1: %% The command below calls the preprint style
2: %% which will produce a one-column, single-spaced document.
3: %% Examples of commands for other substyles follow. Use
4: %% whichever is most appropriate for your purposes.
5: %%
6: % \documentclass[12pt,preprint]{aastex}
7: \documentclass{emulateapj}
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: %% Sometimes a paper's abstract is too long to fit on the
18: %% title page in preprint2 mode. When that is the case,
19: %% use the longabstract style option.
20:
21: %% \documentclass[preprint2,longabstract]{aastex}
22:
23: %% If you want to create your own macros, you can do so
24: %% using \newcommand. Your macros should appear before
25: %% the \begin{document} command.
26: %%
27: %% If you are submitting to a journal that translates manuscripts
28: %% into SGML, you need to follow certain guidelines when preparing
29: %% your macros. See the AASTeX v5.x Author Guide
30: %% for information.
31:
32: \def\FLASH {{\sc flash}}
33: \def\PARAMESH {{\sc paramesh}}
34: \def\mag{{\bf B}}
35: \def\mach{{\cal M}}
36: \def\gsim{\;\lower4pt\hbox{${\buildrel\displaystyle >\over\sim}$}\,}
37: \def\lsim{\;\lower4pt\hbox{${\buildrel\displaystyle <\over\sim}$}\,}
38:
39: \newcommand{\vdag}{(v)^\dagger}
40: \newcommand{\myemail}{skywalker@galaxy.far.far.away}
41:
42: %% You can insert a short comment on the title page using the command below.
43:
44: %\slugcomment{Not to appear in Nonlearned J., 45.}
45:
46: %% If you wish, you may supply running head information, although
47: %% this information may be modified by the editorial offices.
48: %% The left head contains a list of authors,
49: %% usually a maximum of three (otherwise use et al.). The right
50: %% head is a modified title of up to roughly 44 characters.
51: %% Running heads will not print in the manuscript style.
52:
53: \shorttitle{Anisotropic thermal conduction in SNRs}
54: \shortauthors{Orlando et al.}
55:
56: %% This is the end of the preamble. Indicate the beginning of the
57: %% paper itself with \begin{document}.
58:
59: \begin{document}
60:
61: %% LaTeX will automatically break titles if they run longer than
62: %% one line. However, you may use \\ to force a line break if
63: %% you desire.
64:
65: \title{The importance of magnetic-field-oriented thermal conduction
66: in the interaction of SNR shocks with interstellar clouds}
67:
68: \author{S. Orlando\altaffilmark{1},
69: F. Bocchino\altaffilmark{1}}
70: \affil{INAF - Osservatorio Astronomico di Palermo ``G.S.
71: Vaiana'', Piazza del Parlamento 1, 90134 Palermo, Italy}
72: \author{F. Reale\altaffilmark{2,1},
73: G. Peres\altaffilmark{2,1} and
74: P. Pagano}
75: \affil{Dip. di Scienze Fisiche \& Astronomiche, Univ. di
76: Palermo, Piazza del Parlamento 1, 90134 Palermo,
77: Italy}
78:
79: % Notice that each of these authors has alternate affiliations, which
80: % are identified by the \altaffilmark after each name. The actual alternate
81: % affiliation information is typeset in footnotes at the bottom of the
82: % first page, and the text itself is specified in \altaffiltext commands.
83: % There is a separate \altaffiltext for each alternate affiliation
84: % indicated above.
85:
86: \altaffiltext{1}{Consorzio COMETA, via Santa Sofia 64, 95123 Catania, Italy}
87: \altaffiltext{2}{INAF, Viale del Parco Mellini 84, 00136 Roma, Italy}
88:
89: % The abstract environment prints out the receipt and acceptance dates
90: % if they are relevant for the journal style. For the aasms style, they
91: % will print out as horizontal rules for the editorial staff to type
92: % on, so long as the author does not include \received and \accepted
93: % commands. This should not be done, since \received and \accepted dates
94: % are not known to the author.
95:
96: \begin{abstract}
97: We explore the importance of magnetic-field-oriented thermal conduction
98: in the interaction of supernova remnant (SNR) shocks with radiative gas
99: clouds and in determining the mass and energy exchange between the clouds
100: and the hot surrounding medium. We perform 2.5D MHD simulations of a shock
101: impacting on an isolated gas cloud, including anisotropic thermal
102: conduction and radiative cooling; we consider the representative case of
103: a Mach 50 shock impacting on a cloud ten-fold denser than the ambient
104: medium. We consider different configurations of the ambient magnetic
105: field and compare MHD models with or without the thermal conduction. The
106: efficiency of the thermal conduction in the presence of magnetic field is,
107: in general, reduced with respect to the unmagnetized case. The reduction
108: factor strongly depends on the initial magnetic field orientation,
109: and it is minimum when the magnetic field is initially aligned with the
110: direction of shock propagation. The thermal conduction contributes to
111: suppress hydrodynamic instabilities, reducing the mass mixing of the
112: cloud and preserving the cloud from complete fragmentation. Depending
113: on the magnetic field orientation, the heat conduction may determine a
114: significant energy exchange between the cloud and the hot surrounding
115: medium which, while remaining always at levels less than those in the
116: unmagnetized case, leads to a progressive heating and evaporation of
117: the cloud. This additional heating may contrast the radiative cooling
118: of some parts of the cloud, preventing the onset of thermal instabilities.
119: \end{abstract}
120:
121: % The different journals have different requirements for keywords. The
122: % keywords.apj file, found on aas.org in the pubs/aastex-misc directory,
123: % contains a list of keywords used with the ApJ and Letters. These are
124: % usually assigned by the editor, but authors may include them in their
125: % manuscripts if they wish.
126:
127: \keywords{conduction ---
128: magnetohydrodynamics ---
129: shock waves ---
130: ISM: clouds ---
131: ISM: magnetic fields ---
132: ISM: supernova remnants}
133:
134: % That's it for the front matter. On to the main body of the paper.
135: % We'll only put in tutorial remarks at the beginning of each section
136: % so you can see entire sections together.
137:
138: %________________________________________________________________
139: \section{Introduction}
140: \label{sec1}
141:
142: The interaction of the shock waves of supernova remnants (SNRs) with the
143: magnetized and inhomogeneous interstellar medium (ISM) is responsible of
144: the great morphological complexity of SNRs and certainly plays a major
145: role in determining the exchange of mass, momentum, and energy between
146: diffuse hot plasma and dense clouds or clumps. These exchanges may occur
147: through, for example, hydrodynamic ablation and thermal conduction and,
148: among other things, lead to the cloud crushing and to the reduction of
149: the Jeans mass causing star formation.
150:
151: The propagation of hot SNR shock fronts in the ISM and their interaction
152: with local over-dense gas clouds have been investigated with detailed
153: hydrodynamic and MHD modeling. The most complete review of this problem
154: in the unmagnetized, non-conducting, and non-radiative limits is provided
155: by \cite{1994ApJ...420..213K}. These studies have shown that the cloud is
156: disrupted by the action of both Kelvin-Helmholtz (KH) and Rayleigh-Taylor
157: (RT) instabilities after several crushing times, with the cloud material
158: expanding and diffusing into the ambient medium. An ambient magnetic
159: field can both act as a confinement mechanism of the plasma and be
160: modified by the interstellar flow and by local field stretching. Also,
161: a strong magnetic field is known to limit hydrodynamic instabilities
162: developing during the shock-cloud interaction by providing an additional
163: tension at the interface between the cloud and the surrounding medium
164: \citep[e.g][]{1994ApJ...433..757M, 1996ApJ...473..365J}.
165:
166: The interaction of the shock with a {\it radiative} cloud has been
167: only recently analyzed in detail \citep[e.g.][]{2002A&A...395L..13M,
168: 2004ApJ...604...74F}. 2D calculations have shown that the effect of
169: the radiative cooling is to break up the clouds into numerous dense
170: and cold fragments that survive for many dynamical timescales. In the
171: case of the interaction between magnetized shocks and radiative clouds,
172: the magnetic field may enhance the efficiency of the radiative cooling,
173: influencing the size and distribution of condensed cooled fragments
174: \citep{2005ApJ...619..327F}.
175:
176: The role played by the thermal conduction during the shock-cloud
177: interaction has been less studied so far. In a previous paper,
178: \cite{2005A&A...444..505O} (hereafter Paper I) have addressed this point
179: in the unmagnetized limit. In particular, we have investigated the effect
180: of thermal conduction and radiative cooling on the cloud evolution and on
181: the mass and energy exchange between the cloud and the surrounding medium;
182: we have selected and explored two different physical regimes chosen so
183: that either one of the processes is dominant. In the case dominated by
184: the radiative losses, we have found that the shocked cloud fragments into
185: cold, dense, and compact filaments surrounded by a hot corona which is
186: ablated by the thermal conduction. Instead, in the case dominated by
187: thermal conduction, the shocked cloud evaporates in a few dynamical
188: timescales. In both cases, we have found that the thermal conduction
189: is very effective in suppressing the hydrodynamic instabilities that
190: would develop at the cloud boundaries, preserving the cloud from complete
191: destruction. \cite{orlando2} and \cite{2006A&A...458..213M} have studied
192: the observable effects of thermal conduction on the evolution of the
193: shocked cloud in the X-ray band.
194:
195: Here, we extend the previous studies by investigating the effect of
196: the thermal conduction in a magnetized medium, unexplored so far.
197: Of special interest to us is to investigate the role of anisotropic
198: thermal conduction - funneled by locally organized magnetic fields -
199: in the mass and energy exchange between ISM phases. In particular,
200: we aim at addressing the following questions: How and under which
201: physical conditions does the magnetic-field-oriented thermal conduction
202: influence the evolution of the shocked cloud? How do the mass mixing
203: of the cloud material and the energy exchange between the cloud and the
204: surrounding medium depend on the orientation and strength of the magnetic
205: field and on the efficiency of the thermal conduction?
206:
207: To answer these questions, we take as representative the model case of
208: a shock with Mach number $\mach = 50$ (corresponding to a post-shock
209: temperature $T\approx 4.7\times 10^6$ K for an unperturbed medium
210: with $T=10^4$ K) impacting on an isolated cloud ten-fold denser than
211: the ambient medium. Paper I has shown that, in this case, the thermal
212: conduction dominates the evolution of the shocked cloud in the absence of
213: magnetic field. Around this basic configuration, we perform a set of MHD
214: simulations, with different interstellar magnetic field orientations,
215: and compare models calculated with thermal conduction turned either
216: ``on'' or ``off'' in order to identify its effects on the cloud evolution.
217:
218: The paper is organized as follows: in Sect. \ref{sec2} we describe the
219: MHD model and the numerical setup; in Sect. \ref{sec3} we discuss the
220: results; and finally in Sect. \ref{sec4} we draw our conclusions.
221:
222: %
223: %________________________________________________________________
224: \section{The model}
225: \label{sec2}
226:
227: We model the impact of a planar supernova shock front onto an isolated
228: gas cloud. The shock propagates through a magnetized ambient medium and
229: the cloud is assumed to be small compared to the curvature radius of
230: the shock\footnote{In the case of a small cloud, the SNR does not evolve
231: significantly during the shock-cloud interaction, and the assumption of a
232: planar shock is justified \citep[see also][]{1994ApJ...420..213K}.}.
233: The fluid is assumed to be fully ionized with a ratio of specific
234: heats $\gamma = 5/3$. The model includes radiative cooling, thermal
235: conduction (including the effects of heat flux saturation) and resistivity
236: effects. The shock-cloud interaction is modeled by solving numerically
237: the time-dependent non-ideal MHD equations (written in non-dimensional
238: conservative form):
239:
240: \begin{equation}
241: \frac{\partial \rho}{\partial t} + \nabla \cdot (\rho {\bf u}) = 0~,
242: \end{equation}
243:
244: \begin{equation}
245: \frac{\partial \rho {\bf u}}{\partial t} + \nabla \cdot (\rho
246: {\bf uu}-{\bf BB}) + \nabla P_* = 0~,
247: \end{equation}
248:
249: \begin{eqnarray}
250: \lefteqn{\frac{\partial \rho E}{\partial t} +\nabla\cdot [{\bf u}(\rho
251: E+P_*) -{\bf B}({\bf u}\cdot {\bf B})] =} \nonumber \\
252: & \displaystyle ~~~~~~~~~~~~~~~
253: \nabla\cdot [{\bf B}\times(\eta\nabla\times{\bf B})]
254: -\nabla\cdot {\bf F}_{\rm c} -n_{\rm e} n_{\rm H} \Lambda(T)~,
255: \end{eqnarray}
256:
257: \begin{equation}
258: \frac{\partial {\bf B}}{\partial t} +\nabla \cdot({\bf uB}-{\bf
259: Bu}) = -\nabla\times(\eta\nabla\times{\bf B})~,
260: \end{equation}
261:
262: \noindent
263: where
264:
265: \[
266: P_* = P + \frac{B^2}{2}~,~~~~~~~~~~~~~
267: E = \epsilon +\frac{1}{2} u^2+\frac{1}{2}\frac{B^2}{\rho}~,
268: \]
269:
270: \noindent
271: are the total pressure, and the total gas energy (internal energy,
272: $\epsilon$, kinetic energy, and magnetic energy) respectively, $t$
273: is the time, $\rho = \mu m_H n_{\rm H}$ is the mass density, $\mu =
274: 1.26$ is the mean atomic mass (assuming cosmic abundances), $m_H$
275: is the mass of the hydrogen atom, $n_{\rm H}$ is the hydrogen number
276: density, {\bf u} is the gas velocity, $T$ is the temperature, {\mag} is
277: the magnetic field, $\eta$ is the resistivity according to \cite{spi62},
278: ${\bf F}_{\rm c}$ is the conductive flux, and $\Lambda(T)$ represents the
279: radiative losses per unit emission measure \citep[e.g.][]{rs77, mgv85,
280: 2000adnx.conf..161K}. We use the ideal gas law, $P=(\gamma-1) \rho
281: \epsilon$.
282:
283: In order to track the original cloud material, we use a tracer that
284: is passively advected in the same manner as the density. We define
285: $C_{\rm cl}$ the mass fraction of the cloud inside the computational
286: cell. The cloud material is initialized with $C_{\rm cl} = 1$, while
287: $C_{\rm cl} = 0$ in the ambient medium\footnote{We checked that the
288: used numerical scheme ensures that always $0 \leq C_{\rm cl} \leq 1$.}.
289: During the shock-cloud evolution, the cloud and the ambient medium mix
290: together, leading to regions with $0 < C_{\rm cl} < 1$. At any time $t$
291: the density of cloud material in a fluid cell is given by $\rho_{\rm cl}
292: = \rho C_{\rm cl}$.
293:
294: The thermal conductivity in an organized magnetic field is known to be
295: highly anisotropic and it can be extraordinarily reduced in the direction
296: transverse to the field. The thermal flux, therefore, is locally split
297: into two components, along and across the magnetic field lines,
298: ${\bf F}_{\rm c} = F_{\parallel}~{\bf i}+F_{\perp}~{\bf j}$, where
299:
300: \begin{equation}
301: \begin{array}{l}\displaystyle
302: F_{\parallel} = \left(\frac{1}{[q_{\rm spi}]_{\parallel}}+
303: \frac{1}{[q_{\rm sat}]_{\parallel}}\right)^{-1}~,
304: \\ \\ \displaystyle
305: F_{\perp} = \left(\frac{1}{[q_{\rm spi}]_{\perp}}+
306: \frac{1}{[q_{\rm sat}]_{\perp}}\right)^{-1}~,
307: \end{array}
308: \label{cond}
309: \end{equation}
310:
311: \noindent
312: to allow for a smooth transition between the classical and saturated
313: conduction regime. In Eqs. \ref{cond}, $[q_{\rm spi}]_{\parallel}$ and
314: $[q_{\rm spi}]_{\perp}$ represent the classical conductive flux along
315: and across the magnetic field lines \citep{spi62}
316:
317: \begin{equation}
318: \begin{array}{l}\displaystyle
319: [q_{\rm spi}]_{\parallel} = -\kappa_{\parallel} [\nabla T]_{\parallel}
320: \approx - 5.6\times 10^{-7} T^{5/2}~ [\nabla T]_{\parallel}
321: \\ \\ \displaystyle
322: [q_{\rm spi}]_{\perp} = -\kappa_{\perp} [\nabla T]_{\perp}
323: \approx - 3.3\times 10^{-16} \frac{n^2_{\rm H}}{T^{1/2}B^2}~ [\nabla
324: T]_{\perp}
325: \end{array}
326: \label{spit_eq}
327: \end{equation}
328:
329: \noindent
330: where $[\nabla T]_{\parallel}$ and $[\nabla T]_{\perp}$ are the thermal
331: gradients along and across the magnetic field, and $\kappa_{\parallel}$ and
332: $\kappa_{\perp}$ (in units of erg s$^{-1}$ K$^{-1}$ cm$^{-1}$) are the
333: thermal conduction coefficients along and across the magnetic field
334: lines\footnote{For the values of $T$, $n_{\rm H}$ and $B$ used here,
335: $\kappa_{\parallel}/\kappa_{\perp}\approx 10^{16}$ at the beginning of
336: the shock-cloud interaction.}, respectively. The saturated flux along
337: and across the magnetic field lines, $[q_{\rm sat}]_{\parallel}$ and
338: $[q_{\rm sat}]_{\perp}$, are \citep{cm77}
339:
340: \begin{equation}
341: \begin{array}{l}\displaystyle
342: [q_{\rm sat}]_{\parallel} = -\mbox{sign}\left([\nabla T]_{\parallel}\right)~
343: 5\phi \rho c_{\rm s}^3,
344: \\ \\ \displaystyle
345: [q_{\rm sat}]_{\perp} = -\mbox{sign}\left([\nabla T]_{\perp}\right)~
346: 5\phi \rho c_{\rm s}^3,
347: \end{array}
348: \label{therm}
349: \end{equation}
350:
351: \noindent
352: where $c_{\rm s}$ is the isothermal sound speed, and $\phi$ is a number
353: of the order of unity; we set $\phi = 0.3$ according to the values
354: suggested for a fully ionized cosmic gas: $0.24<\phi<0.35$ \citep[][
355: and references therein]{1984ApJ...277..605G, 1989ApJ...336..979B,
356: 2002A&A...392..735F}. As discussed in Paper I, this choice implies that
357: no thermal precursor develops during the shock propagation, consistent
358: with the fact that no precursor is observed in young and middle aged SNRs.
359:
360: The initial unperturbed ambient medium is magnetized, isothermal
361: (with temperature $T_{\rm ism}=10^4$ K, corresponding to an isothermal
362: sound speed $c_{\rm ism} = 11.5$ km s$^{-1}$), and uniform (with hydrogen
363: number density $n_{\rm ism} = 0.1$ cm$^{-3}$; see Table \ref{tab1}).
364: The gas cloud is in pressure equilibrium with its surrounding and has
365: a circular cross-section with radius $r_{\rm cl} = 1$~pc; its radial
366: density distribution is given by
367:
368: \begin{equation}
369: n_{\rm cl}(r) = n_{\rm ism}+\frac{n_{\rm cl0}-n_{\rm ism}}
370: {\cosh\left[\sigma\left(r/r_{\rm cl}\right)^{\sigma}\right]}~~~,
371: \end{equation}
372:
373: \noindent
374: where $n_{\rm cl0}$ is the hydrogen number density at the cloud center, $r$
375: is the radial distance from the cloud center and $\sigma=10$. The above
376: distribution describes a thin transition layer ($\sim 0.3~r_{\rm cl}$)
377: around the cloud that smoothly brings the cloud density to the value
378: of the surrounding medium\footnote{A finite transition layer, in
379: general, is expected in real interstellar clouds due, for instance,
380: to thermal conduction (\citealt{1986ApJ...304..787B}; see also
381: \citealt{2006ApJS..164..477N}).}. The initial density contrast between
382: the cloud center and the ambient medium is $\chi=n_{\rm cl0}/n_{\rm ism}
383: = 10$. The cloud temperature is determined by the pressure balance
384: across the cloud boundary.
385:
386: \begin{deluxetable}{llll}
387: \footnotesize
388: \tablecaption{Summary of the initial physical parameters characterizing the
389: MHD simulations. \label{tab1}}
390: \tablewidth{0pt}
391: \tablehead{
392: \colhead{ } & \colhead{Temperature} & \colhead{Density} & \colhead{Velocity}
393: }
394: \startdata
395: {\it ISM} & $10^4$ K & $0.1$ cm$^{-3}$ & 0.0 \\
396: {\it Cloud} & $10^3$ K & $1.0$ cm$^{-3}$ & 0.0 \\
397: {\it Post-shock medium:} & $4.7\times 10^6$ K & $0.4$ cm$^{-3}$ & $430$
398: km s$^{-1}$ \\
399: \enddata
400: \end{deluxetable}
401:
402: The SNR shock front propagates with a velocity $w= \mach c_{\rm ism}$
403: in the ambient medium, where $\mach$ is the shock Mach number, and
404: $c_{\rm ism}$ is the sound speed in the interstellar medium; we consider
405: a shock propagating with $\mach = 50$, i.e. a shock velocity $w\approx
406: 570$ km s$^{-1}$ and a temperature $T_{psh}\approx 4.7\times 10^6$~K.
407: As discussed in Paper I, in this case (for a cloud with $r_{\rm cl} =
408: 1$~pc and $\chi=10$) the cloud dynamics would be dominated by thermal
409: conduction in the absence of magnetic field. The post-shock conditions
410: of the ambient medium well before the impact onto the cloud are given
411: by the strong shock limit \citep{zel66}.
412:
413: Starting from this basic configuration, we consider a set of simulations
414: with different initial magnetic field orientations. We adopt a 2.5D
415: Cartesian coordinate system $(x,y)$, implying that the simulated clouds
416: are cylinders extending infinitely along the $z$ axis perpendicular to
417: the $(x,y)$ plane. The primary shock propagates along the $y$ axis.
418: In this geometry, we consider three different field orientations: 1)
419: parallel to the planar shock and perpendicular to the cylindrical cloud,
420: 2) perpendicular to both the shock front and the cloud, and 3) parallel
421: to both the shock and the cloud. The magnetic field components along the
422: $x$ and the $z$ axis are enhanced by a factor $(\gamma+1)/(\gamma-1)$
423: (where $\gamma$ is the ratio of specific heats) in the post-shock region
424: \citep[in the strong shock limit;][]{zel66}, whereas the component
425: along the $y$ axis is continuous across the shock. We include runs in
426: the strong and weak magnetic field limits, considering initial field
427: strengths of $|\mag| = 2.63,\,1.31,\,0.26,\,0~\mu$G in the unperturbed
428: ambient medium\footnote{The unmagnetized case (i.e. $|\mag| =0$)
429: described here is analogous to the one studied in Paper I except for
430: the fact that in the present case the cloud is a cylinder rather than
431: a sphere and has smooth boundaries.}, corresponding to $\beta_0 =
432: 1,\,4,\,100,\,\infty$, where $\beta_{\rm 0} = P/(B^2/8\pi)$ is the
433: ratio of thermal to magnetic pressure in the pre-shock region. This
434: range of $\beta_{\rm 0}$ includes typical values inferred for the
435: diffuse regions of the ISM \citep[e.g.][]{2004RvMP...76..125M}
436: and for shock-cloud interaction regions in evolved SNR shells
437: \citep[e.g][]{2000A&A...359..316B}. There is no magnetic field component
438: exclusively associated to the cloud.
439:
440: We follow the shock-cloud interaction for $3.5\,\tau_{\rm cc}$, where
441: $\tau_{\rm cc}\approx \chi^{1/2} r_{\rm cl}/w$ is the cloud crushing time,
442: i.e. the characteristic time of the shock transmission through the
443: cloud; for the conditions considered here ($\chi=10$ and $\mach = 50$),
444: $\tau_{\rm cc} \approx 5.4\times 10^3$ yr. Each simulation is repeated
445: either with or without thermal conduction for each field orientation.
446: Table \ref{tab2} lists the runs and the initial physical parameters of
447: the simulations.
448:
449: \begin{deluxetable}{lcccccc}
450: \footnotesize
451: \tablecaption{Summary of the MHD simulations. In all runs the shock Mach
452: number is $\mach=50$, the density contrast is $\chi=10$, and the
453: cloud crushing time is $\tau_{\rm cc} \approx 5.4\times 10^3$ yr.
454: \label{tab2}}
455: \tablewidth{0pt}
456: \tablehead{
457: \colhead{Run} & \colhead{$|\mag|$} & \colhead{$\beta_0$} & \colhead{Field} &
458: \colhead{Therm.} & \colhead{Rad.} & \colhead{Res.$^a$}\\
459: \colhead{ } & \colhead{$\mu$G} & \colhead{ } & \colhead{Comp.} &
460: \colhead{Cond.} & \colhead{Losses} \\
461: }
462: \startdata
463: NN & 0 & $\infty$ & $-$ & no & no & 132 \\
464: NR & 0 & $\infty$ & $-$ & no & yes & 132 \\
465: TN & 0 & $\infty$ & $-$ & yes & no & 132 \\
466: TR & 0 & $\infty$ & $-$ & yes & yes & 132 \\
467: NN-Bx4 & 1.31 & 4 & $B_{\rm x}$ & no & no & 132 \\
468: NN-By4 & 1.31 & 4 & $B_{\rm y}$ & no & no & 132 \\
469: NN-Bz4 & 1.31 & 4 & $B_{\rm z}$ & no & no & 132 \\
470: TN-Bx4 & 1.31 & 4 & $B_{\rm x}$ & yes & no & 132 \\
471: TN-By4 & 1.31 & 4 & $B_{\rm y}$ & yes & no & 132 \\
472: TN-Bz4 & 1.31 & 4 & $B_{\rm z}$ & yes & no & 132 \\
473: TR-Bx1 & 2.63 & 1 & $B_{\rm x}$ & yes & yes & 132 \\
474: TR-By1 & 2.63 & 1 & $B_{\rm y}$ & yes & yes & 132 \\
475: TR-Bz1 & 2.63 & 1 & $B_{\rm z}$ & yes & yes & 132 \\
476: TR-Bx4 & 1.31 & 4 & $B_{\rm x}$ & yes & yes & 132 \\
477: TR-By4 & 1.31 & 4 & $B_{\rm y}$ & yes & yes & 132 \\
478: TR-Bz4 & 1.31 & 4 & $B_{\rm z}$ & yes & yes & 132 \\
479: TR-Bx100 & 0.26 & 100 & $B_{\rm x}$ & yes & yes & 132 \\
480: TR-By100 & 0.26 & 100 & $B_{\rm y}$ & yes & yes & 132 \\
481: TR-Bz100 & 0.26 & 100 & $B_{\rm z}$ & yes & yes & 132 \\
482: \\
483: TR-Bz4-hr & 1.31 & 4 & $B_{\rm z}$ & yes & yes & 264 \\
484: TR-Bz4-hr2 & 1.31 & 4 & $B_{\rm z}$ & yes & yes & 528 \\
485: \enddata
486: \tablenotetext{a}{Initial number of zones per cloud radius}
487: \end{deluxetable}
488:
489: \begin{figure*}
490: \centering
491: \epsscale{1.07}
492: \plotone{f1.ps}
493: \caption{Mass density distribution (gm cm$^{-3}$) in the $(x,y)$
494: plane, in log scale, in the simulations NN (left half panels) and
495: TR (right half panels), sampled at the labeled times in units of
496: $\tau_{\rm cc}$. The contour encloses the cloud material.}
497: \label{fig1}
498: \end{figure*}
499:
500: We solve numerically the set of MHD equations using \FLASH\
501: \citep{for00}, a multiphysics code including the \PARAMESH\
502: library \citep{mom00} for the adaptive mesh refinement. The MHD
503: equations are solved using the \FLASH\ implementation of the HLLE
504: scheme \citep{einfeldt88}. The code has been extended with additional
505: computational modules to handle the radiative losses and the
506: anisotropic thermal conduction \citep[see][ for the details of the
507: implementation]{2007A&A...464..753P}.
508:
509: The 2.5D Cartesian $(x,y)$ grid extends between $-4$ and 4~pc in the $x$
510: direction and between $-1.4$ and 6.6~pc in the $y$ direction. Initially
511: the cloud is located at $(x,y) = (0,0)$ and the primary shock front
512: propagates in the direction of the $y$ axis. At the coarsest resolution,
513: the adaptive mesh algorithm used in the \FLASH\ code uniformly covers
514: the 2.5D computational domain with a mesh of $4^2$ blocks, each with
515: $8^2$ cells. We allow for 5 levels of refinement, with resolution
516: increasing twice at each refinement level. The refinement criterion
517: adopted \citep{loehner} follows the changes of the density and of the
518: temperature. This grid configuration yields an effective resolution of
519: $\approx 7.6\times 10^{-3}$ pc at the finest level, corresponding to
520: $\approx 132$ cells per cloud radius. In Sect. \ref{sp_resol}, we
521: discuss the effect of spatial resolution on our results, considering the
522: additional runs TR-Bz4-hr and TR-Bz4-hr2 which use an identical setup to
523: run TR-Bz4, but with higher resolution ($\approx 264$ and $\approx 528$
524: cells per cloud radius, respectively; see Table \ref{tab2}).
525:
526: We use a constant inflow boundary condition for the post-shock gas at the
527: lower boundary, with free outflow elsewhere. For runs with zero magnetic
528: field ($\beta_{\rm 0} = \infty$), we use reflecting boundary conditions
529: at $x = 0$ along the symmetry axis of the problem and only evolve half
530: of the grid.
531:
532: %________________________________________________________________
533:
534: \section{Results}
535: \label{sec3}
536:
537: \subsection{Dynamical evolution}
538: \label{d_evol}
539:
540: Figs.~\ref{fig1} and \ref{fig2} show the evolution of the mass density in
541: the $(x,y)$ plane in the simulations with $\beta_0 = \infty$ (runs NN and
542: TR) and with $\beta_0 = 4$ (runs NN-Bx4, NN-By4, NN-Bz4, TR-Bx4, TR-By4,
543: TR-Bz4). The left (right) half panels show the result of models without
544: (with) thermal conduction and radiative losses.
545:
546: \begin{figure*}[!h]
547: \centering
548: \epsscale{1.07}
549: \plotone{f2.ps}
550: \caption{As in Fig.~\ref{fig1} for the simulations with $\beta_0 =
551: 4$ and the magnetic field oriented along $x$ (upper panels), $y$ (middle
552: panels), and $z$ (lower panels). The figure shows the distribution in
553: models either without (left half panels) or with (right half panels)
554: thermal conduction and radiative losses. For runs NN-Bx4, TR-Bx4,
555: NN-By4 and TR-By4, we plot the magnetic field lines; for runs NN-Bz4
556: and TR-Bz4, we include contours of $\log (B^2/8\pi)$.}
557: \label{fig2}
558: \end{figure*}
559:
560: From Fig.~\ref{fig1}, we note that the thermal conduction drives
561: the cloud evolution in the unmagnetized case ($\beta_0 = \infty$;
562: run TR): after the initial compression due to the primary shock, the
563: cloud expands and gradually evaporates due to the heating driven by the
564: thermal conduction in a few dynamical timescales (see right half panels
565: in Fig.~\ref{fig1}). The heat conduction strongly contrasts the radiative
566: cooling of some parts of the cloud and no thermal and hydrodynamic
567: instabilities (visible in run NN; see left-panels in Fig.~\ref{fig1})
568: develops during the cloud evolution, making the cloud more stable and
569: longer-living (the mass mixing is strongly reduced; see Paper I for
570: more details).
571:
572: We now discuss the effect of the magnetic-field-oriented (anisotropic)
573: thermal conduction on the shock-cloud collision when an ambient magnetic
574: field permeates the ISM. We first summarize the expected evolution in the
575: presence of an ambient magnetic field, according to the well-established
576: results of previous models without thermal conduction. We distinguish
577: between fields perpendicular to the cylindrical clouds (i.e. with
578: only $B_{\rm x}$ and $B_{\rm y}$ components; referred to as ``external''
579: fields by \citealt{2005ApJ...619..327F}) and fields parallel to the
580: cylindrical clouds (i.e. with only the $B_{\rm z}$ component; referred to
581: as ``internal'' fields). In the former case, the magnetic field plays
582: a dominant role along the cloud surface and in the wake of the cloud
583: where it reaches its highest strength (and the plasma $\beta$ its lowest
584: values; e.g. \citealt{1994ApJ...433..757M}; \citealt{1996ApJ...473..365J}). In
585: the case of $B_{\rm x}$, the magnetic field is trapped at the nose of
586: the cloud, leading to a continuous increase of the magnetic pressure
587: and field tension there (see upper panels in Fig.~\ref{fig2}); in the
588: case of $B_{\rm y}$, the cloud expansion leads to the increase of magnetic
589: pressure and field tension laterally to the cloud (see middle panels in
590: Fig.~\ref{fig2}). In the case of $B_{\rm z}$ (internal field), the magnetic
591: field, being parallel to the cylindrical cloud, modifies only the total
592: effective pressure of the plasma (\citealt{1996ApJ...473..365J}); in
593: the case of radiating shocks, the additional magnetic pressure may play
594: a crucial role in the shocked cloud, preventing further compression of
595: the cloud material (\citealt{2005ApJ...619..327F}).
596:
597: \subsubsection{External magnetic fields}
598: \label{external_field}
599:
600: In the case of predominantly external magnetic fields,
601: \cite{1994ApJ...433..757M} and \cite{1996ApJ...473..365J} have shown
602: that the hydrodynamic instabilities can be suppressed even in models
603: neglecting the thermal conduction due to the tension of the magnetic
604: field lines which maintain a more laminar flow around the cloud surface
605: (see also \citealt{2005ApJ...619..327F}): for a $\gamma=5/3$ gas,
606: the KH instabilities are suppressed if $\beta < 2/\mach^2$, whereas
607: RT instabilities are suppressed if $\beta < (2/\gamma)(\chi/\mach)^2$
608: (see also \citealt{1961hhs..book.....C}). However, for the parameters
609: used in this paper ($\chi = 10$ and $\mach = 50$), the magnetic field
610: cannot suppress KH instabilities in any of our runs, whereas the RT
611: instabilities are suppressed only in runs that lead to locally very strong
612: field ($\beta < 0.05$). This can be seen in model NN-Bx4 (upper panels in
613: Fig.~\ref{fig2}), presenting a large field increase at the cloud boundary,
614: compared to model NN (Fig.~\ref{fig1}): in the latter case the growth
615: of KH and RT instabilities at the cloud boundary is much more evident
616: than in NN-Bx4. On the other hand, the hydrodynamic instabilities are
617: suppressed more efficiently in models including the thermal conduction
618: (runs TR-Bx4 and TR-By4) even in cases with low field increase (for
619: instance in our $B_{\rm y}$ case) as it is evident in Fig.~\ref{fig2}
620: by comparing models NN-Bx4 and NN-By4 with models TR-Bx4 and TR-By4,
621: respectively.
622:
623: \begin{figure*}
624: \centering
625: \epsscale{1.07}
626: \plotone{f3.ps}
627: \caption{Heat flux (upper panels) and magnetic field
628: strength (lower panels) distributions in the $(x,y)$ plane in the
629: simulations TR-Bx4 (left panels), TR-By4 (center), and TR-Bz4 (right),
630: at time $t=2~\tau_{\rm cc}$. The arrows in the upper panels describe
631: the heat flux and scale linearly with respect to the reference value
632: shown in the upper right corner of each panel. The scale of the
633: magnetic field strength is linear and is given by the bar on the
634: right, in units of 10~$\mu$G. The red contour encloses the cloud
635: material.}
636: \label{fig3}
637: \end{figure*}
638:
639: The thermal exchanges between the cloud and the surrounding
640: medium strongly depend on the initial field orientation.
641: Fig.~\ref{fig3} shows the heat flux and magnetic field strength
642: distributions in the $(x,y)$ plane in runs TR-Bx4, TR-By4, and TR-Bz4,
643: at time $t=2~\tau_{\rm cc}$. In our $B_{\rm x}$ case (upper panels
644: in Fig.~\ref{fig2}), the magnetic field lines gradually envelope
645: the cloud, reducing the heat conduction through the cloud surface
646: (see left panels in Fig.~\ref{fig3}): thermal exchanges
647: between the cloud and the surrounding medium are channelled through
648: small regions located at the side of the cloud. The cloud expansion
649: and evaporation are strongly limited by the confining effect of the
650: magnetic field (cf. the unmagnetized case TR in Fig.~\ref{fig1} with
651: model TR-Bx4 in Fig.~\ref{fig2}) that becomes up to 30 times stronger
652: just outside the cloud than inside it (see, also, the lower left
653: panel in Fig.~\ref{fig3}). The consequent thermal insulation induces
654: the radiative cooling and condensation of the plasma into the cloud
655: during the phase of cloud compression ($t < \tau_{\rm cc}$). At the
656: end of this phase, the cloud material has temperature $T\approx 10^5$~K
657: and density $n_{\rm H}\approx 10$~cm$^{-3}$ where primary and reverse
658: shocks transmitted into the cloud are colliding; for these values of $T$
659: and $n_{\rm H}$, the Field length scale (\citealt{1990ApJ...358..375B})
660: derived from the ratio of cooling timescale over conduction timescale
661: (see Paper I for details) is
662:
663: \begin{equation}
664: l \approx 10^6 \frac{T^2}{n_{\rm H}} \approx 3.2\times
665: 10^{-4}\mbox{~pc}~.
666: \label{ratio}
667: \end{equation}
668:
669: \noindent
670: The radiative cooling dominates over the effects of the thermal conduction
671: in cold and dense regions with dimensions larger than $l$. At variance
672: with our unmagnetized case TR, therefore, thermal instabilities develop
673: in run TR-Bx4. One of this cold and dense structures is evident in
674: Fig.~\ref{fig2} (upper panels) and is located at the cloud boundary
675: near the nose of the cloud (at $x\approx 0.4$ pc and $y\approx 3.0$ pc)
676: at $t=3~\tau_{\rm cc}$.
677:
678: In the $B_{\rm y}$ case, the initial field direction is mostly maintained
679: in the cloud core during the evolution, allowing efficient thermal
680: exchange between the core and the hot medium upwind of the cloud (see
681: center panels in Fig.~\ref{fig3}): the core is gradually heated and
682: evaporates in few dynamical timescales. This is illustrated by run
683: TR-By4 in Fig.~\ref{fig2}. On the other hand, the cloud is thermally
684: insulated laterally where the magnetic field lines prevent thermal
685: exchange between the cloud and the surrounding medium. Also, a strong
686: magnetic field component along the $x$ axis develops in the wake of the
687: cloud and inhibits thermal conduction with the medium downwind of the
688: cloud. The thermal insulation at the side of the cloud determines the
689: growth of thermal instabilities where shocks transmitted into the cloud
690: collide (see middle panels in Fig.~\ref{fig2}).
691:
692: In both external field configurations, elongated structures of strong
693: field concentration are produced on the axis downwind of the cloud
694: due to the focalization of the magnetized fluid flows there (see
695: upper and middle panels of Fig.~\ref{fig2}, and lower panels in
696: Fig.~\ref{fig3}). These filamentary structures, identified as ``flux
697: ropes'' by \cite{1994ApJ...433..757M}, are formed by magnetic field lines
698: stretched around the cloud shape and do not carry a significant amount of
699: cloud material (as shown by the tracer $C_{\rm cl}$) although the plasma
700: there moves with the cloud (see also \citealt{2000ApJ...543..775G}).
701:
702: \subsubsection{Internal magnetic fields}
703:
704: Predominantly internal magnetic fields strongly suppress the heat
705: conduction, providing an efficient thermal insulation of the cloud
706: material (see right panels in Fig.~\ref{fig3}). In the realistic
707: configuration of an elongated cloud with finite length $L$ along the $z$
708: axis, some heat would be conducted along the magnetic field lines. The
709: characteristic timescales for the conduction along magnetic field lines is
710: (see Paper I)
711:
712: \begin{equation}
713: \tau_{\rm cond}\approx 2.6\times 10^{-9} \frac{n_{\rm H}L^2}{T^{5/2}}~.
714: \end{equation}
715:
716: \noindent
717: We estimate that the cloud would thermalize in $\tau_{\rm cond}>
718: 3.5~\tau_{\rm cc}$ (i.e. the physical time covered by our simulations),
719: if the length scale of the cloud along the $z$ axis is $L> 3$~pc. In
720: this case, hydrodynamic instabilities develop at the cloud boundary, being
721: both the magnetic field and the thermal conduction not able to suppress
722: them. The growth of these instabilities is clearly seen in Fig.~\ref{fig2}
723: (lower panels). The combined effect of hydrodynamic instabilities
724: and shocks transmitted into the cloud leads to unstable high-density
725: regions at the cloud boundaries that trigger the development of thermal
726: instabilities there (see lower panels in Fig.~\ref{fig2}). However,
727: as discussed by \cite{2005ApJ...619..327F}, internal magnetic field
728: lines are expected to resist compression in the shocked cloud, thus
729: reducing the cooling efficiency. In fact, in our run TR-Bz4, the cloud
730: material is prevented from cooling below $T\approx 10^3$~K. Since the
731: thermal conduction does not play any significant role in the shock-cloud
732: interaction, our $B_{\rm z}$ case leads to results similar to those obtained
733: by \cite{2005ApJ...619..327F} and we do not discuss further this case.
734:
735: \subsection{Role of thermal conduction}
736:
737: In this section, we study more quantitatively the effect of thermal
738: conduction on the cloud evolution and, in particular, on the cloud
739: compression and on the magnetic field increase. To this end, we use the
740: tracer defined in Sect. \ref{sec2} to identify zones whose content is
741: the original cloud material by more than 90\%. Then, we define the
742: cross-sectional area of cloud material, $A_{\rm cl}(t)$, as the total
743: area in the $(x,y)$ plane occupied by these zones. We define the cloud
744: compression (or expansion) as $A_{\rm cl}/A_{\rm cl0}$, where $A_{\rm
745: cl0}$ is the initial cross-sectional area. We also define an average
746: mass-weighted temperature of the cloud and an average magnetic field
747: strength associated to the cloud as
748:
749: \begin{equation}
750: \langle T \rangle_{\rm cl} = \frac{\displaystyle
751: \int_{A(C_{\rm cl}>0.9)} C_{\rm cl}~\rho T ~da}{\displaystyle
752: \int_{A(C_{\rm cl}>0.9)} C_{\rm cl}~\rho ~da}
753: \end{equation}
754:
755: \begin{equation}
756: \langle B \rangle_{\rm cl} = \frac{\displaystyle
757: \int_{A(C_{\rm cl}>0.9)} C_{\rm cl}~B ~da}{\displaystyle
758: \int_{A(C_{\rm cl}>0.9)} C_{\rm cl}~da}
759: \end{equation}
760:
761: \begin{figure*}
762: \centering
763: \epsscale{1.0}
764: \plotone{f4.ps}
765: \caption{Evolution of cloud compression (upper panels),
766: of average temperature (middle panels), and of average magnetic
767: field strength (lower panels) of the cloud for runs which neglect the
768: thermal conduction and the radiation (dot-dashed lines; left panels),
769: for runs which include the thermal conduction but neglect the radiation
770: (solid; left panels), for runs which include the radiation but neglect
771: the thermal conduction (dotted; right panels) and for runs which
772: include both physical effects (dashed; right panels). The magnetized
773: cases with $\beta_{\rm 0} = 4$ are marked with red (initial magnetic
774: field along $x$), green (initial {\mag} along $y$) and blue (initial
775: {\mag} along $z$) lines; the unmagnetized cases are marked with black
776: lines. The light yellow regions mark the location of solutions which
777: have thermodynamical characteristics in between the cases of maximum
778: efficiency of the thermal conduction (models TN and TR) and the
779: cases without thermal conduction (models NN and NR). By comparing the
780: position of the magnetized models curves inside the yellow region, it
781: is possible to quantitatively assess the degree of suppression of the
782: effects of the thermal conduction by the magnetic fields.}
783: \label{fig4}
784: \end{figure*}
785:
786: \noindent
787: where we integrate on zones with $C_{\rm cl}>0.9$. Note that our
788: choice of considering cells with the value of the passive tracer $C_{\rm
789: cl}>0.9$ is arbitrary. To determine how sensitive the results are to this
790: value and, in particular, to small changes in it, we derive our results
791: also considering the values $C_{\rm cl}>0.85$ and $C_{\rm cl}>0.95$.
792: In all the cases, we find that the results derived with the different
793: thresholds show the same trend with differences lower than $10\%$.
794:
795: Fig.~\ref{fig4} shows the cloud compression, $A_{\rm cl}/A_{\rm cl0}$,
796: the average temperature of the cloud, $\langle T \rangle_{\rm cl}$,
797: normalized to the post-shock temperature of the surrounding medium
798: ($T_{\rm psh}=4.7\times 10^6$ K), and the average magnetic field strength
799: associated to the cloud, $\langle B \rangle_{\rm cl}$, normalized to
800: the initial field strength ($B_{\rm 0} = 1.31\;\mu$G, corresponding to
801: $\beta_{\rm 0}=4$) as a function of time for models neglecting thermal
802: conduction and radiation (hereafter NNs models), for models including
803: conduction but neglecting radiation (TNs models), and for models including
804: both conduction and radiation (TRs models); we also include the results
805: derived from the unmagnetized case NR with radiative cooling and without
806: thermal conduction. The figure shows both the magnetized cases with
807: $\beta_{\rm 0}=4$ and the unmagnetized cases (see Table \ref{tab2}).
808:
809: In all the NNs models either with (NN-Bx4, NN-By4, and NN-Bz4) or without
810: (NN) the magnetic field, the evolution of the cloud compression and
811: of the average cloud temperature is roughly the same (see left panels
812: in Fig.~\ref{fig4}). The cloud is initially compressed over a timescale
813: $t\approx \tau_{\rm cc}$ due to the ambient post-shock pressure; during this
814: phase $\langle T \rangle_{\rm cl}$ rapidly increases. After $t\approx
815: \tau_{\rm cc}$, the cloud partially reexpands, leading to a decrease of
816: $\langle T \rangle_{\rm cl}$. In the last phase ($t > 2.0~\tau_{\rm cc}$),
817: the cloud is compressed again by the interaction with the ``Mach stem''
818: formed during the reflection of the primary shock at the symmetry axis,
819: and $\langle T \rangle_{\rm cl}$ increases; later $A_{\rm cl}/A_{\rm cl0}$
820: continues to decrease, because of the mixing of the cloud material with
821: the ambient medium (see Sect. \ref{mix}; see also Paper I), while $\langle
822: T \rangle_{\rm cl}$ stabilizes at $\approx 0.17\;T_{\rm psh}$.
823:
824: The field increase in the cloud material depends on the initial
825: configuration of {\mag} (see lower left panel in Fig.~\ref{fig4}). In
826: the case of external fields ($B_{\rm x}$ and $B_{\rm y}$ components), {\mag}
827: is mainly intensified due to stretching of field lines due to sheared
828: motion. In the $B_{\rm x}$ case, the magnetic field undergoes the greatest
829: increase and $\langle B \rangle_{\rm cl}$ keeps increasing during
830: the whole evolution. In fact the field is mainly intensified at the nose
831: of the cloud where the background flow continues to stretch the field
832: lines during the evolution (see upper panels in Fig.~\ref{fig2}). In the
833: $B_{\rm y}$ case, the field increase occurs mainly at the side of the
834: cloud where the field lines are stretched along the cloud surface. In the
835: case of internal fields ($B_{\rm z}$ component), the field increase
836: is due to squeezing of field lines through compression. $\langle B
837: \rangle_{\rm cl}$, therefore, follows the changes in $A_{\rm cl}/A_{\rm cl0}$,
838: since the field is locked within the cloud material. Thus the greatest
839: field increase occurs at $t\approx \tau_{\rm cc}$ when the shocks
840: transmitted into the cloud collide.
841:
842: The effects of thermal conduction are greatest in the unmagnetized
843: model (TN) which can be considered an extreme limit case (see left
844: panels in Fig.~\ref{fig4}). During the first stage of evolution
845: ($t < 0.8\;\tau_{\rm cc}$), the cloud is heated efficiently by the
846: thermal conduction and its average temperature increases rapidly to
847: $\sim 0.5~T_{\rm psh}$. As a consequence, the pressure inside the cloud
848: increases and the cloud reexpands earlier than in model NN. Afterwards,
849: the average cloud temperature, $\langle T \rangle_{\rm cl}$, keeps
850: increasing up to $\sim 0.9~T_{\rm psh}$ at $t=3.5\;\tau_{\rm cc}$.
851:
852: In the case of predominantly external magnetic fields (models TN-Bx4
853: and TN-By4), the thermal conduction still plays a significant role in
854: the cloud evolution, although its effects are not as large as in the
855: unmagnetized case (TN). During the initial compression, the thermal
856: conduction contributes to the cloud heating: the average temperature of
857: the cloud reaches values larger than in models neglecting the conduction
858: (compare TNs with NNs models in the left panels of Fig.~\ref{fig4}). This
859: effect is greatest in the $B_{\rm y}$ case which is the configuration
860: of field lines that allows the most efficient thermal exchange
861: between the cloud and the hot environment (see Sect. \ref{d_evol}). At
862: $t=3.5\;\tau_{\rm cc}$, $\langle T \rangle_{\rm cl}$ in TNs models reaches
863: values larger than in NNs models ($\approx 0.21\;T_{\rm psh}$ in the
864: $B_{\rm x}$ case and $\approx 0.33\; T_{\rm psh}$ in the $B_{\rm y}$ case).
865: For internal magnetic fields, the thermal conduction plays no role in
866: the evolution of the shocked cloud, being strongly ineffective due to
867: {\mag} (see Sect. \ref{d_evol}). As a consequence, the TN-Bz4 model
868: leads to the same results as NNs models.
869:
870: In general, therefore, the effects of the thermal conduction in the
871: presence of an ambient magnetic field are reduced with respect to the
872: corresponding unmagnetized case, but not entirely suppressed. This
873: can be seen in Fig. \ref{fig4}, where we have marked in light yellow
874: the region between the fully conductive unmagnetized case (TN) and the
875: case without thermal conduction (NN). The magnetized TNs models are always
876: within this region, meaning that the effects of the thermal conduction
877: are never as large as in the unmagnetized case (TN) but not completely
878: suppressed as in the model NN.
879:
880: We also note that the thermal conduction influences indirectly the
881: magnetic field increase. The main changes are in the $B_{\rm y}$ case and
882: are due to the larger expansion of the cloud that reduces the increase
883: of the field associated to the cloud, being the field locked within the
884: cloud material.
885:
886: In our unmagnetized case TR (including thermal conduction and
887: radiative cooling), the thermal conduction prevents the onset of
888: thermal instabilities, and the evolution of the shocked cloud is
889: the same as found in the TN model. At variance with our unmagnetized
890: case TR, Fig.~\ref{fig4} shows that thermal instabilities develop in
891: all our magnetized TRs runs, being the effects of thermal conduction
892: reduced by the magnetic field. The effects of radiative cooling are
893: very strong for internal fields (our $B_{\rm z}$ case; see run TR-Bz4
894: in Fig.~\ref{fig4}). In this case, the heat conduction is totally
895: suppressed by the magnetic field and the evolution of $A_{\rm cl}/A_{\rm
896: cl0}$ and of $\langle T \rangle_{\rm cl}$ are the same as those found
897: in the unmagnetized case with radiative cooling and without thermal
898: conduction (model NR); at $t=3.5\;\tau_{\rm cc}$, run TR-Bz4 (and NR)
899: shows the largest cloud compression ($A_{\rm cl}/A_{\rm cl0} \approx
900: 0.1$) and the lowest cloud average temperature ($\langle T \rangle_{\rm
901: cl}\approx 0.12\;T_{\rm psh}$). In the case of external fields (runs
902: TR-Bx4 and TR-By4), the effects of heat conduction are reduced but not
903: suppressed and the results are intermediate between those derived for
904: runs NR and TR (i.e. within the light yellow region in the right panels
905: in Fig. \ref{fig4}). The cooling efficiency is largely reduced in our
906: $B_{\rm y}$ case (run TR-By4), namely that with the magnetic field
907: configuration that allows the most effective thermal conduction.
908:
909: \subsection{Mass mixing and energy exchange}
910: \label{mix}
911:
912: We use the tracer to derive the cloud mass, $M_{\rm cl}$, as the total
913: mass in zones whose content is the original cloud material by more
914: than 90\%,
915:
916: \begin{equation}
917: M_{\rm cl} = L~\int_{A(C_{\rm cl}>0.9)} C_{\rm cl}~\rho~da~,
918: \label{mass}
919: \end{equation}
920:
921: \noindent
922: where $L$ is the cloud length along the $z$ axis, and the integral is
923: done on zones with $C_{\rm cl}>0.9$. We investigate the mixing of cloud
924: material with the ambient medium by defining the remaining cloud mass
925: as $M_{\rm cl}/M_{\rm cl0}$, where $M_{\rm cl0}$ is the initial cloud mass.
926:
927: The tracer allows us to investigate also the energy exchange between
928: the cloud and the surrounding medium; we derive the internal energy,
929: ${\cal I}_{\rm cl}$, and the kinetic energy, ${\cal K}_{\rm cl}$ of the
930: cloud as
931:
932: \begin{equation}
933: {\cal I}_{\rm cl} = L~\int_{A(C_{\rm cl}>0.9)} C_{\rm cl}~\rho\epsilon~da~,
934: \label{eint}
935: \end{equation}
936:
937: \begin{equation}
938: {\cal K}_{\rm cl} = \frac{L}{2} \int_{A(C_{\rm cl}>0.9)} C_{\rm cl}~\rho
939: |{\bf u}|^2~da~,
940: \label{ekin}
941: \end{equation}
942:
943: \noindent
944: where again $L$ is the cloud length along the $z$ axis, and the integral
945: is done on zones with $C_{\rm cl}>0.9$. We also define the total energy
946: of the cloud as
947:
948: \begin{equation}
949: E_{\rm cl} = {\cal I}_{\rm cl}+{\cal K}_{\rm cl}~.
950: \end{equation}
951:
952: \begin{figure*}[!ht]
953: \centering
954: \epsscale{1.0}
955: \plotone{f5.ps}
956: \caption{Presentation as in Fig.~\ref{fig4} for the evolution
957: of the cloud mass (upper panels), of the internal energy of the
958: cloud (middle panels), and of the kinetic energy of the cloud (lower
959: panels).}
960: \label{fig5}
961: \end{figure*}
962:
963: Fig.~\ref{fig5} shows the evolution of the cloud mass, $M_{\rm
964: cl}/M_{\rm cl0}$, for NNs, TNs, and TRs models; again we also include the
965: unmagnetized case with radiative cooling and without thermal conduction
966: (model NR). Both unmagnetized cases and magnetized cases with $\beta_{\rm
967: 0}=4$ are shown. In models without thermal conduction and radiation
968: (NNs models), the hydrodynamic instabilities drive the mass mixing of
969: the cloud\footnote{This is also true in our magnetized cases because,
970: for the parameters used in this paper ($\mach = 50$ and $\chi=10$), the
971: hydrodynamic instabilities are partially suppressed by the magnetic field
972: only in runs evolving to strong fields (see Sect. \ref{d_evol}).}. The
973: mass loss rate of the cloud, $\dot{m}_{\rm cl}$, increases significantly
974: after $1.5\;\tau_{\rm cc}$ (i.e. after the hydrodynamic instabilities
975: have fully developed at the cloud boundary), with $\dot{m}_{\rm cl}
976: \approx 1.5\times 10^{-6} L_{\rm pc}~~M_{\odot}$~yr$^{-1}$, where $L_{\rm
977: pc}$ is the cloud length along the $z$ axis in units of pc: $\sim 20$\%
978: of the cloud mass is contained in mixed zones at $t = 3.5\;\tau_{\rm
979: cc}$. The only exception is run NN-Bx4 ($\sim 15$\% of the cloud mass
980: is in mixed zones at $t = 3.5\;\tau_{\rm cc}$), being in this case RT
981: instabilities partially suppressed by the magnetic field (compare run
982: NN-Bx4 with runs NN-By4 and NN-Bz4 in Fig.~\ref{fig2}).
983:
984: In TNs models with external magnetic fields (TN-Bx4 and TN-By4), the
985: mass loss rate of the cloud is less efficient than in NNs models with
986: $\dot{m}_{\rm cl}\approx 6\times 10^{-7} L_{\rm pc}~~M_{\odot}$~yr$^{-1}$
987: ($\sim 10$\% of the cloud mass is in mixed zones at $t = 3.5\;\tau_{\rm
988: cc}$). In fact, in these cases the thermal conduction suppresses most
989: of the hydrodynamic instabilities and the mass loss mainly comes from
990: the cloud evaporation driven by the thermal conduction rather than from
991: hydrodynamic ablation. Note that our unmagnetized TN model is an
992: extreme limit case in which the hydrodynamic instabilities are totally
993: suppressed by the thermal conduction which drives the cloud mixing; in
994: this case, the mass loss rate is $\dot{m}_{\rm cl}\approx 1.5\times
995: 10^{-7} L_{\rm pc}~~M_{\odot}$~yr$^{-1}$ ($\sim 5$\% of the cloud mass is
996: in mixed zones at $t = 3.5\;\tau_{\rm cc}$).
997:
998: In magnetized TRs models, the onset of thermal instabilities
999: increases the mass loss rate of the cloud with respect to the
1000: unmagnetized case ($\dot{m}_{\rm cl}$ ranges between $1.5\times
1001: 10^{-6} L_{\rm pc}~~M_{\odot}$~yr$^{-1}$ and $4\times 10^{-6} L_{\rm
1002: pc}~~M_{\odot}$~yr$^{-1}$) due to the fragmentation of the cloud in
1003: dense and cold cloudlets. We expect, therefore, that the larger the
1004: amount of cloud mass mixed with the surrounding medium at the end of the
1005: evolution, the more limited the thermal exchange between the cloud and
1006: the hot ambient medium (and, therefore, the greater the efficiency of
1007: radiative cooling). In fact, the upper right panel in Fig.~\ref{fig5}
1008: shows that the mass mixing has the greatest efficiency in run TR-Bz4
1009: (i.e. in the case with the thermal conduction totally suppressed) which
1010: shows a mass loss rate of the cloud similar to that derived from the
1011: unmagnetized NR model. On the other hand, in runs TR-Bx4 and TR-By4,
1012: the mass mixing is intermediate between those derived with runs NR and TR.
1013:
1014: Fig.~\ref{fig5} also shows the evolution of internal (middle panels)
1015: and kinetic (lower panels) energy of the cloud, normalized to the
1016: initial total energy of the cloud, $E_{\rm cl0}$. Among the magnetized
1017: cases considered, the greatest values of ${\cal I}_{\rm cl}$ are reached
1018: in our $B_{\rm y}$ case which is the field configuration that allows the
1019: most efficient thermal exchange between the cloud and the environment;
1020: the increase of ${\cal I}_{\rm cl}$ is due to the heat conducted to the
1021: shocked cloud. Also, the $B_{\rm y}$ case leads to the greatest values of
1022: ${\cal K}_{\rm cl}$ because the cloud has a larger cross-sectional area
1023: (because of the larger cloud expansion due to the heating driven by heat
1024: conduction; see upper panels in Fig.~\ref{fig4}) and offers, therefore,
1025: a larger surface to the pressure of the shock front responsible of the
1026: cloud acceleration.
1027:
1028: \subsection{Role of the initial field strength}
1029:
1030: In this section we explore the effects of the initial field strength
1031: on the mass mixing and energy exchange of the cloud. Fig. \ref{fig6}
1032: shows the evolution of the cloud mass, $M_{\rm cl}/M_{\rm cl0}$ (upper
1033: panel), and of the total (internal plus kinetic) energy of the cloud,
1034: $E_{\rm cl}/E_{\rm cl0}$ (lower panel), for magnetized TRs models with
1035: different values of $\beta_{\rm 0}$. We discuss here only the cases
1036: of predominantly external magnetic fields ($B_{\rm x}$ or $B_{\rm y}$
1037: case) since no significant dependence on the initial field strength
1038: has been found in the case of predominantly internal magnetic fields
1039: ($B_{\rm z}$ case).
1040:
1041: \begin{figure}[!t]
1042: \centering
1043: \epsscale{1.1}
1044: \plotone{f6.ps}
1045: \caption{Evolution of the cloud mass (upper panel) and of
1046: the total energy of the cloud (internal plus kinetic; lower panel) for
1047: runs including both the thermal conduction and the radiative cooling
1048: (TRs models). The figure shows the simulations with the magnetic field
1049: oriented along $x$ (red lines) or $y$ (green) and with $\beta_{\rm 0}
1050: = 1$ (dotted lines), 4 (solid), and 100 (dashed).}
1051: \label{fig6}
1052: \end{figure}
1053:
1054: Fig. \ref{fig6} shows that the initial field strength plays a significant
1055: role in the $B_{\rm x}$ case. In particular, models with greater values
1056: of $\beta_{\rm 0}$ show a more efficient mixing of the cloud material
1057: and a less rapid increase of the cloud energy. As discussed in Sect.
1058: \ref{mix}, in the $B_{\rm x}$ case, the rate of mass-loss from the cloud is
1059: mainly driven by ablation through the hydrodynamic instabilities (being
1060: the thermal conduction strongly suppressed by the magnetic field). On
1061: the other hand, in the case of external fields, the instabilities can
1062: be dumped by the magnetic field, depending on its strength (see
1063: Sect. \ref{external_field}). For instance, in the $B_{\rm x}$ case with
1064: $\beta_{\rm 0}=4$, we found that the RT instabilities are mostly suppressed
1065: by the magnetic field (see upper panels in Fig. \ref{fig2}). In the
1066: $B_{\rm x}$ case with $\beta_{\rm 0}=100$, instead the magnetic field is
1067: too weak to dump the hydrodynamic instabilities over the timescales
1068: considered; these instabilities, in turn, lead to the formation of
1069: regions dominated by the radiative cooling, triggering the development
1070: of thermal instabilities. Both the hydrodynamic and
1071: the thermal instabilities determine the cloud mass mixing (which is
1072: higher for higher values of $\beta_{\rm 0}$). In addition, the thermal
1073: instabilities reduce the increase of the cloud energy (which is less
1074: rapid for higher $\beta_{\rm 0}$) due to significant radiative losses.
1075:
1076: In the $B_{\rm y}$ case, the initial field strength has a smaller influence
1077: on the dynamic and thermal evolution of the cloud than in the $B_{\rm x}$
1078: case (see Fig. \ref{fig6}). In addition, at variance with the $B_{\rm x}$
1079: case, models with greater values of $\beta_{\rm 0}$ show a less efficient
1080: mixing of the cloud material and a more rapid increase of the cloud
1081: energy. In the $B_{\rm y}$ case, in fact, the hydrodynamic instabilities
1082: responsible of the mass mixing are mainly suppressed by the thermal
1083: conduction rather than by the magnetic field as in the $B_{\rm x}$
1084: case. As a consequence, the higher the value of $\beta_{\rm 0}$, the
1085: more effective the thermal conduction in suppressing the instabilities
1086: and in heating the plasma, the less efficient the cloud mass mixing
1087: and the more rapid the increase of the cloud energy.
1088:
1089: \subsection{Effect of spatial resolution}
1090: \label{sp_resol}
1091:
1092: The effective resolution adopted in our simulations is $\approx
1093: 132$ cells per cloud radius, a value above the resolution requirements
1094: suggested by \cite{1994ApJ...420..213K} for non-radiative clouds. However,
1095: for radiative clouds, we expect that the details of the plasma radiative
1096: cooling depend on the numerical resolution: a higher resolution
1097: may lead to different peak density and hence influence the cooling
1098: efficiency of the gas, preventing further compression of the cloud. In
1099: the non-conducting regime, \cite{2005ApJ...619..327F} found that the
1100: results generally converge for simulations with resolution larger than
1101: 100 cells per cloud radius ($\lsim 10$\% differences). In the simulations
1102: presented here, the thermal conduction partially contrasts the radiative
1103: cooling in the case of external fields ($B_{\rm x}$ or $B_{\rm y}$),
1104: alleviating the problem of numerical resolution (see also Paper I).
1105:
1106: In order to check if our adopted resolution is sufficient to capture the
1107: basic cloud evolution over the time interval considered, we compare three
1108: simulations (TR-Bz4, TR-Bz4-hr, and TR-Bz4-hr2) with different spatial
1109: resolution (132, 264, and 528 zones per cloud radius, respectively) for
1110: the $B_{\rm z}$ case with $\beta = 4$, namely one of the cases in which
1111: the growth of hydrodynamic and thermal instabilities is most prominent
1112: and the effect of thermal conduction (contrasting the development of
1113: hydrodynamic and thermal instabilities) is negligible. Since this case
1114: is one of the most demanding for resolution, it can be considered a
1115: worst case comparison of convergence.
1116:
1117: \begin{figure}[!t]
1118: \centering
1119: \epsscale{1.1}
1120: \plotone{f7.ps}
1121: \caption{Presentation as in Fig.~\ref{fig6} for runs TR-Bz4
1122: (solid lines), TR-Bz4-hr (dashed), and TR-Bz4-hr2 (dotted).}
1123: \label{fig7}
1124: \end{figure}
1125:
1126: Figure \ref{fig7} compares the evolution of the cloud mass, $M_{\rm
1127: cl}/M_{\rm cl0}$, and of the total energy of the cloud, $E_{\rm cl}/E_{\rm
1128: cl0}$, for the three simulations TR-Bz4, TR-Bz4-hr, and TR-Bz4-hr2. In
1129: general, we find that the results obtained with the three simulations
1130: agree quite well in their qualitative behavior, showing differences
1131: $\lsim 10\%$. In runs TR-Bz4-hr and TR-Bz4-hr2, the remaining cloud
1132: mass and the total energy of the cloud are, in general, systematically
1133: higher than in run TR-Bz4. The larger mass mixing in TR-Bz4 is driven
1134: by the higher diffusion of the low-resolution grid down to the very
1135: small structures which tend to smear out concentrated density peaks,
1136: promoting mass mixing. The slightly lower energy of the cloud in TR-Bz4 is
1137: a consequence of the larger mass mixing derived in this run with respect
1138: to TR-Bz4-hr and TR-Bz4-hr2. Note that, in runs showing the onset of
1139: thermal instabilities (i.e NRs and TRs models), the size of the latters
1140: reaches the resolution limit toward the end of the simulations when the
1141: relevant physical processes are already at a late stage.
1142:
1143: %________________________________________________________________
1144: \section{Summary and conclusion}
1145: \label{sec4}
1146:
1147: We investigated the importance of magnetic-field-oriented thermal
1148: conduction in the interaction between an isolated elongated dense cloud
1149: and an interstellar shock-wave of an evolved SNR shell through numerical
1150: MHD simulations. To our knowledge, these simulations represent the first
1151: attempt to model the shock-cloud interaction that simultaneously considers
1152: magnetic fields, radiative cooling, and anisotropic thermal conduction.
1153: Our findings lead to several conclusions:
1154:
1155: \begin{enumerate}
1156: \item In general, we found that the effects of thermal conduction on the
1157: evolution of the shocked cloud are reduced in the presence of an ambient
1158: magnetic field with respect to the unmagnetized cases investigated in Paper
1159: I. The efficiency of anisotropic thermal conduction strongly
1160: depends on the initial magnetic field orientation and configuration. This
1161: efficiency is the largest when the initial {\mag} is aligned with the
1162: direction of propagation of the shock front, and is the smallest when
1163: {\mag} is aligned with the cylindrical cloud, namely when the heat
1164: conduction is completely suppressed by the magnetic field.
1165:
1166: \item We found that the hydrodynamic instabilities are suppressed
1167: efficiently by the anisotropic thermal conduction when the initial
1168: magnetic field is perpendicular to the cylindrical cloud (a configuration
1169: referred to as ``external fields''). On the contrary, in the case
1170: of {\mag} parallel to the cylindrical axis of the cloud (i.e. when
1171: the field has component only along the $z$ axis - internal field),
1172: hydrodynamic instabilities develop at the cloud boundary. We found that,
1173: for the parameters of the simulations chosen, the magnetic tension is
1174: unable to suppress alone the hydrodynamic instabilities.
1175:
1176: \item As for thermal instabilities, we found that, depending on the
1177: magnetic field orientation, the heat flux contributes to the heating
1178: of some parts of the cloud, reducing the efficiency of radiative
1179: cooling there, and preventing any thermal instability.
1180:
1181: \item The mass loss of the cloud due to mixing with the surrounding
1182: medium is mainly driven by hydrodynamic instabilities; in the case
1183: of external fields (initial {\mag} perpendicular to the cylindrical
1184: cloud) the anisotropic thermal conduction reduces the mass mixing of
1185: the cloud. In any case, the mass loss rate is larger than that in the
1186: corresponding unmagnetized case ($\dot{m}_{\rm cl}\approx 1.5\times 10^{-7}
1187: L_{\rm pc}~~M_{\odot}$~yr$^{-1}$, i.e. $\sim 5$\% of the cloud mass is in mixed
1188: zones at $t = 3.5\;\tau_{\rm cc}$), but can get very high when the thermal
1189: conduction is completely suppressed ($\dot{m}_{\rm cl}\approx 4\times
1190: 10^{-6} L_{\rm pc}~~M_{\odot}$~yr$^{-1}$, i.e. $\sim 45$\% of the cloud mass is
1191: in mixed zones at $t = 3.5\;\tau_{\rm cc}$).
1192:
1193: \item The thermal conduction mostly rules the energy exchange between
1194: the cloud and surrounding medium. The exchange is favored when the
1195: magnetic field configuration is such that the conductive flow is not
1196: suppressed (i.e. external field configurations, $B_{\rm x}$ and $B_{\rm y}$
1197: cases), but it is never as high as in the absence of magnetic field. In
1198: the $B_{\rm y}$ case, the cloud core is efficiently heated and evaporates
1199: in few dynamical timescales.
1200:
1201: \item In general, the initial magnetic field strength has a small
1202: influence on the dynamic and thermal evolution of the shocked cloud for
1203: the ranges of values explored in this paper (namely $0.26~\mu\mbox{G}
1204: \leq |\mag| \leq 2.63~\mu\mbox{G}$).
1205: \end{enumerate}
1206:
1207: It is worth noting that some details of our simulations depend on the
1208: choice of the model parameters. For instance, the onset of thermal
1209: instabilities or the evaporation of the whole cloud depends on the
1210: initial shock Mach number, and on the density and dimensions of the
1211: cloud. The cases that we present here (i.e. $\mach = 50$, $\chi=10$,
1212: and different configurations of \mag) are representative of a regime
1213: in which both the thermal conduction and the radiative cooling play an
1214: important role in the evolution of the shocked cloud. Nevertheless, our
1215: analysis proves that anisotropic thermal conduction can not be neglected
1216: in investigations of the evolution of shocked interstellar clouds.
1217:
1218: In our simulations, we consider laminar thermal conduction,
1219: although regions of strong turbulence of different strength and extent
1220: develop in the system (for instance, at the shear layers along the
1221: cloud boundary or at the vortex sheets in the cloud wake). In fact,
1222: the turbulence in these regions may have a significant effect on
1223: thermal conduction, leading to significant deviations of thermal
1224: conductivity from its laminar values (e.g. \citealt{2001ApJ...562L.129N};
1225: \citealt{2006ApJ...645L..25L}); in some cases, the turbulence may
1226: enhance the heat transfer, exceeding the classical Spitzer value
1227: (\citealt{2006ApJ...645L..25L}). As a result, thermal conduction
1228: may be not only anisotropic (in the presence of the magnetic field)
1229: but also ``inhomogeneous'' due to the presence of turbulence. However,
1230: even modeling accurately the turbulent thermal conductivity, we do not
1231: expect significant changes in the results of our $B_{\rm z}$ case, being
1232: the thermal conduction strongly ineffective in the whole spatial domain;
1233: in the remaining cases ($B_{\rm x}$ and $B_{\rm y}$), our modeled thermal
1234: conductivity could be underestimated in regions of strong turbulence,
1235: affecting some details of the simulations but not the main conclusion
1236: of the paper that, in general, anisotropic thermal conduction can play
1237: an important role in the evolution of the shocked cloud.
1238:
1239: Note also that the field configurations studied in this work
1240: are highly idealized. More realistic fields are expected to have
1241: more complex topologies and, often, the field can be tangled and
1242: chaotic. In the latter case, the thermal conduction will approach
1243: isotropy, whereas the effect of MHD turbulence is expected to
1244: partially suppress the heat transfer within a factor $\sim 5$
1245: below the classical Spitzer estimate\footnote{As already discussed,
1246: the MHD turbulence can even enhance the heat transfer in some cases
1247: (see \citealt{2006ApJ...645L..25L}).} (\citealt{2001ApJ...562L.129N};
1248: \citealt{2006ApJ...645L..25L}). The shock-cloud collision in the presence
1249: of an organized ambient magnetic field, discussed here, and that in
1250: the absence of magnetic field can be considered as extreme cases: the
1251: former leading to highly anisotropic thermal conduction, the latter to
1252: the classical Spitzer thermal conduction. The case of chaotic magnetic
1253: field is expected to fall in between these two.
1254:
1255: Our simulations were carried out in 2.5D Cartesian geometry, implying
1256: that the modeled clouds are elongated along the $z$ axis. This choice
1257: is expected to affect some details of the simulations but not our main
1258: conclusions. Adopting a 3D Cartesian geometry and modeling a spherical
1259: cloud, the highly symmetric shock transmitted into the cloud converging
1260: on the symmetry axis would lead to compression stronger than those found
1261: in our 2.5D simulations, enhancing the radiative cooling. Also, 3D
1262: simulations would provide an additional degree of freedom for hydrodynamic
1263: instabilities, increasing the mass loss rate of the cloud in the cases in
1264: which the mass mixing of cloud material is driven by instabilities. Note
1265: that, for a spherical cloud, our $B_{\rm x}$ and $B_{\rm z}$ cases no
1266: longer differ.
1267:
1268: Finally we assume, in our simulations, that the cloud and the
1269: ambient material have the same composition, implying that microscopic
1270: mass mixing due to shear instabilities would be irrelevant. In a more
1271: realistic condition, a cold dense cloud may have a different composition
1272: from the hot ambient flow and the degree of microscopic mixing may
1273: translate into different spectral signatures of the system. In this case,
1274: species diffusion could also be important, along with thermal conduction,
1275: to determine the degree of microscopic mixing of the materials and,
1276: consequently, one would have to ask about the typical values of the
1277: Lewis number (i.e. the ratio of thermal diffusivity to mass diffusivity)
1278: in the system.
1279:
1280: It is worth emphasizing that the quantitative results of our
1281: simulations depend on the physical parameters of the model (shock Mach
1282: number, density contrast and dimension of the cloud, etc.) as well as on
1283: the basic assumptions of the model (geometry of the cloud, geometry of
1284: the ambient magnetic field, laminar thermal conduction, composition
1285: of the cloud and of the ambient medium, etc.). Nevertheless, our results
1286: undoubtedly show that the magnetic-field-oriented thermal conduction can
1287: play an important role in the evolution of the shock-cloud interaction
1288: (which depends on the magnetic field orientation and configuration) and,
1289: in particular, in the mass and energy exchange between the cloud and the
1290: hot surrounding medium. We conclude, therefore, that a self-consistent
1291: and quantitative description of the interaction between magnetized
1292: shock-waves and interstellar gas clouds should include the effects of
1293: thermal conduction.
1294:
1295: The results presented here are interesting for the study of
1296: middle-aged SNR shells expanding into a magnetized ISM and
1297: whose morphology is affected by ISM inhomogeneities (for
1298: instance, G272.2-3.2, e.g. \citealt{1996rftu.proc..247E};
1299: Cygnus Loop, e.g. \citealt{2002AJ....124.2118P}; Vela SNR, e.g.
1300: \citealt{2005A&A...442..513M}). It will be further interesting to extend
1301: the present study, by modeling the shock-cloud interaction in 3D with
1302: radiative cooling, anisotropic thermal conduction, and magnetic field
1303: included and, even, considering detailed comparisons of model results
1304: with observations.
1305:
1306:
1307: \acknowledgements
1308: The authors thank Timur Linde for his help with the MHD portion of FLASH
1309: and the referee for constructive and helpful criticism. The software
1310: used in this work was in part developed by the DOE-supported ASC /
1311: Alliance Center for Astrophysical Thermonuclear Flashes at the University
1312: of Chicago, using modules for thermal conduction and optically thin
1313: radiation built at the Osservatorio Astronomico di Palermo. Most of the
1314: simulations have been executed at CINECA (Bologna, Italy) in the framework
1315: of the INAF-CINECA agreement on ``High Performance Computing resources for
1316: Astronomy and Astrophysics''. This work makes use of results produced by
1317: the PI2S2 Project managed by the Consorzio COMETA, a project co-funded
1318: by the Italian Ministry of University and Research (MIUR) within the
1319: Piano Operativo Nazionale ``Ricerca Scientifica, Sviluppo Tecnologico,
1320: Alta Formazione'' (PON 2000-2006); more information is available at
1321: http://www.pi2s2.it and http://www.consorzio-cometa.it. This work was
1322: supported in part by Istituto Nazionale di Astrofisica.
1323:
1324:
1325: \bibliographystyle{apj}
1326: \bibliography{references}
1327:
1328: \clearpage
1329:
1330: \end{document}
1331: