exercises Chapter 2
H ( r, t ) =
[ yA xB] e j t jkz
(2.24.1)
where ( x , y, z ) is a rotated coordinate system with respect to a fix coordinate system ( x, y, z ) as shown in
where z = ( z cos + x sin ) is the unit vector in the direction of propagation. Show that the energy
transport velocity is v =
P = cz . w
Solution
The dot products E E and H H can be evaluated as follow:
2 2 E ( r, t ) E ( r, t ) = [ xA + yB] xA + yB = AA + BB = A + B 1 1 1 2 2 H ( r, t ) H ( r, t ) = 2 [ yA xB] yA xB = 2 AA + BB = 2 A + B
and now we can substitute them inside the definition of the timeaveraged energy density:
D.RamacciaandA.Toscano
Pag.53
Exercises Chapter 2
(2.24.3)
y B A
A B
A2 B2 1 = z A 2 + B 2 + 0 = z 0
and, consequently,
v=
where c =
D.RamacciaandA.Toscano
Pag.54
Exercises Chapter 2
E ( x, y, z ) = yE0e j t e
jk( x + z )
k=
0
c0
(2.25.1)
1. Inserting E ( x, y, z ) into Maxwell's equations, work out an expression for the corresponding magnetic field H ( x, y, z ) .
2. What is the direction of propagation and its unit vector k ?
3. working with Maxwell's equations, determine the electric field E ( x, y, z, t ) and the
propagation direction k , if we started with a magnetic field given by:
H ( x, z, t ) = yH 0e j t e
jk
3z x 2
(2.25.2)
Solution
Question n 1 From the first Maxwell's equation, we can find the magnetic field as follow:
H ( x, z, t ) =
The cross product has to be evaluated:
1 E ( x, z, t ) j
(2.25.3)
x E ( x, z, t ) = x 0
jk ( x + z )
y y E ( x, z, t )
z z 0
(2.25.4)
where E ( x, z, t ) = E 0 e j t e
E ( x, z, t ) = x
j t jk ( x + z ) E 0e e z
+ z E e j t e jk( x + z ) 0 x
2
=
2
jk j t jk ( x + z ) = x E 0e e 2
+ z jk E e j t e jk( x + z ) 0 2
(2.25.5)
and consequently
D.RamacciaandA.Toscano
Pag.55
Exercises Chapter 2
j 1 j = E ( x, z, t ) z E ( x, z, t ) = x j 2 2 =z E ( x, z, t )
(2.25.6)
E ( x, z, t )
E ( x, z, t )
where k = , = and z = ( z x )
The direction of propagation can be found as cross product between the direction of oscillation of the electric and magnetic field. So:
x k = y z = 0 1 2 y 1 0 z 0 = 1 2 ( x + z ) = x 2
(2.25.7)
Question n 3
(2.25.8)
where H ( x, z, t ) = H 0e j t e
jk
3z x 2
. So:
D.RamacciaandA.Toscano
Pag.56
Exercises Chapter 2
j t jk ( + z H 0e e x
3z x 2
1 j 3k jk H ( x, z, t ) + z H ( x, z, t ) = x j 2 2 j 1 j 3 H ( x, z, t ) = H ( x, z, t ) + z = x 2 2 j 1 = H ( x, z, t ) 2
where k = , = and x =
(2.25.9)
(
(
1 3x + z = H ( x, z, t ) x 2
3x + z . We can assume a new coordinate system aligned
Fig. 2.25.2: Rotation of the coordinate system. The direction of propagation can be found as cross product between the direction of oscillation of the electric and magnetic field. So:
x k = y z = 0 3 y z 1 0 = x 3z = z 0 1
D.RamacciaandA.Toscano
Pag.57
Exercises Chapter 2
Fig. 2.26.1: Polarizeranalyzer filter combination. Assume that the amplitude attenuation through the first polarizer are a1 , a2 with respect to the x and ydirections. The polarizer transmits primarily the xpolarization, so that a2
a1 . The
analyzer is rotated by an angle so that the same gains a1 , a2 now refer to the x' and y' directions. 1. Ignoring the phase retardance introduced by each polarizer, show that the polarization vectors at the input, and after the first and second polarizer, are: E0 = x cos + y sin E1 = xa1 cos + ya2 sin (2.26.1)
2 2 E2 = x a1 cos cos + a1a2 sin sin + y a2 cos sin a1a2 sin cos
where ( x, y ) are related to ( x, y ) as follow:
) (
x cos sin x = y sin cos y 2. Explain the meaning an usefulness of the matrix operations:
a1 0 0 cos a2 sin sin a1 cos 0 0 cos a2 sin
(2.26.2)
(2.26.3)
and
cos sin sin a1 0 cos cos 0 a2 sin sin a1 0 cos cos 0 a2 sin
(2.26.4)
Exercises Chapter 2
(2.26.5)
4. If the input light were unpolarized, that is incoherent, show that the average of the intensity of part (3) over all angles 0 2 , will be given by the generalized Malus's law:
I= 1 4 4 2 2 a1 + a2 cos 2 + a1 a2 sin 2 2
(2.26.6)
Solution
Question n 1 The electric field E1 after the polarizer is an attenuated form of the field E0 , that is each component is attenuated by a factor a1 or a2 , according to x or ydirections respectively. So we can characterize the polarizer with a own attenuation matrix:
a A= 1 0 0 a2
and, consequently,
0 cos a1 cos a E1 = A E0 = 1 = xa1 cos + ya2 sin = 0 a2 sin a2 sin
(2.26.7)
The electric field E1 passes thought the second polarizer that is rotated of an angle by an angle so that the same gains a1 , a2 now refer to the x' and y'directions. The rotation can be expressed by the matrix (2.26.2) and then the x' and y'components of the field E2 have to be attenuated by a factor a1 or a2 , according to x or ydirections respectively. So:
0 cos a E2 = 1 0 a2 - sin sin E1 cos
(2.26.8)
from which:
0 cos a E2 = 1 0 a2 - sin = x sin a1 cos = cos a2 sin
2 a1 cos cos
) (
2 a2 cos sin
(2.26.9)
Question n 2
D.RamacciaandA.Toscano
Pag.59
Exercises Chapter 2
The matrix operation (2.26.3) is simply the cascade of the matrix that we used to solve the point (1). In fact the first two matrixes are necessary to rotate and attenuate the field E1 that passes through the analyzer, the third matrix represents the attenuation through the polarizer and the fourth represents the tilt of the electric field vector with respect to the x and ydirections. Using these matrixes operation, it is a very simple model system. In the matrix operation (2.26.4), shown here for simplicity, it is possible to note that there is a new matrix M at the beginning of the expression:
cos sin sin a1 0 cos cos 0 a2 sin
M
This has the same structure of a rotation matrix, but the angle is opposite, i.e. . This suggest that (2.26.4) represents a system with another analyzer at the end tilted of an angle , but without any attenuations. Question n 3 The light intensity is the timeaveraged energy density multiplied by the speed of light in the host medium, that is vacuum in this case. From exercise 2.24, we have already demonstrated the expression of the timeaveraged energy density w and so:
I=
2 1 2 1 2 c E = c E x + E y 2 2
(2.26.10)
that is the light intensity is proportional to the sum of the square module of the components of electric field.
2 I = E x + E y = a1 cos cos + a1a2 sin sin
) + ( a22 cos sin a1a2 sin cos ) 4 2 2 3 = ( a1 cos 2 cos 2 + a1 a2 sin 2 sin 2 + 2a1 a2 sin sin cos cos ) + 4 2 2 3 + ( a2 cos 2 sin 2 + a1 a2 sin 2 cos 2 2a1a2 sin sin cos cos ) 2 2 2 2 4 4 = ( a1 cos 2 + a2 sin 2 ) cos 2 + a1 a2 sin 2 + 2a1a2 ( a1 - a2 ) sin sin cos cos
2 2 2
= (2.26.11)
Question n 4
The average of the intensity over all angles 0 2 can be evaluated integrating the eq. (2.26.11) and dividing it by 2 as follow:
D.RamacciaandA.Toscano
Pag.60
Exercises Chapter 2
2 2 a1 a2 sin 2 2 2 0
I=
I d =
cos 2 d +
sin 2 d +
d +
(2.26.12)
cos sin d
(2.26.13)
I d = 2 a1 + a2 cos
4 4
2 2 + a1 a2 sin 2
D.RamacciaandA.Toscano
Pag.61
Exercises Chapter 2
Fig. 2.27.1: Plane wave viewed from stationary and moving frames. First, prove the equivalence of the three relationships given by:
cos = cos 1 cos
1 2 .\
sin =
sin (1 cos )
tan ( 2 ) tan ( 2 )
1+ (2.27.1) 1
where = v c and = 1
(2.27.2)
(1 + cos )
=f
(2.27.3)
Solution
The three relationships in (2.27.1) relate the apparent propagation angles , in the two frames that are different because of the aberration of light due to the motion. They are a consequence of the Lorentz transformation of the frequencywavenumber fourvector ( c, k ) :
= ( ck z ) k = k z z c k = k x x
(2.27.4)
D.RamacciaandA.Toscano
Pag.62
Exercises Chapter 2
in the frame S', k = k cos , k = k sin , with k = c , the Eqs. (2.27.4) may be rewritten in z x the form:
cos =
( cos ) (1 cos )
(2.27.6)
sin =
sin (1 cos )
(2.27.7)
(2.27.8)
sin = cos (1 cos ) 1 + 1 cos sin = = 1 cos + cos (1 cos ) 1 cos sin 1 tan 2 = = (1 )(1 + cos ) (1 )
(2.27.9)
D.RamacciaandA.Toscano
Pag.63
Exercises Chapter 2
In the identity of eq. (2.27.2) between the angle , we note that the last term is equal to 1 2 , that can be expressed using the second identity in eq. (2.27.1):
sin = sin (1 cos ) 1
= 1 2 =
sin 2 sin
2
(1 cos )2
(2.27.11)
Consequently since
sin 2 sin
2
(1 cos )2 = 1 2
(2.27.2)
(1 cos ) = 1 + cos
(2.27.12)
(1 cos ) =
= =
2 (1 cos )
(1 cos ) =
(2.27.13)
(2.27.14)
The second identity in eq. (2.27.2), i.e. (1 + cos )(1 cos ) = 1 2 , is formally identical to the first identity, but the angles , are inverted. This is reasonable if the velocity vector v is directed in the negative zdirection, that is v v . Since the term is defined as v c and it is the square, the sign of v is negligible and the identity is valid. The alternative Doppler formulas in eqs. (2.27.3) are obtained applying (2.27.14) to the relativistic Doppler formula, relating the frequency of the wave as measured by an observer in the moving frame S' to the frequency of a source in the fixed frame S:
f = f (1 cos ) = f 1 cos 1
2
(2.27.15)
D.RamacciaandA.Toscano
Pag.64
S.J. Orfanidis Electromagnetic Waves and Antennas 1 2 1 = 2 (1 cos ) = (1 + cos ) (1 + cos ) 1 1 1 2 = = (1 + cos )(1 cos ) and then substitute them inside (2.27.15): f = f (1 cos ) = f = f 1 cos 1 2 f
Exercises Chapter 2
(1 + cos ) (1 + cos )
1 cos
(2.27.16)
D.RamacciaandA.Toscano
Pag.65
Exercises Chapter 2
(2.28.1)
it was assumed that the plane wave was propagating in the zdirection in all three reference frames S, Sa, Sb.
Fig. 2.28.1: Propagating plane wave along the direction. If in the frame S the wave is propagating along the direction shown in Fig. 2.28.1, show that the Doppler formula may be written in the following equivalent forms:
f b = fa
where
a =
and v is the relative velocity of the observer and source given by v= 1 v b va c2 (2.28.4)
and a , b are the propagation directions in the frame Sa, Sb. Moreover, show the following relations among these angles:
cos a =
cos a , 1 a cos
cos b =
cos b , 1 b cos
cos b =
cos a 1 cos a
(2.28.5)
D.RamacciaandA.Toscano
Pag.66
Exercises Chapter 2
Solution
The relativistic Doppler formula relates the frequency of the wave as measured by an observer in the moving frame S' to the frequency of a source in the fixed frame S:
f = f (1 cos ) = f 1 cos 1
2
(2.28.6)
If we consider separately the frame Sa and the frame Sb, we can write for each frame a relativistic Doppler formula:
f a = f a (1 a cos ) = f f b = f b (1 b cos ) = f 1 a cos
2 1 a
1 b cos
2 1 b
(2.28.7)
From the first equation in (2.28.7), we can write f as a function of fa, and substitute it inside the second equation, in order to obtain:
f b = fa
b (1 b cos ) a (1 a cos )
(2.28.8)
If the observer moves with the same velocity of the frame Sa, he will have the sensation that the frame Sa is fixed and the frame Sb is moving. So it is possible to use the relationships (2.27.16):
f = f (1 cos ) = 1 cos 1 + cos
(1 + cos )
=f
(2.28.9)
where
f f a , f f b a , b
(2.28.10)
The relationships between the angles a , b expressed in (2.28.5) can be obtained from the analogue expression for the angles , :
cos =
cos 1 cos
(2.28.11)
where is the apparent angle along which the wave propagates in the moving frame, is the angle along which the wave propagates in the stationary frame and is the ratio between the velocity v of the moving frame with respect to the fixed frame and the speed of light in vacuum. So
D.RamacciaandA.Toscano Pag.67
Exercises Chapter 2
if we consider first the frame Sa and then the frame Sb with respect to the fixed frame S, we can write the two following relationships:
cos a =
cos a , 1 a cos
cos b =
cos b 1 b cos
(2.28.12)
from which we can obtain the expression of cosb as a function of cosa . From the first identity of (2.28.12) we can write:
cos =
and we can substitute it in the second identity:
a + cos a 1 + a cos a
a + cos a b cos b 1 + a cos a cos b = = = 1 b cos 1 a + cos a b 1 + a cos a a + cos a b (1 + a cos a ) = 1 + a cos a b a + b cos a + cos a b b a cos a = a = 1 + a cos a b a b cos a
= =
(1 b a ) cos a ( b a ) = (1 b a ) ( b a ) cos a
D.RamacciaandA.Toscano
Pag.68
Exercises Chapter 2
= 9 0 , and permeability = 0 . You may use the "weakly lossy dielectric" approximation.
Fig. 2.29.1: Section of the ground with an underground object. 1. Determine the numerical value of the wavenumber k = j in meters-1, and the penetration depth = 1 in meters. 2. Determine the value of the complex refractive index nc = n r jni of the ground at 900 MHz. 3. With reference to the above figure, explain why the electric field returning back to the radar antenna after getting reflected by the buriedpipe is given by:
E ret E0
2
4 h2 + d2 = exp
(2.29.1)
where E0 is the transmitted signal, d is the depth of the pipe, and h is the horizontal displacement of the antenna from the pipe. You may ignore the angular response of the radar antenna and assume it emits isotropically in all directions into the ground. 4. The depth d may be determined by measuring the roundtrip time t ( h ) of the transmitted signal at successive horizontal distances h. Show that t ( h ) is given by:
t (h) =
2n r d2 + h2 c0
(2.29.2)
where n r is the real part of the complex refractive index n c . 5. Suppose t ( h ) is measured over the range 2 h 2 meters over the pipe and its minimum recorded value is t min = 0.2 s . What is the depth d in meters?
D.RamacciaandA.Toscano Pag.69
Exercises Chapter 2
Solution
Question n 1
+ d k = j = d 1 j = d 1 j 2 2 d
where d and d are the real and imaginary part of the dielectric constant , i.e. = d + j d , and = tan = d d is the loss tangent that is a convenient way to quantify the losses. In this case d = 9 0 and d = . So we can evaluate (2.29.3):
k = j = 90 0 1 j = 90 0 j 2 9 0 2
0 9 0
= 3 = 6 900 106 4 107 8.85 1012 = 56.57 rad / m 0 0 4 107 0 103 = = = 0.063 rad / m 2 9 0 2 9 8.85 1012 The corresponding penetration depth = 1 = 15.87 meters. Question n 2
(2.29.4)
and in this case the relative permittivity is complex because the material is lossy, i.e. the permittivity is complex. The complex value of the permittivity of the ground at 900 MHz is:
= 9 0 j
jtan1.11
79.65
and consequently, using (2.29.4), the complex refractive index of the ground is: n=
(2.29.5)
D.RamacciaandA.Toscano
Exercises Chapter 2
Ignoring the angular response of the radar antenna and assuming it emits isotropically in all directions into the ground, the electric field into the ground is:
E ground = E 0 e jkr
k = j
E 0 e r e j r
(2.29.6)
The distance r between the antenna and the buriedpipe can be evaluated using the Pythagoras' theorem:
r = h2 + d2
(2.29.7)
The transmitted signal reaches the object and returns back to the radar antenna after getting reflected by the buriedpipe. So the round trip is two times long and the backward signal, using (2.29.6) and (2.29.7), can be expressed as: E ret = E 0e 2 jkr = E 0e 2 The module of eq. (2.29.8) is:
E ret = E 0 e 2
h 2 +d2
h 2 + d 2 2 j h 2 +d 2
(2.29.8)
from which
E ret E0
2
=e
4 h 2 +d 2
=1
= e
h 2 +d 2
(2.29.9)
Question n 4
The time is the ratio of distance divided by speed that explains the amount of distance covered in a given time:
s = vt
(2.29.10)
where s is the distance in meter, v is the constant speed in meter per second and t is the time in second. The wave propagates into the ground with velocity cg = c0 n r , where n r is the real part of the refraction index of the ground, and the round trip of the wave from the antenna to the buriedpipe is 2r long. This values can be substituted inside (2.29.10) and, using (2.29.7), we obtain:
2r =
Question n 5
c0 t (h) nr
t (h) =
2n r 2n r = r d2 + h2 c0 c0
(2.29.11)
D.RamacciaandA.Toscano
Pag.71
Exercises Chapter 2
It is possible to note from (2.29.11) that the minimum roundtrip time t ( h ) is when the antenna is aligned with the pipe, that is the value of h is zero. Using (2.29.11), we can evaluate the depth d in meters:
d= c0 3 108 t (h) = 0.2 106 = 10 m 2n r 23
(2.29.12)
D.RamacciaandA.Toscano
Pag.72