Solution of the ECE Vacuum Equations

Horst Eckardt, Douglas W. Lindstrom

Alpha Institute for Advanced Studies (A.I.A.S.) Telesio Galilei Association (TGA)

The vacuum of Einstein Cartan Evans (ECE) theory is a curved and twisted space-time lled by electromagnetic potentials. The vacuum potential can be computed by the condition that electromagnetic force elds vanish. In classical electrodynamics, this condition is not sucient to compute a distribution of potential from given boundary conditions. ECE theory has a richer structure with spin connections of Cartan geometry and additional constraints due to the antisymmetry of the connections in general relativity. Expressions for describing the energy and momentum density are given. It can be shown for the rst time that the high vacuum potentials known from quantum eects are consequences of the structure of space-time itself.
Keywords: Einstein Cartan Evans (ECE) eld theory; antisymmetry;

Abstract

vacuum eld.
Publication Date: 25 August 2009

Introduction

The space free of masses and charges has been a subject of physical interpretation for hundreds of years. Sometimes it was assumed to be empty, sometimes to be a medium called Ether, for example to explain electromagnetic wave propagation. The notion of vacuum often describes empty space, but quantum physicists speak of a quantum vacuum that is not empty at all. In this paper we use vacuum synonymously with space-time itself in the sense of general relativity. We will see that space is lled with a potential, called background potential. In Einstein's theory the vacuum is empty, this result is in sharp contrast to quantum physics, where the vacuum is a sea of virtual particles with a huge energy density. This quantum sea gives rise to the radiative corrections of physics such as the Lamb shift or Casimir eect. In standard physics this is a huge and irreconcilable discrepancy. In Einstein-Cartan-Evans (ECE) theory [1]- [2] both worlds are reconciled, the background in ECE is lled with a potential energy dened directly by the tetrad of Cartan. The potential energy is physical and cannot be arbitrarily changed as in gauge theory. The radiative corrections come from uctuations in this potential of the generally covariant unied eld.
e-mail: [email protected] e-mail: [email protected]

Energy transfer in ECE can take place through resonant solutions of the eld equations. This energy from spacetime in ECE comes from the background potential energy. There is a transfer of energy but the total energy is constant. The conventional continuity equation can be generalized in such a way that curvature and torsion of spacetime can even produce charge carriers. There are experimental hints that such eects can actually happen. In the past there have been some attempts to explain the background elds by solving Maxwell-Heaviside equations for electromagnetic elds. However, the eld equations naturally result to a state where the energy is contained in electric and magnetic elds which are not observed in vacuo (with exception of a 2.7 K background radiation which may exist all over the universe or not). In the current paper we avoid this diculty by using the experimental fact that such elds are not existent. Therefore the background energy observed by quantum processes must be contained in the potentials. We will show how these can be calculated from the basic geometrical properties of space. These are dened Instead by Cartan geometry which is used exclusively in ECE theory. Since the force elds are assumed to be zero, the eld equations identically vanish. we use the basic antisymmetry conditions of the Cartan spin connections and the Maurer-Cartan structure equations. We neglect polarization eects and use the vector form of the equations. In this way we are able to present a general solution of the background potential. We calculate the energy and momentum density by extending a suggestion found in the literature. A discussion section concludes the paper. As a result, we nd a plausible explanation why there are huge energy densities in the vacuum. This theory is based on rst principles, leading to a new view on this subject without need for obscure quantum eects.

2
2.1

Antisymmetry conditions and equations of state

The electric and magnetic eld of ECE theory under omission of polarization eects are

E= B=
relation

A 0 A + , t A A.

(1) (2)

The antisymmetry conditions for the potentials are the electric vector-valued

A 0 A = 0, t
A3 x2 A3 x1 A2 x1

(3)

and the three magnetic (scalar-valued) relations, writte in vector form:

C :=

A2 x3 A1 x3 A1 x2

+ + +

+ 2 A3 + 3 A2 + 1 A3 + 3 A1 = 0. + 1 A2 + 2 A1

(4)

The vacuum conditions are obtained by setting

E = 0, B = 0.

(5) (6)

These are twelve equations for eight unknowns

, A, 0 , .

Taking into account

the antisymmetry constraints (3)-(4), this would lead to an over-determined system. Therefore we smplify the constraints by taking the divergence of Eqs. (3) and (4):

A t

(0 A)

() = 0, C = 0.

(7) (8)

The last equation leads to mixed second derivatives and can be tried to simplify by using the Lindstrom constraint which is a simplication of Eq. (4):

A = A.
On the other hand, from Eqs. (2) and (6) we have

(9)

A = A.

(10)

which is not compatible with the Lindstrom constraint (9) (see section 5.2). Therefore we have to use the full condition (4). In total we have to solve the equation set

A t A2 x1 x3 A1 + x2 x3 A1 + x3 x2
2.2

A 0 A + = 0, t A A = 0, () = 0,

(11) (12) (13)

(0 A) +

A3 + 2 A3 + 3 A2 x2 A3 + + 1 A3 + 3 A1 x1 A2 + + 1 A2 + 2 A1 x1

(14)

= 0.

Pre-evaluation of magnetic constraints

The direct equation set derived in the previous section is not very handable to numerics and the structure of the equations and possible solutions is not obvious. Therefore we follow a dierent line of development. By adding Eqs. (4) and (10) we obtain

A2 + 2 A3 = 0, x3 A1 + 1 A3 = 0, x3 A1 + 1 A2 = 0. x2

(15)

(16)

(17)

Subtracting (4) from (10) gives

A3 + 3 A2 = 0, x2 A3 + 3 A1 = 0, x1 A2 + 2 A1 = 0. x1
Comparing (16) with (17) etc. leads to

(18)

(19)

(20)

A1 1 A1 1 = , x2 A2 x3 A3 A2 1 A2 1 2 = = , x1 A1 x3 A3 A3 1 A3 1 3 = = x1 A1 x2 A2 1 =
which can be written as

(21)

(22)

(23)

Aj
with

Ai Ai = Ak xk xj (1, 2, 3).

(24)

(i, j, k)

being permutations of

solving Eqs. (24) as a boundary

value problem and inserting the solution into (21)-(23) gives a full solution of the magnetic constraint (4) in terms of the vector potential

A.

So far we have handled Eqs. (4) and (6) from the given set (3)-(6). Next we have to handle the electric conditions (3) and (5). The only variables remaining to be determined are

and

0 .

From inserting (3) into (1) we have [3]

E=

+ A = 0 A t

(25)

which gives us two vector equations

=0, A + 0 A =0. t
Restricting to the static case, from the latter equation follows

(26) (27)

0 = 0.
we take the divergence of (26):

(28)

Since Eq. (26) is a vector eqation and we only need to determine one variable,

2
or

() = 0

(29)

) = 0.

(30) Solving this

This is identical with the Coulomb law for zero charge density. dierential equation for the static ECE vacuum equations.

(with (A) already given) completes the solution for

Solutions of the vacuum equations

According to Equation set (24) the vacuum solution for the

pled from the electric potential and the spin connections. We rst consider the solutions of this equation set. According to computer algebra (we used Mathematica [8]) there are three solutions with a number of integration constants

Ci ,

ki

and

: A(1) = k 1 (C5 + C6 tanh(k x t)) k3

A(2) = k

1 (C5 + tanh(k x t) (C6 + C7 tanh(k x t))) k3

(32)

A(3) = k

(C5 + tanh(k x t) (C6 + tanh(k x t) (C7 + C8 tanh(k x t))))

(33)

Details can be found in the appendices. In the form presented here some constants

Ci

have been renamed to clarify their physical meaning. The rst three

constants are written as the wave vector proper argument of the

and the fourth constant which is

a phase factor has been interpreted as time dependence

so that we get a

tanh

function. In principle all constants could be time

dependent as shown in the appendices. Comparing the three solutions, it is obvious that these are parts of a more general solution of type

A(m) = k

1 k3

Dn (t) (tanh(k x t))

(34)

with (possibly time-dependent) canstants

Dn .

This is a basis set of the function

space spanning the solutions of Eq. (24). We will return to this point later. Next we will derive the solutions of the other variables. The spin connection

follows directly from (21)-(23):

(1) = k
the corresponding dence of

C6 (sech(k x t)) , C5 + C6 tanh(k x t)

(35)

(2,3) ,

can be found in Appendix A.

The scalar spin connection can be computed from the explicit time depen-

by using Eq. (27). In principle this equation - being a vector equa-

tion - is over-determined, but as shown in Appendix A all three components are consistent, giving three solutions from which the rst is

0 =

(1)

C6 (sech(k x t)) C5 + C6 tanh(k x t)

(36)

which is similar to the components of the vector spin connection.

We have

assumed that the coecients are independent of time. Finally the pontential

can be derived similarly from Eq. (26). Again all three components of this However, as an

equation lead to the same result, achieving self-consistency.

important nding, a part of the solution (denoted by regardless of the choice of

F (i) )

is always the same,

A(1) , A(2) ,

or

A(3) ,

for example:

(1,i) =
where

cosh(k x t) F (i) C5 cosh(k x t) + C6 sinh(k x t)

(37)

F (i)

is an arbitrary function depending on only one coordinate surface:

F (1) = F (y, z, t), F (2) = F (x, z, t), F (3) = F (x, y, t).

(38)

F = const.

makes the solution unique. To obtain an impression 1-4.

of the nature of vacuum solutions, some of them are graphed in Figs. The constants

Ci were chosen 1 or -1 and the wave vector was positioned in z direction: k = (0, 0, k3 ). Then there are only z components of A and . Fig. 1 shows that A3 varies the more the higher the degree of the solution is. A
similar result holds for the vector spin connection (Fig. 2), however the second solution is a pole, indicating that there are discontinuities in the structure of the vacuum. The same holds for the scalar spin connection (Fig. 3) which is similar to for

3 . Most interesting is the potential, see Fig. 4. It is pole-like or diverges z . This may be quite unusual from the classical view, but we know

from experiments that the energy density of the vacuum is very high. Such a structure may give rise e.g. to sponaneous creation of elementary particles; see sect. 5.5 for further discussion. Besides this particular solution, we tried to nd a more general solution. As already shown in Eq. (34), a general series of tanh functions fulls this. We are lead to the conjecture that a general series expansion of the form

A(m) = k
for any complete function set

1 k3 fn

Dn (t)f n (k x t)
n=0
with constants

(39)

Dn

is also a solution for

This has been shown in Appendix B by assuming this form of

A. A and proving all

equations. The Computer Algebra tool was only able to prove this for a nite series, but the general result follows easily by induction. In particular we get:

(m) = k 0
(m)

f f

m n1 (k n=1 nDn f m n (k x Dn f n=0 m n1 (k n=1 nDn f m n (k x Dn f n=0 m n=0

x t) , t) x t) , t)

(40)

(41)

(m) = f
is the derivative of

0 . Dn f n (k x t)

(42)

according to its argument. As an example the general It is seen that for large

solution for the

tanh

function is depicted in Fig. 5 for several maximum indices

m (all constants set to unity).

m the solution is a smooth

step function with growing height. The vector and scalar spin connection only dier in sign and a factor indicating the space-like and time-like character of both connetcions. The potential contains the original function series in the

5 4 3 2 1 0 -1 -2 -4 -2 0 z

solution 1 solution 2 solution 3

A3(z)

Figure 1: Three solutions for vector potential component

A3 .

10

solution 1 solution 2 solution 3

3(z)

-5

-10 -4 -2 0 z 2 4

Figure 2: Three solutions for vector spin connection component

3 .

10

solution 1 solution 2 solution 3

0(z)

-5

-10 -4 -2 0 z 2 4

Figure 3: Three solutions for calalar spin connection component 7

0 .

20 0 -20 (z) -40 -60 -80 -100 -4 -2 0 z

solution 1 solution 2 solution 3

Figure 4: Three solutions for scalar vacuum potential

14 12 10 8 A3(z) 6 4 2 0 -2

m=2 m=5 m=8 m=11 m=14 m=17

-4

-2

0 z

Figure 5: Vector potential component sion.

A3

for several degrees

of series expan-

denominator.

If the the series is oscillating (which normally is the case), the

potential has a high number of resonances. Another important class of solutions is obtained by a Fourier series:

A(m) = k

1 k3

Dn exp (n i(k x t)) .

(43)

This is a Fourier series in one dimension only because there is no variation in the wave vector or the

k.

In other words, this is a plane wave with xed direction. A

variation in direction, however, can occur if the set of rst four constants

(k, )

Dn 's

is time-dependent which is possible for the general solution (see

Appendices). This is further discussed in sections 5.2 and 5.3.

4
4.1

Energy and momentum density

The classical expression for the energy density of the electromagnetic eld is

u(r, t) =

1 1 2 2 B , 0E + 2 20 1 E B. 0

(44)

and the power ux is described by the Pointing vector

S(r, t) =
Physical units of these quantities are

(45)

J , m3 W [S] = 2 . m [u] =
Both are based on the electromagnetic force elds

(46) (47)

E and B and are therefore not

suitable in situations of potentials without elds. The solution to this diculty is to formulate the energy density and power ux in terms of the potentials. This problem has been studied by Ribaric and Sustersic [4]. They found that other expressions can be dened in terms of potentials which give the same total energy taken as an integral over space as the original terms (44, 45). A practical solution has been reported by Putho [5]. For a given current density density written

J,

charge

scalar potential

and vector potential

the energy density can be

u(r, t) = uA u +
with

(48)

uA (r, t) =
and

1 20

1 c2

Ai t

+ | Ai |

(49)

u (r, t) =

1 2

1 c2

+ | |

(50)

Correspondingly, the power ux is dened by

S(r, t) = SA S + J, SA (r, t) = S (r, t) = 1 0

(51)

Ai t .

Ai ,

(52)

(53)

According to Jackson [6] the change in energy is dened by the divergence of the ux and the electric energy of the charges:

u = t

S J E.

(54)

In vacuo we have equivalent

momentum density

= J = E = 0. The Pointing g of the electromagnetic eld gc2 = S

ux is connected with the by Einstein's mass-energy

(55)

and the equation of motion for the eld is

g = t
(written with tensor divergence [7]) where

is the Maxwell stress tensor.

(56)

4.2

ECE theory
(1-2). We apply some reasonable replacement

In ECE theory there are additional terms contained in the elds according to the spin connections, see Eqs. rules for genral relativity:

+ 0 , t t + .
Therefore we add such terms to the denitions (48-53) analogously:

(57) (58)

uAECE (r, t) = uECE (r, t) =

1 20 1 2
0

1 2 2 |0 Ai | + |i Ai | , c2 |i |
i 2

(59)

1 (0 )2 + c2

(60) (61)

u(r, t) = uA + uAECE u uECE ,

SAECE (r, t) = SECE (r, t) =

1 0
0

(0 Ai )Ai ,
i

(62) (63) (64)

(0 ) ,

S(r, t) = SA + SAECE S SECE .

This procedure should be considered as tentative. For example all cross-terms in the products have been omitted. Note that there is a Pointing ux even if and

are not time-dependent.

When inserting the general solutions (39-42) with abbreviations

g :=
n=0 m

Dn f n (k x t),

(65)

g := f
n=1

nDn f n1 (k x t)

(66)

10

into the above expressions of energy density and power ux, we obtain:

uA (r, t) = u (r, t) = uAECE (r, t) = uECE (r, t) =

k 2 (k 2 + ( )2 )2 2 c g 2 20 k3
0

(67)

(k 2 + ( )2 )2 2 0 c g 2g 4

(68)

4 4 4 k1 + k2 + k3 + k 2 ( )2 2 c g 2 20 k3 0

(69)

(k 2 + ( )2 )2 2 0 c g 2g 4 k 2 2 SA (r, t) = k 2g 0 k3 2 S (r, t) = 0 k 40 g 2 g k 2 2 SAECE (r, t) = k 2g 0 k3 2 SECE (r, t) = 0 k 40 g 2 g

(70)

(71)

(72)

(73)

(74)

It can be seen that the energy densities of ECE and standard theory are equal

uA

and

have been graphed in Figs. 6 and 7 for illustration. Interestingly the power ux Within the approximations

The vector potential contributes smooth densities while the scalar potential leads to innities due to its diverging character. is exactly opposite in ECE and standard theory. made, there is no energy transfer in vacuo.

Discussion

In the last section we want to discuss some further points found for the background or vacuum potential.

5.1

Lindstrom constraint

In section 2.1, Eqs. (9-10), it was pointed out that the Lindstrom constraint normally being used to simplify the equations of electrodynamics is not suited to describe the vacuum vector potential consistently. Alternatively, we can try to retain the Lindstrom condition. Then we have to assume the validity of both equations, it follows

A = 0, A = 0.
Eq. (75) could alternatively be used to compute two components of

(75) (76)

but analysis shows that only

can be determined, they depend on the third component

which remains unspecied. Therefore the method of basing the solution on Eqs. (21)-(23) is preferrable as it leads to unique results. It should be noted that the Lindstrom constraint can even be derived from (21)-(23). Therefore it is not a restriction on the solution in case of the background potential.

11

1.4 1.2 1 uA(z) 0.8 0.6 0.4 0.2 0 -4 -2 0 z

solution 1 solution 2 solution 3

Figure 6: Energy density

uA

for three solutions of tanh type.

1.4 1.2 1 u(z) 0.8 0.6 0.4 0.2 0 -4 -2 0 z

solution 1 solution 2 solution 3

Figure 7: Energy density

for three solutions of tanh type.

12

A3 Function 5 4 3 2 1 0 -1 -2 -6 -4 -2 x 0 2 4 6 -6 -4 -2 0 z 2 4 6 5 4 3 2 1 0 -1 -2

Figure 8: Surface plot of

A3

(3)

3 Function 5 0 -5 -10 -15 -20 -6 -4 -6 -4 -2 -2 z 0 0 2 2 4 4 6 6 x

5 0 -5 -10 -15 -20

Figure 9: Surface plot of

(2)

13

5.2

Relations of time dependent constants

As shown in the appendices, all integration constants can be time-dependent. In Appendix A the condition

k1 k k = 2 = 3 k1 k2 k3
was derived. A time-dependent

(77)

would not guarantee energy conservation.

This had to be accounted for by an additional condition like

k(t)2 = k1 (t)2 + k2 (t)2 + k3 (t)2 = const.

(78) (77) that all

is time-independent, it follows from Eq.

components are time-independent. Therefore we surmise that required.

k cannot be time-

dependent in general. Then introducing additional conditions like (78) is not

5.3

Character of solutions
(40)-(41) it can directly be seen that the ratio of the vector and

From Eqs.

scalar spin connection is

k = 0
and similarly

(79)

for all components

1 Ai

(80)

Ai .

This shows some interconnection of the potentials and

spin connections. The scalar spin connection can be written as

0 =
The vector potential

log g . t

(81)

is rotation free (see Eq. (76)) and therefore could be In principle

written as a gradient of a scalar function but this would be a potential of a potential and purely mathematical. arbitrary function

multiplied by a directional vector

A can k:

be dened as an

A = k f (k x t).
This is a representation for plane waves as is depicted in Figs. 8-9 for

(82)

and

Since

can be arbitrarily nonlinear, these are anharmonic waves in general.

The propagation speed (phase velocity) (restricted to one dimension)

can be obtained from the relation

A A dt 1 A = = . x t dx v t
Inserting the form of (82) it follows

(83)

1 kf = f v
14

(84)

or

|v| =

(85)

which is identical to the classical dispersion relation for electromagnetic waves in vacuo. However, the time dependence of constants in the solution (39) also allows other forms of time dependence, for example a product wave function

A = k f (kx) g(t).
The dispersion relation then becomes

(86)

|v| =

g k f

(87)

which allows arbitrary travelling velocities for waves depending on the forms of

and

g.

The use of product wave functions is similar to quantum mechanics Since the vacuum solutions are not constrained

where we have instantaneous interaction eects which appear plausible in the light of the equation above. wave forms. by the ECE (or Maxwell-Heaviside) eld equations, there is more freedom for This subject may be promising for new types of superluminar communication mechanisms and should be investigated further. Another type of wave to be investigated is standing waves in vacuo. A further natural interpretation would be a uctuating background eld. It is known from observations [9] - although not widely recognized - that such uctuations do exist, resulting in a variance of vacuum speed of light. The origin of these uctuations is speculative at the current state of knowlegde, it may be from global motion of galaxies or local distortions of background elds due to matter of massive stars. Further investigation of these eects would probably require usage of uid dynamics models.

5.4

Resonant Coulomb law

It was shown that the resonant Coulomb (Eq. (29)) follows from the vacuum equations. equations, the Coulomb law, here appears although the eld equations vanish identically because of the condition

E = B = 0.

law is deeply anchored in geometry and is a very fundamental law of nature. Secondly there are resonances possible even in the vacuum. elsewhere [10] the resonances of the Coulomb law arise from a variable spin connection and do not necessarily need an oscillatory charge density as in classical Euler-Bernoulli resonances. Charge density is not present in vacuo.

5.5

Topological charge density

In classical eld theory the Coulomb or Newtonian gravitational law can be written in the form of the Poisson equation:

(88)

15

20 0 -20 (z) -40 -60 -80 -100 -4 -2 0 z

solution 1 solution 2 solution 3

Figure 10: Topological charge density

for three solutions of tanh type.

(here for the electromagnetic case). In vacuo there is no charge density, therefore we would expect

= 0.

(89)

Comparing this with Eq. (29), the spin connection terms appear as a topological charge density:

top =
As

().

(90)

has diverging regions so has the topological charge density. It has to be However, it could be the

stressed that this density is not made up of real charges. The topological charge density consists of certain structures of space-time. pairs of particles can appear spontaneously. An example of topological charge densities is shown in Fig. 10. The picture is very similar to the corresponding potential (Fig. 4). The form looks similar to atomic or molecular charge densities. This could give a hint to virtual particles, although ECE theory is completey conned to the wave or eld model, in contrast to the particle model of quantum mechanics. origin of real particle processes which are known from quantum electrodynamics:

Acknowledgments

The AIAS and TGA colleagues, in particular Kerry Pendergast, are thanked for many interesting discussions and contributions.

16

References
[1] M. W. Evans, Generally Covariant Unied Field Theory (Abramis, Suffolk, 2005 onwards), vol. 1-6 (see also www.aias.us). [2] L. Felker, The Evans Equations of Unied Field Theory (Abramis, 2007). [3] www.aias.us, UFT papers 133 and 134 (2009). [4] M. Ribaric and L. Susersic, Conservation Laws and Open Questions of Classical Electrodynamics (World Scientic Pub. Do., singapore, 1990). [5] H. E. Putho, Electromagnetic Potentials Basis for Energy Density and Power Flux, http://arxiv.org/PS_cache/arxiv/pdf/0904/0904.1617v1.pdf [6] J. D. Jackson, Classical Electrodynamics, Third Edition (John Wiley and Sons, 1999), pp.258-262. [7] ibd., pp.608-610. [8] www.wolfram.com. [9] R. T. Cahill, Combining NASA/JPL One-Way Optical-Fiber Light-Speed Data with Spacecraft Earth-Flyby Doppler-Shift Data to Characterise 3Space Flow, Progress in Physics, vol.4, pp.50-64 (2009). [10] H. Eckardt, Devices for Space-Time Resonance Based on ECE Theory, http://aias.us/documents/miscellaneous/Spacetime-Dev.pdf (2008).

17