1: % To appear in proceedings of "Massive Stars in Interacting Binaries",
2: % eds. Nicole St-Louis & Tony Moffat
3:
4: \documentclass[11pt,twoside]{article}
5: \usepackage{asp2004}
6: \usepackage{psfig}
7: \pagestyle{myheadings}
8:
9: \setcounter{equation}{0}
10: \setcounter{figure}{0}
11: \setcounter{section}{0}
12: \setcounter{table}{0}
13:
14: \begin{document}
15: \title{Stellar Collisions and Black Hole Formation in Dense Star
16: Clusters}
17:
18: \author{Stephen L.\ W.\ McMillan}
19: \affil{Department of Physics, Drexel University, Philadelphia, PA
20: 19104, USA}
21: \author{Simon F. Portegies Zwart}
22: \affil{Astronomical Institute `Anton Pannekoek' and Institute for
23: Computer Science, University of Amsterdam, Kruislaan 403, The
24: Netherlands
25: }
26:
27: \begin{abstract}
28: Close encounters and physical collisions between stars in young dense
29: clusters may lead to the formation of very massive stars and black
30: holes via runaway merging. We examine critically some details of this
31: process, using N-body simulations and simple analytical estimates to
32: place limits on the cluster parameters for which it expected to occur.
33: For small clusters, the mass of the runaway is effectively limited by
34: the total number of high-mass stars in the system. For sufficiently
35: dense larger clusters, the runaway mass is determined by the fraction
36: of stars that can mass segregate to the cluster core while still on
37: the main sequence. The result is in the range commonly cited for
38: intermediate-mass black holes, such as that recently reported in the
39: Galactic center.
40: \end{abstract}
41:
42: \section{Introduction}
43:
44: The past decade has seen the discovery of a large number of massive
45: young star clusters throughout the local universe. These systems are
46: large and young enough that they contain statistically significant
47: numbers of massive stars, affording us key insights into their initial
48: mass function and structural properties. Of greatest interest to
49: dynamicists are those systems in which stellar dynamical time scales
50: are short enough that the cluster can undergo significant structural
51: change during the lifetimes of the most massive stars. In such
52: clusters, dynamical evolution opens up entirely new avenues for
53: stellar and binary evolution, allowing the formation of stellar
54: species completely inaccessible by standard stellar and binary
55: evolutionary pathways.
56:
57: From this perspective, the clusters listed by Portegies Zwart and
58: McMillan (these proceedings) represent an ideal combination of
59: properties, having ages of less than a few million years and
60: relaxation times of less than a few tens of millions of years. In
61: these clusters, dynamical evolution, traditionally regarded as a
62: ``slow'' process, actually occurs much more rapidly than the stellar
63: evolution of even the most massive stars. Thus the dynamics controls
64: the early phases of these stars' lives. We focus here on perhaps the
65: most dramatic environmental modification of standard stellar
66: evolution---repeated physical collisions between stars. The scenario
67: described below is fast becoming the ``standard'' paradigm by which
68: runaway stellar mergers might occur, perhaps forming intermediate-mass
69: black holes (IMBHs) in sufficiently dense systems.
70:
71: The possibility of multiple mergers of massive stars in dense stellar
72: systems was first demonstrated in N-body simulations of R\,136 by
73: Portegies Zwart et al.~(1999). Subsequently Ebisuzaki et al.~(2001)
74: suggested that the ultraluminous X-ray source M82-X1 might be the
75: result of such a process; this possibility was explored in detail by
76: Portegies Zwart et al.~(2004). Portegies Zwart \& McMillan (2002) and
77: McMillan \& Portegies Zwart (2003) have explored the possibility of
78: IMBH formation in young clusters in the Galactic Center. Recently,
79: G\"urkan et al.~(2004) have reaffirmed the basic process using
80: Monte-Carlo simulations, and have carried out systematic studies of
81: runaway mergers in galactic nuclei.
82:
83: In this paper we consider how a young star cluster might come to be in
84: such a high-density state, and look critically at the key physical
85: processes needed for a runaway merger to occur. We then turn briefly
86: to the possibility of an IMBH in the center of the Milky Way Galaxy
87: (Maillard et al.~2004), showing how theory and recent observations may
88: provide a consistent picture of the Galactic center.
89:
90: %---------------------------------------------------------------------
91:
92: \section{Stellar Collisions}
93:
94: Consider a massive object of mass $M=mM_\odot$ and radius $R=rR_\odot$
95: moving through a field of background stars of total mass density
96: $\rho=10^6\,\rho_6\,M_\odot/{\rm pc}^3$ and velocity dispersion
97: $v=10\,v_{10}\,{\rm km/s}$. If $M$ and $R$ are large compared to the
98: masses and radii of other stars, and all velocities are small enough
99: that gravitational focusing dominates the total cross section, the
100: object's collision cross section is $\sigma \approx 2\pi G M R / v^2$,
101: nearly independent of the properties of the other stars. The rate of
102: increase of the object's mass due to collisions then is
103: \begin{equation}
104: \frac{dM}{dt} ~\approx~ \rho\sigma v
105: ~\approx~ 2\pi G M R \rho / v
106: ~=~ 6\times10^{-11}\, m \, r \,
107: \rho_6 \, v_{10}^{-1}~ M_\odot/{\rm yr}\,.
108: \end{equation}
109: If the object initially has $m=100\,m_{100}$ and we adopt a simple
110: mass--radius relation $r=3\,m^{1/2}$, then for the object to reach
111: $m\gg10^3$ in 3 Myr (to form an IMBH within the lifetime of a massive
112: star), the local density must satisfy
113: \begin{equation}
114: \rho_6 \ \ga\ 350 ~ m_{100}^{-1/2}\ v_{10}
115: ~=~ \rho_{crit},\ \ {\rm say}\,.
116: \end{equation}
117: Such a density is much higher than the mean density of any known star
118: cluster, young or old. For comparison, the average density of the
119: Arches cluster is $\rho_6\sim0.6$, that of a fairly compact globular
120: cluster is $\rho_6\sim0.01$, while even the most concentrated globular
121: cluster cores have $\rho_6\la 1-10$.
122:
123: Mergers would be enhanced if the cluster were born very centrally
124: concentrated, as suggested by Portegies Zwart et al.~(2004) and
125: Merritt et al.~(2004). As a simple limiting model of such a cluster,
126: consider the nearly isothermal system of total mass $M_c$ and
127: half-mass radius $r_h$, described by the density profile
128: \begin{eqnarray}
129: \rho(r) &=& \frac{M_c}{8\pi r_h r^2}\,,\\
130: M(r) &=& {\textstyle\frac12}M_c\left(\frac{r}{r_h}\right)\,,
131: \end{eqnarray}
132: for $0 \le r \le 2r_h$. Densities exceeding $\rho_{crit}$ are found
133: for $r<r_{crit}$, where
134: \begin{equation}
135: r_{crit} ~=~ \sqrt{\frac{M_c}{8\pi r_h\rho_{crit}}}
136: ~=~ 2.3\times10^{-3} \ v_{10}^{1/2} m_{100}^{1/4}\ \
137: {\rm pc}\,,
138: \end{equation}
139: where the cluster velocity dispersion is $v = \sqrt{GM_c/2r_h}$.
140: However, the total mass contained within this radius is just
141: \begin{equation}
142: M_{crit} \approx 50\ v_{10}^{5/2} \ m_{100}^{1/4}\ M_\odot\,.
143: \end{equation}
144: We conclude that, for reasonable cluster parameters, there is too
145: little initial mass in the high-density region to accomplish the task
146: of forming a $\sim10^3\,M_\odot$ object in the time available.
147:
148: %---------------------------------------------------------------------
149:
150: \section{Cluster Dynamics}
151:
152: Thus collisions in a static cluster core cannot lead to the formation
153: of an ultramassive object. However, cluster dynamical evolution can
154: result in conditions much more favorable for a runaway merger to
155: occur. The evolution of a cluster is governed by its half-mass
156: relaxation time, the time scale on which two-body encounters transport
157: energy around the system:
158: \begin{equation}
159: t_{rh}\ \approx\ \frac{0.14\, M_c^{1/2} r_h^{3/2}}
160: {G^{1/2} \langle m\rangle \ln\Lambda}
161: \ \approx\ 0.5\ v_{10}^3 \,/\, \bar{\rho}_6 \ {\rm Myr}
162: \label{trelax}
163: \end{equation}
164: (Heggie \& Hut 2003). Here, $N$ is the number of stars in the system,
165: $\langle m\rangle = M_c/N$ is the mean stellar mass, taken here to be
166: $0.5M_\odot$, $\bar{\rho} = 3M_c/8\pi r_h^3 =
167: 10^6\,\bar{\rho}_6\,M_\odot\,{\rm pc}^{-3}$ is the mean cluster
168: density, and $\ln\Lambda\sim \ln(0.1 N)\sim10$. For an equal-mass
169: system, the time scale for dynamical evolution---the core collapse
170: time---is about $15 t_{rh}$, too long to cause significant structural
171: change within a few million years. However, the presence of even a modest
172: range in masses greatly accelerates the process of core collapse
173: (Spitzer 1987). The time scale for a star of mass $m$ to sink to the
174: cluster center as equipartition reduces its velocity is
175: \begin{equation}
176: t_s(M) \ \sim\ \frac{\langle m\rangle}{m}\, t_r\,,\label{tseg}
177: \end{equation}
178: where $t_r\sim t_{rh}\,\bar{\rho}/\rho$ is the local relaxation time.
179:
180: Portegies Zwart \& McMillan (2002) find that the most massive ($m \ga
181: 20 M_\odot$) stars segregate rapidly to the cluster center, forming a
182: dense stellar subcore on a time scale $t_{cc}\sim0.2 t_{rh}$. A
183: central density increase of 2--3 orders of magnitude is typical,
184: boosting even a relatively low-density core into the range where
185: collisions become common, and greatly increasing the supply of raw
186: material to form a collision runaway. In systems with $t_{cc} \la 5$
187: Myr ($t_{rh}\la 25$ Myr), essentially all the massive stars in the
188: cluster reach the center before exploding as supernovae, and hence can
189: participate in the runaway process. The Arches and Westerlund I fall
190: into this category; the Quintuplet, NGC 3603, and R\,136 all come
191: close. In these cases, the maximum mass of the runaway is limited
192: primarily by the total number of massive stars in the
193: system---anywhere from a few percent to several tens of percent of the
194: total, depending on the cluster mass function.
195:
196: In less dense or more massive clusters, the longer half-mass
197: relaxation time means that only a fraction of the massive stars
198: initially present in the system can reach the center in the time
199: available, but the total supply of mass may still ensure that a
200: runaway can occur. We can estimate the amount of mass made available
201: by mass segregation as follows. Again adopting an isothermal model to
202: simplify the calculation, and taking the segregation time scale from
203: Eq.~\ref{tseg}, it is easily shown that a star of mass $m$ can sink to
204: the center within time $T$ if its initial distance $r$ from the
205: cluster center satisfies
206: \begin{equation}
207: r \la r_s(m) = r_h
208: \left(\frac{m}{\langle m\rangle}\right)^{1/2}\,
209: \left(\frac{T}{t_{rh}}\right)^{1/2}\,.
210: \end{equation}
211: The fraction of stars of mass $m$ satisfying this relation is
212: \begin{equation}
213: f(m) = {\textstyle\frac12}
214: \left(\frac{m}{\langle m\rangle}\right)^{1/2}\,
215: \left(\frac{T}{t_{rh}}\right)^{1/2}\,.
216: \end{equation}
217: Choosing (again for simplicity) a mass function $dn/dm \sim m^{-2}$
218: for $0.1 M_\odot < m < 100 M_\odot$, we determine the total stellar
219: mass potentially available for mergers as
220: \begin{equation}
221: M_s = \int_{m_{min}}^{100M_\odot}\,\frac{dn}{dm}\,m\,f(m)\ dm\,,
222: \end{equation}
223: where the lower mass limit $m_{min}$ is somewhat greater than the mean
224: stellar mass ($0.7\,M_\odot$ here), but is otherwise unimportant so
225: long as it is much less than the upper limit of $100\,M_\odot$. The
226: result for the chosen mass function is
227: \begin{eqnarray}
228: M_s &=& 1.7 \left(\frac{T}{t_{rh}}\right)^{1/2} M_c\nonumber\\
229: &=& 1.5 \times 10^4 M_\odot\ v_{10}^{3/2}\
230: \left(\frac{T}{3 {\rm\, Myr}}\right)^{1/2}\,,
231: \nonumber
232: \end{eqnarray}
233: % Thus dynamical effects increase the available mass
234: comfortably above the value needed for IMBH formation.
235: % Note, incidentally, that the available mass is not directly
236: % dependent on the cluster mass, but only on the velocity dispersion.
237:
238: We have assumed here that the cluster relaxation time is long enough
239: that $f(m) \le 1$ for all $m$ of interest. This is the case for
240: $t_{rh} \ga 50$ Myr. An appropriately modified version of this
241: analysis for $t_{rh}<50$ Myr shows the available mass leveling off at
242: a fixed fraction of the cluster mass as $t_{rh}\rightarrow0$, as
243: expected.
244:
245: \section{Mergers and Stellar Evolution}
246:
247: Given that dynamical evolution can concentrate enough mass in a
248: cluster core for collisions to occur at a significant rate, we can
249: then ask (i) if the collisions actually lead to mergers, and (ii)
250: under what circumstances a runaway merger can occur. The answer to
251: the first question is provided by the SPH simulations of Freitag \&
252: Benz (2001; see also Lombardi et al.~2003), who find that the low
253: relative velocities typical of these systems ensure that the colliding
254: stars usually merge with minimal mass loss. In small systems
255: (containing less than a few tens of thousands of solar masses),
256: collision rates are significantly enhanced by the fact that the
257: massive object tends to form binaries, which are then perturbed into
258: eccentric orbits by encounters with other stars (Portegies Zwart \&
259: McMillan 2002). Binary-induced mergers increase the collision cross
260: section, but they still require high central densities before the
261: (three-body) binary formation rates become significant. In larger
262: systems, unbound collisions appear to be the norm.
263:
264: Thus collisions naturally involve the most massive stars in the
265: cluster, and lead to the production of even more massive objects. The
266: merger products are generally out of thermal equilibrium and often
267: rapidly rotating, with the result that their subsequent stellar
268: evolution is currently poorly understood. In our simulations we
269: generally assume that the merged object evolves along a suitably
270: rejuvenated (non-rotating) track appropriate to its mass (Portegies
271: Zwart et al.~1997). This is at best a crude approximation, but we
272: note that it probably underestimates the radius of the merger product
273: during the out-of-equilibrium phase and hence the likelihood of a
274: runaway. Acting in the opposite sense is the fact that, while stellar
275: mass loss rates are very uncertain, ultramassive stars probably have
276: very strong winds. Van Beveren (these proceedings) points out that if
277: the wind mass loss rate exceeds the accretion rate due to mergers,
278: then the entire runaway process may fail. Our simulations generally
279: yield net merger accretion rates of $\sim10^{-3} M_\odot\,{\rm
280: yr}^{-1}$, suggesting that a high (but perhaps not impossibly so)
281: mass-loss rate is needed to shut the process down.
282:
283: Of course, it must be conceded that next to nothing is known about the
284: detailed evolution and ultimate fate of stars hundreds or thousands of
285: times more massive than the Sun. Nevertheless, the estimates
286: presented here make it clear that dynamical evolution in dense stellar
287: systems can easily produce conditions suitable for repeated stellar
288: collisions. The collision runaway at the center of such a system
289: should be extremely luminous and eminently observable during its short
290: lifetime. Observations of the cores of dense young star clusters in
291: our Galaxy and beyond may thus shed light on the structure and
292: lifetimes of such ultramassive stellar objects.
293:
294: %---------------------------------------------------------------------
295:
296: \section{An IMBH in the Galactic Center?}
297:
298: If our evolving cluster happens to reside close to the Galactic
299: center, then dynamical friction will tend to drive it inward, raising
300: the possibility that mass segregation can lead to a collision runaway
301: en route, and that the resulting IMBH can be transported rapidly
302: toward the center by the much more massive cluster (McMillan \&
303: Portegies Zwart 2003). Figure \ref{Fig1} illustrates this
304: possibility. It shows the mass $M$ and Galactocentric distance $R$ of
305: the runaway formed in a cluster of initial mass $4\times10^5 M_\odot$,
306: placed in a circular orbit of radius 10 pc, as it spirals inward and
307: is disrupted by the Galactic field. We clearly see the inward
308: transport of the growing merger product, terminating with the
309: dissolution of the cluster at $t\sim7$ Myr. Subsequently, the black
310: hole sinks more slowly to the center, eventually reaching $R=0$ at
311: $\sim24$ Myr. The early ($t\la7$ Myr) portion of the figure is from
312: an N-body simulation; the remainder is based on a simple analytical
313: model of the Galactic potential (Sanders \& Lowinger 1972).
314:
315: \begin{figure}[htbp!]
316: ~~~\psfig{figure=./mcmillan_fig1.ps,width=0.95\linewidth,angle=-90}
317: \caption[]{Evolution of the Galactocentric radius $R$ (dotted line,
318: left axis) and mass $M$ of the runaway merger product/IMBH (solid
319: line, right axis) in a cluster of initial mass $4\times10^5 M_\odot$
320: moving in the Galactic field.}\label{Fig1}
321: \end{figure}
322:
323: The recent report by Maillard et al.~(2004) of a possible IMBH near
324: the Galactic center, at the heart of the swarm of stars known as
325: IRS\,13, provides the exciting prospect of confronting our models
326: directly with reality. Portegies Zwart et al.~(2005, in preparation)
327: find that the observed properties of IRS\,13 are completely consistent
328: with theoretical expectations for a dense cluster remnant. They
329: further estimate that the total IMBH mass resulting from runaway
330: collisions in the cluster population which produced the inner bulge is
331: consistent with the $\sim3\times10^6\,M_\odot$ supermassive black hole
332: at the Galactic center.
333:
334: %---------------------------------------------------------------------
335:
336: \section*{Acknowledgments}
337: This work was supported by NASA ATP grant NAG5-10775, the Royal
338: Netherlands Academy of Sciences (KNAW), the Dutch organization of
339: Science (NWO), and by the Netherlands Research School for Astronomy
340: (NOVA).
341:
342: %---------------------------------------------------------------------
343:
344: \begin{thebibliography}
345:
346: \bibitem{Ebisuzaki2001} Ebisuzaki, T., Makino, J., Tsuru, T.G.,
347: Funato, Y., Portegies Zwart, S., Hut, P., McMillan, S., Matsushita S.,
348: Matsumoto H., \& Wane R. 2001, ApJ, 562, L19
349:
350: \bibitem{Freitag2001} Freitag, M, \& Benz, W. 2001, A\&A, 375, 711
351:
352: \bibitem{Gurkan2004} G\"urkan, M. A., Freitag, M., \& Rasio,
353: F. A. 2004, ApJ, 604, 632
354:
355: \bibitem{hh2003} Heggie, D.C., \& Hut, P. 2003, The Gravitational
356: Million-Body Problem (Cambridge University Press)
357:
358: \bibitem{lombardi2003}Lombardi, J. C., Thrall, A. P., Deneva, J. S.,
359: Fleming, S. W., \& Grabowski, P. E. 2003, MNRAS, 345, 762
360:
361: \bibitem{Maillard2004} Maillard, J. P., Paumard, T., Stolovy, S. R.,
362: \& Rigaut, F 2004, A\&A, 423, 155
363:
364: \bibitem{Merritt2004} Merritt, D., Piatek, S., Portegies Zwart, S., \&
365: Hemsendorf, M. 2004, ApJ, 608, 25
366:
367: \bibitem{McM+PZ2003} McMillan, S., \& Portegies Zwart, S. 2003, ApJ,
368: 596, 314
369:
370: \bibitem{PortegiesZwart1997} Portegies Zwart, S., Hut, P., \&
371: Verbunt, F. 1997, A\&A, 328, 130
372:
373: \bibitem{PortegiesZwart1999} Portegies Zwart, S., Makino, J.,
374: McMillan, S. L. W., \& Hut, P. 1999, A\&A, 348, 117
375:
376: \bibitem{PortegiesZwart2001} Portegies Zwart, S., McMillan, S., Hut,
377: P., \& Makino, J. 2001, MNRAS, 321, 199
378:
379: \bibitem{Portegies Zwart2002} Portegies Zwart, S., \& McMillan,
380: S. 2002, ApJ, 576, 899
381:
382: \bibitem{PZ2004} Portegies Zwart, S., Baumgardt H., Hut, P., Makino,
383: J., \& McMillan, S. 2004, Nature, 428, 724
384:
385: \bibitem{SL1972} Sanders, R.~H., \& Lowinger, T. 1972, AJ, 77, 292
386:
387: \bibitem{Spitzer1987} Spitzer, L. 1987, Dynamical Evolution of
388: Globular Clusters (Princeton University Press)
389:
390: \end{thebibliography}
391:
392: \end{document}
393: