a

r
X
i
v
:
p
h
y
s
i
c
s
/
0
5
0
6
2
0
9
v
1


[
p
h
y
s
i
c
s
.
c
l
a
s
s
-
p
h
]


2
9

J
u
n

2
0
0
5
Singularity dynamics:
Action and Reaction
M. Mazilu
School of Physics and Astronomy,
University of St. Andrews
St. Andrews, KY16 9SS, UK
Abstract
The interaction between singular and regular elds is considered for Lorentz-
invariant scalar and vector wave equations. The singular eld is generated by a Dirac
source term. Its dynamics are deduced from the total eld Lagrangian. At non-
relativistic speeds, the resulting equations of motion are those of a mass in a scalar
potential. Using this method we deduce the relationship between source amplitude
(scalar gravitational mass) and dynamic mass (inertial mass). Generalising this
method implies Lorentz forces for charge singularities in the electromagnetic eld
and describes the dynamics and interaction of hypothetical magnetic monopoles.
1 Introduction
The method presented builds on the action principle which oers a general tool to
describe both particle and eld dynamics. This principle represents one of the great
unication in physics and applies equally well to Newtonian mechanics [1], general
relativity and electromagnetic elds[2]. Using this single principle it is possible
to deduce the Euler-Lagrange equations describing the evolution of one system or
of many interacting systems. However describing these interactions requires the
assumption of an additional action term. In this paper, we show that it is possible
to eliminate this additional assumption and deduce the reaction dynamics of the
singularity. In the case of scalar eld singularities we derive Newtons equation of
motion together with the equivalence between Newtonian gravitational mass and
inertial mass. For charges and magnetic monopoles we deduce their associated
Lorentz force.
Singularities, vortices and topological defects have received considerable atten-
tion over the years; they determine the properties of many interesting materials such
as vortices in superconductors [3], superuids [4] and two-dimensional condensate
[5], dislocations and defects in solid [6] and liquid crystals [7], optical phase singu-
larities and optical vortices [8] and even topological defect in cosmology [9]. In this
paper, we use singularities together with the eld action principle to deduce their
dynamical behaviour. The method presented here is general and can be applied to
all of the elds mentioned above.
1
For clarity reasons we treat rst the simplest case possible that being a Dirac
source singularity in the Lorentz-invariant scalar wave equation. This leads to equa-
tions of motion similar to Newtonian mechanics (section 2). The method is gener-
alised for other singularities and other elds such as a charges (section 3) where the
electromagnetic eld implies the Lorentz force. In the last section, we apply this
principle to deduce the equations of motion of hypothetical magnetic monopoles
where the dynamics concurs with the one obtained through charge-monopole sym-
metry considerations [10].
2 Scalar eld singularity: Inertia
Before considering the anatomy of the eld singularity, let us start by deducing the
Euler-Lagrange equation for a scalar mass-less eld U described by the following
action
S
U
=

L(U) d
3
x dt =
1
2

1
c
2
(
t
U)
2
(U)
2
d
3
x dt (1)
where L(U) corresponds to the Lagrangian density, = (
x
,
y
,
z
) to the nabla
operator and c to the speed of light. The action principle implies that the scalar
eld U is solution of the standard homogenous wave equation

2
U
1
c
2

2
t
U = 0 (2)
where
2
stands for the Laplacian operator. The next step in the singularity dynam-
ics theory is to dene a singular source term that maintains the Lorentz-invariance
of the supporting eld equation. Starting from the wave equation (2), we can dene
the stationary singular eld U
s
as the time independent solution in the presence of
a Dirac source at the origin of the coordinate system

2
U
s

1
c
2

2
t
U
s
= m
s
(r) (3)
where m
s
corresponds for the source amplitude. To dene the Lorentz-invariant
singularity, we consider equation (3) in a reference frame having the constant velocity
v with respect to the stationary singularity. In this reference frame the singularity
has the velocity v. The transformation between the two reference frames is [11]:
r

= (r

+vt)
r

= r

(4)
t

t +
v r
c
2

.
where = (1 v
2
/c
2
)
1/2
and the prime stands for the coordinates in the moving
reference frame. The subscripts and represent the vectorial components parallel
and perpendicular to the velocity v. Applying these transformations to equation (3)
leads us to a constant velocity moving singularity. Beforehand, we remark that the
singularity can be decomposed into two components, (r) = (r

)(r

). Further, the
wave equation and the scalar eld U
s
is invariant with respect to the transformation
that is it remains unchanged in the transition from one reference frame to another.
2
It is only the singularity which changes in the moving reference system

2
U
s

1
c
2

2
t
U
s
= m
s
((r

vt

))(r

)
= m
s

1
v
2
c
2
(r

vt

))(r

)
= m
s

1
v
2
c
2
(r

s
)
where r

s
= vt

is the position of the singularity in the moving reference system. We

have used the scaling property of the single variable Dirac distribution |a|(ax) =
(x).
We can now write the equation of a moving singularity on an arbitrary path r
s
(t)
with the velocity v = r
s
. The singular eld U
s
is solution of

2
U
s

1
c
2

2
t
U
s
= m
s

1 v
2
/c
2
(r r
s
). (5)
and corresponds to a retarded potential.
In the following, we are considering the action principle in the case of the linear
superposition of the singular eld as dened in equation (5) and a given external eld
U
0
solution of (2). We remark here that the only free parameter of the system is the
position of the Dirac singularity and its velocity. The Euler-Lagrange equation of
motion is deduced when the action is stationary with respect path of the singularity.
The total action integral may be expressed as
S
U0+Us
(r
s
(t)) =

d
3
x dt L(U
0
+ U
s
)
=
1
2

d
3
x dt
1
c
2
(
t
U
0
)
2
(U
0
)
2
+
1
c
2
(
t
U
s
)
2
(U
s
)
2
+

d
3
x dt
1
c
2
(
t
U
0
)(
t
U
s
) (U
0
)(U
s
).
This action integral can be decomposed into two parts. The rst integral has no
cross-terms between the external eld and the singular eld. This integral is inde-
pendent with respect to the position of the singularity and its variation is zero. This
leaves only the second integral S
s
for the action principle
S
s
=

d
3
x dt U
0

2
U
s

1
c
2
U
0

2
t
U
s
where we have integrated the action by parts and assumed the singular eld to be
zero on the boundary of the integration region. Next, we use equation (5) to replace
the singular eld by the Dirac singularity and nally we integrate in Lagrangian
density in the spacial domain.
S
s
=

d
3
x dt U
0
m
s

1 v
2
/
2
(r r
s
)
=

dt U
0
m
s

1 v
2
/c
2
3
The equations of movement can be obtained by using the conventional particle Euler-
Lagrange dierential equation where the position and velocity correspond to those
of the singularity. Indeed we have
d
dt
m
i
v = m
g
U
0
(6)
where we have dened an eective inertial m
i
and gravitational m
g
mass
m
i
=
U
0
m
s
c
2

1 v
2
/c
2
m
g
= m
s

1 v
2
/c
2
.
(7)
One can also dene a dierential relationship between the two masses by taking
the scalar product between the velocity and equation (6). This leads to a energy
balance equation that generalises the famous equations E = mc
2
to the dynamic
domain.
d
dt
m
i
c
2
= m
g

t
U
0
(8)
Indeed, the energy balance equation shows that the variation of the inertial mass is
proportional to its potential energy variation.
Finally, we can deduce Newtons equation of motion for non-relativistic velocities
and far from any other singularities
d
dt
m
s
v = 0 (9)
where the scalar eld gauge is chosen to be U
0
() = c
2
.
3 Charge singularity: Lorentz force
In order to deduce the laws of motion of a charge we have to consider a singularity of
the electromagnetic eld described by Maxwells equations. We proceed in a similar
way to the scalar singularity case. In a rst instance, we deduce the homogenous
Maxwells equations using the action principle. In a second step, we introduce a
singular source to these equations making sure that their symmetries are maintained
(i.e. Lorentz-invariance). Then, we write down the Lagrangian integral of the total
eld composed of an external eld and the singular eld. Finally, the action principle
is applied to the total Lagrangian integral and the dynamics of the singularity is
deduced.
The electromagnetic eld action is expressed as:
S
em
=

0
2
E E
1
2
0
B B

d
3
x dt (10)
where the electric and magnetic elds are dened by
E = V
t
A
B = A
0 = A+
0

t
V
(11)
4
and where V and A are the electric scalar and magnetic vector potential respectively.
The vacuum permittivity and permeability is given by
0
and
0
. The third equation
corresponds to Lorentz gauge. Applying the action principle with respect to the
potentials lets us deduce Maxwells equations in absence of charges:
E
t
B = 0 (12a)
1

0
B
0

t
E = 0 (12b)

0
E = 0 (12c)
B = 0. (12d)
Like in the scalar case, we consider a single stationary Dirac singularity q(r) in
equation of the divergence of the electrical eld (12c). The eect on the electromag-
netic eld of this source is equivalent to a stationary point charge q at the origin.
The singular charge elds E
q
and B
q
are solutions of the in-homogenous Maxwells
equations:
E
q

t
B
q
= 0
1

0
B
q

t
E
q
= 0

0
E
q
= q(r)
B
q
= 0
(13)
The main dierence with the scalar case is in that the potential elds V
q
and
A
q
change in the transition from one reference frame to another. Indeed, using the
moving frame dened by (2) we have for the potentials:
A

A
q

+
vV
q
c
2

= A
q

V

q
= (V
q
+v A
q
)
(14)
where the speed of light is given by 1 = c
2

0
. Combining the transformations (2)
and (14) together with the denition of the electric and magnetic elds (11) we can
deduce the transformation relationship between these elds
E

= E
q

= (E
q

v B
q
)
B

= B
q

B
q

+
v E
q
c
2

.
(15)
This implies the following transformation for equations (13):

t
B

q
= 0
1

t
E

q
= qv((r

vt

))(r

0
E

q
= q((r

vt

))(r

q
= 0
5
Applying the scaling property of the single variable Dirac distribution |a|(ax) =
(x) and introducing the denition of the position of the singularity r

q
= vt

in the
moving reference system we get:

q
+
t
B

q
= 0
1

q

0

t
E

q
= qv(r

q
)

0
E

q
= q(r

q
)

q
= 0
where the relativistic coecient in the source terms cancelled out. The reason for
this is the covariant behaviour of the electromagnetic potentials as opposed to the
invariant scalar potential in equation (5).
This procedure gives us the form of the singularity following an arbitrary path
r
q
(t) with the velocity v = r
q
. The resulting vector singularity is equivalent to a
charge and its associated current. The singular charge elds E
q
and B
q
are solutions
of the in-homogenous Maxwells equations:
E
q

t
B
q
= 0
1

0
B
q

0

t
E
q
= qv(r r
q
)

0
E
q
= q(r r
q
)
B
q
= 0.
(16)
To deduce the equations of motion of this moving charge singularity we dene
the total eld as the linear superposition of the singular elds E
q
and B
q
and the
external elds E
0
and B
0
:
E = E
q
+E
0
B = B
q
+B
0
where the external elds are decomposed onto the external scalar and vector poten-
tials
E
0
= V
0

t
A
0
B
0
= A
0
0 = A
0
+
0

t
V
0
.
Introducing the total elds E and B into the electromagnetic Lagrangian (10)
infers the following singularity action:
S
sq
=

0
E
0
E
q

0
B
0
B
q

d
3
x dt (17)
where we have kept only the terms that vary with respect to the trajectory of the
singularity. Indeed, the terms in E
i
E
i
and B
i
B
i
, where (i = q, 0), are not
dependent on the position of the charge singularity. Rewriting the external eld
with the help of the scalar V
0
and vector potential A
0
allows us to integrate by
6
parts:
S
sq
=

0
(V
0

t
A
0
) E
q

0
(A
0
) B
q

d
3
x dt
=

0
(V
0
E
q
+A
0

t
E
q
)
1

0
(B
q
) A
0

d
3
x dt
where we have assumed the elds E
q
and B
q
to be zero on the boundary of the
integration domain. Using equations (16) we can further simplify the action of the
singularity
S
sq
=

(qV
0
(r r
q
) qA
0
v(r r
q
)) d
3
x dt
=

(qV
0
qA
0
v) dt
which leads directly to Lorentz forces for a charge in the external eld dened by E
0
and B
0
. To convince ourselves of this we can rigidly link a scalar mass singularity
introduced in previous section to a charge singularity. The total action of this double
singularity is:
S
s
+ S
sq
=

U
0
m
s

1
v
2
c
2
+ qV
0
qA
0
v

dt (18)
which lead to the following equation of movement
d
dt
m
i
v = m
g
U
0
+ qE
0
+ qv B
0
(19)
where m
i
and m
g
are the dened by (7). The energy balance equation is
d
dt
m
i
c
2
= m
g

t
U
0
+ qv E
0
. (20)
4 Magnetic Monopole singularity
Magnetic monopoles are hypothetical particles that behave in a similarly to charged
particles with magnetic and electric elds inverted ([10]). The monopoles correspond
to sources of the magnetic eld in equation (12d). In the following we are deducing
the equations of motion of magnetic source singularities by considering a Dirac
source q
m
(r) in equations (12d) where q
m
corresponds to the monopole strength.
Before proceeding further, we have to generalise the potential decomposition of
the electromagnetic eld. Indeed, the potentials that we used for the charged singu-
larities does not allow the presence of sources of the magnetic eld. Conventionally
the magnetic eld is by denition a rotational eld. Therefore, we dene a symmet-
ric electromagnetic potential decomposition that includes an additional magnetic
scalar and electric vector potential.
E = V
q

t
A
q
+
1

0
A
m
B = A
q
+
0
V
m
+
0

t
A
m
0 = A
q
+
0

t
V
q
0 = A
m
+
0

t
V
m
(21)
7
where the subscripts (q, m) stand respectively for the charge and monopole poten-
tials or elds. Using all the newly dened potential denitions we can apply again
the action principle the electromagnetic action dened by equation (10). The re-
sulting evolution equations are identical to equations (12). The only dierence is
the possibility to dene a monopole source density in equation (22d).
E
t
B = 0 (22a)
1

0
B
0

t
E = 0 (22b)

0
E = 0 (22c)
B = 0. (22d)
For the monopole singularity, we consider a single stationary Dirac singularity
at the origin. The singularity constitutes the free term of equation (22d) and acts
as a source of magnetic eld. The singular elds E
m
and B
m
are solution of:
E
m

t
B
m
= 0
1

0
B
m

t
E
m
= 0

0
E
m
= 0
B
m
= q
m
(r)
(23)
In this case there are four potential elds V
sq
, V
sm
, A
sq
and A
sm
and the
electric and magnetic eld is dened by equations (21). Taking the same frame
transformation dened by (2) we have for the potentials:
A

sq

= (A
sq

+
vV
sq
c
2
)
A

sq

= A
sq

V

sq
= (V
sq
+v A
sq
)
A

sm
= (A
sm
+
vV
sm
c
2
)
A

sm
= A
sm
V

sm
= (V
sm
+v A
sm
).
(24)
The implied relationship between the electromagnetic elds in the two frames is
identical to the transformation for the charges only eld.
E

m
= E
m
E

m
= (E
m
v B
m
)
B

m
= B
m
B

m
= (B
m
+
v E
m
c
2
)
(25)
Combining all the above transformations implies the singularity in equations (23)
8
to transform as:

t
B

m
= q
m
v((r

vt

))(r

)
1

t
E

m
= 0

0
E

m
= 0

m
= q
m
((r

vt

))(r

)
Applying the scaling property of the single variable Dirac distribution |a|(ax) =
(x) and introducing the denition of the position of the singularity r

m
= vt

in
the moving reference system we get:

t
B

m
= q
m
v(r

m
)
1

t
E

m
= 0

0
E

m
= 0

m
= q
m
(r

m
)
where the relativistic coecient in the source terms cancelled out just like in the
charge singularity case.
Consequently, the electromagnetic singular elds E
m
and B
m
are solution of

0
E
m

t
B
m
= q
m
v(r r
m
)
1

0
B
m

t
E
m
= 0

0
E
m
= 0

0
B
m
= q
m
(r r
m
).
(26)
for a magnetic monopole singularity on an arbitrary path r
m
(t) with a velocity
v = r
m
.
The equation of motion of the monopole singularity is deduced by dening the
total eld as the linear superposition of the singular elds E
m
and B
m
and the
external elds E
0
and B
0
:
E = E
m
+ E
0
B = B
m
+B
0
where the external elds can be decomposed onto four potential elds V
0q
, V
0m
, A
0q
and A
0m
E
0
= V
0q

t
A
0q
+
1

0
A
0m
(27a)
B
0
= A
0q
+
0
V
0m
+
0

t
A
0m
(27b)
0 = A
0q
+
0

t
V
0q
(27c)
0 = A
0m
+
0

t
V
0m
. (27d)
9
Introducing the total eld in the electromagnetic Lagrangian (eq. 10) infers the
following singularity action:
S
sm
=

0
E
0
E
m

0
B
0
B
m

d
3
x dt (28)
where we have kept only the terms that vary with respect to the trajectory of the
singularity. Indeed, the terms in E
i
E
i
and B
i
B
i
, where (i = m, 0), are not
dependent on the position of the charge singularity. Rewriting the external eld
with the help of the potentials dened in equations (27a) and (27b) allows us to
perform an integration by part
S
sm
=

(
0
(V
0q

t
A
0q
+A
0m
) E
m

0
(A
0q
+V
0m
+
t
A
0m
) B
m

d
3
x dt
=

(
0
(V
0q
E
m
+A
0q

t
E
m
+A
0m
E
m
)

0
(A
0q
B
m
V
0m
B
m
A
0m

t
B
m
)

d
3
x dt
where we have assumed the elds E
m
and B
m
to be zero at innity i.e. the bound-
ary of integration. Using equations (26) we can further simplify the action of the
singularity
S
sm
=

(q
m
V
0m
(r r
q
) q
m
A
0m
v(r r
q
)) d
3
x dt
=

(q
m
V
0m
q
m
A
0m
v) dt
which leads directly to a Lorentz type forces. Like in the case of the charge singu-
larity, we can rigidly link a scalar mass singularity introduced in previous section to
a monopole singularity. The total action of the double singularity is:
S
s
+ S
sm
=

U
0
m
s

1
v
2
c
2
+ q
m
V
0m
q
m
A
0m
v

dt (29)
which lead to the following equation of movement
d
dt
m
i
v = m
g
U
0
+ q
m
B
0m
+ q
m
v E
0m
(30)
where m
i
and m
g
are the dened by (7) and
E
0m
=
1

0
A
0m
B
0m
=
0
V
0m
+
0

t
A
0m
correspond to the magnetic monopole part of the electromagnetic eld. We remark
here that the force acting on the monopole is only due to the monopole part of the
electromagnetic eld. Consequently, there is no direct interaction between charges
and magnetic monopoles.
The energy balance equation is
d
dt
m
i
c
2
= m
g

t
U
0
+ q
m
v B
0m
. (32)
10
5 Conclusion
The eld generated by a singularity gives rise to the laws of motion of the singularity.
In other words, the action of the singularity on the eld implies its dynamic reaction
to the eld. We have shown this to be the case for neutral bodies, charges and
hypothetical magnetic monopoles. Using our method we have deduced not only the
dynamics of masses, but also the relationship between the gravitational source
strength and the associated inertial mass. The charge singularity was shown to be
the origin of the Lorentz force. The magnetic monopoles have an equivalent Lorentz
force.
The conventional Lorentz force does not involve the magnetic monopole eld.
Reciprocally, the monopole Lorentz force does not involve elds originating from
charged particles. Consequently, charges and magnetic monopoles do not interact,
making the discovery of monopoles using charges dicult. A major question to be
addressed concerns the role played by electromagnetic radiation. Can it provide the
link between charges and magnetic monopoles?
References
[1] L.D. Landau and E.M. Lifshitz, Mechanics, Butterworth Heinemann, Oxford
(1976)
[2] L.D. Landau and E.M. Lifshitz, it The Classical Theory of Fields, Butterworth
Heinemann, Oxford (2002)
[3] N. Grnbech-Jensen, A. R. Bishop, and D. Domnguez, Metastable Filamentary
Vortex Flow in Thin Film Superconductors, Phys. Rev. Lett. 76, p. 2985 (1996)
[4] Ji-Min Duan, Superuid compressibility and the inertial mass of a moving sin-
gularity, Phys. Rev. B 48, p. 333 (1993)
[5] R.

S asik, L.M.A. Bettencourt and S. Habib, Thermal vortex dynamics in a
two-dimensional condensate, Phys. Rev. B 62, p. 1238 (2000)
[6] N.D. Mermin, The topological theory of defects in ordered media, Rev. Mod.
Phys. 51, p. 591 (1979)
[7] M.J. Stephen and J.P. Stranley, Physics of liquid crystals, Rev. Mod. Phys. 46,
p. 617 (1974)
[8] M.R. Dennis, Local phase structure of wave dislocation lines: twist and twirl,
J. Opt. A: Pure Appl. Opt. 6, p. S202 (2004)
[9] A.M. Srivastava, Topological defects in cosmology, Pramana - journal of physics,
56, p. 1069 (1999)
[10] E. Comay, Has the Last Word Been Said on Classical Electrodynamics? Editors:
A Chubykalo, V Onoochin, A Espinoza, and R Smirnov-Rueda, Rinton Press,
Paramus (2004), see also arXiv:physics/0405050
[11] P. Lorrain, D.P. Corson and F. Lorrain, Electromagnetic Fields and Waves,
W.H. Freeman and Company, New York (1988)
11