Eem Notes Ch4 Rev05
Eem Notes Ch4 Rev05
Eem Notes Ch4 Rev05
Engineering Electromagnetics
Notes for Chapter 4
(Maxwells Equations, Fields
in dielectric and magnetic
media, Boundary Conditions,
and Plane Waves) Revision
005
Prepared by:
Reviewed by:
Reviewed by:
Authors
Date
Place
Raghuvir Tomar
Jan25,
2012
LNMIIT
Jaipur,
India
Revision History
Revision Date
Description
001
Initial draft
002
003
004
005
______________________________________________________________________________________
Table of Contents
1
2.1
2.2
2.3
2.4
2.5
2.6
2.7
2.8
2.9
2.10
2.11
Boundary conditions........................................................................................ 9
2.12
2.13
2.14
2.15
2.16
2.17
2.18
2.19
2.20
2.21
2.22
2.23
______________________________________________________________________________________
ii
______________________________________________________________________________________
iii
CONSTITUTIVE RELATIONSHIPS
(1.1)
D=*E
(1.2)
=0r
(1.3)
(1.5)
where r is the relative permittivity (dielectric constant) of the medium and 0, the
permittivity of free-space, is given by
0=(1/(36*))*10-9 Farad/m
(1.6)
(1.7)
where is the conductivity of the medium, and E is the incident field that causes a surface
current density, J (in A/m2), on the surface of the conductor.
Eq. (1.7) can easily be derived, at least in the case of a rectangular bar made of material with
finite electrical conductivity, by using Ohms law.
_____________________________________________________________________________________
MAXWELLS EQUATIONS
The four Maxwells equations are reproduced below. Please note that bold letters
represent vectors and a dot on top represents ( / t). All possible variations of the four
Maxwellian equations are shown. The last few variations shown (in case of each of the
equations) are for the case when ejt type of time-dependence is assumed (a very
practically used case).
2.1
v(t)= - ( t)
(2.1.1)
In eq. (2.1.1),v(t) is the induced emf and is the magnetic flux whose rate of
change with time is producing v(t). We can rewrite equation (2.1.1) as
E.dl
B.dS
(2.1.2)
where the left hand side (LHS) equates the induced emf to the line integral of the
corresponding electric field (E) and the right hand side (RHS) equates the magnetic flux
to the surface integral of the corresponding magnetic flux density (B). S represents a
two-dimensional surface enclosed by a closed contour C.
Now let us use Stokes theorem
A.dl (curlA).dS
(2.1.3)
- ( B/ t)
(M1)
which is the first Maxwellian equation in its differential form. We can re-write this
equation in several other forms, as shown below. As already mentioned, the dot on top of
a mathematical quantity represents the time-derivative of that quantity.
.
XE=
-B
(M1.1)
= - ( H/ t)
= - r ( H/ t)
(M1.2)
(M1.3)
_____________________________________________________________________________________
.
=-
(M1.4)
.
=-
r H
= - r( H/ t)
= - jrH
(M1.6)
= - jB
(M1.8)
(M1.5)
(M1.7)
Note that the last two representations (M1.7 and M1.8) are applicable only when
jt
2.2
H.dl I
(2.2.1)
Let us now use Stokes theorem (eq. 2.1.3) to convert eq. (2.2.1) into
XH=
(2.2.2)
where J represents the surface current density (in A/m2) associated with I.
Eq. (2.2.2) is good enough when interactions between time-varying electric and
magnetic fields can be assumed to be negligible. As frequency increases, this, however,
becomes less and less of a safe bet. Maxwells genius lied in suggesting (based on
intuitive reasoning which was later verified experimentally) that an additional term be
added to the RHS of eq. (2.2.2) to account for the fact that time-varying E gives rise to
effects in total H. Eq. (2.2.2) was thus modified to
XH=
( D/ t)+J
(M2)
where the first term on the RHS represents the so-called displacement-current density
(in A/m2) that was added-in by Maxwell (for an interesting up-to-date view of
displacement current please visit http://en.wikipedia.org/wiki/Displacement_current).
_____________________________________________________________________________________
Eq. (M2) is the second Maxwellian equation in its differential form. We can write
this equation in several other forms too, as shown below.
.
XH=
D+J
= ( E/ t) + J
= r( E/ t) + J
(M2.1)
(M2.2)
(M2.3)
.
=
E + J
(M2.4)
.
rE + J
= r( E/ t) + J
= jrE + J
(M2.5)
= jD + J
(M2.8)
(M2.6)
(M2.7)
Note that the last two representations (M2.7 and M2.8) apply only when ejt type of
time-dependence is assumed for both E and H vectors.
2.3
The third Maxwellian equation is derivable from Gausss law which states that the
net electric flux passing through a closed surface S is equal to the total electric charge Q
enclosed by that surface. In other words,
D.dS Q
(2.3.1)
A.dS ( DivA)dV
S
(2.3.2)
( DivD)dV Q dV
V
(2.3.3)
_____________________________________________________________________________________
In eq. (2.3.3), represents the volume charge density in C/m3. This equation easily
yields
.D=
(M3)
2.4
(M4)
by assuming, without any loss of generality, that the constant involved can be assumed to
be zero.
Eq. (M4) is the fourth Maxwellian equation in its differential form.
-( /
t)
(2.5.1)
(2.5.2)
By using Stokes and Divergence theorems, the four Maxwells equations can also
be integrated and converted into their equivalent integral representations given below.
_____________________________________________________________________________________
E.dl
H .dl
C
B.dS
t S
(M.I.1)
D.dS I
t S
(M.I.2)
D.dS dV
(M.I.3)
B.dS 0
(M.I.4)
2.7
Types of media
The various types of transmission media encountered in practice are briefly discussed
below.
1. Homogeneous and non-homogeneous media
_____________________________________________________________________________________
D E
B H
(2.7.1)
(2.7.2)
where [D], [E], [B], and [H] are 3x1 matrices and [] and [] are 3x3 matrices.
D 0 E Pe
(2.9.1)
Pe 0 eE
(2.9.2)
where Xe is called the electric susceptibility of the medium and can, in general, be a
complex number. Thus we write
D 0(1 Xe) E
(2.9.3)
or
D E
(2.9.4)
where
(2.9.5)
(1 X )
0
(2.9.6)
(j)E
(2.9.7)
jj('-j')]E
(2.9.8)
The loss tangent tan of the material is defined as the ratio of the imaginary and
real parts of the complex permittivity. In other words,
tan
' '
'
(2.9.9)
and
' (1 j tan )
(2.9.10)
_____________________________________________________________________________________
'
(2.9.11)
In the case of magnetic materials too, an alignment of molecules takes place when
a magnetic field H is applied across the material. The effect is called magnetic
polarization. We define a magnetic polarization vector Pm and a magnetic susceptibility
Xm. The relevant equations follow:
B 0 H Pm
(2.10.1)
Pm 0 mH
(2.10.2)
B 0(1 Xm) H
(2.10.3)
B H
(2.10.4)
(1 X )
0
(2.10.5)
In the above,
(2.10.6)
Consider the plane interface between two media (medium 1 and medium 2) as shown
in Figure 1. Let , be the properties of medium 1 and , be the properties of
medium 2. The quantities of interest to us are:
1)
2)
3)
4)
5)
6)
_____________________________________________________________________________________
Figure 1 Fields, currents, and surface charge at a general interface between two
media
_____________________________________________________________________________________
10
D.dS dV Q
S
to get
(M.I.3)
S ( Dn 2 Dn1) Sidewalls Q
S
where the second term on the LHS represents contribution of the sidewalls of the cylinder
to the total surface integral of D over S.
If we now allow h to go to zero, the second term on the LHS of the above
equation goes to zero (since there are no sidewalls left). Also the RHS collapses into the
term s S. The term Scancels out on both sides and we get
Dn2-Dn1= s
(2.11.1)
in.(D2-D1)= s
(2.11.2)
B.dS 0
(M.I.4)
leads to
Bn2-Bn1=
(2.11.3)
in.(B2-B1)=
(2.11.4)
11
2) The normal component of the magnetic flux density is continuous across the
interface (since free magnetic charge does not exist).
E.dl
B.dS
t S
(M.I.1)
to write
(Et1-Et2) + contributions due to two sides which are h long
B.(h. )
Then let us allow h to go to zero. The second term on the LHS of the above
equation goes to zero since these two sides are not there anymore. Also, the term on the
RHS goes to zero as h goes to zero. Thus we are left with
Et2=Et1
(2.11.5)
(E2-E1)x in.=
(2.11.6)
H .dl
C
D.dS I
t
S
(M.I.2)
to write
Ht2-Ht1= Js
(2.11.7)
inX(H2-H1)= Js
(2.11.8)
12
3) The tangential component of the electric field intensity is continuous across the
interface.
4) The tangential component of the magnetic field intensity is discontinuous by an
amount equal to the value of the surface current density at the interface.
If medium 1 happens to be a perfect electrical conductor ( tending to infinity),
all field components inside the conductor (medium 1) shall go to zero (this will be shown
a bit later in this chapter). However, finite Js and s may still exist at the interface of the
two media. The boundary conditions in such a case become
Dn2= s
(2.11.9)
Bn2=
(2.11.10)
Et2= 0
(2.11.11)
Ht2= Js
(2.11.12)
Dn2= Dn1
(2.11.13)
Bn2= n
(2.11.14)
Et2= Et1
(2.11.15)
Ht2= Ht1
(2.11.16)
_____________________________________________________________________________________
13
Plane waves are those waves whose surfaces of constant phase are infinite planes
parallel to each other. In free-space, an ideal plane wave will be produced at infinity
distance away from a point source. However, many practically used wave patterns are
quite analogous to plane-waves. The study of plane waves, hence, is worth the effort.
2.13 Wave equations in free-space
In a source-free environment and assuming free-space as the medium involved, take the
curl of eq. (M2) on both sides, and then use eq. (M1) to get (assuming ejt type of timedependence)
X xH=j XE= j-jH
(2.13.1)
That is
( .H)
- 2H=k02H
(2.13.2)
k02= 2
(2.13.3)
.H=0
(2.13.4)
( 2+ k02)H=0
(2.13.5)
( 2+ k02)E=0
(2.13.6)
Eqs. (2.13.5) and (2.13.6) are known as the wave equations for H and E, respectively.
2.14 Wave number in free-space
The wave-number in free-space has already defined by eq. (2.13.3). It can further be
simplified by using the well-known equation for the speed of light, c, in free space:
c2=1/((2.13.7)
_____________________________________________________________________________________
14
We can write
(2.13.8)
k0= /c =2f/c=2/0
where 0 is the free-space wavelength involved. Note that the definition of k0 is similar to
that of the propagation constant of a loss-less transmission line.
2.15 Uniform plane wave
A plane wave is one for which phase is constant on a plane. A uniform plane wave is
one for which both magnitude and phase are constant on a plane.
Consider a plane wave traveling in +x direction. If E and H fields are independent of y
and z and depend only on x, we call it a uniform plane wave.
2.16 Transverse electromagnetic (TEM) mode of propagation
In the next section, we will show that, for a uniform plane wave, both E and H fields
have no component along the x-axis. In other words, E and H lie completely in the y-z
plane. Since y-z plane is a plane transverse (perpendicular) to the direction of
propagation (x), a uniform plane wave can be thought of as obeying the so-called
transverse electromagnetic (TEM) mode of propagation.
2.17 Uniform plane wave in free-space
For a uniform plane wave traveling in the +x direction in free-space, assuming that the
electric field intensity has no dependence on y and z, the wave equation for E simplifies
to
(d2E/dx2)+ k02E=0
(2.17.1)
(2.17.2)
(d2Ey/dx2)+ k02Ey=0
(2.17.3)
(d2Ez/dx2)+ k02Ez=0
(2.17.4)
The wave equation for the magnetic field strength, similarly, yields
(d2Hx/dx2)+ k02Hx=0
(2.17.5)
(d2Hy/dx2)+ k02Hy=0
(2.17.6)
(d2Hz/dx2)+ k02Hz=0
(2.17.7)
_____________________________________________________________________________________
15
.E=0
(M3.MSF)
(2.17.8)
(2.17.9)
(2.17.10)
(2.17.11)
Now we are left with eqs. (2.17.3) and (2.17.4) for the remaining unknowns Ey and Ez
and eqs. (2.17.6) and (2.17.7) for the remaining unknowns Hy and Hz. Please be
reminded that exp (jt) is an implicit term in all these equations.
The solutions to eqs. (2.17.3), (2.17.4), (2.17.6), and (2.17.7) can be written in several
forms. Let us pick the following forms (I am exclusively showing the time-dependence
here):
Ey (x,t)= A exp (jt-jk0x)
(2.17.12)
(2.17.13)
(2.17.14)
(2.17.15)
where A,B, C and D are the four arbitrary constants involved. Note that eqs. (2.7.12)(2.17.15) show that all the four non-zero field components (Ey,Ez,Hy and Hz) have
constant phase over surfaces that satisfy the following equation
t - k0x = constant
(2.17.16)
Since t increases constantly (whether we like it or not), the only way that eq. (2.7.16) can
be satisfied is that x should also constantly increase. In other words, the equi-phase
surface (a plane, as already mentioned) shall travel at a speed dx/dt which can be
_____________________________________________________________________________________
16
computed by taking time-derivative of both sides of eq. (2.17.16). This speed dx/dt is
termed (as in the case of transmission lines) phase velocity vp.
We get
vp
dx 2f 0
c
dt k 0
2
(2.17.17)
6. As will be shown in the final section of this chapter, non-uniform plane waves
(field components being functions of all the three variables x, y, and z), are not
TEM.
Using eq. (M2), we can get
Ey= -(1/jdHz/dx
(2.17.18)
Ez= (1/jdHy/dx
(2.17.19)
(2.17.20)
that is
Ey/Hz=A/D=k0/ sqrt =sqrt (=377 ohm
(2.17.21)
_____________________________________________________________________________________
17
(2.17.22)
The quantity 120=377 ohm is known as the intrinsic impedance of free space. Also note
that, for the uniform plane wave, the magnitudes of Ey, Ez, Hy and Hz are constants over
the entire y-z plane.
2.18 Uniform plane wave in lossless dielectric
The plane wave discussion presented above for free-space is valid for any lossless
dielectric material too, with the following modifications:
1.
(2.18.1)
(2.18.2)
(2.18.3)
In this case, the electrical conductivity is not equal to zero In other words, the
dielectric medium has an ohmic loss. For this case, let us use the following equations
XH=( D/ t)+J
(M2)
D=*E
(1.2))
J=E
(1.7)
to write
_____________________________________________________________________________________
18
XH=jE
+E
= (+jE
(2.19.1)
It can be shown that the electric field vector E and the magnetic field vector H satisfy the
following wave equations
+ keff2)E=0
(2.19.2)
+ keff2)H=0
(2.19.3)
1-(j/
(2.19.4)
= 2eff
keff2=
eff=1-(j/
(2.19.6)
If we re-write q. (2.19.6) as
eff=real-jimaginary
(2.19.7)
real=
(2.19.8)
and
imaginary= /
(2.19.9)
Also note that we can define a loss tangent, tan ( ), of the medium by
tan ( )=imaginary/real=/
(2.19.10)
Eqs. (2.7.12) to (2.7.15) still apply after using keff instead of k0. Also, both Ex and Hx can
still be shown to be zero as was done in case of free space.
The solutions to the wave equation can be written as:
Ey (x,t)= A exp (-jkeffx)
(2.19.11)
_____________________________________________________________________________________
19
(2.19.12)
(2.19.13)
(2.19.14)
where A,B, C and D are the four arbitrary constants involved and exp (jt) is understood
to be there.
The effective wave number can, in general, be expressed as a complex propagation
constant using
jkeff= +j
(2.19.15)
This can be treated as a special case of the preceding section in the sense that
>>> (2.20.1)
That is,
imaginary
>>> real
(2.20.2)
1-(j/
= 2eff
keff2=
where
eff 1-(j/
j/ approximately
(2.20.5)
_____________________________________________________________________________________
20
since
=j [sqrt(j]
(2.20.8)
Hence
sqrt (j)= exp (j/4)=cos(/4) + jsin((/4)
=[1/sqrt (2)] {1+j)
(2.20.9)
=j [sqrt(j]
=j[sqrt (2)] (1+j)
(2.20.10)
Eqs. (2.19.21)-(2.19.24) would still apply but the propagation constant will be given by
-jkeff= - --j
= [sqrt (2)] (1+j)
(2.20.11)
yielding
=attenuation constant=[sqrt (2)]
(2.20.12)
and
=phase constant=[sqrt (2)]
(2.20.13)
_____________________________________________________________________________________
21
is defined as the
(2.20.14)
(2.20.15)
Physically, skin depth can be interpreted as the characteristic depth of penetration, i.e.,
the distance the signal travels inside the conductor before its amplitude decays to
1/e=1/2.718 of the original value. The formula for calculating skin depth is
==sqrt [2/()]
(2.20.16)
The electric and magnetic fields inside the conductor can be shown to satisfy the wave
equation. Hence, the field expressions inside the conductor can be written as
Ex (x,t)= 0
(2.20.17)
(2.20.18)
(2.20.19)
Hx (x,t)= 0
(2.20.20)
(2.20.21)
(2.20.22)
where A,B, C and D are the four arbitrary constants involved and exp (jt) is understood
to be there.
For a perfect conductor ( tending to infinity), it is easy to see from the above
equations that all the field components inside the conductor are zero.
For more info on skin effect and skin depth, use Internet-based search engines. One
good link is http://www.mos.org/sln/toe/skineffect.html
_____________________________________________________________________________________
22
PROBLEM : Compute the skin depth of aluminum, copper, gold, and silver at 10GHz.
SOLUTION :
Aluminum : Conductivity 3.816x10**7 S/m, Skin Depth=8.14x10**-7 m
Copper : Conductivity 5.813x10**7 S/m, Skin Depth=6.60x10**-7 m
Gold : Conductivity 4.098x10**7 S/m, Skin Depth=7.86x10**-7 m
Silver : Conductivity 6.17x10**7 S/m, Skin Depth=6.40x10**-7 m
2.23 Generalized solution for plane wave in free space
So far we have assumed a uniform plane wave, i.e., a wave for which the electric and
magnetic fields depend on only the x co-ordinate. Now let us consider the case where the
E and H fields are functions of all the three coordinates, namely, x, y, and z.
Wave equations for E and H fields shall still be the same, namely
2
+ k02)H=0
(2.23.1)
and
+ k02)E=0
(2.23.2)
Ex/
x2)+ (
Ex/
y2)+(
Ex/
z2)+ k02Ex=0
(2.23.3)
Ey/
x2)+ (
Ey/
y2)+(
Ey/
z2)+ k02Ey=0
(2.23.4)
_____________________________________________________________________________________
23
Ez/
x2)+ (
Ez/
y2)+(
(2.23.5)
The wave equation for the magnetic field strength (eq. 2.23.1), similarly, yields
(
Hx/
x2)+ (
Hx/
y2)+(
Hx/
z2)+ k02Hx=0
(2.23.6)
Hy/
x2)+ (
Hy/
y2)+(
Hy/
z2)+ k02Hy=0
(2.23.7)
Hz/
x2)+ (
Hz/
y2)+(
Hz/
z2)+ k02Hz=0
(2.23.8)
To solve such partial differential equations, we use a technique known as the method of
separation of variables. In this we method, we assume that the unknown function,
Ex,Ey,Ez,Hx,Hy, or Hz, can be represented as a product of three independent functions for
each of the three coordinates x,y, and z. That is, we can write
Ex (x,y,z)=X(x)Y(y)Z(z)
(2.23.9)
(2.23.10)
(2.23.11)
(2.23.12)
(Y/Y)= - ky2
(2.23.13)
(Z/Z)= - kz2
(2.23.14)
where
_____________________________________________________________________________________
24
(2.23.15)
(2.23.16)
(d2Y/dy2)+ ky2Y=0
(2.23.17)
(d2Z/dz2)+ kz2Z=0
(2.23.18)
(2.23.19)
where A is the arbitrary constant involved and only the wave traveling in the forward
direction is considered.
Similarly, we can write
Ey (x,y,z)= B exp (-jkxx)exp (-jkyy)exp (-jkzz)
(2.23.20)
and
Ez (x,y,z)= C exp (-jkxx)exp (-jkyy)exp (-jkzz)
(2.23.21)
.E=0
(M3)
(2.23.22)
where the wavenumber vector k represents the direction of propagation and is given by
k=kxix + kyiy + kziz
(2.23.23)
and the amplitude vector E0 represents the amplitude of the electric field and is given by
_____________________________________________________________________________________
25
(2.23.24)
In other words, the electric field amplitude vector E0 must be perpendicular to the vector
k representing the direction of propagation.
We can also write the following vector relationship for the electric field:
E=E0 exp (-jk.r)
(2.23.25)
(2.23.26)
To derive the magnetic field, now we can use the following Maxwells equation
XE=-jB=-jH
(M1)
We get
-jH=
XE=
X(E0exp(-jk.r) )
(2.23.27)
f)xE
+f(
xE)
(2.23.28)
xE)
(2.23.29)
where in is a unit vector in the direction of propagation and is the intrinsic impedance
of free space, 120 ohms.
Carefully go through solved problems 4.1-4.10 (pp. 254-262) in the following book:
ELECTROMAGNETIC FIELD THEORY AND TRANSMISSION LINES by G.S.N.
Raju, Pearson Education, 2005.
_____________________________________________________________________________________
26