1: \documentclass[12pt,preprint]{aastex}

2: 

3: \usepackage{emulateapj5}

4: 

5: \begin{document}

6: 

7: 

8: \title{The effects of thermal conduction on the hot accretion flows}

9: 

10: 

11: \author{Mohsen Shadmehri and Fazeleh Khajenabi

12: \affil{Department of Physics, School of Science, Ferdowsi

13: University, Mashhad,

14: Iran}}\email{mshadmehri@science1.um.ac.ir,fkhajenabi@science1.um.ac.ir}

15: 

16: 

17: 

18: 

19: \begin{abstract}

20: %

21: Thermal conduction has been suggested as a possible mechanism by

22: which sufficient extra heating is provided in radiation-dominated

23: accretion flows. We consider the extreme case in which the

24: generated energy due to the viscosity  and the energy transported

25: by a saturated form of thermal conduction are balanced by the

26: advection cooling. For the steady-state structure of such

27: accretion flows a set of self-similar solution are presented.

28: Based on these solutions while the radial and the rotational

29: velocities are both sub-Keplerian, increasing the level of

30: thermal conduction has the effects of decreasing the rotational

31: velocity, but increasing the radial velocity. Conduction provides

32: extra heating and the temperature of the gas increases with

33: thermal conduction.

34: 

35: 

36: %

37: \end{abstract}

38: 

39: \keywords{accretion, accretion disks - hydrodynamics}

40: 

41: 

42: \section{Introduction}

43: %

44: 

45: The thin accretion disk model describes flows in which the

46: viscous heating of the gas radiates out of the system immediately

47: after generation (Shakura $\&$ Sunyaev 1973). However, another

48: kind of accretion has been studied during recent years where

49: radiative energy losses are small so that most of the energy is

50: advected with the gas. These Adection-Dominated Accretion Flows

51: (ADAF) occur in two regimes depending on the mass accretion rate

52: and the optical  depth. At very high mass accretion rate, the

53: optical depth becomes very high and the radiation can be trapped

54: in the gas. This type of accretion which is known under the name

55: 'slim accretion disk' has been studied in detail by Abramowicz et

56: al. (1988). But when the accretion rate is very small and the

57: optical depth is very low, we may have another type of accretion

58: (Narayan \& Yi 1994; Abramowitz et al. 1995; Chen 1995). However,

59: numerical simulations of radiatively inefficient accretion flows

60: revealed that low viscosity flows are convectively unstable and

61: convection strongly influences the global properties of the flow

62: (e.g., Igumenshchev, Abramowicz, $\&$ Narayan 2000). Thus, another

63: type of accretion flows has been proposed, in which convection

64: plays as a dominant mechanism of transporting angular momentum

65: and the local released viscous energy (e.g., Narayan,

66: Igumenshchev, $\&$ Abramowicz 2000).

67: 

68: This diversity of models  tells us that modeling the hot

69: accretion flows is a challenging and controversial problem. We

70: think, one of the largely neglected physical ingredient in this

71: field, is {\it thermal conduction}. But a few authors tried to

72: study the role of "turbulent" heat transport in ADAF-like flows

73: (Honma 1996; Manmoto et al. 2000) Since thermal conduction acts to

74: oppose the formation of the temperature gradient that causes it,

75: one might expect that the temperature and density profiles for

76: accretion flows in which thermal conduction plays a significant

77: role to appear different compared to those flows in which thermal

78: conduction is less effective. Just recently, Menou (2005) studied

79: the effect of saturated thermal conduction on optically thin

80: ADAFs using an extension of self-similar solutions of Narayan

81: $\&$ Yi (2004). His solutions suggest that thermal conduction may

82: be an important physical factor to understand hot accretion onto

83: dim accreting black holes.

84: 

85: The aim of the work presented here is to investigate whether

86: thermal conduction can affect  the dynamics of an optically {\it

87: thick} radiation-dominated accretion flow. In \S 2, we present a

88: height-integrated set of equations of the model. Self-similar

89: solutions are investigated in \S 3. The paper concludes with a

90: summary of the results in \S4.

91: 

92: 

93: 

94: 

95: 

96: 

97: 

98: 

99: %

100: %

101: \section{FORMULATION}

102: %

103: In order to implement thermal conductivity correctly it is

104: essential to know whether the mean free path is less than (or

105: comparable to) the scale length of the temperature gradient. For

106: electron mean free path which are greater than the scale length

107: of the temperature gradient the thermal conductivity is said to

108: 'saturate' and the heat flux approaches a limiting value (Cowie

109: \& McKee 1977). But when the mean free paths are much less than

110: the temperature gradient the heat flux depends on the coefficient

111: of thermal conductivity and the temperature gradient. Generally,

112: thermal conduction transfers heat so as to oppose the temperature

113: gradient which causes the transfer. Menou (2005) discussed hot

114: accretion likely proceed under weakly-collisional conditions in

115: these systems. Thus, a saturated form of "microscopic" thermal

116: conduction is physically well-motivated, as we apply in this

117: study. However, one of the primary problems for studying the

118: effects of thermal conduction in plasmas is the unknown value of

119: the thermal conductivity.

120: 

121: 

122: 

123: We consider an accretion disk that is axisymmetric and

124: geometrically thin, i.e. $h/r<1$. Here $r$ and $h$ are,

125: respectively, the disk radius and the half-thickness. Our model

126: generalizes the usual slim disks around a neutron star or a black

127: hole (e.g., Muchotrzeb \& Paczy\'{n}ski 1982; Matsumoto et al.

128: 1984; Abramowicz et al. 1988) by including the effect of thermal

129: conduction. The disk is supposed to be turbulent and possesses an

130: effective turbulent viscosity. We assume the generated energy due

131: to viscous dissipation and the heat conducted into the volume is

132: concerned are balanced by the advection cooling. Consider

133: stationary height-integrated equations described an accretion flow

134: onto a central object of mass $M_{\ast}$. In absence of mass

135: outflows, the continuity equation reads

136: %

137: \begin{equation}

138: \dot{M}=-2\pi r v_{\rm r} \Sigma,

139: \end{equation}

140: %

141: where $\dot{M}$ is the accretion rate, $v_{\rm r}$ is the

142: accretion velocity (and $v_{\rm r}<0$) and $\Sigma=2h\rho$ is the

143: surface density at a cylindrical radius. Also, $\rho$ is the

144: midplane density of the disk.

145: 

146: 

147: The equation of motion in the radial direction is

148: %

149: \begin{equation}

150: \Sigma v_{\rm r}\frac{d v_{\rm r}}{d

151: r}-\Sigma\frac{v_{\varphi}^2}{r}=-\frac{d P}{d

152: r}-\Sigma\frac{GM_{\ast}}{r^2},

153: \end{equation}

154: %

155: where $P=2h p$ is the integrated disk pressure written, in a form

156: compatible with radiation-dominated flow, as $P\simeq 2h a T^4/3$,

157: where $a$ is black body constant and $T$ denotes the midplane

158: temperature of the disk. Similarly, integration over $z$ of the

159: azimuthal equation of motion gives

160: %

161: \begin{equation}

162: r\Sigma v_{\rm r}\frac{d}{d r}(rv_{\varphi})=\frac{d}{d

163: r}[r^{3}\nu\Sigma\frac{d}{d r}(\frac{v_{\varphi}}{r})],

164: \end{equation}

165: %

166: where $\nu$ is a kinematic viscosity coefficient and we assume

167: %

168: \begin{equation}

169: \nu=\alpha c_{\rm s} h,

170: \end{equation}

171: %

172: where  $\alpha$ is a constant less than unity (Shakura \& Sunyaev

173: 1973). Moreover, $h=(c_{\rm s}/v_{\rm K})r$, where $c_{\rm s}$ and

174: $v_{\rm K}$ are sound speed and the Keplerian velocity,

175: respectively: $c_{\rm s}=\sqrt{P/\Sigma}$ and $v_{\rm

176: K}=\sqrt{GM_{\ast}/r}$.

177: %

178: 

179: Now, we can write the energy equation considering the cooling and

180: the heating processes of the accretion flow. The locally released

181: energy due to viscous dissipation, $Q_{\rm vis}$, and the energy

182: transported by thermal conduction, $Q_{\rm cond}$, are balanced

183: by the advection cooling, $Q_{\rm adv}$. Thus,

184: %

185: \begin{equation}

186: Q_{\rm adv}=Q_{\rm cond}+Q_{\rm vis}.

187: \end{equation}

188: %

189: The advection cooling reads (Abramowicz et al. 1998)

190: \begin{equation}

191: Q_{\rm adv}=\frac{\dot{M}}{2\pi

192: r^{2}}\frac{p}{\rho}[\frac{4-3\beta}{\Gamma_{3}-1}\frac{d\ln T

193: }{d\ln r}+(4-3\beta)\frac{d\ln\rho}{d\ln r}],

194: \end{equation}

195: %

196: where $\beta$ is the ratio of the gas to the total pressure and

197: \begin{equation}

198: \Gamma_{3}=1+\frac{(4-3\beta)(\gamma -1)}{\beta+12(\gamma

199: -1)(\beta -1)}.

200: \end{equation}

201: %

202: Since we are considering the extreme case of radiation-dominated

203: accretion flow, we have $\beta=0$ and so $\Gamma_{3}=2/3$. Also,

204: we have

205: 

206: \begin{equation}

207: Q_{\rm vis}=\nu\Sigma r^{2}

208: [\frac{d}{dr}(\frac{v_{\varphi}}{r})]^{2},

209: \end{equation}

210: %

211: and (Cowie \& McKee 1977)

212: %

213: \begin{equation}

214: Q_{\rm cond}=-\frac{2h}{r^{2}}\frac{d}{dr}(r^{2}F_{\rm s})

215: \end{equation}

216: %

217: where as we have already mentioned $F_{\rm s}=5\Phi_{\rm s}\rho

218: c_{\rm s}^{3}$ is the saturated conduction flux on the direction

219: of the temperature gradient ($\Phi_{\rm s}<1$).

220: %

221: \section{SELF-SIMILAR SOLUTIONS}

222: %

223: %

224: The self-similar solution is not able to describe the global

225: behaviour of the accretion flow, because no boundary condition

226: has been taken into account. However, as long as we are not

227: interested in the the behaviour of the flow at the boundaries,

228: such solution describes correctly the true solution

229: asymptotically at large radii. We assume that each physical

230: quantity can be expressed as a power law of the radial distance,

231:  i.e.  $r^{\nu}$, where the power index $\nu$ is determined for

232: each  physical quantity self-consistently. The solutions are

233: \begin{equation}

234: \Sigma(r)=a \Sigma_{0}(\frac{r}{r_0})^{-1/2},

235: \end{equation}

236: %

237: \begin{equation}

238: v_{\varphi}(r)=b

239: \sqrt{\frac{GM_{\ast}}{r_0}}(\frac{r}{r_0})^{-1/2},

240: \end{equation}

241: %

242: \begin{equation}

243: v_{\rm r}(r)=- c

244: \sqrt{\frac{GM_{\ast}}{r_0}}(\frac{r}{r_0})^{-1/2},

245: \end{equation}

246: %

247: \begin{equation}

248: P(r)= d \frac{\Sigma_{0}GM_{\ast}}{r_0}(\frac{r}{r_0})^{-3/2},

249: \end{equation}

250: %

251: 

252: %

253: \begin{equation}

254: h(R)=f r_{0} (\frac{r}{r_0}),

255: \end{equation}

256: %

257: %

258: %

259: %

260: where $\Sigma_{\rm 0}$ and $r_{\rm 0}$ provide convenient units

261: with which the equations can be written in the non-dimensional

262: form. Thus, we obtain the following system of dimensionless

263: equations, to be solved for $a$, $b$, $c$, $d$ and $f$:

264: %

265: \begin{equation}

266: ac=\dot{m},

267: \end{equation}

268: %

269: \begin{equation}

270: -\frac{1}{2}ac^{2}-ab^{2}=\frac{3}{2}d-a,

271: \end{equation}

272: %

273: \begin{equation}

274: -\frac{1}{2}abc=-\frac{3\alpha}{4}\sqrt{\frac{d}{a}}fab,

275: \end{equation}

276: %

277: \begin{equation}

278: af^{2}-2d=0,

279: \end{equation}

280: %

281: %

282: \begin{equation}

283: \frac{3}{2}\dot{m}\frac{d}{a}=\frac{9}{4}\alpha\sqrt{\frac{d}{a}}fb^{2}a+5\Phi_{\rm

284: s}d\sqrt{\frac{d}{a}},

285: \end{equation}

286: %

287: where $\dot{m}=\dot{M}/(2\pi \Sigma_{0}\sqrt{GM_{\ast}r_{0}})$ is

288: nondimensional mass accretion rate. After some algebraic

289: manipulations we find

290: %

291: \begin{equation}

292: a=(\frac{2\sqrt{2}}{3\alpha}\dot{m})\frac{1}{f^{2}},\label{eq:a}

293: \end{equation}

294: %

295: \begin{equation}

296: b=\sqrt{1-\frac{3}{4}f^{2}-\frac{9}{16}\alpha^{2}f^{4}},\label{eq:b}

297: \end{equation}

298: %

299: \begin{equation}

300: c=(\frac{3\alpha}{2\sqrt{2}})f^{2},\label{eq:c}

301: \end{equation}

302: %

303: \begin{equation}

304: d=\frac{\sqrt{2}}{3\alpha}\dot{m},\label{eq:d}

305: \end{equation}

306: %

307: and $f$ is obtained from a forth order algebraic equation:

308: %

309: \begin{equation}

310: 9 \alpha^{2} f^{4}+20 f^{2} - \frac{160\Phi_{\rm s}}{9\alpha} f -

311: 16=0.\label{eq:main}

312: \end{equation}

313: We can solve this algebraic equation numerically and clearly only

314: the real root which corresponds to positive $b^{2}$ is a physical

315: solution.

316: 

317: 

318: If we neglect the thermal conduction (i.e., $\Phi_{\rm s }=0$),

319: the above equation reduces to a second order equation which can

320: be solved analytically (Wang $\&$ Zhou 1999; Shadmehri \&

321: Khajenabi 2005). In this case, the typical behaviour of the

322: solutions can be summarized as follows (Shadmehri \& Khajenabi

323: 2005): (1) The surface density increases with the accretion rate,

324: and decreases with viscosity coefficient $\alpha$; (2) But the

325: radial velocity is directly proportional to $\alpha$; (3) The gas

326: rotates with sub-Keplerian angular velocity, more or less

327: independent of the coefficient $\alpha$; and (4) the opening

328: angle of the disk is fixed, independent of $\alpha$ and

329: $\dot{m}$. Note that Shadmehri $\&$ Khajenabi (2005)  applied a

330: diffusive prescription for the viscous stress tensor in their

331: analysis of the self-similar solution of optically thick

332: advection-dominated flows, but Wang $\&$ Zhou (1999) applied the

333: $\alpha p$ prescription. However, the typical behaviours of the

334: solutions are the same, irrespective of some differences in the

335: coefficients.

336: 

337: The equations (\ref{eq:a}), (\ref{eq:b}), (\ref{eq:c}),

338: (\ref{eq:d}) together with equation (\ref{eq:main}) describes

339: self-similar behaviour of the optically thick advection dominated

340: accretion disk with saturated thermal conduction. The algebraic

341: equation (\ref{eq:main}) shows that the variable $f$ which

342: determines the opening angle of the disk depends only on the

343: $\alpha$ and $\Phi_{\rm s}$. This behaviour is similar to the

344: case without thermal conduction. Thus, one can deduce that the

345: surface density and the pressure are directly proportional to the

346: mass accretion rate according to the equations (\ref{eq:a}) and

347: (\ref{eq:d}). But the rotational and the radial velocities are

348: both independent of the mass accretion rate (see equations

349: (\ref{eq:b}) and (\ref{eq:c})). Also, we can say that the

350: rotational velocity is sub-Keplerian according to the equation

351: (\ref{eq:b}).

352: 

353: 

354: The dependence of the properties of the rotational and the radial

355: velocities on $\alpha$ and $\Phi_{\rm s}$ is illustrated in

356: Figure \ref{fig:figure1}. We consider three values of the

357: viscosity parameter, $\alpha=0.2, 0.02$ and $0.002$ for which the

358: velocities are plotted as a function of $\Phi_{\rm s}$. The solid

359: and dashed lines represent the ratios $v_{\varphi}/v_{\rm K}$ and

360: $v_{\rm r}/v_{\rm K}$, respectively. While the solution without

361: thermal conduction are recovered at small $\Phi_{\rm s}$ values,

362: we can see significant deviation as $\Phi_{\rm s}$ increases.

363: Clearly, the rotational velocity decreases with the magnitude of

364: conduction parameter. But for the radial velocity the behaviour

365: is different, i.e. the velocity increases as $\Phi_{\rm s}$

366: increases. For a fixed viscosity parameter $\alpha$, the solution

367: reaches the non-rotating limit at a specific value of $\Phi_{\rm s

368: }$ which we denotes by $\Phi_{\rm s}^{\rm c}$. If we extend our

369: solution to $\Phi_{\rm s}>\Phi_{\rm s}^{\rm c}$, the  equation

370: (\ref{eq:b}) gives negative $b^{2}$ which is clearly

371: unacceptable. The critical value $\Phi_{\rm s}^{\rm c}$ can be

372: calculated easily:

373: %

374: \begin{equation}

375: \Phi_{\rm s}^{\rm

376: c}=\frac{9}{10\sqrt{6}}\sqrt{\sqrt{4\alpha^{2}+1}-1}.

377: \end{equation}

378: %

379: The above equation shows the critical magnitude of conduction

380: parameter for which the solution tends to the non-rotating limit

381: depends only on the viscosity coefficient so that as $\alpha$

382: increases, the breakdown of the solutions occurs at larger values

383: of $\Phi_{\rm s}$. But as Menou (2005) mentioned in his analysis

384: of optically thin ADAFs with saturated conduction, we think this

385: behaviour is a simple pathological feature of 1D

386: height-integrated equations we solved.

387: 

388: Solutions of equation (\ref{eq:main}) show that for a fixed

389: $\alpha$, the variable $f$ increases with conduction parameter

390: $\Phi_{\rm s}$. Since the gas temperature varies directly in

391: proportion to $f$, we can say as the level of thermal conduction

392: is increased the temperature of flow in comparison to solutions

393: without thermal conduction increases. In fact, thermal conduction

394: behaves as an extra heating source. Also, equation (\ref{eq:a})

395: shows the surface density decreases as $\Phi_{\rm s}$ increases.

396: We can compare the relative values of advection cooling, viscous

397: heating and the transported energy by conduction using our

398: self-similar solutions. One can simply show that $Q_{\rm

399: vis}/Q_{\rm adv}=2(b/f)^{2}$ and $Q_{\rm cond}/Q_{\rm adv}=20

400: \Phi_{\rm s}/(9\alpha f)$. While both these ratios don't depend on

401: the mass accretion rate, the viscosity parameter $\alpha$ and the

402: thermal conduction parameter $\Phi_{\rm s}$ determine the

403: relative importance of the various physical processes in the

404: system. Viscous energy dissipation decreases as $\Phi_{\rm s}$

405: increases, but the contribution of the energy transported by the

406: conduction becomes more significant.

407: 

408: %

409: \section{Conclusions}

410: %

411: 

412: In this paper we have studied  an optically thick ADAFs with

413: thermal conduction. Considering the weakly-collisional nature of

414: hot accretion flows, a saturated form of thermal conduction has

415: been used as a possible mechanism of transporting of  energy.

416: Although our self-similar solutions have their own limitations,

417: they can illuminate some possible effects of the saturated thermal

418: conduction on radiation-dominated accretion flows. We can

419: summarize the main features of our solutions: (1) The surface

420: density is in proportion to the mass accretion rate and it

421: increases as the level of the thermal conduction is increased. (2)

422: The temperature of the gas increases with thermal conduction

423: parameter and the conduction term in the energy equation acts as

424: an extra source of heating. (3) The angular velocity is

425: sub-Keplerian, but it decreases with increasing thermal

426: conduction parameter so that it tends to a non-rotating limit.

427: (4) However, the radial velocity increases with thermal

428: conduction parameter. Considering these significant effects of

429: the thermal conduction,  global transonic solutions representing

430: optically thick (or thin) ADAFs with thermal conduction would be

431: an interesting topic for future studies and our results are a

432: step toward this goal.

433: %\acknowledgements

434: 

435: 

436: 

437: \begin{references}

438: 

439: \reference{} Abramowicz, M. A., Chen, X., Kato, S., Lasota, J. P.,

440: Regev, O. 1995, ApJ, 438, L37

441: 

442: \reference{} Abramowicz, M. A., Czerny, B., Lasota, J. -P.,

443: Szuszkiewicz, E. 1988, ApJ, 332, 646

444: 

445: \reference{} Chen, X. 1995, MNRAS, 275, 641

446: 

447: \reference{} Cowie, L. L. $\&$ McKee, C. F. 1977, ApJ, 211, 135

448: 

449: \reference{} Honma, F. 1996, PASJ, 48, 77

450: 

451: \reference{} Igumenshchev, I. V., Abramowicz, M. A., $\&$

452: Narayan, R. 2000, ApJ, 537, L27

453: 

454: \reference{} Manmoto, T. et al. 2000, ApJ, 529, 127

455: 

456: \reference{} Matsumoto, R., Kato, S., Fukue, J., Okazaki, A. T.

457: 1984, PASJ, 36, 71

458: 

459: \reference{} Menou, K. 2005, ApJL, submitted (astro-ph/0507189)

460: 

461: \reference{} Muchotrzeb, B. $\&$ Paczy\'{n}ski, B. 1982, Acta

462: Astron., 32, 1

463: 

464: \reference{} Narayan, R., Igumenshchev, I. V., $\&$ Abramowicz, M.

465: A. 2000, ApJ, 539, 798

466: 

467: \reference{} Narayan, R. $\&$ Yi, I. 1994, ApJ, 428, L13

468: 

469: \reference{} Shadmehri, M. $\&$ Khajenabi, F. 2005, MNRAS, 361,

470: 719

471: 

472: \reference{} Shakura, N. I. $\&$ Sunyaev, R.A. 1973, A\&A,  24,

473: 337

474: 

475: \reference{} Wang, J.-M. $\&$ Zhou, Y.-Y. 1999, ApJ, 516, 420

476: 

477: 

478: 

479: 

480: 

481: 

482: \end{references}

483: 

484: 

485: %\clearpage

486: \begin{figure}

487: \plotone{f1.eps} \caption{Profile of the  rotation velocity

488: ($v_{\varphi}/v_{\rm K}$, solid lines) and the radial velocity

489: ($v_{\rm r}/v_{\rm K}$, dashed lines) as a function of the

490: saturation constant, $\Phi_{\rm s}$. In order to make easier

491: comparison the rotational and the radial velocities are scaled by

492: the Keplerian velocity, $v_{\rm K}$. Each curve is labeled by

493: corresponding viscosity coefficient, $\alpha$.

494: }\label{fig:figure1}

495: \end{figure}

496: 

497: 

498: %

499: %

500: %

501: \end{document}

502: