We will present, in the case of a spherically symmetric distribution of
gaseous fluid satisfying a polytropic equation of state, the complete
static solution, over all the three-dimensional space, of the Einstein
field equations of General Relativity. In this work we will use the
time-like signature , following [#!DiracGravity!#]. We will
start from the same differential system already described
in [#!LiquidShells!#], which we will succinctly review here. Just as
in [#!LiquidShells!#], the solution will be given in terms of the
coefficients of the metric, for an invariant interval given in terms of
the Schwarzschild coordinates
by
where and
are two positive functions of
only
. As was shown in detail in [#!LiquidShells!#], under these
conditions the independent components of the Einstein field equations and
the Bianchi consistency condition are equivalent to the set of three
first-order differential equations
where the primes indicate differentiation with respect to ,
is the energy density of the matter,
is its isotropic pressure, and
where we have the constant
, in which
is the
universal gravitational constant and
is the speed of light. Note that
all the derivatives are written as what we will call homogeneous
derivatives, that is, the product of the derivative by a single power of
. In our case here the matter distribution will be characterized by
four parameters, the two parameters defining the polytropic equation of
state, the total asymptotic gravitational mass
, associated to the
Schwarzschild radius
, and a parameter associated to the value of
the energy density
at its point of maximum. We will assume that
the gas satisfies the polytropic equation of state
over the whole three-dimensional space, involving a positive real constant
and the integer or half-integer
, which we assume to be strictly
larger that one. In principle
could be any real number larger than
one, and we assume that it is either an integer of a half integer just for
simplicity, since this seems to cover all cases of interest. At this point
we will introduce an auxiliary function, also just for simplicity, since
it will appear repeatedly in all that follows,
which is a dimensionless function, so that the polytropic equation of state reduces to
Note that from its definition we immediately have for the derivative of
,
Given this, our system of differential equations shown in Equations (2)-(4) can now be written as
still in terms of homogeneous derivatives. Our problem is therefore that
of finding three functions, ,
and
, that
solve these equations and that satisfy the correct boundary conditions at
asymptotic radial infinity. We will start our analysis by partially
solving some of the equations by analytic means, in order to write all
relevant quantities in terms of a single function. In order to do this we
first change variables from the dimensionless function
to the
equally dimensionless function
, which is defined to be such
that
which then implies that we have for the corresponding homogeneous derivatives
The function is analogous to the function
found in
Equation (2.9) of the paper by Tooper [#!tooper!#], but it is used here
in a completely different context. Note that the asymptotic boundary
condition on
, that it must behave as the exterior
Schwarzschild solution for sufficiently large
, translates here as
under that same condition. Substituting these expressions in
the component field equation shown in Equation (9) a very simple
relation giving the derivative of
in terms of
results,
Therefore, the energy density is given in terms of the
derivative of
, and wherever
, characterizing a
region where there is a vacuum, we have that
is a constant.
Since
and
are both given in terms of
, and since
is given in terms of
, it follows at this point
that, given a function
, the functions
,
,
and
are all determined. The only function that has yet to be
determined in terms of
is
. We can obtain
in
terms of
, and therefore of
, using the consistency
condition in Equation (11), which with the use of
Equation (8) can be written as
This can now be integrated from to
, and recalling that we have
that
, because for
we have the boundary condition
, we get
which determines in terms of
, up to the integration
constant
, that is to be obtained from the asymptotic boundary
conditions at radial infinity, which in the case of
is simply
. Therefore, given
, this determines
in
terms of it, through
. Note, for future use, that the fact that we
also have that
, because for
we have the boundary
condition
, implies that we always have that
. We see therefore that the determination of the
function
leads with no further difficulty to the determination
of all the functions that describe both the matter and the geometry of the
system,
The free parameters of the system are ,
and
, all of which
describe the nature and state of the matter, and the value of
at its point of maximum, which can also be seen as related to the matter,
since it determines the general scale of the matter energy density, as can
be seen from Equation (17).
Both for the subsequent analysis and for the numerical approach, it is
convenient to transform variables at this point, in order to write
everything in terms of dimensionless variables and functions. In order to
do this we must now introduce an arbitrary radial reference position
. For now the value of this parameter remains completely
arbitrary, other than that it must be strictly positive, and has no
particular physical meaning. It is only a mathematical device that allows
us to define a dimensionless radial variable and a dimensionless parameter
associated to the mass
by
as well as to define the dimensionless function of , to assume the
role of
,
Note that the asymptotic condition that for sufficiently
large
translates here as the condition that
for
sufficiently large
. It is important to observe that under the change
of variables from
to
the homogeneous derivatives transform in a
very simple way, for example in the case of
,
We will also introduce at this point a notation for the derivative of
, which will be useful in order to deal with the second-order
differential equation for
that we will arrive at,
where in this case the prime denotes differentiation with respect to
. We adopt the convention that whenever there is a prime indicating a
derivative, it is to be taken with respect to the explicitly indicated
variable of the function. From now on the problem will be formulated in
terms of the two functions
and
, that will be
treated as independent variables. In terms of these new variables we have
for the system of differential equations shown in
Equations (9)-(11), where from now on the auxiliary
function
will be written as
,
where the primes now indicate differentiation with respect to . We
therefore have for all the relevant quantities written in terms of
,
and
,
where
is a dimensionless
constant. We see therefore that the determination of
and
thus of
leads to the complete solution of the problem. We have
therefore reduced the solution of the problem to the determination of the
single function
. In order to determine
, we
obtain an ordinary differential equation for it by eliminating
from Equations (28) and (29). Equation (28) can
be written as
and Equation (29) can be written as
so that equating the two right-hand sides we get
Since is the derivative of
, this can be
interpreted either as a second-order ordinary differential equation for
, or as one of a pair of first-order coupled ordinary
differential equations determining
and
, the other
equation being simply
This second interpretation is the one we will adopt here. This pair of
first-order ordinary differential equations can be used for the numerical
integration of this differential system, as we will in fact do later on,
in Section 4. However, before we plunge into the numerical
approach, let us first show that is in fact a very simple
function, as is
, and that it has some properties which are,
perhaps, somewhat unexpected.