International Journal of Theoretical Physics, Vol. 21, No.

5, 1982

Mass-Energy in Static Gravitational Fields

Cheng-Deng Kuo

Chest Department, Veterans General Hospital, Shih- Pai, Taipei, Taiwan, R.O.C.

Received May 15, 1981

The mass-energy equation in static gravitational fields is shown to be E = g44mc 2,

which agrees with the expression E=moc2(dx~/ds)~ for the energy in a
gravitational field possessing a timelike Killing vector (. For the Schwarzschild
field this leads to E~ -~ mocz + ½toov2 - kmoM/r. For the Reissner-Nordstr6m
field an additional term describing the interaction between the mass and the
charge is found to be 2~rkmoQ2/c2r ~. In the Kerr-Newman case more terms are
found due to the central rotating gravitating mass.

1. I N T R O D U C T I O N

The famous mass-energy equation of Einstein, E = rnc 2, in the special

theory of relativity is associated with force F in the following m a n n e r :

rp
W = JpoFdx rp d /[ m - dx
= Jpo-d~ ~ ) dx = mc 2 - mo c2 (1)

E =W+ mo c2 = m c 2 (2)
where the force acts on the particle to accelerate it. However, according to
the general theory of relativity, a gravitating system must be described in
terms of a Riemannian geometry defined by the metric tensor g~,,. The
expression of F must then be so generalized to accommodate the g~, and
their derivatives as to properly fit in with general relativity. The mass-energy
equation thus obtained m a y therefore contain g~, or their derivatives as its
components. Equation (2) in flat space is the limiting case of this general
mass-energy equation. Moreover, since gravity is a metric p h e n o m e n o n in
general relativity, it m a y also be expected that the classical gravitational
potential energy will be embedded in this general mass-energy equation.

383
0020-7748/82/0500-0383503.00/0 © 1982 Plenum Publishing Corporation
384 IKuo

2. NONVANISHING MINKOWSKI ACCELERATION VECTOR

The gravitational field equations for nonempty space are

1 ~ -8~rk T~ . (3)
a '~p ~-- R/Lv - - -~gP" R = c2

with T ~ denoting the energy-momentum tensor of all kinds of e n e r g y

content in the space except the gravitational one. When only the noninter-
acting incoherent matter is involved, the energy-momentum tensor is m e r e l y
the matter tensor:

dx~ dx ~
T ~ = M ~ = Po ds ds (4)

The trajectory of a material point in this matter field is then determined by

the geodesic equations

T~'~; v = M~; v = pof ~ = 0 (5)

D ( dx ~' ) d2x ~ ( l~ } dx~ dx ~

:
-As = - d 7 + ds
(6)

where D/ds is the covariant derivative with respect to the affine parameter
s, and the f~ is called the Minkowski acceleration vector. If, however, o t h e r
kinds of force field are present, they must be included in the energy-
momentum tensor T ~ to make it complete. Setting the divergence of the
complete energy-momentum tensor equal to zero will give the correct
equations of motion. Denoting all kinds of nongravitational field other t h a n
the matter field by S ~'~, we have

T ~'~ = M ~ + S ~" (7)

and, as a consequence of this,

f . = - Zs '; (8)
P0

A nonvanishing S ~" will not have zero divergence in general since the S ~'" is
not a complete energy-momentum tensor in itself. Hence the trajectory o f a
material particle in the presence of S ~" field is not, generally speaking, given
by the geodesic equations (5). It is, instead, given by (8). The particle is thus
subject to the effect of S ~ field, and accelerated.
Mass-Energy 385

3. E = g44mc 2 IN THE STATIC GRAVITATIONAL F I E L D S

First of all let us take into account only the static gravitational fields

ds 2 = g44( dx 4 )2 __ gijdxidxj

i, j = 1 , 2 , 3 (9)
g4i = gi4 = 0

x4=ct

g44 and gis being independent of x 4. Dividing the above equation by ds z and
taking the covariant derivatives with respect to s over both sides, we get

g4nf4dx 4 - giJidx j = 0 (10)

where f4 and f i are given by equation (6). The f4 in equation (10) can be
further reduced as follows. From the Christoffel 3-index symbols of the
second kind

(4}=0,
44
4=o,
/j
{4, _
0x'

we have, by virtue of equation (6),

f4__ d2x4 ~g44 dXi dx4

ds 2 ~_g44 Ox ~ ds ds

1 .( (11)

Classically the force is defined by

F=d dx
dt ( m--~ ) (12)

In general relativity we must generalize F to include the metric tensor g~,, as

stated previously. The simplest way is to replace the ordinary derivative
before the parenthesis in equation (12) by the covariant derivative
(Sokolnikoff, 1969)

Fj=D[mdX:l dx4 (13)

dt ~ dt ]' m = m° ds
386 Kuo

This equation can be put in a more convenient form which reads

FJ= m°c2( ----~-

dx 4 - I
) fJ (14)

The work done by the generalized force F s in displacing a particle b2¢ a

distance dx j in the S ~ field is therefore

w = fP~FjdxJ=
Pl jp~
""~ g, j F ' dx j

which becomes, using equations (10), (11), and (14),

w l't2 2 ds ,.4 m czft2d( dx4]

:it mo C -~x4 944J dx4 = 0 "tl x g4"4-"~S ]

Thus

2 dx4
W = moC g44---~-
s +const (15)

The constant of integration can be determined by noting the fact that f f = 0 ,

F j ---0 and consequently W = 0 when the S ~ field vanishes. Hence

2 dz4
const = - moC g44"--~- s~'=o (16)

Substitution back into equation (15) gives

2 dx4 2 dx4
W = moC g44---~ s ~ ' ~ o - m ° c g44 ds s,'=o (17)

The lost term in equation (17) is the energy of the test particle at infinity:

2 ax 4
Eo =moC g44 ds s~'=0 (18)

As the particle proceeds towards the center of the S ~ field, it gains an

amount of energy W to arrive at a total energy

dx 4 I
E = e o + IV= m°cZg44-';7--az. I]s~-~0 (19)
Mass-Energy 387

In view of the nearly identical expression of equations (18) and (19), and the
fact that the S ~" field is arbitrary, we can write the total energy of the
particle in any static field as

dx 4
E = moc-g44----
~- = gn4mc 2 (20)

whether the S "v field is present or not. Being a function of space coordi-
dx 4
nates, the expression of g44--~-
s changes as the particle moves towards or
away from a force field, and the total energy content of that particle
changes accordingly. In flat space, equation (20) is evidently the mass-energy
equation of Einstein in special relativity.

4. E = g 4 4 m c 2 EXPANDED IN TAYLOR SERIES

To get a closer insight into equation (20) we try to expand it in terms of

the Taylor series, and see what meaning each term will have from the
view-point of classical mechanics.
From the metric (9) we get

dx4)2 _ 1 V2 dxi dxS (21)

-'~ g44 -- V2//¢2 ' = gij dt dt

Taking the square root and applying it to the mass-energy equation (20), we
obtain

E = moc2 g44 (22)

(g44--V21C2) I/2

where the positive root is adopted since we want this equation to reduce to
that of special relativity for large radial distances g44~1. The two-
dimensional Taylor expansion of the function F(X, Y) around the point
(Xo = 1, ro = o ) is
X
r(x,Y)= (X- y)l/2
--1+ ~1 ( X _ I ) + ~Y+.--
1 (23)

We have by identifying g44 with X and D2//¢ 2 with Y, and by using equations
(22) and (23),
E -~ m o c2 + ½rot v2 + ½moc2(g44 -- 1) (24)
388 IKuo

for low velocity. The Schwarzschild metric is of the standard form (9);
equations (20) and (24) are therefore directly applicable to it. This yields

1 2 kmoM
Es = m°c2 + 2 m ° v r (25)

in which the subscript S stands for the Schwarzschild field. The first atnd
second terms in (25) are the rest and kinetic energies of the test particle,
respectively. The third term is the familiar gravitational potential energy of
Newton.
In the field of Reissner (1916) and Nordstr/3m (1918), the g44 is given
by

2 k M -~ 4~rkQ2
(26)
g~ = 1 c4r 2
rc 2

So the total energy of the test particle in this field is

2TrkmoQ 2
ER_ N ~ E s + c2r z (27)

where the E s is given by equation (25). The additional term in equation (27)
is due to the presence of an electrostatic field; it may therefore be called the
mass-charge interaction energy between the test particle and the source.
As to the Kerr-Newman metric (Kerr, 1963; Newman et al., 1965)
expresses in terms of Boyer-Lindquist (1967) coordinates

ds 2 = 1- c2dt z - y, "sin2Odepcdt

_ ~.~sin20 dtp2 - - -Z
~ d r -,- - ~ d O 2 (28)

with the following abbreviations:

f ~ _ 2 k M r _ 4~rkQ 2
(29a)
¢2 C4

]9.. = r 2 + a 2 c o s 20 (29b)

A=r~+a2--[2 (29c)

A = ( r z + a2) z - a2A sin20 (29d)

Mass-Energy 389

the presence of the cross term (-2a~2/E)sin2Odq~cdt invalidates the as-

sumption that the metric under consideration is static, i.e., equation (9). To
remove this term we must thus introduce a new time differential in the
rotating system. Redefinition of simultaneity (Adler et al., t975)

cdt* = cdt Z af~

- ~2sin20 drp (30)

brings equation (28) into the following form:

ds2-~(1--~)c2(dt*) 2 ~-~-sin28dep2-~drZ-EdO 2 (31)

The line element (31) and equations (20) and (24) state that in the Kerr-
Newman field the total energy of a test particle measured on the rotating
frame is

f~ ~ 2 cdt*
EK-I¢=(1---~ ) m°c ds (32)

which can be expanded to second order in a / r , and reads

km°MaZc°s20 21rkrn°Q2a2c°s20 (33)

EK_ N -~ ER_ N + r3 -- r4c2

with ER_ u given by equation (27). Since the parameter a is a measure of the
angular momentum per unit mass of the source

Ma = - 1 j (34)
C
J being the angular momentum of the source, we see that rotation of the
source can create extra potentials of rotation besides the Newtonian one. It
is noted, however, that these potentials of rotation shown in equation (33)
vanish in the equatorial plane where 8 = 2" Hence the planetory orbits of
most of the planets in our solar system are not affected by these rotation
potentials in the sense of classical mechanics.
Substituting equation (30) into equation (32), we have

EK_u = moc z 1---~ ds

[ dx 4 . dg 3
m o c2 (35)
t g*'---d- + g,3 )
390 Kuo

where x 3 = % The term mocZga4dxa/ds can be interpreted as the e n e r g y

measured in the coordinate time system, and the term g43moc2dx3/ds is due
to the difference of measurements in the coordinate time system and the
rotating system.
In any geometry endowed with a symmetry described by a K i l l i n g
vector field (, motion along any geodesic whatsoever leaves constant the
scalar product of the tangent vector with the Killing vector (Misner et al.,
1973)

PK = P~'~, = g~,KP~'

dx" (36)
E = moc2g4~ ds

T h e energy in the K e r r - N e w m a n n field (35) is seen to agree with expression

(36) exactly.

REFERENCES

Sokolnikoff, I. C. (1969). TensorAnalysis. John Wiley & Sons, New York, pp. 292-296.
Reissner, H. (1916). Annalen der Physik, 50, 106.
Nordstr6m, G. (1918). Proceedings, Koninklijke NederlandseAkademie t)an Wetenschappen, 20,
1238.
Kerr, R. P. (1963). Physical Review Letters, II, 237.
Newman, E. T., et al. (1965). Journal of Mathematical Physics, 6, 918.
Boyer, R. H., and Lindquist., R. W. (1967). Journal of Mathematical Physics, 8, 265.
Adler, R. J., Bazin, M. J., and Schiffer, M. (1975). Introduction to General Relativity. McGraw-
Hill, New York, pp. 124.
Misner, C. W., Thorne, K. S., and Wheeler, J. A. (1973). Gravitation. Freeman, San Francisco,
pp. 651