SSC/Virial/PolytropesEmbedded/FirstEffortAgain/Pt2

From JETohlineWiki
Jump to navigation Jump to search

Virial Equilibrium of Adiabatic Spheres (1st Effort)[edit]


Part I:   Isolated Configurations
 

Part II:  Configurations Embedded in an External Medium
 

Nonrotating Adiabatic Configuration Embedded in an External Medium[edit]

For a nonrotating configuration (C=J=0) that is embedded in, and is influenced by the pressure Pe of, an external medium, the statement of virial equilibrium is,

3Bχeq33γg3Aχeq13Dχeq3=0.


Solution Expressed in Terms of K and M (Whitworth's 1981 Relation)[edit]

This is precisely the same condition that derives from setting equation (3) to zero in Whitworth's (1981, MNRAS, 195, 967) discussion of the Global Gravitational Stability for One-dimensional Polytropes. The overlap with Whitworth's narative is clearer after introducing the algebraic expressions for the coefficients A, B, and D, to obtain,

4π(PePnorm)χeq3

=

3(4π3)1γg(MlimitMtot)γg𝔣A𝔣Mγgχeq33γg35(MlimitMtot)2𝔣W𝔣M2χeq1;

dividing the equation through by (4πχeq3/Pnorm),

Pe

=

Pnorm[(34π)γg(MlimitMtot)γg𝔣A𝔣Mγgχeq3γg(320π)(MlimitMtot)2𝔣W𝔣M2χeq4]

 

=

PnormRnorm4[(34πReq3)γg(MlimitMtot)γg𝔣A𝔣MγgRnorm3γg4(320πReq4)(MlimitMtot)2𝔣W𝔣M2];

and inserting expressions for the parameter normalizations as defined in our accompanying introductory discussion to obtain,

Pe

=

GMtot2[(34πReq3)γg(MlimitMtot)γg𝔣A𝔣MγgKMtotγg2G(320πReq4)(MlimitMtot)2𝔣W𝔣M2]

 

=

K(3Mlimit4πReq3)γg𝔣A𝔣Mγg(3GMlimit220πReq4)𝔣W𝔣M2.

If the structural form factors are set equal to unity, this exactly matches equation (5) of Whitworth, which reads:

Whitworth (1981, MNRAS, 195, 967)
Whitworth (1981, MNRAS, 195, 967)

Notice that, when Pe0, this expression reduces to the solution we obtained for an isolated polytrope, expressed in terms of K and Mlimit (see the left-hand column of our table titled "Two Points of View").

Solution Expressed in Terms of M and Central Pressure[edit]

Beginning again with the relevant statement of virial equilibirum, namely,

A=Bχeq43γgDχeq4,

but adopting the alternate expression for the coefficient, B, given above, that is,

B=4π3[(PcPnorm)χ3γ]eq𝔣A,

we can write,

15[(MlimitMtot)1𝔣M]2𝔣W

=

4π3[(PcPnorm)χ3γ]eq𝔣Aχeq43γg(4π3)PePnormχeq4

320π[(MlimitMtot)1𝔣M]2𝔣W

=

[(PcPnorm)𝔣APePnorm]χeq4

 

=

[𝔣APcPe]Req4GMtot2

320π(GMlimit2Req4)𝔣W𝔣M2

=

𝔣APcPe.

Again notice that, when Pe0, this expression reduces to the solution we obtained for an isolated polytrope, but this time expressed in terms of Pc and Mlimit (see the right-hand column of our table titled "Two Points of View").

Contrast with Detailed Force-Balanced Solution[edit]

As has just been demonstrated, the virial theorem provides a mathematical expression that allows us to relate the equilibrium radius of a configuration to the applied external pressure, once the configuration's mass and either its specific entropy or central pressure are specified. In contrast to this, as has been discussed in detail in another chapter, Horedt (1970), Whitworth (1981) and Stahler (1983) have each derived separate analytic expressions for Req and Pe — given in terms of the Lane-Emden function, Θ, and its radial derivative — without demonstrating how the equilibrium radius and external pressure directly relate to one another. That is to say, solution of the detailed force-balanced equations provides a pair of equilibrium expressions that are parametrically related to one another through the Lane-Emden function. For example — see our related discussion for more details — Horedt derives the following set of parametric equations relating the configuration's dimensionless radius, ra, to a specified dimensionless bounding pressure, pa:

raReqRHoredt

=

ξ~(ξ~2θ~)(1n)/(n3),

paPePHoredt

=

θ~nn+1(ξ~2θ~)2(n+1)/(n3),

where,

RHoredt

=

[4π(n+1)n(GK)nMlimitn1]1/(n3),

PHoredt

=

K4n/(n3)[(n+1)34πG3Mlimit2](n+1)/(n3).

It is important to appreciate that, in the expressions for ra and pa, the tilde indicates that the Lane-Emden function and its derivative are to be evaluated, not at the radial coordinate, ξ1, that is traditionally associated with the "first zero" of the Lane-Emden function and therefore with the surface of the isolated polytrope, but at the radial coordinate, ξ~, where the internal pressure of the isolated polytrope equals Pe and at which the embedded polytrope is to be truncated. The coordinate, ξ~, therefore identifies the surface of the embedded — or, pressure-truncated — polytrope. We also have taken the liberty of attaching the subscript "limit" to M in both defining relations because it is clear that Hoerdt intended for the normalization mass to be the mass of the pressure-truncated object, not the mass of the associated isolated (and untruncated) polytrope. In anticipation of further derivations, below, we note here the ratio of Hoerdt's normalization parameters to ours, assuming γ=(n+1)/n:

(RHoerdtRnorm)n3

=

[4π(n+1)n(GK)nMlimitn1][(KG)nMtot1n]

 

=

4π(n+1)n(MlimitMtot)n1,

(PHoerdtPnorm)n3

=

K4n[(n+1)34πG3Mlimit2]n+1[G3(n+1)Mtot2(n+1)K4n]

 

=

[(n+1)34π(MlimitMtot)2]n+1.

Next, we demonstrate that this pair of parametric relations satisfies the virial theorem and, in so doing, demonstrate how ra and pa may be directly related to each other. Given that Hoerdt's chosen normalization radius and normalization pressure are defined in terms of K and Mlimit, we begin with the virial theorem derived above in terms of K and Mlimit, setting γg=(n+1)/n.

Pe

=

K(3Mlimit4πReq3)(n+1)/n𝔣A𝔣M(n+1)/n(3GMlimit220πReq4)𝔣W𝔣M2.

After setting Req=raRHoredt, a bit of algebraic manipulation shows that the first term on the right-hand side of the virial equilibrium expression becomes,

K(3Mlimit4πReq3)(n+1)/n𝔣A𝔣M(n+1)/n

=

ra3(n+1)/n𝔣A[(3𝔣M)n3(n+1)3n(4π)n](n+1)/[n(n3)][K4nG3(n+1)Mlimit2(n+1)]1/(n3),

while the second term on the right-hand side becomes,

(3GMlimit220πReq4)𝔣W𝔣M2

=

35𝔣W𝔣M2ra4(4π)(n+1)/(n3)(n+1)4n/(n3)[K4nG3(n+1)Mlimit2(n+1)]1/(n3).

But, using Horedt's expression for Pe, the left-hand side of the virial equilibrium equation becomes,

Pe=paPHoerdt

=

pa(4π)(n+1)/(n3)(n+1)3(n+1)/(n3)[K4nG3(n+1)Mlimit2(n+1)]1/(n3).

Hence, the statement of virial equilibrium is,

pa

=

{ra3(n+1)/n𝔣A[(3𝔣M)n3(n+1)3n(4π)n](n+1)/[n(n3)]

 

 

35𝔣W𝔣M2ra4(4π)(n+1)/(n3)(n+1)4n/(n3)}(4π)(n+1)/(n3)(n+1)3(n+1)/(n3)

 

=

𝔣A(3𝔣Mra3)(n+1)/n3(n+1)5𝔣W𝔣M2ra4;

or, multiplying through by ra4 and rearranging terms,

𝔣A(3𝔣M)(n+1)/nra(n3)/npara4

=

3(n+1)5𝔣W𝔣M2.

Now, Hoerdt has given analytic expressions for ra and pa in terms of the Lane-Emden function and its first derivative. The question is, what should the expressions for our structural form factors be in order for this virial expression to hold true for all pressure-truncated polytropic structures? As has been summarized above, in the case of an isolated polytrope, whose surface is located at ξ1 and whose global properties are defined by evaluation of the Lane-Emden function at ξ1, we know that (see the above summary),

Structural Form Factors for Isolated Polytropes

𝔣M

=

[3Θ'ξ]ξ1

𝔣W

=

3255n[Θ'ξ]ξ12

𝔣A

=

3(n+1)(5n)[Θ']ξ12

These same expressions may or may not work for pressure-truncated polytropes, even if the evaluation radius is shifted from ξ1 to ξ~. Let's see …

Inserting Hoerdt's expressions for ra and pa into the viral equilibrium expression, we have,

3(n+1)5𝔣W𝔣M2

=

𝔣A(3𝔣M)(n+1)/n[ξ~(ξ~2θ~)(1n)/(n3)](n3)/nθ~nn+1(ξ~2θ~)2(n+1)/(n3)[ξ~(ξ~2θ~)(1n)/(n3)]4

 

=

𝔣A(3𝔣M)(n+1)/n[ξ~(ξ~2θ~)(1n)/(n3)](n3)/nθ~nn+1ξ~4[(ξ~2θ~)2(n+1)+4(1n)]1/(n3)

 

=

𝔣A(3𝔣M)(n+1)/n[ξ~(n3)(ξ~2θ~)(1n)]1/nθ~nn+1ξ~4[(ξ~2θ~)2]

 

=

𝔣A(3𝔣M)(n+1)/nξ~(n+1)/n(θ~)(1n)/nθ~nn+1(θ~)2.

Assuming that the structural form factor, 𝔣M, has the same functional expression as in the case of isolated polytropes (but evaluated at ξ~ instead of at ξ1), the virial relation further reduces to the form,

θ~nn+1(θ~)2

=

𝔣A(ξ~θ~)(n+1)/nξ~(n+1)/n(θ~)(1n)/n[(n+1)35]ξ~2(θ~')2𝔣W

 

=

𝔣A(θ~)2[(n+1)35]ξ~2(θ~')2𝔣W

θ~n+1

=

𝔣A[(n+1)35]ξ~2𝔣W

𝔣Aθ~n+1𝔣W

=

[(n+1)35]ξ~2.

While this does not give us individual expressions for the form factors, 𝔣W and 𝔣W, the expression derived for the ratio of the form factors makes sense because the term that has been subtracted from 𝔣A in the numerator on the lefthand side, that is, θ~n+1, naturally goes to zero in the limit of ξ~ξ1, producing the correct expression for the ratio, 𝔣A/𝔣W, in isolated polytropes. In summary, then, we have,

Structural Form Factors for Pressure-Truncated Polytropes

𝔣~M

=

[3Θ'ξ]ξ~

𝔣~Aθ~n+1𝔣~W

=

[(n+1)35]ξ~2

Notice that, in an effort to differentiate them from their counterparts developed earlier for "isolated" polytropes, we have affixed a tilde to each of these three form-factors, 𝔣i.

Renormalization[edit]

Grunt Work[edit]

Returning to the dimensionless form of the virial expression and multiplying through by [χeq/(3D)], we obtain,

χeq4=BDχeq43γgAD,

or, after plugging in definitions of the coefficients, A, B, and D, and rewriting χeq explicitly as Req/Rnorm,

(ReqRnorm)4

=

(34π)γg(MlimitMtot)γg(PePnorm)1𝔣~A𝔣~Mγg(ReqRnorm)43γg320π(MlimitMtot)2(PePnorm)1𝔣~W𝔣~M2.

This relation can be written in a more physically concise form, as follows. First, normalize Pe to a new pressure scale — call it Pad — and multiply through by (Rnorm/Rad)4 in order to normalizing Req to a new length scale,Rad:

(ReqRad)4

=

(34π)γg(MlimitMtot)γg(PnormPad)(PePad)1𝔣~A𝔣~Mγg(RnormRad)3γg(ReqRad)43γg

 

 

320π(MlimitMtot)2(PnormPad)𝔣~W𝔣~M2(PePad)1(RnormRad)4,

or,

Xad4

=

(34π)γg(MlimitMtot)γg(PnormPad)(RnormRad)3γg𝔣~A𝔣~MγgXad43γgΠad320π(MlimitMtot)2(PnormPad)(RnormRad)4𝔣~W𝔣~M21Πad,

where,

Xad

ReqRad,

Πad

PePad.

By demanding that the leading coefficients of both terms on the right-hand-side of the expression are simultaneously unity — that is, by demanding that,

(34π)γg(MlimitMtot)γg(PnormPad)(RnormRad)3γg𝔣~A𝔣~Mγg

=

1,

and,

320π(MlimitMtot)2(PnormPad)(RnormRad)4𝔣~W𝔣~M2

=

1,

we obtain the expressions for Rad/Rnorm and Pad/Pnorm as shown in the following table.

Renormalization for Adiabatic (ad) Systems

RadRnorm

[15(4π3)γg1(MlimitMtot)2γg𝔣~W𝔣~A𝔣~M2γg]1/(43γg)

 

=

[15n(4π3)(MlimitMtot)n1𝔣~Wn𝔣~An𝔣~Mn1]1/(n3)

PadPnorm

[𝔣~A4(3534π𝔣~M2𝔣~W3)γg(MlimitMtot)2γg]1/(43γ)

 

=

[𝔣~A4n(3534π)n+1(𝔣~M2𝔣~w3)n+1(MlimitMtot)2(n+1)]1/(n3)

Using these new normalizations, we arrive at the desired, concise virial equilibrium relation, namely,

Πad=Xad3γgXad4

                       

Πad=Xad3(n+1)/nXad4

or,

Xad43γgΠadXad4=1

                       

Xad(n3)/nΠadXad4=1.

In Terms of Free-Energy Coefficients[edit]

Referring back to relations between our free-energy coefficients, as presented earlier, we note that,

RadRnorm

=

(𝒜)n/(n3),

PadPnorm

=

34π[4n𝒜3(n+1)]1/(n3).

Hence, we can write,

Xad

=

χeq(𝒜)n/(n3),

Πad

=

𝒟[𝒜3(n+1)4n]1/(n3).


In Terms of Horedt's Equilibrium Parameters[edit]

For later use it is also worth developing expressions for both Xad and Πad that are in terms of our structural form factors and Horedt's two dimensionless functions, ra and pa. (Adopting a unified notation, we will set γg(n+1)/n.)

XadReqRad

=

ReqRHoerdtRHoerdtRnormRnormRad

 

=

ra[4π(n+1)n(MlimitMtot)n1]1/(n3)[5n(34π)(MlimitMtot)1n𝔣~An𝔣~Wn𝔣~M1n]1/(n3)

Xadn3

=

3ran3[5(n+1)𝔣~A𝔣~W]n𝔣~Mn1.

And,

ΠadPePad

=

PePHoerdtPHoerdtPnormPnormPad

 

=

pa[(n+1)34π(MlimitMtot)2](n+1)/(n3)[𝔣~A4n(4π353𝔣~W3𝔣~M2)(n+1)(MlimitMtot)2(n+1)]1/(n3)

Πadn3

=

pan3𝔣~A4n[(n+1)3353(𝔣~W3𝔣~M2)]n+1.

Plugging in the expressions for ra and pa, as reprinted, for example, above, along with our deduced expressions for 𝔣M and 𝔣A (in terms of 𝔣W), these two relations become:

Xadn3

=

3[ξ~(ξ~2θ~)(1n)/(n3)]n3[5(n+1)𝔣~A𝔣~W]n[3θ~'ξ~]n1

 

=

ξ~[(n3)+2(1n)(n1)][35(n+1)𝔣~A𝔣~W]n

 

=

[35(n+1)ξ~2𝔣~A𝔣~W]n;

and,

Πadn3

=

[θ~nn+1(ξ~2θ~)2(n+1)/(n3)]n3𝔣~A4n[(n+1)35𝔣~W]3(n+1)(3𝔣~M)2(n+1)

 

=

θ~n(n+1)(n3)(ξ~2θ~)2(n+1)𝔣~A4nξ~6(n+1)[(n+1)ξ~235𝔣~W]3(n+1)ξ~2(n+1)(θ~')2(n+1)

 

=

θ~n(n+1)(n3)𝔣~A4n[(n+1)ξ~235𝔣~W]3(n+1).

Summary[edit]

If we define,

aad

(n+1)ξ~235𝔣W,     and,

bad

θ~n+1,

in which case the relationship between 𝔣A and 𝔣A for pressure-truncated polytropes can be rewritten as,

𝔣A

=

aad+bad.

In addition, the expressions for the dimensionless equilibrium radius and the dimensionless external pressure, as just derived, may be written as, respectively,

Xad

=

[1+badaad]n/(n3),     and,

Πad

=

bad[aad3(n+1)(aad+bad)4n]1/(n3)=badaad[1+badaad]4n/(n3).

Using these expressions, it is easy to demonstrate that the virial equilibrium relation is satisfied, namely,

Πad

=

Xad3(n+1)/nXad4.

P-V Diagram[edit]

For an arbitrary value of the adiabatic exponent, γg, it isn't possible to invert this virial relation to obtain an analytic expression for χad as a function of Πad. But, as written, the virial relation dictates the behavior of Πad as a function of Xad. Figure 4 displays this Πad(χad) behavior for a number of different values of γg.

Figure 4: Equilibrium Adiabatic P-V Diagram

The curves shown here, on the right, trace out the function,

Πad=(Xad43γg1)/Xad4,

for six different values of γg — specifically, for 2,5/3,7/5,6/5,1,2/3, as labeled — and show the dimensionless external pressure, Πad, that is required to construct a nonrotating, self-gravitating, pressure-truncated adiabatic sphere with an equilibrium radius Xad. The solid red curve identifies the behavior of an isothermal (γg=1) system. The mathematical solution becomes unphysical wherever the pressure becomes negative.

Equilibrium Adiabatic P-R Diagram
Equilibrium Adiabatic P-R Diagram


For physically relevant solutions, both Xad and Πad must be nonnegative. Hence, as is illustrated by the curves in Figure 4, the physically allowable range of equilibrium radii is,

1Xadforγg<4/3;

0<Xad1forγg>4/3.

Each of the Πad(Xad) curves drawn in Figure 4 exhibits an extremum. In each case this extremum occurs at a configuration radius, Xextreme, given by,

ΠadXad=0,

that is, where,

43γgXad43γg=0Xextreme=[43γg]1/(43γg).

For each value of γg, the corresponding dimensionless pressure is,

Πextreme=(43γ1)[3γg4]4/(43γg).

In terms of the polytropic index, the equivalent limiting expressions are,

Xextreme=[4n3(n+1)]n/(n3)

          and          

Πextreme(n3)=(4n)4n(n3)n3[3(n+1)]3(n+1).

(In a separate, related discussion of the free-energy function, we demonstrate that this "extremum" also serves as a dividing line between dynamically stable and unstable models along a given curve.)


In examining the group of plotted curves, notice that, for γg>4/3, an equilibrium configuration with a positive radius can be constructed for all physically realistic — that is, for all positive — values of Πad. Also, consistent with the behavior of the curves shown in Figure 4, the extremum arises in the regime of physically relevant — i.e., positive — pressures only for values of γg<4/3; and in each case it represents a maximum limiting pressure.



Material that appears after this point in our presentation is under development and therefore
may contain incorrect mathematical equations and/or physical misinterpretations.
|   Go Home   |


Maximum Mass[edit]

n = 5 Polytropic[edit]

When γa=6/5 — which corresponds to an n=5 polytropic configuration — we obtain,

Πmax=Πad|extreme(γg=6/5)=(318210510),

which corresponds to a maximum mass for pressure-bounded n=5 polytropic configurations of,

Mmax=Πmax1/2(35322π)1/2(cs¯8G3Pe)1/2=(31921257π)1/2(cs¯8G3Pe)1/2.

This result can be compared to other determinations of the Bonnor-Ebert mass limit.

See Also[edit]

Tiled Menu

Appendices: | VisTrailsEquations | VisTrailsVariables | References | Ramblings | VisTrailsImages | myphys.lsu | ADS |