1: \documentclass[twocolumn,showpacs,preprintnumbers,amsmath,amssymb]{revtex4}
2:
3: \usepackage{dcolumn}
4: \usepackage{bm}
5:
6: \usepackage{amssymb}
7: \usepackage{epsfig}
8: \usepackage{graphicx}
9:
10: \newcommand{\be}{\begin{equation}}
11: \newcommand{\ee}{\end{equation}}
12:
13: \newcommand{\bn}{\bar{\nabla}}
14:
15: \newcommand{\ba}{\begin{eqnarray}}
16: \newcommand{\ea}{\end{eqnarray}}
17: \newcommand{\nn}{\nonumber\\}
18: \newcommand{\n}[1]{\label{#1}}
19: \newcommand{\bea}{\begin{eqnarray}}
20: \newcommand{\eea}{\end{eqnarray}}
21:
22:
23: \newcommand{\BM}[1]{{\mbox{\boldmath $#1$}}}
24: \newcommand{\BF}[1]{\mbox{\bf #1}}
25: \newcommand{\BFE}[1]{\mbox{\bf e}_{\hat{#1}}}
26: \newcommand{\hn}{\hat{\nabla}}
27:
28:
29: \newcommand{\ind}[1]{\mbox{\tiny #1}}
30:
31: \newcommand{\hh}{\, ,\hspace{0.5cm}}
32: \newcommand{\hhh}{\, ,\hspace{0.2cm}}
33:
34:
35:
36: \def\({\left(} \def\){\right)}
37:
38: \newcommand{\bd}{\begin{displaymath}}
39: \newcommand{\ed}{\end{displaymath}}
40: \newcommand{\ds}{\displaystyle}
41:
42:
43:
44:
45: \newcommand{\non}{\nonumber}
46: \newcommand{\eq}[1]{(\ref{#1})}
47:
48: \newcommand{\bBox}{\bar{\phantom{!}\Box\phantom{!}}}
49: \newcommand{\tBox}{\tilde{\phantom{!}\Box\phantom{!}}}
50: \newcommand{\Tr}{\phantom{!}\mbox{\bf Tr}\phantom{!}}
51: \newcommand{\tr}{{\rm tr}}
52:
53: \newcommand{\e}{{\rm e}}
54: \newcommand{\x}{{\mathbf x}}
55: \newcommand{\y}{{\mathbf y}}
56: \newcommand{\U}{{\cal U}}
57: \newcommand{\J}{{\cal J}}
58: \newcommand{\E}{{\mathbb E}}
59:
60:
61: \newcommand{\pa}{\partial}
62:
63: \newcommand{\lap}{\bigtriangleup}
64:
65:
66: \newcommand{\pr}[5] {\bibitem{#1} {#2}, Phys. Rev. {\bf D#3}, #4 (#5).}
67: \newcommand{\cqg}[5] {\bibitem{#1} {#2}, Class. Quant. Grav. {\bf #3}, #4 (#5).}
68: \newcommand{\PRL}[5] {\bibitem{#1} {#2}, Phys. Rev. Lett. {\bf #3}, #4 (#5).}
69:
70: \topmargin -1cm
71:
72: \begin{document}
73:
74: \title{A Toy Model for Topology Change Transitions: Role of Curvature
75: Corrections}
76:
77: \author{Valeri P. Frolov}
78:
79: \email{frolov@phys.ualberta.ca}
80:
81: \affiliation{Theoretical Physics Institute,\\ University of Alberta,
82: Edmonton, Alberta, Canada T6G 2G7}
83:
84: \author{Dan Gorbonos}
85:
86: \email{gorbonos@phys.ualberta.ca}
87:
88: \affiliation{Theoretical Physics Institute,\\ University of Alberta,
89: Edmonton, Alberta, Canada T6G 2G7}
90:
91:
92:
93: \begin{abstract}
94: We consider properties of near-critical solutions describing a test
95: static axisymmetric $D$-dimensional brane interacting with a bulk
96: $N$-dimensional black hole $(N>D)$. We focus our attention on the
97: effects connected with curvature corrections to the brane action.
98: Namely, we demonstrate that the second order phase transition in
99: such a system is modified and becomes first order. We discuss
100: possible consequences of these results for merger transitions
101: between caged black holes and black strings.
102: \end{abstract}
103:
104:
105:
106: \pacs{04.70.Bw, 04.50.-h} \ \ \
107:
108: \maketitle
109:
110: \section{Introduction}
111:
112:
113: Transitions with a change of Euclidean topology is a subject of wide
114: physical interest. One interesting example is the phase transition
115: connected with a nucleation of a black hole in a thermal bath.
116: Consider a thermal field with temperature $T$ in a flat spacetime.
117: One can use Euclidean fields on a spacetime with the topology
118: $R^{D-1}\times S^1$ to describe a canonical ensemble for such a
119: field. The size of the compact dimension $S^1$ is $\beta=1/T$. A
120: nucleation of a black hole changes the Euclidean topology from
121: $R^{D-1}\times S^1$ to $S^{D-2}\times R^2$. The corresponding
122: Euclidean space after the black hole nucleation is the
123: Gibbons-Hawking instanton~\cite{GH}.
124:
125: Another important example of a similar phenomenon is the so-called
126: merger phase transition which occurs in models with large extra
127: dimensions when a black hole is localized in a spacetime which has
128: additional $k$ compact dimensions ($D=4+k$) (a caged black hole, for
129: reviews see~\cite{review,niarchos,reviewob1,reviewob2,reviewob3}).
130: In the absence of the black hole such a spacetime has the topology
131: $R^4\times T^k$. Kol argued~\cite{conifold-barak} that the black
132: hole-black string phase transition includes a local topology change
133: of the corresponding Euclidean manifold so that the singular
134: geometry is a cone over $S^{D-3}\times S^{2}$~\cite{endnote1}. This
135: topology change is similar to the conifold
136: transition~\cite{conifolds}. In the black hole phase the $S^{D-3}$
137: is contractible while in the black string phase the $S^{2}$ is
138: contractible. In order to achieve this topology change one has to
139: pass a configuration which is singular at the tip of the cone. The
140: ``double-cone'' over $S^{D-3}\times S^{2}$ is given by \be\n{dc}
141: ds^{2}=d\rho^{2}+\frac{\rho^{2}}{D-2}\,\left[d\chi^{2}
142: +\cos(\chi)^{2}\,dt^{2}+(D-4)\,d\,\Omega_{D-3}^{2}\right]. \ee For
143: more details see~\cite{review,Barak2}.
144:
145:
146: Kol~\cite{Barak1} proposed that there exists a relation between
147: merger transitions and Choptuik's critical collapse
148: ~\cite{choptuik-original,chptuik-review}. This correspondence can be
149: achieved by performing two analytic continuations. The physics of
150: the critical collapse and merger transitions have some common
151: features like a singular critical solution which turns out to be an
152: attractor and a self-similar solution in the neighborhood of the
153: singular point. A better understanding of one of the systems may
154: shed light on the other.
155:
156: It is interesting that there also exists a close similarity between
157: the properties of {\em merger transitions} and a toy model
158: proposed some time ago for study transitions during which the
159: Euclidean topology is changed~\cite{CFL,FLC,BBH1}. This model
160: consists of an $N$-dimensional static bulk black hole and a
161: $D$-dimensional brane ($D<N$) interacting with this black hole. The
162: brane is assumed to be a test brane and infinitely thin. The former
163: assumption means that one neglects the effects connected with the
164: gravitational field of the brane, while the latter one implies that
165: the effects of the brane thickness are neglected and its worldsheet
166: is a minimal surface which provides an extremum of the
167: Dirac-Nambu-Goto (DNG) action. It is assumed that the brane is
168: static and axisymmetric, so that the induced geometry on the brane
169: possesses the $O(D-1)$ group of isometry. It is also assumed that
170: far from the black hole the brane surface is parallel to the
171: equatorial plane of the bulk black hole. For such a brane there
172: exists two qualitatively different configurations: One, which is
173: called {\em subcritical}, is a brane which does not intersect the
174: black hole event horizon, and the other, {\em supercritical}, is a
175: brane crossing the horizon. In the latter configuration the induced
176: geometry on the world sheet of the brane is the geometry of
177: a $D$-dimensional black hole, which is called a {\em brane black
178: hole}, or briefly {\em BBH}. Such a black hole is absent for a
179: subcritical configuration. Thus by changing the position of the
180: brane at infinity ({\em asymptotic data}), one generates a
181: transition between BBH and no-BBH phases (see Fig. \ref{3con}).
182:
183: \begin{figure}[htb]
184: \epsfxsize=40mm \epsfbox{subc.eps}
185: \epsfxsize=40mm \epsfbox{superc.eps}
186: \epsfxsize=40mm \epsfbox{critical.eps} \caption{Three possible types
187: of configurations - the {\em subcritical} embedding (left) where the
188: brane does not touch the black hole horizon, the {\em supercritical}
189: (right) where we have an induced black hole on the brane (BBH), and
190: the {\em critical} (center) which is singular at its tip.}
191: \label{3con}
192: \end{figure}
193:
194: If this change is done adiabatically, then one deals with a one
195: parameter set of quasistatic solutions. After Wick's rotation of
196: time one gets a one parameter set of Euclidean induced metrics, with
197: a change of the Euclidean topology of the induced metric at some
198: critical value of the asymptotic data, which plays the role of an
199: order parameter. The Euclidean topology changes from $S^1 \times
200: R^{D-1}$ for the subcritical configuration to $R^2 \times S^{D-2}$
201: for the supercritical. It was demonstrated~\cite{CFL,FLC,BBH1} that
202: when the effects of the brane stiffness are neglected the relation
203: between the asymptotic data and the mass of the induced BBH for the
204: transition between sub- and supercritical configurations is {\em
205: universal}, that is it does not depend on the bulk black hole
206: characteristics. Moreover, there is no mass gap for a creation of a
207: BBH, so that the corresponding phase transition is of the second
208: order. The near-critical solutions possess discrete (for $D\le 6$)
209: or continuous (for $D>6$) self-similarity, which makes this
210: transition formally similar both to the merger transitions and the
211: near-critical collapse discovered by Choptuik~
212: \cite{choptuik-original}. (For a general review of the critical
213: collapse see e.g.~\cite{chptuik-review}. See also a
214: discussion~\cite{sorkin} of the critical collapse in a higher
215: dimensional spacetime.). These properties and close similarity of
216: near-critical solutions for both the BBH model and merger
217: transitions make it interesting to consider in the framework of the
218: BBH model some general problems which exist for this class of
219: models.
220:
221: Before discussing these problems we mention that the universality of
222: the near-critical behavior in the BBH model is a consequence of the
223: following fact: only near-horizon properties of these solutions are
224: important . In this near-horizon domain there is no dimensionful
225: parameter which determines the behavior of the system. As a result
226: of this the system has scaling properties and the related phase
227: transition is of the second order. This is why the model captures
228: many universal features of various physical
229: systems~\cite{Filev:2008xt}. One important example of such a system
230: provides a holographic description of the meson melting phase
231: transition of matter in the fundamental
232: representation~\cite{Mateos:2007yp,Mateos:2007vn,Mateos:2006nu,Hoyos:2006gb,Babington:2003vm}.
233: The configuration consists of $N_c$ color Dp-branes and $N_f$ flavor
234: Dq-branes when $p<q$. The addition of the flavor Dq-branes is dual
235: to the addition of matter in the fundamental representation in the
236: gauge theory. In the limit $N_c \gg N_f$ the Dp-branes are described
237: by a p-brane supergravity action (``black Dp-branes'') while the
238: Dq-branes are described by the DNG action. In other words, we can
239: say that there are Dq-branes in the background of Dp-branes. In this
240: system the point, where the brane touches the black hole horizon
241: corresponds to a certain temperature where the mesons
242: melt~\cite{endnote2}. The holographic description of the melting
243: corresponds to the transition from the subcritical embedding to the
244: supercritical one.
245:
246:
247:
248:
249: Let us consider merger transitions in more detail. It is evident
250: that the double-cone solution \eq{dc} is smooth everywhere except
251: for the tip $\rho=0$ where the curvature becomes singular. Near the
252: tip the Kretchsman curvature scalar ${\cal
253: R}^2=R_{\mu\nu\rho\sigma}\,R^{\mu\nu\rho\sigma}$ infinitely grows
254: \be {\cal R}^2=\frac{4\,(D-3)^{2}\,(D-2)}{(D-4)\,\rho^{4}}\, . \ee
255: The existence of the infinite curvature indicates that the solution
256: obtained in the framework of classical Einstein gravity should be
257: modified by quantum corrections. In other words, the naked
258: singularity that is formed during the merger transition in its
259: classical description might be resolved by the inclusion of quantum
260: corrections into the classical action~\cite{endnote25}. This
261: conclusion is important. It means that if the transition between a
262: black hole and black string phases occurs through the merger
263: transition, one can expect the formation of a region with very high
264: (up to the Planckian) curvature in a system characterized by
265: macroscopic parameters (size of extra dimensions). An important
266: question is how quantum gravity effects modify an adopted picture of
267: classical merger transition.
268:
269: Trying to answer this question one inevitably meets two difficult
270: problems. One is of technical origin, namely, how the near-critical
271: solutions for the merger problem are modified by quantum gravity
272: corrections, for example, by adding to the Einstein-Hilbert action
273: quadratic curvature corrections which arise in one loop
274: computations. One can expect that the corresponding corrections
275: become important when the curvature near the tip reaches the
276: Planckian scale. The other more difficult problem is the following.
277: At the corresponding Planckian scale the higher loop quantum gravity
278: terms might also become important. If this happens, it would
279: indicate that a complete solution of the problem requires the
280: summation of all quantum loops or the use of a more fundamental
281: theory of gravity, such as string theory. All of the above makes the
282: problem very complicated for analysis.
283:
284: For this reason it is interesting to analyze a much simpler BBH
285: model which has qualitatively the same behavior as merger
286: transitions. Its critical solutions also have
287: curvature singularity
288: at the cone tip where it touches the horizon. One can expect that
289: adding terms quadratic in the extrinsic curvature to the classical DNG
290: Lagrangian, which are analogous to local one loop corrections in
291: quantum gravity, may ``cure'' this ``disease.'' Such curvature
292: corrections naturally arise as a result of the stiffness
293: effect~\cite{endnote3}. In the case of strings they were suggested
294: by Polyakov~\cite{polyakov}. We can think about such terms as
295: corrections that come from the finite thickness of the brane which
296: is ignored in the DNG action. The DNG action can be considered as
297: the zeroth order in the expansion in the width over a typical length
298: in the system~\cite{VilShel,CarGre}. Usually the small parameter in
299: such an expansion is the ratio of the thickness of the brane to the
300: characteristic radius of the brane bending. It is instructive first
301: to study effects connected with the leading order terms in this
302: expansion. In this analogy the thickness of the brane plays the role
303: similar to the Planck scale. Moreover, if one describes the brane as
304: a special topologically stable solution of some nonlinear field
305: theory, one may, in principle, answer not only the question of how
306: the quadratic curvature corrections modify the near-critical
307: solutions, but also investigate the complete field theoretical
308: object behavior in the near-critical regime. Effectively this
309: corresponds to the summation of all the stiffness corrections.
310:
311: In the present paper we focus on the first problem, namely we will
312: analyze how the lowest order stiffness corrections modify the phase
313: transition in the BBH systems.
314:
315:
316: A good analogy that helps to understand the effect of stiffness
317: terms is a stiff bar. Consider the bending of a stiff bar. When we
318: take into account the effect of the stiffness of the bar, its
319: bending costs energy. Hence we cannot bend the bar as much as we
320: want and it would eventually break long before a sharp tip is
321: created. The sharp tip corresponds to the singular critical solution
322: of the DNG action. One might expect that inclusion of higher
323: derivative terms to the Lagrangian would prevent the creation of
324: such a singular solution and will form a first order phase
325: transition long before. Indeed, as we will see this is what happens
326: in the BBH system for the subcritical configuration.
327:
328: The higher derivative corrections to the BBH system can serve as a
329: toy model for the singularity resolution of ``small BHs.'' Small BHs
330: are singular limits of BH parameters in which the horizon becomes
331: singular (see for example~\cite{Sen1,Sen2}). If we look at the
332: induced BH on the brane, the critical solution has
333: a singularity exactly of this type.
334:
335: The paper is organized as follows. In Sec.~{\it II} we review the
336: main results concerning the near-critical branes obtained in the
337: absence of stiffness in the BBH model. In Sec.~{\it III} we discuss
338: the curvature corrections for a stiff brane. The stiff brane
339: equations are presented in Sec.~{\it IV}. Section~{\it V} contains
340: the analysis of near-critical solutions in the linear approximation.
341: In Secs.~{\it VI} and~{\it VII} the numerical results for
342: near-critical branes are presented. Section~{\it VIII} contains a
343: summary of the obtained results and their discussion.
344:
345:
346: \section{Nonstiff Branes}
347:
348: In this section we briefly review the main results of~\cite{BBH1}
349: concerning the behavior of near-critical $D$-dimensional branes
350: without stiffness interacting with a bulk static spherically
351: symmetric $N$-dimensional black hole. We do this mainly to explain
352: the set up of the problem and to fix the notations we will use
353: later. The Schwarzschild-Tangherlini metric of the bulk
354: $N$-dimensional spacetime is \be\n{met}
355: dS^2=g_{\mu\nu}dx^{\mu}dx^{\nu}=-F dT^2+F^{-1} dr^2 + r^2
356: d\Omega^2_{N-2}\, , \ee where $F=1-(r_g/r)^{N-3}$ and
357: $d\Omega^2_{N-2}$ is the metric of a $(N-2)$-dimensional unit sphere
358: $S^{N-2}$. We define the coordinates $\theta_i$ ($i=1,\ldots,N-2$)
359: on this sphere by the relations \be d\Omega^2_{i+1}= d\theta^2_{i+1}
360: +\sin^2 \theta_{i+1} d\Omega^2_{i}\, . \ee
361:
362:
363: We denote by $x^\mu$ ($\mu=0,\ldots,N-1$) the bulk spacetime
364: coordinates and by $\zeta^a\;(a=0,\ldots, D-1)$ the coordinates on
365: the brane worldsheet. The functions $x^{\mu}=X^{\mu}(\zeta^a)$
366: determine the brane world sheet describing the embedding of the
367: $(D-1)$-dimensional object (brane) in a bulk $N$-dimensional
368: spacetime. We assume that $D\le N-1$.
369:
370:
371: In the absence of stiffness, the brane configuration in an external
372: gravitational field $g_{\mu\nu}$ can be obtained by solving the
373: equations which follow from the DNG action~\cite{Dirac, Nambu,Goto}
374: \begin{equation}
375: S \: = \int d^D\zeta \sqrt{-\mbox{det}\gamma_{ab}}\, , \label{action}
376: \end{equation}
377: where $\gamma_{ab} \: = \: g_{\mu\nu}X^\mu_{,a} X^\nu_{,b}$ is an
378: induced metric on the brane world sheet. We set the brane tension
379: factor, which does not enter the brane equations, equal to 1. It is
380: well known that an extremum of this action is a minimal surface. Let
381: $n^{\mu}_{(i)}$ be unit normals to the brane, and \be
382: K^{(i)}_{\alpha\beta}=-\frac{\partial X^{\mu}}{\partial
383: \zeta^{\alpha}}\,\frac{\partial X^{\nu}}{\partial
384: \zeta^{\beta}}\,\nabla_{\nu}\,n^{(i)}_{\mu}\, , \label{ext_def}\ee
385: be an extrinsic curvature tensor. ($\nabla_{\nu}$ is a covariant
386: derivative with respect to the bulk metric $g_{\mu\nu}$.) Then the
387: nonstiff brane equations are of the form \be\n{breq}
388: K^{(i)}=g^{\alpha\beta}K^{(i)}_{\alpha\beta}=0\, . \ee
389:
390: For the axially symmetric $D$-dimensional static brane (with the
391: isometry group $O(D-1)$)
392: the induced metric is $(n=D-2$)
393: \be\n{mebr}
394: ds^2=\gamma_{ab}d\zeta^a d\zeta^b=
395: \ee
396: \[
397: -F dT^2+[F^{-1}+r^2 (d\theta/dr)^2] dr^2 + r^2 \sin^2\theta d\Omega^2_{n}\, ,
398: \]
399: and the action \eq{action} reduces to
400: \be\n{raction}
401: S=\Delta T {\cal A}_{n}\int dr {\cal L}\hhh
402: {\cal L}=r^{n}\, \sin^{n}\theta\, \sqrt{1+F r^2 (d\theta/dr)^2}\, .
403: \ee
404: Here $\Delta T$ is the interval of time, and ${\cal
405: A}_n=2\pi^{n/2}/\Gamma(n/2)$ is the surface area of a unit $n$-dimensional
406: sphere.
407:
408: By analyzing the brane equation it is easy to show~\cite{BBH1} that
409: for a brane which asymptotically approaches the equatorial plane
410: $\theta=\pi/2$ one has \be\n{asq} \theta={\pi\over 2}+q(r)\hh
411: q={p\over r}+p'\left\{
412: \begin{array}{cc}
413: r^{-1}\ln r\, ,&\mbox{ for }n=1\, ,\\
414: r^{-n}\, ,&\mbox{ for }n>1\,.
415: \end{array}
416: \right. \ee We call the set of parameters $\{p,p'\}$, which
417: characterizes the solution, the {\em asymptotic data}.
418:
419: The same solution can be determined by its behavior near the
420: horizon. A subcritical brane is uniquely specified by the distance
421: of its tip from the horizon. The condition of the brane surface
422: regularity at this point requires that its tangent plane at the tip
423: is orthogonal to the symmetry axis. This fixes the second constant
424: in the solution. Similarly, a regular brane crossing the horizon is
425: orthogonal to the horizon surface, so that a unique constant fixing
426: the solution is the ``gravitational'' radius of the induced BBH. A
427: solution separating sub- and supercritical solution is a {\em
428: critical solution}. We denote by $\{p_*,p'_*\}$ its asymptotic data.
429:
430:
431:
432:
433:
434: \begin{figure}[htb]
435: \begin{center}
436: \epsfxsize=80mm
437: \epsfbox{zones.eps}
438: \caption{This figure schematically shows a configuration of a
439: supercritical brane in the regime when it is close to the critical
440: one. $Z$ is a proper distance from the horizon as a function of the
441: radius $R$. $R_0$ is the radius of the surface of the intersection
442: of the brane with the horizon. In the region where $Z\ll r_g$ the
443: curvature surface of the horizon can be neglected (the {\em Rindler
444: domain}).} \label{f1}
445: \end{center}
446: \end{figure}
447:
448:
449: Figure \ref{f1} illustrates the near-critical behavior of a
450: supercritical brane. (A similar graph for subcritical branes can be
451: easily obtained from this one by evident changes.) A near-critical
452: brane configuration is characterized by a parameter $R_0$, which is
453: its radius at the intersection with the horizon. For a subcritical
454: brane a similar parameter is $Z_0$, the proper distance of the tip
455: of the brane from the horizon. We consider a case when $R_0$ ($Z_0$)
456: is much smaller than the gravitational radius $r_g$ of the bulk
457: black hole. In the vicinity of the horizon, located at $Z=0$, that
458: is for $Z\ll r_g$, one has \be r-r_g\approx \kappa Z^2/2\hh
459: F\approx\kappa^2 Z^2\, , \ee where $\kappa={1\over 2}(dF/dr)|_{r_g}$
460: is the surface gravity. We call this region a {\em near (or Rindler)
461: zone}. The corresponding induced metric for a near-critical brane
462: in the Rindler zone is \be\n{hmet} ds^2= -\kappa^2 Z^2 dT^2
463: +\left[\left({dZ\over d\lambda}\right)^2+\left({dR\over
464: d\lambda}\right)^2\right]d\lambda^2
465: +R^2 d\Omega^2_{n}\, .
466: \ee
467: Here $(Z(\lambda),R(\lambda))$ is a brane equation written in a
468: parametric form. The action \eq{action} for this induced metric is
469: \ba
470: S&=&\kappa \Delta T {\cal A}_{n} {\cal S}\, ,\\
471: {\cal S}&=&\int d\lambda Z R^n \sqrt{
472: (dZ/d\lambda)^2+(dR/d\lambda)^2}\, .\label{ex_action} \ea This
473: action is evidently invariant under the transformations
474: $\lambda\to\tilde{\lambda}(\lambda)$. In the regions where either
475: $Z$ or $R$ is a monotonic function of $\lambda$, these functions
476: themselves can be used as parameters. As a result, one obtains two
477: other forms of the action which are equivalent to ${\cal S}$ \be
478: {\cal S}=\int dZ {\cal L}_R=\int dR {\cal L}_Z\, , \ee where \be
479: {\cal L}_R=Z R^{n} \sqrt{1+{R'}^2}\hh {\cal L}_Z=Z R^{n}
480: \sqrt{1+{\dot{Z}}^2}\, . \ee Here the prime stands for the
481: derivative with respect to $Z$, while the dot stands for the
482: derivative with respect to $R$. The corresponding Euler-Lagrange
483: equations are \ba\n{eqR}
484: &&Z R R'' +(R R'-nZ)(1+{R'}^2)=0\, ,\\
485: &&R Z \ddot{Z} +(nZ \dot{Z}-R)(1+{\dot{Z}}^2)=0\, .\n{eqZ} \ea It is
486: easy to check that the form of Eqs. \eq{eqR}-\eq{eqZ} is invariant
487: under the following transformations: \ba\n{scR}
488: &&R(Z)=k\tilde{R}(\tilde{Z})\hh
489: Z=k\tilde{Z}\, ,\\
490: &&Z(R)=k\tilde{Z}(\tilde{R})\hh
491: R=k\tilde{R}\, .\n{scZ}
492: \ea
493:
494: Equations \eq{eqR}-\eq{eqZ} have a simple solution \be\n{crit}
495: R=\sqrt{n} Z \ee which plays a special role. We call it a {\em
496: critical solution}. It describes a critical brane which touches the
497: horizon of the bulk black hole at one point, $Z=R=0$. It separates
498: the two different families of solutions, supercritical and
499: subcritical.
500:
501: The relation between $R_0$ and $\{p,p'\}$ is of the form
502: \be
503: \ln R_0=\gamma \ln \Delta p +f(\ln \Delta p)+\ldots\,
504: \ee
505: where
506: \be
507: \gamma ={2\over n+2}\hh
508: \Delta p=\sqrt{(p-p_*)^2+(p'-p'_*)^2}\, ,
509: \ee
510: and the function $f(z)$ is
511: periodic, $f(z+\omega)=f(z)$, with the period
512: \be
513: \omega={\pi (n+2)\over \sqrt{4+4n-n^2}}\, .
514: \ee
515:
516: Our aim is to study how the near-critical solutions are modified
517: when a brane is stiff. We assume that the effective width of the
518: brane is much smaller that the gravitational radius of the black
519: hole. In this approximation, as in the case of a nonstiff brane, the
520: main features of the phase transition in the BBH system are
521: determined by the brane behavior in the near zone, that is close to
522: the event horizon of the bulk black hole, where the Rindler
523: approximation is valid.
524:
525:
526: \section{Stiff Branes}
527:
528: In this work we consider the Dirac-Nambu-Goto action with minimal
529: stiffness correction terms which play the role of higher curvature
530: corrections~\cite{Car}: \be S=-\int d^{n+2}\zeta
531: \sqrt{-\gamma}(1+B\,K^{2}+C\,{\cal K}^{2})\, . \label{theaction}\ee
532: Here $\gamma$ is the determinant of the induced metric given in Eq.
533: ($\ref{action}$), $K=\sum_{i} K_{\mu}^{(i)\mu}$ is the trace of the
534: extrinsic curvature tensor, and ${\cal K}^{2}=\sum_{i}
535: K_{(i)\mu\nu}K^{(i)\mu\nu}$ is its square. A minimal model of
536: stiffness corrections involves only quadratic powers of the
537: extrinsic curvature tensor. In this work we will concentrate on the
538: above ``truncated'' model as a toy model whose solution gives us the
539: static configuration of a stiff brane embedded close to the horizon
540: of the bulk black hole, namely, in the Rindler zone.
541:
542: In the particular
543: case of a domain wall the stiffness coefficients were calculated in
544: the framework of a microscopic model of a vacuum defect in field
545: theory with spontaneous symmetry breaking~\cite{CarGre}. In this
546: work we show that the exact numerical values of the coefficients do
547: not affect the qualitative features of the solution. Nevertheless
548: the sign of the coefficients is important. Positive stiffness
549: coefficients
550: \[ B,C>0
551: \] ensure us that the energy density for the static solution
552: \be \epsilon=-{\cal L}=\sqrt{-\gamma}(1+B\,K^{2}+C\,{\cal K}^{2}),
553: \ee is positive since $K^{2}$ and ${\cal K}^{2}$ are both
554: non-negative.
555:
556: Before going further let us discuss two interesting special cases of
557: the general theory.
558:
559: {\bf $\bf C=0$ case.} In this special case the stiff string
560: equations have a simple exact solution. Namely, any solution of the
561: DNG equations \eq{breq} is at the same time a solution of the stiff
562: string equations. Indeed, a general variation of the action can be
563: split into the variation along the brane and the transverse one.
564: Variations along the brane surface vanish identically. For
565: transverse variations, multiplying the equations of motion by a
566: normal to the brane gives \bea n^{\mu}_{(i)}\,\frac{\delta {\cal
567: L}}{\delta X^{\mu}}=
568: K^{(i)}\,\(1+B\,K^{2}\) \nonumber\\
569: -2\,B\,\sqrt{-\gamma}\,K\,\,n^{\mu}_{(i)}\,\frac{\delta K}{\delta
570: X^{\mu}}=0. \n{eeqq}\eea Now, substituting the DNG equation
571: $K^{(i)}=0$ into the right hand-side of \eq{eeqq}, we see that it
572: vanishes. Hence $K^{(i)}=0$ is a solution of the stiff string
573: equations.
574:
575: {\bf $\bf B+C=0$ case.} This case is not interesting for our
576: consideration since it violates the positive energy condition, but
577: it is of mathematical interest. Let us notice that the Gauss-Codazzi
578: relation for a flat bulk spacetime implies\be {\cal R}=K^{2}-{\cal
579: K}^{2}.\ee Thus one can rewrite the action~(\ref{theaction}) in the
580: following form: \be S=-\int d^{n+2}\zeta \sqrt{-\gamma}(1+B\,{\cal
581: R}+(B+C)\,{\cal K}^{2})\, , \ee For $B+C=0$ the term with ${\cal
582: K}^{2}$ vanishes.
583:
584: Let us return to the discussion of the general case. The units of
585: the stiffness coefficients are length-squared. Therefore under
586: scaling transformation of the spatial coordinates \be R\rightarrow
587: s\,R\;\;,\;\; Z\rightarrow s\,Z, \ee we have \be K^{2}\rightarrow
588: s^{-2}\,K^{2} \;\; ,\;\; {\cal K}^{2} \rightarrow s^{-2}\,{\cal
589: K}^{2}.\ee
590:
591: Using this transformation we can set one of the coefficients to be
592: unit, say $C=1$. We can think about it as taking the basic unit
593: length of the stiff string to be $\sqrt{C}$. Then we are left only
594: with one free parameter $B$ in the action. We will use this choice
595: later in Secs.~$\ref{num1}$ and $\ref{num2}$, when we will discuss
596: the results of the numerical calculations.
597:
598: For completeness we give here the components of the extrinsic
599: curvature for the brane $\(Z(\lambda),R(\lambda)\)$ with the induced
600: metric (\ref{hmet}):
601:
602: \bea K_{\lambda\lambda}&=&{\cal
603: P}^{-1}{\cal A},\\
604: K_{TT}&=&-Z\,\frac{dR}{d\lambda}{\cal P}^{-1}, \nonumber\\
605: K_{\theta \theta}&=&R\,\frac{dZ}{d\lambda}{\cal P}^{-1},\nonumber
606: \eea
607: where
608: \ba
609: {\cal P}&=& \sqrt{
610: (dZ/d\lambda)^2+(dR/d\lambda)^2}\, ,\\
611: {\cal A}&=&\frac{dZ}{d\lambda}\,\frac{d^{2}R}{d\lambda^{2}}
612: -\frac{dR}{d\lambda}\,\frac{d^{2}Z}{d\lambda^{2}}\, .
613: \ea
614:
615: The action (\ref{theaction}) for the induced metric~(\ref{hmet})
616: takes the following form: \ba
617: S&=&-\kappa \Delta T {\cal A}_{n} \int d\lambda {\cal L}\, \label{stiff_action}\\
618: {\cal L}&=&{\cal L}_0+B\,{\cal L}_1+C{\cal L}_2\, ,\nonumber\\
619: {\cal L}_0&=&Z\,R^n \,{\cal P}\, ,\nonumber\\
620: {\cal L}_1&=&Z\,R^n \,{\cal P}\,
621: \(\frac{{\cal A}}{{\cal P}^{3}}+\frac{dR/d\lambda}{Z\,{\cal P}}
622: +\frac{n\,dZ/d\lambda}{R\,{\cal P}}\)^{2},\nonumber\\
623: {\cal L}_2&=&{R^n \({dR/d\lambda}\)^2\over Z{\cal P}}+{n Z {\(d
624: Z/d\lambda\)}^2 R^{n-2}\over {\cal P}}\nonumber\\ &+&{Z R^n {\cal
625: A}^2\over {\cal P}^5}. \nonumber\ea The action (\ref{theaction}) is
626: evidently invariant under transformations
627: $\lambda\to\tilde{\lambda}(\lambda)$ as in the nonstiff case. In a
628: general case, a variation of this action gives equations containing
629: fourth derivatives, while the corresponding constraint equations are
630: of the third order~\cite{endnote4}.
631:
632:
633:
634: \section{Euler-Lagrange equations for stiff branes}
635: In the regions where either $Z$ or $R$ is a
636: monotonic function of $\lambda$, one of the coordinates can be used
637: as a parameter. As a result, one obtains two additional forms of the
638: action:
639:
640:
641: \be {\cal S}=-\int dZ {\cal L}_R=-\int dR {\cal L}_Z\, , \ee where
642: \bea {\cal L}_R=ZR^n {\cal P}\(1+B\left[\frac{R''}{{\cal
643: P}^{3}}+\frac{R'}{Z\,{\cal P}}+\frac{n}{R\,{\cal P}}\right]^2
644: \right. \nonumber
645: \\\left. +C\left[\frac{R''^{2}}{{\cal P}^{6}}+\frac{R'^{2}}{Z^{2}\,{\cal
646: P}^{2}}+\frac{n}{R^{2}\,{\cal P}^{2}}\right]\), \eea
647: \\
648: \bea {\cal L}_Z=ZR^n {\cal P}\(1+B[\frac{1}{Z\,{\cal
649: P}}-\frac{\ddot{Z}}{{\cal P}^{3}}+\frac{n\,\dot{Z}}{R\,{\cal
650: P}}]^{2}\right. \nonumber\\\left. +C[\frac{\ddot{Z}^{2}}{{\cal
651: P}^{6}}+\frac{1}{Z^{2}\,{\cal
652: P}^{2}}+\frac{n\,\dot{Z}^{2}}{R^{2}\,{\cal
653: P}^{2}}]\).\label{lagrangian_sub} \eea
654: ${\cal P}$ in the first
655: relation means ${\cal P}=\sqrt{1+{R'}^2}$, while in the second
656: equation one has ${\cal P}=\sqrt{1+{\dot{Z}}^2}$.
657:
658: As in the nonstiff case,
659: the form with $R(Z)$ is useful for the description of the
660: supercritical brane while $Z(R)$ is more suitable for the
661: description of subcritical branes.
662:
663:
664:
665: The equation for $R(Z)$ takes the following form:
666: \begin{widetext}
667: \be
668: -2\,b\,Z^{3}\,R^{3}\(1+R'^{2}\)^{2}\,R^{(4)}+4\,b\,Z^{2}\,R^{2}\,\(R'^{2}+1\)\left[5\,Z\,R\,R'\,R''-\(R'^{2}+1\)\(n\,Z\,R'+R\)\right]\,R^{(3)}
669: +F\(R,R',R'',Z\)=0, \label{Req}\ee
670:
671: where \bea
672: &&F\(R,R',R'',Z\)=5\,b\,Z^{3}\,R^{3}\(1-6\,R'^{2}\)\,R''^{3}
673: +3\,b\,Z^{2}\,R^{2}\,\(1+R'^{2}\)\left[5\,R\,R'+n\,Z\,\(4\,R'^{2}-1\)\right]\,R''^{2}
674: \nonumber\\
675: &+&Z^{3}\,R^{3}\,\(1+R'^{2}\)^{3}\,R''
676: -Z\,R\,\(1+R'^{2}\)^{2}\(2\,\left[2\,b+3\,B\right]\,n\,Z\,R\,R'
677: +b\,R^{2}\,\left[R'^{2}-2\right]\right.\\
678: &+&\left.
679: n\,Z^{2}\,\left[b+3\,B-3\,B\,n+2\,b\,(n-2)\,R'^{2}\right]\)\,R''+\(1+R'^{2}\)^{3}\,\(R^{3}\,R'\left[Z^{2}+Z^{2}\,R'^{2}\right]-n\,Z\,R^{2}\,\left[Z^{2}+Z^{2}\,R'^{2}\right]\)
680: \nonumber\\
681: &-&\(1+R'^{2}\)^{3}\(\left[b-3\,B\,(n-1)\right]\,n\,Z^{2}\,R\,R'-b\,n\,Z\,R^{2}\,R'^{2}+b\,R^{3}\,R'\,\left[R'^{2}+2\right]+n\,[n-2]\,Z^{3}\left[b+B\,(n-1)+2\,b\,R'^{2}\right]\),
682: \nonumber \eea
683: and $b=B+C$.
684:
685: A similar equation for $Z(R)$ is \be - 2\,b\,R^3\,{Z}^3\,{\left( 1 +
686: {\dot{Z}}^2
687: \right)}^2\,Z^{(4)}+4\,b\,R^{2}\,Z^{2}\,\(1+{\dot{Z}}^{2}\)\,\left[5\,R\,Z\,\dot{Z}\,\ddot{Z}-\(1+\dot{Z}^{2}\)\,\(n\,Z+R\,\dot{Z}\)\right]\,Z^{(3)}
688: +G\(Z,\dot{Z},\ddot{Z},R\)=0, \label{zr}\ee
689:
690: where \bea &&
691: G\(Z,\dot{Z},\ddot{Z},R\)=5\,b\,R^{3}\,Z^{3}\,\(1-6\,\dot{Z}^{2}\)\,\ddot{Z}^{3}
692: +3\,b\,R^{2}\,Z^{2}\,\(1+\dot{Z}^{2}\)\left[5\,n\,Z\,\dot{Z}+R\,\(4\,\dot{Z}^{2}-1\)\right]\,\ddot{Z}^{2}
693: +R^{3}\,Z^{3}\,\(1+\dot{Z}^{2}\)^{3}\,\ddot{Z}\nonumber\\&-&R\,Z\,\(1+\dot{Z}^{2}\)^{2}\,\(2\,\left[3\,B+2\,b\right]\,n\,R\,Z\,\dot{Z}
694: -
695: b\,R^{2}\,\left[2\,\dot{Z}^{2}-1\right]+n\,Z^{2}\,\left[2\,b\,(n-2)+\(3\,B+b-3\,B\,n\)\right]\)\ddot{Z}
696: \nonumber\\
697: &-&R^{2}\,Z^{2}\(R-n\,Z\,\dot{Z}\)\(1+\dot{Z}^{2}\)^{4}
698: +\(1+\dot{Z}^{2}\)^{3}\(-b\,n\,R^{2}\,Z\,\dot{Z}+n\,R\,Z^{2}\left[b-3\,B(n-1)\right]\,\dot{Z}^{2}+b\,R^{3}\left[1+2\,\dot{Z}^{2}\right]\right.\\
699: &+&\left.
700: 2\,b\,n\,(n-2)\,Z^{3}\,\dot{Z}+n\,(n-2)\,Z^{3}\,\dot{Z}^{3}\left[b+B(n-1)\right]\).\nonumber
701: \eea
702: \end{widetext}
703:
704: Let us denote \be l=\max\(\sqrt{B},\sqrt{C}\)\, . \ee $l$ has
705: dimensionality of the length. The extrinsic curvature corrections
706: are dominant for $R,Z\lesssim l$ where the stiff brane differs
707: significantly from the DNG brane. The significant effect of the
708: stiffness is localized in the region where the original DNG brane is
709: extremely bent. This happens in the neighborhood of the point
710: $R=Z=0$ which is defined by the length scale $l$. For $R,Z \gg l$
711: the solution for the stiff brane-BH system approaches the DNG
712: brane-BH system. In particular, $R=\sqrt{n}Z$ is the attractor
713: solution for the DNG brane-BH system and therefore it should be also
714: an attractor for the stiff brane-BH system.
715:
716: \section{Near-Critical Modes}
717: \label{perturbations} Let us study linear perturbations to the
718: attractor solution for the case of a stiff brane. Our objective is
719: to obtain the modes in the neighborhood of the attractor far away
720: from the singular region $Z\gg l$ but still located in the Rindler
721: zone. The modes will guide us later in the setting of the boundary
722: conditions for stiff branes.
723:
724: Let us substitute in the equation for $R$~(\ref{Req}) the following
725: expression: \be R(Z)=\sqrt{n}Z+\rho(Z), \ee and keep only linear
726: terms in $\rho(Z)$. Then we obtain the following linearized
727: equation: \bea
728: a_{0}^{s}(Z)\,\rho&+&a_\Omega{1}^{s}(Z)\,\rho'+a_{2}^{s}(Z)\,\rho''+a_{3}^{s}(Z)\,\rho^{(3)}+a_{4}^{s}(Z)\,\rho^{(4)} \nonumber\\
729: &=&2\,C\,(n-1)\,(n+1)^{2}\,n^{-\frac{1}{2}}Z,
730: \label{pert eq}\eea \bea a_{0}^{s}(Z)&=&(n+1)\,\left[2\,B\(\,n-5\)-C(n+7)\right]+a_{0}(Z),\nonumber\\
731: a_{1}^{s}(Z)&=&(n+1)\,Z\left[2\,B\,\(16-n\)-17\,C\,(n-1)\right]+a_{1}(Z), \nonumber\\
732: a_{2}^{s}(Z)&=&-(n+1)\,Z^{2}\,\left[2\,B\,(n+1)+C(2\,n-1)\right]+a_{2}(Z), \nonumber\\
733: a_{3}^{s}(Z)&=&-4\,(B+C)(n+1)Z^{3},\nonumber\\
734: a_{4}^{s}(Z)&=&-2\,(B+C)Z^{4}. \eea Here $a_{i}(Z)$ are the
735: coefficients in the linearized DNG brane equations:
736: \bea a_{0}(Z)&=&(n+1)^{2}\,Z^{2},\nonumber\\
737: a_{1}(Z)&=&(n+1)^{2}\,Z^{3}, \nonumber\\
738: a_{2}(Z)&=&(n+1)\,Z^{4}.
739: \eea
740: Now let us take the limit $Z\gg l$ and as a result we obtain the following equation:
741: \bea
742: &&(n+1)^{2}\left[\,Z\,\rho+Z^{2}\,\rho'\right]+(n+1)\,Z^{3}\,\rho''\nonumber\\
743: &-&2\,(B+C)\,\left[2(n+1)\,Z^{2}\,\rho^{(3)}+Z^{3}\,\rho^{(4)}\right]
744: \nonumber\\&=&2\,C\,(n-1)\,(n+1)^{2}\,n^{-\frac{1}{2}}.\n{lin}
745: \eea
746:
747:
748: The leading term of the particular solution at large $Z$ is
749: \be\rho_{P}= \frac{C\,(n^{2}-1)}{\sqrt{n}\,Z}+{\cal
750: O}\(\frac{1}{Z^{2}}\),\ee and therefore it does not have an effect
751: on the attractor $R=\sqrt{n}Z$ at large $Z$, as expected. A general
752: solution of \eq{lin} is a sum of this particular solution and a
753: general solution of the homogeneous equation obtained from \eq{lin}
754: by omitting the right-hand side. It is interesting that this
755: homogeneous solution depends only on the sum $B+C$ of the stiffness
756: coefficients.
757:
758: Since the homogeneous equation is of the fourth order,
759: it has four linearly independent solutions (asymptotic modes).
760: Two of the asymptotic modes reproduce the asymptotic solutions for
761: the DNG brane: \be \rho \sim
762: Z^{-\frac{1}{2}(n\pm\sqrt{n^{2}-4\,n-4})}. \label{asym}\ee
763:
764: The other two modes appear only for the stiff brane: \be \rho \sim
765: \exp\({\pm \sqrt{\frac{n+1}{2\,(B+C)}}\,Z}\). \ee The additional two
766: modes above are added due to the stiffness corrections. One of the
767: additional modes is an unstable mode which takes the solution away
768: from the attractor. This mode should be eliminated by appropriate
769: boundary conditions in order to reproduce the DNG solutions at large
770: distances.
771:
772: Thus we arrive to a boundary value problem. Let us take, for
773: example, the subcritical configuration where the solution can be
774: written as $Z(R)$ (a similar discussion is applicable to the
775: supercritical configuration with some evident changes). Since
776: Eq.~(\ref{zr}) is of the fourth order, we need four initial values
777: (in case of an initial value problem). The configuration is axially
778: symmetric with the symmetry axis $R=0$ and it is plausible that the
779: stiff brane solution preserves the same axial symmetry. Consider a
780: brane passing through the point $Z(0)=Z_{0}$ (the proper distance of
781: the tip of the brane from the horizon). The axial symmetry and the
782: regularity of the brane at $R=0$ enforces $\dot{Z}(0)=0$ and
783: $Z^{(3)}(0)=0$, while $\ddot{Z}(0)$ remains a free parameter. This
784: free parameter will allow us to eliminate the unstable mode by
785: finding the specific value for $\ddot{Z}(0)$.
786:
787: In conclusion, we have a boundary value problem defining
788: near-critical solutions of the stiff brane equations: In order to
789: obtain the right asymptotic behavior at large $R$
790: \[Z \rightarrow \frac{R}{\sqrt{n}}\]
791: for a brane which passes at $Z(0)=Z_{0}$ we have to tune the
792: parameter $\ddot{Z}(0)=\ddot{Z}_{0}$.
793:
794: \section{Numerical Results for Stiff Branes: $n=1$ Case}
795: \label{num1} Using a numerical shooting analysis we obtained the
796: values of $\ddot{Z}_0$ for which there is a solution for the
797: subcritical brane that satisfies the boundary conditions: It starts
798: at $Z_0$ and asymptotically goes to the attractor. A similar
799: analysis was performed for the supercritical configuration where
800: values of $R''(0)$ were determined as a function of $R_{0}$ (the
801: radius of the BBH horizon).
802:
803: Let us start by examining the case $n=1$. As we will see, this case
804: is qualitatively different from $n>1$. For $B=0$ the action
805: [Eq.~(\ref{stiff_action})] is completely symmetric for the
806: interchange of $R\leftrightarrow Z$. This discrete symmetry of the
807: equations implies the same results for the subcritical and
808: supercritical configurations. For this reason we give here the
809: results for the subcritical configuration $B=0$ and $n=1$
810: (Fig.~\ref{figB0}) when for the supercritical configuration the
811: graph is the same (up to the interchange of $R\leftrightarrow Z$).
812: The plot at this figure shows $\ddot{Z}_0$ as a function of $Z_0$
813: for near-critical configurations. The finite gap in the neighborhood
814: of the point $R=Z=0$ demonstrates the first order phase transition
815: in the BBH system. A detailed interpretation of this picture is
816: given below in the discussion of the case $n=2$ (where the same
817: picture appears only in the subcritical configuration).
818:
819: \begin{figure} [htb]
820: \begin{center}
821: \epsfxsize=60mm \epsfbox{figB0.eps} \caption{$\ddot{Z}_0$ as a
822: function of $Z_0$ for $n=1$ and $B=0$. The symmetry of the action in
823: this case implies that the same graph is valid to the supercritical
824: configuration as well -- $R''(0)$ as a function of $R_0$.}
825: \label{figB0}
826: \end{center}
827: \end{figure}
828:
829: For the case of $B \neq 0$ the action (\ref{stiff_action}) is no
830: longer symmetric under the reflection
831: $R\leftrightarrow Z$. Nevertheless we find
832: numerically that the results are symmetric within the used accuracy.
833: Evidence for this symmetry we can find in the linearized equations
834: for a perturbation around the attractor. In Sec.~\ref{perturbations}
835: we studied linearized perturbations to the supercritical
836: configuration \be R(Z)=\sqrt{n}Z+\rho(Z). \ee Keeping only linear
837: terms in $\rho(Z)$ gave us Eq. ($\ref{pert eq}$). For the special
838: case of $n=1$ this equation reads \bea &&
839: (B+C)\,Z^{4}\,\rho^{(4)}(Z)+
840: 4\,(B+C)\,Z^{3}\,\rho^{(3)}(Z)\nonumber\\&+&Z^{2}\(4\,B+C-Z^{2}\)\,\rho''(Z)-2\,Z^{3}\,\rho'(Z)\nonumber
841: \\&-&2\(Z^{2}-4\,B-4\,C\)\,\rho(Z)=0 \eea
842:
843: In a similar way for the subcritical configuration substitution of
844: \be Z(R)=\frac{R}{\sqrt{n}}+\zeta(R) \ee into Eq. (\ref{zr}) and
845: keeping only linear terms in $\zeta(R)$ gives the following linear
846: equation ($n=1$)
847: \bea
848: &&(B+C)\,R^{4}\,\zeta^{(4)}(R)+4\,(B+C)\,R^{3}\,\zeta^{(3)}(R)\nonumber\\
849: &+&R^{2}\(4\,B+C-R^{2}\)\,\ddot{\zeta}(R)-2\,R^{3}\,\dot{\zeta}(R)\nonumber \\
850: &-&2\(R^{2}-4\,B-4\,C\)\,\zeta(R)=0.
851: \eea
852:
853: Hence the symmetry $R\leftrightarrow Z$ is demonstrated analytically
854: in the linear approximation. This does not imply an exact reflection
855: symmetry of the solutions, but at least makes it possible.
856:
857: \section{Numerical Results for Stiff Branes: $n>1$ Case}
858: \label{num2} For $n>1$ there is no reflection symmetry of the action
859: anymore, and sub- and supercritical solutions behave quite
860: differently. Let us start with the subcritical configuration and
861: demonstrate that it exhibits the same qualitative features as $n=1$.
862: This is a good place to compare the details of the new picture with
863: the nonstiff case.
864:
865: Consider, for example, the case of $n=2$ and $B=1$ in
866: Fig.~\ref{B1n2}. The value of $\ddot{Z}_0$ is plotted as a function
867: of the position where the brane crosses the axis of symmetry $Z_0$.
868: In the case of DNG branes, i.e. without stiffness, the dependence of
869: $\ddot{Z}_0$ on $Z_0$ is determined from the Euler-Lagrange equation
870: (\ref{eqZ}) to be (see in~\cite{BBH1}): \be
871: \ddot{Z}_0=\frac{1}{(n+1)\,Z_0}. \label{init_DNG}\ee This function
872: is plotted in Fig.~
873: \ref{B1n2} for comparison with the case of stiff
874: branes.
875:
876: \begin{figure}[htb]
877: \epsfysize=50mm \epsfbox{C1+DNG.eps} \caption{$\ddot{Z}_0$ as a
878: function of $Z_0$ for $n=2$. The dashed line is the same function
879: for DNG branes (without stiffness terms).} \label{B1n2}
880: \end{figure}
881:
882: Few features can be observed in the graph:
883: \begin{itemize}
884: \item There is a finite gap $0<Z_0\lesssim 1$ in which the solution for the embedded stiff brane does not exist at
885: all. Hence the singular point is resolved for the subcritical
886: branch. This is a characteristic feature of first order phase
887: transitions.
888: \item $\ddot{Z}_0$ is bounded, unlike DNG branes [Eq.~
889: (\ref{init_DNG})]
890: for which $\ddot{Z}_0$ is unbounded
891: \item For $1 \lesssim Z_0 \lesssim 1.25$ we see coexistence of two
892: branches of solutions. For any $Z_0$ in this range there are two
893: possible values of $\ddot{Z}_0$ and thus two possible configurations
894: of the stiff brane. One can compare the energy [see
895: \eq{lagrangian_sub}] of the two branches. A numerical comparison of
896: the energies [Fig.~\ref{energy fig}] reveals that the branch with
897: the lower values of $\ddot{Z}_0$ is energetically favored. The
898: branch with higher energy corresponds to a local phase at maximum.
899: This solution should be unstable and separates two stable phases in
900: a first order phase transition.
901: \begin{figure}[htb]
902: \epsfxsize=60mm \epsfbox{energy.eps} \caption{The energy density
903: integrated for $0\leq R \leq5$ as a function of $Z_0$ comparing two
904: branches in the segment ($1 \lesssim Z_0 \lesssim 1.25$). Note that
905: the minimal energy is obtained at the point which corresponds
906: approximately to $\ddot{Z}_0=0$.} \label{energy fig}
907: \end{figure}
908:
909:
910: \item There are solutions for stiff branes that satisfy the boundary conditions with negative $\ddot{Z}_0$. Such
911: solutions exist only for $-0.025\lesssim \ddot{Z}_0 \lesssim 0$.
912:
913: \item There exists an ``end point'' in the plot with minimal value
914: of $\ddot{Z}_0$. For $\ddot{Z}_0$ less than this value a solution
915: does not exist.
916:
917: \item At large values of $Z_0$ we see that the effects of stiffness are
918: negligible. The points of the stiff branes approach the DNG branes
919: at large values of $Z_0$.
920: \end{itemize}
921:
922: In order to check that the obtained results are robust we repeated
923: the same calculations for various values of $B$. In all cases we
924: found that the same qualitative behavior repeats itself: A finite
925: gap in the existence of solutions for $0<Z_0\lesssim l$,
926: two branches of solutions in a small neighborhood of $Z_0 \sim
927: l$ etc.
928:
929: For illustration we give in Fig.~\ref{two plots} two graphs for two
930: values of $B$ with four orders of magnitude difference
931: ($B=0.01,\;100$).
932: \begin{figure}[htb]
933: \epsfxsize=55mm \epsfbox{B100.eps} \epsfxsize=60mm
934: \epsfbox{B001.eps}
935: \caption{
936: $\ddot{Z}_0$ as a function of $Z_0$ for $n=2$. $B=100,\; \;
937: 0.01$.}\label{two plots}
938: \end{figure}
939:
940: In addition we checked for various dimensions $n=4,5$ and found the
941: same qualitative behavior (see Fig.~\ref{n=4} for $n=4$ as an
942: example). Despite the fact that $n \leq 4$ is different from $n \geq
943: 5$ since in the former the phase space behavior of the critical
944: solution behaves as of a focal point and in the latter as a node
945: (see~\cite{BBH1}). This type of transition in the near-critical
946: solutions has no influence on the neighborhood of the singular point
947: $R=Z=0$ where the stiffness terms are dominant.
948: \begin{figure}[htb]
949: \epsfxsize=60mm \epsfbox{C1n4.eps}
950: \caption{
951: $\ddot{Z}_0$ as a function of $Z_0$ for $n=4,\, B=1$.} \label{n=4}
952: \end{figure}
953:
954: It is surprising that when we repeated similar calculations for {\em
955: supercritical} stiff branes with $n>1$ we found that for the
956: supercritical configurations there is no singularity resolution.
957: The stiffness terms break the symmetry between the supercritical and
958: subcritical brane-black hole systems. The supercritical solutions
959: show no gap nor double-branch behavior. As an example let us look at
960: the supercritical configuration for $n=2$. For $B<0.906$ we did not
961: find evidence for the existence of a solution in the vicinity of the
962: point $R(0)=R''(0)=0$. See Fig.~\ref{superlow} for $B=1$ and
963: Fig.~\ref{superhigh} for $B=0.5$. We stress that in both cases the
964: curvature singularity still exists.
965:
966: \begin{figure}[htb]
967: \epsfxsize=60mm \epsfbox{C1super.eps}
968: \caption{
969: $R''(0)$ as a function of $R_0$ (supercritical) for $n=2,\quad
970: B=1$.} \label{superlow}
971: \end{figure}
972:
973:
974: \begin{figure}[htb]
975: \epsfxsize=60mm \epsfbox{B05super.eps}
976: \caption{
977: $R''(0)$ as a function of $R_0$ (supercritical) for $n=2,\,\quad
978: B=0.5$.} \label{superhigh}
979: \end{figure}
980:
981: \noindent
982:
983: \section{Discussion}
984:
985: We observed that due to the stiffness corrections the singularity of
986: the critical solution is resolved for $n=1$ in a symmetric form
987: (both in the subcritical and supercritical configurations) and for
988: $n>1$ only for subcritical configurations. We observed this
989: resolution in the creation of a finite gap and a clear signature of
990: a first order phase transition. This signature is observed in a
991: typical hysteresis curve of coexistence of two phases--stable and
992: unstable. For $n=1$ we see a first order phase transition on both
993: sides of the singularity (supercritical and subcritical
994: configurations) when for $n>1$ we see only half of this picture--a
995: first order phase transition in the subcritical configuration.
996:
997: We expect that a similar picture would emerge in merger transitions
998: when higher derivative corrections are included. Inclusion of higher
999: derivative corrections might cause the merger transition to become
1000: first order in nature and create a finite gap between the thin
1001: black string (``the waist'') and the caged black hole. This way the
1002: naked singularity and the violation of cosmic censorship hypothesis
1003: that appear in the classical approximation would be resolved. This
1004: might also be a natural way to resolve the apparent tension between
1005: the suggested scenario for the merger transition and the observation
1006: that such a pinch-off can occur only at infinite affine parameter
1007: along the horizon~\cite{Horowitz:2001cz}. The resolution is the
1008: following. When the system approaches the Planckian scale, at a
1009: finite time, the first order phase transition takes the system to
1010: the second phase. Therefore with quantum corrections the
1011: ``pathologies'' of infinite affine parameter and naked singularity
1012: would be resolved.
1013:
1014: The asymmetry that we found might be a result of the incompleteness
1015: of the truncated model that we used to describe the full effect of
1016: quantum corrections. It might be also a hint on an asymmetry which
1017: is generic in the topology change in general.\\
1018:
1019: \section*{Acknowledgments} One of the authors (V.F.) is
1020: grateful to the Research Program on Gravity and Cosmology at the
1021: Yukawa Institute where this work was started. He also thanks the
1022: Natural Sciences and Engineering Research Council of Canada (NSERC)
1023: and the Killam Trust for financial support. DG thanks NSERC for the
1024: financial support, and thanks Vadim Asnin, Amit Giveon, Barak Kol,
1025: Matthew Lippert and Amos Ori for fruitful discussions.
1026:
1027: \begin{thebibliography}{}
1028:
1029: \bibitem{GH} G.~W Gibbons and S~W Hawking,
1030: "Action integrals and partition functions in quantum gravity",
1031: Phys.\ Rev.\ D {\bf 15}, 2752 (1977).
1032:
1033: \bibitem{review}
1034: B.~Kol,
1035: ``The phase transition between caged black holes and black strings:
1036: A review,''
1037: Phys.\ Rept.\ {\bf 422}, 119 (2006)
1038: [arXiv:hep-th/0411240].
1039: %%CITATION = PRPLC,422,119;%%
1040:
1041: \bibitem{niarchos}
1042: V.~Niarchos,
1043: ``Phases of Higher Dimensional Black Holes,''
1044: arXiv:0808.2776 [hep-th].
1045: %%CITATION = ARXIV:0808.2776;%%
1046:
1047: \bibitem{reviewob1}
1048: N.~A.~Obers,
1049: ``Black Holes in Higher-Dimensional Gravity,''
1050: arXiv:0802.0519 [hep-th].
1051: %%CITATION = ARXIV:0802.0519;%%
1052:
1053: \bibitem{reviewob2}
1054: T.~Harmark, V.~Niarchos and N.~A.~Obers,
1055: ``Instabilities of black strings and branes,''
1056: Class.\ Quant.\ Grav.\ {\bf 24}, R1 (2007)
1057: [arXiv:hep-th/0701022].
1058: %%CITATION = CQGRD,24,R1;%%
1059:
1060:
1061: \bibitem{reviewob3}
1062: T.~Harmark and N.~A.~Obers,
1063: ``Phases of Kaluza-Klein black holes: A brief review,''
1064: arXiv:hep-th/0503020.
1065:
1066:
1067: \bibitem{conifold-barak}
1068: B.~Kol,
1069: ``Topology change in general relativity and the black-hole black-string
1070: transition,''
1071: JHEP {\bf 0510}, 049 (2005)
1072: [arXiv:hep-th/0206220].
1073: %%CITATION = JHEPA,0510,049;%%
1074:
1075: \bibitem{endnote1}
1076: The $S^{2}$ is the time circle fibred over a segment in the $r-z$
1077: plane when $r$ is the radial coordinate and $z$ is the coordinate of
1078: the compact dimension
1079:
1080: \bibitem{conifolds}
1081: P.~Candelas and X.~C.~de la Ossa,
1082: ``Comments on Conifolds,''
1083: Nucl.\ Phys.\ B {\bf 342}, 246 (1990).
1084: %%CITATION = NUPHA,B342,246;%%
1085:
1086: \bibitem{Barak2}
1087: V.~Asnin, B.~Kol and M.~Smolkin,
1088: ``Analytic evidence for continuous self similarity of the critical merger
1089: solution,''
1090: Class.\ Quant.\ Grav.\ {\bf 23}, 6805 (2006)
1091: [arXiv:hep-th/0607129].
1092: %%CITATION = CQGRD,23,6805;%%
1093:
1094: \bibitem{Barak1}
1095: B.~Kol,
1096: ``Choptuik scaling and the merger transition,''
1097: JHEP {\bf 0610}, 017 (2006)
1098: [arXiv:hep-th/0502033].
1099: %%CITATION = JHEPA,0610,017;%%
1100:
1101: \bibitem{choptuik-original}
1102: M.~W.~Choptuik,
1103: ``Universality And Scaling In Gravitational Collapse Of A Massless Scalar
1104: Field,''
1105: Phys.\ Rev.\ Lett.\ {\bf 70}, 9 (1993).
1106: %%CITATION = PRLTA,70,9;%%
1107:
1108: \bibitem{chptuik-review}
1109: C.~Gundlach,
1110: ``Critical phenomena in gravitational collapse,''
1111: Phys.\ Rept.\ {\bf 376}, 339 (2003)
1112: [arXiv:gr-qc/0210101].
1113: %%CITATION = PRPLC,376,339;%%
1114:
1115: \bibitem{CFL}
1116: M.~Christensen, V.~P.~Frolov and A.~L.~Larsen,
1117: ``Soap bubbles in outer space: Interaction of a domain wall with a black
1118: hole,''
1119: Phys.\ Rev.\ D {\bf 58}, 085008 (1998)
1120: [arXiv:hep-th/9803158].
1121: %%CITATION = PHRVA,D58,085008;%%
1122:
1123: \bibitem{FLC}
1124: V.~P.~Frolov, A.~L.~Larsen and M.~Christensen,
1125: ``Domain wall interacting with a black hole: A new example of critical
1126: phenomena,''
1127: Phys.\ Rev.\ D {\bf 59}, 125008 (1999)
1128: [arXiv:hep-th/9811148].
1129:
1130: \bibitem{BBH1}
1131: V.~P.~Frolov,
1132: ``Merger transitions in brane-black-hole systems: Criticality, scaling, and
1133: self-similarity,''
1134: Phys.\ Rev.\ D {\bf 74}, 044006 (2006)
1135: [arXiv:gr-qc/0604114].
1136: %%CITATION = PHRVA,D74,044006;%%
1137:
1138: \bibitem{sorkin}
1139: E.~Sorkin and Y.~Oren,
1140: ``On Choptuik's scaling in higher dimensions,''
1141: Phys.\ Rev.\ D {\bf 71}, 124005 (2005)
1142: [arXiv:hep-th/0502034].
1143: %%CITATION = PHRVA,D71,124005;%%
1144:
1145: \bibitem{Filev:2008xt}
1146: V.~G.~Filev and C.~V.~Johnson,
1147: ``Universality in the Large $N_c$ Dynamics of Flavour: Thermal Vs. Quantum
1148: Induced Phase Transitions,''
1149: arXiv:0805.1950 [hep-th].
1150: %%CITATION = ARXIV:0805.1950;%%
1151:
1152: \bibitem{Mateos:2007yp}
1153: D.~Mateos and L.~Patino,
1154: ``Bright branes for strongly coupled plasmas,''
1155: JHEP {\bf 0711}, 025 (2007)
1156: [arXiv:0709.2168 [hep-th]].
1157: %%CITATION = JHEPA,0711,025;%%
1158:
1159: %\cite{Mateos:2007vn}
1160: \bibitem{Mateos:2007vn}
1161: D.~Mateos, R.~C.~Myers and R.~M.~Thomson,
1162: ``Thermodynamics of the brane,''
1163: JHEP {\bf 0705}, 067 (2007)
1164: [arXiv:hep-th/0701132].
1165: %%CITATION = JHEPA,0705,067;%%
1166:
1167: \bibitem{Mateos:2006nu}
1168: D.~Mateos, R.~C.~Myers and R.~M.~Thomson,
1169: ``Holographic phase transitions with fundamental matter,''
1170: Phys.\ Rev.\ Lett.\ {\bf 97}, 091601 (2006)
1171: [arXiv:hep-th/0605046].
1172: %%CITATION = PRLTA,97,091601;%%
1173:
1174: \bibitem{Hoyos:2006gb}
1175: C.~Hoyos-Badajoz, K.~Landsteiner and S.~Montero,
1176: ``Holographic Meson Melting,''
1177: JHEP {\bf 0704}, 031 (2007)
1178: [arXiv:hep-th/0612169].
1179: %%CITATION = JHEPA,0704,031;%%
1180:
1181: \bibitem{Babington:2003vm}
1182: J.~Babington, J.~Erdmenger, N.~J.~Evans, Z.~Guralnik and I.~Kirsch,
1183: ``Chiral symmetry breaking and pions in non-supersymmetric gauge / gravity
1184: duals,''
1185: Phys.\ Rev.\ D {\bf 69}, 066007 (2004)
1186: [arXiv:hep-th/0306018].
1187: %%CITATION = PHRVA,D69,066007;%%
1188:
1189: \bibitem{endnote2}
1190: Higher derivative corrections to this configuration correspond to
1191: finite 't Hooft coupling corrections in the gauge theory. In our
1192: approximation of Rindler spacetime the higher derivative corrections
1193: to the bulk BH (or the Dp-brane) are suppressed and therefore we
1194: consider only corrections to the brane (or the Dq-brane). Therefore
1195: this work might give guidelines for the necessary modifications to
1196: the holographic picture of meson melting.
1197:
1198: \bibitem{endnote25}
1199: For a discussion of possible quantum gravity corrections in the
1200: similar problem of critical collapse see~\cite{brady,peleg}.
1201:
1202: \bibitem{endnote3}
1203: Inclusion of linear order corrections in the extrinsic curvature
1204: will depend on a choice of sign and thus on an orientation
1205: convention and we restrict ourselves to lagrangians that depend only
1206: on the embedding of the worldsheet. See~\cite{Car} for a discussion
1207: of this point.
1208:
1209: \bibitem{polyakov}
1210: A.~M.~Polyakov,
1211: ``Fine Structure of Strings,''
1212: Nucl.\ Phys.\ B {\bf 268}, 406 (1986).
1213: %%CITATION = NUPHA,B268,406;%%
1214:
1215: \bibitem{VilShel} A. Vilenkin, and E.P.S. Shellard,
1216: "Cosmic Strings and Other Topological Defects",
1217: Cambridge University Press, Cambridge Monographs on
1218: Mathematical Physics (1994).
1219:
1220: \bibitem{CarGre}
1221: B.~Carter and R.~Gregory,
1222: ``Curvature corrections to dynamics of domain walls,''
1223: Phys.\ Rev.\ D {\bf 51}, 5839 (1995)
1224: [arXiv:hep-th/9410095].
1225: %%CITATION = PHRVA,D51,5839;%%
1226:
1227: \bibitem{Sen1}
1228: A.~Sen,
1229: ``Stretching the horizon of a higher dimensional small black hole,''
1230: JHEP {\bf 0507}, 073 (2005)
1231: [arXiv:hep-th/0505122].
1232: %%CITATION = JHEPA,0507,073;%%
1233:
1234: \bibitem{Sen2}
1235: A.~Sen,
1236: ``Black Hole Entropy Function, Attractors and Precision Counting of
1237: Microstates,''
1238: arXiv:0708.1270 [hep-th].
1239:
1240: \bibitem{Dirac}
1241: P.~A.~M.~Dirac,
1242: ``An Extensible model of the electron,''
1243: Proc.\ Roy.\ Soc.\ Lond.\ A {\bf 268}, 57 (1962).
1244:
1245: \bibitem{Nambu}
1246: J. Nambu, Lectures at the Copenhagen Summer Symposium (1970),
1247: unpublished;
1248:
1249: \bibitem{Goto}
1250: T.~Goto,
1251: ``Relativistic quantum mechanics of one-dimensional mechanical continuum and
1252: subsidiary condition of dual resonance model,''
1253: Prog.\ Theor.\ Phys.\ {\bf 46}, 1560 (1971).
1254: %%CITATION = PTPKA,46,1560;%%
1255:
1256: \bibitem{Car}
1257: B.~Carter,
1258: ``Equations of motion of a stiff geodynamic string or higher brane,''
1259: Class.\ Quant.\ Grav.\ {\bf 11}, 2677 (1994).
1260: %%CITATION = CQGRD,11,2677;%%
1261:
1262: \bibitem{endnote4}
1263: In the special case when $B+C=0$, after reduction of the action
1264: $\int \sqrt{-g}{\cal R}$ and integration by parts one obtains the
1265: following contribution of this term to ${\cal L}$
1266: \[
1267: n(n-1){\cal P}ZR^{n-2} +{n R^{n-2}\over {\cal P}}
1268: \frac{dR}{d\lambda}\left[2R\frac{dZ}{d\lambda}+(n-1)Z\frac{dR}{d\lambda}\right].
1269: \]
1270:
1271: \bibitem{Horowitz:2001cz}
1272: G.~T.~Horowitz and K.~Maeda,
1273: ``Fate of the black string instability,''
1274: Phys.\ Rev.\ Lett.\ {\bf 87}, 131301 (2001)
1275: [arXiv:hep-th/0105111].
1276: %%CITATION = PRLTA,87,131301;%%
1277:
1278: \bibitem{brady}
1279: P.~R.~Brady and A.~C.~Ottewill,
1280: ``Quantum corrections to critical phenomena in gravitational collapse,''
1281: Phys.\ Rev.\ D {\bf 58}, 024006 (1998)
1282: [arXiv:gr-qc/9804058].
1283:
1284: \bibitem{peleg}
1285: Y.~Peleg, S.~Bose and L.~Parker,
1286: ``Choptuik scaling and quantum effects in 2D dilaton gravity,''
1287: Phys.\ Rev.\ D {\bf 55}, R4525 (1997)
1288: [arXiv:gr-qc/9608040].
1289:
1290: \end{thebibliography}
1291:
1292:
1293:
1294:
1295: \end{document}
1296:
1297:
1298: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1299:
1300:
1301: \bibitem{Hawking-Gibbons1}
1302: G.~W.~Gibbons and S.~W.~Hawking,
1303: ``Gravitational Multi - Instantons,''
1304: Phys.\ Lett.\ B {\bf 78}, 430 (1978).
1305: %%CITATION = PHLTA,B78,430;%%
1306:
1307: \bibitem{Hawking-Gibbons2}
1308: G.~W.~Gibbons and S.~W.~Hawking,
1309: ``Classification Of Gravitational Instanton Symmetries,''
1310: Commun.\ Math.\ Phys.\ {\bf 66}, 291 (1979).
1311: %%CITATION = CMPHA,66,291;%%
1312: