EECS 117

Lecture 21: Wave Propagation in Lossy Media and

Poyntings Theorem
Prof. Niknejad
University of California, Berkeley
University of California, Berkeley EECS 117 Lecture 21 p. 1/18
Time-Harmonic Wave Equation
Start by taking the curl of Faradays Eq.
(E) = jB
H = E + jE
(E) = j(E+ jE)
In a source free region, E = 0, and thus
(E) = ( E)
2
E =
2
E
We thus have Helmholtz equation

2
E
2
E = 0
Where
2
= j( + j) = + j
University of California, Berkeley EECS 117 Lecture 21 p. 2/18
Lossy Materials
In addition to conductive losses , materials can also
have dielectric and magnetic losses
A lossy dielectric is characterized by a complex
permittivity =
r
+ j
i
, where
i
arises due to the phase
lag between the eld and the polarization. Likewise
=
r
+ j
i
.
Most materials we study are weakly magnetic and thus

r
.
For now assume that ,, and are real scalar
quantities
Thus
=
_
(
2
)(1 +

j
)
University of California, Berkeley EECS 117 Lecture 21 p. 3/18
Propagation Constant and Loss
Lets compute the real and imaginary part of
= j

_
1 j

_
1
2
Consider (1 jh) = re
j
, so that
y =
_
1 jh =

re
j/2
Note that tan = h, and r =

1 + h
2
. Finally
cos

2
=
_
1 + cos
2
=

1 +
1
r
2
=
_
r + 1
2r
University of California, Berkeley EECS 117 Lecture 21 p. 4/18
Propagation Constant and Loss (cont)
Similarly
sin

2
=
_
1 cos
2
=
_
r 1
2r
y =

re
j/2
=
_
r + 1
2
j
_
r 1
2
= a + jb
Using the above manipulations, we can now break
into real and imaginary components
= j

(a + jb) =

b + j

a = + j
=

r 1

2
_
University of California, Berkeley EECS 117 Lecture 21 p. 5/18
Propagation Constant and Loss (nal)
We have now nally shown that
=
_

2
_
_
1 +
_

_
2
1
_
1/2
=
_

2
_
_
1 +
_

_
2
+ 1
_
1/2
Its easy to show that the imaginary part of can be
lumped into an effective conductivity term
In practice, most materials are either low loss, such that

1, or good conductors, such that

e

1
In these extreme cases, simplied versions of the above
equations are applicable
University of California, Berkeley EECS 117 Lecture 21 p. 6/18
Effective Dielectric Constant
We can also lump the conductivity into an effective
dielectric constant
H = E + jE = j
e
E
where
e
= j/. In the low loss case, this is a
good way to include the losses
When or become complex, the wave impedance is
no longer real and the electric and magnetic eld fall out
of phase. Since H = E/
c

c
=
_

e
=
_

j/
=
_

_
1 j

University of California, Berkeley EECS 117 Lecture 21 p. 7/18

Propagation in Low Loss Materials
If

1, then our equations simplify

= j

_
1 j
1
2

_
To rst order, the propagation constant is unchanged by
the losses (
e
= +

)
=

=
1
2

e
_

A more accurate expression can be obtained with a 1st

order expansion of (1 + x)
n
=

_
1 +
1
8
_

_
2
_
University of California, Berkeley EECS 117 Lecture 21 p. 8/18
Propagation in Conductors
As we saw in the previous lecture, this approximation is
valid when

1
= + j =
_
j = e
j45

= =
_

2
The phase velocity is given by v
p
= /
v
p
=
_
2

This is a function of frequency! This is a very dispersive

medium.
University of California, Berkeley EECS 117 Lecture 21 p. 9/18
Waves in Conductors
The wavelength is given by
=
v
p
f
= 2
_

f
Example: Take = 10
7
S/m and f = 100 MHz. Using
the above equations
= 10
4
m
v
p
= 10
4
m/s
Note that
0
= 3 m in free-space, and thus the wave is
very much smaller and much slower moving in the
conductor
University of California, Berkeley EECS 117 Lecture 21 p. 10/18
Energy Storage and Loss in Fields
We have learned that the power density of electric and
magnetic elds is given by
w
m
=
1
2
E D =
1
2
E
2
w
m
=
1
2
H B =
1
2
H
2
Also, the power loss per unit volume due to Joule
heating in a conductor is given by
p
loss
= E J
Using J = H
D
t
, this can be expressed as
E J = E H

t
(D)
University of California, Berkeley EECS 117 Lecture 21 p. 11/18
Poynting Vector
We will demonstrate that the Poynting vector EH
plays an important role in the energy of an EM eld.
(EH) = H (E) E (H)
E J = H (E) (EJ) E
D
t
= H (
B
t
) E
D
t
(EH)
H
B
t
= H
_
H
t
_
=
1
2
H H
t
=
1
2
|H|
2
t
E
D
t
= E
_
E
t
_
=
1
2
E E
t
=
1
2
|E|
2
t
University of California, Berkeley EECS 117 Lecture 21 p. 12/18
Poyntings Theorem
Collecting terms we have shown that
E J =

t
_
1
2
|H|
2
_

t
_
1
2
|E|
2
_
(EH)
Applying the Divergence Theorem
_
V
E JdV =

t
_
V
_
1
2
|H|
2
+
1
2
|E|
2
_
dV
_
S
E HdV
The above equation can be re-stated as
power
dissipated in
volume V (heat)
=
rate of change
of energy
storage in
volume V

a surface
integral over the
volume of
E H
University of California, Berkeley EECS 117 Lecture 21 p. 13/18
Interpretation of the Poynting Vector
We now have a physical interpretation of the last term in
the above equation. By the conservation of energy, it
must be equal to the energy ow into or out of the
volume
We may be so bold, then, to interpret the vector
S = E H as the energy ow density of the eld
While this seems reasonable, its important to note that
the physical meaning is only attached to the integral of
S and not to discrete points in space
University of California, Berkeley EECS 117 Lecture 21 p. 14/18
Current Carrying Wire
I I
B =

0
I
2r
Consider the above wire carrying a uniform current I
From circuit theory we know that the power loss in the
wire is simply I
2
R. This is easily conrmed
P
L
=
_
V
E JdV =
_
V
1

J JdV =
1
A
2

int
V
I
2
dV
P
L
=
A
A
2

I
2
=

A
I
2
University of California, Berkeley EECS 117 Lecture 21 p. 15/18
Energy Stored around a Wire Section
Lets now apply Poyntings Theorem. Since the current
is dc, we can neglect all time variation

t
= 0 and thus
the energy storage of the system is xed in time.
The magnetic eld around the wire is simply given by
H =

I
2r
The electric eld is proportional to the current density.
At the surface of the wire
E =
1

J =
I
A
z
University of California, Berkeley EECS 117 Lecture 21 p. 16/18
Power Loss in Wire
I I
H
E
S
The Poynting vector at the surface thus points into the
conductor
S = E H =
I
A
z

I
2r
=
rI
2
2rA
The power ow into the wire is thus given by
_
s
S ds =
_

0
_
2
0
I
2
2rA
rddz = I
2
R
This result conrms that the energy owing into the wire
from the eld is heating up the wire.
University of California, Berkeley EECS 117 Lecture 21 p. 17/18
Sources and Fields
This result is surprising because it hints that the signal
in a wire is carried by the elds, and not by the charges.
In other words, if a signal propagates down a wire, the
information is carried by the elds, and the current ow
is impressed upon the conductor from the elds.
We know that the sources of EM elds are charges and
currents. But we also know that if the conguration of
charges changes, the elds carry this information
University of California, Berkeley EECS 117 Lecture 21 p. 18/18