This action might not be possible to undo. Are you sure you want to continue?
In
Stellar Astrophysics
by
George W. Collins, II
copyright 2003
To the kindness, wisdom, humanity, and memory of
D. Nelson Limber
and
Uco van Wijk
ii
Table of Contents
Preface to the Pachart Edition
Preface to the WEB Edition
Introduction
1. A brief historical review
2. The nature of the theorem
3. The scope and structure of the book
References
Chapter I Development of the Virial Theorem
1. The basic equations of structure
2. The classical derivation of the Virial Theorem
3. Velocity dependent forces and the Virial Theorem
4. ContinuumField representation of the Virial Theorem
5. The Ergodic Theorem and the Virial Theorem
6. Summary
Notes to Chapter 1
References
Chapter II Contemporary Aspects of the Virial
Theorem
1. The Tensor Virial Theorem
2. Higher Order Virial Equations
3. Special Relativity and the Virial Theorem
4. General Relativity and the Virial Theorem
5. Complications: Magnetic Fields, Internal Energy, and
Rotation
6. Summary
Notes to Chapter 2
References
v
vi
1
1
3
4
5
6
6
8
11
11
14
17
18
19
20
20
22
25
27
33
38
41
45
iii
Chapter III The Variational Form of the Virial Theorem
1. Variations, Perturbations, and their implications for
The Virial Theorem
2. Radial pulsations for selfgravitating systems: Stars
3. The influence of magnetic and rotational energy upon
a pulsating system
4. Variational form of the surface terms
5. The Virial Theorem and stability
6. Summary
Notes to Chapter 3
References
Chapter IV Some Applications of the Virial Theorem
1. Pulsational stability of White Dwarfs
2. The Influence of Rotation and Magnetic Fields on the
White Dwarf Gravitational Instability
3. Stability of Neutron Stars
4. Additional Topics and Final Thoughts
Notes to Chapter 4
References
Symbol Definitions and First Usage
Index
48
48
49
53
60
63
71
72
78
80
80
86
90
93
98
100
102
107
iv
Preface to the Pachart Edition
As Fred Hoyle has observed, most readers assume a preface is written first and thus
contains the author’s hopes and aspirations. In reality most prefaces are written after the fact and
contain the authors' views of his accomplishments. So it is in this case and I am forced to observe
that my own perception of the subject has deepened and sharpened the considerable respect I
have always had for the virial theorem. A corollary aspect of this expanded perspective is an
awareness of how much remains to be done. Thus by no means can I claim to have prepared here
a complete and exhaustive discussion of the virial theorem; rather this effort should be viewed as
a guided introduction, punctuated by a few examples. I can only hope that the reader will
proceed with the attitude that this constitutes not an end in itself, but an establishment of a point
of view that is useful in comprehending some of the aspects of the universe.
A second traditional role of a preface is to provide a vehicle for acknowledging the help
and assistance the author received in the preparation of his work. In addition to the customary
accolades for proof reading which in this instance go to George Sonneborn and Dr. John
Faulkner, and manuscript preparation by Mrs. Delores Chambers, I feel happily compelled to
heap praise upon the publisher.
It is not generally appreciated that there are only a few thousand astronomers in the
United States and perhaps twice that number in the entire world. Only a small fraction of these
could be expected to have an interest in such an apparently specialized subject. Thus the market
for such a work compared to a similar effort in another domain of physical sciences such as
Physics, Chemistry or Geology is miniscule. This situation has thereby forced virtually all
contemporary thought in astrophysics into the various journals, which for economic reasons
similar to those facing the wouldbe book publisher; find little room for contemplative or
reflective thought. So it is a considerable surprise and great pleasure to find a publisher willing to
put up with such problems and produce works of this type for the small but important audience
that has need of them.
Lastly I would like to thank my family for trying to understand why anyone would write
a book that won't make any money.
George W. Collins, II
The Ohio State University
November 15, 1977
v
Preface to the Internet Edition
Not only might one comfortably ask “why one would write a book on this subject?”, but
one might further wonder why anyone would resurrect it from the past. My reasons revolve
around the original reasons for writing the monograph in the first place. I have always regarded
the virial theorem as extremely powerful in understanding problems of stellar astrophysics, but I
have also found it to be poorly understood by many who study the subject. While it is obvious
that the theorem has not changed in the quartercentury that has passed since I first wrote the
monograph, pressures on curricula have reduced the exposure of students to the theorem even
below that of the mid 20
th
century. So it does not seem unreasonable that I make it available to
any who might learn from it. I would only ask that should readers find it helpful in their research,
that they make the proper attribution should they employ its contents.
The original monograph was published by Pachart Press and had its origin in a time
before modern word processors and so lacked many of the cosmetic niceties that can currently be
generated. The equations were more difficult to read and sections difficult to emphasize. The
format I chose then may seem a little archaic by today’s standards and the referencing methods
rather different from contemporary journals. However, I have elected to stay close to the original
style simply as a matter of choice. Because some of the derivations were complicated and
tedious, I elected to defer them to a “notes” section at the end of each chapter. I have kept those
notes in this edition, but enlarged the type font so that they may be more easily followed.
However, confusion arose in the main text between superscripts referring to references and
entries in the notes sections. I have attempted to reduce that confusion by using italicized
superscripts for referrals to the notes section. I have also added some references that appeared
after the manuscript was originally prepared. These additions are in noway meant to be
exhaustive or complete. It is hoped that they are helpful. I have also corrected numerous
typographical errors that survived in the original monograph, but again, the job is likely to be
incomplete. Finally, the index was converted from the Pachart Edition by means of a page
comparison table. Since such a table has an inherent one page error, the entries in the index could
be off by a page. However, that should be close enough for the reader to find the appropriate
reference.
I have elected to keep the original notation even though the Einstein summation
convention has become common place and the vectordyadic representation is slipping from
common use. The reason is partly sentimental and largely not wishing to invest the time required
to convert the equations. For similar reasons I have decided not to rewrite the text even though I
suspect it could be more clearly rendered. To the extent corrections have failed to be made or
confusing text remains the fault is solely mine
iv
Lastly, I would like to thank John Martin and Charlie Knox who helped me through the
vagaries of the soft and hardware necessary to reclaim the work from the original. Continuing
thanks is due A.G. Pacholczyk for permitting the use of the old Copyright to allow the work to
appear on the Internet.
George W. Collins, II
April 9, 2003
vii
Copyright 2003
Introduction
1. A Brief Historical Review
Although most students of physics will recognize the name of the viria1 theorem, few can
state it correcet1y and even fewer appreciate its power. This is largely the result of its diverse
development and somewhat obscure origin, for the viria1 theorem did not spring full blown in its
present form but rather evolved from the studies of the kinetic theory of gases. One of the lasting
achievements of 19th century physics was the development of a comprehensive theory of the
behavior of confined gases which resulted in what is now known as thermodynamics and
statistical mechanics. A brief, but impressive, account of this historical development can be
found in "The Dynamical Theory of Gases" by Sir James Jeans
1
and in order to place the viria1
theorem in its proper prospective, it is worth recounting some of that history.
Largely inspired by the work of Carnot on heat engines, R. J. E. C1aussius began a long
study of the mechanical nature of heat in 1851
2
. This study led him through twenty years to the
formulation of what we can now see to be the earliest clear presentation of the viria1 theorem.
On June 13, 1870, Claussius delivered a lecture before the Association for Natural and Medical
Sciences of the Lower Rhine "On a Mechanical Theorem Applicable to Heat."
3
In giving this
lecture, C1aussius stated the theorem as "The mean vis viva of the system is equal to its viria1."
4
In the 19th century, it was commonplace to assign a Latin name to any special characteristic of a
system. Thus, as is known to all students of celestial mechanics the vis viva integral is in reality
the total kinetic energy of the system. C1aussius also turned to the Latin word virias (the plural
of vis) meaning forces to obtain his ‘name’ for the term involved in the second half of his
theorem. This scalar quantity which he called the viria1 can be represented in terms of the forces
F
i
acting on the system as
∑
•
i
i 2
1
r F
i
and can be shown to be 1/2 the average potential energy
of the system. So, in the more contemporary language of energy, C1aussius would have stated
that the average kinetic energy is equal to 1/2 the average potential energy. Although the
characteristic of the system C1aussius called the viria1 is no longer given much significance as a
physical concept, the name has become attached to the theorem and its evolved forms.
Even though C1aussius' lecture was translated and published in Great Britain in a scant
six weeks, the power of the theorem was slow in being recognized. This lack of recognition
prompted James Clerk Maxwell four years later to observe that ''as in this country the
importance of this theorem seems hardly to be appreciated, it may be as well to explain it a little
1
more fully."
5
Maxwell's observation is still appropriate over a century later and indeed serves as
the "raison d'etre" for this book.
After the turn of the century the applications of the theorem became more varied and
widespread. Lord Rayleigh formulated a generalization of the theorem in 1903
6
in which one can
see the beginnings of the tensor viria1 theorem revived by Parker
7
and later so extensively
developed by Chandrasekhar during the 1960's.
8
Poincare used a form of the viria1 theorem in
1911
9
to investigate the stability of structures in different cosmological theories. During the
1940's Paul Ledoux developed a variational form of the virial theorem to obtain pulsational
periods for stars and investigate their stability.
10
Chandrasekhar and Fermi extended the virial
theorem in 1953 to include the presence of magnetic fields
11
At this point astute students of celestial mechanics will observe that the virial theorem
can be obtained directly from Lagrange's Identity by simply averaging it over time and making a
few statements concerning the stability of the system. Indeed, it is this derivation which is most
often used to establish the virial theorem. Since Lagrange predates Claussius by a century, some
comment is in order as to who has the better claim to the theorem.
In 1772 the Royal Academy of Sciences of Paris published J. L. Lagrange's "Essay on the
Problem of Three Bodies."
12
In this essay he developed what can be interpreted as Lagrange's
identity for three bodies. Of course terms such as "moment of inertia", "potential” and "kinetic
energy" do not appear, but the basic mathematical formulation is present. It does appear that this
remained a special case germane to the threebody problem until the winter of 184243 when
Karl Jacobi generalized Lagrange's result to nbodies. Jacobi's formulation closely parallels the
present representation of Lagrange's identity including the relating of what will later be known as
the virial of Claussius to the potential.
13
He continues on in the same chapter to develop the
stability criterion for nbody systems which bears his name. It is indeed a very short step from
this point to what is known as the Classical Virial Theorem. It is difficult to imagine that the
contemporary Claussius was unaware of this work. However, there are some notable and
important differences between the virial theorem of Claussius and that which can be deduced
from Jacobi's formulation of Lagrange's identity. These differences are amplified by considering
the state of physics during the last half of the 19th century. The passion for unification which
pervaded 20th century physics was not extant in the time of Jacobi and Claussius. The study of
heat and classical dynamics of gravitating systems were regarded as two very distinct disciplines.
The formulation of statistical mechanics which now provides some measure of unity between the
two had not been accomplished. The characterization of the properties of a gas in terms of its
internal and kinetic energy had not yet been developed. The very fact that Claussius required a
new term, the virial, for the theorem makes it clear that its relationship to the internal energy of
the gas was not clear. In addition, although he makes use of time averages in deriving the theory,
it is clear from the development that he expected these averages to be interpreted as phase or
ensemble averages. It is this last point which provides a major distinction between the virial
theorem of Claussius and that obtainable from Lagrange's Identity. The point is subtle and often
overlooked today. Only if the system is ergodic (in the sense of obeying the ergodic theorem) are
phase and time averages the same. We will return to this point later in some detail. Thus it is fair
2
to say that although the dynamical foundation for the virial theorem existed well before
Claussius' pronouncement, by demonstrating its applicability to thermodynamics he made a new
and fundamental contribution to physics.
2. The Nature of the Theorem
By now the reader may have gotten some feeling for the wide ranging applicability of the
virial theorem. Not only is it applicable to dynamical and thermodynamical systems, but we shall
see that it can also be formulated to deal with relativistic (in the sense of special relativity)
systems, systems with velocity dependent forces, viscous systems, systems exhibiting
macroscopic motions such as rotation, systems with magnetic fields and even some systems
which require general relativity for their description. Since the theorem represents a basic
structural relationship that the system must obey, applying the Calculus of Variations to the
theorem can be expected to provide information regarding its dynamical behavior and the way in
which the presence of additional phenomena (e.g., rotation, magnetic fields, etc.) affect that
behavior.
Let us then prepare to examine why this theorem can provide information concerning
systems whose complete analysis may defy description. Within the framework of classical
mechanics, most of the systems I mentioned above can be described by solving the force
equations representing the system. These equations can usually be obtained from the beautiful
formalisms of Lagrange and Hamilton or from the Boltzmann transport equation. Unfortunately,
those equations will, in general, be nonlinear, secondorder, vector differential equations which,
exhibit closed form solutions only in special cases. Although additional cases may be solved
numerically, insight into the behavior of systems in general is very difficult to obtain in this
manner. However, the virial theorem generally deals in scalar quantities and usually is applied on
a global scale. It is indeed this reduction in complexity from a vector description to a scalar one
which enables us to solve the resulting equations. This reduction results in a concomitant loss of
information and we cannot expect to obtain as complete a description of a physical system as
would be possible from the solution of the force equations.
There are two ways of looking at the reason for this inability to ascertain the complete
physical structure of a system from energy considerations alone. First, the number of separate
scalar equations one has at his disposal is fewer in the energy approach than in the force
approach. That is, the energy considerations yield equations involving only energies or 'energy
like' scalars while the force equations, being vector equations, yield at least three separate
'component' equations which in turn will behave as coupled scalar equations. One might sum up
this argument by simply saying that there is more information contained in a vector than in a
scalar.
The second method of looking at the problem is to note that energies are normally first
integrals of forces. Thus the equations we shall be primarily concerned with are related to the
first integral of the defining differential force equations. The integration of a function leads to a
3
loss of 'information' about that function. That is, the detailed structure of the function over a
discrete range is lumped into a single quantity known as integral of the function, and in doing so
any knowledge of that detailed structure is lost. Therefore, since the process of integration results
in a loss of information, we cannot expect the energy equation (representing the first integral of
the force equation) to yield as complete a picture of the system as would the solution of the force
equations themselves. However, this loss of detailed structure is somewhat compensated for,
firstly by being able to solve the resulting equations due to their greater simplicity, and secondly,
by being able to consider more difficult problems whose formulation in detail is at present
beyond the scope of contemporary physics.
3. The Scope and Structure of the Book
Any introduction to a book would be incomplete if it failed to delimit its scope. Initially
one might wonder at such an extensive discussion of a single theorem. In reality it is not possible
to cover in a single text all of the diverse applications and implications of this theorem. All areas
of physical science in which the concepts of force and energy are important are touched by the
virial theorem. Even within the more restricted study of astronomy, the virial theorem finds
applications in the dust and gas of interstellar space as well as cosmological considerations of the
universe as a whole. Restriction of this investigation to stars and stellar systems would admit
discussions concerning the stability of clusters, galaxies and clusters of galaxies which could in
themselves fill many separate volumes. Thus, we shall primarily concern ourselves with the
application of the virial theorem to the astrophysics of stars and starlike objects. Indeed, since
research into these objects is still an open and aggressively pursued subject, I shall not even be
able to guarantee that this treatment is complete and comprehensive. Since, as I have already
noted, the virial theorem does not by its very nature provide a complete description of a physical
system but rather extensive insight into its behavior, let me hope that this same spirit of incisive
investigation will pervade the rest of this work.
With regard to the organization and structure of what follows, let me emphasize that this
is a book for students  young and old. To that end, I have endeavored to avoid such phrases as
"it can easily be shown that.....”, or others designed to extol the intellect of the author at the
expense of the reader. Thus, in an attempt to clarify the development I have included most of the
algebraic steps of the development. The active professional or well prepared student may skip
many of these steps without losing content or continuity. The skeptic will wish to read them all.
However, in order not to burden the casual reader, the more tedious algebra has been relegated to
notes at the end of each chapter. Each chapter of the book has been subdivided into sections (as
has the introduction), which represent a particular logically cohesive unit. At the end of each
chapter, I have chosen to provide a brief summary of what I feel constitutes the major thread of
that chapter. A comfortable rapport with the content of these summaries may encourage the
reader in the belief that he is understanding what the author intended.
4
5
References
1 Jeans, J. H. 1925 Dynamical Theory of Gases, Cambridge University Press, London,
p. 11.
2 Claussius, R. J. E. 1851. Phil. Mag. S. 4 Vol. 2, pp. 121, 102119.
Claussius, R. J. E. 1856. Phil. Mag. S. 4, Vol. 12, p. 81.
Claussius, R. J. E. 1862. Phil. Mag. S. 4, Vol. 24, pp. 8197, 201213.
3 Claussius, R. J. E. 1870. Phil. Mag. S. 4, Vol. 40, p. 122.
4 Claussius, R. J. E. 1870. Phil. Mag. S. 4, Vol. 40, p. 124.
5 Maxwell, J. C. 1874. Scientific Papers. Vol. 2, p. 410, Dover Publications, Inc., N. Y.
6 Rayleigh, L. 1903. Scientific Papers. Vol. 4, p. 491, Cambridge, England.
7 Parker, E. N. 1954. Phys. Rev. 96, pp. 16869.
8 Chandrasekhar, S., and Lebovitz, N.R., (1962) Ap.J. 136, pp. 10371047 and references
therein.
9 Poincare, H. 1911. Lectures on Cosmological Theories, Hermann, Paris.
10 Ledoux, P. 1945, Ap. J. 102, pp. 134153.
11 Chandrasekhar, S. and Fermi, E. 1953, Ap. J. l18, p.116.
12 Lagrange, J. L. 1873. Oeuvres de Lagrange ed: M. J.A. Serret GauthierVillars, Paris,
pp. 240, 241.
13 Jacobi, C. G. J. 1889. Varlesungen uber Dynamik ed: A. Clebsch G. Reimer, Berlin,
pp. 1822.
The Virial Theorem in Stellar Astrophysics
© Copyright 2003
I. Development of the Virial Theorem
1. The Basic Equations of Structure
Before turning to the derivation of the virial theorem, it is appropriate to review the origin
of the fundamental structural equations of stellar astrophysics. This not only provides insight into
the basic conservation laws implicitly assumed in the description of physical systems, but by
their generality and completeness graphically illustrates the complexity of the complete
description that we seek to circumvent. Since lengthy and excellent texts already exist on this
subject, our review will of necessity be a sketch. Any description of a physical system begins
either implicitly or explicitly from certain general conservation principles. Such a system is
considered to be a collection of articles, each endowed with a spatial location and momentum
which move under the influence of known forces. If one regards the characteristics of spatial
position and momentum as being highly independent, then one can construct a multidimensional
space through which the particles will trace out unique paths describing their history.
This is essentially a statement of determinism, and in classical terms is formulated in a
sixdimensional space called phasespace consisting of three spatial dimensions and three
linearly independent momentum dimensions. If one considers an infinitesimal volume of this
space, he may formulate a very general conservation law which simply says that the divergence
of the flow of particles in that volume is equal to the number created or destroyed within that
volume.
The mathematical formulation of this concept is usually called the Boltzmann transport
equation and takes this form:
∑ ∑
= =
=
∂
ψ ∂
+
∂
ψ ∂
+
∂
ψ ∂
3
1 i
3
1 i i
i
i
i
S
p
p
x
x
t
,
6
The Virial Theorem in Stellar Astrophysics
or in vector notation 1.1.1
S
t
p
= ψ ∇ • + ψ ∇ • +
∂
ψ ∂
f v ,
where ψ is the density of points in phase space, f is the vector sum of the forces acting on the
particles and S is the 'creation rate' of particles within the volume. The homogeneous form of this
equation is often called the Louisville Theorem and would be discussed in detail in any good
book on Classical Mechanics.
A determination of ψ as a function of the coordinates and time constitutes a complete
description of the system. However, rarely is an attempt made to solve equation (1.1.1) but rather
simplifications are made from which come the basic equations of stellar structure. This is
generally done by taking 'moments' of the equations with respect to the various coordinates. For
example, noting that the integral of ψ over all velocity space yields the matter density ρ and that
no particles can exist with unbounded momentum, averaging equation (1.1.1) over all velocity
space yields
S ) (
t
= ρ • ∇ +
∂
ρ ∂
u , 1.1.2
where u is the average stream velocity of the particles and is defined by
∫
ψ
ρ
= dv
1
v u . 1.1.3
For systems where mass is neither created nor destroyed 0 S = , and equation (1.1.2) is just a
statement of the conservation of mass. If one multiplies equation (1.1.1) by the particle velocities
and averages again over all velocity space he will obtain after a great deal of algebra the Euler
Lagrange equations of hydrodynamic flow
∫
−
ρ
− • ∇
ρ
− Ψ −∇ = ∇ • +
∂
∂
dv ) (
1 1
) (
t
u v S u u
u
P . 1.1.4
Here the forces f have been assumed to be derivable from a potential Ψ. The symbol is known
as the pressure tensor and has the form
P
∫
− − ψ = dv ) )( ( u v u v P . 1.1.5
These rather formidable equations simplify considerably in the case where many collisions
randomize the particle motion with respect to the mean stream velocity u .Under these conditions
the last term on the right of equation (1.1.4) vanishes and the pressure tensor becomes diagonal
with each element equal. Its divergence then becomes the gradient of the familiar scalar known
as the gas pressure P. If we further consider only systems exhibiting no stream motion we arrive
at the familiar equation of hydrostatic equilibrium
Ψ ∇ ρ − = ∇P . 1.1.6
Multiplying equation (1.1.1) by v and averaging over v, has essentially turned the Boltzmann
transport equation into an equation expressing the conservation of momentum. Equation (1.1.6)
along with Poisson's equation for the sources of the potential
ρ π − = Ψ ∇ G 4
2
, 1.1.7
constitute a complete statement of the conservation of momentum.
7
The Virial Theorem in Stellar Astrophysics
Multiplying equation (1.1.1) by v v • or v
2
and averaging over all velocity space will
produce an equation which represents the conservation of energy, which when combined with
the ideal gas law is
F v • ∇ − χ + ρε = • ∇ ρ + ρ
dt
dE
, 1.1.8
where F is the radiant flux, ε the total rate of energy generation and χ is the energy generated by
viscous motions. If one has a machine wherein no mass motions exist and all energy flows by
radiation, we have a statement of radiative equilibrium;
ρε = • ∇ F . 1.1.9
For static configurations exhibiting spherical symmetry these conservative laws take their
most familiar form:
Conservation of mass ρ π =
2
r 4
dr
) r ( dm
.
Conservation of momentum
2
r
) r ( Gm
dr
) r ( dP ρ
− = . 1.1.10
Conservation of energy ρε π =
2
r 4
dr
) r ( dL
, . F r 4 ) r ( L
2
π =
2. The Classical Derivation of the Virial Theorem
The virial theorem is often stated in slightly different forms having slightly different
interpretations. In general, we shall repeat the version given by Claussius and express the virial
theorem as a relation between the average value of the kinetic and potential energies of a system
in a steady state or a quasisteady state. Since the understanding of any theorem is related to its
origins, we shall spend some time deriving the virial theorem from first principles. Many
derivations of varying degree of completeness exist in the literature. Most texts on stellar or
classical dynamics (e.g. Kurth
1
) derive the theorem from the Lagrange identity. Landau and
Lifshitz
2
give an eloquent derivation appropriate for the electromagnetic field which we shall
consider in more detail in the next section. Chandrasekhar
3
follows closely the approach of
Claussius while Goldstein
4
gives a very readable vector derivation firmly rooted in the original
approach and it is basically this form we shall develop first. Consider a general system of mass
points m
i
with position vectors r
i
which are subjected to applied forces (including any forces of
constraint) f
i
. The Newtonian equations of motions for the system are then
i
i i
i
dt
) m ( d
f
v
p = = . 1.2.1
8
The Virial Theorem in Stellar Astrophysics
Now define 
.

\

=
•
= • = • =
∑ ∑ ∑ ∑
i
2
i i 2
1
i i i
i i
i 2
1
i
i
i i i
r m
dt
d
dt
) ( d
m
dt
d
m G
r r
r
r
r p . 1.2.2
The term in the large brackets is the moment of inertia (by definition) about a point and
that point is the origin of the coordinate system for the position vectors r
i
. Thus, we have
dt
dI
G
2
1
= , 1.2.3
where I is the moment of inertia about the origin of the coordinate system.
Now consider
∑ ∑
• + • =
i
i i i i
dt
dG
r p p r , 1.2.4
but
∑ ∑ ∑
= = • = •
i i i
2
i i i i i i i
T 2 v m m r r p r , 1.2.5
where T is the total kinetic energy of the system with respect to the origin of the coordinate
system. However, since p is really the applied force acting on the system (see equation 1.2.1),
we may rewrite equation (1.2.4) as follows:
i
∑
• + =
i
i i
T 2
dt
dG
r f . 1.2.6
The last term on the right is known as the Virial of Claussius. Now consider the Virial of
Claussius. Let us assume that the forces f
i
obey a power law with respect to distance and are
derivable from a potential. The total force on the ith particle may be determined by summing all
the forces acting on that particle. Thus
∑
≠
=
i j
ij i
F f , 1.2.7
where F
ij
is the force between the ith and jth particle. Now, if the forces obey a power law and
are derivable from a potential then,
n
ij ij i ij i i ij
r a ) r ( m −∇ = Φ ∇ = F . 1.2.8
The subscript on the ∇operator implies that the gradient is to be taken in a coordinate system
having the ith particle at the origin. Carrying out the operation implied by equation (1.2.8), we
have
) ( r a n
j i
) 2 n (
ij ij ij
r r F − − =
−
. 1.2.9
9
The Virial Theorem in Stellar Astrophysics
Now since the force acting on the ith particle due to the jth particle may be paired off
with a force exactly equal and oppositely directed, acting on the jth particle due to the ith
particle, we can rewrite equation (1.2.7) as follows:
∑ ∑
>
+ = =
j i j
ji ij ij i
F F F f . 1.2.10
Substituting equation (1.2.10) into the definition of the Virial of Claussius, we have
∑ ∑∑
>
• + • = •
i i i j
j ji i ij i i
r F r F r f . 1.2.11
It is important here to notice that the position vector r
i
, which is 'dotted' into the force vector,
bears the same subscript as the first subscript on the force vector. That is, the position vector is
the vector from the origin of the coordinate system to the particle being action upon. Substitution
of equation (1.2.9) into equation (1.2.11) and then into equation (1.2.6) yields:
U n T 2
dt
dG
− = , 1.2.12
where U is the total potential energy.
1.1
For the gravitational potential n = 1, and we arrive at
a statement of what is known as Lagranges’ Identity:
Ω + = = T 2
dt
I d
dt
dG
2
2
2
1
. 1.2.13
To arrive at the usual statement of the virial theorem we must average over an interval of
time (T
0
). It is in this sense that the virial theorem is sometimes referred to as a statistical
theorem. Therefore, integrating equation (1.2.12), we have
∫ ∫ ∫
− =
0 0
0 0
0 0
0
0
dt ) t (
n
dt ) t ( T
2
dt
dt
dG 1
T T
T T T
U
0
T
. 1.2.14
and, using the definition of average value we obtain:
  U n T 2 ) 0 ( G ) ( G
1
0
0
− = − T
T
. 1.2.15
If the motion of the system over a time T
0
is periodic, then the lefthand side of equation (1.2.15)
will vanish. Indeed, if the motion of the system is bounded [i.e., G(t) < ∞], then we may make
the left hand side of equation (1.2.15) as small as we wish by averaging over a longer time. Thus,
if a system is in a steady state the moment of inertia ( I ) is constant and for systems governed by
gravity
0 T 2 = Ω + . 1.2.16
It should be noted that this formulation of the virial theorem involves time averages of
indeterminate length. If one is to use the virial theorem to determine whether a system is in
accelerative expansion or contraction, then he must be very careful about how he obtains the
average value of the kinetic and potential energies.
10
The Virial Theorem in Stellar Astrophysics
3. Velocity Dependent Forces and the Virial Theorem
There is an additional feature of the virial theorem as stated in equation (1.2.16) that
should be mentioned. If the forces acting on the system include velocity dependent forces, the
result of the virial theorem is unchanged. In order to demonstrate this, consider the same system
of mass points m
i
subjected to forces f
i
which may be divided into velocity dependent ) (
i
w and
velocity independent forces (z
i
). The equations of motion may be written as:
i i i i
z w f p + = = . 1.3.1
Substituting into equation (1.2.6), we have
∑ ∑
• + = • −
i i
i i i i
T 2
dt
dG
r z r w . 1.3.2
Remembering that the velocity dependent forces may be rewritten as
dt
d
i
i i i i
r
v w α = α = . 1.3.3
We may again average over time as in equation (1.2.12). Thus
U n T 2 dt
dt
d 1
dt
dt
dG 1
i
0
i
i
i
0
0
0
0 0
− = • α −
∫
∑
∫
r
r
T T
T T
, 1.3.4
where U is the average value of the potential energy for the "nonfrictional" forces. Carrying out
the integration on the left hand side we have
    U n T 2 ) 0 ( r ) ( r
1
) 0 ( G ) ( G
1
i
2
i 0
2
i i
0
0
0
− = − α + −
∑
T
2T
T
T
. 1.3.5
Thus, if the motion is periodic, both terms on the left hand side of equation (1.3.5) will vanish in
a time T
0
equal to the period of the system. Indeed both terms can be made as small as required
providing the "frictional" forces
i
w do not cause the system to cease to be in motion over the
time for which the averaging is done. This apparently academic aside has the significant result
that we need not worry about any Lorentz forces or viscosity forces which may be present in our
subsequent discussion in which we shall invoke the virial theorem.
4. ContinuumField Representation of the Virial Theorem
Although nearly all derivations of the virial theorem consider collections of masspoints
acting under forces derivable from a potential, it is useful to look at this formalism as it applies
to a continuum density field of matter. This is particularly appropriate when one considers
applications to stellar structure where a continuum representation of the material is always used.
In the interests of preserving some rigor let us pass from equation (1.2.1) to its analogous
representation in the continuum. Let the mass m
i
be obtained by multiplying the density ρ(r) by
an infinitesimal volume ∆V so that 1.2.1 becomes
11
The Virial Theorem in Stellar Astrophysics
( )
dt
) V ( d
dt
d
V
dt
d
V V
dt
d
i
∆
ρ + ∆ ρ +
ρ
∆ = ∆ ρ = v
v
v v f . 1.4.1
Conservation of mass requires that
( ) 0
dt
) V ( d
dt
d
V V
dt
d
dt
dm
i
=
∆
ρ +
ρ
∆ = ∆ ρ = . 1.4.2
Multiplying this expression by v we see that the first and last terms on the right hand side of
equation (1.4.1) are of equal magnitude and opposite sign. Thus, if we define a "force density",
f, so that f ∆V = f
i
, we can pass to this continuum representation of equation (1.2.1):
  ) ( ) (
dt
d
) ( ) ( r r v r r p f = ρ = , 1.4.3
where p(r) by analogy to 1.2.1 is just the local momentum density.
We can now define G in terms of the continuum variables so that
( )
∫ ∫ ∫ ∫
ρ = • ρ = • ρ = • =
V
2
2
1
V
2
1
V V
dV
dt
dr
dV
dt
d
dV
dt
d
dV G r r r
r
r p , 1.4.4
so that
( ) dV
dt
d
r dV r
dt
d
G
V
2
2
1
V
2
2
1
∫ ∫
ρ
− ρ = . 1.4.5
Once again, one uses conservation of mass requiring that the mass within any sub volume V' is
constant with time so that 0
dt
) ' V (
=
dm
with that subvolume V' defined such that
0 dV
dt
d
dV
dt
d
' V ' V
=
ρ
=


.

\

ρ
∫ ∫
. 1.4.6
Thus, the second integral in equation (1.4.5) after integration by parts is zero. If we take the
original volume V to be large enough so as to always include all the mass of the object, we may
write equation (1.4.5) as
∫
= ρ =
V
2
1
2
dt
dI
dV ) r (
dt
d
2
1
G . 1.4.7
With these same constraints on V we may differentiate equation (1.4.4) with respect to time and
obtain
∫ ∫
• + ρ =
(
¸
(
¸
• + • =
V
2
V
dV ) v ( dV
dt
d
dt
d
dt
dG
f
p
p r r
r
. 1.4.8
The first term under the integral is just kinetic energy density and hence its volume
integral is .just the total kinetic energy of the configuration and
∫
• + =
V
2
2
2
1
dV T 2
dt
I d
r f . 1.4.9
12
The Virial Theorem in Stellar Astrophysics
Considerable care must be taken in evaluating the second term in equation (1.4.9) which
is basically the virial of Claussius. In the previous derivation we went to some length [i.e.,
equation (1.2.10)] to avoid "double counting" the forces by noting that the force between any two
particles A and B can be viewed as a force at A due to B, or a force at B resulting from A. The
contributions to the virial, however, are not equal as they involve a 'dot' product with the position
vector. Thus, we explicitly paired the forces and arranged the sum so pairs of particles were only
counted once. Similar problems confront us within continuum derivation. Thus, each force at a
field point f (r) will have an equal and opposite counterpart at the source points r . '
After some algebra, direct substitution of the potential gradient into the definition of the
Virial of Claussius yields
1.2
(
¸
(
¸
− ρ ρ =
− − • − ρ ρ − = •
∫ ∫
∫ ∫ ∫
−
V
n
' V
2
1
V ' V
2 n
2
n
V
dV ' dV ) ' )( ' ( ) ( n
dV ' dV ) ' )( ' ( ) ' )( ' ( ) ( dV
r r r r
r r r r r r r r r f
. 1.4.10
Since V = V', the integrals are fully symmetric with respect to interchanging primed with non
primed variables. In addition the double integral represents the potential energy of ρ(r) with
respect to ρ(r ') , and ρ(r ') with respect to ρ(r); it is just twice the total potential energy. Thus,
we find that the virial has the same form as equation (1.2.12), namely,
∫
− = •
V
n dV U r f . 1.4.11
Substitution of this form into equation (1.4.9) and taking n = 1 yields the same expression for
Lagrange's identity as was obtained in equation (1.2.13), specifically,
Ω + = T 2
dt
I d
2
2
2
1
1.4.12
Thus Lagrange's identity, the virial theorem and indeed the remainder of the earlier arguments,
are valid for the continuum density distributions as we might have guessed.
Throughout this discussion it was tacitly assumed that the forces involved represented
"gravitational" forces insofar as the force was ρ∇Φ. Clearly, if the force depended on some
other property of the matter (e.g., the charge density, ε(r) the evaluation of would go
as before with the result that the virial would again be –nU where U is the total potential energy
of the configuration.
∫
•
V
dV r f
13
The Virial Theorem in Stellar Astrophysics
5. The Ergodic Theorem and the Virial Theorem
Thus far, with the exception of a brief discussion in Section 2, we have developed
Lagrange's identity in a variety of ways, but have not rigorously taken that finial step to produce
the virial theorem. This last step involves averaging over time and it is in this form that the
theorem finds its widest application. However, in astrophysics few if any investigators live long
enough to perform the timeaverages for which the theorem calls. Thus, one more step is needed.
It is this step which occasionally leads to difficulty and erroneous results. In order to replace the
time averages with something observable, it is necessary to invoke the ergodic theorem.
The Ergodic Theorem is one of those fundamental physical concepts like the Principle of
Causality which are so "obvious" as to appear axiomatic. Thus they are rarely discussed in the
physics literature. However, to say that the ergodic theorem is obvious is to belittle an entire area
of mathematics known as ergodic theory which uses the mathematical language of measure
theory. This language alone is enough to hide it forever from the eye of the average physical
scientist. Since this theorem is central to obtain what is commonly called the virial theorem, it is
appropriate that we spend a little time on its meaning. As noted in the introduction, the
distinction between an ensemble average and an average of macroscopic system parameters over
time was not clear at the time of the formulation of the virial theorem. However, not too long
after, Ludwig Boltzmann
6
formulated an hypothesis which suggested the criterion under which
ensemble and phase averages would be the same. Maxwell later stated it this way: "The only
assumption which is necessary for a direct proof is that the system if left to itself in its actual
state of motion will, sooner or later, pass through every phase which is consistent with the
equation of energy".
7
Essentially this constitutes what is most commonly meant by the ergodic theorem.
Namely, if a dynamic system passes through every point in phase space then the time average of
any macroscopic system parameter, say Q, is given by
s
t
t
t
Q dt ) t ( Q
1 Lim
Q
0
0
> =<
(
¸
(
¸
∞ →
= > <
∫
+T
T T
, 1.5.1
where <Q>
s
is some sort of instantaneous statistical average of Q over the entire system.
The importance of this concept for statistical mechanics is clear. Theoretical
considerations predict <Q>
s
whereas experiment provides something which might be construed
to approximately <Q>
t
. No matter how rapid the measurements of something like the pressure or
temperature of the gas, it requires a time which is long compared to characteristic times for the
system. The founders of statistical mechanics, such as Boltzmann, Maxwell and Gibbs, realized
that such a statement as equation (1.5.1) was necessary to enable the comparison of theory with
experiment and thus a great deal of effort was expended to show or at least define the conditions
under which dynamical systems were ergodic (i.e., would pass through every point in phase
space).
14
The Virial Theorem in Stellar Astrophysics
Indeed, as stated, the ergodic theorem is false as was shown independently in 1913 by
Rosentha1
8
and Plancherel
9
more modern version of this can be seen easily by noting that no
system trajectory in phase space may cross itself. Thus, such a curve may have no multiple
points. This is effectively a statement of system boundary conditions uniquely determining the
system's past and future. It is the essence of the Louisville theorem of classical mechanics. Such
a curve is topologically known as a Jordan curve and it is a well known topological theorem that
a Jordan curve cannot pass through all points of a multidimensional space. In the language of
measure theory, a multidimensional space filling curve would have a measure equal to the space
whereas a Jordan curve being onedimensional would have measure zero. Thus, the ergodic
hypothesis became modified as the quasiergodic hypothesis. This modification essentially states
that although a single phase trajectory cannot pass through every point in phase space, it may
come arbitrarily close to any given point in a finite time. Already one can sense confusion of
terminology beginning to mount. Ogorodnikov
10
uses the term quasiergodic to apply to systems
covered by the Lewis theorem which we shall mention later. At this point in time the
mathematical interest in ergodic theory began to rise rapidly and over the next several years
attracted some of the most, famous mathematical minds of the 20th century. Farquhar
11
points
out that several noted physicists stated without justification that all physical systems were quasi
ergodic. The stakes were high and were getting higher with the development of statistical
mechanics and the emergence of quantum mechanisms as powerful physical disciplines. The
identity of phase and time averages became crucial to the comparison of theory with observation.
Mathematicians largely took over the field developing the formidable literature currently
known as ergodic theory; and they became more concerned with showing the existence of the
averages than with their equality with phase averages. Physicists, impatient with mathematicians
for being unable to prove what appears 'reasonable', and also what is necessary, began to require
the identity of phase and time averages as being axiomatic. This is a position not without
precedent and a certain pragmatic justification of expediency. Some essentially adopted the
attitude that since thermodynamics “works”, phase and times averages must be equal. However,
as Farquhar observed “such a pragmatic view reduces statistical mechanics to an ad hoc
technique unrelated to the rest of physical theory.”
12
Over the last half century, there have been many attempts to prove the quasiergodic
hypothesis. Perhaps the most notable of which are Birkhoff's theorem
13
and the generalization of
a corollary known as Lewis' theorem.
14
These theorems show the existence of time averages and
their equivalence to phase averages under quite general conditions. The tendency in recent years
has been to bypass phase space filling properties of a dynamical system and go directly to the
identification of the equality of phase and time averages. The most recent attempt due to Siniai
15
,
as recounted by Arnold and Avez
16
proves that the BoltzmannGibbs conjecture is correct. That
is, a "gas" made up of perfectly elastic spheres confined by a container with perfectly reflecting
walls is ergodic in the sense that phase and time averages are equal.
At this point the reader is probably wondering what all this has to do with the virial
theorem. Specifically, the virial theorem is obtained by taking the time average of Lagrange's
identity. Thus
15
The Virial Theorem in Stellar Astrophysics
t t
t
t
2
2
T 2 dt
dt
I d
2
1 Lim 0
0
> < − > =<


.

\

(
¸
(
¸
∞ →
∫
+
U
T
T
, 1.5.2
and for systems which are stable the left hand side is zero. The first problem arises with the fact
that the time average is over infinite time and thus operationally difficult to carryout
l.3
.
Farquhar
17
points out that the time interval must at least be long compared to the relaxation time
for the system and in the event that the system crossing time is longer than the relaxation time,
the integration in equation(1.5.2) must exceed that time if any statistical validity is to be
maintained in the analysis of the system. It is clear that for stars and starlike objects these
conditions are met. However, in stellar dynamics and the analysis of stellar systems they
generally are not. Indeed, in this case, the astronomer is in the envious position of being in the
reverse position from the thermodynamicists. For all intents and purposes he can perform an
'instantaneous' ensemble average which he wishes to equate to a 'theoretically determined' time
average. This interpretation will only be correct if the system is ergodic in the sense of satisfying
the 'quasiergodic hypothesis'. Pragmatically if the system exhibits a large number of degrees of
freedom then persuasive arguments can be made that the equating of time and phase averages is
justified. However, if isolating integrals of the motion exist for the system, then it is not justified,
as these integrals remove large regions of phase space from the allowable space of the system
trajectory. Lewis' theorem allows for ergodicity in a subspace but then the phase averages must
be calculated differently and this correspondence to the observed ensemble average is not clear.
Thus, the application of the virial theorem to a system with only a few members and hence a few
degrees of freedom is invalid unless care is taken to interpret the observed ensemble averages in
light of phase averages altered by the isolating integrals of the motion. Furthermore, one should
be most circumspect about applying the virial theorem to large systems like the galaxy which
appear to exhibit quasiisolating integrals of the motion. That is, integrals which appear to
restrict the system motion in phase space over several relaxation times. However, for stars and
starlike objects exhibiting 10
50
or more particles undergoing rapid collisions and having short
relaxation times, these concerns do not apply and we may confidently interchange time and
phase averages as they appear in the virial theorem. At least we may do it with the same
confidence of the thermodynamicist. For those who feel that the ergodic theorem is still "much
ado about nothing", it is worth observing that by attempting to provide a rational development
between dynamics and thermodynamics, ergodic theory must address itself to the problems of
irreversible processes. Since classical dynamics is fully reversible and thermodynamics includes
processes which are not, the nature of irreversibility must be connected in some sense to that of
ergodicity and thus to the very nature of time itself. Thus, anyone truly interested in the
foundations of physics cannot dismiss ergodic theory as mere mathematical 'nitpicking'.
16
The Virial Theorem in Stellar Astrophysics
6. Summary
In this chapter, I have tried to lay the groundwork for the classical virial theorem by first
demonstrating its utility, then deriving it in several ways and lastly, examining an important
premise of its application. An underlying thread of continuity can be seen in all that follows
comes from the Boltzmann transport equation. It is a theme that will return again and again
throughout this book. In section 1, we sketched how the Boltzmann transport equation yields a
set of conservation laws which in turn supply the basic structure equations for stars. This sketch
was far from exhaustive and intended primarily to show the informational complexity of this
form of derivation. Being suitably impressed with this complexity, the reader should be in an
agreeable frame of mind to consider alternative approaches to solving the vector differential
equations of structure in order to glean insight into the behavior of the system. The next two
sections were concerned with a highly classical derivation of the virial theorem with section 2
being basically the derivation as it might have been presented a century ago. Section 3 merely
updated this presentation so that the formalism may be used within the context of more
contemporary field theory. The only 'tricky' part of these derivations involves the 'pairing' of
forces. The reader should make every effort to understand or conceptualize how this occurs in
order to understand the meaning of the virial itself. The assumption that the forces are derivable
from a potential which is described by a power law of the distance alone, dates back at least to
Jacobi and is often described as a homogeneous function of the distance.
In the last section, I attempted to provide some insight into the meaning of a very
important theorem generally known as the ergodic theorem. Its importance for the application of
the virial theorem cannot be too strongly emphasized. Although almost all systems of interest in
stellar astrophysics can truly be regarded as ergodic, many systems in stellar dynamics cannot. If
they are not, one cannot replace averages over time by averages over phase or the ensemble of
particles without further justification.
17
The Virial Theorem in Stellar Astrophysics
Notes to Chapter 1
1.1 Since a
ij
= a
ji
for all known physical forces, we may substitute equation (1.2.9) in
equation (1.2.11) as follows:
 
 
∑∑ ∑∑
∑ ∑∑
>
−
>
−
>
−
− = − • − − =
• − + • − − = •
i i j
2
ij
) 2 n (
ij ij
i i j
j i j i
) 2 n (
ij ij
i i i j
j i j i j i
) 2 n (
ij ij i i
r r a n ) ( ) ( r a n
) ( ) ( r na
r r r r
r r r r r r r f
. N 1.1.1
Thus
∑∑ ∑∑ ∑
> >
Φ − = − = •
i i j
ij
i i j
n
ij ij
i
i i
) r ( n r a n r f . N 1.1.2
Since the second summation is only over j > i, there is no "doublecounting" involved, and the
double sum is just the total potential energy of the system.
1.2 As in Section 2, let us assume that the force density is derivable from a potential which
is a homogeneous function of the distance between the source and field point.
5
Then, we can
write the potential as
( ) 0 n ' dV ' ) ( ) (
n
' V
< ∀ − ρ = Φ
∫
r r r r , N 1.2.1
and the force density is then
( )
∫
− ∇ ρ ρ − = Φ ∇ ρ − =
' V
n
r r
' dV ' ) ' ( ) ( ) ( ) ( ) ( r r r r r r r f , N 1.2.2
while the force density at a source point due to all the field points is
( )
∫
− ∇ ρ ρ − = Φ ∇ ρ − =
V
n
' r ' r
dV ' ) ( ) ' ( ) ' ( ) ' ( ) ' ( r r r r r r r f , N 1.2.3
where ∇
r
and∇ denote the gradient operator evaluated at the field point r and the source point
r’ respectively. Since the contribution to the force density from any pair of sources and field
points will lie along the line joining the two points,
' r
( ) ( ) ( ) ( ) ' ' n ' '
2 n n
' r
n
r
r r r r r r r r − − = − −∇ = − ∇
−
. N 1.2.4
Now , so multiplying equation (N 1.2.2) by r and integrating over
V produces the same result as multiplying equation (N 1.2.3) by r' and integrating over V'. Thus,
doing this and adding equation (N 1.2.2) to equation (N 1.2.3). we get
∫ ∫
• = •
' V V
' dV ' ) ' ( dV ) ( r r r r f f
( ) ( ) dV ' dV ) ( ) ' ( dV ' dV ) ( ) ( dV 2
n
' r
' V
n
r
V V ' V
r' r r r r r' r r r r r
V
− ∇ ρ ρ − = − ∇ • ρ ρ − = •
∫ ∫ ∫ ∫ ∫
f . N 1.2.5
1.3 It should be noted that the left hand side of 1.5.2 is zero if the system is periodic
and the integral is taken over the period.
18
The Virial Theorem in Stellar Astrophysics
19
References
1 Kurth, Rudolf (1957), Introduction to the Mechanics of Stellar Systems, Pergamon Press,
N. Y., London, Paris, pp. 69.
2 Landau, L. D. & Lifshitz, E. M. (1962), The Classical Theory of Fields, Trans. M.
Hamermesh, AddisonWesley Pub. Co., Reading, Mass. USA, p. 9597.
3 Chandrasekhar, S. (1957), An Introduction to the Study of Stellar Structure, Dover Pub.,
Inc., pp. 4951.
4 Goldstein, H. (1959), Classical Mechanics, AddisonWesley Pub. Co., Reading, Mass. ,
pp. 6971.
5 Landau, L. D. & Lifshitz, E. M. (1960), Mechanics, Trans. J. B. Sykes & J. S. Bell,
AddisonWesley Publ. Co., Reading, Mass., pp. 2223.
6 Boltzmann, L. (1811), Sitzler Akad. Wiss. Wien 63, 397, 679.
7 Maxwell, J. C. (1879), Trans. Cam. Phil. Soc. 12, p.547.
8 Rosenthal, A. (1913), Ann. der Physik 42, p.796.
9 Plancherel, M. (1913), Ann. der Physik 42, p.1061.
10 Ogorodnikov, K. F. (1965), Dynamics of Stellar Systems, Pergamon Press, The
McMillan Co., New York, p. 153.
11 Farquhar, I. E. (1964), Ergodic Theory in Statistical Mechanics, Interscience Pub.,
John Wiley & Sons, Ltd., London, New York, Sydney, p. 77.
12 .ibid., p. 3.
13 Birkhoff, G. D. (1931), Proc. of Nat'l. Acad. of Sci.: U.S. 17, p. 656.
14 Lewis, R. M. (1966), Arch. Rational Mech. Anal 5, p. 355.
15 Sinai, Ya. (1962), Vestnik Mossovskova Gosudrastvennova Universitata Series Math. 5.
16 Arnold, V. I., and Avez, A. (1968), Ergodic Problem of Classical Mechanics, W. A.
Benjamin, Inc., New York, Amsterdam, p. 78.
17 Farquhar, I. E. (1964), Loc. cit. pp. 2332.
The Virial Theorem in Stellar Astrophysics
© Copyright 2003
II Contemporary Aspects of the Virial Theorem
1. The Tensor Virial Theorem
The tensor representation of the virial theorem is an attempt to restore some of the
information lost in reducing the full vector equations of motion described in Chapter I, section 1 to
scalars. Although the germ of this idea can be found developing as early as 1903 in the work of
Lord Rayleigh
l
, it wasn't until the 1950's that Parker
2, 3
, and Chandrasekhar and Fermi
4
found the
concept particularly helpful in dealing with the presence of magnetic fields. The concept was
further expanded by Lebovitz
5
and in a series of papers by Chandrasekhar and Lebovitz
6, 7
during
the 1960's, for the investigation of the stability of various gaseous configurations. Chandrasekhar
33
has given a fairly comprehensive recounting of his efforts on this subject after the original version
of this monograph was prepared. However, the most lucid derivation is probably that presented by
Chandrasekhar
8
in 1961 and it is a simplified version of that derivation which I shall give here.
As previously mentioned the motivation for this approach is to regain some of the
information lost in forming the scalar virial theorem by keeping track of certain aspects of the
system associated with its spatial symmetries. If one recalls the fullblown vector equations of
motion in Chapter I, section 1, this amounts to keeping some of the component information of those
equations, but not all. In particular, it is not surprising that since system symmetries inspire this
approach that the information to be kept relates to motions along orthogonal coordinate axes.
At this point, it is worth pointing out that the derivation in Chapter I, section 2, essentially
originates from the equations of motion of the system being considered. The derivations take the
form of multiplying those equations of motions by position vectors and averaging over the spatial
volume. The final step involves a further average over time. That is to say that the virial theorem
results from taking spatial moments of the equations of motion and investigating their temporal
behavior. (Recall that the equations of motion themselves are moments of the Boltzmann transport
equation.) Since moment analysis of this type also yields some of the most fundamental
conservation laws of physics (i.e., momentum, mass and energy), it is not surprising that the virial
theorem should have the same power and generality as these laws. Indeed, it is rather satisfying to
one who believes that "all that is good and beautiful in physics" can be obtained from the
20
The Virial Theorem in Stellar Astrophysics
Boltzmann equation that the virial theorem essentially arises from taking higher order moments of
that equation. With that in mind let us consider a collisionless pressurefree system analogous with
that considered in Chapter I, section 2 and neglect viscous forces and macroscopic forces such as
net rotation and magnetic fields as we shall consider them later. Under these conditions, equation
(1.1.4) becomes
dt
d
) (
t
u
u u
u
ρ = Φ ∇ ρ − = ∇ • ρ +
∂
∂
ρ , 2.1.1
which is simply the vector representation of either equations (1.1.4) or (1.2.1). In Chapter I, section
2, we essentially took the inner product of equation (1.1.4) with the position vector r and integrated
over the volume to produce a scalar equation. Here we propose to take the outer product of
equation (1.1.1) with the position vector r producing a tensor equation which can be regarded as a
set of equations relating the various components of the resulting tensors. Cursory dimensional
arguments should persuade one that this procedure should produce relationships between the
various moments of inertia of the system and energylike tensors. Thus, our starting point is
∫ ∫
Φ ∇ ρ = ρ
V V
dV dV
dt
d
r
u
r . 2.1.2
Although some authors choose a slightly different convention, the term on the right hand side of
equation (2.1.2) can properly be called the virial tensor. Now, as before let the potential be
( ) 0 n ' dV ' ) ' ( ) (
n
' V
< ∀ − ρ = Φ
∫
r r r r . 2.1.3
Then following exactly the same manipulation as in Chapter I, only taking into account outer
products instead of inner products with the position vectors, we get the virial tensor.
2.1
( )( )( )
)
`
¹
¹
´
¦
− − − ρ ρ = Φ ∇ ρ
∫ ∫ ∫
−
V ' V
2 n
2
1
V
dV ' dV ' ' ' ) ' ( ) ( n dV r r r r r r r r r . 2.1.4
If we define
( )( )( ) dV ' dV ' ' ' ) ' ( ) (
dV ) (
dV ) (
2 n
V ' V
2
1
V
2
1
V
−
− − − ρ ρ =
ρ =
ρ =
∫ ∫
∫
∫
r r r r r r r r
uu
rr
U
T
I
, 2.1.5
equation (2.1.2 becomes)
U T
I
n 2
dt
d
2
2
2
1
+ = . 2.1.6
which is essentially the tensor representation of Lagrange's identity
2.2
where I is sometimes called
the moment of inertia tensor, T the kinetic energy tensor and U the potential energy tensor. By
eliminating additional external forces such as magnetic fields and rotation we have lost much of the
power of the tensor approach. However, some insight into this power can be seen by considering in
component form one term in the expansion of the virial tensor
2.1
.
21
The Virial Theorem in Stellar Astrophysics
dV
dt
dx
x
dt
d
dV
dt
d
dt
d
j
i
V V


.

\

ρ = 
.

\

ρ
∫ ∫
r
r . 2.1.7
Since this tensor is clearly symmetric we find, by using the same conservation of mass arguments
discussed earlier, that
0 dV
dt
dx
x
dt
dx
x
dt
d
V
i
j
j
i
=


.

\

− ρ
∫
, 2.1.8
which simply says the angular momentum about x
k
is conserved. Thus the tensor virial theorem
leads us to a fundamental conservation law which would not have been apparent from the scalar
form derived earlier.
2 Higher Order Virial Equations
In the last section it became clear that both the scalar and tensor forms of the virial theorem
are obtained by taking spatial moments of the equations of motion. Chandrasekhar
9
was apparently
the first to note this and to inquire into the utility of taking higher moments of the equations of
motion. There certainly is considerable precedent for this in mathematical physics. As already
noted, moments in momentumspace of the Boltzmann transport equation yield expressions for the
conservation of mass, momentum and energy. Spatial moments of the transport equation of a
photon gas can be used to obtain the equation of radiative transfer. Approximate solutions to the
resulting equations can be found if suitable assumptions such as the existence of an equation of
state are made to "close" the moment equations. Such is the origin of such diverse expressions as
the Eddington approximation in radiative transfer, the diffusion approximation in radiative transfer,
the diffusion approximation in gas dynamics and many others. Usually, the higher the order of the
moment expressions, the less transparent their physical content. Nevertheless, in the spirit of
generality, Chandrasekhar investigated the properties of the first several moment equations. In a
series of papers, Chandrasekhar and Lebovitz
10, 11
and later Chandrasekhar
12, 13
developed these
expressions as far as the fourthorder moments of the equations of motion.
Since for no moment expressions other than those of the first moment do any terms ever
appear that can be identified with the Virial of Claussius, it is arguable as to whether they should be
called virial expressions at all. However, since it is clear that this investigation was inspired by
studies of the classical virial theorem, I will briefly review their development. Recall the Euler
Lagrange equation of hydrodynamic flow developed in Chapter I, equation (1.1.4)
∫
−
ρ
− • ∇
ρ
− Ψ −∇ = ∇ • +
∂
∂
dv ) (
1 1
) (
t
u v S u u
u
P . 2.2.1
22
The Virial Theorem in Stellar Astrophysics
Quite simply the nth order "virial equations" of Chandrasekhar are generated by taking (nl) outer
tensor products of the radius vector r and equation (2.2.1). The result is then integrated over all
physical space. This leads to a set of tensor equations containing tensors of rank n. If we assume
that particle collisions are isotropic, then the source term of equation (2.2.1) vanishes
and∇ . The symbolic representation for the nth order "virial equation" can then be written
as:
P ∇ = • P
∫ ∫ ∫
= ∇ + Φ ∇ ρ + ρ
− − −
V
) 1 n (
V V
) 1 n ( ) 1 n (
0 PdV dV
dt
d
r r
u
r . 2.2.2
Recalling our arguments in Chapter I, section 4, about conservation of mass, it follows from
equation (1.4.6) that
∫ ∫
ρ =


.

\

ρ
V V
dV
dt
dQ
QdV
dt
d
, 2.2.3
where Q is any pointdefined property of the medium. Thus, the first order "virial equations",
correspond to the integrals of the equation of motion themselves over volume. So
∫ ∫ ∫
= ∇ + Φ ∇ ρ + ρ
V V V
0 PdV dV dV
dt
d
u . 2.2.4
Noting that ∇ , the second and third integrals are zero by the divergence theorem and
equation (2.2.4) becomes
( 3 / Q Q 1 • ∇ = )
0 dV
dt
d
V
2
2
2
1
=


.

\

ρ
∫
r , 2.2.5
essentially telling us that the center of mass
is not being accelerated. Setting n = 2 we
arrive at the second order "virial equations" that are the tensor version of Lagrange's identity that
we discussed in the previous section.


.

\

ρ
∫
V
dV r
As in equation (2.2.2), we can represent the nth order "virial equations" as
0 PdV dV dV
dt
d
V
) 1 n ( ) 1 n (
V V
) 1 n (
= ∇ + Φ ∇ ρ + ρ
∫ ∫ ∫
− − −
r r
u
r , 2.2.6
which after use of continuity and the divergence theorem becomes
( ) ( 0 dV P dV dV
dt
d
dV
dt
d
V V
) 1 n ( ) 1 n (
V
) 1 n (
V
) 1 n (
= ∇ • + Φ ∇ ρ + ρ − ρ
∫ ∫ ∫ ∫
− − − −
r r u r u r 1 ) . 2.2.7
Since the outer product in general does not commute, the integrals of the second and fourth term
become strings of tensors of the form
23
The Virial Theorem in Stellar Astrophysics
( ) ( ) ( ) uu r u r u r u r u u r
) 2 n ( ) 3 n ( ) 2 n ( ) 1 n (
dt
d
− − − −
+ ⋅ ⋅ ⋅ + + = ρ ,
and 2.2.8
( ) ( ) 1 1 1
) 2 n ( ) 3 n ( ) 2 n ( ) 1 n (
P P
− − − −
+ ⋅ ⋅ ⋅ + + = ∇ r r r r r .
However, the first term can be written as
dV
dt
d
! n
1
dV
dt
d
V
) n (
2
2
V
) 1 n (
∫ ∫
ρ = ρ
−
r u r . 2.2.9
Thus, each term in equation (2.2.7) represents one or more tensors of rank n, the first of
which is the second time derivative of a generalized moment of inertia tensor and the last three are
all 'energylike' tensors. From equation (2.2.8), it is clear that the second integral will generate
tensors which are spatial moments of the kinetic energy distributions while the last term will
produce moment tensors of the pressure distribution. The third integral is, however, the most
difficult to rigorously represent. For n = 2 we know it is just the total potential energy.
Chandrasekhar
8
shows how these tensors can be built up from the generalized Newtonian tensor
potential or alternatively from a series of scalar potentials which obey the equations
.
.
.
G 32
G 8
G 4
6
4
2
ρ π − = ℜ ∇
ρ π − = ℑ ∇
ρ π − = Ψ ∇
. 2.2.10
Thus, we have formulated a representation of what Chandrasekhar calls the "higher order
virial equations". They are, in fact, spatial tensor moments of the equations of motion. We may
expect them to be of importance in the same general way as the virial theorem itself. That is, in
stationary systems the left hand side of equation (2.2.7) vanishes and the result is a system of
identities between the various tensor energy moments. Keeping in mind that any continuum
function can be represented in terms of a moment expansion, equation (2.2.7) must thus contain all
of the information concerning the structure of the system. These equations thus represent an
alternate form to the solution of the equation of motions. Like most series expansions, it is devoutly
to be wished that they will converge rapidly and the "higher order" tensors can usually be
neglected.
24
The Virial Theorem in Stellar Astrophysics
3. Special Relativity and the Virial Theorem
So far we have considered only the virial theorem that one obtains from the Newtonian
equations of motion. Since there are systems such as white dwarfs, wherein the dynamic pressure
balancing gravity is supplied by particles whose energies are very much larger than their rest
energy, it is appropriate that we investigate the extent to which we shall have to modify the virial
theorem to include the effects of special relativity. For systems in equilibrium, the virial theorem
says 2T = Ω. One might say that it requires a potential energy equal to 2T to confine the motions of
particles having a total kinetic energy T. As particles approach the velocity of light the kinetic
energy increases without bound. One may interpret this as resulting from an unbounded increase of
the particle's mass. This increase will also affect the gravitational potential energy, but the effect is
quadratic. Thus we might expect in a relativistic system that a potential energy less than 2T would
be required to maintain equilibrium. This appears to be the conclusion arrived at by Chandrasekhar
when, while investigating the internal energy of white dwarfs he concludes that as the system
becomes more relativistically degenerate, T approaches Ω and this "must be the statement of the
virial theorem for material particles moving with very nearly the velocity of light."
14
This is indeed the case and is the asymptotic limit represented by a photon gas or polytrope
of index n = 3 (see Collins
32
). In order to obtain the somewhat more general result of a relativistic
form of Lagrange's identity, we shall turn to the discussion of relativistic mechanics of Landau and
Lifshitz
15
mentioned in Chapter I. As most discussions in field mechanics generally start from a
somewhat different prospective, let us examine the correspondence with the starting point of the
equations of motion adopted in the earlier sections. Generally most expositions of field mechanics
start with the statement that
0 = ℑ • , 2.3.1
where is the Maxwell stressenergy tensor and is known as the fourgradient operator. This is
equivalent to saying that there exists a volume in spacetime sufficiently large so that outside that
volume the stressenergy tensor is zero. This equivalence is made obvious by applying Gauss'
divergence theorem so that
ℑ
∫
R
• . 2.3.2 0 d dR
S
= • ℑ = ℑ
∫
s
In short, equation (2.3.1) is a conservation law. We have already seen that the fundamental
conservation laws of physics are derivable from the Boltzmann transport equation as are the
equations of motion. Indeed, the operation of taking moments is quite similar in both cases. Thus,
both starting points are equivalent as they have their origin in a common concept.
Although the conceptual development for this derivation is inspired by Landau and Lifshitz
the subscript notation will be largely that employed by Misner, Thorne, and Wheeler
16
. Tempting
as it is to use the coordinate free geometry of these authors, the concept of taking moments at this
point is most easily understood within the context of a coordinate representation so for the moment
we will keep that approach. In a Lorentz coordinate system, Landau and Lifshitz give components
of the 4velocity of a particle as
25
The Virial Theorem in Stellar Astrophysics
3 0 .
ds
dx
u ⋅ ⋅ ⋅ = α =
α
α
, 2.3.3
where c ) c / v 1 ( c
dt
2 / 1 2 2
γ = − =
ds
. Note this is a somewhat unconventional definition of γ.
The components of the energymomentum tensor are
( )
dt
ds
u cu
β α αβ
ρ = ℑ , 2.3.4
and specifically ,
j
2
j 0
u c iρ = ℑ
which are clearly symmetric in α and β. Since , the trace of equation (2.3.4) is 1 u
3
0
2
− =
∑
= α
α
γ ρ − = ℑ
∑
= α
αα
2
3
0
c . 2.3.5
In terms of the threedimensional components the conservation law expressed by equation (2.3.1)
can be written as
0
x t
) u (
c
x t ic
1
3
1 k k
jk j
3
1 k k
jk 0 j
=
∂
∂ℑ
+
∂
ρ ∂
=
∂
∂ℑ
+
∂
∂ℑ
∑ ∑
= =
. 2.3.6
Substituting for the components of and repeating the algebra of earlier derivations we have ℑ
2.3
( )
∑
∫ ∫
=
γτ + + Ω = γ
ρ
3
1 j
V V
j j
dV T dV / x x
dt
d
2 dt
d
, 2.3.7
where τ is the kinetic energy density of relativistic particles
17
. Again, using the conservation of
mass arguments in Chapter I, this becomes
( )
∑
∫ ∫
=
γτ + + Ω = γ ρ
3
1 j
V V
j j
2
2
2
1
dV T dV / x x
dt
d
. 2.3.8
where, if we define the volume integral on the left to be the relativistic moment of inertia, we can
write
( )
∫
γτ + + Ω =
V
r
2
2
2
1
dV T I
dt
d
. 2.3.9
In the low velocity limit γ → 1 so that I
r
→ I and we recover the ordinary Lagrange's identity. In the
relativistic limit as γ → 0 we recover for stable systems the Chandrasekhar result that T + Ω = 0
(i.e., the total energy E = 0). Thus, it is fair to say that equation (2.3.9) is an expression of the
Lagrange's identity including the effects of special relativity.
It is worth noting by analogy with section 1 that the tensor relativistic theorem can be
derived by taking the outer or 'tensor' product of the spacelike position vector with equation (2.3.6)
and integrating over the volume. Following the same steps that lead to equation (N.2.3.3), we get
∫ ∫
= ℑ −
∂
ρ ∂
V
ij
V
j
i
0 dV dV
t
) u (
x c . 2.3.10
26
The Virial Theorem in Stellar Astrophysics
Since is symmetric, we can add equation (2.3.10) to its counterpart with the indices
interchanged, and get
ij
ℑ
∫ ∫
= ℑ −


.

\

∂
ρ ∂
+
∂
ρ ∂
V V
ij
i
j
j
i
0 dV 2 dV
t
u (
x
t
) u (
x c , 2.3.11
or finally
∫ ∫
ℑ =


.

\

γ ρ
V
ij
V
i j
2
2
2
1
dV dV / ) x x ( [
dt
d
. 2.3.12
The integral on the left can be viewed on the relativistic moment of inertia tensor while the right
hand side is just the volume integral of the components of the energy momentum tensor. Following
the prescription used to generate equation (2.3.11), but subtracting the transpose of equation
(2.3.11) from itself, yields
∫
=


.

\

∂
ρ ∂
−
∂
ρ ∂
V
i
j
j
i
0 dV
t
u (
x
t
) u (
x , 2.3.13
which becomes by application of Leibniz’s law
0 dV
t
u (
x
t
) u (
x
dt
d
V
i
j
j
i
=
(
(
¸
(
¸


.

\

∂
ρ ∂
+
∂
ρ ∂
γ
ρ
∫
, 2.3.14
and is the relativistic form of the expression for the conservation of angular momentum obtained in
section 1, equation (2.1.8).
4. General Relativity and the Virial Theorem
The development of quantum theory and the formulation of the general theory of relativity
probably represent the two most significant advances in physical science in the twentieth century.
In light of the general nature and wide applicability of the virial theorem it is surprising that little
attempt was made during that time to formulate it within the context of general relativity. Perhaps
this was a result of the lack of physical phenomena requiring general relativity for their description
or possibly the direction of mathematical development undertaken for theory itself. For the last
twenty years, there has been a concerted effort on the part of relativists to seek coordinatefree
descriptions of general relativity in order to emphasize the connection between the fundamental
geometrical properties of space and the description of associated physical phenomena. Although
this has undoubtedly been profitable for the development of general relativity, it has drawn
attention away from that technique in theoretical physics known as 'moment analysis'. This
technique produces results which are in principle coordinate independent but usually utilize some
specific coordinate frames for the purpose of calculation.
27
The Virial Theorem in Stellar Astrophysics
Another point of difficulty consists of the nature of the theory itself. General relativity, like
so many successful theories, is a field theory and is thus concerned with functions defined at a
point. Virtually every version of the virial theorem emphasizes its global nature.
†
That is, some
sort of symmetrical volume is integrated or summed over to produce the appropriate physical
parameters. This difference becomes a serious problem when one attempts to assign a physical
operational interpretation to the quantities represented by the spatial integrals. The problem of
operational definition of macroscopic properties in general relativity has plagued the theory since
its formulation. Although continuous progress has been made, there does not exist any completely
general formulation of the virial theorem within the framework of general relativity at this time.
This certainly is not to say that such a formulation cannot be made. Indeed, what we have seen so
far should convince even the greatest skeptic that such a formulation does exist since its origin is
basically that of a conservation law (see footnote at the end of the chapter). Even the general theory
of relativity recognizes conservation laws although their form is often altered. Let us take a closer
look at the origin of some of these problems. This can be done by taking into account in a self
consistent manner in the Einstein field equations all terms of order 1/c
2
. This is the approach
adopted by Einstein, Infield, and Hoffman
19
in their approach to the relativistic nbody problem and
successfully applied by Chandrasekhar
20
to hydrodynamics. Although more efficient approximation
techniques exist for the calculation of higher order relativistic phenomena such as gravitational
radiation, this time honored approach is adequate for calculating the first order (i.e. c
2
) terms
commonly known as the postNewtonian terms.
††
During the first half of 1960s, in studying the
hydrodynamics of various fluid bodies, Chandrasekhar developed the virial theorem to an
extremely sophisticated level. The most comprehensive recognition of this work can be found in his
excellent book on the subject
9
. One of the highlights not dealt with in the book are his efforts to
include the firstorder effects of general relativity. In an impressive and lengthy paper during 1965,
Chandrasekhar developed the postNewtonian equations of hydrodynamics including a formulation
of the virial theorem
20
. It is largely this effort which we shall summarize here. One of the
fundamental difficulties with the general theory of relativity is its nonlinearity. The physical
properties of matter are represented by the geometry of space and in that turn determines the
geometry of space. It is this nonlinearity that causes so much difficulty with approximation theory
and with which the Einstein, Infield, Hoffman theory (EIH) deals directly.
†__________________________________________
It is worth noting that in order to 'test' any field theory against observation, it is necessary to
compare integrals of the field quantities with the observed quantities. Even something as
elementary as density is always "observed" by comparing some mass to some volume. It is
impossible in principle to measure anything at a point. This obvious statement causes no trouble as
long as we are dealing with concepts well within the range of our experience where we can expect
our intuition to behave properly. However, beyond this comfortable realm, we are liable to
attribute physical significance and testability to quantities which are in principle untestable. The
result is to restrict the range of possibility for a theory unnecessarily.
††___________________________
For a beautifully concise and complete summary of the postNewtonian approximation, see Misner,
Thorne, and Wheeler, Gravitation (1973), w. H. Freeman & Co., San Francisco, Chapter 39.
28
The Virial Theorem in Stellar Astrophysics
The basic approach assumes that the metric tensor can be written as being perturbed slightly from
the flatspace or Euclidean metric so that
αβ αβ αβ
+ = h g g
) 0 (
, 2.4.1
where the h are small terms of the order of c
αβ
2
or smaller. This enables one to determine the
elements of the energymomentum tensor up to terms of the order of c
2
from its definition so that
αβ β α αβ
− + ε = ℑ g P u u ) P ( . 2.4.2
The Einstein field equations can be written in terms of the Ricci tensor and the energymomentum
tensor as
( )
αβ αβ αβ
− ℑ
π
− = ℜ g J
c
G 8
2
1
4
, 2.4.3
where J is just the trace . The Ricci tensor is essentially a geometric tensor and contains
information relating to the metric alone.
αβ
ℑ
The EIH approximation provides a prescription for solving the field equations in various
powers of 1/c
2
given the information concerning
αβ
ℑ and . In general the procedure determines
the metric coefficients to one higher order than was originally specified. This procedure can be
repeated but there remain some unsolved problems as to convergence of the scheme in general. At
any point one may use the perturbed metric and the prescription for obtaining the equations of
motion to generate a set of perturbed equations of motion. The relativistic prescription that free
particles follow geodesic paths is logically equivalent to stating the fourspace divergence of the
energymomentum tensor is zero. That is
αβ
g
0 = ℑ •
αβ
. 2.4.4
Indeed it is this condition that in the flatspace metric yields of the EulerLagrange equations of
hydrodynamic flow. It was this condition that we needed in section 3 to obtain a form of the virial
theorem appropriate for special relativity. Unfortunately the process of taking the divergence looses
one order of approximation and thus it is not possible to go directly from the perturbed metric to the
equations of motion and maintain the same level of accuracy. One must first pass through the field
equations and the EIH approximation scheme. In order to follow this prescription one must first
start with an approximation to the metric g . Here it is traditional to invoke the principle of
equivalence that requires that
αβ
21
) c / 1 (
c
2
3
2
< + δ
Ψ
− =
αβ αβ
O h . 2.4.5
With this as a starting point one may proceed with the approximation scheme and obtain the
equations of motion. Like many approximation schemes the mathematical manipulations are
formidable and physical insight easily lost. However, progress is rapid in a field such as this and
what was an original research effort by Chandrasekhar in 1965 becomes a 'homework' problem for
Misner, Thorne and Wheeler
16
in 1972 (e.g. M.T.W. exercise 39.13.) Although it is contrary to the
spirit of this book to quote results without derivation, I find in this case I must. We have laid
neither a sufficient mathematical framework nor developed the general theory of relativity
sufficiently to present the derivation in detail without consuming excessive space. Instead, consider
29
The Virial Theorem in Stellar Astrophysics
the types of effects we might expect general relativity to introduce and see if these can be identified
in the resulting equations of motion.
First, energy is matter and therefore its motion must be followed in the equations of motion
as well as that of matter. This is really a consequence of special relativity but insofar as this 'added'
mass affects the metric, we should find its effects present. The distortion of space also changes or at
least complicates what is meant by a volume and thus it is useful to define a density ρ* which
obeys a continuity equation
0 ) (
t
= ρ • ∇ +
∂
ρ ∂
∗
∗
u ,
2.4.6
where  
2 2
0
c / ) 3 u ( 1 Ψ + + ρ = ρ
∗
.
For purposes of simplification, Chandrasekhar finds it convenient to define a slightly altered form
of the density which explicitly contains the internal energy of the gas and has a slightly altered
continuity equation
0
t t c
1
t
0
0
2
=
(
¸
(
¸

.

\

∂
ρ ∂
−
∂
Ψ ∂
ρ + σ • ∇ +
∂
σ ∂
u , 2.4.7
where . ] c / ) / P 2 u ( 1 [
2
0
2
0
ρ + Π + Ψ + + ρ = σ
Here is the internal energy of the gas and P is the local pressure. It is worth noting that ρ
2
c / Π
o
is
the density one would find in the absence of general relativity but where relativistic effects are
important it is essentially a nonobservable quantity since one can not devise a test for measuring it.
The general theory of relativity is a nonlinear theory and thus we should expect terms to
appear which reflect this nonlinearity. They will be of different types. Firstly one should expect
effects of the Newtonian potential Ψ affecting the metric directly which in turn modifies Ψ. These
terms are indeed present but Misner, Thorne and Wheeler show that they can be represented by
direct integrals over the mass distribution
22
. Secondly, since the matter and the metric are
inexorably tied together, motion of matter will 'drag' the metric which will introduce velocity
dependent terms in the 'potentials' used to represent those terms.
Both these effects can indeed be represented by 'potentials' but not just the Newtonian
potential. Thus, various authors introduce various kinds of potentials to account for these nonlinear
terms. With this in mind the equations of motion as derived by Chandrasekhar become
23
( )   ( )   ( )
)
`
¹
¹
´
¦
∇ − ∇ • − ∇ • + ∇ − − Ψ
ρ
+ Ψ + ∇ + Ψ ∇ ρ −
∂
σ ∂
] )[ ( ) ( 4 4
dt
d
c
P c / 2 1
t
) (
2
1
2
1
2
1
2
7
2
0 2
0
Y Y u Y u Y Y u
u
  0
Gc 2
1
2 2
2
= Φ ∇ Ψ ∇ + Ψ ∇ Φ ∇
π
+ . 2.4.8
30
The Virial Theorem in Stellar Astrophysics
Here, the various potentials which we have introduced can be defined by the fact that they satisfy a
Poisson's equation of the form
u Y
0
2
0
2
0
2
G 4
G 4
G 4
ρ π − = ∇
ϕ ρ π − = Φ ∇
ρ π − = Ψ ∇
, 2.4.9
where ) / P ( u
0 2
3
2
1
2
ρ + Π + Ψ + = ϕ .
Y is a vector potential whose source is the same as that of the Newtonian potential weighted
by the local velocity field. Similarly, Φ is a scalar potential whose source is again that of the
Newtonian potential but weighted by a function ϕ related to the total internal energy field.
Expansive as the equations of motion are we may still derive some comprehension for the
meaning of the various terms in equation (2.4.8). The first two terms are basically Newtonian,
indeed neglecting the contribution to the mass from energy σ = ρ
o
and they are identical to the first
term of the NewtonianEulerLagrange equations of hydrodynamic flow. The first part of the third
term is just the pressure gradient and thus also to be expected on Newtonian grounds alone. The
remaining contribution to the pressure gradient results from the space curvature introduced by the
presence of the matter and is perhaps the most likely relativistic correction to be expected. The
remaining tensors are the nonlinear interaction terms alluded to earlier. The lengthy expression in
braces contains the effects of the 'dragging' of the metric by the matter and the induced velocity
dependent terms. The last term represents the direct effect of the matterenergy potentials on the
metrics and this effect, in turn, is propagated to the potentials themselves.
Having obtained the equations of motion for the system, the procedure for obtaining the
general relativistic form of Lagranges' identity is the same as we have used repeatedly in earlier
sections. For simplicity, we shall compute the scalar version of Lagranges' identity by taking the
inner product of equation (2.4.8) with the position vector r .We should expect this procedure to
yield terms similar to the classical derivation but with differences introduced by differences
between ρ
o
, ρ*, and σ. In addition we shall take our volumes large enough so that volume integrals
of divergence vanish. In this regard it is worth noting that if volume contains the entire system, then
by the divergence theorem
∫ ∫
= • ∇ = ∇
V V
3
1
0 dV ) A ( AdV 1 . 2.4.10
Thus, by integrating the equations of motion over the volume yields
( )
∫ ∫
≡ =
)
`
¹
¹
´
¦
(
¸
(
¸
∇ − − Ψ
ρ
+ σ
V V
2
0
dV
dt
d
0 dV
2
7
4
c dt
d
K Y Y u u , 2.4.11
after noting that remaining integrals in the braces { }of equation (2.4.8) can be integrated by parts
to zero. Equation (2.4.11) is a statement of conservation of linear momentum. This is a useful result
for simplifying equation (2.4.8). Now we are prepared to write down Lagrange's identity by letting
31
The Virial Theorem in Stellar Astrophysics
the integral of equation (2.4.11) be the local linear momentum density K and taking the scalar
product of r with equation (2.4.8).
  ( )  
( ) 0
c
2
) ( )] ( [
c
4
) c / 2 1 ( P
dt
d
2
2
1
2
2
0
= Φ ∇ • + Ψ ∇ • ϕ
ρ
−
∇ − ∇ • • − ∇ • •
ρ
+ Ψ + + Ψ ∇ • ρ − •
r r
Y Y u r Y u r
K
r
. 2.4.12
After multiple integration by parts and liberal use of the divergence theorem, this becomes
( ) ( )  
∫ ∫
− > Φ < + + Ψ + + Ω + = •
V
4
7
2
2
V
4
c
1
PdV c / 2 1 3 T 2 dV
dt
d
Z  Y W
4
1
K r , 2.4.13
where
dV
dV
dV u
dV
dV u T
V
0
V
0
V
2
0
V
0 2
1
V
2
2
1
∫
∫
∫
∫
∫
• ρ =
Ψ ϕ ρ >= Φ <
Ψ ρ =
Ψ ρ − = Ω
σ =
Y u Y
W , 2.4.14
and
  
dV ' dV
'
) ' ( ' ) ' (
V ' V
3 0 0
∫ ∫
¦
)
¦
`
¹
¦
¹
¦
´
¦


.

\

−
− • − •
ρ′ ρ =
r r
r r u r r u
Z .
This can be made somewhat more familiar if we rewrite the left hand side of equation (2.4.13) so
that
( ) ( )  
  Z Y W
4
1
2
7
2
V
2
V
2
7
0
2
V
2
2
2
2
1
4
c
1
PdV ) c / 2 1 ( 3 T 2 dV r 4
dt
d
c
1
dV r
dt
d
− − > Φ < + +
Ψ + + Ω + = ∇ • − • − • Ψ ρ +


.

\

σ
∫ ∫ ∫
Y Y r u r
. 2.4.15
The first term on the left hand side of equation (2.4.15) is just
2
2
2
1
dt
I d
in the Newtonian
limit. The second term arises from the correction to the metric resulting from the potential and the
'dragging' of the metric due to internal motion. The first two terms on the right are just what one
would expect in the Newtonian limit while the next term can be related to the total internal energy.
32
The Virial Theorem in Stellar Astrophysics
This term contains a relativistic correction resulting directly from the change in metric due to the
presence of matter. The remaining terms are all energy like and the first two (W and <Φ>)
represent relativistic corrections arising from the change in the potential caused by the metric
modification by the potential itself. The last two involve metric dragging.
We have gone to some length to show the problems injected into the virial theorem by the
nonlinear aspects of general relativity. Writing Lagrange's identity as in equation (2.4.15)
emphasizes the origin of the various terms  whether they are Newtonian or Relativistic. Although
terms to this order should be sufficient to describe most phenomena in stellar astrophysics, we can
ask if higher order terms or other metric theories of gravity provide any significant corrections to
the virial theorem. The Einstein, Infield Hoffman approximation has been iterated up to 2 1/2
times
24, 25
in a true tourdeforce by Chandrasekhar and coworkers looking for additional effects.
At the 2 1/2 level of approximation, radiation reaction terms appear which could be significant for
nonspherical collapsed objects which exist over long periods of time. Using a parameterized
version of the postNewtonian approximation, Ni
26
has developed a set of hydrodynamic equations
which must hold in nearly all metric theories of gravity and that depend on the values (near unity)
of a set of dimensionless parameters. This latter effort is useful for relating various terms in the
equations to the fundamental assumptions made by different theories.
Perhaps the most obvious lesson to be learned from the EIH approach to this problem is that
continued application of the theory is not the way to approach the general results. However, of
some consequence is the result that conservation laws for energy, momentum, and angular
momentum exist and are subject to an operational interpretation at all levels of approximation.
Thus it seems reasonable to conjecture that these laws as well as the virial theorem remained well
posed in the general theory.
5. Complications: Magnetic Fields, Internal Energy, and Rotation
The full power and utility of the virial theorem does not really become apparent until one
realizes that we need not be particularly specific about the exact nature of the potential and kinetic
energies that appear in the earlier derivations. Thus the presence of complicating forces can be
included insofar as they are derivable from a potential. Similarly as long as the total kinetic energy
can be expressed in terms of energies arising from macroscopic motions and internal thermal
motions, it will be no trouble to express the virial theorem in terms of these more familiar
parameters of the system. One may proceed in just this manner or return to the original equations of
motion for the system. We shall discuss both approaches.
In Chapter I, we derived the EulerLagrange equations of hydrodynamic flow. These
equations of motion are completely general and are adequate to describe the effects of rotation and
magneticfields if some care is taken with the coordinate frame and the ‘pressure tensor’P. With
this in mind, we may rewrite equation (1.1.4), noting that the left hand side is a total time derivative
33
The Virial Theorem in Stellar Astrophysics
and that the pressure tensor can be explicitly split to include the presence of large scale
electromagnetic fields. Thus
( )
∫
−
ρ
− ℑ + • ∇
ρ
− Φ −∇ =
∂
∂
dv ) (
1 1
t
g
u v S
u
P , 2.5.1
where the tensor P
g
refers to the gas pressure alone and the tensor ℑ represents the Maxwell
stress tensor for the electromagnetic field which has components
27
∑ ∑
δ − + δ − = ℑ
k
k k
k
ij 2
1
j i k k ij 2
1
j i ij
B H B H D E D E , 2.5.2
or in dyadic notation
] [ ] [
2
1
2
1
B H HB D E ED • − − • − = ℑ 1 1 . 2.5.3
For almost all cases in astrophysics, it is appropriate to ignore electrostriction and
magnetostriction effects which complicate the relationships between E and D and B and H. In the
absence of these body forces, the divergence of equation (2.5.3) yields
2.4
( ) (   ) { } H H D D D × ∇ × + × ∇ × − ρ
π
= ℑ • ∇
2
e
c
4
1
, 2.5.4
which is just the Lorentz force on the medium. It is useful when considering a configuration in
uniform rotation to transform the problem into a corotating coordinate frame. This enables one to
see the effect of macroscopic mass motion explicitly in the formalism and thus assess its interaction
with other large scale properties of the system. In addition such systematic motion is represented by
the stream velocity u in the collision term of the Boltzmann equation making its meaning clearer.
Therefore, just as we have separated the effects of the electromagnetic field from the pressure
tensor, let us explicitly represent the effects of rotation.
In transforming the inertial coordinate frame to a noninertial rotating frame attached to the
system, we must allow for temporal changes in any vector seen in one frame but not the other.
Goldstein
28
gives a particularly lucid account of how this is to be accomplished by use of the
operator
× +

.

\

=

.

\

w
inertial  non inertial
dt
d
dt
d
, 2.5.5
where w is the angular velocity appropriate for the point function upon which the operator acts (i.e.
the angular velocity of the rotating frame).
If we let
) ( r w u × + = w , 2.5.6
then the equations of motion for uniform rotation (i.e., w = constant) become
∫
ρ
− ℑ + • ∇
ρ
− Φ −∇ = × × + × +
V
d (
1
)] ( [
1
) ( ) ( 2
dt
d
v S r w
w
w) w w w P . 2.5.7
It can be shown that for uniform rotation the term
2.5
.
)] ( ) [( ) (
2
1
r r r × • × ∇ = × × w w w w , 2.5.8
34
The Virial Theorem in Stellar Astrophysics
so that it may be combined with the gravitational potential and
2
2
1
) ( r × w may be considered a
'rotational potential'. The term 2 is known as the "coriolis force". Since both the pressure
tensor and Maxwell tensor are normally "fixed" to the body, we should expect their formulation in
the rotating frame to be simpler. It is difficult to proceed much further with the equations of motion
without making some simplifying assumptions. The most helpful and also reasonable of these is to
assume that all collisions processes are isotropic in space. This has two results. Firstly, the collision
term on the right hand side of the equation of motion averages to zero when integrated over all
velocity space. Secondly, the gas pressure tensor P
) ( w × w
g
becomes diagonal with all elements equal and
can thus , where P is the scalar gas pressure. This effectively guarantees the existence of
a scalar equation of state which will be useful later when relating P to the internal energy. Since
almost all astrophysical situations relate to plasmas, we may consider the configurations to have a
very large conductivity and can therefore neglect the contribution to the Maxwell Field tensor of
electric fields. Using these assumptions, and equations (2.5.4) and (2.5.8), the equation of motion
becomes
P
g
= • ∇ P
  (   H H r w
w
× ∇ ×
πρ
− ρ ∇ − × + Φ −∇ = × +
4
1
/ P ) ( ) ( 2
dt
d
2
w w ) . 2.5.9
As we have done before we shall multiply the equations of motion by ρr, and integrate over the
volume and generate a general tensor expression for Lagrange's identity including rotation and
magnetic fields.
2.6
Thus,
( ) ( ) dV ] [
4
1
PdV ) ( n 2 2
dt
d
V V
2
2
2
2
1
H H r r × ∇ ×
π
− ∇ ω − • + + = • +
∫ ∫
 U T L I 1 1w w w . 2.5.10
where I , T, and U are moment of inertia, kinetic and potential energy tensors respectively. L is an
angular momentumlike tensor
2.6
. In order to simplify the last two terms it is worth noting that they
are derived from the divergence of two tensors. So, we may use the chain rule followed by the
divergence theorem to simplify these terms. Thus
dV ) ( ) ( dV ) ( [ dV ) (
V V V
ℑ + • ∇ − ℑ + • ∇ = ℑ + • ∇
∫ ∫ ∫
P P P r r r . 2.5.11
The divergence theorem guarantees that the first integral can be written as a surface integral
and if the volume is taken to be large enough can always be made to be zero. However, since
magnetic fields always extend beyond the surface of what one normally considers the surface of the
configuration, we will keep these terms for the moment. Thus
S P P ≡ • ℑ + = ℑ + • ∇
∫ ∫
S r r d ) ( dV ) ( [
V S
, 2.5.12
or, in terms of the components, the surface terms can be written as
∫ ∫ ∫
∑
−
π
−
π
=
S S
i j 0 i
2
j
S
k
k k i j ij
dS x P dS H x
8
1
dS H H ( x
4
1
S . 2.5.12a
Keeping in mind that ∇ (i.e., a second rank tensor with components δ 1 = r
ij
), the second
integral in equation (2.5.11) becomes
35
The Virial Theorem in Stellar Astrophysics
( )
∫ ∫ ∫
π
+ π + = ℑ + •
V V
2
V
dV
4
1
dV ) 8 / H P ( dV ) ( HH 1 1 P . 2.5.13
Defining ( )
∫
π
≡
V
dV
8
1
HH M , 2.5.14
we arrive at the final tensor form for Lagrange's Identity, including rotation and magnetic fields.
  S U M  T L
I
+


.

\

π + + ω − • + + = • +
∫
V
2 2
2
2
2
1
dV ) 8 / H P ( ) ( n 2 2
dt
d
1 1 1w w w , 2.5.15
where L . ( )
∫
ϖ − ϖ ρ =
V
j i k j i k ijk
dV r r r r
We have suffered through the tensor derivation in order to show the complete generality of
this formalism. The tensor component equations are essential for investigating nonradial
oscillations and other such phenomena which cannot be represented by a simple scalar approach.
However, it is easier to appreciate the physical significance of this approach by looking at the
scalar counterpart of equation (2.5.15). In section 1, we pointed out that the scalar form is derived
by taking "inner" products of the position vector with the equation of motion while the tensor virial
theorem involves "outer" or tensor products. One may either repeat the derivation of equation
(2.5.15) taking "inner" products or contract, the component form of equation (2.5.15) over indices i
and j.
The contraction of the tensors I, T , M, U, and 1 yield the moment of inertia about the
origin of the coordinate system I, the kinetic energy due to internal motion T, the total magnetic
energy M , the total potential energy Φ respectively. Some care must be taken in contracting the
tensors L, and 1w. From the definition of L, in equation (N2.6.3), it is clear that the contracted form
of that expression can be written as
( )
∫ ∫
• = × ρ • − = × • ρ
V V
l w w w dV 2 dV 2 dV ) ( 2
V
w r w r
∫
, 2.5.16
where l is the net volume angular momentum density on the material due to coriolis forces. We can
again choose our rotating frame so that ldV is zero and this term must vanish from the contracted
equation. The simplest method for deriving the value of the contracted form of in
equation (2.5.15) is again to examine the contracted form of the term giving rise to it. Since
w = w x r, we can expand the left hand side of equation (2.5.16) by means of identities relating to
the vector triple product [see note 2.6, specifically equation (N2.6.4)], and obtain
∫
) ( 1 1 w w w  •
( )      
∫ ∫ ∫ ∫
ℜ = ρ = • × ρ = × • ρ = × × • ρ
V V V V
2
2 dV v dV dV dV v r v r r r w w w w , 2.5.17
where ℜ is just the total energy due to rotation. Substitution of the contraction of these tensors into
equation (2.5.15) yields a much simpler result
36
The Virial Theorem in Stellar Astrophysics
∫
∑
+ + + Ω − + ℜ + =
V
i
ii
2
2
2
1
PdV 3 n ) T ( 2
dt
I d
S U M . 2.5.18
The contraction of the surface terms yield the scalar integrals
( )( ) ( ) S r S H H r d 8 / H P d
4
1
S
2
0 0 0
i
S
0 0 ii
• π + − • •
π
=
∫
∑
∫
S . 2.5.19
Here, the subscript "0" indicates the value of the variables on the surface. In practice these integrals
are usually small compared with the magnitude of the volume integrals found in the remainder of
the expression. So far, we have said little about the contribution of the volume integral of the
pressure. Since even in the tensor representation this term appears as a scalar, there was no loss of
generality in deferring the evaluation of the integral until now. Dimensional analysis will lead one
to the result that the pressure integral is an 'energylike' integral. From thermodynamic
consideration, we can write the internal "heat energy" U as
dV c
V
v
T U
∫
ρ = , 2.5.20
where c
v
is the specific heat of constant volume and is the temperature. We also know that the
kinetic energy associated with the material is
( ) P c c Nk
2
3
v p 2
3
2
3
− − ρ = = T T E . 2.5.21
Combining these two equations we get the total internal "heat energy" as
( ) ( )
∫ ∫
−
ρ
=
−
=
V v p V v p
3
2
1 c / c
dV
1 c / c
dV E
U . 2.5.22
It is traditional to let γ = c
p
/c
v
and if we let this be constant throughout the volume and neglect the
surface term in equation (2.5.18), then we can write the scalar form of Lagrange's identity as
U M ) 1 ( 3 ) T ( 2
dt
I d
2
2
2
1
− γ + Ω − + ℜ + = . 2.5.23
At the beginning of this discussion, I said that one could either derive Lagrange's identity
from the equations of motion or from careful consideration of meanings of potential and kinetic
energy. For example, the first and last terms on the right hand side of equation (2.5.23) are just the
contribution to the total kinetic energy of the system from macroscopic motions, rotation and
thermal motion respectively. The remaining terms are just a specification of the nature of the total
potential energy. Thus, by realizing that 3(γl)U is just twice the kinetic energy due to thermal
motion we could have written equation (2.5.23) down immediately. However, it is unlikely that this
could have been done for equation (2.5.15). Since the variational form of this equation will be
useful in the next chapter, our efforts have not been wasted.
Before leaving this section on complicating phenomena, there is one last aspect to be
investigated. In Chapter I, section 3, we noted that the inclusion of velocity dependent forces such
as frictional force do not alter the results of the virial theorem as they can be averaged to zero given
sufficient time. This result was apparently first noted by E. H. Milne
29
in 1925. Although no
modification is made to the virial theorem some effects can be seen in Lagrange's identity and so,
37
The Virial Theorem in Stellar Astrophysics
let us take a moment to recapitulate these arguments. If we have a volume force which can be
derived from a "friction" tensor, then one would add terms to the equations of motion of the form
f • ρ = w
f
f , 2.5.24
which make the following contribution to the tensor form of Lagrange's identity:
( ) dV
dr
d
dV dV
V
2
1
V V
f
f f •
(
¸
(
¸
ρ = • ρ =
∫ ∫ ∫
rr rw rf . 2.5.25
If f were indeed constant throughout the configuration, the righthand side of equation (2.5.25)
would just become
f
I
:

.

\

dt
d
2
1
, 2.5.26
and Lagrange's identity in its full generality would be
( )   S U M T L f
I I
+ + γ − + − • + + − = • + 
.

\

+ ] ) 1 [( n 2 2
dt
d
dt
d
2
1
2
2
2
1
M U 1 1 1 w w w w : . 2.5.27
6. Summary
In this chapter we have continued the development of the virial theorem as it appears in
more contemporary usage. The tensor virial theorem is a more general form of Lagrange's identity
which when averaged over time provides rather general expressions for the coordinated behavior of
some energy like tensors of the system. Further insight into the nature of this process is discovered
in the second section where we find that taking higher order spatial moments of the equation of
motion is equivalent to recovering the information selectively lost in the classical derivation of the
virial theorem. In principle this approach could be used in a prescription for the complete solution
of the equations of motion. However, it seems likely that in practice it would be more difficult than
implementing a direct numerical solution of the original equations themselves. The importance of
the method lies in the fact that such moment expressions for stable systems are normally rapidly
convergent. Thus, the largest amount of information can be recovered with the least effort.
In the next two sections, we considered the effects of a relativity principle on the
development of the virial theorem. We found that large velocities require large gravitational fields
to keep them in check and thus one might argue that a separate discussion of the effects of special
relativity is not warranted. However, there are at least two dynamically stable systems for which
this is not true (i.e., pure radiation spheres which approximate some models of super massive stars
and white dwarfs where lowmass, high velocity electrons, were kept in check by the highmass,
38
The Virial Theorem in Stellar Astrophysics
low velocity nucleons). In addition, Lagrange's identity is applicable to systems which are not in
equilibrium and hence may be relativistic. For this reason, we have developed the special and
general relativistic versions of the theorem separately and will return in the last chapter to discuss
some specific applications of them. I have attempted throughout the chapter to emphasize the
similarity of the derivation of the virial theorem and particularly in section 3, the perceptive reader
may have noticed that the derivation is equivalent to carrying out
∫
•
V
r 0 dV = ℑ • , 2.6.1
where r is a four vector in the Lorentz metric. However, we ignored all contributions from the
timelike part. These would have been of the form
dV
x
t dV t
dt
d
V
j j
j 0
V
00 2
1
∫
∑
∫


.

\

∂
∂ℑ
= ℑ , 2.6.2
which after appropriate application of the divergence theorem, becomes
S
x
d
dt
d c
dt
dE
dV
dt
d
S
2
V
•
γ
ρ
− = =


.

\

ε
∫ ∫
. 2.6.3
Because of the linear independence of the time and space coordinates, this is not a new
result but rather an expression of the conservation of energy. Essentially it states that the time rate
of change of the total energy of the system equals the momentum flux across the system boundary
multiplied by c
2
. When the metric is no longer the 'flat' Lorentz metric as in the case in section 4,
things are no longer simple. It is this loss of simplicity which caused me to stray from the more
rigorous approaches of other sections. Thus rather than stress the manipulative complexity of the
postNewtonian approximation, I have attempted to provide physical motivation from the existence
of terms arising in the equations of motion that results from the nonlinear nature of general
relativity. The derivation presented in this section follows exactly the prescription of earlier
sections, but for simplicity I presented the development of the scalar version of the virial theorem
only. Sticking with the postNewtonian approximation avoided some difficult problems of
uniqueness and interpretation.
However, I remain convinced that a general formulation of Lagrange's identity and the virial
theorem which is compatible with the field equation of Einstein exists and its formulation would be
most rewarding
†
.
†_______________________________________
Since the Pachart edition of this effort was written, there has been some significant
progress in this area. Bonazzola
30
, formulated the virial theorem for the spherical, stationary case
in full general relativity. He notes that in the nonstationary, nonspherical case, the existence of
gravitational radiation destroys any strict equivalence between the general relativistic and
Newtonian cases so that there can be no unique formulation of the virial theorem. Thus any
formulation of a viriallike relation will depend on the specific nature of the configuration. This
approach was extended by Vilain
31
who applied a similar formulation of a general relativistic virial
theorem to the stability of perfect fluid spheres.
39
The Virial Theorem in Stellar Astrophysics
For most stellar astrophysical applications the postNewtonian result is probably sufficient. In the
last section of this chapter some of the powers of the virial theorem to deal with difficult situations
became apparent. The results which have been generalized to include additional effects are not the
result of any new physical concepts. Rather, they are the result of the specific identification of the
physical contributions to the system made by such attributes as magnetic fields and macroscopic
motion. Although I included only magnetic fields throughout most of the discussion, the inclusion
of electric fields in the Maxwell field tensor makes it clear how to proceed should they be present.
Lastly we looked again at velocity dependent forces not so much with an eye to their effect on the
virial theorem, but rather with a view to their persistent presence in the variational form of
Lagrange's identity. The presence of all these complicating aspects is included only to make their
interplay explicit. The basic theorem must hold; all the rest is done to glean more insight.
40
The Virial Theorem in Stellar Astrophysics
Notes to Chapter 2
2.1 The term on the left of equation (2.1.2) becomes
dV
dt
d
dt
d
dV
dt
d
dt
d
dV
dt
d
dV
dt
d
V V V
2
2
V
∫ ∫ ∫ ∫
ρ − 
.

\

ρ = ρ = ρ
r r r r
r
r
r
u
r . N2.1.1
The third term can further be simplified So that
( ) dV
dt
d
dV
dt
d
dV
dt
d
dt
d
V
2
2
2
1
V
2
2
2
1
V


.

\

ρ = ρ = 
.

\

ρ
∫ ∫ ∫
rr rr
r r
r . N2.1.2
In obtaining the third term in equation (2.1.4) we have assumed that the volume V is large
enough to contain all matter and the conservation of mass argument explicitly developed in
Chapter I has the form
( )
∫ ∫ ∫
Φ ∇ ρ + ρ =


.

\

ρ
V V V
2
2
2
1
dV dV dV
dt
d
r uu rr . N2.1.3
2.2 If one writes equation (2.1.6) in component form, one gets
ij ij
ij
U T
I
n 2
dt
d
2
2
2
1
+ = , N2.2.1
where in Cartesian Coordinates the tensor components are
( ) dV ' dV ' ) x x )( x x ( ) ' ( ) (
dV u u
dV x x
2 n
j j i i
V ' V
2
1
ij
V
j i ij
V
j i ij
−
− ′ − ′ − ρ ρ =
ρ =
ρ =
∫ ∫
∫
∫
r r r r U
T
I
. N2.2.2
By contracting these tensors and letting the potential be that of the gravitational field, we recover
the scalar form of Lagrange's identity as given in Chapter I, equation (1.4.12).
2.3 Taking the scalar product of a spacelike position vector with equation (2.3.6) and
integrating over a volume sufficient to contain the system we get
∑∑
∫ ∫
∑
=
∂
∂ℑ
=
∂
ρ ∂
j k k
jk
j
V V
j
j
j
0 dV
x
x dV
t
) u (
x c . N2.3.1
From the chain rule the second integral can be written as
( ) dV
x
x
dV x
x
dV
x
x
jk
j k j k j k
V k
j
jk j
k V k
jk
j
V
ℑ
∂
∂
− ℑ
∂
∂
=
∂
∂ℑ
∑∑ ∑∑ ∑∑
∫ ∫ ∫
. N2.3.2
The first integral on the right is just the integral of the divergence of over V and if the
volume is chosen to enclose the entire system the integral must vanish as on the surface
jk j
x ℑ
jk
ℑ 0 =
41
The Virial Theorem in Stellar Astrophysics
enclosing V (i.e., Gauss's Law applies). Since
ij j i
x / x δ = ∂ ∂ , the second integral
becomes . Equation (N2.3.1) is then
∑
∫
ℑ
j
V
jj
dV
∫
= ℑ
V
jj
0 dV
0
t
x
j
=

.

∂
∂
3
1 j
= ℑ +
V V
00
dV dV
) 1
1
−
γ
2
c
( ρ − ε − γτ −
2 2
c
∫
γτ + + Ω =
V
T
)
2 2 2
H c +
(D H ∇ − • ∇
2
1
)
0
( D D + • ∇ −
2
1
c
A G ) ( + × ∇ × +
( )
∑ ∑
∫
− ρ
∂
∂
j j
V
j j
dV u
t
x c . N2.3.3
Now,
\

from the orthogonality of the Lorentz frame and , so we
can write
∑ ∑
= = α
αα
ℑ − ℑ = ℑ
3
0
00 jj
∫ ∫
∑
γ ρ +


.

\

ρ
∂
∂
V
2
j
j j
0 c dV u x
t
c . N2.3.4
∫
e
With the sign convection is the negative of the total energy density, Bergmann
00
ℑ
17
, among
others, shows us that the "relativistic" kinetic energy density τ is given by
( c
2
γ ρ = τ
−
or ρ = γτ . N2.3.5 ρ −
2
c
Thus, using equation (2.3.3) to rewrit the first term of equation (N2.3.4), we get
( )
∫
∑
∫
= ρ +
(
¸
(
¸
γ
ρ
∂
∂
=
V
j j
2
1
3
1 j
V
0 dV c dV
dt
/ x x d
t
, N2.3.6 )
where ε is the potential energy density. Applying Leibniz’s law for the differentiation of definite
integrals
l8
to the first term in equation (N2.3.6) and rewriting the second one, we get
( )
∑
∫
(
¸
(
¸
γ
ρ
∂
∂
=
j j
2
1
3
1 j
V
dV dV
dt
/ x x d
t
. N2.3.7
2.4
( 
2
1
2
D c
4
1
− +
π
= ℑ 1 HH DD . N2.4.1
If we take the divergence of we get ℑ
) { } H H D H H H D D D D • + • + ∇ • + • ∇ + ∇ •
π
= ℑ • ∇
2 2 2
c ( c ) c ( ) ( ) (
4
1
, N2.4.2
and invoking Maxwell's laws that
e
ρ = • D ∇ and = • ∇ H , this becomes
) { } H H H H D D D • ∇ • + ρ + ∇ •
π
= ℑ • ∇
2 2
e
) c ( ) (
4
1
. N2.4.3
Making use of the vector identity
( ) A G G A G A G A ) ( ) ( ) ( ∇ • ∇ • + × ∇ × = • ∇ , N2.4.4
equation (N2.4.3) takes on the more familiar form
42
The Virial Theorem in Stellar Astrophysics
{ } )] ( [ c ) ( (
4
1
2
e
H H D D D × ∇ × + × ∇ × − ρ
π
= ℑ • ∇ . N2.4.5
2.5 By use of the vector identity equation (N2.4.4), it is clear that
( ) ( ) ( )   ( )  ( ) r r r r r × ∇ • × + × × ∇ × × = × ∇ w w w w w
2
2
1
, N2.5.1
but
( ) ( ) ( ) ) ( ) ( w w w w w • ∇ + • ∇ − ∇ • − ∇ • = × × ∇ r r r r r . N2.5.2
The constancy of w causes the second and fourth term to vanish, while the first third terms are
equal but of opposite sign. Again, since w is constant.
( )  ( ) ( )   ) (r r r r ∇ • × × = × ∇ • × w w w w . N2.5.3
In component form
( )   ( r r r × = ω =
∂
)
∂
ω = ∇ • ×
∑∑∑ ∑∑
w w
k j
i j k j k
jk
i
k j ijk
r
x
x
r ) (
A
A
e e , N2.5.4
where e
ijk
in a completely antisymmetric tensor of rank 3 sometimes called the Levi Civita tensor
density. Thus
( ) ( ) r r × × = × ∇ w w w
2
2
1
. N2.5.5
2.6
( )   ( )  dV
4
1
dV / P dV ) ( 2 dV
dt
d
V V
2
V V
∫ ∫ ∫ ∫
× ∇ ×
π
− ρ + × + Φ ∇ ρ − = × ρ + ρ H H r r w r
w
r w w  . N2.6.1
As in section 3, the first term on the left becomes
dV ) ( dV ) (
dt
d
dV
dt
d
V V
2
2
2
1
V
∫ ∫ ∫
ρ −


.

\

ρ = ρ ww rr
w
r . N2.6.2
The second term is more difficult to simplify. Let the local velocity field by defining a r w × = ϖ
local angular velocity field . Then, by expanding the resulting vector triple product in the second
term we can write
ϖ
( )    dV ( ) ( 2 dV 2 dV ) (
V V V
∫ ∫ ∫
• − • ρ = × × ρ = × ρ 2 ) w w w w ϖ ϖ ϖ r r r r r w r N2.6.3
or   L w w w 2 dV ) ( ) ( 2 dV ) (
V V
= − ρ • = × ρ 2
∫ ∫
rr r r w r ϖ ϖ ,
where the tensor L is an angular momentlike three index tensor representing the various
components of volume net angular momentum within the body. Thus, w• L is a kinetic energy like
tensor resulting from the net motions induced by the coriolis forces. However, since there can be no
net motions about any axis other than that defined by the total angular momentum of the body all
components of w L must be zero except those associated with the axis of rotation. In addition we
can choose our rotating frame w so that in that frame the net angular momentum is zero and all
•
43
The Virial Theorem in Stellar Astrophysics
contributions from the term w L will vanish. For the sake of generality we shall keep the term for
the present. The first term on the right of equation (N2.6.1) is given by equation (2.1.4). Since we
just went to some length to show that,
•
( ) ( ) r r × × = × ∇ w w w
2
2
1
, we shall use the earlier version and
an expansion of the vector triple product to evaluate the second term on the right.
V
∫
w
Thus
( )   dV dV ) ( dV ) ( dV
2
V V V
2
2
1
ω ρ − • ρ = × × ρ = × ∇ ρ
∫ ∫ ∫
rr r r r r r r w w w w w ,
N2.6.4
or ( ) dV dV dV
V
2
V V
2
2
1
∫ ∫ ∫
ρ ω −


.

\

ρ • = × ∇ ρ rr rr r r w w .
It is worth noting at this point that all terms dealt with so far involve only the volume integrals
and which are the same as the tensors defined in Section 1. Thus by combining
equations (N2.6.2), (N2.6.3), (N2.6.4), and (2.1.4), we may assess our progress in simplifying
equation (N2.6.1) so far [see equation (2.5.10)].
∫
ρ
V
dV rr
∫
ρ
V
dV ww
44
The Virial Theorem in Stellar Astrophysics
References
1 Rayleigh, L. (1903), Scientific Papers, Vol. 4, Cambridge, England, p. 491.
2 Parker, E. N. (1954), Phys. Rev. 96, pp. 16869.
3 . (1957), Ap. J. Suppl. 3. p. 51.
4 Chandrasekhar, S., and Fermi, E. (1953), Ap. J. 118, p. 116.
5 Lebovitz, N. R. (1961), Ap. J. 634, pp. 500536.
6 Chandrasekhar, S., and Lebovitz, N. R. (1962), Ap. J. 135, pp. 238, 248, 1032.
7 . (1962), Ap. J. 136, pp. 10371047.
8 Chandrasekhar, S. (1961), Hydrodynamic and Hydromagnetic Stability, Oxford University
Press, London. pp. 571581.
9 . (1969), Ellipsoidal Figures of Equilibrium, Yale University Press, Chapt. 2, pp.1537.
10 Chandrasekhar, S., and Lebovitz, N. R. (1962), Ap. J. 135, pp. 238247.
(1962), Ap. J. 136, pp. 10321036.
12 Chandrasekhar, S. (1962), Ap. J. 136, pp. 10691081.
13 Chandrasekhar, S., and Lebovitz, N. R. (1968), Ap. J. 152. pp. 293304.
14 Chandrasekhar, S. (1957), An Introduction to the Study of Stellar Structure, Dover Pub.,
Inc. , p. 424, eq. 81.
15 Landau, L. D. and Lifshitz, E. M. (1962), The Classical Theory of Fields, Trans. M.
Hamermesh, AddisonWesley Pub. Co., Reading, Mass. USA, pp. 9597.
16 Misner, C. W., Thorne, K. S., and Wheeler, J. A. (1973), Gravitation, W. H. Freeman & J
Co., San Francisco.
17 Bergmann, P. G. (1942), Introduction to the Theory of Relativity, PrenticeHall, Inc.,
New York. p. 92.
18 Sokolnikoff, I. S., and Redheffer, .(1958), Mathematics of Physics and Modern
Engineering, McGrawHill Book Co.. Inc., New York, p. 262.
19 Einstein, A., Infeld, E.. and Hoffman, (1938), Ann. Math 39, pp. 65100.
20 Chandrasekhar, S. (1965). Ap. J. 142, pp. 14881512.
21 Schiff, I. I. (1960), American Journal of Physics 28, p. 340.
22 Misner, C. W., Thorne, K. S., and Wheeler, J. A. (1973), Gravitation, W. H. Freeman &
Co. , San Francisco, p. 1082.
23 Chandrasekhar. S. (1965), Ap. J. 142, p. 1499, eq. 80.
24 Chandrasekhar, S., and Nutku, Y. (1969), Ap. J. 158, pp. 5579.
25 Chandrasekhar, S., and Esposita, P. (1970), Ap. J. 160, pp. 153179.
26 Ni. W. (1973), Ap. J. 181, pp. 957976.
27 Panofsky, W. K. H. and Phillips, M. (1955), Classical Electricity and Magnetism,
AddisonWesley Pub. Co., Reading, Mass. pp. 162.
28 Goldstein, H. (1959), Classical Mechanics, AddisonWesley Pub. Co., Reading. Mass.
pp.133.
29 Milne, E. A. (1925), Phil. Mag. S. 6, Vol. 50, pp. 409414.
30 Bonazzola, S. (1973), Ap. J. 182, pp. 335340.
45
The Virial Theorem in Stellar Astrophysics
31 Vilain, C. (1979), Ap. J. 227, pp. 307318.
32 Collins, G.W.,II, (1989) The Fundamentals of Stellar Astrophysics, W.H. Freeman and Co.,
New York, pp. 160167.
33 Chandrasekhar, S. (1987), Ellipsoidal Figures of Equilibrium, Dover Publications Inc.,
New York, pp.1537.
46
The Virial Theorem in Stellar Astrophysics
47
The Virial Theorem in Stellar Astrophysics
© Copyright 2003
III The Variational Form of the Virial Theorem
1. Variations and Perturbations and their Implications for the
Virial Theorem
Perturbation analysis is truly an old mechanism in which one explores the behavior of a
system in a known state by assuming there are small variations of the independent variables
describing the system and determining the individual variation in the independent variables. The
vehicle for determining the independent variable changes is found in the very equations which
describe the initial state of the system. The equations usually chosen for this type of analysis are
the equations of motion for the system. For example, consider the equations of motion for an
object moving under the influence of a point potential Φ.
Φ −∇ =
2
2
dt
d r
. 3.1.1
Assume a solution r
o
(t) is known, which satisfies equation (3.1.1) for a particular
potential Φ
o
. Since equation (3.1.1) is valid for any system where Φ is known, one could define
)
`
¹
δ + =
Φ δ + Φ = Φ
r r r
0
0
, 3.1.2
and equation (3.1.1) would require
) (
dt
) ( d
2
2
Φ δ + Φ −∇ =
δ + r r
. 3.1.3
However, since both ∇ and d/dt are linear operators, equation (3.1.3) becomes
Φ δ ∇ − Φ −∇ =
δ
+
2
2
2
0
2
dt
d
dt
d r r
, 3.1.5
but we already know that
0
2
0
2
dr
d
Φ −∇ =
r
, 3.1.6
so that by subtracting equation (3.1.5) from equation (3.1.4), we get
Φ δ −∇ =
δ
2
2
dr
d r
, 3.1.7
48
The Virial Theorem in Stellar Astrophysics
which we called the perturbed equations of motions where δΦ is the perturbation that involves
the perturbation δr. A short approach which leads to the same result is to "take the variation" of
equation (3.1.1) wherein the operator δ is not affected by time or space derivatives. This
technique "works" because the time and space operators in the equation of motion are linear,
hence any linear perturbation or departure from a given solution will produce the sum of the
original equations of motion on the perturbed equations of motion. In general, I shall use the
variational operator δ in this sense, that is, it represents a small departure of a variable from the
value it had which satisfies the equations governing the system. It is not necessary that one
perturbs the equations of motion in order to gain information about the system. Clearly any
equations which describe the structure of the system are subject to this type of analysis. Thus, if
taking variations of the equations of motion produces useful results, might not the variational
form of the moments of those equations also be expected to contain interesting information? It
was in this spirit that Paul Ledoux developed the variational form of the scalar virial theorem
l
,
and was able to predict the pulsational period of a star.
The variational approach yields differential equations which describe parameter
relationships for a system disturbed from an initial state. If that state happens to be an
equilibrium state, variational analysis of the equations of motion would yield a description of the
system motion about the equilibrium configuration. Variational analysis of spatial moments
could then be expected to yield macroscopic properties of that motion. This is indeed the case, as
Ledoux
1
demonstrated by determining perhaps the most obvious macroscopic property of such
motion, the pulsational period of the system. Chandrasekhar
2
found the tensor form of the virial
theorem useful in determining nonradial modes of oscillation of stars. In addition, he and Fermi
3
investigated the effects of a magnetic field on the pulsation of a star. An additional macroscopic
property closely connected with the pulsational period, with which this approach deals, is the
global stability of the system. We shall examine this aspect of the analysis later. For now, let us
be content with observing in some detail how the variational approach yields the pulsational
periods of stars.
2. Radial Pulsations for SelfGravitating Systems: Stars
In this section we shall use the virial theorem, to obtain an expression for the frequency
of radial pulsations in a gas sphere. The approach will be to apply a small variation to the virial
theorem and by making use of several conservation laws, obtain expressions for the variation of
the moment of inertia, kinetic energy, and potential energy as a function of time.
Remember from the earlier section that Lagrange's identity is
Ω + = T 2
dt
I d
2
2
2
1
. 3.2.1
49
The Virial Theorem in Stellar Astrophysics
In this form, no time averaging has been carried out and the equation must apply to a dynamic
system at any point in time. Now, consider a star with radius R. Let r be the distance from the
center of symmetry to any point in the configuration and or be the displacement of a point mass
from the equilibrium position r
o
. Conservation of mass requires that for a spherical shell of radius
) r ( m ) r r ( m
0 0
= δ + . 3.2.2
We wish to find the variations δI, δT, δΩ of the quantities I, T, Ω, from the equilibrium values I
0
,
T
0
, and Ω
0
. The variational form of the virial theorem then becomes
Ω δ + δ =
δ
T 2
dt
) I ( d
2
2
2
1
. 3.2.3
Since I was defined as the moment of inertia about the center of the coordinate system, we have
by definition that,
∫
=
M
0
2
) r ( dm r I . 3.2.4
Thus, we have
∫ ∫
δ + δ = δ
M
0
M
0
2
)] r ( dm [ r ) r ( rdm r 2 I . 3.2.5
Since, by the conservation of mass [equation (3.2.2)], δm(r) = 0 for all r, d(δm(r) ) = δ(dm(r) ) =0
for all r and the second integral of (3.2.5) vanishes leaving
∫
δ = δ
M
0
) r ( rdm r 2 I . 3.2.6
Now from equation (3.2.4), we have
2
r
) r ( dm
dI
= . 3.2.7
Since this must always be true, it is true at the equilibrium point r
o
. Therefore,
) r ( dm r dI
2
0 0
= . 3.2.8
So, to first order accuracy in r, we may rewrite equation (3.2.6) as
∫
δ
= δ
0
I
0
0
0
dI
r
r
2 I . 3.2.9
In a similar manner we may evaluate the variation of the gravitational potential energy
with respect to small variation in r
3.1
and obtain
∫
Ω
Ω
δ
= Ω δ
0
0
0
0
d
r
r
2 . 3.2.10
All that now remains to be determined in equation (3.2.3) is the variation of the total
kinetic energy T. To first order only the variation of the thermal kinetic energy will contribute to
equation (3.2.3).
3.2
50
The Virial Theorem in Stellar Astrophysics
) r ( dm ) 1 (
P
3 T 2
0
0
M
0
0
0
ρ
δρ
− γ
ρ
≅ δ
∫
. 3.2.11
In order to facilitate obtaining an expression for
0
/ ρ δρ we shall now specify a time
dependence for the pulsation about r
o
. For simplicity, let us assume the motion is simply
periodic. Thus, defining a quantity ξ as
t i
0
0
e
r
r
σ
ξ =
δ
= ξ , 3.2.12
where 2π/σ is the period of oscillation, we may rewrite the variations of I and Ω as follows:
¦
)
¦
`
¹
Ω ξ − = Ω δ
ξ = δ
∫
∫
Ω
σ
σ
0
0
0
0 0
t i
I
0
0 0
t i
d e
dI e 2 I
. 3.2.13
Conservation of mass requires that
3.3
t i
0
0
0 0
0
e
dr
d
r 3
σ


.

\
 ξ
+ ξ − =
ρ
δρ
. 3.2.14
Substitution of this back into the expression for the variation of the kinetic energy yields
) r ( dm e
dr
d
r 3 ) 1 (
P
3 T 2
0
t i
M
0
0
0
0 0
0
0 σ
∫ 

.

\
 ξ
+ ξ − γ
ρ
− = δ . 3.2.15
Equation (3.2.15) may be simplified to yield
3.4
∫ ∫
Ω − γ ξ +
ρ
ξ
− = δ
Ω
σ σ
M
0
0
0
0
t i
0
0 0
0 0 t i
d ) 1 ( e 3 ) r ( dm
dr
d P
e 3 T 2
0
. 3.2.16
We now have all the material necessary to evaluate the variational form of the virial theorem to
first order accuracy. Substituting equations (3.2.13) and (3.2.16) into equation (3.2.3), we obtain
∫ ∫ ∫ ∫
Ω Ω
σ σ σ σ
Ω ξ − Ω − γ ξ +
γ
ρ
ξ
= ξ σ −
0 0 0
0 0
0
t i
0 0
t i
M
0
0 0
0 0 0 t i
I
0
0 0
t i 2
d e d ) 1 ( e 3 ) r ( dm
dr
d
r P
e 3 dI e . 3.1.17
Solving for σ
2
, which is related to the pulsation period, we have
∫
∫ ∫
ξ
γ
ρ
ξ
+ Ω ξ − γ −
= σ
Ω
0
0
I
0
0 0
M
0
0 0
0 0 0
0 0
0
2
dI
) r ( dm
dr
d
r P
3 d ) 4 3 (
. 3.2.18
For a model of known equilibrium structure, the integrals in equation (3.2.18) may be
evaluated and the frequency for which it is stable to radial pulsations may be computed.
However, for purposes of examining the behavior of a pulsating star we may assume the star is
sufficiently homogeneous so that γ is constant. Also, let us assume the pulsation increases
radially outward in a linear manner. Under these admittedly ad hoc assumptions, equation
(3.2.18) reduces to the extremely simple form
51
The Virial Theorem in Stellar Astrophysics
0
0 2
I
) 4 3 ( Ω − γ
− = σ . 3.2.19
In order to obtain a feeling for the formula we have developed, we shall attempt to
estimate some approximation pulsation frequencies. For a sphere of uniform density
0
2
0
R
GM
5
3
= Ω . 3.2.20
The moment of inertia for a sphere about an axis is equal to 3/2 the moment of inertia about its
center and is given by
0
2
0 Z
I
2
3
MR
5
2
I = = . 3.2.21
Therefore,
15
MR 4
I
2
0
0
= . 3.2.22
Prior theory concerning stellar structure implies that γ > 4/3. If we take γ = 5/3, appropriate for a
fully convective star, we obtain
3
0
2
R
GM
4
9
= σ , 3.2.23
or ρ π = σ G 3
2
.
Remembering that the period T is just 2π/σ we have
2
1
G 3
4
−
ρ


.

\

π
= T . 3.2.24
Thus, we see that the theory does produce a period which is inversely proportional to the
square root of the mean density. This law has been found to be experimentally correct in the case
of the Classical Cepheids. It should be noted that this property will be preserved even for the
integral form equation (3.2.18), only the constant of proportionality will change. If we evaluate
the constant of proportionality from equation (3.2.24), we have
2
1
3
10 92 . 7
−
ρ × ≅ T sec. 3.2.25
where ρ is given in (gm/cc).
Taking an observed value for the mean density of a Cepheid variable to be between 10
3
and 10
6
cm/cc (eg. Ledoux and Walraven)
4
, we arrive at the following estimate for the periods
of these stars.
0.3 days < T < 90 days 3.2.26
52
The Virial Theorem in Stellar Astrophysics
It is freely admitted that this estimate is arrived at in the crudest way, however, and it is
comforting that the result nicely brackets the observed periods for Cepheid variables. It should
also be noted that for most stars the expression arrived at in equation (3.2.23) for σ
2
is a lower
limit. As the mass becomes more centrally concentrated the magnitude of the gravitational
energy will increase while the moment of inertia will decrease. Even for reasonable density
distributions the value arrived at in equation (3.2.23) will not differ by more than an order of
magnitude. This would imply that a value for the period calculated in this manner should be
correct within a factor of 2 or 3. Thus, without solving the force equations, an estimate for a very
important parameter in describing the pulsation of a gas sphere may be obtained which is the
period for which that sphere is stable to radial pulsation.
3 The Influence of Magnetic and Rotational Energy upon a
Pulsating System
We shall now consider what the effect of introducing magnetic and rotational energies
into a pulsating system will be upon the frequency of pulsation of that system. It is worth noting
that solution of such a problem in terms of the force equations would be difficult indeed as it
would require detailed knowledge of the geometry of the magnetic field throughout the star.
However, since our approach expresses the pulsation frequency in terms of volume integrals,
only knowledge of the total magnetic and rotational energies will be required.
In order to simplify the mathematical development we shall make some of the
assumptions which were made during previous sections. These assumptions are listed below:
1. A first order theory will be adequate. That is, all deviations from
equilibrium shall be small.
2. Radiation pressure will be considered to be negligible (i.e., Γ
l
= γ).
3. γ will be constant throughout the system.
We have already seen that it is possible to write Lagrange's identity so as to include the
effects of rotational and magnetic energy. One of the points of that derivation that required some
care was the inclusion of surface terms arising from the fact that stellar magnetic fields usually
extend well beyond the normal surface of the star. However, for the moment let us neglect these
terms since they usually will be small and as such will not affect the general character of the
solution. Thus, as in Chapter II, we may write the scalar form of Lagrange's identity as follows:
M + Ω + = T 2
dt
I d
2
2
2
1
, 3.3.1
53
The Virial Theorem in Stellar Astrophysics
where T is the total kinetic energy including rotation and M is the total magnetic energy. Now let
us break up the total kinetic energy of the system into the sum of three energies T
l
, T
2
and T
3
where T
l
is the kinetic energy of the pulsating system due to the pulsating motion, T
2
is the
kinetic energy of the gas due to thermal energy, and T
3
is the kinetic energy of rotation. The
contribution to the kinetic energy (of particle motion) due to an element of mass is
dm ) c c (
2
3
dm
2
3
dn k
2
3
d
v p
T RT T − = = =
2
T , 3.3.2
where T and R are the gas temperature and constant respectively. But the internal energy dU of
the element of mass is
dm c d
v
T = U , 3.3.3
Combining equation (3.3.2) and equation (3.3.3), with the definition of γ, we have
U T d ) 1 (
2
3
d
2
− γ = . 3.3.4
Integrating this over the entire system we obtain
U 2T ) 1 ( 3
2
− γ = . 3.3.5
Thus, we may write the virial theorem for the system as
M U 2T T + Ω + − γ + + = ) 1 ( 3 2
dt
I d
3 1
2
2
2
1
. 3.3.6
The variational form becomes
M U T δ + Ω δ + δ − γ + δ = δ ) 1 ( 3 2 ) I (
dt
d
3
2
2
2
1
. 3.3.7
There is no term containing a variation in γ as it is assumed to be constant throughout the system.
In section 2, it was shown that variation of the pulsational kinetic energy was of second order
and could therefore be neglected. Thus
0 2
1
= δT . 3.3.8
Since we have already computed the variation of the kinetic energy of the gas, we may
most easily find the variation of the total internal energy (δU) in terms of this quantity. As a
result of the assumption of constant γ, we may take the variation of equation (3.3.5), and obtain
U T δ − γ + = δ ) 1 ( 3 2
2
. 3.3.9
Now, if we further assume a periodic form for the pulsation and linearly increasing amplitude
(ξ
o
=const.), equation (3.2.15), gives the following expression for the variation of the kinetic
energy of the gas:
∫
σ
− γ
ρ
ξ
− = δ
M
0
0
t i
0
0 0
2
) r ( dm e ) 1 (
P 3
3 2 T . 3.3.10
54
The Virial Theorem in Stellar Astrophysics
or .
∫
σ
ξ − γ − = δ
V
0
t i
0 2
dV P 3 e ) 1 ( 3 2 T
Combining equation (3.3.10) and equation (3.3.9), we obtain
∫
σ
ξ − = δ
V
0
t i
0
dV P 3 e U . 3.3.11
In order to express this variation in terms of the other energies present in the system, we
shall assume that the system is in quasisteady state. With this assumption, the relevant quantities
are averaged over one pulsation period so that the 0
dt
I d
2
2
>= < . We shall assume that the
remaining values are the equilibrium values of the configuration. The virial theorem as expressed
in equation (3.3.6) then becomes
0 ) 0 ( ) 0 ( ) 0 ( 2
0 3 2 1
= + Ω + + + M 2T 2T T . 3.3.12
Now, from the elementary kinetic theory of gases we have
∫
=
V
0 2
3
2
dV P T . 3.3.13
Also, since the system is neither expanding or contracting,
0 ) 0 (
1 1
>= =< T T . 3.3.14
Making use of these two results and replacing the average values of the quantities in equation
(3.3.12) with the equilibrium values we find
∫
+ Ω + = −
V
0 0 3 0
) 0 ( 2 dV P 3 M T . 3.3.15
Identifying the left side of equation (3.3.15) with the right side of equation (3.3.11), we finally
obtain
( )
0 0 3
t i
0
) 0 ( 2 e M T U + Ω + ξ = δ
σ
. 3.3.16
We have already determined expressions for the variations of the gravitational energy and
moment of inertia in the previous section [equation (3.2.13)]. Under the assumption used above,
of a constant perturbation ξ
o
, these equations become
¦
)
¦
`
¹
Ω ξ − = Ω δ
ξ = δ
σ
σ
0 0
t i
0 0
t i
e
I e 2 I
. 3.3.17
Thus, we need only obtain expressions for the variation of the magnetic and rotational energies
in order to evaluate equations (3.3.7).
Consider first the rotational energy. Now an element of mass rotating about an axis with a
velocity ω will possess an elemental angular momentum
( ) ) r ( dm sin r ) r ( dm y x d
2 2 2 2
θ ω = + ω = L . 3.3.18
55
The Virial Theorem in Stellar Astrophysics
Here the xy plane is the plane perpendicular to axis of rotation and θ is the polar angle measured
from the axis of rotation. Such an elemental mass possessing such an angular momentum will
have a rotational kinetic energy given by
L T d d
2
1
3
ω = . 3.3.19
Thus, the total rotational kinetic energy is just
∫
ω =
L
L T
0
2
1
3
d . 3.3.20
We may use this expression to obtain the variation δT
3
. The first term on the right of equation
(3.3.7) becomes
) ( d d d 2
0 0 0
3
∫ ∫ ∫
δ ω + ω δ = ω δ = δ
L L L
L L L T . 3.3.21
However, the conservation of angular momentum requires that L remain constant during the
pulsation. Thus, the variation of L is zero and the last integral on the right of equation (3.3.21)
vanishes, so that
∫ ∫
ω
ω
δω
+ ω δ = δ
L L
L L T
0
0
0
0
3
d d 2 . 3.3.22
where ω
o
is the rotational velocity of the equilibrium configuration.
Now, again making use of the conservation of angular momentum we see that
. const sin r
2
= θ ω 3.3.23
Since we are only considering radial pulsation so that δθ is zero, the variation equation (3.3.23)
yields
0 sin r r 2 sin r
2
= θ δ ω + θ δω . 3.3.24
This is equivalent to "conserving angular momentum in shells."
If we evaluate equation (3.3.24) at the equilibrium position, we obtain an expression of
first order accuracy,
t i
0
0 0
0
e 2
r
r 2
σ
ξ − =
δ
=
ω
δω
. 3.3.25
Substitution of this expression into equation (3.3.22) yields
( )
∫
ω ξ − = δ
σ
L
L T
0
0 0 0
t i
3
d e 2 2 . 3.3.26
If, for simplicity, we further assume that the rotational velocity is a constant throughout the
configuration. We obtain a very simple form of the variation of the rotational energy.
( )
0 0 0
t i
3
e 2 2 L T ω ξ − = δ
σ
, 3.3.27
where L
o
is the total angular momentum for the system. Thus, only the variation of the magnetic
energy remains to be determined.
56
The Virial Theorem in Stellar Astrophysics
In order to determine the variation of the total magnetic energy it is necessary to establish
a coordinate system appropriate to the geometry of the field and to the geometry of the
configuration. Although the configuration is spherically symmetric, the geometry of the
magnetic field present is not known. Thus, we shall consider the variations in Cartesian
coordinates and later reduce our result to a form which is compatible with our previous results.
Now the total magnetic energy of the configuration is defined (in c.g.s. units).
∫
πρ
=
M
0
2
) r ( dm
8
H
M . 3.3.28
Thus, denoting the Cartesian coordinates as x
1
, x
2
, and x
3
, the variational form of the magnetic
energy is
) r ( dm
8
1
) r ( dm
4
1
2
2
M
0
M
0
ρ
δρ
π
−
ρ
δ •
π
= δ
∫ ∫
H
H H
M , 3.3.29
or in Cartesian coordinates
3 2 1
i
2
3 2 1 i i
dx dx dx
8
1
dx dx dx H H
4
1
∫∫∫
∑
∫∫∫
ρ
δρ
π
− δ
π
= δ H M . 3.3.30
Although we have already obtained an expression for δρ/ρ in section 3, due to the introduction of
Cartesian coordinates it is convenient to express this variation in terms of the variation of the
coordinates η
i
3.5
, namely
∑
=
∂
η ∂
− =
ρ
δρ
3
1 i i
i
x
, 3.3.31
Before we can evaluate the expression for the variation of the magnetic energy, we must
first determine the variation of the magnetic field δH
i
. A rather lengthy argument
3.6
shows we can
express this in terms of the coordinate variations, so
∑


.

\

∂
η ∂
−
∂
η ∂
= δ
j i
j
i
j
i
j i
x
H
x
H H . 3.3.32
If we substitute (3.3.32) and (3.3.31) into (3.3.29), we have
∫∫∫
∑
∫∫∫
∑∑
∫∫∫
∑∑
∂
η ∂
π
+
∂
η ∂
π
−
∂
η ∂
π
= δ
j
3 2 1
j
j
2
i j
3 2 1
j
j
2
i
i j
3 2 1
j
i
j i
dx dx dx
x
H
8
1
dx dx dx
x
H
4
1
dx dx dx
x
H H
4
1
M
. 3.3.33
The second and third terms combine to yield
∫∫∫
∑
∫∫∫
∑∑
∂
η ∂
π
+
∂
η ∂
π
= δ
j
3 2 1
j
j
2
i j
3 2 1
j
i
j i
dx dx dx
x
H
8
1
dx dx dx
x
H H
4
1
M . 3.3.34
57
The Virial Theorem in Stellar Astrophysics
Now, if we assume that η
i
is only a function of x
i
then the sum on j in the first term collapses and
the remaining terms result in
3 2 1
i i
i
2
j
j
2
i
dx dx dx
x x
H 2
8
1
∫∫∫
∑


.

\

∂
η ∂
−
∂
η ∂
π
= δ H M . 3.3.35
At this point, it is appropriate to reintroduce the assumption concerning the nature of the
variation η
i
. It was earlier assumed that ξ
o
was constant. The equivalent assumption for the s '
i
η
is that
η
i
= const x
i
. 3.3.36
Substitution of this explicit variation into equation (3.3.35) yields
∫∫∫
π
− = δ
3 2 1
2
dx dx dx
8
. const
H M . 3.3.37
Now, since we wish to consider the same type of pulsation from the s '
i
η , as we have
assumed in the earlier section, we require that
η r δ = . 3.3.38
or
ξ
η
=
r
rˆ e ξ
t i
0
σ
= . 3.3.39
Making use of our definition for s '
i
η , equation (3.3.36), we have
t i
0
k j i
e
r
x x x
. const
σ
ξ = ξ =


.

\
 + +
. 3.3.40
This relation can only be true if the pulsation in the three coordinates (η
i
) are in phase and of
equal amplitude, and if
t i
0
e . const
σ
ξ = . 3.3.41
Now, using the definition for the mass in a given volume in Cartesian coordinates and the value
for the constant in equation (3.3.37), we may rewrite the variation of the magnetic energy as
follows:


.

\

ρ π
ξ − = δ
∫
σ
M
0
2
t i
0
) r ( dm
H
8
1
e M . 3.3.42
Making use of equation (3.3.28) we may rewrite the variation in terms of the total magnetic
energy of the equilibrium configuration
0
t i
0
e M M
σ
ξ − = δ . 3.3.43
Thus, we have obtained an expression for the last variation required to evaluate the variational
form of the virial theorem (equations 3.3.7). We may, therefore, substitute equations (3.3.16),
(3.3.27), and (3.3.43), into equations (3.3.7), and obtain
58
The Virial Theorem in Stellar Astrophysics
( ) ( )
0
t i
0 0
t i
0 0 0 0 0
t i
0 0 0 0
t i
0 0
t i
2
2
2
1
e e ) ( e ) 1 ( 3 e 2 I e 2
dt
d
M M L L
σ σ σ σ σ
ξ − Ω ξ − + Ω + ω ξ − γ + ω ξ − = ξ . 3.3.44
Simplifying equation (3.3.44), we find that
0
0 0 0 0 2
I
) 3 5 ( ) )( 4 3 ( L M ω γ − + + Ω − γ −
= σ . 3.3.45
Although we have made some strict assumptions in deriving equation (3.3.45), one
should not feel that they are all of paramount importance. The assumption of constancy of ξ
o
and
γ were only made so that the resultant integrals could be integrated in terms of the original
parameters. If necessary, these assumptions may be omitted and an integral expression similar to
equation (3.2.18) may be derived. However, the work required to obtain this expression is non
trivial and in order for it to be useful one must have a detailed model in mind. One must also
know the detailed geometry of the magnetic fields and of the star in order to evaluate the
integrals that result. Also, to study the behavior of σ
2
, a great deal of numerical work will be
required. For purposes of studying the effects of changes in γ, Ω
o
, M
o
, ω
0
, and L
0
, equation
(3.3.43) will be quite adequate and is much easier to handle. Equation (3.3.45) contains many
aspects which one may check for 'reasonableness'. If we let η
o
and ω
0
be zero, then equation
(3.3.45) becomes identical to the previously derived equation (3.2.19). Letting only ω
0
be zero
we obtain an expression identical to one arrived at by Chandrasekhar and Limber
5
(1954). If ω
0
is nonzero, while M
o
is zero, the expression is that of Ledoux
l
(1945). It should not be
surprising to find the magnetic energy entering in an additive manner to the gravitational energy.
Both are potential energies and since the basic equations are scalar in nature, we should expect
the final result to merely 'modify' the gravitational energy. However, the rotational energy is
kinetic in nature and hence would not enter into the final result in the same manner as the
magnetic energy. Let us briefly investigate the effect upon σ
2
, and hence on the pulsational
period of the presence of magnetic and rotational energy. Since,
σ π = / 2 T , 3.3.46
an increase in σ indicates a decrease in the period and viceversa. Now, since γ > 4/3 and the
gravitational potential energy is defined as being negative, the first term in the numerator of
equation (3.3.45) will be positive only if
0 0
M > Ω . 3.3.47
Thus, the introduction of magnetic fields only serves to reduce σ
2
and thereby lengthen the
period of pulsation. However, the addition of rotational energy (ω
0
L
0
) will tend to increase σ
2
as
long as γ > 5/3. When γ = 5/3, the introduction of rotation has no effect on the period of
pulsation. If γ < 5/3, the influence of rotation is similar to that of magnetic energy.
59
The Virial Theorem in Stellar Astrophysics
4. Variational Form of the Surface Terms
In deriving the virial theorem, we noted earlier that the use of the divergence theorem
yields some surface integrals which are generally ignored. Formally they may be ignored by
taking the bounding surface of the configuration to be at infinity. However, in reality, this
generally proves to be inconvenient for stars as they usually have a reasonably welldefined
surface or boundary. For stars possessing general magnetic fields which extend beyond the
surface, these surface contributions should be included. They are usually wished away by
assuming they are small compared to the total magnetic energy arising from the volume
integration. Although this may be true for simple fields in stars, it is unlikely to be true for other
gaseous configurations such as flares and in any event a numerical estimate of their importance
is far more reassuring than an intuitive feeling. For this reason, let us consider the way in which
these surface terms affect the variational formalism of the previous section. To facilitate the
calculations, we will assume the star is nearly spherical and the pulsations are radial. If the
magnetic field is strong this will clearly not be the case and the full tensor virial theorem must be
used. However, the simplicity generated by the use of the scalar virial theorem justifies the
approach for purposes of illustration. Let us begin by sketching the origin of the virial theorem as
rigorously presented by Chandrasekhar
6
. The equations of motion for a gas with zero resistivity
are
( ) H H
u
× × ∇
π
+ Φ ∇ ρ + ρ −∇ = ρ
4
1
dt
d
. 3.4.1
Employing the identity ( 2 / ) ( ) ( ) H H H H H H • ∇ − ∇ • = × × ∇ and taking the scalar product of
equation (3.4.1) with the position vector r then integrating over all space enclosed by the
bounding surface, we get
∫ ∫ ∫ ∫ ∫
∇ •
π
− ∇ • •
π
+ Φ ∇ • ρ + ∇ • − = • ρ
V
2
V V V V
dV ) H (
8
1
dV ) (
4
1
dV PdV dV
dt
d
r H H r r r
u
r . 3.4.2
As in section 3, this becomes
) d )( (
4
1
d H
8
1
d P ) 1 ( 3 T 2
dt
I d
S
0 0 0
S
2
0
S
0
2
2
2
1
∫ ∫ ∫
• •
π
+ •
π
− • − + Ω + − Γ = − S H H r S r S r M U , 3.4.3
where P
o
and H
o
are the gas pressure and magnetic field present at the surface r
o
. It is the
behavior of the three integrals in equation (3.4.3) that will interest us as hopefully the remaining
terms are by now familiar. Consider first the effect of a pulsation on the surface term arising
from the pressure by taking the variations of the surface pressure integral.
) ( d P d P d P d P
S
0 0
S
0 0
S
0 0
S
0 0
∫ ∫ ∫ ∫
δ • + • δ + • δ = • δ S r S r S r S r . 3.4.4
For radial variations only
dS 2 d d sin r 2 d d sin r r 2 ) ( d
2
0 0 0
ξ = φ θ θ ξ = φ θ θ δ = δS , 3.4.5
where, as in section 2, ξ = δr/r. In Chapter III, section 2 [equations (N3.2.13) and (3.2.14)], we
already have shown that for adiabatic pulsations
60
The Virial Theorem in Stellar Astrophysics

.

\
 ξ
+ ξ γ − =
ρ
δρ
γ =
δ
dr
d
r 3
P
P
. 3.4.6
Combining equation (3.4.5) and equation (3.4.6) with equation (3.4.4), we get
S r S r S r d
dr
d
r d P ) 1 ( 3 d P
0
S r
0
S
0 0
S
0 0
0
•


.

\

ξ
γ − • ξ − γ = • δ
∫ ∫ ∫
. 3.4.7
Earlier we assumed that ξ was constant throughout the star and hence its derivative vanished.
Here, we only require the derivative to vanish at the surface in order to simplify equation (3.4.7)
to get
∫ ∫
ξ − γ = • δ
S
0 0
S
0 0
dS r P ) 1 ( 3 d P S r . 3.4.8
Now consider the variation of the two magnetic integrals in equation (3.4.3).
∫ ∫
∫ ∫ ∫
∫ ∫ ∫ ∫
δ •
π
− • δ
π
−
• δ •
π
+ δ • •
π
+ • δ •
π
+
• δ •
π
+ • • δ
π
=
(
¸
(
¸
•
π
− • •
π
δ
S S
0
2
0 0
2
0
S
0 0 0
S
0 0 0
S
0 0 0
S
0 0 0
S
0 0 0
S S
0
2
0 0 0 0
d H
8
1
d H
8
1
) d )( ( 2
8
1
) d )( ( 2
8
1
) d )( ( 2
8
1
) d )( ( 2
8
1
) d )( ( 2
8
1
d H
8
1
) d )( ( 2
8
1
S r S r
S r H H S H H r S H H r
S H H r S H H r S r S H H r
3.4.9
This truly horrendous expression does indeed simplify
3.7
by using
0
2 H H ξ − = δ . 3.4.10
Using this result, equation (3.4.5) and the definition of ξ, equation (3.4.9) becomes:
∫ ∫
• −
(
¸
(
¸
• •
π
ξ
− = δ
S
0
2
0
S
0 0 0 m
d H d )( ( 2
8
Q S r S H H r . 3.4.11
where Q
m
stands for the original magnetic surface term that appears in equation (3.4.3). Thus, the
variation of the surface terms can be represented as
)
`
¹
ξ − γ = δ
ξ − = δ
P P
m m
Q ) 1 ( 3 Q
Q Q
. 3.4.12
If we assume a linear or homologous pulsation, as was done in the previous two sections,
then the expression for the pulsational frequency [equation (3.3.45)], becomes
0
m P 0 0 0 0 2
I
Q Q ) 1 3 ( ) 3 5 ( ) )( 4 3 ( − − γ + ω γ − − + Ω − γ
− = σ
L M
. 3.4.13
Since γ > 4/3, the contribution of the surface pressure term is such as to increase σ
2
and thereby
improve the stability of the system. Basically, this results because an unstable system will have
to do work against the surface pressures either in expanding or contracting the surface. This
energy is thus not available to feed the instability. The situation is not as obvious for the
61
The Virial Theorem in Stellar Astrophysics
magnetic contribution Q
m
, since Q
m
is the difference between two positive quantities. Thus, the
result depends entirely on the geometry of the field. The effect of the field geometry can be made
somewhat clearer by considering a spherical star so that the radius vector is parallel to the
surface normal. Under these conditions Q
m
becomes
 
∫
β
π
= − •
π
=
S S
2
0
2
0
2
0 m
rdS H cos
8
1
rdS H 1 ) ˆ
ˆ
( 2
8
1
Q r H
∫
, 3.4.14
where β is the local angle between the field and the radius vector. Thus, the average of cos2β
weighted by over the surface will determine the sign of Q
2
0
H
m
. In any event, it is clear that
∫
=
π
<
S
2
0
3
0 2
1
0
2
0 m
H R dS r H
8
1
Q . 3.4.15
It is worth noting that in the case where the magnetic field increases slowly with depth,
this term can be of the same order of magnitude as the internal magnetic field energy and must
be included. Furthermore, whether or not the local contribution to Q
m
is positive or negative
depends on whether or not the local value of β is greater or less than π/4. Since a positive value
of Q
m
increases the value of σ
2
, fields exhibiting a local angle to the radius vector greater than
π/4 tend to stabilize the object, whereas more radial fields enhance the instability. This simply
results from the fact that a radial motion will tend to compress fields more nearly tangential to
the motion than 45
o
, thereby removing energy from the motion. Conversely, more nearly radial
fields will tend to feed the perturbation leading to a decrease in stability.
One thing becomes immediately clear from this discussion. If Q
m
is an important term in
equation (3.4.13), radial pulsation will not occur. Since a magnetic field cannot exhibit spherical
symmetry, the departures from symmetry will yield a variable "restoring force" over the surface
inferring that nonradial modes will be excited. In this case the tensor virial theorem must be
used and the field geometry known. Lastly, for purposes of simplicity, we have assumed no
coupling between the gas pressures and magnetic pressures. Unless the system is rather bizarre,
the gas will be locally relaxed on a time scale less than the pulsation period and hence the two
cannot be treated independently. This assumption was made merely for the sake of simplicity
and doesn't affect the illustrative aspects of the effects. However, unless Q
m
and Q
P
are
comparable the coupling between the two will be weak and we may expect equation (3.4.13) to
give good quantitative results.
5. The Virial Theorem and Stability
In the last section, I alluded to the effects that the surface terms have on the stability of
the system being considered. This concept deserves some amplification as it represents one of
the most productive applications of the virial theorem. However, before embarking on a detailed
development of the virial theorem for this purpose, it is appropriate to review the use of the word
stability itself.
62
The Virial Theorem in Stellar Astrophysics
When inquiring into the meaning of the word, it is customary to consult a dictionary. This
approach provides the following definition:
Stability: "That property of a body which causes it, when disturbed
from a condition of equilibrium or steady motion, to develop forces or
moments which tend to restore the body to its original condition."
This definition is subject to several interpretations and serves to illustrate the danger of
consulting an English dictionary to learn the meaning of a technical term. The word stability is
usually associated with the word equilibrium. This is primarily because the concept of stability
normally is first encountered during the study of statistics. However, there are many dynamical
situations, which are not equilibrium situations that even the most skeptical person would call
stable. One of the most obvious examples to an astronomer, are the stars themselves. Not all stars
would be regarded as stable, but certainly most of the main sequence stars are. Since stars are not
really equilibrium configurations, but rather steady state configurations, we see that we must
extend our conceptualization of stability to include some dynamical systems. The normal
definition of equilibrium requires that the sum of all forces acting on the system is zero. This
concept may be broadened to dynamical systems if one requires that the generalized forces (Q
i
)
acting on the systems are zero. Here the concept of the generalized force may be most simply
stated as
∑
∂
∂
• =
j i
j i
q
Q
r
F , 3.5.1
where F
j
represents the physical forces of the system acting on the jth particle and the q
i
's
represent any set of linearly independent 'coordinates' adequate to describe the system. In a
conservative system all the forces are derivable from a potential Φ .Thus, the generalized forces
may be written as
∑ ∑
∂
Φ ∂
− =


.

\

∂
∂
•


.

\

∂
Φ ∂
= • Φ ∇ − =
j i i
j
j
j i
j
j
j i
q q r dq
d
Q
r
r
r
. 3.5.2
Thus, saying the generalized forces must vanish is equivalent to saying the potential energy must
be in extremum.
0
q
Q
) 0 ( q q
i
i
i i
=


.

\

∂
Φ ∂
− =
=
. 3.5.3
Now, in terms of this definition of equilibrium we may proceed to a definition of stable
equilibrium. If the potential extremum implied by equation (3.5.3) is a minimum, then the
equilibrium is said to be stable. The conditions thus imposed on the potential are
0
q
Q
) 0 ( q q
2
i
2
i
i i
>


.

\

∂
Φ ∂
− =
=
. 3.5.4
63
The Virial Theorem in Stellar Astrophysics
In order to see that this definition of stability is consistent with our dictionary definition, consider
the following argument. Suppose a system is disturbed from equilibrium by an increase in the
total energy dE above the total energy at equilibrium. If Φ is a minimum, any disturbances from
equilibrium will produce an increase in the potential energy. Since the conservation of energy
will apply to the system after the incremental energy dE has been applied, the kinetic energy
must decrease. This implies that the velocities will decrease for all particles and eventually
become zero. Thus, the motion of the system will be bounded (Note: the bound may be
arbitrarily large). If, however, the departure from equilibrium brings about a decrease in the
potential energy, then the velocities may increase without bound. We would certainly call such
motion unstable motion.
However, simple and clearcut as this definition of stability may seem, it is still
inadequate to serve the needs of mathematical physicists in describing the behavior of systems of
particles. Thus it is not uncommon to find modifying adjectives or compound forms of the word
"stable" appearing in the literature. A few common examples are: Secular stability, global
stability, quasistable, bistable, and overstable. These terms are usually used without definition
in the hope that the reader will be able to discern the correct meaning from the context. The
introduction of these modifiers as often as not arises from the mode of analysis used to describe
the system. It is a common practice to examine the response of the system to a continuous
spectrum of perturbations. If any of these perturbations grow without bound the system is said to
be unstable. This would seem in full accord with our dictionary definition and thereby wholly
satisfying. Unfortunately, one is rarely able to calculate the response in general. It is usually
necessary to linearize the equations describing the system in order to solve them. Analysis of this
type is called linear stability theory, and is actually the basis for most stability criterion. Thus,
when analyzing a system not only must one correctly carry out the stability analysis, he must
also decide on the applicability of the analysis to the system.
Recently, it has been quite fashionable to use the virial theorem as the vehicle to carry out
linearized normal mode analysis of systems in order to determine their state of stability.
However, the determination of a system state of stability seems to have inspired Jacobi to
develop the nbody representation of Lagrange's identity from which it is a short step to the virial
theorem. To see now how closely tied the virial theorem is connected to stability; let us
summarize some of Jacobi's arguments. In Chapter I, [i.e. equation (1.4.12)], we arrived at
simple statements of Lagrange's identity for selfgravitating systems as:
Ω + = T 2
dt
I d
2
2
2
1
. 3.5.5
One could say with some confidence that if d
2
I/dt
2
> 0 for all t the system would have to
have at least one particle whose position coordinates increased without bound. That is to say, the
system would be unstable. However, since both T and Ω vary with time it would be difficult to
say something a priori about d
2
I/dt
2
from Lagrange's identity alone. Thus, Jacobi employed the
64
The Virial Theorem in Stellar Astrophysics
constancy of the total energy (i.e., E = T + Ω = const.), and the fact that for selfgravitating
systems Ω > 0, to modify equation (3.5.5) to give:
E 2 E 2
dt
I d
2
2
2
1
> Ω − = . 3.5.6
So, if E > 0, d
2
I/dt
2
> 0 and the system is unstable. This is known as Jacobi's stability
criterion and provides a sufficient (but not necessary) condition for a system to be called
unstable.
It is the constancy of E with time that makes this a valuable criterion for stability and is
the reason Jacobi used it. There is another approach to the problem of temporal variation which
was not available to Jacobi. In Chapter 2 we discussed the extent to which time averages of
quantities may be identified within their phase averages, through use of the Ergodic theorem.
Integrating equation (3.5.5) over some time t , we get
Ω + = 
.

\

T 2
dt
dI
t
1
0
t
0
0
0
. 3.5.7
If the system is to remain bounded, dI/dt must always be finite. If the system is to be always
stable, then the limit of the lefthand side must tend to zero. Furthermore, the time averages on
the righthand side of (3.5.7) will tend to phase averages if the system is Ergodic. Thus
2<T>+ <Q> = 0, 3.5.8
constitutes a stability criterion for Ergodic systems. That is, equation (3.5.8) must be satisfied in
a stable Ergodic system and failure of equation (3.5.8) is sufficient for instability. It is not
uncommon to find the statement in the literature that (2T + Q > 0) insures the instability of a
system citing the virial theorem as the justification. Actually, it is Lagrange's identity that is the
relevant expression and it only guarantees that at the moment the system is acceleratively
expanding. What is really meant is that if (T + Q > 0) the system is indeed unstable as this is just
a statement of Jacobi's stability criterion concerning the total energy of the system.
Now, let us turn the applications of the variational form of the virial theorem to stability,
keeping in mind that the variational approach is essentially a first order or linearized analysis.
The majority of Chapter III has been devoted to obtaining expressions for the frequency of a
pulsating system. We obtained a value for the square of the frequency in terms of the equilibrium
energies of the configuration. However, these expressions could be neither positive nor negative.
In the earlier sections, we discussed only the meaning of the positive values, as negative squares
of frequencies had no apparent physical meaning. Let us look again at the nature of the assumed
pulsation in order to further investigate the meaning of these pulsation frequencies. In section 2
[i.e. equations (3.2.12)], we assumed that the pulsation would be periodic and of the form
t i
0
e
r
r
σ
ξ =
δ
. 3.5.9
65
The Virial Theorem in Stellar Astrophysics
where σ was the frequency of pulsation and ξ
o
did not depend on time. Now, if we make the
formal identification between σ and the period and take σ to be purely imaginary, we may write
0
t / i 2π ± = σ , 3.5.10
where t
o
is a real number. Combining equation (3.5.10) and equation (3.5.9), we have
0
t / t 2
0
e
r
r
π ±
ξ =
δ
. 3.5 .11
Thus, the pulsation becomes exponential in nature. If the sign of σ is negative, then the sign of
the exponential in equation (3.5.11) will be positive and the pulsation will grow without bound
with a rate of growth determined by t
o
.
One might be tempted to choose the negative sign of equation (3.5.11) saying that the
system is stable as the pulsation will die out, even though σ
2
< 0. This would be wrong. All
classical equations of dynamical symmetry exhibit fulltime symmetry, thus those solutions
which damp out in the future were unstable in the past and viceversa. A specific solution would
be fully determined by the boundary conditions at t = 0. Further, we assumed that a full
continuum of perturbations are present, resulting from small but inevitable, departures from
perfection produced by statistical fluctuations. Thus, if there exists even one mode with σ
2
< 0,
the instability associated with that mode will grow without bound. Therefore, this becomes a
sufficient condition for the system to be unstable in the strictest sense of the word
σ
2
< 0. 3.5.12
It is also worth noting that this criterion applies to the entire system, and thus is a "global"
stability condition. However, it can be made into a local condition by taking an infinitesimal
volume and including the surface terms discussed in section 4.
Now, let us see what implications this analysis has for the stability of stars. In section 2,
we established an expression for the pulsation frequency of a gravitating gas sphere [equation
(3.2.19)]. Now, applying the instability criterion equation (3.5.12), we see that the sphere will
become unstable, when
0
I
) 4 3 (
0
0
<
Ω − γ
− . 3.5.13
Since the moment of inertia (I
0
) is intrinsically positive, while the gravitational potential energy
is intrinsically negative, equation (3.5.13) becomes
)
`
¹
< γ
< − γ
3
4
0 ) 4 3 (
. 3.5.14
66
The Virial Theorem in Stellar Astrophysics
Thus, a star will become unstable when γ is less than 4/3. This is the familiar instability
criterion demonstrated by Chandrasekhar in Stellar Structure.
7
He further demonstrates that a gas
with γ equal to 4/3 corresponds to a gas where the total pressure is entirely due to radiation. If we
consider the other stability criterion [equation (3.5.14)], we can see that a necessary condition for
stability of a homogeneous nonrotating gas sphere is
γ > 4/3. 3.5.15
Thus, having extracted as much information as possible from the pulsation expression developed
in section 2, let us turn to the more general formulae resulting from our analysis in section 3.
Remember the final expression for the pulsational frequency was
0
0 0 0 0 2
I
) 3 5 ( ) )( 4 3 ( L M ω γ − + + Ω − γ −
= σ . 3.5.16
Consider first a gas sphere which is not rotating, but which has a magnetic field. Equation
(3.5.16) becomes then
0
0 0 2
I
) )( 4 3 ( M + Ω − γ −
= σ . 3.5.17
Substituting this into the instability criterion equation (3.5.12), we have
0 ) )( 4 3 (
0 0
< + Ω − γ M . 3.5.18
Now, if we assume γ > 4/3, we have a sufficient condition for instability due to the presence of
magnetic energy as follows:
0 0
Ω > M . 3.5.19
In the following manner, we may obtain a crude estimate of the magnitude of the magnetic fields
necessary to disrupt a star. The gravitational potential energy for a sphere of uniform density is
R
GM
5
3
2
− = Ω , 3.5.20
while the magnetic energy is
∫∫∫
= =
6
R
dx dx dx
8
1
2
3
3 2 1
2
H
H M . 3.5.21
Combining equation (3.5.20) and equation (3.5.21), we see that the root mean square value of the
magnetic field sufficient to disrupt a uniformly dense sphere is
gauss
R
M
10 2
2
8
2
× > H , 3.5.22
where M and R are given in solar units. Thus, for a main sequence A star with M = 4M
⊙
and
R = 5R
⊙,
we have
H
rms
>3x10
7
gauss . 3.5.23
However, for a star like VV Cephei with M = l00 M
⊙
, and R = 2600 R
⊙
, we have
H
rms
> 3000 gauss . 3.5.24
67
The Virial Theorem in Stellar Astrophysics
We may conclude from these arguments that for a main sequence star, an extremely large
magnetic field would be sufficient to cause the star to become unstable. However, for an
unusually large star, the required field becomes much smaller. In the case of VV Cephei,
Babcock has measured a field ranging from +2000 to 1200 gauss. Thus, it would appear that VV
Cephei is on the verge of being magnetically unstable. One might argue that our crude estimates
of Ω are so crude as to be meaningless due to the large central concentration of the mass in giant
stars. However, it should be pointed out that the magnetic field one can observe is, of necessity, a
surface field and, therefore, provides us with a lower limit on the magnetic energy. Thus, we may
have some hope that our limiting field values are not too far from being realistic.
It is interesting to note that the instability criterion equation (3.5.18) permits the existence
of a gas with γ < 4/3 providing the magnetic energy exceeds the gravitational energy. Indeed, the
stability criterion (3.5.18) would require a necessary condition for the stability of any
configuration where M > Ω that γ be less than 4/3. However, it is also true that the physical
meaning of a gas having a γ < 4/3 is a little obscure.
If we now consider a rotating configuration with no magnetic field, equation (3.5.14)
combined with the stability criterion equation (3.5.12) becomes
0 0 0
) 4 3 ( ) 3 5 ( Ω − γ > ω γ − L . 3.5.25
If we restrict γ to be less than 5/3 we have
) 3 5 (
) 4 3 (
0
0 0
γ −
Ω − γ
> ω L . 3.5.26
Since Ω
o
is intrinsically negative, we see that the stability condition will always be
satisfied with any ω
o
. Thus, for all known stars the stability criterion for rotation is not
particularly useful.
However, all this is not meant to imply that the rotational terms are unimportant. Indeed,
Ledoux
l
has shown that rotational velocities encountered in stars may lead to a variation in the
pulsational period by as much as 20%. Let us briefly consider the instability criterion when both
magnetic and rotational energy are present. This may be obtained by combining equation
(3.5.15) and
0 ) 3 5 ( ) )( 4 3 (
0 0 0 0
> ω γ − − + Ω − γ L M . 3.5.27
As before, this condition may never be satisfied unless
0 0
M > Ω . However, even in the event
that
0 0
M > Ω , the condition may still not be satisfied because of the presence of the rotational
term. Thus, it is evident that if 4/3 < γ < 5/3, then the presence of rotation will help stabilize
stars. This result is certainly not intuitive. A physical explanation of the result might be supplied
by the following argument.
68
The Virial Theorem in Stellar Astrophysics
Consider a pulsating configuration containing both rotational and magnetic energy. As
the system expands or contracts, a certain amount of energy will be required to slow down or
speed up the rotation in order to keep the angular momentum constant. This energy must be
supplied by the kinetic energy of the gas itself, and, since this is supplied by the potential
energies present, ultimately must come from the gravitational and magnetic energies. Therefore,
the amount of this energy transferred from the magnetic and gravitational energies will depend
on γ. Also since the gravitational and magnetic energies must supply this energy to the rotation,
the energy is no longer available to "feed" the pulsation and disrupt the star.
We may now ask what sort of increase in the maximum magnetic field can this additional
rotational "stability" supply. From our previous investigation with the rotational stability criteria
we might expect the result to be small. That is, since the rotational stability criteria did not
supply us with as important a result as did the magnetic instability, we would expect the effects
of rotation to be small compared with the magnetic energy. If one considers a uniform model
with γ = 3/2, rotating at critical velocity, he will find the magnetic field may only be increased by
about 0.3% before instability will again set in. Thus, even though stability is increased by the
presence of rotation, it is not increased a great deal.
It is appropriate at this point to make some comments regarding all of the stability criteria
relating to the stability of radial pulsations. It would have been more correct to employ the
integral form of the expressions for the frequency of pulsation. However, the result one would
obtain by using the integral expression and a specific model would only differ in degree from
those derived here. It is hoped that the degree of differences would not be large. There is one
respect in which the differences between the derived criteria and the 'correct' ones may result in a
difference in kind. It must be remembered that the expressions developed for the pulsational
frequencies were based on a first order theory as are the stability criteria developed in this
section. However, the conditions at which one wishes to apply an instability criteria are generally
such that the second, and higher, order terms are not small and should not be neglected.
Chandrasekhar and Fermi
3
have shown that a sphere under the influence of a strong dipole field
will tend to be "flattened" in much the same way as it will be by rotation. Once the spherical
symmetry has been destroyed, either by the presence of a strong, magnetic field or rapid rotation,
the concept of radial pulsation becomes inconsistent. As mentioned before, analysis of such
systems would require the use of the tensor virial theorem and considerable insight into the types
of perturbations to employ.
I would be remiss if I left the subject of the virial theorem and stability without some
discussion of the recent questions raised with regard to the appropriateness of the approach. To
me, these questions appear to be partly substantive and partly semantic and revolve largely
around one of these modifiers mentioned earlier, namely, secular stability. So far, our discussion
has been restricted to problems involving dynamical stability about which there seems to be little
argument. A dynamically unstable system will disintegrate exponentially, usually on a time scale
related to the hydrodynamical time scale for the system. Such destruction is usually so
unambiguous that no complications arise in the use of the word unstable. Such is not the case for
the term secular stability.
69
The Virial Theorem in Stellar Astrophysics
The notion of secular stability involves the response of the system to small dissipative
forces, such as viscosity and thus must depend to some extent, on the nature of those forces.
Time scales for development of instabilities will be governed by the forces and hence may be
very long. Perhaps one of the clearest contemporary discussions of the term is given by Hunter
8
who notes that there is less than universal agreement on this meaning of the term. He points out
that difficulties arise in rotating systems resulting from the presence of the coriolis forces, which
lead to a clear distinction between dynamically and secularly stable systems. As we saw in
Chapter II, section 5, the terms associated with the coriolis forces can be made to vanish by the
proper choice of a coordinate frame and they would appear to play no role in the energy balance
of the system. However, their variation does not vanish and hence they will affect the pulsational
analysis. Since globally the forces are conservative the first result is not surprising and since
radially moving mass in a rotating frame must respond to the conservation of angular
momentum, neither is the second. Now if dissipative forces are present such as viscosity then it
may be possible to redistribute local angular momentum while conserving it globally so that no
equilibrium configuration is ever reached. In addition, except for the global constraint on the
total angular momentum, no constraints are placed on the transfer of energy from the rotational
field to the thermal field. Indeed the presence of dissipative forces guarantees that this must
happen. Thus, instabilities associated with these forces might exist which would otherwise go
undetected. This line of reasoning demonstrates a qualitative difference between the cases of
uniform rotation and differential rotation in that in the former dissipative forces will be inactive
and the analysis will be appropriate while in latter cases they must be explicitly included. This
point is central to a lengthy series of papers
9, lO, 11, 12, 13
by Ostriker and others, which discuss the
stability of a variety of differentially rotating systems. However, the majority of these papers
clearly state that the authors are dealing with systems with zero viscosity and so the problem is
not one of the accuracy of the analysis but rather of the applicability of the analysis to physical
systems. In practice, the viscosity of the gas in most stars is so extremely low that the time scales
for the development of instability arising from viscosity driven instabilities will be very long.
One cannot hope to untangle in a few paragraphs a controversy which has taken more
than a decade to develop and at a formal mathematical level is quite subtle. However, it is worth
noting that recent
14
statements which essentially say that the tensor virial approach to stability is
wrong do nothing to clarify the situation. The presence of dissipative forces can be included in
the equations of motion and thus in the resulting tensor representation of Lagrange's identity. The
resulting stability analysis would then correctly reflect the presence of these forces and thus be
dependent on their specific nature. Insofar as the time averaged form of Lagrange's identity,
which is technically the virial theorem, is used the arguments of Milne
15
as presented in
Chapter I still apply. The presence of velocity dependent forces does not affect the virial theorem
unless those forces stop or destroy the system during the time over which the average is taken. At
this level assailing the virial theorem is as useful an enterprise as denying the validity of a
conservation law.
70
The Virial Theorem in Stellar Astrophysics
6. Summary
In this chapter we have explored the results of applying a specific analytical technique to
the virial theorem. As in other chapters, we began with the simple and moved to the more
complex. Having discussed the implications of the variational approach to the virial theorem we
moved to develop the explicit form for the simple scalar theorem appropriate for selfgravitating
systems. We recreated the pulsational formula [equation (3.2.19)] originally due to Ledoux. One
implication of this result is that the fundamental mode of oscillation depends only on the square
root of the density and when coupled with the stability criterion in section 5, leads immediately
to the Jean's stability criterion. This is not surprising as both results have as the derivational
origin the same concept (i.e., the equations of motion). However, it is reassuring when a different
approach yields results already well accepted.
In section 3 we expanded the variational approach to include the effects of magnetic
fields and rotation. In spite of many distractions dealing with the variation of magnetic fields,
etc., the influence of these added features on the pulsation frequency and hence stability became
clear. Rotation can either enhance or reduce the stability of a configuration depending on
whether or not the value of γ for the gas permits net energy to be fed to the pulsation. The
influence of an internal magnetic field is to destabilize the star for all realistic values of γ.
However, the effect of a surface field proves to be more complex. Here the result depends
critically on the geometry of the field. In the last section we dealt briefly with the overall
question of stability and showed explicitly how the virial theorem provided an excellent basis for
a linear stability analysis of a symmetric system. Throughout the chapter we confined ourselves
to spherically symmetric systems exhibiting radial pulsations only. As mentioned, this is
inappropriate when considering either rotation or magnetic fields as neither can exhibit spherical
symmetry and thus one would expect nonradial oscillation to be excited. However, unless the
field energies become quite large one would expect the pulsational frequencies not to differ
greatly from the purely radial theory.
This line of reasoning becomes particularly dangerous when one turns to a discussion of
stability. First the interesting situations of marginal stability are liable to involve substantial
magnetic fields or rapid rotation. If these aspherical properties are large, the departure from
spherical systems of the mass distribution will also be large invalidating every aspect of the
analysis. In addition, for the stability analysis to be valid all possible modes of perturbation must
be included. Limiting oneself to only the radial modes is to invite a misleading result.
Fortunately the techniques for dealing with these problems exist and have been developed here
as well as the literature. The tensor virial theorem as is presented in Chapter II, section 1 allows
one to follow perturbations in independent spatial coordinates. In principle, a complete
variational analysis of perturbations to all independent spatial coordinates will allow one to
compute the nonradial as well as radial modes of oscillation and thereby obtain a much more
secure analysis of the system's stability.
71
The Virial Theorem in Stellar Astrophysics
Notes to Chapter 3
3.1 Remember the defining expression for the gravitational potential energy is
∫
− = Ω
M
0
r
) r ( dm ) r ( m
G . N3.1.1
We may again use the fact that the variation of m(r) and dm(r) are both zero, to obtain
) r ( dm ) r ( m
r
r
G
M
0
2 ∫
δ
− = Ω δ , N3.1.2
which can be written as an energy integral by noting from equation (N3.1.1) that
r
) r ( Gm
) r ( dm
d
− =
Ω
. N3.1.3
Evaluating the above expression at r = r
o
and using the result in equation (3.2.11) to first order
accuracy we get
Ω
δ
− = Ω δ
∫
Ω
d
r
r 0
0
0
. N3.1.4
3.2 We may write the total energy as the sum of two energies T
l
and T
2
, where T
1
is the
kinetic energy due to the mass motion of the gas arising from the pulsations themselves. Now the
total kinetic energy of mass motion is given by
∫ ∫

.

\

= 
.

\

ρ π =
R
0
M
0
2
2
1
2
2
2
1
1
) r ( dm
dt
dr
dr
dt
dr
r 4 T . N3.2.1
Thus, the variation of T
1
∫

.

\
 δ

.

\

= δ
M
0
2
1
1
) r ( dm
dt
r d
dt
dr
T . N3.2.2
However, from the definition of δr we see that
r = r
o
+ δr . N3.2.3
Since the equilibrium point r
0
cannot vary with time by definition, equation (N3.2.2) becomes
∫

.

\
 δ
= δ
M
0
2
2
1
1
) r ( dm
dt
) r ( d
T . N3.2.4
The largest term in the integral of equation (N3.2.4) is second order in δr and may be
neglected with respect to the first order terms of equation (3.2.13), and equation (3.2.6). Thus, to
first order we have
2 2 1
T T T T δ ≅ δ + δ = δ . N3.2.5
Now consider the kinetic energy of a small volume of an ideal gas.
dV Nk d
2
3
2
T = T . N3.2.6
72
The Virial Theorem in Stellar Astrophysics
However, the gas pressure is given by P
g
= NkT and dm(r) = ρdV. Therefore,
) r ( dm
P
d
g
2
3
2
ρ
= T . N3.2.7
Thus, twice the total kinetic energy of the gas sphere arising from thermal sources is
∫
ρ
=
M
0
g
2
) r ( dm
P
3 2T . N3.2.8
Now neglecting radiation pressure, so that the total pressure is equal to the gas pressure, and
remembering that the variation of dm(r) is zero, we have
∫ 

.

\

ρ
δ = δ = δ
M
0
g
2
) r ( dm
P
3 2 T 2 T . N3.2.9
We shall now assume that the pulsations are adiabatic so that
ρ
δρ
γ =
δ
P
P
, N3.2.10
where γ is the ratio of specific heats c
p
/c
v
. Now
( )
 
0
2
2
2
) 1 (
P
P
P
P
P P P
δρ − γ ρ
=

.

\
ρ
δρ − ρ
δ
=
ρ
δρ − ρδ
=


.

\

ρ
δ . N3.2.11
Therefore, again evaluating at the equilibrium position, substituting into equation (N3.2.9),
And keeping only terms up to first order, we have
∫
ρ
δρ
− γ
ρ
≅ δ
M
0
0 0
0
) r ( dm ) 1 (
P
3 T 2 . N3.2.12
3.3 In out attempt to obtain an expression for δρ/ρand thereby determining the variation
energy, we shall invoke the following argument. From the conservation of mass, we have
∫ ∫ ∫ ∫
δ ρ π + δρ π + ρ δ π = ρ π = = δ
' r
0
' r
0
' r
0
2 2
' r
0
2
) r ( d r 4 dr r 4 dr ) r r 2 ( 4 dr r 4 0 ) ' r ( m . N3.3.1
Rewriting equation (3.2.28) and evaluating at r = r
0
, we have
∫ ∫
(
¸
(
¸
δ +
δ
ρ − = δρ
' r
0
' r
0
0
0 2
0 0 0
2
0
) r ( d
r
rdr 2
r dr r . N3.3.2
From the definition of ξ (equation 3.2.12), we have
2
0
0
0
2
0
0
0
r
rdr
r
) r ( d
d or,
r
dr dr / ) r ( d r
dr
d δ
−
δ
= ξ
− δ
=
ξ
. N3.3.3
73
The Virial Theorem in Stellar Astrophysics
Eliminating d(δr) from equation (N3.3.2) with the aid of equation (N3.3.3), we have again to the
first
∫ ∫
(
¸
(
¸
ξ
+ ρ − = δρ
' r
0
0
' r
0
0
0
2
0 0 0
2
0
dr
dr
d
r 3 r dr r . N3.3.4
Equation (N3.3.4) must hold for all values of r'. This can only be true if the integrands equal.
Thus,


.

\
 ξ
+ ξ − =
ρ
δρ
0
0
0
dr
d
r 3 . N3.3.5
3.4 In an attempt to simplify equation (3.2.15), let us consider the second integral.
∫ ∫
π
ξ
− γ
ρ
=
ξ
− γ
ρ
σ σ
R
0
0
2
0
t i
0
0
0
0
0
M
0
t i
0
0
0
0
0
dr r 4 e
dr
d
r ) 1 (
P
3 ) r ( dm e
dr
d
r ) 1 (
P
3 . N3.4.1
Integrating the right hand side by parts we obtain

∫ ∫
− γ ξ π − ξ − γ π = − γ
ρ


.

\
 ξ
π
σ σ σ
R
0
0
3
0 0 0
t i
R
0
0
3
0 0
t i
R
0
0
3
0
0
0
0
0 t i
dr r ) 1 ( P
dr
d
e 12 r ) 1 ( P e 12 dr r ) 1 (
P
dr
d
e 12
0
 . N3.4.2
At this point we shall impose the boundary conditions that P
o
→ 0 at r
o
= R and ξ
o
→ 0 at
r
o
= 0. The first of these conditions is the familiar condition of stellar structure which essentially
defines the surface of the gas sphere. The second condition is required by assuming that the
radial pulsation is a continuous function. With these conditions, the integrated part of equation
(N3.4.2) vanishes and the remaining integral becomes:
 
∫ ∫ ∫ ∫
ξ − γ + ξ − γ + ξ = − γ ξ
R
0
R
0
R
0
0
2
0 0 0 0
3
0
0
0
0
3
0 0 0
R
0
0
3
0 0 0
dr r ) 1 ( P 3 dr r
dr
dP
) 1 ( r P dr r ) 1 ( P
dr
d
. N3.4.3
Conservation of momentum (i.e., hydrostatic equilibrium) requires that
2
0
0
0
0
r
) r ( Gm
dr
dP ρ
− = . N3.4.4
Therefore,
0
0
0
0
0
0
0
0
2
0
0
0 3
0
dr
d
dr
) r ( dm
r
) r ( Gm
r
) r ( m Gr 4
dr
dP
r 4
Ω
= − =
ρ π
− = π . N3.4.5
Making use of the second form of equation (N3.4.5) to simplify the second integral in
equation (N3.4.3), and the definition of dm(r) to simplify the other two integrals in equation
(N3.4.3), we obtain the following expression for the variation of the kinetic energy [i.e. equation
(3.2.15)].
∫
∫ ∫ ∫
− γ
ρ
ξ
+
Ω − γ ξ +
γ
ρ
ξ
+ − γ
ρ
ξ
− ≅ δ
σ
Ω
σ σ σ
M
0
0
0
0 0 t i
0
0 0
t i
M
0
0
0 0
0 0 0 t i
M
0
0
0
0 0 t i
) r ( dm ) 1 (
P
e 9
d ) 1 ( e 3 ) r ( dm
dr
d r P
e 3 ) r ( dm ) 1 (
P
e 9 T 2
0
. N3.4.6
74
The Virial Theorem in Stellar Astrophysics
The above expression is obtained by making use of the previously mentioned definition and
substituting it into equation (N3.4.3), then into equation (N3.4.2), and finally into equation
(3.2.15). Since the first and last integrals are identical except for the difference in sign, they
vanish from the expression and
d ) 1 ( e 3 ) r ( dm
dr
d r P
e 3 T 2
0
0
0 0
t i
M
0
0
0 0
0 0 0 t i
∫ ∫
Ω
σ σ
Ω − γ ξ +
γ
ρ
ξ
≅ δ . N3.4.7
3.5 Let us define
i i
x δ ≡ η . N3.5.1
Now the definition of the mass within a given volume in Cartesian coordinates becomes
∫∫∫
ρ =
3 2 1
dx dx dx ) V ( m . N3.5.2
The conservation of mass requires, as it did in section 2, that the variation of the mass is zero.
Thus taking the variations of equation (N3.5.2), we obtain
) x ( d dx dx dx ) x ( d dx dx dx ) x ( d dx dx dx ) V ( m
3 2 1 3 2 1 3 2 1 3 2 1
∫∫∫ ∫∫∫ ∫∫∫ ∫∫∫
δ ρ δ ρ + δ ρ + δρ = δ . N3.5.3
Rewriting the last three integrals, we have
∫∫∫
∑
∫∫∫
=
δ
ρ − = δρ
3 2 1
3
1 i i
i
3 2 1
dx dx dx
dx
) x ( d
dx dx dx . N3.5.4
Since δx
i
= η
i
, we have by the chain rule
∑
=
∂
η ∂
=
η
=
δ
3
1 j j
i
j
i
i
i
i
i
x dx
dx
dx
d
dx
) x ( d
, N3.5.5
However, since the x
i
's are linearly independent and the η
i
's are just the variation of these
coordinates, not only is the second term in the product [under the summation sign of equation
(N3.5.5)] zero if i= j but, so is the first term.
Thus as might be expected from the orthogonality of the x
i
's we have
i
i
i
i
x dx
d
∂
η ∂
=
η
. N3.5.6
Substitution of this into equation (N3.5.4), yields
∫∫∫
∑
∫∫∫
=
∂
η ∂
ρ − = δρ
3 2 1
3
1 i i
i
3 2 1
dx dx dx
x
dx dx dx . N3.5.7
Equation (N3.5.7) must hold for any volume of integration. This can only be true if the
integrands themselves are equal. Thus, we finally obtain
∑
=
∂
η ∂
− =
ρ
δρ
3
1 i i
i
x
. N3.5.8
75
The Virial Theorem in Stellar Astrophysics
3.6 Suppose a displacement n takes place with the slow continuous movement so that
dt
dn
u = . N3.6.1
Then, if the electrical conductivity of the medium is infinite, the timevariation of the electric
field is
H E × − = ∆ u . N3.6.2
However, Maxwell's equations for an infinitely conducting medium require that
H E ∆
∂
∂
− = ∆ × ∇
t
) ( . N3.6.3
Combining equation (N3.6.2) and equation (N3.6.3), we have
) (
t
) (
H
H
× − × ∇ =
∂
∆ ∂
− u . N3.6.4
The integral form of equation (N3.6.4) is just
) ( H H × × ∇ = ∆ n . N3.6.5
But, from the definition of the time and space variations and the equation relating to the total
time derivatives, we know that
dt
t ∂
∂
= ∆
H
H , N3.6.6
and
∇ • − =
∂
∂
v
dt
d
t
, N3.6.7
which results in
dt
dt
d
dt
d
j
j


.

\

∇ • − = ∆
∑
H
x
H
H . N3.6.8
However,
∑
∂
∂
=
i
i
i
dt
dx
x dt
d
, N3.6.9
while definition of the space variation δH is
∑
∇ •
∂
∂
= δ
i
dt
x
H
H
H . N3.6.10
Combining equations (N3.6.l0), (N3.6.9), and equation (N3.6.8), we have
H
x
H H ∇ • − δ = ∆
∑
i
i
dt
dt
d
. N3.6.11
Noting that the variation of a linearly independent quantity may be interpreted as the total
differential of the quantity
∑
δ =
∂
∂
=
i
i j
j
i
i
x dx
x
x
dx , N3.6.12
we obtain
H H H ∇ • − δ = ∆ n . 3.6.13
76
The Virial Theorem in Stellar Astrophysics
Combining equation (N3.6.l3) and equation (N3.6.S), we finally obtain an equation for the
variation of the magnetic field (δH).
H H H ) ( ) ( ∇ • − × × ∇ = δ n n . N3.6.14
Now the curl of a crossproduct may be written as
) ( ) ( ) ( ) ( ) ( H H H H H • ∇ + ∇ • − • ∇ − ∇ • = × × ∇ n n n n n . N3.6.15
Since the divergence of H is always zero, the last term vanishes. Combining this expression for
the curl of a crossproduct with equation (N3.6.l4) yields
H H H H H ) ( ) ( ) ( ) ( ∇ • + ∇ • − • ∇ − ∇ • = δ n n n n . N3.6.16
The last two terms are identical except for opposite signs and thus cancel out. The remaining
expression may be written in component form as follows:
∑


.

\

∂
η ∂
−
∂
η ∂
= δ
j i
j
i
j
i
j i
x
H
x
H H . N3.6.17
3.7 In the section 3.2 we went through an extensive argument (see Note 3.6, equation
N3.6.l6), to show that
) ( ) ( r H r H H δ • ∇ − δ ∇ • = δ , N3.7.1
where δr plays the role of η in that discussion. The easiest way to evaluate the first term is in
Cartesian coordinates, remembering that ξ
0
is constant. Then,
( )
∑
ξ =
∂
ξ ∂
= δ ∇ •
i i
i
i
x
x
H ) ( H r H . N3.7.2
The second term can be evaluated the same way, so that
( )
∑
ξ =
∂
ξ ∂
= δ • ∇
i i
i
j
3
x
x
H ) ( H r H . N3.7.3
So, the variation of the magnetic field has taken the particularly simple form
0
2 H H ξ − = δ . N3.7.4
77
The Virial Theorem in Stellar Astrophysics
References
1 Ledoux, P. (1945), Ap. J. 102, 143.
(1958), Handbuch der Physik, SpringerVer1sg, Berlin, Gottingen,
Heide1berg, pp. 605687.
2 Chandrasekhar, S. (1963), Ap. J. 137, p. ll85.
3 Chandrasekhar, S. and Fermi, E. (1953), Ap. J. ll8, p. ll6.
4 Ledoux, P. and Walraven, (1958), Handbook der Physik, Vol. 51, pp. 431592, Springer
Ver1ag, Berlin, Gottingen, Heidelberg.
5 Chandrasekhar, S. and Limber, D. N. (1959), Ap. J. ll9, p. 10.
6 Chandrasekhar, S. (1961), Hydrodynamic and Hydromagnetic Stability, Oxford
University Press, London, p.
7 . (1939), An Introduction to the Study of Stellar Structure, Dover Pub.,
Inc. (1957), p. 53.
8 Hunter, C. (1977), Ap. J. 213, p.497.
9 LydenBell, D. and Ostriker, J. P. (1967), M.N.R.A.S. 136, p. 293.
10 Tassoul, J. L. and Ostriker, J. P. (1968), Ap. J. 154, p. 613.
II Ostriker, J. P. and Tassoul, J. L. (1969), Ap. J. 155, p. 987.
12 Ostriker, J. P. and Bodenheimer (1973), Ap. J. 180, p. 171.
13 Ostriker, J. P. and Peebles, P. J. E. (1973), Ap. J. 186, p. 467.
14 Bardeen, J. M., Friedman, J. L., Schutz, B. F. and Sarkin, R. (1977), Ap. J. Lett.
217, p. L49.
15 Milne, E. A. (1925), Phil. Mag. S. 6, Vol. 50, pp. 409419.
78
The Virial Theorem in Stellar Astrophysics
79
The Virial Theorem in Stellar Astrophysics
© Copyright 2003
IV Some Applications of the Virial Theorem
1. Pulsational Stability of White Dwarfs
By now, I hope the reader has been impressed by the wide range of problems which can
be dealt with by the virial theorem. Some of the problems mentioned in the last chapter indicate
the type of insight which can be achieved through use of virial theorem, however, the type of
objects which were explicitly discussed, notably normal stars, are currently judged by the naive
to be well understood. In order to illustrate the power of this remarkable theorem, I cannot resist
discussing some objects about which even less is known. We shall see that at least one
commonly held tenant of stellar structure, while leading to nearly the correct numerical result, is
conceptually wrong.
During the 1960's advances in observational astronomy presented problems requiring
theoreticians to postulate the existence of a wide range of objects previously considered only of
academic interest. These terms, like supermassive stars, neutron stars, and black holes became
'household' words in the literature of astrophysics. Many of the objects were clearly so
condensed as to require the application of the General Theory of Relativity or some other
gravitational theory for their description. When one postulates the existence of a "new" object it
is always wise to subject that object to a stability analysis. This is particularly important for
highly collapsed objects as the time scale for development of the instability will be very short.
Since general relativistic effects can usually be viewed as an effective increase in the
gravitational force, one would expect its presence to decrease the stability of objects in which it
is important. What came as a surprise is the importance of these effects where one would
normally presume them to be of little or no importance.
80
The Virial Theorem in Stellar Astrophysics
Apparently inspired by a comment of R. P. Feynman in 1963, W. A. Fowler noted that
effects of general relativity would lead to previously unexpected instabilities in supermassive
stars
l
.
†
Noting that the conditions for this instability also exist in massive white dwarfs,
Chandrasekhar and Tooper
2
showed by means of rather detailed calculations that a white dwarf
would become unstable when its radius shrank to about 246 Schwarzschild radii or on the order
of 1000 km. This corresponds to a mass about 1.5% below the wellknown Chandrasekhar
limiting mass for degenerate objects. During the next 15 years, this instability received a great
deal of attention and I will not attempt to fully recount it here. Rather, let us examine with the aid
of hindsight and the virial theorem, how this result could be anticipated without the need of
detailed calculations.
One can see that the stability analysis coupled with the postNewtonian form of the virial
theorem given in Chapter II [equation (2.4.15)] would serve as the basis for investigating this
effect. However the estimation or calculation of the relativistic terms on the right hand side of
equation (2.4.15) is extremely difficult. Instead, by assuming spherical symmetry we may start
with the spherically symmetric equation of motion given by Meltzer and Thorne
3
as did Fowler
4
and follow the formalism of Chapter III. Thus
2 2
2
2
2
2
2
c
r G 4
r
) r ( Gm
c
P
1
rc
) r ( Gm 2
dt
dr
c
1
dr
dP 1
dt
d ρ π
− −
(
(
(
(
(
¸
(
¸
ρ
+
− 
.

\

+
ρ
− = 
.

\

y
dt
dr
y y , 4.1.1
where
( )
0
2
c / P
ρ
+ ρ
= y .
If we confine our attention to objects nearly in equilibrium, no large scale radial motions
can exist. Thus the term involving (dr/dt)
2
can be neglected and equation (4.1.1) becomes
2 2 2
2
2
2
2
c
Pr G 4
r
) r ( Gm
c / P 1
rc / ) r ( Gm 2 1
dr
dP 1
dt
r d π
− −
(
¸
(
¸
ρ +
−
ρ
= y , 4.1.2
or in the postNewtonian approximation (i.e.. keeping only terms of the order 1/c
2
)
2 2 2 2 2
2
2
c
r GP 4
r
) r ( Gm
rc
) r ( Gm 2
c
P
1
dr
dP
dt
r d ρ π
−
ρ
=
(
¸
(
¸
− ⋅ ⋅ ⋅ − −
ρ
− = ρy . 4.1.3
†__________________________________________
See Fricke
9
who also uses a postNewtonian virial approach to this problem.
81
The Virial Theorem in Stellar Astrophysics
In hydrostatic equilibrium, dP/dr is given by OppenheimerVolkoff as
( ) 

.

\

−
π
+ + ρ
− =
2
2
2
3
2
rc
) r ( Gm 2
1 r
c
P r 4
) r ( m
c
P
G
dr
dP
, 4.1.4
or

.

\

+
(
¸
(
¸
ρ π +
ρ
+
ρ
+
ρ
− =
4
2
2 4
2 2
2 2
c
1
O rG 4
r
) r ( Gm
r
) r ( m G 2
c
1
r
) r ( Gm
dr
dP
. 4.1.5
If we retain dP/dr explicitly for the expansion of the first term of equation (4.1.3) all other
relativistic terms of equation (4.1.5) will be of the order l/c
4
in the product in equation (4.1.3).
Thus, equation (4.1.3) becomes
2 2 2 2 2
2
2
c
rGP 4
rc
) r ( Gm 2
c
P
1
r
) r ( Gm
dr
dP
dt
r d ρ π
−
(
¸
(
¸
+ ⋅ ⋅ ⋅ + +
ρ
+
ρ
− − = ρy . 4.1.6
Now if we again form Lagrange's identity by multiplying by r and integrating over all volume,
we get
dV
c
rGP 4
dV
rc
) r ( m G
2
rc
) r ( Gm
dV
r
) r ( Gm
dr
dr
dP
r 4 dV
dt
r d
r
V
2
V
2
2
2 2
V V
2
R
0
3
V
2
2
2
∫ ∫ ∫ ∫ ∫ ∫
ρ π
−


.

\

−
ρ
−
ρ
− π − ==


.

\

ρy . 4.1.7
The last integral can be integrated by parts,
4.1
so that
dV
c r
) r ( m G
dV
c
rGP 4
V
2 2
2 2
V
2 ∫ ∫
ρ
=
ρ π
. 4.1.8
With somewhat less effort the first integral becomes
∫ ∫ ∫ ∫
− = π − π = π =

.

\

π
V
R
0
2
R
0
3
R
0
3
R
0
3
PdV 3 Pdr r 12 P r 4 dP r 4 dr
dr
dP
r 4 . 4.1.9
Putting the results of equation (4.1.8) and equation (4.1.9) into equation (4.1.7), noting that the
first term on the right hand side of equation (4.1.9) is zero, and rewriting the left hand side in
terms of a relativistic moment of inertia we get
∫ ∫ ∫
ρ
− − Ω − =
V V
2
2 2
2 2
V
2
r
2
2
1
dV
r
) r ( m G
c
3
dV
r
P ) r ( Gm
c
1
PdV 3
dt
I d
. 4.1.10
which is equivalent to equation (2.4.15) of Chapter II for spherical stars but vastly simpler.
Although this approach, which is basically due to Fowler
4
, lacks the rigor of the EIH post
Newtonian approach, it does yield the same results for spherical stars nearly in hydrostatic
equilibrium. It is worth noting that the relativistic correction terms are of the same mixed energy
integrals as those that appear in equation (2.4.15).
82
The Virial Theorem in Stellar Astrophysics
Taking the variation of equation (4.1.10) as we did in Chapter III, we have
( )
(
¸
(
¸
ρ
δ −
(
¸
(
¸
δ − Ω δ −
(
¸
(
¸
δ = δ
∫ ∫ ∫
dV
r
) r ( m G
c
3
dV
r
P ) r ( Gm
c
1
PdV 3 I
dt
d
V
2
2 2
2
V
2
V
r
2
2
2
1
. 4.1.11
As before, let us suppose that the variation of these quantities results from a variation of the
independent variable δr. Further assume that δr/r = ξ
o
e
iσt
where ξ
o
is constant and the variation is
adiabatic. Since we can write the internal heat energy density as (Γ
1
−1)u = P, the first term
becomes
U δ > − Γ < = δ
∫
1 3 PdV 3
1
V
. 4.1.12
Equation (3.2.10) (i.e. Chapter III) leads to
Ω ξ − = Ω δ . 4.1.13
The variation of the relativistic correction terms can be computed as follows: Let
∫
= Ω
M
0
2 2
2 2
2
3
1
) r ( dm
c r
) r ( m G
. 4.1.14
so that the variation of the last term in equation (4.1.11) is
1
M
0
2
2 2
2
1
2 ) r ( dm
r
1
) r ( m G
c
3
2 Ω ξ − =

.

\

δ = Ω δ
∫
. 4.1.15
It is convenient (particularly for the relativistic terms) to normalize by the dimensionless quantity
(2GM/Rc
2
). Thus,
∫ ∫

.

\

=

.

\



.

\


.

\

= Ω
1
0
2
2
2
2
8
3
2
2
2
2 4
1
2
2
3
1
dq ) x / q (
Rc
GM 2
Mc
M
) r ( dm
r
R
M
) r ( m
Rc
GM 2
Mc , 4.1.16
where the dimensionless variables are q = [m(r)/M], and x = r/R. The remaining terms in 4.1.11
can be normalized in a similar way by making use of the homologous dependence of P. That is
4 2
r / ) r ( Gm P η = , 4.1.17
where η is a dimensionless scale factor. Therefore, we can let
∫ ∫ ∫

.

\

πη
(
¸
(
¸
=
πη
= =
V
1
0
3 2
2
2
R
0
5
2 3 2
2 2 2
1
dx
x
q
Rc
GM 2
Mc
r
dr r ) r ( m G 4
c
1
r
PdV ) r ( Gm
c
1
P . 4.1.18
As in equation (4.1.16), the integral in equation (4.1.18) is dimensionless and determined by the
equilibrium model. Thus the remaining equation is
1 1
P 2 P ξ = δ . 4.1.19
Replacing P with u(Γ
1
−1) as with the first term and letting
∫
> − Γ <
= =
V
1
1
2
1
1
1
dV
r
) r ( uGm
c
1
P U . 4.1.20
Equation (4.1.11) then becomes:
83
The Virial Theorem in Stellar Astrophysics
( )
1 1 1 r
2
2
2
1
2 1 1 3 I
dt
d
Ω δ − δ > − Γ < + Ω δ δ > − Γ < = δ U  U
1
. 4.1.21
Now, since the internal energy U is coupled with all other terms including the relativity terms we
shall eliminate it in a somewhat different fashion than in Chapter III. Since the total energy must
be constant, its variation is zero. Thus
1 1
0 E Ω δ − δ + Ω δ − δ = = δ U U , 4.1.22
and equation (4.1.21) becomes
( )
1 1 1 1 1 r
2
2
2
1
5 3 1 2 4 3 I
dt
d
Ω δ > − Γ < − δ > − Γ < − Ω δ > − Γ =< δ U . 4.1.23
Substituting in the variations from equations (4.1.13), (4.1.15, and (4.1.19) into equation (4.1.23)
and noting that two time differentiations of the perturbation will give a σ
2
in the first term,
equation (4.1.23) becomes
1 1 1 1 0 1 r
2
5 3 2 1 4 4 3 I Ω > − Γ < + > − Γ < − Ω > − Γ =< σ U y . 4.1.24
Making one last normalization of Ω
0
which for polytropes is
2
2
0
Mc
Rc
GM 2
2
3
n 5
1
R
GM
n 5
3

.

\


.

\

−
=

.

\

−
= Ω , 4.1.25
and calling the dimensionless integrals in equation (4.1.16) and equation (4.1.18), ζ
1
and ζ
2
respectively, equation (4.1.24) becomes
( )
(
¸
(
¸
> Γ − < ς + > − Γ < ς

.

\

− > − Γ <
−

.

\

= σ
1 2 1 1
2
1
2
2
r
2
3 5 2 1 4
Rc
GM 2
4 3
) n 5 ( 2
3
Rc
GM 2
Mc I y . 4.1.26
Since the average relativity factor y is always positive, this expression can be used as a stability
criterion as in Chapter III. That is
( )


.

\

−
> − Γ <
< > Γ − < ς + > − Γ < ς 
.

\

) n 5 ( 2
4 3 3
3 5 2 1 4
R
R
1
1 2 1 1
S
. 4.1.27
where we have used the fact that the Schwarzschild radius (R
S
) is 2GM/c
2
. We can now use this
to investigate the stability of white dwarfs as they approach the Chandrasekhar limiting mass. As
this happens, the equation of state approaches that of relativistically degenerate electron gas and
the internal structure, that of a poly trope of index, n = 3.
84
The Virial Theorem in Stellar Astrophysics
As n → 3, Γ
l
→ 4/3, and the system becomes unstable. Thus let
ε + = Γ
3
4
1
, 4.1.28
and equation (4.1.27) becomes (using Fowler’s
4
values forζ
1
and ζ
2
),
( )
¦
¦
¦
)
¦
¦
¦
`
¹
ε
>
≥ − ε = ς + ς −
ε
1.13
R
R
or
0
R
R
5 . 2
4
9
2
R
R
4
9
0
S
0
S
2 1 3
4
0
S
. 4.1.29
Thus, if you imagine a sequence of white dwarfs of increasing mass, the value of (R
o
/R
s
)
will monotonically decrease as a result of the mass radius relation for white dwarfs and (l/ε) will
monotonically increase as the configuration approaches complete relativistic degeneracy.
Clearly, the point must come where the system becomes unstable and collapses. However, in
order to find that point, we need an estimate of how ε changes with increasing mass. For that we
turn to an interesting paper by Faulkner and Gribben
5
who show
4.2
3
x 2
2 −
≅ ε , 4.1.30
where x is the Chandrasekhar degeneracy parameter.
So, our instability condition can be written as
2
0
S
x 7 . 1
R
R
> . 4.1.31
All that remains is to estimate an average value of the degeneracy parameter x which we
can expect to be much larger than 1. From Chandrasekhar
6
( )
3 3 3 4
e
3
e
x h 3 / c m 8 Bx π = = ρ . 4.1.32
Now neglecting inverse β decay the local density will be roughly given by ρ = m
p
ρ
e
/m
e
and
( )
ρ
(
(
¸
(
¸
π
=
p
3 3
e
3
3
m c m 8
h 3
x . 4.1.33
Let ρ be given by its average value so that
( )
2
3
2
3
2
p
3 3
e
2
3
2
R
M
m c m 32
h 9
x
(
(
¸
(
¸
π
= . 4.1.34
Normalizing R by Schwarzschild radius we get
85
The Virial Theorem in Stellar Astrophysics
2
x = 7x10
6
(M
⊙
/M)
4/3
(R
S
/R
0
)
2
. 4.1.35
This can be rigorously combined equation (4.1.31) to provide value for (Ro/Rs).
However, since this entire argument is illustrative we also assume that the mass is roughly the
limiting mass for white dwarfs. Then equation (4.1.31) becomes:
(R
0
/R
S
) > 228(M
⊙
/M)
4/9
≅ 200 , 4.1.36
which is in remarkable agreement with the more precise figure of Chandrasekhar and Tooper
2
of
246. It is most likely that the discrepancy arises from the rather casual way of estimating
2
x
since it will be affected by both the type of volume averaging to determine ρ and the details of
the equation of state used in relating ρ to ρ
e
. However, it should be remembered that the result is
also only correct in the post Newtonian approximation and is an inequality setting a lower limit
on instability. The interesting result is that General Relativity becomes important, indicating that
instability sets in, at many times the Schwarzschild radius. This is the same result that Fowler
found for supermassive stars supported by radiation pressure and serves as some justification
for using the postNewtonian approximation. It should be noted that substituting R
0
into the
massradius relation for white dwarfs suggests that the critical mass should only be reduced by
1.5 percent. Hence the Chandrasekhar limiting mass for white dwarfs, while being somewhat too
large, is still an excellent approximation.
2. The Influence of Rotation and Magnetic Fields on White
Dwarf Gravitational Instability
At this point the reader is likely to complain that the derivation indicating the presence of
an instability resulting from general relativity has been anything but brief. The length results
largely from a somewhat different approach to the general relativistic term than used earlier.
That the approach succeeds at all is largely a result of presumed spherical symmetry. However,
to further demonstrate the efficacy of this approach let us consider what impact rotation and
magnetic fields may have on the results of the last section. Fowler found that a very small
amount of rotation would stabilize larger supermassive stars against the gravitational instability
so one might wonder what would be the effect in white dwarfs. However, the situation for white
dwarfs is quite different. Here the gravitational field is proportionally much stronger with γ being
driven to 4/3 by the equation of state and not the radiation field. Thus we may expect that a much
larger rotational energy field is required to bring about stability than is the case for supermassive
stars. In spite of this expectation, we shall assume that the effects of rotation and magnetic fields
are not so extreme as to significantly alter the spherical symmetry.
Under these conditions, the Newtonian approach of Chapter III will suffice to calculate
the terms to be added to the equations of motion and to perform the required variational analysis.
In Chapter III, we defined the rotational kinetic energy T
3
and magnetic energy M
o
as
86
The Virial Theorem in Stellar Astrophysics
¦
¦
¦
)
¦
¦
¦
`
¹
π
=
ω =
∫
∫
V
2
0
2
1
3
dV
8
H
d
M
L T
L
, 4.2.1
which have variations
)
`
¹
ξ − = δ
ξ − = δ
M M
T T ) 0 ( 2
3 3
. 4.2.2
Adding this to the variational form of Lagrange's identity in section 1 [equation (4.1.21)], we get
( )
1 1 3 1 r
2
2
2
1
2 1 2 1 3 I
dt
d
Ω δ − δ > − Γ < + δ + δ + Ω δ δ > − Γ < = δ U M T  U
1
. 4.2.3
.
Now the condition on the variation of the total energy becomes
1 1 3
0 E Ω δ − δ + δ + δ + Ω δ − δ = = δ U M T U , 4.2.4
which enables us to rewrite 4.2.3 as
( ) ) ( 3 5 1 2 ( 4 3 I
dt
d
1 1 1 1 r
2
2
2
1
Ω δ − δ > Γ − < + δ > − Γ < − δ − δ > − Γ =< δ
3 1
T U M) U . 4.2.5
Substituting in the values for the variations we get an expression analogous to equation (4.1.24)
)] 0 ( [ 5 3 2 1 4 ) ( 4 3 I
3 1 1 1 1 0 0 1 r
2
T U M − Ω > − Γ < + > − Γ < − − Ω > − Γ =< σ y . 4.2.6
These expressions differ from those in Chapter III only because the gravitational potential
energy is taken here to be positive. In order for us to proceed further it will be necessary to
normalize both the rotational energy and magnetic field by something. Let us consider the case
for ridged rotation so that
I I
2
3
1
z
2
2
1
3
ω = ω = T . 4.2.7
Here we are ignoring the relativistic corrections to I and take I = αMR
2
. In addition, let us
normalize the angular velocity ω by the critical value for a Roche model. Then


.

\

=


.

\

= ω
2
0 0
2 2
3
0
2 2
c R
GM 2
R 27
c w 4
R 27
GM 8
w . 4.2.8
87
The Virial Theorem in Stellar Astrophysics
This certainly does not imply that we are in any way assuming that white dwarfs are represented
by a Roche model but rather that it merely provides us with a convenient scale factor. Thus


.

\
 α
=
0
S 2 2
R
R
Mc w
81
4
3
T . 4.2.9
In a similar manner let us normalize the magnetic energy M
o
by the energy sufficient to bring
about disruption of the star. In Chapter III we showed that if other effects were absent then
M
o
> Ω would disrupt the star. Using this as the normalization constant we have


.

\

(
¸
(
¸
−
=
0
S 2 2
0
R
R
Mc
) n 5 ( 2
3
H M . 4.2.10
Under these conditions we can expect the maximum values
for w and H to be
)
`
¹
=
=
1
1 w
H
, 4.2.11
and in any event the assumption of sphericity will probably break down for w > 0.8 and H > 0.3.
Putting these values for T
3
and M
o
along with the previously determined values for Ω, U
l
, and
Ω
l
, into equation (4.2.6) we can arrive at stability conditions analogous to equation (4.1.27).
Namely
( )
81
w 8
) n 5 ( 2
) 1 ( 4 3 3
3 5 2 1 4
R
R
2 2
1
1 2 1 1
S
α
+


.

\

−
− > − Γ <
< > Γ − < ς + > − Γ < ς 
.

\
 H
. 4.2.12
As before, let us pass to the case where n = 3, so that
( )
81
w 8
) 1 ( 2
R
R
2
2
2
9
2 1 3
4 S
α
+ − ε < ς + ς 
.

\

H . 4.2.13
With α= 0.113 for polytropes of n = 3 and again using Fowler's
4
values for ζ
1
and ζ
2
this
becomes
2 3 2
0
S
w 10 4 . 4 ) 1 ( 89 . 0
R
R
−
× + − ε <


.

\

H . 4.2.14
Using the same analysis for ε as before
) 1 ( 10 4 . 8
R
R
2 8
0
S
H − × <


.

\

−
(M/M
⊙
)
4/3
2 3
0
S
w 10 4 . 4
R
R
−
× +


.

\

, 4.2.15
and taking M to be near the Chandrasekhar limit, we have
88
The Virial Theorem in Stellar Astrophysics
0
) 1 (
10 3 . 9
) 1 (
w 10 1 . 4
R
R
R
R
2
6
2
2 4
S
0
3
S
0
>
−
×
−
−
×


.

\

+


.

\

−
H H
. 4.2.16
For a point of reference it is worth mentioning the size of the normalization quantities so
that various values of w and H may be held in perspective. If we assume that we are dealing with
objects on the order of 10
3
km then the disruption field is of the order of 3 x 10
15
gauss while the
critical equatorial velocity would be about 10
4
km/sec. The largest observed fields in white
dwarfs reach 10
8
gauss and although it can be argued that larger fields may be encountered in
more massive white dwarfs, fields in excess of 10
12
gauss would not seem to be supported by
observations. Thus, a plausible upper value of H would be on the order of 10
3
. If, for the
moment we neglect rotation, equation (4.2.14) can be written as
3
1
2
S
0
) 1 ( 210
R
R
−
− >


.

\

H . 4.2.17
So it is clear that the only effect the field has is to act with general relativity to further destabilize
the star. However, for the field to make any appreciable difference it will have to be truly large.
For our plausible upper limit of H ≈10
3
the effect is to increase the radius at which the
instability sets in by about 1%.
The situation regarding rotation is slightly more difficult to deal with as the resulting
inequality is a cubic. With H = 0, equation (4.2.14) becomes
0 10 3 . 9 w 10 1 . 4
R
R
R
R
6 2 4
S
0
3
S
0
> × − ×


.

\

+


.

\

−
. 4.2.18
An additionally useful expression for the equatorial velocity which corresponds to a given w is
(km/sec)
R
R
w 10 15 . 1
R
R
3 3
cw 2
R v
2
1
0
S 5
2
1
0
S
eq eq


.

\

× =


.

\

= ω = . 4.2.19
If we pick a few representative values of w and solve equation (4.2.17) by means of the general
cubic we get the results below for the associated values of (R
o
/R
S
) and V
eq
.
Rotational Effects on the White Dwarf Instability Limit
w
0 0.1 0.5 1.0
v
eq
(km/sec)
0 794 4128 9453
R
o
/R
S
210 209 194 148
Even these few values are sufficient to indicate that although rotation helps to stabilize
the star in the sense of allowing it to attain a smaller radius before collapse, nevertheless, like
89
The Virial Theorem in Stellar Astrophysics
magnetic fields, the effect is small. One may choose to object to the assumption of rigid rotation
as being too conservative. However, it is clear from the development that for either rotation or
magnetic fields to really play an important role the total energy stored by either mechanism must
approach that in the gravitational field. In order to do this with differential rotation, the
differential velocity field would have to be alarmingly high. It seems likely that the resulting
shear would produce significant dynamical instabilities.
Thus, we have seen that neither magnetic fields nor rotation can significantly alter the
fact that a white dwarf will become unstable at or about 1000 km. Classically, the star reaches
this point when it is within, but less than, a few percent of the Chandrasekhar limiting mass. So it
is not the limiting mass resulting from the change in the equation of state that keeps us from
observing more massive white dwarfs. Rather it is the presence of general relativistic instability
that destroys any more massive objects.
It is quite simple to dismiss this argument as 'nitpicking' as the mass at which the
instability occurs is nearly identical to the Chandrasekhar limiting mass. However, when one
tries to generalize the results of one problem to another, it is conceptual errors such as this that
may lead to much more serious errors in the generalization. As we shall see in the next section,
this is indeed the case with neutron stars.
3. Stability of Neutron Stars
A second class of objects whose existence became well established during the 1960's is
the neutron stars. It is a commonly held misconception that a neutron star is nothing more than a
somewhat collapsed white dwarf, since the masses are thought to be similar. In reality the ratios
of typical white dwarf radii to neutron stars is suspected to be nearly 1000 which is just about the
ratio of the sun's radius to that of a typical white dwarf. A similar misconception relates to the
notion of the neutron star’s limiting mass. It is popular to suggest that since the mass limit in
white dwarfs arises as a result of the change in the equation of state so a similar change in the
equation of state for a neutron star yields a limiting mass for these objects. It is true that a
limiting mass exists for neutron stars but this limit does not primarily arise from a change in the
equation of state.
Let us consider a very simple argument to dramatize this point. The equation of the state
change that results in the Chandrasekhar limit occurs because the electrons achieve relativistic
velocities. If this were to happen in neutron stars, the configuration would still have to satisfy the
virial theorem.
For the moment let us ignore the effects of general relativity and just consider the special
relativistic virial theorem as we derived in Chapter II [equation (2.3. 9)].
90
The Virial Theorem in Stellar Astrophysics
∫
γ τ + + Ω =
V
2
r
2
2
1
dV ) / ( T
dt
I d
, 4.3.1
where now γ = (1 v
2
/c
2
)
1/2
.
As the neutrons become relativistic γ
1
→ 0 and T = αMc
2
where α >>1. We may write
the gravitational potential energy as


.

\
 η
− = η − = Ω
0
S
2
0
2
R
R
2
Mc
R
GM
. 4.3.2
The variational form of the virial theorem will require that
T + Ω < 0 , 4.3.3
so that
Mc
2
[α−(η/2)(R
S
/R
0
)] < 0 , 4.3.4
or (R
0
/R
S
) > η/2α .
Since η is of the order of unity and α >> 1, this would require that the object have a radius less
than the Schwarzschild radius in order to be stable against radial pulsations. This is really
equivalent to invoking Jacobi's stability condition on the total energy.
This simplistic argument can be criticized on the grounds that it ignores general relativity
which can be viewed as increasing the efficiency of gravity. Perhaps the "increased gravity"
would help stabilize the star against the rapidly increasing internal energy. This is indeed the
case for awhile. However, based on the analysis in section 1, as the value of Γ approaches 4/3 the
same type of instability which brought about the collapse of the white dwarfs will occur in the
neutron stars. The exact value of R
o
/R
S
for which this happens will depend on the exact nature of
the equation of state as well as details of model construction. However, since the general
relativistic correction terms will be much larger than in the case of white dwarfs, we should
expect the value of Γ to depart farther from the relativistic limit of 4/3 than before. That this is
indeed the case is clearly shown by Tooper
7
in considering the general properties of relativistic
adiabatic fluid spheres. He concludes that the instability always occurs before the gas has
become relativistic at high pressures. Unfortunately we cannot quantitatively apply the results of
Section 1 since the way in which Γ approaches 4/3 (more properly the way in which Γ departs
from 5/3) depends in detail on the equation of state. However, we may derive some feeling for
the way in which the instability sets in by assuming the compression has driven the value of Γ
down from 5/3 to 3/2 (i.e., just half way to its relativistic value. Substitution of Γ = 3/2 in
equation (4.1.27) and using Fowler’s
4
values of ζ
l
and ζ
2
for a polytrope of index 2 gives a
stability limit of
3 . 4
R
R
S
0
>


.

\

. 4.3.5
Thus R
o
for a neutron star would have to be greater than about 12 km. Since typical model radii
are of the order of 10 km, 3/2 is probably a representative value of Γ, yet it is still far from the
relativistic value of 4/3. This argument further emphasizes the fact that it is the general
91
The Virial Theorem in Stellar Astrophysics
relativistic instability which places an upper limit on the size of the configuration, not the
equation of state becoming relativistic.
The discussion in section 2 would lead us to believe that neither rotation nor magnetic
fields can seriously modify the onset of the general relativistic instability. This can be made
somewhat quantitative by evaluating equation (4.2.10) for a polytrope of index n = 2. However,
in order to do this, we must reevaluate the moment of inertia weighting factor a. A crude
estimate here will suffice since we are neglecting an increase of perhaps a factor of 2 due to
general relativistic terms. One can show by integrating 4 by parts in Emden polytropic
variables that
∫
ρ π
R
0
2
dr r
(
(
(
(
(
¸
(
¸
ξ
θ
ξ
ξ ξθ
ξ
+ = α
ξ
ξ
∫
1
1
d
d
d
6
1
2
1
0
2
1
, 4.3.6
which for n = 2 very approximately gives α = 0.345. Substitution into 4.2.13 then yields
2 2 2
0
S
w 10 2 . 3 ) 1 ( 234 . 0
R
R
−
× + − <


.

\

H . 4.3.7
The effects of magnetic fields and rotation are qualitatively the same for neutron stars as
for white dwarfs. However, as R
o
is only a few times the Schwarzschild radius, the normalizing
fields and rotational velocities are truly immense. For a neutron star with a 10 km radius the
magnetic field corresponding to H = 1 in of the order of 10
18
gauss while the rotational velocity
corresponding to w = 1 would be about 10% the velocity of light. These values vastly exceed
those for the most extreme pulsar. Thus, barring modification to the equation of state resulting
from these effects, they can largely be ignored in investigating neutron star stability. This result
is exactly in accord with what one might have expected on the basis of the white dwarf analysis.
We began this discussion by indicating that the analysis would be very simplistic and yet
we have attained some very useful qualitative results. Since the massradius law for any
degenerate equation of state (excepting small technical wiggles) will provide for stars whose
radius decreases with increasing mass, we can guarantee that the resulting decrease in Γ will give
rise to an unstable configuration at a few Schwarzschild radii. Thus, there will exist an upper
limit to the mass allowable for a neutron star. The origin of this limit is conceptually identical to
that for white dwarfs. Furthermore, as for white dwarfs, this limit can be modified by the
presence of magnetic fields and rotation only for the most extreme values of each. One may
argue that in discussing effects of general relativity we have included terms of O (1/c
2
) and that
higher order effects may be important. While this is true regarding such items as gravitational
radiation, none of these terms should be important unless the configuration becomes smaller than
several Schwarzschild radii.
92
The Virial Theorem in Stellar Astrophysics
Even then, they are unlikely to affect the qualitative behavior of the results. Very little
has been said about the large volume of work relating to the equation of state for neutron
degenerate matter. This is most certainly not to deny its existence, just its relevance. One of the
strong points of this approach is that insight can be gained into the global behavior of the object
without undue concern regarding the microphysics. This type of analysis is a probe intended to
ascertain what effects are important in the construction of a detailed model and what may be
safely ignored. It cannot hope to provide the information of a detailed structural model but only
point the way toward successful model construction.
4. Additional Topics and Final Thoughts
It would be possible and perhaps even tempting to continue demonstrating the efficacy of
the virial theorem in stellar astrophysics. However, attempting to exhaust the possible
applications of the virial theorem is like trying to exhaust the applicability of the conservation of
momentum. I would be remiss if I did not indicate at least some other possible areas in which the
virial theorem can lend insight.
In Chapter III, section 4 we discussed the variational effect of the surface terms resulting
from the application of the divergence theorem. These terms are generally neglected and for
good reason. In most cases the bounding surface can be chosen so as to include the entire
configuration. In instances where this is not the case, such as with magnetic fields, the term is
still generally negligible. If one considers the form of the surface terms given by equation
(2.5.19) (Chapter II), compared to the volume contribution [i.e. equation (2.5.18)], the ratio of
the scalar value is
> <
=
•
= ℵ
∫
∫
P
P
PdV
d P
0
V
S
0
S r
, 4.4.1
where <P> is the average value of the internal pressure. Since, in any equilibrium configuration
the pressure must be a monotone increasing function as one moves into the configuration, ℵ< 1.
In general it is very much less than one. However, in the case where γ→ 4/3 the variational
contribution of the surface pressure approaches
∫
• S r d P 3
0
, while the internal pressure and
Newtonian gravity contributions vanish. Such terms are then available to combine with the
effects of general relativity. Since they are of the same sign as the general relativistic terms the
surface terms will only serve to increase the instability of the entire configuration. This results in
an increase of the radius at which white dwarfs become unstable. Thus, the accretion of matter
onto the surface of such objects may cause them to collapse sooner than one might otherwise
expect.
93
The Virial Theorem in Stellar Astrophysics
Since magnetic fields also are usually assumed to increase inward, the influence of
magnetic surface terms will be similar to that of a nonzero surface pressure. Like the surface
pressure terms, they will in general be small compared to the internal contributions. Only in case
of a system with an effective γ approaching 4/3 could these small terms be expected to exert a
trigger effect on the resulting configuration.
There is one instance in which the use of the surface terms can be significant. One of the
most attractive aspects of this entire approach is that it can deal with the properties of an entire
system. Indeed the spatial moments are taken in order to achieve that result. However it is
interesting to consider the effects of applying the virial theorem to a subvolume of a larger
configuration. Clearly, as one shrinks the volume to zero he recovers the equations of motion
themselves multiplied by the local positional coordinate. If one considers a case intermediate to
these limits and investigates the stability of a subvolume which could include the surface of the
star, it would be possible to analyze the outer layer for instabilities which might not be globally
apparent. It is true that a local stability criterion would be sufficient to locate such instabilities
but one could not be sure how the instabilities would propagate without carrying out a large
structural analysis. This latter effect can be avoided by utilizing a subglobal form of the virial
theorem. Under these conditions one might expect the surface terms to be the dominant terms of
the resulting expression.
In discussing some of the more bizarre and contemporary aspects of stellar structure it is
easy to overlook the role played by the virial theorem in the development of the classical theory
of stellar structure. It is the virial theorem which provides the theoretical basis for the definition
of the KelvinHelmholtz contraction time. This is just the time required for a star to radiate away
the available gravitational potential energy at its present luminosity. It is the virial theorem
which essentially tells how much of the gravitational energy is available. Thus if the contraction
liberating the potential energy is uniform and d
2
I/dt
2
is zero then the total kinetic energy always
must be
T = 1/2 Ω . 4.4.2
This makes the other half of the gravitational energy available to be radiated away. The Kelvin
Helmholtz contraction time for poly tropes is thus
n 5
10 5 . 4
RL ) n 5 ( 2
GM 3
7 2
−
×
≅
−
= T (M/M
⊙
)
2
(R
⊙
L
⊙
/RL) (years). 4.4.3
Reasoning that this provided an upper limit to the age of the sun Lord Kelvin challenged the
Darwinian theory of evolution on the sound ground that 23 million years (i.e. KHT for a
polytrope with an internal density distribution of n=3)
8
was not long enough to allow for the
evolutionary development of the great diversity of life on the planet. The reasoning was flawless,
only the initial assumption that the sun derived its energy from gravitational contraction which
was plausible at the time, failed to withstand the development of stellar astrophysics.
94
The Virial Theorem in Stellar Astrophysics
Another aspect of classical stellar evolution theory is clarified by application of the virial
theorem. All basic courses in astronomy describe postmain sequence evolution by pointing out
that the contraction of the core is accompanied by an expansion of the outer envelope. Most
students find it baffling as to why this should happen and are usually supplied with
unsatisfactory answers such as "it's obvious" or "it's the result of detailed model calculations"
which freely translated means "the computer tells me it is so." However, if the virial theorem is
invoked, then once again any internal rearrangement of material that fails to produce sizable
accelerative changes in the moment of inertia will require that
2T+Ω = 2EΩ = 0 . 4.4.4
Since the only way that the star can change its total energy E without outside intervention is by
radiating it away to space, any internal changes in the mass distribution which take place on a
time scale less than the Kelvin Helmholtz contraction time will have to keep the total energy and
hence the gravitational potential energy constant. Now
Ω = αM
2
/R . 4.4.5
where α is a measure of the central condensation of the object, so, as the core contracts and α
increases, R will have to increase in order to keep Ω constant. In general the evolutionary
changes in a star do take place on a time scale rather less than the contraction time and thus we
would expect a general expansion of the outer layer to accompany the contraction of the core.
The microphysics which couples the core contraction to the envelope expansion is indeed
difficult and requires a great deal of computation to describe it in detail. However the mass
distribution of the star places constraints on the overall shape it may take on during rapid
evolution processes. It is in the understanding of such global problems that the virial theorem is
particularly useful.
I have attempted throughout this book to emphasize that global properties are the very
essence of the virial theorem. The centrality of taking spatial moments of the equations of motion
to the entire development of the theorem demonstrates this with more clarity than any other
aspect. Although this global structure provides certain problems when the development is
applied to continuum mechanics nothing is encountered within the framework of Newtonian
mechanics which is insurmountable. Only within the context of general relativity may there lie
fundamental problems with the definition of spatial moments. Even here the first order theory
approximation to general relativity yields an unambiguous form of the virial theorem for
spherical objects. In addition, certain specific time independent or at least slowly varying cases
of the nonapproximated equations also yield unique results. Thus one can realistically hope that
a general formulation of the virial theorem can be made although one must expect that the
interpretation of the resultant spacetime moments will not be intuitively obvious.
The rather recent development of the virial theorem provides us with a dramatic example
of the fact that theories do not develop in an intellectual vacuum. Rather they are pushed and
shoved into shape by the passage of the time. Thus we have seen the virial theorem born in an
effort to clarify thermodynamics and arising in parallel form in classical dynamics. However the
similarity did not become apparent until the implications of the ergodic theorem inspired by
95
The Virial Theorem in Stellar Astrophysics
statistical mechanics were understood. Although sparsely used by the early investigators of
stellar structure, the virial theorem did not really attract attention until 1945 when the global
analysis aspect provided a simple way to begin to understand stellar pulsation. The attendant
stability analysis implied by this approach became the main motivation for further development
of the tensor and relativistic forms and provides the primary area of activity today. Only recently
has the similarity of virial theorem development to that of other conservation laws been clearly
expounded.
Recent criticism of some work utilizing the virial theorem, incorrectly attacks the
theorem itself as opposed to analyzing the application of the theorem and the attendant
assumptions. This is equivalent to attacking a conservation law and serves no useful purpose.
Indeed it may, by rhetorical intimidation, turn some less sophisticated investigators aside from
consideration of the theorem in their own problems. This would be a most unfortunate result as
by now even the most skeptical reader must be impressed by the power of the virial theorem to
provide insight into problems of great complexity. Although there is a tradeoff in that a
complete dynamical description of the system is not obtainable, certain general aspects of the
system are analyzable. Even though some might claim a little knowledge to be a dangerous
thing, I prefer to believe that a little knowledge is better than none at all. Thus, the perceptive
student of science will utilize the virial theorem to provide a 'first look' at problems to see which
are of interest. Used well this first look will not be the last.
Through the course of this book we have examined the origin of the virial theorem, noted
its development and applicability to a wide range of astrophysical problems, and it is irresistible
to contemplate briefly its future growth. In my youth the course of future events always seemed
depressingly clear but turned out to be generally wrong. Now, in spite of a better time base on
which to peer forward, the new future seems at best "seen through a glass darkly", and I am
mindful that astronomers have not had an exemplary record as predictors of future events
†
.
Nevertheless, there may be one or two areas of growth for the virial theorem on which we can
count with some certainty.
Immediate problems which seem ideally suited to the application of the virial theorem
certainly include exploration into the nature of the energy source in QSO's. Perhaps one will
finally observe that the gravitational energy of assembly of a galaxy or its components is of the
same order as the estimated energy liberated by a QSO during its lifetime. The virial theorem
implies that half of this energy may be radiated away. Thus, it would appear that one need not
look for the source of such energy but rather be concerned with the details of the "generator".
†___________________________
It is said that the great American astronomer, Simon Newcomb "proved" that heavier than air
flight was impossible and that after the Wright Brothers flew, it was rumored that he maintained
it would never be practical as no more than two people could be carried by such means.
96
The Virial Theorem in Stellar Astrophysics
Perhaps future development will consider applications of the virial theorem as
represented by Lagrange's identity. To date the virial theorem has been applied to systems in or
near equilibrium. It is worth remembering that perhaps the most important aspect of the theorem
is that it is a global theorem. Thus systems in a state of rapid dynamic change are still subject to
its time dependent form.
In the mid twentieth century, as a consequence of discovering that the universe is not a
quiet place, theoreticians became greatly excited about the properties of objects undergoing
unrestrained gravitational collapse. It is logical to suppose that sooner or later they will become
interested in the effects of such a collapse upon fields other than gravitation (i.e. magnetic or
rotational), that may be present. The virial theorem provides a clear statement on how the energy
in such a system will be shifted from one form to another as soon as one has determined d
2
I/dt
2
.
Future investigation in this area may be relevant to phenomena ranging from novae to quasars.
Perhaps the most exciting and at the same time least clear and speculative development in which
the virial theorem may play a role involves its relationship to general relativity. This is a time of
great activity and anticipatory excitement in fundamental physics and general relativity in
particular.
Perhaps through the efforts of Stephan Hawking and others, and as Denis Sciama has
noted, we are on the brink of the unification of general relativity, quantum mechanics, and
thermodynamics. Thermodynamics is the handmaiden of statistical mechanics and it is here
through the application of the ergodic theorem that the virial theorem may play its most
important future role.
You may remember that in Chapter II, difficulty in the interpretation of moments taken
over spacetime frustrated a general development of the virial theorem in general relativity and it
was necessary to invoke first order approximations to the relativistic field equations. In addition
the ergodic theorem seems inexorably tied to the nature of reversible and irreversible processes.
The advances in relating general relativity to thermodynamics bring these areas and theorems
into direct conceptual confrontation and may perhaps provide the foundations for the proper
understanding of time itself.
97
The Virial Theorem in Stellar Astrophysics
Notes to Chapter 4
4.1 The last integral can be integrated by parts so that
∫ ∫ ∫
∫ ∫
∫ ∫ ∫
π
+
π
−
π
+
ρ
=
(
¸
(
¸
+
π
+
π
−
ρ
=
π
−
π
−
π
=
ρ π
R
0
R
0
2 2
R
0
2
V
2 2
2 2
R
0
2
R
0
2
V
2 2
2 2
R
0
V
R
0
2
2
R
0
2
2
2
2
dr
dr
dP
) r ( m
c
G 8
dr
dr
dP
) r ( m
c
G 8
) r ( P ) r ( m
c
G 8
dV
c r
) r ( m G
dr
dr
dP
P ) r ( m
c
G 8
Pr ) r ( m
c
G 8
dV
c r
) r ( m G
rdr
c
P ) r ( Gm 8
dr
dr
dP
c
) r ( Gm r 4
c
) r ( m Pr G 4
dV
c
Pr G 4
N4.1.1
The third term vanishes since m(0) = 0 and P(R) = 0, the last two integrals cancel so that
dV
c r
) r ( m G
dV
c
Pr G 4
V
2 2
2 2
V
2
2
∫ ∫
ρ
=
ρ π
. N4.1.2
4.2 Start1ng w1th the polytropic equation of state
p=Kρ
γ
. N4.2.1
It is not hard to convince yourself that
ρ
ρ
=
ρ
= γ
d
dP
P n d
nP d
A
A
. N4.2.2
This can be reduced to a sing1e parameter by considering Chandrasekhar's parametric equation
of state for a nearly re1ativistic degenerate gas
6
3
Bx ), x ( Af P = ρ = , N4.2.3
where f(x) = x(2x
2
3) (x
2
+1) + 3 Sinh
1
(x). The limit of the hyperbolic sine is:
  x 2 ln ) x ( sinh
x
Lim
1
=
∞ →
−
. N4.2.4
Now consider the behavior of f(x) as x → ∞ .
¦
)
¦
`
¹
− ≅
+ ⋅ ⋅ ⋅ + − − + = ⋅+ ⋅ ⋅ + + − − ≅
− −
) x 2(x f(x)
or
x x 3 x x 2 ) x x )( 3 x )( 3 x 2 ( x ) x ( f
2 4
1
2
3
2 2 4 1
2
1
2 2
. N4.2.5
98
The Virial Theorem in Stellar Astrophysics
Simi1ar1y
( )
2
3
x 3
x 4 x 8
B
A
d
dx
dx
dP
d
dP −
=
ρ
=
ρ
. N4.2.6
Thus
( )
¦
¦
¦
¦
)
¦
¦
¦
¦
`
¹
≅ − + ⋅ ⋅ ⋅ + + + = −


.

\

−
+ = ε
−
(
¸
(
¸
−
−
=
(
¸
(
¸
−
−
≅ − γ = ε
−
− −
−
3
x 2
3
4
x x 2
3
2
3
4
1 x
x
1
3
2
or
3
4
1 x
) 1 x 2 (
3
2
x 3
) x 4 x 8 (
b
a
) x x ( A 2
Bx
2
4 2
2
2
2
2
3
4
2
3
2 4
3
3
4
. N4.2.7
99
The Virial Theorem in Stellar Astrophysics
References
1 Thorne, K. S. (1972), Stellar Evolution, (ed. HangYee Chin and Amador Muriel)
M.I.T. Press, Cambridge, Mass. p. 616.
2 Chandrasekhar, S. and Tooper, R. F. (1964), Ap. J. 139, p. 1396.
3 Meltzer, D. W. and Thorne, K. S. (1966), Ap. J. 145, p. 514.
4 Fowler, W. A. (1966), Ap. J. 144, p. 180.
5 Faulkner, J. and Gribbin, J. R. (1966), Nature, Vol. 218, p. 7347.
6 Chandrasekhar, S. (1957), An Introduction to the Study of Stellar Structure,
Dover Pub., pp. 360361.
7 Tooper, R. F. (1965), Ap. J. 142, p. 1541.
8. Chandrasekhar, S. (1957), An Introduction to the Study of Stellar Structure,
Dover Pub., pp. 454455.
9. Fricke, K.J. (1973). Ap. J. 183, pp. 941958.
100
The Virial Theorem in Stellar Astrophysics
101
Symbol Definitions and First Usage
Since this work contains a large number of concepts symbolically expressed, I felt it
might be useful if a brief definition of these symbols existed in some place for purposes of
reference. In general, the use of bold face type denotes a vector quantity; while an Old English
Text type font used for tensorlike quantities of rank 2 or higher. The exception is an outlined
font used for the unit tensor. Subscripted Old English type is used to represent the components of
these tensors. Various other type faces have been employed to provide symbolic representation
of scalar quantities which appear throughout this work. What follows is s list of the meaning of
these symbols and where they first appear.
Scalars and Special Parameters
Symbol Meaning First Used Symbol Meaning First Used
A An arbitrary scalar 2.4.10
A Constant in the degenerate equation
of state N4.2.3
B Constant in the degenerate equation
of state N4.2.3
D The magnitude of the electric
displacement vector N2.4.1
E Total energy of a system 1.1.8
F The magnitude of the radiative flux
1.1.10
G Gravitational constant 1.1.7
G A temporary quantity 1.2.2
H The magnitude of the magnetic field
intensity N2.4.1
I Moment of inertia about a coordinate
origin 1.2.3
I
r
Moment of inertia including
relativistic terms 2.3.15
I
z
Moment of inertia about the Zaxis
3.2,21
J Trace of the Maxwell tensor 2.4.3
K The constant of proportionality in the
polytropic equation of state N4.2.1
L(r) Stellar luminosity 1.1.10
Le Solar luminosity 4.4.3
L The magnitude of the angular
momentum vector 3.3.19
M Total internal magnetic energy 2.5.18
M
⊙
The mass of the sun 3.5.22
N The particle number density 2.5.21
P Total scalar pressure 1.1.6
P
g
Total gas pressure N3.2.7
P
1
A relativistic correction term 4.1.18
Q Arbitrary macroscopic system
parameter 1.5.1
<Q> Average Q 1.5.1
Q Arbitrary pointdefined system
property 2.2.3
Q
p
Surface pressure term 3.4.12
Defined in 3.4.8
Q
m
Magnetic surface term 3.4.11
Q
i
The generalized forces 3.5.1
R The Rydberg constant 3.3.2
R Stellar or configuration radius 3.2.20
R
S
The Schwarzschild 4.1.27
R
⊙
The solar radius 3.5.22
ℜ The total rotational kinetic energy
2.5.17
102
Symbol Meaning First Used Symbol Meaning First Used
S The 'creation rate' or collision term in
the Boltzmann transport equation
1.1.1
S The velocityaveraged ‘creation rate’
1.1.2
S The surface enclosing a volume V
2.5.12
T The total kinetic energy of the system
1.2.5
T The timeaveraged kinetic energy of
the system 1.2.15
<T> The phaseaveraged kinetic energy of
the system 3.5.8
T
0
A period of time 1.2.14
T KelvinHelmholtz contraction time
4.4.3
T Pulsation period 3.2.24
T The kinetic or gas temperature 2.5.20
T
1
The kinetic energy of radial motion
3.3.6
T
2
Thermal kinetic energy 3.3.2
T
3
Rotational kinetic energy 3.3.6
U Total potential energy 1.1.12
U The total internal heat energy 2.5.20
U The timeaveraged potential energy
of the system 1.2.15
U
1
The postNewtonian correction x c
2
to the Newtonian internal energy
4.1.20
V Volume enclosing the system 1.4.1
W A relativistic superpotential 2.4.13
Y A relativistic superpotential 2.4.13
Z A relativistic superpotential 2.4.13
a
ij
forcelaw proportionality constant
1.2.8
c The speed of light 2.3.3
c
p
Specific heat of constant pressure
2.5.21
c
v
Specific heat of constant volume
2.5.21
h Planck’s constant 4.1.32
H dimensionless magnetic field
intensity 4.2.10
k Boltzmann’s constant 2.5.21
m
i
mass of the ith particle 1.2.1
m
e
mass of the electron 4.1.32
m
p
mass of the proton 4.1.33
m(r) mass interior to a sphere of radius r
1.1.10
m(V) mass interior to a sphere of volume
V N3.5.2
n forcelaw exponent 1.2.8
n polytropic index 4.1.25
q A dimensionless mass 4.1.16
q
i
ith linearly independent coordinate
3.5.1
r radial coordinate 1.1.10
r
i
radial coordinate of the ith particle
1.2.2
r
i j
separation between the ith and jth
particles 1.2.8
s The proper length 2.3.3
t time 1.1.1
t
0
an initial time 1.5.1
u stream speed 2.4.6
u thermal energy density 4.1.20
v magnitude of a velocity vector 1.1.3
w Fractional angular velocity 4.2.8
x
α
components of Minkowski space
2.3.3
x Parametric variable in the degenerate
equation of state 4.1.30
x a dimensionless length 4.1.16
x a Cartesian coordinate 3.3.18
y a Cartesian coordinate 3.3.18
y normalized relativistic density 4.1.1
Γ
1
1
st
adiabatic constant 4.1.31
2
c / Π internal energy of a relativistic gas
2.4.7
Φ An arbitrary potential 1.2.8
<Φ> A relativistic superpotential 2.4.13
103
Symbol Meaning First Used Symbol Meaning First Used
χ Local energy of nonconservative
forces 1.1.8
ζ
1
A dimensionless measure of
polytropic structure 4.1.26
ζ
2
A dimensionless measure of
polytropic structure 4.1.26 ℑ A high order superpotential
2.2.10
η A dimensionless scale factor 4.1.17
ℜ A high order superpotential
2.2.10
η A proportionality constant 4.3.2
θ Polar angle in Spherical
Ψ The Newtonian potential 1.1.4
Coordinates 3.3.18
ℵ Ratio of average to surface pressure
θ The Polytropic Temperature 4.3.6
4.4.1
Ω Newtonian Gravitational potential
energy 1.2.13
ξ The Polytropic radial coordinate in
Emden Variables 4.3.6
ξ
1
The Polytropic Emden Radius 4.3.6
Ω Timeaveraged Gravitational
potential energy 1.2.16
ρ The local matter density 1.1.2
ρ
e
The electric charge density 2.5.4
<Ω> Phaseaveraged Gravitational
potential energy 3.5.8
ρ
e
The electric mass density 4.1.32
ρ* Modified matterenergy density 2.4.6
Ω
1
/c
2
PostNewtonian correction to the
Gravitational potential energy 4.1.14
σ Modified matterenergy density 2.4.6
σ Pulsational frequency 3.2.12
α parameter measuring central mass
concentration 4.2.8
τ Relativistic kinetic energy density
β The angle between the local Hfield
and r 3.4.14
N2.3.5
ϕ The source for a relativistic super
potential 2.4.9
β
i
proportionality constant for velocity
dependant forces 1.3.3
φ Azimuthal polar coordinate 3.4.5
ψ Phasespace point density 1.1.1
γ
2
1
2 2
) c / v 1 ( − 2.3.3
ω Magnitude of the angular velocity
N2.6.4, 3.3.18
γ
2
1
2 2
) c / v 1 (
−
− 4.3.1
γ ratio of specific heats 2.5.23
ε Energy generation rate from non
viscous sources 1.1.9
ε The total energy density 2.6.3
ε The potential energy density N2.3.6
ε A perturbation parameter 4.1.28
E Thermal kinetic energy density2.5.21
104
Vectors and Vector Components
Symbol Meaning First Used Symbol Meaning First Used
p
i
Momentum vector of the ith particle A An arbitrary vector N2.4.9
B The magnetic field vector 2.5.3 1.2.1
B
i
The components of B 2.5.2
p
i
Components of the momentum
vector of a particle 1.1.1 D The electric displacement vector
2.5.3 p The local momentum density 1.4.3
r
i
Radius vector to the ith particle 1.2.1 D
i
The components of D 2.5.2
E The electric field vector 2.5.3 r
i
Components of radius vector 2.5.23
E
i
The components of E 2.5.2 rˆ Unit vector in the rdirection page 2
F The radiative flux 1.1.8 u The local streamvelocity 1.1.2
F
i
Total force on the ith particle page 2 u
i
Components of the stream velocity
F
ij
Force between the ith & jth particles 2.1.4
1.2.7
u
α
Components of the 4velocity 2.3.3
G An arbitrary vector N2.4.9
u The Lagrangian displacement
velocity vector N3.6.1
H The magnetic field intensity 2.5.3
H
i
Cartesian components of H 2.5.2
v The local velocity vector 1.1.1
0
ˆ
H A unit vector along H 3.4.14
v
i
Velocity vector of the ith particle
1.2.1
K The relativistic linear momentum
density in the post Newtonian
approximation 2.4.11
w The local peculiar velocity in a
rotating coordinate frame 2.5.6
w Velocitydependant force vector1.3.1
S The vector “creation” rate 1.1.4
w The angular velocity field
Y A relativistic super potential 2.4.8
vector 2.5.5
dS The differential surface normal
vector 2.3.2
w
i
Cartesian components of the
angular velocity field vector N2.5.4
f The local force vector 1.1.1
ϖ The local residual velocity field in a
rotating coordinate frame N2.6.2
f
i
Force acting on the ith particle 1.2.1
f Local force density 1.4.3
ϖ
i
Cartesian components of the local
residual velocity field 2.5.15
f
f
The frictional force density 2.5.24
l The net local angular momentum
density 2.5.16
x A Cartesian coordinate vector 2.6.3
x
i
Components of the Cartesian
coordinate vector 1.1.1
n The Lagrangian displacement
vector 3.3.38
z
i
A velocity independent force vector
on the ith particle 1.3.1
η
i
The Cartesian components of the
Lagrangian displacement vector
N3.5.1
105
Tensors and Tensor Components
Symbol Meaning First Used Symbol Meaning First Used
f The frictional tensor acting as the
source for frictional forces 2.5.24
I The moment of inertia tensor 2.1.5
I
i j
Components of the moment of inertia
tensor 2.1.11
ℑ The Maxwell stressenergy tensor or
energy momentum tensor 2.3.1
ij
ℑ Components of ℑ 2.3.4
L The volume angular momentum
tensor 2.5.10, N 2.6.3
L
i j k
Components of L 2.5.15
M The magnetic energy tensor 2.5.14
M
i j
Components of M
P The pressure tensor 1.1.4
P
g
The gas pressure tensor 2.5.1
R
i j
Components of the Ricci tensor 2.4.3
S The surface energy tensor 2.5.12
S
i j
Components of S 2.5.12
T The kinetic energy tensor 2.1.5
T
i j
Components of T 2.1.11
U The potential energy tensor 2.1.5
U
i j
Components of U N2.2.1
1 The unit tensor 2.2.7
δ
ij
Components of 1 , the Kronecker
delta 2.4.5
ε
i j k
The components of the LeviCivita
tensor density N2.5.4
h
i j
Components of the metric
perturbation tensor 2.4.1
g
i j
Components of the metric tensor
2.4.1
106
Index
Angular Momentum, 43, 57
conservation, 22, 27, 56
Angular Velocity, 34
critical, 87
local, 43
Arnold, V. I., 15
Average Energy
kinetic, 1
potential, 1
Average stream velocity, 6
Averages:
phases, 1416, 65
time, 10, 1416
velocity, 6
Avez, A., 15
Birkhoff's Theorem, 15
Black holes, 79
Boltzmann Transport Equation, 3, 6, 21
Boltzmann, L., 14
Carnot, N. L. S. 1
Center of Mass
acceleration, 23
Chandrasekhar limit, 81, 84, 88, 90
Chandrasekhar, S., 1, 9, 20, 23, 24, 29, 47,
5759, 81, 85, 86
Charge density, 13
Claussius, R. J. E., 1
Conservation laws, 8, 25, 33
Conservation of:
angular momentum, 22, 27, 70
energy, 8
linear momentum, 31
mass, 6, 12, 23, 26, 48, 73
momentum, 74
Coriolis force, 35, 36, 44, 70
Creation rate, 6
Critical angular velocity, 87
Density
charge, 13
electron, 85
force, 12
kinetic energy, 12
matter, 6, 12, 30
momentum, 12
relativistic matter, 30
EIH approximation, 29, 33, 82
Einstein field equations, 29
Einstein, Infeld, Hoffman
approximation, (see EIH)
Electron density, 85
Electrostriction, 33
Energy
a11 forms (see specific forms)
conservation, 8
Equations of motion:
Newtonian, 9
perturbed, 49
relativistic, 30
for a zero resistivity gas,60
Equatorial velocity, 89
Ergodic hypothesis, 15
Euclidean metric, 29
EulerLagrange Equations, 7, 23, 29
with magnetic fields, 33
Farquhar, I. E., 14, 15
Faulkner, J., 86
Fermi, E., 2, 20, 49, 69
Feynmann, R. P., 81
Force,
Coriolis, 35, 36, 44, 70
friction, 11, 37
generalized, 63
Lorentz, 12, 34
Force density, 12
Fowler, W. A., 81, 82, 85, 86, 91
Friction forces, 11, 37
Generalized forces, 63
Goldstein, H., 9, 34
Gravitational potential, 10
Gravitational potential energy, 25, 72, 87, 95
variation, 51
Gribben, J. R., 85
107
Hawking, S. 97
Heat energy, 37 (see also Internal energy)
Hunter, C., 70
Hydrostatic equilibrium, 7, 82
(see also Conservation of momentum)
Instability, (see Stability)
Internal energy, 25, 32, 37, 83
variation, 54
Jacobi's stability criterion, 65, 91
Jacobi, K., 2, 64
Jeans' stability criterion, 65
Jeans, Sir J., 1
KelvinHelmholtz Contraction Time, 94
Kelvin, Lord, 93
Kinetic energy, 1, 9, 12, 25, 33, 55, 72
average, 1
pulsational, 55
relativistic, 42, 91
rotational, 54
thermal, 54
variation, 51
Kinetic energy density, 12
Kurth, R., 8
Lagrange, J. L., 2
Lagrange's identity, 2, 8, 10, 13, 31, 35, 36,
38, 54, 66, 82
specialrelativistic form, 25
Landau, L. D., 8, 25
Lebovitz, N., 20, 23
Ledoux, P., 1, 49, 70
Lewis' Theorem, 15
Leibniz law, 27, 42
Lifshitz, E. M., 8, 25
Limber, D. N., 59
Linear momentum
conservation, 31
Local angular velocity, 43
Lorentz force, 12, 34
Lorentz frame, (see Lorentz space)
Lorentz metric, (see Lorentz space)
Lorentz space, 25, 39, 41
Louisville Theorem, 6, 15
Magnetic energy, 54, 87
variation, 5759
Magnetic disruption energy, 68, 87
Magnetostriction, 33
Mass,
acceleration of center, 23
conservation, 6, 12, 23, 26, 50, 74
Matter density, 6, 12, 30
Maxwell, J. C., 1, 14
Maxwell's laws, 42, 76
Meltzer, D. W., 80
Milne, E. H., 37, 70
Misner, C., 25
Moment of inertia, 9
relativistic, 26, 82
variation, 50
about an axis, 52, 87
Momentum, (see angular and linear)
conservation, 74
Momentum density, 12
Neutron stars, 79
stability, 9091
Newcomb, S., 96
Noninertial coordinate frames, 34
Ogorodnikov, K. F., 14
Oppenheimer, R. J., 82
Ostriker, J., 71
Parker, E. N., 1, 20
Phase average, 1416
Phase space, 6, 15
Plancherel, K., 14
Poincare, H. I., 1
Poisson's equation, 7, 30
Poly tropes, 25, 84, 88, 92
KelvinHelmholtz contraction
time for, 94
PostNewtonian approximation (see EIH)
Potential, 7, 21
Gravitational, 10
Rotational, 35
Scalar, 23, 31
Vector, 31
Potential energy, 1, 10, 18, 33, 64
average, 1
Gravitational, 72, 87, 95
variation, 51
108
Potentials,
relativistic, 31
relativistic interpretation, 31, 32
scalar, 23
Pulsation,
effects of surface terms, 6062
Pulsation energy, 54
variation, 54
Pulsation frequency, 66
rotational effects, 59
Pulsation periods, 53
effects of rotation and magnetism, 59
Pulsational stability, 67
Pulsational frequency,
for determining stability, 66
Rayleigh, Lord, 1, 66
Relativistic matter density, 30
Relativistic moment of inertia, 26, 82
Relativistic correction terms to
energies, 32, 82
Relativistic equations of motions, 30
Relativistic kinetic energy, 43, 91
Relativistic form of Lagrange's identity
(see Lagrange's identity)
Relativistic potentials, 31, 32
Roche model, 87
Rosenthal, A. 14
Rotational energy, 37
variation, 5557
kinetic, 54, 56
Rotational potential, 35
Scalar potentials, 23, 31
Schwarzschild radius, 81, 84, 86, 91
Sciama, D., 97
Secular stability, 64, 70
Siniai, Y., 15
Space,
Lorentz, 25, 39, 41
phase, 6, 15
velocity, 6
Spacetime, 25, 96
Stability:
global, 66
secular, 15, 70
against magnetic fields, 6870
against rotation, 6870
of neutron stars, 9293
of white dwarfs, 8486
Stability criterion,
Jacobi's, 65, 91
Jeans’, 71
pulsational, 67
Surface terms:
magnetic, 62
pressure, 62, 93
Thermal energy, 54
Thorne, K. S., 25, 81
Time averages, 10, 1416
Time variation, 76, 77
Tooper, R. F., 81, 86, 91
Total energy, 40, 64, 65
variation, 83, 87
Variation of (see specific term)
Vector potential, 31
Velocity,
angular, 34
average stream, 6
equatorial, 89
local angular, 43
Velocity space, 6
Virial, 1, 3, 12, 13
Virial equations, 2324
Virial tensor, 21
Vis viva, 1
Viscosity, 12, 70
Volkoff, G., 82
Volume average, 86
Wheeler, J. A., 25
White dwarfs, 25, 81, 85
stability, 8386
109
To the kindness, wisdom, humanity, and memory of
D. Nelson Limber and Uco van Wijk
ii
Table of Contents
Preface to the Pachart Edition Preface to the WEB Edition Introduction 1. 2. 3. A brief historical review The nature of the theorem The scope and structure of the book References Development of the Virial Theorem v vi 1 1 3 4 5 6 6 8 11 11 14 17 18 19 20 20 22 25 27 33 38 41 45
Chapter I 1. 2. 3. 4. 5. 6.
The basic equations of structure The classical derivation of the Virial Theorem Velocity dependent forces and the Virial Theorem ContinuumField representation of the Virial Theorem The Ergodic Theorem and the Virial Theorem Summary Notes to Chapter 1 References Contemporary Aspects of the Virial Theorem
Chapter II 1. 2. 3. 4. 5. 6.
The Tensor Virial Theorem Higher Order Virial Equations Special Relativity and the Virial Theorem General Relativity and the Virial Theorem Complications: Magnetic Fields, Internal Energy, and Rotation Summary Notes to Chapter 2 References
iii
The Variational Form of the Virial Theorem 48 48 49 53 60 63 71 72 78 80 80 86 90 93 98 100 102 107 Variations. 2. 3. 4. 4. 3. 2. and their implications for The Virial Theorem Radial pulsations for selfgravitating systems: Stars The influence of magnetic and rotational energy upon a pulsating system Variational form of the surface terms The Virial Theorem and stability Summary Notes to Chapter 3 References Some Applications of the Virial Theorem Chapter IV 1. 6. Perturbations. Pulsational stability of White Dwarfs The Influence of Rotation and Magnetic Fields on the White Dwarf Gravitational Instability Stability of Neutron Stars Additional Topics and Final Thoughts Notes to Chapter 4 References Symbol Definitions and First Usage Index iv . 5.Chapter III 1.
It is not generally appreciated that there are only a few thousand astronomers in the United States and perhaps twice that number in the entire world. which for economic reasons similar to those facing the wouldbe book publisher. So it is in this case and I am forced to observe that my own perception of the subject has deepened and sharpened the considerable respect I have always had for the virial theorem. most readers assume a preface is written first and thus contains the author’s hopes and aspirations. Delores Chambers. I can only hope that the reader will proceed with the attitude that this constitutes not an end in itself. I feel happily compelled to heap praise upon the publisher.Preface to the Pachart Edition As Fred Hoyle has observed. Collins. 1977 v . A corollary aspect of this expanded perspective is an awareness of how much remains to be done. punctuated by a few examples. and manuscript preparation by Mrs. John Faulkner. A second traditional role of a preface is to provide a vehicle for acknowledging the help and assistance the author received in the preparation of his work. Thus the market for such a work compared to a similar effort in another domain of physical sciences such as Physics. In reality most prefaces are written after the fact and contain the authors' views of his accomplishments. This situation has thereby forced virtually all contemporary thought in astrophysics into the various journals. but an establishment of a point of view that is useful in comprehending some of the aspects of the universe. II The Ohio State University November 15. Lastly I would like to thank my family for trying to understand why anyone would write a book that won't make any money. Only a small fraction of these could be expected to have an interest in such an apparently specialized subject. So it is a considerable surprise and great pleasure to find a publisher willing to put up with such problems and produce works of this type for the small but important audience that has need of them. find little room for contemplative or reflective thought. George W. Thus by no means can I claim to have prepared here a complete and exhaustive discussion of the virial theorem. In addition to the customary accolades for proof reading which in this instance go to George Sonneborn and Dr. rather this effort should be viewed as a guided introduction. Chemistry or Geology is miniscule.
To the extent corrections have failed to be made or confusing text remains the fault is solely mine iv . I have elected to stay close to the original style simply as a matter of choice. I have also corrected numerous typographical errors that survived in the original monograph. I have kept those notes in this edition. I have elected to keep the original notation even though the Einstein summation convention has become common place and the vectordyadic representation is slipping from common use. The equations were more difficult to read and sections difficult to emphasize. Because some of the derivations were complicated and tedious. Since such a table has an inherent one page error. I elected to defer them to a “notes” section at the end of each chapter. However. However. I would only ask that should readers find it helpful in their research. the entries in the index could be off by a page. I have attempted to reduce that confusion by using italicized superscripts for referrals to the notes section. The original monograph was published by Pachart Press and had its origin in a time before modern word processors and so lacked many of the cosmetic niceties that can currently be generated. but I have also found it to be poorly understood by many who study the subject. I have always regarded the virial theorem as extremely powerful in understanding problems of stellar astrophysics. that they make the proper attribution should they employ its contents. the job is likely to be incomplete. So it does not seem unreasonable that I make it available to any who might learn from it. confusion arose in the main text between superscripts referring to references and entries in the notes sections. It is hoped that they are helpful. However. the index was converted from the Pachart Edition by means of a page comparison table. For similar reasons I have decided not to rewrite the text even though I suspect it could be more clearly rendered. While it is obvious that the theorem has not changed in the quartercentury that has passed since I first wrote the monograph. pressures on curricula have reduced the exposure of students to the theorem even below that of the mid 20th century. These additions are in noway meant to be exhaustive or complete. but again. Finally. I have also added some references that appeared after the manuscript was originally prepared.Preface to the Internet Edition Not only might one comfortably ask “why one would write a book on this subject?”. but enlarged the type font so that they may be more easily followed. but one might further wonder why anyone would resurrect it from the past. that should be close enough for the reader to find the appropriate reference. The reason is partly sentimental and largely not wishing to invest the time required to convert the equations. The format I chose then may seem a little archaic by today’s standards and the referencing methods rather different from contemporary journals. My reasons revolve around the original reasons for writing the monograph in the first place.
2003 vii . II April 9. Pacholczyk for permitting the use of the old Copyright to allow the work to appear on the Internet.and hardware necessary to reclaim the work from the original. Continuing thanks is due A.G. Collins. I would like to thank John Martin and Charlie Knox who helped me through the vagaries of the soft. George W.Lastly.
So. This is largely the result of its diverse development and somewhat obscure origin."4 In the 19th century. it was commonplace to assign a Latin name to any special characteristic of a system. Even though C1aussius' lecture was translated and published in Great Britain in a scant six weeks. A brief. the name has become attached to the theorem and its evolved forms. in the more contemporary language of energy. account of this historical development can be found in "The Dynamical Theory of Gases" by Sir James Jeans1 and in order to place the viria1 theorem in its proper prospective. C1aussius also turned to the Latin word virias (the plural of vis) meaning forces to obtain his ‘name’ for the term involved in the second half of his theorem. Claussius delivered a lecture before the Association for Natural and Medical Sciences of the Lower Rhine "On a Mechanical Theorem Applicable to Heat. Largely inspired by the work of Carnot on heat engines. it may be as well to explain it a little 1 . C1aussius would have stated that the average kinetic energy is equal to 1/2 the average potential energy. This scalar quantity which he called the viria1 can be represented in terms of the forces F i acting on the system as 1 2 ∑F • r i i i and can be shown to be 1/2 the average potential energy of the system. Thus. as is known to all students of celestial mechanics the vis viva integral is in reality the total kinetic energy of the system. E. for the viria1 theorem did not spring full blown in its present form but rather evolved from the studies of the kinetic theory of gases. R. This lack of recognition prompted James Clerk Maxwell four years later to observe that ''as in this country the importance of this theorem seems hardly to be appreciated. J. it is worth recounting some of that history. On June 13. A Brief Historical Review Although most students of physics will recognize the name of the viria1 theorem. Copyright 2003 Introduction 1. C1aussius began a long study of the mechanical nature of heat in 18512. This study led him through twenty years to the formulation of what we can now see to be the earliest clear presentation of the viria1 theorem. One of the lasting achievements of 19th century physics was the development of a comprehensive theory of the behavior of confined gases which resulted in what is now known as thermodynamics and statistical mechanics. few can state it correcet1y and even fewer appreciate its power. the power of the theorem was slow in being recognized. but impressive."3 In giving this lecture. 1870. Although the characteristic of the system C1aussius called the viria1 is no longer given much significance as a physical concept. C1aussius stated the theorem as "The mean vis viva of the system is equal to its viria1.
10 Chandrasekhar and Fermi extended the virial theorem in 1953 to include the presence of magnetic fields11 At this point astute students of celestial mechanics will observe that the virial theorem can be obtained directly from Lagrange's Identity by simply averaging it over time and making a few statements concerning the stability of the system. it is this derivation which is most often used to establish the virial theorem. Thus it is fair 2 . The characterization of the properties of a gas in terms of its internal and kinetic energy had not yet been developed.8 Poincare used a form of the viria1 theorem in 19119 to investigate the stability of structures in different cosmological theories."12 In this essay he developed what can be interpreted as Lagrange's identity for three bodies. L. although he makes use of time averages in deriving the theory. Of course terms such as "moment of inertia".more fully. Jacobi's formulation closely parallels the present representation of Lagrange's identity including the relating of what will later be known as the virial of Claussius to the potential. "potential” and "kinetic energy" do not appear. In addition. During the 1940's Paul Ledoux developed a variational form of the virial theorem to obtain pulsational periods for stars and investigate their stability.13 He continues on in the same chapter to develop the stability criterion for nbody systems which bears his name. It is indeed a very short step from this point to what is known as the Classical Virial Theorem. Lord Rayleigh formulated a generalization of the theorem in 19036 in which one can see the beginnings of the tensor viria1 theorem revived by Parker7 and later so extensively developed by Chandrasekhar during the 1960's. In 1772 the Royal Academy of Sciences of Paris published J. The passion for unification which pervaded 20th century physics was not extant in the time of Jacobi and Claussius. it is clear from the development that he expected these averages to be interpreted as phase or ensemble averages. some comment is in order as to who has the better claim to the theorem. We will return to this point later in some detail. The point is subtle and often overlooked today. Lagrange's "Essay on the Problem of Three Bodies. Indeed. Only if the system is ergodic (in the sense of obeying the ergodic theorem) are phase and time averages the same. Since Lagrange predates Claussius by a century. The formulation of statistical mechanics which now provides some measure of unity between the two had not been accomplished. the virial. The study of heat and classical dynamics of gravitating systems were regarded as two very distinct disciplines. It is difficult to imagine that the contemporary Claussius was unaware of this work. These differences are amplified by considering the state of physics during the last half of the 19th century. However. The very fact that Claussius required a new term. but the basic mathematical formulation is present. for the theorem makes it clear that its relationship to the internal energy of the gas was not clear. It is this last point which provides a major distinction between the virial theorem of Claussius and that obtainable from Lagrange's Identity. After the turn of the century the applications of the theorem became more varied and widespread. It does appear that this remained a special case germane to the threebody problem until the winter of 184243 when Karl Jacobi generalized Lagrange's result to nbodies."5 Maxwell's observation is still appropriate over a century later and indeed serves as the "raison d'etre" for this book. there are some notable and important differences between the virial theorem of Claussius and that which can be deduced from Jacobi's formulation of Lagrange's identity.
Unfortunately. systems with velocity dependent forces. It is indeed this reduction in complexity from a vector description to a scalar one which enables us to solve the resulting equations. There are two ways of looking at the reason for this inability to ascertain the complete physical structure of a system from energy considerations alone. but we shall see that it can also be formulated to deal with relativistic (in the sense of special relativity) systems. Let us then prepare to examine why this theorem can provide information concerning systems whose complete analysis may defy description. This reduction results in a concomitant loss of information and we cannot expect to obtain as complete a description of a physical system as would be possible from the solution of the force equations. rotation. First. The integration of a function leads to a 3 . by demonstrating its applicability to thermodynamics he made a new and fundamental contribution to physics.) affect that behavior. the virial theorem generally deals in scalar quantities and usually is applied on a global scale. yield at least three separate 'component' equations which in turn will behave as coupled scalar equations. magnetic fields. These equations can usually be obtained from the beautiful formalisms of Lagrange and Hamilton or from the Boltzmann transport equation. the energy considerations yield equations involving only energies or 'energylike' scalars while the force equations. Since the theorem represents a basic structural relationship that the system must obey.g. However. The Nature of the Theorem By now the reader may have gotten some feeling for the wide ranging applicability of the virial theorem. systems exhibiting macroscopic motions such as rotation. exhibit closed form solutions only in special cases. 2. That is. applying the Calculus of Variations to the theorem can be expected to provide information regarding its dynamical behavior and the way in which the presence of additional phenomena (e. Thus the equations we shall be primarily concerned with are related to the first integral of the defining differential force equations. vector differential equations which. systems with magnetic fields and even some systems which require general relativity for their description. The second method of looking at the problem is to note that energies are normally first integrals of forces. be nonlinear. One might sum up this argument by simply saying that there is more information contained in a vector than in a scalar. Although additional cases may be solved numerically. most of the systems I mentioned above can be described by solving the force equations representing the system. Within the framework of classical mechanics. being vector equations. insight into the behavior of systems in general is very difficult to obtain in this manner.to say that although the dynamical foundation for the virial theorem existed well before Claussius' pronouncement. etc. those equations will. Not only is it applicable to dynamical and thermodynamical systems. viscous systems. secondorder.. the number of separate scalar equations one has at his disposal is fewer in the energy approach than in the force approach. in general.
or others designed to extol the intellect of the author at the expense of the reader. Therefore. Thus. The active professional or well prepared student may skip many of these steps without losing content or continuity. in order not to burden the casual reader. this loss of detailed structure is somewhat compensated for. since the process of integration results in a loss of information. let me hope that this same spirit of incisive investigation will pervade the rest of this work. The skeptic will wish to read them all. That is. the virial theorem does not by its very nature provide a complete description of a physical system but rather extensive insight into its behavior. 4 . as I have already noted.. With regard to the organization and structure of what follows. Indeed.. A comfortable rapport with the content of these summaries may encourage the reader in the belief that he is understanding what the author intended. in an attempt to clarify the development I have included most of the algebraic steps of the development. the detailed structure of the function over a discrete range is lumped into a single quantity known as integral of the function. we shall primarily concern ourselves with the application of the virial theorem to the astrophysics of stars and starlike objects. since research into these objects is still an open and aggressively pursued subject. galaxies and clusters of galaxies which could in themselves fill many separate volumes. Initially one might wonder at such an extensive discussion of a single theorem. However. Restriction of this investigation to stars and stellar systems would admit discussions concerning the stability of clusters. Thus. To that end. At the end of each chapter. the more tedious algebra has been relegated to notes at the end of each chapter. Each chapter of the book has been subdivided into sections (as has the introduction). we cannot expect the energy equation (representing the first integral of the force equation) to yield as complete a picture of the system as would the solution of the force equations themselves.young and old.. In reality it is not possible to cover in a single text all of the diverse applications and implications of this theorem. and secondly.loss of 'information' about that function. let me emphasize that this is a book for students . I have chosen to provide a brief summary of what I feel constitutes the major thread of that chapter. and in doing so any knowledge of that detailed structure is lost. the virial theorem finds applications in the dust and gas of interstellar space as well as cosmological considerations of the universe as a whole. I shall not even be able to guarantee that this treatment is complete and comprehensive. 3. Since.”. firstly by being able to solve the resulting equations due to their greater simplicity.. All areas of physical science in which the concepts of force and energy are important are touched by the virial theorem. by being able to consider more difficult problems whose formulation in detail is at present beyond the scope of contemporary physics. Even within the more restricted study of astronomy. The Scope and Structure of the Book Any introduction to a book would be incomplete if it failed to delimit its scope. However. I have endeavored to avoid such phrases as "it can easily be shown that. which represent a particular logically cohesive unit.
124.R. Mag. Ledoux. S. R. 96. Phil. 10371047 and references therein. Clebsch G.A. 1856. pp. L. J. Vol. and Lebovitz. 410. 102119. N. pp. Reimer. J. Chandrasekhar. J. Vol. 1889. Phys. C. E. pp. S. Claussius. 4. Rev. p. N. 1870. p. Claussius. 1903. 491. E. C. Vol. Oeuvres de Lagrange ed: M. Inc. 4. 8197. R. 40. 2. S. S. p. J. Mag. England. E. 1822. 2. L. Scientific Papers. p. Paris. Phil. 1925 Dynamical Theory of Gases. 1874. 1862. 40. J. E. Maxwell. (1962) Ap. Claussius. S. Paris. J. Phil. E. and Fermi. pp. 136. 4. E. Lectures on Cosmological Theories. p. 12. J. Phil. Parker. P. Serret GauthierVillars. Vol. Jacobi. S. 1870. 134153. Poincare. 122. 121.. Cambridge University Press. 4. Ap. J. Claussius. R. Mag. 1945. N.. 201213. 24. Hermann. 3 4 5 6 7 8 9 10 11 12 13 5 . 1851. 1954. pp. E. Cambridge. Ap. G. R. p. J. pp. Chandrasekhar. Mag. 16869. Mag.116. Berlin. Claussius. p. Dover Publications. H. pp.. l18. 1873. H. 241. Y. J. 1911. Lagrange. Rayleigh. 81. S. Vol. Vol. 240.J. London. R. 102. Phil. 4. J.References 1 2 Jeans. J. Scientific Papers. 1953. 11. 4 Vol. Varlesungen uber Dynamik ed: A.
Any description of a physical system begins either implicitly or explicitly from certain general conservation principles. it is appropriate to review the origin of the fundamental structural equations of stellar astrophysics. each endowed with a spatial location and momentum which move under the influence of known forces. he may formulate a very general conservation law which simply says that the divergence of the flow of particles in that volume is equal to the number created or destroyed within that volume. If one considers an infinitesimal volume of this space. This not only provides insight into the basic conservation laws implicitly assumed in the description of physical systems. then one can construct a multidimensional space through which the particles will trace out unique paths describing their history. The Basic Equations of Structure Before turning to the derivation of the virial theorem. Such a system is considered to be a collection of articles. This is essentially a statement of determinism. + ∑ pi + ∑ xi ∂t i =1 ∂x i i =1 ∂p i 6 . Development of the Virial Theorem 1. our review will of necessity be a sketch. The mathematical formulation of this concept is usually called the Boltzmann transport equation and takes this form: 3 ∂ψ ∂ψ ∂ψ 3 =S . but by their generality and completeness graphically illustrates the complexity of the complete description that we seek to circumvent.The Virial Theorem in Stellar Astrophysics Copyright 2003 I. Since lengthy and excellent texts already exist on this subject. and in classical terms is formulated in a sixdimensional space called phasespace consisting of three spatial dimensions and three linearly independent momentum dimensions. If one regards the characteristics of spatial position and momentum as being highly independent.
1) over all velocity space yields ∂ρ . Equation (1.1.The Virial Theorem in Stellar Astrophysics or in vector notation ∂ψ + v • ∇ψ + f • ∇ p ψ = S . 1.2) is just a statement of the conservation of mass.1 A determination of ψ as a function of the coordinates and time constitutes a complete description of the system. f is the vector sum of the forces acting on the particles and S is the 'creation rate' of particles within the volume. 1. has essentially turned the Boltzmann transport equation into an equation expressing the conservation of momentum.1) but rather simplifications are made from which come the basic equations of stellar structure.1. averaging equation (1. However. If we further consider only systems exhibiting no stream motion we arrive at the familiar equation of hydrostatic equilibrium ∇P = −ρ∇Ψ .1.7 constitute a complete statement of the conservation of momentum.Under these conditions the last term on the right of equation (1. For example.1.1. 1.6 Multiplying equation (1. 7 . Its divergence then becomes the gradient of the familiar scalar known as the gas pressure P.3 ρ For systems where mass is neither created nor destroyed S = 0 . 1.4) vanishes and the pressure tensor becomes diagonal with each element equal.1.1.2 + ∇ • (uρ) = S ∂t where u is the average stream velocity of the particles and is defined by 1 u = ∫ ψvdv . The symbol P is known as the pressure tensor and has the form P = ∫ ψ ( v − u)( v − u)dv .1) by the particle velocities and averages again over all velocity space he will obtain after a great deal of algebra the EulerLagrange equations of hydrodynamic flow ∂u 1 1 + (u • ∇)u = −∇Ψ − ∇ • P − ∫ S ( v − u)dv .5 These rather formidable equations simplify considerably in the case where many collisions randomize the particle motion with respect to the mean stream velocity u .1.1.6) along with Poisson's equation for the sources of the potential ∇ 2 Ψ = −4πGρ .1. This is generally done by taking 'moments' of the equations with respect to the various coordinates.1.1. If one multiplies equation (1.1. rarely is an attempt made to solve equation (1. and equation (1. The homogeneous form of this equation is often called the Louisville Theorem and would be discussed in detail in any good book on Classical Mechanics. ∂t where ψ is the density of points in phase space. 1. 1.1) by v and averaging over v. 1. noting that the integral of ψ over all velocity space yields the matter density ρ and that no particles can exist with unbounded momentum.4 ∂t ρ ρ Here the forces f have been assumed to be derivable from a potential Ψ.1.
The Newtonian equations of motions for the system are then pi = d(m i v i ) = fi dt . ∇ • F = ρε . which when combined with the ideal gas law is dE 1. Kurth1) derive the theorem from the Lagrange identity. Chandrasekhar3 follows closely the approach of Claussius while Goldstein4 gives a very readable vector derivation firmly rooted in the original approach and it is basically this form we shall develop first. 1.8 ρ + ρ∇ • v = ρε + χ − ∇ • F . Many derivations of varying degree of completeness exist in the literature.1. we have a statement of radiative equilibrium. dt where F is the radiant flux. Consider a general system of mass points mi with position vectors ri which are subjected to applied forces (including any forces of constraint) fi. we shall repeat the version given by Claussius and express the virial theorem as a relation between the average value of the kinetic and potential energies of a system in a steady state or a quasisteady state. dr . ε the total rate of energy generation and χ is the energy generated by viscous motions.1.The Virial Theorem in Stellar Astrophysics Multiplying equation (1. Since the understanding of any theorem is related to its origins.g.1. If one has a machine wherein no mass motions exist and all energy flows by radiation.9 For static configurations exhibiting spherical symmetry these conservative laws take their most familiar form: dm(r ) Conservation of mass = 4πr 2 ρ .1 8 .1. L(r ) = 4πr 2 F . 1. dr Conservation of momentum dP(r ) Gm(r )ρ =− dr r2 dL(r ) = 4πr 2ρε .1) by v • v or v2 and averaging over all velocity space will produce an equation which represents the conservation of energy.10 Conservation of energy 2. The Classical Derivation of the Virial Theorem The virial theorem is often stated in slightly different forms having slightly different interpretations. Landau and Lifshitz2 give an eloquent derivation appropriate for the electromagnetic field which we shall consider in more detail in the next section. 1.2. we shall spend some time deriving the virial theorem from first principles. Most texts on stellar or classical dynamics (e. In general.
7 j≠ i where Fij is the force between the ith and jth particle. Now consider dG = ∑ ri • p i + ∑ p i • ri dt i .8 The subscript on the ∇operator implies that the gradient is to be taken in a coordinate system having the ith particle at the origin.2.2. 1. we may rewrite equation (1. Thus f i = ∑ Fij . Now.2. However.1).2. we have 1.The Virial Theorem in Stellar Astrophysics Now define G = ∑ p i • ri = ∑ m i i i dri • ri = dt 1 2 ∑m i i d (ri • ri ) 1 d = 2 ∑ m i ri2 . Carrying out the operation implied by equation (1.2.2 The term in the large brackets is the moment of inertia (by definition) about a point and that point is the origin of the coordinate system for the position vectors ri. The total force on the ith particle may be determined by summing all the forces acting on that particle. 1.8). 1.4 = 2T . 1.2. 2 i 1.2.9 Fij = − n a ij rij( n − 2) (ri − r j ) . Now consider the Virial of Claussius. 9 .2.2.2.3 where I is the moment of inertia about the origin of the coordinate system. if the forces obey a power law and are derivable from a potential then. dt dt i 1. since p i is really the applied force acting on the system (see equation 1. Let us assume that the forces fi obey a power law with respect to distance and are derivable from a potential. 1. Thus.2.5 but ∑r • p = ∑ m r • r = ∑ m v i i i i i i i i i where T is the total kinetic energy of the system with respect to the origin of the coordinate system. we have G= 1 2 dI dt . Fij = ∇ i m i Φ (rij ) = −∇ i a ij rijn .6 dt i The last term on the right is known as the Virial of Claussius.4) as follows: dG = 2T + ∑ f i • ri .
1. It should be noted that this formulation of the virial theorem involves time averages of indeterminate length.10) into the definition of the Virial of Claussius.2.2. then we may make the left hand side of equation (1.2. and we arrive at a statement of what is known as Lagranges’ Identity: dG 1 d 2 I = = 2T + Ω . bears the same subscript as the first subscript on the force vector. 1. That is.6) yields: dG 1.13 dt 2 dt 2 To arrive at the usual statement of the virial theorem we must average over an interval of time (T0). 10 .11 It is important here to notice that the position vector ri.12).15 T0 If the motion of the system over a time T0 is periodic.1 For the gravitational potential n = 1.11) and then into equation (1. using the definition of average value we obtain: 1 [G (T0 ) − G (0)] = 2T − nU .2.9) into equation (1. j j>i Substituting equation (1.2.2.2. 1.e.15) as small as we wish by averaging over a longer time. 1. Therefore.12 = 2T − nU .2..14 and.2.15) will vanish. the position vector is the vector from the origin of the coordinate system to the particle being action upon. integrating equation (1. then the lefthand side of equation (1.2. we have 1 T0 ∫ T0 0 dG 2 dt = T0 dt ∫ T0 0 T ( t )dt − n T0 ∫ T0 0 U ( t )dt . dt where U is the total potential energy. Substitution of equation (1. If one is to use the virial theorem to determine whether a system is in accelerative expansion or contraction.7) as follows: 1. we have ∑f i i • ri = ∑∑ Fij • ri + F ji • r j i j >i .16 2T + Ω = 0 . which is 'dotted' into the force vector.1. we can rewrite equation (1. It is in this sense that the virial theorem is sometimes referred to as a statistical theorem. acting on the jth particle due to the ith particle. Indeed.The Virial Theorem in Stellar Astrophysics Now since the force acting on the ith particle due to the jth particle may be paired off with a force exactly equal and oppositely directed.2. if the motion of the system is bounded [i.2.10 f i = ∑ Fij = ∑ Fij + Fji . if a system is in a steady state the moment of inertia ( I ) is constant and for systems governed by gravity 1.2. G(t) < ∞]. then he must be very careful about how he obtains the average value of the kinetic and potential energies.2.2. Thus.
4 where U is the average value of the potential energy for the "nonfrictional" forces.3.3.3. Carrying out the integration on the left hand side we have 1 [G (T0 ) − G (0)] + 1 ∑ α i ri2 (T0 ) − ri2 (0) = 2T − nU .2. if the motion is periodic.2 1. the result of the virial theorem is unchanged.3.3.1 becomes 11 . The equations of motion may be written as: pi = fi = wi + z i .6).2.12). Substituting into equation (1.5) will vanish in a time T0 equal to the period of the system.16) that should be mentioned. Let the mass mi be obtained by multiplying the density ρ(r) by an infinitesimal volume ∆V so that 1. dt i i Remembering that the velocity dependent forces may be rewritten as dr wi = αi vi = αi i . Thus dri 1 T0 dG 1 T0 ∫0 dt dt − T0 ∫0 ∑ α i dt • ri dt = 2T − nU . ContinuumField Representation of the Virial Theorem Although nearly all derivations of the virial theorem consider collections of masspoints acting under forces derivable from a potential.2.5 T0 2T0 i Thus.3.The Virial Theorem in Stellar Astrophysics 3. both terms on the left hand side of equation (1. we have dG − ∑ w i • ri = 2T + ∑ z i • ri . If the forces acting on the system include velocity dependent forces. dt We may again average over time as in equation (1.1 1.2. In order to demonstrate this. T0 i 1.1) to its analogous representation in the continuum. consider the same system of mass points mi subjected to forces fi which may be divided into velocity dependent ( w i ) and velocity independent forces (zi). 1. This is particularly appropriate when one considers applications to stellar structure where a continuum representation of the material is always used.3 1. This apparently academic aside has the significant result that we need not worry about any Lorentz forces or viscosity forces which may be present in our subsequent discussion in which we shall invoke the virial theorem. In the interests of preserving some rigor let us pass from equation (1. [ ] 4. it is useful to look at this formalism as it applies to a continuum density field of matter. Indeed both terms can be made as small as required providing the "frictional" forces w i do not cause the system to cease to be in motion over the time for which the averaging is done.2. Velocity Dependent Forces and the Virial Theorem There is an additional feature of the virial theorem as stated in equation (1.
8 dt V dt dt V d ∫ dt (ρr )dV − ∫ r 2 1 2 2 The first term under the integral is just kinetic energy density and hence its volume integral is .4) with respect to time and obtain dG dr dp = ∫ p • + r • dV = ∫ (ρv 2 + r • f ) dV . so that f ∆V = fi .4.2. 1. the second integral in equation (1. 2 dt V With these same constraints on V we may differentiate equation (1. dt V V Once again. we can pass to this continuum representation of equation (1.4.just the total kinetic energy of the configuration and 1 2 d2I = 2T + ∫ f • r dV . 1.4. we may write equation (1.4.2. one uses conservation of mass requiring that the mass within any sub volume V' is constant with time so that dm(V ' ) = 0 with that subvolume V' defined such that dt d dρ ∫ ρdV = ∫ dV = 0 . if we define a "force density". 1.1): d f (r ) = ρ(r ) [v(r )] = p(r ) .5 dV .4.4. dt dt dt V V V V so that G= 1 2 1.4. Thus. 1.5) after integration by parts is zero.1) are of equal magnitude and opposite sign. dt 2 V 1.7 G= ∫ (ρr )dV = 12 dt .4.4 dρ 1.4.6 dt V ' V ' dt Thus.4.4.1 dt dt dt dt Conservation of mass requires that dm i d dρ d (∆V) = (ρ∆V ) = ∆V +ρ =0 .5) as 1 d dI 2 1. If we take the original volume V to be large enough so as to always include all the mass of the object.4.3 dt where p(r) by analogy to 1.9 12 .The Virial Theorem in Stellar Astrophysics d (ρv∆V ) = v∆V dρ + ρ∆V dv + ρv d(∆V) . 1.2 dt dt dt dt Multiplying this expression by v we see that the first and last terms on the right hand side of equation (1. f. fi = We can now define G in terms of the continuum variables so that dr d dr 2 1 1 G = ∫ p • rdV = ∫ ρ • r dV = 2 ∫ ρ (r • r )dV = 2 ∫ ρ dV .4.1 is just the local momentum density.
2. The contributions to the virial. we find that the virial has the same form as equation (1.4.13).4. After some algebra. are valid for the continuum density distributions as we might have guessed. if the force depended on some other property of the matter (e. In the previous derivation we went to some length [i. we explicitly paired the forces and arranged the sum so pairs of particles were only counted once.2. the virial theorem and indeed the remainder of the earlier arguments. Throughout this discussion it was tacitly assumed that the forces involved represented "gravitational" forces insofar as the force was ρ∇Φ. V V V V' Substitution of this form into equation (1. 2 1 d I = 2T + Ω 1. equation (1.. In addition the double integral represents the potential energy of ρ(r) with respect to ρ(r ') . Clearly..2. 13 . however.11 ∫ f • rdV = −nU .4.e.9) which is basically the virial of Claussius.10 1 n = n 2 ∫ ∫ ρ(r )ρ(r ' )( r − r ' ) dV' dV V V' Since V = V'. the integrals are fully symmetric with respect to interchanging primed with nonprimed variables.12 2 dt 2 Thus Lagrange's identity. 1.12). Thus.The Virial Theorem in Stellar Astrophysics Considerable care must be taken in evaluating the second term in equation (1.4. 1. each force at a field point f (r) will have an equal and opposite counterpart at the source points r ' . specifically. direct substitution of the potential gradient into the definition of the Virial of Claussius yields 1. Thus.4.10)] to avoid "double counting" the forces by noting that the force between any two particles A and B can be viewed as a force at A due to B. Thus. are not equal as they involve a 'dot' product with the position vector. the charge density.2 n −2 ∫ f • rdV = − n2 ∫ ∫ ρ(r)ρ(r' )(r − r' ) • (r − r' )( r − r' ) dV' dV . Similar problems confront us within continuum derivation.g. or a force at B resulting from A. it is just twice the total potential energy.9) and taking n = 1 yields the same expression for Lagrange's identity as was obtained in equation (1. namely. ε(r) the evaluation of ∫ f • rdV would go V as before with the result that the virial would again be –nU where U is the total potential energy of the configuration. and ρ(r ') with respect to ρ(r).
sooner or later.e. the distinction between an ensemble average and an average of macroscopic system parameters over time was not clear at the time of the formulation of the virial theorem. pass through every phase which is consistent with the equation of energy". No matter how rapid the measurements of something like the pressure or temperature of the gas. The Ergodic Theorem is one of those fundamental physical concepts like the Principle of Causality which are so "obvious" as to appear axiomatic. Theoretical considerations predict <Q>s whereas experiment provides something which might be construed to approximately <Q>t. it requires a time which is long compared to characteristic times for the system. we have developed Lagrange's identity in a variety of ways. 14 . would pass through every point in phase space).1 T → ∞ T ∫t 0 where <Q>s is some sort of instantaneous statistical average of Q over the entire system. is given by Lim 1 t 0 + T < Q >t = Q( t )dt =< Q > s . The importance of this concept for statistical mechanics is clear.. not too long after. Since this theorem is central to obtain what is commonly called the virial theorem.5. in astrophysics few if any investigators live long enough to perform the timeaverages for which the theorem calls. Thus they are rarely discussed in the physics literature. but have not rigorously taken that finial step to produce the virial theorem. However. The founders of statistical mechanics. This last step involves averaging over time and it is in this form that the theorem finds its widest application. Thus. such as Boltzmann. Namely. one more step is needed. In order to replace the time averages with something observable.7 Essentially this constitutes what is most commonly meant by the ergodic theorem. This language alone is enough to hide it forever from the eye of the average physical scientist. with the exception of a brief discussion in Section 2.The Virial Theorem in Stellar Astrophysics 5. it is necessary to invoke the ergodic theorem. say Q. it is appropriate that we spend a little time on its meaning. It is this step which occasionally leads to difficulty and erroneous results. Ludwig Boltzmann6 formulated an hypothesis which suggested the criterion under which ensemble and phase averages would be the same. However. realized that such a statement as equation (1. Maxwell and Gibbs. The Ergodic Theorem and the Virial Theorem Thus far. However. to say that the ergodic theorem is obvious is to belittle an entire area of mathematics known as ergodic theory which uses the mathematical language of measure theory. Maxwell later stated it this way: "The only assumption which is necessary for a direct proof is that the system if left to itself in its actual state of motion will. 1.1) was necessary to enable the comparison of theory with experiment and thus a great deal of effort was expended to show or at least define the conditions under which dynamical systems were ergodic (i.5. if a dynamic system passes through every point in phase space then the time average of any macroscopic system parameter. As noted in the introduction.
Physicists.The Virial Theorem in Stellar Astrophysics Indeed. Thus 15 . Already one can sense confusion of terminology beginning to mount. However. there have been many attempts to prove the quasiergodic hypothesis. and they became more concerned with showing the existence of the averages than with their equality with phase averages. a "gas" made up of perfectly elastic spheres confined by a container with perfectly reflecting walls is ergodic in the sense that phase and time averages are equal. the ergodic hypothesis became modified as the quasiergodic hypothesis. and also what is necessary. as Farquhar observed “such a pragmatic view reduces statistical mechanics to an ad hoc technique unrelated to the rest of physical theory. as recounted by Arnold and Avez16 proves that the BoltzmannGibbs conjecture is correct. At this point in time the mathematical interest in ergodic theory began to rise rapidly and over the next several years attracted some of the most. It is the essence of the Louisville theorem of classical mechanics. The identity of phase and time averages became crucial to the comparison of theory with observation. Farquhar11 points out that several noted physicists stated without justification that all physical systems were quasiergodic. Ogorodnikov10 uses the term quasiergodic to apply to systems covered by the Lewis theorem which we shall mention later. the ergodic theorem is false as was shown independently in 1913 by Rosentha18 and Plancherel 9 more modern version of this can be seen easily by noting that no system trajectory in phase space may cross itself. The tendency in recent years has been to bypass phase space filling properties of a dynamical system and go directly to the identification of the equality of phase and time averages. That is. This is a position not without precedent and a certain pragmatic justification of expediency. famous mathematical minds of the 20th century. This is effectively a statement of system boundary conditions uniquely determining the system's past and future. the virial theorem is obtained by taking the time average of Lagrange's identity. Some essentially adopted the attitude that since thermodynamics “works”. a multidimensional space filling curve would have a measure equal to the space whereas a Jordan curve being onedimensional would have measure zero. In the language of measure theory. Thus. such a curve may have no multiple points. Mathematicians largely took over the field developing the formidable literature currently known as ergodic theory. it may come arbitrarily close to any given point in a finite time. This modification essentially states that although a single phase trajectory cannot pass through every point in phase space. as stated.14 These theorems show the existence of time averages and their equivalence to phase averages under quite general conditions. phase and times averages must be equal.” 12 Over the last half century. At this point the reader is probably wondering what all this has to do with the virial theorem. Perhaps the most notable of which are Birkhoff's theorem13 and the generalization of a corollary known as Lewis' theorem. Such a curve is topologically known as a Jordan curve and it is a well known topological theorem that a Jordan curve cannot pass through all points of a multidimensional space. Specifically. impatient with mathematicians for being unable to prove what appears 'reasonable'. Thus. began to require the identity of phase and time averages as being axiomatic. The most recent attempt due to Siniai15. The stakes were high and were getting higher with the development of statistical mechanics and the emergence of quantum mechanisms as powerful physical disciplines.
integrals which appear to restrict the system motion in phase space over several relaxation times. T → ∞ 2 ∫t 0 dt 2 1. the application of the virial theorem to a system with only a few members and hence a few degrees of freedom is invalid unless care is taken to interpret the observed ensemble averages in light of phase averages altered by the isolating integrals of the motion. Thus. in this case. At least we may do it with the same confidence of the thermodynamicist. these concerns do not apply and we may confidently interchange time and phase averages as they appear in the virial theorem. 16 . if isolating integrals of the motion exist for the system. for stars and starlike objects exhibiting 1050 or more particles undergoing rapid collisions and having short relaxation times.The Virial Theorem in Stellar Astrophysics Lim 1 t 0 + T d 2 I dt =< 2T > t − < U > t . the integration in equation(1. However.2) must exceed that time if any statistical validity is to be maintained in the analysis of the system. This interpretation will only be correct if the system is ergodic in the sense of satisfying the 'quasiergodic hypothesis'.5. The first problem arises with the fact that the time average is over infinite time and thus operationally difficult to carryout l. However. the nature of irreversibility must be connected in some sense to that of ergodicity and thus to the very nature of time itself. Farquhar17 points out that the time interval must at least be long compared to the relaxation time for the system and in the event that the system crossing time is longer than the relaxation time.3. one should be most circumspect about applying the virial theorem to large systems like the galaxy which appear to exhibit quasiisolating integrals of the motion.5. the astronomer is in the envious position of being in the reverse position from the thermodynamicists. That is. Thus. as these integrals remove large regions of phase space from the allowable space of the system trajectory.2 and for systems which are stable the left hand side is zero. Indeed. Lewis' theorem allows for ergodicity in a subspace but then the phase averages must be calculated differently and this correspondence to the observed ensemble average is not clear. in stellar dynamics and the analysis of stellar systems they generally are not. anyone truly interested in the foundations of physics cannot dismiss ergodic theory as mere mathematical 'nitpicking'. Pragmatically if the system exhibits a large number of degrees of freedom then persuasive arguments can be made that the equating of time and phase averages is justified. For all intents and purposes he can perform an 'instantaneous' ensemble average which he wishes to equate to a 'theoretically determined' time average. For those who feel that the ergodic theorem is still "much ado about nothing". it is worth observing that by attempting to provide a rational development between dynamics and thermodynamics. Furthermore. ergodic theory must address itself to the problems of irreversible processes. However. then it is not justified. It is clear that for stars and starlike objects these conditions are met. Since classical dynamics is fully reversible and thermodynamics includes processes which are not.
The next two sections were concerned with a highly classical derivation of the virial theorem with section 2 being basically the derivation as it might have been presented a century ago. I have tried to lay the groundwork for the classical virial theorem by first demonstrating its utility. An underlying thread of continuity can be seen in all that follows comes from the Boltzmann transport equation. examining an important premise of its application. the reader should be in an agreeable frame of mind to consider alternative approaches to solving the vector differential equations of structure in order to glean insight into the behavior of the system. Summary In this chapter. The only 'tricky' part of these derivations involves the 'pairing' of forces. The reader should make every effort to understand or conceptualize how this occurs in order to understand the meaning of the virial itself. 17 . If they are not. This sketch was far from exhaustive and intended primarily to show the informational complexity of this form of derivation. The assumption that the forces are derivable from a potential which is described by a power law of the distance alone. It is a theme that will return again and again throughout this book. Section 3 merely updated this presentation so that the formalism may be used within the context of more contemporary field theory. In the last section. dates back at least to Jacobi and is often described as a homogeneous function of the distance. we sketched how the Boltzmann transport equation yields a set of conservation laws which in turn supply the basic structure equations for stars. Its importance for the application of the virial theorem cannot be too strongly emphasized. one cannot replace averages over time by averages over phase or the ensemble of particles without further justification. In section 1.The Virial Theorem in Stellar Astrophysics 6. I attempted to provide some insight into the meaning of a very important theorem generally known as the ergodic theorem. many systems in stellar dynamics cannot. Being suitably impressed with this complexity. then deriving it in several ways and lastly. Although almost all systems of interest in stellar astrophysics can truly be regarded as ergodic.
1 Thus ∑f i i • ri = − n ∑∑ a ij rijn = − n ∑∑ Φ (rij ) i j> i i j> i . Thus. we get 2 ∫ f • rdV = − ∫ ρ(r ) ∫ ρ(r )r •∇ r ( r − r' ) dV' dV = − ∫ ρ(r ' ) ∫ ρ(r )r∇ r ' ( r − r' ) dV' dV .11) as follows: ∑ f i • ri = −∑∑ na ij rij( n −2) (ri − rj ) • ri + (rj − ri ) • rj [ ] i i j> i = − n ∑∑ a ij rij( n − 2) (ri − r j ) • (ri − r j ) = − n ∑∑ a ij rij( n − 2) rij2 i j> i i j> i [ ] .2 is zero if the system is periodic and the integral is taken over the period.2.1 Since aij = aji for all known physical forces. Since the contribution to the force density from any pair of sources and field points will lie along the line joining the two points. N 1.3).2 Since the second summation is only over j > i.2. doing this and adding equation (N 1.1.5 Then. we may substitute equation (1. V' N 1.2.2.2 As in Section 2.2.3) by r' and integrating over V'. we can write the potential as Φ (r ) = ∫ ρ(r )( r − r ' ) dV' ∀ n < 0 .4 Now V ∫ f (r) • rdV = ∫ f (r' ) • r' dV' .2. N 1. ∇ r ( r − r ' ) = −∇ r ' ( r − r ' ) = n ( r − r ' ) n n n −2 (r − r') .3 where ∇r and ∇ r ' denote the gradient operator evaluated at the field point r and the source point r’ respectively.1.2.2.2) to equation (N 1. so multiplying equation (N 1.2. 18 .3 It should be noted that the left hand side of 1.The Virial Theorem in Stellar Astrophysics Notes to Chapter 1 1.5 n n V V V' V' V 1. N 1.2. V N 1.1 n and the force density is then f (r ) = −ρ(r )∇ r Φ (r ) = −ρ(r ) ∫ ρ(r ' )∇ r ( r − r ' ) dV' .2 while the force density at a source point due to all the field points is n f (r ' ) = −ρ(r ' )∇ r ' Φ (r ' ) = −ρ(r ' ) ∫ ρ(r )∇ r ' ( r − r ' ) dV .2) by r and integrating over V' V produces the same result as multiplying equation (N 1. let us assume that the force density is derivable from a potential which is a homogeneous function of the distance between the source and field point.9) in equation (1.2. 1. and the double sum is just the total potential energy of the system. there is no "doublecounting" involved. n V' N 1.5. N 1.
M. Landau. New York. (1962). . p. E. (1913).. (1964).. Trans. Hamermesh. Co. (1964). pp. Co.: U. D.. Rational Mech. Ya. 679.. Ogorodnikov. S. M. (1879)..S. (1957). 78. of Sci. Arnold. Acad. Wiss. pp.. L. pp. Wien 63. London. A. Mass. J. R. Classical Mechanics. p. I. 77. Mechanics. Ergodic Theory in Statistical Mechanics.. Y. Proc. 69. Sinai. I. Plancherel. Rosenthal.. 17. S. Dynamics of Stellar Systems. AddisonWesley Publ. E.. Sitzler Akad. 3. pp. p. F. New York. Vestnik Mossovskova Gosudrastvennova Universitata Series Math. (1965). 2223. Inc. K. B. G. Ann. I. pp. Bell. Phil. H. J. Trans. 4951.796. p.. Pergamon Press. Sydney. p. Reading. Inc. Mass. Sykes & J. W.. AddisonWesley Pub. 397. der Physik 42. & Lifshitz. C. 656. AddisonWesley Pub.. L. Ltd. Pergamon Press. M. A. der Physik 42. 12. and Avez.ibid. A. & Lifshitz. p. 153. Mass. Anal 5. Cam. 6971. USA. N. 9597. . Dover Pub. D. (1960). 355. John Wiley & Sons. Goldstein. (1968). Birkhoff. 2332. (1966). Ann.1061.The Virial Theorem in Stellar Astrophysics References 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 Kurth. Introduction to the Mechanics of Stellar Systems. Trans. of Nat'l. M. (1959). Chandrasekhar. Co. p. Reading. Arch. Maxwell. p. (1931). L.547. Interscience Pub. M. p. Benjamin. Lewis. D. Farquhar. An Introduction to the Study of Stellar Structure. E. p. Reading. (1962). London. Soc. Paris. Boltzmann. Landau.. 5. The Classical Theory of Fields. E. Amsterdam. (1811). V. Ergodic Problem of Classical Mechanics. The McMillan Co. New York. 19 . (1913). Loc. cit. Rudolf (1957). Farquhar.
Although the germ of this idea can be found developing as early as 1903 in the work of Lord Rayleighl. the most lucid derivation is probably that presented by Chandrasekhar8 in 1961 and it is a simplified version of that derivation which I shall give here. 7 during the 1960's. The final step involves a further average over time.e. it is not surprising that since system symmetries inspire this approach that the information to be kept relates to motions along orthogonal coordinate axes. 3. and Chandrasekhar and Fermi4 found the concept particularly helpful in dealing with the presence of magnetic fields.. but not all. it is worth pointing out that the derivation in Chapter I. it is rather satisfying to one who believes that "all that is good and beautiful in physics" can be obtained from the 20 . If one recalls the fullblown vector equations of motion in Chapter I. section 1. section 1 to scalars. Chandrasekhar33 has given a fairly comprehensive recounting of his efforts on this subject after the original version of this monograph was prepared.The Virial Theorem in Stellar Astrophysics Copyright 2003 II Contemporary Aspects of the Virial Theorem 1. As previously mentioned the motivation for this approach is to regain some of the information lost in forming the scalar virial theorem by keeping track of certain aspects of the system associated with its spatial symmetries. The concept was further expanded by Lebovitz5 and in a series of papers by Chandrasekhar and Lebovitz6. Indeed. essentially originates from the equations of motion of the system being considered. this amounts to keeping some of the component information of those equations. The Tensor Virial Theorem The tensor representation of the virial theorem is an attempt to restore some of the information lost in reducing the full vector equations of motion described in Chapter I.) Since moment analysis of this type also yields some of the most fundamental conservation laws of physics (i. That is to say that the virial theorem results from taking spatial moments of the equations of motion and investigating their temporal behavior. for the investigation of the stability of various gaseous configurations. section 2. However. At this point. mass and energy). (Recall that the equations of motion themselves are moments of the Boltzmann transport equation. In particular. momentum. it wasn't until the 1950's that Parker2. it is not surprising that the virial theorem should have the same power and generality as these laws. The derivations take the form of multiplying those equations of motions by position vectors and averaging over the spatial volume.
2. 2.1. Under these conditions. Here we propose to take the outer product of equation (1.1.1 n −2 ρr∇ΦdV = n 1 ∫ ∫ ρ(r )ρ(r ' )(r − r ')(r − r ')( r − r ' ) dV' dV . our starting point is du 2. However. n −2 2.The Virial Theorem in Stellar Astrophysics Boltzmann equation that the virial theorem essentially arises from taking higher order moments of that equation. ∫ V Although some authors choose a slightly different convention. section 2 and neglect viscous forces and macroscopic forces such as net rotation and magnetic fields as we shall consider them later.1.1.1. 2 ∫ V V V' If we define I = ∫ (ρrr )dV V 2.3 V' Then following exactly the same manipulation as in Chapter I.2.1.1. only taking into account outer products instead of inner products with the position vectors. 21 . By eliminating additional external forces such as magnetic fields and rotation we have lost much of the power of the tensor approach.4) or (1.2. equation (1. we get the virial tensor.1) with the position vector r producing a tensor equation which can be regarded as a set of equations relating the various components of the resulting tensors. With that in mind let us consider a collisionless pressurefree system analogous with that considered in Chapter I.1 + ρ(u • ∇)u = −ρ∇Φ = ρ ∂t dt which is simply the vector representation of either equations (1. as before let the potential be n Φ (r ) = ∫ ρ(r ' )( r − r ' ) dV' ∀n < 0 . some insight into this power can be seen by considering in component form one term in the expansion of the virial tensor2.6 dt 2 which is essentially the tensor representation of Lagrange's identity2. the term on the right hand side of equation (2.2 ∫ ρr dt dV = V ρr∇ΦdV . 2.1.2 becomes) d 2I = 2T + nU .2 where I is sometimes called the moment of inertia tensor.1.4) becomes ∂u du ρ . Cursory dimensional arguments should persuade one that this procedure should produce relationships between the various moments of inertia of the system and energylike tensors.1.1. T the kinetic energy tensor and U the potential energy tensor.2) can properly be called the virial tensor.4) with the position vector r and integrated over the volume to produce a scalar equation.1.1).1. section 2. Thus. we essentially took the inner product of equation (1.4 T= U= 1 2 1 2 V ∫ (ρuu)dV V V' .5 ∫ ∫ ρ(r)ρ(r' )(r − r')(r − r')( r − r' ) 1 2 dV' dV equation (2. Now. In Chapter I.
Chandrasekhar and Lebovitz10. dt V which simply says the angular momentum about xk is conserved. moments in momentumspace of the Boltzmann transport equation yield expressions for the conservation of mass. it is arguable as to whether they should be called virial expressions at all. 13 developed these expressions as far as the fourthorder moments of the equations of motion. by using the same conservation of mass arguments discussed earlier. However. As already noted.2. 11 and later Chandrasekhar12. since it is clear that this investigation was inspired by studies of the classical virial theorem. Since for no moment expressions other than those of the first moment do any terms ever appear that can be identified with the Virial of Claussius. the higher the order of the moment expressions. momentum and energy.The Virial Theorem in Stellar Astrophysics V ∫ρ d dr d dx j dV . Such is the origin of such diverse expressions as the Eddington approximation in radiative transfer.7 Since this tensor is clearly symmetric we find.1 22 . I will briefly review their development. Recall the EulerLagrange equation of hydrodynamic flow developed in Chapter I. in the spirit of generality.1.4) ∂u 1 1 + (u • ∇)u = −∇Ψ − ∇ • P − ∫ S( v − u)dv . There certainly is considerable precedent for this in mathematical physics. Chandrasekhar investigated the properties of the first several moment equations.1. ∂t ρ ρ 2. 2 Higher Order Virial Equations In the last section it became clear that both the scalar and tensor forms of the virial theorem are obtained by taking spatial moments of the equations of motion. the diffusion approximation in gas dynamics and many others. r dV = ∫ ρ x i dt dt dt dt V 2. Approximate solutions to the resulting equations can be found if suitable assumptions such as the existence of an equation of state are made to "close" the moment equations.8 ∫ ρ x i dt − x j dt dV = 0 . Usually. In a series of papers. Thus the tensor virial theorem leads us to a fundamental conservation law which would not have been apparent from the scalar form derived earlier. Chandrasekhar9 was apparently the first to note this and to inquire into the utility of taking higher moments of the equations of motion. the diffusion approximation in radiative transfer.1. Nevertheless. Spatial moments of the transport equation of a photon gas can be used to obtain the equation of radiative transfer. the less transparent their physical content. equation (1. that dx i d dx j 2.
7 Since the outer product in general does not commute.2. Setting n = 2 we V arrive at the second order "virial equations" that are the tensor version of Lagrange's identity that we discussed in the previous section. the second and third integrals are zero by the divergence theorem and equation (2.6) that d dQ ∫ ρQdV = ∫ ρ 2.2 ∫ r ρ dt dV + V r ρ∇Φ + V r ∇PdV = 0 .6 ( ) ( ) 2. If we assume that particle collisions are isotropic. the integrals of the second and fourth term become strings of tensors of the form 23 .5 2 ∫ dt 2 V essentially telling us that the center of mass ∫ ρrdV is not being accelerated. 2. we can represent the nth order "virial equations" as ( n −1) du ( n −1) ( n −1) ∫ ρr dt dV + V ρr ∇ΦdV + V r ∇PdV = 0 .3 dV . ∫ ∫ V which after use of continuity and the divergence theorem becomes d d ( n −1) ( n −1) ( n −1) ( n −1) ∫ ρr udV −V ρ dt r udV + V ρr ∇ΦdV + V P1 • ∇ r dV = 0 .2. So d 2. about conservation of mass. ∫ ∫ dt V Noting that ∇Q = ∇ • (1 Q / 3) .2. As in equation (2. it follows from equation (1.2.2.4) becomes 2 1 d ρrdV = 0 .4.1).2. dt V dt V where Q is any pointdefined property of the medium.2. then the source term of equation (2. ∫ ∫ ∫ dt V 2.2).2. correspond to the integrals of the equation of motion themselves over volume.The Virial Theorem in Stellar Astrophysics Quite simply the nth order "virial equations" of Chandrasekhar are generated by taking (nl) outer tensor products of the radius vector r and equation (2. ∫ ∫ V Recalling our arguments in Chapter I.4 ∫ ρudV + V ρ∇ΦdV + V ∇PdV = 0 . The symbolic representation for the nth order "virial equation" can then be written as: ( n −1) du ( n −1) ( n −1) 2. Thus.1) vanishes and ∇ • P = ∇P . the first order "virial equations".2. This leads to a set of tensor equations containing tensors of rank n.2. section 4. The result is then integrated over all physical space.
dt ( ) ( ) ( ) P∇ r = P 1r + r 1r + ⋅⋅⋅ + r 1 However.7) vanishes and the result is a system of identities between the various tensor energy moments.7) represents one or more tensors of rank n.2. in fact. We may expect them to be of importance in the same general way as the virial theorem itself.2. equation (2. however. Like most series expansions.7) must thus contain all of the information concerning the structure of the system. . in stationary systems the left hand side of equation (2. dt V n! dt 2 V ( ( n −1) ) ( ( n −2) ( n −3) ( n −2) ) 2.2. each term in equation (2. Chandrasekhar8 shows how these tensors can be built up from the generalized Newtonian tensor potential or alternatively from a series of scalar potentials which obey the equations ∇ 2 Ψ = −4πGρ ∇ 4 ℑ = −8πGρ ∇ 6 ℜ = −32πGρ . 2.2. The third integral is.8 .9 Thus. These equations thus represent an alternate form to the solution of the equation of motions. 24 . the first of which is the second time derivative of a generalized moment of inertia tensor and the last three are all 'energylike' tensors.The Virial Theorem in Stellar Astrophysics ρ and d ( n −1) r u = u r ( n − 2) u + r u r ( n −3) u + ⋅ ⋅ ⋅ + r ( n − 2 ) uu . That is. They are. 2. the first term can be written as d 1 d2 (n ) ρr ( n −1) udV = ∫ ∫ ρr dV .2. From equation (2. spatial tensor moments of the equations of motion. For n = 2 we know it is just the total potential energy.2.2. . Keeping in mind that any continuum function can be represented in terms of a moment expansion. we have formulated a representation of what Chandrasekhar calls the "higher order virial equations". it is clear that the second integral will generate tensors which are spatial moments of the kinetic energy distributions while the last term will produce moment tensors of the pressure distribution.8).10 Thus. it is devoutly to be wished that they will converge rapidly and the "higher order" tensors can usually be neglected. the most difficult to rigorously represent. .
Special Relativity and the Virial Theorem So far we have considered only the virial theorem that one obtains from the Newtonian equations of motion. the virial theorem says 2T = Ω. This is equivalent to saying that there exists a volume in spacetime sufficiently large so that outside that volume the stressenergy tensor is zero. Generally most expositions of field mechanics start with the statement that • ℑ = 0. One might say that it requires a potential energy equal to 2T to confine the motions of particles having a total kinetic energy T. This equivalence is made obvious by applying Gauss' divergence theorem so that 2. Although the conceptual development for this derivation is inspired by Landau and Lifshitz the subscript notation will be largely that employed by Misner.1 where ℑ is the Maxwell stressenergy tensor and is known as the fourgradient operator. but the effect is quadratic. we shall turn to the discussion of relativistic mechanics of Landau and Lifshitz15 mentioned in Chapter I. T approaches Ω and this "must be the statement of the virial theorem for material particles moving with very nearly the velocity of light. R S In short. This increase will also affect the gravitational potential energy.The Virial Theorem in Stellar Astrophysics 3.3.2 ∫ • ℑdR = ∫ ℑ • ds = 0 . equation (2. As most discussions in field mechanics generally start from a somewhat different prospective. One may interpret this as resulting from an unbounded increase of the particle's mass.3. Thus we might expect in a relativistic system that a potential energy less than 2T would be required to maintain equilibrium. the concept of taking moments at this point is most easily understood within the context of a coordinate representation so for the moment we will keep that approach. This appears to be the conclusion arrived at by Chandrasekhar when. As particles approach the velocity of light the kinetic energy increases without bound. Thus. Landau and Lifshitz give components of the 4velocity of a particle as 25 . let us examine the correspondence with the starting point of the equations of motion adopted in the earlier sections. Thorne. wherein the dynamic pressure balancing gravity is supplied by particles whose energies are very much larger than their rest energy. We have already seen that the fundamental conservation laws of physics are derivable from the Boltzmann transport equation as are the equations of motion. it is appropriate that we investigate the extent to which we shall have to modify the virial theorem to include the effects of special relativity. Indeed. In a Lorentz coordinate system. Since there are systems such as white dwarfs. and Wheeler16. Tempting as it is to use the coordinate free geometry of these authors. For systems in equilibrium. while investigating the internal energy of white dwarfs he concludes that as the system becomes more relativistically degenerate. 2. In order to obtain the somewhat more general result of a relativistic form of Lagrange's identity."14 This is indeed the case and is the asymptotic limit represented by a photon gas or polytrope of index n = 3 (see Collins32). both starting points are equivalent as they have their origin in a common concept.1) is a conservation law.3. the operation of taking moments is quite similar in both cases.
j=1 V V where.3 3 d ρ d 2. 2.The Virial Theorem in Stellar Astrophysics where ds dt = c(1 − v 2 / c 2 )1 / 2 dx α .3.6 +∑ +∑ ∂t ic ∂t k =1 ∂x k k =1 ∂x k Substituting for the components of ℑ and repeating the algebra of earlier derivations we have2.3 The components of the energymomentum tensor are ℑ αβ = ρ cu α u β ds . α = 0⋅⋅⋅3 . Note this is a somewhat unconventional definition of γ.3. using the conservation of mass arguments in Chapter I.4 which are clearly symmetric in α and β. we can write 2 1 d (I r ) = Ω + T + ∫ γτdV .3. 2.3.3.2.4) is ∑ℑ α =0 αα = −ρc 2 γ . this becomes 2 3 1 d 2..3. the trace of equation (2.3.9) is an expression of the Lagrange's identity including the effects of special relativity. it is fair to say that equation (2. 2. j=1 V V where τ is the kinetic energy density of relativistic particles17.9 2 dt 2 V In the low velocity limit γ → 1 so that Ir → I and we recover the ordinary Lagrange's identity.10 c∫ x i ∂t V V 26 . 2.8 ∑ 2 dt 2 ∫ (ρx j x j / γ )dV = Ω + T + ∫ γτdV .3).5 In terms of the threedimensional components the conservation law expressed by equation (2.6) and integrating over the volume. the total energy E = 0).3. In the relativistic limit as γ → 0 we recover for stable systems the Chandrasekhar result that T + Ω = 0 (i.e.3. Since 3 ∑u α =0 3 = −1 . Again. ds = γc .3. Thus. we get ∂ (ρu j ) dV − ∫ ℑij dV = 0 .3.1) can be written as 3 ∂ℑ ∂ (ρu j ) 3 ∂ℑ jk 1 ∂ℑ j0 jk =c =0. It is worth noting by analogy with section 1 that the tensor relativistic theorem can be derived by taking the outer or 'tensor' product of the spacelike position vector with equation (2. dt ℑ0 j = iρc 2 u j .3. if we define the volume integral on the left to be the relativistic moment of inertia. and specifically ( ) 2 α 2.3. Following the same steps that lead to equation (N. uα = 2.7 ∑ dt ∫ 2 dt (x j x j / γ )dV = Ω + T + ∫ γτdV .
yields ∂ (ρu j ) ∂ (ρu i 2.3. 1 2 d2 dt 2 4.3.12 V V The integral on the left can be viewed on the relativistic moment of inertia tensor while the right hand side is just the volume integral of the components of the energy momentum tensor. we can add equation (2. V which becomes by application of Leibniz’s law ∂ (ρu i d ρ ∂ (ρu j ) dV = 0 . This technique produces results which are in principle coordinate independent but usually utilize some specific coordinate frames for the purpose of calculation. V 2.10) to its counterpart with the indices interchanged. 2. there has been a concerted effort on the part of relativists to seek coordinatefree descriptions of general relativity in order to emphasize the connection between the fundamental geometrical properties of space and the description of associated physical phenomena.3.3. For the last twenty years.8).3.3.14 ∫ x i ∂t ∂t dt V γ and is the relativistic form of the expression for the conservation of angular momentum obtained in section 1.1.11). Perhaps this was a result of the lack of physical phenomena requiring general relativity for their description or possibly the direction of mathematical development undertaken for theory itself.11) from itself. General Relativity and the Virial Theorem The development of quantum theory and the formulation of the general theory of relativity probably represent the two most significant advances in physical science in the twentieth century. 27 . Following the prescription used to generate equation (2. it has drawn attention away from that technique in theoretical physics known as 'moment analysis'.3.13 ∫ x i ∂t − x j ∂t dV = 0 . In light of the general nature and wide applicability of the virial theorem it is surprising that little attempt was made during that time to formulate it within the context of general relativity. but subtracting the transpose of equation (2. equation (2. and get ∂ (ρu j ) ∂ (ρu i c∫ x i + xj ∂t ∂t V dV − 2 ∫ ℑij dV = 0 .11 or finally ∫ [ρ( x j x i ) / γ dV = ∫ ℑij dV . Although this has undoubtedly been profitable for the development of general relativity.The Virial Theorem in Stellar Astrophysics Since ℑij is symmetric. + xj 2.
Gravitation (1973). Virtually every version of the virial theorem emphasizes its global nature. Infield. In an impressive and lengthy paper during 1965. see Misner. Chandrasekhar developed the postNewtonian equations of hydrodynamics including a formulation of the virial theorem20. and Wheeler. w. Infield. One of the fundamental difficulties with the general theory of relativity is its nonlinearity. The problem of operational definition of macroscopic properties in general relativity has plagued the theory since its formulation.†† During the first half of 1960s. It is this nonlinearity that causes so much difficulty with approximation theory and with which the Einstein. It is impossible in principle to measure anything at a point.The Virial Theorem in Stellar Astrophysics Another point of difficulty consists of the nature of the theory itself. what we have seen so far should convince even the greatest skeptic that such a formulation does exist since its origin is basically that of a conservation law (see footnote at the end of the chapter). some sort of symmetrical volume is integrated or summed over to produce the appropriate physical parameters. San Francisco. is a field theory and is thus concerned with functions defined at a point. †__________________________________________ It is worth noting that in order to 'test' any field theory against observation. This is the approach adopted by Einstein. This certainly is not to say that such a formulation cannot be made. General relativity.e. Although continuous progress has been made. This difference becomes a serious problem when one attempts to assign a physical operational interpretation to the quantities represented by the spatial integrals. The physical properties of matter are represented by the geometry of space and in that turn determines the geometry of space. Even something as elementary as density is always "observed" by comparing some mass to some volume. Indeed. This can be done by taking into account in a selfconsistent manner in the Einstein field equations all terms of order 1/c2.† That is. this time honored approach is adequate for calculating the first order (i. Hoffman theory (EIH) deals directly. Although more efficient approximation techniques exist for the calculation of higher order relativistic phenomena such as gravitational radiation. beyond this comfortable realm. Chapter 39. Chandrasekhar developed the virial theorem to an extremely sophisticated level. there does not exist any completely general formulation of the virial theorem within the framework of general relativity at this time. The most comprehensive recognition of this work can be found in his excellent book on the subject9. This obvious statement causes no trouble as long as we are dealing with concepts well within the range of our experience where we can expect our intuition to behave properly. and Hoffman19 in their approach to the relativistic nbody problem and successfully applied by Chandrasekhar20 to hydrodynamics. It is largely this effort which we shall summarize here. 28 . However. The result is to restrict the range of possibility for a theory unnecessarily. we are liable to attribute physical significance and testability to quantities which are in principle untestable. Thorne. in studying the hydrodynamics of various fluid bodies. like so many successful theories.. Let us take a closer look at the origin of some of these problems. Even the general theory of relativity recognizes conservation laws although their form is often altered. H. One of the highlights not dealt with in the book are his efforts to include the firstorder effects of general relativity. c2) terms commonly known as the postNewtonian terms. ††___________________________ For a beautifully concise and complete summary of the postNewtonian approximation. it is necessary to compare integrals of the field quantities with the observed quantities. Freeman & Co.
4 • ℑαβ = 0 .The Virial Theorem in Stellar Astrophysics The basic approach assumes that the metric tensor can be written as being perturbed slightly from the flatspace or Euclidean metric so that (0 gαβ = gαβ) + hαβ . 2. 2 c where J is just the trace ℑαβ . It was this condition that we needed in section 3 to obtain a form of the virial theorem appropriate for special relativity. This procedure can be repeated but there remain some unsolved problems as to convergence of the scheme in general. M. However. The Ricci tensor is essentially a geometric tensor and contains information relating to the metric alone. 2.1 where the hαβ are small terms of the order of c2 or smaller. Instead. In order to follow this prescription one must first start with an approximation to the metric gαβ . At any point one may use the perturbed metric and the prescription for obtaining the equations of motion to generate a set of perturbed equations of motion. The EIH approximation provides a prescription for solving the field equations in various powers of 1/c2 given the information concerning ℑαβ and gαβ .3 ℜ αβ = − 4 (ℑ αβ − 1 Jgαβ ) . consider 29 . Unfortunately the process of taking the divergence looses one order of approximation and thus it is not possible to go directly from the perturbed metric to the equations of motion and maintain the same level of accuracy. exercise 39.4. In general the procedure determines the metric coefficients to one higher order than was originally specified. Thorne and Wheeler16 in 1972 (e.4. This enables one to determine the elements of the energymomentum tensor up to terms of the order of c2 from its definition so that ℑαβ = (ε + P)u α u β − Pgαβ . We have laid neither a sufficient mathematical framework nor developed the general theory of relativity sufficiently to present the derivation in detail without consuming excessive space. That is 2.W.13.4. progress is rapid in a field such as this and what was an original research effort by Chandrasekhar in 1965 becomes a 'homework' problem for Misner. I find in this case I must.2 The Einstein field equations can be written in terms of the Ricci tensor and the energymomentum tensor as 8πG 2. The relativistic prescription that free particles follow geodesic paths is logically equivalent to stating the fourspace divergence of the energymomentum tensor is zero. Here it is traditional to invoke the principle of equivalence that requires that21 2Ψ hαβ = − 2 δ αβ + O (< 1 / c 3 ) . One must first pass through the field equations and the EIH approximation scheme.) Although it is contrary to the spirit of this book to quote results without derivation.4. Like many approximation schemes the mathematical manipulations are formidable and physical insight easily lost.T. 2.5 c With this as a starting point one may proceed with the approximation scheme and obtain the equations of motion.g.4. Indeed it is this condition that in the flatspace metric yields of the EulerLagrange equations of hydrodynamic flow.
where [ ] For purposes of simplification. The distortion of space also changes or at least complicates what is meant by a volume and thus it is useful to define a density ρ* which obeys a continuity equation ∂ρ ∗ + ∇ • (ρ ∗u) = 0 . Thus. It is worth noting that ρo is the density one would find in the absence of general relativity but where relativistic effects are important it is essentially a nonobservable quantity since one can not devise a test for measuring it. Chandrasekhar finds it convenient to define a slightly altered form of the density which explicitly contains the internal energy of the gas and has a slightly altered continuity equation 1 ∂Ψ ∂ρ 0 ∂σ 2. Thorne and Wheeler show that they can be represented by direct integrals over the mass distribution22. With this in mind the equations of motion as derived by Chandrasekhar become23 ρ0 d ∂(σu) − ρ 0 ∇Ψ + ∇ 1 + 2Ψ / c 2 P + 2 [4uΨ − 7 Y − 1 ∇( Y )] + 4u • (∇Y) − 1 (u • ∇)[Y − 1 ∇( Y )] 2 2 2 2 ∂t c dt 1 + ∇ 2 Φ ∇Ψ + ∇ 2 Ψ ∇Φ = 0 . ∂t 2.6 ∗ 2 2 ρ = ρ 0 1 + (u + 3Ψ ) / c .4. These terms are indeed present but Misner. They will be of different types. energy is matter and therefore its motion must be followed in the equations of motion as well as that of matter.7 + ∇ • σu + 2 ρ 0 − = 0 . First. we should find its effects present.The Virial Theorem in Stellar Astrophysics the types of effects we might expect general relativity to introduce and see if these can be identified in the resulting equations of motion. Here Π / c 2 is the internal energy of the gas and P is the local pressure.8 2πGc 2 [( )] [ ] 30 . various authors introduce various kinds of potentials to account for these nonlinear terms. The general theory of relativity is a nonlinear theory and thus we should expect terms to appear which reflect this nonlinearity. Firstly one should expect effects of the Newtonian potential Ψ affecting the metric directly which in turn modifies Ψ. Secondly. This is really a consequence of special relativity but insofar as this 'added' mass affects the metric. Both these effects can indeed be represented by 'potentials' but not just the Newtonian potential.4. motion of matter will 'drag' the metric which will introduce velocity dependent terms in the 'potentials' used to represent those terms. ∂t ∂t ∂t c where σ = ρ 0 [1 + (u 2 + 2Ψ + Π + P / ρ 0 ) / c 2 ] . since the matter and the metric are inexorably tied together.4. 2.
∫ c dt V 2 dt V after noting that remaining integrals in the braces { }of equation (2. Similarly. 2 2 Y is a vector potential whose source is the same as that of the Newtonian potential weighted by the local velocity field. Now we are prepared to write down Lagrange's identity by letting 31 . V V 2.11) is a statement of conservation of linear momentum. ∇ 2 Y = −4πGρ 0 u where ϕ = u 2 + Ψ + 1 Π + 3 (P / ρ 0 ) . This is a useful result for simplifying equation (2. ρ*.4.4.4.11 σu + 2 4uΨ − Y − ∇( Y ) dV = 0 ≡ ∫ K dV . the various potentials which we have introduced can be defined by the fact that they satisfy a Poisson's equation of the form ∇ 2 Ψ = −4πGρ 0 ∇ 2 Φ = −4πGρ 0 ϕ .8).8). we shall compute the scalar version of Lagranges' identity by taking the inner product of equation (2. In this regard it is worth noting that if volume contains the entire system. the procedure for obtaining the general relativistic form of Lagranges' identity is the same as we have used repeatedly in earlier sections. In addition we shall take our volumes large enough so that volume integrals of divergence vanish.4. The lengthy expression in braces contains the effects of the 'dragging' of the metric by the matter and the induced velocity dependent terms. For simplicity.4.10 ∫ ∇AdV = 13 ∫ ∇ • (1 A)dV = 0 .We should expect this procedure to yield terms similar to the classical derivation but with differences introduced by differences between ρo .9 Thus. The remaining contribution to the pressure gradient results from the space curvature introduced by the presence of the matter and is perhaps the most likely relativistic correction to be expected. then by the divergence theorem 2.4. indeed neglecting the contribution to the mass from energy σ = ρo and they are identical to the first term of the NewtonianEulerLagrange equations of hydrodynamic flow. The last term represents the direct effect of the matterenergy potentials on the metrics and this effect. Having obtained the equations of motion for the system. The remaining tensors are the nonlinear interaction terms alluded to earlier. The first two terms are basically Newtonian.8) with the position vector r . Equation (2.4.4. by integrating the equations of motion over the volume yields ρ0 d 7 d 2.The Virial Theorem in Stellar Astrophysics Here.8) can be integrated by parts to zero. is propagated to the potentials themselves. in turn. Expansive as the equations of motion are we may still derive some comprehension for the meaning of the various terms in equation (2. The first part of the third term is just the pressure gradient and thus also to be expected on Newtonian grounds alone. Φ is a scalar potential whose source is again that of the Newtonian potential but weighted by a function ϕ related to the total internal energy field. and σ.
15) is just d2I 32 .4.14 < Φ >= ∫ ρ 0 ϕΨdV V Y = ∫ ρ 0 u • YdV V and Z = ∫ ∫ ρ 0 ρ′ [u • (r − r ' )][u'•(r − r ' )] 3 dV ' dV .The Virial Theorem in Stellar Astrophysics the integral of equation (2.4. The first two terms on the right are just what one would expect in the Newtonian limit while the next term can be related to the total internal energy. The second term arises from the correction to the metric resulting from the potential and the 'dragging' of the metric due to internal motion.4.4. 2.8). this becomes d 1 2 1 ∫ (r • K )dV = 2T + Ω + 3V 1 + 2Ψ / c PdV + c 2 [4W + < Φ > − 74 Y . 2.4.4.13 where T= 1 2 V ∫ σu V 2 dV Ω = − 1 ∫ ρ 0 ΨdV 2 W = ∫ ρ 0 u 2 ΨdV V . 1 2 The first term on the left hand side of equation (2.4. 0 r − r ' V V' This can be made somewhat more familiar if we rewrite the left hand side of equation (2.12 2ρ − 2 (ϕr • ∇Ψ + r • ∇Φ ) = 0 c [ ] After multiple integration by parts and liberal use of the divergence theorem. 2. ∫ dt V ( ) 2.4 Z] . r• 4ρ dK − ρ 0 • ∇ Ψ + P(1 + 2Ψ / c 2 ) + 2 r • [u • (∇Y)] − 1 r • (u • ∇)[Y − ∇( Y )] 2 dt c .13) so that 2 1 d 1 d σr 2 dV + 2 ∫ ρ 0 [4Ψ(r • u ) − 7 r • Y − r • ∇( Y )]dV = 2T + Ω + 3∫ (1 + 2Ψ / c 2 )PdV 2 2 2 ∫ c dt dt V V V .11) be the local linear momentum density K and taking the scalar product of r with equation (2.4.15 1 + 2 [4W + < Φ > − 7 Y − 1 Z] 2 4 c in the Newtonian dt 2 limit.
These equations of motion are completely general and are adequate to describe the effects of rotation and magneticfields if some care is taken with the coordinate frame and the ‘pressure tensor’P. Perhaps the most obvious lesson to be learned from the EIH approach to this problem is that continued application of the theory is not the way to approach the general results. we derived the EulerLagrange equations of hydrodynamic flow. Writing Lagrange's identity as in equation (2. noting that the left hand side is a total time derivative 33 . Similarly as long as the total kinetic energy can be expressed in terms of energies arising from macroscopic motions and internal thermal motions. we may rewrite equation (1. 5.The Virial Theorem in Stellar Astrophysics This term contains a relativistic correction resulting directly from the change in metric due to the presence of matter.4.15) emphasizes the origin of the various terms .1. Thus it seems reasonable to conjecture that these laws as well as the virial theorem remained well posed in the general theory. Thus the presence of complicating forces can be included insofar as they are derivable from a potential. 25 in a true tourdeforce by Chandrasekhar and coworkers looking for additional effects. Ni26 has developed a set of hydrodynamic equations which must hold in nearly all metric theories of gravity and that depend on the values (near unity) of a set of dimensionless parameters. Internal Energy. In Chapter I. radiation reaction terms appear which could be significant for nonspherical collapsed objects which exist over long periods of time. and Rotation The full power and utility of the virial theorem does not really become apparent until one realizes that we need not be particularly specific about the exact nature of the potential and kinetic energies that appear in the earlier derivations. it will be no trouble to express the virial theorem in terms of these more familiar parameters of the system.4). and angular momentum exist and are subject to an operational interpretation at all levels of approximation. momentum. One may proceed in just this manner or return to the original equations of motion for the system. At the 2 1/2 level of approximation. Although terms to this order should be sufficient to describe most phenomena in stellar astrophysics. With this in mind. The Einstein. We shall discuss both approaches. However.whether they are Newtonian or Relativistic. The remaining terms are all energy like and the first two (W and <Φ>) represent relativistic corrections arising from the change in the potential caused by the metric modification by the potential itself. of some consequence is the result that conservation laws for energy. Using a parameterized version of the postNewtonian approximation. This latter effort is useful for relating various terms in the equations to the fundamental assumptions made by different theories. Complications: Magnetic Fields. Infield Hoffman approximation has been iterated up to 2 1/2 times24. we can ask if higher order terms or other metric theories of gravity provide any significant corrections to the virial theorem. The last two involve metric dragging. We have gone to some length to show the problems injected into the virial theorem by the nonlinear aspects of general relativity.
This enables one to see the effect of macroscopic mass motion explicitly in the formalism and thus assess its interaction with other large scale properties of the system.5.5. u = w + (w × r) .5.2 2. we must allow for temporal changes in any vector seen in one frame but not the other.5 It can be shown that for uniform rotation the term .7 2.5. w = constant) become dw 1 1 + 2( w × w ) + w × ( w × r ) = −∇Φ − [∇ • (P + ℑ)] − ∫ S( w)dv .4 1 ∇•ℑ= Dρ e − D × (∇ × D) + c 2 [H × (∇ × H )] .3 or in dyadic notation ℑ = ED − 1 1[E • D] − HB − 1 1[H • B] . 2 If we let 2. Thus ∂u 1 1 = −∇Φ − ∇ • (Pg + ℑ) − ∫ S ( v − u)dv .5. In addition such systematic motion is represented by the stream velocity u in the collision term of the Boltzmann equation making its meaning clearer. Goldstein28 gives a particularly lucid account of how this is to be accomplished by use of the operator d d = + w× . it is appropriate to ignore electrostriction and magnetostriction effects which complicate the relationships between E and D and B and H.5. 2 2 For almost all cases in astrophysics.6 2. then the equations of motion for uniform rotation (i. dt ρ ρV 2. 2 2 k k 2.5 dt inertial dt non inertial where w is the angular velocity appropriate for the point function upon which the operator acts (i.. In the absence of these body forces. let us explicitly represent the effects of rotation.e. the divergence of equation (2.4 4π which is just the Lorentz force on the medium. 2.e. w × (w × r ) = 1 ∇[( w × r ) • ( w × r )] .5.8 34 . just as we have separated the effects of the electromagnetic field from the pressure tensor.1 ∂t ρ ρ where the tensor Pg refers to the gas pressure alone and the tensor ℑ represents the Maxwell stress tensor for the electromagnetic field which has components27 ℑij = E i D j − 1 δ ij ∑ E k D k + H i B j − 1 δ ij ∑ H k B k . It is useful when considering a configuration in uniform rotation to transform the problem into a corotating coordinate frame. { } In transforming the inertial coordinate frame to a noninertial rotating frame attached to the system. Therefore. 2. 2. the angular velocity of the rotating frame).5.5.3) yields2.The Virial Theorem in Stellar Astrophysics and that the pressure tensor can be explicitly split to include the presence of large scale electromagnetic fields.
T.The Virial Theorem in Stellar Astrophysics so that it may be combined with the gravitational potential and 1 ( w × r ) 2 may be considered a 2 'rotational potential'.5.5.5. L is an angular momentumlike tensor2. The most helpful and also reasonable of these is to assume that all collisions processes are isotropic in space. kinetic and potential energy tensors respectively.8). However.10 2 2 4π V dt V [ ] where I . the collision term on the right hand side of the equation of motion averages to zero when integrated over all velocity space. where P is the scalar gas pressure.6 Thus. the gas pressure tensor Pg becomes diagonal with all elements equal and can thus ∇ • Pg = P .4) and (2. Since almost all astrophysical situations relate to plasmas. and equations (2.9 dt 4πρ As we have done before we shall multiply the equations of motion by ρr. the equation of motion becomes dw 1 [H × (∇ × H )] .∫ r∇PdV − 1 ∫ r[H × (∇ × H )]dV . Since both the pressure tensor and Maxwell tensor are normally "fixed" to the body. we will keep these terms for the moment. + 2( w × w ) = −∇ Φ + ( w × r ) 2 − ∇P / ρ − 2. In order to simplify the last two terms it is worth noting that they are derived from the divergence of two tensors. Secondly. in terms of the components. Thus V ∫ r∇ • (P + ℑ) dV = ∫ ∇ • [r(P + ℑ) dV − ∫ (∇r) • (P + ℑ) dV V V . and integrate over the volume and generate a general tensor expression for Lagrange's identity including rotation and magnetic fields. 4π S k S S 2. It is difficult to proceed much further with the equations of motion without making some simplifying assumptions. Firstly.11) becomes 35 . we should expect their formulation in the rotating frame to be simpler. 2 1 d (I ) + 2w • L = 2T + nU + w • (1w ) − ω2 1 .12a Keeping in mind that ∇r = 1 (i. Thus 2..e.5. V S or.2. we may consider the configurations to have a very large conductivity and can therefore neglect the contribution to the Maxwell Field tensor of electric fields.5. This has two results. a second rank tensor with components δij ). 2. This effectively guarantees the existence of a scalar equation of state which will be useful later when relating P to the internal energy. 2. since magnetic fields always extend beyond the surface of what one normally considers the surface of the configuration.5. we may use the chain rule followed by the divergence theorem to simplify these terms.5. and U are moment of inertia. Using these assumptions.5.6. the second integral in equation (2. So.12 ∫ ∇ • [r(P + ℑ) dV = ∫ r(P + ℑ) • dS ≡ S .11 The divergence theorem guarantees that the first integral can be written as a surface integral and if the volume is taken to be large enough can always be made to be zero. The term 2( w × w ) is known as the "coriolis force". the surface terms can be written as 1 1 2 Sij = ∫ x j (H i ∑ H k dS k − 8π ∫ x j H dSi − ∫ P0 x jdSi .
5. in equation (N2.16 V V where l is the net volume angular momentum density on the material due to coriolis forces. the kinetic energy due to internal motion T.5. In section 1.6. and 1 yield the moment of inertia about the origin of the coordinate system I.5.w 1 ) in equation (2. we pointed out that the scalar form is derived by taking "inner" products of the position vector with the equation of motion while the tensor virial theorem involves "outer" or tensor products. U. We can again choose our rotating frame so that ∫ ldV is zero and this term must vanish from the contracted equation. 8π V we arrive at the final tensor form for Lagrange's Identity. 4π V 2.15). 2 1 d I + 2w • L = 2[T . it is clear that the contracted form of that expression can be written as 2 ∫ ρr • ( w × w )dV = −2 w • ∫ ρ(r × w )dV = 2 w • ∫ ldV . V V V V where ℜ is just the total energy due to rotation.5.The Virial Theorem in Stellar Astrophysics V ∫ 1 • (P + ℑ)dV = ∫ 1(P + H V 2 / 8π)dV + 1 ∫ (HH )dV .5.14 ∫ (HH )dV .5. and 1w.M ] + nU + w • ( 1w ) − ω 2 1 + 1 ∫ (P + H 2 / 8π) dV + S .16) by means of identities relating to the vector triple product [see note 2.17 ∫ ρr • [w × (w × r )]dV = ∫ ρr • [w × v]dV = ∫ ρ[r × w ] • vdV = ∫ ρv dV = 2ℜ .5.15) over indices i and j. M. Some care must be taken in contracting the tensors L.6.15) is again to examine the contracted form of the term giving rise to it. Substitution of the contraction of these tensors into equation (2. The simplest method for deriving the value of the contracted form of w • (1 w . The tensor component equations are essential for investigating nonradial oscillations and other such phenomena which cannot be represented by a simple scalar approach. Defining M≡ V We have suffered through the tensor derivation in order to show the complete generality of this formalism. The contraction of the tensors I.5. Since w = w x r.15) taking "inner" products or contract.5. it is easier to appreciate the physical significance of this approach by looking at the scalar counterpart of equation (2. the component form of equation (2. From the definition of L. specifically equation (N2.5. we can expand the left hand side of equation (2. 2. T .13 1 2. the total magnetic energy M . One may either repeat the derivation of equation (2. including rotation and magnetic fields.5. and obtain 2 2. V 2.15 2 dt 2 V where Lijk = ∫ ρ(rk ri ϖ j − ϖ k ri r j )dV .4)].6.3).15) yields a much simpler result 36 . However. the total potential energy Φ respectively.
37 . In practice these integrals are usually small compared with the magnitude of the volume integrals found in the remainder of the expression.21 2 2 2 Combining these two equations we get the total internal "heat energy" as EdV ρdV =∫ U= 2∫ 3 (c p / c v − 1) V (c p / c v − 1) . Before leaving this section on complicating phenomena.5. we can write the internal "heat energy" U as 1 2 U = ∫ ρc v TdV . 2.23 2 dt 2 At the beginning of this discussion.23) down immediately.20 where cv is the specific heat of constant volume and is the temperature. We also know that the kinetic energy associated with the material is E = 3 NkT = 3 ρ(c p − c v )T − 3 P . However. by realizing that 3(γl)U is just twice the kinetic energy due to thermal motion we could have written equation (2. 2.5.5. For example. our efforts have not been wasted. it is unlikely that this could have been done for equation (2.23) are just the contribution to the total kinetic energy of the system from macroscopic motions. rotation and thermal motion respectively. H. V 2. The remaining terms are just a specification of the nature of the total potential energy. Thus.The Virial Theorem in Stellar Astrophysics d 2I = 2(T + ℜ) + M − Ω + nU + 3∫ PdV + ∑ Sii . In Chapter I. then we can write the scalar form of Lagrange's identity as 2 1 d I = 2(T + ℜ) + M − Ω + 3( γ − 1)U .22 It is traditional to let γ = cp /cv and if we let this be constant throughout the volume and neglect the surface term in equation (2. Although no modification is made to the virial theorem some effects can be seen in Lagrange's identity and so. I said that one could either derive Lagrange's identity from the equations of motion or from careful consideration of meanings of potential and kinetic energy. This result was apparently first noted by E.19 ∑ Sii = 4π ∫ (r0 • H 0 )(H 0 • dS ) − ∫ (P0 + H 02 / 8π)r • dS .15).18).5. i S S Here.5.5. there is one last aspect to be investigated. Milne29 in 1925. From thermodynamic consideration. 2. Since even in the tensor representation this term appears as a scalar. V 2.5. So far. we noted that the inclusion of velocity dependent forces such as frictional force do not alter the results of the virial theorem as they can be averaged to zero given sufficient time.5. we have said little about the contribution of the volume integral of the pressure.5.5. there was no loss of generality in deferring the evaluation of the integral until now. Dimensional analysis will lead one to the result that the pressure integral is an 'energylike' integral. the first and last terms on the right hand side of equation (2.18 dt 2 i V The contraction of the surface terms yield the scalar integrals 1 2. section 3. Since the variational form of this equation will be useful in the next chapter. the subscript "0" indicates the value of the variables on the surface.
Further insight into the nature of this process is discovered in the second section where we find that taking higher order spatial moments of the equation of motion is equivalent to recovering the information selectively lost in the classical derivation of the virial theorem. 2.e. high velocity electrons. In principle this approach could be used in a prescription for the complete solution of the equations of motion. Thus.24 which make the following contribution to the tensor form of Lagrange's identity: d 2. pure radiation spheres which approximate some models of super massive stars and white dwarfs where lowmass.25 ∫ rf f dV = V ρrw • fdV = 12 V ρ dr (rr ) • fdV . However. were kept in check by the highmass..The Virial Theorem in Stellar Astrophysics let us take a moment to recapitulate these arguments. 38 .5. Summary In this chapter we have continued the development of the virial theorem as it appears in more contemporary usage. it seems likely that in practice it would be more difficult than implementing a direct numerical solution of the original equations themselves. The importance of the method lies in the fact that such moment expressions for stable systems are normally rapidly convergent.26 :f.5.5. the righthand side of equation (2. We found that large velocities require large gravitational fields to keep them in check and thus one might argue that a separate discussion of the effects of special relativity is not warranted. If we have a volume force which can be derived from a "friction" tensor.25) would just become 1 dI 2. ∫ ∫ V If f were indeed constant throughout the configuration. there are at least two dynamically stable systems for which this is not true (i. the largest amount of information can be recovered with the least effort.5.27 1 dt 2 dt 6. In the next two sections. 2 dt and Lagrange's identity in its full generality would be 1 2 d 2 I 1 dI + 2 : f + 2 w • L = 2(T − M ) + nU + w • [ w − w1 ] + 1[(1 − γ )U + M ] + S . However. then one would add terms to the equations of motion of the form f f = ρw • f . we considered the effects of a relativity principle on the development of the virial theorem. 2. The tensor virial theorem is a more general form of Lagrange's identity which when averaged over time provides rather general expressions for the coordinated behavior of some energy like tensors of the system.5.
39 . the perceptive reader may have noticed that the derivation is equivalent to carrying out 2. but for simplicity I presented the development of the scalar version of the virial theorem only.3 Because of the linear independence of the time and space coordinates. V where r is a four vector in the Lorentz metric. However. Bonazzola30. ∫ j dt V which after appropriate application of the divergence theorem.2 2 ∫ tℑ00 dV = V t ∑ ∂x j dV .6. this is not a new result but rather an expression of the conservation of energy. I have attempted to provide physical motivation from the existence of terms arising in the equations of motion that results from the nonlinear nature of general relativity. Lagrange's identity is applicable to systems which are not in equilibrium and hence may be relativistic. Thus any formulation of a viriallike relation will depend on the specific nature of the configuration. things are no longer simple. I have attempted throughout the chapter to emphasize the similarity of the derivation of the virial theorem and particularly in section 3. we ignored all contributions from the timelike part. Thus rather than stress the manipulative complexity of the postNewtonian approximation. Sticking with the postNewtonian approximation avoided some difficult problems of uniqueness and interpretation. It is this loss of simplicity which caused me to stray from the more rigorous approaches of other sections. we have developed the special and general relativistic versions of the theorem separately and will return in the last chapter to discuss some specific applications of them.6. These would have been of the form ∂ℑ0 j 1 d 2. the existence of gravitational radiation destroys any strict equivalence between the general relativistic and Newtonian cases so that there can be no unique formulation of the virial theorem. nonspherical case.1 ∫ r • • ℑ dV = 0 . He notes that in the nonstationary. stationary case in full general relativity.6. †_______________________________________ Since the Pachart edition of this effort was written. The derivation presented in this section follows exactly the prescription of earlier sections.The Virial Theorem in Stellar Astrophysics low velocity nucleons). This approach was extended by Vilain31 who applied a similar formulation of a general relativistic virial theorem to the stability of perfect fluid spheres. becomes dE d ρc 2 dx ∫ εdV = = −∫ • dS . In addition. there has been some significant progress in this area. Essentially it states that the time rate of change of the total energy of the system equals the momentum flux across the system boundary multiplied by c2. For this reason. However. formulated the virial theorem for the spherical. When the metric is no longer the 'flat' Lorentz metric as in the case in section 4. dt dt V γ dt S 2. I remain convinced that a general formulation of Lagrange's identity and the virial theorem which is compatible with the field equation of Einstein exists and its formulation would be most rewarding†.
Lastly we looked again at velocity dependent forces not so much with an eye to their effect on the virial theorem. The results which have been generalized to include additional effects are not the result of any new physical concepts. Although I included only magnetic fields throughout most of the discussion. all the rest is done to glean more insight. Rather.The Virial Theorem in Stellar Astrophysics For most stellar astrophysical applications the postNewtonian result is probably sufficient. The presence of all these complicating aspects is included only to make their interplay explicit. but rather with a view to their persistent presence in the variational form of Lagrange's identity. the inclusion of electric fields in the Maxwell field tensor makes it clear how to proceed should they be present. they are the result of the specific identification of the physical contributions to the system made by such attributes as magnetic fields and macroscopic motion. 40 . The basic theorem must hold. In the last section of this chapter some of the powers of the virial theorem to deal with difficult situations became apparent.
2.1. N2.1. equation (1. j k V j k V j k V k k k The first integral on the right is just the integral of the divergence of x j ℑ jk over V and if the volume is chosen to enclose the entire system the integral must vanish as ℑ jk = 0 on the surface 41 .1 The term on the left of equation (2.1.1 N2.3.1 Tij = ∫ ρu i u j dV V .2.4.6) in component form.1.1 ∂t ∂x k j k V V j From the chain rule the second integral can be written as ∂x j ∂ℑ jk ∂ N2. 2 2 ∫ dt dt dt V V V dt N2.3.2 If one writes equation (2. 2 dt 2 where in Cartesian Coordinates the tensor components are Iij = ∫ ρx i x j dV V N2. ∫ ∫ ∫ V The third term can further be simplified So that d2 d2 d rdr ρr dV = 1 ∫ ρ 2 (rr ) dV = 1 2 ∫ ρrr dV .3 Taking the scalar product of a spacelike position vector with equation (2. N2.2.3 2 ∫ dt 2 V V V 2. we recover the scalar form of Lagrange's identity as given in Chapter I.1.4) we have assumed that the volume V is large enough to contain all matter and the conservation of mass argument explicitly developed in Chapter I has the form 2 1 d ρrr dV = ∫ (ρuu )dV + ∫ ρr∇ΦdV . i N2.12).6) and integrating over a volume sufficient to contain the system we get ∂ (ρu j ) ∂ℑ jk c∫ ∑ x j dV = ∑∑ ∫ x j dV = 0 . one gets d 2 Iij 1 = 2Tij + nUij .3.2 ∑∑ ∫ x j ∂x dV = ∑∑ ∫ ∂x (x j ℑ jk )dV − ∑∑ ∫ ∂x ℑ jk dV .2 Uij = 1 2 V V' ∫ ∫ ρ(r)ρ(r' )(x − x ′ )( x j − x ′j )( r − r ' ) i n −2 dV' dV By contracting these tensors and letting the potential be that of the gravitational field.The Virial Theorem in Stellar Astrophysics Notes to Chapter 2 2.1.2) becomes dr dr d 2r d rdr du ∫ ρr dt dV = V ρr dt 2 dV = V ρr dt dt dV − V ρ dt dt dV .2 In obtaining the third term in equation (2.
2 { ( )} N2. ∂t j V N2. j=1 V V 3 N2.7 2.3.3. j = 0 from the orthogonality of the Lorentz frame and ∑ ℑ jj = ∑ ℑαα − ℑ00 .4.3. this becomes 1 ∇•ℑ= (D • ∇)D + Dρ e + (c 2 H • ∇)H − 1 ∇ D • D + c 2 H • H .3.3.1 { ( )} N2.3.4 42 .3.1) is then ∑ c∫ x j V j ∂ (ρu j )dV − ∑ ∫ ℑ jjdV = 0 . Applying Leibniz’s law for the differentiation of definite integralsl8 to the first term in equation (N2. Since ∂x i / ∂x j = δ ij .3 3 3 ∂x Now.3.6 ∑ ∫ ∂t 12 ρ dt dV + ∫ (ρc 2 − γτ − ε − ρc 2 )dV = 0 .4.3.6) and rewriting the second one. 2 4π and invoking Maxwell's laws that ∇ • D = ρ e and ∇ • H = 0 .5 Thus. 2 4π Making use of the vector identity ∇(A • G ) = A × (∇ × G ) + ( A • ∇)G + G × (∇ × A) + (G • ∇) A .The Virial Theorem in Stellar Astrophysics enclosing V j V (i. among others.4 ℑ= 1 DD + c 2 HH − 1 1 D 2 + c 2 H 2 .3) takes on the more familiar form [ ( )] N2. shows us that the "relativistic" kinetic energy density τ is given by τ = ρc 2 ( γ −1 − 1) or γτ = ρc 2 − ρc 2 γ .4).4. equation (N2. we get 3 ∂ d (x j x j / γ ) N2. we get ∂ 1 d (x j x j / γ ) ∑ ∫ ∂t 2 ρ dt dV = Ω + T + ∫ γτ dV . the second integral becomes ∑ ∫ ℑ jj dV . using equation (2. Equation (N2. 2 4π If we take the divergence of ℑ we get 1 ∇•ℑ= (D • ∇)D + D(∇ • D) + (c 2 H • ∇)H + c 2 H(∇ • H ) − 1 ∇ D • D + c 2 H • H . Gauss's Law applies). Bergmann17. so we ∂t j=1 α =0 can write ∂ N2. ∂t j V V V With the sign convection ℑ00 is the negative of the total energy density..e.3) to rewrite the first term of equation (N2. j=1 V V where ε is the potential energy density.4 c ∫ ∑ ρx j u j dV + ∫ ρc 2 γdV + ∫ ℑ00 dV = 0 .3 N2.4.4. N2.
4 ∂x i i j k j k where eijk in a completely antisymmetric tensor of rank 3 sometimes called the Levi Civita tensor density. since w is constant. since there can be no net motions about any axis other than that defined by the total angular momentum of the body all components of w • L must be zero except those associated with the axis of rotation.6. Thus. w • L is a kinetic energy like tensor resulting from the net motions induced by the coriolis forces.4.4. Again. N2. 4π { } N2. by expanding the resulting vector triple product in the second term we can write 2 ∫ ρr ( w × w )dV = 2 ∫ ρr[w × (ϖ × r )]dV = 2∫ ρr[ϖ ( w • r ) − r ( w • ϖ )]dV N2.2 The constancy of w causes the second and fourth term to vanish.6. N2.4).5 2.3 ∫ ρr [ ] or 2 ∫ ρr ( w × w )dV = 2 w • ∫ ρ[r (r ϖ ) − ϖ (rr)]dV = 2 wL . N2.5 2 ∇ w×r but 2.5. N2. N2.3 In component form [(w × r ) • ∇](r) = ∑∑∑ eijk ω j rk ∂x = ∑∑ e jk ω j rk = (w × r ) . In addition we can choose our rotating frame w so that in that frame the net angular momentum is zero and all 43 . [(w × r ) • ∇](w × r ) = w × [(w × r ) • ∇](r ) .1 ∇ × (w × r ) = (w • ∇ )r − (r • ∇) w − w (∇ • r ) + r (∇ • w ) . N2.5. However. Thus 1 ( )2 = w × (w × r ) . 2 N2.The Virial Theorem in Stellar Astrophysics ∇•ℑ= 1 (Dρ e − D × (∇ × D) + c 2 [H × (∇ × H )] .5. it is clear that 1 2 ∇(w × r ) = (w × r ) × [∇ × (w × r )] + [(w × r ) • ∇ ](w × r ) .5 By use of the vector identity equation (N2.5. Let the local velocity field w = ϖ × r by defining a local angular velocity field ϖ . Then.6.2 2 ∫ dt dt V V V The second term is more difficult to simplify.5. V V V V V where the tensor L is an angular momentlike three index tensor representing the various components of volume net angular momentum within the body. the first term on the left becomes d2 dw ρr dV = 1 2 ∫ ρ(rr )dV − ∫ ρ(ww )dV . while the first third terms are equal but of opposite sign.6 dw 1 2 dV + 2 ∫ ρr ( w × w )dV = − ∫ ρ∇ Φ + (w × r ) + P / ρ dV − ∫ r[H × (∇ × H )]dV .1 dt 4π V V V V As in section 3.
4 or 2 ρr∇(w × r ) dV = w • ∫ ρrrwdV − ω 2 ∫ ρrrdV .2).6. (N2.6.6. (N2.4).6. 2 2 V V V N2.10)].6.5. and (2. Thus 1 2 V ∫ ρr∇(w × r ) dV = ∫ ρr[w × (w × r)]dV = ∫ ρrw (w • r)dV − ∫ ρrrω dV .1. Since we 2 just went to some length to show that. For the sake of generality we shall keep the term for the present.4). 44 . Thus by combining 1 2 V V equations (N2. 1 ∇(w × r ) = w × (w × r ) .6. The first term on the right of equation (N2. we shall use the earlier version and 2 an expansion of the vector triple product to evaluate the second term on the right.1) so far [see equation (2.4). we may assess our progress in simplifying equation (N2.The Virial Theorem in Stellar Astrophysics contributions from the term w • L will vanish. ∫ V V V It is worth noting at this point that all terms dealt with so far involve only the volume integrals ∫ ρrrdV and ∫ ρww dV which are the same as the tensors defined in Section 1.1) is given by equation (2.1.3).
. 2. M. pp. Lebovitz. and Wheeler. Ap. Ap. pp. Panofsky. pp.. Ap. pp.. Reading. 135. The Classical Theory of Fields. J.. E. pp.The Virial Theorem in Stellar Astrophysics References 1 2 3. p. Ap. Freeman & Co. pp... USA. p. Bonazzola. 96. Phys. 1032. S. 80. 152. K. 136. PrenticeHall. 81. R. (1959). Chandrasekhar. 293304. Ap. p. Co. S. (1925). p. and Fermi. Mass. J. L. S. Goldstein. Chandrasekhar. pp. Vol. (1938). 162. American Journal of Physics 28. 116. Suppl. W. Ni. and Lifshitz. Chandrasekhar. 136. pp. and Phillips. D. 335340. (1961). W. N. Y. 6. Chapt. Chandrasekhar. Misner. S. Ap. 340. (1960). eq. (1962). A. 136. Ellipsoidal Figures of Equilibrium. N. 424. J. (1969). W. 135. Vol. Chandrasekhar. J. Phil. p. (1962). pp. R. C. R. S. (1962). 158. Oxford University Press. S. Introduction to the Theory of Relativity. 4. W. I. S. 3. pp. J. S. Ann. Co. p. E. London. Classical Mechanics. 160. pp. and Nutku.1537. J. Inc. Parker. Schiff. S. C. pp. Mag. S. J. H. P. 1499. 153179. 4 5 6 7. . Reading.. McGrawHill Book Co. H. 50. Chandrasekhar. J. M. p. p.. (1962). . H. pp. (1973).. eq. S. 10 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 Rayleigh. and Lebovitz. Chandrasekhar. I. Einstein. (1961). 181. 957976. M. New York. N. 409414. Ap. AddisonWesley Pub. (1965). (1953). 14881512. (1962). and Wheeler. G. N. 10691081. 45 . Hydrodynamic and Hydromagnetic Stability. Ap. Chandrasekhar. J. J. S. E. p. Ap. A. 142. E. Math 39. Chandrasekhar. (1973). Ap. pp. Infeld. and Lebovitz. San Francisco. 571581. A. pp. Gravitation. Reading. Scientific Papers... (1942). S.. I. Ap. pp... (1969). 10321036. J. Bergmann. N. Ap. 16869. Inc. 142. S. New York. 1082. Mass. J. Ap. (1954). (1973). pp. J. H. 10371047. Inc. and Hoffman. pp. 262. Misner. 9597.. 118. (1968). and Lebovitz. J. (1957). 51. J.. K. J. and Redheffer. Gravitation. Rev. 238247. Chandrasekhar. (1903). 491. 65100. 92. Milne. Ap. Cambridge.. P. (1957). S. Thorne. Sokolnikoff. (1973). L. Classical Electricity and Magnetism. England. pp. (1962). (1970). San Francisco. pp. Dover Pub. Yale University Press. A. 182.. 500536. 8 9. Landau. J. 634. (1955).133. Co. Freeman & J Co.. AddisonWesley Pub. Mathematics of Physics and Modern Engineering. and Esposita. 248..(1958). K. An Introduction to the Study of Stellar Structure. Ap. Mass. R. W. S. Trans.. (1965). Hamermesh. 238. AddisonWesley Pub. Thorne.. Chandrasekhar. W. E. 5579.
Freeman and Co.. Ap. Ellipsoidal Figures of Equilibrium. (1989) The Fundamentals of Stellar Astrophysics. G. 46 . Collins. 307318. C.The Virial Theorem in Stellar Astrophysics 31 32 33 Vilain.H. New York. Chandrasekhar. (1987).II. Dover Publications Inc. W. pp.. J. pp. 227.1537. New York.. pp. (1979).W. S. 160167.
The Virial Theorem in Stellar Astrophysics 47 .
r = r0 + δr 3.1. dr 2 so that by subtracting equation (3.1. Since equation (3.3) becomes d 2 r0 d 2 δr + 2 = −∇Φ − ∇δΦ . consider the equations of motion for an object moving under the influence of a point potential Φ.1.1.1) would require d 2 (r + δr ) = −∇(Φ + δΦ ) . dr 2 3.1.1.6 3. dt 2 dt but we already know that d 2 r0 = −∇Φ 0 .3 3. since both ∇ and d/dt are linear operators.1.5) from equation (3. we get d 2 δr = −∇δΦ . which satisfies equation (3.1) is valid for any system where Φ is known. equation (3.1) for a particular potential Φo.4). The vehicle for determining the independent variable changes is found in the very equations which describe the initial state of the system. For example.1.5 3. The equations usually chosen for this type of analysis are the equations of motion for the system.1.The Virial Theorem in Stellar Astrophysics Copyright 2003 III The Variational Form of the Virial Theorem 1.1.1. Variations and Perturbations and their Implications for the Virial Theorem Perturbation analysis is truly an old mechanism in which one explores the behavior of a system in a known state by assuming there are small variations of the independent variables describing the system and determining the individual variation in the independent variables.1.7 48 .1 dt 2 Assume a solution ro(t) is known. one could define Φ = Φ 0 + δΦ . 3. dt 2 However. d 2r = −∇Φ .2 and equation (3.
and potential energy as a function of time. with which this approach deals. might not the variational form of the moments of those equations also be expected to contain interesting information? It was in this spirit that Paul Ledoux developed the variational form of the scalar virial theoreml.2. A short approach which leads to the same result is to "take the variation" of equation (3.1. Thus. obtain expressions for the variation of the moment of inertia.1 49 . variational analysis of the equations of motion would yield a description of the system motion about the equilibrium configuration. if taking variations of the equations of motion produces useful results. Remember from the earlier section that Lagrange's identity is 2 1 d I = 2T + Ω .The Virial Theorem in Stellar Astrophysics which we called the perturbed equations of motions where δΦ is the perturbation that involves the perturbation δr. as Ledoux1 demonstrated by determining perhaps the most obvious macroscopic property of such motion. to obtain an expression for the frequency of radial pulsations in a gas sphere. An additional macroscopic property closely connected with the pulsational period. hence any linear perturbation or departure from a given solution will produce the sum of the original equations of motion on the perturbed equations of motion. the pulsational period of the system. This technique "works" because the time and space operators in the equation of motion are linear. It is not necessary that one perturbs the equations of motion in order to gain information about the system. If that state happens to be an equilibrium state. Variational analysis of spatial moments could then be expected to yield macroscopic properties of that motion. it represents a small departure of a variable from the value it had which satisfies the equations governing the system. and was able to predict the pulsational period of a star.1) wherein the operator δ is not affected by time or space derivatives. Radial Pulsations for SelfGravitating Systems: Stars In this section we shall use the virial theorem. We shall examine this aspect of the analysis later. that is. is the global stability of the system. Chandrasekhar2 found the tensor form of the virial theorem useful in determining nonradial modes of oscillation of stars. In general. The approach will be to apply a small variation to the virial theorem and by making use of several conservation laws. In addition. he and Fermi3 investigated the effects of a magnetic field on the pulsation of a star. kinetic energy. let us be content with observing in some detail how the variational approach yields the pulsational periods of stars. Clearly any equations which describe the structure of the system are subject to this type of analysis. 2 dt 2 3. I shall use the variational operator δ in this sense. 2. For now. The variational approach yields differential equations which describe parameter relationships for a system disturbed from an initial state. This is indeed the case.
The Virial Theorem in Stellar Astrophysics
In this form, no time averaging has been carried out and the equation must apply to a dynamic system at any point in time. Now, consider a star with radius R. Let r be the distance from the center of symmetry to any point in the configuration and or be the displacement of a point mass from the equilibrium position ro. Conservation of mass requires that for a spherical shell of radius m(r0 + δr ) = m(r0 ) . 3.2.2
We wish to find the variations δI, δT, δΩ of the quantities I, T, Ω, from the equilibrium values I0, T0, and Ω0. The variational form of the virial theorem then becomes
1 2
d 2 (δI) = 2δT + δΩ . dt 2
3.2.3
Since I was defined as the moment of inertia about the center of the coordinate system, we have by definition that,
I = ∫ r 2 dm(r ) .
0
M
3.2.4
Thus, we have
δI = ∫ 2rδrdm(r ) + ∫ r 2 δ[dm(r )] .
0 0
M
M
3.2.5
Since, by the conservation of mass [equation (3.2.2)], δm(r) = 0 for all r, d(δm(r) ) = δ(dm(r) ) =0 for all r and the second integral of (3.2.5) vanishes leaving δI = ∫ 2rδrdm(r ) .
0 M
3.2.6
Now from equation (3.2.4), we have dI = r2 . dm(r ) Since this must always be true, it is true at the equilibrium point ro. Therefore, dI 0 = r02 dm(r ) . So, to first order accuracy in r, we may rewrite equation (3.2.6) as I 0 δr δI = 2 ∫ dI 0 . 0 r 0 3.2.7
3.2.8
3.2.9
In a similar manner we may evaluate the variation of the gravitational potential energy with respect to small variation in r 3.1 and obtain Ω 0 δr δΩ = 2∫ dΩ 0 . 3.2.10 0 r0 All that now remains to be determined in equation (3.2.3) is the variation of the total kinetic energy T. To first order only the variation of the thermal kinetic energy will contribute to equation (3.2.3).3.2
50
The Virial Theorem in Stellar Astrophysics
P0 δρ ( γ − 1) 0 dm(r ) . 3.2.11 0 ρ ρ0 0 In order to facilitate obtaining an expression for δρ / ρ 0 we shall now specify a time dependence for the pulsation about ro. For simplicity, let us assume the motion is simply periodic. Thus, defining a quantity ξ as δr ξ= = ξ 0 eiσ t , 3.2.12 r0 where 2π/σ is the period of oscillation, we may rewrite the variations of I and Ω as follows: 2δT ≅ 3∫
M
δI = 2e i σ t ∫ ξ 0 dI 0
0
Conservation of mass requires that3.3
. Ω0 δΩ = −e i σ t ∫ ξ 0 dΩ 0 0
I0
3.2.13
dξ δρ 3.2.14 = − 3ξ 0 + r0 0 e i σ t . ρ0 dr0 Substitution of this back into the expression for the variation of the kinetic energy yields M P dξ 3.2.15 2δT = −3∫ 0 ( γ − 1) 3ξ 0 + r0 0 e i σ t dm(r0 ) . 0 ρ dr0 0 Equation (3.2.15) may be simplified to yield3.4 M P ξ Ω0 d 3.2.16 2δT = −3 e i σ t ∫ 0 0 dm(r0 ) + 3 e i σ t ∫ ξ 0 ( γ − 1)dΩ 0 . 0 0 ρ 0 dr0 We now have all the material necessary to evaluate the variational form of the virial theorem to first order accuracy. Substituting equations (3.2.13) and (3.2.16) into equation (3.2.3), we obtain I0 M P ξ r dγ Ω0 Ω0 dm(r ) + 3e i σ t ∫ ξ 0 ( γ − 1)dΩ 0 − e i σ t ∫ ξdΩ 0 . 3.1.17 − σ 2 e i σ t ∫ ξ 0 dI 0 = 3e i σ t ∫ 0 0 0 0 0 0 0 ρ 0 dr0 Solving for σ2, which is related to the pulsation period, we have M P ξ r dγ Ω0 dm(r ) − ∫ (3γ − 4)ξ 0 dΩ 0 + 3∫ 0 0 0 0 0 ρ 0 dr0 2 . 3.2.18 σ = I0 ∫ ξ 0 dI 0
0
For a model of known equilibrium structure, the integrals in equation (3.2.18) may be evaluated and the frequency for which it is stable to radial pulsations may be computed. However, for purposes of examining the behavior of a pulsating star we may assume the star is sufficiently homogeneous so that γ is constant. Also, let us assume the pulsation increases radially outward in a linear manner. Under these admittedly ad hoc assumptions, equation (3.2.18) reduces to the extremely simple form
51
The Virial Theorem in Stellar Astrophysics
σ2 = −
(3γ − 4)Ω 0 . I0
3.2.19
In order to obtain a feeling for the formula we have developed, we shall attempt to estimate some approximation pulsation frequencies. For a sphere of uniform density Ω0 = 3 GM 2 . 5 R0 3.2.20
The moment of inertia for a sphere about an axis is equal to 3/2 the moment of inertia about its center and is given by 2 3 2 I Z = MR 0 = I 0 . 3.2.21 5 2 Therefore, 2 4MR 0 I0 = . 3.2.22 15 Prior theory concerning stellar structure implies that γ > 4/3. If we take γ = 5/3, appropriate for a fully convective star, we obtain 9 GM σ2 = , 3.2.23 4 R3 0
or
σ 2 = 3πG ρ .
Remembering that the period T is just 2π/σ we have 4π − 1 2 T = 3G ρ .
3.2.24
Thus, we see that the theory does produce a period which is inversely proportional to the square root of the mean density. This law has been found to be experimentally correct in the case of the Classical Cepheids. It should be noted that this property will be preserved even for the integral form equation (3.2.18), only the constant of proportionality will change. If we evaluate the constant of proportionality from equation (3.2.24), we have T ≅ 7.92 × 10 ρ
3 − 1 2
sec.
3.2.25
where ρ is given in (gm/cc). Taking an observed value for the mean density of a Cepheid variable to be between 103 and 10 cm/cc (eg. Ledoux and Walraven)4, we arrive at the following estimate for the periods of these stars. 0.3 days < T < 90 days 3.2.26
6
52
In order to simplify the mathematical development we shall make some of the assumptions which were made during previous sections. That is. 3. γ will be constant throughout the system. however. Thus. only knowledge of the total magnetic and rotational energies will be required.e. 3. This would imply that a value for the period calculated in this manner should be correct within a factor of 2 or 3. As the mass becomes more centrally concentrated the magnitude of the gravitational energy will increase while the moment of inertia will decrease. Radiation pressure will be considered to be negligible (i. A first order theory will be adequate. we may write the scalar form of Lagrange's identity as follows: 2 1 d I = 2T + Ω + M . for the moment let us neglect these terms since they usually will be small and as such will not affect the general character of the solution. and it is comforting that the result nicely brackets the observed periods for Cepheid variables. One of the points of that derivation that required some care was the inclusion of surface terms arising from the fact that stellar magnetic fields usually extend well beyond the normal surface of the star. Γl = γ). without solving the force equations. We have already seen that it is possible to write Lagrange's identity so as to include the effects of rotational and magnetic energy. Thus.2.23) for σ2 is a lower limit. 2. since our approach expresses the pulsation frequency in terms of volume integrals. It is worth noting that solution of such a problem in terms of the force equations would be difficult indeed as it would require detailed knowledge of the geometry of the magnetic field throughout the star. Even for reasonable density distributions the value arrived at in equation (3.23) will not differ by more than an order of magnitude. as in Chapter II. However. These assumptions are listed below: 1.The Virial Theorem in Stellar Astrophysics It is freely admitted that this estimate is arrived at in the crudest way. It should also be noted that for most stars the expression arrived at in equation (3..1 2 dt 2 53 . 3 The Influence of Magnetic and Rotational Energy upon a Pulsating System We shall now consider what the effect of introducing magnetic and rotational energies into a pulsating system will be upon the frequency of pulsation of that system.3. However.2. all deviations from equilibrium shall be small. an estimate for a very important parameter in describing the pulsation of a gas sphere may be obtained which is the period for which that sphere is stable to radial pulsation.
6 2 dt 2 The variational form becomes 1 2 d2 (δI) = 2δT3 + 3( γ − 1)δU + δΩ + δM .7 There is no term containing a variation in γ as it is assumed to be constant throughout the system.). dt 2 3. we may take the variation of equation (3.3. equation (3.3. 3. we may write the virial theorem for the system as 2 1 d I = 2T1 + 2T3 + 3( γ − 1)U + Ω + M .2) and equation (3.8 Since we have already computed the variation of the kinetic energy of the gas. 3.2 where T and R are the gas temperature and constant respectively.15).3. 3. Now let us break up the total kinetic energy of the system into the sum of three energies Tl. Thus 2δT1 = 0 .3. we have 3 3.3. 2 Integrating this over the entire system we obtain 2T2 = 3( γ − 1)U . But the internal energy dU of the element of mass is dU = c vTdm .4 dT2 = ( γ − 1)dU . 3. we may most easily find the variation of the total internal energy (δU) in terms of this quantity. The contribution to the kinetic energy (of particle motion) due to an element of mass is dT2 = 3 3 3 kTdn = RTdm = (c p − c v )Tdm . gives the following expression for the variation of the kinetic energy of the gas: M 3P ξ 0 0 2δT2 = −3∫ ( γ − 1)e i σ t dm(r0 ) . 2 2 2 3. 3.3. As a result of the assumption of constant γ. and obtain 2δT2 = +3( γ − 1)δU .3.3. it was shown that variation of the pulsational kinetic energy was of second order and could therefore be neglected.2.3. if we further assume a periodic form for the pulsation and linearly increasing amplitude (ξo=const.5).3.3 Combining equation (3. T2 is the kinetic energy of the gas due to thermal energy. T2 and T3 where Tl is the kinetic energy of the pulsating system due to the pulsating motion.3. with the definition of γ. 3. In section 2.9 Now.The Virial Theorem in Stellar Astrophysics where T is the total kinetic energy including rotation and M is the total magnetic energy.3.5 Thus. and T3 is the kinetic energy of rotation.3).10 0 ρ0 54 .
we finally obtain 3. from the elementary kinetic theory of gases we have T2 = 3 ∫ P0 dV .13 2 V Also. 3.3.15) with the right side of equation (3.11).10) and equation (3.3. δΩ = −e ξ 0 Ω 0 δI = 2e i σ t ξ 0 I 0 iσ t 3.3. Consider first the rotational energy.3. ( ) 3. we obtain δU = −ξ 0 e i σ t ∫ 3P0 dV . Now an element of mass rotating about an axis with a velocity ω will possess an elemental angular momentum dL = ω x 2 + y 2 dm(r ) = ωr 2 sin 2 θ dm(r ) . the relevant quantities 2 are averaged over one pulsation period so that the < d I 2 >= 0 . we need only obtain expressions for the variation of the magnetic and rotational energies in order to evaluate equations (3.6) then becomes 2T1(0) + 2T2 (0) + 2T3 (0) + Ω 0 + M = 0 .3. T1 (0) =< T1 >= 0 .3.18 55 .14 Making use of these two results and replacing the average values of the quantities in equation (3. these equations become .11 In order to express this variation in terms of the other energies present in the system. V 3.3. 3. we shall assume that the system is in quasisteady state. We shall assume that the dt remaining values are the equilibrium values of the configuration.3.The Virial Theorem in Stellar Astrophysics or 2δT2 = −3( γ − 1)ξ 0 e i σ t ∫ 3P0 dV . Under the assumption used above.16 δU = ξ 0 e i σ t (2T3 (0) + Ω 0 + M0 ) . 3.3. With this assumption. We have already determined expressions for the variations of the gravitational energy and moment of inertia in the previous section [equation (3.2.3.12) with the equilibrium values we find − 3∫ P0 dV = 2T3 (0) + Ω 0 + M0 . since the system is neither expanding or contracting.7).9).13)].3. 3.3. V Combining equation (3. of a constant perturbation ξo.15 V Identifying the left side of equation (3.3.17 Thus.3. The virial theorem as expressed in equation (3.12 Now.3.
The Virial Theorem in Stellar Astrophysics
Here the xy plane is the plane perpendicular to axis of rotation and θ is the polar angle measured from the axis of rotation. Such an elemental mass possessing such an angular momentum will have a rotational kinetic energy given by dT3 = 1 ωdL . 3.3.19 2 Thus, the total rotational kinetic energy is just
T3 =
1 2
∫
L
L
0
ωdL .
3.3.20
We may use this expression to obtain the variation δT3. The first term on the right of equation (3.3.7) becomes 2δT3 = δ ∫ ωdL = ∫ δωdL + ∫ ωd (δL ) .
0 0 0
L
L
3.3.21
However, the conservation of angular momentum requires that L remain constant during the pulsation. Thus, the variation of L is zero and the last integral on the right of equation (3.3.21) vanishes, so that L L δω 2δT3 = ∫ δωdL + ∫ ω0 dL . 3.3.22 0 ω 0 0 where ωo is the rotational velocity of the equilibrium configuration. Now, again making use of the conservation of angular momentum we see that ωr 2 sin θ = const. 3.3.23 Since we are only considering radial pulsation so that δθ is zero, the variation equation (3.3.23) yields δωr 2 sin θ + 2ωrδr sin θ = 0 . 3.3.24
This is equivalent to "conserving angular momentum in shells." If we evaluate equation (3.3.24) at the equilibrium position, we obtain an expression of first order accuracy, δω0 2δr = = −2ξ 0 e i σ t . 3.3.25 ω0 r0 Substitution of this expression into equation (3.3.22) yields
L
0
2δT3 = − ∫ 2e i σ t ξ 0 ω0 dL0 .
(
)
3.3.26
If, for simplicity, we further assume that the rotational velocity is a constant throughout the configuration. We obtain a very simple form of the variation of the rotational energy. 3.3.27 2δT3 = − 2e i σ t ξ 0 ω0 L0 , where Lo is the total angular momentum for the system. Thus, only the variation of the magnetic energy remains to be determined.
(
)
56
The Virial Theorem in Stellar Astrophysics
In order to determine the variation of the total magnetic energy it is necessary to establish a coordinate system appropriate to the geometry of the field and to the geometry of the configuration. Although the configuration is spherically symmetric, the geometry of the magnetic field present is not known. Thus, we shall consider the variations in Cartesian coordinates and later reduce our result to a form which is compatible with our previous results. Now the total magnetic energy of the configuration is defined (in c.g.s. units). 2 M H dm(r ) . 3.3.28 M=∫ 0 8πρ Thus, denoting the Cartesian coordinates as x1, x2, and x3, the variational form of the magnetic energy is 1 M H • δH 1 M 2 δρ dm(r ) − H 2 dm(r ) , 3.3.29 δM = ρ 4π ∫0 8π ∫0 ρ or in Cartesian coordinates δM = 1 1 2 δρ ∫∫∫ ∑ H i δH i dx 1dx 2 dx 3 − 8π ∫∫∫ H ρ dx 1dx 2 dx 3 . 4π i 3.3.30
Although we have already obtained an expression for δρ/ρ in section 3, due to the introduction of Cartesian coordinates it is convenient to express this variation in terms of the variation of the coordinates ηi 3.5, namely 3 ∂η δρ 3.3.31 = −∑ i , ρ i =1 ∂x i Before we can evaluate the expression for the variation of the magnetic energy, we must first determine the variation of the magnetic field δHi. A rather lengthy argument3.6 shows we can express this in terms of the coordinate variations, so ∂η ∂η j . δH i = ∑ H j i − H i ∂x ∂x i j j If we substitute (3.3.32) and (3.3.31) into (3.3.29), we have ∂η j ∂ηi 1 1 2 δM = ∫∫∫ ∑∑ H i H j ∂x j dx 1dx 2 dx 3 − 4π ∫∫∫ ∑∑ H i ∂x j dx 1dx 2 dx 3 4π i j i j
∂η j 1 + H2 ∑ dx 1dx 2 dx 3 8π ∫∫∫ j ∂x j The second and third terms combine to yield ∂η j ∂ηi 1 1 2 δM = ∫∫∫ ∑∑ H i H j ∂x j dx 1dx 2 dx 3 + 8π ∫∫∫ H ∑ ∂x j dx 1dx 2 dx 3 . 4π j i j
3.3.32
. 3.3.33
3.3.34
57
The Virial Theorem in Stellar Astrophysics
Now, if we assume that ηi is only a function of xi then the sum on j in the first term collapses and the remaining terms result in ∂η j 1 2 ∂η i 2 3.3.35 δM = ∫∫∫ ∑ 2H i ∂x j − H ∂x i dx 1dx 2 dx 3 . i 8π At this point, it is appropriate to reintroduce the assumption concerning the nature of the variation ηi. It was earlier assumed that ξo was constant. The equivalent assumption for the ηi ' s is that 3.3.36 ηi = const xi . Substitution of this explicit variation into equation (3.3.35) yields δM = − const. 2 ∫∫∫ H dx 1dx 2 dx 3 . 8π 3.3.37
Now, since we wish to consider the same type of pulsation from the ηi ' s , as we have assumed in the earlier section, we require that η = δr . 3.3.38 or η ˆ 3.3.39 = ξ = ξ 0 e iσt r . r Making use of our definition for ηi ' s , equation (3.3.36), we have xi + x j + xk = ξ = ξ 0 eiσt . 3.3.40 const. r This relation can only be true if the pulsation in the three coordinates (ηi) are in phase and of equal amplitude, and if const. = ξ 0 e i σ t . 3.3.41 Now, using the definition for the mass in a given volume in Cartesian coordinates and the value for the constant in equation (3.3.37), we may rewrite the variation of the magnetic energy as follows: 1 M H2 3.3.42 δM = −ξ 0 e i σ t ∫ 8π 0 ρ dm(r ) . Making use of equation (3.3.28) we may rewrite the variation in terms of the total magnetic energy of the equilibrium configuration δM = −ξ 0 e i σ t M0 . 3.3.43 Thus, we have obtained an expression for the last variation required to evaluate the variational form of the virial theorem (equations 3.3.7). We may, therefore, substitute equations (3.3.16), (3.3.27), and (3.3.43), into equations (3.3.7), and obtain
58
and L0. the introduction of magnetic fields only serves to reduce σ2 and thereby lengthen the period of pulsation. the work required to obtain this expression is nontrivial and in order for it to be useful one must have a detailed model in mind.44).3.47 Thus.43) will be quite adequate and is much easier to handle. equation (3. the first term in the numerator of equation (3.18) may be derived. I0 3. Also.The Virial Theorem in Stellar Astrophysics d2 (2e iσ t ξ 0 I 0 ) = −2e iσ t ξ 0 ω0 L0 + 3(γ − 1)(ξ 0 e iσ t )(ω0 L0 + Ω 0 + M0 ) − ξ 0 e iσ t Ω 0 − ξ 0 e iσ t M0 . Let us briefly investigate the effect upon σ2. one should not feel that they are all of paramount importance. 3.45) becomes identical to the previously derived equation (3. 59 .3. Since. Mo. Now. It should not be surprising to find the magnetic energy entering in an additive manner to the gravitational energy.3. since γ > 4/3 and the gravitational potential energy is defined as being negative. ω0. these assumptions may be omitted and an integral expression similar to equation (3.2. and hence on the pulsational period of the presence of magnetic and rotational energy.3.3.3.19). If ω0 is nonzero. However. Both are potential energies and since the basic equations are scalar in nature.46 an increase in σ indicates a decrease in the period and viceversa.45). One must also know the detailed geometry of the magnetic fields and of the star in order to evaluate the integrals that result. When γ = 5/3. to study the behavior of σ2 . If we let ηo and ω0 be zero. the influence of rotation is similar to that of magnetic energy. then equation (3.44 2 dt Simplifying equation (3. However.45) will be positive only if Ω 0 > M0 . For purposes of studying the effects of changes in γ.2.3. If γ < 5/3. while Mo is zero. the rotational energy is kinetic in nature and hence would not enter into the final result in the same manner as the magnetic energy. the expression is that of Ledouxl (1945). If necessary. 3. the addition of rotational energy (ω0 L0) will tend to increase σ2 as long as γ > 5/3. we should expect the final result to merely 'modify' the gravitational energy.45 Although we have made some strict assumptions in deriving equation (3.3. However.45) contains many aspects which one may check for 'reasonableness'. 3. Ωo. a great deal of numerical work will be required. The assumption of constancy of ξo and γ were only made so that the resultant integrals could be integrated in terms of the original parameters.3.3. we find that 1 2 σ2 = − (3γ − 4)(Ω 0 + M0 ) + (5 − 3γ )ω 0 L0 . Equation (3. the introduction of rotation has no effect on the period of pulsation. T = 2π / σ . Letting only ω0 be zero we obtain an expression identical to one arrived at by Chandrasekhar and Limber5 (1954).
3 8π S dt S S where Po and Ho are the gas pressure and magnetic field present at the surface ro.4.4.1) with the position vector r then integrating over all space enclosed by the bounding surface.4. where.2 ∫ ∫ ∫ ∫ V As in section 3. For this reason. Let us begin by sketching the origin of the virial theorem as rigorously presented by Chandrasekhar6.5 d(δS) = 2r0 δr0 sin θ dθdφ = 2ξr02 sin θ dθdφ = 2ξdS .3) that will interest us as hopefully the remaining terms are by now familiar.4. Formally they may be ignored by taking the bounding surface of the configuration to be at infinity. we will assume the star is nearly spherical and the pulsations are radial. ξ = δr/r.4. this generally proves to be inconvenient for stars as they usually have a reasonably welldefined surface or boundary. 3. If the magnetic field is strong this will clearly not be the case and the full tensor virial theorem must be used. The equations of motion for a gas with zero resistivity are du 1 (∇ × H ) × H . However.The Virial Theorem in Stellar Astrophysics 4. Although this may be true for simple fields in stars. as in section 2.13) and (3. we noted earlier that the use of the divergence theorem yields some surface integrals which are generally ignored.4. 3. let us consider the way in which these surface terms affect the variational formalism of the previous section.2. In Chapter III. these surface contributions should be included.1 ρ = −∇ρ + ρ∇Φ + dt 4π Employing the identity (∇ × H ) × H = (H • ∇)H − ∇(H • H) / 2 and taking the scalar product of equation (3. They are usually wished away by assuming they are small compared to the total magnetic energy arising from the volume integration. we already have shown that for adiabatic pulsations 60 . To facilitate the calculations.4 S S S S For radial variations only 3.2.4. δ ∫ P0 r0 • dS = ∫ δP0 r0 • dS + ∫ P0 δr0 • dS + ∫ P0 r0 • d (δS) . 3. section 2 [equations (N3. Variational Form of the Surface Terms In deriving the virial theorem. However. Consider first the effect of a pulsation on the surface term arising from the pressure by taking the variations of the surface pressure integral. 3. in reality.14)]. it is unlikely to be true for other gaseous configurations such as flares and in any event a numerical estimate of their importance is far more reassuring than an intuitive feeling. we get du 1 1 2 ∫ ρr • dt dV = −V r • ∇PdV + V ρr • ∇ΦdV + 4π V r • (H • ∇)HdV − 8π V r • ∇(H )dV . It is the behavior of the three integrals in equation (3. the simplicity generated by the use of the scalar virial theorem justifies the approach for purposes of illustration. this becomes 2 1 1 2 1 d I − 2T = 3(Γ − 1)U + Ω + M − ∫ P0 r • dS − 2 2 ∫ H 0 r • dS + 4π ∫ (r0 • H 0 )(H 0 • dS) . For stars possessing general magnetic fields which extend beyond the surface.
4. P ρ dr Combining equation (3. 3.4.4. Here.4.12 .4. 3. equation (3. becomes (3γ − 4)(Ω 0 + M0 ) − (5 − 3γ )ω 0 L0 + (3γ − 1)Q P − Q m . this results because an unstable system will have to do work against the surface pressures either in expanding or contracting the surface.7 dr r 0 S S S 0 Earlier we assumed that ξ was constant throughout the star and hence its derivative vanished.4.8 S S Now consider the variation of the two magnetic integrals in equation (3.45)].4. δQ P = 3( γ − 1)ξQ P If we assume a linear or homologous pulsation.The Virial Theorem in Stellar Astrophysics dξ δP δρ 3.10 Using this result. the contribution of the surface pressure term is such as to increase σ2 and thereby improve the stability of the system. This energy is thus not available to feed the instability. we only require the derivative to vanish at the surface in order to simplify equation (3.4.4.11 8π S S where Qm stands for the original magnetic surface term that appears in equation (3.3).5) and equation (3. The situation is not as obvious for the σ2 = − 61 . we get dξ r • dS .3. equation (3.9 This truly horrendous expression does indeed simplify3. the variation of the surface terms can be represented as δQ m = −ξQ m 3. 3. 1 1 1 1 2 δ ∫ 2(r0 • H 0 )(H 0 • dS) − ∫ H 0 r0 • dS = ∫ 2(δr0 • H 0 )(H 0 • dS) + ∫ 2(r0 • δH 0 )(H 0 • dS) 8π S 8π S 8π S 8π S + 1 1 1 ∫ 2(r0 • H 0 )(δH 0 • dS) + 8π ∫ 2(r0 • H 0 )(H 0 • dδS) + 8π ∫ 2(H 0 • δH 0 )(r0 • dS) 8π S S S − 1 1 2 2 ∫ H 0 δr0 • dS − 8π ∫ H 0r0 • dδS 8π S S 3.4. then the expression for the pulsational frequency [equation (3. 3.4).4.13 I0 Since γ > 4/3.7) to get δ ∫ P0 r0 • dS = 3( γ − 1)ξ ∫ P0 r0 dS . Thus.4.5) and the definition of ξ.4.9) becomes: ξ 2 δQ m = − ∫ 2(r0 • H 0 )(H 0 • dS − ∫ H 0 r0 • dS . δ ∫ P0 r0 • dS = 3( γ − 1) ∫ ξP0 r0 • dS − γ ∫ r0 3.6) with equation (3.4.3). Basically. as was done in the previous two sections.7 by using δH = −2ξH 0 .4.6 =γ = − γ 3ξ + r .4.
5. the average of cos2β 2 weighted by H 0 over the surface will determine the sign of Qm. since Qm is the difference between two positive quantities.13) to give good quantitative results.15 Qm < ∫ H 0 r0 dS = 12 R 0 H 0 .4. However. However. unless Qm and QP are comparable the coupling between the two will be weak and we may expect equation (3.The Virial Theorem in Stellar Astrophysics magnetic contribution Qm . whereas more radial fields enhance the instability. If Qm is an important term in equation (3.4. Unless the system is rather bizarre. Since a magnetic field cannot exhibit spherical symmetry. for purposes of simplicity. this term can be of the same order of magnitude as the internal magnetic field energy and must be included. more nearly radial fields will tend to feed the perturbation leading to a decrease in stability. we have assumed no coupling between the gas pressures and magnetic pressures. it is appropriate to review the use of the word stability itself. One thing becomes immediately clear from this discussion. whether or not the local contribution to Qm is positive or negative depends on whether or not the local value of β is greater or less than π/4. This assumption was made merely for the sake of simplicity and doesn't affect the illustrative aspects of the effects.4. the gas will be locally relaxed on a time scale less than the pulsation period and hence the two cannot be treated independently. Thus. fields exhibiting a local angle to the radius vector greater than π/4 tend to stabilize the object. Since a positive value of Qm increases the value of σ2. 8π S S [ ] where β is the local angle between the field and the radius vector. 62 . Under these conditions Qm becomes 1 1 2 2 ˆ ˆ 2 3. This simply results from the fact that a radial motion will tend to compress fields more nearly tangential to the motion than 45o. thereby removing energy from the motion.14 Qm = ∫ 2(H 0 • r) − 1 H 0 rdS = 8π ∫ cos β H 0 rdS . In any event. the result depends entirely on the geometry of the field.4. it is clear that 1 2 3 2 3. The effect of the field geometry can be made somewhat clearer by considering a spherical star so that the radius vector is parallel to the surface normal. radial pulsation will not occur. This concept deserves some amplification as it represents one of the most productive applications of the virial theorem. 8π S It is worth noting that in the case where the magnetic field increases slowly with depth. before embarking on a detailed development of the virial theorem for this purpose. In this case the tensor virial theorem must be used and the field geometry known. the departures from symmetry will yield a variable "restoring force" over the surface inferring that nonradial modes will be excited. The Virial Theorem and Stability In the last section. Conversely.13). Furthermore. Lastly. Thus. I alluded to the effects that the surface terms have on the stability of the system being considered.
Here the concept of the generalized force may be most simply stated as ∂r Q i = ∑ Fj • . there are many dynamical situations. ∂q i q i = q i ( 0) Now. The normal definition of equilibrium requires that the sum of all forces acting on the system is zero. This concept may be broadened to dynamical systems if one requires that the generalized forces (Qi) acting on the systems are zero. However.5. If the potential extremum implied by equation (3.The Virial Theorem in Stellar Astrophysics When inquiring into the meaning of the word. in terms of this definition of equilibrium we may proceed to a definition of stable equilibrium. which are not equilibrium situations that even the most skeptical person would call stable. 3. Since stars are not really equilibrium configurations. saying the generalized forces must vanish is equivalent to saying the potential energy must be in extremum." This definition is subject to several interpretations and serves to illustrate the danger of consulting an English dictionary to learn the meaning of a technical term. ∂q 2 i q i = q i ( 0) 63 . to develop forces or moments which tend to restore the body to its original condition. we see that we must extend our conceptualization of stability to include some dynamical systems.5.1 ∂q i j where Fj represents the physical forces of the system acting on the jth particle and the qi's represent any set of linearly independent 'coordinates' adequate to describe the system.4 Q i = − > 0.5. ∂Φ 3. This approach provides the following definition: Stability: "That property of a body which causes it. but rather steady state configurations. The word stability is usually associated with the word equilibrium. but certainly most of the main sequence stars are. This is primarily because the concept of stability normally is first encountered during the study of statistics. when disturbed from a condition of equilibrium or steady motion. One of the most obvious examples to an astronomer.2 Q i = −∑ ∇ j Φ • = ∑ ∂q dq i ∂q i j j ∂r j i Thus. In a conservative system all the forces are derivable from a potential Φ .Thus. are the stars themselves. Not all stars would be regarded as stable.5.3) is a minimum. The conditions thus imposed on the potential are ∂ 2Φ 3. 3. it is customary to consult a dictionary. the generalized forces may be written as ∂Φ dr j ∂r r j • j = − ∂Φ .5. then the equilibrium is said to be stable.3 Q i = − = 0.
In Chapter I.4. Thus. any disturbances from equilibrium will produce an increase in the potential energy.5 One could say with some confidence that if d2I/dt2 > 0 for all t the system would have to have at least one particle whose position coordinates increased without bound. Jacobi employed the 64 . However. he must also decide on the applicability of the analysis to the system. the determination of a system state of stability seems to have inspired Jacobi to develop the nbody representation of Lagrange's identity from which it is a short step to the virial theorem. we arrived at simple statements of Lagrange's identity for selfgravitating systems as: 1 2 d2I dt 2 = 2T + Ω . equation (1. since both T and Ω vary with time it would be difficult to say something a priori about d2I/dt2 from Lagrange's identity alone. and overstable. Thus. It is a common practice to examine the response of the system to a continuous spectrum of perturbations. and is actually the basis for most stability criterion. Thus it is not uncommon to find modifying adjectives or compound forms of the word "stable" appearing in the literature. Suppose a system is disturbed from equilibrium by an increase in the total energy dE above the total energy at equilibrium. consider the following argument. [i. Analysis of this type is called linear stability theory. when analyzing a system not only must one correctly carry out the stability analysis. bistable. That is to say. it has been quite fashionable to use the virial theorem as the vehicle to carry out linearized normal mode analysis of systems in order to determine their state of stability. one is rarely able to calculate the response in general. the system would be unstable.12)]. then the velocities may increase without bound. Since the conservation of energy will apply to the system after the incremental energy dE has been applied. It is usually necessary to linearize the equations describing the system in order to solve them.e. global stability.5. the departure from equilibrium brings about a decrease in the potential energy. let us summarize some of Jacobi's arguments. If any of these perturbations grow without bound the system is said to be unstable. quasistable. Thus. 3. however. This implies that the velocities will decrease for all particles and eventually become zero. However.The Virial Theorem in Stellar Astrophysics In order to see that this definition of stability is consistent with our dictionary definition. Recently. If Φ is a minimum. This would seem in full accord with our dictionary definition and thereby wholly satisfying. The introduction of these modifiers as often as not arises from the mode of analysis used to describe the system. To see now how closely tied the virial theorem is connected to stability. the kinetic energy must decrease. These terms are usually used without definition in the hope that the reader will be able to discern the correct meaning from the context. A few common examples are: Secular stability. However. it is still inadequate to serve the needs of mathematical physicists in describing the behavior of systems of particles. We would certainly call such motion unstable motion. If. simple and clearcut as this definition of stability may seem. the motion of the system will be bounded (Note: the bound may be arbitrarily large). Unfortunately.
It is the constancy of E with time that makes this a valuable criterion for stability and is the reason Jacobi used it. we discussed only the meaning of the positive values. In section 2 [i. to modify equation (3. d2I/dt2 > 0 and the system is unstable.8) is sufficient for instability. It is not uncommon to find the statement in the literature that (2T + Q > 0) insures the instability of a system citing the virial theorem as the justification.12)].5) to give: 2 1 d I = 2E − Ω > 2E . it is Lagrange's identity that is the relevant expression and it only guarantees that at the moment the system is acceleratively expanding. let us turn the applications of the variational form of the virial theorem to stability.5. if E > 0.7 = 2T + Ω . E = T + Ω = const. In Chapter 2 we discussed the extent to which time averages of quantities may be identified within their phase averages. Furthermore. 3. keeping in mind that the variational approach is essentially a first order or linearized analysis.5.8 t0 constitutes a stability criterion for Ergodic systems.5. as negative squares of frequencies had no apparent physical meaning. If the system is to be always stable. However. The majority of Chapter III has been devoted to obtaining expressions for the frequency of a pulsating system. the time averages on the righthand side of (3.7) will tend to phase averages if the system is Ergodic. 2 2 dt 3. and the fact that for selfgravitating systems Ω > 0.2. through use of the Ergodic theorem. This is known as Jacobi's stability criterion and provides a sufficient (but not necessary) condition for a system to be called unstable. we get 1 dI 0 3. equations (3.6 So. Let us look again at the nature of the assumed pulsation in order to further investigate the meaning of these pulsation frequencies. That is.5.5. equation (3.8) must be satisfied in a stable Ergodic system and failure of equation (3. Actually.. then the limit of the lefthand side must tend to zero.). In the earlier sections.5. Thus 2<T>+ <Q> = 0.5. What is really meant is that if (T + Q > 0) the system is indeed unstable as this is just a statement of Jacobi's stability criterion concerning the total energy of the system. t 0 dt 0 If the system is to remain bounded.e.5) over some time t . Now. these expressions could be neither positive nor negative. There is another approach to the problem of temporal variation which was not available to Jacobi.5.e.The Virial Theorem in Stellar Astrophysics constancy of the total energy (i. 3. We obtained a value for the square of the frequency in terms of the equilibrium energies of the configuration. Integrating equation (3.5.9 r 65 . we assumed that the pulsation would be periodic and of the form δr = ξ0eiσt . dI/dt must always be finite.
resulting from small but inevitable. when (3γ − 4)Ω 0 − <0 . the pulsation becomes exponential in nature.14 . it can be made into a local condition by taking an infinitesimal volume and including the surface terms discussed in section 4. if there exists even one mode with σ2 < 0. the instability associated with that mode will grow without bound. we see that the sphere will become unstable. equation (3.12 It is also worth noting that this criterion applies to the entire system.5 . If the sign of σ is negative. where to is a real number.10 δr = ξ 0 e ± 2πt / t0 .5. All classical equations of dynamical symmetry exhibit fulltime symmetry. In section 2. departures from perfection produced by statistical fluctuations. r 3. 3. applying the instability criterion equation (3. One might be tempted to choose the negative sign of equation (3. Thus. let us see what implications this analysis has for the stability of stars. even though σ2 < 0. Now.11) will be positive and the pulsation will grow without bound with a rate of growth determined by to.The Virial Theorem in Stellar Astrophysics where σ was the frequency of pulsation and ξo did not depend on time. we may write σ = ±2πi / t 0 .19)].12). Therefore. Now.5. A specific solution would be fully determined by the boundary conditions at t = 0. Now. while the gravitational potential energy is intrinsically negative. γ < 43 66 . Combining equation (3. we established an expression for the pulsation frequency of a gravitating gas sphere [equation (3. we assumed that a full continuum of perturbations are present.9).5.11) saying that the system is stable as the pulsation will die out. 3.2.5. we have 3.13 I0 Since the moment of inertia (I0) is intrinsically positive.11 Thus.10) and equation (3. and thus is a "global" stability condition. Further.5.5.5.5. thus those solutions which damp out in the future were unstable in the past and viceversa. this becomes a sufficient condition for the system to be unstable in the strictest sense of the word σ2 < 0.5. if we make the formal identification between σ and the period and take σ to be purely imaginary.13) becomes (3γ − 4) < 0 3. This would be wrong.5. However. then the sign of the exponential in equation (3.
and R = 2600 R⊙.5. but which has a magnetic field.5.22 H > 2 × 10 8 2 gauss . I0 3. let us turn to the more general formulae resulting from our analysis in section 3.20 5 R while the magnetic energy is R3 H 1 2 M = ∫∫∫ H dx 1dx 2 dx 3 = .5.14)].5.5.5.17 σ2 = I0 Substituting this into the instability criterion equation (3. 3. having extracted as much information as possible from the pulsation expression developed in section 2.18 Now. This is the familiar instability criterion demonstrated by Chandrasekhar in Stellar Structure. for a main sequence A star with M = 4M⊙ and R = 5R⊙.5. we have (3γ − 4)(Ω 0 + M0 ) < 0 . if we assume γ > 4/3. we can see that a necessary condition for stability of a homogeneous nonrotating gas sphere is γ > 4/3.21).5.16 Consider first a gas sphere which is not rotating. we have Hrms >3x107 gauss .12).5. a star will become unstable when γ is less than 4/3. 67 . 3.16) becomes then − (3γ − 4)(Ω 0 + M0 ) . 3. Equation (3. If we consider the other stability criterion [equation (3.23 3.21 6 8 Combining equation (3. 3. Remember the final expression for the pulsational frequency was σ2 = − (3γ − 4)(Ω 0 + M0 ) + (5 − 3γ )ω 0 L0 . R 2 where M and R are given in solar units.7 He further demonstrates that a gas with γ equal to 4/3 corresponds to a gas where the total pressure is entirely due to radiation. we have a sufficient condition for instability due to the presence of magnetic energy as follows: M0 > Ω 0 .5. for a star like VV Cephei with M = l00 M⊙.5.5. 3.24 However.The Virial Theorem in Stellar Astrophysics Thus.15 Thus. we have Hrms > 3000 gauss . we may obtain a crude estimate of the magnitude of the magnetic fields necessary to disrupt a star.20) and equation (3. we see that the root mean square value of the magnetic field sufficient to disrupt a uniformly dense sphere is M 2 3.5.5. 3. Thus. The gravitational potential energy for a sphere of uniform density is 3 GM 2 Ω=− . 3.5.19 In the following manner.
5. it would appear that VV Cephei is on the verge of being magnetically unstable. equation (3. 3. for an unusually large star. 3. One might argue that our crude estimates of Ω are so crude as to be meaningless due to the large central concentration of the mass in giant stars. However. It is interesting to note that the instability criterion equation (3. all this is not meant to imply that the rotational terms are unimportant. the condition may still not be satisfied because of the presence of the rotational term. Thus.18) permits the existence of a gas with γ < 4/3 providing the magnetic energy exceeds the gravitational energy. therefore.14) combined with the stability criterion equation (3. the required field becomes much smaller.26 ω0 L0 > (5 − 3γ ) Since Ωo is intrinsically negative. 3. an extremely large magnetic field would be sufficient to cause the star to become unstable.5. the stability criterion (3. even in the event that Ω 0 > M0 .The Virial Theorem in Stellar Astrophysics We may conclude from these arguments that for a main sequence star. we see that the stability condition will always be satisfied with any ωo. Thus.5.12) becomes (5 − 3γ )ω0 L0 > (3γ − 4)Ω 0 . However. of necessity.5. we may have some hope that our limiting field values are not too far from being realistic. 68 . it is evident that if 4/3 < γ < 5/3. This result is certainly not intuitive. In the case of VV Cephei. Thus. However. Ledouxl has shown that rotational velocities encountered in stars may lead to a variation in the pulsational period by as much as 20%. If we now consider a rotating configuration with no magnetic field.18) would require a necessary condition for the stability of any configuration where M > Ω that γ be less than 4/3. it should be pointed out that the magnetic field one can observe is.5. this condition may never be satisfied unless Ω 0 > M0 . This may be obtained by combining equation (3. then the presence of rotation will help stabilize stars. a surface field and.25 If we restrict γ to be less than 5/3 we have (3γ − 4)Ω 0 . However.5.5. Let us briefly consider the instability criterion when both magnetic and rotational energy are present. provides us with a lower limit on the magnetic energy. for all known stars the stability criterion for rotation is not particularly useful.15) and (3γ − 4)(Ω 0 + M0 ) − (5 − 3γ )ω0 L0 > 0 . it is also true that the physical meaning of a gas having a γ < 4/3 is a little obscure. However. A physical explanation of the result might be supplied by the following argument. Thus. Indeed.27 As before.5. Babcock has measured a field ranging from +2000 to 1200 gauss. Indeed.
the concept of radial pulsation becomes inconsistent. It would have been more correct to employ the integral form of the expressions for the frequency of pulsation. I would be remiss if I left the subject of the virial theorem and stability without some discussion of the recent questions raised with regard to the appropriateness of the approach. rotating at critical velocity. and. Also since the gravitational and magnetic energies must supply this energy to the rotation. and higher. It is appropriate at this point to make some comments regarding all of the stability criteria relating to the stability of radial pulsations. Chandrasekhar and Fermi3 have shown that a sphere under the influence of a strong dipole field will tend to be "flattened" in much the same way as it will be by rotation. these questions appear to be partly substantive and partly semantic and revolve largely around one of these modifiers mentioned earlier.3% before instability will again set in. 69 . We may now ask what sort of increase in the maximum magnetic field can this additional rotational "stability" supply. From our previous investigation with the rotational stability criteria we might expect the result to be small. Such destruction is usually so unambiguous that no complications arise in the use of the word unstable. However. If one considers a uniform model with γ = 3/2. It is hoped that the degree of differences would not be large. It must be remembered that the expressions developed for the pulsational frequencies were based on a first order theory as are the stability criteria developed in this section. However. either by the presence of a strong. the conditions at which one wishes to apply an instability criteria are generally such that the second. even though stability is increased by the presence of rotation. a certain amount of energy will be required to slow down or speed up the rotation in order to keep the angular momentum constant. magnetic field or rapid rotation. usually on a time scale related to the hydrodynamical time scale for the system. As mentioned before. So far. analysis of such systems would require the use of the tensor virial theorem and considerable insight into the types of perturbations to employ. namely. our discussion has been restricted to problems involving dynamical stability about which there seems to be little argument. the energy is no longer available to "feed" the pulsation and disrupt the star. There is one respect in which the differences between the derived criteria and the 'correct' ones may result in a difference in kind. the amount of this energy transferred from the magnetic and gravitational energies will depend on γ. order terms are not small and should not be neglected.The Virial Theorem in Stellar Astrophysics Consider a pulsating configuration containing both rotational and magnetic energy. since the rotational stability criteria did not supply us with as important a result as did the magnetic instability. A dynamically unstable system will disintegrate exponentially. since this is supplied by the potential energies present. the result one would obtain by using the integral expression and a specific model would only differ in degree from those derived here. secular stability. it is not increased a great deal. Such is not the case for the term secular stability. That is. ultimately must come from the gravitational and magnetic energies. Once the spherical symmetry has been destroyed. As the system expands or contracts. Thus. This energy must be supplied by the kinetic energy of the gas itself. Therefore. To me. we would expect the effects of rotation to be small compared with the magnetic energy. he will find the magnetic field may only be increased by about 0.
is used the arguments of Milne15 as presented in Chapter I still apply. Time scales for development of instabilities will be governed by the forces and hence may be very long. lO. In practice. except for the global constraint on the total angular momentum. He points out that difficulties arise in rotating systems resulting from the presence of the coriolis forces. One cannot hope to untangle in a few paragraphs a controversy which has taken more than a decade to develop and at a formal mathematical level is quite subtle. Indeed the presence of dissipative forces guarantees that this must happen. no constraints are placed on the transfer of energy from the rotational field to the thermal field. section 5. it is worth noting that recent14 statements which essentially say that the tensor virial approach to stability is wrong do nothing to clarify the situation. 12. However. Since globally the forces are conservative the first result is not surprising and since radially moving mass in a rotating frame must respond to the conservation of angular momentum. instabilities associated with these forces might exist which would otherwise go undetected. the majority of these papers clearly state that the authors are dealing with systems with zero viscosity and so the problem is not one of the accuracy of the analysis but rather of the applicability of the analysis to physical systems. their variation does not vanish and hence they will affect the pulsational analysis. The resulting stability analysis would then correctly reflect the presence of these forces and thus be dependent on their specific nature. In addition. This line of reasoning demonstrates a qualitative difference between the cases of uniform rotation and differential rotation in that in the former dissipative forces will be inactive and the analysis will be appropriate while in latter cases they must be explicitly included. However. the viscosity of the gas in most stars is so extremely low that the time scales for the development of instability arising from viscosity driven instabilities will be very long. which is technically the virial theorem. 13 by Ostriker and others. At this level assailing the virial theorem is as useful an enterprise as denying the validity of a conservation law. such as viscosity and thus must depend to some extent. Insofar as the time averaged form of Lagrange's identity. 70 . Now if dissipative forces are present such as viscosity then it may be possible to redistribute local angular momentum while conserving it globally so that no equilibrium configuration is ever reached. which lead to a clear distinction between dynamically and secularly stable systems. The presence of dissipative forces can be included in the equations of motion and thus in the resulting tensor representation of Lagrange's identity. This point is central to a lengthy series of papers9. neither is the second. The presence of velocity dependent forces does not affect the virial theorem unless those forces stop or destroy the system during the time over which the average is taken. the terms associated with the coriolis forces can be made to vanish by the proper choice of a coordinate frame and they would appear to play no role in the energy balance of the system. 11. which discuss the stability of a variety of differentially rotating systems. Thus. Perhaps one of the clearest contemporary discussions of the term is given by Hunter8 who notes that there is less than universal agreement on this meaning of the term. However. on the nature of those forces. As we saw in Chapter II.The Virial Theorem in Stellar Astrophysics The notion of secular stability involves the response of the system to small dissipative forces.
a complete variational analysis of perturbations to all independent spatial coordinates will allow one to compute the nonradial as well as radial modes of oscillation and thereby obtain a much more secure analysis of the system's stability. unless the field energies become quite large one would expect the pulsational frequencies not to differ greatly from the purely radial theory. First the interesting situations of marginal stability are liable to involve substantial magnetic fields or rapid rotation. As in other chapters. The tensor virial theorem as is presented in Chapter II.. Throughout the chapter we confined ourselves to spherically symmetric systems exhibiting radial pulsations only. However. 71 .e. the departure from spherical systems of the mass distribution will also be large invalidating every aspect of the analysis. In spite of many distractions dealing with the variation of magnetic fields. the influence of these added features on the pulsation frequency and hence stability became clear. This is not surprising as both results have as the derivational origin the same concept (i. However. In principle. the effect of a surface field proves to be more complex.The Virial Theorem in Stellar Astrophysics 6. In section 3 we expanded the variational approach to include the effects of magnetic fields and rotation. If these aspherical properties are large. etc. In addition. As mentioned. we began with the simple and moved to the more complex. Fortunately the techniques for dealing with these problems exist and have been developed here as well as the literature. Rotation can either enhance or reduce the stability of a configuration depending on whether or not the value of γ for the gas permits net energy to be fed to the pulsation.2.. leads immediately to the Jean's stability criterion. Summary In this chapter we have explored the results of applying a specific analytical technique to the virial theorem. Having discussed the implications of the variational approach to the virial theorem we moved to develop the explicit form for the simple scalar theorem appropriate for selfgravitating systems. Here the result depends critically on the geometry of the field.19)] originally due to Ledoux. We recreated the pulsational formula [equation (3. this is inappropriate when considering either rotation or magnetic fields as neither can exhibit spherical symmetry and thus one would expect nonradial oscillation to be excited. However. for the stability analysis to be valid all possible modes of perturbation must be included. This line of reasoning becomes particularly dangerous when one turns to a discussion of stability. the equations of motion). section 1 allows one to follow perturbations in independent spatial coordinates. it is reassuring when a different approach yields results already well accepted. In the last section we dealt briefly with the overall question of stability and showed explicitly how the virial theorem provided an excellent basis for a linear stability analysis of a symmetric system. Limiting oneself to only the radial modes is to invite a misleading result. One implication of this result is that the fundamental mode of oscillation depends only on the square root of the density and when coupled with the stability criterion in section 5. The influence of an internal magnetic field is to destabilize the star for all realistic values of γ.
2.4 0 r0 3. N3.2. N3. the variation of T1 1 2 ∫ R 0 dr 4πr ρ dr = dt 2 2 1 2 ∫ M 0 dr dm(r ) .2.13). 2 N3.2.3 dm(r ) r Evaluating the above expression at r = ro and using the result in equation (3. dT2 = 3 NkTdV . dt 2 N3.The Virial Theorem in Stellar Astrophysics Notes to Chapter 3 Remember the defining expression for the gravitational potential energy is M m ( r )dm ( r ) . N3.11) to first order accuracy we get Ω 0 δr δΩ = − ∫ dΩ .1 Ω = −G ∫ 0 r We may again use the fact that the variation of m(r) and dm(r) are both zero.1) that dΩ Gm(r ) =− .2 We may write the total energy as the sum of two energies Tl and T2. N3. Thus.1.2 dm(r ) . Since the equilibrium point r0 cannot vary with time by definition.2.3 r = ro + δr .1.6).2) becomes δT1 = 1 2 ∫ M 0 δT1 = 1 2 ∫ M 0 d(δr ) dm(r ) .1. and equation (3. equation (N3. to obtain M δr δΩ = −G ∫ 2 m(r )dm(r ) . Now the total kinetic energy of mass motion is given by T1 = Thus. dt 2 N3.2.4 The largest term in the integral of equation (N3.2. to first order we have δT = δT1 + δT2 ≅ δT2 . dt dt However.1 dr dδr N3.2.2.4) is second order in δr and may be neglected with respect to the first order terms of equation (3.1.2. where T1 is the kinetic energy due to the mass motion of the gas arising from the pulsations themselves.6 72 .2. N3. from the definition of δr we see that N3.5 Now consider the kinetic energy of a small volume of an ideal gas.1 3.1.2 0 r which can be written as an energy integral by noting from equation (N3.
The Virial Theorem in Stellar Astrophysics However.2.7 2 ρ Thus. From the definition of ξ (equation 3.9). again evaluating at the equilibrium position. Pg dT2 = 3 dm(r ) .2. we have r' r' 2 2 2δrdr0 ∫0 r0 δρdr0 = − ∫0 ρ 0 r0 r0 + d(δr) .2.2.2. we have 2 P ρδP − Pδρ = δ = ρ ρ2 (δP P)ρ − δρ = P . we have M Pg N3.10 =γ . ρ 0 We shall now assume that the pulsations are adiabatic so that δP δρ N3.2. we shall invoke the following argument.9 2δT = 2δT2 = 3∫ δ dm(r ) . twice the total kinetic energy of the gas sphere arising from thermal sources is M Pg 2T2 = 3∫ dm(r ) . ρ0 ρ0 N3. N3. and remembering that the variation of dm(r) is zero. Therefore.2 N3. dξ = = − 2 .2. P ρ where γ is the ratio of specific heats cp /cv .2. From the conservation of mass.3.8 0 ρ Now neglecting radiation pressure.12 3. 0 0 0 0 r' r' r' r' N3.1 Rewriting equation (3. so that the total pressure is equal to the gas pressure.3 73 .3. N3.12). Now N3. we have δm(r ' ) = 0 = ∫ 4πr 2 ρdr = ∫ 4π(2rδr )ρdr + ∫ 4πr 2 δρdr + ∫ 4πr 2 ρd (δr ) . 2 dr0 r0 r0 r0 N3. substituting into equation (N3. ρ [( γ − 1)δρ 0 ] 2δT ≅ 3∫ M 0 P0 δρ ( γ − 1) dm(r ) .3.11 ρ2 P Therefore.2.28) and evaluating at r = r0 . we have dξ r0 d (δr ) / dr − dr d (δr ) δrdr0 or. And keeping only terms up to first order.3 In out attempt to obtain an expression for δρ/ρand thereby determining the variation energy. the gas pressure is given by Pg = NkT and dm(r) = ρdV.
6 M P ξ iσt 0 0 + 9e ∫ ( γ − 1)dm(r0 ) 0 ρ0 74 . With these conditions. N3.The Virial Theorem in Stellar Astrophysics Eliminating d(δr) from equation (N3. hydrostatic equilibrium) requires that dP0 Gm(r0 )ρ =− . the integrated part of equation (N3.4 dr0 r02 Therefore. 4πGr02 m(r0 )ρ Gm(r0 ) dm(r0 ) dΩ 0 3 dP0 .4.2) with the aid of equation (N3.4.4.e.4.5 = − 3ξ + r0 ρ0 dr0 In an attempt to simplify equation (3. This can only be true if the integrands equal.5) to simplify the second integral in equation (N3.3.3).15)].4.1 3∫ 0 ( γ − 1)r0 0 e i σ t dm(r ) = 3∫ 0 ( γ − 1)r0 0 e i σ t 4πr02 dr0 .4 ∫0 r0 δρdr0 = − ∫0 ρ 0 r0 3 + r0 dr0 dr0 .2) vanishes and the remaining integral becomes: R R R R dP0 3 d 3 2 3 ∫0 ξ 0 dr P0 (γ − 1)r0 dr0 = ∫0 P0 ξ 0 r0 + ∫0 (γ − 1)ξ 0 dr0 r0 dr0 + 3∫0 P0 (γ − 1)ξ 0 r0 dr0 . δρ dξ . N3. we have again to the first r' r' dξ 2 2 N3.2. Equation (N3.3. N3. The first of these conditions is the familiar condition of stellar structure which essentially defines the surface of the gas sphere. we obtain the following expression for the variation of the kinetic energy [i. M P R P dξ dξ N3.4 [ ] At this point we shall impose the boundary conditions that Po → 0 at ro = R and ξo → 0 at ro = 0.3. M P ξ M P ξ r dγ Ω0 2δT ≅ −9e i σ t ∫ 0 0 ( γ − 1)dm(r0 ) + 3e i σ t ∫ 0 0 0 dm(r0 ) + 3e i σ t ∫ ξ 0 ( γ − 1)dΩ 0 0 0 0 ρ 0 dr0 ρ0 . The second condition is required by assuming that the radial pulsation is a continuous function.4. N3.3.4.4.2.3).2 0 0 dr ρ 0 dr 0 0 3. 0 ρ 0 ρ dr0 dr0 0 0 Integrating the right hand side by parts we obtain R 0 dξ P R R d 12πe i σ t ∫ 0 0 (γ − 1)r03 dr0 = 12πe i σ t P0 (γ − 1)r03 ξ 0 − 12πe i σ t ∫ ξ 0 P0 (γ − 1)r03 dr0 .4.5 4πr0 =− =− = dr0 r0 dr0 dr0 r0 [ ] Making use of the second form of equation (N3. equation (3.3). and the definition of dm(r) to simplify the other two integrals in equation (N3.4. N3.3 Conservation of momentum (i. Thus..e.3.15). N3. let us consider the second integral.4) must hold for all values of r'.
Thus as might be expected from the orthogonality of the xi's we have dηi ∂ηi . we finally obtain 3 ∂η δρ = −∑ i .4.The Virial Theorem in Stellar Astrophysics The above expression is obtained by making use of the previously mentioned definition and substituting it into equation (N3. we obtain δm(V) = ∫∫∫ δρdx 1dx 2 dx 3 + ∫∫∫ ρd (δx 1 )dx 2 dx 3 + ∫∫∫ ρdx 1d(δx 2 )dx 3 ∫∫∫ ρdx 1dx 2 d(δx 3 ) . so is the first term. N3.5 However. dx i dx i j=1 dx j ∂x j N3.6 = dx i ∂x i Substitution of this into equation (N3. N3.5 Let us define N3.4).7 0 0 ρ 0 dr0 3.5. as it did in section 2.2 The conservation of mass requires.5. Since the first and last integrals are identical except for the difference in sign.5.5.2).5.5.5.5)] zero if i= j but. then into equation (N3. dx i N3. Now the definition of the mass within a given volume in Cartesian coordinates becomes m(V) = ∫∫∫ ρdx 1dx 2 dx 3 . N3.1 ηi ≡ δx i .4.4 Since δxi = ηi.15).7 ∫∫∫ i =1 ∂x i Equation (N3.5. that the variation of the mass is zero.8 75 . yields 3 ∂η δρdx 1dx 2 dx 3 = − ∫∫∫ ρ∑ i dx 1dx 2 dx 3 .2. ρ i =1 ∂x i N3. This can only be true if the integrands themselves are equal. and finally into equation (3. N3. they vanish from the expression and M P ξ r dγ Ω0 2δT ≅ 3e i σ t ∫ 0 0 0 dm(r0 ) + 3e i σ t ∫ ξ 0 ( γ − 1)dΩ 0 . we have by the chain rule 3 d(δx i ) dηi dx ∂ηi = =∑ i .2).5.5.3 Rewriting the last three integrals. since the xi's are linearly independent and the ηi's are just the variation of these coordinates.5. not only is the second term in the product [under the summation sign of equation (N3. Thus. Thus taking the variations of equation (N3.4. N3.7) must hold for any volume of integration. we have 3 ∫∫∫ δρdx 1dx 2 dx 3 = − ∫∫∫ ρ∑ i =1 d (δx i ) dx 1dx 2 dx 3 .5.3).
and equation (N3.6.9). we know that ∂H N3. Maxwell's equations for an infinitely conducting medium require that ∂ ∇ × (∆E) = − ∆H . we have ∂ (∆H ) − = ∇ × (−u × H ) . ∂t and ∂ d = − v•∇.The Virial Theorem in Stellar Astrophysics Suppose a displacement n takes place with the slow continuous movement so that dn .6.6. N3. dt dt j However.6.6 76 . we have dx ∆H = δH − ∑ i dt •∇H . i ∂x Combining equations (N3.9 =∑ dt i ∂x i dt while definition of the space variation δH is ∂H N3. 3.11 dt i Noting that the variation of a linearly independent quantity may be interpreted as the total differential of the quantity ∂x dx i = ∑ i dx j = δx i .6.1 u= dt Then.4 ∂t The integral form of equation (N3.12 i ∂x j we obtain ∆H = δH − n • ∇H . N3. N3.6.6.6.8).13 3. N3.6. N3.6. N3.6. d ∂ dx i .6.l0).10 δH = ∑ dt • ∇H . N3.6. from the definition of the time and space variations and the equation relating to the total time derivatives. N3. if the electrical conductivity of the medium is infinite.6.2) and equation (N3.3 ∂t Combining equation (N3. the timevariation of the electric field is ∆E = −u × H . N3.3).2 However.6.4) is just ∆H = ∇ × (n × H ) .6.6 ∆H = dt .6.6.7 ∂t dt which results in dx j dH N3.5 But. (N3.6.8 ∆H = −∑ • ∇H dt .
N3.3 ∂x i i So.l4) yields δH = (H • ∇)n − H (∇ • n) − (n • ∇)H + (n • ∇)H . remembering that ξ0 is constant.16 The last two terms are identical except for opposite signs and thus cancel out.2 we went through an extensive argument (see Note 3.7.7.The Virial Theorem in Stellar Astrophysics Combining equation (N3. the variation of the magnetic field has taken the particularly simple form δH = −2ξH 0 .6. we finally obtain an equation for the variation of the magnetic field (δH).17 δH i = ∑ H j i − H i ∂x ∂x i j j 3.l6).7 In the section 3. Combining this expression for the curl of a crossproduct with equation (N3.14 Now the curl of a crossproduct may be written as ∇ × (n × H ) = (H • ∇)n − H (∇ • n) − (n • ∇)H + n(∇ • H ) . ∂ (ξx i ) (H • ∇)δr = ∑ H i = ξH .6. N3. Then.15 Since the divergence of H is always zero.6. equation N3.S).6. N3.6. δH = ∇ × (n × H ) − (n • ∇)H . The remaining expression may be written in component form as follows: ∂η ∂η j .6.2 ∂x i i The second term can be evaluated the same way.l3) and equation (N3. N3. The easiest way to evaluate the first term is in Cartesian coordinates.6.6.1 where δr plays the role of η in that discussion. N3.4 77 . N3.7. N3. so that ∂ (ξx i ) H (∇ • δr ) = H j ∑ = 3ξH .6.7. N3. the last term vanishes. to show that δH = (H • ∇) δr − H (∇ • δr ) .
M. 137. p. Ap. J. 102. (1939). P. 613. S. 987. Ledoux. pp. ll6. J. C. and Limber. Ap. J. Chandrasekhar. and Ostriker. J. (1925). E. J. Ostriker. (1957). J. and Tassoul. Berlin. E. p. L. London. J. and Walraven. B. p. S. Milne. F. L49. (1973). Vol.The Virial Theorem in Stellar Astrophysics References 1 2 3 4 5 6 7. S. Ap. 180. 50. 213.N. 217. 78 . 143. Ostriker. Heide1berg. J. Friedman. (1945). (1961). p. Berlin. L. 53. (1968). L. 293. Lett. E. 154. J. J. 431592. R. 155. Tassoul. (1958). Dover Pub. Oxford University Press. (1977). J. P. Bardeen. pp. J. (1967).S. 605687. Ap. Ap. J. p. M. SpringerVer1sg. J. ll8. J. 8 9 10 II 12 13 14 15 Ledoux.497.. LydenBell. (1969). Vol.. pp. and Sarkin. Ap. 10. Schutz. Ap. 6. Heidelberg. and Bodenheimer (1973). 51. Chandrasekhar. p. 186. p. Chandrasekhar. P. (1977). ll85. Handbuch der Physik.R. p. Ap. 409419. Phil. Ap. D. S. P. SpringerVer1ag. Ap. p. Inc.A. (1963). Gottingen. J. D. An Introduction to the Study of Stellar Structure. J. Ostriker. 171. (1959). 467. S. A. ll9. J. (1958). 136. J.. N. p. and Fermi. P. and Ostriker. Mag. J. P. Handbook der Physik. Hunter. Gottingen. Hydrodynamic and Hydromagnetic Stability. p. (1953). p. P. and Peebles. P. Chandrasekhar.
The Virial Theorem in Stellar Astrophysics 79 .
notably normal stars. while leading to nearly the correct numerical result. Many of the objects were clearly so condensed as to require the application of the General Theory of Relativity or some other gravitational theory for their description. is conceptually wrong. the type of objects which were explicitly discussed. I cannot resist discussing some objects about which even less is known. neutron stars. like supermassive stars.The Virial Theorem in Stellar Astrophysics Copyright 2003 IV Some Applications of the Virial Theorem 1. I hope the reader has been impressed by the wide range of problems which can be dealt with by the virial theorem. however. and black holes became 'household' words in the literature of astrophysics. During the 1960's advances in observational astronomy presented problems requiring theoreticians to postulate the existence of a wide range of objects previously considered only of academic interest. Pulsational Stability of White Dwarfs By now. Since general relativistic effects can usually be viewed as an effective increase in the gravitational force. We shall see that at least one commonly held tenant of stellar structure. When one postulates the existence of a "new" object it is always wise to subject that object to a stability analysis. one would expect its presence to decrease the stability of objects in which it is important. In order to illustrate the power of this remarkable theorem. This is particularly important for highly collapsed objects as the time scale for development of the instability will be very short. 80 . What came as a surprise is the importance of these effects where one would normally presume them to be of little or no importance. Some of the problems mentioned in the last chapter indicate the type of insight which can be achieved through use of virial theorem. are currently judged by the naive to be well understood. These terms.
The Virial Theorem in Stellar Astrophysics Apparently inspired by a comment of R. One can see that the stability analysis coupled with the postNewtonian form of the virial theorem given in Chapter II [equation (2. 4. P.1. − = − dt 2 ρ dr 1 + P / ρc 2 r2 c2 or in the postNewtonian approximation (i.1. Instead.e.3 †__________________________________________ See Fricke9 who also uses a postNewtonian virial approach to this problem.1. this instability received a great deal of attention and I will not attempt to fully recount it here. by assuming spherical symmetry we may start with the spherically symmetric equation of motion given by Meltzer and Thorne3 as did Fowler4 and follow the formalism of Chapter III. If we confine our attention to objects nearly in equilibrium. = − ⋅ ⋅ ⋅ − = − 1 − 2 − 2 2 dr ρc dt rc r2 c2 4.† Noting that the conditions for this instability also exist in massive white dwarfs. Chandrasekhar and Tooper2 showed by means of rather detailed calculations that a white dwarf would become unstable when its radius shrank to about 246 Schwarzschild radii or on the order of 1000 km. Fowler noted that effects of general relativity would lead to previously unexpected instabilities in supermassive starsl.4. W.15)] would serve as the basis for investigating this effect. Thus 2 2Gm(r ) y 2 dr 1 + 2 − d dr 1 dP c dt rc 2 Gm(r ) 4πGρr .2 ρy 2 Gm(r )ρ 4πGPρr d 2 r dP P 2Gm(r ) . Feynman in 1963.4. 81 . Thus the term involving (dr/dt)2 can be neglected and equation (4.1 − − y y =− P dt dt ρ dr r2 c2 1+ 2 ρc where y= (ρ + P / c ) 2 ρ0 .5% below the wellknown Chandrasekhar limiting mass for degenerate objects. During the next 15 years.. let us examine with the aid of hindsight and the virial theorem.1. how this result could be anticipated without the need of detailed calculations. keeping only terms of the order 1/c2) y2 4. no large scale radial motions can exist.1) becomes d 2 r 1 dP 1 − 2Gm(r ) / rc 2 Gm(r ) 4πG Pr . Rather.15) is extremely difficult. A. This corresponds to a mass about 1. However the estimation or calculation of the relativistic terms on the right hand side of equation (2.
1.4.1.The Virial Theorem in Stellar Astrophysics In hydrostatic equilibrium. 4. lacks the rigor of the EIH postNewtonian approach.9 Putting the results of equation (4.1. it does yield the same results for spherical stars nearly in hydrostatic equilibrium.1.15) of Chapter II for spherical stars but vastly simpler.1. 0 0 0 dr V 4. 4. It is worth noting that the relativistic correction terms are of the same mixed energy integrals as those that appear in equation (2. ∫ c2 r 2c2 V V With somewhat less effort the first integral becomes 4.5 If we retain dP/dr explicitly for the expansion of the first term of equation (4.1.1. 82 .6 ρy =− − + ⋅ ⋅ ⋅ + − 1 + 2 + 2 2 2 dr dt r rc c2 ρc Now if we again form Lagrange's identity by multiplying by r and integrating over all volume.8) and equation (4.1.4. equation (4.15).7). 4.1. 4 2 r r c 4.4 dP Gm(r )ρ 1 =− + 2 dr r2 c 2G 2 m 2 (r )ρ Gm(r )ρ 1 + + 4πrGρ 2 + O 4 . Thus.3) all other relativistic terms of equation (4.1.7 0 dr r rc V V V V V The last integral can be integrated by parts.1 so that G 2 m2 (r)ρ 4πrGPρ dV = ∫ dV . which is basically due to Fowler4. we get R G 2 m2 (r) 2 2 d2r Gm(r)ρ Gm(r)ρ dP 4πrGPρ ρy r 2 dV == −∫ 4πr 3 dr − ∫ dV − ∫ − 2∫ ∫ dt 2 rc2 dV − ∫ c 2 dV . c 3 2 2 4.3) becomes 2 4πrGPρ dP Gm(r )ρ P 2Gm(r ) 2 d r .5) will be of the order l/c4 in the product in equation (4. Although this approach.4.1. dP/dr is given by OppenheimerVolkoff as dP =− dr or Gρ+ P ( r 2 1 − 2Gm(r ) 2 rc )[m(r) + 4πr P c ] .9) into equation (4.9) is zero.10 2 r dt 2 c V c V r2 V which is equivalent to equation (2.8 ∫ R 0 R R R dP 4πr 3 dr = ∫ 4πr 3 dP =4πr 3 P − ∫ 12πr 2 Pdr = −3∫ PdV . and rewriting the left hand side in terms of a relativistic moment of inertia we get d2Ir 1 Gm(r )P 3 G 2 m 2 (r )ρ 1 = 3∫ PdV − Ω − 2 ∫ dV − 2 ∫ dV .1.1.1. noting that the first term on the right hand side of equation (4.3).1.
18 c V r c 0 r Rc 0 x As in equation (4. Since we can write the internal heat energy density as (Γ1−1)u = P.20 Equation (4. 2 2 2 2 1 2 1 2GM m ( r ) R dm( r ) 2 2GM 3 3 Ω1 = 2 Mc ∫ 4 = 8 Mc (q / x ) 2 dq . Chapter III) leads to δΩ = −ξΩ .1. 4.1.11) is 3 M 1 2δΩ1 = 2 ∫ G 2 m 2 (r )δ 2 dm(r ) = −2ξΩ1 .1. Further assume that δr/r = ξoeiσt where ξo is constant and the variation is adiabatic.1. Thus.1. the first term becomes 3δ ∫ PdV = 3 < Γ1 − 1 > δU .1. the integral in equation (4. and x = r/R. 4.1.1.14 2 0 r 2c2 so that the variation of the last term in equation (4.19 Replacing P with u(Γ1−1) as with the first term and letting U1 = 1 c2 1 uGm(r ) dV = P1 .16 2 ∫0 2 r Rc M M Rc where the dimensionless variables are q = [m(r)/M].15 0 c r It is convenient (particularly for the relativistic terms) to normalize by the dimensionless quantity (2GM/Rc2).11 can be normalized in a similar way by making use of the homologous dependence of P. 4.10) as we did in Chapter III.The Virial Theorem in Stellar Astrophysics Taking the variation of equation (4.1. That is P = ηGm 2 (r ) / r 4 . 4. we have 1 2 1 Gm(r )P 3 G 2 m 2 (r )ρ d2 (δI r ) = 3δ ∫ PdV − δΩ − 2 δ ∫ dV .13 The variation of the relativistic correction terms can be computed as follows: Let 2 2 M G m (r ) Ω1 = 3 ∫ dm(r ) .2.18) is dimensionless and determined by the equilibrium model.12 V Equation (3.17 1 Gm(r )PdV 1 R 4πηG 2 m 3 (r )r 2 dr 2GM 1 q = 2∫ = Mc 2 P1 = 2 ∫ 2 2 5 ∫ πη dx . < Γ1 − 1 > r V ∫ 4.16). The remaining terms in 4. let us suppose that the variation of these quantities results from a variation of the independent variable δr.1.1.1.1. 4.11 As before. we can let 2 3 4. Therefore.1.1. where η is a dimensionless scale factor. Thus the remaining equation is δP1 = 2ξ P1 .1.10) (i. 4.11) then becomes: 83 . 4. dV − 2 δ ∫ r c V r2 dt 2 V c V 4.e.
23) and noting that two time differentiations of the perturbation will give a σ2 in the first term.27 where we have used the fact that the Schwarzschild radius (RS) is 2GM/c2.21) becomes 1 2 1 2 d2 (δI r ) =< 3Γ1 − 4 > δΩ − 2 < Γ1 − 1 > δU1 − < 3Γ1 − 5 > δΩ1 . since the internal energy U is coupled with all other terms including the relativity terms we shall eliminate it in a somewhat different fashion than in Chapter III.1.22 and equation (4. equation (4.1. (4. 2 5−n R 5 − n 2 Rc 4. R 4.1. Thus δE = 0 = δU − δΩ + δU1 − δΩ1 . 4.23) becomes σ 2 I r y =< 3Γ1 − 4 > Ω 0 − 4 < Γ1 − 1 > U1 + 2 < 3Γ1 − 5 > Ω1 .25 and calling the dimensionless integrals in equation (4.1.21 dt 2 Now.1.15.26 Since the average relativity factor y is always positive. this expression can be used as a stability criterion as in Chapter III.24 4.1.1.1.The Virial Theorem in Stellar Astrophysics d2 (δI r ) = 3 < Γ1 − 1 > δU . that of a poly trope of index.1.19) into equation (4. That is 3 < 3Γ1 − 4 > RS (4ς1 < Γ1 − 1 > +2ς 2 < 5 − 3Γ1 > ) < 2(5 − n ) .23 Substituting in the variations from equations (4. As this happens. We can now use this to investigate the stability of white dwarfs as they approach the Chandrasekhar limiting mass. ζ1 and ζ2 respectively. its variation is zero. Making one last normalization of Ω0 which for polytropes is 3 GM 1 3 2GM 2 Ω0 = = Mc .1.16) and equation (4.24) becomes 2GM 2GM 3 (4ς1 < Γ1 − 1 > +2ς 2 < 5 − 3Γ1 > ) . dt 2 4. equation (4.1. 84 .δΩ+ < Γ1 − 1 > δU1 − 2δΩ1 .18). the equation of state approaches that of relativistically degenerate electron gas and the internal structure.13).1. 4.1. < 3Γ1 − 4 > − σ 2 I r y = Mc 2 2 2 Rc Rc 2(5 − n ) 4. n = 3. Since the total energy must be constant.1.1. and (4.1.
1. in order to find that point.1.1.13 > ε R0 ≥ 0 R0 . ( ) 4. From Chandrasekhar6 4 ρ e = Bx 3 = 8πm e c 3 / 3h 3 x 3 .34 Normalizing R by Schwarzschild radius we get 85 .5 R S 3 4 R S 1.1.The Virial Theorem in Stellar Astrophysics As n → 3.30 3 where x is the Chandrasekhar degeneracy parameter. Clearly. Thus let Γ1 = 4 + ε . our instability condition can be written as RS > 1 . For that we turn to an interesting paper by Faulkner and Gribben5 who show4. R0 4.32 Now neglecting inverse β decay the local density will be roughly given by ρ = mpρe/me and 3h 3 x3 = ρ .27) becomes (using Fowler’s values forζ1 and ζ2).31 All that remains is to estimate an average value of the degeneracy parameter x which we can expect to be much larger than 1.1. 3 4 and equation (4.33 ( ) 3 M 3 . 4. the value of (Ro/Rs) will monotonically decrease as a result of the mass radius relation for white dwarfs and (l/ε) will monotonically increase as the configuration approaches complete relativistic degeneracy. Γl → 4/3.7 x 2 . if you imagine a sequence of white dwarfs of increasing mass.2 2 x −2 ε≅ .28 ( 4 ς1 + 2ς 2 ) = 9 ε − 2.1.29 Thus.1.1. So. 4. and the system becomes unstable. 3 3 (8πm e c m p ) Let ρ be given by its average value so that 9h 3 x2 = 2 3 3 32π m e c m p 4. However. 9ε R S − 4 R0 or 4. 2 R 2 2 4. we need an estimate of how ε changes with increasing mass. the point must come where the system becomes unstable and collapses.
Then equation (4.1. It is most likely that the discrepancy arises from the rather casual way of estimating x 2 since it will be affected by both the type of volume averaging to determine ρ and the details of the equation of state used in relating ρ to ρe. It should be noted that substituting R0 into the massradius relation for white dwarfs suggests that the critical mass should only be reduced by 1. 4. is still an excellent approximation.1.35 This can be rigorously combined equation (4. However.1. the situation for white dwarfs is quite different. This is the same result that Fowler found for supermassive stars supported by radiation pressure and serves as some justification for using the postNewtonian approximation. Hence the Chandrasekhar limiting mass for white dwarfs. Here the gravitational field is proportionally much stronger with γ being driven to 4/3 by the equation of state and not the radiation field. The Influence of Rotation and Magnetic Fields on White Dwarf Gravitational Instability At this point the reader is likely to complain that the derivation indicating the presence of an instability resulting from general relativity has been anything but brief. it should be remembered that the result is also only correct in the post Newtonian approximation and is an inequality setting a lower limit on instability. 2. However. In spite of this expectation. at many times the Schwarzschild radius. 4. while being somewhat too large. Thus we may expect that a much larger rotational energy field is required to bring about stability than is the case for supermassive stars. since this entire argument is illustrative we also assume that the mass is roughly the limiting mass for white dwarfs. The interesting result is that General Relativity becomes important. indicating that instability sets in.5 percent. to further demonstrate the efficacy of this approach let us consider what impact rotation and magnetic fields may have on the results of the last section. The length results largely from a somewhat different approach to the general relativistic term than used earlier. However. Under these conditions.36 which is in remarkable agreement with the more precise figure of Chandrasekhar and Tooper2 of 246. In Chapter III. Fowler found that a very small amount of rotation would stabilize larger supermassive stars against the gravitational instability so one might wonder what would be the effect in white dwarfs.The Virial Theorem in Stellar Astrophysics x 2 = 7x106 (M⊙/M)4/3 (RS/R0)2 .1. the Newtonian approach of Chapter III will suffice to calculate the terms to be added to the equations of motion and to perform the required variational analysis. That the approach succeeds at all is largely a result of presumed spherical symmetry. we defined the rotational kinetic energy T3 and magnetic energy Mo as 86 . we shall assume that the effects of rotation and magnetic fields are not so extreme as to significantly alter the spherical symmetry. However.31) to provide value for (Ro/Rs).31) becomes: (R0/RS) > 228(M⊙/M)4/9 ≅ 200 .
4 d2 (δI r ) =< 3Γ1 − 4 > (δU − δM) − 2 < Γ1 − 1 > δU1 + < 5 − 3Γ1 > (δT3 − δΩ1 ) . we get 1 2 d2 (δI r ) = 3 < Γ1 − 1 > δU .6 These expressions differ from those in Chapter III only because the gravitational potential energy is taken here to be positive.2. let us normalize the angular velocity ω by the critical value for a Roche model.2.5 Substituting in the values for the variations we get an expression analogous to equation (4. Now the condition on the variation of the total energy becomes δE = 0 = δU − δΩ + δT3 + δM + δU1 − δΩ1 .1. In addition. which enables us to rewrite 4.7 T3 = 1 ω 2 I z = 1 ω 2 I .δΩ + 2δT3 + δM + < Γ1 − 1 > δU1 − 2δΩ1 .2.3 as 1 2 4. In order for us to proceed further it will be necessary to normalize both the rotational energy and magnetic field by something.2. 2 3 Here we are ignoring the relativistic corrections to I and take I = αMR2.1.2 .1 δT3 = −2ξT3 (0) 4.2.The Virial Theorem in Stellar Astrophysics T3 = ∫ 1 0 2 L ωdL M=∫ which have variations H2 dV 8π V . 0 0 0 4.2.24) σ 2 I r y =< 3Γ1 − 4 > (Ω 0 − M0 ) − 4 < Γ1 − 1 > U1 + 2 < 3Γ1 − 5 > [Ω1 − T3 (0)] . 4.2. dt 2 4. 4.21)]. Then 8GM 4 w 2 c 2 2GM ω2 = w 2 27 R 3 = 27 R R c 2 .2. δM = −ξM Adding this to the variational form of Lagrange's identity in section 1 [equation (4.2. dt 2 4. Let us consider the case for ridged rotation so that 4.8 87 .3 .
4 × 10 w .The Virial Theorem in Stellar Astrophysics This certainly does not imply that we are in any way assuming that white dwarfs are represented by a Roche model but rather that it merely provides us with a convenient scale factor.2. Ul.3.113 for polytropes of n = 3 and again using Fowler's4 values for ζ1 and ζ2 this becomes RS 2 −3 2 4. 0 4. Putting these values for T3 and Mo along with the previously determined values for Ω. 0 Using the same analysis for ε as before RS 4/3 −8 2 R < 8. let us pass to the case where n = 3. Namely 2 2 RS (4ς1 < Γ1 − 1 > +2ς 2 < 5 − 3Γ1 > ) < 3 < 3Γ1 − 4 > (1 − H ) + 8αw .89ε(1 − H ) + 4.2.4 × 10 (1 − H ) (M/M⊙) 0 and taking M to be near the Chandrasekhar limit. into equation (4.6) we can arrive at stability conditions analogous to equation (4.2.8 and H > 0. 4. 4. Using this as the normalization constant we have 3 2 2 RS M0 = H Mc R .10 Under these conditions we can expect the maximum values for w and H to be w =1 4.4 × 10 w . we have RS −3 2 R + 4. H =1 and in any event the assumption of sphericity will probably break down for w > 0.27). 2(5 − n ) 0 4.2.2.2.1. In Chapter III we showed that if other effects were absent then Mo > Ω would disrupt the star.2. so that 8αw 2 RS 4 .15 88 .9 In a similar manner let us normalize the magnetic energy Mo by the energy sufficient to bring about disruption of the star.11 . Thus T3 = 4α 2 2 R S w Mc R .12 2(5 − n ) 81 R As before.13 ( 3 ς1 + 2ς 2 ) < 9 ε(1 − H 2 ) + 2 81 R With α= 0. and Ωl.14 R < 0. 81 0 4.2.
2.2.17 R > 210(1 − H ) . S So it is clear that the only effect the field has is to act with general relativity to further destabilize the star.2. The situation regarding rotation is slightly more difficult to deal with as the resulting inequality is a cubic. like 89 .18 R + R 4.14) becomes R0 R0 4 2 −6 4.3 × 10 > 0 . for the moment we neglect rotation. S S 3 4. 4.17) by means of the general cubic we get the results below for the associated values of (Ro/RS) and Veq.19 If we pick a few representative values of w and solve equation (4.2.1 × 10 w − 9. However. With H = 0.2.2.3 × 10 −6 R + R (1 − H 2 ) − (1 − H 2 ) > 0 .2. equation (4. for the field to make any appreciable difference it will have to be truly large. fields in excess of 1012 gauss would not seem to be supported by observations. For our plausible upper limit of H ≈ 103 the effect is to increase the radius at which the instability sets in by about 1%. If.1 794 209 0. nevertheless.The Virial Theorem in Stellar Astrophysics R 0 R 0 4.1 × 10 4 w 2 9. a plausible upper value of H would be on the order of 103. If we assume that we are dealing with objects on the order of 103 km then the disruption field is of the order of 3 x 1015 gauss while the critical equatorial velocity would be about 104 km/sec.0 9453 148 Even these few values are sufficient to indicate that although rotation helps to stabilize the star in the sense of allowing it to attain a smaller radius before collapse.15 × 10 5 w S = R eq ω = R 3 3 R0 0 1 1 3 v eq (km/sec) . Thus. Rotational Effects on the White Dwarf Instability Limit w veq(km/sec) Ro/RS 0 0 210 0.14) can be written as −1 R0 2 3 4. S S An additionally useful expression for the equatorial velocity which corresponds to a given w is R 2 2cw R S 2 = 1.5 4128 194 1. equation (4. The largest observed fields in white dwarfs reach 108 gauss and although it can be argued that larger fields may be encountered in more massive white dwarfs.16 For a point of reference it is worth mentioning the size of the normalization quantities so that various values of w and H may be held in perspective.
The Virial Theorem in Stellar Astrophysics magnetic fields. It is popular to suggest that since the mass limit in white dwarfs arises as a result of the change in the equation of state so a similar change in the equation of state for a neutron star yields a limiting mass for these objects. However. A similar misconception relates to the notion of the neutron star’s limiting mass. the star reaches this point when it is within. So it is not the limiting mass resulting from the change in the equation of state that keeps us from observing more massive white dwarfs. It is quite simple to dismiss this argument as 'nitpicking' as the mass at which the instability occurs is nearly identical to the Chandrasekhar limiting mass. Thus. the configuration would still have to satisfy the virial theorem. However. It seems likely that the resulting shear would produce significant dynamical instabilities. we have seen that neither magnetic fields nor rotation can significantly alter the fact that a white dwarf will become unstable at or about 1000 km. If this were to happen in neutron stars. Let us consider a very simple argument to dramatize this point. In order to do this with differential rotation. a few percent of the Chandrasekhar limiting mass. Classically. since the masses are thought to be similar. it is clear from the development that for either rotation or magnetic fields to really play an important role the total energy stored by either mechanism must approach that in the gravitational field. it is conceptual errors such as this that may lead to much more serious errors in the generalization. It is true that a limiting mass exists for neutron stars but this limit does not primarily arise from a change in the equation of state. 3. One may choose to object to the assumption of rigid rotation as being too conservative. when one tries to generalize the results of one problem to another. In reality the ratios of typical white dwarf radii to neutron stars is suspected to be nearly 1000 which is just about the ratio of the sun's radius to that of a typical white dwarf. 9)].3. For the moment let us ignore the effects of general relativity and just consider the special relativistic virial theorem as we derived in Chapter II [equation (2. As we shall see in the next section. but less than. Rather it is the presence of general relativistic instability that destroys any more massive objects. this is indeed the case with neutron stars. The equation of the state change that results in the Chandrasekhar limit occurs because the electrons achieve relativistic velocities. It is a commonly held misconception that a neutron star is nothing more than a somewhat collapsed white dwarf. the effect is small. Stability of Neutron Stars A second class of objects whose existence became well established during the 1960's is the neutron stars. the differential velocity field would have to be alarmingly high. 90 .
However.4 Mc2[α−(η/2)(RS/R0)] < 0 . based on the analysis in section 1.The Virial Theorem in Stellar Astrophysics 1 2 d2Ir = Ω + T + ∫ (τ / γ )dV .3. This argument further emphasizes the fact that it is the general 91 . However. or (R0/RS) > η/2α . Substitution of Γ = 3/2 in equation (4. He concludes that the instability always occurs before the gas has become relativistic at high pressures.3.1.3.1 where now γ = (1. That this is indeed the case is clearly shown by Tooper7 in considering the general properties of relativistic adiabatic fluid spheres.5 R > 4 .3 .3. Perhaps the "increased gravity" would help stabilize the star against the rapidly increasing internal energy.v2/c2)1/2.2 Ω = −η =− R0 2 R0 The variational form of the virial theorem will require that T+Ω<0 .3 so that 4. However. Since η is of the order of unity and α >> 1.27) and using Fowler’s4 values of ζl and ζ2 for a polytrope of index 2 gives a stability limit of R0 4.3. this would require that the object have a radius less than the Schwarzschild radius in order to be stable against radial pulsations. We may write the gravitational potential energy as ηMc 2 R S GM 2 . 4. Unfortunately we cannot quantitatively apply the results of Section 1 since the way in which Γ approaches 4/3 (more properly the way in which Γ departs from 5/3) depends in detail on the equation of state. S Thus Ro for a neutron star would have to be greater than about 12 km. we should expect the value of Γ to depart farther from the relativistic limit of 4/3 than before. As the neutrons become relativistic γ 1 → 0 and T = αMc2 where α >>1. we may derive some feeling for the way in which the instability sets in by assuming the compression has driven the value of Γ down from 5/3 to 3/2 (i. This is really equivalent to invoking Jacobi's stability condition on the total energy. as the value of Γ approaches 4/3 the same type of instability which brought about the collapse of the white dwarfs will occur in the neutron stars. since the general relativistic correction terms will be much larger than in the case of white dwarfs..e. yet it is still far from the relativistic value of 4/3. Since typical model radii are of the order of 10 km. 4. The exact value of Ro/RS for which this happens will depend on the exact nature of the equation of state as well as details of model construction. This is indeed the case for awhile. dt 2 V 4. This simplistic argument can be criticized on the grounds that it ignores general relativity which can be viewed as increasing the efficiency of gravity. 3/2 is probably a representative value of Γ. just half way to its relativistic value.
The origin of this limit is conceptually identical to that for white dwarfs. not the equation of state becoming relativistic. A crude estimate here will suffice since we are neglecting an increase of perhaps a factor of 2 due to general relativistic terms. we must reevaluate the moment of inertia weighting factor a.2 × 10 w . in order to do this.3.345. However. This result is exactly in accord with what one might have expected on the basis of the white dwarf analysis. there will exist an upper limit to the mass allowable for a neutron star. 0 4. Thus.7 The effects of magnetic fields and rotation are qualitatively the same for neutron stars as for white dwarfs.13 then yields RS 2 −2 2 R < 0. The discussion in section 2 would lead us to believe that neither rotation nor magnetic fields can seriously modify the onset of the general relativistic instability. barring modification to the equation of state resulting from these effects. This can be made somewhat quantitative by evaluating equation (4. While this is true regarding such items as gravitational radiation. we can guarantee that the resulting decrease in Γ will give rise to an unstable configuration at a few Schwarzschild radii. Since the massradius law for any degenerate equation of state (excepting small technical wiggles) will provide for stars whose radius decreases with increasing mass. Substitution into 4. none of these terms should be important unless the configuration becomes smaller than several Schwarzschild radii. this limit can be modified by the presence of magnetic fields and rotation only for the most extreme values of each. Thus. These values vastly exceed those for the most extreme pulsar.2. However. We began this discussion by indicating that the analysis would be very simplistic and yet we have attained some very useful qualitative results. the normalizing fields and rotational velocities are truly immense. One can show by integrating 4π∫ r 2 ρdr by parts in Emden polytropic 0 R variables that 6 ξ1 ξθdξ ξ 2 ∫0 . as for white dwarfs. One may argue that in discussing effects of general relativity we have included terms of O (1/c2) and that higher order effects may be important. Furthermore.2.234(1 − H ) + 3.3. they can largely be ignored in investigating neutron star stability. 92 .10) for a polytrope of index n = 2. α = 1 + 1 2 dθ ξ1 dξ ξ1 which for n = 2 very approximately gives α = 0. For a neutron star with a 10 km radius the magnetic field corresponding to H = 1 in of the order of 1018 gauss while the rotational velocity corresponding to w = 1 would be about 10% the velocity of light. as Ro is only a few times the Schwarzschild radius.The Virial Theorem in Stellar Astrophysics relativistic instability which places an upper limit on the size of the configuration.6 4.
These terms are generally neglected and for good reason. such as with magnetic fields.5. In general it is very much less than one. Since they are of the same sign as the general relativistic terms the surface terms will only serve to increase the instability of the entire configuration.18)]. This results in an increase of the radius at which white dwarfs become unstable.The Virial Theorem in Stellar Astrophysics Even then. However. 93 . Additional Topics and Final Thoughts It would be possible and perhaps even tempting to continue demonstrating the efficacy of the virial theorem in stellar astrophysics. 4. Very little has been said about the large volume of work relating to the equation of state for neutron degenerate matter. the ratio of the scalar value is ∫ P0 r • dS P0 ℵ= S = . In Chapter III. they are unlikely to affect the qualitative behavior of the results. attempting to exhaust the possible applications of the virial theorem is like trying to exhaust the applicability of the conservation of momentum. It cannot hope to provide the information of a detailed structural model but only point the way toward successful model construction. in any equilibrium configuration the pressure must be a monotone increasing function as one moves into the configuration. In most cases the bounding surface can be chosen so as to include the entire configuration. the term is still generally negligible. In instances where this is not the case.1 <P> PdV ∫ V where <P> is the average value of the internal pressure. equation (2. I would be remiss if I did not indicate at least some other possible areas in which the virial theorem can lend insight. Thus. just its relevance. in the case where γ→ 4/3 the variational contribution of the surface pressure approaches 3P0 ∫ r • dS . Such terms are then available to combine with the effects of general relativity. compared to the volume contribution [i. section 4 we discussed the variational effect of the surface terms resulting from the application of the divergence theorem. ℵ< 1. the accretion of matter onto the surface of such objects may cause them to collapse sooner than one might otherwise expect.19) (Chapter II). 4. If one considers the form of the surface terms given by equation (2.e. This is most certainly not to deny its existence. This type of analysis is a probe intended to ascertain what effects are important in the construction of a detailed model and what may be safely ignored. One of the strong points of this approach is that insight can be gained into the global behavior of the object without undue concern regarding the microphysics.5. while the internal pressure and Newtonian gravity contributions vanish.4. Since. However.
There is one instance in which the use of the surface terms can be significant. This is just the time required for a star to radiate away the available gravitational potential energy at its present luminosity.e. It is true that a local stability criterion would be sufficient to locate such instabilities but one could not be sure how the instabilities would propagate without carrying out a large structural analysis. Indeed the spatial moments are taken in order to achieve that result.4. it would be possible to analyze the outer layer for instabilities which might not be globally apparent. as one shrinks the volume to zero he recovers the equations of motion themselves multiplied by the local positional coordinate.The Virial Theorem in Stellar Astrophysics Since magnetic fields also are usually assumed to increase inward. 2(5 − n )RL 5−n 4. they will in general be small compared to the internal contributions. This latter effect can be avoided by utilizing a subglobal form of the virial theorem.2 This makes the other half of the gravitational energy available to be radiated away. In discussing some of the more bizarre and contemporary aspects of stellar structure it is easy to overlook the role played by the virial theorem in the development of the classical theory of stellar structure. One of the most attractive aspects of this entire approach is that it can deal with the properties of an entire system. Only in case of a system with an effective γ approaching 4/3 could these small terms be expected to exert a trigger effect on the resulting configuration. However it is interesting to consider the effects of applying the virial theorem to a subvolume of a larger configuration. 4. The reasoning was flawless. Thus if the contraction liberating the potential energy is uniform and d2I/dt2 is zero then the total kinetic energy always must be T = 1/2 Ω . Under these conditions one might expect the surface terms to be the dominant terms of the resulting expression. It is the virial theorem which provides the theoretical basis for the definition of the KelvinHelmholtz contraction time. If one considers a case intermediate to these limits and investigates the stability of a subvolume which could include the surface of the star. Clearly. It is the virial theorem which essentially tells how much of the gravitational energy is available. The KelvinHelmholtz contraction time for poly tropes is thus T= 3GM 2 4. 94 . failed to withstand the development of stellar astrophysics.5 × 10 7 ≅ (M/M⊙)2(R⊙L⊙/RL) (years). only the initial assumption that the sun derived its energy from gravitational contraction which was plausible at the time. Like the surface pressure terms.3 Reasoning that this provided an upper limit to the age of the sun Lord Kelvin challenged the Darwinian theory of evolution on the sound ground that 23 million years (i. KHT for a polytrope with an internal density distribution of n=3)8 was not long enough to allow for the evolutionary development of the great diversity of life on the planet.4. the influence of magnetic surface terms will be similar to that of a nonzero surface pressure.
5 where α is a measure of the central condensation of the object. However the similarity did not become apparent until the implications of the ergodic theorem inspired by 95 . if the virial theorem is invoked. 4.4. R will have to increase in order to keep Ω constant. so. then once again any internal rearrangement of material that fails to produce sizable accelerative changes in the moment of inertia will require that 2T+Ω = 2EΩ = 0 . I have attempted throughout this book to emphasize that global properties are the very essence of the virial theorem. as the core contracts and α increases.The Virial Theorem in Stellar Astrophysics Another aspect of classical stellar evolution theory is clarified by application of the virial theorem. certain specific time independent or at least slowly varying cases of the nonapproximated equations also yield unique results. Even here the first order theory approximation to general relativity yields an unambiguous form of the virial theorem for spherical objects. Now Ω = αM2/R . Although this global structure provides certain problems when the development is applied to continuum mechanics nothing is encountered within the framework of Newtonian mechanics which is insurmountable. The rather recent development of the virial theorem provides us with a dramatic example of the fact that theories do not develop in an intellectual vacuum. Rather they are pushed and shoved into shape by the passage of the time. The microphysics which couples the core contraction to the envelope expansion is indeed difficult and requires a great deal of computation to describe it in detail. In general the evolutionary changes in a star do take place on a time scale rather less than the contraction time and thus we would expect a general expansion of the outer layer to accompany the contraction of the core. In addition. However the mass distribution of the star places constraints on the overall shape it may take on during rapid evolution processes. Thus we have seen the virial theorem born in an effort to clarify thermodynamics and arising in parallel form in classical dynamics. Only within the context of general relativity may there lie fundamental problems with the definition of spatial moments. Thus one can realistically hope that a general formulation of the virial theorem can be made although one must expect that the interpretation of the resultant spacetime moments will not be intuitively obvious." However. 4.4.4 Since the only way that the star can change its total energy E without outside intervention is by radiating it away to space. Most students find it baffling as to why this should happen and are usually supplied with unsatisfactory answers such as "it's obvious" or "it's the result of detailed model calculations" which freely translated means "the computer tells me it is so. The centrality of taking spatial moments of the equations of motion to the entire development of the theorem demonstrates this with more clarity than any other aspect. All basic courses in astronomy describe postmain sequence evolution by pointing out that the contraction of the core is accompanied by an expansion of the outer envelope. It is in the understanding of such global problems that the virial theorem is particularly useful. any internal changes in the mass distribution which take place on a time scale less than the Kelvin Helmholtz contraction time will have to keep the total energy and hence the gravitational potential energy constant.
incorrectly attacks the theorem itself as opposed to analyzing the application of the theorem and the attendant assumptions. †___________________________ It is said that the great American astronomer. the new future seems at best "seen through a glass darkly". Even though some might claim a little knowledge to be a dangerous thing. Only recently has the similarity of virial theorem development to that of other conservation laws been clearly expounded. by rhetorical intimidation. The attendant stability analysis implied by this approach became the main motivation for further development of the tensor and relativistic forms and provides the primary area of activity today. This is equivalent to attacking a conservation law and serves no useful purpose. certain general aspects of the system are analyzable. Through the course of this book we have examined the origin of the virial theorem. Indeed it may. Now. Thus. Thus. the virial theorem did not really attract attention until 1945 when the global analysis aspect provided a simple way to begin to understand stellar pulsation. Simon Newcomb "proved" that heavier than air flight was impossible and that after the Wright Brothers flew. The virial theorem implies that half of this energy may be radiated away. Although sparsely used by the early investigators of stellar structure. it was rumored that he maintained it would never be practical as no more than two people could be carried by such means. Recent criticism of some work utilizing the virial theorem. turn some less sophisticated investigators aside from consideration of the theorem in their own problems. Although there is a tradeoff in that a complete dynamical description of the system is not obtainable. Immediate problems which seem ideally suited to the application of the virial theorem certainly include exploration into the nature of the energy source in QSO's. and I am mindful that astronomers have not had an exemplary record as predictors of future events†. This would be a most unfortunate result as by now even the most skeptical reader must be impressed by the power of the virial theorem to provide insight into problems of great complexity. it would appear that one need not look for the source of such energy but rather be concerned with the details of the "generator". noted its development and applicability to a wide range of astrophysical problems. In my youth the course of future events always seemed depressingly clear but turned out to be generally wrong. Nevertheless. there may be one or two areas of growth for the virial theorem on which we can count with some certainty. Perhaps one will finally observe that the gravitational energy of assembly of a galaxy or its components is of the same order as the estimated energy liberated by a QSO during its lifetime. I prefer to believe that a little knowledge is better than none at all. Used well this first look will not be the last. the perceptive student of science will utilize the virial theorem to provide a 'first look' at problems to see which are of interest. 96 . and it is irresistible to contemplate briefly its future growth.The Virial Theorem in Stellar Astrophysics statistical mechanics were understood. in spite of a better time base on which to peer forward.
Thus systems in a state of rapid dynamic change are still subject to its time dependent form. we are on the brink of the unification of general relativity.The Virial Theorem in Stellar Astrophysics Perhaps future development will consider applications of the virial theorem as represented by Lagrange's identity. that may be present. Perhaps the most exciting and at the same time least clear and speculative development in which the virial theorem may play a role involves its relationship to general relativity. and as Denis Sciama has noted. magnetic or rotational). Future investigation in this area may be relevant to phenomena ranging from novae to quasars. In the mid twentieth century. It is logical to suppose that sooner or later they will become interested in the effects of such a collapse upon fields other than gravitation (i. The virial theorem provides a clear statement on how the energy in such a system will be shifted from one form to another as soon as one has determined d2I/dt2. Perhaps through the efforts of Stephan Hawking and others. In addition the ergodic theorem seems inexorably tied to the nature of reversible and irreversible processes. as a consequence of discovering that the universe is not a quiet place. To date the virial theorem has been applied to systems in or near equilibrium. theoreticians became greatly excited about the properties of objects undergoing unrestrained gravitational collapse. The advances in relating general relativity to thermodynamics bring these areas and theorems into direct conceptual confrontation and may perhaps provide the foundations for the proper understanding of time itself. You may remember that in Chapter II. difficulty in the interpretation of moments taken over spacetime frustrated a general development of the virial theorem in general relativity and it was necessary to invoke first order approximations to the relativistic field equations. Thermodynamics is the handmaiden of statistical mechanics and it is here through the application of the ergodic theorem that the virial theorem may play its most important future role. and thermodynamics. It is worth remembering that perhaps the most important aspect of the theorem is that it is a global theorem. This is a time of great activity and anticipatory excitement in fundamental physics and general relativity in particular. 97 . quantum mechanics.e.
3 2 2 1 where f(x) = x(2x 3) (x +1) + 3 Sinh (x).2. N4.2.The Virial Theorem in Stellar Astrophysics Notes to Chapter 4 4. x→∞ Now consider the behavior of f(x) as x → ∞ .2 . N4.1.4 sinh −1 ( x ) = ln 2 x . N4. N4.2.5 98 . ρ = Bx 3 . [ ] f ( x ) ≅ x (2x 2 − 3)( x 2 − 3)( x + 1 x −1 + ⋅ ⋅ ⋅+) = 2x 4 + x 2 − 3x 2 − 3 x −1 + ⋅ ⋅ ⋅ + 2 2 or f(x) ≅ 2(x − x ) 4 2 .2.1.1 The last integral can be integrated by parts so that 2 R 4πr Gm( r ) dP R 8πGm ( r ) P 4πG Pr 2 ρ 4πG Pr 2 m(r ) dV = dr − ∫ rdr −∫ 2 2 ∫ c2 0 0 dr c c c V 0 R G 2 m 2 (r )ρ 8πG 8πG R dP dV − 2 m(r ) Pr + 2 ∫ m(r ) P + dr =∫ 2 2 0 dr r c c c 0 V G 2 m 2 (r )ρ 8πG 8πG R dP 8πG R dP dV + 2 m(r )P(r ) − 2 ∫ m(r ) dr + 2 ∫ m(r ) dr =∫ 2 2 0 0 dr dr r c c c c 0 V R R N4.1 The third term vanishes since m(0) = 0 and P(R) = 0.2 Start1ng w1th the polytropic equation of state p=Kργ .2.1 It is not hard to convince yourself that d nP ρ dP γ= = . The limit of the hyperbolic sine is: Lim N4. N4. the last two integrals cancel so that 4πG Pr 2 ρ G 2 m 2 (r )ρ dV = ∫ dV ∫ c2 r 2c2 V V 4.2 d nρ P dρ This can be reduced to a sing1e parameter by considering Chandrasekhar's parametric equation of state for a nearly re1ativistic degenerate gas6 P = Af ( x ).
N4.7 99 .The Virial Theorem in Stellar Astrophysics Simi1ar1y dP dP dx A 8x 3 − 4 x = = . dρ dx dρ B 3x 2 ( ) N4.6 Thus a (8x 3 − 4 x ) Bx 3 ε=γ−4 ≅ 3 2A ( x 4 − x 2 ) b 3x 2 or ε= 4 2x −2 x2 4 2 2 1 + 2 − = 2 + x −2 + x −4 + ⋅ ⋅ ⋅ + − ≅ 3 3 3 x −1 3 3 − 4 3 2 (2 x 2 − 1) 4 = 2 − 3 x −1 3 ( ) .2.2.
W. 616.. Ap. 360361. p. J. J. Chandrasekhar. Fowler. (1966). R. pp. (1966). Chandrasekhar. An Introduction to the Study of Stellar Structure. J. and Tooper. Ap. p. Chandrasekhar. 1541. Dover Pub. Tooper. J. (1972). Dover Pub. W. Mass.J. S.I. A. (1964). R. Cambridge. 183. Stellar Evolution. J. (1973). 514. D. 142. p. An Introduction to the Study of Stellar Structure. F. pp. K. (1966). p. (ed.The Virial Theorem in Stellar Astrophysics References 1 2 3 4 5 6 7 8. 145. J. J. (1965). Vol. Ap. K. Ap. p. and Gribbin. HangYee Chin and Amador Muriel) M. p. Thorne.. and Thorne. Nature. Ap. S. 100 . 218. 1396.T. 180. 7347. R. S. Faulkner. 9. Press. S. 941958. K. pp. 139. Fricke. 144. S. F. 454455. (1957). Meltzer. (1957).
The Virial Theorem in Stellar Astrophysics 101 .
17 102 .2.4. In general.3 Constant in the degenerate equation of state N4.6 Total gas pressure N3.1 Total energy of a system 1.2.5.1.10 Gravitational constant 1.19 Total internal magnetic energy 2.1 Q Arbitrary pointdefined system property 2.7 A temporary quantity 1.18 Arbitrary macroscopic system parameter 1.1.21 Total scalar pressure 1.15 Moment of inertia about the Zaxis 3.11 Qi The generalized forces 3.12 Defined in 3. while an Old English Text type font used for tensorlike quantities of rank 2 or higher.3. Scalars and Special Parameters Symbol A A B D E F G G H I Ir Iz J K L(r) Le Meaning First Used Symbol L M M⊙ N P Pg P1 Q Meaning First Used An arbitrary scalar 2.2.2.2 The magnitude of the magnetic field intensity N2.5.2.3 The magnitude of the angular momentum vector 3. Subscripted Old English type is used to represent the components of these tensors.1.1.3 The magnitude of the electric displacement vector N2.4.22 The particle number density 2.1 R The Rydberg constant 3.4.18 The mass of the sun 3.1.20 RS The Schwarzschild 4.5.1.3.10 Constant in the degenerate equation of state N4.27 R⊙ ℜ The solar radius 3.22 The total rotational kinetic energy 2.5.4.5.5.1 Stellar luminosity 1.10 Solar luminosity 4.3 Moment of inertia including relativistic terms 2. Various other type faces have been employed to provide symbolic representation of scalar quantities which appear throughout this work. The exception is an outlined font used for the unit tensor.2.4.2. I felt it might be useful if a brief definition of these symbols existed in some place for purposes of reference.3 Qp Surface pressure term 3.4.4.1 Moment of inertia about a coordinate origin 1.2.3.1 <Q> Average Q 1.7 A relativistic correction term 4.5.4.5. What follows is s list of the meaning of these symbols and where they first appear.8 The magnitude of the radiative flux 1.1.Symbol Definitions and First Usage Since this work contains a large number of concepts symbolically expressed.8 Qm Magnetic surface term 3.3 The constant of proportionality in the polytropic equation of state N4.2.2 R Stellar or configuration radius 3. the use of bold face type denotes a vector quantity.21 Trace of the Maxwell tensor 2.
1 radial coordinate 1.1.5.13 A relativistic superpotential 2.15 The phaseaveraged kinetic energy of the system 3.10 Boltzmann’s constant 2.4.1.20 magnitude of a velocity vector 1.5.5.8 The speed of light 2.13 forcelaw proportionality constant 1.3 time 1.2.13 103 .10 radial coordinate of the ith particle 1.2 forcelaw exponent 1.1 an initial time 1.1.8 polytropic index 4.21 Specific heat of constant volume 2.5.3.8 components of Minkowski space 2.30 a dimensionless length 4.2.3.5 The timeaveraged kinetic energy of the system 1.1.16 ith linearly independent coordinate 3.5.8 The proper length 2.2.12 The total kinetic energy of the system 1.1 A relativistic superpotential 2.3.2.5.1.1.8 A relativistic superpotential 2.13 A relativistic superpotential 2.2.2.1.2.24 The kinetic or gas temperature 2.3 Parametric variable in the degenerate equation of state 4.25 A dimensionless mass 4.3.2.1.2.5.1.21 mass of the ith particle 1.33 mass interior to a sphere of radius r 1.4.2.1.1.14 KelvinHelmholtz contraction time 4.18 normalized relativistic density 4.5.1 The velocityaveraged ‘creation rate’ 1.1.3 Pulsation period 3.2 separation between the ith and jth particles 1.2 Rotational kinetic energy 3.5.2.2.3.20 The kinetic energy of radial motion 3.10 mass interior to a sphere of volume V N3.1.3.1.31 internal energy of a relativistic gas 2.3 Specific heat of constant pressure 2.1.1.18 a Cartesian coordinate 3.6 Total potential energy 1.6 thermal energy density 4.8 A period of time 1.1 stream speed 2.1.6 Thermal kinetic energy 3.4.21 Planck’s constant 4.16 a Cartesian coordinate 3.3.32 dimensionless magnetic field intensity 4.12 The total internal heat energy 2.5.4.4.2 The surface enclosing a volume V 2.20 Volume enclosing the system 1.4.15 The postNewtonian correction x c2 to the Newtonian internal energy 4.7 An arbitrary potential 1.4.3 Fractional angular velocity 4.4.Symbol S Meaning First Used Symbol h H k mi me mp m(r) m(V) n n q qi r ri ri j s t t0 u u v w xα x x x y y Γ1 Π / c2 Φ <Φ> Meaning First Used S S T T <T> T0 T T T T1 T2 T3 U U U U1 V W Y Z aij c cp cv The 'creation rate' or collision term in the Boltzmann transport equation 1.3.1.1 1st adiabatic constant 4.32 mass of the proton 4.1 mass of the electron 4.20 The timeaveraged potential energy of the system 1.2.
4 The electric mass density 4.2.4.8 A high order superpotential 2.21 1 2 − 2 A dimensionless measure of polytropic structure 4.3 (1 − v 2 / c 2 ) 1 2 ℜ Ψ ℵ Ω Ω <Ω> Ω1/c2 α β βi γ γ γ ε ε ε ε E η η θ θ ξ ξ1 ρ ρe ρe ρ* σ σ τ ϕ φ ψ ω 2.23 Energy generation rate from nonviscous sources 1.Symbol χ ℑ Meaning First Used Symbol ζ1 ζ2 Meaning First Used Local energy of nonconservative forces 1.2.6 A perturbation parameter 4.6.18 104 .6 The Polytropic Emden Radius 4.17 A proportionality constant 4.3.2.3.6.12 Relativistic kinetic energy density N2.1.1.2.4.1.4.1.1.1.4 Ratio of average to surface pressure 4.8 PostNewtonian correction to the Gravitational potential energy 4.2 The electric charge density 2.26 A dimensionless scale factor 4.5 Phasespace point density 1.32 Modified matterenergy density 2.1.2.3.4.1.3 (1 − v 2 / c ) 4.6 Pulsational frequency 3.18 The Polytropic Temperature 4.1 Newtonian Gravitational potential energy 1.2.4.5 The source for a relativistic super potential 2.1.10 A high order superpotential 2.5.5. 3.1.3.1 ratio of specific heats 2.13 Timeaveraged Gravitational potential energy 1.2 Polar angle in Spherical Coordinates 3.4.6 Modified matterenergy density 2.6 The local matter density 1.16 Phaseaveraged Gravitational potential energy 3.1.3.1 Magnitude of the angular velocity N2.5.3.3.3.3.14 parameter measuring central mass concentration 4.3.4.8 The angle between the local Hfield 3.26 A dimensionless measure of polytropic structure 4.6 The Polytropic radial coordinate in Emden Variables 4.5.9 The total energy density 2.14 and r proportionality constant for velocitydependant forces 1.3 The potential energy density N2.10 The Newtonian potential 1.9 Azimuthal polar coordinate 3.28 Thermal kinetic energy density2.3.
4.5.6.23 Unit vector in the rdirection page 2 The local streamvelocity 1.3 The components of D 2.5.2 Components of the stream velocity 2.1 Components of the momentum vector of a particle 1.5.5.9 The magnetic field vector 2.1.1 pi p ri ri ˆ r u ui uα u v vi w w w wi ϖ ϖi x xi zi Momentum vector of the ith particle 1.4.4.3 The components of E 2.2.6.1 A velocity independent force vector on the ith particle 1.5.Vectors and Vector Components Symbol A B Bi D Meaning First Used Symbol pi Meaning First Used Di E Ei F Fi Fij G H Hi ˆ H0 K S Y dS f fi f ff l n ηi An arbitrary vector N2.2 The electric displacement vector 2.1.4.3 Cartesian components of H 2.4.2 The local force vector 1.4 The local residual velocity field in a rotating coordinate frame N2.1 The angular velocity field vector 2.3.5.1.11 The vector “creation” rate 1.4 Components of the 4velocity 2.1 The local peculiar velocity in a rotating coordinate frame 2.14 The relativistic linear momentum density in the post Newtonian approximation 2.2.4.1.6.5.3 The components of B 2.3.1.1.2.3.6 Velocitydependant force vector1.3.5.16 The Lagrangian displacement vector 3.38 The Cartesian components of the Lagrangian displacement vector N3.3 Components of the Cartesian coordinate vector 1.2.2.5.5 Cartesian components of the angular velocity field vector N2.2 The radiative flux 1.4 A relativistic super potential 2.3 Radius vector to the ith particle 1.5.2 Cartesian components of the local residual velocity field 2.1 Components of radius vector 2.8 Total force on the ith particle page 2 Force between the ith & jth particles 1.1 Force acting on the ith particle 1.4.1 Local force density 1.3 The frictional force density 2.5.2 The electric field vector 2.1 Velocity vector of the ith particle 1.1 The local momentum density 1.3 The Lagrangian displacement velocity vector N3.1 105 .1.5.5.24 The net local angular momentum density 2.8 The differential surface normal vector 2.9 The magnetic field intensity 2.3.1.1 The local velocity vector 1.5.15 A Cartesian coordinate vector 2.5.5.7 An arbitrary vector N2.2 A unit vector along H 3.
15 The magnetic energy tensor 2.1.1 The unit tensor 2.11 The Maxwell stressenergy tensor or energy momentum tensor 2.1.1.3 Components of L 2.5.2.5 The components of the LeviCivita tensor density N2.5.1 106 .11 The potential energy tensor 2.5.3 The surface energy tensor 2.1 Components of ℑ 2.4. the Kronecker delta 2.12 Components of S 2.5.4 The volume angular momentum tensor 2.5.4 The gas pressure tensor 2.1.1.Tensors and Tensor Components Symbol Meaning First Used Symbol Ri j S Si j T Ti j U Ui j 1 δij Meaning First Used f I Ii j ℑ ℑij L Li j k M Mi j P Pg The frictional tensor acting as the source for frictional forces 2.1.4. N 2.5.5 Components of U N2.12 The kinetic energy tensor 2.4.5.5 Components of the moment of inertia tensor 2.3.5 Components of T 2.4 Components of the metric perturbation tensor 2.1 ε ijk hi j gi j Components of the Ricci tensor 2.1 Components of the metric tensor 2.2.6.4.7 Components of 1 .3.24 The moment of inertia tensor 2.14 Components of M The pressure tensor 1.5.10.
25. 49. R. 57 conservation. 33 Farquhar. 81 Force.. 73 momentum. 85 Electrostriction. 2. 34 Force density. 95 variation. 87. 86 Fermi. 34 critical. 13 electron. 29. 15 Average Energy kinetic. 87 local. 11. 9 perturbed. R. E... 44. E. A. 1416. 43.60 Equatorial velocity. 35. V. 12 matter. 1 potential. 21 Boltzmann. 69 Feynmann. 31 mass. 33. 65 time. 6 Critical angular velocity.Index Angular Momentum. 70 friction. 8. 89 Ergodic hypothesis. 91 Friction forces. 27. 74 Coriolis force.. 11. 63 Goldstein.. Coriolis. 90 Chandrasekhar. 12. 29 with magnetic fields. 86 Charge density. 8 linear momentum. J. S. S. 15 Euclidean metric. 8 Equations of motion: Newtonian. 1416 velocity. 10. Infeld.. 26. 30 for a zero resistivity gas. I. Hoffman approximation. 13 Claussius. A. 51 Gribben. J. 25. 37 generalized. (see EIH) Electron density. 7. 87 Density charge. 86. 85 force. 34 Gravitational potential.. 49 relativistic. 23 Chandrasekhar limit. 1 Average stream velocity. 44. 47. W. J. 6 Averages: phases. 12. 37 Generalized forces. 6. 85. 85. 9. N.. 10 Gravitational potential energy. H. L. 30 EIH approximation. 12. 1 Center of Mass acceleration. 24. 70 energy. 30 momentum. 6. 14.. 9. 23.. I. P. 82. 72. 23. 22. R. 48. E. 20. 56 Angular Velocity. L. 6. 20. 81. 15 Black holes. 43 Arnold. 84. 88. 81.. 14 Carnot. 3. 36. 35. 29 Einstein. 23. 27. 1 Conservation laws. 15 Faulkner. 22. 5759. 12 relativistic matter. 33 Conservation of: angular momentum. 82 Einstein field equations. 29. 85 107 . 79 Boltzmann Transport Equation. 15 Birkhoff's Theorem. 33 Energy a11 forms (see specific forms) conservation. 70 Creation rate. 12 kinetic energy. 1. 81. 29 EulerLagrange Equations.. 12 Fowler. 36. 6 Avez. 63 Lorentz.
33 Mass. 33. 31 Potential energy. R. 65 Jeans. 2. 39. L. M. I. 10 Rotational. 82 variation.. 10. 37... 9 relativistic. C. D. 91 rotational. 14 Maxwell's laws. K. 32. 31. Sir J. D. (see Lorentz space) Lorentz metric. 37. 52. 83 variation.. 2 Lagrange's identity. 91 Jacobi. 25 Moment of inertia. (see Lorentz space) Lorentz space. 70 Lewis' Theorem. D. 12 Kurth. 96 Noninertial coordinate frames. 68. 59 Linear momentum conservation. W. 34 Lorentz frame. 37 (see also Internal energy) Hunter. 7. 42. 88. 35 Scalar. 10. 2. 9091 Newcomb. 1. 66. 71 Parker. 31 Local angular velocity. 82 (see also Conservation of momentum) Instability. 30 Maxwell. 36. 25. 7. 64 average. 1 Gravitational. 76 Meltzer. 54 variation. E.. 8. N. E. 54 thermal.Hawking. 23. K. 54 Jacobi's stability criterion. 80 Milne.. 8. 12.. 23 Ledoux. 65. 1. 79 stability. 94 PostNewtonian approximation (see EIH) Potential. 51 Kinetic energy density. 1 pulsational.. 42. 12. J. S. 30 Poly tropes.. 6. 1 Poisson's equation.. 13. 50 about an axis. R. 64 Jeans' stability criterion. J. 6. 34 Ogorodnikov. 20 Phase average. 94 Kelvin. C. 15 Plancherel.. E. 6. 51 108 . 12 Neutron stars. 87 Momentum.. 70 Hydrostatic equilibrium. 25 Landau. N. 49. 27. 12. 55. 55 relativistic. 12. 21 Gravitational. 72 average. 72. 1. 26. J. S. 35. 87 Magnetostriction. 92 KelvinHelmholtz contraction time for. 5759 Magnetic disruption energy. 38.. 9. 87 variation. 1416 Phase space. H... 97 Heat energy. 1 KelvinHelmholtz Contraction Time. 25 Limber. 25. Lord. 82 specialrelativistic form. 25. C. 54.. 8 Lagrange. (see Stability) Internal energy. 33. 18. 23. 74 Momentum density. 8. 1. 95 variation. F. 25 Lebovitz. 50. 31 Vector. N. 14 Poincare. P. K. 25. H. 93 Kinetic energy. 15 Leibniz law.. 14 Oppenheimer. 82 Ostriker. 70 Misner.. 26. 43 Lorentz force. 15 Magnetic energy.. 20.. 23 conservation.. 84. 74 Matter density. 54. 42 Lifshitz. 87. 41 Louisville Theorem. 1. 7. J. (see angular and linear) conservation. L. acceleration of center. 6.
1. 1. 15 Space. 8486 Stability criterion. Lorentz. 43. 34 average stream. 87 Variation of (see specific term) Vector potential. 23 Pulsation. 15. K. 31 relativistic. 6. 25. 91 Total energy. 76. 65 variation. 6062 Pulsation energy. 21 Vis viva. 32. 31. F. Y. 81.. 1416 Time variation. 91 Sciama. 3. 54 Thorne. 32 Roche model. 70 against magnetic fields.. 66 Relativistic matter density. A. 82 Volume. for determining stability. 6870 of neutron stars. 96 Stability: global. 54 variation. 25. 67 Pulsational frequency. 1 Viscosity. 89 local angular. 85 stability. 56 Rotational potential. A. effects of surface terms. 12. 83. 59 Pulsation periods. 65. Lord. 23. 2324 Virial tensor. 84. 97 Secular stability. 6 Spacetime. 26. 67 Surface terms: magnetic. 82 Relativistic correction terms to energies. 13 Virial equations. 6 equatorial. 54 Pulsation frequency. 35 Scalar potentials. G. 41 phase. 54. 31 Schwarzschild radius. 86. 40. 70 Volkoff. 37 variation. 25. 32 scalar. 86 Wheeler.interpretation. 62. 82 Relativistic equations of motions. 64. 87 Rosenthal. 66 rotational effects. 39. 66 Rayleigh.. 62 pressure. 59 Pulsational stability. 6870 against rotation. 25 White dwarfs. 14 Rotational energy. 53 effects of rotation and magnetism. 70 Siniai. 81. S. 86. 31. 8386 109 . D. R. angular. relativistic. 31 Velocity. 15 velocity. 77 Tooper.average. 43 Velocity space. 66 secular. 71 pulsational. 30 Relativistic kinetic energy. 81 Time averages.Potentials. 81. 64. 9293 of white dwarfs. J. 30 Relativistic moment of inertia. Jacobi's... 5557 kinetic. 6 Virial. 93 Thermal energy.. 91 Jeans’. 91 Relativistic form of Lagrange's identity (see Lagrange's identity) Relativistic potentials. 12. 10. 25.
This action might not be possible to undo. Are you sure you want to continue?
We've moved you to where you read on your other device.
Get the full title to continue reading from where you left off, or restart the preview.