Special Relativity 3
Special Relativity 3
Special Relativity 3
Introduction
In this note, an elementary account of special relativity is given. The knowledge of basic
calculus is enough to understand the theory. In fact we do not use dierentiation until
we get to the part of composition of velocities. If you can accept the two hypotheses
by Einstein, you may skip directly to Section 5. The following three sections are to
explain that light is somehow dierent from sound, and thus compels us to make further
assumptions about the nature of matter, if we stick to the existence of static ubiquitous
ether as the medium of light.
There is nothing original here. I have collected from various sources the parts which
seem within the capacity of the average economics students. I heavily draws upon Harrison
([4]). And yet, something new in the way of exposition may be found in the subsections
5.4 and 6.1. Thus the reader can reach an important formula by Einstein that E = mc2
with all the necessary mathematical steps displayed explicitly.
Thanks are due to a great friend of mine, Makoto Ogawa, for his comments. He read
this note while he was travelling on the train between Tokyo and Niigata at an average
speed of 150km/h.
s = speed of sound;
x = position in x-coordinate;
0
= value in another frame;
l
2sl
l
+
= 2
.
sV
s+V
s V2
(1)
The first term in the middle of the above eq.(1) stands for the time required to reach
wall A, and the second for that time to came back to the ears. Then, for the sound wave
directed to wall B, we obtain this:
r
V tB 2
2 l2 + (
)
2
tB =
.
(2)
s
Pythagoras theorem is here used to get the length of the hypotenuse. From eq.(2), it
follows
(s tB )2 = 4 l2 + (V tB )2 .
Thus,
2l
.
tB = 2
s V2
tA
s
1
= 2
=r
>1
2
tB
s V
V2
1 2
s
(3)
( 0 < V < s ).
Now two sound waves are displaced from each other because their arrival times are
dierent. (Note that no Doppler eect is involved since the source of sound, the walls,
and the ears are on the same stage, keeping the same mutual distance among them.) Two
waves may be in phase or out of phase when we change slightly the distance between the
ears and the wall A. That is, there can be interference. When the eect of interference
is visualized through an optical apparatus after converting it into electric current, we can
observe bright and dark stripes or rings called fringes.
l1
2l1
l1
2cl1
+
= 2
=
.
2
cV
c+V
c V
c 2
tB =
2l2
2l2
.
=
c
c2 V 2
l1
l2
.
tA
=
tB
The time dierence is:
t tA tB =
2l1
2l2
r
.
V 2
V
2
c(1 ( ) ) c 1 ( )
c
c
(4)
(5)
(6)
Michelson rotated the whole interferometer by 90 degrees, and the time dierence
would be then (with l1 and l2 interchanged):
t0 =
2l2
2l1
r
.
V 2
V
2
c(1 ( ) ) c 1 ( )
c
c
The whole time dierence between the two cases before and after the rotation by 90
degrees becomes
t + t0 =
2
V
V
1
((l1 + l2 ) (1 ( )2 )1 ) (l1 + l2 ) (1 ( )2 ) 2 )
c
c
c
6
2
V
1 V
((l1 + l2 ) (1 + ( )2 + ) (l1 + l2 ) (1 + ( )2 + ))
c
c
2 c
2
V 2
1 V 2
+
((l1 + l2 ) (1 + ( ) ) (l1 + l2 ) (1 + ( ) ))
c
c
2 c
l1 + l2 V 2
=
( ) .
c
c
=
(The formula (1 + x)n = 1 + nx + is used, and the terms of orders not lower than two
are dropped. Why then addition, t + t0 , ? When the mirror for adjustment is slid
in order to annul the eect of interference, we simply add the adjustment shifts whether
they are forward or backward.)
Hence, the dierence in light path lengths between the two cases is
V
c (t + t0 ) = (l1 + l2) ( )2 ,
c
where the speed of light c = 3 105 km/s = 3 108 m/s, and the orbital speed of the
earth V = 30 km/s = 3 104 m/s, and l1 + l2 + 11 m. (See Fig.4 to note that the light
goes along the diagonal 15 times.) We can calculate
V
c (t + t0 ) = (l1 + l2 ) ( )2 + 22 108 = 2.2 107 .
c
On the other hand, the light used in the experiments was the yellow light from heated
sodium whose wave length = 5.9 107 m, and so we have
c (t + t0 )
2.2 107
+
+ 3.7 101 + 0.4.
7
5.9 10
Michelson wished to observe the displacements of interference fringes, sometimes and
somewhere larger than 0.4
The results up to 1887 was reported in Michelson and Morley[9]. The results were to
their disappointment, and the speed of ether wind was estimated less than one sixth of
the orbital speed of the earth. Note that the experiments were conducted day and night,
each of four seasons, and on the top of mountains so that they could avoid the o-setting
of the orbital velocity of the earth by some other movements such as the solar system,
which is at present estimated as large as 250 km/s. (The rotational speed of the earth
near the equator is 500 m/s.)
Lorentz Contraction
In 1892, H.A.Lorentz published a paper which explains the negative results of MichelsonMorley experiments. (See Janssen[5] for the information on the papers written by Lorentz,
and on how Lorentz improved on his ideas for a considerable period.) The idea was
very simple: the things shrink or contract in the direction of motion, and the size of
7
contraction depends on the speed of the movement, and is just enough to make the above
two time durations tA (eq.(4)) and tB (eq.(5)) equal. (G.F.FitzGerald arrived at the
same contraction theory in 1894.) That is, when a certain body is moving at a velocity
v, its length l is contracted in the direction of motion as
r
V
l1 = 1 ( )2 l2 = l2 < l2 .
(7)
c
It is obvious from eq.(6) that no ether wind could be observed.
Along with his theory of contraction, he had to prepare many formulas for conversion
or transformation concerning the time and the motion of matter on a moving frame.
A set of these formulas were named by Poincar Lorentz transformation, and most of
them re-emerge in Einsteins theory of relativity as a necessary consequence of his two
hypotheses.
Special Relativity
5.1
First, we define inertial reference frames in a less rigorous way. Suppose that there are
two observer, O and O 0 , who set up their respective three dimensional Euclidean spaces,
and the three axes x-, y-, and z-axis for O are all parallel to those for O0 , x 0 -, y 0 -, and
z 0 -axis. These two observers are stand at their respective origin, called also O and O 0 .
When the observer O0 is moving parallel to the x-axis at a constant velocity V , or so
observed by O, and the observer O is seen by observer O 0 to move parallel to the x 0 -axis
at a constant velocity V , then these two sets of coordinates form two inertial reference
frames. On these two frames, no external force is working to change their velocity. We
simply call them systems: system O and system O0 .
Einstein made two hypotheses in [1].
Hypothesis 1. There can be no experiment to judge which system is really moving.
Or there is no absolute space.
Hypothesis 2. Light has a constant speed in a vacuum on each system independent
of the motion of the emitting body.
Einstein then defined simultaneity, and derived formulas, most of which had been
established by Lorentz. For Einstein, however, things do not contract, but each observer
on a dierent system simply makes dierent observations.
5.2
Simultaneity or synchronism
At two points, A and B , on a system, two local times coincide with each other, if light
starts form point A at As local time tA , and reaches point B at Bs local time tB , then
reflected back to A, arriving at A at As local time t0A , and if the following equation
holds:
tB tA = t0A tB .
This is the definition of synchronism by Einstein [1, p.894]. Almost needless to say,
we assume that two local times once synchronized proceed uniformly for ever on the same
system. This definition enables us to determine a particular local time using light, thanks
to Hypothesis 2. That is, suppose that on system O, the time is synchronized at the
outset, time 0. Then we have
Rule 1. When light start at point A at time tA and travelled the distance x and reaches
point B , then the time of arrival at B is
x
tA + .
c
One more consequence is:
Rule 2. When on a system two light beams start at point A at the same time, and reach
point B again at the same time, the distances (light path lengths) which the two light
beams have travelled are the same.
5.3
Lorentz Transformation
Let us consider the Michelson and Morley experiment. Observer O is, say, on the sun,
while observer O0 is on the Michelson interferometer, which is moving along x 0 -axis at
velocity V . For observer O 0 , since two light beams start and return, through the identical
path, at the same time, the two distances, l10 and l20 , should be equal by Rule 2: l10 = l20 .
On the other hand, for observer O, light goes out and comes back through dierent paths,
and yet by Rule 2 the two light path lengths must be equal, ctA = ctB , hence from eqs.(4)
and (5)
l1 = l2 < l2 ,
i.e., Lorentz contraction, eq.(7). Naturally, we take for granted the equality l20 = l2 .,
because this direction is at right angles to that of motion. And so, we get l1 = l10 along
the direction of motion. When we set the two origins O = O0 at time t = 0, in general,
l10 = x 0 and l1 = x V t. Hence
x V t = x0 , or
x0 =
xV t
, y0 = y, z0 = z .
(8)
x0 + V t 0
, y = y 0 , z = z0 .
(9)
These two eqs. (8) and (9) are Lorentz transformations of coordinates between two
systems O and O 0 .
5.4
Time
t =
V
c2
(10)
t0 +
V
c2
x0
Eq.(10) can yield fascinating stories. At the origin O0 on system O0 , that is, the point
x = vt on system O, eq.(10) becomes
0
t =
V
c2
Vt
t (1
=
V2
)
c2
= t<t .
Time flows more slowly at the origin on system O 0 than on system O. This is called
time dilation, and has been confirmed in experiments, especially by the greater longevity
of muons, showering down on the surface of the earth. (See Harrison [4, time dilation].)
It should be noted that time goes more quickly on system O 0 than on system O at the
origin O, i.e., at x = 0. To remain at the point O on system O 0 , an observer on system
O 0 has to run continuously to the left at velocity V . )
Eq.(10) is somewhat disturbing because time on system O 0 depends on not only time
in system O but location therein as well. This can be made more understandable if we
derive eq.(10) by using Rule 1 above. Look at Fig.6. At time 0, the origins O and O0 were
at the same place, and two light beams were emitted. One beam went along O A x,
reaching x at time t , and the other O0 A0 x , reaching x at time t0 . For the light
path O A x, we have
r
x2
= ct .
(11)
2 l2 +
4
10
(12)
Note that to derive eq.(12) we have used eq.(8) because, on system O, a certain length of
system O0 is observed as contracted by factor , thus we need to correct this. Eliminating
l from these two eqs.(11) and (12) gives us again eq.(10)
t0 =
5.5
V
c2
Composition of Velocities
Now in this section, we use dierentiation of functions, following Landau and Lifshitz [7].
First, we derive the formulas for converting velocities. From eqs. (8) and (10), it follows
dt cV2 dx
dx V dt
0
0
0
dx =
, dy = dy , dz = dz , dt =
.
Define wx
dx
,
dt
wy
dy
,
dt
and wz
wx0
dz
,
dt
wx V
dx0
dx V dt
=
=
,
V
0
dt
dt c2 dx
1 cV2 wx
11
dy0
dy
wy
=
,
=
V
0
dt
dt c2 dx
1 cV2 wx
dz0
dz
wz
=
=
.
V
0
dt
dt c2 dx
1 cV2 wx
wy0
wz0
Dually,
wx0 + V
dx
=
,
dt
1 + cV2 wx0
wy0
dy
=
,
dt
1 + cV2 wx0
dz
wz0
.
=
dt
1 + cV2 wx0
wx
wy
wz
(13)
As a special case, we have from eq.(13) the formula for the composition of velocities.
When wy0 = wz0 = 0, and wx0 = w 0 , eq.(13) is written as
w=
w0 + V
.
1 + cV2 w0
(14)
This tells us that when an object is moving on system O 0 at velocity w0 while system O 0
itself moving at velocity V relative to system O, the object is moving, for the observer
O, at the velocity described by eq.(14), which is not greater than c when w0 and V
are themselves not greater than c. For example, when w0 = V = 0.8c, w = 1.6c/1.64 +
0. 9756c. Thus,
Rule 3. When two velocities are not greater than c, the composition of these two is
not greater than c, either. The composition of two velocities is equal to c only when the
two velocities are themselves c.
6
6.1
in Fig.7. The centre of mass in system O 0 does not move, and so it is moving at the
velocity V in system O. Then, by putting
v1
u+V
u + V
and v2
,
uV
1 + c2
1 uV
c2
the equation
m v1 + m v2 = 2m V .
should hold. This, however, fails to be valid: the LHS is smaller than the RHS when
0 |V | < c and 0 |u| < c, because
2
2V (1 uc2 )
2V (1 uc2 )
u+V
u + V
+
=
=
< 2V ,
2
2
1 + uV
1 uV
(1 + uV
) (1 uV
)
(1 uc2 Vc2 )
c2
c2
c2
c2
supposing now V > 0. This could be a fatal flaw: the two momenta, one calculated with
two balls separately (LHS) and the other calculated using the centre of mass of two balls
(RHS), do not coincide on system O. This is, however, because we have assumed the
mass is independent of its motion.
Now we suppose that the mass depends on its velocity in a system, and denote by
m(v) the mass as a function of velocity. we should have
13
(15)
2V
2 , and
1 + Vc2
(16)
(17)
m0 m(0).
Eq.(17) leads to a quadratic equation of V
wV 2 2c2 V + c2 w = 0,
which yields
V =
c2
c2 c c2 w2
c4 c2w 2
=
w
w
(18)
m0 V
m0
= w
.
wV
1
V
(19)
Using eq.(18) ,
r
w2 (c2 + c c2 w2 )
w2
w
w2
c2 w2 + w2 c c2 w2
=
=
1
+
1
.
= 2
=
V
c4 c2 (c2 w 2)
c2 w2
c2
c c c2 w2
Hence, eq.(19) becomes
m0
m(w) = q
w2
c2
m0
m(v) = q
v2
c2
(20)
The quantity m0 is called rest mass. Based on this new formula(20), the reader can
verify the general case, i.e., eq.(15), noting the following identities:
14
u+V 2
)
1 + uV
(c2 u2 )(c2 V 2 )
2
c
1
=
,
c2
(c2 + uV )2
(
u + V 2
)
1 uV
(c2 u2)(c2 V 2 )
c2
1
=
.
c2
(c2 uV )2
(
An experiment was reported in Nacken[10] ,which more or less confirmed the formula
(20). What eq.(20) tells us is:
Rule 4. If an object is moving at a velocity less than c, we cannot accelerate it up
to c in a continuous way. (See, however, Feinberg [3] for tachyons, which are supposed
to fly faster than light.)
6.2
dp
d(mv)
=
,
dt
dt
where p is the momentum, v the velocity of a body, m the mass, and t time. Be careful
not to pull out m to the front of d, because now the mass is not independent of velocity,
hence of time. The change in energy with respect to time is:
dE
f dx
d(mv)
dv dm 2
=
v = mv
+
v .
dt
dt
dt
dt
dt
On the other hand, dierentiating eq.(20) with respect to v leads to
dm
m0
v dv
1
v dv
1
dv
= q
=m
=
m
.
3 2
2
v
2
2 v2
dt
c
dt
c
dt
c
dt
1
2
2
c
1 vc2
From this, it follows
d(mc2)
dm 2
dv dm 2
dE
=
c = mv
+
v =
.
dt
dt
dt
dt
dt
It seems natural to regard E = 0 when m = 0. Therefore, we may write
mc2 = E or E = mc2.
15
This relation has been used, unfortunately much for military purposes, in nuclear
fission, where the weight of an atom is greater than the sum of the weights of the parts
after dividing that atom. And within the sun, lost mass is believed to be converted to
energy as nuclear fusion, where the weight of an atom(of helium) created is less than the
sum of the weights of the parts(of 4 hydrogen atoms) which have formed that atom.
Final Remarks
Some of the students may have wondered whether the denominator in eq.(4), c+V , admits
a speed greater than that of light, contradicting Rule 3 in Section 5. The magnitude, c+V ,
l1
is not a speed of anything, but the magnitude, c+V
, is the time required for light to travel
between two mirrors, and so measured by the observer.
In retrospect, the special theory starts by asking how we can be sure about the same
common time duration, say a second or a hour, between two places in this universe.
Einsteins answer is that light has the absolutely same common velocity in a vacuum
everywhere in the universe. Then, given a common distance, somehow, we can designate
a common time duration, when a suitable apparatus is available to observe the departure
and the arrival of light in experiments. This common time duration system can, however,
force two people on dierent frames to observe dierent times for the same phenomenon.
References
[1] Einstein, Albert:
Zur Elektrodynamik der bewegter Krper,
Annalen der Physik, 17, 891-921 (1905). (Available from: http://www.physik.uniaugsburg.de/annalen/history/ )
[2] Einstein, Albert:
Ist die Trgheit eines Krpers von seinem Energieinhalt abhngig?,
Annalen der Physik, 18, 639-641 (1905). (Available from:
http://www.physik.uni-augsburg.de/annalen/history/ )
[3] Feinberg, Gerald: Possibility of faster-than-light particles, Physical Review, 159,
1089-1105 (1967).
[4] Harrison, David M.: The Special Theory of Relativity, 1999.
( http://www.upscale.utoronto.ca/GeneralInterest/Relativity.html )
[5] Janssen, Michael H.P.: A Comparison between Lorentzs Ether Theory and Special Relativity in the Light of the Experiments of Trouton and Noble, Ph.D.
Thesis, University of Pittsburgh, 1995. (Available from: http://www.mpiwgberlin.mpg.de/litserv/diss/janssen_diss/ )
[6] Kakiuchi, Yoshinobu: Basic Physics I, (in Japanese), Tokyo: Gakujutsu-Tosho Shuppan, 1967. NB. This book contains many typos.
16
[7] Landau, Lev D. and Lifshitz, Evgeny M.: The Classical Theory of Field, 2nd ed.,
London: Pergamon Press, 1962. (The original Russian edition was published in 1941.)
[8] Maehara, Shouji: Why can we not run faster than light?(in Japanese), Mathematical Science (Suuri-Kagaku), 72-77 (1968).
[9] Michelson, Albert.A. and Morley, Edward.W.: On the relative motion of the earth
and the luminiferous ether, American Journal of Science, 34, 333-345 (1887). (Available from http://www.aip.org/history/gap/PDF/michelson.pdf)
[10] Nacken, M.: Measurements and mass changeability of electrons in quick cathode
rays, (in German), Annalen der Physik, 23, 313-329 (1935).
[11] Panofsky, Wolfgang K.H. and Phillips, Melba: Classical Electricity and Magnetism,
Reading: Addison-Wesley, 1962. Chapters 15 and 16.
[12] Planck, Max.: Treatise on Thermodynamics, 3rd English ed., New York: Dover, 1926.
[13] Prigogine, Ilya. and Defay, Raymond: Chemical Thermodynamics, London: Longman, 1954.
[14] Wilson, Allan H.: Thermodynamics and Statistical Mechanics, Cambridge: Cambridge University Press, 1957.
[15] Zemansky, Mark W.: Heat and Thermodynamics, 4th ed., New York: McGraw-Hill,
1957. 7th with Richard H. Dittman, 1997.
17