1:
2: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
3: % This is a sample input file for your contribution to a
4: % multi-author book ("proceedings-like) to be published
5: % by Springer Verlag.
6: %
7: % Please use it as a template for your own input, and please
8: % follow the instructions for the formal editing of your
9: % manuscript as described in the file "1readme".
10: %
11: % Please send the Tex and figure files of your manuscript
12: % together with any additional style files as well as the
13: % PS file to the editor of your book.
14: %
15: % He or she will collect all contributions for the planned
16: % book, possibly compile them all in one go and pass the
17: % complete set of manuscripts on to Springer.
18: %
19: % Vers. 07/99
20: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
21:
22:
23:
24: %RECOMMENDED%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
25:
26: \documentclass[runningheads]{svmult}
27:
28: \usepackage{makeidx} % allows index generation
29: \usepackage{graphicx} % standard LaTeX graphics tool
30: % for including eps-figure files
31: \usepackage{subeqnar} % subnumbers individual equations
32: % within an array
33: \usepackage{multicol} % used for the two-column index
34: %\usepackage{cropmark} % cropmarks for pages without
35: % pagenumbers - only needed when manuscript
36: % is printed from paper and not from data
37: \usepackage{physprbb} % modified textarea for proceedings,
38: % lecture notes, and the like.
39: \makeindex % used for the subject index
40: % please use the style sprmidx.sty with
41: % your makeindex program
42:
43: %%upright Greek letters (example below: upright "mu")
44: \newcommand{\greeksym}[1]{{\usefont{U}{psy}{m}{n}#1}}
45: \newcommand{\umu}{\mbox{\greeksym{m}}}
46: \newcommand{\udelta}{\mbox{\greeksym{d}}}
47: \newcommand{\uDelta}{\mbox{\greeksym{D}}}
48: \newcommand{\uPi}{\mbox{\greeksym{P}}}
49: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
50:
51:
52: %OPTIONAL%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
53: %
54: %\usepackage{amstex} % useful for coding complex math
55: %\mathindent\parindent % needed in case "Amstex" is used
56: %
57: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
58:
59: %AUTHOR_STYLES_AND_DEFINITIONS%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
60: %
61: %\DeclareGraphicsRule{.ps.gz}{ps}{.ps.bb}{'gunzip -c #1}
62: %Please reduce your own definitions and macros to an absolute
63: %minimum since otherwise the editor will find it rather
64: %strenuous to compile all individual contributions to a
65: %single book file
66: %
67: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
68:
69: \begin{document}
70: %
71: \title*{The Formation of Stellar Clusters in Turbulent Molecular Clouds:
72: Effects of the Equation of State}
73: %
74: %
75: %\toctitle{The Formation of Stellar Clusters in Turbulent Molecular
76: %Clouds:
77: %\protect\newline Effects of the Equation of State}
78: % allows explicit linebreak for the table of content
79: %
80: %
81: \titlerunning{Formation of Stellar Clusters}
82: % allows abbreviation of title, if the full title is too long
83: % to fit in the running head
84: %
85: \author{Yuexing Li\inst{1, 2}
86: \and Ralf S. Klessen\inst{3}
87: \and Mordecai-Mark Mac Low\inst{1,2}}
88: %
89: \authorrunning{Li, Klessen \& Mac Low}
90: % if there are more than two authors,
91: % please abbreviate author list for running head
92: %
93: %
94: \institute{Department of Astronomy, Columbia University, New York, NY
95: 10027, USA
96: \and Department of Astrophysics, American Museum of Natural History, New
97: York
98: \and Astrophysikalisches Institut Potsdam, D-14482 Potsdam, Germany}
99:
100: \maketitle % typesets the title of the contribution
101:
102: \begin{abstract}
103: We study numerically the effect of varying the equation of state (EOS)
104: on the
105: formation of stellar clusters in turbulent molecular clouds. Our
106: results show that the EOS plays an important role in the fragmentation
107: of the
108: clouds, and the determination of the initial mass function (IMF) of the
109: protostellar cores. Fragmentation decreases with increasing adiabatic
110: index $\gamma$ in the range $0.2 < \gamma < 1.4$, although the total
111: amount of mass accreted appears to remain roughly constant through
112: that range, resulting in more massive cores at higher $\gamma$.
113: Fragmentation and collapse ceases entirely for $\gamma > 1.4$ as
114: expected from analytic arguments. Primordial gas may have effective
115: $\gamma > 1$, in which case these results may help explain why models of
116: the formation of the first stars produce isolated, massive objects.
117: \end{abstract}
118:
119: \section{INTRODUCTION}
120: Fragmentation of molecular clouds is key to the formation of star
121: clusters, yet the conditions for fragmentation are poorly
122: understood. There have been a few analytical approaches (Jeans 1902;
123: Penston 1966; Low \& Lynden-Bell 1976), and numerical investigations
124: of the effects of various physical processes on the collapse, such as
125: geometry and rotation of the clouds (see Bonnell \& Bastien 1993 for a
126: review), and magnetic fields (Galli et al. 2001).
127: Recently, the effects of turbulence have been studied extensively in a
128: series of 3D simulations (Klessen, Burkert, \& Bate 1998, Paper I;
129: Klessen, Heitsch, \& Mac Low 2000, Paper II; Heitsch, Mac Low, \&
130: Klessen 2001, Paper III; and Klessen 2001, Paper IV). However, these
131: results are based on isothermal models with $\gamma=1$. The effect of
132: the equation of state (EOS), which is essential in understanding the
133: physical structure and stability of the turbulent clouds, remains
134: uncertain.
135:
136: The balance of heating and cooling in a molecular cloud can be
137: approximately described by the polytropic EOS, $P = K\rho^{\gamma}$,
138: where $K$ is a constant, and $P, \rho$ and $\gamma$ are thermal
139: pressure, gas density and polytropic index, respectively. The density
140: structures generated by supersonic turbulence depends on the EOS
141: (Scalo et al. 1998; Passot \& Vazquez-Semadeni 1998; Passot \&
142: Vazquez-Semadeni 1999). A detailed analysis by Spaans \& Silk (2000)
143: suggests that $0.2 < \gamma < 1.4$ in the interstellar medium.
144:
145: In this paper, we perform a 3D numerical survey over this range of
146: $\gamma$, to investigate the effects of the EOS on the fragmentation
147: of turbulent clouds and the subsequent formation of a protostellar
148: cluster.
149:
150: \section{SIMULATIONS}
151: \label{sec_result}
152: We carry out smoothed particle hydrodynamics (SPH) simulations to
153: determine the effects of EOS on fragmentation. The code is
154: implemented with periodic boundary conditions (Klessen 1997) and
155: ``sink'' particles (Bate, Bonnell, \& Price 1995) to replace
156: high-density cores. The driven turbulence is set up following the $k =
157: $1--2 model in Mac Low (1999). For detailed descriptions of the code,
158: see Paper II. The treatment of the polytropic cloud in the code is the
159: same as in Bonnell \& Bastien (1991). We vary $\gamma$ in steps of 0.1
160: in otherwise identical simulations. Self-gravity is included in the
161: calculation at $t=2.0$ after turbulence is fully established (see
162: Paper II). Our models use 200,000 particles. The 13 simulations were
163: performed simultaneously using a serial version of the code on
164: processors of the Parallel Computing Facility of the AMNH.
165:
166:
167: \subsection{Fragmentation}
168: \label{subsec_frag}
169: Figure 1 shows that the fragmentation properties and the accretion of
170: the sinks strongly depend on the value of $\gamma$. The lower the
171: $\gamma$, the earlier fragmentation occurs, and the more fragments
172: form. It takes longer for gas with high $\gamma$ to fragment. The
173: slope of the accretion curves are roughly the same, but the number of
174: sinks is very different for different $\gamma$ at the same time, which
175: suggests that \textit{individual} protostellar cores is different
176: for different $\gamma$. In low-$\gamma$ gas, there are more
177: cores, but the mass of each core is much smaller than
178: in high-$\gamma$ gas.
179:
180: \begin{figure}[tb]
181: \leavevmode
182: \includegraphics[height=1.6in]{liF1a.ps}
183: \includegraphics[height=1.6in]{liF1b.ps}
184: \caption{\label{fig_sinknum}(\textit{left}) Number of sink particles
185: representing collapsed star-forming cores as a function of time, and
186: (\textit{right}) fraction of total mass in the box accreted onto cores
187: as a function of time, for different $\gamma$.}
188: \end{figure}
189:
190: The Jeans mass for a
191: polytropic gas sphere
192: $M_J = (K\pi/G)^\frac32 \gamma^\frac32 \rho^{\frac32(\gamma-\frac43)}$
193: so
194: \begin{equation}
195: \frac{\partial M_J}{\partial \rho}
196: \propto \frac{3\gamma - 4}{2}\rho^{(3\gamma - 6)/2},
197: \end{equation}
198: which implies that $\gamma = 4/3$ is a critical value during collapse.
199: When $\gamma < 4/3$, the Jeans mass decreases as density increases,
200: promoting fragmentation, but when $\gamma > 4/3$, the Jeans mass
201: \textit{increases} with density, stabilizing collapsing gas. Indeed,
202: our model with $\gamma = 1.4$ shows no collapse.
203:
204: \subsection{The Mass Spectra}
205: \label{subsec_mas}
206: Observations show that the mass distribution of gas clumps in various
207: clouds follows a simple power law, $dN/dM = M^{\nu}$, with $\nu
208: \approx -1.5$ (Blitz 1993), while the stellar IMF follows $\nu =
209: -2.33$ at the high-mass end (Salpeter 1955).
210:
211: \begin{figure}[tb]
212: \leavevmode
213: \includegraphics[height=1.2in]{liF2a.ps}
214: \includegraphics[height=1.2in]{liF2b.ps}
215: \includegraphics[height=1.2in]{liF2c.ps}
216: %
217: \includegraphics[height=1.1in]{liF2d.ps}
218: \includegraphics[height=1.1in]{liF2e.ps}
219: \includegraphics[height=1.1in]{liF2f.ps}
220:
221: \caption{\label{fig_mas}Top: 3-D distribution of the gas and protostars for
222: different $\gamma$. Bottom: Mass spectra of gas clumps
223: (\textit{thin
224: lines}) and of collapsed cores (\textit{hatched thick-lined
225: histograms}) for the corresponding cube above. The
226: percentage shows the fraction of total mass accreted onto cores. The
227: vertical line shows the SPH resolution limit. Shown also are two
228: power-law spectra with $\nu = -1.5$ (dashed-line) and $\nu =
229: -2.33$ (dotted line).}
230: \end{figure}
231: Figure \ref{fig_mas} shows that the spectra of both the clumps and the
232: cores change with $\gamma$, but that the ones of cores change more
233: dramatically. In low-$\gamma$ models, the mass distribution of the
234: cores at the high-mass end is roughly log-normal. As $\gamma$
235: increases, fewer but more massive cores emerge. When $\gamma > 1.0$,
236: the distribution is dorminated by high mass cores only, and the
237: spectrum tends to flatten out. It is no longer fitted by either a
238: log-normal or a power-law. The clump mass spectra, on the other hand,
239: do show power-law behavior at the high mass side, even for $\gamma >
240: 1.0$.
241:
242: Our results suggest
243: that stars tend to form in clusters in a low-$\gamma$
244: environment. Protostellar cores are low mass. The apparent lack of
245: power-law
246: behavior for the cores in the protostellar cluster might imply that
247: simple accretion is unable to generate as many high-mass stars as
248: predicted by the observations, which further suggests that other
249: mechanisms such as collisions (Bonnell, Bate \& Zinnecker 1998) may be
250: at work to produce the massive stars in a cluster. Higher resolution
251: models will be necessary to confirm this, however.
252:
253: On the other hand, our results also imply that massive stars can form
254: in small groups or alone in gas with $\gamma > 1.0$.
255: Spaans \& Silk (2000) suggest that a stiffer EOS ($\gamma > 1.0$) leads
256: to a peaked IMF, biased toward massive stars, while an EOS with
257: $\gamma < 1.0$ results in a power-law IMF, in general agreement with our
258: simulations.
259:
260: \section{ISOLATED STARS}
261: \label{sec_discussion}
262: The formation of isolated massive stars is of great interest, since in
263: most cases, massive stars are found in clusters. But recently, Lamers
264: et al. (2002) reported observations of isolated massive
265: stars or very small groups of a few isolated massive stars in the bulge
266: of M51. Massey (2002) also reported many field massive stars
267: in both the Large and Small Magellanic Clouds. From our
268: simulations, we see that when $\gamma > 1$, only very few or possibly
269: only one fragment occurs. This cloud core is massive, and would result in the
270: formation of a high-mass star.
271:
272: High resolution simulations by Abel, Bryan \& Norman (2002) of the
273: formation of the first star suggest that initially only one massive
274: metal-free star forms per pregalactic halo. In the early Universe,
275: inefficient cooling due to the lack of metals may result in high
276: $\gamma$. Our models would then suggest weak fragmentation, resulting
277: in the formation of only one massive star per cloud.
278:
279: \vspace{0.1in}{\small We thank M. Fall, H. Lamers, and H. Zinnecker for
280: valuable discussions, and D. Janies and W. Wheeler for their work on the
281: Parallel Computing Facility. YL thanks the AIP for its warm hospitality,
282: and the Kade Foundation for support of her visit there. This research
283: was partially funded by NASA grant NAG5-10103, and by NSF CAREER grant
284: AST99-85392. }
285:
286:
287: \begin{thebibliography}{8.}
288: \addcontentsline{toc}{section}{References}
289: Abel, T, Bryan, G., \& Norman, M. L. 2002, Science, 295, 5552, 93\\
290: %Benz, W. 1990, in The Numerical Modeling of Nonlinear Stellar
291: %Pulsations,
292: % ed. J. R. Buchler (Dordrecht: Kluwer), 269
293: Bate, M. R., Bonnell, I. A., \& Price, N. M. 1995, MNRAS, 277, 362\\
294: Bonnell, I. A., Bate, M. R. \& Zinnecker, H. 1998, MNRAS, 298, 93\\
295: Bonnell, I. \& Bastien, P. 1993, ApJ, 406, 614\\
296: Galli, D., Shu, F. H., Laughlin, G., \& Lizano, S. 2001, ApJ, 551, 367\\
297: Heitsch, F., Mac Low, M.-M. \& Klessen, R. S. 2001, ApJ, 547, 280,
298: Paper III\\
299: Jeans, J. H. 1902, Phil. Trans.A., 199, 1\\
300: Klessen, R. S. 1997, MNRAS, 292, 11\\
301: Klessen, R. S., Burkert, A., \& Bate, M. R. 1998, ApJ, 501, L205, Paper
302: I\\
303: Klessen, R. S., Heitsch, F., \& Mac Low, M.-M. 2000, ApJ, 535, 887,
304: Paper II\\
305: Klessen, R. S. 2001, ApJ, 556, 837, Paper IV\\
306: Lamers, H. J. G. L. M. et al. 2002, ApJ, 566, 818\\
307: Mac Low, M.-M. 1999, ApJ, 524, 169\\
308: Massey, P. 2002, ApJS, 141, 81\\
309: Passot, T., \& Vázquez-Semadeni, E. 1998, Phys. Rev. E, 58, 4501\\
310: Salpeter, E. E. 1955, ApJ, 121, 161\\
311: %Scalo, J. 1998, in ASP Conf. Ser. 142, The Stellar Initial Mass
312: %Function (38th
313: % Herstmonceux Conference), ed. G. Gilmore \& D. Howell (San
314: %Francisco:
315: % ASP), 201
316: Spaans, M., \& Silk, J. 2000, ApJ, 538, 115\\
317:
318: \end{thebibliography}
319:
320:
321: %INDEX%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
322: % Please check with the editor of your book whether he plans to
323: % include a "mutual" subject index - if so, please code your entries
324: % in the standard syntax. For your own purposes you may print your
325: % "personal" index by using the following commands:
326: %
327: %\clearpage
328: %\addcontentsline{toc}{section}{Index}
329: %\flushbottom
330: %\printindex
331: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
332:
333: \end{document}
334: