Abstract. We discuss the concept of wave energy packets and demonstrate that this
physical quantity can be also used in physical problems involving the relativistic effects.
We show that simple definition of energy density could lead to relativistic paradoxes and
contradict the time dilation and length contraction when applied to the free
electromagnetic waves.
1. Introduction
2. The concept of wave energy packet in the classic approach
The plane electromagnetic wave travels in vacuum as fleeting orthogonal
electric and magnetic fields. The field intensities oscillate spatially along the
direction of propagation and temporary normal to the direction of
propagation as depicted by figure 1.
measure the local amplitude of the electric and magnetic field. When
comparing their measurements made at any arbitrary time t these observers
identify a longitudinally oscillating electric and magnetic field having the
same period = cT and the same phase. In the non-dissipative vacuum these
fleeting oscillations are mathematically described by the familiar wave
equations:
x
E ( x, t ) E max sin t
c
B ( x, t ) Bmax sin t
c
(1)
In vacuum
1
1
0 E 2 ( x, t ) 0 H 2 ( x, t ) 0 E 2 ( x, t ) c 2 B 2 ( x, t )
2
2
1
E ( x, t )
with H ( x, t ) 0 and c we get E ( x, t ) cB( x, t )
0 0
0
( x, t )
( x, t ) 0 E 2 ( x, t ) 0 H 2 ( x, t ) 0 c 2 B 2 ( x, t )
(2)
and:
(3)
Replacing the expressions for wave fields (1) we get the explicit form as:
x
x
2
( x, t ) 0 Emax
sin 2 t max sin 2 t
c
c
2
max 0 E max . This equation shows that the
(4)
with
energy carried by the plane
wave oscillates temporary at any arbitrary location and longitudinally at any
arbitrary time instant. The energy density changes between zero and a peak
value also called wave crest as further symbolically depicted by figure 2.
This figure represents intuitively the spatial-temporal oscillations of the
wave energy density as a travelling tread of spindles.
The average energy density in any arbitrary point x along the propagation
axis is given by a temporal integration of (4) as:
med ( x)
2
max 0 E max
2
x
sin
dt
max
T T/ 4
c
2
2
T /4
(5)
/4
2
max 0 E max
2
x
sin
dt
max
/ 4
c
2
2
(6)
(7)
In a plane wave this vector is always pointing in the direction of wave
propagation and its amplitude is given by the expression:
S ( x, t ) c ( x, t ) c 0 E 2 ( x, t )
(8)
Figure 2 shows the Poynting vector S changing periodically in amplitude
from peak values at wave crests to zero values in between.
This behaviour supports us to assume that the wave energy is travelling as
discrete energy packets that we call wave energy packets - WEP. The
concept of wave energy packet is used here to delimit temporary or spatially
the periodic variations in energy density and should not be yet confronted
with the wave packet concept as known from quantum mechanics.
We are now concerned with calculating the average energy that the plane
wave transports by means of a single wave energy packet throughout a
stationary surface of area A orthogonal to the direction of wave propagation
at some arbitrary position x. This may be for example the average energy
stored into a photo-detector having a transversal sensitive window of area A
subsequent to the absorption of all light photons within a single wave
packet. We calculate the average energy by a time integrating of (4) for the
duration of the wave packet. In a non-dissipative propagation medium
(perfect vacuum), we have no attenuation during propagation and therefore
the average energy in the wave packet is the same in any point along the
propagation direction.
W packet
T /2
T /2
( x, t ) Acdt Ac 0
2
E ( x, t )dt AcT 0
2
E max
E2
A 0 max
4
4
(9)
For consistency we obtain the same result by a spatial integration along the
packet length at any arbitrary time instant t:
W packet
/2
/2
( x, t ) Adx A 0
2
E ( x, t )dx A 0
2
E max
4
E y
B x B x
E z
E y V Bz
1 2
Ez V By
1 2
By
B z
V
Ez
c2
1 2
By
(10)
V
Ey
c2
1 2
Bz
where = V/c. For the plane and linear polarised wave we have
Ex=Bx=Ez=By=0 and therefore E'x=B'x=E'z=B'y=0 thus when seen from K' the
wave remains also linear polarised with the wave field intensities given by:
E ( x , t )
E ( x, t ) V B ( x, t )
B ( x , t )
1 2
V
E ( x, t )
c2
1 2
B ( x, t )
(11)
Using the expressions (1) for the stationary field components under the
vacuum condition E ( x, t ) cB( x, t ) and the inverse Lorenz transformation
for the space and time coordinates
1
( x Vt )
1
1
(t
V
x )
c2
(12)
x
sin t
B ( x , t ) Bmax
c
(13)
with
1
E max
1
E max
Bmax
1
Bmax
1
1
1
(14)
Therefore for the moving observers the wave remains plain polarised but the
wave frequency and the amplitudes of the electric and magnetic oscillations
decrease by the Doppler factor. The wavelength however increases by the
Doppler factor.
Y
Y'
S c
Wave A
volum
e
V
c
PD
A
PD'
'
K
O
Wave
volum
e
K'
O'
X, X'
Figure 3. The wave energy packets as measured by two similar photo detectors each stationary in its own reference frame
With these equations we calculate the energy of a single wave energy packet
as measured by the two similar photo detectors each stationary in its
reference frame using (6):
5
W packet
2 T 1 1
T Emax
2
W packet T Emax
T 1 1
(15)
The detector moving in respect to the light source is further subject to the
Doppler effect. Including the equations for the relativistic Doppler shift
T
1
1
W packet
W packet
we have finally:
1
1
(16)
W packet
h
and N
W packet
(14)
N W packet
T
1
N W packet T
(15)
cT
In equations (5-7) we make the notations 2 A 2
and m 0
2
E max
2
W packet
W packet
m
m
(16)
2 1
m Emax
2
m Emax
1
(17)
2 W packet 1
m W packet Emax
2
m W packet
1
Emax W packet
(18)
Y'
S2
S1
/2
K
-L
K'
O'
X, X'
Ei ( x, t ) E max sin t
E max sin t
c
c
xL
xL
Bi ( x, t ) Bmax sin t
Bmax sin t
c
c
(20)
2x
L
sin t
(21)
Bi ( x, t ) 2 Bmax sin
2x
L
cos t
2 Bmax sin
2x
2 E max cos
2x
and
2xe
2xe
1 sin
0 where the interference wave has only an electric
component with the magnetic component being permanent zero and other
positions xb with cos
2xb
2xb
0 sin
1 where the interference wave has
0 2
Ei ( x, t ) c 2 Bi2 ( x, t )
2
(22)
10
For each of the interfering waves we have E max cBmax , with the peak energy
2
density in each wave being max 0 E max
and the average spatial-temporal
2
0 E max
. With this (22) becomes:
2
2x
L
2x
L
i ( x, t ) 2 max cos 2
sin 2 t sin 2
cos 2 t
1
2
(23)
Figure 5. The pulsing of the energy density within the interference volume
2xn
2xn 1
sin 2
where the energy density is constant all the time and
sin 2 t 0 cos 2 t 0
c
c
2
period T/2 when the energy density is constant in the whole interference
volume and equal to i ( x, t 0 ) max . Finally we identify a set of particular
positions xa called anti-nodes that is actuality the reunion of subsets xe and xb
where the energy density is pulsing in between 0 and an absolute maximum
equal to i ( xa , t ) 2 max . For all other axial positions in-between the nodes
and anti-nodes the energy density is pulsing periodically in between two
extreme values symmetric around the average value max .
We conclude our analysis by underling that the nodes and anti-nodes are
simply points in medium which build up the interference pattern. Contrary to
wave crests that travel at light speed along the propagation direction, the
positions of nodes and anti-nodes remain the same in the stationary reference
frame of the light sources. The distance between two adjacent nodes is /4
equal to the distance between two adjacent anti-nodes.
11
waves travel in opposite directions. After a half wave period the photon
density in both interfering waves reaches zero at this anti-node and therefore
the wave density at this position also reaches zero. In quantum mechanics
approach the probability to detect a photon at this moment and at this point
is infinite small.
At the particular interference node located at x=/8 the photon density never
reaches zero. Here we find always some photons belonging to either of the
interfering waves. As resulted by calculation actually the total number of
photons at the node position is invariant with time. However the relative
composition of these photons changes continuously from 100% photons
travelling leftward, throughout 50%-50%, to 100% photons travelling
rightward. Therefore the net energy transport at any node position reverses
direction periodically forth and back depending on which interfering wave is
dominant at once i.e. it just happens to embrace many photons than the
other.
- Sa calculez energia totala intr-un pachet (fus) de interferenta si numarul de
fotoni din el. Care este energia fotonului in unda de interferenta?
E
1
and
1
1
V
E)
c2
(25)
13
With (25) we calculate the energy density measured by observers in K' as:
1 2 2 2 1 2 1
(x t ), 0 E (x t ), c B (x t ), 0E tx ),( 2
2
2 1
2
(26)
Therefore the local energy density for the interference volume changes as:
( x , t ) 1 2
( x, t ) 1 2
(27)
We obtain the same result considering that the energy density of interference
volume in frame K' is build up by the superposition of two plane waves
generated by two travelling sources moving in opposite directions at same
speed and creating the energy densities given by:
1 2 1
1 2 1
and
1 2 1
1 2 1
(28)
For the stationary frame we have + = - = 1/2 where is the energy density
in the interference volume. Appling further the principle of energy
conservation in frame K' we calculate the energy density of the interference
pattern as seen from K' with:
1 1
1 2
2 1 1
1 2
(29)
1 2
1
1
14
( x, t ) 1 2
3. The increase of energy density by ( x, t ) 1 2 . This result is the
L
1 2
L
2. The energy packet length dilation in vacuum (outside the laser cavity)
as demonstrated above:
1
1
????
Our special concern is now the energy inside the cavity (which according to
our previous definition is a particular form of bounded energy) and how it is
seen by observers of the moving frame. This experiment could be also
realized using a resonant microwave cavity that is excited at the left side by
a microwave oscillator whose frequency is fine tuned such that a stationary
electromagnetic wave occurs along the cavity length.
15
Y'
Resonant
cavity
Resonant
cavity
L'
o'
K'
X O'
X'
Figure 4. Bounded and unbounded energy packets for the Fabry-Perot laser
experiment as seen from two inertial reference frames.
1 2
m
1
m
1
????
(19)
Thus the bounded and unbounded energy transform differently and this
could be explained by the Heisenbergs uncertainty principle stating that we
cant measure simultaneously the energy and the bounding volume of the
electromagnetic wave.
8. The wave packet concept applied to the Margaritondos experiment1
In this experiment a photo-detector stationary in a reference frame K counts
the number of incident photons. The same experiment is contemplated by
the observers of a second inertial reference frame K' that is moving at
constant speed V relative to the first frame.
16
Y'
Y
K at t = t0
b
K' at t = t0
PD
c
PD
c
'
O'
X'
ct
Y'
ut
A
A
K at t = t0+t
PD
K' at t = t0+t
PD
O'
X'
Figure 5. Margaritondos experiment used for deriving the transformation
equation for the energy of a photon. Figure 4a depicts the experiment as detected from
the rest frame of the photomultiplier whereas Figure 4b illustrates the experiment as
detected from a reference frame that moves uniformly in respect to the first one.
We accept that the number of photon counts is the same when viewed by the
observers of the two inertial frames and we apply the concept of wave
packet that we presented before. We know that the average number of
photons per wave packet is invariant. Therefore if the photo-detector is
counting let say two complete wave packets in frame K then it shall count
also two complete wave packets when seen from frame K' to get the same
number of counts. We calculate the time window required to perform the
counting process. In frame K we have
t
were Vrel is the relative speed between the photo-detector and the
Vrel
wave packets as seen by observes in frame K'. This is the apparent speed
between the wave packets (the wave front of the light signal) and a moving
object in frame K' and not the speed of light c as properly seen by the photodetector from its own reference frame. Using the Lorentz transformation for
the measurement time intervals
t
1
t
1 2
shift
as:
Vrel c
c 1 c V
t
(20)
18