1: \documentclass[12pt,preprint]{aastex}
2:
3: %\documentclass{aastex}
4: %\usepackage{emulateapj5}
5: %\usepackage{apjfonts}
6: %\usepackage{epsf}
7: %\slugcomment{submitted to ApJ}
8: %\slugcomment{D~R~A~F~T \hskip .5cm \today \hskip .5cm D~R~A~F~T}
9:
10: \shorttitle{GRMHD Simulations of Jet Formation} \shortauthors{Mizuno
11: et al.}
12:
13: \usepackage{graphicx}
14:
15: \begin{document}
16:
17: \title{General Relativistic Magnetohydrodynamic Simulations of \\
18: Jet Formation with a Thin Keplerian Disk}
19:
20: \author{
21: Yosuke Mizuno\altaffilmark{1,6}, Ken-Ichi
22: Nishikawa\altaffilmark{1,2}, Shinji Koide\altaffilmark{3}, Philip
23: Hardee\altaffilmark{4} \\ and Gerald J. Fishman\altaffilmark{5} }
24:
25: \altaffiltext{1}{National Space Science and Technology Center, 320
26: Sparkman Drive, VP 62, Huntsville, AL 35805, USA;
27: Yosuke.Mizuno@msfc.nasa.gov}
28: \altaffiltext{2}{Center for Space
29: Plasma and Aeronomic Research, University of Alabama in Huntsville}
30: \altaffiltext{3}{Department of Physics, Kumamoto University,
31: Kurokami, Kumamoto, 860-8555, Japan}
32: \altaffiltext{4}{Department of
33: Physics and Astronomy, The University of Alabama, Tuscaloosa, AL
34: 35487, USA}
35: \altaffiltext{5}{NASA-Marshall Space Flight Center,
36: National Space Science and Technology Center, 320 Sparkman Drive, VP
37: 62, Huntsville, AL 35805, USA}
38: \altaffiltext{6}{NASA Postdoctoral
39: Program Fellow/ NASA Marshall Space Flight Center}
40:
41: \begin{abstract}
42:
43: We have performed several simulations of black hole systems
44: (non-rotating, black hole spin parameter $a=0.0$ and rapidly
45: rotating, $a=0.95$) with a geometrically thin Keplerian disk using
46: the newly developed RAISHIN code. The simulation results show the
47: formation of jets driven by the Lorentz force and the gas pressure
48: gradient. The jets have mildly relativistic speed ($\gtrsim 0.4~c$).
49: The matter is continuously supplied from the accretion disk and the
50: jet propagates outward until each applicable terminal simulation
51: time (non-rotating: $t/\tau_{\rm S} = 275$ and rotating:
52: $t/\tau_{\rm S} = 200$, $\tau_{\rm S} \equiv r_{\rm S}/c$). It
53: appears that a rotating black hole creates an additional, faster,
54: and more collimated matter-dominated inner outflow ($\gtrsim
55: 0.5~c$) formed and accelerated by the twisted magnetic field
56: resulting from frame-dragging in the black hole ergosphere.
57: This is the first known simulation confirming the formation of an inner
58: magnetically-driven, matter-dominated jet by the frame-dragging
59: effect from a black hole co-rotating with a thin
60: Keplerian disk threaded by a vertical magnetic field. This
61: result indicates that jet kinematic structure depends on black hole
62: rotation and on the initial magnetic field configuration and
63: strength.
64:
65:
66: \end{abstract}
67: \keywords{accretion, accretion disks - black hole physics -
68: magnetohydrodynamics: (MHD) - method: numerical -relativity}
69:
70: \section{Introduction}
71:
72: Both magnetic and gravitational fields play an important role in the
73: dynamics of matter in many astrophysical systems. In a highly
74: conducting plasma, the magnetic field can be amplified by gas
75: contraction or shear motion. Even when the magnetic field is weak
76: initially, the magnetic field can grow on short time scales and
77: influence the gas dynamics of the system. The magnetic influence on
78: gas dynamics is particulary important for a compact object such as a
79: black hole or a neutron star. Relativistic jets have been observed
80: or postulated in various astrophysical objects, including active
81: galactic nuclei (AGNs) (e.g., Urry \& Pavovani 1995; Ferrari 1998),
82: microquasars in our galaxy (e.g., Mirabel \& Rodiriguez 1999), and
83: gamma-ray bursts (GRBs) (e.g., Zhang \& M\'{e}sz\'{a}ros 2004; Piran
84: 2005; M\'{e}sz\'{a}ros 2006). The most promising mechanisms for
85: producing the relativistic jets involve magnetohydrodynamic
86: centrifugal acceleration and/or magnetic pressure driven
87: acceleration from an accretion disk around the compact objects
88: (e.g., Blandford \& Payne 1982; Fukue 1990), or involve the extraction of
89: rotating energy from a rotating black hole (Penrose 1969; Blandford
90: \& Znajek 1977).
91:
92: General relativistic magnetohydrodynamics (GRMHD) codes with
93: fixed spacetimes have been developed to investigate relativistic magnetorotators
94: (RMRs) (e.g., Koide et al. 1998; De Villiers \& Hawley 2003; Gammie et al.\ 2003;
95: Komissarov 2004; Ant\'{o}n et al.\ 2005; Anninos et al.\ 2006).
96: These codes have been used to study the Blandford-Znajek
97: effect near a rotating black hole (Koide 2003; Komissarov 2005;
98: McKinney 2005), and the formation of GRB jets in collapsars (Mizuno
99: et al.\ 2004a, 2004b; De Villiers et al.\ 2005b).
100:
101: Recently, in order to investigate the properties of accretion flows
102: onto a black hole associated with the magneto-rotational instability
103: (MRI) (Balbus \& Hawley 1991), many simulations have been performed
104: using a thick torus-like disk (the disk thickness $H/r > 0.1$, where
105: $H$ is height of the disk and $r$ is radius from a black hole) with
106: weak poloidal magnetic fields in a torus (the plasma beta, $\beta =
107: p_{\rm gas}/p_{\rm mag} > 100$, where $p_{\rm gas}$ is gas pressure
108: and $p_{\rm mag}$ is magnetic pressure) (De Villiers et al.\ 2003,
109: 2005a; Krolik et al.\ 2005; Hawley \& Krolik 2006; Beckwith et al.\
110: 2006; McKinney \& Gammie 2004; McKinney 2006). The initial
111: ``poloidal-loop'' magnetic fields in the torus contribute to the
112: generation of MRI, diffusion of matter and magnetic field, and jet
113: generation. However, in their simulations the structure of magnetic
114: fields that piled up and twisted near the black hole is different from
115: the magnetic fields which are twisted by a thin Keplerian disk
116: and/or the frame-dragging effect of the rotating black hole ($H/r
117: \sim 0.06$) with a stronger initial vertical magnetic field ($\beta
118: < 40$) (Koide, Shibata, \& Kudoh 1998, 1999; Koide et al. 2000;
119: Nishikawa et al. 2005). In the thin disk simulations MRI does not
120: grow since the wavelength at the maximum growth rate is larger
121: than the height of the thin disk. Koide et al.\ (2000) have found
122: that jets are formed from thin Keplerian accretion disks for both
123: counter- and co-rotating black holes. In the co-rotating disk case,
124: the jet had a two-layered structure: an inner gas pressure-driven jet,
125: and an outer magnetically driven jet (Koide et al. 2000).
126:
127: In this Letter we report on new simulations showing jet formation
128: from the black hole magnetosphere co-rotating with a thin Keplerian
129: disk threaded by a vertical magnetic field using the recently
130: developed three-dimensional, GRMHD code RAISHIN (Mizuno et al.
131: 2006). Our new result shows
132: that an inner matter-dominated jet is generated by the
133: magnetic field twisted in the ergosphere of the rotating black hole
134: in addition to the outer jet formed by the co-rotating accretion
135: disk seen by Koide et al.\ (2000). While our inner jet is similar
136: the inner jet created by a counter-rotating disk found by Koide et al.
137: (2000), in this case the inner jet is located closer to the rotation
138: axis.
139:
140:
141: \section{Numerical Method}
142:
143: In order to study the formation of relativistic jets from a geometrically thin
144: Keplerian disk, we use a 2.5-dimensional GRMHD code with Boyer-Lindquist
145: coordinates $(r, \theta, \phi)$. The method is based on a 3+1 formalism of the
146: general relativistic conservation laws of particle number and energy momentum,
147: Maxwell equations, and Ohm's law with no electrical resistance
148: (ideal MHD condition) in a curved spacetime (Mizuno et al. 2006).
149:
150: In the RAISHIN code, a conservative, high-resolution shock-capturing scheme is
151: employed. The numerical fluxes are calculated using the Harten, Lax, \& van Lee
152: (HLL) approximate Riemann solver scheme. The flux-interpolated, constrained
153: transport scheme is used to maintain a divergence-free magnetic field. The
154: RAISHIN code has proven to be accurate to second order and has passed numerical
155: tests including highly relativistic cases, and highly magnetized cases in both
156: special and general relativity. Code details and the results of code tests can
157: be found in Mizuno et al.\ (2006). In the simulations presented here we use
158: minmod slope limiter reconstruction, HLL approximate Riemann solver, flux-CT
159: scheme and Noble's 2D method.
160:
161: In our present simulations: a geometrically thin Keplerian disk
162: rotates around a black hole (non-rotating, $a=0.0$ or rapidly
163: rotating, $a=0.95$, here $a$ is black hole spin parameter), where
164: the disk density is 100 times higher than the coronal density. The
165: thickness of the disk is $H/r \sim 0.06$ at $r=10 r_{\rm S}$. In the
166: rotating black hole case the disk is co-rotating with the black
167: hole. The background corona is free-falling into the black hole
168: (Bondi flow). The initial magnetic field is assumed to be uniform
169: and parallel to the rotational axis i.e., the Wald solution (Wald
170: 1974). Our scale-free simulations are normalized by the speed of
171: light, $c$, and the Schwarzschild radius, $r_{\rm S}$, with
172: timescale, $\tau_{\rm S} \equiv r_{\rm S}/c$. Values of the magnetic
173: field strength and gas pressure depend on the normalized density,
174: $\rho_{0}$. In these simulations the magnetic field strength,
175: $B_{0}$, is set to $0.05 \sqrt{\rho_{0} c^{2}})$. These initial
176: conditions are similar to Koide et al.\ (1999, 2000) but with a
177: weaker initial magnetic field. Koide et al.\ (2000) used $B_{0}=0.3
178: \sqrt{\rho_{0} c^{2}}$. The simulations are performed in the region
179: $1.1 r_{\rm S} \le r \le 20.0 r_{\rm S}$ (non-rotating black hole
180: case) and $0.75 r_{\rm S}\le r \le 20.0 r_{\rm S}$ (rapidly rotating
181: black hole case) and $0.03 \le \theta \le \pi/2$. We use $128
182: \times 128$ computational zones with logarithmic zone spacing in the
183: radial direction (Koide et al.\ 1999). We assume axisymmetry with
184: respect to the $z$-axis and mirror symmetry with respect to the
185: equatorial plane. We employ a free boundary condition at the inner
186: and outer boundaries in the radial direction through which waves,
187: fluids, and magnetic fields can pass freely.
188:
189: In our simulations, $\beta$ is higher than 1 in the disk initially
190: ($\beta \sim 40$ at $r=5r_{\rm S}$). Therefore, in principle, MRI
191: grows. However, the wavelength at the maximum growth rate is
192: $\lambda_{\rm MRI} \sim v_{\rm Az}/\Omega $ where $\Omega$ is the
193: angular velocity and $v_{\rm Az}$ is the $z$-component of the
194: Alfv\'{e}n velocity (e.g., Balbus \& Hawley 1991, Gammie 2004). With
195: our initial conditions, the wavelength at the maximum growth rate is
196: larger than $0.6 r_{\rm S}$ at $r= 5.0 r_{\rm S}$ and larger than
197: the disk height ($0.3 r_{\rm S}$). Therefore the thin disk does not
198: support the growth of MRI at this wavelength. Also, shorter wavelengths
199: associated with MRI have slower growth rate and may not grow
200: significantly on our simulation time scale with our grid resolution.
201: Thus, our simulations investigate the physics of jet formation and
202: the properties of the jets in the context of a black hole with a
203: thin Keplerian disk threaded by vertical magnetic fields (using Wald
204: solution) in the absence of significant MRI.
205:
206: In contrast, previous GRMHD simulations with a thick torus
207: have been used to study disk evolution via MRI and the
208: resulting outflows (e.g., De Villiers et al. 2003; McKinney \&
209: Gammie 2004; McKinney 2006; De Villiers et al. 2005; Hawley \&
210: Krolik 2006). Here MRI creates turbulent structure in the thick disk,
211: fluctuating accretion into the black hole, and unsteady
212: outflows near the funnel wall, e.g., Hawley \&
213: Krolik (2006).
214:
215: \section{Results}
216:
217: Figure \ref{f1a} shows snapshots of the density (panels (a) and
218: (b)), plasma beta ($\beta=p_{\rm gas}/p_{\rm mag}$) distribution
219: (panels (c) and (d)), and total velocity (panels (e) and (f)) for
220: the non-rotating black hole case, $a=0.0$ (left panels); and the
221: rapidly rotating black hole case, $a=0.95$ (right panels); at each
222: simulation's terminal time (non-rotating: $t = 275\tau_{\rm S}$ and
223: rotating: $t = 200\tau_{\rm S}$). The Keplerian disk at the
224: marginally stable circular orbit ($r = 3r_{\rm S}$) rotates around
225: the black hole in about $40 \tau_{\rm S}$. In the non-rotating black
226: hole case about 7 inner disk rotations occurred during the
227: simulation.
228:
229: The numerical results show that matter in the disk loses angular
230: momentum to the magnetic field and falls into the black hole. A
231: centrifugal barrier decelerates the falling matter and produces a shock
232: around $r=2 r_{\rm S}$. Matter near the shock region is
233: accelerated by the $\mathbf{J} \times \mathbf{B}$ force and the gas
234: pressure gradient in the z-direction (see the arrows in the panels in
235: Figure 1). In the simulations, matter is continuously supplied from the accretion disk
236: and the jet propagates outward through the outer boundary of the
237: grid. In general, the results are similar to the previous work
238: of Koide et al.\ (1998, 1999, 2000) and Nishikawa et al.\
239: (2005).
240:
241: Contours of the toroidal magnetic field strength are shown by the
242: white lines in Figs.\ 1c and 1d and indicate where the magnetic field
243: is most twisted.
244: In the non-rotating black hole case the magnetic field is twisted by
245: the rotation of the Keplerian disk near the black hole region,
246: propagates outwards along the poloidal magnetic field as an Alfv\'{e}n
247: wave and forms a jet. In the rapidly
248: rotating black hole case the magnetic field is strongly twisted by
249: the frame-dragging effect of the rotating black hole near the black hole
250: region e.g., Figure 1d contours and Figure 2, propagates outwards as an
251: Alfv\'{e}n wave along the poloidal magentic field lines shown in Fig.\ 1b
252: and forms an additional inner jet component closer to the black hole along the
253: rotation axis.
254:
255: The total velocity distribution of non-rotating and rapidly rotating
256: black hole cases are shown in Figs. 1e and 1f. The jets in both
257: cases have speeds greater than $0.4 c$ (mildly relativistic) and the
258: speeds are comparable to the Alfv\'{e}n speeds. In the jets,
259: toroidal velocity is the dominant velocity component (see Figs. 3a
260: and 3d). In the rapidly rotating black hole case the velocity
261: distribution indicates a two-component jet. The outer jet is similar to
262: that of the non-rotating black hole case but the inner jet is not
263: seen in the non-rotating black hole case. The inner jet is faster
264: than the outer jet (over $0.5c$).
265:
266:
267: Figure 3 shows the distribution of various
268: physical quantities on the $z/r_{\rm S} = 2$ surface at the terminal
269: simulation time and allows us to examine the relationship between
270: the twisted magnetic field and the jet velocity. In the
271: non-rotating black hole case the jet is generated in the region $r
272: \ge 2.7 r_{\rm S}$ as shown in Figs.\ 1a and 3a. In this region, the
273: toroidal velocity ($v_{\phi}$) is the dominant velocity component.
274: The magnetic field is twisted by the rotation of the Keplerian
275: disk. As the jet propagates outward, the magnetic field is
276: concentrated towards the rotation axis and the vertical component of
277: the magnetic field, $B_{\rm z}$, becomes larger near the rotation
278: axis. In the rapidly rotating black hole case, a jet is generated in
279: the region $r \ge 0.5 r_{\rm S}$. Now, two jet components
280: can be seen in the velocity distribution. An inner jet is located
281: at $0.7 r_{\rm S} \le r \le 2.0 r_{\rm S} $ and an outer jet is
282: located at $r \ge 2.0 r_{\rm S}$ (Fig.\ 3d). For both jet
283: components, the toroidal velocity component is dominant. The
284: toroidal velocity of the inner jet is larger than that of the outer
285: jet. Here the magnetic field near the black hole region is
286: strongly twisted by the frame-dragging effect as shown in Fig. 2.
287: Since the magnetic field is twisted and pinched toward the rotation
288: axis, the $z$ component of the magnetic field becomes large near the
289: rotation axis (Fig.\ 3e), and the inner jet accelerates more rapidly
290: than the outer jet.
291:
292: To confirm the jet acceleration mechanism, we evaluate the vertical
293: components of the Lorentz force, $\mathbf{F}_{\rm EM}= \rho_{\rm e}
294: \mathbf{E} + \mathbf{J} \times \mathbf{B}$ and the gas pressure
295: gradient, $\mathbf{F}_{\rm gp} = - \nabla p$ at the $z=2.0 r_{\rm
296: S}$ surface (e.g., Figs.\ 3c and 3f). Although this analysis is
297: simpler than that of Hawley \& Krolik (2006), it illustrates which
298: force dominates jet formation, and has been used in previous work
299: (e.g., Koide et al 1999, 2000, 2006; Mizuno et al.
300: 2004a,b; Nishikawa et al. 2005).
301:
302: This simple analysis clearly shows that the outer jet is accelerated mainly by
303: the gas pressure gradient. The inner part of the outer jet may be
304: accelerated partially by the Lorentz force but the Lorentz force in the
305: outer part of the outer jet in the rapidly rotating black hole case is
306: lower than the gas pressure gradient by an order of magnitude. The inner jet
307: is accelerated by the Lorentz force. Therefore the acceleration
308: mechanism is different in the inner and outer jets.
309:
310: \section{Summary and Discussion}
311:
312: We have performed simulations of jet formation from a geometrically
313: thin accretion disk near both non-rotating and rotating black holes
314: using the newly developed RAISHIN code (Mizuno et al. 2006).
315: Simulation results show the formation of magnetically-driven jets
316: accelerated by the Lorentz force and the gas pressure. In general, results
317: are similar to those found in previous GRMHD
318: simulations with a thin Keplerian disk (Koide et al.\ 1998, 1999,
319: 2000; Nishikawa et al. 2005). However, in this study the rotating
320: black hole creates a second, faster, and more collimated inner
321: matter-dominated jet formed by the twisted magnetic field resulting
322: from frame-dragging in the black hole ergosphere. Thus, kinematic
323: jet structure depends on black hole rotation. While an inner jet is
324: created by a counter-rotating thin
325: Keplerian disk (Koide et al. 2000), the inner jet seen in
326: our simulations is a new result not found in a previous co-rotating
327: thin Keplerian disk case (Koide et al.\ 2000). This difference appears to
328: be the result of the weaker initial magnetic field strength used in
329: the present simulations. With stronger magnetic field (Koide et al.\ 2000),
330: transport of angular momentum from the disk is more rapid,
331: accreting matter falls
332: more quickly, is decelerated by the centrifugal
333: barrier near the black hole strongly, and generates a stronger shock.
334: As a result, jets are accelerated by the gas-pressure rather than
335: the Lorentz force of magnetic field twisted by frame-dragging. The
336: jet seen in the co-rotating disk simulation in Koide et al. (2000)
337: corresponds to the outer jet seen in our present co-rotating disk
338: simulation. Therefore this is a first simulation confirming the
339: formation of an inner matter-dominated jet driven by the magnetic
340: field twisted by the frame-dragging effect from a black hole
341: co-rotating with a thin Keplerian disk threaded by a vertical
342: magnetic field.
343:
344: A two component jet structure has also been seen in GRMHD
345: simulations of a black hole co-rotating with a thick torus (e.g.,
346: Hawley \& Krolik 2006; McKinney 2006). One component is a
347: matter-dominated outflow (funnel wall jet) with a mildly
348: relativistic velocity ($\sim 0.3c$) along the centrifugal barrier
349: accelerated and collimated by magnetic and gas pressure forces in
350: the inner torus and the surrounding corona. This formation mechanism
351: and jet are the same as the outer jet seen in our simulations.
352: Therefore the funnel wall jet in Hawley \& Krolik (2006) is the same
353: as the outer jet in our GRMHD simulations. Their other component is
354: a highly-relativistic Poynting flux dominated jet that is produced
355: from the formation of a large scale radial magnetic field within the
356: funnel. In our simulation, such a highly-relativistic Poynting flux
357: dominated jet is not seen. This is likely caused by the difference
358: in the initial magnetic field configuration. In Hawley \& Krolik
359: (2006) and McKinney (2006), there are initial ``poloidal-loop''
360: magnetic fields inside the torus. In the simulations, the magnetic field is
361: twisted and expands from the torus as a magnetic tower and fills the
362: funnel region (Hirose et al. 2004). In the late stage of the
363: simulations a highly-relativistic Poynting flux dominated jet is
364: formed in the low density regions in the funnel. On the other hand,
365: our magnetic field vertically threads the disk and ergosphere
366: initially. The vertical magnetic field near the black hole region is
367: twisted by the frame-dragging effect (see Fig. 2) and forms the
368: inner jet. The inner jet is much denser than the highly-relativistic
369: Poynting flux dominated jet of Hawley \& Krolik (2006) and McKinney
370: (2006) because in the inner jet the matter is supplied from the
371: accretion disk and the free-falling corona. While our jets are
372: formed by the same basic mechanisms as in Hawley \& Krolik (2006)
373: and McKinney (2006), the different initial magnetic configuration
374: has led to different jet properties and a slower matter dominated
375: jet spine. Previous results with the co-rotating disk case of Koide
376: et al. (2000) found that a stronger magnetic field suppressed the
377: formation of the presently observed matter-dominated jet spine.
378:
379:
380: \acknowledgments
381:
382: Y. M. is a NASA Postdoctoral Program fellow at NASA Marshall Space
383: Flight Center. He thanks G.\ Richardson, D.\ Hartmann, C.\ Fendt,
384: and M.\ Camenzind for useful discussions. K.\ N. acknowledges
385: partial support by National Science Foundation awards ATM-0100997,
386: INT-9981508, and
387: AST-0506719, and the National Aeronautic and Space Administration
388: award NNG05GK73G to the University of Alabama in Huntsville. P.\
389: H. acknowledges partial support by National Space Science and
390: Technology Center (NSSTC/NASA) cooperative agreement NCC8-256 and
391: NSF award AST-0506666 to the University of Alabama. The simulations
392: have been performed on the IBM p690 at the National Center for
393: Supercomputing Applications (NCSA) which is supported by the NSF and
394: Altix3700 BX2 at YITP in Kyoto University.
395:
396: \begin{thebibliography}{}
397:
398: \bibitem[Anninos et al.(2005)]{Ann05} Anninos, P., Fragile, P. C., \& Salmonson,
399: J. D. 2005, \apj, 635, 723
400:
401: \bibitem[Ant\'{o}n et al.(2006)]{Ant06} Ant\'{o}n, L., et al. 2006,
402: \apj, 637, 296
403:
404: \bibitem[Balbus \& Hawley(1991)]{Balb91} Balbus, S. A. \& Hawley, J. F. 1991,
405: \apj,376, 214
406:
407: \bibitem[Beckwith et al.(2006)]{Bec06} Beckwith, K., Hawley, J. F., \& Krolik, J. H. 2006, \apj, submitted (astro-ph/0605295)
408:
409: \bibitem[Blandford \& Znajek(1977)]{Bla77} Blandford, R. D. \& Znajek, R. L.
410: 1977, \mnras, 179, 433
411:
412: \bibitem[Blandford \& Payne(1982)]{Bla82} Blandford, R. D. \& Payne, D. G. 1982,
413: \mnras, 199, 883
414:
415: \bibitem[De Villiers \& Hawley(2003)]{DeVH03} De Villiers, J.-P. \& Hawley, J. F. 2003,
416: \apj, 589, 458
417:
418: \bibitem[De Villiers et al.(2003)]{DeV03} De Villiers, J.-P., Hawley, J. F.,
419: \& Krolik, J. H. 2003, \apj, 599, 1238
420:
421: \bibitem[De Villiers et al.(2005a)]{DeV05a} De Villiers, J.-P., Hawley, J. F.,
422: Krolik, J. H., \& Hirose, S. 2005a, \apj, 620, 878
423:
424: \bibitem[De Villiers et al.(2005b)]{DeV05b} De Villiers, J.-P., Staff, J., \&
425: Ouyed, R. 2005b, in preparation (astro-ph/0502225)
426:
427: \bibitem[Ferrari(1998)]{Fer98} Ferrari, A. 1998, \araa, 36, 539
428:
429: \bibitem[Gammie et al.(2003)]{Gam03} Gammie, C. F., McKinney, J. C., \& T\'{o}th, G. 2003, \apj, 589, 444
430:
431: \bibitem[Gammie(2004)]{Gam04} Gammie, C. F. 2004, \apj, 614, 309
432:
433: \bibitem[Fukue(1990)]{Fuk90} Fukue, J. 1990, \pasj, 42, 793
434:
435: \bibitem[Hawley \& Krolik(2006)]{Haw06} Hawley \& Krolik 2006, \apj, 614, 103
436:
437: \bibitem[Hirose et al.(2004)]{hiro04} Hirose, S., Krolik, J., De Villiers, J. P., \&
438: Hawley, J. F. 2004, ApJ, 606, 1083
439:
440: \bibitem[Koide et al.(1998)]{Koi98} Koide, S., Shibata, K.,
441: \& Kudoh, T. 1998, \apjl, 495, L63
442:
443: \bibitem[Koide et al.(1999)]{Koi99} -----. 1999, \apj, 522, 727
444:
445: \bibitem[Koide et al.(2000)]{Koi00} Koide, S., Meier, D. L., Shibata, K.,
446: \& Kudoh, T. 2000, \apj, 536, 668
447:
448: \bibitem[Koide(2003)]{Koi03} Koide, S. 2003, \prd, 67, 104010
449:
450: \bibitem[Koide et al.(2006)]{Koi06} Koide, S., Kudoh, T., \& Shibata, K. 2006, \prd, 74,
451: 044005
452:
453: \bibitem[Komissarov(2004)]{Kom04} Komissarov, S. S. 2004, \mnras, 350,
454: 1431
455:
456: \bibitem[Komissarov(2005)]{Kom05} -----. 2005, \mnras, 359, 801
457:
458: \bibitem[Krolik et al.(2005)]{Kro05} Krolik, J. H., Hawley, J. F., \& Hirose, S. 2005,
459: \apj, 622, 1008
460:
461: \bibitem[McKinney(2005)]{McK05} McKinney, J. C. 2005, \apjl, L5
462:
463: \bibitem[McKinney(2006)]{McK06} -----. 2006, \mnras, 368, 1561
464:
465: \bibitem[McKinney \& Gammie(2004)]{McK04} McKinney, J. C. \& Gammie,
466: C. F. 2004, \apj, 977
467:
468: \bibitem[M\'{e}sz\'{a}ros(2006)]{Mez06} M\'{e}sz\'{a}ros, P. 2006, Rep. Prog.
469: Phys., in press (astro-ph/0605208)
470:
471: \bibitem[Mirabel \& Rodr\'{i}guez(1999)]{Mir99} Mirabel, I. F. \& Rodr\'{i}guez,
472: L. F. 1999, \araa, 37, 409
473:
474: \bibitem[Mizuno et al.(2004a)]{Miz04a} Mizuno, Y., Yamada, S., Koide, S.,
475: \& Shibata, K. 2004a, \apj, 606, 395
476:
477: \bibitem[Mizuno et al.(2004b)]{Miz04b} -----. 2004b, \apj, 615, 389
478:
479: \bibitem[Mizuno et al.(2006a)]{Miz06} Mizuno, Y., Nishikawa, K.-I., Koide, S.,
480: Hardee, P., \& Fishman, G. J. 2006, \apjs, submitted (astro-ph/0609004)
481:
482: \bibitem[Nishikawa et al.(2005)]{Nis05} Nishikawa, K.-I., Richardson, G., Koide, S.,
483: Shibata, K., Kudoh, T., Hardee, P., \& Fishman, G. J. 2005, \apj,
484: 625, 60
485:
486: \bibitem[Penrose(1969)]{Pen69} Penrose, R. 1969, Nuovo Cimento, 1, 252
487:
488: \bibitem[Piran(2005)]{Pir05} Piran, T. 2005, Reviews of Modern Physics, 76,
489: 1143
490:
491: \bibitem[Urry \& Padovani(1995)]{Urr95} Urry, C. M. \& Padovani,
492: P. 1995, \pasp, 107, 803
493:
494: \bibitem[Wald(1974)]{Wal74} Wald, R. M. 1974, \prd, 10, 1680
495:
496: \bibitem[Zhang \& M\'{e}sz\'{a}ros(2004)]{Zha05} Zhang, B. \& M\'{e}sz\'{a}ros, P.
497: 2004, Int. J. Mod. Phys., A19, 2385
498:
499: \end{thebibliography}
500:
501: \newpage
502:
503: \begin{figure}
504: \epsscale{0.70} \plotone{f1.eps} \caption{Snapshots of density and
505: plasma beta of the non-rotating black hole case ($a=0.0$; {\it a, c,
506: e}) and the rapidly rotating black hole case ($a=0.95$; {\it b, d,
507: f}) at the applicable terminal simulation time. The color scales
508: show the logarithm of density (upper panels), plasma beta ($\beta =
509: p_{\rm gas} / p_{\rm mag}$; middle panels) and total velocity (lower
510: panels). A negative velocity indicates inflow towards the black
511: hole. The white lines indicate magnetic field lines (contour of the
512: poloidal vector potential; upper panels) and contours of the toroidal
513: magnetic field stength (middle panels). Arrows depict the poloidal
514: velocities normalized to light speed, as indicated above each panel
515: by the arrow. \label{f1a}}
516: \end{figure}
517:
518: \begin{figure}
519: \epsscale{0.7} \plotone{f2.eps} \caption{Bent and twisted magnetic
520: field lines obtained from the co-rotating black hole case at $t=133
521: \tau_{\rm S}$.
522: The field lines are traced from $z = - 15r_{\rm S}$
523: to $z = + 15r_{\rm S}$ beginning at $r = 6 r_{\rm S}$ (red lines)
524: and $r = 12 r_{\rm S}$ (blue lines). The black hole is indicated by
525: the green circle.}
526: \end{figure}
527:
528: \newpage
529:
530: \begin{figure}
531: \epsscale{0.85} \plotone{f3.eps} \caption{Various physical
532: quantities on the $z/r_{\rm S} = 2$ surface at the applicable
533: terminal simulation time. Panels {\it (a), (d)} show the velocity
534: components $v_{\rm r}$ ({\it solid line}), $v_{\phi}$ ({\it dashed
535: line}), and $v_{\rm z}$ ({\it dot-dashed line}). Panels {\it (b),
536: (e)} show the magnetic field components $B_{\rm r}$ ({\it solid
537: line}), $B_{\phi}$ ({\it dashed line}), and $B_{\rm z}$. ({\it
538: dot-dashed line}). Panels {\it (c), (f)} show the $z$ component of
539: the Lorentz force $\mathbf{F}_{\rm EM}$ ({\it solid line}) and the
540: gas pressure gradient $\mathbf{F}_{\rm gp}$ ({\it dashed line}).
541: \label{f3}}
542: \end{figure}
543:
544: \newpage
545:
546:
547:
548: \end{document}
549: