1: %draft 3 - 12 Jun 08
2: %---------------------------------------------------------------------------
3: %Use this line for submission to ApJ
4: %\documentclass[12pt,preprint]{aastex}
5: %Use this line for a more compact single-column version to submit to LANL
6: %\documentclass[10pt,preprint]{aastex}
7: %Use this line for a compact two-column version
8: %\documentclass[10pt,preprint2]{aastex}
9: \documentclass{emulateapj}
10:
11: %\usepackage{emulateapj5}
12: %\usepackage{amsmath,natbib,graphicx,euscript} %,epsfig}
13: \usepackage{epstopdf}
14: \usepackage{graphics,graphicx}
15:
16: \newcommand{\myemail}{fragilep@cofc.edu}
17: \slugcomment{To appear in ApJ}
18:
19: %-----------------------------------------------------------------------------
20: \begin{document}
21: %------------------------------------------------------------------------------
22:
23: \title{Epicyclic Motions and Standing Shocks in Numerically Simulated Tilted Black-Hole Accretion Disks}
24:
25: \author{P. Chris Fragile}
26: \affil{Department of Physics \& Astronomy, College of Charleston,
27: Charleston, SC 29424}
28: \email{fragilep@cofc.edu}
29:
30: \and
31:
32: \author{Omer M. Blaes}
33: \affil{Department of Physics, University of California, Santa Barbara, CA 93106}
34:
35: \date{{\small \today}}
36: \date{{\small \LaTeX-ed \today}}
37: %-----------------------------------------------------------------------------
38:
39: \begin{abstract}
40: This work presents a detailed analysis of the overall flow structure and unique features of the inner region of the tilted disk simulations described in Fragile et al. (2007). The primary new feature identified in the main disk body is a latitude-dependent radial epicyclic motion driven by pressure gradients attributable to the gravitomagnetic warping of the disk. The induced motion of the gas is coherent over the scale of the entire disk and is fast enough that it could be observable in features such as relativistic iron lines. The eccentricity of the associated fluid element trajectories increases with decreasing radius, leading to a crowding of orbit trajectories near their apocenters. This results in a local density enhancement akin to a compression. These compressions are sufficiently strong to produce a pair of weak shocks in the vicinity of the black hole. These shocks are roughly aligned with the line-of-nodes between the black-hole symmetry plane and disk midplane, with one shock above the line-of-nodes on one side of the black hole and the other below on the opposite side. These shocks enhance angular momentum transport and energy dissipation near the hole, forcing some material to plunge toward the black hole from well outside the innermost stable circular orbit. A new, extended simulation, which was evolved for more than a full disk precession period, allows us to confirm that these shocks and the previously identified ``plunging streams'' precess with the disk in such a way as to remain aligned relative to the line-of-nodes, as expected based on our physical understanding of these phenomena. Such a precessing structure would likely present a strong quasi-periodic signal.
41:
42: \end{abstract}
43:
44: \keywords{accretion, accretion disks --- black hole physics ---
45: galaxies: active --- MHD --- relativity --- X-rays: stars}
46:
47:
48: \section{Introduction}
49: \label{sec:intro}
50:
51: The purpose of this paper is to present a more detailed analysis of
52: the numerical simulation results described in \citet{fra07b}
53: (hereafter Paper I). That paper described a global numerical
54: simulation of an accretion disk subject to the magneto-rotational
55: instability (MRI) that was misaligned (tilted) with respect to the
56: rotation axis of a modestly fast rotating ($a/M=0.9$) Kerr black
57: hole. It was the first such numerical simulation of a tilted disk to
58: fully incorporate the effects of the black hole spacetime and not
59: rely on an {\em ad hoc} prescription of angular momentum transport.
60:
61: A clear identification of a tilted black-hole accretion disk in Nature has yet to be made. Nevertheless, there is reason to believe they may be quite common. In any accretion disk system, the orientation of the disk is set by the net angular momentum of the gas reservoir on large scales. The orientation of the black hole, on the other hand, can either be set by its formation or its evolution. For stellar-mass black holes in low-mass binary systems, only the formation is likely to matter (any evolution subsequent to the formation is unlikely to change the orientation of the black hole significantly). Since the formation of the black hole is largely independent of the angular momentum of the gas reservoir, alignment should generally not be expected in this case. For high mass binaries and active galactic nuclei (AGN), on the other
62: hand, a large episode of misaligned accretion could reorient the spin of
63: the black hole \citep{nat99}. Therefore a more detailed
64: understanding of the evolutionary history of the system, including
65: mergers
66: in the case of AGN, would be needed to know if a tilted configuration is
67: expected.
68:
69: A tilted disk is subject to differential warping due to the effect of Lense-Thirring precession. In disks with a large ``viscous'' stress (parametrized by the \citet{sha73} $\alpha$ parameter) the warping results in the Bardeen-Petterson configuration \citep{bar75,kum85}, characterized by an
70: alignment of the disk with the equatorial plane of the black hole inside some characteristic warp radius. The Bardeen-Petterson effect has been invoked by a number of authors to explain peculiar observations, such as misaligned jets in AGN \citep{kon05,cap06,cap07} and X-ray binaries \citep{fra01a,mac02}.
71:
72: However, we did not see evidence for the Bardeen-Petterson effect in our simulation in Paper I. This
73: was not surprising since our simulation was carried out in the
74: low stress regime,
75: with $\alpha < H/r$, where $H$ is the half-height of the disk.
76: In this limit warps produced in the disk propagate as waves
77: \citep{pap95a}, rather than diffusively as in the Bardeen- Petterson
78: case. Instead of a smooth transition between an untilted disk at small radii and tilted disk at large radii, as with the Bardeen-Petterson effect, we found the tilt to be a long-wavelength oscillatory function of radius.
79: %For X-ray binaries, the thick-disk regime most likely corresponds to the Hard or possibly Quiescent %states, in which no thermal (thin) disk component is observed \citep{mcc06}.
80:
81: The tilted simulation in Paper I also showed other dramatic
82: differences from comparable simulations of untilted disks. Accretion
83: onto the hole occurred predominantly through two opposing plunging
84: streams that started from high latitudes with respect to both the
85: black-hole and disk midplanes.
86: %Fundamentally, this was due to the
87: %aspherical nature of the gravitational spacetime around the rotating
88: %black hole.
89: These plunging streams also started from a larger radius than the
90: innermost stable circular orbit (ISCO), which is often assumed to
91: represent the inner edge of untilted disks. We interpreted this as
92: being due to the fact that the tilted disk encounters a
93: generalized ISCO surface at a larger cylindrical radius than an
94: untilted disk \citep{fra07a}. In this regard the tilted black hole
95: effectively acts like an untilted black hole of lesser spin.
96:
97: Because of the fast sound-crossing time in the disk, the torque of
98: the black hole acted globally rather than differentially. Instead of
99: strongly warping the disk, the torque caused the disk to experience
100: global (solid-body) precession. The precession had a frequency of $\nu_\mathrm{prec} = 3
101: (M_\odot/M)$ Hz, a value consistent with many observed low-frequency
102: quasi-periodic oscillations (QPOs). However, this value is strongly
103: dependent on the size of the disk ($\nu_\mathrm{prec} \propto r_o^{-5/2}$, where $r_o$ is the outer radius), so this frequency may be expected
104: to be vary over a large range. In the limit of a very large disk or in a case where the disk is strongly coupled to a gas reservoir at large radii, this precession frequency may be
105: expected to drop to zero.
106:
107: In the present work we expound on some of the features of the
108: tilted-disk simulation which were not described in Paper I. This
109: paper is organized as follows: In \S \ref{sec:methods} we briefly
110: review the details of the simulations presented in Paper I, in
111: particular focusing on models 90h (untilted) and 915h (tilted),
112: which are identical other than the initial tilt of the black hole
113: relative to the disk ($\beta_0=0$ and $15^\circ$, respectively), and
114: model 915m, which is a lower resolution version of 915h. In \S
115: \ref{sec:bending}, we describe the large-scale epicyclic motion that
116: appears in the tilted simulation but is absent in the comparable
117: untilted simulation. This epicyclic motion is driven by radial pressure gradients associated with the stationary bending
118: wave created by the warping action of the gravitomagnetic torque of
119: the black hole (e.g. \citealt{nel99}).
120: %Interestingly, we find that the resulting bending
121: %wave is quasistationary and leads to the excitation of radial and
122: %azimuthal motions that are odd functions of $z$.
123: Next, in \S \ref{sec:shock}, we describe a pair of standing shocks
124: that again have no counterparts in
125: untilted simulations. The shocks form roughly
126: along the line-of-nodes between the disk midplane and black-hole
127: symmetry plane, one shock on each side of the
128: black hole. We verify that this is not a chance alignment by following the long-term evolution of
129: model 915m and confirming that the shocks remain aligned with the line-of-nodes as it precesses with the disk. Finally, in \S \ref{sec:discussion}, we consider some of
130: the implications of our discoveries, particularly in the context of
131: relativistic iron lines.
132:
133:
134: \section{Numerical Methods}
135: \label{sec:methods}
136:
137: The simulations in Paper I were carried out using the Cosmos++
138: astrophysical magnetohydrodynamics (MHD) code \citep{ann05}.
139: Cosmos++ includes several schemes for solving the GRMHD equations;
140: in Paper I, the artificial viscosity formulation was used. The
141: magnetic fields were evolved in an advection-split form, while using
142: a hyperbolic divergence cleanser to maintain an approximately
143: divergence-free magnetic field. The GRMHD equations were evolved in
144: a ``tilted'' Kerr-Schild polar coordinate system
145: $({t},{r},{\vartheta},{\varphi})$. This coordinate system is related
146: to the usual (untilted) Kerr-Schild coordinates
147: $({t},{r},{\theta},{\phi})$ through a simple rotation about the
148: ${y}$-axis by an angle $\beta_0$, as described in \citet[][see also
149: Fragile \& Anninos 2007]{fra05b}.
150:
151: The simulations were carried out on a spherical polar mesh with
152: nested resolution layers. The base grid contained $32^3$ mesh zones
153: and covered the full $4\pi$ steradians. Varying levels of refinement
154: were added on top of the base layer; each refinement level doubling
155: the resolution relative to the previous layer. The main simulations,
156: models 90h and 915h from Paper I, had two levels of refinement, thus
157: achieving peak resolutions equivalent to a $128^3$ simulation. To
158: demonstrate convergence, in Paper I we also presented results at
159: higher and lower resolutions. In this paper we also discuss results
160: of model 915m, which used a single level of refinement and had a
161: peak resolution equivalent to a $64^3$ simulation, but was run for
162: significantly longer.
163:
164: In the radial direction a logarithmic coordinate of the form $\eta
165: \equiv 1.0 + \ln (r/r_{\rm BH})$ was used, where $r_{\rm BH}=1.43 r_G$ is the black-hole horizon radius and $r_G=GM/c^2$ is the gravitational radius. The spatial resolution
166: near the black-hole horizon was $\Delta r \approx 0.05 r_G$; near
167: the initial pressure maximum of the torus, the resolution was
168: $\Delta r \approx 0.5 r_G$. In the angular direction, in addition to
169: the nested grids, a concentrated latitude coordinate $x_2$ of the
170: form $\vartheta = x_2 + \frac{1}{2} (1 - h) \sin (2 x_2)$ was used
171: with $h = 0.5$, which concentrates resolution toward the midplane of
172: the disk. As a result $r_{\rm center} \Delta \vartheta = 0.3 r_G$
173: near the midplane while it is a factor of $\sim 3$ larger for the
174: fully refined zones near the pole. This grid is shown in Figure 1 of
175: Paper I.
176:
177: The simulations started from the analytic solution for an
178: axisymmetric torus around a rotating black hole \citep{cha85}. The
179: initial torus was identical to model KDP of \citet{dev03c}, which is
180: the relativistic analog of model GT4 of \citet{haw00}. As in model
181: KDP, the spin of the black hole was $a/M=0.9$; the inner radius of
182: the torus was $r_{\rm in}=15 r_G$; the radius of the initial
183: pressure maximum of the torus $r_{\rm center}=25 r_G$; and the
184: power-law exponent used in defining the initial specific angular
185: momentum distribution was $q=1.68$. An adiabatic equation of state
186: was assumed, with $\Gamma=5/3$. The torus was seeded with a weak
187: dipole magnetic field in the form of poloidal loops along the
188: isobaric contours within the torus. The field was normalized such
189: that initially $\beta_{\rm mag} =P/P_B \ge \beta_{\rm mag,0}=10$
190: throughout the torus. For the tilted simulations (915h and 915m) the black
191: hole was inclined by an angle $\beta_0=15^\circ$ relative to the
192: disk (and the grid) through a transformation of the Kerr metric.
193: From this starting point, simulations 90h and 915h were allowed to
194: evolve for a time equivalent to 10 orbits at the initial pressure
195: maximum, $r_{\rm center}$, corresponding to hundreds of orbits at
196: the ISCO. Model 915m was evolved for 100 orbital times at the
197: initial pressure maximum. Table 1 summarizes the parameters of
198: all three simulations.
199:
200: \begin{deluxetable}{cccccc}
201: \tabletypesize{\scriptsize}
202: \tablecaption{Simulation Parameters \label{tbl-1}}
203: \tablewidth{0pt}
204: \tablehead{
205: \colhead{Simulation} & \colhead{Tilt} &
206: \colhead{$a/M$} & \colhead{Equivalent} &
207: \colhead{$r_{\rm center}$\tablenotemark{a}}
208: & \colhead{Duration\tablenotemark{b}} \\
209: \colhead{} & \colhead{Angle} & \colhead{} &
210: \colhead{Peak} & \colhead{} & \colhead{} \\
211: \colhead{} & \colhead{} & \colhead{} &
212: \colhead{Resolution} & \colhead{} & \colhead{} \\
213: }
214: \startdata
215: 90h & 0 & 0.9 & $128^3$ & 25 $r_{\rm G}$ & 10 \\
216: 915m & $15^\circ$ & 0.9 & $64^3$ & 25 $r_{\rm G}$ & 100 \\
217: 915h & $15^\circ$ & 0.9 & $128^3$ & 25 $r_{\rm G}$ & 10 \\
218: \enddata
219:
220: \tablenotetext{a}{Radius of initial pressure maximum.}
221: \tablenotetext{b}{In units of $t_{\rm orb}$, the geodesic orbital period
222: at the initial pressure maximum $r_{\rm center}$.}
223:
224: \end{deluxetable}
225:
226: \section{Epicyclic Motion}
227: \label{sec:bending}
228:
229: As noted above, Lense-Thirring precession causes differential warping in tilted black-hole accretion disks. In the thick-disk regime, appropriate for the simulations described in Paper I, warp disturbances are expected to propagate in a wave-like, rather than diffusive, manner. For nearly Keplerian disks, such as those resulting from MRI turbulent simulations, the resulting bending wave is expected to produce large horizontal motion within the disk \citep{nel99,tor00}. The radial and azimuthal components of this motion are odd functions of $z$, leading to large vertical shear ($\partial V^r/\partial z$, $\partial V^\phi/\partial z$).
230:
231: %Before describing the numerical results we want to review some of
232: %what is understood about tilted disks in the thick-disk regime. From
233: %\citet{nel99}, the form of the perturbed velocity field for a warped
234: %disk is
235: %\begin{eqnarray}
236: %\frac{\partial v_r^\prime}{\partial t} + i \Omega v_r^\prime -
237: %2\Omega
238: %v_\phi^\prime = - \frac{\partial W}{\partial r} ~, \nonumber \\
239: %\frac{\partial v_\phi^\prime}{\partial t} + i \Omega v_\phi^\prime +
240: %v_r^\prime r^{-1} \frac{d(r^2 \Omega)}{dr} = -i \frac{W}{r} ~,
241: %\nonumber \\
242: %\frac{\partial v_z^\prime}{\partial t} + i \Omega v_z^\prime = -
243: %\frac{\partial W}{\partial z} ~,
244: %\end{eqnarray}
245: %where the primes denote perturbations, $\Omega$ is the orbital
246: %frequency, and $W=P^\prime/\rho=\rho^\prime c_s^2/\rho$, with $c_s$
247: %being the sound speed. According to \citep{nel99} ``It can be seen
248: %that, as a bending wave travels through the disk, it lead to the
249: %excitation of radial motions that are odd functions of $z$, such
250: %that a vertical shear is induced.'' ``Slowly varying warps may induce large
251: %horizontal motions and vertical shear.''
252:
253: Figures \ref{fig:rhoVr1} and \ref{fig:rhoVph1} compare the radial and azimuthal motions of the gas for our tilted and untilted simulation (915h and 90h). Clearly
254: the fluid velocity in the tilted disk is no longer dominated by
255: turbulent motion as it is for untilted disks \citep{dev03b}, but has
256: an ordered sense about it. For instance, in Figure
257: \ref{fig:rhoVr1}{\em a} we see that in the right-hand side of the
258: image the gas in the upper layers of the disk is moving radially
259: outward, whereas the gas in the lower layers is moving radially
260: inward. The sense of motion is reversed on the left-hand side of the
261: image. Similarly, on the right-hand side of Figure
262: \ref{fig:rhoVph1}{\em a} the gas in the upper half of the disk is
263: moving slower than the bulk angular velocity, whereas the gas in the
264: lower half is moving faster than this average. Again, the sense of
265: the flow is reversed on the left-hand side of the image. This
266: pattern of motion is reminiscent of epicyclic motion, with the upper
267: and lower halves of the disk executing epicycles that are
268: $180^\circ$ out of phase with one another. Notice that no such
269: organized motion is apparent in our untilted simulation (Figures
270: \ref{fig:rhoVr1}{\em b} and \ref{fig:rhoVph1}{\em b}).
271: %This is simply a
272: %manifestation of the bending wave, as described in \citet{nel99}.
273:
274: We find that the pattern of the epicyclic motion is tied to
275: the orientation of the disk relative to the black hole. As our
276: simulated tilted disk 915m precesses, the velocity pattern of the epicyclic
277: motion changes as represented in Figures \ref{fig:rhoVr2} and
278: \ref{fig:rhoVph2}, where we show the velocity patterns at $t=10$ and
279: $t=50t_\mathrm{orb}$, approximately 1/2 precession period apart.
280: Note that the sense of the epicyclic motion is reversed between the
281: two different evolution times.
282:
283:
284: %\clearpage
285: \begin{figure*}
286: %\plotone{torus3d.m_rho_Vr_xz_10.eps}
287: \includegraphics[scale=.6]{f1.jpg} \caption{Meridional plots
288: ($\varphi=0$) through the final dumps ($t=10t_\mathrm{orb}$) of
289: simulations 915h ({\em left}) and 90h ({\em right}) showing a
290: pseudocolor representation of $V^r$ for outflowing ($V^r > 0$) and
291: inflowing ($V^r < 0$) gas as {\em hot} and {\em cold} colors,
292: respectively. The velocity scale is normalized to the speed of light
293: $c$. The plots are overlaid with isocontours of density at $\rho =
294: 0.4$, 0.04, and $0.004\rho_\mathrm{max,0}$.
295: %(0.01, 0.001, and 0.0001 in code units)
296: Material with $\rho<0.004\rho_\mathrm{max,0}$ is excluded from the
297: figure. The figure is oriented in the sense of the grid, such that
298: the black hole is tilted $15^\circ$ to the left in the {\em left}
299: panel. By this time in simulation 915h the disk has precessed such
300: that the angular momentum axis of the disk is no longer in the plane
301: of this figure. At this point $\mathbf{J}_\mathrm{disk}$ has
302: acquired a positive $y$-component and a negative $x$-component.
303: \label{fig:rhoVr1}}
304: \end{figure*}
305: %\clearpage
306:
307: %\clearpage
308: \begin{figure*}
309: %\plotone{torus3d.m_rho_Vphi_yz_10.eps}
310: \includegraphics[scale=.66]{f2.jpg}\caption{Meridional plots
311: ($\varphi=90^\circ$) through the final dumps ($t=10t_\mathrm{orb}$)
312: of simulations 915h ({\em left}) and 90h ({\em right}) showing a
313: pseudocolor representation of $(V^\phi-\Omega)/\Omega$ for
314: superorbital ($V^\phi
315: > \Omega$) and suborbital ($V^\phi < \Omega$) gas as {\em hot} and
316: {\em cold} colors, respectively, where $\Omega = (M/r^3)^{1/2}/[1+a(M/r^3)^{1/2}]$ is the particle orbital angular frequency. The plots are overlaid with
317: isocontours of density at $\rho = 0.4$, 0.04, and
318: $0.004\rho_\mathrm{max,0}$. Material with
319: $\rho<0.004\rho_\mathrm{max,0}$ is excluded from the figure. The
320: figure is oriented in the sense of the grid, such that the black
321: hole is tilted $15^\circ$ away from the viewer's vertical in the {\em left}
322: panel. \label{fig:rhoVph1}}
323: \end{figure*}
324: %\clearpage
325:
326: %\clearpage
327: \begin{figure*}
328: %\plotone{torus3d.m.915m_rho_Vr_xz.eps}
329: \includegraphics[scale=.66]{f3.jpg} \caption{Meridional plots
330: ($\varphi=0$) at times $t=10$ ({\em left}) and $50t_\mathrm{orb}$ ({\em right})
331: (representing approximately 1/2 precession period difference) from simulation
332: 915m showing a
333: pseudocolor representation of $V^r$ for outflowing ($V^r > 0$) and
334: inflowing ($V^r < 0$) gas as {\em hot} and {\em cold} colors,
335: respectively. The color scale and contours are normalized the same as in Fig. \ref{fig:rhoVr1}. The figure is oriented in the sense of the grid, such that
336: the black hole is tilted $15^\circ$ to the left. Note that the sense of the radial motion as seen from
337: this fixed viewing direction is reversed between the two frames. This is consistent with the epicyclic motion tracking the precession of the disk.
338: \label{fig:rhoVr2}}
339: \end{figure*}
340: %\clearpage
341:
342: %\clearpage
343: \begin{figure*}
344: %\plotone{torus3d.m.915m_rho_Vphi_yz.eps}
345: \includegraphics[scale=.66]{f4.jpg} \caption{Meridional plots
346: ($\varphi=90^\circ$) at times $t=10$ ({\em left}) and $50t_\mathrm{orb}$ ({\em right})
347: (representing approximately 1/2 precession period difference) from simulation
348: 915m showing a
349: pseudocolor representation of $(V^\phi-\Omega)/\Omega$ for
350: superorbital ($V^\phi
351: > \Omega$) and suborbital ($V^\phi < \Omega$) gas as {\em hot} and
352: {\em cold} colors, respectively. The color scale and contours are normalized the same as in Fig. \ref{fig:rhoVph1}. The
353: figure is oriented in the sense of the grid, such that the black
354: hole is tilted $15^\circ$ away from the viewer's vertical.
355: Note that the sense of the residual azimuthal motion as seen from this fixed viewing direction is
356: reversed between the two frames. This is consistent with the epicyclic motion tracking the precession of
357: the disk. \label{fig:rhoVph2}}
358: \end{figure*}
359: %\clearpage
360:
361: Figure \ref{fig:rho_stream} shows the epicyclic motion caused by the
362: bending wave from a different perspective. In the figure, a
363: $4\times4$ lattice of streamlines starts in the $xz$-plane. The
364: lattice is centered above and below the symmetry plane of the black
365: hole and the viewer is looking almost down the black-hole spin axis (the black-hole spin axis is tilted $10^\circ$ away from the viewer's line-of-sight to give a better perspective). As
366: expected from the previous figures of simulation 915h at
367: $t=10t_\mathrm{orb}$, the streamlines that begin above the symmetry
368: plane of the disk are initially moving radially outward, whereas
369: those that begin below are moving radially inward. After
370: approximately one-quarter of an orbit, the upper streamlines have
371: reached their apocenter and the radial motion changes direction.
372: This is also where material encounters one of the standing shocks,
373: described in the next section, which explains the sudden change in
374: direction of the streamlines. Much of this material then begins
375: plunging toward the hole, ultimately accreting within a couple
376: orbits. Thus, as noted in \citet{fra05b}, high-latitude material is
377: preferentially being drained from the disk.
378:
379: %\clearpage
380: \begin{figure}
381: \begin{center}
382: %\plotone{torus3d.m.915h_rho_V_stream.eps}
383: \includegraphics[scale=.675]{f5.jpg} \caption{Isosurface plot
384: of density ({\em semitransparent blue}) plus selected streamlines for simulation 915h. The
385: isosurface at $\rho=0.1\rho_\mathrm{max,0}$ is chosen to highlight
386: the plunging streams. There are 16 streamlines that begin in a $4
387: \times 4$ lattice in the $xz$-plane. The lattice is centered about
388: the symmetry plane of the black hole and the figure is oriented
389: looking almost directly down the spin axis of the hole (red arrow). The spin axis is actually tilted $10^\circ$ away from the viewer's line-of-sight to offer a better perspective.
390: \label{fig:rho_stream}}
391: \end{center}
392: \end{figure}
393: %\clearpage
394:
395: Figure \ref{fig:rho_stream} also reveals that the eccentricity $e$ of the particle streamlines depends on the height of the streamline above the disk midplane (the top streamlines are more eccentric that those one row below). This is consistent with the expectation that $e \propto \xi$ \citep{iva97}, where $\xi$ is the vertical distance measured perpendicular to the midplane of the disk ($\xi$ is the $z$-component of a cylindrical coordinate system ($R$, $\psi$, $\xi$) aligned with each concentric ring of the disk, what is sometimes called the twisting coordinate system).
396:
397: \section{Standing Shocks}
398: \label{sec:shock}
399:
400: %SEE SHAPIRO AND TEUKOLSKY 1983 FOR A DISCUSSION OF STANDING SHOCKS
401: %IN ACCRETION DISKS
402:
403: The full expression for the eccentricity of the orbit of each fluid element is \citep{iva97}
404: %
405: \begin{equation}
406: e = \frac{R\xi}{6M} \Psi = \frac{R\xi}{6M} \frac{\partial(\beta \cos \gamma)}{\partial R} ~,
407: \end{equation}
408: where $\beta$ and $\gamma$ are the tilt and twist of the disk, respectively, defined for each concentric ring. In Figure \ref{fig:eccentricity} we plot $\beta(r)$, $\gamma(r)$, and $r\Psi(r)$ (for computational convenience we have replaced the cylindrical twisting coordinate $R$ with the spherical-polar radius $r$). It is important to note the increase in $r\Psi$ with decreasing $r$. Such behavior, with orbits getting more eccentric closer to the black hole, results in a concentration of fluid element trajectories near their respective apocenters \citep[see Fig. 5a of][]{iva97}. Notice in Figure \ref{fig:rho_stream}, for instance, that each of the top four streamlines makes its closest approach to the next streamline out when it is at its own apocenter.
409:
410: %\clearpage
411: \begin{figure}
412: %\plotone{eccentricity.eps}
413: \plotone{f6} \caption{Plot of the tilt $\langle\beta \rangle_t$, twist $\langle \gamma \rangle_t$, and radial dependence of the orbital eccentricity $r\Psi$ as a function of radius through the disk. The data for this plot has been time-averaged from $t=9t_\mathrm{orb}$ to $10t_\mathrm{orb}$. The initial tilt and twist were 0.2618 and 0, respectively. \label{fig:eccentricity}}
414: \end{figure}
415: %\clearpage
416:
417: The crowding of fluid element trajectories near their apocenters produces a local density enhancement. This will be most pronounced away from the disk midplane because of the dependence of $e$ on $\xi$. Because the fluid elements are traveling supersonically, this local density enhancement produces a weak shock in the flow. In Figure \ref{fig:rhoV} we
418: identify the standing shocks by the sudden change in the magnitude
419: and direction of the velocity vectors along a roughly linear feature
420: associated with the leading edge of the each plunging stream.
421: %The combined plunging stream -- standing shock structures
422: %nearly coincide with the line-of-nodes of the disk -- black-hole
423: %system.
424:
425: %\clearpage
426: \begin{figure}
427: \begin{center}
428: %\plotone{torus3d.m.915h_rho_V_10.eps}
429: \includegraphics[scale=.298]{f7.jpg} \caption{Isosurface plot of
430: density ({\em semitransparent blue}) plus selected velocity vectors for simulation 915h at
431: $t=10t_\mathrm{orb}$. The density isosurface is the same as in Fig.
432: \ref{fig:rho_stream}. The figure is oriented
433: looking directly down the spin axis of the hole. Velocity vectors are only included for gas
434: within $\pm 0.5 r_G$ of the initial disk midplane to prevent
435: overcrowding the image. The sudden change in direction and magnitude
436: of the velocity vectors (highlighted by the black lines) is indicative of a shock.
437: \label{fig:rhoV}}
438: \end{center}
439: \end{figure}
440: %\clearpage
441:
442: Another way to identify the standing shocks is by plotting the
443: magnitude of the vorticity, indicated by $\vert\mathrm{curl~}
444: \mathbf{V}\vert = \vert\nabla \times \mathbf{V}\vert$, which
445: increases at a shock. Figure \ref{fig:rho_curlV} shows an isosurface
446: of $\vert\mathrm{curl~} \mathbf{V}\vert$ overlaid on a density
447: isosurface. Very similar results are obtained if the gradient of
448: the entropy, another indicator of the presence of a shock, is
449: plotted instead. Notice that one of the shock surfaces lies mostly above
450: the chosen isodensity surface, while the other lies below it.
451: This is consistent with the linear dependence of $e$ on $\xi$. We don't expect the shocks to extend into the disk midplane where epicyclic motion ceases. The location of the shocks in azimuth is also consistent with the apocenters of the fluid elements trajectories, as inferred from Figures \ref{fig:rhoVr1} -- \ref{fig:rho_stream}.
452:
453: %\clearpage
454: \begin{figure}
455: %\plotone{torus3d.m.915h_rho_curlV_10.eps}
456: \begin{center}
457: \includegraphics[scale=.298]{f8.jpg} \caption{Isosurface plots
458: of density ({\em semitransparent blue}) and $\vert\mathrm{curl~}
459: \mathbf{V}\vert$ ({\em red}) for simulation 915h at
460: $t=10t_\mathrm{orb}$. $\vert\mathrm{curl~} \mathbf{V}\vert$ is a
461: good tracer of the location of a shock. The plot is restricted to
462: the region $\pi/4<\vartheta<3\pi/4$ to prevent overcrowding of the
463: image from shocks associated with the outflowing jets. The density
464: isosurface is the same as in Fig. \ref{fig:rho_stream}. The figure is oriented
465: looking directly down the spin axis of the hole.
466: \label{fig:rho_curlV}}
467: \end{center}
468: \end{figure}
469: %\clearpage
470:
471: %The alignment of the standing shocks so close to the line-of-nodes
472: %of the disk -- black-hole system gives us a clue as to their origin.
473: %It appears the shocks are caused by supersonic compressions of the
474: %disk that happen twice per orbit.
475: %The importance of the line-of-nodes can be seen by analyzing the
476: %lowest post-Newtonian order force per unit
477: %mass on the fluid: $\mathbf{V}
478: %\times \mathbf{h}$, where $\mathbf{V}$ is the fluid velocity and
479: %\begin{equation}
480: %\mathbf{h} = \frac{2\mathbf{J}_\mathrm{BH}}{r^3} -
481: %\frac{6(\mathbf{J}_\mathrm{BH}\cdot\mathbf{r})\mathbf{r}}{r^5} ~.
482: %\end{equation}
483: %Assuming $\mathbf{V} = (0, V \hat{\mathbf{\phi}}, 0)$ and
484: %$\mathbf{J}_\mathrm{BH} = (0, 0, aM\hat{\mathbf{z}})$ in cylindrical coordinates $(R,\phi,z)$ then
485: %\begin{equation}
486: %\mathbf{V} \times \mathbf{h} = \frac{2aMV}{r^3} \hat{R} -
487: %\frac{6aMV\cos \theta \sin \theta}{r^3} \hat{\theta} ~. \label{eqn:LT}
488: %\end{equation}
489: %assuming $R>>z$.
490: %Notice that the second term in equation (\ref{eqn:LT}) implies a
491: %forcing of the fluid away from the black-hole symmetry plane ($\theta=\pi/2$)
492: %for fluid in a prograde orbit ($\mathbf{J}_\mathrm{BH}$ and
493: %$\mathbf{J}_\mathrm{disk}$ nearly parallel as opposed to nearly
494: %anti-parallel). This is the source of the Lense-Thirring precession
495: %of the disk. Importantly, this term has a dependence on
496: %$\theta$, being strongest near $\theta=\pi/4$ and $3\pi/4$ and going to zero at the black-hole symmetry plane and at the poles. This means that high-latitude disk material near $\theta=\pi/4$ and $3\pi/4$ (in the black-hole frame) potentially can precess faster than the rest of the disk. This peculiar behavior could be inferred from Figure \ref{fig:rhoVph1}{\em b}, where the high-latitude disk material appears to be orbiting faster than the bulk of the disk. In fact, for the tilted-disk simulation, we find that near $\theta=\pi/4$ and $3\pi/4$, the acceleration time, defined as $t_\mathrm{accel} = \vert \mathbf{V} \vert/\vert \mathbf{V} \times \mathbf{h} \vert$, is shorter than the sound-crossing time, $t_\mathrm{cs}$ (see Figure \ref{fig:taccel}). Coincidentally, the standing shocks we have been describing appear to form near the ends of the ``acceleration zones,'' the regions where $t_\mathrm{accel}/t_\mathrm{cs} < 1$. This suggests that the standing shocks are driven by strong gravitomagnetic forcing in the vicinity of the black hole. Even so these
497: %are relatively weak shocks, with Mach numbers typically no larger
498: %than $\mathcal{M} \lesssim 1.5$.
499:
500: %\clearpage
501: %\begin{figure}
502: %\plotone{torus3d.m.915h_rho_taccel_10.eps} \caption{Isosurface plots of
503: %density ({\em semitransparent blue}) and $t_\mathrm{accel}/t_\mathrm{cs}$ ({\em yellow}) for
504: %simulation 915h at $t=10t_\mathrm{orb}$. The density isosurface is
505: %the same as in Fig. \ref{fig:rho_stream}. The isosurface of $t_\mathrm{accel}/t_\mathrm{cs} = 1$
506: %shows the acceleration zone where gas is brought up to supersonic velocities prior to passing through the standing shock. \label{fig:taccel}}
507: %\end{figure}
508: %\clearpage
509:
510: %The second term in equation (\ref{eqn:LT}) also has another effect: it offers additional vertical support to a disk of finite thickness. The angular dependence of this term means that a tilted disk gets the least support when it passes through the black-hole symmetry plane, which occurs along the line-of-nodes. To maintain vertical equilibrium then, gas or magnetic pressure must compensate for the loss of gravitomagnetic support in this region. We can see the extra gas pressure clearly aligned with the line-of-nodes in the isobar plot of Figure \ref{fig:rho_P}.
511:
512: %\clearpage
513: %\begin{figure}
514: %\plotone{torus3d.m.915h_rho_P_10.eps} \caption{Isosurface plots of
515: %density ({\em semitransparent blue}) and pressure ({\em green}) for
516: %simulation 915h at $t=10t_\mathrm{orb}$. The density isosurface is
517: %the same as in Fig. \ref{fig:rho_stream}. The pressure isosurface
518: %shows a region of enhanced pressure nearly aligned with the line-of-nodes
519: %and centered near the midplane of the disk. \label{fig:rho_P}}
520: %\end{figure}
521: %\clearpage
522:
523: %For half an orbit
524: %this is the material on ``top'' of the disk; for the other half of
525: %the orbit it is the material on the ``bottom''. After passing
526: %through the line-of-nodes the same high-latitude material goes from
527: %being furthest from the black-hole symmetry plane to being closest
528: %and therefore precessing the slowest. It is this transition from
529: %fast precession to slow leads to a compression of the fluid as it
530: %passes through the line-of-nodes. Close to the black hole the
531: %compression is strong enough to develop into a shock. Again, the
532: %compression only happens once per orbit for gas near the ``top'' of
533: %the disk and once per orbit for gas on the ``bottom'', only
534: %occurring when the gas goes from fast precession to slow precession.
535: %Those transitions happen $180^\circ$ out of phase, explaining why we
536: %see a single shock above the disk along one side of the
537: %line-of-nodes and below the disk along the other side.
538:
539: Based on our physical understanding of the plunging streams identified in Paper I and the standing shocks identified in this paper, we expect both features to maintain a constant orientation vis a vis the line-of-nodes between the disk midplane and black-hole symmetry plane. To confirm this we have evolved our ``medium''
540: resolution simulation (model 915m of Paper I) for
541: $100t_\mathrm{orb}$, which is more than a full precession period.
542: Figure \ref{fig:915m_shock} shows the plunging streams and standing
543: shocks at 0.1, 0.35, 0.6, and 1.1 precession periods. The
544: orientations of both features closely track the precession of the
545: disk and maintain roughly constant alignments relative to the line-of-nodes.
546: %This is to
547: %be expected since the black-hole spacetime itself is axisymmetric;
548: %any non-axisymmetric features, such as the plunging streams and
549: %standing shocks, must arise from the interplay between the black
550: %hole and the tilted disk. As the disk precesses, the line-of-nodes necessarily precesses with it.
551:
552: %\clearpage
553: \begin{figure}
554: \begin{center}
555: %\plotone{torus3d.m.915m_rho_curlV.eps}
556: \includegraphics[scale=.298]{f9.jpg} \caption{Same as Fig.
557: \ref{fig:rho_curlV} except at times $t=10$ (panel {\em a}), 32
558: (panel {\em b}), 50 (panel {\em c}), and $100t_\mathrm{orb}$ (panel
559: {\em d}) (representing approximately 0.1, 0.35, 0.6, and 1.1
560: precession period) from simulation 915m. Also, the isodensity
561: surfaces are at $\rho=0.04$, 0.01, 0.02, and
562: $0.02\rho_\mathrm{max,0}$, respectively, instead of
563: $0.1\rho_\mathrm{max,0}$ as in previous figures. The arrows are
564: included to give a visual reference of the precession of the
565: line-of-nodes. By noting which shock is above the black-hole
566: symmetry plane and which is below we see that: the shocks in panels
567: {\em a} and {\em b}, {\em a} and {\em c}, and {\em a} and {\em d}
568: are $90^\circ$, $180^\circ$, and $360^\circ$ out of phase,
569: respectively. It is apparent that both the plunging stream and shock
570: precess with the disk. \label{fig:915m_shock}}
571: \end{center}
572: \end{figure}
573: %\clearpage
574:
575: \subsection{Shock Effects}
576: The presence of non-axisymmetric standing shocks in the accretion
577: disk can have important consequences, principally including enhanced
578: angular momentum transport and dissipation. To
579: illustrate the former, in Figure \ref{fig:ang_mom} we plot angular
580: momentum residual weighted by the circular orbit value, i.e. we plot
581: $(\langle\langle\ell\rangle_A\rangle_t -
582: \ell_\mathrm{cir})/\ell_\mathrm{cir}$ as a function of radius, where
583: $\ell_\mathrm{cir}$ is the angular momentum of circular orbits with
584: inclinations of $15^\circ$ and $0^\circ$ for simulations 915h and 90h, respectively, calculated
585: from the following expression (equation 26 of Paper I)
586: %
587: \begin{equation}
588: \ell = \frac{N_1 + \Delta (Mr)^{1/2} N_2^{1/2} \cos i}{D} ~,
589: \end{equation}
590: %
591: where
592: %
593: \begin{equation}
594: N_1 = -aMr \left(3r^2 + a^2 - 4Mr \right) \cos^2 i ~,
595: \end{equation}
596: %
597: \begin{equation}
598: N_2 = r^4 + a^2 \sin^2 i \left(a^2 + 2r^2 - 4Mr \right) ~,
599: \end{equation}
600: %
601: \begin{equation}
602: D = a^2 \left( 2r^2 + a^2 - 3Mr \right) \sin^2 i + r^4 + 4M^2r^2 -
603: 4r^3M - Mra^2 ~,
604: \end{equation}
605: %
606: and
607: %
608: \begin{equation}
609: \Delta=r^2-2Mr+a^2 ~;
610: \end{equation}
611: %
612: and $\langle\ell\rangle_A =
613: \langle\rho\ell\rangle_A/\langle\rho\rangle_A$ is the
614: density-weighted shell average of the specific angular momentum.
615: Shell averaged quantities are computed as:
616: %
617: \begin{equation}
618: \langle\mathcal{Q}\rangle_A(r,t) = \frac{1}{A} \int^{2\pi}_0
619: \int^{\vartheta_2}_{\vartheta_1} \mathcal{Q} \sqrt{-g}
620: \mathrm{d}\vartheta \mathrm{d}\varphi ~,
621: \end{equation}
622: %
623: where $A = \int^{2\pi}_0 \int^{\vartheta_2}_{\vartheta_1} \sqrt{-g}
624: \mathrm{d}\vartheta \mathrm{d}\varphi$ is the surface area of the
625: shell. The data have also been time-averaged over the interval,
626: $7t_{\rm orb} = t_{\rm min} \le t \le t_{\rm max} = 10t_{\rm orb}$,
627: where time averages are defined as
628: %
629: \begin{equation}
630: \langle\mathcal{Q}\rangle_t = \frac{1}{t_{\rm max} - t_{\rm min}}
631: \int^{t_{\rm max}}_{t_{\rm min}} \mathcal{Q} \mathrm{d}t ~.
632: \end{equation}
633: %
634: The sharp down-turn of the specific angular momentum inside
635: $r\lesssim 10 r_G$ in simulation 915h is indicative of extra angular
636: momentum being removed from the flow. This suggests that the
637: standing shock plays a significant role in the transport of angular
638: momentum in the tilted disk.
639:
640: %\clearpage
641: \begin{figure}
642: %\plotone{Ell_res.eps}
643: \plotone{f10.eps} \caption{Plot of the weighted residual
644: specific angular momentum $(\langle\langle\ell\rangle_A\rangle_t -
645: \ell_\mathrm{cir})/\ell_\mathrm{cir}$ as a function of radius for
646: simulations 915h ({\em solid line}) and 90h ({\em dashed line}). The
647: data has been time-averaged over the interval $t=7$ to $10t_{\rm
648: orb}$. The residuals are calculated from the specific angular
649: momenta of circular orbits with inclinations of $15^\circ$ and
650: $0^\circ$, respectively. \label{fig:ang_mom}}
651: \end{figure}
652: %\clearpage
653:
654: The shock also enhances the energy dissipation in the disk. This is
655: shown in Figure \ref{fig:entropy}, where we plot the
656: density-weighted shell averages of the fluid entropy in the tilted
657: and untilted disks. Actually, since we are only interested in
658: relative changes in entropy, for simplicity we plot $\mathcal{S} =
659: \ln (P/\rho^\Gamma)$. The
660: significant enhancement in entropy generation inside $r \lesssim 10
661: r_G$ for the tilted disk indicates the extra dissipation provided by the shock. This up-turn nicely coincides with the down-turn in Figure \ref{fig:ang_mom}, further supporting the association of both effects with the shock.
662:
663: %\clearpage
664: \begin{figure}
665: %\plotone{Entropy_rad.eps}
666: \plotone{f11.eps} \caption{Plot of the density-weighted
667: shell average of the fluid entropy as a function of radius for
668: simulations 915h ({\em solid line}) and 90h ({\em dashed line}). The
669: data has been time-averaged over the interval $t=7$ to $10t_{\rm
670: orb}$. \label{fig:entropy}}
671: \end{figure}
672: %\clearpage
673:
674: %\clearpage
675: \begin{figure*}[htbp]
676: %\plottwo{velocity_915h.eps}{velocity_90h.eps}
677: \plottwo{f12a.eps}{f12b.eps} \caption{Plot of
678: characteristic velocities as functions of radius within the tilted
679: ({\em left panel}) and untilted ({\em right panel}) disks. All data
680: have been time-averaged over the interval $t=7$ to $10t_{\rm orb}$.
681: Of particular importance is the sharp upturn in $V^r$, indicating
682: the start of the plunging region, which happens at a larger radius
683: for the tilted simulation. \label{fig:velocities}}
684: \end{figure*}
685: %\clearpage
686:
687: The extra energy dissipation and angular momentum transport
688: significantly affect the inner
689: regions of the disk. In Figure \ref{fig:velocities}, we see that the
690: radial plunging region, defined by the sharp upturn in
691: $\overline{V}^r$, begins at a considerably larger radius for the
692: tilted simulation (915h) than for the untilted (90h). This can have
693: important implications for disk observations \citep{kro02} since most of the radiated energy from a disk comes from just outside the plunging radius. Of special concern is
694: the common use of the inner edge of the disk as a direct indicator
695: of the spin of the black hole. Clearly tilt must be taken into
696: account if one wants to relate the plunging radius of the disk to
697: the spin of the black hole for tilted disks. Figure
698: \ref{fig:velocities} also compares other characteristic velocities
699: associated with the disks. All of the velocities are density
700: weighted shell averages $\overline{V}=\langle \langle \rho V
701: \rangle_A/\langle \rho \rangle_A \rangle_t$. The local sound speed
702: is recovered from $c_s^2=\Gamma (\Gamma-1)P/[(\Gamma-1)\rho + \Gamma
703: P]$. The Alfv\'en speed is
704: %
705: \begin{equation}
706: v_A = \sqrt{\frac{\vert\vert B \vert\vert^2}{4 \pi \rho h +
707: \vert\vert B \vert\vert^2}} ~.
708: \end{equation}
709: %
710: We approximate the turbulent velocity as $V_\mathrm{turb}\approx
711: \alpha^{1/2}c_s$, where
712: \begin{equation}
713: \alpha = \left\langle \frac{ \vert u^r u^\varphi \vert\vert B
714: \vert\vert^2 - B^r B^\varphi \vert}{4 \pi P} \right\rangle_A
715: \end{equation}
716: is the dimensionless stress parameter. The most notable difference
717: is in the radial inflow velocity $\overline{V}^r$, although there is also some indication that the turbulent stress rises faster in the interior of the tilted disk.
718:
719: %EXTRA DISSIPATION FROM SHOCKS? PERHAPS TRY PLOTTING INTEGRATED
720: %ENERGY PER RADIAL SHELL AND LOOK FOR DROP INSIDE THE SHOCK?
721: %PLOT OF TOTAL KINETIC AND MAGNETIC ENERGIES AS
722: %A FUNCTION OF TIME FOR TILTED AND UNTILTED SIMS. SEE IF THESE ARE
723: %LOWER FOR THE TILTED SIM THAN UNTILTED?
724: %There is considerably less kinetic and internal energy on
725: %the grid in the tilted sim than the untilted. However, because
726: %there is also considerably less mass left on the grid, it is not
727: %possible to definitively attribute this to extra dissipation.
728:
729:
730: \subsection{Post-Shock Magnetic Field}
731: We found previously that magnetic fields remain subthermal
732: everywhere in the disk throughout the simulation (see Figure 11 of
733: Paper I). They, therefore, must not play a significant role in the
734: dynamics of the epicyclic motion, plunging streams, or standing shocks.
735: This statement is further supported by the fact that very similar
736: features were seen in earlier {\em hydrodynamic} simulations of
737: tilted black-hole accretion disks \citep{fra05b}. Despite the minor
738: role of the magnetic fields in the dynamics, the presence of a
739: standing shock can enhance the strength of the magnetic field. From
740: \citet{fra05a} we note that the post-shock value of
741: $\beta_\mathrm{mag}$ in the Newtonian limit can depend sensitively on the strength of the
742: shock
743: %
744: \begin{equation}
745: \beta_\mathrm{mag,ps} = 2\Gamma(\Gamma-1)^2/(\Gamma+1)^3
746: \mathcal{M}^2 \beta_{\mathrm{mag},i}
747: \end{equation}
748: where $\beta_{\mathrm{mag},i}$ is the pre-shock value and we assume the field is oriented perpendicular to the shock normal. For
749: $\Gamma=5/3$, as in our work, this gives
750: %
751: \begin{equation}
752: \beta_\mathrm{mag,ps} = 0.078 \mathcal{M}^2 \beta_{\mathrm{mag},i}
753: ~.
754: \end{equation}
755: Therefore, for $\mathcal{M} < 3.6$, the post-shock
756: $\beta_\mathrm{mag}$ will be lower than the pre-shock value. In
757: Figure \ref{fig:beta} we show that there is, indeed, a thin layer of
758: magnetically dominated plasma just behind the shock. To get
759: $\beta_\mathrm{mag,ps}=0.1$ as Figure \ref{fig:beta} starting from
760: $\beta_{\mathrm{mag},i}=1$ (a reasonable guess for the high latitude
761: material), we need $\mathcal{M}=1.13$, which is consistent with what
762: we see in the simulation. Again, these do not appear to be
763: tremendously strong shocks.
764:
765: %\clearpage
766: \begin{figure}[b]
767: \begin{center}
768: %\plotone{torus3d.m.915h_rho_invBeta_10.eps}
769: \includegraphics[scale=.298]{f13.jpg} \caption{Isosurface plot
770: of density ({\em semitransparent blue}) plus an isosurface of $\beta_\mathrm{mag}=P/P_B =
771: 0.1$ ({\em orange}), indicating a thin layer of magnetically dominated plasma behind
772: the shock. The density isosurface is the same as in Fig. The figure is oriented
773: looking directly down the spin axis of the hole.
774: \ref{fig:rho_stream}. \label{fig:beta}}
775: \end{center}
776: \end{figure}
777: %\clearpage
778:
779: \section{Discussion}
780: \label{sec:discussion}
781:
782: In this paper we have compared two simulations that are identical to
783: one another in every respect except one: the initial tilt between
784: the black-hole and disk angular momenta. Despite the apparent
785: similarity, we observed remarkable differences in the evolution of
786: these two simulations.
787:
788: The primary new feature that we describe in the main disk body is a
789: strong epicyclic driving attributable to the gravitomagnetic torque
790: of the misaligned (tilted) black hole. The induced motion of the gas
791: is coherent over the scale of the entire disk. An interesting point
792: about this epicyclic motion that has not been made before is that it
793: could be detectable, for instance in the profile of relativistically
794: broadened iron lines. We have previously pointed out the importance
795: of the iron lines in directly probing tilted accretion disks
796: \citep{fra05c}, but at the time we were not in a position to
797: recognize the importance of the epicyclic motion. From Figure
798: \ref{fig:rhoVph1} we get that the velocities associated with the
799: epicyclic motion represent a significant fraction ($\lesssim40\%$)
800: of the orbital velocity of the gas. The corresponding shift in a
801: reflection feature such as the iron line should be of a similar
802: magnitude. The interesting thing is to note that an observer viewing
803: the disk from the vantage point of Figure \ref{fig:rhoVph1}{\em a}
804: and seeing the ``top'' of the disk would see a smaller than expected
805: red shift (the gas going away is not moving as fast as expected) and
806: a larger than expected blue shift (the approaching gas is moving
807: faster than expected). An observer viewing the same disk from the
808: same vantage point but seeing the ``bottom'' of the disk would see
809: exactly the opposite shift. Therefore, depending on the viewing
810: angle, the entire line profile of a misaligned disk could be shifted
811: toward the blue {\em or} the red relative to an aligned disk. On the
812: redshifted side this effect might be confused with gravitational
813: redshifting, making it appear that the line is coming from deeper in
814: the potential well than is actually the case.
815:
816: If the disk actually precesses then there is a clear and simple way
817: to disentangle this effect because the epicyclic motion is phased
818: with the orientation of the disk, as discussed in \S
819: \ref{sec:bending} above. This means that the line shift we are
820: describing would reverse itself twice per precession period,
821: appearing for half a precession period as an overall blueshift and
822: for the other half as an overall redshift.
823: If the integration time of the detector is longer than the
824: precession period, the net effect would generally be to smear
825: or broaden the line. This may provide an alternative explanation for an exceptionally broad line profile such as GX 339-4 \citep{mil04}, without requiring a high black-hole spin or small disk radius. On the other hand, if the integration time is
826: shorter than the precession period, then one would expect the iron line to vary in phase with changes in the X-ray flux (i.e. in conjunction with a low frequency QPO that would correspond to the precession frequency of the disk). Such behavior has been identified in at least one source, GRS 1915+105 \citep{mil05}.
827: %say for a large disk with a very
828: %long precession time, then one must be careful not to misinterpret
829: %this effect. Nevertheless, we again find the iron line could be a
830: %powerful diagnostic of tilted disks, in this case to directly
831: %measure the speed of the epicyclic motion.
832:
833: The second new feature of our tilted disk simulation that we
834: described is a pair of standing shocks, roughly aligned with the
835: line-of-nodes between the disk and black hole symmetry planes.
836: %The
837: %presence of a standing shock indicates a supersonic compression of
838: %the gas. We have already shown in Figure \ref{fig:rho_P} that there
839: %is indeed a compression of the fluid as it passes near the
840: %line-of-nodes. We have suggested that this compression is
841: %attributable to the latitude dependence of the Lense-Thirring
842: %precession. In modestly thick tilted accretion disks, such as the
843: %one we simulated, high-latitude gas precesses fastest when it is
844: %furthest from the black-hole symmetry plane and slowest when it is
845: %closest. As gas orbits past the line-of-nodes, on one passing it
846: %will transition from slow to fast precession; on the other passing
847: %it will transition from fast to slow. It is this second transition
848: %that results in a compression. Close to the black hole the
849: %compression is strong enough to become a shock.
850: This asymmetric shock provides additional angular momentum transport and energy dissipation in the tilted disk, relative to the untilted one. This enhanced dissipation may help compensate for the loss of radiative efficiency (relative to an untilted disk) due to the plunging region
851: starting further out than the equatorial ISCO radius of the spinning black hole.
852: In collisionless accretion flows, such as those thought to be relevant for
853: Sgr~A* \citep[e.g.][]{qua03}, it is conceivable that such shocks could
854: also be sites of particle acceleration. The precession of these shocks
855: might then result in periodic variations of nonthermal radiation.
856:
857: In considering the magnetic fields in the inner part of the tilted
858: disk, as noted in Paper I, the magnetic fields remain largely
859: subthermal. They do not play an important role in the physics of the
860: epicyclic motion, plunging streams, or standing shocks, although we do
861: find some enhancement of $\beta_\mathrm{mag}$ associated with the
862: shocks.
863:
864: We finish with a figure (\ref{fig:schematic}) which provides a
865: visual summary of our findings. We note the overall consistency of
866: our interpretation of the results: 1) The observed epicyclic motion
867: is in agreement with expectations for warped disks in the wave-like propagation limit; and 2) The shocks are located near
868: the apocenters of the epicyclic motion, as one expects when the eccentricity increases with decreasing radius.
869:
870: %\clearpage
871: \begin{figure}
872: %\plotone{schematic.eps}
873: \plotone{f14.eps} \caption{Schematic diagram of the inner
874: region of the tilted accretion disk, showing the pattern of
875: epicyclic motion, the standing shock, and the plunging streams.
876: \label{fig:schematic}}
877: \end{figure}
878: %\clearpage
879:
880: \begin{acknowledgements}
881: We would like to recognize Christopher Lindner and Joseph Niehaus
882: for their contributions to this work. We also thank Steve Balbus
883: and Gordon Ogilvie for very useful discussions and comments and the anonymous referee for suggested improvements.
884: PCF gratefully acknowledges
885: the support of a Faculty R\&D grant from the College of Charleston
886: and a REAP grant from the South Carolina Space Grant Consortium.
887: This work was supported under the following National Science
888: Foundation grants and programs: AST~0707624,
889: Partnerships for Advanced Computational
890: Infrastructure, Distributed Terascale Facility (DTF), and Terascale
891: Extensions: Enhancements to the Extensible Terascale Facility.
892: \end{acknowledgements}
893:
894:
895:
896: \clearpage
897: %\bibliographystyle{apj}
898: %\bibliography{myrefs}
899:
900: \begin{thebibliography}{}
901:
902: \bibitem[\protect\citeauthoryear{{Anninos}, {Fragile}, \&
903: {Salmonson}}{{Anninos} et~al.}{2005}]{ann05}
904: {Anninos}, P., {Fragile}, P.~C., \& {Salmonson}, J.~D. 2005, \apj, 635, 723
905:
906: \bibitem[\protect\citeauthoryear{{Bardeen} \& {Petterson}}{{Bardeen} \&
907: {Petterson}}{1975}]{bar75}
908: {Bardeen}, J.~M., \& {Petterson}, J.~A. 1975, \apjl, 195, L65
909:
910: \bibitem[\protect\citeauthoryear{{Caproni} et~al.}{{Caproni}
911: et~al.}{2007}]{cap07}
912: {Caproni}, A., {Abraham}, Z., {Livio}, M., \& {Mosquera Cuesta}, H.~J. 2007,
913: \mnras, 379, 135
914:
915: \bibitem[\protect\citeauthoryear{{Caproni}, {Abraham}, \& {Mosquera
916: Cuesta}}{{Caproni} et~al.}{2006}]{cap06}
917: {Caproni}, A., {Abraham}, Z., \& {Mosquera Cuesta}, H.~J. 2006, \apj, 638, 120
918:
919: \bibitem[\protect\citeauthoryear{{Chakrabarti}}{{Chakrabarti}}{1985}]{cha85}
920: {Chakrabarti}, S.~K. 1985, \apj, 288, 1
921:
922: \bibitem[\protect\citeauthoryear{{De Villiers} \& {Hawley}}{{De Villiers} \&
923: {Hawley}}{2003}]{dev03b}
924: {De Villiers}, J., \& {Hawley}, J.~F. 2003, \apj, 592, 1060
925:
926: \bibitem[\protect\citeauthoryear{{De Villiers}, {Hawley}, \& {Krolik}}{{De
927: Villiers} et~al.}{2003}]{dev03c}
928: {De Villiers}, J., {Hawley}, J.~F., \& {Krolik}, J.~H. 2003, \apj, 599, 1238
929:
930: \bibitem[\protect\citeauthoryear{{Fragile} \& {Anninos}}{{Fragile} \&
931: {Anninos}}{2005}]{fra05b}
932: {Fragile}, P.~C., \& {Anninos}, P. 2005, \apj, 623, 347
933:
934: \bibitem[\protect\citeauthoryear{{Fragile} et~al.}{{Fragile}
935: et~al.}{2007a}]{fra07a}
936: {Fragile}, P.~C., {Anninos}, P., {Blaes}, O.~M., \& {Salmonson}, J.~D. 2007a,
937: in proceedings of the 11th Marcel Grossmann Meeting on General Relativity
938: (astro-ph/0701272)
939:
940: \bibitem[\protect\citeauthoryear{{Fragile} et~al.}{{Fragile}
941: et~al.}{2005}]{fra05a}
942: {Fragile}, P.~C., {Anninos}, P., {Gustafson}, K., \& {Murray}, S.~D. 2005,
943: \apj, 619, 327
944:
945: \bibitem[\protect\citeauthoryear{{Fragile} et~al.}{{Fragile}
946: et~al.}{2007b}]{fra07b}
947: {Fragile}, P.~C., {Blaes}, O.~M., {Anninos}, P., \& {Salmonson}, J.~D. 2007b,
948: \apj, 668, 417
949:
950: \bibitem[\protect\citeauthoryear{{Fragile}, {Mathews}, \& {Wilson}}{{Fragile}
951: et~al.}{2001}]{fra01a}
952: {Fragile}, P.~C., {Mathews}, G.~J., \& {Wilson}, J.~R. 2001, \apj, 553, 955
953:
954: \bibitem[\protect\citeauthoryear{{Fragile}, {Miller}, \&
955: {Vandernoot}}{{Fragile} et~al.}{2005}]{fra05c}
956: {Fragile}, P.~C., {Miller}, W.~A., \& {Vandernoot}, E. 2005, \apj, 635, 157
957:
958: \bibitem[\protect\citeauthoryear{{Hawley}}{{Hawley}}{2000}]{haw00}
959: {Hawley}, J.~F. 2000, \apj, 528, 462
960:
961: \bibitem[\protect\citeauthoryear{Ivanov \& Illarionov}{Ivanov \&
962: Illarionov}{1997}]{iva97}
963: Ivanov, P.~B., \& Illarionov, A.~F. 1997, \mnras, 285, 394
964:
965: \bibitem[\protect\citeauthoryear{{Kondratko}, {Greenhill}, \&
966: {Moran}}{{Kondratko} et~al.}{2005}]{kon05}
967: {Kondratko}, P.~T., {Greenhill}, L.~J., \& {Moran}, J.~M. 2005, \apj, 618, 618
968:
969: \bibitem[\protect\citeauthoryear{Krolik \& Hawley}{Krolik \&
970: Hawley}{2002}]{kro02}
971: Krolik, J.~H., \& Hawley, J.~F. 2002, \apj, 573, 754
972:
973: \bibitem[\protect\citeauthoryear{{Kumar} \& {Pringle}}{{Kumar} \&
974: {Pringle}}{1985}]{kum85}
975: {Kumar}, S., \& {Pringle}, J.~E. 1985, \mnras, 213, 435
976:
977: \bibitem[\protect\citeauthoryear{{Maccarone}}{{Maccarone}}{2002}]{mac02}
978: {Maccarone}, T.~J. 2002, \mnras, 336, 1371
979:
980: \bibitem[\protect\citeauthoryear{{Miller} \& {Homan}}{{Miller} \&
981: {Homan}}{2005}]{mil05}
982: {Miller}, J.~M., \& {Homan}, J. 2005, \apjl, 618, L107
983:
984: \bibitem[\protect\citeauthoryear{{Miller} et~al.}{{Miller}
985: et~al.}{2004}]{mil04}
986: {Miller}, J.~M., et~al. 2004, \apj, 601, 450
987:
988: \bibitem[\protect\citeauthoryear{{Natarajan} \& {Armitage}}{{Natarajan} \&
989: {Armitage}}{1999}]{nat99}
990: {Natarajan}, P., \& {Armitage}, P.~J. 1999, \mnras, 309, 961
991:
992: \bibitem[\protect\citeauthoryear{Nelson \& Papaloizou}{Nelson \&
993: Papaloizou}{1999}]{nel99}
994: Nelson, R.~P., \& Papaloizou, J. C.~B. 1999, \mnras, 309, 929
995:
996: \bibitem[\protect\citeauthoryear{{Papaloizou} \& {Lin}}{{Papaloizou} \&
997: {Lin}}{1995}]{pap95a}
998: {Papaloizou}, J.~C.~B., \& {Lin}, D.~N.~C. 1995, \apj, 438, 841
999:
1000: \bibitem[\protect\citeauthoryear{{Quataert}}{{Quataert}}{2003}]{qua03}
1001: {Quataert}, E. 2003, Astronomische Nachrichten Supplement, 324, 435
1002:
1003: \bibitem[\protect\citeauthoryear{{Shakura} \& {Sunyaev}}{{Shakura} \&
1004: {Sunyaev}}{1973}]{sha73}
1005: {Shakura}, N.~I., \& {Sunyaev}, R.~A. 1973, \aap, 24, 337
1006:
1007: \bibitem[\protect\citeauthoryear{{Torkelsson} et~al.}{{Torkelsson}
1008: et~al.}{2000}]{tor00}
1009: {Torkelsson}, U., {Ogilvie}, G.~I., {Brandenburg}, A., {Pringle}, J.~E.,
1010: {Nordlund}, {\AA}., \& {Stein}, R.~F. 2000, \mnras, 318, 47
1011:
1012: \end{thebibliography}
1013:
1014:
1015:
1016: \end{document}
1017: