Uniform Plane Waves Page 1

Uniform Plane Waves

1 The Helmholtz Wave Equation
Let’s rewrite Maxwell’s equations in terms of E and H exclusively. Let’s assume the medium is
lossless (σ = 0). Let’s also assume we are operating in a source-free region, which means that
there are no current sources (J src = 0) or free charge densities (ρv = 0). Maxwell’s equations in
phasor form then reduce to

∇·E = 0 (1)
∇·H = 0 (2)
∇×E = −jωµH (3)
∇×H = jωεE. (4)

We note with the source-free assumption that the first two of Maxwell’s equations are no longer
independent relations, because they can be derived by taking the divergence of the last two
equations. Therefore, we only really need the curl equations in this derivation.
Taking the curl of Equation (3) and substituting in Equation (4), we obtain

∇ × ∇ × E = −jωµ∇ × H = ω 2 µεE. (5)

There is a vector identity that says that

∇ × ∇ × A = ∇(∇ · A) − ∇2 A, (6)

i.e., the curl of a curl of a vector is equal to the gradient of the divergence of the vector minus
the Laplacian of the vector. Using this identity, we obtain

∇2 E + ω 2 µεE = 0, (7)

an equation known as the vector Helmholtz equation for E. Repeating for H yields

∇2 H + ω 2 µεH = 0. (8)

Let’s focus our behaviour on the wave equation for E for now. Let’s rewrite the equation as

∇2 E + k 2 E = 0, (9)
p
where k = ω 2 µε is called the wavenumber (units of m−1 ). The vector wave equation is an
amazingly compact equation and is formidable when expanded: there are three Laplacians to
expand in three dimensions (x, y, z in the case of a Cartesian coordinate system). We can expand
the Laplacian operator to obtain
∂ 2E ∂ 2E ∂ 2E
∇2 E + k 2 E = + + + k 2 E = 0. (10)
∂x2 ∂y 2 ∂z 2
Prof. Sean Victor Hum ECE422: Radio and Microwave Wireless Systems
Uniform Plane Waves Page 2

2 Case 1: Plane Waves in a Lossless Medium

We’ll start off by assuming spatial variation in only one direction, say z. We will also assume that
an electric field with only an x component. The resulting equation is

∂ 2 Ex
+ k 2 Ex = 0. (11)
∂z 2
This is an ordinary differential equation with solutions

Ex (z) = E + e−jkz + E − ejkz . (12)

This is a phasor solution with a real-time representation of

E(z, t) = E + cos(ωt − kz) + E − cos(ωt + kz). (13)

Each component of the wave solution is known as a uniform plane wave since the wave is uniform
in the xy-plane, and the wave is infinite in extent. We see that the first term represents a wave
travelling in the +z direction. This is because if we consider a single point (a fixed phase point) on
the cosine curve, and try to stay with that point (i.e., keep the argument of the cosine constant),
as time increases, z must increase to maintain this condition. Hence, the wave travels in the +z
direction, while the second term refers to one travelling in the −z direction.
1
t = t0
t = t1
0.8
t = t2

0.6

0.4

0.2

−0.2

−0.4

−0.6

−0.8

−1
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35
z

The speed at which this phase point travels (the wave’s phase velocity ) is also determined using
the constant phase condition:
ωt − K
ωt − kz = K ⇒ z = (14)
k
and
dz ω 1
vp = = =√ . (15)
dt k µε
Prof. Sean Victor Hum ECE422: Radio and Microwave Wireless Systems
Uniform Plane Waves Page 3

If we use the permittivity and permeability of free space, we obtain the speed of light c =
√
1/ ε0 µ0 = 2.998 × 108 m/s.
The spatial period of the wave (the distance between two successive points of identical phase) is
the wavelength of the wave. It can be measured at a fixed point in time. Since the argument of
the cosine is periodic (i.e., it repeats every 2π),

[ωt − kz] − [ωt − k(z + λ)] = 2π, (16)

and the wavelength λ is equal to

2π 2πvp vp
λ= = = . (17)
k ω f
The first equality is known as the dispersion relation since it relates the wavenumber to the
wavelength. For the case considered here, there is a linear (proportional) relationship between the
two.
Finally, let’s find the relationship between the electric and magnetic fields for a plane wave. H
can be found using the appropriate curl equation (3) from Maxwell’s equations. The curl operator
is quite straightforward since there is no variation in x and y:
     
∂Ez ∂Ey ∂Ex ∂Ez ∂Ey ∂Ex
∇×E =  − x̂ + − ŷ + − ẑ (18)
∂y ∂z ∂z ∂x ∂x ∂y
∂Ey ∂Ex
= − x̂ + ŷ, (19)
∂z ∂z
and since Ey = 0,
∂Ex
∇×E = ŷ = −jωµH (20)
∂z
j ∂Ex
Hy = . (21)
ωµ ∂z
Evaluating the derivative,
j
(−jk) E + e−jkz − E − ejkz


Hy = (22)
ωµ
k
E + e−jkz − E − ejkz .


= (23)
ωµ
√ q
k ω µε ε
Since ωµ = ωµ
= µ
, we can write the equation as

1  + −jkz
− E − ejkz ,

Hy = E e (24)
η
where r
µ
η= (25)
ε
Prof. Sean Victor Hum ECE422: Radio and Microwave Wireless Systems
Uniform Plane Waves Page 4

is defined as the wave impedance of the plane wave in ohms. Hence, E and H for each travelling
wave are related through a simple multiplication operation: the magnetic field is equal to the
electric field divided by the wave impedance. The sign the product depends on the direction of
power flow according to the sign of the Poynting vector. This is analogous to the relationship
between voltage and current in guided-wave (transmission line) scenarios whereby we use the
characteristic
q impedance to relate the two quantities. Note that the wave impedance of free
space is µε0o = 377 Ω.

3 Case 2: Plane Waves in a Lossy Medium

In a lossless medium, σ = 0. However, a nonzero conductivity changes the Helmholtz equation
slightly and hence the plane wave solution. Maxwell’s curl equation for H becomes

∇ × H = jωD + J (26)
= jωεE + σE (27)
 σ
= jwεE 1 − j . (28)
ωε
Note the quantity σ/jωε is the ratio of the conduction current to the displacement current. The
σ
quantity tan δ = ωε is often called the loss tangent of the medium. The angle δ is the angle by
which the displacement current leads the total current density.
The resulting wave equation for E becomes
 σ
∇2 E + ω 2 µεE 1 − j = 0. (29)
ωε
p
Then effectively, the expression k 2 = ω 2 µε is been replaced with ω 2 µε[1 − j(σ/ωε)]. Let us
define a complex propagation constant of the medium as
r
√ σ
γ = α + jβ = jω µε 1 − j . (30)
ωε
The wave equation becomes
∂ 2 Ex
− γ 2 Ex = 0, (31)
∂z 2
which has solutions
Ex (z) = E + e−γz + E − eγz . (32)
Examining the positive travelling wave component, we see it has the form

E + e−γz = E + e−αz e−jβz , (33)

which in the time domain is represented as

E + e−αz cos(ωt − βz). (34)

We see that it is similar to the positive travelling wave in a lossless medium, with the exception
that there is an exponential damping factor in front of the harmonic term, indicating that the wave
Prof. Sean Victor Hum ECE422: Radio and Microwave Wireless Systems
Uniform Plane Waves Page 5

is attenuated as z increases. A similar damping exists for the negative travelling wave. Hence, in
a lossy medium, the wave is dissipated in the medium and is attenuated with distance.
If the loss is removed (σ = 0), α = 0 and the solution is the same as the lossless case with
√
γ = jk = jω µε. Therefore, k = β for lossless media and the two terms are often used
interchangeably.
The wave impedance for the lossy case is evaluated from Equation 21 and is found as follows:
j
(−γ) E + e−γz − E − eγz


Hy = (35)
ωµ
jωµ
η = , (36)
γ
which we note is a complex quantity since γ is complex.

4 Case 3: Plane Waves in a Good Conductor

σ
In the case of a medium with high conductivity such as a metal, ωε
 1 and
r r
√ σ √ σ
γ = jω µε 1 − j ≈ jω µε −j (37)
ωε ωε
p √
= j ωσµ (38)
r
ωσµ
= (1 + j) , (39)
2
√
since j = 16 45◦ = √12 + j √12 . The damping introduced by the e−αz clearly dominates now due
to the magnitude of α (the real part of γ). In fact, the plane wave will only penetrate the medium
to a depth r
1 2
δs = = , (40)
α ωσµ
which is defined as the skin depth of the conductor. At this depth, the amplitude of the plane wave
has decayed by an amount 1/e of 36.8% of its original value. At high frequencies this depth is vary
small and hence waves propagate along the “skin” of the conductor. This has practical importance
since the ohmic loss introduced by a metal can be exaggerated by the fact that microwave signals
only propagate in a portion of the conductor, versus the entire cross section.

5 General Plane Wave Solutions

We now re-consider the lossless case (Case 1), which developed a plane wave that was 1) polarized
in the x̂ direction, and 2) propagated along the z-axis. We now want to consider a general plane
wave that has arbitrary linear polarization, and travels in any direction. To solve the Helmholtz
equation
∂ 2E ∂ 2E ∂ 2E
∇2 E + k 2 E = + + + k 2 E = 0, (41)
∂x2 ∂y 2 ∂y 2

Prof. Sean Victor Hum ECE422: Radio and Microwave Wireless Systems
Uniform Plane Waves Page 6

we first notice that it reduces to three scalar equations,

∂ 2 Ex ∂ 2 Ex ∂ 2 Ex
∇2 Ex + k 2 Ex = + + + k 2 Ex = 0 (42a)
∂x2 ∂y 2 ∂z 2
∂ 2 Ey ∂ 2 Ey ∂ 2 Ey
∇2 Ex + k 2 Ey = + + + k 2 Ey = 0 (42b)
∂x2 ∂y 2 ∂z 2
∂ 2 Ez ∂ 2 Ez ∂ 2 Ez
∇2 Ex + k 2 Ez = + + + k 2 Ez = 0. (42c)
∂x2 ∂y 2 ∂z 2
The solution to the first equation, for example, is assumed to be written in the form

Ex (x, y, z) = f (x)g(y)h(z), (43)

that is, the wave solution can be expressed as a product of three independent functions, each acting
on a separate spatial variable. We are free to make this assumption and if any inconsistencies
arise in assuming such a form, we could attempt another form for the solution. Using the assumed
form, the wave equation for Ex becomes

∂ 2f ∂ 2g ∂ 2h
gh + f h + f g + k 2 f gh = 0. (44)
∂x2 ∂y 2 ∂z 2
Since f , g, and h are only functions of one spatial variable each, we can replace the partial
derivatives with ordinary derivatives. Doing this and dividing by f gh,

1 d2 f 1 d2 g 1 d2 h
+ + + k2 = 0 (45)
f dx2 g dy 2 h dz 2
or
1 d2 f 1 d2 g 1 d2 h
2
+ 2
+ 2
= −k 2 . (46)
f dx g dy h dz
Since each term on the left hand side is dependent on only one spatial variable, the sum of these
terms can only equal −k 2 if each term is a constant. Hence, we write

d2 f
= −kx2 f (47a)
dx2
d2 g
2
= −ky2 f (47b)
dy
d2 h
2
= −kz2 f (47c)
dz
We have already solve versions of these equations in our 1D example earlier. The only additional
constraint then, is that in addition our solution satisfying all three equations simultaneously, the
wavenumber components must also satisfy

kx2 + ky2 + kz2 = k 2 , (48)

Prof. Sean Victor Hum ECE422: Radio and Microwave Wireless Systems
Uniform Plane Waves Page 7

so that (46) is satisfied. This relation is called the consistentency relation or dispersion relation.
Most generally, we can write a solution satisfying both (46) and the dispersion relation as

Ex (x, y, z) = Ae−j(kx x+ky y+kz z) . (49)

Let’s simplify this expression by defining a vector wavenumber,

k = kx x̂ + ky ŷ + kz ẑ = k n̂. (50)

If we define a position vector

r = x x̂ + y ŷ + z ẑ, (51)
then we can express Ex as
Ex (x, y, z) = Ae−jk·r . (52)
Similarly,

Ey (x, y, z) = Be−jk·r (53)

Ez (x, y, z) = Ce−jk·r . (54)

Finally, let’s combine all the field components into a vector expression for the electric field,

E = (Ax̂ + B ŷ + C ẑ)e−jk·r = E 0 e−jk·r . (55)

With a potential solution for the Helmholtz equation, let us check it satisfies the divergence
relation in Maxwell’s equations by evaluating ∇ · E. To do this, we will make the use of the
following vector identity:
∇ · f A = A · ∇f + f ∇ · A (56)

∇ · E = E 0 · ∇ e−jk·r + e−jk·r (∇ · E 0 )


(57)
= E0 · (−jk) e−jk·r = 0. (58)

From this equation, we can see that that k · E 0 = 0 – that is, the wavenumber k and polarization
of the electric field are orthogonal. Since k indicates the direction the plane wave is travelling,
an important conclusion is that the electric field is orthogonal to the direction of propagation.
What about the magnetic field? Knowing E = E 0 e−jk·r , we can find the magnetic field by
evaluating
j
H= ∇ × E, (59)
ωµ
with the assistance of the vector identity

∇ × f A = (∇f ) × A + f ∇ × A. (60)

Prof. Sean Victor Hum ECE422: Radio and Microwave Wireless Systems
Uniform Plane Waves Page 8

j 
∇e−jk·r × E 0 + e−jk·r (∇ × E 0 )


H = (61)
ωµ
j
= (−jk)e−jk·r × E 0 (62)
ωµ
k n̂ −jk·r
= e × E0 (63)
ωµ0
1
= n̂ × E 0 e−jk·r (64)
η
1
= n̂ × E. (65)
η
This is not a surprising result: E and H are related through the wave impedance, and are
perpendicular to each other. Furthermore, H is also perpendicular to the direction of propagation.
Hence, E, H, and k all form a triplet related through the right-hand rule: that the direction of
k is equal to the direction of E × H. The right-handed triplet is illustrated below.

Since the electric and magnetic fields are orthogonal to the direction of propagation, we call such
a wave a transverse electromagnetic wave, which you may have heard about in an earlier field
and waves course. This contrasts to transverse electric (TE) and transverse magnetic (TM) wave
solutions that also exist in wave problems.

Prof. Sean Victor Hum ECE422: Radio and Microwave Wireless Systems