Exponential Metric Fields
arXiv:0710.5071v2 [gr-qc] 15 Jun 2010
Wytler Cordeiro dos Santos
Abstract The Laser Interferometer Space Antenna
(LISA) mission will use advanced technologies to
achieve its science goals: the direct detection of gravitational waves, the observation of signals from compact
(small and dense) stars as they spiral into black holes,
the study of the role of massive black holes in galaxy
evolution, the search for gravitational wave emission
from the early Universe. The gravitational red-shift,
the advance of the perihelion of Mercury, deflection of
light and the time delay of radar signals are the classical tests in the first order of General Relativity (GR).
However, LISA can possibly test Einstein’s theories in
the second order and perhaps, it will show some particular feature of non-linearity of gravitational interaction.
In the present work we are seeking a method to construct theoretical templates that limit in the first order
the tensorial structure of some metric fields, thus the
non-linear terms are given by exponential functions of
gravitational strength. The Newtonian limit obtained
here, in the first order, is equivalent to GR.
so different from others fields, for example the electromagnetic field, the scalar field, etc., by themselves obey
linear equations in a given spacetime, they form a nonlinear system when their mutual interactions are taken
into account. The distinctive feature of the gravitational field is that it is self-interacting (as the YangMills field): it is non-linear even in the absence of other
fields. This is because it defines the spacetime over
which it propagates (Hawking & Ellis 1973).
Linearized gravity is any approximation to General
(0)
(0)
Relativity obtained from gµν = gµν + hµν (where gµν is
any curved background spacetime) in Einstein’s equation and retaining only the terms linear in hµν (Wald
1984). The weakness of the gravitational field means
in the context of general relativity that the spacetime is nearly flat. Small gravitational perturbations
in Minkowski space can be treated in the simplest linearized version of General Relativity,
Keywords Linearized gravity, Newtonian limit, Gravitational waves
as describing a theory of a symmetric tensor field hµν
propagating on at background spacetime. This theory is Lorentz invariant in the sense of Special Relativity. If one wants to obtain a solution of the nonlinear equations, it is necessary to employ an iterative
method on approximate linear equations whose solutions are shown to converge in a certain neighbourhood
of initial surface. It should be possible to avoid some
of these difficulties of non-linearity by working in some
spacetimes that shall be described in this paper. The
proposal is that some metric fields can be separated
into the parts carrying the dynamical information and
those parts characterizing the coordinates system. In
this proposal, terms of the coordinates system will have
tensorial structure limited only in the first order. The
tensors that will describe gµν have linear behavior. Naturally, there is a price that must be paid for the linear
1 Introduction
The extreme difficulties which arise if one tries to draw
physically important conclusions from the basic assumptions of Einstein’s theory are mainly due to the
non-linearity of the field equations. Moreover, the fact
that the spacetime topology is not given a priori, and
the impossibility to integrate tensors over finite regions
cause difficulties unknown in other branches of mathematical physics. Actually in this respect they are not
Wytler Cordeiro dos Santos
e-mail: wytler@fis.unb.br
Universidade de Brası́lia, Faculdade do Gama, Brası́lia DF
gµν = ηµν + hµν ,
(1)
2
tensors. The dynamic terms that carry the gravitational strength have exponential structure. The principal idea came from the basic principle that one should
interpret (1) as separation between pure mathematical and physical terms in metric field tensor. In spite
of the fact that ηµν plays a key role as empty flat and
background spacetime of the Standard Model in the description of fundamental interactions, this background
tensor metric ηµν is an object wholly mathematical and
entirely geometrical, while hµν contains the physical information. The strength of gravity is tied in the components of hµν . The proposal of this paper is a working
hypothesis to untie the strength of gravity from geometrical tensors. This proposal is valid for a family of
metric field tensors gµν , and some basic examples such
Newtonian limit and gravitational plane waves of low
amplitudes are described.
This paper is outlined as follows: in Sec. II, we
present the basic mathematic concepts of the (quasi)idempotent tensors that compose the structure of metric fields approached in this work. In Sec. III, we
propose how to link the strength of gravity with the
tensors from Sec. II, then is defined a family of exponential metrics. In Sec. IV, we present some examples of these exponential metrics, such as: Yilmaz metric, circularly polarized wave and rotating bodies. In
Sec. V, we present exponential metrics (‘adjoint metric field’) that are non-physical, but which help us to
compute Christoffel symbols, and consequently the curvature tensors, Ricci tensors and determinant of metric
field. In Sec. VI, we verify the Newtonian limit and
also we obtain gravitational waves. In Sec. VII, we
present a general conclusion.
We assume spacetime (M, g) to be a C ∞ 4−dimensional, globally hyperbolic, pseudo-Riemannian manifold M with Lorentzian metric tensor g (whose components are gµν ) associated with the line element
ds2 = gµν (x)dxµ dxν ,
assumed to have signature (+ − −−) (Landau & Lifschitz 1984). Lower case Greek indices refer to coordinates on M and take the values 0, 1, 2, 3. The relation
between the metric field gµν and the material contents
of spacetime is expressed by Einstein’s field equation,
8πG
1
Rµν − gµν R = 4 Tµν ,
2
c
(2)
Tµν being the stress-energy-momentum tensor, Rµν the
contract curvature tensor (Ricci tensor) and R its trace.
In an empty region of spacetime we have Rµν = 0, such
a region is called vacuum field.
2 Pure Mathematical terms of spacetime
geometry
The Minkowski flat spacetime (R4 , η), where the components of η are ηµν = diag(1, −1, −1, −1), is the simplest empty spacetime and it is in fact the spacetime of
Special Relativity. One can obtain spaces locally identical to (R4 , η) but with different (large scale) topology
properties by identifying points in R4 which are equivalent under a discrete isometry without a fixed point.
Any local Lorentz frames is merely the statement that
any curved space has the Minkowski flat space ‘tangent’ to it at any point. The Minkowski spacetime is
the universal covering space for all such derived spaces
(Hawking & Ellis 1973). In this sense it should be reasonable to one choose (R4 , η) as background spacetime
and important compound piece of some spacetimes. In
fact, the most straightforward approach to linear gravitation is realized in Minkowski spacetime. The conformal structure of Minkowski space is what one would
regard as the ‘normal’ behavior of a spacetime at infinity.
The metric tensor η from background Minkowski
spacetime in any coordinates is an object wholly mathematical and entirely geometrical. No strength of gravity is linked up with the mathematical structure of η.
Then, we choose this metric tensor as a principal descriptive piece of some spacetimes built below, and it
does rise and lower the indices in the same way as in
Special Relativity with ηµν η να = δµ α . The aim of this
paper is to obtain some spacetimes (M, g) that their
non-linearity are less hard, softer at least in a tensorial
descriptive way. Therefore, it is defined only a symmetric tensor Υ, which like η is an object wholly mathematical and entirely geometrical. This (metric) tensor
Υ that will be a piece of metric g can have non-static
terms from spacetime in their components, however in
this approach, this tensor purely will not have gravitational strength. The components Υµν of tensor Υ are
raised and lowered by η µν and ηµν ,
η µν Υνα = Υµ α ,
Υµν η να = Υµ α ,
µν
η Υνα η αβ = Υµβ .
(3)
In this context it is adopted the point of view, that Υ
is a tensor on a background Minkowski spacetime, similar to deviation hµν from linearized version of general
relativity (1). But instead one has the infinitesimal condition for |hµν | ≪ 1, it is accepted that the magnitude
of Υµν can be equal to the magnitude of empty flat
spacetime (|Υµν | ≈ |ηµν |). Moreover, it might be defined as an important mathematical relationship among
3
Υµν themselves,
Υµν Υνρ = −2Υµ ρ .
(4)
The above equation is an important argument to shape
some spacetimes that are described below. This equation will improve linearity in the tensorial sense. The
contracting indices of tensor Υ by themselves show that
Υµν are (quasi-)idempotent elements (Υ · Υ ∝ Υ; otherwise a factor −2 in the operation), and the equation
(4) improves at least to a linear tensorial template of
some tensor metric fields. It is possible to rise or lower
indices of Υµν operating themselves,
ρµ
Υ Υµν Υ
νσ
ρσ
= 4Υ .
(5)
One can also verify the expression Υµ ν Υνρ = −2Υµρ .
The trace of Υ is obtained when ρ = µ in the expression (4),
Υµν Υνµ = −2Υµ µ ,
(6)
and realize derivative calculation of trace Υµ µ from (6)
since ηµν = diag(1, −1, −1, −1),
µ
µν
−2∂α Υµ = 2Υµν ∂α Υ ,
(7)
g(Φ, x) = eΦ η + sinh(Φ)Υ,
with components
gµν = eΦ ηµν + sinh(Φ)Υµν .
Υµν ∂α Υµν = Υµν ∂α Υµν = −∂α Υµ µ .
(8)
Only both these tensors, η from background Minkowski
spacetime and Υ, will become the mathematical and
geometrical basis to physical spacetimes described as
follows.
3 Strength of gravity terms tied in spacetime
geometry
A dimensionless parameter Φ characterizing the strength of gravity at a spacetime point ℘ with coordinates
(t, x) = (xα ) due to a gravitating source is the ratio of
the potential energy, mϕN (due to this source), to the
inertial mass-energy mc2 of a test body at ℘, i.e.,
ϕN (xα )
.
c2
(9)
Here ϕN (xα ) is the gravitational potential. For a point
source with mass M in Newtonian gravity,
Φ(xα ) = −
GM
,
c2 r
(10)
where r is the distance to the source. So, for a nearly
Newtonian system, we can use Newtonian potential for
ϕN .
(11)
The inverse tensor is just
g µν = e−Φ η µν − sinh(Φ)Υµν .
(12)
If the tensor Υ is diagonal, then the inverse tensor of g
is
g−1 (Φ, x) = g(−Φ, x) = e−Φ η − sinh(Φ)Υ.
(13)
From the above definitions it follows that gµν g να = δµ α ,
in fact
gµν g
thus,
Φ(xα ) =
To construct spacetimes with the basis η from background Minkowski spacetime and Υ defined in previous
section, it is proposed to tie the strength of gravity Φ
to theses tensors. While the tensor Υ can be a function
of the coordinates (t, x) = (xα ), the strength of gravity
Φ will be a function of the coordinates and also of the
Newtonian gravitational constant G and of a parameter
of mass M . Thus, I do propose:
να
=
Φ
−Φ να
να
e ηµν + sinh(Φ)Υµν
e
η
− sinh(Φ)Υ
where the rightside is
δµ α + sinh(Φ) e−Φ Υµν η να − eΦ ηµν Υνα − sinh(Φ)Υµν Υνα
that is
δµ
α
i
h
α
να
Φ
−Φ
+ sinh(Φ) −(e − e
)Υµ − sinh(Φ)Υµν Υ
α
να
α
2
δµ − sinh (Φ) 2Υµ + Υµν Υ
=
(14)
Using the expression (4) in the second term in the
parenthesis (Υµν Υνα = −2Υµ α ) it results in,
gµν g να
=
δµ α − sinh2 (Φ) [2Υµ α + (−2)Υµ α ] ,
(15)
finally one did prove this identity gµν g να = δµ α . Because of equation (4), there are no quadratic values naturally in Υ and consequently the non-linearity of metric
field will be less complicated.
If one considers some spacetime (M, g) such as the
same of g from definition (11), one can observe that the
first term γµν = eΦ ηµν is a background spacetime conformally flat (where the Weyl tensor vanishes). In the
previous section one has accepted that |Υµν | ≈ |ηµν |,
but now because of sinh (Φ) (that can be small) multiplying this tensor Υ, it can be understood as (small)
disturbance Sµν = sinh(Φ)Υµν , such as,
gµν = γµν + Sµν ,
(16)
4
if one is dealing with the Einstein’s vacuum equation,
the (small) disturbance that represents the gravitational wave can be separated from the conformally flat
background spacetime γµν (Birrell & Davies 1982).
4 Some Examples
4.1 Yilmaz Metric
As a first application of the metric field defined in previous sections, let us take Υµν to be:
Υµν
0
0
=
0
0
0
2
0
0
0
0
2
0
0
0
0
, Υµ ν = 0
0
0
0
2
0
2
0
0
0
0
2
0
0
0
0
2
0
−2
0
0
0
0
−2
0
0
0
0
−2
and
Υµν
0
0
=
0
0
(17)
that Υµν Υνα = −2Υµ α , with trace given Υµν Υνµ =
−2Υµ µ = −2TrΥ, that Tr(Υ) = −6 . Now we can
display the tensor metric (11),
gµν
gµν
=
=
0
1
0
0
0
0
−1
0
0
0
+
sinh(Φ)
eΦ
0
0
0
−1
0
0
0
0
0
−1
Φ
e
0
0
0
0
−e−Φ
0
0
,
0
0
−e−Φ
0
0
0
0
−e−Φ
0
2
0
0
0
0
2
0
0
0
0
2
(18)
GR:
g00
=
1+Φ+
Φ2
3Φ3
+
+ ···
2
16
g11 = −1 + Φ −
3Φ2
− ··· ,
8
g11 coefficients differing only in the second order terms,
while g00 differing in the third order. Both, Yilmaz and
GR, give observational indistinguishable predictions for
red-shift, light bending and perihelion advance, but
the Yilmaz metric does not admit black holes. Citing this property and assuming that Yilmaz theory is
correct, Clapp (Clapp 1973) has suggest that a significant component of quasar red-shift may be gravitational. Robertson (Robertson 1999a,b) has suggested
that some neutron stars and black hole candidates may
be like ‘Yilmaz stars’. Robertson argues that neutron
star with mass ∼ 10M⊙ is found for Yilmaz metric
while that an object of nuclear density greater than
∼ 2.8M⊙ should be a black hole in Schwarzschild metric. Ibison has tested Yilmaz theory by working out
the corresponding Friedmann equations generated by
assuming the Friedmann-Robertson-Walker cosmological metrics (Ibison 2006). There are a series of claims
and counter-claims involving Fackerell (Fackerell 1996;
Yilmaz 1994; Alley & Yilmaz 2000), and also Misner
and Wyss (Wyss & Misner 1999; Misner 1999; Alley &
Yilmaz 1999) about Yilmaz theory. At the present time
both Yilmaz and Schwarzschild solutions give results
in agreement with observation (Rosen 1974). However,
it may be possible in the future, with LISA mission
(NASA 2010; Baker et al. 2007), to distinguish between
Yilmaz and Schwarzschild.
since −eΦ + 2 sinh(Φ) = −e−Φ . Then the line element
is,
4.2 Circularly Polarized Wave
ds2 = eΦ c2 dt2 − e−Φ (dx2 + dy 2 + dz 2 ).
Gravitational waves are one of the most important predictions of General Relativity. Now we can try a solution of gravitational waves in z direction,
(19)
This metric field (18,19) has been proposed by Yilmaz
(Yilmaz 1958, 1992, 1976, 1982, 1973, 1977, 1997). In
2GM
the case of a mass singularity, Φ = − 2
≪ 1 we
c r
have the far-field metric,
ds2 = (1 −
2GM
2GM 2 2
)c dt − (1 + 2 )(dx2 + dy 2 + dz 2 ). (20)
c2 r
c r
This is to be contrasted with the Schwarzschild (in
General Relativity, GR) line element in isotropic coordinates (Landau & Lifschitz 1984),
Υµν
Υµ ν
ds =
1 + Φ/4
1 − Φ/4
2
2
2
c dt −
Φ
1−
4
4
2
2
2
2
2
2
(dr + r dθ + r sin θdφ ),
(21)
if we compare expansions with the line element of Yilmaz theory (19),
Yilmaz:
g00
=
1+Φ+
Φ2
Φ3
+
+ ···
2
6
g11 = −1 + Φ −
Φ2
+ ···
2
=
0
1 + cos ζ
sin ζ
0
0
0
Υµα η αν =
0
0
0
sin ζ
1 − cos ζ
0
0
−1 − cos ζ
− sin ζ
0
0
0
,
0
2
0
1 + cos ζ
sin ζ
0
0
sin ζ
1 − cos ζ
0
0
− sin ζ
−1 + cos ζ
0
0
0
0
−2
(22)
and
Υ
2
=
0
0
0
0
µν
= Υαβ η
αµ βν
η
0
0
=
0
0
0
0
0
2
(23)
so Υµν Υνα = −2Υµ α is verified for the tensor above
and Tr(Υ) = −4. Then, if one chooses ζ = ωt − kz,
the metric field is a solution for a gravitational plane
5
wave gµν = eΦ ηµν + sinh(Φ)Υµν ,
gµν
=
+
eΦ
0
0
0
the presence of g00 and g33 as static terms, weakly pertubed in these coordinates.
0
0
0
− cosh(Φ)
0
0
0
− cosh(Φ)
0
0
0
−eΦ
0
0
0
0
0 cos ζ
sin
ζ
0
sinh(Φ)
0 sin ζ − cos ζ 0 .
0
0
0
2
4.3 Rotating Bodies
(24)
Where the first term can be the background spacetime
(asymptotically flat) and the second term is the disturbance in this background or in other words the circularly polarized radiation hTµνT with amplitude sinh(Φ).
The gravitational wave polarization is important from
astrophysical and cosmological viewpoints. A binary
system of two stars in circular orbit one around the
other is expected to emit circularly polarized waves in
the direction perpendicular to the plane of the orbit
(Schutz 1990). Moreover, the Big Bang left behind an
echo in the electromagnetic spectrum, the cosmic microwave background, but the Big Bang most likely also
left cosmological gravitational waves that will be possible to observe with the help of LISA (NASA 2010; Baker
et al. 2007). Since cosmological gravitational waves
propagate without significant interaction after they are
produced, once detected they should provide a powerful
tool for studying the early Universe at the time of gravitational wave generation (Buonanno 2004). Various
mechanisms for cosmological gravitational wave generation have been proposed, and many of these state
that the cosmological gravitational wave are circularly
polarized. T. Kahniashvili et al (Kahniashvili et al.
2005) argued that helical turbulence produced during a
first-order phase transition generated circularly polarized cosmological gravitational waves. Other physicists
have said that the parity violation due to the gravitational Chern-Simons term in superstring theory can
produce the primordial gravitational waves with circular polarization (Lue et al. 1999; Choi et al. 2000;
Alexander et al. 2006; Satoh et al. 2007; Saito et al.
2007).
If we assume long distance from source, we can obtain plane wave solution with Φ ≪ 1 so that,
gµν
=
=
(1 + Φ)ηµν + ΦΥµν
1+Φ
0
0
0
−1
0
0
0
0
0
−1
0
0
0
0
−1 + Φ
0
0
0
0
sin ζ
0
0 cos ζ
+Φ
0 sin ζ
− cos ζ 0
0
0
0
0
the second term is the circularly polarized radiation
hTµνT . Because of the far distance from source, we have
The Kerr metric is important astrophysically since it
is a good approximation to the metric of a rotating
star at large distances. It is possible to obtain a kind
Kerr metric from gµν = eΦ ηµν + sinh(Φ)Υµν , that in
coordinates (t, x, y, z) the tensor Υµν is:
cosh2 Λ
− sinh Λ cosh Λ sin φ
sinh Λ cosh Λ cos φ
− sinh Λ cosh Λ sin φ
sinh2 Λ sin2 φ
− sinh2 Λ cos φ sin φ
0
0
0
0
0
sinh Λ cosh Λ cos φ
− sinh2 Λ cos φ sin φ
sinh2 Λ cos2 φ
0
0
(25)
satisfying Υµν Υνρ = −2Υµ ρ with Tr(Υ) = −2. In
this example we choose Υ33 = 0, but if the choice was
Υ33 = 2, the above tensor still satisfy the algebra (4).
One can change coordinates (t, x, y, z) to the BoyerLindquist coordinates (t, r, θ, φ), with spatial part as
flat space in ellipsoidal coordinates,
t
x
y
z
=
=
=
=
t,
p
2
2
pr + a sin θ cos φ,
2
r + a2 sin θ sin φ,
r cos θ,
(26)
The Minkowski tensor metric related to them is:
1
ηµν
0
=
0
0
0
0
2
2 cos2 θ
− r +a
r2 +a2
0
0
0
−r 2 − a2 cos2 θ
0
0
0
− r 2 + a2 sin2 θ
0
,
(27)
we are assuming that the angle φ from tensor Υµν of
(25) can be the same angle from transformations (26).
So, this coordinate transformations will become tensor
(25) in:
Υµν = (−2)
cosh2 Λ
0
0
sinh Λ cosh ΛR sin θ
0
0
0
0
0
0
0
0
sinh Λ cosh ΛR sin θ
0
0
sinh2 ΛR2 sin2 θ
(28)
√
where R = r2 + a2 . At the appendices, it is verified
with details that the above tensor obeys the algebra (4)
in the background Minkowski spacetime (27). Now we
can choose a particular solution for this tensor choosing
the geometric terms sinh Λ and cosh Λ:
√
a sin θ
r 2 + a2
sinh Λ = −
and
cosh Λ =
, (29)
ρ
ρ
6
(0)
with
ρ2 = r2 + a2 cos2 θ
(30)
satisfying cosh2 Λ − sinh2 Λ = 1.
The physical terms that contain the strength of gravity eΦ and sinh Φ can be: 1
Φ=
Mr
≪ 1,
+ a2 )
(31)
(r2
so that:
eΦ ≈ 1 + Φ = 1 +
Mr
(r2 + a2 )
and
sinh(Φ) ≈ Φ =
Mr
.
+ a2 )
(r2
We can compute each term of metric field gµν (for more
details see appendices),
g00
=
g03
=
g11
=
g22
=
g33
=
Mr
∆ − a2 sin2 θ
+ 2
,
ρ2
(r + a2 )
2
2M ra sin θ
g30 =
,
ρ2
2
2
ρ
M rρ
− +
,
∆
∆(r 2 + a2 )
2
M rρ
−ρ2 − 2
,
r h + a2
i
2
2
2
− sin θ r + a2 − ∆a2 sin2 θ
ρ2
4.4 Deformed Schwarzschild spacetime
− M r sin2 θ,
where we use the definition ,
∆ = r2 − 2M r + a2 .
(32)
The metric tensor gµν in the matrix form is:
∆−a2 sin2 θ
2M ra sin2 θ
ρ2
ρ2
0
0
0
− ρ∆
0
0
−ρ2
2M ra sin2 θ
ρ2
0
0
+
0
2
with
0
0
Υµ ν = Υµα η αν
0
ρ2
0
0
M rρ
∆(r 2 +a2 )
0
0
0
0
0
rρ
− rM2 +a
2
0
2
0
−M r sin2 θ
.
(33)
and
The above first matrix is just ‘exact Kerr solution’.
While the second matrix can be seen as deviation or
deformity from ‘exact solution’. Then, it makes sense
to see this deviation as approximate solution of
(0)
gµν = gµν
+ hµν ,
1 In
Another example can be given in spherical coordinates
(t, r, θ, φ) where the components of metric tensor of
Minkowski flat spacetime is:
ηµν = diag(1, −1, −r2 , −r2 sin θ),
with the respective inverse
1
η µν = diag(1, −1, − r12 , − r2 sin
θ ). As we have seen before, let us describe a simpler (quasi-)idempotent tensor
in spherical coordinates given by
0 0
0
0
0 2
0
0
Υµν =
(35)
2
2
0 0
r
−r sin θ
0 0 −r2 sin θ r2 sin2 θ
0
0
i
h
2
− sin2 θ (r 2 +a2 ) −∆a2 sin2 θ
Mr
(r 2 +a2 )
2
where gµν is some known exact solution (that in this
case is Kerr solution) and hµν is the perturbation. For
Einstein’s vacuum equation it is possible to obtain explicitly the vacuum perturbation equations from an arbitrary exact solution (Wald 1984).
Gravitational wave observations of extreme-massratio-inspirals (EMRIs) by LISA will provide unique
evidence for the identity of the supermassive objects in
galactic nuclei. It is commonly assumed that these objects are indeed Kerr black holes. K. Glampedakis and
S. Babak argue that from the observed signal, LISA will
have the potential to prove (or disprove) this assumption, by extracting the first few multipole moments
of the spacetime outside these objects. The possibility of discovering a non-Kerr object should be taken
into account when constructing waveform templates for
LISA’s data analysis tools. They provide a prescription
for building a ‘quasi-Kerr’ metric, that is a metric that
slightly deviates from Kerr, and present results on how
this deviation impacts orbital motion and the emitted
waveform (Glampedakis & Babak 2006).
this paragraph we assume G = 1 and c = 1.
(34)
Υµ ν = η µα Υαν
0
0
−1
sin θ
0
0
1
sin θ
−1
0
0
−1
0
0
.
sin θ
−1
0 0
0 −2
=
0 0
0 0
0 0
0 −2
=
0 0
0 0
1
sin θ
(36)
(37)
7
Then Tr(Υ) = Υµ µ = −4. One can show that Υµν
given by η µα Υαβ η βν is,
Υµν
0
0
=
0
0
0
2
0
0
0
0
1
r2
1
− r2 sin
θ
0
0
1
− r2 sin
θ
1
r 2 sin2 θ
(38)
These examples suggest that the Tr(Υ) ∈ Z are constant numbers. Forward it will be necessary to calculate
the derivative of Tr(Υ) in any situations, thus we shall
assume that,
∂α Tr(Υ) = 0
(42)
in this paper.
that satisfies the algebraic relation (4),
Υµα Υαν
=
=
0
0
0
0
0
0
0
4
0
0
0
2
− sin2 θ
0 −2 sin θ
2
0
0
0
0
0 −2
0
0
ν
−2
1 = −2Υµ . (39)
0
0
−1
sin θ
0
0
sin θ −1
From expression (11) we have the metric field gµν =
eΦ ηµν + sinh ΦΥµν , it follows that,
g00
grr
gθθ
gφφ
gθφ
= eΦ ,
1
= − Φ,
e
= −r2 cosh Φ,
= −r2 sin2 θ cosh Φ,
= gφθ = −r2 sin θ sinh Φ,
(40)
in the case that Φ = − 2GM
c2 r ≪ 1 with line element
expanded in the first order of the strength of gravity,
2
ds
=
+
dr 2
2GM
− r 2 dθ 2 − r 2 sin2 θdφ2
c2 dt2 −
1− 2
c r
1 − 2GM
2
c r
{z
}
|
Schwarzschild spacetime
GM
4
r sin θdθdφ,
(41)
c2
This metric field is asymptotically flat, the components
approach those of Minkowski spacetime in spherical coordinates. If only gθφ would be vanished, we could obtain the Schwarzchild spacetime. However, it is necessary the terms Υθφ = Υφθ = −r2 sin θ in the symmetric
tensor of equation (35) for this tensor may obey the algebraic relation (4). Nevertheless, it raises a question:
what kind of gravitating source should distort a spherically symmetric spacetime? The main purpose of this
section is to illustrate some (quasi-)idempotent tensors
Υ that must satisfy the algebraic relation (4). There
are many works with discussions about Yilmaz metric field and their sources. However, this paper does
not discuss the physics of gravitating sources of circularly polarized wave, rotating bodies and deformed
Schwarzschild spacetime purposed here. In forthcoming work, it will be necessary to analyse principally the
gravitating source that distort the static Schwarzschild
solution.
5 Adjoint Metric Field, Christoffel Symbols
and determinant
5.1 Adjoint Metric Field
One can see that spacetime (11) is asymptotically flat.
The Minkowski spacetime is the universal covering
space for all such derived spacetimes, and in this sense
it is possible to restore Minkowski spacetime in (11) if
one turns off the gravitational strength in metric field,
or in other words, if Φ = 0 then gµν = ηµν . However
one can construct a kind of spacetime with hyperbolic
cosine instead of hyperbolic sine that is not asymptotically flat,
ğµν = eΦ ηµν + cosh(Φ)Υµν ,
(43)
with respective inverse,
ğ µν = e−Φ η µν + cosh(Φ)Υµν ,
(44)
and with ğµν ğ να = δµ α since Υµν Υνα = −2Υµ α from
equation (4). A map between metrics field (11) and
(43) is obtained through partial derivative in Φ:
ğµν =
∂gµν
∂Φ
and
ğ µν = −
∂g µν
.
∂Φ
(45)
∂ 2 gµν
∂ 2 g µν
µν
and
g
=
.
∂Φ2
∂Φ2
This ‘adjoint metric field’ (43) helps us to simplify
the Christoffel symbols. But before, it is necessary to
establish some relationships between g and ğ,
also one has gµν =
gµν ğνα
=
=
+
(eΦ ηµν + sinh(Φ)Υµν )(e−Φ η να + cosh(Φ)Υνα )
δµ α + [eΦ cosh(Φ) + e−Φ sinh(Φ)]Υµ α
να
sinh(Φ) cosh(Φ)Υµν Υ ,
(46)
with Υµν Υνα = −2Υµ α and hyperbolic expressions
eΦ = cosh(Φ) + sinh(Φ) and e−Φ = cosh(Φ) − sinh(Φ)
we have:
gµν ğ να = δµ α + Υµ α ,
(47)
and,
g µν ğνα = δ µ α + Υµ α .
(48)
8
where,
It is important to observe that:
gµν ğ µν = g µν ğµν = Tr(η) + Tr(Υ).
(49)
=
Γβ(1)
µν
Further we have:
gµν Υνα = (eΦ ηµν + sinh(Φ)Υµν )Υνα = e−Φ Υµ α , (50)
g
µν
Φ
Υνα = e Υ
µ
α,
(51)
ğµν Υνα = −e−Φ Υµ α ,
(52)
ğ µν Υνα = −eΦ Υµ α .
(53)
Nevertheless for this ‘adjoint metric field’, when
there is not gravitational strength (Φ = 0) it is not possible to reorder Minkowski spacetime. Because if Φ = 0
then gµν = ηµν + Υµν . Unless all Υµν are vanished, one
can not obtain Minkowski spacetime. Moreover, with
Υµν 6= 0 a spacetime (M, ğ) has a different topology of
physical spacetimes. For example, if one chooses Υµν
from equation (17), the correspondent line element is
ds̆2 = eΦ c2 dt2 + e−Φ (dx2 + dy 2 + dz 2 ),
(54)
and
Γβ(2)
µν =
1
sinh(Φ)g αβ (∂µ Υαν + ∂ν Υαµ − ∂α Υµν ). (61)
2
Since Γ is not a tensor, it can not have intrinsic geometrical meaning as measures of how much a manifold
is curved. Below we shall compute intrinsic objects as
Ricci tensor and scalar curvature.
5.3 Determinant g
From identity (Schutz 1990),
√
Γνµν = ∂µ (ln −g),
=
5.2 Christoffel Symbols
We want a manifold M with Levi-Civita connection,
since the connection coefficient Γ is given by
Γβµν =
1 αβ
g (∂µ gαν + ∂ν gαµ − ∂α gµν ).
2
(55)
=
g αν ∂µ Υαν
=
+
=
+
where we used (8) and (42), thus,
√
1
[Tr(η) + Tr(Υ)] ∂µ Φ = ∂µ (ln −g). (63)
2
(57)
and we finally find,
√
Φ
[Tr(η) + Tr(Υ)] .
−g = exp
2
1 αβ h
g
ğαν ∂µ Φ + ğαµ ∂ν Φ − ğµν ∂α Φ
2
i
sinh(Φ)(∂µ Υαν + ∂ν Υαµ − ∂α Υµν )
i
h
1
(δ β ν + Υβ ν )∂µ Φ + (δ β µ + Υβ µ )∂ν Φ − gαβ ğµν ∂α Φ
2
1
sinh(Φ)gαβ (∂µ Υαν + ∂ν Υαµ − ∂α Υµν ).
(58)
2
Let us consider Christoffel symbols as a combination of
two terms: the first is dependent of ∂α Φ and the second
is dependent of ∂α Υ,
β(2)
Γβµν = Γβ(1)
µν + Γµν
e−Φ η αν ∂µ Υαν − sinh(Φ)Υαν ∂µ Υαν
e−Φ ∂µ Tr(Υ) − sinh(Φ)[−∂µ Tr(Υ)] = 0,
Γνµν
from which we obtain that Christoffel symbols are
Γβ
µν
=
=
(56)
that in term of (43) is
∂α gµν = ğµν ∂α Φ + sinh(Φ)∂α Υµν ,
1h
Tr(η)∂µ Φ + ∂µ Φ + Tr(Υ)∂µ Φ + Υα µ ∂α Φ
2
i
−(δ α µ + Υα µ )∂α Φ
1
+ sinh(Φ)g αν (∂µ Υαν + ∂ν Υαµ − ∂α Υµν )
2
1
1
[Tr(η) + Tr(Υ)] ∂µ Φ + sinh(Φ)g αν ∂µ Υαν ,
2
2
we note that the second term with g αν ∂µ Υαν is vanished,
We have that,
∂α gµν = (eΦ ηµν + cosh(Φ)Υµν )∂α Φ + sinh(Φ)∂α Υµν ,
(59)
(62)
where g = det(gµν ), it is possible to obtain the determinant g by contracting the indices in (59) with β = ν
and the equation (48):
Γνµν
that has signature (+ + ++).
1 β α
δ ν δ µ + δ β µ δ α ν + δ α µ Υβ ν + δ α ν Υβ µ − gαβ ğµν ∂α Φ
2
(60)
=
(64)
In fact, since ln(det gµν ) = Tr(ln gµν ). Also, one should
√
use the identity dg = g g µν dgµν and to compute −g.
One can verify the examples of above section in coordinates (t, x, y, z):
Yilmaz metric:
circularly polarized wave:
rotating bodies:
g = − exp {Φ [4 + (−6)]} = −e−2Φ
g = − exp {Φ [4 + (−4)]} = −1
g = − exp {Φ [4 + (−2)]} = −e2Φ
The natural volume element of manifold M with
√
metric tensor (11), dv = −gd4 x, is invariant under
9
coordinate transformation. An action with Lagrangian
L in spacetime (M, g) is given by,
Z
√
L −g d4 x,
S=
(65)
M
then we have,
Z
Φ
[Tr(η) + Tr(Υ)] d4 x.
L exp
S=
2
M
For Newtonian limit, we must retain only the terms in
the first order in the strength of gravity Φ, or in other
words, we have ∂Φ∂Φ ≪ ∂∂Φ and Φ∂∂Φ ≪ ∂∂Φ , with
this assumption we have the second term of (68) and
(69) such as,
1
2∂µ ∂ν Φ + Υβ ν ∂µ ∂β Φ + Υβ µ ∂ν ∂β Φ
2
1
−ğµν g αβ ∂α ∂β Φ − [Tr(η) + Tr(Υ)]∂µ ∂ν Φ.
2
(1)
Rµν
=
(66)
6 Newtonian Limit and Gravitational Waves
6.1 Newtonian Limit
One can choose the simplest Υ, and it should be from
(17) such as Tr(Υ) = −6, where the Ricci tensor is now
given by
(1)
Rµν = 2∂µ ∂ν Φ +
In the Newtonian limit one assumes that velocities are
v
≪ 1, that gravitational potentials are near
small,
c
their Minkowski values, 2 and that pressures or other
mechanical stresses are negligible compared to the energy densities |P | ≪ ρc2 .
The description of Einstein’s field equation (2) will
use values from Christoffel symbols from equation (59),
β(1)
where this has two terms: Γµν dependent of ∂α Φ and
β(2)
Γµν dependent of ∂α Υ. In this section we shall verify
the Einstein tensor Gµν = Rµν − 21 gµν R, only in the case
β(2)
that ∂α Υ = 0, then Γµν = 0, and in any coordinates
system, the Ricci tensor is given by
α(1)
α(1)
(1)
β(1)
α(1)
β(1)
Rµν
= ∂α Γα(1)
µν −Γµβ Γνα −∂ν Γµα +Γαβ Γµν . (67)
(1)
1
ğµν ğ αβ − gµν g αβ ∂α Φ∂β Φ
2
1
+ 2∂µ ∂ν Φ + Υβ ν ∂µ ∂β Φ + Υβ µ ∂ν ∂β Φ
2
−ğµν g αβ ∂α ∂β Φ ,
(68)
=
the second term of Ricci tensor is:
α(1)
Γµβ Γβ(1)
να
=
1
3
∂µ Φ ∂ν Φ + (∂µ ΦΥγ ν + ∂ν ΦΥγ µ )∂γ Φ
2
2
1
1
− gµν g λγ ∂γ Φ ∂λ Φ + Υλ µ Υγ ν ∂γ Φ ∂λ Φ,
2
2
with (63) we can obtain the third term:
1
= [Tr(η) + Tr(Υ)]∂µ ∂ν Φ,
2
∂ν Γα(1)
µα
(69)
and the fourth is
1
[Tr(η) + Tr(Υ)] 2∂µ Φ∂ν Φ
4
β
+∂µ ΦΥ ν ∂β Φ + ∂ν ΦΥβ µ ∂β Φ − ğµν g αβ ∂α Φ∂β Φ . (70)
α(1)
Γβα Γβ(1)
µν =
2 Observe
ηµν
g µν
1 β
1 β
1
αβ
Υ ν ∂µ ∂β Φ + Υ µ ∂ν ∂β Φ − ğµν g ∂α ∂β Φ.
2
2
2
(72)
The scalar curvature is obtained by further contracting
(1)
indices R(1) = g µν Rµν ,
R(1) = 3g µν ∂µ ∂ν Φ + g µν Υβ ν ∂µ ∂β Φ,
(73)
g µν = (1 − Φ)η µν − ΦΥµν in terms of the first order in
Φ, we find that the scalar curvature is
R(1) = 3η µν ∂µ ∂ν Φ + η µν Υβ ν ∂µ ∂β Φ.
(74)
Since the Newtonian limit approach is in the background Minkowski spacetime, we find that
R(1) = 32Φ + Υβµ ∂µ ∂β Φ.
(75)
From the field equations (2) we have
Since ∂α Υ = 0 the first term of Rµν is given by
∂α Γα(1)
µν
(71)
that gµν = eΦ ηµν +sinh(Φ)Υµν ≈ (1+Φ)ηµν +ΦΥµν =
+ Φ(ηµν + Υµν ) = ηµν + hµν and
≈ ηµν − Φ(ηµν + Υµν ) = ηµν − hµν .
R=−
8πG
T,
c4
(76)
in the situations where Newtonian theory can be applicable T ≈ ρc2 . Now we can combine equations (76)
and (75) such as
32Φ + Υβµ ∂µ ∂β Φ = −
8πG
ρ.
c2
(77)
We assume that velocities are small, then
2=
∂2
c2 ∂t2
− ∇2 ≈ −∇2 .
(78)
And in particular from (17) we have Υµν ⇒ 2δij , where
δij is the Kronecker delta in R3 . The field equation (77)
results in:
−3∇2 Φ + 2δ ij ∂i ∂j Φ = −
8πG
ρ,
c2
(79)
or
∇2 Φ =
8πG
ρ.
c2
(80)
10
2ϕN
We can introduce Φ = 2 , where ϕN is the Newtoc
nian potential. Thus we can obtain the Poisson’s equation of Newton’s law of gravitation
∇2 ϕN = 4πGρ,
(81)
in this way for a point source with mass M we have
ϕN = −
GM
.
r
(2)
Rµν
=
(82)
2GM
Consequently Φ = − 2 . In this discussion that folc r
lows (71) we assumed that Υµν is from (17), it results
that the line element is given by (20) at thefirst order
in Φ.
Gravitational waves are one of the most important
physical phenomena associated with the presence of
strong and dynamic gravitational fields. Though such
gravitational radiation has not yet been detected directly. There is strong indirect evidence for its existence
around the famous binary pulsar PSR 1913+16 (Taylor & Weisberg 1982), that in 1974 it was discovered by
R.A. Hulse and J.H. Taylor (Hulse & Taylor 1975), a
discovery for which they were awarded the 1993 Nobel
Prize.
Here, for a description of plane gravitational waves
we can assume that Φ is constant so ∂α Φ = 0, implying
β(2)
β(1)
in Γµν = 0 in (59), thus we have Γβµν = Γµν given by
equation (61):
1
sinh(Φ)g αβ (∂µ Υαν + ∂ν Υαµ − ∂α Υµν ). (83)
2
One can obtain the field equation writing the Ricci tensor in terms of Christoffel symbols above,
α(2)
1
sinh(Φ) g αβ (∂α ∂µ Υβν + ∂α ∂ν Υβµ − ∂α ∂β Υµν ),
2
(87)
with sinh(Φ) g αβ ≈ Φ[(1 − Φ)η αβ − ΦΥαβ ] = Φη αβ , one
can obtain the Ricci tensor in Minkowski background
spacetime, which this yields,
(2)
Rµν
=
1
Φ (∂ β ∂µ Υβν + ∂ β ∂ν Υβµ − 2Υµν ),
2
(88)
The gauge choice ∂ β Υβν = 0, simplifies Ricci tensor in:
6.2 Gravitational Waves
Γβ(2)
µν =
The study of gravitational waves involves essentially
the approximation of Einstein’s weak field equation,
from these results one can obtain wave equation with
low amplitudes, Φ ≪ 1. This can only be brought in
the first order of the gravitational potential to find,
α(2)
(2)
β(2)
α(2)
β(2)
Rµν
= ∂α Γα(2)
µν −Γµβ Γνα −∂ν Γµα +Γαβ Γµν . (84)
From equation (63) we have
1
Γα
αν = [Tr(η) + Tr(Υ)]∂ν Φ = 0, such as,
2
α(2)
(2)
β(2)
Rµν
= ∂α Γα(2)
µν − Γµβ Γνα ,
(89)
One should observe that scalar curvature
(2)
R(2) = g µν Rµν
is vanished for the first order in Φ, because
R(2) =
1
1 µν
η Φ 2Υµν = 2Tr(Υ) = 0,
2
2
since the trace of tensor Υ is a constant number, in accordance with equation (42). Then, for the field equation (2) the gravitational wave equation obtained in the
vacuum is only,
2Υµν = 0.
(90)
It is necessary that the tensor Υµν satisfies the condition (4) and the above wave equation. Its shape should
be, for example, the tensor (22) with ζ = κµ xµ =
ωt − kz, a solution for gravitational plane waves with
circular polarization traveling in the z direction, with
amplitude Φ.
7 Conclusion
(85)
As shown in appendix, the The above Ricci tensor is
given by,
1
(2)
Rµν
= sinh2 (Φ) ∂α Υαβ (∂µ Υβν + ∂ν Υβµ − ∂β Υµν )
2
1
+ sinh(Φ) g αβ (∂α ∂µ Υβν + ∂α ∂ν Υβµ − ∂α ∂β Υµν )
2
1
− sinh2 (Φ)[g αγ g βλ ∂µ Υγβ ∂ν Υαλ
4
+2(g αγ g βλ − g αβ g γλ )∂β Υγµ ∂α Υνλ ].
1
(2)
Rµν
= − Φ 2Υµν .
2
(86)
This paper deals with a tensorial structure that assumes
a (quasi-)idempotent feature to be able to improve at
least the linear tensorial template of some tensor metric
fields. It is clear that Einstein’s field equations are nonlinear, however, with these (quasi-)idempotent tensorial
structure, without quadratic tensorial values, the nonlinearity becomes more moderate although there is a
price to pay. The part that carries the dynamical information, the strength of gravity is tied to the tensorial
11
structure by exponential functions. In this approach
the metric field can be characterized by a background
spacetime conformally flat affected by a disturbance.
We have approached some examples in this tensorial
structure that results in exponential metric fields, we
can point out as the main exponential metric obtained
in this paper which has been extensively explored: the
Yilmaz exponential metric (Yilmaz 1958, 1992, 1976,
1982, 1973, 1977, 1997; Clapp 1973; Robertson 1999a,b;
Ibison 2006). H. Yilmaz has argued that in his theory,
the gravitational field can be quantized via Feynman’s
method (Yilmaz 1995; Alley 1995). Further, it has been
found that the quantized theory is finite. Incidentally
in the exponential metric fields approached in this work
just as in the Yilmaz theory there are no black holes in
the sense of event horizons, but there can be stellar collapse (Robertson 1999a,b). However, there are no point
singularities.
Interesting results obtained in this work from exponential metric fields are: circularly polarized wave; rotating bodies that in the first order is a deformation of
Kerr metric and also we have a deformed static spherically symmetric spacetime. Many discussions around
massive stellar objects have suggested, for example,
that Kerr metric should be slightly deviated from Kerr.
The possibility of discovering a non-Kerr object should
be taken into account when constructing waveform templates for LISA’s data analysis tools (Glampedakis &
Babak 2006; White 2006). The technological development is ripe enough so much so in the years to come we
might be able to test the second order relativist-gravity
effects and may lead to answers to some important questions of gravity.
In this work, we have obtained a simple and general expression for the volume element of a manifold
in coordinates (t, x, y, z) given in terms of strength of
gravity and of traces of tensors η and Υ. It is possible that an analysis of any Lagrangian of field interacting with gravity will become easer. An interesting observation is the spacetime of circularly polarized plane wave, in this spacetime the volume element
√
−g d4 x is the same of Minkowski spacetime, in this
sense this gravitational radiation obtained from exponential metric field does not modify the volume element of background Minkowski spacetime where this
plane wave travels onto. Moreover, it was purposed
and verified the Newtonian limit as solution for Einstein’s equation, since we can assume that the trace of
stress-energy tensor is T ≈ ρc2 . Other important solution of Einstein’s equation analysed in this paper was
the plane gravitational wave for the empty space since
we have considered the vanished stress-energy tensor to
the first order in Φ. Both solved Einstein’s equations
for Newtonian limit and plane gravitational wave propagating in the vacuum are cases that the strength of
gravity is small, Φ ≪ 1. We have analysed the Newtonian limit in the case that ∂α Υ = 0, and analysed
the plane gravitational wave considering the strength of
gravity as a constant term, thus we had two indepen(1)
dent Ricci tensors, Rµν which ∂α Υ = 0 (for Newtonian
(2)
limit) and Rµν which ∂α Φ = 0 (for plane gravitational
wave). In a forthcoming work, an analysis of Einstein’s
equations with both non-vanished ∂α Υ and ∂α Φ, will
be considered.
It is missing a discussion about quantities of physical
interest in the solutions of Einstein’s equations which
describe the exterior and interior gravitational field.
Yilmaz has argued the existence of the matter part in
the right-hand side of the field equations correspondent
to field energy in the exterior. This paper lacks a discussion about the interior and the exterior field energies
denoted by a total stress-energy tensor. An analysis
about the total stress-energy tensor will be the object of
a forthcoming study, where the physical consequences
of terms of deformity in Kerr and Schwarzschild solutions could be analysed.
We know that the dark energy and the dark matter problems are challenges to modern astrophysics and
cosmology; as a typical example, we could mention the
galactic rotation curves of spiral galaxies, that probably, indicates the possible failure of both Newtonian
gravity and General Relativity on galactic and intergalactic scales. To explain astrophysical and cosmological problems with arguments against dark energy
and dark matter many works have been devoted to the
possibility that the Einstein-Hilbert Lagrangian, linear
in the Ricci scalar R, should be generalized. In this
sense, the choice of a generic function f (R) can be derived by matching the data and the requirement that
no exotic ingredient have to be added (Allemandi et al.
2004; Barrow et al. 1983; Capozziello 2002; Capozziello
et al. 2003, 2005; Carroll et al. 2004, 2005; Faraoni
2008; Flanagan 2003; Koivisto 2006; Nojiri et al. 2007,
2003a,b; Sotiriou et al. 2010). This class of theories
when linearized exhibits others polarization modes for
the gravitational waves, of which two correspond to the
massless graviton and others such as massive scalar and
ghost modes in f (R) gravity (Bellucci et al. 2009; Bogdanos et al. 2009). In this way, analyses in any order
to f (R) gravity with ‘exponential metrics’ proposed in
the present work could give a positive contribution to
the debate of astrophysical and cosmological questions.
Acknowledgements It is a pleasure to acknowledge
many stimulating and helpful discussions with my colleagues and friends André L. A. Penna and Caio M. M.
Polito.
12
A Details of Calculation for Rotating Bodies
A.1 Multiplicative Properties of Υ in Boyer-Lindquist coordinates
The metric tensor ηµν of Minkowski flat spacetime in Boyer-Lindquist coordinates is given by:
ηµν
1
0
0
r 2 +a2 (cos(θ))2
0 −
r 2 +a2
=
0
0
0
0
0
0
0
2
−r2 − a2 (cos (θ))
0
0
2
− r2 + a2 (sin (θ))
while the (quasi-)idempotent tensor Υµν from (25)
2
(cosh (Λ))
0
0
0
Υµν = (−2)
0
0
√
sinh (Λ) cosh (Λ) r2 + a2 sin (θ) 0
,
(A1)
in the same coordinates is:
0
0
0
0
√
sinh (Λ) cosh (Λ) r2 + a2 sin (θ)
0
.
0
2
2
(sinh (Λ)) r2 + a2 (sin (θ))
(A2)
Let us verify that above tensor Υµν satisfies the algebraic relation (4). We begin with the inverse metric tensor of
Minkowski space:
η
µν
1
0
0
0 − 2 r22 +a2 2
r +a (cos(θ))
=
0
0
0
0
0
0
0
−1
2
− r2 + a2 (cos (θ))
0
0
1
− (r2 +a2 )(sin(θ))
2
such as Υµ ν e Υµ ν are calculated with contracting indices,
Υµ ν = η µα Υαν
and
= (−2)
Υµ ν = Υµα η αν
= (−2)
2
(cosh (Λ))
0 0
0
0 0
0
0 0
cosh(Λ)
√
− sinh(Λ)
r 2 +a2 sin(θ)
0 0
,
(A3)
√
sinh (Λ) cosh (Λ) r2 + a2 sin (θ)
0
0
2
− (sinh (Λ))
2
(cosh (Λ))
0
0 0
0 0
cosh(Λ)
√
− sinh(Λ)
r 2 +a2 sin(θ)
0
0
0 0
0
√
sinh (Λ) cosh (Λ) r2 + a2 sin (θ)
0 0
− (sinh (Λ))
2
(A4)
.
(A5)
13
Notice that it follows the contravariant tensor Υµν given by:
2
cosh(Λ)
√
(cosh (Λ))
0 0 − sinh(Λ)
r 2 +a2 sin(θ)
0
0 0
0
µν
µα
βν
Υ = η Υαβ η = (−2)
0
0 0
0
(sinh(Λ))2
cosh(Λ)
√
0 0 (r2 +a
− sinh(Λ)
2 )(sin(θ))2
r 2 +a2 sin(θ)
.
(A6)
Finally one can obtain that Υµν Υνα = −2Υµ α , this is proved by observing that,
2
Υµν Υνα
= (−2) · (−2)
(cosh (Λ))
0
0 0
cosh(Λ)
√
− sinh(Λ)
r 2 +a2 sin(θ)
0 0
0
0
0 0
0
√
sinh (Λ) cosh (Λ) r2 + a2 sin (θ)
0 0
− (sinh (Λ))
2
= −2Υµ α .
(A7)
It is easy to see that the above tensor is the same of equation (A5) as claimed.
A.2 Calculation of components
Let us compute the components of metric tensor gµν = eΦ ηµν + sinh(Φ)Υµν for rotating bodies. First we may
calculate g00 = eΦ η00 + sinh(Φ)Υ00 ≈ (1 + Φ)η00 + ΦΥ00 ,
Mr
r 2 + a2
Mr
,
−2· 2
·
2
2
+a )
(r + a )
ρ2
√
r 2 + a2
Mr
since Φ = 2
from (31), and also cosh Λ =
from (29),
2
(r + a )
ρ
g00
=
1 + Φ − 2Φ cosh2 Λ = 1 +
g00
=
1−
(r2
Mr
Mr
ρ2 − 2M r
2M r
+
+ 2
=
,
2
2
2
2
ρ
(r + a )
ρ
(r + a2 )
(A8)
(A9)
with ρ = r2 + a2 cos2 θ then,
g00
=
=
=
r2 + a2 cos2 θ − 2M r
Mr
r2 + a2 (1 − sin2 θ) − 2M r
Mr
+
=
+ 2
2
2
2
2
ρ
(r + a )
ρ
(r + a2 )
2
2
2
2
Mr
r + a − 2M r − a sin θ
+ 2
ρ2
(r + a2 )
Mr
∆ − a2 sin2 θ
+ 2
,
2
ρ
(r + a2 )
(A10)
where ∆ = r2 − 2M r − a2
Calculation of g03 :
g03
=
=
p
−2 sinh(Φ) sinh Λ cosh Λ r2 + a2 sin θ !
√ 2
a sin θ
2M ra sin2 θ
Mr
r + a2 p 2
r + a2 sin θ =
−
,
−2 2
2
r +a
ρ
ρ
ρ2
again we used the definition (29).
(A11)
14
Calculation of g22 :
g22 = −eΦ ρ2 = − 1 +
Mr
(r2 + a2 )
ρ2 = −ρ2 −
M rρ2
r 2 + a2
(A12)
Calculation of g33 :
g33
=
=
=
=
=
=
=
=
=
2
2
2
−eΦ (r2 + a2 ) sin2θ − 2 sinh(Φ) sinh2 Λ(r
+ a ) sin
θ2 2
Mr
Mr
a sin θ
(r2 + a2 ) sin2 θ − 2 2
(r2 + a2 ) sin2 θ
− 1+ 2
(r + a2 )
r + a2
ρ2
2M ra2 sin4 θ
−(r2 + a2 ) sin2 θ −
− M r sin2 θ
ρ2
ρ2 (r2 + a2 ) sin2 θ + 2M ra2 sin4 θ
−
− M r sin2 θ
ρ2
(r2 + a2 cos2 θ)(r2 + a2 ) sin2 θ + 2M ra2 sin4 θ
−
− M r sin2 θ
ρ2
(r2 + a2 − a2 sin2 θ)(r2 + a2 ) sin2 θ + 2M ra2 sin4 θ
− M r sin2 θ
−
ρ2
2
(r + a2 ) − a2 sin2 θ (r2 + a2 ) sin2 θ + 2M ra2 sin4 θ
−
− M r sin2 θ
ρ2
−(r2 + a2 )2 sin2 θ + (r2 − 2M r + a2 )a2 sin4 θ
− M r sin2 θ
ρ2
i
− sin2 θ h 2
2
2
2 2
−
∆a
sin
θ
− M r sin2 θ
r
+
a
ρ2
(A13)
Calculation of g11 :
g11 = −eΦ
r2
ρ2
,
+ a2
(A14)
we have that ∆ = r2 − 2M r − a2 can be given by:
∆ =
∆ =
2
r2 − 2M r− a2 = (r2 + a
) − 2M r
2M r
2
2
(r + a ) 1 − 2
,
(r + a2 )
(A15)
which implies that
1
2M r
1
1
−
.
=
r 2 + a2
∆
(r2 + a2 )
(A16)
Now, the component g11 is given by:
2
2M r
ρ2
ρ
Mr
Φ
g11 = −e 2
1− 2
.
=− 1+ 2
r + a2
(r + a2 ) ∆
(r + a2 )
Mr
≪ 1, the component g11 is given just in the first order:
+ a2
Mr
1− 2
(r + a2 )
M rρ2
+
.
∆(r2 + a2 )
Hence we have that
g11
=
g11
=
ρ2
∆
ρ2
−
∆
−
(A17)
r2
(A18)
(2)
B Calculation of Ricci tensor Rµν
The Ricci tensor from section VI, used to evaluated gravitational waves, is given by:
α(2)
α(2)
(2)
β(2)
α(2)
β(2)
Rµν
= ∂α Γα(2)
µν − Γµβ Γνα − ∂ν Γµα + Γαβ Γµν .
(B1)
15
Let Φ be constant. Then Γα
αν =
1
[Tr(η) + Tr(Υ)]∂ν Φ are vanished such that
2
α(2)
(2)
β(2)
Rµν
= ∂α Γα(2)
µν − Γµβ Γνα .
(B2)
Let us calculate the first term:
∂α Γα(2)
µν =
1
1
sinh(Φ) ∂α g αβ (∂µ Υβν + ∂ν Υβµ − ∂β Υµν ) + sinh(Φ) g αβ (∂α ∂µ Υβν + ∂α ∂ν Υβµ − ∂α ∂β Υµν ).
2
2
(B3)
If Φ is constant, then we have ∂α g αβ = sinh(Φ)∂α Υαβ , such that,
∂α Γα(2)
µν =
1
1
sinh2 (Φ) ∂α Υαβ (∂µ Υβν + ∂ν Υβµ − ∂β Υµν ) + sinh(Φ) g αβ (∂α ∂µ Υβν + ∂α ∂ν Υβµ − ∂α ∂β Υµν ). (B4)
2
2
Let us calculate the second term from Ricci tensor (B2):
α(2)
Γµβ Γβ(2)
να
=
=
1
1
sinh(Φ)g αγ (∂µ Υγβ + ∂β Υγµ − ∂γ Υµβ ) sinh(Φ)g βλ (∂α Υνλ + ∂ν Υαλ − ∂λ Υαν )
2
2
1
sinh2 (Φ)g αγ g βλ [(∂µ Υγβ ∂α Υνλ − ∂µ Υγβ ∂λ Υαν ) + (∂β Υγµ ∂α Υνλ + ∂γ Υµβ ∂λ Υαν )
4
+(∂β Υγµ ∂ν Υαλ − ∂γ Υµβ ∂ν Υαλ ) − (∂β Υγµ ∂λ Υαν + ∂γ Υµβ ∂α Υνλ ) + ∂µ Υγβ ∂ν Υαλ ].
The contracting indices α, β, γ and λ can be manipulated as following: α ↔ λ and β ↔ γ in the second term of
each above parentesis, that result in
α(2)
Γµβ Γβ(2)
να =
1
sinh2 (Φ)g αγ g βλ [ 2∂β Υγµ ∂α Υνλ − 2∂γ Υµβ ∂α Υνλ + ∂µ Υγβ ∂ν Υαλ ]
4
(B5)
1
sinh2 (Φ)[g αγ g βλ ∂µ Υγβ ∂ν Υαλ + 2(g αγ g βλ − g αβ g γλ )∂β Υγµ ∂α Υνλ ].
4
(B6)
or
α(2)
Γµβ Γβ(2)
να =
Finally we have the Ricci tensor given by:
(2)
Rµν
=
1
sinh2 (Φ) ∂α Υαβ (∂µ Υβν + ∂ν Υβµ − ∂β Υµν )
2
1
+ sinh(Φ) g αβ (∂α ∂µ Υβν + ∂α ∂ν Υβµ − ∂α ∂β Υµν )
2
1
− sinh2 (Φ)[g αγ g βλ ∂µ Υγβ ∂ν Υαλ + 2(g αγ g βλ − g αβ g γλ )∂β Υγµ ∂α Υνλ ].
4
(B7)
16
References
Alexander, S., Peskin, M. E. & Sheikh-Jabbari, M. M., 2006, Phys. Rev. Lett. 96, 081301,
(arXiv:hep-th/0403069)
Allemandi, G., Borowiec, A. & Francaviglia, M., 2004, Phys. Rev., D 70:103503,
Alley, C.O. & Yilmaz, H., 1999, Refutation of C.W. Misner’s claims in his article “Yilmaz Cancels Newton”, Nuovo Cimento
B 114 1087-98
Alley, C.O. & Yilmaz, H., 2000, Response to Fackerell’s Article, prepint, (gr-qc/0008040)
Alley, C.O. , 1995, The Yilmaz Theory of Gravity and its Compatibility with Quantum Theory, Annals of the New York
Academy of Sciences, 755, 464
Baker, J., et al., 2007, LISA: Probing the Universe with Gravitational Waves, LISA-LIST-RP-436 Version 1.0
Barrow, J. & Ottewill, A.C., 1983, J. Phys. A: Math. Gen., 16:2757
Bellucci, S., Capozziello, S., De Laurentis, M. & Faraoni, V., 2009, Phys. Rev. D 79, 104004
Birrell, N.D. & P.C.W. Davies, 1982, Quantum Fields in Curved Space, Cambridge University Press
Bogdanos, C., Capozziello, S., De Laurentis, M. & Nesseris, S., 2009, Massive, massless and ghost mo des of gravitational
waves from higher-order gravity, prepint,
arXiv:0911.3094v3[gr-qc]
Buonanno, A., 2004, Particle Physics And Cosmology: The Quest For Physics Beyond The Standard Model(s), eds. H. E.
Haber and E. Nelson, (World Scientific, Singapore), p. 855
Capozziello, S., 2002, Int. J. Mod. Phys., D 11:483
Capozziello, S., Carloni, S. & Troisi. A., 2003, Rec. Res. Dev. in Astron. and Astroph., 1:1
Capozziello, S., Cardone, V. F. & Troisi, A., 2005, Phys. Rev., D 71:043503
Carroll, S.M., Duvvuri, V., Trodden, M. & Turner, M., 2004, Phys. Rev., D 70:043528
Carroll, S. M., De Felice A., Duvvuri, V., Easson, D.A., Trodden, M. & Turner, M., 2005, Phys. Rev., D 71:063513
Choi,K., Hwang, J. C. & Hwang, K. W., 2000, Phys. Rev. D 61, 084026 [arXiv:hep-ph/9907244]
Clapp, R.E., 1973, Preliminary Quasar Model Based on the Yilmaz Exponential Metric, Phys.Rev.D 7 345-55
Fackerell, E.D., 1996, Remarks on the Yilmaz and Alley Papers, Proceedings of the First Australian Conference on General
Relativity and Gravitation, ed D.L. Wiltshire 117-31
Faraoni, Valerio, 2008, f (R) gravity: Successes and Challenges, preprint, arXiv 0810.2602 v1 [gr-qc]
Flanagan, E. E., 2003 Class. Quant. Grav., 21:417
Glampedakis, K. & Babak, S., 2006 Mapping spacetimes with LISA: inspiral of a test body in a ‘quasi-Kerr’ field, Class.
Quantum Grav. 23, 4167-88
Hawking, S.W. & Ellis, G.F.R., 1973, The Large Scale Structure of Space-time, Cambridge University Press
Hulse, R.A.& Taylor, J.H., 1975, Discovery a Pulsar in a Binary System, Astrophys.J., 195, L51-L53
Ibison, M., 2006, Cosmological Test of the Yilmaz Theory of Gravity, Class. Quantum Grav. 23 577-89
Kahniashvili, T., Gogoberidze, G. & Ratra, B., 2005 Polarized Cosmological Gravitational Waves from Primordial Helical
Turbulence, Phys.Rev.Lett., 95, 151301, (arXiv:astro-ph/0505628)
Koivisto, T., 2006, Phys. Rev., D 73:083517
Landau, L.D. & Lifschitz, E.M., 1984, Classical Theory of Fields, Oxford: Pergamon Press
Lue, A., Wang, L. M. & Kamionkowski, M., 1999, Phys. Rev. Lett. 83, 1506, (arXiv:astro-ph/9812088)
Misner, C.W., 1999, Yilmaz Cancels Newton, Nuovo Cimento B 114 1079-85, (arXiv:gr-qc/9504050)
NASA, 2010, http://lisa.nasa.gov/index.html
S. Nojiri and S. D. Odintsov., 2007, Int. J. Geom. Meth. Mod. Phys., 4:115
Nojiri S. & Odintsov, S. D., 2003, Phys. Lett., B 576:5
Nojiri S. & Odintsov, S. D., 2003, Phys. Rev., D 68:123512
Robertson, S.L., 1999a, Bigger Bursts from Merging Neutron Stars, Astro.J.Lett. 517 117-9
Robertson, S.L., 1999b, X-ray Novae, Event Horizons, and the Exponential Metric, Astro.J. 515 365-80
Rosen, N., 1974, A Theory of Gravitation, Annals of Physics 84 455-73
Saito, S., Ichiki, K. & Taruya, A., 2007, Probing polarization states of primordial gravitational waves with CMB anisotropies,
preprint, arXiv:astro-ph/0705.3701
Satoh, M., Kanno, S. & Soda, J., 2007, Circular Polarization of Primordial Gravitational Waves in String-inspired Inflationary Cosmology, prepint, (arXiv:astro-ph/0706.3585)
Schutz, B.F., 1990, A First Course in General Relativity, Cambridge University Press
Sotiriou, T. P., Faraoni, V., 2010, f (R) theories of gravity Rev. Mod. Phys. 82, 451497, arXiv 0805.1726v3[gr-qc]
Taylor, J.H. & Weisberg, J.M., 1982 A New Test of General Relativity: Gravitational Radiation and the Binary Pulsar PSR
1913+16, Astrophys.J., 253, 908-20
Wald, R.M., 1984, General Relativity, The University of Chicago Press
White, F., 2006, A quasi-Kinnersley tetrad for the quasi-Kerr metric, preprint arXiv:gr-qc/0601094
Wyss, W. & Misner, C.W., 1999 Comments on the Yilmaz Theory of Gravity, Nuovo Cimento B 114 1099-101
Yilmaz, H., 1958, New approach to General Relativity,
Phys.Rev.111 1417-26
Yilmaz, H., 1973, Correspondence Paradox in General Relativity, Lettere al Nuovo Cimento 7 337-40
17
Yilmaz, H., 1976, Physical Foundations of the New Theory of Gravitation, Annals of Physics 101 413-32
Yilmaz, H., 1977, Hughes-Drever Experiment and Theories of Gravitation, Lettere al Nuovo Cimento 20 681-5
Yilmaz, H., 1982, Relativity and Quantum Mechanics,
Int.J.Theory. Phys. 21 871-902
Yilmaz, H., 1992, Towards a Field Theory of Gravitational, Il Nuovo Cimento 107 B 941-60
Yilmaz, H., 1994, Did the Apple Fall?, Frontiers of Fundamental Physics, eds M. Barone and F. Selleri (New York: Plenum)
116-24
Yilmaz,H., 1995, Gravity and Quantum Field Theory, a Modern Synthesis, Annals of the New York Academy of Sciences,
755, 476
Yilmaz, H., 1997, Towards a Comprehensible Physical Theory: Gravity and Quantum Mechanics, The Present Status of the
Quantum Theory of Light, eds S Jeffers and et al. (Dordrecht: Kluwer Academic) 503-25
This manuscript was prepared with the AAS LATEX macros v5.2.
View publication stats