Kopal 1972

Download as pdf or txt
Download as pdf or txt
You are on page 1of 25

TIDAL EVOLUTION IN CLOSE BINARY SYSTEMS*

ZDENI~K KOPAL
The Lunar Science Institute, Houston, Tex., U.S.A.**

(Received 7 March, 1972)

Abstract. The aim of the present paper will be to give a mathematical outline of the theory of tidal
evolution in close binary systems of secularly constant total momentum - an evolution activated
by viscous friction of dynamical tides raised by the two components on each other. The first section
contains a general outline of the problem; and in Section 2 we shall establish the basic expressions
for the energy and momenta of close binaries consisting of components of arbitrary internal structure.
In Section 3 we shall investigate the maximum and minimum values of the energy (kinetic and
potential) which such systems can attain for given amount of total momentum; while in Section 4
we shall compare these results with the actual facts encountered in binaries with components whose
internal structure (and, therefore, rotational momenta) are known to us from evidence furnished
by the observed rates of apsidal advance.
The results show that all such systems - be these of detached or semi-detached type - disclose
that more than 99 ~ of their total momenta are stored in the orbital momentum. The sum of the
rotational momenta of the constituent components amounts to less than a percent of the total
- a situation characteristic of a state close to the minimum energy for given total momentum. This
appears, moreover, to be true not only of the systems with both components on the Main Sequence,
but also of those possessing evolved components in contact with their Roche limits.
Under such conditions, a synchronism between rotation and revolution (characteristic of both
extreme states of maximum and minimum energy) is not only possible, but appears to have been
actually approached - if not attained - in the majority of cases. In other words, it would appear
that - in at least a large majority of known cases - the existing close binaries have already attained
orbits of maximum distension consistent with their momenta; and tidal evolution alone can no
longer increase the present separations of the components to any appreciable extent.
The virtual absence, in the sky, of binary systems intermediate between the stages of maximum
and minimum energy for given momentum leads us to conjecture that the process of dynamical
evolution activated by viscous tides may enroll on a time-scale which is relatively short in comparison
with their total age - even for systems like Y Cygni or A G Persei, whose total age can scarcely
exceed 107 yr. A secular increase of the semi-majur axes of relative orbits is dynamically coupled
with a corresponding variation in the velocity of axial rotation of both components through the
'tidal lag' arising from the viscosity of stellar material. The differential equations of so coupled a
system are given in Section 5; but their solution still constitutes a task for the future.

1. Introduction

C l o s e b i n a r y s y s t e m s - i.e., d o u b l e s t a r s i n w h i c h t h e d i m e n s i o n s o f t h e i r c o n s t i t u e n t
components amount to an appreciable fraction of their separation - represent con-
f i g u r a t i o n s w h i c h a r e f a r f r o m e x c e p t i o n a l i n t h e sky. T h o u s a n d s o f s u c h o b j e c t s h a v e
b e e n d i s c o v e r e d so f a r ; a n d d i f f e r e n t effects e x h i b i t e d b y v i r t u e o f t h e p r o x i m i t y o f
their components have made the latter the best known stars to us in many respects.
Moreover, t h e e x t e n t o f t h e i r s y m b i o s i s as e v i d e n c e d b y t h e i r o b s e r v e d p r o p e r t i e s

* The Lunar Science Institute Contribution No. 90. The Lunar Science Institute is operated by
the Universities Space Research Association under Contract No. NSR 09-051-001 with the National
Aeronautics and Space Administration.
** Normally at the Department of Astronomy, University of Manchester, England.

Astrophysics and Space Science 17 (1972) 161-185. All Rights Reserved


Copyright 9 1972 by D. Reidel Publishing Company, Dordrecht-Holland
162 ZDENEK KOPAL

offers a potential insight into the past as well as the future of such a binary; for whatever
instantaneous properties of their components may be, certain properties of the system
are bound to be preserved over long intervals of time.
The most important of these is their total m o m e n t u m - representing a sum of the
rotational m o m e n t a of the components and of the orbital m o m e n t u m of relative
motion. As close binaries represent isolated dynamical systems in the sky - effectively
shielded from gravitational interaction with other stars by the scarcity of their close
encounters in space - the total initial momenta remain preserved over long intervals
of time to a farreaching degree of exactitude as long as their total mass is preserved;
the only 'leakage' being possible through removal of m o m e n t u m by photons of the
escaping radiation; but this should represent a totally insignificant fraction of the
total.
Another quantity which should be likewise preserved is the total energy of the
system - representing a sum of the kinetic, potential, and thermal energy of the con-
stituent components as well as the orbital energy of the system. Escape of radiant
energy in the form of light represents a more serious leakage than was the case with
the momentum. A continuous exchange between different types of energy should,
however, be operative within the total, as a result of the evolutionary changes in the
interiors of the constituent components, as well as within the system as a whole. In
particular, dissipative processes connected with dynamical tides should gradually
tend to degrade kinetic energy of motion into thermal energy through the medium of
viscosity. As a result, the kinetic (and potential) energy of a close binary system whose
components interact through dynamical tides should be expected secularly to dimin-
ish, while the total m o m e n t u m remains constant. Therefore, a ratio of the energy-to-
m o m e n t u m in any given system could, in principle, serve as an indicator of its evolu-
tionary stage.
The first investigator who voiced this idea almost a hundred years ago was that
great pioneer of bygone days, George Howard Darwin (1879). He considered, to be
sure, a simplified case of a binary in which only one component was finite in size and
angular momentum, but remained spherical in form; while its companion was re-
garded as a mass-point of finite mass, but zero angular momentum. The relative orbit
of two such bodies was, moreover, constrained to remain circular.
Within the scheme of such an approximation (and with the E a r t h - M o o n system in
mind as a model) Darwin showed that, for systems of constant momentum, the states
of maximum and minimum energy coincided with the establishment of synchronism
between axial rotation and orbital revolution. Moreover, under these conditions, Dar-
win proved that these are the only cases in which synchronism between rotation and
revolution can be encountered. Therefore, such tidal evolution as may occur in the
course of time must be accommodated between these two extremes - at least as long
as the mass of the system remains conserved. We may also add that such numerical
work as Darwin did in this connection was limited to applications to the E a r t h - M o o n
system; and the more general character was outlined by him only in graphical form.
Our present investigation will depart from where Darwin left off, and be concerned
TIDAL EVOLUTION IN CLOSE BINARY SYSTEMS 163

with a generalization of his analysis in several major respects:


(1) Both components of our binary will be allowed to possess finite amounts of
angular momenta; or, should one become a mass-point, the eccentricity of the relative
orbit can be taken into account regardless of its magnitude.
(2) A process will be outlined by which the effects of mutual tidal distortion of both
components can be taken into account in forming expressions for the total energy
as well as momentum.
We shall, moreover, replace Darwin's graphical approach to a solution of the
problem by a numerical investigation of the limiting energy- momentum ratios; and,
in order to pursue the actual evolutionary process between the two extreme cases, a
system of differential equations will be set up governing the changes in the size of the
orbit in the permissible course of the tidal evolution. Furthermore, profiting from
some of our earlier work on the effects of dynamical tides on the axial rotation of the
individual components (Kopal, 1972) we shall be able to treat the semi-major axis A
of the relative orbit as well as the rates c01,2 of axial rotation of the two components
as dependent variables of the same problems. No particular solutions of the respec-
tive system will, to be sure, be attempted within the scope of this paper; but with all
ground work duly laid out, such a solution becomes henceforward a purely numerical
problem.

2. Close Binary Systems: Energy and Momentum

In order to investigate the range through which a binary system of constant total
momentum can evolve in the course of a continuing decrease of its kinetic and poten-
tial energy, consider first the total amount of momentum which must be conserved in
the course of time. Let, in what follows, rn~, 2 denote the masses of the two components
(in general, we shall assume m~, >~mz); and A, the semi-major axis of their relative
orbit of eccentricity e. If so, the angular momentum H o associated with the orbital
motion of the two components about their common center of gravity is given by

Ho - ml + m2 \ dt]

where r denotes the instantaneous radius-vector and v, the true anomaly measured
from the periastron passage. Kepler's second law of (unperturbed) elliptic motion, of
the form

r Zdv
~ = x /- - G m
(1+rnz)A(1-eZ), (2.2)

where G denotes the constant of gravitation, enables us to reduce (2.1) to

(2.3)
l m l + rn2)
164 ZDENI~KKOPAL

which by virtue of Kepler's third law

02 _ G ( m l + rn2)
A3 (2.4)
can be rewritten as
Ho = G 2 / a m l m 2 ( m l +/'/22) -1/3 0 -1/3 (1 - - e2) 1/2 . (2.5)

The angular moments H1, 2 of axial rotation of the two components are obviously
given by
H i = Cico i + CiO, i = 1, 2, (2.6)
where col, 2 stand for the angular velocities of axial rotation of the respective configu-
rations; and C1,2, for their moments of inertia about the axis of rotation; while C1,2
represent the contributions to rotational momenta arising from the tides which the
two components mutually raise on each other.
In what follows we shall assume that the axes of rotation of both components are
perpendicular to the plane of their orbit. If so, neither of the two Keplerian laws as
represented by Equations (2.2) and (2.4) is subject to any perturbations of first order
caused by mutual distortion of the components in close binary systems; while the
values of C1,2 as well as C1,2 are, to this order of accuracy, given by
2 (k2), co2R5
C i = mi~ 2 + ~-... (2.7)
9G
and

C~ = ~-m
1 3_~ ( k 2 ) ~ R~2 +..., (2.8)

i = 1,2; where ~3~ denotes the radius of gyration of the respective component of exter-
nal radius R~, defined by the equation
Ri

rni~ 2 = -}re f oa4da ; (2.9)


iiJ

while the 'apsidal-motion' constants (kj)~, associated with j-th spherical harmonic
distortion, are (for configurations of sufficiently high degree of central condensation)
given by
Ri

8U+2) f 0a 2J+3 d a ,
(kj), - (2j + 1)m,R~5 (2.10)
0

where O(a) denotes the internal density distribution of the respective star.
The total angular momentum H of the system, which should be secularly conserved
to a very high degree of exactitude, will then be given by the equation
H = 1to + Hi + HE, (2.11)
TIDAL EVOLUTION IN CLOSE BINARY SYSTEMS 165

where Ho and 1[-/1, 2 c a n be inserted from (2.5) and (2.6). In order to reduce this ex-
pression to a nondimensional form, let the combined mass rn s + m2 of the system be
hereafter adopted as our unit of mass, and L as the unit of length. If, moreover, Co
denotes the corresponding unit of the time, Equation (2.11) can more explicitly be
rewritten as
H _ ( m l +mz) L2 ~ G2/3mimz ~/1-e 2
Co ((ms + m2) ~" L~ r (Coa) - ' ~ +
Cicocol Cz%c% (Cs + C2) (Co0)~ (2.12)
+ (mi + m~) L2 + (ms + m~) L2 + (ms + m2) L2 ~"
In order to simplify this expression, let us set

z~/3G2/3mlm2
= 1, (2.13)
(ms + m2) 4/3 L~
yielding
(ms + m2) L3/2
co = GI/2(msm2)3/4, (2.14)

as our unit of time, and

(ml + mz) Lz -- { G ( m l m 2 ) 3/2 L} 1/2 (2.15)


TO

as the unit of momentum.


If, lastly, we set

(Co~)-'3 = x, (2.16)
CoCOs = y , (2.17)
ToO) 2 = Z ; (2.18)
and
el, 2
(ml + m2) L2 - tq, 2 (2.19)
while
C1 + C2
- 2, (2.20)
(ml + m 2 ) L 2

Equation (2.12) can be rewritten as

h = x ~ / 1 - e 2 + 2x -3 + ~lY + ~:2z, (2.21)

where h denotes now the normalized value of the total m o m e n t u m H, and x, y, z as


well as #~, z, 2 are nondimensional variables and parameters.
Having formulated the complete expression for the total angular m o m e n t u m
(rotational plus orbital) of a close binary system, let us proceed to formulate similarly
a complete expression for the total energy of the system - kinetic, potential, and ther-
166 ZDENEK KOPAL

mal. Let, in what follows, To denotes the kinetic energy of orbital motion, and Ta, 2
that of axial rotation. As is well known, the kinetic energy of the orbital motion can
be generally represented by an expression of the form
To = lzm i (22 + j)2 + ~ ) + 89 2 (222 + ~ + ~2), (2.22)

where xa, 2, Yl, 2, z~, 2 denote the rectangular coordinates of the centres of gravity of
the two components of mass m~, 2 in a system with origin at the center of gravity of
the binary; and dots denote their time-derivatives.
In order to evaluate the kinetic energy of relative orbital motion, we shift the origin
of coordinates from the center of gravity of the system to that of one (say, the primary)
component by setting

x 2 -- X 1 = X, Y2 -- Yi = Y, Z2 -- Zl = Z, (2.23)
while, by the integrals of the mass center,

mlxl + m 2 x 2 = m l y l + m2Y2 = m l z l + m 2 z 2 = 0. (2.24)


Eliminating x , , 2, Y,, 2, 21,2 from (2.22) with the aid of (2.23) and (2.24) we find that

1 mlrn 2 ~2
To _ _ _ (22 _1_ _1_ ~,2), (2.25)
2 ml + m 2
where
2 2 _{_ j)2 ..~ ~2 ~ V 2 (2.26)

represents the square of the velocity V of motion of the secondary in its relative orbit
around the primary star. On the other hand, by a well-known integral of the problem
of two bodies,

V2=G(rni + m2) { ! - 1A} ; (2.27)

and as the mean value of r - a by the laws of elliptic motion is given by


P
1 [' dt 1
(2.28)
P Jr
0
A

over the orbital cycle of period P, the kinetic energy of orbital motion

To - - Gmim2-- 22(.~. 2. /.3 ~ 1,,~2(m


~
1 q- m2) -1/3 ~-~2/3
(2.29)
2A

by (2.4); while the kinetic energy of axial rotation

Ti --- zCicol
1 2 + ~C~O2 .
(2.30)

The total potential energy W of our binary system (equal to the negative value o
gravitational energy liberated when the configurations contract from infinite disten-
TIDAL EVOLUTION IN CLOSE BINARY SYSTEMS 167
sion to their present size) can be expressed as

- W = Wo + W~ + 14"2, (2.31)

where Wo represents the relative potential energy of the two bodies, while W~,2 stands
for the separate potential energy of each component. As is well known (cf., e.g.,
Kopal, 1959; Section II.2) the relative potential energy can be expressed as

Wo - Gmlrn2 {1 + $12}, (2.32)


7-

where, correctly to quantities of first order in superficial distortion,


2 4 2

(2.33)
i=1 j=2 i=1

in which Oi stands for the mean density of the respective configurations; and (kj)r for
their 'apsidal-motion constants' depending on internal structure.
In addition, the potential energies of the individual components are given by the
mass-integrals
lni

W~= G
f (a)
0
a
din, (2.34)

d m = 4rcoa 2 da sin 0 dO dqS,

which depend only on the internal distribution of density ~, and are unaffected by
distortion to the first order in small quantities (cf. again Kopal, 1959; p. 24).
The internal thermal energy E T of the individual components is given by
mi

ET J~7-1' (2.35)
0

where p denotes the pressure inside the respective configuration; and y, the ratio of
specific heats of stellar material. This thermal energy (being fed mainly by sources of
nuclear origin) is - like the potential energy W~ - obviously unaffected by orbital
parameters (except at the time of Kelvin contraction). The only exception is the heat
source arising from viscous friction of tidal origin; and the latter - being quadratic
in the velocity components of tidal displacements - can be regarded as a small quan-
tity of second order. The same is, moreover, true of the kinetic energy of nonradial
oscillations produced by tides in systems with eccentric orbits, which will likewise
be disregarded in the sequel.
Within the scheme of this approximation - and if we omit parts like W~ or (Er) i
which are independent of the orbital elements - the total energy E of the system
168 ZDENEKKOPAL

should be expressible by

E=T+ W= 89 ~+C2co 2+(61+C2) O 2} +

+
1 17"lj_m2
- -
Gmlm2
AZf22 {1 + g12}, (2.36)
2 rn 1 + m 2 A

in which we replaced r -1 by its time-average A -~ over a cycle, and where g12 repre-
sents the average value of (2.33), in which
P
1 eZ,,312, F (2-n 3-o e2) 237,
P J7 = 2 ' 2 ' '
0

where 2F1 denotes an ordinary hypergeometric series, which terminates (i.e., reduces
to the respective Jacobi polynomials) for integral values of n.
To begin with, let us disregard the small term involving g~ 2 on the right-hand side of
(2.36), and on elimination of A rewrite this latter equation as
f G2Q2 ~1/3
2 ~ = c l o ~ + c ~ o ~ + (c~ + c2) o 2 - ~ m 2 ---- (2.38)
ira 1 --}-m2. } "

In order to normalize this equation in an appropriate manner, let us (consistent with


2.14) adopt

(ml + rnz) L2 G(mlm2) 3:z


-- (2.39)
"Co
2 (ml -t- m 2 ) L '

as our unit of energy, while all others remain the same as before. I f so, Equation
(2.38) can obviously be rewritten in the nondimensional form

2E = ;tx -6 - x -2 + tClY2 q- K72Z2 , (2.40)

where the variables x, y, z as well as the parameters tq, 2 and 2 continue to be given by
Equations (2. l 6)-(2.20).

3. Evolution of Systems with Constant Momentum: Limiting Cases

In order to ascertain the range through which a sum of the kinetic and potential
energy of a binary system can vary while preserving the constancy of the momentum,
let us depart from the normalized Equations (2.21) and (2.40) for the m o m e n t u m and
the energy; and regard the system to be wide enough not only for the distortion S12
occurring on the right-hand side of Equation (2.36), but also the parameter 2 as
negligible to a first approximation. In other words, we shall regard both components -
though of finite size - to remain spherical. If, accordingly, the rotational momenta of
TIDAL EVOLUTION IN CLOSE BINARY SYSTEMS 169

both components, as given by Equation (2.7) reduce to their leading terms /T/ 1351, 2 the

remaining parameters ~1,2 of our p r o b l e m can be regarded as constant; and Equations


(2.21) with (2.40) simplified (for orbits which are circular, or whose eccentricity is
small enough for its squares and higher powers to be negligible) to

h = x + tciy + ~c2z , (3.1)


2E = - x -2 ~- K l y 2 --l- 1(,2z 2 , (3.2)

which will constitute the basis for our present work in this section.
If we eliminate x f r o m the equation for the energy with the aid of that for the mo-
mentum,
2E = / e l y 2 q- /~2 Z2 - - (h - ~lY - / ~ 2 z ) - 2 , (3.3)

and the conditions for the extrema

dE dE
- 0, (3.4)
~y 0z

will be fulfilled if

y = z = (h -/r -- /r -3" (3.5)

Setting z = y in the foregoing equation we find that

y = [h - (~:l +/s y]-3, (3.6)

and similarly (if we wish) for z. If, moreover, we abbreviate

h - (to1 + Ic2) y = Y , (3.7)


so that
h-Y
y - , (3.8)
/r q- K2

Equation (3.6) assumes the f o r m

y 4 _ b y 3 + (x~ + x2) = 0. (3.9)

If, lastly, we set

r = (~1 + ~2)~/'Y ' (3.10)


and
h = (to I + tc2)~/4h', (3.11)

Equation (3.9) can be reduced to a bi-quadratic


y , 4 _ h , y , 3 + 1 = 0, (3.12)

containing only one constant arbitrary parameter. This equation will be fundamental
to all arguments which we propose to develop in this section.
170 ZDENEK KOPAL

In order to solve the foregoing bi-quadratic, let us introduce an auxiliary quantity


9 defined as a solution of
r _ 40 = h '2 . (3.13)
if so, then
y,4_h,y,3 + l= y,2 + 2y, 03/ - h ' + ~-~i72 -j x

(3.14)
20112 J '
which by virtue of the indentity
( 0 3/2 -1- h') ( 0 3/2 - - h') = 40 (3.15)

allows us to rewrite our bi-quadratic in the form of the product

{i-Y' + 88 3/2 - h')] 2 + [88 _ h') d l + 2h'~-3/2] 2} x


X {[Y' -- 8 8 Jr- h ' ) ] 2 -Jr- [ 8 8 q_ h i ) d l - 2h'r 2} = 0.
(3.16)
This result makes it obvious that the nature of the roots of this bi-quadratic depends
on the value ofh'. If

4
h' > 33/4 , (3.17)

our bi-quadratic admits of two real and distinct roots of the form

Y;.2 = 1(~312 + h') {1 _+ x/2h'0 -3/2 - 1}, (3.18)

and two complex roots (which are of no physical interest); while if


h' < 4/3 3/4, (3.19)

all four roots of (3.12) are purely imaginary. Accordingly, by (3.8), (3.10) and (3.18),
h 312 + h' r t _
Y l , 2 -- - -+ + 3/4 {1 +- - , / 2 h 3/2 -- 1}., (3.20)

and, by (3.5),
zl, z = Yl, 2. (3.21)
Consistent with the extremal conditions (3.4) which led to our bi-quadratic, the
roots (3.20) and (3.21) obviously correspond to the states of maximum and minimum
energy consistent with a given momentum h. It is, furthermore, easy to show that each
one of the real roots of our bi-quadratic corresponds to the state of synchronism
between rotation and revolution of the two components. For, in accordance with
Equations (2.16)-(2.18) of the preceding section,
(,01 fD2
-- x 3 y and - - = x3z ; (3.22)
f2 12
TIDAL EVOLUTION IN CLOSE BINARY SYSTEMS 171

and if 0)1 = 0)2 = ~'~,we should have

x3y = i and x 3 z = 1. (3.23)

If, however, we insert in these equations for x from (3.1) and remember that - in
accordance with (3.5) - y = z at the extrema, the foregoing Equations (3.23) will reduce
to (3.5) obtaining as a consequence of (3.4). In other words, the conditions (3.4) re-
quiring that the energy E be m a x i m u m or m i n i m u m f o r constant m o m e n t u m h are
consistent with Equations (3.23). Both extrema correspond, therefore, to synchronism
between rotation and revolution; the maximum corresponding to the case in which
the bulk of the total m o m e n t u m H is represented by the angular momenta H~, 2 of
axial rotation; and the minimum, to a case in which most part of H has been trans-
ferred to the orbital m o m e n t u m H o. In the former case, the components are close to
each other and rotate fast; in the latter, an increase in the size of the orbit has rendered
the orbital m o m e n t u m the dominant part of the whole; but the axial rotation of both
components has been slowed down to the mean daily motion in a wide orbit. More-
over, in the limiting case hm~n= 43/4, these two states coalesce into one and leave no
r o o m for evolution from one into the other; while if h' <~3/4, synchronism between
rotation and revolution can never be attained.
Before we proceed to confront these consequences of the foregoing simple theory
with the observations, let us consider still a simpler case to which our preceding ana-
lysis will reduce if one of the two coefficients ~:1,2 (say, x2) reduces to zero or a quan-
tity very small in comparison with ~:1 - as it may happen even in systems whose com-
ponents are comparable in mass, if the secondary's internal density concentration
becomes so high that its radius of gyration ~ 2 ~ 0 . If so, all of our foregoing analysis
not only continues to hold good for ~c2 = 0, but can be readily generalized to cover the
case of arbitrary eccentricity e of the relative orbit of the two stars.
For if e > 0, Equation (3.1) follows then from (2.21) as

h = x ~ / 1 - e 2 + ~lY ; (3.24)

and putting this through the mill of our procedure, we recover the bi-quadratic
Equation (3.12) with its roots (3.18) provided that the foregoing definitions (3.10) and
(3.11) for Y' and h' are replaced by
Y = [~:1 (1 - e Z ) ] l / 4 Y ' (3.25)
and
h = [tq (1 - e2)]l/4h ' . (3.26)

After this digression, let us return to establish the maximum and minimum values
of the energy E attainable for a given value o f h (or h'). Since these extrema correspond
to the two roots (3.20) for Yl,2 while 21,2--=Y1.2 and, by (3.23), x l , 2 = Y l , 2 -~/3, on
insertion in (3.2) it follows that
2/3 2
2El, 2 = - - Y l , 2 -~- (/s "q- K 2 ) Y l , 2 , (3.27)

where y~, 2 is given by (3.20). Moreover, since for h ' > 43/4 the cubic Equation (3.13)
172 ZDENEK KOPAL

admits of only one real root of the form

4
= ~ 7 cosh89 (3.28)

where
{ h' "]2
cosh ~b = 3 ~/3 h, 2 = , ~ - , (3.29)
16 ~hmin]
it follows that the factor 03/2 which occurs repeatedly in Equation (3.20) will be given
by
0 3/2 = (8/3 3/4) cosh3/2 lgb = h' {1 + (h'min/h') 2 c o s h l t ~ } 1/2 . (3.30)

The two values of El, 2 as given by Equation (3.27) are distinct (and such that
E 1 ~ E2) provided that h ' > ~3/, in accordance with (3.17); and we wish now briefly to
consider what happens as E 2 ~ E 1 . For E = E2, the axial rotation of both components
starts in synchronism with their revolution in close proximity of each other. As E < E2,
the rate of revolution begins to slow down as a result of increasing separation; and
the products
x3 y = lc~ l x 3 ( h - x - ~c2z), (3.31)
x3 z = i c 2 1 x 3 ( h - x - t q y ) , (3.32)

obtained by an insertion from (3.1) for y or z would be maximized (for constant z or


y) when
x= 88 ) or 3 ( h _ ~ q y ) (3.33)

which by (3.1) can be rewritten as

x = 3x2z or 3~:1y, (3.34)

Moreover, the sum Kl(x3y) + tc2(x3 z) will be maximized for


x = 88 (3.35)

implying that
[}r (x3y) + tr (x3Z)]max = 89 4" (3.36)

If, for example, ~2 (or tq) were equal to zero - such as would be the case if one compo-
nent could be regarded as a mass-point - the maximum number of axial rotations per
orbital revolution (i.e., the maximum number of 'days' in a 'month' would be given
by the ratios
1
- --
(gO1/~c'~)max ~ (x3y)max -- (88 4 (3.37)
3tq
or
1
(~02/a)max ------(X 3 z ) .... _
-- - - (88 4, (3.38)
3~c2
respectively.
TIDAL EVOLUTION IN CLOSE BINARY SYSTEMS 173

4. Comparison with Observations

Returning to the main theme of our analysis, we note that Equations (3.17) and (3.27)
of the preceding section impose certain restraints on the physical properties of close
binary systems which we wish to confront now with values furnished by the actual
observations. In order to do so, let us first specify completely our system of units - of
which a choice of the unit L of length has so far been left open. For investigations of
any processes during which the total momentum H as given by Equations (2.11) or
(2.12) is conserved, the possibility suggests itself to normalize our variables in such a
way as to render its normalized value to be equal to

h = 1. (4.1)

If so, however, then it follows from Equation (2.15) that

H2
L - G ( m , m 2 ) 3/2' (4.2)

which inserted in (2.14) yields

(ml + m 2 ) H 3
"Co= G 2 ( m l m 2 ) 3 (4.3)

as our unit of time; while from (2.16)

(4.4)

rendering x 2 proportional to the semi-major axis A of the relative orbit. Lastly, by


(2.19), the constants

(4.5)

Notice that the time-invariance of these units requires a constancy, not only of the
sum m~ + m 2 which has been adopted as the unit of mass, but of m~ and m 2 separately.
With these facts in mind let us return now to the inequality (3.17), disclosing - in
the light of (3.11) and (4.1) - that

(tel q_ tO2)-1/4 t> 4/33/4, (4.6)

if synchronism between rotation and revolution is to be possible at all. If, in particular,


~1 = x2, it would follow that

~a, 2 ~< ~z72 ; (4.7)


174 ZDENI~KKOPAL

or, by (4.5), that the moments of inertia Ca, 2 of the two components

G 2 (mlm2) 3
(4.8)
(ma + n "

If the density concentrations of the two components are pronounced and their radii
of gyration small, the foregoing relations (4.7) and (4.8) should represent strong in-
equalities; but with decreasing central condensations (or increasing size) of the com-
ponents the limit given by the equality sign may be approached.
Next, let us consider the limiting values El, 2 of the total energy expressed in terms of

ma + ms L2 = (4.9)
9o + m )n

This fractional energy has already been defined by Equation (3.27), which for h = 1
can be rewritten as
2E1, 2 = (Ya, 2/h'2) z - yl,2,'
2/3 (4.10)

where (from 3.20)

Ya,2 = h'4{ 1 - 88 + 1) [1 __ x / 2 ~ - ~ - 1]}, (4.11)

in which we have abbreviated

03/2
0~- - - - 2fl- 1 cosh3/2 89(cosh- a f12) (4.12)
h'
and
fl = h ' / h ' i , . (4.13)

For the minimum value ,~ a, rain= ~43/4 in accordance with (3.17), c~=2. On the other
hand, as h ' ~ oo (corresponding to both components being regarded as mass-points),
Equation (3.13) makes it evident that O~/2/h = e ~ 1. Therefore, as

h~in ~< h' ~<oo, (4.13)


the ratio e will be constrained to obey the inequality

2 >~ ~ >/1. (4.14)


Accordingly, the values of Ya, 2 will vary fromthe double-root (4)3/4 to Yl = 1, Y2 = oo and
c~= 1. Moreover, since

c~ 2 = l~ 2 + x/~- 1} (4.15)

and, therefore,

2 cosh 89cosh-a f12 = [f12 + x/~--- 131/3 + Eft2 + X / Y - - 1]-1/3;


(4.16)
TIDAL EVOLUTION IN CLOSE BINARY SYSTEMS 175

so that, b y (4.12)

= 2/?z 7 {1 § [/?2 § Jfl4__ 13-2/3}3/2

= 1 + 6(4/?) -4/3 + 6 (4/?) - s / 3 - 36 (4/?) - ' 2 / 3 + 70(4/?) -26/3 + . . . ;


(4.17)
or, inversely,
(4/?) -4/3 = ~ ( ~ - 1) -- ~ 6 ( ~ - - 1) 2 + ~!Tv(e - 1) 3 -- 9 @ z ( e - 1) 4 + . . . .
(4.18)
M o r e o v e r , o n i n s e r t i o n o f these results in (4.11) it follows t h a t

YI = 1 - 3 (4/?) -8/3 + 116(4/?) -12/3 + ...


14
= 1 - ~z(c~ - 1) 2 + ~-(c~ - 1) 3 + ... (4.19)
and
Y2 = 3 . 3 (4/?) 4 {1 - 3 (4/?) -4/3 + 3 (4fl) - s / 3 - 9 (4/?) - ' 2 / 3 § ...} =
= 3 - 3 (4/?)4 {1 - 89(cz - 1) + { (~z - 1) 2 - 89(~ - 1) 3 4 " " } , (4.20)

c o r r e c t for p o w e r s o f fi u p to, a n d i n c l u d i n g , the -4th.


A six-digit t a b u l a t i o n o f e as a f u n c t i o n o f fi, c o n s t r u c t e d o n the basis o f E q u a t i o n
(4.17), is c o n t a i n e d in the a c c o m p a n y i n g T a b l e I; while T a b l e I I c o n t a i n s the v a l u e s
o f fi, y~,z=-Zl,z (cf. E q u a t i o n 4.11); x~,2 ( f r o m E q u a t i o n 3.23) a n d Et, z ( E q u a t i o n
4.10) c o r r e s p o n d i n g to e = 1 (0.1)2. I n a s m u c h as the p r o d u c t h'x satisfies the s a m e bi-
q u a d r a t i c E q u a t i o n (3.12) as Y' - i.e.,
3
X1, 2 - - X1, 2 §
h,-4 = 0 (4.21)

which, c o m b i n e d w i t h (3.23), furnishes a l i n e a r r e l a t i o n

Y,, 2 = h'4 (1 - x,, 2) (4.22)

b e t w e e n the r o o t s o f o u r b i - q u a d r a t i c s listed in T a b l e II.


W i t h these results in o u r p o s s e s s i o n let us t u r n n o w to real b i n a r y systems such as we

TABLE I *

1 2.00000 5 1.11234
1.25 1.74105 5.5 1.09877
1.5 1.57869 6 1.08782
1.75 1.46904 7 1.07134
2 1.39087 8 1.05961
2.25 1.33278 9 1.05088
2.5 1.28819 10 1.04416
3 1.22052 20 1.01746
3.5 1.18221 50 1.00513
4 1J5199 100 1.00204
4.5 1.12956 oo 1.00000

* The data presented in this table were evaluated by M.E. Alexander.


176 ZDEN~K KOPAL

TABLE II*
fl-1 yl y2 xl x~ E1 E2

1 0 1 oo 1 0 --0.5 +co
1.1 0.183503 1.00052 7942.96 0.999880 0.050119 --0.500114 3753.38
1.2 0.305967 1.00252 972.682 0.999073 0.100927 --0.500374 388.171
1.3 0.411945 1.00914 278.850 0.996935 0.153075 --0.501495 96.7427
1.4 0.508526 1.02200 112.404 0.992792 0.207211 --0.503622 32.9111
1.5 0.598840 1.04371 54.2524 0.985844 0.264156 --0.507080 12.7951
1.6 0.684632 1.07892 29.1307 0.975000 0.325000 --0.512484 5.09788
1.7 0.767001 1.13539 16.6722 0.958557 0.391443 --0.520642 1.80988
1.8 0.846689 1.22996 9.83963 0.933333 0.466667 --0.532981 -t-0.327986
1.9 0.924225 1.41217 5.73489 0.891327 0.558673 --0.552622 --0.336486
2. 1.0O0000 2.37037 2.37037 0.750000 0.750000 --0.592592 --0.592592

* The data presented in this table were evaluated by M. E. Alexander.

know to exist in the sky, and inquire about the stage which they may have already
reached in the course of their evolution. In order to do so we must, of course, be in
possession of all requisite basic data defining, not only the orbital momentum of the
system (expressible in terms of the masses ml,a of the two components and their
separation A), but also the rotational momenta of the two stars (i.e., their moments
of inertia C1,2 and angular velocities ml, 2 of axial rotation); and it is the latter two
quantities which are more difficult to come by.
The angular velocities o)2, 2 can, in principle, be measured from the rotational
broadening of the spectral lines of the respective components (or, for eclipsing system,
from the extent of the 'rotational effect' within minima). Such observations as are
available (for their recent survey see, e.g. Slettebak, 1970) indicate that the components
in most known close binaries describing orbits which deviate but little from circles
rotate essentially in synchronism with their revolution. Noticeable deviations from
synchronism, with components rotating faster than they revolve, are encountered only
in systems with markedly eccentric orbits (such as ~ CrB, e = 0.33; A R Cas, e = 0.25;
or Y Cyg, e=0.14; the latter being a very young system) which are not fully covered
by the discussion contained in the present paper.
Therefore, in what follows we shall restrict the subject of our inquiry to the follow-
ing question: are the data available to us on close binary systems consistent with an
assumption that these systems have already attained their state of maximum disten-
sion - in which o)1,2 = O and the total energy of the system is minimum for given
total momentum?
In an attempt to answer this question, let us return to Equation (4.4) from which,
by hypothesis,

G (mlm2) 2 A (4.23)
= ( m l + m2) H 2'

where xl is the root whose numerical values are listed in column (5) of Table II, and
where (for rotation of both components synchronized with their revolution). It follows
TIDAL EVOLUTION IN CLOSE BINARY SYSTEMS 177

then from Equations (2.3), (2.4), (2.6) and (2.11) that

H = H 0 + s (C 1 + C2) = (1 + v) Ho, (4.24)


where Ho continues to be given by (2.3), C1, 2 by (2.7), and v represents a non-dimen-
sional parameter

s 1+ h2 + 1+ h2 (4.25)
v= Ho

where hi, 2 denote the fractional radii of gyration of the two components expressed in
terms of the semi-axis A of the relative orbit.
The radii of gyration of our configurations could be evaluated by quadratures from
(2.9) if we knew their individual internal structure; or determined empirically for
systems exhibiting evidence of apsidal advance, from the rate of which the constants
k 2 given by Equation (2.10) can be deduced (cf., e.g., Kopal, 1959; Sections II.6 and
VI.9. That these apsidal-motion constants should be related with the fractional radii
of gyration ~/R of the respective stars is evident from Equation (2.9), which shows
(at least for configurations of pronounced central condensation) k 2 to be merely a
moment of inertia of higher order. Numerical work by Motz (1952) has disclosed that,
very approximately, numerical relationship between 5j/R and k2 assumes the form*
log (~/R) = 0.24 log k 2 - - 0.13, (4.26)
which rewritten in terms of the ratios

~i Ri
-hi and - - = r i, i=1,2, (4.27)
A A

transforms into a relation disclosing that

loghi = log ri + 0.24 logk z - 0.13, (4.28)


where both the fractional radii rl, 2 and apsidal-motion constants k 2 c a n be deduced
from an appropriate kind of photometric evidence. And once this has been done, the
values of the ratio v of the sum of the rotational momenta of both components to the
orbital momentum of the system can be evaluated from (4.25).
If we combine now (4.23) with (4.24) and insert for s and Ho from (2.3) and (2.4),
we arrive at a simple relation

1
x - 1 + v' (4.29)

which in combination with the numerical data compiled in Table II discloses that, in
the limiting case of greatest distension (i.e., x = xl) corresponding to minimum energy
per given total momentum, the maximum value of xl = 1 corresponds (by Equation

* It should be noted that the values of k2 used by Motz were one-half of those defined by our Equa-
tion (2.10).
178 ZDENEK KOPAL

4.28) to v = 0 and, therefore (by 4.25) to zero gyration radii (i.e., infinite density con-
centration) of both components; while the minimum value of xl = 43. corresponds to
v = 89 Therefore, in order that our binary should have reached the synchronous stage
of maximum distension, it is necessary that

0 <~ v ~< 1. (4.30)


On the other hand, an excess of the actual value of v over one-third would - in the
light of our analysis - be sufficient to justify a conclusion that the respective system is
still in the process of dynamical evolution, and has not yet reached its ultimate stage
of maximum distension.
As regards the initial stage of maximum energy for given m o m e n t u m - correspond-
ing to the case when x = x 2 - the numerical values of this latter parameter are con-
strained by the inequality

88> x2 > 0 (4.31)


and, therefore,
89< v < Go. (4.32)
The limit v = oo corresponds then to the case of infinite disparity between the masses of
the two components, one of which (say, m2) becomes zero. In such a case, the second-
ary's rotational m o m e n t u m as well as its orbital m o m e n t u m (for the centre of mass
of the primary component of mass ml coincides with that of the orbit) also become
zero; the total m o m e n t u m (4.24) of the system consists only of Crab.
Close binary systems for which sufficiently reliable observational data are on hand
for computation of the ratios v of the rotational to the orbital momenta are listed in
the accompanying Table III. It includes six eclipsing systems of the 'detached' type,
and four 'semi-detached' systems - all characterized by an orbital eccentricity small
enough for its squares and higher powers to be negligible. The individual columns of
Table I I I successively indicate: (1)the system; (2) the spectra of its components; (3)
orbital eccentricity; (4) the mean apsidal-motion constant k2 of the system; (5) and
(6), the mean fractional radii r~,2 of the respective components; (7) and (8), their
fractional radii hl,2 of gyration evaluated with the aid of Equation (4.28); (9), the
mass-ratio of the system; and (10), the parameter v as given by (4.25).
The basic data for rl, 2 and ma/m ~ have been taken from K o p a l and Shapley's
Catalogue (1956), augmented by subsequent spectroscopic data by Batten (1967);
while the mean apsidal-motion constants/~2 have been taken from their most recent
compilation by Alexander (1972).* The ultimate column of Table I I I then contains the
values of v following from Equation (4.25) from the data listed in the preceding co-
lumns - with the exception of that for Y Cyg, which was multiplied by the factor 1.75
giving the excess of its spectroscopically determined angular velocity of axial rotation
(cf. Luyten et al., 1939) over the Keplerian angular velocity of orbital revolution.
A glance at the data listed in the last column of Table I I I discloses at once that, in all
* It should be noted that Alexander (in conformity with most previous investigators) used the
symbol ks to denote one-half of the value of k2 as defined by our Equation (2.10).
TABLE IlI
Comparison with observations
Star Spectra e 89 rl r2 89 89 m2/mx kv

Detached systems
RS CVn F4 -1-K0 IV 0.033 0.0079 0.092 0.28 0.010 0.031 0.96 0.0021
4-7 4.17 • 4-1 4-2 >
G L Car B3 4- B4 0,16 0.0125 0.217 0.217 0.027 0,027 1.0 0.0029
4-1 • 4-8 4-8 g
Y Cyg 09.5 4- 09.5 0.14 0.0093 0.208 0.202 0.024 0,023 0.99 0.0039
4,1 4-13 4-8 4-8 4-1
CO Lac B8,5 4- B9.5 0.027 0.0042 0.252 0.223 0.025 0.022 0.82 0.0023 :z
3 • 4-25 4-21 4-7
V 451 Oph A0 4- [A2] 0.025 0.0082 0.212 0.170 0.024 0.019 0.83 0.0019
4-10 • • • 4-1 r~
A G Per B5 4- [B7] 0.067 0.0059 0.211 0.194 0.022 0.020 088. 0.0017
• 4-20 • 4-7 4-3 z

r~
Semi-detached systems
R Z Cas A2 + [g K0] 0.013 0.0043 0.241 0.284 0.023 0,027 0.35 0.0030
4-6 4-20 :52 -t-2 4.1 fi
W Del A0 4- gG5 0.036 0.0039 0.153 0.244 0.023 0.036 0.21 0,0045
4-2 • • q-2 4.1
fl Per B8 + gG8 0.010 0.0029 0.227 0.239 0.020 0.021 0.19 0,0029
4.1 4.10 • • 4.1
TX U M a B8 + [gG3] 0,022 0.0033 0,158 0,277 0.014 0.025 0.30 0.0017
4-2 4-6 4.1 4-1 4-2
180 ZDENEK KOPAL

systems considered by us, the sum of the angular momenta of axiaI rotation turns out to
be less than one percent of the angular momentum of orbital motion; and, therefore,
a synchronism between rotation and revolution is possible - though not necessary -
for each system listed in our table. Some - like Y Cygni, for instance - may not have
yet attained it; but none are probably very far from it at the present time. In other
words, these (and probably most other known) close binary systems have already
come pretty close to the 'end of their tether'; and tidal evolution alone would no
longer increase the present separation of their components to any appreciable extent.
But, on the other hand, close binary systems could scarcely have originated in the
form of such limiting configurations to begin with, and must at one time have been
much closer together. That this must have been so if their origin is to be sought in any
kind of fission of an originally single configuration is self-evident. But the same
would also be true if they originated by capture; for a dissipation of the requisite
amount of kinetic energy could have been accomplished only at a very close range -
close to actual contact. F r o m whichever direction we attempt to approach our prob-
lem, we find that the initial state of the system should have been close to our 'state 2'
(i.e., synchronous rotation at a close range); the initial degree of proximity being
limited mainly by the physical dimensions of the constituent components; and
'state 1' could have been approached only by some kind of evolutionary process.
A manifest lack of intermediate cases (for which the ratio v would be, say, greater
than 0.1) suggests that the process of dynamical evolution through viscous tides,
which tends to increase separation, may be very effective and operate on a (cosmically)
short time-scale. Such a possibility should focus attention on the quantitative aspects
of the time-scale of tidal evolution in close binary systems - an aspect to which we
shall turn our attention in the concluding section of this paper.

5. Tidal Evolution

In the preceding section of this paper we investigated the range within which the total
energy of a binary system can vary for the constant value of its total momentum. This
analysis has, however, left the actual process by which the energy E can vary within
the interval E 2 >1E >~E 1 completely open; and it is not till this has been specified that
we can assign this progression any definite time-scale. The aim of the present section
will be to investigate the processes which can cause the total energy of the system to
vary - monotonously or otherwise - within the permitted interval between El, 2; and
to attempt to trace the evolutionary course of the individual components for the case
in which the total m o m e n t u m of the respective binary system remains preserved in the
course of time.
In order to investigate the specific nature of constraints imposed by the postulated
constancy of m o m e n t u m on the evolution of close binaries, let us depart from the
expression for H as given by Equation (2.11) and assume - for the time being - that
the masses ml. 2 of both components as well as their internal structure (i.e., the mo-
ments of inertia C~, z) remain constant in time. Let us, moreover, assume that the
TIDAL EVOLUTION IN CLOSE BINARY SYSTEMS 181

relative orbit of the two components is nearly circular - so that quantities of the
order of the squares (or time-derivatives) of its eccentricity e can hereafter be ignored.
If so, a time-derivative of (2.11) can be expressed as

/:I = c ~ + c2~ + ~ ((ms + m2) A


}.2
A = 0, (5.1)

and represents an algebraic relation which the time-derivatives (5~, 2 and A m t


satisfy in the course of time.
An obvious way in which Equation (5.1) can be satisfied is to set cbl, 2 = 0 and A = 0
- corresponding to a secularly stationary case - in which all terms in (5.1) vanish
identically. Such a solution would, indeed, be admissible if the fluid (and, therefore,
deformable) components of a close binary were to consist of inviscid gas, in which
case there would be no dissipation of energy through dynamical tides; and the tidal
bulge produced by the attraction of the companion would follow exactly the direction
of the attracting force.
Inviscid fluids represent, however, only a mathematical abstraction which is not
fulfilled exactly by any material in nature; and so are conservative dynamical systems
in which the conditions cb~,2 = 0 and A = 0 could be fulfilled for any length of time.
All materials constituting celestial bodies possess a certain viscosity (which, for
planetary configurations, can become very large) capable of degrading kinetic energy
into heat; and dissipative forces arising from viscosity are bound to render both time-
derivatives A as well as o51,2 different from zero in the course of time - the latter,
because of the dissipation of kinetic energy of axial rotation through viscous tides;
and the former, because viscosity will render the crest of the tidal wave to deviate
from the direction of the attracting force, and thus give rise to the phenomenon of
'tidal lag'.
Let, in what follows, e denote the angle by which the axis of symmetry of a tidally-
distorted configuration deviates from the radius-vector joining the centres of the two
stars on account of viscosity. If so, and if we restrict ourselves (to begin with) to tides
characterized by second-harmonic symmetry, then we know from the equations for
the perturbation of the Keplerian elements (cf., e.g., Brouwer and Clemence, 1961)
that, within the scheme of our approximation,

dA G(rnl+m2){m~(~) 5
dt - ~ nA 2 (k2) 1 sin2g, +

+ -- (k2) 2 sin2e 2 + . . . , (5.2)


rn 2
where (k2)~,2 are 'apsidal-motion' constants (2.10) associated with the second-har-
monic tidal distortion and depending on the internal structure of the respective con-
figuration; e~, 2, the tidal lags of each star; and n, their mean daily motion, given by

Q2 = n 2 _ G(m I + m2)
A3 (5.3)
182 ZDENEKKOPAL

I f we insert (5.2) in (5.1), we find the former to be satisfied if

Cfo i + { \ A6 ] (k2) i sin 2el = 0, i = 1, 2. (5.4)

These conditions are sufficient, not necessary.


On the other hand, in our earlier study of axial rotation of viscous configurations
(Kopal, 1972) we established that, within the scheme of our approximation, for confi-
gurations in radiative equilibrium
Ri
Ci@ + ~3-~ (co~ - n) Gm~A
~ co2m3-,f #r s dr = 0, (5.5)
0
i = 1,2, where # (r) denotes the distribution of viscosity in the interior; and

C,~, + CrC#c,(m3-[~ R~, (co,- n)=O (5.6)


\ mil~
if the configuration possesses a convective core of radius Rci, with a viscosity/~c~ =
=/~ (Rci) at the interface.
Eliminating the terms C/5 i between Equations (5.4) and (5.5)-(5.6) we find that, in
the case of radiative equilibrium, the angle e~ of tidal lag should be given by
Ri
sin2e i = 130
~ ~ #r s d r =
0
Ri
#r s dr

while in the convective case,


l
0
or 7 dr
; (5.7)

8"A (5.8)

12 Grn3_ i f or 7 dr
0
implying that el > 0 if col > n and vice versa.
Let us now introduce the foregoing expressions (5.7) and (5.5) for tidal lag in
Equation (5.2) for A in which we make use of (5.3); the result will be
R1 R2
aN7~2 767"ciml G+3 l~12{co2(col
d~ - 5 _~3-
- tl) f #r 8 dr +o
co2(co2- t/) f
__~3- #r 8 dr
}
/~/1 W/2
0 0
(5.9)
TIDAL EVOLUTION IN CLOSE BINARY SYSTEMS 183

for radiative equilibrium, and

dA7/2at 14n5 rnl G


+ m2 #clR6cl (c~ - n) + rn~- (c~ - n) (5.10)

in the presence of a convective core. We may note that A is positive if cox> n (i.e., if the
'day' is shorter than the 'month'), and negative if the opposite is the case.
If we normalize all variables in the foregoing equations in accordance with the
scheme adopted in Section IV, we find that Equation (5.9) appropriate for radiative
equilibrium can be rewritten as
dx 7
-- fIY~(Yl - x-3) + f2y2(y2 - x-3), (5.11)
dz

and Equation (5.10) applicable to convective equilibrium as


dx 7
- - f ~ ( Y l - - X - 3 ) q- f 2 ( Y 2 -- X - 3 ) , (5.12)
dz
where
Ri
76n (rnl+m[~m~m2
fi=5H~oo \
f
rn, / rrt~ #rS dr' (5.13)
0
and
14n l~mlm2~ n6
s,' = 5;;L o t,-d-, ) #o,1,c, (5.14)

are nondimensional parameters, in which we have abbreviated


(m 1 q- m 2 ) H 2
Lo= G(mlma)2 ; (5.15)

and "c denotes the normalized time. Moreover, Equations (5.5) and (5.6) governing
axial rotation can be similarly normalized to

x6 dyj~ + g,y~ (Yi - x-3) = 0 (5.16)


dz

for radiative equilibrium, and

x6 dyz (5.17)
d-~ + gi (Yi - x-3) = 0,

for convective core, where


Ri

g~-35HL 6 mi / rn~ ~ #r 8 d r (5.18)


0
184 ZDENEK KOPAL

and
2~z (m3-~) #tiRe6-, (5.19)
9~ - 5HL3o \ mi / ~
represent another set of nondimensional parameters, in which the tc~'s continue to be
given by Equation (4.5).
Equations (5.11) and (5.16) or (5.12) and (5.17) constitute sets of simultaneous
differential equations of second order, which govern the tidal evolution of close
binary systems motivated by viscous friction; and these provide us with the means for
bridging the gap between the two extremes of synchronous rotation investigated in
Section IV. In order to utilize them to this end, all that is necessary is to integrate
(numerically or otherwise) Equations (5.11)-(5.12) and (5.16)-(5.17) from the initial
conditions
x=x2 and Yi=X;3 (5.20)

for values of x z tabulated in column (6) of Table II as a function of e corresponding


to the state of maximum positive energy. Since, for synchronous rotation, the angles
el, 2 of tidal lag vanish, both time-derivatives cbl, 2 as well as A are then zero, thus
satisfying (5.1) identically. However, any arbitrary deviation from these initial con,
ditions - no matter how small - will set in motion a course of tidal evolution, which
can be followed by integration of Equations (5.11),(5.16) and (5.12)-(5.t7) in the
direction of diminishing energy, until a state has been reached when Equations (3.23)
are fulfilled again for

x = x 1 and y=z=yi=xl3, (5.21)

for which the axial rotation of both components becomes once more synchronized
with revolution in an orbit of maximum distension.
As the energy dissipation through tidal friction is bound to be operative in close
binary systems - be these planetary (satellite) or stellar - the tidal evolution the
mechanics of which has been outlined in this section must be expected to occur in
nature; and numerical integrations to illustrate its course will be presented in a
subsequent communication. In saying so we wish, however, also to stress the limita-
tions of the evolution scheme developed so far. To mention the least serious one first,
throughout this paper (with small exceptions) we considered the initial relative orbit
of the two components to be very nearly circular (the squares of the eccentricity
being neglected), and to have remained so throughout its entire subsequent history.
A breakdown of the validity of such an assumption should, however, entail no serious
consequences; and a more detailed analysis of such a case will be given by us in a
subsequent paper to be published in this journal.
A much more serious limitation inherent in our work so far has been the assumed
constancy of the masses of both components, as well as of their internal structure.
While the constancy of mass is likely to be fulfilled to a high degree of approximation
in planetary or satellite cases, the moments of inertia of the planets may undergo
TIDAL EVOLUTION IN CLOSE BINARY SYSTEMS 185

slow secular changes. In the realm of the stars, the masses are likely to remain constant
(within 1 ~ or less) only while both components of a binary system remain on the
Main Sequence. However, as soon as one (or both) components of such a system
has evolved away from the Main Sequence, a mass-exchange between components
(cf., e.g., Plavec, 1970) or mass-loss to the entire system (Kopal, 1971) can take place
which would profoundly influence its evolution.
One of the immediate consequences of the increased degrees of freedom will be
to make it no longer possible to deduce the angles el, 2 of tidal lag directly from the
condition of the constancy of the m o m e n t u m as we have done in this section. The
extent to which tides can lag under more general conditions will then have to be
deduced from appropriate hydrodynamical equations (cf. Kopal, 1968) and solved
separately before the corresponding evolutionary course can be traced with any degree
of confidence. An analysis of so generalized a problem transcends, however, widely
the scope of the present paper; and will be taken up in subsequent communications.

References

Alexander, M. E. : 1972, M.Sc. Thesis, Univ. of Manchester, (unpublished).


Batten, A. H. : 1967, Publ. Dominion Astrophys. Obs. 13, 119.
Brouwer, D. and Clemence, G. M. : 1961, Celestial Mechanics, Acad. Press, New York.
Darwin, G. H.: 1879, Proc. Roy. Soc. London 29, 168.
Kopal, Z.: 1959, Close Binary Systems (Internat. Astrophys. Series, Vol. 5), Chapman-Hall and
John Wiley, London and New York.
Kopal, Z.: 1968, Astrophys. Space Sci. 1, 284, 411.
Kopal, Z. : 1971, Publ. Astron. Soe. Pacific 83, 521.
Kopal, Z. : 1972, Astrophys. Space Sci. 16, 3.
Kopal, Z. and Shapley, M. B. : 1956, Jodrell Bank Ann. 1, 141.
Luyten, W. J., Struve, O., and Morgan, W. W.: 1939, Publ. Yerkes Obs. 7, pt. 4, p. 251.
Motz, L. : 1952, Astrophys. J. 115, 562.
Plavec, M. : 1970, Publ. Astron. Soc. Pacific 82, 957.
Slettebak, A. (ed.): 1970, Stellar Rotation, D. Reidel Pub/. Co., Dordrecht, pp. 178-189.

You might also like