Electromagnetic Waves: H, Maxwell's Equations (

CHAPTER 14
Electromagnetic Waves
14.1 Introduction
Some wave solutions to Maxwell’s equations have already been
encountered in the Solved Problems of Chapter 13. The present chapter will
extend the treatment of electromagnetic waves. Since most regions of
interest are free of charge, it will be assumed that charge density ρ = 0.
Moreover, linear isotropic materials will be assumed, with D = E, B = μΗ,
and J = σΕ.
14.2 Wave Equations

With the above assumptions and with time dependence ejωt for both E and
H, Maxwell’s equations (Table 13-1) become
Taking the curl of (1) and (2),

Now, in Cartesian coordinates only, the Laplacian of a vector
satisfies the identity
Substitution for the “curl curls” and use of (3) and (4) yields the vector
wave equations
The propagation constant γ is that square root of γ2 whose real and

imaginary parts are positive:
with
14.3 Solutions in Cartesian Coordinates

The familiar scalar wave equation in one dimension,
has solutions of the form F = f(z — ut) and F = g(z + ut), where f and g are
arbitrary functions. These represent waves traveling with speed u in the +z
and – z directions, respectively. In Fig. 14-1 the first solution is shown at t =
0 and t = t1; the wave has advanced in the +z direction a distance of ut1 in
the time interval t1 For the particular choices
Fig. 14-1
harmonic waves of angular frequency ω are obtained:
in which β ≡ ω/u. Of course, the real and imaginary parts are also solutions
to the wave equation. One of these solutions, F = C sin(ωt) – βz), is shown
in Fig. 14-2 at t = 0 and t = π/2ω. In this interval the wave has advanced in
the positive z direction a distance d = u(π/2ω) = π/2β. At any fixed t, the
waveform repeats itself when x changes by 2π/β; the distance
Fig. 14-2
is called the wavelength. The wavelength and the frequency f ≡ ω/2π enjoy
the relation
where T ≡ 1/f = 2π/ω is the period of the harmonic wave.

The vector wave equations of Section 14.2 have solutions similar to
those just discussed. Because the unit vectors ax, ay, and az in Cartesian
coordinates have fixed directions, the wave equation for H can be rewritten
in the form
14.4 Plane Waves

Of particular interest are solutions (plane waves) that depend on only one
spatial coordinate, say z. Then the equation becomes
which, for an assumed time dependence ejωt, is the vector analog of the one-
dimensional scalar wave equation. Solutions are as above, in terms of the
propagation constant γ.
The corresponding solutions for the electric field are
The fixed unit vectors aH and aE are orthogonal and neither field has a
component in the direction of propagation. This being the case, one can
rotate the axes to put one of the fields, say E, along the x axis. Then from
Maxwell’s equation (2) it follows that H will lie along the ±y axis for
propagation in the ±z direction.
EXAMPLE 1. Given the field E = E0e–γzaE (time dependence suppressed),

show that E can have no component in the propagation direction, +az.
The Cartesian components of aE are found by projection:
From ∇ · E = 0,
which can hold only if aE . az = 0. Consequently, E has no component in az.

The plane wave solutions obtained above depend on the properties μ,
, and σ of the medium, because these properties are involved in the
propagation constant γ.
14.5 Solutions for Partially Conducting Media

For a region in which there is some but not much conductivity (e.g., moist
earth, seawater), the solution to the wave equation in E is taken to be
Then, from (2) of Section 14.2,
The ratio E/H is characteristic of the medium (it is also frequency-

dependent). More specifically for waves E = Exax H = Hyay which
propagate in the +z direction, the intrinsic impedance, η, of the medium is
defined by
Thus,
where the correct square root may be written in polar form, |η| , with
(If the wave propagates in the –z direction, Ex/Hy = –η. In effect, γ is

replaced by –γ and the other square root used.)
Inserting the time factor ejωt and writing γ = α+jβ results in the
following equations for the fields in a partially conducting region:
The factor e–αz attenuates the magnitudes of both E and H as they propagate
in the +z direction. The expression for α, (5) of Section 14.2, shows that
there will be some attenuation unless the conductivity σ is zero, which
would be the case only for perfect dielectrics or free space. Likewise, the
phase difference θ between E(z, t) and H(z, t) vanishes only when σ is zero.
The velocity of propagation and the wavelength are given by
If the propagation velocity is known, λf = u may be used to determine the

wavelength λ. The term (σ/ω )2 has the effect of reducing both the velocity
and the wavelength from what they would be in either free space or perfect
dielectrics, where σ = 0. Observe that the medium is dispersive: waves with
different frequencies ω have different velocities u.
14.6 Solutions for Perfect Dielectrics

For a perfect dielectric, σ = 0, and so
Since α = 0, there is no attenuation of the Ε and Η waves. The zero angle

on η results in Η being in time phase with Ε at each fixed location.
Assuming Ε in ax and propagation in az, the field equations may be
obtained as the limits of those in Section 14.5
The velocity and the wavelength are
Solutions in Free Space.

Free space is nothing more than the perfect dielectric for which
For free space, η = η0 ≈ 120π Ω and u = c ≈ 3 × 108 m/s.
14.7 Solutions for Good Conductors; Skin Depth

Materials are ordinarily classified as good conductors if σ ω in the
range of practical frequencies. Therefore, the propagation constant and the
intrinsic impedance are
It is seen that for all conductors the Ε and Η waves are attenuated.
Numerical examples will show that this is a very rapid attenuation. α will
always be equal to β. At each fixed location Η is out of time phase with Ε
by 45° or π/4 rad. Once again assuming Ε in ax and propagation in az, the
field equations are, from Section 14.5,
Moreover,
The velocity and wavelength in a conducting medium are written here in

terms of the skin depth or depth of penetration,
EXAMPLE 2. Assume a field E = 1.0e–αzej(ωt –βz) ax (V/m), with f = ω/2π

= 100 MHz, at the surface of a copper conductor, σ = 58 MS/m, located at z
> 0, as shown in Fig. 14-3. Examine the attenuation as the wave propagates
into the conductor.
Fig. 14-3
At depth z the magnitude of the field is

where
Thus, after just 6.61 micrometers the field is attenuated to e–1 = 36.8% of its
initial value. At 5δ or 33 micrometers, the magnitude is 0.67% of its initial
value—practically zero.
14.8 Interface Conditions at Normal Incidence

When a traveling wave reaches an interface between two different regions,
it is partly reflected and partly transmitted, with the magnitudes of the two
parts determined by the constants of the two regions. In Fig. 14-4, a
traveling E wave approaches the interface z = 0 from region 1, z < 0. Ei and
Er are at z = – 0, while Et is at z = +0 (in region 2). Here, i signifies
“incident,” r “reflected” and t “transmitted.” Normal incidence is assumed.
The equations for E and H can be written as
Fig. 14-4
One of the six constants—it is almost always —may be taken as real.
Under the interface conditions about to be derived, one or more of the
remaining five may turn out to be complex.
With nominal incidence, E and H are entirely tangential to the interface,

and thus are continuous across it. At z = 0 this implies
Furthermore, the intrinsic impedance in either region is equal to ±Ex/Hy (see

Section 14.5).
The five equations above can be combined to produce the following ratios
in terms of the intrinsic impedances:
The intrinsic impedances for various materials were examined earlier. They
are repeated here for reference.
EXAMPLE 3. Traveling E and H waves in free space (region 1) are
normally incident on the interface with a perfect dielectric (region 2) for
which r = 3.0. Compare the magnitudes of the incident, reflected, and
transmitted E and H waves at the interface.
14.9 Oblique Incidence and Snell’s Laws

An incident wave that approaches a plane interface between two different
media generally will result in a transmitted wave in the second medium and
a reflected wave in the first. The plane of incidence is the plane containing
the incident wave normal and the local normal to the interface; in Fig. 14-5
this is the xz plane. The normals to the reflected and transmitted waves also
lie in the plane of incidence. The angle of incidence θi, the angle of
reflection θr, and the angle of transmission θt—all defined as in Fig. 14-5—
obey Snell’s law of reflection, and Snell’s law of refraction,
Fig. 14-5
and Snell’s law of refraction,
EXAMPLE 4. A wave is incident at an angle of 30° from air to Teflon, r

= 2.1. Calculate the angle of transmission, and repeat with an interchange of
the regions.
Since μ1 = μ2,
From Teflon to air,

Supposing both media of the same permeability, propagation from the
optically denser medium ( 1 > 2) results in θt > θi. As θi increases, an
angle of incidence will be reached that results in θt = 90°. At this critical
angle of incidence, instead of a wave being transmitted into the second
medium there will be a wave that propagates along the surface. The critical
angle is given by
EXAMPLE 5. The critical angle for a wave propagating from Teflon into
free space is
14.10 Perpendicular Polarization

The orientation of the electric field E with respect to the plane of incidence
determines the polarization of a wave at the interface between two different
regions. In perpendicular polarization, E is perpendicular to the plane of
incidence (the xz plane in Fig. 14-6) and is thus parallel to the (planar)
interface. At the interface,
Fig. 14-6
and
Note that for normal incidence, θi = θt = 0° and the expressions reduce to

those found in Section 14.9.
It is not difficult to show that if μ1 = μ2,
Hence, a perpendicularly polarized incident wave suffers either partial or

total reflection.
14.11 Parallel Polarization

For parallel polarization, the electric field vector E lies entirely within the
plane of incidence, the xz plane as shown in Fig. 14-7. (Thus E assumes the
role played by H in perpendicular polarization.) At the interface,
Fig. 14-7
and
In contrast to perpendicular polarizations, if μ1 = μ2 there will be a

particular angle of incidence for which there is no reflected wave. This
Brewster angle is given by
EXAMPLE 6. The Brewster angle for a parallel-polarized wave traveling

from air into glass for which r = 5.0 is
14.12 Standing Waves
When waves traveling in a perfect dielectric (σ1 = α1 = 0) are normally
incident on the interface with a perfect conductor (σ2 = ∞, η2 = 0), the
reflected wave in combination with the incident wave produces a standing
wave. In such a wave, which is readily demonstrated on a clamped taut
string, the oscillations at all points of a half-wavelength interval are in time
phase. The combination of incident and reflected waves may be written
Since η2 = 0, Er0/Ei0 = – 1 and
Taking the real part,
The standing wave is shown in Fig. 14-8 at time intervals of T/8, where T =
2π/ω is the period. At t = 0, E = 0 everywhere; at t = 1(T/8), the endpoints
of the E vectors lie on sine curve 1; at t = 2(T/8), they lie on sine curve 2;
and so forth. Sine curves 2 and 6 form an envelope for the oscillations; the
amplitude of this envelope is twice the amplitude of the incident wave. Note
that adjacent half-wavelength segments are 180° out of phase with each
other.
Fig. 14-8
14.13 Power and the Poynting Vector

Maxwell’s first equation for a region with conductivity σ is written and then
E is dotted with each term
where, as usual, E2 = E. E. The vector identity ∇ · (A × B) = B · (∇ × A) —

A · (∇ × B) is employed to change the left side of the equation.
By Maxwell’s second equation,

Similarly,
Substituting and rearranging terms,
Integration of this equation throughout an arbitrary volume v gives
where the last term has been converted to an integral over the surface of v
by use of the divergence theorem.
The integral on the left has the units of watts and is the usual ohmic
term representing energy dissipated per unit time as heat. This dissipated
energy has its source in the integrals on the right. Because E2/2 and μH2/2
are the densities of energy stored in the electric and magnetic fields,
respectively, the volume integral (including the minus sign) gives the
decrease in this stored energy. Consequently, the surface integral (including
the minus sign) must be the rate of energy entering the volume from
outside. A change of sign then produces the instantaneous rate of energy
leaving the volume:
where = E × H is the Poynting vector, the instantaneous rate of energy

flow per unit area at a point.
In the cross product that defines the Poynting vector, the fields are
supposed to be in real form. If, instead, E and H are expressed in complex

Electromagnetic Waves: H, Maxwell's Equations (

Uploaded by

Copyright:

Available Formats

Electromagnetic Waves: H, Maxwell's Equations (

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Electromagnetic Waves: H, Maxwell's Equations (

Uploaded by

Copyright:

Available Formats

CHAPTER 14

14.2 Wave Equations

Taking the curl of (1) and (2),

satisfies the identity

The propagation constant γ is that square root of γ2 whose real and

14.3 Solutions in Cartesian Coordinates

harmonic waves of angular frequency ω are obtained:

where T ≡ 1/f = 2π/ω is the period of the harmonic wave.

14.4 Plane Waves

The corresponding solutions for the electric field are

EXAMPLE 1. Given the field E = E0e–γzaE (time dependence suppressed),

which can hold only if aE . az = 0. Consequently, E has no component in az.

14.5 Solutions for Partially Conducting Media

Then, from (2) of Section 14.2,

The ratio E/H is characteristic of the medium (it is also frequency-

(If the wave propagates in the –z direction, Ex/Hy = –η. In effect, γ is

If the propagation velocity is known, λf = u may be used to determine the

14.6 Solutions for Perfect Dielectrics

Since α = 0, there is no attenuation of the Ε and Η waves. The zero angle

The velocity and the wavelength are

Solutions in Free Space.

For free space, η = η0 ≈ 120π Ω and u = c ≈ 3 × 108 m/s.

14.7 Solutions for Good Conductors; Skin Depth

The velocity and wavelength in a conducting medium are written here in

EXAMPLE 2. Assume a field E = 1.0e–αzej(ωt –βz) ax (V/m), with f = ω/2π

At depth z the magnitude of the field is

14.8 Interface Conditions at Normal Incidence

With nominal incidence, E and H are entirely tangential to the interface,

Furthermore, the intrinsic impedance in either region is equal to ±Ex/Hy (see

14.9 Oblique Incidence and Snell’s Laws

and Snell’s law of refraction,

EXAMPLE 4. A wave is incident at an angle of 30° from air to Teflon, r

From Teflon to air,

14.10 Perpendicular Polarization

Note that for normal incidence, θi = θt = 0° and the expressions reduce to

It is not difficult to show that if μ1 = μ2,

Hence, a perpendicularly polarized incident wave suffers either partial or

14.11 Parallel Polarization

In contrast to perpendicular polarizations, if μ1 = μ2 there will be a

EXAMPLE 6. The Brewster angle for a parallel-polarized wave traveling

Since η2 = 0, Er0/Ei0 = – 1 and

Taking the real part,

14.13 Power and the Poynting Vector

where, as usual, E2 = E. E. The vector identity ∇ · (A × B) = B · (∇ × A) —

By Maxwell’s second equation,

Substituting and rearranging terms,

Integration of this equation throughout an arbitrary volume v gives

where = E × H is the Poynting vector, the instantaneous rate of energy

You might also like