1: %\documentclass[12pt,preprint]{aastex}
2: \documentclass{emulateapj}
3: \usepackage{apjfonts,amsmath}
4:
5: \defcitealias{KK}{Paper I}
6: \newcommand{\paperI}{\citetalias{KK}}
7: \newcommand{\vct}[1]{\mathbf{#1}}
8: \newcommand{\bol}[1]{\mbox{\boldmath{$#1$}}}
9: \newcommand{\mach}{\mathcal{M}}
10: \newcommand{\step}{\mathcal{H}}
11: \newcommand{\Msun}{\rm M_\odot}
12: \newcommand{\kms}{\rm\;km\;s^{-1}}
13: \newcommand{\pcc}{\rm\;cm^{-3}}
14: \newcommand{\D}{\mathcal{D}}
15: \newcommand{\I}{\mathcal{I}}
16: \newcommand{\IRp}{\I_{1, R}}
17: \newcommand{\ITp}{\I_{1, \varphi}}
18: \newcommand{\IRc}{\I_{2, R}}
19: \newcommand{\ITc}{\I_{2, \varphi}}
20: \newcommand{\X}{f\,}
21: \newcommand{\rmin}{r_{\rm min}}
22: \newcommand{\ext}{{\rm ext}}
23: \newcommand{\cs}{c_{s}}
24:
25: \shorttitle{DYNAMICAL FRICTION OF DOUBLE PERTURBERS}
26: \shortauthors{KIM, KIM, \& S{\'A}NCHEZ-SALCEDO}
27:
28: \begin{document}
29: \title{Dynamical Friction of Double Perturbers in a Gaseous Medium}
30: \author{Hyosun Kim\altaffilmark{1}, Woong-Tae Kim\altaffilmark{1},
31: and F.~J.~S{\'a}nchez-Salcedo\altaffilmark{2}}
32: \altaffiltext{1}{Department of Physics and Astronomy, FPRD, Seoul National
33: University, Seoul 151-742, Korea; hkim@astro.snu.ac.kr, wkim@astro.snu.ac.kr}
34: \altaffiltext{2}{Instituto de Astronom\'ia, Universidad Nacional Aut\'onoma
35: de M\'exico, Ciudad Universitaria, 04510 Mexico City, Mexico;
36: jsanchez@astroscu.unam.mx}
37:
38: \begin{abstract}
39: In many astrophysical situations, as in the coalescence of supermassive
40: black hole pairs at gas rich galactic nuclei,
41: the dynamical friction experienced by an object is a combination of its
42: own wake as well as the wakes of its companions.
43: Using a semi-analytic approach,
44: we investigate the composite wake due to, and the resulting drag forces on,
45: double perturbers that are placed at the opposite sides of the
46: orbital center and move on a circular orbit in a uniform gaseous medium.
47: The circular orbit makes the wake of each perturber asymmetric, creating
48: an overdense tail at the trailing side.
49: The tail not only drags the perturber backward
50: but it also exerts a positive torque on
51: the companion. For equal-mass perturbers, the positive torque
52: created by the companion wake is, on average, a fraction $\sim40-50$\% of
53: the negative torque created by its own wake, but this fraction may be even
54: larger for perturbers moving subsonically. This suggests that
55: the orbital decay of a perturber in a double system, especially in the
56: subsonic regime, can take considerably longer than in isolation.
57: We provide the fitting formulae for the
58: forces due to the companion wake and discuss our results in light of
59: recent numerical simulations for mergers of binary black holes.
60: \end{abstract}
61:
62: \keywords{binaries : general
63: --- black hole physics
64: --- hydrodynamics
65: --- waves}
66:
67:
68: %------------------------------------------------------------------------------
69: \section{INTRODUCTION}
70:
71: Understanding the nature of the dynamical friction (DF) force
72: is of great importance to describe the evolution of
73: gravitational systems.
74: Although the concept of DF was introduced by \citet{chandra}
75: studying collisionless backgrounds, it is also of astrophysical interest
76: when considering gaseous media
77: (e.g, \citealt{dok64,rud71,rep80,ost99,san99,san01}).
78: In a seminal paper, \citet{ost99} derived the analytic formulae for
79: drag forces on a perturber moving straight
80: in a uniform gaseous medium, which have found a variety
81: of astrophysical applications,
82: from accretion disks (e.g. \citealt{nar00,kar01,cha01}) to
83: the intracluster medium (e.g., \citealt{kim05,kim07,con08}).
84: Recently, \citet[hereafter \paperI]{KK} extended the work of
85: \citet{ost99} to a more realistic case where the
86: perturber moves on a circular orbit (see \citealt{bar07} for
87: the relativistic case).
88:
89: Although the works mentioned above have improved our understanding on
90: DF force on a single object,
91: there are many astronomical situations involving double or multiple bodies
92: in which one needs an analytical estimate of the net drag force.
93: While the orbital evolution of a binary system due to the DF in
94: a collisionless system has been studied in great detail
95: (e.g., \citealt{heg75}), it is still lacking for the gaseous case.
96: The latter is crucial to describe the formation and hardening
97: of close binary systems \citep{bat02},
98: merging of double black holes at galactic centers
99: \citep{esc04,esc05,dot06,dot07,may07}, and
100: orbital decays of kpc-sized giant clumps formed
101: in primordial ``clump cluster'' galaxies \citep{imm04,bou07}.
102: In particular, angular momentum loss to gas provides a plausible mechanism
103: to explain the relatively rapid coalescence of supermassive
104: black hole pairs in galaxy centers (e.g., \citealt{beg80,gou00,arm02,arm05}).
105: \citet{esc04,esc05} found, through numerical simulations, that
106: the black holes produce a composite wake of an inclined ellipsoidal
107: shape that exerts a net torque on the black holes.
108: This clearly indicates that the in-spiral of one black hole is
109: affected also by the wake from its companion.
110: In this \textit{Letter}, we consider a system composed
111: of two perturbers on coplanar circular orbits
112: in order to assess quantitatively the effect of the companion wake.
113: Using a semi-analytic
114: approach, we study the structure of the combined density wake,
115: evaluate the resulting drag force on each object,
116: and apply our results to the cases considered
117: in numerical simulations of DF-induced mergers of black holes.
118:
119: %---------------------------------------------------------------------------
120: \section{FORMULATION}
121:
122: We consider two point-mass perturbers moving on coplanar
123: circular orbits around
124: the same orbital center in an inviscid gaseous medium.\footnote{The
125: perturbers may orbit under a common external gravitational potential
126: and/or under their mutual gravity.
127: They constitute a binary if the latter dominates.}
128: In order to isolate the effects of the companion, we assume a perfectly
129: uniform background with density $\rho_0$ and ignore the
130: complications from density gradients.
131: To simplify the presentation further, we consider that both perturbers
132: orbit with a fixed radius $R_p$ and at a constant velocity $V_p$,
133: and are located at the opposite sides of the orbital center,
134: but the extension to general case is straightforward.
135: To study the response of gas to the perturbers, we employ the same
136: formalism as in Paper I for a time-dependent
137: linear perturbation analysis.
138: The reader is referred to Paper I for a more detailed description
139: of the method.
140:
141: Assuming that the perturbed density $\alpha = (\rho-\rho_0)/\rho_0$
142: is adiabatic and very small, we linearize the equations of hydrodynamics
143: to obtain a three-dimensional wave equation
144: \begin{equation}\label{eq:wave}
145: \vct{\nabla}^2\alpha-\frac{1}{\cs^2}\frac{\partial^2\alpha}{\partial t^2}
146: = -\frac{4\pi G}{\cs^2}\rho_\ext(\vct{x},t),
147: \end{equation}
148: where $\cs$ is the adiabatic speed of sound in the unperturbed medium
149: and $\rho_\ext$ denotes the mass density of the perturbers
150: (e.g., \citealt{ost99}).
151:
152: We work in cylindrical coordinates $(R,\varphi, z)$ with its origin lying
153: at the orbital center and the $z$-axis perpendicular to the orbital plane.
154: Assuming that the perturbers with mass $M_p$ and $fM_p$ each
155: (with $f$ denoting the mass ratio) are introduced at
156: $(R_p, 0, 0)$ and $(R_p, \pi, 0)$ at $t=0$, respectively, one can write
157: \begin{eqnarray}\label{eq:rho_ext}
158: \rho_\ext(\vct{x},t)&=& M_p\,\step (t)\,\delta (R-R_p)\,\delta (z)\nonumber\\
159: &\times&
160: \{\delta[R_p\,(\varphi-\Omega t)]+\X \delta[R_p\,(\varphi-\pi-\Omega t)]\},
161: \end{eqnarray}
162: where $\step (t)$ is the Heaviside step function
163: and $\Omega \equiv V_p/R_p$ is the angular speed of the perturbers.
164: Since equation (\ref{eq:wave}) is linear, $\alpha$ is given
165: by a simple superposition of the wakes of both perturbers.
166: By solving equation (\ref{eq:wave}) based on the retarded Green's function
167: technique and simplifying the resulting integral analytically,
168: one can show that the perturbed density is reduced to
169: \begin{equation}\label{eq:alpha}
170: \alpha (\vct{x},t) = \frac{G M_p}{\cs^2 R_p}\, \D (R,\varphi,z,t),
171: \end{equation}
172: where $\D=\D_1+ \X \D_2$ with
173: $\D_1$ and $\D_2$ denoting the dimensionless wake,
174: given by equation (8) in Paper I, of the perturber with
175: mass $M_p$ and $fM_p$, respectively.
176: Note that $\D_2(R,\varphi,z,t)=\D_1(R,\varphi-\pi,z,t)$
177: for the perturbers in consideration.
178:
179: The gravitational drag force exerted on the perturber of mass $M_p$
180: located at the position $\vct{x}_p$
181: can be obtained by directly evaluating the integral
182: \begin{equation}\label{eq:DF}
183: \vct{F}_\mathrm{DF} = G M_p \rho_0 \int d^3\vct{x}\
184: \frac{\alpha(\vct{x},t)\ (\vct{x}-\vct{x}_p)}
185: {\vert \vct{x}-\vct{x}_p \vert^3} =
186: \vct{F}_{\rm DF,1} +
187: \vct{F}_{\rm DF,2},
188: \end{equation}
189: where $\vct{F}_{\rm DF,1} = -\mathcal{F}
190: (\IRp\,\hat{\vct{R}} + \ITp\,\hat{\bol{\varphi}})$
191: and $\vct{F}_{\rm DF,2} = -\mathcal{F}\X
192: (\IRc\,\hat{\vct{R}} + \ITc\,\hat{\bol{\varphi}})$,
193: with $\mathcal{F} \equiv 4\pi\rho_0\,(GM_p/V_p)^2$.
194: Here, $\hat{\vct{R}}$ represents the
195: unitary direction vector along $R$, and
196: $\IRp$ and $\ITp$ are the dimensionless
197: drag forces on a perturber by its own wake
198: in the radial and azimuthal directions, respectively,
199: while $\IRc$ and $\ITc$ refer to those
200: from the wake of its companion with mass $fM_p$.
201: The former was evaluated and widely discussed in \paperI.
202: We here focus on $\IRc$ and $\ITc$ defined by equations (12) of
203: \paperI\ except $\D_2$ replacing $\D$ for the perturbed density.
204:
205: As in Paper I, we calculate $\D(\vct{x},t)$ and
206: $\vct{F}_\mathrm{DF}$ on a three-dimensional Cartesian mesh centered
207: at the center of the orbit.
208: We checked that the grid spacing of $\sim R_p/640$ and the box
209: size of $\sim$(20--100) $R_p$ are sufficient to give converged
210: results.
211:
212:
213: %------------------------------------------------------------------------------
214: \section{RESULTS}
215:
216: %------------------------------------------------------------------------------
217: \subsection{Density Wake}\label{sec:wake}
218:
219: %fig1 : 14
220: \begin{figure}%fig1
221: \epsscale{1.2}
222: \plotone{f1.eps}
223: \caption{\label{fig:wake}
224: Distributions of the dimensionless perturbed density $\D$
225: in a steady state for
226: (\textit{a}) $\mach=0.6$,
227: (\textit{b}) 1.2, (\textit{c}) 2.0, and (\textit{d}) 3.2
228: on the orbital plane ($z=0$).
229: The perturbers of equal mass located at $(x,\,y)=(\pm R_p,\,0)$
230: are moving in the counterclockwise direction
231: along a circular orbit marked by the black circle in each frame.
232: Colorbar labels $\log \D$.}
233: \end{figure}
234:
235: In this section, we limit our presentation to the cases with equal-mass
236: perturbers; the cases with $f\neq 1$ will be
237: briefly discussed in \S\ref{sec:dis}.
238: As the perturbers introduced at $t=0$ move along a circular orbit,
239: they continuously launch sound waves that propagate and affect the
240: surrounding medium that would otherwise be uniform and static.
241: Any location inside the causal region is able to
242: receive sonic perturbations from both perturbers, possibly multiple times,
243: creating a density wake that differs significantly depending on the
244: Mach number $\mach\equiv V_p/\cs$ of the perturbers.
245: Figure \ref{fig:wake} displays the distributions of
246: the dimensionless wake $\D$ on the orbital plane ($z=0$),
247: when a steady state is reached ($t \gg R_p/\cs$)
248: for $\mach=0.6$, 1.2, 2.0, and 3.2.
249: In each frame, the perturbers located at $(x,y)=(\pm R_p,0)$ are
250: rotating counterclockwise.
251:
252: For subsonic perturbers ($\mach<1$), the perturbed density is smooth
253: without involving a shock. The bending of the wakes caused by the circular
254: motions leads to slight over-densities at the trailing sides
255: (see Fig.~\ref{fig:wake}a), producing nonvanishing drag forces.
256: In the steady state, which is achieved at $t\rightarrow\infty$,
257: a parcel of gas at any position receives one
258: sonic perturbation from each perturber.
259: For supersonic cases ($\mach>1$), on the other hand,
260: the wake of each perturber initially consists of a sonic sphere and
261: a Mach cone, the interiors of which are influenced
262: by sonic disturbances once and twice, respectively.
263: Because of the circular motion, a perturber (and the head of its Mach cone)
264: is able to overtake its own sonic sphere and subsequently the other
265: sonic sphere from the companion, both of which are expanding radially outward.
266: This in turn provides additional perturbations to the wakes and thus forms
267: long high-density tails that loosely wrap the
268: perturbers in a trailing spiral fashion, as Figure~\ref{fig:wake} shows.
269: The tails in fact trace the regions bounded by shock discontinuities
270: where the gas has received sonic signals four times
271: (three from one perturber and one from the other) and
272: do not overlap with each other,
273: provided the Mach number is less than 2.972 (see below).
274: Note that the densest parts of
275: a tail are located at the immediate trailing side of a perturber,
276: which indicates that the companion wake generally tends to reduce
277: the net DF force, as we shall show in \S\ref{sec:force}.
278:
279: The wake tails thicken as the Mach number increases from unity.
280: \paperI\ showed that a tail from a single perturber becomes fat
281: enough to make the inner edge contact with the outer edge at
282: $\mach_1=4.603$. The self-overlapping of a tail develops
283: a new thin tail that becomes thicker with
284: increasing $\mach$ and again overlaps itself at $\mach_2=7.790$, and so on.
285: In double-perturber cases, however, one tail is able to
286: mutually overlap with the other even before it undergoes
287: self-overlapping.
288: The critical Mach numbers $\mach_n$ for the mutual
289: overlapping of tails are determined by equation (B3) in \paperI\
290: for half-integer $n$. A few critical Mach numbers are
291: $\mach_{1/2}=2.972$, $\mach_{3/2}=6.202$, and $\mach_{5/2}=9.371$.
292: The high-density narrow tails shown in Figure~\ref{fig:wake}d for
293: $\mach=3.2$ are constructed by combining six sonic disturbances
294: emitted by the perturbers.
295:
296:
297: %------------------------------------------------------------------------------
298: \subsection{Gravitational Drag Force}\label{sec:force}
299:
300: Once the perturbed density $\D(\vct{x},t)$ is constructed,
301: it is straightforward to calculate the DF forces exerted on each perturber.
302: While the volume of
303: space influenced by the perturbers steadily increases with time,
304: the resulting drag forces quickly converge to their steady-state
305: values typically within one orbital period.
306: Figure~\ref{fig:force} plots the various drag forces in a steady state
307: on a perturber as functions of the Mach number.
308: In order to avoid a divergence of the force integral,
309: only the region with the distance $r>\rmin=R_p/10$ from the perturber
310: is taken into account in the force computation.
311: Note that only $\ITp$ depends on the Coulomb logarithm $\ln (\rmin/R_p)$ for
312: supersonic perturbers; the other three forces
313: ($\IRp$, $\IRc$, and $\ITc$) are independent of the adopted value for
314: $\rmin$.
315: The local bumps in $\IRc$ and $\ITc$ are caused by
316: the overlapping of the tails occurring at the critical Mach numbers,
317: as discussed in \S\ref{sec:wake}.
318:
319: Figure~\ref{fig:force} shows that $\ITc$ has opposite sign to
320: $\ITp$ for all $\mach$. This implies that, regardless of the Mach number,
321: one perturber in a double-perturber system gains
322: angular momentum from the gravitational torque exerted by the companion
323: wake, while its own wake always takes away angular momentum from it.
324: For equal-mass perturbers,
325: the net drag force in the azimuthal direction is thus smaller than the
326: isolated counterpart.
327: The contribution of the companion wake to the DF force in the azimuthal
328: direction is delineated in the inset of Figure~\ref{fig:force}.
329: In the supersonic range, the ratio $-\ITc/\ITp$ is on average $\sim40\%$,
330: and drops to $\sim12\%$ at $\mach\sim1.2-1.4$
331: where the net azimuthal drag force is maximized.
332: Since $\ITp$ increases with decreasing $\ln(\rmin/R_p)$ for $\mach>1$,
333: the effect of the companion wake on the orbital
334: decay of supersonic double perturbers would diminish
335: as the perturber size nominally represented by $\rmin$
336: decreases relative to the orbital radius.
337: Interestingly, in the subsonic case, the ratio $-\ITc/\ITp$,
338: which does not depend on $\rmin$,
339: is $\sim 50\%$ at $\mach \sim 0.7$ and steeply
340: approaches unity as $\mach\rightarrow 0$,
341: suggesting that the effect of the companion
342: wake is larger as the speed of perturbers decreases.
343:
344: %fig2 : 14
345: \begin{figure}%fig2
346: \epsscale{1.04}
347: \plotone{f2.eps}
348: \caption{\label{fig:force}
349: Gravitational drag forces on a perturber in a double system
350: in the radial (\textit{blue}) and azimuthal (\textit{red}) directions
351: as functions of the Mach number $\mach$.
352: The dashed curves adopted from Paper I with $\rmin/R_p=0.1$
353: represent the forces ($\I_1$)
354: originated from the wake of the perturber itself,
355: while those ($\I_2$) from the companion wake are plotted as dotted lines.
356: The solid lines give the net DF forces ($\I_1+\I_2$) for
357: equal-mass perturbers. The inset plots the ratio $-\ITc/\ITp$
358: which is positive and less than unity for all $\mach$.}
359: \end{figure}
360:
361: On the other hand, $\IRp$ and $\IRc$ are of comparable amplitude
362: over a wide range of $\mach$, and thus give rise to a net radial drag
363: on double perturbers that is about twice larger than in the
364: corresponding single-perturber
365: cases\footnote{When $\mach=0$,
366: the steady-state solution of equation (\ref{eq:wave}) is simply
367: $\alpha=(GM_p/\cs^2) (|r-R_p|^{-1}+f|r+R_p|^{-1})$, for which
368: equation (\ref{eq:DF}) yields
369: $\IRc/\mach^2 = 0.5$ and $\ITc/\mach^2 =0$.}.
370: They affect the orbital eccentricity rather than removing angular momentum
371: much (e.g., Paper I), which
372: may have a gravitational wave signature detectable with
373: \textit{Laser Interferometer Space Antenna (LISA)} \citep{arm05}.
374:
375:
376: For practical purposes, we fit our results for $\IRc$ and $\ITc$ using
377: \begin{equation}\label{eq:IRc}
378: \frac{\IRc}{\mach^2} =
379: \left\{\begin{array}{l@{\ \textrm{if}\ }l}
380: 0.5-0.43\Big(1-\cosh^{-0.36} (2.2\mach)\Big) & \mach\!\!<\! 2.97,\\
381: 0.76-0.08\Big(\mach+(\mach-2.76)^{-1}\Big) & 2.97\!\!\leq\!\mach\!\!<\!6.2,\\
382: 0.56-0.027\Big(\mach+(\mach-6)^{-1}\Big) & \mach\!\!\geq\! 6.2,
383: \end{array}\right.
384: \end{equation}
385: and
386: \begin{equation}\label{eq:ITc}
387: \frac{\ITc}{\mach^2} =
388: \left\{\begin{array}{l@{\quad\textrm{if}\ \mach}l@{\,2.97,}}
389: -0.022\ (10-\mach)\;\tanh\,(3\mach/2) &<\\
390: -0.13+0.07\,\tan^{-1} (5\mach-15) &\geq
391: \end{array}\right.
392: \end{equation}
393: which are accurate within 6\% of the numerical results for all $\mach$.
394: The algebraic fits to $\IRp$ and $\ITp$ are given by equations (13) and (14)
395: of \paperI. It can be shown that
396: $\ITp\approx-\ITc\rightarrow \mach^3/3$
397: in the limit of small $\mach$.
398:
399:
400: %------------------------------------------------------------------------------
401: \section{DISCUSSION}\label{sec:dis}
402:
403: For a system composed of two perturbers moving on coplanar circular orbits
404: at the opposite sides of the system center,
405: we have found that a perturber is dragged backward by its own
406: induced wake, while it is simultaneously pulled forward by the
407: wake of its companion.
408: For equal-mass perturbers,
409: the ratio of the positive torque from the companion wake to the
410: negative torque from its own wake varies between 0.1 to 1.0,
411: and has a mean value at about 0.4.
412: This indicates that, since the wake tails are a large-scale
413: perturbation, the effect of a companion wake on the
414: orbital decay of a double system is by no means negligible
415: except perhaps at $\mach\sim 1.2$--$1.4$.
416: When the perturbers do not have the same orbital radius,
417: the positive torque generated by the companion wake
418: is expected to be enhanced (thus making the net azimuthal drag reduced)
419: if the companion has a larger orbital radius.
420: In order to quantify this effect, we have computed the drag force,
421: for instance, on a body with $\mach=1$ at $R_p$ when the companion
422: is at $2R_p$ and $\mach=2$ (in order to have the same $\Omega$),
423: and found that the positive torque increases by a factor of $\sim1.4$.
424: If the perturbers have different orbital frequencies so that one perturber
425: completes several orbits in the orbital period of the companion,
426: the orbit-averaged torque exerted by the wake of the companion is likely to
427: be reduced.
428: In the limit of very different frequencies, the orbit-averaged
429: torque by the companion wake becomes negligible.
430:
431: The results of this \textit{Letter} can be immediately applied to the
432: numerical models considered in \citet{esc04} for DF-induced mergers
433: of supermassive black holes in a gaseous medium.
434: In their model, two black holes of equal mass are initially
435: separated from each other widely and orbit at near-transonic speed
436: under an external gravitational potential and begin to undergo
437: gaseous drag.
438: Our results shown in Figure~\ref{fig:force} suggest that
439: the companion wake presumably plays a minor role in this early phase
440: of the orbital decay.
441: \citet{esc04} reported that at some point when the binary separation
442: is reduced to $\sim7$ pc, the binary produces a wake in the surrounding
443: medium that is well approximated by an ellipsoid with an axis ratio of 2:1.
444: The major axis of the ellipsoid lags behind the binary axis
445: by $\sim22.5\arcdeg$. Since the wake configuration of this sort
446: can be obtained by blurring the perturbed density shown in
447: Figure~\ref{fig:wake}a, we infer that the black holes have
448: $\mach\sim 0.6$ at this time.
449: The binary keeps hardening by an ellipsoidal torque, and
450: the effect of the companion wake is now non-negligible at all.
451:
452: %fig3 : 13
453: \begin{figure}%fig3
454: \epsscale{1.1}
455: \plotone{f3.eps}
456: \caption{\label{fig:ex}
457: Decay of the separation of a black hole binary of equal mass caused
458: by dynamical friction due to a background gas.
459: The thick and thin
460: curves correspond to the cases when the wakes of
461: both binary components are considered and when the companion wake
462: is ignored, respectively. In both cases, the solid lines plot the
463: results based on both radial and azimuthal forces, while those with
464: only azimuthal forces are given as dashed lines.
465: }
466: \end{figure}
467:
468: Figure 10 of \citet{esc04} shows that it takes the binary about
469: 1.5 Myr to decay from 7 to 0.7 pc. To check whether this is
470: consistent with our predictions, we consider a
471: binary black hole with mass $M_p=5\times 10^8\Msun$ each,
472: embedded in a uniform medium with number density $n_0=1.5\times 10^5\pcc$
473: and sound speed $\cs=650\kms$, a condition similar to those in \citet{esc04}.
474: Initially, the binary has $\mach=0.6$ and a separation of $s_0=7$ pc.
475: We follow the orbital decay of the binary subject to the DF forces
476: found in \S\ref{sec:force}.
477: Figure~\ref{fig:ex} plots the resulting temporal changes of the
478: separation $s$.
479: The thick curves correspond to the cases
480: when the wakes of both binary components are considered, which show that
481: the binary decays to $s=0.7$ pc in $\sim1.5$ Myr, consistent with the
482: results of \citet{esc04}.
483: When the effects of the companion wake are neglected,
484: the decay time becomes shorter by about a factor of 1.5,
485: as indicated by the thin lines.
486: Note that the close agreement between the two cases with and without
487: the radial force demonstrates that it does not affect the decay much.
488:
489: The fact that the companion wake always acts against the orbital decay of
490: a double system suggests an interesting possibility
491: that a less massive perturber in an unequal-mass system
492: may be able to experience a forward net force and move radially outward
493: by acquiring (instead of losing) angular momentum from the wake of
494: a more massive body, resembling a dynamical barrier,
495: while the latter is little influenced by the former.
496: This \textit{dynamical boost} happens if the mass ratio $f$ of two
497: components exceeds $|\ITp/\ITc|$.
498: The evolution of unequal-mass perturbers is more complex
499: since the assumption of identical orbits for both perturbers
500: will fail soon. In practice, one expects that the orbit of the light
501: perturber will be quenched to the massive perturber;
502: the former is trapped on a radius where the drag of
503: its own wake is slightly larger than the forward force of the
504: companion wake, until the orbit of the massive perturber decays
505: and the wakes decouple. We will discuss this case somewhere else.
506:
507:
508:
509: %------------------------------------------------------------------------------
510: \acknowledgments
511:
512: We are grateful to an anonymous referee for a thoughtful report.
513: This work was supported by KICOS through the grant
514: K20702020016-07E0200-01610 provided by MOST,
515: and partly for H.K.~by the BK21 project of the Korean Government.
516: The computations were performed on the
517: Linux cluster at the KASI built with funding from KASI and
518: the ARCSEC.
519: F.J.S.S.~acknowledges financial support from projects
520: PAPIIT IN114107 and CONACyT 60526.
521:
522:
523: \begin{thebibliography}{}
524: \bibitem[Armitage \& Natarajan(2002)]{arm02}
525: Armitage, P.\ J., \& Natarajan, P.\ 2002, \apj, 567, L9
526: \bibitem[Armitage \& Natarajan(2005)]{arm05}
527: Armitage, P.\ J., \& Natarajan, P.\ 2005, \apj, 634, 921
528: \bibitem[Barausse(2007)]{bar07} Barausse, E.\ 2007, \mnras, 382, 826
529: \bibitem[Bate et al.(2002)]{bat02} Bate, M.~R., Bonnell, I.~A., \& Bromm, V.\ 2002, \mnras, 336, 705
530: \bibitem[Begelman et al.(1980)]{beg80}
531: Begelman, M.\ C., Blandford, R.\ D., \& Rees, M.\ J.\ 1980, Nature, 287, 307
532: \bibitem[Bournaud et al.(2007)]{bou07} Bournaud, F., Elmegreen, B.~G., \& Elmegreen, D.~M.\ 2007, \apj, 670, 237
533: \bibitem[Chandrasekhar(1943)]{chandra} Chandrasekhar, S.\ 1943, \apj, 97, 255
534: \bibitem[Chang(2001)]{cha01} Chang, H.-Y.\ 2001, \apjl, 551, L159
535: \bibitem[Conroy \& Ostriker(2008)]{con08}
536: Conroy, C., \& Ostriker, J.\ P.\ 2008, \apj, in press; astro-ph/0712.0824
537: \bibitem[Dokuchaev(1964)]{dok64} Dokuchaev, V.~P.\ 1964, Soviet Astron., 8, 23
538: \bibitem[Dotti et al.(2006)]{dot06} Dotti, M., Colpi, M., \& Haardt, F.\ 2006, \mnras, 367, 103
539: \bibitem[Dotti et al.(2007)]{dot07} Dotti, M., Colpi, M., Haardt, F.,
540: \& Mayer, L.\ 2007, \mnras, 369, 956
541: \bibitem[Escala et al.(2004)]{esc04} Escala, A., Larson, R.~B., Coppi, P.~S., \& Mardones, D.\ 2004, \apj, 607, 765
542: \bibitem[Escala et al.(2005)]{esc05} Escala, A., Larson, R.~B., Coppi, P.~S., \& Mardones, D.\ 2005, \apj, 630, 152
543: \bibitem[Gould \& Rix(2000)]{gou00}
544: Gould, A.,\& Rix, H.-W.\ 2000, \apj, 532, L29
545: \bibitem[Heggie(1975)]{heg75}
546: Heggie, D.\ C.\ 1975, \mnras, 173, 729
547: \bibitem[Immeli et al.(2004)]{imm04} Immeli, A., Samland, M., Gerhard, O., \& Westera, P.\ 2004, \aap, 413, 547
548: \bibitem[Karas \& \v{S}ubr(2001)]{kar01}
549: Karas, V., \v{S}ubr, L.\ 2001, \aap, 376, 686
550: \bibitem[Kim(2007)]{kim07} Kim, W.-T.\ 2007, \apj, 667, L5
551: \bibitem[Kim et al.(2005)]{kim05} Kim, W.-T., El-Zant, A.~A., \& Kamionkowski, M.\ 2005, \apj, 632, 157
552: \bibitem[Kim \& Kim(2007)]{KK} Kim, H., \& Kim, W.-T.\ 2007, \apj, 665, 432 (\paperI)
553: \bibitem[Mayer et al.(2007)]{may07}
554: Mayer, L., Kazantzidis, S., Madau, P., Colpi, M., Quinn, T.,
555: \& Wadsley, J.\ Science, 316, 1874
556: \bibitem[Narayan(2000)]{nar00} Narayan, R.\ 2000, \apj, 536, 663
557: \bibitem[Ostriker(1999)]{ost99} Ostriker, E.~C.\ 1999, \apj, 513, 252
558: \bibitem[Rephaeli \& Salpeter(1980)]{rep80} Rephaeli, Y., \& Salpeter, E.~E.\ 1980, \apj, 240, 20
559: \bibitem[Ruderman \& Spiegel(1971)]{rud71} Ruderman, M.~A., \& Spiegel, E.~A.\ 1971, \apj, 165, 1
560: \bibitem[S{\'a}nchez-Salcedo \& Brandenburg(1999)]{san99} S{\'a}nchez-Salcedo, F.~J., \& Brandenburg, A.\ 1999, \apjl, 522, L35
561: \bibitem[S{\'a}nchez-Salcedo \& Brandenburg(2001)]{san01} S{\'a}nchez-Salcedo, F.~J., \& Brandenburg, A.\ 2001, \mnras, 322, 67
562: \end{thebibliography}
563:
564: \clearpage
565:
566:
567:
568: \end{document}
569: