1: \documentclass[12pt,preprint]{aastex}
2: %
3: \def\bJ{{\bf J}}
4: \def\bS{{\bf J}}
5: \def\bL{{\bf L}}
6: \def\<{\langle}
7: \def\>{\rangle}
8: %
9: \def\alf{Alfv\'en }
10: \def\detg{{\sqrt{g}}}
11: \def\bnabla{{\bf \nabla}}
12: \def\bB{{\bf B}}
13: \def\bP{{\bf P}}
14: \def\bU{{\bf U}}
15: \def\bF{{\bf F}}
16: \def\bva{{\bf v}_{\rm A}}
17: \def\kdv{({\bf k} \cdot {\bf v}_{\rm A})}
18: \def\khdv{(\hat{\bf k} \cdot {\bf v}_{\rm A})}
19: \def\va{{\bf v}_{\rm A}}
20: \def\bk{{\bf k}}
21: %
22: \def\msun{{\rm\,M_\odot}}
23: \def\yr{{\rm\,yr}}
24: \def\au{{\rm\,AU}}
25: \def\del{{\partial}}
26: \def\gm{{\rm\,g}}
27: \def\cm{{\rm\,cm}}
28: \def\sec{{\rm\,s}}
29: \def\erg{{\rm\,erg}}
30: \def\kev{{\rm\,keV}}
31: \def\ev{{\rm\,eV}}
32: \def\K{{\rm\,K}}
33: \def\eps{\epsilon}
34: %
35: \def\sA{{\mathcal{A}}}
36: \def\sD{{\mathcal{D}}}
37: \def\sE{{\mathcal{E}}}
38: \def\order{{\mathcal{O}}}
39: \def\Lone{{\mathcal{L}}_1}
40: %
41: \def\lta{\lesssim}
42: \def\gta{\gtrsim}
43: %
44: \shortauthors{Gammie et al.}
45: \shorttitle{Spin Evolution}
46: %
47: \begin{document}
48: %
49: \title{Black Hole Spin Evolution}
50:
51: \author{Charles F. Gammie\altaffilmark{1,2}, Stuart L.
52: Shapiro\altaffilmark{1,2,3}, and Jonathan C. McKinney\altaffilmark{2}}
53:
54: \affil{Department of Physics, University of Illinois at Urbana-Champaign\\
55: 1110 West Green St., Urbana, IL 61801, USA; }
56:
57: \email{gammie@uiuc.edu, shapiro@astro.physics.uiuc.edu, jcmcknny@uiuc.edu}
58:
59: \altaffiltext{1}{Department of Astronomy, University of Illinois at
60: Urbana-Champaign, Urbana, IL 61801}
61: \altaffiltext{2}{Center for Theoretical Astrophysics, University of
62: Illinois at Urbana-Champaign, Urbana, IL 61801}
63: \altaffiltext{3}{National Center for Supercomputing Applications,
64: University of Illinois at Urbana-Champaign, Urbana, IL 61801}
65:
66: \begin{abstract}
67:
68: We consider a subset of the physical processes that determine the spin
69: $j \equiv a/M$ of astrophysical black holes. These include: (1) Initial
70: conditions. Recent models suggest that the collapse of supermassive
71: stars are likely to produce black holes with $j \sim 0.7.$ (2) Major
72: mergers. The outcome of a nearly equal mass black hole-black hole
73: merger is not yet known, but we review the current best guesses and
74: analytic bounds. (3) Minor mergers. We recover the result of Blandford
75: \& Hughes that accretion of small companions with isotropically
76: distributed orbital angular momenta results in spindown, with $j \sim
77: M^{-7/3}$. (4) Accretion. We present new results from fully
78: relativistic magnetohydrodynamic accretion simulations. These show
79: that, at least for one sequence of flow models, spin equilibrium ($dj/dt
80: = 0$) is reached for $j \sim 0.9$, far less than the canonical value
81: $0.998$ of Thorne that was derived in the absence of MHD effects. This
82: equilibrium value may not apply to all accretion flows, particularly
83: thin disks. Nevertheless, it opens the possibility that black holes
84: that have grown primarily through accretion are not maximally rotating.
85:
86: \end{abstract}
87:
88: \keywords{accretion, accretion disks, black hole physics,
89: Magnetohydrodynamics: MHD, Methods: Numerical}
90:
91: \section{Introduction}
92:
93: The massive, dark objects observed in the centers of galaxies (e.g.
94: \citealt{miy95,mag98}) and the stellar-mass compact objects observed in
95: binary systems systems \citep{mr03} are most readily interpreted as
96: black holes. Alternative models require the introduction of exotic
97: physics \citep{bahc90} or modification of Einstein's equations for the
98: gravitational field (recently \citealt{dp03}). More conventional models
99: such as clusters of compact stars are strongly constrained by
100: observations. Most remarkably, proper motion and radial velocity
101: studies of stars near the putative black hole in Sgr A$^*$
102: \citep{sch02,gez02} require that approximately $3 \times 10^6 \msun$ be
103: concentrated within a region $120 \au$ in radius. There is no stable
104: configuration of normal matter with such a large mass in such a small
105: volume; cluster lifetimes are too short \citep{mao98}. Black holes are
106: therefore the ``most conservative'' model for massive dark objects and
107: galactic black hole candidates (hereafter GBHCs).
108:
109: Black hole solutions of Einstein's equations have three parameters: mass
110: $M$, spin $\bS$, and charge $Q$ (by the ``no-hair,'' or uniqueness,
111: theorem; see \citealt{wal84}). Of these, $Q$ is likely to be negligible
112: in astrophysical contexts because electric charge is shorted out by the
113: surrounding plasma \citep{bz77}. Thus while much of the variation in
114: the observational appearance of black holes is likely due to variation
115: in external parameters such as the angle between black hole spin vector
116: and line of sight, the gas accretion flow geometry and accretion rate
117: $\dot{M}$, and other environmental factors, some might also be due to
118: variation in black hole spin $j \equiv J/M^2 = a/M$.
119:
120: Several features of supermassive black holes (hereafter SMBHs) and GBHCs
121: have been interpreted as evidence for black hole spin:
122:
123: (1) Some SMBHs and GBHCs show broad, skewed Fe K$\alpha$ lines, for
124: example in MCG-6-30-15 \citep{tan95,fab02}, Cyg X-1 \citep{mil02a}, and
125: XTE J1650-500 (\citealt{mil02b}; for a review see \citealt{rn02}). If
126: one assumes that these lines originate in plasma on nearly circular,
127: equatorial geodesics within a few $M$ of the black hole, then the line
128: shape is sensitive to the spin of the hole (e.g. \citealt{lao91}).
129: Within the context of this model rotating holes are required to explain
130: the observed red wing of the line.
131:
132: (2) The ratio $R$ of observed quasar radiative energy per unit comoving
133: volume to the current mass density of black holes is directly related to
134: the mean radiative efficiency $\eps$ of accretion onto black holes
135: \citep{sol82}: $\eps > R$. Estimates suggest $R > 0.1$
136: \citep{yt02,erz02}. If one assumes that accretion occurred through a
137: classical thin disk in which the binding energy of the innermost stable
138: circular orbit (hereafter ISCO) determines $\eps$ \citep{bar70}, then $R
139: > 0.1$ requires $j > 0.67$.
140:
141: (3) In GBHCs, quasi-periodic oscillations (QPOs) are observed in X-ray
142: light curves at frequencies ranging from a fraction of a Hz to $450$ Hz
143: in GRO J1655-40 \citep{str01}. Assuming that these QPO frequencies are
144: bounded above by the rotation frequency of the ISCO and that the QPO is
145: not an overtone, one can place a limit on the mass and spin of the black
146: hole. In GRO J1655-40, $95\%$ confidence limits on the mass
147: \citep{sha99} require $M > 5.5 \msun$, or $j > 0.15$ \citep{str01}. A
148: physical or phenomenological model for the QPO can provide more
149: stringent constraints, but requires additional assumptions.
150:
151: (4) The shape of the X-ray continuum from an accreting black hole may
152: depend on the spin. Calculating an expected continuum requires the
153: black hole mass, spin, flow geometry (usually, but not always, a thin
154: disk) and a model for the accretion flow atmosphere. Models have been
155: applied to a number of objects by, e.g., \cite{zcc97} and \cite{gme01},
156: and usually suggest $j \sim 1$.
157:
158: %
159: %
160: %(5) Winged or X-type radio sources may reflect sudden changes in the
161: %orientation of a central black hole's spin axis due to mergers, leading to
162: %a sudden flip in the observed direction of an associated jet
163: %\citep{me02}.
164: %
165:
166: This list is necessarily incomplete, and in each case the evidence for
167: black hole spin is open to debate. Models for the dynamics of the
168: plasma surrounding the black hole and its radiative properties must be
169: invoked. These models describe an intrinsically complex physical system
170: and use approximations of unknown accuracy. Future calculations,
171: particularly numerical models of the accretion flows, may help reduce
172: the uncertainties. Analysis of gravitational waveforms emitted by
173: perturbed black holes undergoing mergers can reveal their masses and
174: spins and may prove less ambiguous once these signals can be measured
175: reliably \citep{tho95,flan98,drey03}.
176:
177: Given the existing evidence for black hole spin, it is useful to
178: consider the physical processes governing spin evolution. In this paper
179: we consider initial conditions (\S\ref{initial}), mergers with black
180: holes of comparable mass (\S\ref{major}), mergers with smaller objects
181: (\S\ref{minor}), and accretion (\S\ref{accretion}), then summarize our
182: results in \S\ref{finale}. Throughout the paper we adopt geometrized
183: units and set $G = c = 1$.
184:
185: \section{Initial Conditions}\label{initial}
186:
187: Nonprimordial black holes form from gravitational collapse and in
188: general are born with nonzero spin. If subsequent accretion is
189: negligible, the initial spin state will be preserved and the black hole
190: spin will be determined by the dynamics of the initial collapse.
191:
192: If the initial collapse occurs from a massive star, the spin depends on
193: the angular momentum profile of the progenitor star and the
194: (magneto)hydrodynamics of core collapse in evolved, spinning stars. The
195: dependence is not fully understood at this time, although detailed
196: Newtonian simulations of the collapse of spinning stars with $M \lesssim
197: 300 M_{\odot}$ have been performed and suggest how spinning black holes
198: may arise during core collapse (see, e.g. \citealt{heg02} for a review
199: and references). The simulations show that the fate of the collapse
200: depends critically on the mass, spin, metallicity and magnetic field of
201: the progenitor, as well as details of the equation of state and neutrino
202: transport, so it is not surprising that the issue is not resolved.
203:
204: There are also results for idealized versions of the general
205: relativistic collapse problem. \cite{shib02} followed the collapse of a
206: marginally unstable supermassive star in full general relativity. They
207: considered the case of a uniformly rotating star supported by radiation
208: pressure and spinning at the mass-shedding (maximal spin) limit.
209: (Mass-shedding is the likely situation by the time the star has cooled
210: and contracted quasistatically to the point of onset of collapse
211: \citep{baum99}, provided that the star can sustain solid body rotation
212: during the contraction phase.) They found that the final object is a
213: Kerr--like black hole surrounded by a disk of orbiting gaseous debris.
214: The final black hole mass and spin were determined to be $M_{\rm h}/M
215: \approx 0.9$ and $J_{\rm h}/M^2_{\rm h} = j = a/M \approx 0.75$, for an
216: arbitrary progenitor star mass $M$. The remaining mass goes into the
217: disk of mass $M_{\rm disk}/M \approx 0.1$. In fact, the final black
218: hole and disk parameters can be calculated analytically from the initial
219: stellar density and angular momentum distribution \citep{shap02}. The
220: results obtained here apply to the collapse of {\it any} marginally
221: unstable $n \approx 3$ polytrope spinning uniformly at mass-shedding.
222: Hence these results may be applicable to core collapse in very massive
223: stars $\gtrsim 300 {\rm M_{\odot}}$ and to collapsar models of
224: long-duration gamma-ray bursts \citep{mw99}. This work suggests,
225: therefore, that black holes formed during core collapse of massive stars
226: should be born rapidly rotating, but well below the Kerr limit.
227:
228: Black holes may also arise in other dynamical scenarios, like the
229: coalescence of binary neutron stars. The most detailed calculations of
230: binary neutron star mergers in full general relativity are the
231: hydrodynamic simulations of \cite{su02}. They considered the coalescence
232: of irrotational binaries (physically the most likely case), modeled as
233: equal-mass polytropes with adiabatic indices $\Gamma = 2$ and $2.25$,
234: and considered a range of initial masses below the maximum mass limit.
235: They followed the merger from the ISCO to coalescence. For intermediate
236: mass stars, the merged remnant is a differentially rotating,
237: ``hypermassive'' neutron star. For high mass stars, the remnant is a
238: rotating black hole. For the binaries which formed black holes, the
239: precollapse $J/M^2$ ranged from $0.9$ -- $1.0$, with the higher values
240: associated with the smaller masses. The black hole products have $J/M^2
241: \sim 0.8$ -- $0.9$ due to the loss of $\sim 10\%$ of the initial angular
242: momentum through gravitational radiation. In these calculations most of
243: the mass is conserved and goes into the black hole, and no disk forms
244: about the hole.
245:
246: \section{Spin-up By Major Mergers}\label{major}
247:
248: Black hole spin will also change when a black hole merges with a black
249: hole of comparable mass (a ``major merger''). The outcome of this
250: merger and the partitioning of energy and angular momentum between
251: internal and radiative degrees of freedom is an area of active research.
252: Here we summarize some current estimates.
253:
254: Consider binary black holes inspiralling due to gravitational radiation
255: from initially circular orbits (gravitational radiation reduces the
256: eccentricity of the orbits on a timescale short compared to the orbital
257: evolution timescale). Once the black holes reach the
258: ISCO, they will plunge together and merge on an orbital
259: timescale. The location of the ISCO, as well as the global parameters
260: characterizing the binary at this critical separation, are not known to
261: high precision except for black holes with test-particle companions.
262: Several different approaches have been formulated and have yielded
263: approximate solutions (for a recent review and references, see \citealt{bs03}).
264: These range from high-order post-Newtonian calculations to fully
265: nonlinear numerical solutions of the initial value vacuum Einstein
266: equations. We compare some of the available results for nonspinning
267: black holes in Table 1, adapted from \cite{bs03}.
268:
269: In constructing this table, we identify the mass of each black hole with
270: the irreducible mass
271: \begin{equation}
272: M_{\rm BH} = M_{\rm irr} = \left(\frac{A}{16 \pi}\right)^{1/2},
273: \end{equation}
274: where $A$ is the proper area of the black hole's event horizon
275: \citep{chr70}. The binding energy can then be defined as
276: \begin{equation}
277: E_b = M - 2 M_{\rm BH},
278: \end{equation}
279: where $M$ is the total (ADM) mass of the system measured at large
280: distance from the holes (see, e.g., \citealt{my74}). Values for the
281: nondimensional binding energy $\bar E_b \equiv E_b/\mu$, the orbital
282: angular velocity $\bar \Omega \equiv m \Omega$ and the angular
283: momentum $\bar J \equiv J/(\mu m)$ at the ISCO are listed. Here $\mu$
284: is the reduced mass, $\mu = M_{\rm BH}/2$, and $m$ is the sum of the black
285: hole masses, $m = 2 M_{\rm BH}$. The fractional losses from the system of
286: mass and angular momentum due to gravitational radiation emission during
287: the plunge are expected to be small (see, e.g., \citealt{kh99}), but precise
288: values await more reliable relativistic calculations. Meanwhile, a reasonable
289: first approximation is to assume that the final black hole
290: will have a mass and angular momentum nearly equal to the binary system
291: at the ISCO. With this assumption, the spin parameter of the final black
292: hole is given by
293: \begin{equation}
294: \label{spin}
295: {J \over M^2} = {\bar J \over {4 \left [1 + {\bar E_b \over 4} \right ]^2}}
296: \end{equation}
297: Results of the numerical calculations of \cite{coo94}, \cite{bau00}, and
298: \cite{ggb02} are tabulated, as well as the third-order Post-Newtonian
299: results of \cite{djs00}. The final black hole spin parameter computed
300: according to equation (\ref{spin}) is listed in the fifth column in the
301: table. A strict upper limit to the final spin parameter, $(J/M^2)_{\rm
302: max}$, is provided by the black hole area theorem and is listed in the
303: sixth column for comparison (see equation (\ref{max}) and discussion
304: below.) In the table we also include the analytical values for a test
305: particle orbiting a Schwarzschild black hole, $\bar E_b = \sqrt{8/9} - 1
306: = -0.0572$, $\bar J = 2 \sqrt{3} = 3.464$, $\bar \Omega = 1/6^{3/2} =
307: 0.0680$, with $J/M^2$ evaluated according to equation (\ref{spin}) and
308: $(J/M^2)_{\rm max}$ calculated according to equation (\ref{max}).
309:
310: Despite the differences in the ISCO calculations (for a recent detailed
311: comparison, see \citealt{cook03}), the values obtained for the expected
312: final spin parameter based on mass and angular momentum conservation are
313: all comparable and high, $J/M^2 \gtrsim 0.8$. Note that the range of
314: these estimated final spin parameters is far narrower than the range of
315: calculated ISCO orbital frequencies. Note also that these results are
316: for nonspinning holes.
317:
318: For spinning black holes one knows from test particle orbits that the
319: location of the ISCO depends strongly on spin. Most numerical
320: treatments of this problem use the conformal flatness approximation in
321: constructing the metric. This is problematic because isolated Kerr
322: black holes are not conformally flat, and forcing them to be so is
323: tantamount to adding in a compensating gravitational radiation field.
324: Conformally flat treatments of spinning binary black holes are therefore
325: contaminated with spurious gravitational radiation, and are probably not
326: as reliable as treatments of zero spin black holes.
327:
328: Nevertheless, the calculations of \cite{pfe00}, who generalized the
329: numerical calculation of \cite{coo94} by allowing for spin, but again
330: assumed conformal flatness, are revealing. They considered binaries
331: consisting of holes of equal mass and equal spin magnitude in circular
332: orbits. Their study was restricted to binaries in the range $--$0.50 to
333: $++$0.17. where the $+$ or $-$ sign denotes whether each hole is co- or
334: counter-rotating, respectively, and the numerical coefficient indicates
335: the magnitude of the spin parameter $J/M_{\rm BH}^2$ of each hole.
336: Assuming that the total mass and angular momentum are conserved during
337: the plunge from the ISCO, one finds from their numerical data that the
338: spin $J/M^2$ of the final, merged hole varies from 0.63 for the
339: $--$0.50 binary to 0.82 for the $++$0.17 binary.
340:
341: These values can be compared with the strict upper limit, $(J/M^2)_{\rm
342: max}$, provided by the area theorem, combined with the fact that the
343: final angular momentum cannot exceed the total angular momentum of the
344: system at the ISCO:
345: \begin{equation}
346: \label{max}
347: \left({J\over{M^2}}\right)_{\rm max} = {2 \over x_{\rm max}}
348: {\left ( 1 - {1 \over x^2_{\rm max}} \right )^{1/2}}
349: \end{equation}
350: where
351: \begin{equation}
352: \label{irr}
353: x_{max}^2 = 1 + {{\bar J}^2 \over
354: 4 \left( 1 + \sqrt {1 - \left (J/M_{\rm BH}^2 \right)} \right )^2}
355: \end{equation}
356: \citep{pct02}. The values of $(J/M^2)_{\rm max}$ reach 0.92 for the $--$0.50
357: binary and 0.97 for the $++$0.17 binary. We expect that the values given
358: by assuming that mass and angular momentum at the ISCO are conserved
359: during the plunge provide more realistic estimates. The reason is that
360: the amount of radiation allowed by the area theorem is far larger than
361: actually found by numerical computations, where such computations are
362: available. For example, the area theorem allows $29 \%$ of the
363: mass-energy to be radiated in a head-on collision of identical,
364: nonspinning black holes falling from rest at large separation, while
365: the numerical calculations yield $ 0.1 \%$ (\citealt{sma79,ann93}).
366:
367: Typically, therefore, the merger of two black holes of comparable mass
368: will immediately drive the spin parameter of the merged hole to $\gta
369: 0.8$.
370:
371: \section{Spin-Down by Minor Mergers}\label{minor}
372:
373: Recently, \cite{hb03} have considered the spin evolution of a black
374: hole due to mergers with smaller companions. Here we briefly revisit
375: the problem. Our approach is slightly simplified, does not use the
376: Fokker-Planck formalism, and evaluates the power law for spin decay
377: exactly in the limit of small $j$. Along the way, we provide a small
378: $j$ expansion for the radius and specific energy of the ISCO.
379:
380: Consider a merger between a large black hole of mass $M$ and a small
381: black hole of mass $m$, with $q \equiv m/M \ll 1$. The large black hole
382: has spin angular momentum $\bS$. The change in the total spin angular
383: momentum of the black hole is
384: \begin{equation}
385: \Delta (\bS^2) = (\bS + \Delta \bS)^2 - \bS^2 = 2 \bS \cdot \Delta \bS +
386: (\Delta \bS)^2.
387: \end{equation}
388: We are interested in how the spin evolves due to a large number of
389: mergers, so we take an ensemble average:
390: \begin{equation}
391: \<\Delta (\bS^2)\> = \< 2 \bS \cdot \Delta \bS \> + \< (\Delta \bS)^2 \>.
392: \end{equation}
393: The first term on the right is the ``resistive'' or ``dynamical
394: friction'' term, due to correlations between the black hole spin and its
395: change in angular momentum due to the accreted object. The second term
396: on the right is the ``random walk'' term. What is perhaps surprising is
397: that the dynamical friction term does not vanish.
398:
399: To evaluate $\Delta \bS$ we will assume that the merger occurs through
400: the slow, gravity-wave driven inspiral of the smaller hole onto the
401: larger hole. The larger hole is assumed to be slowly rotating, $j \ll
402: 1$, and the smaller hole is assumed to have isotropically distributed
403: orbital angular momentum $\bL$. The inspiral may be regarded as
404: progressing through a series of nearly circular orbits of fixed
405: inclination \citep{hug01}. The change in total angular momentum of the
406: larger black hole is then approximately $\bL$ evaluated at the ISCO (we
407: assume that the radiation of energy and angular momentum during the
408: plunge is negligible).
409:
410: With this picture in mind, the random walk term is
411: \begin{equation}\label{randwalk}
412: \< (\Delta \bS)^2 \> \approx \bL^2 = (m l M)^2,
413: \end{equation}
414: where $l \approx 2 \sqrt{3} + \order(j)$ is the specific orbital angular
415: momentum of a particle on the ISCO around a hole of unit mass.
416:
417: Now consider the dynamical friction term. One might naively expect
418: $\bL$ to be uncorrelated with $\bS$ if $\bL$ is isotropically
419: distributed, as we have assumed. But consider a particle with orbital
420: inclination angle $i$, $\mu = \cos(i)$. By expanding the fundamental
421: equations for the ISCO to lowest order in $j$ (see \citealt{hb03}
422: equations 1 and 2), one can show that
423: \begin{equation}
424: {L_z\over{m M}} = 2 \sqrt{3}\, \mu - {2\sqrt{2}\over{3}}\, j \mu^2 +
425: \order(j^2),
426: \end{equation}
427: \begin{equation}
428: \sE = {2 \sqrt{2}\over{3}} - {1\over{18 \sqrt{3}}}\, j \mu + \order(j^2),
429: \end{equation}
430: and
431: \begin{equation}
432: {r\over{M}} = 6 - {4 \sqrt{2}\over{\sqrt{3}}} \, j \mu + \order(j^2).
433: \end{equation}
434: Here $L_z$ is the component of orbital specific angular momentum
435: parallel to the black hole spin axis, $\sE$ is the particle specific
436: energy, and $r$ is the ISCO orbital radius measured in Boyer-Lindquist
437: coordinates. Evidently prograde ($\mu = 1$) orbits have orbital angular
438: momentum of smaller magnitude than retrograde ($\mu = -1$) orbits; their
439: ISCO lies ``closer'' to the black hole. Thus $\bS$ is correlated with
440: $\Delta \bS$.
441:
442: Using this result,
443: \begin{equation}
444: 2 \bS \cdot \Delta \bS = 2 j M^2 \Delta \bS_z \approx 2 j M^3 m
445: (2 \sqrt{3} \mu - {2 \sqrt{2}\over{3}} \mu^2 j).
446: \end{equation}
447: Averaging over $\mu$,
448: \begin{equation}
449: \< 2 \bS \cdot \Delta \bS \> \approx {1\over{2}} \int_{-1}^{1} d\mu
450: \left( 2 j M^3 m (2 \sqrt{3} \mu - {2 \sqrt{2}\over{3}} \mu^2 j)\right)
451: = -{ 4 \sqrt{2} M^3 m j^2\over{9}},
452: \end{equation}
453: since the first term vanishes under integration.
454:
455: We are now in a position to develop an expression for the evolution of
456: $j$:
457: \begin{equation}
458: {M \<\Delta (\bS^2)\>\over{\sE m}}
459: \approx {M d(j^2 M^4)\over{ d M }}
460: = {d (j^2 M^4)\over{ d\ln M }}
461: \end{equation}
462: where $\sE m$ is the change in mass of the large hole in each event,
463: and we can pass to the continuum limit only if $q \ll 1$. Using
464: equation (\ref{randwalk}),
465: \begin{equation}
466: {d (j^2 M^4)\over{ d\ln M }}
467: = {M\over{\sE m}} \left(
468: -{4 \sqrt{2}\over{9}} j^2 M^3 m + l^2 m^2 M^2 \right),
469: \end{equation}
470: and solving for $d \ln j/d\ln M$,
471: \begin{equation}
472: {d \ln j \over{d \ln M}} = - 2 - {2 \sqrt{2}\over{9 \sE}}
473: + {l^2 m\over{2 \sE M j^2}}.
474: \end{equation}
475: The first term describes conservation of spin angular momentum (Hughes
476: \& Blandford's ``doctrine of original spin''), the second term
477: dynamical friction, and the third term the random walk. Substituting
478: for $\sE$ and $l$,
479: \begin{equation}
480: {d \ln j \over{d \ln M}} = - {7\over{3}} + {9 m\over{\sqrt{2} M j^2}}.
481: \end{equation}
482: The final term can be ignored whenever $j^2/q \gg 27/(7 \sqrt{2})$; then
483: $j \sim M^{-7/3}$, in agreement with the \cite{hb03} result $j \sim
484: M^{-2.4}$. At late times, when the final term becomes comparable to the
485: first term, the hole will fluctuate around $j \sim q^{1/2}$. Thus minor
486: mergers with smaller objects with isotropically distributed orbital
487: angular momentum will spin down a hole.
488:
489: \section{Spin-up By Gas Accretion}\label{accretion}
490:
491: Once formed, black holes may grow through accretion of the surrounding
492: plasma. Accretion onto GBHCs in X-ray binaries is well established.
493: In the case of SMBHs,
494: appreciable growth of black hole seeds by gas accretion is
495: supported by the consistency between the total energy density in QSO
496: light and the BH mass density in local galaxies, adopting a reasonable
497: accretion rest-mass--to--energy conversion efficiency
498: (\citealt{sol82,yt02,erz02}). But quasars have been discovered out to
499: redshift $z \sim 6$, so it follows that the first SMBHs must have formed
500: by $z_{\rm BH} \gtrsim 6$ or within $t_{\rm BH} \lesssim 10^9$ yrs after
501: the Big Bang. This timescale provides a tight constraint on SMBH seed
502: formation scenarios. For example, it has been argued that if they
503: indeed grew by accretion, seeds of mass $ \gtrsim 10^5 {\rm ~M_{\odot}}$
504: must have formed by $ z \sim 9$ to have sufficient time to reach a mass
505: of $ \sim 10^9 {\rm ~M_{\odot}}$ \citep{gne01}. For a discussion of
506: plausible scenarios for forming these black hole seeds, see \cite{shap03}.
507:
508: Accretion will cause the spin $j = a/M = J/M^2$ of a black hole to
509: evolve in both magnitude and direction. In this section we will assume
510: that the orientation of the spin vector is fixed. Adopting Kerr-Schild
511: coordinates $t,r,\theta,\phi$, the angular momentum accretion rate is
512: \begin{equation}
513: \dot{J} \equiv \int d\theta d\phi\, \sqrt{-g}\, {T^{r}}_\phi
514: \end{equation}
515: where the integral is taken on the horizon, $g$ is the metric
516: determinant, and ${T^\mu}_\nu$ is the stress-energy tensor of the
517: accreting material.\footnote{We assume that the mass and angular
518: momentum of the disk is small compared to that of the black hole.} We
519: will consistently use Kerr-Schild coordinates here and below, but notice
520: that Kerr-Schild $r$ and $\theta$ are identical to the more familiar
521: Boyer-Lindquist $r$ and $\theta$. Mass-energy is accreted at a rate
522: \begin{equation}
523: \dot{M} = \dot{E} \equiv \int d\theta d\phi\, \sqrt{-g}\, {T^{r}}_t,
524: \end{equation}
525: and finally spin evolution is governed by
526: \begin{equation}
527: {d j\over{d t}} = {\dot{J}\over{M^2}} - {2 j \dot{E}\over{M}}.
528: \end{equation}
529: It is useful to define the dimensionless spinup parameter $s$,
530: \begin{equation}
531: s \equiv {d j\over{d t}} {M\over{\dot{M}_0}},
532: \end{equation}
533: where
534: \begin{equation}
535: \dot{M}_0 \equiv \int d\theta d\phi \, \sqrt{-g} \, \rho_0 u^r
536: \end{equation}
537: is the rest-mass accretion rate, $\rho_0$ is the rest-mass density, and
538: $u^r$ is the radial component of the plasma four-velocity. If $s < 0$
539: the black hole is spinning down.
540:
541: It was first noted by \cite{bar70} that a black hole accreting
542: through a cold disk of constant orientation could achieve maximal
543: rotation $j = 1$ in finite time. Bardeen assumed that $\dot{J} = l
544: \dot{M}_0 M$ and $\dot{E} = \sE \dot{M}_0$ where $l = u_{\phi}$ and $\sE =
545: -u_t$ are the mean angular momentum and energy per unit rest mass of a
546: particle on the ISCO, respectively. This is equivalent to the ``no
547: torque boundary condition'': no torque is exerted on the disk by the
548: plasma in the plunging region.
549:
550: \cite{tho74} noted that a thin disk inevitably radiates, and some of
551: this radiation will be accreted by the hole. Preferential accretion of
552: low angular momentum photons then limits the spin of the hole to a
553: maximum value $j = 0.998$.
554:
555: Exceptions to the \cite{tho74} result have been noted by several authors
556: in succeeding years. \cite{ajs78} pointed out that material accreted
557: from a thick, partially pressure supported disk would have {\it greater}
558: specific angular momentum than that accreted from a thin disk, and so
559: one might obtain $j > 0.998$. \cite{pg98} considered the evolution of a
560: relativistic hot flow using a viscous model for angular momentum
561: transport. They typically found spin equilibrium at $j \simeq 0.7$.
562: Finally, \cite{tho74}, quoting Bardeen, noted that magnetic fields could
563: connect material in the disk and the plunging region, and thus sharply
564: reduce the equilibrium spin.
565:
566: Magnetic fields in the plunging region have received some attention in
567: recent years, beginning with the work of \cite{kro99} and \cite{gam99},
568: based on earlier work by \cite{tak90} and others. The inflow model of
569: \cite{gam99} considered an inflowing plasma near the equatorial plane.
570: It assumed that a well-ordered magnetic field threaded both the plunging
571: region and the inner edge of the disk, and integrated the resulting one
572: dimensional steady flow equations. \cite{li00,li02} considered a model
573: in which a thin disk is connected to the black hole via high latitude
574: field lines, rather than through the inflow itself. \cite{ak00}
575: considered the implications of the magnetic accretion torques, including
576: the possible change in surface brightness of the disk.
577:
578: Recently it has become possible to study the dynamics of nonspherical
579: black hole accretion numerically in full general relativity
580: \citep{gam03,dvh03}. Given an appropriate set of initial and boundary
581: conditions, this allows one to calculate $dj/dt$ directly for a
582: nonradiative flow, within the magnetohydrodynamic (MHD) approximation.
583:
584: Here we consider a sequence of initial conditions in which $j$ alone is
585: varied and ask for which model is spin equilibrium, $s = 0$, achieved.
586: We use HARM \citep{gam03} to integrate the equations of ideal, general
587: relativistic magnetohydrodynamics (GRMHD) in a stationary, Kerr
588: background spacetime in a variant of Kerr-Schild coordinates.
589: Axisymmetry is assumed. Details of the scheme, the coordinate system,
590: and an extensive series of convergence tests may be found in
591: \cite{gam03}.
592:
593: Our initial conditions contain a \cite{fm76} torus with inner radius at
594: $r = 6 M$ and pressure maximum at $r = 12 M$. The initial magnetic
595: field is purely poloidal and field lines follow isodensity contours. It
596: is derived from a vector potential $A_\phi \propto {\rm
597: MAX}(\rho_0/\rho_{0,min} - 0.2, 0)$. The field is constrained to have a
598: minimum ratio of gas to magnetic pressure of $100$. Our equation of
599: state is $p = (\gamma - 1) u$, where $p$ is the gas pressure, $u$ is the
600: internal energy, and $\gamma = 4/3$.
601:
602: We set the inner boundary of the computational domain at $r_{in} = 0.98
603: r_h$ where $r_h = (1 + \sqrt{1 - j^2})$ is the event horizon radius.
604: Because the inner boundary is inside the event horizon it is causally
605: isolated from the rest of the flow. The outer boundary of the
606: computational domain is located at $r_{out} = 40 M$. This is distant
607: enough that the influence of the outer boundary on the inner accretion
608: flow ($r \sim M$) is negligible. Outflow boundary conditions (zero
609: order extrapolation of primitive variables with a switch forbidding
610: inflow) are used at the outer boundary. Tests indicate that all results
611: are independent of $r_{in}$ and $r_{out}$, unless $r_{in}$ is outside
612: the horizon. Our numerical resolution is $256 \times 256$ in zones
613: equally spaced in the coordinates $x_1 = \ln r$ and $x_2$ such that
614: $\theta = \pi x_2 + (1/2) (1 - H) \sin(2\pi x_2)$, where $H$ is a
615: parameter that gradually concentrates zones toward to the equator as $H$
616: is decreased from $1$ to $0$. Here we use $H = 0.3$. We find that
617: varying the resolution by a factor of $2$ in either direction leaves our
618: results unchanged.
619:
620: Figure 1 illustrates the outcome of a typical evolution of this
621: configuration around a black hole with $j = 0.75$. It shows the
622: logarithm of the density field in the $R = r \cos(\theta), Z = r
623: \sin(\theta)$ plane. The disk has become turbulent due to the
624: magnetorotational instability \citep{bh91}, and angular momentum
625: transport by MHD turbulence leads to gradual inflow along the equator.
626: The overall structure of the flow is similar to that observed by
627: \cite{dvhk}, with an evacuated funnel near the poles, outflow at
628: intermediate latitudes, a nearly-Keplerian equatorial torus, and a
629: plunging region between the torus and the event horizon.
630:
631: Figure 2 shows the evolution of $\dot{M}_0, \dot{E}/\dot{M}_0$, and
632: $\dot{J}/(\dot{M}_0 M)$ for a model with $j = 0.9375$. The cyan lines
633: show the values expected for the classical thin disk in which the
634: specific energy and angular momentum of accreted material is equal to
635: that of a particle on the ISCO, as in \cite{bar70}. While the specific
636: energy of accreted material is accurately predicted by the thin disk
637: model, the specific angular momentum is substantially lower. This is a
638: result of ordered magnetic fields in the plunging region which transport
639: angular momentum outward into the bulk of the disk.
640:
641: A similar suppression of accreted angular momentum has been observed in
642: fully relativistic MHD simulations of accretion onto black holes by
643: \cite{dvh} and \cite{dvhk}. In particular, Table 2 of the latter can be
644: used to estimate $s = 1.64$ for their model with $j = 0.5$. This is in
645: very close agreement with $s = 1.66$ for our model. Given that
646: different initial conditions and completely different numerical methods
647: were used for these two calculations, and that our calculation is two
648: dimensional while de Villiers et al.'s calculation is three dimensional,
649: the agreement is remarkable. We also note that the $j = 0.998$
650: calculation by \cite{dvhk} shows $s < 0$.
651:
652: Figure 3 shows the dimensionless spin evolution factor $s =
653: (dj/dt)(M/\dot{M}_0)$ for four separate models with $j = 0.5, 0.75,
654: 0.88, 0.97$. At $j = 0.97$ the hole is spinning down ($s < 0$). Figure
655: 4 shows $s(j)$, including models that were excluded from Figure 3 for
656: clarity. This sequence of accretion models reaches equilibrium ($s =
657: 0$) for $j \simeq 0.93$. This suggests that magnetic interactions can
658: lead to spin equilibration at lower $j$ than the canonical $j = 0.998$
659: of \cite{tho74}.
660:
661: We are not suggesting that spin equilibrium is {\it always} reached at
662: $j \simeq 0.93$. Models with different initial conditions may produce
663: different results. For example, the models we have considered here have
664: a ratio of scale height to local radius $H/r \sim 0.2$--$0.3$ at
665: pressure maximum. Thinner disks are likely to produce different
666: results; indeed, in a later publication we will present results from the
667: self-consistent evolution of a thin disk that closely match the
668: predictions of a classical thin disk model. But thinner disks imply
669: lower accretion rates, so thin disk accretion will have lower weight in
670: determining the black hole spin than thick disk accretion over a
671: comparable timescale. To sum up, what we have shown is that there exist
672: self-consistent GRMHD models with $j \gta 0.93$ that are unambiguously
673: spinning down the black hole: accretion need not necessarily lead to
674: near-maximal rotation.
675:
676: \section{Conclusions}\label{finale}
677:
678: In this paper, we have considered astrophysical processes that influence
679: the spin evolution of black holes. Fully relativistic collapse
680: calculations suggest that the initial spin of a newborn black hole $j
681: \lesssim 0.75$--$0.9$, where the upper limit applies to the collapse of
682: maximally and uniformly rotating massive stars. The outcome of
683: subsequent mergers with black holes of comparable mass is not yet fully
684: understood, but current best estimates suggest a final spin $j \sim
685: 0.8$--$0.9$ (see Table 1). Mergers with black holes of much smaller
686: mass can be treated in the test-particle approximation and, following
687: \cite{hb03}, we have presented an argument showing that such mergers
688: tend to spin down the black hole to $j \ll 1$, provided that the small
689: black holes have isotropically distributed orbital angular momentum.
690:
691: We have also presented results from fully relativistic
692: magnetohydrodynamic models of accretion onto a rotating hole. These
693: show that, at least for the particular series of thick disk models we
694: consider, spin equilibrium is reached at $j \approx 0.93$. This
695: demonstrates that accretion need not lead to near-maximal rotation. Our
696: models have a ratio of scale height to local radius $H/r \sim
697: 0.2$--$0.3$, and thus correspond to near-Eddington accretion rates.
698: Accretion at lower rates (through thinner disks) may be capable of
699: producing higher spin, provided that the orbital angular momentum of the
700: accreting material remains aligned with the black hole spin. If the
701: orbital plane of the accreting material varies, as seems likely (see the
702: discussion of \citealt{np98}), even thin disk accretion may be unable to
703: produce $j \approx 1$.
704:
705: All these results suggest that near-maximal rotation of black holes is
706: neither necessary nor likely. Black hole spins $j \sim 0.7$ -- $0.95$
707: are produced in a variety of scenarios. This corresponds to thin disk
708: radiative efficiencies of $10\%$--$19\%$, which is broadly consistent
709: with the radiative efficiencies required by Soltan-type arguments
710: \citep{yt02,erz02}. Such modest spins are also not in conflict with the
711: idea that radio galaxies are powered by black hole spindown. The
712: \cite{bz77} luminosity of a black hole scales as $j^2 (B^r)^2$ where
713: $B^r$ is the mean radial magnetic field on the event horizon. Unless
714: $B^r$ is a sharply increasing function of $j$, the Blandford-Znajek
715: luminosity of a black hole with $j \simeq 0.9$ is not very different
716: from that of a nearly maximally rotating black hole. Only if black
717: holes were built up mainly through thin disk (sub-Eddington) accretion
718: of material in a fixed orbital plane would near-maximal rotation be the
719: norm.
720:
721: \acknowledgments
722:
723: This work was supported in part by NSF ITR Grant PHY 02-05155, NSF
724: PECASE Grant AST 00-93091, NSF Grant PHY-0090310, NASA Grant NAG5-8418
725: and NASA Grant NAG5-10881 to the University of Illinois at
726: Urbana-Champaign. JCM was supported by a NASA GSRP fellowship. CFG is
727: pleased to acknowledge a National Center for Supercomputing Applications
728: (NCSA) faculty fellowship. Some of the computations described here were
729: performed at NCSA.
730:
731: \clearpage
732:
733: \begin{thebibliography}{}
734:
735: \bibitem[Abramowicz, Jaroszinski, \& Sikora(1978)]{ajs78} Abramowicz,
736: M., Jaroszinski, M., \& Sikora, M. 1978, AA, 63, 221
737:
738: \bibitem[Agol \& Krolik(2000)]{ak00} Agol, E.~\& Krolik, J.~H.\ 2000,
739: \apj, 528, 161
740:
741: \bibitem[Anninos(1993)]{ann93} Anninos, P., Hobill, D., Seidel, E.,
742: Smarr, L., Suen, W.-M. 1993, Phys. Rev. Lett, 71, 2851
743:
744: \bibitem[Balbus \& Hawley(1991)]{bh91} Balbus, S.A., \& Hawley, J.F.
745: 1991, ApJ, 376, 214
746:
747: \bibitem[Bahcall et al.(1990)]{bahc90} Bahcall, S., Lynn, B.W., \&
748: Selipsky, S.B. 1990, ApJ, 362, 251
749:
750: \bibitem[Bardeen(1970)]{bar70} Bardeen, J.M. 1970, Nature, 226, 64
751:
752: \bibitem[Baumgarte(2000)]{bau00} Baumgarte, T.W. 2000, Phys. Rev. D.,
753: 62, 084020
754:
755: \bibitem[Baumgarte \& Shapiro(1999)]{baum99} Baumgarte, T.W., \& Shapiro,
756: S.L. 2003, ApJ, 526, 941
757:
758: \bibitem[Baumgarte \& Shapiro(2003)]{bs03} Baumgarte, T.W., \& Shapiro,
759: S.L. 2003, Phys. Rep. 376, 41
760:
761: \bibitem[Blandford \& Znajek(1977)]{bz77} Blandford, R.D., \& Znajek, R.
762: 1977, MNRAS, 179, 433
763:
764: \bibitem[Christodoulou(1970)]{chr70} Christodoulou, D. 1970 Phys. Rev. Lett., 25, 1596
765:
766: \bibitem[Cook(1994)]{coo94} Cook, G.B. 1994, Phys. Rev. D., 50, 5025
767:
768: \bibitem[Cook(2003)]{cook03} Cook, G. B., talk presented at KITP Conference
769: on Gravitational Interaction of Compact Objects, May 13 2003,
770: ($http://online.itp.ucsb.edu/online/gravity\_c03/cook/$).
771:
772: \bibitem[Damour et al.(2000)]{djs00} Damour, T., Jaranowski, P.,
773: \& Sch\"afer, G. 2000, Phys. Rev. D. 62, 084011
774:
775: \bibitem[DeDeo \& Psaltis(2003)]{dp03} DeDeo, S., \& Psaltis, D. 2003,
776: astro-ph/0302095
777:
778: \bibitem[Dreyer et al.(2003)]{drey03} Dreyer, O., Kelly, B., Krishnan,
779: B., Finn, L.S., Garrison, D., Lopez-Aleman, R. 2003, gr-qc/0309007
780:
781: \bibitem[Gezari et al.(2002)]{gez02} Gezari, S., Ghez, A.~M., Becklin,
782: E.~E., Larkin, J., McLean, I.~S., \& Morris, M.\ 2002, \apj, 576, 790
783:
784: \bibitem[Gierli{\' n}ski, Macio{\l}ek-Nied{\' z}wiecki, \&
785: Ebisawa(2001)]{gme01} Gierli{\' n}ski, M., Macio{\l}ek-Nied{\' z}wiecki,
786: A., \& Ebisawa, K.\ 2001, \mnras, 325, 1253
787:
788: \bibitem[Grandecl\'ement et al.(2002)]{ggb02} Grand\-cl\'{e}ment, P.,
789: Gourgoulhon, E., \& Bonazzola, S., 2002, Phys. Rev. D., 65, 044021
790:
791: \bibitem[Heger et al.(2002)]{heg02} Heger, A., Woosley, S., Baraffe, I.,
792: \& Abel, T.\ 2002, Lighthouses of the Universe: The Most Luminous
793: Celestial Objects and Their Use for Cosmology Proceedings of the
794: MPA/ESO/, p.~369, 369
795:
796: \bibitem[Elvis, Risaliti, \& Zamorani(2002)]{erz02} Elvis, M.,
797: Risaliti, G., \& Zamorani, G.\ 2002, \apjl, 565, L75
798:
799: \bibitem[Evans \& Hawley(1988)]{eh88} Evans, C.R., \& Hawley, J.F. 1988,
800: \apj, 332, 659
801:
802: \bibitem[Fabian et al.(2002)]{fab02} Fabian, A.~C.~et al.\
803: 2002, \mnras, 335, L1
804:
805: \bibitem[Fishbone \& Moncrief(1976)]{fm76} Fishbone, L.G., \& Moncrief,
806: V. 1976, ApJ, 207, 962
807:
808: \bibitem[Font(2000)]{font00} Font, J. A., 2000, Liv. Rev. in Rel., 3,
809: 2000-2font
810:
811: \bibitem[Flanagan \& Hughes(1998)]{flan98} Flanagan, E. E., \& Hughes, S. A.,
812: 1998, \prd, 57, 4535
813:
814: \bibitem[Gammie(1999)]{gam99} Gammie, C.F. 1999, ApJL, 522, L57
815:
816: \bibitem[Gammie, McKinney, \& T{\' o}th(2003)]{gam03} Gammie, C.~F.,
817: McKinney, J.~C., \& T{\' o}th, G.\ 2003, \apj, 589, 444
818:
819:
820: \bibitem[Gnedin(2001)]{gne01} Gnedin, O. Y. 2001, Class. \& Quan. Grav., 18, 3983
821:
822: \bibitem[Harten et al.(1983)]{hll83} Harten, A., Lax, P.D., \& van Leer,
823: B. 1983, SIAM Rev. 25, 35
824:
825: \bibitem[Hawley, Smarr, \& Wilson(1984)]{hsw} Hawley, J.F., Smarr, L.L.,
826: \& Wilson, J.R. 1984, ApJS, 55, 211
827:
828: \bibitem[Hughes(2001)]{hug01} Hughes, S.~A.\ 2001, \prd, 64, 64004
829:
830: \bibitem[Hughes \& Blandford(2003)]{hb03} Hughes, S.~A.~\& Blandford,
831: R.~D.\ 2003, \apjl, 585, L101
832:
833: \bibitem[Khanna et al.(1999)]{kh99} Khanna, G., Baker, J., Gleiser, R.,
834: Laguna, P., Nicasio, C., Nollert, H-P., Price, R., \& Pullin, J.
835: 1999, \prl, 83, 3581
836:
837: \bibitem[Krolik(1999)]{kro99} Krolik, J.~H.\ 1999, \apjl,
838: 515, L7
839:
840: \bibitem[Laor(1991)]{lao91} Laor, A.\ 1991, \apj, 376, 90
841:
842: \bibitem[Li(2000)]{li00} Li, L.\ 2000, \apjl, 533, L115
843:
844: \bibitem[Li(2002)]{li02} Li, L.\ 2002, \apj, 567, 463
845:
846: \bibitem[MacFadyen \& Woosley(1999)]{mw99} MacFadyen, A. I. \& Woosley, S. E.\
847: 1999, \apj, 524, 262
848:
849: \bibitem[Magorrian et al.(1998)]{mag98} Magorrian, J.~et al.\ 1998, \aj,
850: 115, 2285
851:
852: \bibitem[Maoz(1998)]{mao98} Maoz, E.\ 1998, \apjl, 494, L181
853:
854: \bibitem[McClintock \& Remillard(2003)]{mr03} McClintock, J.E., \&
855: Remillard, R.A. 2003, astro-ph/0306213
856:
857: \bibitem[Merritt \& Eckers(2002)]{me02} Merritt, D. \& Eckers, R. D.
858: 2002, Science, 297, 1310
859:
860: \bibitem[Miller et al.(2002)]{mil02a} Miller, J.~M.~et al.\
861: 2002, \apj, 578, 348
862:
863: \bibitem[Miller et al.(2002)]{mil02b} Miller, J.~M.~et al.\
864: 2002, \apjl, 570, L69
865:
866: \bibitem[Miller et al.(2001)]{mil01} Miller, J.~M.~et al.\
867: 2001, \apj, 563, 928
868:
869: \bibitem[Miyoshi et al.(1995)]{miy95} Miyoshi, M., Moran,
870: J., Herrnstein, J., Greenhill, L., Nakai, N., Diamond, P., \& Inoue, M.\
871: 1995, \nat, 373, 127
872:
873: \bibitem[\'{O}Murchadha and York (1974)]{my74} \'{O}Murchadha, N.
874: \& York Jr., J. W., 1974 Phys. Rev. D. 10, 2345
875:
876: \bibitem[Natarajan \& Pringle(1998)]{np98} Natarajan, P.~\& Pringle,
877: J.~E.\ 1998, \apjl, 506, L97
878:
879: \bibitem[Pfeiffer et al.(2000)]{pfe00} Pfeiffer, H. P., Cook, G.B.,
880: \& Teukolsky, S. A. 2002, Phys. Rev. D., 62, 104018
881:
882: \bibitem[Pfeiffer, Cook, \& Teukolsky(2002)]{pct02} Pfeiffer,
883: H.~P., Cook, G.~B., \& Teukolsky, S.~A.\ 2002, \prd, 66, 24047
884:
885: \bibitem[Popham \& Gammie(1998)]{pg98} Popham, R.~\& Gammie, C.~F.\
886: 1998, \apj, 504, 419
887:
888: \bibitem[Reynolds \& Nowak(2002)]{rn02} Reynolds, C.S., \& Nowak, M.A.
889: 2002, Phys. Rep. 377, 389
890:
891: \bibitem[Sch{\" o}del et al.(2002)]{sch02} Sch{\" o}del, R.~et al.\
892: 2002, \nat, 419, 694
893:
894: \bibitem[Shahbaz et al.(1999)]{sha99} Shahbaz, T., van der Hooft, F.,
895: Casares, J., Charles, P.~A., \& van Paradijs, J.\ 1999, \mnras, 306, 89
896:
897: \bibitem[Shapiro (2003)]{shap03} Shapiro, S. L. 2003, in Carnegie
898: Observatories Astrophysics Series, Vol 1: Coevolution of Black Holes and
899: Galaxies, ed. L. C. Ho (Cambridge: Cambridge Univ. Press), in press
900: astro-ph/ 0304202.
901:
902: \bibitem[Shapiro \& Teukolsky(1983)]{st83} Shapiro, S.L., \& Teukolsky,
903: S.A. 1983, Black Holes, White Dwarfs, and Neutron Stars: The Physics of
904: Compact Objects (New York: Wiley)
905:
906: \bibitem[Shapiro \& Shibata(2002)]{shap02} Shapiro, S.~L.~\& Shibata, M.\
907: 2002, \apj, 577, 904
908:
909: \bibitem[Shibata \& Shapiro(2002)]{shib02} Shibata, M.~\& Shapiro, S.~L.\
910: 2002, \apjl, 572, L39
911:
912: \bibitem[Shibata \& Uryu(2002)]{su02} Shibata, M. \& Uryu, K.
913: 2002, Prog. Theor. Phys., 107, 265
914:
915: \bibitem[Smarr(1979)]{sma79} Smarr, L. 1979, in Sources of
916: Gravitational Radiation, ed. L. Smarr (Cambridge University Press, Cambridge)
917:
918: \bibitem[Soltan(1982)]{sol82} Soltan, A.\ 1982, \mnras, 200, 115
919:
920: \bibitem[Strohmayer(2001)]{str01} Strohmayer, T.~E.\ 2001,
921: \apjl, 552, L49
922:
923: \bibitem[Takahashi et al.(1990)]{tak90} Takahashi, M. , Nitta, S. ,
924: Tatematsu, Y. \& Tomimatsu, A. 1990, \apj, 363, 206
925:
926: \bibitem[Tanaka et al.(1995)]{tan95} Tanaka, Y.~et al.\
927: 1995, \nat, 375, 659
928:
929: \bibitem[Thorne(1995)]{tho95} Thorne, K.~S.\ 1995, Seventeeth Texas
930: Symposium on Relativistic Astrophysics and Cosmology, 759, 127
931:
932: \bibitem[Thorne(1974)]{tho74} Thorne, K.~S.\ 1974, \apj, 191, 507
933:
934: \bibitem[De Villiers \& Hawley(2003)]{dvh03} De Villiers, J.~\& Hawley,
935: J.~F.\ 2003, \apj, 589, 458
936:
937: \bibitem[de Villiers \& Hawley(2003)]{dvh} de Villiers, J.-P., \&
938: Hawley, J.F., 2003, astro-ph/0303241
939:
940: \bibitem[de Villiers et al.(2003)]{dvhk} de Villiers, J.-P.,
941: Hawley, J.F., \& Krolik, J.H. 2003, astro-ph/0307260
942:
943: \bibitem[Wald(1984)]{wal84} Wald, R.~M.\ 1984, Chicago:
944: University of Chicago Press, 1984, 313
945:
946: \bibitem[Yu \& Tremaine(2002)]{yt02} Yu, Q.~\& Tremaine, S.\ 2002,
947: \mnras, 335, 965
948:
949: \bibitem[Zhang, Cui, \& Chen(1997)]{zcc97} Zhang, S.~N.,
950: Cui, W., \& Chen, W.\ 1997, \apjl, 482, L155
951:
952:
953: \end{thebibliography}
954:
955: \clearpage
956:
957: \begin{figure}
958: \epsscale{0.5}
959: \plotone{f1.eps}
960: \caption{
961: A snapshot from an evolution of a weakly magnetized Fishbone-Moncrief
962: torus around a $j = 0.75$ black hole. Color corresponds to
963: $\log(\rho_0)$. Red is high rest-mass density, and black is low.
964: }
965: \end{figure}
966:
967: \begin{figure}
968: \plotone{f2.eps}
969: \caption{
970: Evolution of the mass, energy, and angular momentum accretion rate for a
971: weakly magnetized Fishbone-Moncrief tori around a black hole with $j =
972: 0.9375$.
973: }
974: \end{figure}
975:
976: \begin{figure}
977: \plotone{f3.eps}
978: \caption{
979: Evolution of $s = (dj/dt)(M/\dot{M}_0)$ for a series of four
980: Fishbone-Moncrief tori.
981: }
982: \end{figure}
983:
984: \begin{figure}
985: \plotone{f4.eps}
986: \caption{
987: Time-averaged value of $s$ for a sequence of Fishbone-Moncrief tori.
988: The squares indicate data points from simulations, while the thin line
989: indicates values expected for a thin disk.
990: }
991: \end{figure}
992:
993: \clearpage
994:
995: \begin{table}
996: \begin{center}
997: \caption{Representative values for nonspinning binary black holes}
998: \begin{tabular}{llllll}
999: \tableline
1000: \tableline
1001: Reference & $\bar E_b$\tablenotemark{a} & $ \bar
1002: J$\tablenotemark{b} &$\bar \Omega$\tablenotemark{c}
1003: &$ J/M^2$\tablenotemark{d} &$(J/M^2)_{\rm max}$\tablenotemark{e} \\
1004: \hline
1005: Schwarzschild & -0.0572 & 3.464 & 0.068
1006: & 0.8913 & 0.9897\\
1007: Cook(1994) & -0.09030 & 2.976 & 0.172
1008: & 0.7788 & 0.9578\\
1009: Baumgarte (2000) & -0.092 & 2.95 & 0.18
1010: & 0.773 & 0.955\\
1011: Grandclement et al. (2002) & -0.068 & 3.36 & 0.103
1012: & 0.869 & 0.985\\
1013: Damour et al. (2000) & -0.0668 & 3.27 & 0.0883
1014: & 0.846 & 0.980\\
1015: \tableline
1016: \end{tabular}
1017: \tablenotetext{a}{Binding energy per unit reduced mass at the ISCO.}
1018: \tablenotetext{b}{Angular momentum per unit reduced mass at the ISCO.}
1019: \tablenotetext{c}{Orbital angular velocity at the ISCO.}
1020: \tablenotetext{d}{Estimated spin parameter of final black hole.}
1021: \tablenotetext{e}{Maximum spin parameter of final black hole (see
1022: text).
1023: }
1024: \end{center}
1025: \end{table}
1026:
1027: \clearpage
1028:
1029: \end{document}
1030:
1031: \bye
1032:
1033:
1034: