Solutions to Maxwell’s equations

• let’s try a simple guess at a solution

– “time harmonic” solutions
?   ? 
E  Es  x, y, z  e jt H  H s  x, y, z  e jt
• let’s substitute into Maxwell’s equations, and see if this
could work 
 B
 E  
t

• left hand side of equation

 

  E s  x, y , z  e 
jt

 e   Es  x , y , z 
jt

 Solutions to Maxwell’s ?   ? 
E  Es  x , y , z  e jt
H  H s  x, y, z  e jt
equations  
  
  Es  x, y , z  e jt  e jt   Es  x, y, z  
• right hand side of curl E equation
 


B

  r o H  
   r o

 H s  x, y, z  e jt 
t t t
   e jt  
   r  o H s  x, y , z    jr o H s  x, y, z  e jt
t  

this is just H

• now substitute
 
 
e   Es  x, y, z    jr o H s  x, y, z  e jt
j t

 
 
  Es  x, y, z    jr o H s  x, y, z 

• if the time independent forms Es and Hs satisfy this

equation, they will satisfy Maxwell
 Solutions to Maxwell’s equations
?   ? 
E  Es  x, y , z  e jt H  H s  x, y, z  e jt

• let’s keep going

– substitute into next Maxwell equation

  D
 H  J 
t
• left hand side of equation
 

  H s  x, y , z  e jt
 
 e   H s  x, y , z 
jt

• left side
  
J   E   Es  x, y, z  e jt
  notice I took the material
Js properties out of the time
derivative
 


D   r  o E


  r o

 Es  x, y, z  e jt  
 j r  o Es  x, y, z  e jt
t t t
     
  H s  r   J s  r   j r  o Es  r 
 Maxwell’s equations, time harmonic form
• so assuming that none of the materials change in time, time
independent fields Es and Hs that satisfy
   
 
 
  Es  r    jr o H s  r  E  x, y, z , t   Es  x, y, z  e jt
       
  H s  r   J s  r   j r  o Es  r  H  x, y, z, t   H s  x, y, z  e jt

– when multiplied by exp(jt), the product will satisfy the full

time dependent set of Maxwell equations
• so now what???
• can we use these equations to predict/understand any
other phenomena?
– let’s try to solve this set of coupled partial differential
equations…
 Summary of electromagnetics: Maxwell’s equations
• summarizing everything we have so far, valid even if things are
changing in time


 B  
Faraday’s law  E   D  v Gauss’s law

t

  D  
Ampere’s law  H  J  B  0
t
• plus material properties

     
 E
D   H
B  J E
 r o r o
 Summary of electromagnetics: Maxwell’s equations
in “time harmonic form”
• summarizing everything we have so far, assuming time
harmonic behavior, and using Ohm’s law for J
   
E  x, y, z, t   Es  x, y, z  e jt H  x, y, z, t   H s  x, y, z  e jt
     
 
  Es  r    j H s  r  D  v
     
  H s  r     j  Es  r  B  0
• plus (time independent) material properties

     
 E
D   H
B  J E
 r o r o
 Spatial forms that solve Maxwell
• let’s assume the time dependence is taken care of using the
phasor approach (i.e., time dependence is exp(jt))
• what about the Es and Hs (the spatial part)?
– as usual, let’s guess…
– special case: let’s assume there are no components of the fields in
the z direction, i.e.,
• NOTE: we are NOT assuming that E & H are independent of z!!
• the fields are “transverse” to the z direction
• the field is “contained in the x-y plane
• the only field components are x^ and/or y^ directed:
– Ex(x, y, z), Ey(x, y, z), Hx(x, y, z), Hy(x, y, z)
– or ^ and/or ^ if its cylindrical
– under these conditions, what happens to Maxwell’s equations?
• let’s look at the curl first

 0
  0

    Fz Fy   Fx  Fz   Fy Fx 
curl F    F     xˆ   z  x  yˆ   x  y  zˆ
 y z     
   
 Maxwell for transverse-to-z fields
   Fy   Fx   Fy Fx 
curl F    Ftrans  z     xˆ    ˆ
y     zˆ
 z   z   x y 
• but from the first Maxwell equation
   
 
  Es  r    j H s  r 
– for the transverse-to-z case this reduces to

 
 
  E y   Ex   E y Ex 
  Etrans    xˆ    yˆ     zˆ   j  H x xˆ  H y yˆ   0 zˆ 

 z   z   x y 
 by
assumption 
 that H is trans 

  E   E
  Etrans     y    j H x   Etrans    x 
   j H y
 x   y  z
 z  
  E E 
  Etrans    y  x   0
 z
 x y 
 Maxwell for transverse-to-z fields
   Fy   Fx   Fy Fx 
curl F    Ftrans  z     xˆ    ˆ
y     zˆ
 z   z   x y 
• from the second Maxwell equation
   
  H s  r     j  Es  r 
– for the transverse-to-z case this reduces to

 
 
  H y   H x   H y H x  
  H trans    xˆ    yˆ     zˆ    j   Ex xˆ  E y yˆ   0 zˆ 
 z   z   x y  
 by
assumption 
 that E is trans 

  H y    H x 
  H trans    
 x      j  E x

   H 
trans    z     j  E y
 z  y
 
  H y H x 
  H trans     0
 z
 x y 
 Maxwell, time harmonic, transverse-to-z
• collecting all the terms,
– assuming time harmonic solutions
– using Ohm’s law
– assuming there is no component of either E or H in the z direction
• Maxwell’s equations reduce to
E y H y E y Ex
 j H x     j  E x  0
z z x y
Ex H x H y H x
  j H y    j  E y  0
z z x y
• things to notice
– Ey is connected to Hx via d/dz and 
– Ex is connected to Hy via d/dz and 
– Hy is connected to Ex via d/dz and 
– Hx is connected to Ey via d/dz and 
– Ey and Ex are connected via d/dx and d/dy
– Hy and Hx are connected via d/dx and d/dy
 A little more fiddling with time harmonic transverse-
to-z Maxwell
• time harmonic solutions, using Ohm’s law, assuming there is
no component of either E or H in the z direction
• can we get rid of the “mixture” or E and H?

  E y  2 Ey H x H x
  j H x   j    j  E y
z  z  z 2 z z

2 Ey 2 Ey
 j    j  E y   j   j  E y  0
z 2
z 2

  Ex   2 Ex H y H y
  j H   j     j  Ex
z  z 
y
 z 2 z z

 2 Ex  2 Ex
  j     j  Ex   j   j  Ex  0
z 2 z 2
 A little more fiddling with time harmonic transverse-
to-z Maxwell
• time harmonic solutions, using Ohm’s law, assuming there is
no component of either E or H in the z direction
• can we get rid of the “mixture” or E and H?

  H x  2H x E y E y
    j  E     j   j H x
z  z
y
 z 2 z z

2H x 2H x
   j  j H x   j   j  H x  0
z 2 z 2

  H y  2H y Ex Ex

     j  Ex      j    j H y
z  z  z 2 z z

2H y 2H y
    j    j H y   j   j  H y  0
z 2
z 2
 “Wave equation” form of the time harmonic
transverse-to-z Maxwell’s equations
• assuming time harmonic solutions, using Ohm’s law, and that
there is no component of either E or H in the z direction
– Maxwell’s equations reduce to

2 Ey  2 Ex
 j   j  E y  0  j   j  Ex  0
z 2
z 2

E y Ex
 0
x y

2H x 2H y
 j   j  H x  0  j   j  H y  0
z 2 z 2

H y H x
 0
x y
 “Wave equation” form of the time harmonic
transverse-to-z Maxwell’s equations
• to get a better handle on all this, let’s make another simplifying
assumption
– E points in the x^ direction only
• then the only thing we have left is

 2 Ex
 j   j  Ex  0
z 2

• do we recognize a solution to an equation of this form?

2 F
  2
F 0 F  (constant )  e z
z 2
 Transverse “plane waves”
• “plane waves”
– “all” the fields are “contained” in a plane perpendicular to a
particular direction
• here we made that direction z^
• this still looks pretty abstract…
• let’s try a simple guess at a solution, and check to see if we can
satisfy these new equations
– proposal: uniform plane wave
 ?? z
Etrans  Exo e xˆ
– E points in the x^ direction only
– Exo is a CONSTANT, independent of time, place, frequency!
• this is what makes it a “uniform” plane wave
–  is also a “constant,” independent of time and place, but I’m not
sure about 
– substitute and see what we get…
 Uniform plane wave substitution

• in the transverse-to-z version of Maxwell we needed the derivative

  Ex    Exo e z 
 Exo e z
z z
 assumed
Etrans  Exo e z xˆ

• and one of Maxwell’s equations was

Ex
  j H y Exo e z   j H y
z
Exo e z 
Hy   j Exo e z
 j  
 
H yo

H yo  j Exo


• so we have the y component of H, what about the x component?

  assumed
Uniform plane wave substitution Etrans  Exo e z xˆ

E y
• the x component of H is given by  j H x
z

• but by assumption, Ey = 0, so Hx = 0!

 assumed
• so far we have assumed Etrans  Exo e z xˆ

– and from this Maxwell requires

 z    z
H trans  H yo e y   j
ˆ Exo  e yˆ
  
– so the presence of an electric field requires the presence of a magnetic
field!
• we still have a couple of equations to check!
– maybe we can figure out what the constant  is?
  assumed
Uniform plane wave substitution: Etrans  Exo e z xˆ
what is ?    
H trans  H yo e z yˆ   j Exo  e z yˆ
• another of Maxwell’s equations was   

 
Hy

H y  
    j  E x  z 
z  j Exo  e 
    
 
      j E e z
 xo
z E x

 1
j Exo e z     j  E xo e z j  2     j 
 

 2  j   j 
 Uniform plane wave solution to Maxwell’s equations

• so we now have the whole spatial solution!!!

 assumed     z
Etrans  Exo e z xˆ H trans   j Exo  e yˆ
  
 2  j   j 
• and the complete, time harmonic solution is
 assumed     jt  z
E plane  Exo e jt  z xˆ H plane   j Exo  e yˆ
wave wave   
   j   j 
–  is called the “complex propagation constant”
– note that E and H are perpendicular to each other
• this is really a big deal!!
 Lossless uniform plane wave solution to Maxwell’s equations
 assumed    
E plane  Exo e jt  z xˆ H plane   j Exo  e jt  z yˆ    j   j 
wave wave   
• to understand what this thing looks like, let’s make an
additional assumption: the region of space we are in is an
insulator
– =0
• then we get a simplification in 

   j  0  j     2    j  

b
– gamma is purely imaginary,  = jb !

    j   jt  z   
E plane  E xo e j t  b z  xˆ H plane   j Exo  e yˆ    Exo  e j t  b z  yˆ
   
wave wave
   
 Lossless, uniform plane wave solution to Maxwell’s equations
     j t  b z 
E plane  E xo e j t  b z  xˆ H plane    Exo  e yˆ b   
wave wave   
• this is also a really big deal!!
• what does this thing look like????
• b is called the phase or propagation constant
– note the units on b are inverse length
– every time you move a distance of 2p/b along z, the exp function
repeats itself
   2p   
exp  j  t  b  z      exp  j t   b z  2p     exp  j t  b z  
   b   
– so this function is periodic in space
– the spatial repeat interval of a periodic function is usually called
the wavelength l
2p 2p
l 
b  
 Lossless, uniform plane wave solution to Maxwell’s
equations
   j t  b z  b   
 
E plane  E xo e j t  b z  xˆ H plane    Exo  e yˆ 2p 2p
wave wave    l 
b  
• what does this thing look like????
• what if we wanted to “follow” a plane of constant phase as the
clock ticks
– in other words, if time t is increasing, how should we move along
the z axis to keep the phase of the exponential a constant?
– let’s look at the –b case first, where zo is the observation point
needed to keep the total phase at the constant value o

 o
 exp  j t  b zo    t  b zo  o zo  t
b b
– the “constant phase point” speed (the “phase velocity” vp) is just
dzo/dt
dzo   1
 vp   
dt b   
 Lossless, uniform plane wave solution to Maxwell’s
equations
b   
     j t  b z 
E plane  E xo e j t  b z  xˆ H plane    Exo  e yˆ
l
2p

2p
wave wave    b  
• let’s continue to “follow” a plane of constant phase as time t is
increasing, but this time consider the +b case
o 
 exp  j t  b zo    t  b zo  o zo   t
b b
dzo 
– the phase velocity vp is still dzo/dt  vp  
dt b

• so if we consider the (t + bz) solution, the phase front is

“moving” in the -z^ direction
• if we consider the (t – bz) solution, the phase front is “moving”
in the +z^ direction
 1
• regardless, the speed is v p  
b 
 Lossless, uniform plane wave solution to Maxwell’s
equations
• so in a region of space that has zero conductivity, one possible
solution to Maxwell’s equations is
 2p 2p
E plane  E xo e j t  b z  xˆ b    l
b

wave  

    j t  b z   1
H plane    Exo  e yˆ vp  
 b 
wave  
• this is called a uniform transverse electromagnetic plane wave
– a wave because it is periodic in time and space
– a plane wave because a “surface of constant phase” is a flat plane,
and is “propagating” in the z^ direction
– electromagnetic because E and H are intimately connected
– transverse because the E and H fields are contained completely in
the x-y plane, transverse to the direction of propagation
– uniform because the magnitude of the field is constant wrt x and y
 Units check
• we have 2p 2p
– : 1/sec b    l 
– : farad/meter
b  
– : henry/meter
• : (henry)·(farad)/m2 = (Ω·sec)·(sec/Ω)/m2 = sec2/m2
– b: [sec/m]/[sec] = 1/m = 1/distance
 1
• we have vp  
b 
– : 1/sec
– b: [sec/m]/[sec] = 1/m = 1/distance
– vp: [1/sec]/[1/m] = m/sec (speed)
– : farad/meter
– : henry/meter
• : (henry)·(farad)/m2 = (Ω·sec)·(sec/Ω)/m2 = sec2/m2
– vp: 1/[sec/m] = m/sec
• numbers: free space, o = 8.854x10-12 F/m, o = 4px10-7 H/m
– vp = 3x108 m/sec
– @  = 2px1010sec-1 (f = 10GHz): l = 3 cm = 0.03m
 Lossless, uniform plane wave solution to Maxwell’s
equations
• let’s look at the relationship between E and H a little more
closely
     j t  b z 
E plane  E xo e j t  b z  xˆ H plane    Exo  e yˆ
wave wave   
• we found that if we pick the (t - bz) the “wave” is “traveling” (or
“propagating”) in the +z^ direction
     j t  b z 
E plane  E xo e j t  b z  xˆ H plane    Exo  e yˆ
wave wave   
– for this choice, assuming Exo is positive
• E points in the +x^ direction
• H points in the +y^ direction
• note that the direction of propagation is x^  y^ = z^
i.e., its the same as  
Eo  H o
 Lossless, uniform plane wave solution to Maxwell’s
equations
• if we pick the (t + bz) the “wave” is “traveling” (or
“propagating”) in the -z^ direction
     j  t  b z 
E plane  E xo e j t  b z  xˆ H plane    Exo  e yˆ
wave wave   
– for this choice, assuming Exo is positive
• E points in the +x^ direction
• H points in the -y^ direction
• note that the direction of propagation is x^  (-y^) = -z^
i.e., its still the same as  
Eo  H o

• this type of solution to Maxwell’s equations, with no field

components in the direction of propagation, is a TEM (transverse
electromagnetic) wave
 Lossless, uniform plane wave solution to Maxwell’s
equations
• let’s look at the magnitudes of E and H

     j  t  b z  
E plane  Exo e 
j t  b z 
xˆ  Exo H plane    Exo  e yˆ  Exo
wave wave    
– notice that the ratio of E to H is

E plane
wave Exo 
 
  
H plane Exo
wave 

– units check:
• E: volt/meter : farad/meter
• H: amp/meter : henry/meter
• E/H: volt/amp = ohm /: henry/farad = (ohm)2
 Lossless, uniform plane wave solution to Maxwell’s
equations

E plane
• the ratio of E to H is wave 
 
 
– where  is called the H plane
wave
“wave impedance”
• if the  = o and  = o then we are in “free space,” and the
“impedance of free space” is

  o
E plane H plane   o  377 
wave wave o
 Uniform plane wave summary
 
• the direction of propagation is given by Eo  H o

• the propagation constant is b   

2p 2p
• the wavelength is l 
– inversely dependent on frequency
b  
– inversely dependent on square root of  and 

 1
• the phase velocity is vp  
– independent of frequency
b 
– inversely dependent on square root of  and 
• and the fields are
 j t  b z      j t  b z    
E plane  E xo e xˆ H plane    Exo  e yˆ E plane H plane  
  
wave wave   wave wave
 Summary of electromagnetics: time harmonic form
of Maxwell’s equations
• summarizing everything we have so far, assuming time
harmonic behavior, and using Ohm’s law for J
     
 
  Es  r    j H s  r  D  v
     
  H s  r     j  Es  r  B  0
– “dielectric”: displacement current dominates,  >> /
– “conductor”: conduction current dominates,  << /
• plus (time independent) material properties

     
 E
D   H
B  J E
 r o r o
 Uniform plane wave solution to Maxwell’s equations

• the complete, time harmonic solution is

 assumed     jt  z
E plane  Exo e jt  z xˆ H plane   j Exo  e yˆ
wave wave   
– E and H are perpendicular to each other

 2  j   j     j   j     j b

–  is called the “complex propagation constant”
1
2p  d  d b 
l vp  vg   
b b d b  d 

• direction of propagation
 
Eo  H o
 Power flow
• is there anything more general we can say about what it means
for a wave to be “propagating”?

• let’s start with   D
 H  J 
t
– now take dot product with E on both sides

    D 
E   H  J  
 t 
– left hand side
     

E   H   
 H    E   E  H 
vector
ID

– so now we have

       D
   
H    E   E  H  E J  E 

  t

B

t
 Power flow
• so far using Maxwell’s equations and a vector ID
 

  H        E  
H 
t

  E  H  E J  E   t
• or  
     E  H
 
 E  H  E J   E    H 
t t
 
 E  H
• let’s look closely at E  and H 
t t
              
– using the chain rule
t

 t 

E E   E E  E  E   2 E  E 
 t   t 
– or
    1    1  2
E  E  
 t  2  t
E E 
2
t
E   
• so now we have
    1  2 1 
 
 E  H  E J  
2 t
 E   
2 t
 H2
 Power flow
• we now have, using the general form of Maxwell’s equations
    1  2 1 

 E  H  E J  
2 t
E 
2 t
H2    
• let’s integrate over some volume of space
     1 2 1 2
   
 E  H dv   E Jdv   
t volume  2
 E 
2
 H  dv

volume volume

• the divergence theorem let’s us convert the volume integral of

div(ExH) into a surface integral of ExH

     
   1 2 1
 E  H dS        2
E Jdv E H dv
surface volume
t 2
volume  
2
enclosing  energy energy
V ohmic power stored stored
dissipation in E field in H field
 Power flow: the Poynting vector
     
   1 2 1
 E  H dS        2
E Jdv E H dv
surface volume
t volume 
2 2

enclosing  energy energy
V ohmic power stored stored
dissipation in E field in H field

• the right hand side represents the power flowing into the
volume
• so the left side must represent the same thing
  
 surface

enclosing
 
E  H dS  power flow into the volume
V

• then getting rid of the minus sign tells us that

  
surface
enclosing
 
E  H dS  power flow out of the volume
V

• so we interpret the Poynting vector P as the instantaneous

power density   
P  EH
– units: E·H: (V/m)·(Amp/m) = Watt/m2
 Poynting vector in phasor form
• when using phasors we need to remember to take the real part
to get a physically meaningful result (as opposed to a
mathematically convenient result)
• for our time harmonic form we also have to do the time average
over one period
• so the actual power flow would be   
Ps ,ave
1

 Re ES  H S 
2

– where H* is the complex conjugate of H
• the imaginary part of H is replaced by its negative
• notes:
– the complex conjugate of a product is the product of the complex
conjugates 
 a  jb    c  jd     a  jb    c  jd 
 

– the complex conjugate of exp(a+jb) is exp(a-jb)


e a  jb    e a  jb 
 