astro-ph0404160/ms.tex
1: %% Beginning of file 'ms.tex'
2: %%
3: %% Modified 04 Mar 20
4: %%
5: 
6: \documentclass[12pt,preprint]{aastex}
7: %\usepackage{emulateapj5}
8: 
9: %% manuscript produces a one-column, double-spaced document:
10: 
11: % \documentclass[manuscript]{aastex}
12: 
13: %% preprint2 produces a double-column, single-spaced document:
14: 
15: % \documentclass[preprint2]{aastex}
16: 
17: \newcommand{\myemail}{ asai@astro.s.chiba-u.ac.jp}
18: 
19: \shorttitle{MHD Simulations of Cold Fronts}
20: \shortauthors{Asai, Fukuda, \& Matsumoto}
21: 
22: 
23: \begin{document}
24: 
25: \title{Magnetohydrodynamic Simulations of the Formation of Cold Fronts
26: 	in Clusters of Galaxies including Heat Conduction}
27: 
28: \author{N. Asai\altaffilmark{1}, N. Fukuda\altaffilmark{2}, 
29: and R. Matsumoto\altaffilmark{3}}
30: 
31: \altaffiltext{1}{Graduate School of Science and Technology, 
32: Chiba University, 1-33 Yayoi-cho, Inage-ku, Chiba 263-8522, Japan;
33: \myemail}
34: 
35: \altaffiltext{2}{Department of Computer Simulation,
36: Faculty of Informatics, Okayama University of Science,
37: 1-1 Ridai-cho, Okayama 700-0005, Japan; fukudany@sp.ous.ac.jp}
38: 
39: \altaffiltext{3}{Department of Physics, Faculty of Science, 
40: Chiba University, 1-33 Yayoi-cho, Inage-ku, Chiba 263-8522, Japan; 
41: matumoto@astro.s.chiba-u.ac.jp }
42: 
43: 
44: \begin{abstract}
45: Recent {\it Chandra} observations of clusters of galaxies 
46: revealed the existence of a sharp ridge in the X-ray surface brightness 
47: where the temperature drops across the front. 
48: This front is called the cold front. We present 
49: the results of two-dimensional magnetohydrodynamic simulations
50: of the time evolution of a dense subcluster plasma moving in a cluster of 
51: galaxies. 
52: Anisotropic heat conduction along the magnetic field lines is included.
53: In the models without magnetic fields, the numerical results 
54: indicate that the heat conduction from the hot ambient plasma 
55: heats the cold dense plasma of the subcluster and diffuses 
56: out the cold front. When magnetic fields exist in a cluster of galaxies, 
57: however, cold fronts can be maintained because the heat conduction across the
58: magnetic field lines is suppressed. 
59: We found that, even when the magnetic fields in a cluster of galaxies are 
60: disordered, heat conduction across the front is restricted 
61: because the magnetic field lines are stretched along the front. 
62: Numerical results reproduced the X-ray intensity distribution 
63: observed in the A3667 cluster of galaxies.
64: 
65: \end{abstract}
66: 
67: \keywords{conduction---MHD---galaxies: cluster: magnetic fields---
68: X-ray: galaxies---plasmas}
69: 
70: 
71: \section{INTRODUCTION}
72: High-resolution {\it Chandra} observations have revealed various 
73: substructures in the hot X-ray emitting plasmas in clusters of galaxies. 
74: For example, sharp discontinuities of X-ray surface brightness 
75: distributions have been found around subclusters moving in a cluster 
76: of galaxies such as A2142 \citep{mar00} and A3667 
77: (Vikhlinin, Markevitch, \& Murray 2001b).
78: The temperature drops sharply across this boundary between the 
79: gravitationally confined, cold dense plasmas in the moving subcluster 
80: and the hot, less-dense ambient plasma.
81: This sharp discontinuity is called the cold front. 
82: It is not a shock front, because {\it Chandra} images of the 
83: X-ray surface brightness show that the temperature decreases 
84: in the denser side.
85: 
86: \citet{ett00} suggested that the sharp temperature gradient 
87: observed by {\it Chandra} in A2142 requires the heat 
88: conductivity near the front to be reduced by several 
89: orders of magnitude from the classical Spitzer value \citep{spi62}, 
90: % $\kappa_{\rm Sp}= 5 \times 10^{-7} T^{5/2} 
91: % \, {\rm erg \, s^{-1} \, cm^{-1} \, K^{-1}}$.
92: $\kappa_{\rm Sp}= 5 \times 10^{-7} T^{5/2}$ 
93: $\, {\rm erg \, s^{-1} \, cm^{-1} \, K^{-1}}$.
94: 
95: Based on the observations of A3667, 
96: Vikhlinin, Markevitch, \& Murray (2001a) suggested that 
97: turbulent, intergalactic magnetic fields stretched along 
98: the cold front reduce the growth rate of the Kelvin-Helmholtz (K-H) 
99: instability, which possibly destructs the front. 
100: Furthermore, heat conduction across the magnetic fields 
101: can be suppressed. It has been recognized that heat conduction 
102: in a cluster of galaxies can be very efficient (e.g., \citealt{tak77}) 
103: because the Coulomb mean free path, 
104: $l_{\rm c} \sim 4 \, (kT / 5 \, {\rm keV})^{2}(n / 10^{-3}{\rm cm^{-3}})^{-1} 
105: \, {\rm kpc}$, is large. Here, $n$ is the number density. 
106: The time required for heat to diffuse by conduction across a length  
107: $L$ is given by $\tau_{\rm Sp} \sim \rho L^{2}/ \kappa_{\rm Sp}
108: \sim 10^{7} (kT/ 5 \, {\rm keV})^{-5/2}
109: (n / 10^{-3} \, {\rm cm^{-3}})(L/100 \, {\rm kpc})^{2} \, {\rm yr}$, 
110: where $\rho$ is the density.
111: However, if magnetic fields exist, the characteristic length of the 
112: heat exchange across magnetic field lines is reduced to the Larmor radius, 
113: % $r_{\rm L} = 2500 \, (B/1 \, {\rm \mu G})^{-1} (T/5 \, {\rm keV})^{1/2} 
114: % \, {\rm km}$.
115: $r_{\rm L} = 2500 \, (B/1 \, {\rm \mu G})^{-1}$ $(T/5 \, {\rm keV})^{1/2} 
116: \, {\rm km}$.
117: Since the typical magnetic field strength in a cluster of galaxies is 
118: $\sim {\rm \mu G}$ (e.g., \citealt{kro94,car02}), 
119: the heat conductivity across magnetic field lines is drastically reduced.
120: 
121: In this Letter, we present the results of numerical simulations of the 
122: formation of cold fronts. These simulations include the effects of 
123: heat conduction.
124: We adapted the simulation code, originally developed for
125: solar flare simulations \citep{yok97,yok01}, 
126: which includes anisotropic heat conduction along magnetic field lines. 
127: 
128: 
129: \section{SIMULATION MODELS}
130: 
131: We simulated the time evolution of the cluster plasma in 
132: a frame comoving with the subcluster. The basic equations are as follows, 
133: 
134: \begin{equation}
135: \frac{\partial \rho}{\partial t} + \mbox{\boldmath $\nabla$ $\cdot$}
136:  (\rho \mbox{\boldmath $v$}) = 0,
137: \end{equation}
138: 
139: \begin{equation}
140: \rho \left[
141: \frac{\partial \mbox{\boldmath $v$}}{\partial t} 
142: + (\mbox{\boldmath $v$ $\cdot$ $\nabla$}) 
143: \mbox{\boldmath $v$}
144: \right] =
145: -\mbox{\boldmath $\nabla$}{\it p} + 
146: \frac{(\mbox{\boldmath $\nabla$} \times 
147: \mbox{\boldmath $B$}) \times \mbox{\boldmath $B$}}{4 \pi} 
148: - \rho \mbox{\boldmath $\nabla$} \psi, 
149: \end{equation}
150: 
151: \begin{equation}
152: \frac{\partial \mbox{\boldmath $B$}}{\partial t} 
153: = \mbox{\boldmath$\nabla$} \times (\mbox{\boldmath $v$} 
154: \times \mbox{\boldmath $B$}),
155: \end{equation}
156: 
157: \begin{equation}
158: \frac{\partial}{\partial t} 
159: \left[
160: \frac{1}{2} \rho v^{2} 
161: + \frac{B^{2}}{8 \pi} + \frac{p}{\gamma -1}
162: \right]
163: + \mbox{\boldmath$\nabla$$\cdot$}
164: \left[
165: \left(\frac{1}{2} \rho v^{2} + \frac{\gamma p}{\gamma -1}
166: \right) \mbox{\boldmath$v$} + \frac{1}{4 \pi} (-\mbox{\boldmath $v$} 
167: \times \mbox{\boldmath$B$}) 
168: \times \mbox{\boldmath $B$} - \kappa_{\parallel}  
169: \mbox{\boldmath $\nabla_{\parallel}$} T
170: \right]
171: = - \rho \mbox{\boldmath $v$ $\cdot$ $\nabla$} \psi,
172: \end{equation}
173: where $\rho$, $\mbox{\boldmath $v$}$, $p$, $\mbox{\boldmath $B$}$, and 
174: $\psi$ are the density, velocity, pressure, magnetic fields, 
175: and gravitational potential, respectively. We use the specific heat ratio 
176: $\gamma = 5/3$. The subscript $\parallel$ denotes the components parallel 
177: to the magnetic field lines. We assume that heat is conducted only 
178: along the field lines.
179: We solved the magnetohydrodynamic (MHD) equations by a modified 
180: Lax-Wendroff method with artificial viscosity. 
181: The heat conduction term in the energy equation is 
182: solved by the SOR method (see \citealp{yok01} for details).
183: The units of length, velocity, density, pressure, temperature, and time 
184: in our simulations are 
185: $r_{0}=250 \, {\rm kpc}$, 
186: $v_{0}=1000 \, {\rm km \, s^{-1}}$,
187: $\rho_{0}=5 \times 10^{-27} \, {\rm g \, cm^{-3}}$, 
188: $p_{0} = 3 \times 10^{-11} \, {\rm erg \, cm^{-3}}$, 
189: $T_{0}= 4 \, {\rm keV}$, and 
190: $t_{{0}}=r_{0}/v_{0}= 3 \times 10^{8} \, {\rm yr}$,
191: respectively. We adopted two-dimensional Cartesian coordinates $(x, y)$
192: by assuming translational symmetry in the $z$-direction.
193: The typical size of the simulation box and the number of grid points are 
194: $2.5 \, {\rm Mpc} \times 2.5 \, {\rm Mpc}$, and 
195: $(N_{x}, N_{y}) = (1201,1201)$, respectively.
196: 
197: Figure \ref{fig1} shows the initial density distribution. 
198: We assume that the subcluster is a sphere having a radius 
199: $r=500 \, {\rm kpc}$, 
200: and it is isothermal with temperature $kT_{\rm in} = 4 \, {\rm keV}$.
201: We assume that the gravitational potential at $r = (x^{2}+y^{2})^{1/2}$ 
202: is given by $\psi (r) = (c_{\rm s}^{2}/\gamma) \ln[1+(r/r_{\rm c})^{2}]$,
203: % \begin{equation}
204: % \psi (r)=  
205: % \frac{c_{\rm s}^{2}}{\gamma}
206: % \ln \left[1+
207: % \left(\frac{r}{r_{\rm c}}
208: % \right)^{2}
209: % \right],
210: % \end{equation}
211: where we use the core radius $r_{\rm c} = 290 \, {\rm kpc}$, 
212: and the sound speed inside 
213: the subcluster $c_{\rm s \, in} = 1000 \, {\rm km \, s^{-1}}$, 
214: where the subscript $in$ denotes the values inside the subcluster.
215: The density distribution is determined by the hydrostatic equation 
216: $\mbox{\boldmath$\nabla$}p = - \rho \mbox{\boldmath $\nabla$} \psi$, as 
217: $\rho_{\rm in} (r) = \rho_{\rm c} / [1 + (r/ r_{\rm c})^{2}]$,
218: where we adopt 
219: $\rho_{\rm c}= 10^{-26} \, {\rm g \, cm^{-3}}$. 
220: We also assume that the subcluster is surrounded by 
221: hot $(kT_{\rm out} = 8 \, {\rm keV})$, less-dense 
222: $(\rho_{\rm out} = 0.25 \, \rho_{0})$ 
223: plasma, where the subscript $out$ denotes the values outside the subcluster. 
224: The gravitational potential is taken to be fixed throughout
225: the simulation.
226: 
227: In order to simulate the moving subcluster in a cluster of galaxies, 
228: we assume that ambient plasmas initially have a uniform speed with 
229: Mach number $M = v_{x}/c_{\rm s \, out} = 1$, 
230: where $c_{\rm s \, out}$ is the sound speed of the ambient plasmas. 
231: The Mach number 
232: with respect to the sound speed inside the subcluster is 
233: $ M^{\prime}= v_{x}/c_{\rm s \, in}= \sqrt{2}$. 
234: 
235: Table \ref{tbl-1} shows the model parameters.
236: Models H and HC are without magnetic fields. In models M and MC,
237: we assume uniform magnetic fields parallel to the $y$-axis. In these models, 
238: the initial ratio of the ambient gas pressure to the magnetic pressure 
239: ($\beta= p_{\rm gas} / p_{\rm mag}$) is taken to be $\beta_{0} = 100$, 
240: which corresponds to the magnetic field strength 
241: $ B \sim 3 \, {\rm \mu G}$.
242: 
243: In models HC, MC, and MTC we included heat conduction. 
244: When the magnetic field does not exist, heat is conducted isotropically.
245: When the Coulomb mean free path $l_{\rm c}$ becomes comparable to 
246: or greater than the temperature scale height 
247: $l_{T} = T /|\mbox{\boldmath $\nabla$} T|$, the heat flux $Q$ 
248: is saturated as 
249: $Q_{\rm sat}= - [l_{T}/(4.2 \, l_{\rm c})] \kappa_{\rm Sp} 
250: \mbox{\boldmath $\nabla$} T$ when $l_{T} < 4.2 \, l_{\rm c}$ 
251: (e.g., \citealt{cow77}). 
252: In our initial model, since $l_{\rm c} \simeq 10 \, {\rm kpc}$ and 
253: $l_{T} < 6 \, {\rm kpc}$ around the cold front,
254: $Q_{\rm sat}$ at the front is smaller than 15\% of the classical value.
255: However, this reduced heat flux is large enough to smear the cold front 
256: within $10^{7} \, {\rm yr}$ because the conduction time scale across the 
257: cold front is 
258: $\tau_{\kappa} \sim (4.2 \, l_{\rm c}/ l_{T}) \, \rho \, {l_{T}}^{2} 
259: / \kappa_{\rm Sp} = 2 \times 10^{5} \, (kT/ 5 \, {\rm keV})^{-5/2} 
260: (n / 10^{-3} \, {\rm cm^{-3}}) (l_{T}/6 \, {\rm kpc})
261: (4.2 \, l_{\rm c}/42 \, {\rm kpc}) \, {\rm yr}$.
262: 
263: When the magnetic field exists, heat is conducted only in the 
264: direction parallel to the magnetic field lines. 
265: The conductivity parallel to the field lines is  
266: % $\kappa_{\parallel} = 5 \times 10^{-7} \, T^{5/2} \, {\rm erg \, s^{-1} 
267: % \, cm^{-1} \, K^{-1}}$. 
268: $\kappa_{\parallel} = 5 \times 10^{-7} \, T^{5/2}$ $ \, {\rm erg \, s^{-1} 
269: \, cm^{-1} \, K^{-1}}$. 
270: Conversely, the conductivity across the field lines is $ \kappa_{\perp} = 0$. 
271: The conduction time scale along the field lines is 
272: % $\tau_{\kappa} \sim \rho L^{2}/ \kappa_{\parallel}
273: % \sim 10^{7} \, (kT/ 5 \, {\rm keV})^{-5/2} (n / 10^{-3} \, {\rm cm^{-3}})
274: % (L/100 \, {\rm kpc})^{2} \, {\rm yr}$.
275: $\tau_{\kappa} \sim \rho L^{2}/ \kappa_{\parallel}
276: \sim 10^{7} \, (kT/ 5 \, {\rm keV})^{-5/2} (n / 10^{-3} \, {\rm cm^{-3}})$
277: $(L/100 \, {\rm kpc})^{2} \, {\rm yr}$. 
278: 
279: In model MTC, we simulated a subcluster moving in an ambient plasma with initially disordered magnetic 
280: fields, and we included the effects of heat conduction.
281: In this model, we superposed magnetic fields created by thin currents 
282: flowing in the $z$-direction. The location and strength of the thin currents are 
283: chosen randomly. In this particular model, the size of the simulation box is 
284: $7.5 \, {\rm Mpc} \times 2.5 \, {\rm Mpc}$, and the number of grid 
285: points is $(N_{x}, N_{y}) = (901, 301)$. 
286: 
287: The left boundary at $x = -5$ is taken to be a fixed boundary, except in 
288: model MTC. 
289: In model MTC, the magnetic fields at the left boundary are extrapolated from 
290: magnetic fields 
291: in internal points. Other quantities are fixed at the left boundary. 
292: The effects of the left boundary condition is negligible in the time 
293: range we report in this Letter.  
294: Other boundaries are free boundaries where waves can be transmitted.
295: 
296: \section{RESULTS}
297: 
298: \subsection{Effects of Magnetic Fields on Heat Conduction}\label{sec:mc}
299: 
300: Figure \ref{fig2} shows the distributions of density, temperature and 
301: X-ray intensity for model MC (upper panels) and model HC (lower panels) 
302: at $t=3.0$ (0.9 Gyr). 
303: Model MC includes the effects of anisotropic heat conduction; that is, 
304: heat is conducted only in the direction parallel to the magnetic field lines. 
305: The X-ray intensity distribution is visualized from 
306: the simulation results as the logarithm of $\rho^{2}$.
307: In model HC with isotropic heat conduction, 
308: the subcluster plasma is heated by the ambient hot plasma 
309: and evaporates rapidly, 
310: because heat is conducted from the ambient plasma to the cold subcluster 
311: plasma. Thus, the sharp front of the temperature distribution disappears.
312: We also carried out simulations that included the saturation effects of heat 
313: flux, and we confirmed that the effects are negligible. 
314: In model MC, 
315: the heat conduction and plasma diffusion across the front are 
316: suppressed because the magnetic field lines are parallel to the cold front.
317: The X-ray intensity distribution in model MC shows 
318: a steeper gradient at the leading edge of the subcluster than that in 
319: model HC. 
320: The numerically obtained X-ray intensity distribution of model MC 
321: reproduces the overall structures of X-ray intensity observed in A3667.
322: At $t=3.0$, the magnetic field strength near the front increases 
323: by a factor of 8, and $\beta$ decreases by a factor of 100 from the 
324: initial state.
325: In this model, the lifetime of the cold front is 
326: more than $10^{9} \,{\rm yr}$. 
327: 
328: Numerical results also reproduced the bow shock upstream of the cold front.
329: The interaction between the inflowing hot ambient plasma and the 
330: cold subcluster plasma drives another shock propagating inside the subcluster. 
331: This shock is located at $(x,y)=(2.5,0)$ in the upper panels of Figure \ref{fig2}. 
332: The shock propagating inside the subcluster initially smears the cold front by 
333: shock heating. Subsequently, adiabatic expansion 
334: following the shock compression cools the subcluster plasma. 
335: This cold plasma is pulled back by gravity and the cold front is recovered.
336: 
337: 
338: \subsection{Comparison with Models without Heat Conduction}\label{sec:m}
339: 
340: Figure \ref{fig3} shows the temperature distribution at $t=3.0$ in 
341: models M (left) and H (right). 
342: In these models, heat conduction is not included. 
343: By comparing model H with HC, we confirm that 
344: when the magnetic field does not exist, cold fronts 
345: can be maintained only if the heat conductivity is small enough. 
346: 
347: The results of model H indicate that the K-H instability 
348: grows at the contact discontinuity between the cold, dense subcluster 
349: plasma and the hot, less-dense ambient plasma. The growth of 
350: the K-H instability is not prominent near the stagnation 
351: point $(x,y) \sim (-1,0)$, but the K-H vortices grow 
352: around $(x,y) \sim (2,\pm 2)$. 
353: The K-H instability does not disrupt the cold front, because the velocity 
354: difference inside and outside of the contact surface is small in the 
355: upstream side of the dense subcluster.
356: In model M, the magnetic tension force along the front 
357: suppresses the K-H instability, even in the sides of the dense 
358: subcluster, 
359: % $(x, y) = (2, \pm 2)$ 
360: where the velocity shear is large. 
361: Except for this stability for the K-H instability, 
362: the numerical results of model M are similar to these of model H.
363: 
364: \subsection{Effects of Disordered Magnetic Fields}\label{sec:mtc}
365:  
366: Figure \ref{fig4} shows the results of model MTC, 
367: in which we initially assumed disordered magnetic fields. 
368: The top and bottom panels show the time evolution of the density and 
369: temperature distributions at $t=0$ and $t= 3.0$ ($0.9 \, {\rm Gyr}$), 
370: respectively. 
371: The white curves and arrows show the magnetic field lines and velocity vectors, 
372: respectively. 
373: Numerical results show that the tangled magnetic field lines in the vicinity 
374: of the front are compressed and stretched along the front. 
375: The temperature distributions 
376: show that heat is conducted along the tangled field lines.
377: As a result, the low-temperature subcluster plasma is heated and diffuses 
378: along the field lines. 
379: Since the tangled magnetic field lines are stretched along the front, 
380: suppression of heat conduction across the field lines enables 
381: the existence of cold fronts.
382: 
383: 
384: \section{DISCUSSION}
385: 
386: \citet{hei03} carried out hydrodynamical simulations of 
387: the interaction of cold subcluster plasma and hot ambient matter. 
388: They suggested that a cold front can be formed after the passage of 
389: shock waves through the subcluster plasma, when the cold plasma, initially 
390: pushed forward by the shock, falls back. However, in their simulations 
391: heat conduction and magnetic fields were neglected.
392: In this Letter, we presented the results of numerical simulations 
393: which included both magnetic fields and heat conduction. 
394: We confirmed that, since the Spitzer conductivity is very large 
395: in hot ambient plasma, cold fronts cannot be maintained longer 
396: than $0.3 \, {\rm Gyr}$, unless the heat conduction across 
397: the front is suppressed by the magnetic fields. 
398: 
399: We also carried out numerical simulations which included heat conduction 
400: starting from the same initial condition as \citet{hei03}.
401: We found that, when we adopt the Spitzer conductivity, the subcluster 
402: evaporates rapidly and the cold front is not formed.  
403: When we included magnetic fields and the effects of anisotropic heat 
404: conduction, the numerical results resembled Heinz's results. 
405: As suggested by \citet{hei03}, after the passage of shock waves, 
406: the cold plasma pulled back by gravity helps sharpen the cold front. 
407: 
408: In \S \ref{sec:m}, we presented numerical results that showed the 
409: K-H instability does not grow near the cold front, 
410: because the velocity of inflowing plasmas is reduced after the plasmas pass through 
411: the bow shock. In the magnetohydrodynamical models, the magnetic fields, which stretched 
412: along the contact surface, also stabilized the K-H instability. 
413: Thus, the K-H instability does not threaten the existence of 
414: cold fronts. 
415: 
416: In \S \ref{sec:mtc} we showed results of a model with initially disordered 
417: magnetic fields. Numerical results show that the magnetic fields near 
418: the front are gradually stretched. Thus, the initially tangled magnetic field 
419: lines become parallel to the front and suppress the heat conduction 
420: across the front. We confirmed the suggestion by \citet{vik01a} that 
421: the stretching of turbulent intergalactic magnetic fields enables 
422: the maintenance of cold fronts.
423: %Let us discuss the effects of smaller scale turbulence. 
424: If the scale of magnetic turbulence is smaller than the mean free path, 
425: heat can be conducted almost isotropically. 
426: The heat conductivity in such a turbulent MHD medium 
427: is as large as $1/5$ of the Spitzer value \citep{nar01} and still large 
428: enough to diffuse out the cold front. We expect that coherent fluid motions 
429: and the stretching of the magnetic fields along the cold front increase 
430: the coherent length of field lines and make that region 
431: less turbulent. 
432: 
433: Let us now discuss the three-dimensional (3D) effects. Since magnetic fields 
434: compressed in front of the cold front can expand in the third direction, 
435: the strength of the magnetic fields in this region will be reduced in 3D. 
436: However, even such weakened magnetic fields suppress the heat conduction 
437: across them. Thus, we expect that cold fronts can be sustained even in 3D. 
438: We hope to report the results of such 3D MHD simulations in subsequent 
439: papers.
440: 
441: In this Letter, we assumed that the ambient plasma has magnetic fields. 
442: Recently, \citet{oka03} pointed out the possibility that 
443: magnetic fields are spontaneously generated by the Ramani-Laval instability 
444: around the cold front. Such magnetic fields can also help to suppress 
445: the heat conduction across the front.
446: 
447: 
448: \acknowledgments 
449: 
450: We thank T. Hanawa, S. Miyaji, N. Okabe, M. Hattori, and U. Fujita 
451: for their discussions.
452: The simulation code CANS that we used was developed as 
453: a product of the ACT-JST project (PI. R. Matsumoto). 
454: We thank T. Yokoyama for code development.
455: Numerical computations were carried out 
456: by VPP5000 at the Astronomical Data Analysis Center of 
457: the National Astronomical Observatory Japan (PI. N. Asai, project-ID yna23). 
458: 
459: 
460: \begin{thebibliography}{} 
461: 
462: \bibitem[Carilli \& Taylor (2002)]{car02} 
463: Carilli, C. L., \& Taylor, G. B. 2002, \araa, 40, 319
464: 
465: \bibitem[Cowie \& McKee (1977)]{cow77} 
466: Cowie, L. L., \& McKee, C. F. 1977, \apj, 211, 242
467: 
468: 
469: \bibitem[Ettori \& Fabian (2000)]{ett00} 
470: Ettori, S., \&  Fabian, A. C. 2000, \mnras, 317, L57
471: 
472: \bibitem[Heinz et al.(2003)]{hei03} Heinz, S., Churazov, E., Forman, W., 
473: Jones, C., \& Briel, U. G. 2003, \mnras, 346, 13
474: 
475: \bibitem[Kronberg (1994)]{kro94} 
476: Kronberg, P. P. 1994, Rep. Prog. Phys., 57, 325 
477: 
478: \bibitem[Markevitch et al.(2000)]{mar00} 
479: Markevitch, M., et al. 2000, \apj, 541, 542
480: 
481: \bibitem[Narayan \& Medvedev (2001)]{nar01} 
482: Narayan, R., \& Medvedev, M. V. 2001, \apj, 562, L129
483: 
484: \bibitem[Okabe \& Hattori (2003)]{oka03} 
485: Okabe, N., \& Hattori, M. 2003, \apj, 599, 964
486: 
487: \bibitem[Spitzer (1962)]{spi62} Spitzer, L. 1962, Physics of Fully 
488: Ionized Gases (2d ed.; New York: Intersciense)
489: 
490: \bibitem[Takahara \& Ikeuchi (1977)]{tak77} 
491: Takahara, F., \& Ikeuchi, S. 1977, Prog. Theor. Phys., 58, 1728
492: 
493: \bibitem[Vikhlinin et al.(2001a)]{vik01a} 
494: Vikhlinin, A., Markevitch, M., \& Murray, S. S. 2001a, \apj, 549, L47
495: 
496: \bibitem[Vikhlinin et al.(2001b)]{vik01b} 
497: Vikhlinin, A., Markevitch, M., \& Murray, S. S. 2001b, \apj, 551, 160
498: 
499: \bibitem[Yokoyama \& Shibata (1997)]{yok97} 
500: Yokoyama, T., \& Shibata, K. 1997, \apj, 474, L61
501: 
502: \bibitem[Yokoyama \& Shibata (2001)]{yok01} 
503: Yokoyama, T., \& Shibata, K. 2001, \apj, 549, 1160
504: \end{thebibliography}
505: 
506: \clearpage
507: 
508: \begin{deluxetable}{cccc}
509: \tabletypesize{\scriptsize}
510: \tablecaption{Simulation models \label{tbl-1}}
511: \tablewidth{0pt}
512: \tablehead{
513: \colhead{Model} & \colhead{$\beta_{0}$} & \colhead{$\kappa$} 
514: & \colhead{simulation box}
515: }
516: \startdata
517: H & $\infty$  & $0$ & $2.5 \, {\rm Mpc} \times 2.5 \, {\rm Mpc}$\\
518: M & 100  & $0$ & $2.5 \, {\rm Mpc} \times 2.5 \, {\rm Mpc}$\\
519: HC & $\infty$ & $\kappa$ & $2.5 \, {\rm Mpc} \times 2.5 \, {\rm Mpc}$\\
520: MC &100 & $\kappa_{\parallel}$ & $2.5 \, {\rm Mpc} \times 2.5 \, {\rm Mpc}$\\
521: MTC & nonuniform & $\kappa_{\parallel}$& $7.5 \, {\rm Mpc} \times 2.5 \, {\rm Mpc}$\\
522: \enddata
523: \end{deluxetable}
524: 
525: %%
526: \begin{figure}
527: %\plotone{f1.eps} 
528: \caption{
529: The initial condition of models M and MC. The color contour in the 
530: left panel shows the density. Arrows show the velocity vectors 
531: and white lines show the magnetic field lines. The right panel shows the 
532: distribution of density (solid curve), pressure (dotted curve), 
533: and temperature (dashed line) along $y = 0$.
534: The plasma beta in the ambient plasma is $\beta_{\rm 0} = 100$. 
535: \label{fig1}}
536: \end{figure}
537: %%
538: \begin{figure}
539: %\plotone{f2.eps} 
540: \caption{
541: The effects of heat conduction and magnetic fields. 
542: The panels show the distributions of density, temperature, and  X-ray 
543: intensity at $t=3.0$ from left to right. The upper panels show 
544: the results for model MC (with magnetic fields and heat conduction), 
545: the lower panels show the results for model HC 
546: (without magnetic fields and  with heat conduction). 
547: The X-ray intensity distribution is visualized from the simulation 
548: results as the logarithm of $\rho^{2}$. 
549: The white curves and arrows show the magnetic field 
550: lines and velocity vectors, respectively. \label{fig2}} 
551: \end{figure}
552: %%
553: \begin{figure}
554: %\plotone{f3.eps} 
555: \caption{
556: Results of the simulations without heat conduction 
557: for models M (left, with magnetic fields) and 
558: H (right, without magnetic fields). 
559: The color contours show the temperature at $t=3.0$. 
560: The solid curves and arrows show 
561: the magnetic field lines and velocity vectors, respectively. 
562: \label{fig3}} 
563: \end{figure}
564: %%
565: \begin{figure}
566: %\plotone{f4.eps} 
567: \caption{
568: Time evolutions of model MTC (with disordered magnetic fields) 
569: at the central region of $5.0 \, \rm Mpc \times 2.5 \, \rm Mpc$ 
570: at $t =0$ and $t=3.0$. The top panels show the density distribution 
571: and the bottom panels show the temperature distribution. 
572: The white curves and arrows show the magnetic field lines 
573: and velocity vectors, respectively. \label{fig4}} 
574: \end{figure}
575: %%
576: 
577: \end{document}
578: 
579: %% end of file 'ms.tex'
580: 
581: 
582: 
583: 
584: 
585: 
586: 
587: 
588: 
589: 
590: 
591: 
592: 
593: 
594: 
595: