1: \documentclass [12pt] {book}

2: %\topmargin 0in \textheight 9in \oddsidemargin 0in

3: 

4: 

5: \usepackage{graphicx}

6: \usepackage{amsmath}

7: \usepackage{amsfonts}

8: \usepackage{amssymb}

9: 

10: 

11: \def\f{{\rm p}}

12: \def\vd{{(5)}}

13: \def \vu {{(5)}\!}

14: \def \D {\mbox{D}}

15: \def \curl {\mbox{curl}\,}

16: \def \ep {\varepsilon}

17: \def \cu{{\cal U}}

18: \def \cq{ Q}

19: \def \cp{ \Pi}

20: \def \bcq { \bar{Q}}

21: \def \bcp { \bar{\Pi}}

22: \def\bea{\begin{eqnarray}}

23: \def\eea{\end{eqnarray}}

24: \newcommand{\be}{\begin{equation}}

25: \newcommand{\ee}{\end{equation}}

26: \newcommand{\ba}{\begin{eqnarray}}

27: \newcommand{\ea}{\end{eqnarray}}

28: \newcommand{\half}{\frac{1}{2}}

29: \newcommand{\pre}{\frac{1}{4\pi\alpha'}}

30: \newcommand{\four}{^{(4)}}

31: \newcommand{\five}{\tilde}

32: \newcommand{\mb}[1]{\mbox{\boldmath $#1$}}

33: \evensidemargin 0in

34: %\renewcommand{\baselinestretch}{1.5}

35: \textwidth  6.0in

36: 

37: \begin{document}

38: \pagestyle{empty}

39: \input{intestazione.tex}

40: \pagestyle{headings}

41: \pagenumbering{roman}

42: \tableofcontents

43: \chapter{Foreword}

44: \footnote{This is a rough and informal view of quantum gravity,

45: using \cite{R1}.}Just one year after the discovery of general

46: relativity, Einstein pointed out that quantum effects must lead to

47: modifications in the theory of general relativity \cite{E1}. The

48: problems of compatibility between gravity and quantum mechanics were further

49: considered by Heisenberg \cite{H1}. He realized that since the

50: gravitational coupling constant is dimensional, a theory which

51: quantizes gravity will have serious problems. Indeed in the

52: seventies, t'Hooft and Veltman as well as Deser and Van

53: Nieuwenhuizen \cite{tH1}, confirmed that the quantum theory of

54: gravity coupled with matter has non-renormalizable divergences.

55: This was disappointing for a quantum theory of gravity. Indeed it

56: showed that the old techniques of quantization could not work for

57: general relativity. Soon after, Hawking \cite{H2} discovered that

58: a black hole is not cold, but thanks to quantum effects emits

59: radiation with temperature

60: \begin{equation}

61:    T=\frac{\hbar c^3}{8\pi kGM}.

62: \end{equation}

63: Then Unruh \cite{U1} proposed a relation between accelerated

64: observers, quantum theory, gravity and thermodynamics. This

65: suggested new profound relations between quantum and classical

66: worlds. The idea was that a full quantum gravity description must

67: reproduce these relations.

68: 

69: A connection between classical gravity and quantum field theory

70: was first seen in the seventies.

71: In the late sixties indeed, Veneziano \cite{V1}, trying to understand why the theoretically

72: expected amplitude divergences did not appear when high

73: energy elastic scattering of baryons produced

74: particles with higher spins, introduced a duality between the Mandelstam

75: variables $s$ and $t$ of the process \cite{IZ}. In this

76: way the divergences in one channel for higher spins, were

77: exactly cancelled out by the other channel.

78: The Veneziano theory had some peculiarities

79: such as the existence of

80: a massless spin-two particle, which did not

81: correspond to any renormalizable quantum theory. This particle was

82: then recognized as the graviton.

83: Indeed several years later, people realized that the phenomenological

84: model of Veneziano could be derived from a more fundamental

85: theory. The idea was to quantize one dimensional objects (strings)

86: instead of point-like objects (particles). Interpreting the massless spin-two

87: particle as the graviton, it was

88: conjectured that string theory leads to a non-divergent theory of

89: quantum gravity.

90: As well as producing exciting insights,

91: the introduction of strings implied also the existence of extra spatial dimensions, which was

92: at that time considered to be a problem.

93: 

94: In the eighties Polyakov \cite{P1} showed that the extra

95: dimensions were linked to the necessity of keeping the classical

96: symmetries for this action at the quantum level.

97: 

98: Later works showed also that string theory provides a consistent

99: theory at perturbed level. The aim of this theory was to unify all

100: the forces without introducing phenomenological coupling

101: constants.

102: 

103: Another important concept was developed by t'Hooft \cite{tH2} and

104: promoted by Susskind \cite{S1}. They proposed that the information

105: of a physical state in the interior of a region can be represented

106: on the region's boundary and is limited by the area of this

107: boundary. This was motivated to explain why the entropy of a black

108: hole (which contains all the information of

109:  this object) depends only on the area of its horizon and not on

110:  the volume inside the horizon. This principle was called

111: the ``Holographic principle''. The holography was a tentative way to connect

112: classical physics with quantum physics in a purely geometrical

113: way. In 1998 Maldacena \cite{M1} applied this concept in string

114: theory. Soon after, Witten \cite {W2} clarified the holographic

115: features of anti-deSitter spacetimes which was the basis for the

116: Maldacena work. The concept was that a conformal quantum field

117: theory (CFT) on a boundary of a spacetime with anti-deSitter

118: background can be described as a classical gravitational theory.

119: This was later called the AdS/CFT correspondence.

120: 

121: A sociological event was also born with string theory. A dialogue

122: between General Relativity

123: and Quantum Physics communities reemerged.

124: 

125: But, as Rovelli explains \cite{R1}, there are still profound

126: divergences. From the point of view of the General Relativity

127: community, quantum field theory is problematic. Up to now indeed,

128: quantum physics with gravity is consistent only for fixed

129: spacetime backgrounds. This implies that it is inadequate for a

130: full understanding of a dynamical theory of spacetime. On the

131: other side, for the Quantum Mechanics community, General

132: Relativity is only a low energy limit of a much more complex

133: theory, and thus cannot be taken too seriously as an indication of

134: the deep structure of Nature.

135: 

136: A bridge between the two approaches must be found. One possible way is to study the

137: dynamics of gravitational theories inspired by quantum gravity.

138: 

139: If one follows the direction of string theory and the

140: holographic principle, one must find a mechanism to reduce the dimensions of the

141: spacetime to four, at least at low energies.

142: 

143: One possibility was introduced in 1999 by Randall and Sundrum

144: \cite{RS}, where the extra dimension is infinitely large and the

145: matter is trapped on a four-dimensional submanifold called the

146: ``brane''. The graviton is instead localized on the brane only at

147: low energies, reproducing the correct Newtonian limit. This is

148: possible thanks to a negative cosmological constant which away from

149: the brane induces an anti-deSitter spacetime.

150: 

151: In this thesis I will discuss the astrophysical and cosmological implications

152: of this mechanism for simple phenomenological set ups.

153: 

154: I use the signature $(-,+,+,+)$ and $(-,+,+,+,+)$ for the

155: four and five dimensional Lorentzian manifold, the natural

156: units $c=\hbar=1$ and the definition of the Ricci tensor

157: $R^\alpha{}_{\mu\alpha\nu}=R_{\mu\nu}$, where $R_{\alpha\mu\nu\beta}$ is the

158: Riemann tensor.

159: 

160: \mainmatter

161: \newpage

162: \chapter{Extra dimensions: motivations}

163: 

164: To give a flavour of how extra dimensions appear in modern

165: theoretical physics, I will describe simple examples in the

166: context of String Theory and Holography. Since this chapter must

167: be seen as only as a motivation for the study of extra

168: dimensions, I will not directly connect these theories with

169: the Randall-Sundrum mechanism. I will anyway discuss throughout

170: the thesis, the influences that these theories have in the

171: Randall-Sundrum-type models.

172: 

173: \section{String Theory}

174: 

175: Currently the most promising theory for the unification of all the

176: forces seems to be superstring theory. There are actually five

177: anomaly-free perturbative string theories which are: type I, type

178: IIA, type IIB, SO(32) and $E_8\times E_8$ heterotic theories

179: \cite{GSW,PO}. These are all supersymmetric and require in general

180: ten dimensions. Another interesting theory is supergravity. In

181: eleven dimensions it is unique and it has been proved that its

182: compactification in ten dimensions reproduces the low energy limit

183: of the type IIA superstring theory. This was the starting point to

184: conjecture that a more general theory, whose low energy limit is

185: the eleven-dimensional supergravity \cite{SU}, is actually the

186: ultimate theory. This concept was also reinforced thanks to the

187: evidence that the superstring theories are related to each other

188: by dualities. The ultimate theory was called M-Theory

189: \cite{MT1,MT2,MT3,MT4}.

190: 

191: String theory contains an infinite tower of massive states,

192: corresponding to the oscillations of the string. The massless

193: states are separated from the massive ones by a gap of energy of

194: order $1/\sqrt{\alpha'}$. $T=1/{2\pi\alpha'}$ is called the string

195: tension, which is usually taken to be close to the Planck scale.

196: This is the only arbitrary parameter of the theory. In the limit

197: of an infinite tension, the massless excitations decouple from

198: the massive ones and the theory is described by the low energy

199: limit which in the effective action contains the usual

200: Einsteinian gravity. The low energy limit, more geometrically, is

201: dictated by the comparison between the curvature radius of the spacetime and the

202: string length. This means that the low energy limit is for

203: $R^{-1/2}\gg \sqrt{\alpha'}$ where $R$ is the Ricci scalar.

204: 

205: \subsection{Classical string action and equation of motion}

206: 

207: \subsubsection{String action in flat background}

208: 

209: In this section I give a taste of how the extra dimensions arise in string theory.

210: A complete treatment of bosonic and

211: supersymmetric string theories can be found in \cite{GSW,PO}.

212: 

213: The motion of a free falling relativistic particle is described by

214: the maximum spacetime length from an event $P_1$ to $P_2$

215: \cite{Wald}. The spacetime length is described by the integral \be

216: {\cal L}=-m\int^{P_2}_{P_1} ds, \ee which depends on the path,

217: where $m$ is the mass of the particle and \be

218: ds^2=g_{\alpha\beta}dx^\alpha dx^\beta, \ee where

219: $g_{\alpha\beta}$ is the spacetime metric.

220: 

221: Supposing the particle is massive ($m>0$), we can define the

222: proper time $\tau$ such that \be \frac{ds}{d\tau}=-1. \ee With

223: this parameter we can rewrite the total length in terms of the

224: particle four velocity $u^\alpha=dx^\alpha/d\tau$, as \be {\cal

225: L}=-m\int^{\tau_2}_{\tau_1}\sqrt{-u^\alpha u_\alpha} d\tau. \ee

226: 

227: Now we introduce a one-dimensional object instead of a point-like

228: one, i.e. a string. With the same logic we can say that the motion

229: of a string is described by the extremal area covered from a curve

230: $S_1$ to $S_2$. So the action for a string is \be {\cal

231: A}=-\frac{1}{2\pi\alpha'}\int^{S_2}_{S_1} dA, \ee where

232: $T=1/2\pi\alpha'$ is the tension of the string.

233: 

234: Since the string is a two-dimensional spacetime object, we can

235: describe its motion as a two-dimensional sheet. This sheet can be

236: described by two internal coordinates. We use the proper time

237: $ \tau $ and the proper spatial length $ \sigma $. Of course the

238: metric which describes this sheet in spacetime is (with signature

239: $(-,+)$) \be h_{ab}{}^{\mu\nu}=\partial_a x^\mu \partial_b x^\nu,

240: \ee where the Latin letters $a,b,..$ are the internal coordinates

241: $(\sigma,\tau)$. This is just the generalization of the point-like

242: four velocity $u^\alpha=\partial_\tau x^\alpha$. Indeed we have

243: this simple scheme: \ba

244: &\mbox{particle}\rightarrow\mbox{string}\cr\nonumber\cr\nonumber

245: &\sqrt{-u^\alpha

246: u_\alpha}\rightarrow \sqrt{-\mbox{det}(h_{ab}{}^\alpha{}_\alpha)}

247: =\sqrt{-\mbox{det}(h_{ab})}.\nonumber\cr\nonumber \ea Then finally

248: the action for a string is \be\label{ng} {\cal

249: A}=-\frac{1}{2\pi\alpha'}\int^{\tau_2,\sigma_2}_{\tau_1,\sigma_1}

250: \sqrt{-\mbox{det}(h_{ab})}d\sigma d\tau. \ee This is called the

251: Nambu-Goto action. From now on $\mbox{det}(h_{ab})=h$ and the

252: surface with boundaries $(\sigma_1,\tau_1)$ and

253: $(\sigma_2,\tau_2)$ is denoted by $M$.

254: 

255: Since we are interested in the quantum theory of strings, we need

256: to manipulate the action so that it appears in a more ``linear"

257: form. In this way we can use the standard quantization rules for

258: non-linear sigma models. To do that we introduce an auxiliary

259: metric $\gamma^{ab}(\sigma,\tau)$ and rewrite eq. (\ref{ng})

260: as \be\label{polact} {\cal A}=-\frac{1}{4\pi\alpha'}\int_M d\tau

261: d\sigma \sqrt{-\gamma}\gamma^{ab}h_{ab}. \ee Then we can see under

262: which conditions this action is equivalent to the Nambu-Goto one.

263: If $\gamma_{ab}$ is an auxiliary metric, then the variation of the

264: action with respect to it must vanish. Varying the action we get

265: the equation \be h_{ab}=\frac{1}{2}\gamma_{ab}\gamma^{cd}h_{cd},

266: \ee from which follows

267: \be

268: \sqrt{-h}=\sqrt{-\gamma}\gamma_{ab}h^{ab}. \ee

269: 

270: The action (\ref{polact}) is called the Polyakov action. This

271: action is invariant under the worldsheet ($(\sigma,\tau)$-space)

272: diffeomorphism

273: \begin{eqnarray}

274: &\delta x^\mu=\pounds_{\xi}x^\mu=\xi^a\partial_a x^\mu\ ,\\

275: &\delta \gamma^{ab}=\pounds_{\xi}\gamma^{ab}=\xi^c\partial_c

276: \gamma^{ab}- 2\gamma^{c(a}\partial_c\xi^{b)},

277: \end{eqnarray}

278: and the two-dimensional Weyl invariance \be \delta

279: \gamma^{ab}=\omega \gamma^{ab}. \ee

280: Here $\xi^a$ and $\omega$

281: are small vector and scalar parameters.

282: Of course we have also the

283: spacetime Poincar\'e invariance

284: \be \delta

285: x^\mu=\omega^{\mu\nu}x_\nu+b^\mu\ , \ee

286: where $\omega^{\mu\nu}$ represent a rotation and $b^\mu$ a translation in spacetime.

287: 

288: Now $\gamma^{ab}$ is actually the metric for the gravitational

289: field on the worldsheet, so we can use it to raise or lower the

290: worldsheet indices.

291: 

292: In general in the action (\ref{polact}) we could add a term which

293: describes the field equation for it. In two dimensions (the

294: worldsheet) the only possibility is to add a cosmological

295: constant, since the Ricci scalar is a total derivative, and so

296: therefore does not affect the equations of motion. However this

297: breaks the Weyl invariance, which is a key point to quantize the

298: string, and so it is set to zero.

299: 

300: Since the action is invariant under variations of $\gamma^{ab}$,

301: the energy-momentum tensor associated with it,

302: \be

303: T_{ab}=\partial_a x^\mu\partial_b x_\mu-\half

304: \gamma_{ab}\partial^cx^\mu\partial_cx_\mu\ , \ee

305: must vanish. In

306: particular its trace is zero. This is a consequence of the

307: conformal invariance (Weyl) of the action. We will see that, in

308: general, quantum mechanically the trace of this tensor does not

309: vanish. In order to keep this symmetry we will need in general

310: extra-dimensions.

311: 

312: Thanks to the classical Weyl and Poincar\'e symmetries we can locally choose the

313: worldsheet metric. We use the Minkowski one \be

314: \gamma_{ab}=\eta_{ab}=\mbox{diag}(-1,1). \ee Since strings are

315: one-dimensional objects they can be open or closed. We will see

316: that in the first case we have some boundary conditions to impose.

317: The variation of the action with respect to the fields

318: $x^\mu(\tau,\sigma)$ is \be -4\pi\alpha'\delta{\cal A}=\int_M

319: d^2\sigma \partial_a x^\mu \partial^a \delta x^\mu, \ee where we

320: used $\delta\partial_a x^\mu\simeq\partial_a\delta x^\mu$ and

321: $d\sigma d\tau=d^2\sigma$.

322: 

323: Integrating by parts we get

324: \be -4\pi\alpha'\delta{\cal A}=\int_M

325: d^2\sigma \partial _a (\eta^{ab}\partial _b x^\mu \delta

326: x_\mu)-\int_M d^2\sigma \delta x_\mu \Box x^\mu\ . \ee

327: If the string

328: is closed the first integral is zero. So the equations of motion

329: are \be \label{eqm} \Box

330: x^\mu=(\partial^2_\tau-\partial^2_\sigma)x^\mu(\tau,\sigma)=0. \ee

331: For an open string the first integral is \be \int_M d^2\sigma

332: \partial _a (\eta^{ab}\partial _b x^\mu \delta

333: x_\mu)=\oint_{\partial M}d\sigma_a\eta^{ab}\partial_b x^\mu\delta

334: x_\mu. \ee

335: We can choose $\partial M$ to be a space-like boundary,

336: so that $d\sigma_a=d\tau\delta^\sigma_a$. Normalizing the length

337: of the string such that $0\leq \sigma\leq\pi$, in order to have

338: zero variation we can consider a combination of the following

339: possibilities for the ends of the string:

340: 

341: {\it Neumann boundary conditions} \ba

342: \partial_\sigma x^\mu\Big|_\pi=0,\ \ \partial_\sigma x^\mu\Big|_0=0.

343: \ea

344: 

345: {\it Dirichlet boundary conditions} \ba

346: x^\mu\Big|_\pi=\mbox{const}\Rightarrow\delta x^\mu\Big|_\pi=0,\ \

347: x^\mu\Big|_0=\mbox{const}\Rightarrow\delta x^\mu\Big|_0=0. \ea

348: 

349: \subsubsection{General string action}

350: 

351: The string action can be generalized in the following way

352: \cite{GSW}:

353: \be\label{coupled} {\cal S}=-\pre \int d^2\sigma

354: \left[\sqrt{-\gamma}\gamma^{ab}g_{\mu\nu}(x^\alpha)\partial_a

355: x^\mu\partial_b x^\nu+\epsilon^{ab}B_{\mu\nu} (x^\alpha)\partial_a

356: x^\mu \partial_b x^\nu\right]. \ee

357: The coupling functions

358: $g_{\mu\nu}$ and $B_{\mu\nu}$ can be identified as the background

359: spacetime graviton and antisymmetric tensor fields in which the

360: string is propagating. $\epsilon^{ab}$ is the two-dimensional

361: Levi-Civita symbol. The coupling of these tensors with the string

362: fields is well justified also at the quantum level. At quantum

363: level we also get a massless scalar field in the spectrum. Such a

364: field in general breaks the Weyl invariance. Fradkin and Tseytlin

365: \cite{F1} have suggested that one should add to the string theory

366: action the renormalizable but not Weyl invariant term \be

367: S_{dil}=\frac{1}{4\pi}\int d^2\sigma

368: \sqrt{\gamma}R^{(2)}\Phi(x^\alpha), \ee where $R^{(2)}$ is the

369: two-dimensional Ricci scalar and the scalar field $\Phi(x^\alpha)$

370: is called the dilaton. Now it is true that at a classical level

371: this term breaks the conformal symmetry. Since string theory is a

372: quantum theory, we actually can require the weak condition that

373: the Weyl anomaly is cancelled at least at quantum level.

374: 

375: From now we use the following definition of total string action

376: \be

377: I[x,\gamma]={\cal S}+S_{dil}\ .

378: \ee

379: In the next

380: section we show how the quantization of this action leads to the Weyl anomalies.

381: 

382: 

383: \subsection{Quantization of the string action}

384: 

385: As an example of how extra-dimensions appear in string theory we first consider

386: the simplest background in which $\Phi=0$.

387: 

388: A modern concept of quantization is to introduce the partition function via the path

389: integral

390: \be

391: Z=\int {\cal D}\gamma(\sigma){\cal D}x(\sigma)e^{-I[x,\gamma]}\ ,

392: \ee

393: where ${\cal D}f$ means the integration over all the possible functions $f$.

394: 

395: Since Weyl symmetry of the classical action we can restrict ourselves in considering

396: \be

397: \gamma_{ab}=e^\phi\eta_{ab}\ .

398: \ee

399: In this way we can try to make explicit the measure ${\cal D}\gamma(\sigma)$ as the integration

400: of all the possible reparameterizations of the worldsheet metric. This is possible by introducing

401: auxiliary fields that basically fix the gauge for any choice of $\phi$ under integration.

402: These fields, called

403: Faddeev-Popov ghosts ($b,c$), are anticommuting \cite{GSW}.

404: Using the fact that the partition function can be rewritten as

405: \be

406: Z=\int {\cal D}\phi(\sigma)\int {\cal D}x(\sigma){\cal D}b(\sigma) {\cal D}c(\sigma)

407: e^{-I[x,b,c]}\ ,

408: \ee

409: where now the effective action becomes, in complex coordinates ($ds^2=dzd\overline z$)

410: \be

411: I[x,b,c]=I[x,\gamma]-\frac{1}{2\pi\alpha'}\int d^2\sigma \left[b_{zz}\nabla_{\overline z}

412: c^z+ \mbox{c.c.} \right]\ ,

413: \ee

414: the ghost $c^z$($c^{\overline z}$) is a holomorphic (antiholomorphic) vector and the

415: antighost $b_{zz}$($b_{\overline z\overline z}$) is an holomorphic (antiholomorphic) quadratic

416: differential.

417: 

418: The effective action $I[x,b,c]$ is classically conformal invariant.

419: In order to keep this invariance at the quantum level, we have to show under which conditions

420: the product of the measures ${\cal D}x{\cal D}b {\cal D} c$ and $I[x,b,c]$

421: are quantum Weyl invariant. Considering the

422: quantum fluctuations of the string and ghost fields as gaussian, because of their

423: dependence on the

424: worldsheet metric, we obtain, under rescaling

425: $\eta_{ab}\rightarrow e^{\xi} \eta_{ab}$, \cite{Ginsparg2}

426: \be\label{poly}

427: {\cal D}_{e^{\xi}\eta}x {\cal D}_{e^\xi\eta}[\mbox{ghost}]=\exp\left(\frac{D-26}{48\pi}S_L(\xi)

428: \right){\cal D}_{\eta}x{\cal D}_\eta[\mbox{ghost}]\ ,

429: \ee

430: where $S_L(\xi)$ is known as the Liouville action

431: (see \cite{Ginsparg2} for the explicit form) and

432: $D$ is the spacetime dimension.

433: Then in order to keep Weyl invariance at the quantum level, the first requirement is that

434: $D=26$.

435: 

436: Considering now a non vanishing $\Phi$, to see under what conditions $I[x,b,c]$ is quantum scale invariant,

437: one can calculate the quantum trace of the renormalized energy-momentum tensor of

438: the string, defined as

439: \be

440: \gamma^{ab}\frac{\delta \ln Z}{\delta \gamma_{ab}}\Big|_{\rm ren}=

441: i\gamma^{ab}\langle T_{ab}\rangle\ .

442: \ee

443: The renormalized result is \cite{C1}

444: \ba 2\pi

445: \langle T^a{}_a\rangle=\beta^\Phi\sqrt{-\gamma}R^{(2)}+\beta_{\mu\nu}^g\sqrt{\gamma}\gamma^{ab}

446: \partial_ax^\mu\partial_b

447: x^\nu+\beta_{\mu\nu}^B \epsilon^{ab}\partial_ax^\mu\partial_b

448: x^\nu, \ea

449: where $\beta^\Phi$, $\beta^g$ and $\beta^B$ are local

450: functionals of the coupling functions $\Phi$, $g_{\mu\nu}$ and

451: $B_{\mu\nu}$, and in the limit $R^{(2)-1}\sqrt{\alpha'}\ll

452: 1$,

453: \ba \beta^g_{\mu\nu}=

454: R_{\mu\nu}+2\nabla_\mu\nabla_\nu\Phi-\frac{1}{4}H_{\mu\rho\sigma}

455: H^{\rho\sigma}_\nu+O(\alpha')\label{noncri1}\\

456: \beta^B_{\mu\nu}=\nabla_\lambda H^\lambda_{\mu\nu}-2(\nabla_\lambda\Phi)H^\lambda_{\mu\nu}+

457: O(\alpha')\\

458: \beta^\Phi=\frac{D-D_c}{48\pi^2}+\frac{\alpha'}{16\pi^2}\left\{4(\nabla\Phi)^2-4\nabla^2\Phi-

459: R+\frac{1}{12}H^2\right\}

460: +O(\alpha'^2).\label{noncri2} \ea

461: Here

462: $H_{\mu\nu\lambda}=3\nabla_{[\mu}B_{\nu\lambda]}$, and $D_c=26$ is the number of

463: critical dimensions. In the beta function for the Weyl

464: anomaly $\beta^\Phi$, the leading term for bosonic strings was

465: discovered by Polyakov \cite{P1} in a similar way as we did in finding (\ref{poly}).

466: The graviton part was found by Friedan et

467: al. \cite{F2} and the $H$-fields by Witten \cite{W1} and

468: Curtright and Zachos \cite{C2}. In order to keep the Weyl symmetry at quantum level, we

469: have to make $\beta^\Phi$, $\beta^g$ and $\beta^B$ vanish.

470: 

471: It is important to note that since the coefficient of

472: $R^{(2)}\Phi$ is smaller by a factor $\alpha'$ than the other

473: couplings, its {\it classical} contribution is of the same order

474: as the one-loop {\it quantum contribution} of the $g_{\mu\nu}$ and

475: $B_{\mu\nu}$ couplings. This is because $R^{(2)}\Phi$ is scale

476: non-invariant at the classical level, while the other couplings

477: only lose scale-invariance at the quantum level. It is very simple

478: to prove that $\beta^\Phi$ is actually a constant in space time.

479: Indeed applying the Bianchi identities to $\beta^g$ and $\beta^B$

480: we get \be \nabla_\mu \beta^\Phi=0. \ee Therefore once we have solved the

481: equations for $\beta^{G,B}$, $\beta^\Phi$ is determined up to a

482: constant.

483: 

484: From eqs. (\ref{noncri1}-\ref{noncri2}) we can already argue that

485: string theory

486:  does not have to live in critical

487: dimensions. This kind of String theory is called non-critical

488: string theory \cite{PO}. In the $\beta^\Phi$ function, even if

489: the correction to the critical theory with $D=D_c$ is at first

490: order in $\alpha'$, it is actually possible to solve the equation

491: consistently to the one-loop correction. Myers \cite{M2} indeed

492: found that at least for the bosonic string in flat spacetime, where

493: $D_c=26$, a consistent solution is possible. This is

494: \ba

495: B_{\mu\nu}=0\ ,\ \Phi(x^\mu)=V_\mu x^\mu, \ea

496: where

497: \ba V_\mu

498: V^\mu=\frac{26-D}{6\alpha'}. \ea

499: This solution is compatible with

500: any dimension of the spacetime. There are not any other

501: anomaly-free non-critical string theories known.

502: 

503: Introducing supersymmetry in the string

504: action, which is a natural way to consider matter fields, one can prove

505: that $D_c=10$

506: instead of $D_c=26$ in eq. (\ref{noncri2}) \cite{C1}.

507: The bosonic sector of the supersymmetric string still satisfies

508: equations (\ref{noncri1}-\ref{noncri2}). These can be described by an

509: effective action \cite{M3}

510: \ba S_D=\frac{1}{2\kappa^2_D}\int d^D x

511: \sqrt{-g}e^{-\Phi}\left[ R+(\nabla\Phi)^2-\frac{1}{12}

512: H^2-\frac{2(D-10)}{3}\right].

513: \ea

514: Since the $\beta^\Phi$ function is at

515: order $\alpha'$, we can consistently consider the $\alpha'$

516: corrections to the $\beta^g_{\mu\nu}$ function as well. This

517: calculation can be done in heterotic string theory, which seems

518: to contain the gauge group of the Standard Model \cite{G1}. If we

519: choose a background with $B_{\mu\nu}=0$, the effective action then

520: becomes \cite{BoDe:85}

521: \be S_D=\frac{1}{2\kappa^2_D}\int d^Dx

522: \sqrt{-g}e^{-\Phi}\left[

523: R+(\nabla\Phi)^2-\frac{2(D-10)}{3}-\frac{\alpha'}{8}

524: \left\{L_{GB}-(\nabla\Phi)^4 \right\}\right], \ee

525: where

526: $L_{GB}=R^{\mu\nu\alpha\beta}R_{\mu\nu\alpha\beta}-4R^{\mu\nu}R_{\mu\nu}+R^2$

527: is the Gauss-Bonnet term.

528: 

529: Applying a conformal transformation

530: $g_{\mu\nu}\rightarrow e^{\frac{4}{D-2}\Phi}g_{\mu\nu}$ we can rewrite the

531: action in a more familiar form \cite {GBE,M3}

532: \ba

533: S_D=\frac{1}{2k^2_D}\int d^Dx \sqrt{-g}[

534: R&-&\frac{4}{D-2}(\nabla\Phi)^2-\frac{2(D-10)}{3}

535: e^{\frac{4}{D-2}\Phi}+\cr\nonumber

536: &+&\frac{\alpha'}{8}e^{-\frac{4}{D-2}\Phi}(L_{GB}+4\frac{D-4}{(D-2)^3}(\nabla\Phi)^4)].

537: \ea

538: This frame is called the Einstein frame. At this level it is

539: explicit that string theory contains, as effective low energy

540: theory, the generalization of the Einstein theory in extra

541: dimensions, called the

542: Lovelock theory of gravity \cite{Lov}, when $\Phi$ is constant.

543: 

544:     \section{Holography}

545: The idea of Holography is to relate quantum physics on a spacetime boundary with

546: classical geometrical properties of the spacetime.

547: 

548: Initially the aim of this theory was to understand the quantum physics of black holes.

549: Indeed when a black hole is formed,

550: classical and quantum physics encounter each other at its boundary, the horizon.

551: In particular all the quantum degrees of

552: freedom live holographically on the horizon. This arises from the study of the entropy

553: of a black hole. We know that the entropy is

554: the measure of the degrees of freedom of a physical state. Applying covariant quantization and

555: the analogies between thermodynamics and

556: black hole physics, Bekenstein and then Hawking \cite{holo1,holo2,holo3,H2} discovered

557: that the entropy of a black hole

558: is proportional to its area $A$:

559: \begin{equation}

560: S=\frac{A}{4G_N}.

561: \end{equation}

562: Quantum mechanically this was a breakthrough. Indeed quantum field theory (QFT)

563: gives an estimation of entropy

564: proportional to the volume and not to the area of a physical object.

565: This seems to say that QFT lives only on the horizon. Consider a spherical

566: region $\Gamma$ of volume $V$ in an asymptotically flat spacetime.

567: The boundary $\delta \Gamma$ has area $A$. The maximal entropy is

568: defined by

569: \begin{equation}

570: S_{max}=\ln N_{states}

571: \end{equation}

572: where $N_{states}$ is the total number of possible states of $\Gamma$.

573: Since we are considering gravitational objects, the number of states should

574: be linked with the degrees of freedom of the spacetime.

575: Suppose we consider that, at each fixed time, the space can be quantized. This means

576: that we divide it into elementary cells of width $\alpha$. Each of these cells store

577: the local information of a physical state. Moreover, assuming that for each cell

578: we have $m$ possible states, we get approximately

579: \begin{equation}

580: N_{states}=m^{V/\alpha^3}\ .

581: \end{equation}

582: This implies that the maximum entropy is

583: \begin{equation}

584: S_{max}\propto V.

585: \end{equation}

586: Now we consider a static star of energy $E$ and radius $R$. A static equilibrium implies that

587: the energy of the star $E$ must be bounded,

588: $E<M$. Where $M=R/2$ is the maximum mass one can fit in a static sphere or radius $R$.

589: The second law of thermodynamics tells us that an increase in energy implies an increment of

590: entropy. The maximum entropy will

591: then be reached when the black hole is formed. Indeed after the formation of a black hole,

592: all the degrees of freedom inside

593: the horizon are causally disconnected from the exterior, so that they

594: cannot be taken into account. This means that the maximum entropy is the

595: black hole entropy

596: \begin{equation}

597: S_{max}=\frac{A}{4G_N}.

598: \end{equation}

599: If we believe that we can calculate the entropy of a star by simple QFT calculations, we

600: implicitly break unitarity

601: of the quantum theory, losing predictability  \cite{predict}. This is because there is a

602: transition of the number of states

603: from the collapsing star ($\ln N_{states}\propto V$) to the black hole ($\ln N_{states}

604: \propto A$). If instead

605: we accept that the maximum entropy of a

606: spatial region is proportional to the area of its boundary, rather then its volume, then we

607: can retain unitarity in the collapse process.

608: This is how the holographic principle was first formulated. In the extended version this

609: principle states that

610: for any Lorentzian

611: manifold it is possible to find a submanifold (screen) where all the quantum degrees of

612: freedom are present \cite{Bousso}.

613: The AdS/CFT correspondence \cite{M1} is a specific and explicit example. Here the screen

614: is identified as the AdS boundary.

615: In this holographic correspondence, a quantum conformal field theory on the screen,

616: can be

617: described as a boundary effect of a classical geometrical theory of spacetime.

618: If we believe seriously in this principle,

619: the evidence

620: that our world is described by quantum degrees of freedom, makes the study of gravity

621: in higher dimensions encouraging. In the following

622: I describe how the relationship between quantum physics and classical

623: gravity arises in an anti de Sitter manifold (AdS), which is

624: the geometry in which the Randall-Sundrum mechanism is constructed.

625: 

626:         \subsection{AdS/CFT correspondence}

627: This test of the holographic principle was developed in particular string theories.

628: Here it has been shown

629: that on $D$-dimensional AdS backgrounds there is a correspondence between a classical

630: perturbed $D$-dimensional super-gravity theory and a

631: $(D-1)$-dimensional super-conformal field theory.

632: This duality was then argued to be present also

633: in deformations of AdS, such as the generalization of the Randall-Sundrum model, where the

634: quantum theory of the boundary was not necessarily

635: a conformal field theory (for a very nice introduction to the topic see \cite{padilla}).

636: In order to have the flavour of this

637: duality, I give a simple

638: example of AdS/CFT correspondence. Here I review the result that the zero

639: point energy of a CFT theory is actually described

640: by the boundary energy of an anti de Sitter spacetime \cite{BK}.

641: 

642: \subsubsection{The Brown-York tensor}

643: To define the energy of a manifold boundary we are going to use the definition of the

644: quasi-local energy given in \cite{BY}. The idea

645: is very simple and clear. We make an analogy with a classical non-relativistic system.

646: Suppose this system has an action $S$.

647: It satisfies the Hamilton-Jacobi equation $H=-\partial S/\partial t$,

648: where $H$ is the Hamiltonian of the system which describes the

649: energy and $t$ is the time. We would like now to generalize this concept for a gravitational

650:  system. Take a $D$ dimensional manifold

651: $M$ which can be locally described by a product of a $D-1$ dimensional space $\Sigma$ and

652: a real line interval. The

653: boundary $\partial \Sigma$ need not be simply connected. The product

654: of $\partial \Sigma$ with the real line

655: orthogonal to

656: $\Sigma$ will be denoted by $B$. By analogy with classical mechanics, the quasi-local

657: energy associated with the spacelike

658: hypersurface $\Sigma$, is defined as minus the variation of the action with respect to a unit

659: increase in proper time separation between

660: $\partial \Sigma$ and its neighboring $D-2$ surface, as measured orthogonally to $\Sigma$

661: at $\partial \Sigma$. This basically measures

662: the variational rate of the action on $B$. In order to define this splitting of spacetime

663: we will use the ADM \cite{ADM}

664: decomposition which has a global splitting (at least in the absence of singularities or null

665: surfaces). Here we can naturally define an

666: Hamiltonian and therefore an energy.

667: 

668: The spacetime metric is $g_{\mu\nu}$ and $n^\mu$ is the outward pointing spacelike unit

669: normal to the boundary $B$. The metric and

670: the extrinsic curvature of $B$ are denoted respectively by $\gamma_{\mu\nu}$ and

671: $\Theta_{\mu\nu}$.

672: Now denote by $u^\mu$ the future pointing timelike unit normal to a family of spacelike

673: hypersurfaces $\Sigma$ that

674: foliate the spacetime. The metric and the extrinsic curvature of $\Sigma$ are given by the

675: spacetime tensors $h_{\mu\nu}$ and

676: $K_{\mu\nu}$, respectively. $h^\mu_\nu$ is also the projection tensor on $\Sigma$. The

677: spatial coordinates $i,j,...=0,...,D-2$ are adapted

678: coordinates on $\Sigma$. We then define the momentum $P^{ij}$ conjugate to the spatial

679: metric $h_{ij}$. The ADM decomposition is

680: simply

681: \begin{equation}

682: ds^2=g_{\mu\nu}dx^\mu dx^\nu=-N^2dt^2+h_{ij}(dx^i+V^idt)(dx^j+V^jdt),

683: \end{equation}

684: where $N$ is the lapse function and $V^i$

685: the shift vector.

686: 

687: For our purposes we use the fact that the foliation $\Sigma$ is orthogonal to $B$, which implies

688: $(u\cdot n)\Big|_B=0$. Because at this restriction,

689: the metric at $B$ can be decomposed as

690: \begin{equation}

691: \gamma_{\mu\nu}dx^\mu dx^\nu=-N^2dt^2+\sigma_{ab}(dx^a+V^adt)(dx^b+V^bdt),

692: \end{equation}

693: where $a,b,...=0,...,D-3$ are adapted coordinates on $\partial \Sigma$.

694: 

695: For general relativity coupled to matter, consider first the action suitable for fixation

696: of the metric on the boundary \cite{York}

697: \begin{equation}

698: S=\frac{1}{2 \kappa}\int_M d^Dx \sqrt{-g} R+\frac{1}{\kappa}\int^{t_2}_{t_1}d^{D-1}x\sqrt{h}K-

699: \frac{1}{\kappa}\int_Bd^{D-1}x\sqrt{-\gamma}\Theta+S^m\ ,

700: \end{equation}

701: where $S^m$ is the matter action, including a possible cosmological term. Now the variation

702: of this action gives

703: \begin{eqnarray}

704: \delta S&=&(\mbox{terms giving the equations of motion})\cr

705: &+&(\mbox{boundary terms coming from the matter action})\cr &+&\int^{t_2}_{t_1}d^{D-1}

706: x P^{ij}\delta h_{ij}+\int_Bd^{D-1}x\pi^{ij}\delta\gamma_{ij}.

707: \end{eqnarray}

708: Here $\pi^{ij}$ is the conjugate momentum to $\gamma_{ij}$. We assume that the

709: matter action contains no derivatives of the metric. For the gravitational variables, the

710: boundary three-metric $\gamma_{ij}$ is

711: fixed on $B$, and the hypersurface metric $h_{ij}$ is fixed in $t_1$ and $t_2$. Since we are

712: interested in the variation of the action,

713: this will have an ambiguity in its definition. The ambiguity in $S$ is taken into account by

714: subtracting an arbitrary function of the

715: fixed boundary data. Thus we define the action

716: \begin{equation}

717: A=S-S^0,

718: \end{equation}

719: where $S^0$ is a functional of $\gamma_{ij}$. The variation in $A$ just differs from the

720: variation of $S$ by the term:

721: \begin{equation}\label{S0}

722: -\delta S^0=-\int_B d^{D-1}x\frac{\delta S^0}{\delta \gamma_{ij}}\delta\gamma_{ij}=-

723: \int_Bd^{D-1}x\pi_0^{ij}\delta \gamma_{ij},

724: \end{equation}

725: where it is clear that $\pi_0^{ij}$ is a function only of $\gamma_{ij}$. We call the

726: classical action $S_{cl}$ the action evaluated

727: on the classical solution of $A$. This will be of course a functional of the fixed boundary

728: data consisting of $\gamma_{ij}$,

729: $h_{ij}(t_1)$, $h_{ij}(t_2)$ and the matter fields. Then the variation of $S_{cl}$ among the

730: possible classical solutions give

731: \begin{eqnarray}

732: &\delta S_{cl}=(\mbox{terms involving variations in the matter fields})\cr \nonumber

733: &+\int^{t_2}_{t_1}d^{D-1}

734: x P^{ij}_{cl}\delta h_{ij}+\int_Bd^{D-1}x(\pi^{ij}_{cl}-\pi_{0}^{ij})\delta\gamma_{ij}.

735: \end{eqnarray}

736: The generalization of the Hamilton-Jacobi equation

737: for the momentum are the equations

738: \begin{equation}

739: P^{ij}_{cl}\Big|_{t_2}=\frac{\delta S_{cl}}{\delta h_{ij}(t_2)}.

740: \end{equation}

741: The generalization of the energy equation is

742: \begin{equation}\label{BY}

743: T^{ij}=\frac{2}{\sqrt{-\gamma}}\frac{\delta S_{cl}}{\delta\gamma_{ij}}\ ,

744: \end{equation}

745: which is the quasi-local stress-energy tensor for gravity.

746: 

747: \subsubsection{Holographic zero point energy of AdS}

748: 

749: In general, if the spacetime is not asymptotically flat the Brown-York tensor (\ref{BY})

750: diverges at infinity. For asymptotically AdS spacetime,

751: there is a resolution of this difficulty. If we believe there is an

752: holographic duality between AdS and CFT, the

753: fact that the Brown-York tensor diverges, it can be interpreted as the standard ultraviolet

754: divergences of

755: quantum field theory, and may be classically removed

756: by adding local counter-terms to the action. These subtractions depend only on the intrinsic

757: geometry of the boundary. Once

758: we

759: renormalize the energy in this way, we should be able via classical and quantum renormalization

760: schemes to obtain the same

761: zero point energy of the spacetime. The idea is to renormalize the stress-energy tensor by

762: adding a finite

763: series

764: of boundary invariants to the classical action.

765: The essential terms are fixed uniquely by requiring finiteness of the stress tensor. For

766: simplicity we will work it out in three dimensions,

767: but this is true in higher dimensions as well \cite{BK}.

768: In particular, global $\mbox{AdS}_5$, with an $S^3\times R$ boundary, has a positive

769: mass \cite{positive}.

770: This result is beautifully explained via the proposed

771: duality with a boundary CFT. The dual super Yang-Mills (SYM) theory on a sphere has a Casimir

772:  energy that precisely matches this spacetime mass.

773: 

774: The action we are considering now is the three-dimensional Einstein-Hilbert with negative

775: cosmological constant $\Lambda=-1/l^2$

776: and boundary terms

777: \begin{equation}

778: S=\frac{1}{2 \kappa}\int_M d^3x \sqrt{-g} \left(R-\frac{1}{l^2}\right)+

779: \frac{1}{\kappa}\int^{t_2}_{t_1}d^2x\sqrt{h}K-

780: \frac{1}{\kappa}\int_Bd^2x(\sqrt{-\gamma}\Theta)+S_{ct},

781: \end{equation}

782: where $S_{ct}$ is the counter-term action added in order to obtain a finite stress tensor.

783: In order

784: to preserve the equation

785: of motion, this extra term must be a boundary term. In particular it lives in $B$.

786: The Brown-York tensor is then

787: \begin{equation}

788: T^{ij}=\frac{1}{\kappa}\left[\Theta^{ij}-\Theta\gamma^{ij}+\frac{2}{\sqrt{-\gamma}}\frac{\delta

789: S_{ct}}{\delta\gamma_{ij}}\right],

790: \end{equation}

791: with everything calculated on the classical solution. Since $S_{ct}$ must only

792: depend on the boundary geometry, it describes a local

793: gravitational action in two dimensions. In particular, we wish to cancel out the divergences

794: coming from the negative energy from the

795: cosmological constant. Therefore the most general form is

796: \begin{equation}

797: S_{ct}=-\int_B\frac{1}{l}\sqrt{-\gamma}d^2 x\ \Rightarrow\ T^{ij}=\frac{1}{\kappa}\left[\Theta^{ij}-

798: \Theta\gamma^{ij}-\frac{1}{l}\gamma^{ij}\right].

799: \end{equation}

800: Now consider $\mbox{AdS}_3$ spacetime in light-cone coordinates

801: \begin{equation}\label{ads}

802: ds^2=\frac{l^2}{r^2}dr^2-r^2dx^+dx^-.

803: \end{equation}

804: In this case $T^{ij}=0$. What we would like to prove is that a dual CFT is living on the

805: surface $ds^2=-r^2dx^+dx^-$ with $r$

806: eventually taken to infinity (AdS boundary). If this is true, asymptotically conformal

807: perturbations of AdS

808: correspond to excitations of CFT fields on its boundary.

809: In particular, on the CFT side even if classically $T^i_i=0$, because quantum excitation

810: of the vacuum, $\gamma^{ij}<T_{ij}>\neq 0$. The scale of

811: this energy is linked to the renormalization scale. Connecting it with the

812: gravitational scale $l$ of AdS, we identify the Brown-York tensor on the boundary

813: of AdS as corresponding to the quantum CFT energy-momentum tensor on flat background.

814: In \cite{Brohe} Brown and Henneaux proved that a perturbation of the AdS metric with

815: asymptotic

816: AdS behaviour must have the following expansion

817: \begin{eqnarray}

818: \delta g_{+-}=O(1),\ \ \delta g_{++}=O(1),\ \ \delta g_{--}=O(1)\\

819: \delta g_{rr}=O\left(\frac{1}{r^4}\right),\ \ \delta g_{+r}=O\left(\frac{1}{r^3}\right),\ \

820: \delta g_{-r}=

821: O\left(\frac{1}{r^3}\right).

822: \end{eqnarray}

823: We rewrite this expansion using the following diffeomorphism

824: \begin{eqnarray}

825: x^+\rightarrow x^+-\xi^+-\frac{l^2}{2r^2}\partial^2_-\xi^-\ ,\label{diff1}\\

826: x^-\rightarrow x^--\xi^--\frac{l^2}{2r^2}\partial^2_+\xi^+\ ,\\\

827: r\rightarrow r+\frac{r}{2}(\partial_+\xi^++\partial_-\xi^-)\ ,\label{diff2}

828: \end{eqnarray}

829: where $\xi^\pm=\xi^\pm(x^\pm)$, which yields at

830: first order

831: \begin{equation}

832: ds^2=\frac{l^2}{r^2}dr^2-r^2dx^+dx^--\frac{l^2}{2}\partial^3_+\xi^+ (dx^+)^2-

833: \frac{l^2}{2}\partial^3_- \xi^-(dx^-)^2.

834: \end{equation}

835: Then we obtain the following non-zero components for the Brown-York tensor

836: \begin{equation}\label{grav}

837: \delta_{\xi^{\pm}}T_{\pm\pm}=-\frac{l}{2k}\partial_\pm^3\xi^\pm.

838: \end{equation}

839: This therefore represents the zero point energy of the boundary of AdS

840: due to perturbations of this spacetime that preserve the AdS symmetries on the boundary.

841: Now we study the

842: CFT side. The boundary of AdS is just a Minkowski spacetime with metric

843: \begin{equation}

844: ds^2=-r^2dx^+dx^-.

845: \end{equation}

846: A conformal theory is such that under conformal transformation of the type

847: $g_{\alpha\beta}\rightarrow \Omega \tilde g_{\alpha\beta}$, it is

848: classically invariant; then its energy momentum tensor satisfies $T^{\alpha\beta}=\tilde

849: T^{\alpha\beta}$. In particular this means that

850: classically $T^\alpha_\alpha=0$. If we now introduce a conformal transformation under

851: the diffeomeorphism (\ref{diff1}-\ref{diff2}), when

852: $r\rightarrow \infty$ we get quantum mechanically (starting again from a zero energy

853: momentum tensor)

854: \footnote{It

855: is beyond the scope of this thesis to prove this result; a very good introduction can be found

856: in \cite{ginsparg}.}

857: \begin{equation}\label{cft}

858: \delta_{\xi^\pm} T_{\pm\pm}=-\frac{c}{24\pi}\partial_\pm^3\xi^\pm.

859: \end{equation}

860: This term comes from the commutation rules.

861: In particular since this does not depend on the particular

862: energy-momentum tensor used, it is called the zero point energy.

863: The constant $c$ is called the

864: central charge and encode the renormalization scale of

865: the theory. Surprisingly eqs. (\ref{grav}) and (\ref{cft}) coincide if we measure

866: the scale of the renormalization with the AdS

867: scale, or $c=12\pi l/\kappa$. This means that the quantum vacuum energy, due

868: to the excitation of the fields in the CFT side, is nothing else than the residual

869: energy due to classical gravitational perturbations of AdS that preserve the AdS boundary symmetries.

870: 

871: \newpage

872: 

873: \chapter{A dynamical alternative to compactification}

874: 

875: The possibility that spacetime has more than four dimensions is an

876: old idea. Originally the extra dimensional spaces were

877: thought to be compactified. This kind of model is usually

878: called Kaluza-Klein compactification (see for example \cite{KK}).

879: Here, to have a correct Newtonian and Standard Model limit, the

880: size of the extra dimensions must be less than the Electroweak

881: scale ($ \sim 1\ \mbox{TeV}^{-1}$). These theories were introduced to

882: incorporate geometrically scalar and vector fields in the

883: projected four-dimensional gravity, thus giving a geometrical interpretation of electromagnetism.

884: 

885: Later, extra dimensions arose very naturally in string theory and associated phenomenology.

886: 

887: In this thesis I will discuss only models with an infinitely large extra

888: dimension. In particular, I am interested in models where

889: gravity is dynamically localized at low energies \cite{RS}.

890: Another type of localization is also possible \cite{DGP,GRS}, in which

891: gravity has the correct Newtonian limit

892: at short distances. This arises from introducing an induced

893: gravity on the four-dimensional submanifold. In this way, at short

894: distances, the induced effects dominate, generating an effective

895: four-dimensional Einstein gravity consistent with Newtonian gravity

896: experiments. At large distances, by contrast, the

897: extra-dimensional gravity becomes more and more important. This

898: modifies, at large distances, the effective four-dimensional

899: gravity.

900: 

901: The fact that particles can be trapped gravitationally in a submanifold,

902: was argued already by Visser \cite{Visser}. He realized that

903: a particle living in a five-dimensional spacetime with warped time component,

904: \begin{equation}\label{w1}

905: ds^2=-e^{-2\Phi(\xi)}dt^2+d{\vec x}\cdot d\vec x+d\xi^2,

906: \end{equation}

907: ($\xi$ is the extra-dimensional coordinate) is exponentially

908: trapped by the scalar field $\Phi(\xi)>0$. In 1999 Randall and

909: Sundrum \cite{RS} used a similar mechanism to localize gravity. In

910: order to do that, the whole four-dimensional metric must be warped.

911: In this way a curved background supports a ``bound state'' of the

912: higher-dimensional graviton with respect to the co-dimensions.

913: So although space is indeed infinite in extent, the

914: graviton is confined to a small region within this space.

915: 

916: Suppose for simplicity we have only one extra dimension

917: \footnote{The mechanism is general even with more than one co-dimension.}.

918: Moreover suppose the five-dimensional gravity is governed by the Einstein-Hilbert action.

919: We then consider a domain wall with

920: vacuum energy $\lambda$ where, as a macroscopic approximation, all the matter fields live.

921: The action is then

922: \begin{equation}\label{act}

923: S=\frac{1}{\tilde\kappa^2}\int d^5x\sqrt{-\tilde

924: g}\{\tilde R-2\tilde\Lambda\}+\int d^4x\sqrt{-g}\{{\cal L}_m-2\lambda\},

925: \end{equation}

926: where $\tilde \kappa$ and $\tilde g_{AB}$ are respectively the five

927: dimensional Planck mass and metric, $\tilde\Lambda$ and $\tilde R$ are the five

928: dimensional negative cosmological constant and Ricci scalar, ${\cal L}_m$ and

929: $g_{\mu\nu}$ are the four-dimensional Lagrangian for matter and

930: the four-dimensional metric.  The three-dimensional submanifold is

931: called the ``brane'' and the total spacetime will be

932: called the ``bulk''. We use the notation $A,B,...=0,...,4$ for

933: the bulk coordinates and $\mu,\nu,...=0,...,3$ for the brane

934: coordinates. The role of the five-dimensional cosmological

935: constant is very important. Physically, we can understand that

936: such a term will induce a positive ``pressure''. This localizes the

937: gravitons produced in the four-dimensional brane on the brane

938: itself. Indeed, as we shall see, the warp factor

939: is produced by the cosmological constant.

940: 

941: In order to simplify the system we introduce

942: $Z_2$-symmetry in the co-dimension. This can be also motivated from the orbifold structure

943: appearing in M-theory to

944: incorporate heterotic string theory. In this way M-theory contains the

945: gauge groups of the standard model of particle physics \cite{HW}.

946: 

947: We start with an empty brane, ${\cal L}_m=0$. We would like to find a warped

948: solution which has, as a slice, a Minkowski

949: spacetime. In the next section we will introduce the

950: general formalism in the presence of matter, then we will study small perturbations

951: around Minkowski, obtaining the correct Newtonian limit.

952: We start with the metric:

953: \begin{equation}\label{ansatz}

954: ds^2=e^{-2\sigma(y)}\gamma_{\mu\nu}(x^\alpha)dx^\mu dx^\nu+dy^2,

955: \end{equation}

956: where $y$ is the extra-dimensional coordinate and

957: $\gamma_{\mu\nu}(x^\alpha)$ is the four-dimensional metric. In order to

958: solve the variational problem $\delta S=0$, we have two ways. The first one is

959: to solve the Einstein equations together with junction

960: conditions between the brane and the bulk. Since the brane is a singular object, with this

961: approach we need to involve the theory of

962: distributions. We will discuss this technique in the next section and study it extensively

963: in the last chapter and Appendix \ref{Appjc}.

964: Here we follow another way. Instead of introducing a

965: thin shell (brane), we consider a spacetime with a boundary at

966: $y=0$. This means that we will allow the four-dimensional

967: metric to vary only on the boundary, whereas we will keep it fixed in the

968: bulk as Minkowski slices, in such a way that the global spacetime is AdS.

969: If we would like to

970: implement the $Z_2$ symmetry, we can just make a copy of the

971: same spacetime on the two sides of the boundary. In this way we

972: don't have to deal with distributions. Now if we substitute the

973: ansatz (\ref{ansatz}) into the action (\ref{act}), we are left

974: with a variational problem for the scalar field $\sigma(y)$ in the

975: bulk and a variational problem for the metric $\gamma_{\mu\nu}$ on

976: the brane. Then we use the constraint

977: $\gamma_{\mu\nu}=\eta_{\mu\nu}$. The

978: Lagrangian for gravity will be \cite{Wald}\footnote{I use the notation $\partial_y f= f'$.}

979: \begin{equation}

980: \sqrt{-\tilde g}\tilde R=12e^{-4\sigma}(\sigma')^2+2(K\sqrt{-g})'+e^{-2\sigma}{\cal R}\ ,

981: \end{equation}

982: where ${\cal R}$ is the Ricci scalar associated with $\gamma_{\mu\nu}$ and $K={1\over 2}

983: g^{\mu\nu}\pounds_n g_{\mu\nu}$,

984: is the trace of the extrinsic curvature orthogonal to the brane. Using this notation we get

985: \begin{equation}

986: \int d^5x\sqrt{-\tilde g}\tilde R=12\int d^5x \{e^{-4\sigma}(\sigma')^2+e^{-2\sigma}

987: {\cal R}\}+2\int d^4x K \sqrt{-g},

988: \end{equation}

989: where the boundary is at $y=0$. The last term is called the Gibbons-Hawking term \cite{GW}.

990: 

991: Since the degrees of freedom of the four-dimensional Ricci scalar and the degrees of freedom

992: of the scalar field $\sigma$ are now

993: completely separate, we can set ${\cal R}=0$ before the variation.

994: Considering only the first integral, since $\delta \sigma=0$ on the boundary we have

995: \begin{equation}

996: 12\delta\int d^5x e^{-4\sigma}(\sigma')^2=-12\int d^5x \{2\sigma''-(4\sigma')^2\}e^{-4\sigma}

997: \delta\sigma.

998: \end{equation}

999: The variation of the five dimensional cosmological constant is

1000: \begin{equation}

1001: -2\delta\int d^5x\sqrt{-\tilde g}\tilde\Lambda=12\int d^5x \frac{2}{3}\tilde\Lambda

1002: e^{-4\sigma}\delta\sigma.

1003: \end{equation}

1004: Combining the two variations we get the equation

1005: \begin{equation}

1006: \sigma''-2(\sigma')^2=\frac{\tilde\Lambda}{3}.

1007: \end{equation}

1008: This equation has two separate branches, one is the trivial

1009: \begin{equation}

1010: \sigma'=\pm\sqrt{-\frac{\tilde\Lambda}{6}}\ ,

1011: \end{equation}

1012: and the other is

1013: \begin{equation}

1014: \sigma'=\pm\sqrt{1+\tanh \left(2\sqrt{-\frac{\tilde\Lambda}{6}}y+C\right)}.

1015: \end{equation}

1016: To implement the $Z_2$-symmetry we have to cut and paste the solution from one side to the

1017: other of the boundary. The second solution does not have parity. Moreover, the Einstein tensor

1018: describing this system contains another constraint equation

1019: hidden in the integration process we are using here.

1020: One can show that the second solution does not satisfy this additional constraint (see the $\tilde G_{yy}$ equation in \ref{second}

1021: with $e^{-\sigma(y)}=a(y)$, and considering the static case).

1022: This implies that we can only use the first solution

1023: \begin{equation}\label{jcRS}

1024: \sigma'=\sqrt{-\frac{\tilde\Lambda}{6}}\mbox{sgn}(y).

1025: \end{equation}

1026: Here we introduced the sign function $\mbox{sgn}(y)$, in order to describe the full spacetime.

1027: In particular we choose the positive

1028: branch of the square root. This is because we want a decreasing warp factor.

1029: Now we have to make the variations of the boundary action with

1030: respect to $g_{\mu\nu}$ vanish, keeping $\partial_y g_{\mu\nu}$ fixed. We have

1031: the combination

1032: \begin{equation}

1033: \frac{1}{\tilde\kappa^2}\int

1034: d^4x\sqrt{-g}\left\{ K_{\mu\nu}-g_{\mu\nu}K-\frac{\tilde\kappa^2\lambda}{2}

1035: g_{\mu\nu}\right\} \delta g_{\mu\nu},

1036: \end{equation}

1037: using the constraint $\gamma_{\mu\nu}=\eta_{\mu\nu}$, this boundary action vanishes for

1038: \begin{equation}\label{jcRS2}

1039: \sigma'(0)=\frac{\tilde\kappa^2\lambda}{6}.

1040: \end{equation}

1041: Combining (\ref{jcRS}) with (\ref{jcRS2}) we get the fine tuning

1042: \begin{equation}\label{min}

1043: \frac{\tilde \kappa^4\lambda^2}{6}+\tilde\Lambda=0.

1044: \end{equation}

1045: We can now write down the global geometry. This is an AdS spacetime in the bulk and

1046: a Minkowski

1047: spacetime on the brane:

1048: \begin{equation}\label{Minkmetric}

1049: ds^2=e^{-2\sqrt{-\tilde\Lambda/6}\mid y\mid}\eta_{\mu\nu}dx^\mu dx^\nu+dy^2.

1050: \end{equation}

1051: In the next paragraph we show how

1052: generalize this model in the presence of matter.

1053: Moreover we will see that if the fine tuning (\ref{min}) does not hold, an

1054: effective four-dimensional cosmological constant appears.

1055: 

1056: \section{The Randall-Sundrum braneworld scenario}

1057: The mechanism provided by Randall-Sundrum can be generalized to include

1058: matter. We will follow

1059: \cite{SMS}. In this work a geometrical approach is used. We suppose that

1060: the bulk spacetime is governed by the Einstein

1061: equations and then we impose the junction conditions.

1062: 

1063: The normal to the brane is $n^A$. The induced metric on

1064: the brane is $g_{AB}= \tilde g_{AB}-n_A n_B$. The Gauss equation

1065: which relates the $D-$dimensional Riemann tensor to the

1066: $(D-1)$-dimensional one is\footnote{I will use the notation with

1067: tilde for five-dimensional objects and without for

1068: four-dimensional ones.}

1069: \begin{equation}\label{gauss}

1070: R^A{}_{BCD}=\five R^E{}_{FGH}\ g^{A}_{~E}\ g^{F}_{~B}\ g^{G}_{~C}\ g^{H}_{~D}\ +K^A{}_C K_{BD}

1071: -K^A{}_D K_{BC}

1072: \end{equation}

1073: where

1074: \begin{equation}

1075: K_{AB}=\frac{1}{2}\pounds_n \tilde g_{AB}

1076: \end{equation}

1077: is the extrinsic curvature of the brane. Then we have the Codacci equations

1078: \begin{equation}\label{codacci}

1079: g^{B}_{~C}\nabla_B K^{C}{}_A- g^{B}_{~A}\nabla_B K=\five R_{CD}n^C g^{D}_{~A},

1080: \end{equation}

1081: where we use the notation $K^A{}_A=K$. From the Gauss equation we

1082: can build up the four-dimensional Einstein tensor

1083: \ba

1084: G_{AB}&=&\five G_{CD}\ g^C_{~A}\ g^D_{~B}\ +\five R_{CD}\ n^C\ n^D\ g_{AB}+KK_{AB}\cr

1085: &-&K^C{}_A\ K_{BC}-

1086: \frac{1}{2}g_{AB}(K^2-K^{AB}\ K_{AB})-\tilde

1087: E_{AB}\ ,

1088: \ea

1089: where we introduced

1090: \begin{equation}

1091: \tilde E_{AB}\equiv\five R^C{}_{DFG}\ n_C\ n^F\ g^{D}_{~A}\ g^{G}_{~B}.

1092: \end{equation}

1093: Now we require that the bulk spacetime is governed by the five-dimensional Einstein equations

1094: \begin{equation}\label{einstein5}

1095: \tilde G_{AB}=\tilde\kappa^2 \tilde T_{AB}\ ,

1096: \end{equation}

1097: where $\tilde T_{AB}$ is the five dimensional energy-momentum tensor. We

1098: then use the decomposition of the Riemann tensor

1099: into Weyl curvature, Ricci tensor and scalar curvature,

1100: \begin{equation}\label{decRie}

1101: \five R_{ABCD}=\frac{2}{3}\left(\tilde g_{A[C}\five R_{D]B}-\tilde g_{B[C}

1102: \five R_{D]A}\right)-\frac{1}{6}\tilde g_{A[C}\tilde g_{D]B}\five R+

1103: \five C_{ABCD}\ ,

1104: \end{equation}

1105: where $\tilde C_{ABCD}$ is the Weyl tensor or the traceless part

1106: of the Riemann tensor. Finally the four-dimensional Einstein

1107: tensor in adapted coordinates to the brane is

1108: \ba

1109: G_{\mu\nu}

1110: &=& {2 \tilde \kappa^2 \over 3}\left[T_{\mu\nu}

1111: +\left(T_{\rho\sigma}n^\rho n^\sigma-{1 \over 4}T^\rho_{~\rho}\right)

1112:  g_{\mu\nu} \right]\cr

1113: &+& KK_{\mu\nu}

1114: -K^{~\sigma}_{\mu}K_{\nu\sigma} -{1 \over 2}g_{\mu\nu}

1115:   \left(K^2-K^{\alpha\beta}K_{\alpha\beta}\right) - \tilde {\cal E}_{\mu\nu},

1116: \label{4dEinstein}

1117: \ea

1118: where

1119: \begin{equation}

1120: \tilde {\cal E}_{AB} \equiv \tilde C^C_{~DEF}n_C n^E\

1121: g_A^{~D}\ g_B^{~F},

1122: \end{equation}

1123: and it is traceless. From the Codacci equations (\ref{codacci}) and the five

1124: dimensional Einstein equations (\ref{einstein5}) we get

1125: \begin{equation}\label{quasicon}

1126: g^{~B}_C\nabla_B K^{~C}_A- g^{~B}_A\nabla_B K=\five T_{CD}n^C g^{~D}_A.

1127: \end{equation}

1128: Up to this point we have not introduced a braneworld. A braneworld is

1129: a four-dimensional hypersurface which can be described by the

1130: equation $y=0$ where $y$ is the extra-dimensional

1131: coordinate. Close to the brane we can use Gaussian normal

1132: coordinates such that the metric has the form

1133: \begin{equation}

1134: ds^2=g_{\mu\nu}dx^\mu dx^\nu+dy^2.

1135: \end{equation}

1136: The energy momentum tensor is

1137: \begin{equation}

1138: \tilde T_{AB}=-\frac{\tilde\Lambda}{\tilde \kappa^2} g_{AB}+\delta(y)\left(-\lambda g_{AB}+

1139: T_{AB}\right)\ ,

1140: \end{equation}

1141: where we assume that, at macroscopic level, the matter ($T_{AB}$), is defined only on the brane.

1142: This is

1143: mathematically achieved by introducing a

1144: Dirac distribution $\delta(y)$. $\lambda$ is the vacuum energy of the brane which

1145: coincides with its tension for $T_{AB}=0$\footnote{The tension of a hypersurface is defined as its

1146: extrinsic curvature.}. In particular

1147: we also suppose that there is not a energy-momentum exchange between bulk and brane, or

1148: \begin{equation}

1149: T_{AB}\ n^A=0.

1150: \end{equation}

1151: As I will show in Appendix \ref{Appjc}, reasonable junction conditions require that the singular

1152: behaviour of the metric is encoded in the singular behaviour of the Lie

1153: derivative of the extrinsic curvature along the normal. From that one can get

1154: Israel's junction conditions \cite{israel},

1155: \begin{eqnarray} \label{jcsms}

1156: \left[g_{\mu\nu}\right]

1157: &=&0\,,

1158: \nonumber\\

1159: \left[K_{\mu\nu}\right]&=& -\tilde\kappa^2

1160: \Bigl(T_{\mu\nu}-\frac{1}{3}g_{\mu\nu}(T-\lambda)\Bigr),

1161: \end{eqnarray}

1162: where $[X]:=\lim_{y \to +0}X - \lim_{y \to -0}X=X^+-X^-$.

1163: Substituting then the junction conditions (\ref{jcsms}) into (\ref{4dEinstein}), we get the

1164: effective four-dimensional gravitational

1165: equations

1166: \begin{eqnarray}

1167: G_{\mu\nu}=-\Lambda g_{\mu\nu}

1168: + 8 \pi G_N T_{\mu\nu}+\frac{48\pi G_N}{\lambda}\,S_{\mu\nu}

1169: -{\cal E}_{\mu\nu}\,, \label{eq:effective}

1170: \end{eqnarray}

1171: where

1172: \begin{eqnarray}

1173: {\cal E}_{\mu\nu}&=&{1\over 2}\left[\tilde {\cal E}_{\mu\nu}\right]\,,\\

1174: \Lambda &=&\frac{1}{2}\left(\tilde\Lambda

1175: +\frac{1}{6}\tilde\kappa^4\,\lambda^2\right)\,,

1176: \label{Lamda4}\\

1177: G_N&=&{\tilde \kappa^4\,\lambda\over48 \pi}\,,

1178: \label{GNdef}\\

1179: S_{\mu\nu}&=&

1180: -\frac{1}{4} T_{\mu\alpha}T_\nu^{~\alpha}

1181: +\frac{1}{12}T T_{\mu\nu}

1182: +\frac{1}{8}g_{\mu\nu}T_{\alpha\beta}T^{\alpha\beta}-\frac{1}{24}

1183: g_{\mu\nu}T^2\ .

1184: \label{pidef}

1185: \end{eqnarray}

1186: From the effective equations (\ref{eq:effective}), the localization of gravity is clearly

1187: obtainable in the limit

1188: $\lambda\rightarrow\infty$, $\tilde\kappa\rightarrow 0$, such that $G_N$ remains finite,

1189: and ${\cal E}_{\mu\nu}\rightarrow 0$. Indeed

1190: in this limit we recover Einsteinian gravity. The fact that $G_N>0$ constrains

1191: $\lambda$ to be positive \cite{CGKT}.

1192: 

1193: We now study the matter conservation

1194: equations. Since $T_{AB}n^A=0$ from (\ref{quasicon}) and using the

1195: junction conditions

1196: \begin{equation}

1197: \nabla^\mu T_{\mu\nu}=0.

1198: \end{equation}

1199: This implies that matter follows the standard conservation equations of general relativity.

1200: Thanks to the

1201: four-dimensional Bianchi identities we get differential equations for the projection of the

1202: Weyl tensor, ${\cal E}_{\mu\nu}$. These effective gravitational

1203: equations do not in general close the system.

1204: Indeed the full determination of the projected Weyl tensor requires in general the study of the

1205: five-dimensional gravitational problem.

1206: As we will see, we can sometimes

1207: get around that by introducing particular symmetries for the metric. Then this formalism

1208: becomes a very

1209: powerful tool.

1210: 

1211: \section{1+3 formalism in the braneworld}

1212: In the last section we commented that the four-dimensional effective equations are not in

1213: general closed. This is because the tensor ${\cal E}_

1214: {\mu\nu}$ is not in general completely constrained by the projected equations. However,

1215: often four-dimensional symmetries constrain sufficiently

1216: the five-dimensional geometry, and families of solutions can be found.

1217: To understand better this concept it is very useful to

1218: use the formalism developed by Maartens (see for example \cite{maacov}) which I explain in the

1219: following. In particular since we will be

1220: interested only in perfect fluids, we can drastically simplify all the equations, using the

1221: perfect fluid energy-momentum

1222: tensor

1223: \begin{equation}\label{decT}

1224: T_{\mu\nu}=\rho u_\mu u_\nu+ p h_{\mu\nu},

1225: \end{equation}

1226: where $u^\mu$ is the unit four-velocity of the matter ($u^\mu u_\mu=-1$) and $h_{\mu\nu}$

1227: is the space-like metric that projects

1228: orthogonal to $u^\mu$ ($h_{\mu\nu}=g_{\mu\nu}+u_\mu u_\nu$). $\rho$

1229: and $p$ are respectively the energy density

1230: and the pressure of the perfect fluid.

1231: In the same way as (\ref{decT}), we decompose the local correction $S_{\mu\nu}$ and the

1232: non-local one ${\cal E}_{\mu\nu}$. ${\cal E}_{\mu\nu}$

1233: gives a non-local contribution in the sense that it codifies all the bulk geometrical

1234: back-reaction on the brane.

1235: The decomposition of the matter-correction is

1236: \be

1237: S_{\mu\nu}={{1\over12}}\rho\left[\rho u_\mu u_\nu +\left(\rho+2

1238: p\right)h_{\mu\nu}\right]\,.

1239:  \ee

1240: Using the fact that ${\cal E}_{\mu\nu}$ is traceless we can decompose it as

1241: \footnote{ $\kappa^2=8\pi G_N$ .}

1242:  \be

1243: -{1\over \kappa^2} {\cal E}_{\mu\nu} = \cu\left(u_\mu u_\nu+{ {1\over3}}

1244: h_{\mu\nu}\right)+ {\cq_\mu} u_{\nu} + {\cq_\nu} u_{\mu}+\cp_{\mu\nu}\, ,

1245:  \ee

1246: where we introduced an effective ``dark" radiative

1247: energy-momentum on the brane, with energy density $\cal U$,

1248: pressure ${\cal U}/3$, momentum density $Q_\mu$ and anisotropic

1249: stress ${\Pi}_{\mu\nu}$.

1250: 

1251: The brane-world corrections can conveniently be consolidated into

1252: an effective total energy density, pressure, momentum density and

1253: anisotropic stress \cite{m}:

1254: \begin{eqnarray}

1255: \rho^{\rm eff} &=& \rho\left(1 +\frac{\rho}{2\lambda} +

1256: \frac{\cal U}{\rho} \right)\,,\label{reff} \\ \label{peff}

1257: p^{\rm eff } &=& p  + \frac{\rho}{2\lambda}

1258: (2p+\rho)+\frac{\cal U}{3}\;, \\ q^{\rm eff }_\mu &=& Q_\mu\;, \\

1259: \label{e:pressure2} \pi^{\rm eff }_{\mu\nu} &=& \Pi_{\mu\nu}\;.

1260: \end{eqnarray}

1261: Note that nonlocal bulk effects can contribute to effective

1262: imperfect fluid terms even when the matter on the brane has

1263: perfect fluid form: there is in general an effective momentum

1264: density and anisotropic stress induced on the brane by

1265: the 5D graviton.

1266: 

1267: The effective total equation of state and sound speed follow from

1268: eqs.~(\ref{reff}) and (\ref{peff}) as

1269:  \bea

1270: w^{\rm eff} &\equiv & {p^{\rm eff}\over\rho^{\rm eff}} =

1271: {w+(1+2w)\rho/2\lambda+ \cu/3\rho \over 1+\rho/2\lambda +\cu/\rho}

1272: \,,\label{vh1}\\ c_{\rm eff}^2 &\equiv & {\dot{p}^{\rm

1273: eff}\over\dot{\rho}^{\rm eff}}= \left[c_{\rm s}^2+{\rho+p \over

1274: \rho+\lambda} +{4\cu\over 9(\rho+p)(1+\rho/\lambda)}\right]

1275: \left[1+ {4\cu\over 3(\rho+p)(1+\rho/\lambda)}\right]^{-1}

1276: \, \label{vh2}

1277:  \eea

1278: where $w=p/\rho$ and $c_{\rm s}^2=\dot p/\dot\rho$. We also used the notation

1279: $\dot f=d f/d\tau$ where $\tau$ is the proper

1280: time of the perfect fluid.

1281: 

1282: \subsection{Conservation equations}

1283: ${\cal E}_{\mu\nu}$, the projection of the bulk Weyl tensor on the

1284: brane, encodes corrections from the 5D graviton effects (often called Kaluza-Klein or KK

1285: modes). From

1286: the brane-observer viewpoint, the energy-momentum corrections in

1287: $S_{\mu\nu}$ are local, whereas the KK corrections in ${\cal

1288: E}_{\mu\nu}$ are nonlocal, since they incorporate 5D gravity wave

1289: modes. These nonlocal corrections cannot be determined purely from

1290: data on the brane. In the perturbative analysis

1291: which leads

1292: to the corrections in the gravitational potential,

1293: the KK modes that generate this correction are

1294: responsible for a nonzero ${\cal E}_{\mu\nu}$; this term is what

1295: carries the modification to the weak-field field equations. An equivalent picture is that

1296: these modes arise as a geometrical bulk

1297: back-reaction to the variations of

1298: the matter fields on the brane. We can see how the matter can source these modes as follows.

1299: 

1300: The standard conservation equations

1301: \begin{equation}

1302: \nabla^\mu T_{\mu\nu}=0

1303: \end{equation}

1304: together with the four-dimensional Bianchi identities applied to the effective

1305: four-dimensional gravitational equations

1306: (\ref{eq:effective}) lead to the following non-local equations

1307: \begin{equation}\label{non-local}

1308: \nabla^\mu {\cal E}_{\mu\nu}=\frac{6 \kappa^2}{\lambda}\nabla^\mu S_{\mu\nu}.

1309: \end{equation}

1310: It is clear that the projection of the Weyl tensor, which encodes the curvature of

1311: the co-dimension, can be sourced (as

1312: a back-reaction) by the variation of the matter fields.

1313: 

1314: It is useful to rewrite the local and non-local conservation equation in a $1+3$ formalism.

1315: Introducing the covariant projected Levi-Civita

1316: tensor $\ep_{abc}$, the spatial covariant

1317: derivative $D_a$ (where $D_a S^{b...}{}_{...c}=h^e{}_a h^b{}_f...h^g{}_c \nabla_e S^{f...}

1318: {}_{...g}$),

1319: the volume expansion $\Theta=\nabla^\alpha u_\alpha$, the

1320: proper time derivative $\dot S^{a...}{}_{...b}=u^\alpha

1321: \nabla_\alpha S^{a...}{}_{...b}$, the acceleration $A_a=\dot u_a$, the shear

1322: $\sigma_{ab}=D_{(a} u_{b)}-\Theta/3 h_{ab}$ and the

1323: vorticity $\omega_a=-{1 \over 2}\mbox{curl}\

1324: u_a$, we get the local conservation equations

1325: \begin{eqnarray}

1326: &&\dot{\rho}+\Theta(\rho+p)=0\,,\label{pc1}\\ && \D_ a

1327: p+(\rho+p)A_ a =0\,,\label{pc2}

1328: \end{eqnarray}

1329: and the nonlocal equations

1330: \begin{eqnarray}

1331: && \dot{\cu}+{{4\over3}}\Theta{\cu}+\D^ a\cq_ a+2A^ a\cq_

1332: a+\sigma^{ ab }\cp_{ ab }=0\,, \label{pc1'}\\&& \dot\cq_{

1333: a}+{{4\over3}}\Theta\cq_ a +{{1\over3}}\D_

1334: a{\cu}+{{4\over3}}{\cu}A_ a +\D^ b\cp_{ ab }+A^ b\cp_{ ab

1335: }+\sigma_{ a}{}^b\cq_b-\ep_{a}{}^{bc}\omega_b\cq_c

1336: \nonumber\\&&~~{} =-{(\rho+p)\over\lambda} \D_a\rho\,.\label{pc2'}

1337: \end{eqnarray}

1338: The non-local equations do not contain evolution equations for the anisotropic part of the

1339: projected Weyl tensor $\Pi_{ab}$.

1340: This makes the system of ``brane'' equations not closed. However if we

1341: ask for spatially homogeneous and isotropic solutions, the system becomes closed.

1342: 

1343: \section{Newtonian limit: localization of gravity}\label{secNewton}

1344: 

1345: Equations (\ref{eq:effective})

1346: show that in the limit $\lambda\rightarrow \infty$ and in the case of a

1347: bulk with ${\cal E}_{\mu\nu}=0$,

1348: we recover

1349: General Relativity. In this section we discuss the correction to the Newtonian

1350: potential due to a point-like particle. In order to do that we have to study the full five

1351: dimensional

1352: equations because such a perturbation is due to purely

1353: five dimensional effects. In doing that we follow \cite{DD}.

1354: 

1355: We start by considering the unperturbed Randall-Sundrum braneworld.

1356: The solution for the metric in Gaussian normal

1357: coordinates is given by eq. (\ref{Minkmetric}), i.e.

1358: \be

1359: ds^2=dy^2+e^{-2\mid y\mid/l}\eta_{\mu\nu}dx^\mu dx^\nu\ ,

1360: \ee

1361: where ${l}$ is the curvature scale of the AdS spacetime, ${l}=\sqrt{-6/\tilde \Lambda}$.

1362: In order to solve the perturbations of this metric it is

1363: better to use a conformally Minkowskian coordinate system.

1364: Consider only one side of the spacetime, say $y\geq 0$. In this

1365: side, we can use the transformation of coordinates

1366: \be

1367: e^{y/l}dy=dz,

1368: \ee

1369: and then the metric transforms to

1370: \be

1371: ds^2=\left(\frac{l}{z}\right)^2\eta_{AB}dx^A dx^B.

1372: \ee

1373: In these coordinates the brane is at $z={l}$.

1374: 

1375: We now perturb the Minkowskian metric. We also expect that,

1376: due to these perturbations, the position of the brane will change, generating a ``bending''.

1377: The metric is

1378: \be\label{metan}

1379: ds^2=\left(\frac{l}{z}\right)^2\left(\eta_{AB}+\gamma_{AB}\right)dx^A dx^B,

1380: \ee

1381: where $\gamma_{AB}$ is a perturbation and the position of the brane is at

1382: \be

1383: z={l}+\xi,

1384: \ee

1385: where $\xi\ll {l}$ is a function describing the bending. We have now the freedom

1386: to choose the transversal gauge

1387: $\gamma_{Az}=0$, and in addition we can use the traceless gauge

1388: $\gamma^\mu{}_\mu=\partial_\rho \gamma^\rho{}_\mu=0$.

1389: In particular we expect that the bending function will depend only on the four-dimensional

1390: brane coordinates, since the bending is due

1391: to a four-dimensional perturbation, therefore we require $\partial_z \xi=0$. Under these

1392: conditions, the unit normal to the brane

1393: is

1394: \be

1395: n^A=-\frac{z}{l}\left(\delta^A_z-\delta^A_\mu\partial^\mu \xi\right).

1396: \ee

1397: From this we can calculate the extrinsic curvature to first order in perturbations

1398: \footnote{$\partial_{\mu\nu}=\partial_\mu

1399: \partial_\nu$.}

1400: \be K^A{}_B=-\frac{1}{2} \tilde g ^{AC}\pounds_n g_{CB}\simeq \delta^\nu{}_B

1401: \delta^A{}_\mu\left(\frac{1}{l}\eta^\mu{}_\nu-\partial^\mu{}_\nu\xi

1402: -\frac{1}{2}\partial_z \gamma^\mu{}_\nu\right). \ee From the

1403: junction conditions (\ref{jcsms}) and using the fine tuning

1404: (\ref{min}) we get on the brane $\Sigma$

1405: \be\label{pluba}

1406: \frac{\tilde\kappa^2}{2}\left(T_{\mu\nu}-{1 \over 3}

1407: g_{\mu\nu}T\right)=\left(\partial_{\mu\nu} \xi-\frac{1}{2}\partial_z

1408: \gamma_{\mu\nu}\right)\Big|_\Sigma. \ee

1409: Taking now the trace and using the

1410: traceless condition for the perturbation we get

1411: \be

1412: -\frac{\tilde\kappa^2}{6} T=\Box_4 \xi\ . \ee

1413: Using a pointlike

1414: particle with mass $M$ \be T_{00}=M\delta(\vec r),\ \ \ \

1415: T_{0i}=T_{ij}=0, \ee we obtain \be\label{xi}

1416: \xi=-\frac{\tilde\kappa^2}{24\pi}\frac{M}{r}. \ee Applying this to

1417: (\ref{pluba}) we get the boundary conditions \be \label{boco}

1418: \partial_z\gamma_{00}\Big|_\Sigma=-{2\tilde\kappa^2 M\over 3}\delta (\vec r)\quad,\quad

1419: \partial_z\gamma_{0i}\Big|_\Sigma=0\quad,\quad

1420: \partial_z\gamma_{ij}\Big|_\Sigma=-{\tilde\kappa^2  M\over 3}\delta (\vec r)\delta_{ij}-

1421: {\tilde\kappa^2 M\over

1422: 12\pi}\partial_{ij}{1\over r}\,.

1423: \ee

1424: Now the procedure is to solve the bulk equations and then impose the boundary conditions

1425: (\ref{boco}).

1426: The bulk equations are nothing else than the linearization

1427: of the Einstein equations

1428: \be

1429: \tilde G_{AB}=-\tilde \Lambda \tilde g_{AB},

1430: \ee

1431: that, using the metric ansatz (\ref{metan}), give \cite{5}

1432: \be

1433: \Box_4 \gamma_{\mu\nu}+\partial_z^2\gamma_{\mu\nu}-\frac{3}{z}\partial_z\gamma_{\mu\nu}=0.

1434: \ee

1435: Finally  the general solution of these equations in the static case is a

1436: superposition of Fourier modes~:

1437: \be

1438: \gamma_{\mu\nu}(x^\mu,z)=\int\! {d^3\vec

1439: k\over(2\pi)^{{3\over2}}} e^{i \vec k\,.\vec r}\hat\gamma_{\mu\nu}(\vec k, z)

1440: \ee

1441: with

1442: \be

1443: \hat\gamma_{\mu\nu}(\vec k, z)=z^2

1444: \left[e_{\mu\nu}^{(1)}(\vec k)H_2^{(1)}(ikz)+e_{\mu\nu}^{(2)}(\vec k)H_2^{(2)}(ikz)\right]

1445: \ee

1446: where (because of the traceless conditions) the polarization tensors $e_{\mu\nu}^{(1,2)}

1447: (\vec k)$ are transverse and traceless,

1448: and  where

1449: $H_2^{(1,2)}(ikw)$ are the Hankel functions of first and second kind of order $2$.

1450: The junction conditions (\ref{boco}) therefore determine a combination of the polarization

1451: tensors such that

1452: \begin{eqnarray}

1453: e_{00}^{(1)}(\vec k)H_1^{(1)}(ik{l})+e_{00}^{(2)}(\vec

1454: k)H_1^{(2)}(ik{l})=-{2\tilde\kappa^2 M\over 3}

1455: {1\over(2\pi)^{3\over2}}{1\over ik{l}^2}\ ,\cr e_{0i}^{(1)}(\vec

1456: k)H_1^{(1)}(ik{l})+e_{0i}^{(2)}(\vec k)H_1^{(2)}(ik{l})=0\ ,\cr e_{ij}^{(1)}(\vec k)H_1^{(1)}

1457: (ik{l})+e_{ij}^{(2)}(\vec k)H_1^{(2)}(ik{l})=-{\tilde\kappa^2

1458: M\over 3} {1\over(2\pi)^{3\over2}}{1\over ik{l}^2}\left(\delta_{ij}-{k_ik_j\over4\pi k^2}

1459: \right)\,.

1460: \end{eqnarray}

1461: We are now interested in the solution which converges in the limit

1462: $z\rightarrow\infty$. This is possible by choosing

1463: $e_{\mu\nu}^{(2)}=0$. Then we have the following complete solution

1464: \be \gamma_{\mu\nu}(\vec

1465: r,z)=\int\!{d^3k\over(2\pi)^{3\over2}}\,e^{i\vec k\,.\vec

1466: r}\hat\gamma_{\mu\nu}(\vec k, z)\quad

1467: ,\quad\hat\gamma_{\mu\nu}(\vec k,z)= {\tilde\kappa^2 M\over3{l}(2\pi)^{3\over2}} \,z^2\,

1468: {K_2(kz)\over k{l}\,K_1(k{l})}\,c_{\mu\nu}\ , \ee with $c_{00}=2$, $c_{0i}=0$ and

1469: $c_{ij}=\delta_{ij}-k_ik_j/4\pi k^2$, and where $K_\nu(z)$ is the

1470: modified Bessel function defined as

1471: $K_\nu(z)=i{\pi\over2}e^{i\nu{\pi\over2}}H_\nu^{(1)}(iz)$. Near

1472: the brane this metric reduces to, setting $\epsilon={z\over{l}}-1$,

1473: \be \label{gamma}

1474: \hat\gamma_{\mu\nu}(\vec k,\epsilon)=

1475: {\tilde\kappa^2 M{l}\over3(2\pi)^{3\over2}}

1476: \left\{{K_2(k{l})\over k{l}\,K_1(k{l})}-\epsilon-{k{l}\over4}\epsilon^2\left[{K_2(k{l})

1477: \over K_1(k{l})}-3{K_0(k{l})\over K_1(k{l})}\right]+{\cal O}(\epsilon^3)\right\}c_{\mu\nu}

1478: . \ee

1479: The

1480: appearance of Dirac distributions in the expansion of

1481: $\gamma_{\mu\nu}(\vec r,z)$ does not however necessarily mean that

1482: $\gamma_{\mu\nu}(\vec r,z)$ is singular at $\vec r=0$ as the sum

1483: may be regular. We now see what the perturbed metric

1484: on the brane is. We know that the brane is at $z={l}+\xi$.

1485: Therefore if we want to expand the metric around $z={l}$ we

1486: have

1487: \be ds^2\Big|_{\Sigma}=\left[\frac{l}{{l}+\xi}\left(\eta_{AB}+\gamma_{AB}\right)dx^A

1488: dx^B\right]\Big|_{\Sigma}\simeq\left(\eta_{AB}+h_{AB}\right) dx^A dx^B, \ee where \be

1489: h_{AB}=\gamma_{AB}\Big|_{\Sigma}-2\frac{\xi}{l}\eta_{AB}. \ee

1490: Therefore

1491: the actual perturbed four-dimensional metric is \be \label{actual}

1492: g_{\mu\nu}=\eta_{\mu\nu}+h_{\mu\nu}. \ee The Newtonian potential

1493: is then $h_{00}$. If we now Fourier transform $h_{00}$ considering

1494: (\ref{xi}) and (\ref{gamma}), we get

1495: \be \hat h_{00}(\vec

1496: k)=\hat{\tilde h}_{00}(\vec

1497: k)=\hat\gamma_{00}\Big|_\Sigma+2{\hat\xi\over{l}}=

1498: {\tilde\kappa^2 M\over k^2{l}(2\pi)^{3\over2}}\left[1-{2k{l}\over3}{K_0(k{l})\over

1499: K_1(k{l})}\right]\,. \ee

1500: Taking the inverse-Fourier transform

1501: and integrating over angles, we obtain, setting $\alpha=r/{l}$,

1502: \be h_{00}(\vec r)={\tilde\kappa^2 M\over4\pi{l}}{1\over

1503: r}\left(1+{4\pi\over3}{\cal K}_\alpha\right)\quad\hbox{with}\quad

1504: {\cal K}_\alpha=\lim_{\epsilon\to0}\int_0^{+\infty}\!

1505: du\,\sin(u\alpha)\,{K_0(u)\over K_1(u)}e^{-\epsilon u}\,. \ee

1506: We have a short and long distance limit,

1507: $\lim_{\alpha\to0}{\cal K}_\alpha=\alpha^{-1}={l}/r$,

1508: $\lim_{\alpha\to\infty}{\cal K}_\alpha=\pi/2\alpha^2=\pi ({l}/r)^2/2$. We hence recover

1509: that at short distances the

1510: correction to Newton's law is ${l}/r$, whereas at

1511: distances large compared with the AdS curvature scale ${l}$,

1512: the correction is reduced by another

1513: ${l}/r$ factor, in agreement with \cite{RS,bunchNew}

1514: \be\label{weakexp} \lim_{r/{l}\to\infty}h_{00}(\vec

1515: r)={2G_NM\over r}\left[1+{2\over3}\left({{l}\over

1516: r}\right)^2\right]\,, \ee where, as before, we identify the

1517: Newtonian constant as $8\pi G_N=\tilde\kappa^2/l$. This

1518: expansion can be taken as a ``quality-test'' that solutions of the

1519: four-dimensional effective field equations have to pass to lead to

1520: a regular bulk. Indeed, as we will discuss in the next chapter,

1521: stellar solutions which do not have the weak regime expansion

1522: (\ref{weakexp}), probably lead to a non-regular Cauchy horizon for

1523: the bulk \cite{static}.

1524: 

1525: \subsection{Quantum correction to the Newtonian potential and holographic interpretation}

1526: 

1527: \footnote{To show the importance of studying braneworld scenarios from the holographic

1528: point of view, I will

1529: occasionally insert sections on the quantum side of the correspondence (e.g. AdS/CFT

1530: and its deformations). The main idea of this thesis

1531: is to study the classical astrophysics and cosmology of braneworlds. Therefore these

1532: sections aim to give the reader only a

1533: flavour of the

1534: quantum counterpart of the correspondence.} Following

1535: \cite{DL} we show that the four-dimensional quantum gravity correction to the

1536: Newtonian law

1537: corresponds to the classical correction (\ref{weakexp}) found in the context of the

1538: braneworld scenario. This is a test of the

1539: deformed AdS/CFT

1540: correspondence at perturbative level.

1541: 

1542: We start with the linearized Einstein equations in four dimensions, using the metric

1543: (\ref{actual}),

1544: where $h_{\mu\nu}$ is a perturbation of the Minkowskian metric, the harmonic gauge

1545: $\partial_\mu g^{\mu\nu}=0$ implies

1546: \be

1547: \partial_\mu(h^{\mu\nu}-{1\over 2} \eta^{\mu\nu}h)=0.

1548: \ee

1549: Defining $\tilde h_{\mu\nu}=h^{\mu\nu}-{1\over 2} \eta^{\mu\nu}h$

1550: we get the

1551: linearized Einstein equations

1552: \be

1553: \Box \tilde h_{\mu\nu}=-16\pi G_N T_{\mu\nu},

1554: \ee

1555: so that in Fourier space

1556: \be

1557: \tilde h_{\mu\nu}(p)=16\pi G_N {1\over p^2}T_{\mu\nu}(p).

1558: \ee

1559: For quantum corrections, the total perturbation will be of form

1560: \begin{equation}

1561: \tilde h_{\mu\nu}=h^c_{\mu\nu}+h^q_{\mu\nu}\ ,

1562: \end{equation}

1563: where the superscript $c$ means ``classical'' and the superscript $q$ means ``quantum".

1564: The quantum spin-two tensor $h^q_{\mu\nu}$

1565: must vanish with the vanishing of the

1566: classical perturbations. We have

1567: then two other

1568: ingredients, the graviton propagator and the self-energy of the gravitons.

1569: These two objects must be represented by tensors of rank 4. This is because they have

1570: to be applied to the classical perturbation of

1571: the metric and they have to reproduce a rank 2 tensor (the quantum perturbation of the metric).

1572: 

1573: We start with the propagator, calling it

1574: $D^{\alpha\beta\gamma\delta}$. Taking two points $x$ and

1575: $x'$ at the same proper time, the propagator describes the short-distance quantum effects

1576: of the products $h_{\alpha\beta}(x)h_{\gamma\delta}(x')$. This

1577: implies that it should be proportional to $\delta(x-x')$. Another

1578: ingredient is that it must depend on the unperturbed metric. In

1579: particular since it must be local, it cannot depend on the derivatives

1580: of the background metric. So in general the propagator will depend

1581: only on the products $\eta^{\alpha\beta}\eta^{\gamma\delta}$.

1582: Bearing in mind that $D_{\mu\nu\alpha\beta}\sim

1583: h_{\mu\nu}h_{\alpha\beta}$, by symmetry we can already guess the

1584: form of this propagator as

1585: \be D^{\mu\nu\alpha\beta}={1 \over

1586: 2}\delta(x-x')\left(\eta^{\mu\alpha}\eta^{\nu\beta}+\eta^{\mu\alpha}\eta^{\nu\alpha}+

1587: \lambda\eta^{\mu\nu}\eta^{\alpha\beta}\right),

1588: \ee

1589: where $\lambda$ is a constant that must be determined.

1590: 

1591: For the Hamiltonian constraint to be non-singular, the propagator in Fourier

1592: space becomes \cite{BdW}

1593: \be D^{\mu\nu\alpha\beta}(p^2)=\frac{1}{2

1594: p^2}\left(\eta^{\mu\alpha}\eta^{\mu\beta}+\eta^{\mu\alpha}\eta^{\nu\alpha}-

1595: \eta^{\mu\nu}\eta^{\alpha\beta}\right).

1596: \ee

1597: Up to now we have

1598: \be

1599: h_q^{\mu\nu}=D^{\mu\nu\alpha\beta}\Pi_{\alpha\beta\gamma\delta}h_c^{\gamma\delta},

1600: \ee

1601: where $\Pi_{\alpha\beta\gamma\delta}$ is the graviton self

1602: energy. Now in Fourier space the self energy is a function of the

1603: momentum $p_\alpha$ of the graviton and of the unperturbed metric

1604: (we still are at the first order in quantum corrections). In

1605: particular since the tensor is a rank four tensor and

1606: the metric is dimensionless, the self energy must have all the

1607: possible combinations of $p_\alpha$ products of order four. This is

1608: because in general we will have at least a term $p_\alpha p_\beta

1609: p_\gamma p_\delta$. If we would like to preserve the classical

1610: isometries at quantum level, we can also use the Slanov-Ward

1611: gravitational identity \cite{SWi}

1612: \be p_\mu p_\nu

1613: D^{\mu\nu\alpha\beta}\Pi_{\alpha\beta\gamma\delta}D^{\gamma\delta\rho\sigma}=0.

1614: \ee

1615: This at first linearized level determines the self-energy up to

1616: two functions $\Pi_1(p)$ and $\Pi_2(p)$. Combining the classical

1617: and the one-loop quantum results at the linearized level, we get

1618: \ba

1619: h_{\mu\nu}&=&16 \pi \frac{G_N}{p^2}\left[T_{\mu\nu}-{1\over

1620: 2}\eta_{\mu\nu}T(p)\right]\cr\nonumber &-&16\pi

1621: G_N\left[2\Pi_2(p)T_{\mu\nu}(p)+\Pi_1(p)\eta_{\mu\nu}T(p)\right].

1622: \ea

1623: The actual form of the $\Pi_i$'s depend on the theory we

1624: consider. However for any massless theory in four dimensions,

1625: after cancelling the infinities with the appropriate counterterms,

1626: the finite remainder must have the form \cite{CLR} \be\label{self}

1627: \Pi_i(p)=32\pi G_N\left(a_i\ln \frac{p^2}{\mu^2}+b_i\right), \ee

1628: where $\mu$ is an arbitrary subtraction mass linked with the

1629: renormalization energy. Using now a point source

1630: $T_{00}=M\delta(\vec r)$, we obtain the perturbed time component

1631: of the metric

1632: \be g_{00}=-(1-\frac{2 G_N M}{r}-\frac{2\alpha G_N^2

1633: M}{r^3}), \ee

1634: where $\alpha=4\cdot 32\pi(a_1+a_2)$. Explicit

1635: calculations of the self-energy (\ref{self}) for different spins

1636: give \cite{DL,245Duff}

1637: \be a_i(s=1)=4a_i(s=1/2)=12a_i

1638: (s=0)=\frac{1}{120(4\pi)^2}(-2,3). \ee

1639: Considering the number of

1640: particle species of spin $s$ going around the loop $N_s$ and

1641: considering that for a single CFT these numbers can be rewritten

1642: in terms of the dimensionality of the gauge group of the CFT ($N$)

1643: \be (N_1,N_{1/2},N_0)=(N^2,4N^2,6N^2), \ee

1644: we have that the

1645: one-loop correction to the Newtonian potential is

1646: \be

1647: V(r)=\frac{G_N M}{r}\left(1+\frac{2N^2 G_N}{3\pi r^2}\right). \ee

1648: Using the AdS/CFT relation \cite{M1} $N^2=\pi{l}^3/2\tilde\kappa^2$ we get exactly

1649: the result (\ref{weakexp}),

1650: which was found considering a five dimensional classical

1651: braneworld scenario.

1652: 

1653: \newpage

1654: \chapter{Stars in the braneworld}

1655: 

1656: The first step in considering infinitely large extra dimensions is

1657: the discovery that gravity can be localized at low energies \cite{RS}.

1658: This basically gives the strongest constraint for such theories.

1659: The second step is therefore to see how deviations from general relativity may

1660: explain the nature of our Universe.

1661: 

1662: The natural scenario,

1663: where deviations from general relativity occur, is cosmology.

1664: Indeed cosmological implications of these braneworld models have been

1665: extensively investigated (see e.g. the review~\cite{maacov} for

1666: further references). But this is not all. Significant deviations from Einstein's theory

1667: in fact occur also in astrophysics. Indeed

1668: very compact objects and gravitational collapse to black holes, can leave

1669: traces of the extra dimensions.

1670: For example, when an horizon

1671: forms, even if the high-energy

1672: effects eventually become disconnected

1673: from the outside region on the brane, they could leave a

1674: signature on the brane \cite{collapse}.

1675: 

1676: In addition to local high-energy effects,

1677: there are also nonlocal corrections arising from the imprint on

1678: the brane of Weyl curvature in the bulk, i.e. from 5-dimensional

1679: graviton stresses. These nonlocal Weyl stresses arise on the brane

1680: whenever there is inhomogeneity in the density; the inhomogeneity

1681: on the brane generates Weyl curvature in the bulk which

1682: `backreacts' on the brane. Note that we can have these nonlocal Weyl

1683: stresses even if the density is homogeneous \cite{static}.

1684: 

1685: The high-energy (local) and bulk graviton stress (nonlocal)

1686: effects combine to significantly alter the matching problem on the

1687: brane, compared with the general relativistic case. For spherical

1688: compact or collapsing objects (uncharged and non-radiating), matching in general

1689: relativity shows that the asymptotically flat exterior spacetime

1690: is Schwarzschild. High-energy corrections to the pressure,

1691: together with Weyl stresses from bulk gravitons, mean that on the

1692: brane, matching no longer leads to a Schwarzschild exterior in

1693: general \cite{static, collapse}. These stresses also mean that the matching conditions do

1694: not have unique solution on the brane \cite{static}; knowledge of the

1695: 5-dimensional Weyl tensor is needed as a minimum condition for

1696: uniqueness.

1697: 

1698: \section{The static star case}

1699: 

1700: \footnote{I will base this section essentially on \cite{static}.}In

1701: this section we consider the simplest case of a static spherical

1702: star with uniform density. We find an exact interior solution,

1703: thus generalizing the Schwarzschild interior solution of general

1704: relativity. We show that the general relativity compactness limit

1705: given by $GM/R<{4\over9}$ is reduced by high-energy 5-dimensional

1706: gravity effects. The existence of neutron stars allows us to put a

1707: lower bound on the brane tension, which is stronger than the bound

1708: from big bang nucleosynthesis, but weaker than the bound from

1709: experiments probing Newton's law on sub-millimetre scales. We also

1710: give two different exact exterior solutions, both of which satisfy

1711: the braneworld matching conditions and field equations and are

1712: asymptotically Schwarzschild, but neither of which is the

1713: Schwarzschild exterior. One of these solutions is the

1714: Reissner-N\"ordstrom-type solution found in~\cite{dmpr}, in which

1715: there is no electric charge, but instead a Weyl `charge' arising

1716: from bulk graviton tidal effects. The other is a new solution.

1717: Both of these exterior solutions carry the imprint of bulk

1718: graviton stresses, and each has an horizon on the brane which is

1719: larger than the Schwarzschild horizon.

1720: 

1721: Both of our solutions (i.e. the full solution, interior plus exterior) are consistent

1722: braneworld solutions, but we do not know the bulk solutions of

1723: which they are boundaries. In fact, no exact 5-dimensional

1724: solution for astrophysical brane black holes is known, and the

1725: uniform star case is even more complicated.

1726: 

1727: We have seen in the section (\ref{secNewton}) that the Newtonian potential on the brane is

1728: modified as follow

1729: \begin{equation}\label{pert}

1730: \Phi={GM\over r}\left(1+{2l^2\over3r^2}\right)\,,

1731: \end{equation}

1732: where $l$ is the curvature scale of AdS. This result assumes that the bulk

1733: perturbations are bounded in conformally Minkowski coordinates,

1734: and that the bulk is nearly AdS. It is not clear whether there

1735: is a covariant way of uniquely characterizing these perturbative

1736: results~\cite{DD}, and therefore it remains unclear what the

1737: implications of the perturbative results are for very dense stars

1738: on the brane. However, it seems reasonable to conjecture that the

1739: bulk should be asymptotically AdS, and that its Cauchy horizon

1740: should be regular. Then perturbative results suggest that on the

1741: brane, the weak-field potential should behave as in

1742: eq.~(\ref{pert}). In fact, perturbative analysis also constrains

1743: the weak-field behaviour of other metric components on the

1744: brane~\cite{bunchNew}, as well as the nonlocal stresses on the

1745: brane induced by the bulk Weyl tensor~\cite{ssm}.

1746: This is also supported at non-linear level if one assumes a

1747: bounded bulk in conformally Minkowskian coordinates.

1748: In this case indeed, it is possible to integrate numerically simple models of

1749: homogeneous and isotropic stars \cite{wiseman}.

1750: 

1751: \subsection{Field equations and matching conditions}\label{sectionField}

1752: 

1753: We have already discussed that the local and nonlocal extra-dimensional modifications to

1754: Einstein's equations on the brane may be consolidated into an

1755: effective total energy-momentum tensor:

1756: \begin{equation}

1757: G_{\mu\nu}=\kappa^2 T^{\rm eff}_{\mu\nu}\,, \label{6'}

1758: \end{equation}

1759: where $\kappa^2=8\pi G_N$ and the bulk cosmological constant is

1760: chosen so that the brane cosmological constant vanishes. The

1761: effective total energy density, pressure, anisotropic stress and

1762: energy flux for a fluid are given by eqs. (\ref{reff}-\ref{e:pressure2}).

1763: 

1764: From big bang nucleosynthesis constraints, $\lambda \gtrsim

1765: 1$~MeV$^4$, but a much stronger bound arises from null results of

1766: sub-millimetre tests of Newton's law\footnote{From the definition (\ref{GNdef}) and the fine-tuning (\ref{min}) we get $\lambda=3/4\pi (l^2 G_N)^{-1}$.

1767: Table-top tests of Newton's laws currently find no deviation down to about $0.2$ mm. This implies from (\ref{pert}) that $l\lesssim 0.2$ mm, or

1768: $\lambda \gtrsim

1769: 10$~TeV$^4$.}: $\lambda \gtrsim

1770: 10$~TeV$^4$.

1771: 

1772: The local effects of the bulk, arising from the brane extrinsic

1773: curvature, are encoded in the quadratic terms, $\sim

1774: (T_{\mu\nu})^2/\lambda$, which are significant at high energies,

1775: $\rho\gtrsim\lambda$. The nonlocal bulk effects, arising from the

1776: bulk Weyl tensor, are carried by nonlocal energy density ${\cal

1777: U}$, nonlocal energy flux ${Q}_\mu$ and nonlocal anisotropic

1778: stress ${\Pi}_{\mu\nu}$. Five-dimensional graviton stresses are

1779: imprinted on the brane via these nonlocal Weyl terms.

1780: 

1781: Static spherical symmetry implies ${Q}_\mu=0$ and

1782: \begin{equation}\label{p}

1783: {\Pi}_{\mu\nu}={\Pi}(r_\mu r_\nu-{\textstyle{1\over3}}

1784: h_{\mu\nu})\,,

1785: \end{equation}

1786: where $r_\mu$ is a unit radial vector. For static spherical symmetry, the

1787: conservation equations (\ref{pc1}-\ref{pc2'}) reduce to

1788: \begin{eqnarray}

1789: &&\D_\mu p+(\rho+p)A_\mu=0\,,\label{c1}\\ && {\textstyle{1\over3}}

1790: \D_\mu{\cal U}+ {\textstyle{4\over3}}{\cal U}A_\mu+\D^\nu {\Pi}_{\mu\nu}=

1791: -\frac{(\rho+p)}{\lambda} \D_\mu\rho\, .

1792: \label{c2}

1793: \end{eqnarray}

1794: In static coordinates the metric is

1795: \begin{eqnarray}\label{m}

1796:  ds^2=-A^2(r)dt^2+B^2(r)dr^2+r^2d\Omega^2\,,

1797: \end{eqnarray}

1798: and eqs.~(\ref{6'})--(\ref{m}) imply

1799: \begin{eqnarray}

1800: &&{1\over r^2}-{1\over B^2}\left({1\over r^2}-{2\over r}{B'\over

1801: B}\right) = 8\pi G_N\rho^{\rm eff}\,,\label{f1}\\ &&-{1\over

1802: r^2}+{1\over B^2}\left({1\over r^2}+{2\over r}{A'\over A}\right) =

1803: 8\pi G_N\left(p^{\rm eff}+ {2\over 3}{\Pi}

1804: \right)\,,\label{f2}\\&&p'+{A'\over A}(\rho+p)=0\,,\label{f3}\\

1805: &&{\cal U}'+4{A'\over A} {\cal U}+ 2{\Pi}'+2{A'\over A}{\Pi}+

1806: {6\over r}{\Pi}=-3\frac{(\rho+p)}{\lambda}\rho'\,. \label{f4}

1807: \end{eqnarray}

1808: 

1809: The exterior is characterized by

1810: \begin{equation}

1811: \rho=0=p\,,~~{\cal U}={\cal U}^+\,,~{\Pi}={\Pi}^+\,,

1812: \end{equation}

1813: so that in general $\rho^{\rm eff}$ and $p^{\rm eff}$ are nonzero

1814: in the exterior: there are in general Weyl stresses in the

1815: exterior, induced by bulk graviton effects. These stresses are

1816: radiative, since their energy-momentum tensor is traceless. The system of equations

1817: for the exterior is not closed until a further condition is given

1818: on ${\cal U}^+$, ${\Pi}^+$ (e.g., we could impose ${\Pi}^+=0$ to close the system).

1819: In other words, from a brane

1820: observer's perspective, there are many possible static spherical

1821: exterior metrics, including the simplest case of Schwarzschild

1822: (${\cal U}^+=0={\Pi}^+$).

1823: 

1824: The interior has nonzero $\rho$ and $p$; in general, ${\cal U}^-$

1825: and ${\Pi}^-$ are also nonzero, since by eq.~(\ref{f4}), {\em

1826: density gradients are a source for Weyl stresses in the interior}.

1827: For a uniform density star, we can have ${\cal U}^-=0={\Pi}^-$,

1828: but nonzero ${\cal U}^-$ and/ or ${\Pi}^-$ are possible,

1829: subject to eq.~(\ref{f4}) with zero right-hand side.

1830: 

1831: From eq.~(\ref{f1}) we obtain

1832: \begin{eqnarray}

1833: B^2(r)=\left[1-\frac{2Gm(r)}{r}\right]^{-1}\,,

1834: \end{eqnarray}

1835: where the mass function is

1836: \begin{eqnarray}

1837: m(r)=4\pi\int^r_a\rho^{\rm eff}(r')r'^2dr'\,,

1838: \end{eqnarray}

1839: and $a=0$ for the interior solution, while $a=R$ for the exterior

1840: solution.

1841: 

1842: The Israel matching conditions at the stellar surface

1843: $\Sigma$ give~\cite{israel}

1844: \begin{eqnarray}

1845: [G_{\mu\nu}r^\nu]_{\Sigma}=0\,,

1846: \end{eqnarray}

1847: where $[f]_\Sigma\equiv f(R^+)-f(R^-)$. By the brane field

1848: equation~(\ref{6'}), this implies $[T^{\rm

1849: eff}_{\mu\nu}r^\nu]_\Sigma=0$, which leads to

1850: \begin{eqnarray}  \label{m2}

1851: \left[p^{\rm eff}+ {2\over 3}{\Pi}\right]_\Sigma=0\,.

1852: \end{eqnarray}

1853: Even if the physical pressure vanishes at the surface, the

1854: effective pressure is nonzero there, so that in general a radial

1855: stress is needed in the exterior to balance this effective

1856: pressure.

1857: 

1858: The general relativity limit of eq.~(\ref{m2}) implies

1859: \begin{equation}\label{p0}

1860: p(R)=0\,.

1861: \end{equation}

1862: This can also be obtained from a slightly different

1863: point of view. Consider the conservation equation (\ref{f3})

1864: \be

1865: p'+(\rho+p)\frac{A'}{A}=0.

1866: \ee

1867: The junction conditions (see Appendix \ref{Appjc}) require that on the stellar

1868: surface the metric and

1869: its first derivative along the orthogonal direction to the surface must be continuous and

1870: non-singular.

1871: This implies that $A'/A$ is a continuous and non-singular function across the boundary of

1872: the star. The model we

1873: are going to consider is such that

1874: \be

1875: \rho=\tilde\rho\ \theta(r-R)\ ,\ p=\tilde p\ \theta(r-R)\ ,

1876: \ee

1877: where the functions with tilde are continuous and non-singular functions and

1878: \ba

1879: \theta(r-R)=\begin{cases} 1\ \mbox{for}\ r<R\cr 0\ \mbox{for}\ r>R\end{cases}\ ,

1880: \ea

1881: is the Heaviside function. Now we consider the first derivative of the pressure

1882: \be

1883: p'=\tilde p'\ \theta(r-R)+\tilde p\ \delta(r-R)\ ,

1884: \ee

1885: where $\delta(r-R)$ is the Dirac distribution. In order to balance this distribution

1886: in eq. (\ref{f3}), we have two possibilities. The first one is that $p(R)=0$ and the

1887: second is that the metric becomes singular\footnote{Even if we relax the standard junction

1888: conditions we are left with the non trivial problem of defining the product of distributions

1889: $\delta(r-R)\theta(r-R)$.}. If we consider the junction conditions, requiring $A'/A$

1890: to be continuous and non-singular, the second choice becomes unavailable. Then the only

1891: physically sensible model is with $p(R)=0$. We have only used the

1892: conservation equations (\ref{f3}) and the geometrical junction conditions, valid for

1893: both general relativity and braneworld effective gravity.

1894: This implies that unlike general relativity, where (\ref{p0}) and (\ref{m2})

1895: correspond to the same

1896: constraint, they become separate conditions in the braneworld.

1897: 

1898: Using $p(R)=0$, the constraint (\ref{m2}) becomes

1899: \begin{equation}\label{m3}

1900: 3\frac{\rho^2(R)}{\lambda}+{\cal U}^-(R) +2 {\Pi}^-(R)= {\cal U}^+(R)

1901: +2 {\Pi}^+(R)\,.

1902: \end{equation}

1903: Equation~(\ref{m3}) gives the matching condition for any static

1904: spherical star with vanishing pressure at the surface. If there

1905: are no Weyl stresses in the interior, i.e. ${\cal U}^-=0= {\Pi}^-$,

1906: and if the energy density is non-vanishing at the surface,

1907: $\rho(R)\neq0$, then there must be Weyl stresses in the exterior,

1908: i.e. the exterior cannot be Schwarzschild. Equivalently, {\em if

1909: the exterior is Schwarzschild and the energy density is nonzero at

1910: the surface, then the interior must have nonlocal Weyl stresses}.

1911: 

1912: 

1913: \subsection {Braneworld generalization of exact uniform-density solution}

1914: 

1915: Here we are interested in the most simple model of a compact star (such as

1916: for example a neutron star \cite{gravitation}). In this model the density is considered

1917: constant inside the star and zero outside and the geometry is isotropic.

1918: This of course implies that $\rho'=0$ everywhere. Moreover, consistently with the

1919: isotropy of the star, we set ${\Pi}^-=0$.

1920: 

1921: Equation~(\ref{f3}) integrates in the interior for

1922: $\rho=$\,const to give

1923: \begin{equation}

1924: A^-(r)={\alpha\over \rho+p(r)}\,,

1925: \end{equation}

1926: where $\alpha$ is a constant.

1927: Eq. (\ref{f4}) reads

1928: \be

1929: {\cal U}'+4\frac{A'}{A}{\cal U}=0\ ,

1930: \ee

1931: with solution

1932: \begin{equation}\label{u}

1933: {\cal U}^-(r)={\beta\over \left[A^-(r)\right]^4}\,,

1934: \end{equation}

1935: where $\beta$ is a constant. The matching condition in

1936: eq.~(\ref{m3}) then reduces for a {\em uniform} star to

1937: \begin{equation}\label{m4}

1938: 3\frac{\rho^2}{\lambda}+{\beta\over\alpha^4}\rho^4= {\cal U}^+(R) +2

1939: {\Pi}^+(R)\,.

1940: \end{equation}

1941: It follows that in general the exterior of a uniform star cannot be

1942: Schwarzschild.

1943: 

1944: Combining eqs. (\ref{f2},\ref{f3}) we obtain the generalization of the Tolman-Oppenheimer-

1945: Volkoff

1946: equation of hydrostatic equilibrium for the braneworld

1947: \be \label{hydro}

1948: \frac{dp}{dr}=-(p+\rho)\frac{G_Nm(r)+4\pi r^3 p^{\rm eff}}{r(r-2G_Nm(r))},

1949: \ee

1950: where the interior mass function is

1951: \begin{eqnarray}\label{mass}

1952: m^-(r)=M\left[1+\frac{3M}{8\pi\lambda R^3}\right]\left({r\over R}

1953: \right)^3+4\pi\frac{\beta}{\alpha^4}\int^r_0 r'{}^2{(p(r')+\rho)^4} dr'\,,

1954: \end{eqnarray}

1955: with $M=4\pi R^3\rho/3$, and the effective pressure is

1956: \be

1957: p^{\rm eff}=p+\frac{\rho}{2\lambda}(2p+\rho)+\frac{\beta}{3\alpha^4}(p+\rho)^4.

1958: \ee

1959: Eq. (\ref{hydro}) can be analytically solved if and only if

1960: $\beta=0$. Indeed in this case it is possible to separate the variables $p$ and $r$.

1961: 

1962: Here we are interested in typical astrophysical stars. Their density

1963: $\rho\sim 10^{-3}{}\mbox{GeV}^4$ is much smaller then the tension of the brane

1964: $\lambda\sim 10\mbox{~TeV}^4$. Non-local corrections are produced

1965: by back-reaction of the five-dimensional gravitational field on the brane. Therefore we

1966: naively expect

1967: that their order of magnitude is much smaller than the local effects.

1968: This is also supported by numerical models \cite{wiseman}.

1969: Since we expect ${\cal U}^-\ll \rho^2/\lambda$, we assume in the following $\beta=0$

1970: and

1971: then we can analytically solve (\ref{hydro}).

1972: 

1973: With uniform density and ${\cal U}^-=0= {\Pi}^-$, we have the

1974: case of purely local (high-energy) modifications to the general

1975: relativity uniform-density solution, i.e. to the Schwarzschild

1976: interior solution~\cite{exact}.

1977: 

1978: We can now calculate the pressure. Considering that

1979: \begin{eqnarray}

1980: B^-(r)={1\over\Delta(r)}\,,

1981: \end{eqnarray}

1982: the pressure is given by

1983: \begin{eqnarray} \label{3}

1984: {p(r)\over\rho}=\frac{[{\Delta(r)}-{\Delta(R)}] (1+\rho/\lambda)}

1985: {[3{\Delta(R)}-{\Delta(r)}]+[3{\Delta(R)}-2{\Delta(r)}]\rho/

1986: \lambda}\,,

1987: \end{eqnarray}

1988: where

1989: \begin{eqnarray}

1990: \Delta(r)=\left[1-{2G_NM\over r}\left({r\over R}\right)^3\left\{ 1+

1991: {\rho\over 2\lambda }\right\}\right]^{1/2}\,.

1992: \end{eqnarray}

1993: In the general relativity limit, $\lambda^{-1}\to 0$, we regain

1994: the known exact solution~\cite{exact}. The high-energy corrections

1995: considerably complicate the exact solution.

1996: 

1997: Since $\Delta(R)$ must be real, we find {\em an astrophysical

1998: lower limit on $\lambda$, independent of the Newton-law and

1999: cosmological limits:}

2000: \begin{equation}\label{al}

2001: {\lambda}\geq \left(\frac{G_NM}{R-2G_NM}\right)\rho~~\mbox{for all

2002: uniform stars}\,.

2003: \end{equation}

2004: In particular, since $\lambda,\rho>0$, $\Delta(R)^2>0$ implies $R>2GM$, so that the Schwarzschild

2005: radius is still a limiting radius, as in general relativity.

2006: Taking a typical neutron star (assuming uniform density) with

2007: $\rho\sim 10^{9}$~MeV$^4$ and $M\sim 4\times 10^{57}$~GeV, we find

2008: \begin{eqnarray}\label{al'}

2009: \lambda > 5\times 10^{8}~\mbox{MeV}^4\,.

2010: \end{eqnarray}

2011: This is the astrophysical limit, below which stable neutron stars

2012: could not exist on the brane. It is much stronger than the

2013: cosmological nucleosynthesis constraint, but much weaker than the

2014: Newton-law lower bound. Thus stable neutron stars are easily

2015: compatible with braneworld high-energy corrections, and the

2016: deviations from general relativity are very small. If we used the

2017: lower bound in eq.~(\ref{al'}) allowed by the stellar limit, then

2018: the corrections to general relativistic stellar models would be

2019: significant, as illustrated in fig. (\ref{g1}).

2020: 

2021: \begin{figure}

2022: \begin{center}

2023: \includegraphics[width=6cm,height=7cm,angle=-90]{g1.ps}

2024: \end{center}

2025: \caption{ \label{g1} Qualitative comparison of the pressure $p(r)$,

2026: in general relativity (upper curve), and in a braneworld model

2027: with $\lambda= 5\times 10^{8}$~MeV$^4$ (lower curve).}

2028: \end{figure}

2029: 

2030: 

2031: We can also obtain an upper limit on compactness from the

2032: requirement that $p(r)$ must be finite. Since $p(r)$ is a

2033: decreasing function, this is equivalent to the condition that

2034: $p(0)$ is finite and positive, which gives the condition

2035: \begin{eqnarray}\label{comp}

2036: \frac{G_NM}{R} \leq {4\over 9}\left[{1+7\rho/4\lambda +

2037: 5\rho^2/8\lambda^2

2038: \over(1+\rho/\lambda)^2(1+\rho/2\lambda)}\right]\,.

2039: \end{eqnarray}

2040: It follows that high-energy braneworld corrections {\em reduce}

2041: the compactness limit of the star. For the stellar bound on

2042: $\lambda$ given by eq.~(\ref{al'}), the reduction would be

2043: significant, but for the Newton-law bound, the correction to the

2044: general relativity limit of $4\over9$ is very small. The lowest

2045: order correction is given by

2046: \begin{equation}\label{comp2}

2047: {G_NM\over R}\leq  \frac{4}{9}\left[1-\frac{3\rho}{4\lambda}

2048: +O\left({\rho^2\over\lambda^2}\right)\right]\,.

2049: \end{equation}

2050: For $\lambda\sim10$~TeV$^4$, the minimum allowed by

2051: sub-millimetre experiments, and $\rho\sim 10^{9}$~MeV$^4$, the

2052: fractional correction is $\sim 10^{-16}$.

2053: 

2054: 

2055: \subsection{Two possible non-Schwarzschild exterior solutions}

2056: 

2057: 

2058: The system of equations satisfied by the exterior spacetime on the

2059: brane is not closed. Essentially, we have two independent unknowns

2060: ${\cal U}^+$ and ${\Pi}^+$ satisfying one equation, i.e.

2061: eq.~(\ref{f4}) with zero right-hand side. Even requiring that the

2062: exterior must be asymptotically Schwarzschild does not lead to a

2063: unique solution. Further investigation of the 5-dimensional

2064: solution is needed in order to determine what the further

2065: constraints are. We are able to find two exterior solutions for a

2066: uniform-density star (with

2067: ${\cal U}^-=0$) that are consistent with all equations

2068: and matching conditions on the brane, and that are asymptotically

2069: Schwarzschild.

2070: 

2071: The first is the Reissner-N\"ordstrom-like solution given

2072: in~\cite{dmpr}, in which a tidal Weyl charge plays a role similar

2073: to that of electric charge in the general relativity

2074: Reissner-N\"ordstrom solution. We stress that there is {\em no}

2075: electric charge in this model: nonlocal Weyl effects from the 5th

2076: dimension lead to an energy-momentum tensor on the brane that has

2077: the same form as that for an electric field, but without any

2078: electric field being present. The formal similarity is not

2079: complete, since the tidal Weyl charge gives a {\em positive}

2080: contribution to the gravitational potential, unlike the negative

2081: contribution of an electric charge in the general relativistic

2082: Reissner-N\"ordstrom solution.

2083: 

2084: The braneworld solution is~\cite{dmpr}

2085: \begin{eqnarray}

2086: &&\left(A^+\right)^2=\left(B^+\right)^{-2}=1-\frac{2G_N{\cal

2087: M}}{r}+\frac{q}{r^2}\,,\label{rn1} \\&& {\cal U}^+=-\frac{{\Pi}^+}{2} = \frac{4}{3}\pi G_N

2088: q\lambda \,{1\over r^4}\,,\label{rn2}

2089: \end{eqnarray}

2090: where the matching conditions imply

2091: \begin{eqnarray}

2092: q&=&-3G_NMR\,{\rho\over\lambda}\,,\\ {\cal M}&=& M \left(

2093: 1-\frac{\rho}{\lambda }\right)\,, \\ \alpha&=& \rho\Delta (R)\,.

2094: \end{eqnarray}

2095: Note that the Weyl energy density in the exterior is {\em

2096: negative}, so that 5-dimensional graviton effects lead to a

2097: strengthening of the gravitational potential (this is discussed

2098: further in~\cite{dmpr,ssm}). Since ${\cal M}>0$ is required for

2099: asymptotic Schwarzschild behaviour, we have a slightly stronger

2100: condition on the brane tension:

2101: \begin{equation}

2102: \lambda>\rho \,.

2103: \end{equation}

2104: However, this still gives a weak lower limit, $\lambda>

2105: 10^{9}~\mbox{MeV}^4$. In this solution the horizon is at

2106: \begin{equation}

2107: r_{\rm h}=G_N{\cal M}\left[1+\left\{1+\left({3R\over

2108: 2G_NM}-2\right){\rho\over \lambda} +{\rho^2\over

2109: \lambda^2}\right\}^{1/2}\right].

2110: \end{equation}

2111: Expanding this exact expression shows that the horizon is slightly

2112: beyond the general relativistic Schwarzschild horizon:

2113: \begin{equation}

2114: r_{\rm h}=2G_NM\left[1+{3(R-2G_NM)\over

2115: 4G_NM}\,{\rho\over\lambda}\right] +O\left({\rho^2\over

2116: \lambda^2}\right)

2117: >2GM\,.

2118: \end{equation}

2119: The exterior curvature invariant ${\cal R}^2=

2120: R_{\mu\nu}R^{\mu\nu}$ is given by

2121: \begin{eqnarray}\label{r2}

2122: {\cal R}= 8\pi G_N\left({\rho\over \lambda}\right)^2\left({R\over

2123: r}\right)^4\,.

2124: \end{eqnarray}

2125: Note that for the Schwarzschild exterior, ${\cal R}=0$.

2126: 

2127: The second exterior is a new solution. Like the above solution, it

2128: satisfies the braneworld field equations in the exterior, and the

2129: matching conditions at the surface of the uniform-density star. It

2130: is given by

2131: \begin{eqnarray}

2132: \left(A^+\right)^2&=&1-\frac{2G_N{\cal N}}{r}\,, \label{n1} \\

2133: \left(B^+\right)^{-2}&=&\left(A^+\right)^2

2134: \left[1+\frac{C}{\lambda(r-{\textstyle{3\over2}} G_N{\cal

2135: N})}\right]\,,\label{n2} \\ {\cal U}^+&=& {2\pi G_N^2{\cal

2136: N}C\over(1-3G_N{\cal N}/2r)^2}\,{1\over r^4}\,,\label{n3}\\ {\Pi}^+&=& \left({2\over3}-

2137: {r \over G_N{\cal N}}\right){\cal U}^+\,.

2138: \label{n4}

2139: \end{eqnarray}

2140: From the matching conditions:

2141: \begin{eqnarray}

2142: {\cal N} &=& M\left[\frac{1+2\rho/\lambda}{1+3G_NM\rho/

2143: R\lambda)}\right]\,,\\ C &=& 3G_NM\rho\left[ \frac{1-3G_NM/2R}{ 1

2144: +3G_NM\rho/R\lambda}\right]\,,\\ \alpha &=&{\rho\Delta(R)\over (1+3

2145: G_NM\rho/R\lambda)^{1/2}}\,.

2146: \end{eqnarray}

2147: The horizon in this new solution is at

2148: \begin{equation}

2149: r_{\rm h}=2G_N{\cal N}\,,

2150: \end{equation}

2151: which leads to

2152: \begin{equation}

2153: r_{\rm h}=2G_NM\left[1+\left({2R-3G_NM\over 2R}\right){\rho\over

2154: \lambda}\right]+O\left({\rho^2\over\lambda^2}\right)>2G_NM\,.

2155: \end{equation}

2156: The curvature invariant is

2157: \begin{eqnarray}

2158: {\cal R}&=&\sqrt{{\textstyle{3\over2}}}RC\left({4\pi R\over

2159: 3M}\right)^2{(1-8G_N{\cal N}/3r+2G_N^2{\cal N}^2 /r^2)^{1/2}\over

2160: 1-3G_N{\cal N}/2r}\times\nonumber\\ &&~~{}\times \left({\rho\over

2161: \lambda}\right)^2\left({R\over r}\right)^3 \,.

2162: \end{eqnarray}

2163: Comparing with eq.~(\ref{r2}), it is clear that these two

2164: solutions are different. The difference in their curvature

2165: invariants is illustrated in fig. (\ref{g3}).

2166: 

2167: \begin{figure}

2168: \begin{center}

2169: \includegraphics[width=6cm,height=7cm,angle=-90]{g3.ps}

2170: \end{center}

2171: \caption{ \label{g3} Qualitative behaviour of the curvature

2172: invariant ${\cal R}^2$: the upper curve is the

2173: Reissner-N\"ordstrom-like solution given by eqs.~(\ref{rn1}) and

2174: (\ref{rn2}); the lower curve is the new solution given by

2175: eqs.~(\ref{n1})--(\ref{n4}) ($\lambda= 5\times 10^{8}$~MeV$^4$).}

2176: \end{figure}

2177: 

2178: 

2179: \subsection{Interior solution with Weyl contribution}

2180: 

2181: We now consider the case where the contribution of the projected Weyl tensor

2182: is important in the interior or when $\beta\neq 0$.

2183: 

2184: Following \cite{m} we can calculate the tidal acceleration on the brane measured by

2185: a co-moving observer with four velocity $u^A$.

2186: Its modulus is

2187: \ba

2188: \tilde A=-\lim_{y\rightarrow 0}n_A\tilde R^A_{BCD}u^B n^C u^D\ .

2189: \ea

2190: Using the decomposition of the Riemann tensor (\ref{decRie}) and recalling that

2191: $T_{AB}n^A=0$, we have

2192: \ba

2193: \tilde A=\kappa^2{\cal U}+\frac{\tilde \Lambda}{6}.

2194: \ea

2195: Now if ${\cal U}<0$ the localization of the gravitational field near the brane is enhanced

2196: reinforcing the Newtonian potential.

2197: In this case therefore a negative Weyl energy contributes to binding the star to

2198: the brane. Since this effect is independent of the brane tension, it makes

2199: the star more stable than

2200: in the general relativistic case. We see it in a very special limiting case when

2201: the exterior is Schwarzschild. In this case

2202: (${\cal U}^+=0={\Pi}^+$), eqs.~(\ref{u}) and (\ref{m4})

2203: imply that the interior must have dark radiation

2204: density:

2205: \begin{eqnarray}

2206: {\cal U}^-(r)=-\left(\frac{4\pi G_N}{\rho}\right)^2[\rho+p(r)]^4\,.

2207: \end{eqnarray}

2208: It follows that the mass function in eq.~(\ref{mass}) becomes

2209: \begin{eqnarray}

2210: m^-(r)&=& M\left(1+\frac{\rho}{2\lambda

2211: }\right) \left({r\over R} \right)^3

2212: -\frac{6\pi}{\lambda\rho^2}\int^r_0 [\rho+p(r')]^4 r'^2

2213: dr'\,,

2214: \end{eqnarray}

2215: which is {\em reduced} by the negative Weyl energy density,

2216: relative to the solution in the previous section and to the

2217: general relativity solution. The effective pressure is given by

2218: \begin{equation}\label{pschw}

2219: p^{\rm eff}=p-{\rho\over2\lambda}(2+6w+4w^2+w^3)\,,

2220: \end{equation}

2221: where $w=p/\rho$. Thus $p^{\rm eff}<p$, so that 5-dimensional

2222: high-energy effects reduce the pressure in comparison with general

2223: relativity. This means, as we already anticipated, that this star is more

2224: stable than the general relativistic one.

2225: 

2226: 

2227: \subsection{Unique exterior solution: a conjecture}

2228: 

2229: We found that the Schwarzschild solution is no longer the unique asymptotically

2230: flat vacuum exterior; in general, the exterior carries an imprint

2231: of nonlocal bulk graviton stresses. The exterior is not uniquely

2232: determined by matching conditions on the brane, since the

2233: 5-dimensional metric is involved via the nonlocal Weyl stresses.

2234: We demonstrated this explicitly by giving two exact exterior

2235: solutions, both asymptotically Schwarzschild. Each exterior which

2236: satisfies the matching conditions leads to different bulk metrics.

2237: Without any exact or

2238: approximate 5-dimensional solutions to guide us, we do not know

2239: how the properties of the bulk metric, and in particular its

2240: global properties, will influence the exterior solution on the

2241: brane.

2242: 

2243: Guided by perturbative analysis of the static weak field

2244: limit~\cite{RS,bunchNew,ssm,DD}, we make the following conjecture:

2245: {\em if the bulk for a static stellar solution on the brane is

2246: asymptotically AdS and has regular Cauchy horizon, then the

2247: exterior vacuum which satisfies the matching conditions on the

2248: brane is uniquely determined, and agrees with the perturbative

2249: weak-field results at lowest order.} An immediate implication of

2250: this conjecture is that the exterior is not Schwarzschild, since

2251: perturbative analysis shows that there are nonzero Weyl stresses

2252: in the exterior~\cite{ssm} (these stresses are the manifestation

2253: on the brane of the massive Kaluza-Klein bulk graviton modes). In

2254: addition, the two exterior solutions that we present would be

2255: ruled out by the conjecture, since both of them violate the

2256: perturbative result for the weak-field potential,

2257: eq.~(\ref{pert}).

2258: 

2259: The static problem is already complicated, so that analysis of

2260: dynamical collapse on the brane will be very difficult. However,

2261: the dynamical problem could give rise to more striking features.

2262: Energy densities well above the brane tension could be reached

2263: before horizon formation, so that high-energy corrections could be

2264: significant. We expect that these corrections, together with the

2265: nonlocal bulk graviton stress effects, will leave a non-static,

2266: but transient, signature in the exterior of collapsing matter.

2267: This what the next section considers.

2268: 

2269: \section{Gravitational collapse}

2270: 

2271: \footnote{This section is based on \cite{collapse}.}The study of gravitational collapse

2272: in general relativity is

2273: fundamental to understanding the behaviour of the theory at high

2274: energies. The Oppenheimer-Snyder model still provides a

2275: paradigmatic example that serves as a good qualitative guide to

2276: the general collapse problem in general relativity. It can be solved analytically,

2277: as it simply assumes a collapsing homogeneous dust cloud of finite

2278: mass and radius, described by a Robertson-Walker metric and

2279: surrounded by a vacuum exterior. In general relativity, this exterior is

2280: necessarily static and given by the Schwarzschild

2281: solution~\cite{s}. In other theories of gravity that differ from

2282: general relativity at high energies, it is natural to look for similar examples.

2283: In this section we analyze an Oppenheimer-Snyder-like collapse in the

2284: braneworld scenario, in order to shed light on some fundamental differences

2285: between collapse in general relativity and on the brane.

2286: 

2287: Braneworld gravitational collapse is complicated by a number of

2288: factors. The confinement of matter to the brane, while the

2289: gravitational field can access the extra dimension, is at the root

2290: of the difficulties relative to Einstein's theory, and this is

2291: compounded by the gravitational interaction between the brane and

2292: the bulk. Matching conditions on the brane are more complicated to

2293: implement, and one also has to impose regularity and

2294: asymptotic conditions on the bulk, and it is not obvious what

2295: these should be.

2296: 

2297: In general relativity, the Oppenheimer-Snyder model of collapsing dust has a Robertson-Walker

2298: interior matched to a Schwarzschild exterior. We show that even

2299: this simplest case is much more complicated on the brane. However,

2300: it does have a striking new property, which may be part of the

2301: generic collapse problem on the brane. The exterior is not

2302: Schwarzschild, and nor could we expect it to be, as discussed

2303: in the previous section, but the exterior is not even {\em static}, as shown by our

2304: no-go theorem. The reason for this lies in the nature of the

2305: braneworld modifications to general relativity.

2306: 

2307: The dynamical equations for the projected Weyl tensor (\ref{non-local})

2308: are the complete set of

2309: equations on the brane. They are not closed, since ${\cal

2310: E}_{\mu\nu}$ contains 5D degrees of freedom that cannot be

2311: determined on the brane. However, using only the 4D

2312: projected equations, we prove a no-go theorem valid for the full

2313: 5D problem: {\em given the standard matching conditions on the

2314: brane, the exterior of a collapsing homogeneous dust cloud cannot be static}.

2315: We are not able to determine the non-static exterior metric, but

2316: we expect on general physical grounds that the non-static

2317: behaviour will be transient, so that the exterior tends to a

2318: static form.

2319: 

2320: The collapsing region in general contains dust and also energy

2321: density on the brane from KK stresses in the bulk (``dark radiation"). We show that in

2322: the extreme case where there is no matter but only collapsing

2323: homogeneous KK energy density, there is a unique exterior which is

2324: static for physically reasonable values of the parameters. Since

2325: there is no matter on the brane to generate KK stresses, the KK

2326: energy density on the brane must arise from bulk Weyl curvature.

2327: In this case, the bulk could be pathological. The collapsing KK

2328: energy density can either bounce or form a black hole with a 5D

2329: gravitational potential, and the exterior is of the Weyl-charged

2330: de\,Sitter type \cite{dmpr}, but with

2331: no mass.

2332: 

2333: \subsection{Gravitational collapse: a no-go theorem}\label{star}

2334: 

2335: A spherically symmetric collapse region has a Robertson-Walker metric

2336: \begin{equation}\label{1}

2337: ds^2=-d\tau^2+a(\tau)^2(1+ {\textstyle {1\over4}}

2338: kr^2)^{-2}\left[dr^2+r^2d\Omega^2\right]\, ,

2339: \end{equation}

2340: where $\tau$ is the proper time of the perfect fluid. This implies that the

2341: four velocity is $u^\alpha=\delta^\alpha_\tau$, so that

2342: \be

2343: T^\alpha_\beta=\mbox{diag}\left[-\rho(\tau),p(\tau),p(\tau),p(\tau)\right]\ .

2344: \ee

2345: Moreover the symmetries of the geometry force the four-dimensional Einstein tensor

2346: to have the following properties

2347: \ba

2348: G^r_r=G^\theta_\theta=G^\phi_\phi\ ,

2349: \ea

2350: and to have the off-diagonal terms identically zero. This implies

2351: \ba

2352: Q_\mu=0=\Pi_{\mu\nu}\ .

2353: \ea

2354: Whether or not this model has a regular bulk solution is not possible

2355: to determine here. However if the metric (\ref{1}) applies throughout the brane (i.e. there is

2356: no collapse and no exterior region) we can fully integrate the five dimensional equations

2357: \cite{bdl} obtaining a regular bulk. We will see this in the last chapter dedicated to cosmology.

2358: 

2359: In this case the only non-local non-zero equation left is

2360: \be

2361: \dot{\cal U}+\frac{4}{3}\Theta{\cal U}=0\ ,

2362: \ee

2363: where $\Theta=3\ \dot a/a$. Then

2364: \be

2365: {\cal U}=\frac{C}{\lambda a^4}\ ,

2366: \ee

2367: where $C$ is an integration constant.

2368: 

2369: Then from the $G^0_0$ components of (\ref{eq:effective}) we get the modified Friedmann equation

2370: \begin{eqnarray} \label{evol}

2371: \frac{\dot{a}^2}{a^2}= {\textstyle{8\over3}}\pi G_N\rho\left(1 +

2372: \frac{\rho}{2\lambda}\right) + \frac{C}{\lambda a^4} -\frac{k}{a^2}+

2373: {\textstyle{1\over3}} {\Lambda}\, .

2374: \end{eqnarray}

2375: The $\rho^2$ term,

2376: which is significant for $\rho\gtrsim \lambda$, is the high-energy

2377: correction term, following from $S_{\mu\nu}$. Standard Friedmann

2378: evolution is regained in the limit $\lambda^{-1} \to 0$.

2379: 

2380: At this point we would like to use this geometry as the interior of a collapsing region and try to

2381: match it with a static exterior. Since we are in comoving coordinates, the boundary of the

2382: collapsing region is described by the implicit function

2383: \be

2384: \Phi=r-r_0=0\ ,

2385: \ee

2386: so that its normal unit vector is $n_\alpha=a(1+kr^2/4)^{-1}\delta_\alpha^r$.

2387: From the junction conditions (\ref{Appjc2})

2388: \ba\label{jc}

2389: \left[G^r_r\right]=\left[p^{\rm eff}\right]=0\ ,

2390: \ea

2391: where again $\left[f\right]:=f(r_0^+)-f(r_0^-)$, and $r_0^\pm$ is respectively the limit

2392: from the exterior and the interior to the boundary. Using the same discussion of par.

2393: (\ref{sectionField}) we can argue also that

2394: \be

2395: \left[p\right]=0\ ,

2396: \ee

2397: so that for a vacuum exterior $p^\pm=0$. Therefore the collapsing region must be dust.

2398: 

2399: In this case in the interior

2400: \be

2401: p^{\rm eff}=\frac{\rho^2}{2\lambda}+\frac{\cal U}{3}\ ,

2402: \ee

2403: implying that the exterior can not be static.

2404: 

2405: In the following we find a ``measure" of the non-static behaviour of the exterior solution.

2406: This is encoded in a Ricci anomaly. As we will see in the final section this anomaly can be

2407: interpreted holographically as the Weyl anomaly due to the Hawking evaporation for a collapsing

2408: body.

2409: 

2410: The conservation equations (\ref{pc1}) for a dust cloud gives

2411: \ba

2412: \rho=\rho_0\left(\frac{a_0}{a}\right)^3 ,

2413: \ea

2414: where $a_0$ is

2415: the epoch when the cloud started to collapse. The proper radius

2416: from the centre of the cloud is $R(\tau)=r a(\tau)/(1+ {\textstyle

2417: {1\over4}} kr^2)$. We call the collapsing boundary surface $\Sigma$,

2418: which has as a proper radius $R_\Sigma(\tau)= r_0 a(\tau)/(1+

2419: {\textstyle {1\over4}} kr_0^2)$.

2420: 

2421: We can rewrite the modified Friedmann equation on the interior

2422: side of $\Sigma$ as

2423: \begin{equation}\label{geo1}

2424: \dot{R}^2= \frac{2G_NM}{R}+ 3\frac{G_NM^2}{4\pi\lambda R^4}+ \frac{Q}{\lambda

2425: R^2}+ E+ \frac{\Lambda}{3} R^2\,,

2426: \end{equation}

2427: where the ``physical mass" $M$ (total energy per proper star

2428: volume) and the total ``tidal charge" $Q$ are

2429: \begin{equation}\label{mq}

2430: M={\textstyle{4\over3}}\pi a_0^3 r_0^3\frac{\rho_0}{(1+ {\textstyle

2431: {1\over4}}kr_0^2)^3},\ \ Q=C\frac{r_0^4}{(1+ {\textstyle

2432: {1\over4}}kr_0^2)^4},

2433: \end{equation}

2434: and the ``energy" per unit mass is given by

2435: \begin{equation}\label{en}

2436: E=-\frac{kr_0^2}{(1+ {\textstyle {1\over4}}kr_0^2)^2}>-1 \,.

2437: \end{equation}

2438: 

2439: Now we assume that the exterior is static, and satisfies the

2440: standard 4D junction conditions. Then we check whether this

2441: exterior is physical consistent by imposing the modified Einstein

2442: equations (\ref{eq:effective}) for vacuum, i.e.\ for $T_{\mu\nu} =0 =

2443: S_{\mu\nu}$. The standard 4D Israel matching conditions, which we

2444: assume hold on the brane, require that the metric and the

2445: extrinsic curvature of $\Sigma$ be continuous. The extrinsic

2446: curvature is continuous if the metric is continuous and if $\dot

2447: R$ is continuous~\cite{s}. We therefore need to match the metrics

2448: and $\dot R$ across $\Sigma$.

2449: 

2450: The most general static spherical metric that could match the

2451: interior metric on $\Sigma$ is

2452: \begin{eqnarray}

2453: ds^2= -F(R)^2A(R)dt^2 +\frac{dR^2}{A(R)} +R^2d\Omega^2,\label{s1}

2454: \end{eqnarray}

2455: where

2456: \be \label{A}

2457: A(R) =1-2G_Nm(R)/R.

2458: \ee

2459: 

2460: We need two conditions to determine the functions $F(R)$ and

2461: $m(R)$. Now $\Sigma$ is a freely falling surface in both metrics.

2462: Therefore the first condition comes from the comparison of the geodesic

2463: equations of the two metrics. The geodesic equation for a radial trajectory seen

2464: from the exterior is calculated as follows. For a time-like geodesic we have

2465: \be\label{geodesic}

2466: \frac{ds^2}{d\tau^2}=-1=-F(R)^2A(R)\dot t^2+\frac{\dot R^2}{A}\ .

2467: \ee

2468: Now since we are requiring that the exterior metric is static, it exist a

2469: time-like Killing vector $\xi^\alpha$ such that

2470: \ba

2471: \pounds_\xi g_{\alpha\beta}&=\xi_{\alpha}{}_;{}_\beta+\xi_{\beta}{}_;{}_\alpha=0\ ,\

2472: \xi^\alpha \xi_\alpha<0\nonumber

2473: \ea

2474: where $\pounds$ denotes the Lie derivative. If $u^\alpha$ is a geodesic vector,

2475: $u^\alpha\nabla_\alpha u_{\beta}=0$, we have that

2476: \be\label{tildeE}

2477: \tilde E^{1/2}=-\xi^\alpha u_{\alpha}=-F(R)^2A(R)\dot t\ ,

2478: \ee

2479: is a constant along the geodesic motion, or $u^\alpha\nabla_\alpha \tilde E=0$.

2480: Using (\ref{geodesic}) and (\ref{tildeE}), the radial geodesic

2481: equation for the exterior metric gives

2482: $\dot{R}^2=-A(R)+\tilde{E}/F(R)^2$.

2483: Comparing this with eq.~(\ref{geo1}) gives one condition. The

2484: second condition is easier to derive if we change to null

2485: coordinates\footnote{These coordinates, denoted by $(v,r,\theta,\phi)$ are

2486: such that the light-cones are described by the equation $v=\mbox{const}$.}. The

2487: exterior static metric, with

2488: $dv=dt+dR/[F(1-2Gm/R)]$, becomes

2489: \begin{eqnarray}

2490: ds^2&=& -F^2Adv^2 +2FdvdR+R^2d\Omega^2\,.\label{s1'}

2491: \end{eqnarray}

2492: The interior Robertson-Walker metric takes the form~\cite{tesi}

2493: \begin{eqnarray}

2494: ds^2&=& -\tau_{,v}^2\frac{1-(k+\dot{a}^2)R^2/a^2}{1-kR^2/a^2} dv^2

2495: +2 \tau_{,v} \frac{dvdR}{\sqrt{1-kR^2/a^2}}

2496: +R^2d\Omega^2\,,\label{s1''}

2497: \end{eqnarray}

2498: where

2499: \be

2500: d\tau=\tau_{,v}dv+\frac{1+{1\over4}kr^2}{r\dot

2501: a-1+{1\over4}kr^2} dR\ .

2502: \ee

2503: Comparing eqs.~(\ref{s1'}) and (\ref{s1''})

2504: on $\Sigma$ gives the second condition. The two conditions

2505: together imply that $F$ is a constant, which we can take as

2506: $F(R)=1$ without loss of generality (choosing $\tilde E=E+1$), and

2507: that

2508: \begin{equation}\label{s3}

2509: m(R)= M+\frac{3M^2}{8\pi\lambda R^3}+ \frac{Q}{2G_N\lambda R}+ \frac{\Lambda

2510: R^3}{6G_N}\,.

2511: \end{equation}

2512: In the limit $\lambda^{-1}\to 0$, we recover the 4D general relativity

2513: Schwarzschild-de\,Sitter solution. However, we note that the above

2514: form of $m(R)$ violates the weak-field perturbative limit in

2515: eq.~(\ref{pert}), and this is a symptom of the problem with a

2516: static exterior. Equations~(\ref{s1}) and (\ref{s3}) imply that

2517: the brane Ricci scalar is

2518: \begin{eqnarray}\label{m1}

2519: R^\mu{}_\mu=4\Lambda+\frac{9G_N M^2}{2\pi\lambda R^6}\,.

2520: \end{eqnarray}

2521: However, the modified Einstein equations (\ref{eq:effective}) for a vacuum exterior imply that

2522: \ba

2523: R_{\mu\nu}&=&\Lambda g_{\mu\nu}- {\cal E}_{\mu\nu}\\

2524: R^\mu{}_\mu &=& 4\Lambda\,. \label{vac}

2525: \ea

2526: Comparing with eq.~(\ref{m1}), we see that a static exterior is

2527: only possible if $M/\lambda=0\,$. We can therefore interpret $R^\mu{}_\mu$

2528: as a kind of ``potential energy" that must be released from the star during the collapse, due

2529: only to braneworld effects. We emphasize

2530: that this no-go result does not require any assumptions on the

2531: nature of the bulk spacetime.

2532: 

2533: In summary, we have explored the consequences for

2534: gravitational collapse of braneworld gravity effects, using the

2535: simplest possible model, i.e.\ an Oppenheimer-Snyder-like collapse on a

2536: generalized Randall-Sundrum type brane. Even in this simplest

2537: case, extra-dimensional gravity introduces new features. Using

2538: only the projected 4D equations, we have shown, independent of the

2539: nature of the bulk, that the exterior vacuum on the brane is

2540: necessarily {\em non-static}. This contrasts strongly with general relativity,

2541: where the exterior is a static Schwarzschild spacetime. Although

2542: we have not found the exterior metric, we know that its non-static

2543: nature arises from (a)~5D bulk graviton stresses, which transmit

2544: effects nonlocally from the interior to the exterior, and (b)~the

2545: non-vanishing of the effective pressure at the boundary, which

2546: means that dynamical information on the interior side can be

2547: conveyed outside. Our results suggest that gravitational collapse

2548: on the brane may leave a signature in the exterior, dependent upon

2549: the dynamics of collapse, so that astrophysical black holes on the

2550: brane may in principle have KK hair.

2551: 

2552: We expect that the non-static exterior will be transient and partially {\em

2553: non-radiative}, as follows from a perturbative study of non-static

2554: compact objects, showing that the Weyl term ${\cal E}_{\mu\nu}$ in

2555: the far-field region falls off much more rapidly than a radiative

2556: term~\cite{ssm}. It is reasonable to assume that the exterior

2557: metric will become static at late times and tend to Schwarzschild,

2558: at least  at large distances.

2559: 

2560: 

2561: \subsection{Gravitational collapse of pure Weyl energy}

2562: 

2563: The one case that escapes the no-go theorem is $M=0$.  In general relativity,

2564: $M=0$ would lead to vacuum throughout the spacetime, but in the

2565: braneworld, there is the tidal KK stress on the brane, i.e.\ the

2566: $Q$-term in eq.~(\ref{geo1}).  The possibility of black holes

2567: forming from KK energy density was suggested in~\cite{dmpr}. The

2568: dynamics of a Friedmann universe (i.e.\ without exterior),

2569: containing no matter but only KK energy density (``dark

2570: radiation") has been considered in~\cite{bcg}. In that case, there

2571: is a black hole in the Schwarzschild-AdS bulk, which sources

2572: the KK energy density. Growth in the KK energy density corresponds

2573: to the black hole and brane moving closer together; a singularity

2574: on the brane can arise if the black hole meets the brane. Here we

2575: investigate the collapse of a bound region of homogeneous KK

2576: energy density within an inhomogeneous exterior. It is not clear

2577: whether the bulk black hole model may be modified to describe this

2578: case, and we do not know what the bulk metric is. However, we know

2579: that there must be 5D Weyl curvature in the bulk, and that the

2580: bulk could be pathological, with a more severe singularity than

2581: Schwarzschild-AdS. Even though such a bulk would be unphysical

2582: (as in the case of the Schwarzschild black string \cite{blackstring}), it is

2583: interesting to explore the properties of a brane with collapsing

2584: KK energy density, since this idealized toy model may lead to

2585: important physical insights into more realistic collapse with

2586: matter and KK energy density.

2587: 

2588: The exterior is static and unique, and given by the Weyl-charged

2589: de\,Sitter metric

2590: \begin{equation}\label{wd}

2591: ds^2=-Adt^2+\frac{dR^2}{A}+R^2d\Omega^2\,,~M=0\,,

2592: \end{equation}

2593: if $A>0$ ($A$ is given by (\ref{A}) together with the definition (\ref{s3})). For $Q=0$ it is de\,Sitter, with horizon

2594: $H^{-1}=\sqrt{3/\Lambda}$. For $\Lambda=0$ it is the special case

2595: $M=0$ of the solutions given in~\cite{dmpr}, and the length scale

2596: $H_Q^{-1}=\sqrt{|Q|/\lambda}$ is an horizon when $Q>0$; for $Q<0$,

2597: there is no horizon. As we show below, the interplay between these

2598: scales determines the characteristics of collapse.

2599: 

2600: For $\Lambda=0$, the exterior gravitational potential is

2601: \begin{equation}

2602: \Phi=\frac{Q}{2\lambda R^2}\,,

2603: \end{equation}

2604: which has the form of a purely 5D potential when $Q>0$. When

2605: $Q<0$, the gravitational force is repulsive. We thus take $Q>0$ as

2606: the physically more interesting case, corresponding to {\em

2607: positive} KK energy density in the interior. However we note the

2608: remarkable feature that $Q>0$ also implies {\em negative} KK

2609: energy density in the exterior:

2610: \begin{equation}

2611: -{\cal E}_{\mu\nu}u^\mu u^\nu=\left\{ \begin{array}{ll}

2612: +3Q/(\lambda R_\Sigma^4)\,, & R<R_\Sigma\,, \\ %&\\

2613: -Q/(\lambda R^4)\,, & R> R_\Sigma\,. \end{array}\right.

2614: \end{equation}

2615: Negativity of the exterior KK energy density in the general case

2616: with matter has been previously noted~\cite{ssm,dmpr}.

2617: 

2618: The boundary surface between the KK ``cloud" and the exterior has

2619: equation of motion $\dot{R}^2=E-V(R)$, where $V=A-1$.

2620: 

2621: For $\Lambda=0$, the cases are:\\

2622: $\bf Q>0$:\, The cloud collapses

2623: for all $E$, with horizon at $R_{\rm

2624: h}=H_Q^{-1}=\sqrt{Q/\lambda}$. For $E<0$, given that $E>-1$, the

2625: collapse can at most  start from rest at

2626: \be

2627: R_{\rm

2628: max}=\sqrt{\frac{Q}{\lambda|E|}}>R_{\rm h}\ .

2629: \ee

2630: {$\bf Q<0$}:\, It follows that if

2631: $E>0$, there is no horizon, and the cloud bounces at

2632: \be

2633: R_{\rm

2634: min}=\sqrt{|Q|/(\lambda E)}\ .

2635: \ee

2636: 

2637: For $\Lambda> 0$, the potential is given by

2638: \be

2639: V/ V_{\rm c}=-\left(\frac{R}

2640: {R_{\rm c}}\right)^2\left[1+\epsilon(R_{\rm c}/ R)^4\right]\,,

2641: \ee

2642: where

2643: $V_{\rm c}={H/ H_Q}\,,$ $R_{\rm c}=1/\sqrt{HH_Q}\,,$ and

2644: $\epsilon={\rm sgn}\,Q$ (see fig. (\ref{fig3})). The horizons are given by

2645: \begin{equation}

2646: \left(R_{\rm h}^\pm\right)^2= \frac{R_{\rm c}^2}{2V_{\rm c}} \left[1\pm

2647: \sqrt{1-4\epsilon V_{\rm c}^2}\right]\,.\label{hor}

2648: \end{equation}

2649: If $\epsilon>0$ there may be two horizons; then $R_{\rm h}^-$ is

2650: the black hole horizon and $R_{\rm h}^+$  is a modified de\,Sitter

2651: horizon. When they coincide the exterior is no longer static, but

2652: there is a black hole horizon. If $\epsilon<0$ there is always one

2653: de\,Sitter-like horizon, $R_{\rm h}^+$.

2654: \begin{figure}[t]

2655: \begin{center}

2656: \includegraphics[width=8cm,height=4cm,angle=0]{L.eps}

2657: \end{center}

2658: \caption{ \label{fig3} The potential $V(R)$ for $\Lambda> 0$, with

2659: $R$ given in units of $R_{\rm c}$ and $V$ given in units of

2660: $V_{\rm c}$.}

2661: \end{figure}

2662: {$\bf Q>0$}:\, The potential has a

2663: maximum $-2V_{\rm c}$ at $R_{\rm c}$. If $E>-2V_{\rm c}$ the cloud

2664: collapses to a singularity. For $V_{\rm c}>{1\over2}$, i.e.\

2665: $Q>3\lambda/4\Lambda$, there is no horizon, and a naked

2666: singularity forms. For for $V_{\rm c}={1\over2}$ there is one

2667: black hole horizon $R_{h}^{-}=R_{h}^{+}=H^{-1}/\sqrt{2}$. If

2668: $E\leq -2V_{\rm c}$, then eq.~(\ref{en}) implies $V_{\rm

2669: c}<{1\over2}$, so there are always two horizons in this case.

2670: Either the cloud collapses from infinity down to $R_{\rm min}$ and

2671: bounces, with $R_{\rm min}<R_{\rm h}^+$ always, or it can at most

2672: start from rest at $R_{\rm max} (>R_{\rm h}^-)$, and collapse to

2673: a black hole, where ($\epsilon=1$)

2674: \begin{eqnarray}

2675: R^2_{\stackrel{\scriptstyle \rm min}{\rm max}} =\frac{R_{\rm c}^2}{2V_{\rm c}}

2676: \left[-E \pm \sqrt{E^2- 4\epsilon V_{\rm c}^2}\right]

2677: \,.\label{mini}

2678: \end{eqnarray}

2679: {$\bf Q<0$}:\, The potential is

2680: monotonically decreasing, and there is always an horizon, $R_{\rm

2681: h}^+$. For all $E$, the cloud collapses to $R_{\rm min} (<R_{\rm

2682: h}^+)$, and then bounces, where $R_{\rm min}$ is given by

2683: eq.~(\ref{mini}) with $\epsilon=-1$.

2684: 

2685: In summary we have analyzed the idealized collapse of homogeneous KK energy

2686: density whose exterior is static and has purely 5D gravitational

2687: potential. The collapse can either come to a halt and bounce, or

2688: form a black hole or a naked singularity, depending on the

2689: parameter values. This may be seen as a limiting idealization of a

2690: more general spherically symmetric but inhomogeneous case. The

2691: case that includes matter may be relevant to the formation of

2692: primordial black holes in which nonlinear KK energy density could

2693: play an important role.

2694: 

2695: \subsection{Holographic limit for $\lambda$ via Hawking process}

2696: 

2697: Recently it has been suggested that Hawking evaporation of a four dimensional black hole

2698: may be described holographically by the five-dimensional classical black hole metric \cite{hawholo}.

2699: Here we see how this correspondence works in the case of an Oppenheimer-Snyder collapse.

2700: It seems that the Hawking effect does not have memory of

2701: the collapsing process \cite{bida}, and we expect that the conclusions

2702: we obtain are general.

2703: 

2704: A Schwarzschild black hole produces quantum

2705: mechanically a trace anomaly for the vacuum energy-momentum tensor. This is due

2706: to the gravitational energy via curvature of spacetime which excites the vacuum.

2707: We discuss the analogy between the

2708: quantum stress-tensor anomaly and the Ricci anomaly found for a dust cloud

2709: gravitational collapse, (\ref{m1}).

2710: 

2711: In order to simplify the problem we will calculate explicitly the conformal anomaly only

2712: for a massless scalar field in a curved background. We will then comment on the general

2713: vacuum excitations and give a measure for the tension of the brane using the correspondence.

2714: 

2715: \subsubsection{Green function for a massless scalar field}

2716: 

2717: Here we will basically follow \cite{bida}.

2718: 

2719: A general local Lagrangian for a scalar field $\Phi$ coupled to gravity is

2720: \be\label{lagmassless}

2721: {\cal L}=\frac{1}{2}\sqrt{-g}\left[\partial_\mu \Phi\partial^\mu \Phi-(m^2-\xi R)\Phi^2\right]\ ,

2722: \ee

2723: where $\xi$ is an a dimensionless constant, $R$ is the Ricci scalar and $m$ is the mass

2724: of the field. The classical equations follow as

2725: \be\label{evolscalar}

2726: \left[\Box + m^2-\xi R\right]\Phi=0\ .

2727: \ee

2728: Since in the following we will consider vacuum excitations,

2729: we make the Lagrangian invariant under conformal transformation of the metric. The only non-trivial possibility is for

2730: $m=0$ and $\xi=1/6$ in four dimensions. Now as usual we define the momentum

2731: \be

2732: \pi=\frac{\partial {\cal L}}{\partial(\partial_t\Phi)}\ ,

2733: \ee

2734: and the quantum commutation rules are

2735: \ba

2736: \left[\Phi(t,x),\Phi(t,x')\right]&=&0\ ,\cr

2737: \left[\pi(t,x),\pi(t,x')\right]&=&0\ ,\cr

2738: \left[\Phi(t,x),\pi(t,x')\right]&=&i\delta^3(x-x')/\sqrt{-g}\ .\nonumber

2739: \ea

2740: If we define the Green function as the two-point correlation function

2741: \be

2742: iG_F(x-x')=\langle 0|T(\Phi(x)\Phi(x'))|0\rangle\ ,

2743: \ee

2744: where $T$ is the temporal order operator and $|0\rangle$ is the vacuum, we have

2745: \be\label{scalarGreen}

2746: \left[\Box -\frac{1}{6} R\right]G_F(x-x')=-\frac{\delta^4(x-x')}{\sqrt{-g}}\ .

2747: \ee

2748: In order to define a meaningful concept of a particle in curved space, we have to work only

2749: with local quantities. This means that we consider, as proper particles, only the high

2750: frequencies in the Fourier space of $\Phi$. Therefore we will be interested in having

2751: a solution for (\ref{scalarGreen}) only

2752: in the limit $x\rightarrow x'$.

2753: 

2754: Introducing the Riemann coordinates $y^\mu=x^\mu-x'^\mu$, we have \cite{kre,pet}

2755: \ba

2756: g_{\mu\nu}&=\eta_{\mu\nu}+\frac{1}{3}R_{\mu\alpha\nu\beta}y^\alpha y^\beta-

2757: \frac{1}{6}R_{\mu\alpha\nu\beta}{}_{;\gamma}y^\alpha y^\beta y^\gamma\cr

2758: &+\left[\frac{1}{20}R_{\mu\alpha\nu\beta}{}_{;\gamma\delta}+\frac{2}{45}R_{\alpha\mu\beta\lambda}

2759: R^\lambda{}_{\gamma\nu\delta}\right]y^\alpha y^\beta y^\gamma y^\delta+...\ ,\nonumber

2760: \ea

2761: where all the coefficients are evaluated at $y^\alpha=0$.

2762: 

2763: We can now solve (\ref{scalarGreen}) around $y^\alpha=0$

2764: in Fourier space and apply an anti-Fourier-transform, to obtain

2765: \be

2766: G_F(x)=\frac{-i}{(4\pi)^2\sqrt{-g}}\int^\infty_0i ds (is)^{-2}F(x,s)\ ,

2767: \ee

2768: where

2769: \be

2770: F(x,s)=1+a(x)(is)^2\ ,

2771: \ee

2772: and

2773: \be

2774: a(x)=\frac{1}{180}\left[R_{\alpha\beta\gamma\delta}R^{\alpha\beta\gamma\delta}-R^{\alpha\beta}

2775: R_{\alpha\beta}+\Box R\right]\ .

2776: \ee

2777: 

2778: \subsubsection{Trace anomaly and the Hawking effect}\label{holocollapse}

2779: 

2780: In semiclassical gravity the gravitational field is treated classically and the matter

2781: fields are

2782: treated quantum mechanically. In this case what actually sources the gravitational field is the

2783: expectation value of the energy-momentum tensor,

2784: \be\label{R}

2785: R_{\mu\nu}-{1\over 2}R g_{\mu\nu}+\Lambda_R g_{\mu\nu}=8\pi G_R \langle T_{\mu\nu}\rangle\ ,

2786: \ee

2787: where the subscript $R$ indicates that the cosmological and Newton constants must be

2788: renormalized. It is possible to prove that for a conformal theory we have

2789: $G_R=G_N$, whereas $\Lambda_R$ must be experimentally evaluated \cite{bida}

2790: \footnote{In general the left hand side can have additional higher curvature and derivative

2791: terms

2792: due to the renormalization process. It is possible to find exactly the expression of these

2793: corrections up to multiplicative constants that must be evaluated experimentally \cite{bida}.

2794: We do not really need in the following the equation (\ref{R}), but we wish to comment that

2795: such corrections are not divergent-free, and therefore allow the energy-momentum tensor to be

2796: non-conserved. Since we believe in the Einstein principle that the stress tensor is conserved

2797: by gravitational identities (e.g. Bianchi) and not by additional constraints introduced by hand,

2798: we set these constants zero.}. Now

2799: the classical energy-momentum tensor

2800: is connected to the variation of the action $S_m$ of matter,

2801: \be

2802: \frac{2}{\sqrt{-g}}\frac{\delta S_m}{\delta g^{\mu\nu}}=T_{\mu\nu}\ .

2803: \ee

2804: Given the introduction of quantum degrees of freedom, the semiclassical energy-momentum tensor

2805: differs from the classical one.

2806: This implies that we should find an effective action to encode the

2807: quantum degrees freedom as well. Calling it $W$ we can define

2808: \be

2809: \frac{2}{\sqrt{-g}}\frac{\delta W}{\delta g^{\mu\nu}}=\langle T_{\mu\nu}\rangle\ .

2810: \ee

2811: From path integral formalism, the $n$-point correlation function can be obtained by

2812: the generating functional

2813: \be

2814: Z[J]=\int{\cal D}[\Phi]\exp\left[iS_m[\Phi]+i\int J(x)\Phi(x)\sqrt{-g}d^4x\right]\ ,

2815: \ee

2816: where ${\cal D}[\Phi]$ indicates the integration over all the functions $\Phi$ and

2817: \be

2818: \left(\frac{\delta^n Z}{\delta J(x^1)...\delta J(x^n)}\right)\Big|_{J(x^i)=0}=\langle \mbox{out}

2819: ,0|T(\Phi(x^1)...\Phi(x^n))|0,\mbox{in}\rangle\ .

2820: \ee

2821: Now we have \cite{schwinger}

2822: \be\label{schwinger}

2823: \delta Z[0]=i\int {\cal D}[\Phi]\delta S_m e^{iS_m[\Phi]}=i\langle \mbox{out},0|

2824: \delta S_m|0,\mbox{in}\rangle\ ,

2825: \ee

2826: and therefore we obtain

2827: \be

2828: \frac{2}{\sqrt{-g}}\frac{\delta Z[0]}{\delta g^{\mu\nu}}=i\langle \mbox{out},0|

2829: T_{\mu\nu}|0,\mbox{in}\rangle\ .

2830: \ee

2831: Then

2832: \be

2833: \frac{2}{\sqrt{-g}}\frac{\delta\ln Z[0]}{\delta g^{\mu\nu}}=i\frac{\langle \mbox{out},0|

2834: T_{\mu\nu}|0,\mbox{in}\rangle}{\langle \mbox{out},0|0,\mbox{in}\rangle}=

2835: i\langle T_{\mu\nu}\rangle\ .

2836: \ee

2837: This means that we can identify $Z[0]=e^{iW}$.

2838: 

2839: Integrating by parts the action for a massless scalar field with Lagrangian (\ref{lagmassless}),

2840: we get\footnote{The boundary term is taken to be zero.}

2841: \be

2842: S_m=-\frac{1}{2}\int d^4 x\sqrt{-g}\Phi \left(\Box -\frac{1}{6}R\right)\Phi=

2843: -\frac{1}{2}\int d^4 x d^4 y\sqrt{-g}\Phi(x) K_{xy}\Phi(y)\ ,

2844: \ee

2845: where

2846: \be

2847: K_{xy}=\left(\Box-\frac{1}{6}R\right)\delta^4(x-y)\sqrt{-g}\ ,

2848: \ee

2849: so that

2850: \be

2851: K_{xy}^{-1}=-G_F(x,y)\ .

2852: \ee

2853: If we change now the variable in (\ref{schwinger}), using

2854: \be

2855: \Phi'(x)=\int d^4y K_{xy}^{1/2}\Phi(y)\ ,

2856: \ee

2857: we obtain

2858: \be

2859: Z[0]\propto (\mbox{det} K_{xy}^{1/2})^{-1}=\sqrt{\mbox{det}(-G_F)}\ .

2860: \ee

2861: Then

2862: \be

2863: W=-i\ln Z[0]=-\frac{1}{2}i\mbox{tr}[\ln(-G_F)]\ .

2864: \ee

2865: We are here interested in the trace of the energy-momentum tensor, which is

2866: \be

2867: \frac{2}{\sqrt{-g}}g^{\mu\nu}\frac{\delta W}{\delta g^{\mu\nu}}=g^{\mu\nu}\langle T_{\mu\nu}\rangle

2868: =\langle T\rangle\ .

2869: \ee

2870: After laborious calculations \cite{duff} and using renormalization techniques,

2871: one obtains the following trace anomaly for the energy-momentum

2872: tensor

2873: \be\label{traceanomaly}

2874: \langle T\rangle=\frac{1}{2880\pi^2}\left[R_{\alpha\beta\gamma\delta}R^{\alpha\beta\gamma\delta}

2875: -R_{\alpha\beta}R^{\alpha\beta}+\Box R\right]\ .

2876: \ee

2877: We now specialize this anomaly to the Schwarzschild background \cite{chrifu}, obtaining

2878: \be

2879: \langle T\rangle =\frac{G_N^2}{60\pi^2}\frac{M^2}{R^6}\ .

2880: \ee

2881: These calculations are made having in mind that the vacuum is only described by a scalar field,

2882: but for more general fields one can find \cite{chridu}

2883: \be\label{anomalyTq}

2884: \langle T\rangle =-u\frac{G_N^2}{60\pi^2}\frac{M^2}{R^6}\ .

2885: \ee

2886: where

2887: \be

2888: u=12N_1-N_0-14N_{1/2}\ ,

2889: \ee

2890: and $N_{0,1/2,1}$ are respectively the number of species of spins $0,1/2,1$ of the theory considered.

2891: 

2892: Now we go back to the gravitational collapse in the braneworld. We can interpret the anomaly

2893: (\ref{m1}) of the Ricci scalar as due to a non-zero energy momentum tensor for the exterior

2894: which is extracting energy (via evaporation) from the collapsing object\footnote{This

2895: in fact is what happens quantum mechanically \cite{bida}.}. Indeed

2896: if we are not too close to the body and we consider only short periods in the evaporation

2897: process, as in standard

2898: semiclassical calculations, we can neglect the

2899: back-reaction on the metric and consider this process as nearly time-independent \cite{DeWitt}.

2900: 

2901: The braneworld calculation tells us that in order to release the potential energy

2902: of eq. (\ref{m1}), we must have an effective non-zero energy momentum tensor with trace

2903: \be

2904: T^{\rm eff}=-\frac{9}{16\pi^2\lambda}\frac{M^2}{R^6}\ .

2905: \ee

2906: But this is the same quantum anomaly as (\ref{anomalyTq})

2907: if we identify

2908: \be

2909: \lambda=\frac{135}{4uG_N^2}\ .

2910: \ee

2911: If we take this analogy seriously, then we have an indirect measure for the tension of the brane

2912: \be

2913: \lambda\sim 10 M_p^4 u^{-1}\ ,

2914: \ee

2915: where $M_p\sim 10^{19}\ \mbox{GeV}$ is the Planck mass. Moreover we obtain

2916: information on the theory which describes the Hawking process. Indeed since $\lambda$ is

2917: positive we should have

2918: \be\label{emp}

2919: N_1>\frac{N_0}{12}+\frac{7}{6}N_{1/2}\ .

2920: \ee

2921: As pointed out by \cite{Emparan}, this is incompatible with a SYM theory with ${\cal N}=4$ at

2922: large $N$, which is the quantum counterpart of the AdS/CFT correspondence \cite{M1}.

2923: This means that this process could be a new test of the holographic principle.

2924: 

2925: \newpage

2926: \chapter{Cosmology in a generalized braneworld}

2927: 

2928: \footnote{This chapter is based on \cite{barcelo, GS1, GS2, DG}.}In recent decades,

2929: developments in cosmology have been strongly

2930: influenced by high-energy physics.  A remarkable example of this

2931: is the inflationary scenario and all its variants.

2932: This influence has been growing and becoming more and more important.

2933: 

2934: It is a general belief that Einstein gravity is a

2935: low-energy limit of a quantum theory of gravity which is still unknown.

2936: Among promising candidates we have string theory, which suggests that

2937: in order to have a ghost-free action, quadratic curvature corrections to

2938: the Einstein-Hilbert action must be proportional to the Gauss-Bonnet

2939: term~\cite{BoDe:85}. An example has been given already in the first chapter.

2940: This term also plays a fundamental role in Chern-Simons

2941: gravitational theories~\cite{Cha:89}.

2942: However, although being a string-motivated scenario, the RS model and

2943: its generalizations~\cite{SMS} do not include these terms.

2944: From a geometric point of view, the combination of the

2945: Einstein-Hilbert and Gauss-Bonnet term constitutes, for 5D spacetimes,

2946: the most general Lagrangian producing second-order field

2947: equations~\cite{Lov} (see also~\cite{lovelock}).

2948: 

2949: These facts provide a strong motivation for the study of braneworld

2950: theories including a Gauss-Bonnet term.

2951: Recent investigations of this issue have shown~\cite{MeOl:00}

2952: that the metric for a vacuum 3-brane (domain wall) is, up to a redefinition

2953: of constants, the warp-factor metric of the RS scenarios. This is because AdS is conformally

2954: Minkowskian and the Gauss-Bonnet term is topological on the boundary. The existence

2955: of a KK zero-mode localized on the

2956: 3-brane producing Newtonian gravity at low energies, has also been demonstrated

2957: ~\cite{local} (see

2958: also~\cite{Ne:01a,ChNeWe:01}). This can be simply proved by considering that higher

2959: curvature terms in the action can produce only

2960: higher order corrections to the Newtonian gravity.

2961: Properties of black hole solutions in

2962: AdS spacetimes have been studied in~\cite{bhads,Ca:01}.  The cosmological

2963: consequences of these scenarios are less well understood.  This issue has

2964: been studied in~\cite{AbMo:01}. However, only simple ans\"atze for

2965: the 5D metric (written in Gaussian coordinates as

2966: in~\cite{bdl})  were considered, e.g. the separability of the

2967: metric components in the time and extra-dimension coordinates.

2968: One can see that this assumption is too strong even in

2969: RS cosmological scenarios~\cite{bdl}, where it leads to a

2970: very restrictive class of cosmological models, not representative of

2971: the true dynamics.

2972: Other results with higher-curvature terms in braneworld scenarios

2973: are considered in~\cite{others}.

2974: 

2975: In this chapter we study the equations governing the dynamics of

2976: FRW cosmological models in braneworld

2977: theories with a Gauss-Bonnet (GB) term (Lanczos gravity \cite{Lanczos}). In doing that

2978: we study the cosmological behaviour of shells

2979: (or branes) that are thin but still of a finite

2980: thickness $T$. In this way we want to shed some light on how

2981: the zero thickness is attained in the presence of GB interactions.

2982: This limit has been studied for Einstein gravity in \cite{Mounaix}.

2983: Thick shells in the context of GB interactions

2984: have been already studied in \cite{Corradini} and \cite{Giovannini}, but

2985: with a focus on different aspects than those here.

2986: The conclusion of our analysis here is twofold.

2987: On the one side, our results show that there is a generalized Friedmann equation

2988:  \cite{Charmousis} that can be found by using

2989: a completely general procedure, in which  the energy density of the brane

2990: in the thin-limit is related to the averaged density.  However,

2991: considering specific geometric configurations, one can find another form for

2992: the Friedmann equation, such as in \cite{GS1},

2993: with a procedure in which

2994: the energy density of the brane in the thin-limit comes from

2995: the value of the boundary density in the thick-brane model.

2996: We also argue that the information lost when treating a

2997: real thin shell

2998: as infinitely thin, is in a sense larger in Lanczos

2999: gravity than in the analogous situation in standard General Relativity.

3000: 

3001: Let us explain further this last point. From a physical point of view,

3002: in the process of passing from the notion of function to that of

3003: distribution, one loses information. Many different series of functions

3004: define the same limiting distribution. For example, the series

3005: %

3006: \begin{equation}

3007: f_T(y)=\left\{

3008: \begin{array}[c]{l}

3009: 0 ~~~{\rm for}~~~ |y| > T/2\,, \\

3010: \mbox{} \\

3011: {1 \over T}  ~~~{\rm for}~~~ |y| < T/2\,;

3012: \end{array}\right.

3013: ~~~~~~~~

3014: g_T(y)=\left\{

3015: \begin{array}[c]{l}

3016: 0 ~~~{\rm for}~~~ |y| > T/2\,, \\

3017: \mbox{} \\

3018: {12 y^2 \over T^3} ~~~{\rm for}~~~ |y| < T/2\,,

3019: \end{array}\right.

3020: \end{equation}

3021: %

3022: define the same limiting Dirac delta distribution. The distribution

3023: only takes into account the total conserved area delimited by the

3024: series of functions. The gravitational field equations relate geometry

3025: with matter content.  If we take the matter content to have some

3026: distributional character, the geometry will acquire also a

3027: distributional character. When analyzing the thin-limit of branes in

3028: Einstein gravity, by constructing series or families of solutions

3029: parametrized by their thickness, we observe that the divergent parts

3030: of the series of functions that describe the density-of-matter profile

3031: transfer directly to the same kind of divergent parts in the

3032: description of the associated geometry. Very simple density profiles

3033: [like the above function $f_T(y)$] are associated with very simple

3034: geometric profiles, and vice-versa.  However, when considering

3035: Lanczos gravity this does not happen. The divergent parts

3036: of the series describing the matter density and the geometry are

3037: inequivalent. A simple density profile does not correspond to a very

3038: simple geometric profile and vice-versa; on the contrary, we observe

3039: that they have some sort of complementary behaviour. This result

3040: leads us to argue that the distributional description of the

3041: cosmological evolution of a brane in Lanczos gravity

3042: is hiding important aspects of the microphysics, not present when

3043: dealing with pure Einstein gravity. Also, we find that for simple models

3044: of the geometry, one can make compatible the two seemingly distinct

3045: generalized Friedmann equations found in the literature.

3046: We will then extend these concepts for a domain wall formed by a scalar field.

3047: 

3048: 

3049: %---------------------------------------------------------------------

3050: \section{Static thick shells in Einstein and Lanczos gravity}

3051: \label{S:static}

3052: %---------------------------------------------------------------------

3053: 

3054: %---------------------------------------------------------------------

3055: \subsection{Einstein gravity}

3056: %---------------------------------------------------------------------

3057: 

3058: To fix ideas and notation let us first describe the simple case

3059: in which we have a static thick brane in an AdS

3060: bulk. We take an ansatz for the metric of the form

3061: %

3062: \begin{eqnarray}

3063: ds^2=  e^{-2A(y)} \eta_{\mu\nu}dx^\mu dx^\nu +dy^2\,,

3064: \label{brane}

3065: \end{eqnarray}

3066: %

3067: where $\eta_{\mu\nu}$ is the four-dimensional Minkowski metric.

3068: Comparing with the formulas given in Appendix~\ref{appa} this means

3069: taking $a(t,y)=n(t,y)=\exp(-2A(y))$, and $b(t,y)=1$.

3070: The energy-momentum tensor has the form

3071: %

3072: \begin{eqnarray}

3073: \tilde\kappa^2\, \tilde T_{AB} = \rho u_{A}u_{B} + p_Lh_{AB} + p_Tn_{A}n_{B}\,,

3074: \label{tAB}

3075: \end{eqnarray}

3076: %

3077: where $u_A=(-e^{2A},\mb{0},0)$ and $n_A=(0,\mb{0},1)$. Here, $\rho$, $p_L$

3078: and $p_T$ represent respectively the energy density, the longitudinal pressure

3079: and the transverse pressure, and are taken to depend only on $y$.

3080: The Einstein equations $\tilde G_{AB}=-\tilde\Lambda \tilde g_{AB}+\tilde\kappa^2 \tilde T_{AB}$

3081: with a negative cosmological constant,

3082: $\tilde\Lambda\equiv -6/l^2$,

3083: result in the following independent equations for the metric function

3084: $A(y)$:

3085: %

3086: \begin{eqnarray}

3087: &&3A''-6A'^2=\rho-{6 \over l^2}\,, \label{eins1}  \\

3088: &&6A'^2=p_T+{6 \over l^2}\,, \label{eins2} \\

3089: &&p_L=-\rho\,.

3090: \label{E-equations}

3091: \end{eqnarray}

3092: %

3093: For convenience, we will hide the $\tilde\kappa^2$ dependence inside the matter magnitudes,

3094: $\rho=\tilde\kappa^2\rho_{\rm true}$, etc.

3095: We also consider a $Z_2$-symmetric geometry around $y=0$.

3096: The brane extends in thickness from $y=-T/2$ to $y=+T/2$. Outside

3097: this region $\rho=p_T=0$, so we have a purely AdS spacetime:

3098: $A(y)=-y /l+b$ for $y\in (-\infty,-T/2)$ and $A(y)=y /l+b$ for

3099: $y\in (T/2,+\infty)$. The junction conditions at $y=-T/2,+T/2$ [see Appendix~\ref{appa}]

3100: tell us that

3101: %

3102: \begin{eqnarray}

3103: && A(-T^-/2)=A(-T^+/2),~~~~ A(T^-/2)=A(T^+/2) \,, \\

3104: && A'(-T^-/2)=A'(-T^+/2),~~~~ A'(T^-/2)=A'(T^+/2)\,.

3105: \label{E-junctions}

3106: \end{eqnarray}

3107: %

3108: From this and using (\ref{eins2}), we deduce that the

3109: transversal pressure is zero at the brane boundaries

3110: $p_T(-T/2)=p_T(T/2)=0$. Since we are imposing

3111: $Z_2$-symmetry with $y=0$ as {\em fixed point}, hereafter we

3112: will only specify the value of the different functions in the interval

3113: $(-T/2,0)$.

3114: 

3115: The function $A'$ is odd and therefore interpolates

3116: from $A'(-T/2)=-1/l$ to $A'(0)=0$. If in addition

3117: we impose that the null-energy condition $\rho+p_T=3A'' \geq 0$ be satisfied

3118: everywhere inside the brane, then $p_T$ has to be a negative and monotonically

3119: decreasing function from $p_T(-T/2)=0$ to $p_T(0)=-6/l^2$.

3120: This condition will turn out to be fundamental in defining

3121: a thin-shell limit.

3122: 

3123: By isolating $A''$ from equations (\ref{eins1}) and (\ref{eins2})

3124: we can relate the

3125: total bending of the geometry on passing through the brane

3126: with its total $\rho+p_T$

3127: %

3128: \begin{eqnarray}

3129: {6 \over l} =3A'\Big|_{-T/2}^{T/2}=\int_{-T/2}^{T/2}  (\rho+p_T)\; dy.

3130: \end{eqnarray}

3131: %

3132: At this stage of generality,

3133: one can create different one-parameter families of thick-brane versions

3134: of the Randall-Sundrum thin brane geometry, by parameterizing each

3135: member of a given family by its thickness $T$. The only requirement needed

3136: to do this is that the value of the previous integral must be kept fixed independently

3137: of the thickness of the particular thick-brane geometry.   Thus, each particular

3138: family can be seen as a regularization of Dirac's delta distribution.

3139: 

3140: 

3141: We can realize that, provided the condition $\rho+p_T \geq 0$ is satisfied,

3142: there exists a constant $C$, independent of the thickness $T$, such that

3143: $p_T<C$, that is, the profile for $p_T$ is bounded and will not

3144: diverge in the thin-shell limit.

3145: Therefore, in the limit in which the thickness of the branes goes to

3146: zero, $T\rightarrow 0$, the integral of $p_T$ goes to zero with the

3147: thickness. (Strictly speaking, the thin-shell limit is reached when

3148: $T/l\rightarrow 0$ but throughout this paper we are going to

3149: maintain $l$ as a finite constant.)  Instead, the profile of $\rho$ has to

3150: develop arbitrarily large values in order to fulfil

3151: %

3152: \begin{eqnarray}

3153: {6 \over l} =\lim_{T\rightarrow 0} \int_{-T/2}^{T/2}  \rho\;dy.

3154: \label{average}

3155: \end{eqnarray}

3156: %

3157: 

3158: 

3159: In the thin-shell limit, we can think of the Einstein equations as

3160: providing a relation between the characteristics of the density

3161: profile and the shape of the internal geometry. A very complicated

3162: density profile will be associated with very complicated function

3163: $A(y)$.  Physically we can argue that when a shell becomes very thin,

3164: one does not care about its internal structure and, therefore, one

3165: tries to describe it in the most simple terms. But what exactly is

3166: the meaning of simple? Here we will adopt two different definitions of

3167: simple: The first is to consider that the internal density is

3168: distributed homogeneously throughout the shell when the shell becomes

3169: very thin; the second is to consider that the profile for $A'$ is

3170: such that it interpolates from $A'(-T/2)=-1/l$ to $A'(0)=0$ through

3171: a straight line, or what is the same, that the internal profile of

3172: $A''$ is constant. Again, we require this for very thin shells. This

3173: geometric prescription is equivalent to asking for a constant internal

3174: scalar curvature, since $R=8A''-20A'^2$, and for every thin shell the term

3175: $A'^2$ is negligible relative to the constant $A''$

3176: term. Hereafter, we will use interchangeably the terms {\em straight

3177: interpolation} or {\em constant curvature} for these models. In

3178: building arbitrarily thin braneworld models, one needs the profiles

3179: for the internal density $\rho$ and the internal $A''$ to acquire

3180: arbitrarily high values (they will become distributions in the limit of

3181: strictly zero thickness). In the first of the two simple models

3182: described, the simplicity applies to the divergent parts of the

3183: matter content; in the second, the simplicity applies to the divergent parts of the geometry.

3184: From the physical point of view advocated in the introduction,

3185: simple profiles are those that do not involve losing information in the

3186: process of taking the limit of strictly zero thickness.

3187: 

3188: We analyze each case independently.

3189: 

3190: %---------------------------------------------------------------------

3191: \subsubsection{Constant density profile}

3192: %---------------------------------------------------------------------

3193: 

3194: We first define for convenience $z\equiv y/T$ as a scale invariant

3195: coordinate inside the brane. Then,

3196: mathematically, the idea that the density profile, which we will assume

3197: to be analytic inside the brane for simplicity, becomes constant

3198: in the thin-shell limit, can be expressed as follows:

3199: %

3200: \begin{eqnarray}

3201: \rho(z)=\sum_n \beta_n(T) \; z^{2n},

3202: \label{constant-density-est}

3203: \end{eqnarray}

3204: %

3205: where

3206: %

3207: \begin{eqnarray}

3208: \lim_{T\rightarrow 0} T ~ \beta_n(T)\rightarrow 0,

3209: ~~~~\forall n\neq 0;~~~~

3210: \lim_{T\rightarrow 0} T ~ \beta_0(T)\rightarrow \rho_b = {\rm constant}\,.

3211: \label{condi}

3212: \end{eqnarray}

3213: %

3214: For these density profiles, the Einstein equations in the

3215: thin-shell limit tell us that

3216: %

3217: \begin{eqnarray}

3218: 3A''=\beta_0(T)-{6 \over l^2}+6A'^2\,.

3219: \end{eqnarray}

3220: %

3221: From here we get the profile for $A'$:

3222: %

3223: \begin{eqnarray}

3224: A'=\sqrt{{\beta_0(T) \over 6}-{1 \over l^2}}

3225: \tan \left(2\sqrt{{\beta_0(T) \over 6}-{1 \over l^2}}\; y \right)\, .

3226: \label{apri}

3227: \end{eqnarray}

3228: %

3229: Notice that this expression only makes sense for $\beta_0(T)>6/l^2$,

3230: but this is just the regime we are interested in. We have to impose now

3231: the boundary condition $A'(T/2)=1/l$ on the previous expression (\ref{apri}),

3232: %

3233: \begin{eqnarray}

3234: {1 \over l}=\sqrt{{\beta_0(T) \over 6}-{1 \over l^2}}

3235: \tan \left(\sqrt{{\beta_0(T) \over 6}-{1 \over l^2}}\; T \right).

3236: \end{eqnarray}

3237: %

3238: In this manner, we have implicitly determined the form of the function

3239: $\beta_0(T)$. In the limit

3240: in which $T\rightarrow 0$ with $T\beta_0(T)\rightarrow \rho_b$,

3241: we find the relation

3242: %

3243: \begin{eqnarray}

3244: {6 \over l}=\rho_b.

3245: \end{eqnarray}

3246: %

3247: This condition is just what we expected from the average

3248: condition (\ref{average}).

3249: 

3250: 

3251: 

3252: 

3253: 

3254: 

3255: %---------------------------------------------------------------------

3256: \subsubsection{Straight interpolation}

3257: %---------------------------------------------------------------------

3258: 

3259: 

3260: In this case, the mathematical idea that in thin-shell limit

3261: the profile for $A'$ corresponds to a straight interpolation,

3262: can be formulated as

3263: %

3264: \begin{eqnarray}

3265: A''(z)=\sum_n \gamma_n(T) \; z^{2n},

3266: \label{straight-interpolation}

3267: \end{eqnarray}

3268: %

3269: with

3270: %

3271: \begin{eqnarray}

3272: \lim_{T\rightarrow 0} T ~ \gamma_n(T)\rightarrow 0,

3273: ~~~~\forall n\neq 0;~~~~

3274: \lim_{T\rightarrow 0} T ~ \gamma_0(T)\rightarrow {2 \over l}\ .

3275: \label{straight-interpolation-cond}

3276: \end{eqnarray}

3277: %

3278: For these geometries, we find that the associated profiles for

3279: $p_T$ and $\rho$ in the thin-shell limit have the form,

3280: %

3281: \begin{eqnarray}

3282: &&p_T=-{6 \over l^2}\left(1 - 4 z^2 \right) + \omega_1(T,z) \,, \\

3283: &&\rho={6 \over l T}+{6 \over l^2}\left(1 - 4 z^2 \right) +

3284: \omega_2(T,z)\,,

3285: \end{eqnarray}

3286: %

3287: where here and throughout this chapter, $\omega_n(T,z)$ denotes functions

3288: that vanish in the limit $T\rightarrow 0$.

3289: Now, from this density profile we can see that

3290: %

3291: \begin{eqnarray}

3292: \lim_{T\rightarrow 0} \int_{-T/2}^{T/2}  \rho\;dy = {6 \over l},

3293: \end{eqnarray}

3294: %

3295: as we expected.  Moreover, we can see that the boundary value of the

3296: density satisfies $T\rho\Big|_{T/2} \rightarrow 6/l$ in the thin

3297: shell limit, which is the same condition satisfied by the averaged

3298: density, $T \langle \rho \rangle \rightarrow 6/l$.

3299: 

3300: An additional interesting observation for what follows is the following.

3301: The set of profiles that yield constant density in the thin-shell limit

3302: (\ref{constant-density-est}) and straight interpolation for the

3303: geometric profile (\ref{straight-interpolation}) coincide. Therefore,

3304: in the thin-shell limit one can assume at the same time a constant

3305: internal structure for the density and a straight-interpolation for

3306: the geometry.

3307: 

3308: 

3309: 

3310: 

3311: 

3312: 

3313: 

3314: 

3315: %---------------------------------------------------------------------

3316: \subsection{Lanczos gravity}\label{EinsteinGB}

3317: %---------------------------------------------------------------------

3318: 

3319: Let us move now to the analysis of the same ideas in the presence

3320: of the Gauss-Bonnet term. The field equations are now \cite{Lanczos}

3321: \begin{eqnarray}

3322: \tilde G_{AB}+\alpha \tilde H_{AB}=-\tilde \Lambda \tilde g_{AB}+\tilde \kappa^2 \tilde T_{AB}

3323: \,, \label{fieldeq}

3324: \end{eqnarray}

3325: where $\tilde H_{AB}$ is the Lanczos tensor \cite{Lanczos}:

3326: %

3327: \begin{eqnarray}

3328: \tilde H_{AB}=2\tilde R_{ACDE} \tilde R_{B}^{\; CDE}-4\tilde R_{ACBD}\tilde R^{CD}

3329: -4\tilde R_{AC} \tilde R_{B}^{\; C}+2\tilde R\tilde R_{AB}-\frac{1}{2}\tilde g_{AB}L_{GB}

3330: \,, \label{habterm}

3331: \end{eqnarray}

3332: %

3333: where $\sqrt{-\tilde g}L_{GB}$ is the Gauss-Bonnet Lagrangian density,

3334: \begin{eqnarray}

3335: L_{GB} = \tilde R^2 -4\tilde R^{AB}\tilde R_{AB}+\tilde R^{ABCD}\tilde R_{ABCD} \,.

3336: \end{eqnarray}

3337: For the ansatz (\ref{brane}) we obtain (see Appendix~\ref{appa})

3338: %

3339: \begin{eqnarray}

3340: &&3A''(1 - 4\alpha A'^2)-6A'^2(1 - 2\alpha A'^2)=\rho-{6 \over l^2},

3341: \label{EGB-equations-rho} \\

3342: &&6A'^2(1 - 2\alpha A'^2)=p_T+{6 \over l^2}, \label{EGB-equations-pt} \\

3343: &&p_L=-\rho.

3344: \label{EGB-equations-pl}

3345: \end{eqnarray}

3346: %

3347: The junction conditions for the geometry are the same as before, eq.

3348: (\ref{E-junctions}), implying again the vanishing of the transversal

3349: pressure at the boundaries, $p_T=0$.

3350: 

3351: In the outside region the solution is a pure AdS spacetime

3352: but with a modified length scale

3353: %

3354: \begin{eqnarray}\label{modlength}

3355: {1 \over \tilde l} \equiv \sqrt{ {1 \over 4\alpha}

3356: \left(1-\sqrt{1- {8\alpha \over l^2}}\right) }.

3357: \end{eqnarray}

3358: %

3359: Now, isolating $A''$ from~(\ref{EGB-equations-rho}) and~(\ref{EGB-equations-pt}),

3360: we can relate the total bending of the geometry on passing through the brane

3361: with the integral of $\rho+p_T$ :

3362: %

3363: \begin{eqnarray}

3364: {6 \over \tilde l}\left(1- {4 \over 3}{\alpha \over {\tilde l}^2}\right)

3365: =(3A'-4 \alpha A'^3)\Big|_{-T/2}^{T/2}=\int_{-T/2}^{T/2}  (\rho+p_T)\; dy.

3366: \end{eqnarray}

3367: %

3368: Again, if the condition $\rho+p_T \geq 0$ is fulfilled throughout the brane

3369: we have that in the thin shell limit,

3370: %

3371: \begin{eqnarray}

3372: {6 \over \tilde l}\left(1- {4 \over 3}{\alpha \over {\tilde l}^2}\right)

3373: =\lim_{T \rightarrow 0} \int_{-T/2}^{T/2}  \rho\; dy.

3374: \label{gb-average-condition}

3375: \end{eqnarray}

3376: %

3377: At this point we can pursue this analysis in the two simple

3378: cases of constant density profile and straight interpolation.

3379: %

3380: 

3381: %---------------------------------------------------------------------

3382: \subsubsection{Constant density profile}

3383: %---------------------------------------------------------------------

3384: 

3385: 

3386: 

3387: Following the same steps as before for a constant density profile

3388: (\ref{constant-density-est})-(\ref{condi}), the equation that one

3389: has to solve in the thin-shell limit is

3390: %

3391: \begin{eqnarray}

3392: 3A''(1-4\alpha A'^2)=\beta_0(T)-{6 \over l^2}+6A'^2(1-2\alpha A'^2)\,.

3393: \end{eqnarray}

3394: %

3395: Introducing the notation $B\equiv A'$ we reduce this equation to the following

3396: integral

3397: %

3398: \begin{eqnarray}

3399: y={1 \over 4\alpha } \int_0^B {(4\alpha B^2-1) \; dB

3400: \over B^4 - {1 \over 2\alpha} B^2

3401: - {1 \over 2\alpha} \left({\beta_0(T) \over 6}-{1 \over l^2}\right)}\,.

3402: \end{eqnarray}

3403: %

3404: The result of performing the integration is

3405: %

3406: \begin{eqnarray}

3407: y={1 \over 2}

3408: \left[

3409: {1 \over \sqrt{-R_{-}}}\tan^{-1}\left({B \over \sqrt{-R_{-}}} \right)

3410: -{1 \over \sqrt{R_{+}}}\tanh^{-1}\left({B \over \sqrt{R_{+}}}\right)

3411: \right],

3412: \end{eqnarray}

3413: %

3414: where

3415: %

3416: \begin{eqnarray}

3417: R_{\pm}={1 \over 4\alpha }

3418: \left[ 1\pm \sqrt{1+ 8\alpha \left({\beta_0(T) \over 6}-{1 \over l^2}\right) }

3419: \right]\,.

3420: \end{eqnarray}

3421: %

3422: Again, by imposing the boundary condition

3423: %

3424: \begin{eqnarray}

3425: {T \over 2}={1 \over 2}

3426: \left[

3427: {1 \over \sqrt{-R_{-}}}\tan^{-1}\left({1 \over \tilde l \sqrt{-R_{-}}} \right)

3428: -

3429: {1 \over \sqrt{R_{+}}}\tanh^{-1}\left({1 \over \tilde l \sqrt{R_{+}}}\right)

3430: \right],

3431: \end{eqnarray}

3432: %

3433: we find the appropriate form for $\beta_0(T)$. With a lengthy but

3434: straightforward calculation, we can check that in the limit $T\rightarrow 0$,

3435: $\beta_0(T)\rightarrow \infty$, we have

3436: %

3437: \begin{eqnarray}

3438: T\beta_0(T)\rightarrow

3439: {6 \over \tilde l}\left(1- {4 \over 3}{\alpha \over {\tilde l}^2}\right),

3440: \end{eqnarray}

3441: %

3442: in agreement with condition (\ref{gb-average-condition}).

3443: 

3444: Using this same asymptotic expansion, we can see that, in the

3445: thin-shell limit, the profile for $A'(y)$ satisfies

3446: %

3447: \begin{eqnarray}

3448: A'(y)-{4 \alpha\over 3}A'(y)^3={1 \over 3} \beta_0(T)\; y.

3449: \end{eqnarray}

3450: %

3451: Recursively, one can create a Taylor expansion for $A'(y)$.

3452: The first two terms are given by

3453: %

3454: \begin{eqnarray}

3455: A'(y)&=&{1 \over 3} \beta_0(T)\; y+

3456: {4 \alpha \over 81} \beta_0(T)^3\; y^3+ {\cal O}(y^5)\cr&=&

3457: {1 \over 3} T \;\beta_0(T)\; z+

3458: {4 \alpha \over 81}T^3 \; \beta_0(T)^3\; z^3 + {\cal O}(z^5).

3459: \end{eqnarray}

3460: %

3461: By differentiating this expression we find

3462: %

3463: \begin{eqnarray}

3464: A''(y)={1 \over 3} \beta_0(T)+{4 \alpha \over 27}T^2 \; \beta_0(T)^3\; z^2

3465: +{\cal O}(z^4).

3466: \end{eqnarray}

3467: %

3468: Now, contrary to what happens in Einstein theory, this profile

3469: does not correspond to the set considered in the straight interpolation before

3470: (see

3471: fig. (\ref{F:fig1})).

3472: By looking at (\ref{straight-interpolation}) we can identify

3473: %

3474: \begin{eqnarray}

3475: \gamma_0(T)\equiv{1 \over 3} \beta_0(T), ~~~~

3476: \gamma_1(T)\equiv{4 \alpha \over 27} T^2 \; \beta_0(T)^3.

3477: \end{eqnarray}

3478: %

3479: Then, we can see that

3480: %

3481: \begin{eqnarray}

3482: \lim_{T \rightarrow 0} T\gamma_0(T)=

3483: {2 \over \tilde l}\left(1- {4 \over 3}{\alpha \over {\tilde l}^2}\right)

3484: \neq {2 \over \tilde l}, ~~~~

3485: \lim_{T \rightarrow 0} T\gamma_1(T)=

3486: {32 \alpha \over \tilde l^3}

3487: \left(1- {4 \over 3}{\alpha \over {\tilde l}^2}\right)^3 \neq 0 .

3488: \end{eqnarray}

3489: %

3490: The coefficients $\gamma_n$ do not satisfy the conditions in

3491: (\ref{straight-interpolation-cond}). Therefore, the scalar

3492: curvature does not have a constant profile as does the energy density.

3493: %====================================================

3494: \begin{figure}[htb]

3495: \vbox{

3496: \hfil

3497: \scalebox{0.5}{\includegraphics{constant-density1.eps}}

3498: \hfil

3499: }

3500: \bigskip

3501: \caption{\label{F:fig1} Family of constant density profiles with decreasing

3502: thickness and associated geometric profile for $A'$ and $A''$.}

3503: \end{figure}

3504: %====================================================

3505: 

3506: 

3507: 

3508: 

3509: %---------------------------------------------------------------------

3510: \subsubsection{Straight interpolation}

3511: %---------------------------------------------------------------------

3512: 

3513: As in the Einstein case, the straight interpolation profile for

3514: $A''$ corresponds to

3515: %

3516: \begin{eqnarray}

3517: A''(z)=\sum_n \gamma_n(T) \; z^{2n},

3518: \label{straight-interpolation2}

3519: \end{eqnarray}

3520: %

3521: with

3522: %

3523: \begin{eqnarray}

3524: \lim_{T\rightarrow 0} T ~ \gamma_n(T)\rightarrow 0,

3525: ~~~~\forall n\neq 0;~~~~

3526: \lim_{T\rightarrow 0} T ~ \gamma_0(T)\rightarrow {2 \over \tilde l}.

3527: \label{straight-interpolation2-cond}

3528: \end{eqnarray}

3529: %

3530: From here we can deduce the associated profiles for $p_T$ and $\rho$

3531: by substituting in (\ref{EGB-equations-rho}) and (\ref{EGB-equations-pt}).

3532: 

3533: In the limit $T \rightarrow 0$, the dominant part in the density profile

3534: is

3535: %

3536: \begin{eqnarray}

3537: \rho=3\gamma_0(T)(1-4\alpha \gamma_0(T)^2 \; T^2 \; z^2).

3538: \end{eqnarray}

3539: %

3540: Identifying

3541: %

3542: \begin{eqnarray}

3543: \beta_0(T) \equiv 3\gamma_0(T)\,, ~~~~

3544: \beta_1(T) \equiv -12 \alpha  T^2 \; \gamma_0(T)^3\,,

3545: \end{eqnarray}

3546: %

3547: we find that

3548: %

3549: \begin{eqnarray}

3550: \lim_{T \rightarrow 0} T \; \beta_0(T) = {6 \over \tilde l}, ~~~~

3551: \lim_{T \rightarrow 0} T \beta_1(T) \neq 0\,.

3552: \end{eqnarray}

3553: %

3554: Therefore, even in the thin-shell limit, a straight interpolation

3555: in the geometry does not correspond to a constant density profile (see fig.

3556: (\ref{F:fig2})).

3557: In the presence of a Gauss-Bonnet term it is not compatible to

3558: impose a simple description for the

3559: interior density profile and for the geometric warp factor

3560: at the same time.

3561: In the limit of strictly zero thickness (distributional limit), one will

3562: unavoidably lose some information on the combined matter-geometry system.

3563: %

3564: %====================================================

3565: \begin{figure}[htb]\label{straightfig}

3566: \vbox{

3567: \hfil

3568: \scalebox{0.5}{\includegraphics{straight-int1.eps}}

3569: \hfil

3570: }

3571: \bigskip

3572: \caption{\label{F:fig2} Plot of the straight interpolation profile

3573: for the geometric factor $A'$ and its associated density profile.}

3574: \end{figure}

3575: %====================================================

3576: 

3577: To finish this section let us make an additional observation.

3578: From expressions (\ref{straight-interpolation2}) and

3579: (\ref{straight-interpolation2-cond}),

3580: we can see that

3581: %

3582: \begin{eqnarray}

3583: A''= \frac{2}{T} A'\Big|^{}_{T/2} + \nu(T,z) ~~~~\mbox{with}~~~~~

3584: \lim_{T\rightarrow 0} T\nu(T,z) = 0   \,. \label{propsi}

3585: \end{eqnarray}

3586: Using this property in~(\ref{EGB-equations-rho},\ref{EGB-equations-pt}),

3587: %

3588: \begin{eqnarray}

3589: \rho +p_T=3A''(1 - 4 \alpha A'^2)={6 \over T} A'(T/2)(1 - 4 \alpha A'^2)\ ,

3590: \end{eqnarray}

3591: %

3592: and evaluating at $y=T/2$, we find

3593: %

3594: \begin{eqnarray}

3595: T\rho\Big|_{T/2}=6A'(1 - 4 \alpha A'^2)\Big|_{T/2}.

3596: \label{boundary-value}

3597: \end{eqnarray}

3598: %

3599: We can see that contrary to what happens in the Einstein case, this condition

3600: is different from the averaged condition (\ref{gb-average-condition}),

3601: %

3602: \begin{eqnarray}

3603: T\langle \rho \rangle=\lim_{T \rightarrow 0} \int_{T/2}^{-T/2}  \rho\; dy=

3604: (3A'-4 \alpha A'^3)\Big|_{-T/2}^{T/2}=

3605: 6A'(1-{4 \over 3} \alpha A'^2)\bigg|_{T/2}.

3606: \label{average3}

3607: \end{eqnarray}

3608: %

3609: Therefore, the averaged density and the boundary density are

3610: different, and this is independent of the brane thickness.  For this simple

3611: model, in thin-shell limit one can define two different internal

3612: density parameters characterizing the thin brane.

3613: One represents

3614: the total averaged internal density, and can be defined as

3615: %

3616: \begin{eqnarray}

3617: \rho_{\rm av} \equiv \lim_{T \rightarrow 0} T \langle \rho \rangle.

3618: \end{eqnarray}

3619: %

3620: The other represents an internal density parameter calculated by

3621: extrapolating to the interior the value of the density on the

3622: boundary. This density can be defined as

3623: %

3624: \begin{eqnarray}

3625: \rho_{\rm bv} \equiv \lim_{T \rightarrow 0} T \rho\Big|_{T/2}.

3626: \end{eqnarray}

3627: %

3628: The junction conditions for a thin shell

3629: \be

3630: \rho=\frac{6}{\tilde l}\left(1-\frac{4}{3}\frac{\alpha}{\tilde l^2}\right)

3631: \ee

3632: given in \cite{Low:2000pq},

3633: corresponds to the averaged condition (\ref{gb-average-condition}), or

3634: (\ref{average3})

3635: and therefore relates the total bending of the geometry

3636: in passing through the brane to its total averaged density.

3637: Instead, the particular condition analyzed for the boundary value

3638: of $\rho\Big|_{T/2}$ in eq. (\ref{boundary-value}) yields in the thin-shell

3639: limit the junction condition \cite{Germani}

3640: \be

3641: \rho=\frac{6}{\tilde l}\left(1-4\frac{\alpha}{\tilde l^2}\right)\ .

3642: \ee

3643: This condition is only considering information about the boundary value

3644: of the density and not about its average.

3645: 

3646: In summary, what this analysis suggests is that in the presence

3647: of the Gauss-Bonnet term, we can not ignore the interior

3648: structure of the brane, by modelling it by a simple model, even

3649: in the thin-shell limit. This point was first made in general terms by \cite{DeruDole}, and we have

3650: made it explicit. We will see again this feature

3651: in the next section on the cosmological dynamics of thick shells.

3652: 

3653: 

3654: 

3655: %---------------------------------------------------------------------

3656: \section{Dynamical thick shells in Einstein and Lanczos gravity}

3657: \label{S:dynamic}

3658: %---------------------------------------------------------------------

3659: 

3660: We use the class of spacetime metrics given in (\ref{themetric}),

3661: which contain a FRW metric in

3662: every hypersurface $\{y={\rm const}.\}$, with a matter content described by

3663: an energy-momentum tensor of the form (\ref{tAB}).  We consider

3664: the additional assumption of a {\em static} fifth dimension: $\dot{b} = 0$.

3665: We can rescale the coordinate $y$ in such a way that $b=1$.

3666: Then the line element (\ref{themetric}) becomes

3667: %

3668: \begin{equation}

3669: ds^2=-n^2(t,y)dt^2+a^2(t,y)h_{ij}dx^idx^j+dy^2\, .

3670: \end{equation}

3671: %

3672: In Appendix~\ref{appa} we show that the $\{ty\}-$component of the

3673: Lanczos field equations, for the case with a well-defined

3674: limit in Einstein gravity, leads to the equation~(\ref{keyrel}).

3675: In our case it implies the following relation:

3676: %

3677: \begin{equation}

3678: n(t,y)=\xi(t)\dot a(t,y)\,. \label{0y}

3679: \end{equation}

3680: %

3681: The remaining field equations can be written in the form

3682: given in~(\ref{ttcom},\ref{ijcom},\ref{yycom}).  In our case they

3683: become\footnote{The coupling constant $\alpha$ used here is one half the

3684: one used in \cite{GS1}.}

3685: %

3686: \begin{eqnarray}

3687: \left[a^4\left(\Phi+2\alpha\Phi^2+\frac{1}{l^2}\right)\right]' =

3688: \frac{1}{6}(a^4)'\rho  \,, \label{ttc}

3689: \end{eqnarray}

3690: \begin{eqnarray}

3691: \frac{n'}{n}\frac{\dot{a}}{a'}

3692: \left[a^4\left(\Phi+2\alpha\Phi^2+\frac{1}{l^2}\right)\right]'

3693: - {\left[a^4\left(\Phi+2\alpha\Phi^2+\frac{1}{l^2}\right)\right]^{\displaystyle{\cdot}}\,}' =

3694: 2\dot{a}a'a^2p_L\,, \label{ijc}

3695: \end{eqnarray}

3696: %

3697: %

3698: \begin{eqnarray}

3699: \left[a^4\left(\Phi+2\alpha\Phi^2+\frac{1}{l^2}\right)\right]^\cdot =

3700: -\frac{1}{6}(a^4)^{\displaystyle{\cdot}} p_T \,, \label{yyc}

3701: \end{eqnarray}

3702: %

3703: where now $\Phi$ is given by

3704: %

3705: \begin{eqnarray}

3706: \Phi=\frac{\dot a^2}{n^2 a^2}+\frac{k}{a^2}-\frac{a'^2}{a^2}

3707: = H^2 +\frac{k}{a^2} -\frac{a'^2}{a^2} \,,

3708: \end{eqnarray}

3709: %

3710: and we define the Hubble function

3711: associated with each $y={\rm const}.$ slice as

3712: %

3713: \begin{eqnarray}

3714: H(t,y)\equiv \frac{\dot a}{n a}\,.

3715: \end{eqnarray}

3716: %

3717: With the assumption $\dot{b}=0$, the field equation (\ref{ijc}) leads to a

3718: conservation equation for matter of the standard form

3719: [see eq.(\ref{5ceq})]:

3720: \begin{eqnarray}

3721: \dot{\rho} = -3\frac{\dot{a}}{a}(\rho+p_L) \,. \label{mconeq}

3722: \end{eqnarray}

3723: 

3724: 

3725: 

3726: Following the approach in the static scenario, we consider here the situation

3727: in which there is a $Z_2$ symmetry and a fixed proper thickness $T$

3728: for the brane.  Then one has to solve separately the equations for

3729: the {\it bulk}

3730: ($|y|>T/2$) and the equations for the thick {\it brane} ($|y|<T/2$).  The first

3731: step has already been done, and the result is~\cite{GS1}:

3732: \begin{eqnarray}

3733: \Phi+2\alpha\Phi^2+\frac{1}{l^2} = \frac{M}{a^4}\,,~~~~

3734: \mbox{for $|y|>\textstyle{T\over2}$}\,,

3735: \end{eqnarray}

3736: where, as we show in Appendix \ref{B2}, $M$ is a constant that can be identified

3737: with the mass of a black

3738: hole present in the bulk.  Once the solution inside the thick brane has been

3739: found, one has to impose the junction conditions~(\ref{cindm},\ref{cincm}) at

3740: $y=\pm T/2$.

3741: %

3742: 

3743: The first thing we can deduce from the junction conditions is that the

3744: quantity $\Phi$ is continuous across the two boundaries $y=\pm T/2\,.$

3745: But in general, its transversal derivative, $\Phi'$, will be discontinuous.

3746: Then, using equation (\ref{yyc}) it follows that the transversal pressure

3747: has to be zero on the boundary, $p_T(t,\pm T/2)=0$.  At the same time, from

3748: (\ref{yyc}) we deduce that we must always have

3749: %

3750: \begin{eqnarray}

3751: a^4\left(\Phi+2\alpha\Phi^2+\frac{1}{l^2}\right)\bigg|_{y=\pm T/2}=M\,.

3752: \label{intm}

3753: \end{eqnarray}

3754: %

3755: On the other hand, using again the relation~(\ref{keyrel}), we find that

3756: %

3757: \begin{eqnarray}

3758: H' = -\frac{a'}{a}H~~~~\Rightarrow~~~~\Phi' = -2\frac{a'}{a}

3759: \left(\Phi + \frac{a''}{a}\right)\,,

3760: \end{eqnarray}

3761: %

3762: and then, expanding (\ref{ttc}), we obtain

3763: %

3764: \begin{eqnarray}

3765: (1+4\alpha\Phi)\frac{a''}{a} = \Phi -\frac{2}{l^2} -\frac{1}{3}\rho

3766: \,.

3767: \label{mult}

3768: \end{eqnarray}

3769: %

3770: 

3771: 

3772: 

3773: 

3774: 

3775: 

3776: In the limit $T \rightarrow 0$, the profiles of the density $\rho$ and of

3777: $a''$ diverge, so that these dominant terms in expression~(\ref{mult})

3778: have to be equated. This results in

3779: %

3780: \begin{eqnarray}

3781: (1+4\alpha\Phi)\left({a' \over a}\right)'= -\frac{1}{3}\rho \,.

3782: \label{00-thin-limit}

3783: \end{eqnarray}

3784: %

3785: In what follows we consider the analysis of the Einstein and

3786: Lanczos theories separately.

3787: 

3788: 

3789: %---------------------------------------------------------------------

3790: \subsection{Einstein gravity}

3791: %---------------------------------------------------------------------

3792: 

3793: 

3794: In Einstein gravity it is not difficult to write down an equation describing

3795: the dynamics of every layer in the interior of a thick shell.

3796: To that end we take $\alpha=0$ in the equations above.

3797: By integrating (\ref{ttc}) over the interval $(-T/2,y_\ast)$,

3798: and using (\ref{intm},\ref{00-thin-limit}), we arrive at

3799: %

3800: \begin{eqnarray}

3801: \left(H^2 +{k \over a^2} + {1 \over l^2}  \right)&=&

3802: \left({a' \over a}\right)^2+

3803: {M \over a^4}+{1 \over 6 a^4}\int_{-T/2}^{y_\ast} (a^4)'\rho\, dy\cr&=&

3804: {1 \over 36} \left(\int_{-y_\ast}^{y_\ast} \rho\, dy \right)^{2}+

3805: {M \over a^4}+{1 \over 6a^4}\int_{-T/2}^{y_\ast} (a^4)'\rho\, dy\,.

3806: \label{layer-Einstein}

3807: \end{eqnarray}

3808: %

3809: A particular layer of matter inside the shell, located at $y=y_\ast$,

3810: can be seen as separating an internal spacetime from a piece of

3811: external spacetime.  From the previous equation, we can see that the

3812: cosmological evolution of each layer $y=y_\ast$ in the thick shell

3813: depends on the balance between the integrated density beyond the layer

3814: (external spacetime), and a weighted contribution of the integrated

3815: density in the internal spacetime. Therefore, the dynamics of each

3816: shell layer will be influenced by the particular characteristics of

3817: the internal density profile inside the shell. However, by looking at

3818: this same equation, we can see that the dynamics of the boundary layer

3819: $y_\ast=-T/2$ are only influenced by the total integrated density

3820: throughout the shell:

3821: %

3822: \begin{eqnarray}

3823: \left(H^2 +{k \over a^2} + {1 \over l^2}-

3824: {M \over a^4}  \right)\bigg|_{-T/2}=

3825: {1 \over 36}\left(\int_{-T/2}^{T/2} \rho dy\right)^{2}=

3826: {1 \over 36}\left(T\langle \rho \rangle \right)^{2}={1 \over 36}\rho_{\rm av}^2 .

3827: \end{eqnarray}

3828: %

3829: This is the modified Friedmann equation for the cosmological evolution

3830: of the brane \cite{bdl}.

3831: 

3832: 

3833: %---------------------------------------------------------------------

3834: %\subsubsection{Constant density profile}

3835: %---------------------------------------------------------------------

3836: 

3837: In the same manner as with static shells, we analyze

3838: the case in which the density profile tends to a time-dependent value

3839: in the thin-shell limit:

3840: %

3841: \begin{eqnarray}

3842: \rho =\beta_0(T,t)+ \omega_3(T,t,z), ~~~~

3843: \lim_{T \rightarrow 0} T \beta_0(T,t)=\rho_{\rm av}(t)\,, ~~~~

3844: \lim_{T \rightarrow 0} T\omega_3(T,t,z) =0\,.

3845: \label{constant-density}

3846: \end{eqnarray}

3847: %

3848: When $\alpha=0$, eq. (\ref{00-thin-limit}) tells us that if the

3849: density profile depends only on $t$ in the thin-shell limit, then, in

3850: this same limit, the divergent part of the geometry $(a'/a)'$ is also

3851: constant through the brane interior, describing what we called before a

3852: straight interpolation. A simple density profile amounts to a simple

3853: and equivalent geometric profile, and vice-versa. In this same case, but

3854: including the Gauss-Bonnet term, $\alpha \neq 0$, the geometrical factor $(a'/a)'$

3855: will exhibit a non-trivial profile in $y$, even in the thin-shell limit, as discussed

3856: in more detail in the next subsection.

3857: 

3858: 

3859: In the case in which $\rho$ depends only on time,

3860: eq. (\ref{layer-Einstein}) reads

3861: %

3862: \begin{eqnarray}

3863: \left.\left(H^2 +{k \over a^2} + {1 \over l^2}  \right)\right|_{y_\ast}=

3864: \left.\left({a' \over a}\right)^2\right|_{y_\ast}+ {M \over a^4(y_\ast)}+

3865: \frac{1}{6}\rho(t)\left.\left(1+ {a^4_{T/2} \over a^4}\right)\right|_{y_\ast}\,.

3866: \end{eqnarray}

3867: %

3868: From Equation (\ref{00-thin-limit}) we deduce that

3869: %

3870: \begin{eqnarray}

3871: \left.{a' \over a}\right|_{y_\ast}={1 \over 6}\rho_{\rm av} -

3872: {1 \over 3}\int_{-T/2}^{y_\ast} \rho \, dy

3873: =- {1 \over 3}\rho_{\rm av} z_\ast\,,

3874: \end{eqnarray}

3875: %

3876: and integrating we obtain

3877: %

3878: \begin{eqnarray}

3879: a(t,y)=a_0(t)\exp\left(-{1 \over 6}\rho_{\rm av}(t)Tz^{2}  \right)\ .

3880: \end{eqnarray}

3881: %

3882: (Remember that $z\equiv y/T$.)

3883: Therefore, at the lowest order in $T$ we have an equation for the

3884: internal geometry of the form

3885: %

3886: \begin{eqnarray}

3887: H_0^2+{k \over a_0^2}+{1 \over l^2}={1 \over 9}\rho_{\rm av}^2 z^2 +

3888: {1 \over 36}\rho_{\rm av}^2(1-4z^2)+\frac{M}{a_0^4}

3889: ={1 \over 36}\rho_{\rm av}^2+\frac{M}{a_0^4},

3890: \end{eqnarray}

3891: %

3892: which is exactly the standard braneworld generalized Friedmann equation~\cite{bdl}.

3893: 

3894: %---------------------------------------------------------------------

3895: \subsection{Lanczos gravity}

3896: %---------------------------------------------------------------------

3897: 

3898: In the general Lanczos case, eq. (\ref{00-thin-limit}) can be

3899: written as

3900: %

3901: \begin{eqnarray}

3902: \left[1+4\alpha\left(H^2+{k \over a^2}\right)\right]\left({a' \over a}\right)'

3903: -4\alpha \left({a' \over a}\right)^2 \left({a' \over a}\right)'

3904: =-{1 \over 3}\rho.

3905: \label{gb-expanded}

3906: \end{eqnarray}

3907: %

3908: Integrating between $-T/2$ and $T/2$ yields

3909: %

3910: \begin{eqnarray}

3911: 2\left[1+4\alpha\left(H^2+{k \over a^2}\right)\right]

3912: \left({a' \over a}\right)\bigg|_{T/2}

3913: -{8\alpha \over 3}\left({a' \over a}\right)^3 \bigg|_{T/2}

3914: =-{1 \over 3}\langle \rho \rangle T=-{1 \over 3}\rho_{\rm av}\,.

3915: \label{gb-blowup}

3916: \end{eqnarray}

3917: %

3918: The boundary equation (\ref{intm}) can be written as

3919: %

3920: \begin{eqnarray}

3921: \left.\left(H^2+{k \over a^2}\right)\right|_{T/2}-

3922: \left.\left({a' \over a}\right)^2\right|_{T/2}

3923: +2\alpha\left[\left(H^2+{k \over a^2}\right)-\left({a' \over a}\right)^2

3924: \right]^2_{T/2}-\cr-{M \over a^4(T/2)}+{1 \over l^2}=0\,. \label{bcegb}

3925: \end{eqnarray}

3926: %

3927: This is a quadratic equation for $(a'/a)^2\Big|_{T/2}$, with solutions

3928: %

3929: \begin{eqnarray}

3930: \left({a' \over a}\right)^2\bigg|_{T/2}={1 \over 4 \alpha}

3931: \left.\left[1+4\alpha\left(H^2+{k \over a^2}\right)\right|_{T/2}

3932: \pm

3933: \sqrt{1+{8\alpha \over l^2}-{8\alpha M \over a^4}} \right]\,.

3934: \end{eqnarray}

3935: %

3936: From these two roots we take only the minus sign, as it is the only

3937: one with a well-defined limit when $\alpha$ tends to zero.

3938: By squaring (\ref{gb-blowup}) and substituting the above solution

3939: we arrive at a cubic equation for $H^2+k/a^2$, first found

3940: in~\cite{Charmousis}. This cubic equation has a real root

3941: that can be expressed as~\cite{Gregory-Padilla}

3942:  %

3943: \begin{eqnarray}

3944: H^2+{k \over a^2}={1 \over 8 \alpha}

3945: \left[

3946: (\sqrt{\lambda^2+\zeta^3}+\lambda)^{2/3}

3947: +

3948: (\sqrt{\lambda^2+\zeta^3}-\lambda)^{2/3}

3949: -2

3950: \right]\,, \label{feqchar}

3951: \end{eqnarray}

3952: %

3953: where

3954: %

3955: \begin{eqnarray}

3956: \lambda \equiv \sqrt{\alpha \over 2} \rho_{\rm av}\,,

3957: ~~~~

3958: \zeta \equiv  \sqrt{1+8\alpha V(a) } \equiv

3959: \sqrt{1+{8\alpha \over l^2}-{8\alpha M \over a^4}}\,.

3960: \end{eqnarray}

3961: %

3962: In addition, we need the conservation equation

3963: %

3964: \begin{eqnarray}

3965: \dot \rho = 3 H (\rho+p_L)\,,

3966: \end{eqnarray}

3967: %

3968: which is valid for each section $y=y_*$, and in particular,

3969: for the boundary, $y=T/2$.

3970: This equation can be averaged to give

3971: %

3972: \begin{eqnarray}

3973: T\langle \dot \rho \rangle = 3 \langle H T (\rho+p_L) \rangle = 3 H\Big|_{T/2}

3974: (T \langle \rho \rangle+T \langle p_L \rangle)+{\cal O}(T),

3975: \end{eqnarray}

3976: %

3977: or written in another way,

3978: %

3979: \begin{eqnarray}

3980: \dot \rho_{\rm av}= 3 H\Big|_{T/2}(\rho_{\rm av}+p_{{\rm av}L})\ .

3981: \end{eqnarray}

3982: %

3983: This happens because

3984: %

3985: \begin{eqnarray}

3986: H(t,y)\rightarrow H_0(t,y_0)+{\cal O}(T)

3987: \end{eqnarray}

3988: %

3989: for any $y_0 \in[-T/2,T/2]$, which we have taken as $y_0=T/2$ for convenience.

3990: 

3991: We analyze now the simple case of a constant density profile

3992: (\ref{constant-density}). For consistency with the $T \rightarrow 0$

3993: case, we know that

3994: %

3995: \begin{eqnarray}

3996: a(t,y)=a_0(t)[1+T{\tilde a}(t,z)]+ {\cal O}(T^2)\,,

3997: \end{eqnarray}

3998: and therefore, from~(\ref{0y}),

3999: \begin{eqnarray}

4000: n(t,y)= \xi(t)\left[a_0\left(1+T\tilde{a}(t,z)\right)\right]^{\displaystyle{\cdot}}\,.

4001: \end{eqnarray}

4002: %

4003: In the same limit, eq. (\ref{00-thin-limit}) becomes

4004: %

4005: \begin{eqnarray}

4006: {\tilde a}_{,zz}= -{1\over 3}

4007: {\beta_0(T,t)T  \over \left[1+4\alpha

4008: \left(H_0^2+{k \over a_0^2}-{\tilde a}_{,z}^2 \right)\right]}.

4009: \end{eqnarray}

4010: %

4011: (Here the subscript $,z$ denotes differentiation with respect to $z$.)

4012: A necessary condition to have a straight interpolation for the geometry

4013: is that ${\tilde a}(t,z)=b(t)Z(z)$. To check whether or not a simple density

4014: profile corresponds to a simple geometrical profile, we can therefore try

4015: to solve this equation by separation of variables.

4016: It is not difficult to see that in order to find a solution

4017: with a well defined Einstein limit, we need

4018: %

4019: \begin{eqnarray}

4020: b(t) = \mu \,,

4021: ~~~~ \mu^{-1}\beta_0(T,t)T = \mu^{-1}\beta_0(T)T = \rho_{\rm av}= {\rm constant},

4022: ~~~~ H_0^2+{k \over a_0^2}=\Lambda_4\,,

4023: \end{eqnarray}

4024: %

4025: where $\mu$ is a constant that can be absorbed into the function $Z(z)$,

4026: and we can take it to be $\mu=1$.

4027: In this way we recover the AdS and dS solutions

4028: for the brane (depending on the sign of the effective four-dimensional

4029: cosmological constant).  To find the specific $y$ profile, we have to solve

4030: %

4031: \begin{eqnarray}

4032: Z_{,zz}= -{1\over 3}

4033: {\rho_{\rm av}  \over \left[1+4\alpha

4034: \left(\Lambda_4-Z_{,z}^2\right)\right]}.

4035: \end{eqnarray}

4036: %

4037: This equation can be integrated to give

4038: %

4039: \begin{eqnarray}

4040: (1+4\alpha\Lambda_4)Z_{,z}-{4\alpha \over 3}Z_{,z}^3=-{1\over 3}\rho_{\rm av}\, z\,.

4041: \end{eqnarray}

4042: %

4043: For our purposes the specific solution of this cubic equation is

4044: not important.  What we want to point out, is that the solution does not

4045: correspond to a straight interpolation as happened in the Einstein case.

4046: So, in general, simple solutions for the matter profile lead to non-trivial

4047: profiles for the scalar curvature even in the thin-shell limit.

4048: 

4049: Taking a simple model for the

4050: geometry, the straight interpolation model,

4051: %

4052: \begin{equation}

4053: a(t,z)=a_0(t)-\frac{1}{2}b(t)z^2T\ ,

4054: \label{req}

4055: \end{equation}

4056: %

4057: we deduce the density profile,

4058: using eq. (\ref{gb-expanded}),

4059: %

4060: \begin{eqnarray}

4061: \lim_{T \rightarrow 0}T \rho=

4062: 3\left[1+4\alpha\left(H_0^2+{k \over a_0^2}\right)\right]

4063: \left({b \over a_0}\right)-12\alpha\left({b \over a_0}\right)^3 z^2.

4064: \end{eqnarray}

4065: %

4066: As in the static case, even for very small thickness the density

4067: profile has now a non-trivial structure.

4068: We observe that

4069: %

4070: \begin{eqnarray}

4071: {a' \over a}\bigg|_{T/2}= -{1 \over 2}{b \over a_0} + {\cal O}(T) \,,

4072: ~~~~

4073: \left({a' \over a}\right)'= {2 \over T}\left[ {a' \over a}\bigg|_{T/2}

4074: +{\cal O}(T)\right]\,. \label{extraas}

4075: \end{eqnarray}

4076: %

4077: The second relation and eq.~(\ref{propsi}) coincide in the thin-shell

4078: limit. Therefore, evaluating~(\ref{gb-expanded}) on $y=T/2$, we obtain

4079: %

4080: \begin{eqnarray}

4081: \left[1+4\alpha\left(H^2+{k \over a^2}\right)\right]\left.\left({a' \over a}\right)

4082: \right|_{T/2}

4083: -4\alpha \left.\left({a' \over a}\right)^3\right|_{T/2}

4084: =-{1 \over 6}T \rho\Big|_{T/2}=-{1 \over 6}\rho_{\rm bv}\,. \label{cubic2}

4085: \end{eqnarray}

4086: %

4087: Following the same steps as before, but using this condition

4088: instead of eq. (\ref{gb-blowup}), we arrive at a cosmological generalized

4089: Friedmann equation \cite{GS1} different from that in~\cite{Charmousis} in its form

4090: and in the fact that it depends on the quantity associated with the boundary value of

4091: the energy density, $\rho_{\rm bv}$, instead of the value associated with the

4092: average of the energy density, $\rho_{\rm av}$.  Remarkably, the cubic equation

4093: that results from combining the last equation~(\ref{cubic2}) with the boundary

4094: condition~(\ref{bcegb}), becomes in this case linear.  That is, the coefficients of the

4095: terms quadratic and cubic in $H^2+k/a^2$ vanish~\cite{GS1}.  The modified

4096: Friedmann equation found in this case is:

4097: %

4098: \begin{eqnarray}\label{nonconv}

4099: H^2+{k \over a^2}=

4100: {1 \over \left(1+{8\alpha \over l^2}-{8\alpha M \over a^4}\right)}

4101: {1 \over 36} \rho_{\rm bv}^2 +{1 \over 4 \alpha}\left(

4102: \sqrt{1+{8\alpha \over l^2}-{8\alpha M \over a^4}}-1 \right)\,.

4103: \end{eqnarray}

4104: %

4105: In contrast with the modified Friedmann equation~(\ref{feqchar}), which

4106: was obtained by using a completely general procedure, in order to obtain

4107: this equation we had to use a procedure which required an

4108: extra assumption, namely equation~(\ref{extraas}). Hence it will

4109: not work  for profiles of the metric function $a(t,z)$ that do not

4110: satisfy these requirements or equivalent ones.   On the other hand,

4111: by looking at the developments here presented, we can conclude that the

4112: different results found in the literature

4113: for the dynamics of a distributional shell have their origin in the

4114: additional internal richness introduced in the brane by the presence

4115: of the GB term.

4116: 

4117: We have analyzed and compared how the thin-shell limit of static and

4118: cosmological braneworld models is attained in Einstein and

4119: Lanczos gravitational theories. We have seen that the

4120: generalized Friedmann equation proposed in~\cite{Charmousis} is always

4121: valid and relates the dynamical behaviour of the shell's boundary with

4122: its total internal density (obtained by integrating transversally the

4123: density profile).  Instead, the generalized Friedmann equation

4124: proposed in~\cite{GS1} relates the dynamical behaviour of the

4125: shell's boundary with the boundary value of the density within the

4126: brane. This equation is not always valid, only for specific

4127: geometrical configurations.

4128: 

4129: Einstein's equations in these models transfer the divergent

4130: contributions of the thin-shell internal density profile to the

4131: structure of the internal geometry in a faithful way. If we do not know

4132: the internal structure of the shell, we can always model it in simple

4133: terms by assuming an (almost) constant density profile and an (almost)

4134: constant internal curvature. However, the GB term makes it incompatible

4135: to have both magnitudes (almost) constant.  If the density is (almost)

4136: constant, then the curvature is not, and vice-versa. Therefore

4137: the particular structure of the Lanczos theory

4138: introduces important microphysical features into the matter-geometry

4139: configurations, beyond those in Einstein gravity, that are hidden in the

4140: distributional limit. We see this more clearly in the next section.

4141: 

4142: \section{Smooth flat brane model}

4143: 

4144: In this section, using an adequate definition of the brane stress-energy tensor,

4145: we confirm the results obtained previously for a Gauss-Bonnet brane model, extending

4146: the straight interpolation case.

4147: To do so, we use an approach

4148: directly based on the field

4149: equations for a smooth flat brane.

4150: 

4151: Suppose we can solve the problem of a five-dimensional scalar field with the metric

4152: \be

4153: ds^2=-e^{-2A_B(y)}d\eta^2+dy^2\ ,

4154: \ee

4155: where $d\eta^2$ is the four-dimensional Minkowski metric. Now the field equations read (see

4156: par. (\ref{EinsteinGB})):

4157: \ba

4158: 3A''_B\left(1-4\alpha A'^2_B\right)-p_B^t=\rho_B\ ,

4159: 6A'^2_B(1-2\alpha A'^2_B)-\frac{6}{l^2}=p^t_B\ ,\nonumber

4160: \ea

4161: where $A'_B(y)$ is a family of solutions parameterized by $B$, subject to the boundary

4162: condition

4163: \be

4164: \lim_{y\rightarrow\infty}A_B'(y)=\frac{1}{\tilde l}\ ,

4165: \ee

4166: ($\tilde l$ is defined in (\ref{modlength})), $\rho_B$ is the energy-density of the matter and

4167: $p_B^t$ is the transversal pressure in the $y$ direction. Moreover we have the ``total bending"

4168: condition

4169: \be

4170: \int^\infty_{-\infty}A''_B(y)dy=\frac{2}{\tilde l}\ .

4171: \ee

4172: Now the domain wall limit implies that

4173: \be

4174: \lim_{B\rightarrow\infty}A'_B(y)=\frac{1}{\tilde l}\ \mbox{for}\ \mid y\mid\geq T/2\ ,

4175: \ee

4176: where $T$ defines a ``proper thickness"\footnote{This is not uniquely defined, and can be

4177: for example interpreted as the proper variance of the distribution $A''_B(y)$,

4178: \be

4179: \sigma_B=\left[\int^\infty_{-\infty} A''_B(y)y^2dy/\int^\infty_{-\infty} A''_B(y)dy\right]^{1/2}

4180: \ .\nonumber

4181: \ee}

4182: of the smooth model and is eventually taken to zero.

4183: From it we have

4184: \be\label{totalbending}

4185: \lim_{B\rightarrow\infty}\int^{T/2}_{-T/2}A''_B(y)dy=\frac{2}{\tilde l}\ ,

4186: \ee

4187: and

4188: \be

4189: \lim_{B\rightarrow \infty}p_B^t=0\ \mbox{for}\ \mid y\mid\geq T/2\ .

4190: \ee

4191: The ``total bending" junction condition follows as

4192: \be

4193: \rho_{\rm av}=\lim_{T\rightarrow 0}T\langle \rho\rangle=\lim_{T\rightarrow 0,B\rightarrow

4194: \infty}\int^{T/2}_{-T/2}\rho_B(y)dy=\frac{6}{\tilde l}\left(1-\frac{4\alpha}{3\tilde l^2}

4195: \right)\ .

4196: \ee

4197: We now define another possibility for the junction conditions that we call ``holographic"

4198: junction

4199: conditions. From the integral (\ref{totalbending}) we have, using the average theorem for

4200: integrals,

4201: \be

4202: \lim_{B\rightarrow\infty}\int^{T/2}_{-T/2} A''_B(y)dy=\lim_{B\rightarrow\infty}A''_B(y_s)

4203: T=\frac{2}{\tilde l}\ ,

4204: \ee

4205: where $\mid y_s\mid\leq T/2$ and we call the hypersurface $y=y_s$ the ``screen".

4206: Therefore we have

4207: \be

4208: \lim_{B\rightarrow\infty}TA''_B(y_s)=\frac{2}{\tilde l}=2\lim_{B\rightarrow\infty}

4209: A'_B(T/2)\ .

4210: \ee

4211: Now considering the straight interpolation case we have that $y_s\sim T/2$ for $T\rightarrow 0$, then

4212: \be

4213: \rho_s=\lim_{T\rightarrow 0, B\rightarrow\infty} T\rho_B(y_s)=\frac{6}{\tilde l^2}\left(

4214: 1-\frac{4\alpha}{\tilde l^2}\right)\ .

4215: \ee

4216: This confirms that, even in the smooth model,

4217: \be

4218: \rho_{\rm av}\neq \rho_s\ .

4219: \ee

4220: \subsection{A simple explicit example}

4221: 

4222: Consider the following family of solutions

4223: \be

4224: A'_B(y)=\frac{1}{\tilde l}\tanh(By)\ ,

4225: \ee

4226: for which

4227: \be

4228: \rho_{\rm av}=\frac{6}{\tilde l^2}\left(1-\frac{4\alpha}{3\tilde l^2}\right)\ .

4229: \ee

4230: To find the screen, we can try to solve the equation

4231: \be

4232: \frac{2}{\tilde l}=TA''_B(y_s)\ ,

4233: \ee

4234: for $y_s$. This has a solution

4235: \be

4236: y_s=\frac{1}{B}\tanh^{-1}\sqrt{1-\frac{2}{BT}}\ .

4237: \ee

4238: Then we have

4239: \be

4240: A'_B(y_s)=\frac{1}{\tilde l}\sqrt{1-\frac{2}{TB}}\ ,

4241: \ee

4242: so that

4243: \be

4244: \lim_{B\rightarrow\infty}A'_B(y_s)=\lim_{B\rightarrow\infty}A'_B(T/2)=\frac{1}{\tilde l}\ .

4245: \ee

4246: We can therefore conclude

4247: \be

4248: \rho_s=\frac{6}{\tilde l}\left(1-\frac{4\alpha}{\tilde l^2}\right)\ .

4249: \ee

4250: 

4251: \section{Holographic description of the dark radiation term}

4252: 

4253: The holographic principle has been applied extensively to cosmological cases in the original

4254: Randall-Sundrum model (see \cite{padilla} for a review). At the time of writing this

4255: thesis, this concept

4256: has also been applied to the Gauss-Bonnet case (see e.g. \cite{Padilla,Gregory-Padilla}).

4257: Since this field is

4258: relatively new, in this section we concentrate on the more simple Randall-Sundrum scenario.

4259: 

4260: Here we show an example, without going too much into details, of how one can interpret,

4261: holographically,

4262: the Weyl contribution to the non-conventional Friedmann equation

4263: (\ref{nonconv}), in the case

4264: $\alpha=0$, when the Gauss-Bonnet term is switched off\footnote{See also \cite{SV} for a different perspective. Here

4265: the quantum mechanical properties of the 5D Schwarzschild-AdS black hole have been used.}. Splitting the energy density into the

4266: matter energy density ($\rho$)

4267: and brane tension ($\lambda$),

4268: $\rho_{\rm bv}=\rho+\lambda$, the modified Friedmann equation, replacing all the constants,

4269: reads (see also sec. \ref{star})

4270: \begin{equation}

4271: H^2=\frac{8\pi G_N}{3}\rho\left(1+\frac{\rho}{2\lambda}\right)+\frac{M}{a^4}-\frac{k}{a^2}

4272: +\frac{1}{3}\Lambda\ .

4273: \end{equation}

4274: In order to apply the AdS/CFT description, we will set $k=\Lambda=0$. We follow again

4275: \cite{bida}.

4276: 

4277: We consider a four-dimensional cosmological model with a Friedmann geometry,

4278: \begin{equation}

4279: ds^2=a(\eta)^2\left[-d\eta^2+d\vec{x}\cdot d\vec{x}\right]\ .

4280: \end{equation}

4281: Since the Universe is accelerating (or decelerating) it produces particles by quantum

4282: vacuum fluctuations. These particles will

4283: back-react on the metric, producing a non-isotropic perturbation of the type

4284: \be

4285: ds^2=a(\eta)^2\left[-d\eta^2+\sum_{i=1}^3\left[1+h_i(\eta)\right] (dx^i)^2\right]\ ,

4286: \ee

4287: where $\mbox{max}\mid h_i(\eta)\mid\ll 1$. For simplicity we will use the constraint

4288: $\sum_{i=1}^3 h_i(\eta)=0$. Considering

4289: the vacuum as a massless scalar field $\phi(x)$, we can decompose it in Fourier space as

4290: \be

4291: \phi(x)=\int d^3k \left[a_k u_k(x)+a_k^\dag u^*_k(x)\right]\ ,

4292: \ee

4293: where $\dag$ means the Hermitian conjugate and $*$ the complex conjugate.

4294: 

4295: Given the symmetries of the problem, we can use the following separation of variables,

4296: \be

4297: u_k=(2\pi)^{-3/2}e^{ik\cdot x}\frac{\chi_k(\eta)}{a(\eta)}\ .

4298: \ee

4299: Now the evolution equation (\ref{evolscalar}) in the conformal case and in conformal

4300: time reduces to

4301: \be

4302: \frac{d^2\chi_k}{d\eta^2}-\sum^3_{i=1}k_i^2\chi_k=0\ .

4303: \ee

4304: If we also impose the orthonormality of the $u_k$,$u^*_k$,

4305: we have the conditions

4306: \be

4307: \chi_k\partial_\eta\chi^*_k-\chi^*_k\partial_\eta\chi_k=i\ .

4308: \ee

4309: Since we would like an asymptotically flat spacetime, we also impose

4310: \be

4311: \lim_{\eta\rightarrow\infty}h_i(\eta)=0\ .

4312: \ee

4313: The normalized positive frequency solution for $\eta\rightarrow -\infty$ is

4314: \be

4315: \chi_k^{\rm in}(\eta)=(2k)^{-1/2}e^{-ik\eta}\ .

4316: \ee

4317: Then we can write the integral equation

4318: \be\label{chi}

4319: \chi_k(\eta)=\chi_k^{\rm in}(\eta)+k^{-1}\int_{-\infty}^\eta V_k (\eta')\sin\left[k(\eta-\eta')

4320: \right]

4321: \chi_k(\eta')d\eta'\ ,

4322: \ee

4323: with $V_k=\sum_{i=1}^3 k_i^2 h_i(\eta)$. The production of particles will be in the

4324: late region ($\eta\rightarrow\infty$),

4325: and

4326: \be

4327: \chi_k^{\rm out}(\eta)=\alpha_k\chi^{\rm in}_k(\eta)+\beta_k\chi^{\rm in\ *}_k(\eta)\ ,

4328: \ee

4329: where the Bogolubov coefficients to first order in $V_k$ are

4330: \ba

4331: \alpha_k &=& 1+i\int^\infty_{-\infty}\chi^{\rm in\ *}_k(\eta)V_k(\eta)\chi_k(\eta)d\eta\cr

4332: \beta_k &=& -i\int^\infty_{-\infty}\chi^{\rm in}_k(\eta)V_k(\eta)\chi_k(\eta)d\eta\

4333: .\nonumber

4334: \ea

4335: The number of particles created per proper volume ($n$) will be linked to the probability

4336: of non-zero values for $\chi_k$ in the

4337: output region. Then

4338: \be

4339: n=(2\pi a)^{-3}\int\mid\beta_k\mid^2d^3k\ ,

4340: \ee

4341: with an associated energy density

4342: \be

4343: \rho=(2\pi)^{-3}a^{-4}\int \mid \beta_k\mid^2 k d^3 k\ .

4344: \ee

4345: The second-order approximation in $V_k$ gives

4346: \be

4347: \rho=-(3840\pi^2 a^4)^{-1}\int^\infty_{-\infty}d\eta_1\int^{\infty}_{-\infty}

4348: d\eta_2\ln\left[2i(\eta_1-\eta_2)\right]

4349: \sum_i(\partial^3_\eta h_i(\eta_1))(\partial^3_\eta h_i(\eta_2))\ .

4350: \ee

4351: Physically the perturbation will be a composition of damped oscillating functions of the form

4352: \be

4353: h_i(\eta)=e^{-\alpha\eta^2}\cos(\beta\eta^2+\delta_i)\ ,

4354: \ee

4355: where $\delta_i-\delta_{i+1}=2/3\pi$. Then we finally have

4356: \be

4357: \rho=\frac{1}{2880\pi}\frac{(\alpha^2+\beta^2)^{1/2}}{\alpha^3 a^4}\ .

4358: \ee

4359: If we now associate

4360: \be

4361: \frac{1}{2880\pi}\frac{(\alpha^2+\beta^2)^{1/2}}{\alpha^3}=M\ ,

4362: \ee

4363: we have an holographic description of the projected Weyl tensor on the brane. Since this

4364: quantum

4365: effect is due only to the geometry, independently of the matter content, it is physically

4366: reasonable

4367: to connect it with the projected Weyl tensor, which is a purely geometrical quantity

4368: dependent only on the bulk

4369: curvature.

4370: 

4371: \newpage

4372: \chapter{Conclusions}

4373: 

4374: In this thesis I studied the Randall-Sundrum mechanism of localization of gravity in the presence

4375: of an infinitely large extra dimension, from the astrophysical and cosmological point of view.

4376: I focused on the exact basic models. As a

4377: motivation for studying these scenarios in light of the holographic principle,

4378: I showed their relations with known quantum

4379: solutions in four-dimensional gravity.

4380: 

4381: \subsection*{Astrophysics}

4382: I investigated how 5-dimensional gravity can affect static

4383: stellar solutions on the brane. I found exact braneworld

4384: generalizations of the uniform-density stellar solution, and used

4385: this to estimate the local (high-energy) effects of bulk gravity.

4386: I derived astrophysical lower limits on the brane tension.

4387: I also found that the star is less

4388: compact than in general relativity. The smallness of high-energy corrections to

4389: stellar solutions flows from the fact that $\lambda$ is well above

4390: the energy density $\rho$ of stable stars. However nonlocal

4391: corrections from the bulk Weyl curvature (5-dimensional graviton

4392: effects) have qualitative implications that are very different

4393: from general relativity.

4394: 

4395: The Schwarzschild solution is no longer the unique asymptotically

4396: flat vacuum exterior; in general, the exterior carries an imprint

4397: of nonlocal bulk graviton stresses. The exterior is not uniquely

4398: determined by matching conditions on the brane, since the

4399: 5-dimensional metric is involved via the nonlocal Weyl stresses.

4400: I demonstrated this explicitly by giving two exact exterior

4401: solutions, both asymptotically Schwarzschild. Each exterior which

4402: satisfies the matching conditions leads to a bulk metric, which

4403: could in principle be determined locally by numerical integration.

4404: Without any exact or

4405: approximate 5-dimensional solutions to guide us, we do not know

4406: how the properties of the bulk metric, and in particular its

4407: global properties, will influence the exterior solution on the

4408: brane.

4409: 

4410: Guided by perturbative analysis of the static weak field

4411: limit, I made the following conjecture:

4412: {\em if the bulk for a static stellar solution on the brane is

4413: asymptotically AdS and has regular Cauchy horizon, then the

4414: exterior vacuum which satisfies the matching conditions on the

4415: brane is uniquely determined, and agrees with the perturbative

4416: weak-field results at lowest order.} An immediate implication of

4417: this conjecture is that the exterior is not Schwarzschild, since

4418: perturbative analysis shows that there are nonzero Weyl stresses

4419: in the exterior (these stresses are the manifestation

4420: on the brane of the massive Kaluza-Klein bulk graviton modes). In

4421: addition, the two exterior solutions that I present would be

4422: ruled out by the conjecture, since both of them violate the

4423: perturbative result for the weak-field potential.

4424: 

4425: The static problem is already complicated, so that analysis of

4426: dynamical collapse on the brane can be very difficult. However,

4427: the dynamical problem gives rise to more striking features.

4428: Energy densities well above the brane tension could be reached

4429: before horizon formation, so that high-energy corrections could be

4430: significant.

4431: In this direction I explored the consequences for

4432: gravitational collapse of braneworld gravity effects, using the

4433: simplest possible model, i.e.\ an Oppenheimer-Snyder-like collapse on a

4434: generalized Randall-Sundrum type brane. Even in this simplest

4435: case, extra-dimensional gravity introduces new features. Indeed using

4436: only the projected 4D equations, I have shown, independent of the

4437: nature of the bulk, that the exterior vacuum on the brane is

4438: necessarily {\em non-static}. This contrasts strongly with GR,

4439: where the exterior is a static Schwarzschild spacetime. Although

4440: I have not found the exterior metric, I know that its non-static

4441: nature arises from (a)~5D bulk graviton stresses, which transmit

4442: effects nonlocally from the interior to the exterior, and (b)~the

4443: non-vanishing of the effective pressure at the boundary, which

4444: means that dynamical information on the interior side can be

4445: conveyed outside. My results suggest that gravitational collapse

4446: on the brane may leave a signature in the exterior, dependent upon

4447: the dynamics of collapse, so that astrophysical black holes on the

4448: brane may in principle have KK hair.

4449: 

4450: I expect that the non-static property of the exterior will be transient and {\em

4451: non-radiative}, as follows from a perturbative study of non-static

4452: compact objects, showing that the Weyl term ${\cal E}_{\mu\nu}$ in

4453: the far-field region falls off much more rapidly than a radiative

4454: term. It is reasonable to assume that the exterior

4455: metric will tend to be static at late times and tend to Schwarzschild,

4456: at least  at large distances.

4457: 

4458: Moreover I showed that this non-static behaviour is due to an anomaly of the Ricci scalar.

4459: Indeed assuming a static exterior,

4460: I found that in the absence of the cosmological constant, it does not vanish as one must expect

4461: from

4462: a vacuum solution. This means that there is a ``potential energy" stored on the boundary of

4463: the star. This energy must

4464: be released in some way producing a time signature in the exterior. Since this anomaly resembles

4465: very much the Weyl anomaly due to the Hawking process, it is reasonable to conjecture

4466: that the time signal can be

4467: holographically reproduced by an ``evaporation process". Using this holographic point of view

4468: I derived an indirect measure for the brane vacuum energy, $\lambda\sim 10 M_p^4 u^{-1}$,

4469: where $u$ depends on the number of fields involved in the Hawking process.

4470: 

4471: \subsection*{Cosmology}

4472: 

4473: I analyzed and compared how the thin-shell limit of static and

4474: cosmological braneworld models is attained in Einstein and

4475: Lanczos gravitational theories. I showed that the

4476: generalized Friedmann equation proposed in~\cite{Charmousis} is always

4477: valid and relates the dynamical behaviour of the shell's boundary with

4478: its total internal density (obtained by integrating transversally the

4479: density profile).  By contrast, the generalized Friedmann equation

4480: proposed in~\cite{Germani} relates the dynamical behaviour of the

4481: shell's boundary with the boundary value of the density within the

4482: brane. This equation is not always valid, but only for specific

4483: geometrical configurations.

4484: 

4485: Einstein equations in these models transfer the divergent

4486: contributions of the thin-shell internal density profile to the

4487: structure of the internal geometry in a faithful way. If one does not know

4488: the internal structure of the shell, one can always model it in simple

4489: terms by assuming an (almost) constant density profile and an (almost)

4490: constant internal curvature. However, the Gauss-Bonnet term makes it incompatible

4491: to have both magnitudes (almost) constant.  If the density is (almost)

4492: constant, then the curvature is not, and vice-versa. Therefore, one

4493: can say that the particular structure of the Lanczos theory

4494: introduces important microphysical features to the matter-geometry

4495: configurations beyond those in Einstein gravity, that are hidden in the

4496: distributional limit.

4497: 

4498: Studying a smooth flat brane model I generalized these conclusions for a more physical model.

4499: In the particular example I gave, one can introduce two types of junction condition

4500: which relate the

4501: ``total bending" of the brane with the matter content. The one I called the

4502: ``total bending"

4503: junction condition that relates the bending to the total matter inside the brane. The second

4504: junction condition relates instead the matter content in a particular

4505: hypersurface called the ``screen",

4506: with the total bending. In this screen all the information of the

4507: total bending is stored. I call this a ``holographic" junction condition.

4508: 

4509: Independently of the gravitational

4510: theory used, I showed how to derive the modified Friedmann equation and how it is related

4511: to the black hole solution of the theory.

4512: 

4513: In particular for the simplest case of the

4514: Randall-Sundrum scenario I showed how to interpret holographically the black hole mass using quantum particle

4515: production with an FRW geometry.

4516: 

4517: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

4518: 

4519: 

4520: \appendix

4521: \newpage

4522: \chapter{Junction conditions}\label{Appjc}

4523: 

4524: In this appendix we describe the junction conditions of a non-null hypersurface $\Sigma$,

4525: following

4526: \cite{Synge}.

4527: We use the principle of the continuity of geodesics across any

4528: hypersurface. Since the geodesic equation involves at most first-order derivatives of the metric,

4529: we require the continuity of the metric together with its first derivative.

4530: 

4531: If $n$ is the unit normal

4532: vector of the hypersurface $\Sigma$, it is always possible, at least locally, to define Gaussian

4533: normal coordinates such that $x^n$ is the adapted coordinate to $n$ and

4534: $a,b,c...$ are the adapted coordinates to $\Sigma$. Then the following are continuous quantities:

4535: \be

4536: g_{\alpha\beta},\ g^{\alpha\beta}\ ,\ \partial_\gamma g_{\alpha\beta}\ ,\ \partial_{\gamma}

4537: g^{\alpha\beta}\ ,

4538: \ \partial_{an}g^{\alpha\beta}\ ,\ \partial_{ab}

4539: g_{\alpha\beta}\ ,

4540: \ \partial_{ab}g^{\alpha\beta}{}\ .

4541: \ee

4542: It follows that the $G^n_\alpha$ component must be continuous.

4543: In a covariant form,

4544: \be\label{Appjc2}

4545: [G^\alpha{}_\beta n^\beta]\Big|_{\Sigma}=0\ ,

4546: \ee

4547: where $[f]\Big|_\Sigma\equiv f\Big|_{\Sigma^+}-f\Big|_{\Sigma^-}$,

4548: and $\Sigma^\pm$ are respectively

4549: the outer and inner face of $\Sigma$. Equivalently

4550: \ba

4551: [g_{\alpha\beta}]\Big|_\Sigma = 0\ ,\

4552: [K_{\alpha\beta}]\Big|_\Sigma = 0\ ,\label{jcK}

4553: \ea

4554: where $K_{\alpha\beta}=\pounds_n g_{\alpha\beta}/2$ is the extrinsic curvature of $\Sigma$.

4555: 

4556: We now work out the so-called Israel junction conditions for thin shells in this case.

4557: Since we are interested in $Z_2$-symmetric branes, this considerably simplifies the calculations.

4558: 

4559: Suppose we have two parallel hypersurfaces $\Sigma_1$ and $\Sigma_2$. We again

4560: introduce Gaussian normal coordinates, where the coordinate $x^n$ defines the orthogonal

4561: direction of both $\Sigma_1$ and $\Sigma_2$ . Moreover we fix the point $x^n=0$ as the centre

4562: point, and we call $T$ the proper distance between the hypersurfaces.

4563: In these coordinates, the $Z_2$ symmetry means

4564: \ba

4565: g_{\alpha\beta}(x^n)&=&g_{\alpha\beta}(-x^n)\ ,\cr

4566: K_{\alpha\beta}(x^n)&=&-K_{\alpha\beta}(-x^n)\ . \nonumber

4567: \ea

4568: Therefore from (\ref{jcK}) we have

4569: \ba

4570: K_{\alpha\beta}\Big|_{\Sigma_1}=-K_{\alpha\beta}\Big|_{\Sigma_2}\ ,

4571: \ea

4572: or equivalently, making explicit only the $x^n$ dependence of $K_{\alpha\beta}$,

4573: \be

4574: K_{\alpha\beta}(T/2)=-K_{\alpha\beta}(-T/2)\ .

4575: \ee

4576: We consider the incremental ratio in the thin shell limit ($T\rightarrow 0$),

4577: \be

4578: \lim_{T\rightarrow 0}\frac{K_{\alpha\beta}(T/2)-K(-T/2)}{T}=\partial_{x^n}K_{\alpha\beta}(0)\ .

4579: \ee

4580: This behaves like a Dirac delta function. It is a straightforward exercise to show

4581: that if $\Phi(x^n)$ is a smooth test function, then

4582: \be

4583: \int^{T/2}_{-T/2}\partial_{x^n}K_{\alpha\beta} \Phi(x)dx=2K_{\alpha\beta}(0) \Phi(0)\ .

4584: \ee

4585: Therefore in the thin limit, $K_{\alpha\beta}$ ``jumps" or more rigorously\footnote{Since

4586: in this limit $\Sigma_1\rightarrow \Sigma_2$, $\Sigma$ denotes one of the two equivalent

4587: hypersurfaces.}

4588: \be

4589: [K_{\alpha\beta}]=2K_{\alpha\beta}\Big|_{\Sigma}\ ,

4590: \ee

4591: where here we define $[f]=f(0^+)-f(0^-)$.

4592: 

4593: \newpage

4594: \chapter{5D geometry}

4595: \section{Lanczos gravity}\label{appa}

4596: In this appendix we present the main geometrical quantities and field

4597: equations associated with the 5D metric

4598: \begin{eqnarray}

4599: ds^2 = \tilde g_{AB}dx^Adx^B = -n^2(t,y)dt^2 + a^2(t,y)h_{ij}(x^k)dx^idx^j + b^2(t,y)dy^2\,,

4600: \label{themetric}

4601: \end{eqnarray}

4602: where $h_{ij}$ is the metric of the three-dimensional maximally symmetric

4603: surfaces $\{y=\mbox{const.\}}$, whose spatial curvature is parametrized

4604: by $k=-1,0,1$.  A particular representation of $h_{ij}$ is

4605: \begin{eqnarray}

4606: h_{ij}dx^idx^j = \frac{1}{\left(1+\frac{k}{4}r^2\right)^2}\left(dr^2 +

4607: r^2d\Omega^2_2\right)\,,

4608: \end{eqnarray}

4609: where $d\Omega^2_2$ is the metric of the 2-sphere.  The metric~(\ref{themetric})

4610: contains as particular cases the metrics used in this thesis.

4611: 

4612: The non-zero components of the Einstein tensor $\tilde G_{AB}$ corresponding to this line

4613: element are given by ($\dot{Q}=\partial_tQ$, $Q'=\partial_yQ$):

4614: \begin{eqnarray}\label{second}

4615: \tilde G_{tt} & = & 3\left\{ n^2\Phi + \frac{\dot{a}}{a}\frac{\dot{b}}{b}-

4616: \frac{n^2}{b^2}\left[\frac{a''}{a}-\frac{a'}{a}\frac{b'}{b}\right]\right\}

4617: \,, \cr

4618: \tilde G_{ty} & = & 3\left( \frac{\dot{a}}{a}\frac{n'}{n}+\frac{a'}{a}\frac{\dot{b}}{b}

4619: -\frac{\dot{a}'}{a}\right) \,, \cr

4620: \tilde G_{ij} & = & \frac{a^2}{b^2}h_{ij}\left\{\frac{a'}{a}\left(\frac{a'}{a}

4621: +2\frac{n'}{n}\right)-\frac{b'}{b}\left(\frac{n'}{n}+2\frac{a'}{a}\right)

4622: +2\frac{a''}{a} + \frac{n''}{n}\right\}

4623: \cr

4624: & - & \frac{a^2}{n^2}h_{ij}\left\{\frac{\dot{a}}{a}\left(\frac{\dot{a}}{a}-

4625: 2\frac{\dot{n}}{n}\right) - \frac{\dot{b}}{b}\left(\frac{\dot{n}}{n}

4626: -2\frac{\dot{a}}{a}\right) + 2\frac{\ddot{a}}{a}+\frac{\ddot{b}}{b} \right\}

4627: - kh_{ij}\,, \cr

4628: \tilde G_{yy} & = & 3\left\{-b^2\Phi + \frac{a'}{a}\frac{n'}{n} - \frac{b^2}{n^2}

4629: \left[\frac{\ddot{a}}{a}-\frac{\dot{a}}{a}\frac{\dot{n}}{n}\right]\right\} \,,

4630: \end{eqnarray}

4631: where

4632: \begin{equation}

4633: \Phi(t,y) = \frac{1}{n^2}\frac{\dot{a}^2}{a^2}-\frac{1}{b^2}\frac{a'^2}{a^2}

4634: +\frac{k}{a^2}\,. \label{defphi}

4635: \end{equation}

4636: Apart from the metric and the Einstein tensor, the field equations in

4637: Lanczos gravity~(\ref{fieldeq}) contain a term quadratic in the

4638: curvature, namely $\tilde H_{AB}$ [see eq.~(\ref{habterm})].

4639: The non-zero components of this tensor can be written as

4640: \begin{eqnarray}

4641: \tilde H_{tt} & = & 6\Phi\left[

4642: \frac{\dot{a}}{a}\frac{\dot{b}}{b}+\frac{n^2}{b^2}\left(

4643: \frac{a'}{a}\frac{b'}{b}-\frac{a''}{a} \right) \right]\,, \nonumber \\

4644: \tilde H_{ty} & = &

4645: 6\Phi\left(\frac{\dot{a}}{a}\frac{n'}{n}+\frac{a'}{a}\frac{\dot{b}}{b}

4646: -\frac{\dot{a}'}{a}\right)\,, \nonumber \\

4647: \tilde H_{ij} & = & 2a^2h_{ij}\left\{ \Phi\left[\frac{1}{n^2}\left(\frac{\dot{n}}{n}

4648: \frac{\dot{b}}{b} - \frac{\ddot{b}}{b}\right) - \frac{1}{b^2}\left(

4649: \frac{n'}{n}\frac{b'}{b}-\frac{n''}{n}\right)\right] \nonumber \right. \\

4650: & + & \frac{2}{a^2bn}\left[ \frac{\dot{a}^2\dot{b}\dot{n}}{n^4}

4651: +\frac{a'^2b'n'}{b^4} + \frac{\dot{a}a'}{b^2n^2}\left(

4652: b'\dot{n}-\dot{b}n'\right) \right] \nonumber \\

4653: & - &  2\left[ \frac{1}{n^2}\frac{\ddot{a}}{a}\left(

4654: \frac{1}{n^2}\frac{\dot{a}}{a}

4655: \frac{\dot{b}}{b} + \frac{1}{b^2}\frac{a'}{a}\frac{b'}{b}\right)-\frac{1}{b^2}

4656: \frac{a''}{a}\left(\frac{1}{n^2}\frac{\dot{a}}{a}\frac{\dot{n}}{n}

4657: +\frac{1}{b^2}\frac{a'}{a}\frac{n'}{n}\right)\right] \nonumber \\

4658: & + & \left. \frac{2}{b^2n^2}\left[\frac{\ddot{a}}{a}\frac{a''}{a}-

4659: \frac{\dot{a}^2}{a^2}\frac{n'^2}{n^2}-\frac{a'^2}{a^2}\frac{\dot{b}^2}{b^2}

4660: -\frac{\dot{a}'}{a}\left(\frac{\dot{a}'}{a}-2\frac{\dot{a}}{a}\frac{n'}{n}

4661: -2\frac{a'}{a}\frac{\dot{b}}{b}\right) \right] \right\}\,, \nonumber \\

4662: \tilde H_{yy} & = & 6\Phi\left[\frac{a'}{a}\frac{n'}{n}+\frac{b^2}{n^2}\left(

4663: \frac{\dot{a}}{a}\frac{\dot{n}}{n}-\frac{\ddot{a}}{a}\right)\right]\,.

4664: \end{eqnarray}

4665: 

4666: In this thesis we consider the situation in which a thick brane is embedded

4667: in the five-dimensional spacetime described by (\ref{themetric}), whose boundaries

4668: are located at $y=\mbox{const.}$ hypersurfaces.  Consider the usual junction

4669: conditions at a hypersurface $\Sigma_{y_c}\equiv\{y=y_c\}$,

4670: that is, the continuity of the induced metric, $g_{AB}=\tilde g_{AB}-n_An_B$ and the

4671: extrinsic curvature, $K_{AB} = -g^{C}_{(A}g^D_{B)}\nabla^{}_C n^{}_D$,

4672: of $\Sigma_{y_c}$:

4673: \begin{eqnarray}

4674: n(t,y^+_c) = n(t,y^-_c) \,, ~~~~ a(t,y^+_c) = a(t,y^-_c) \,, \label{cindm}

4675: \end{eqnarray}

4676: \begin{eqnarray}

4677: \frac{n'(t,y^+_c)}{b(t,y^+_c)} = \frac{n'(t,y^-_c)}{b(t,y^-_c)} \,, ~~~~

4678: \frac{a'(t,y^+_c)}{b(t,y^+_c)} = \frac{a'(t,y^-_c)}{b(t,y^-_c)} \,. \label{cincm}

4679: \end{eqnarray}

4680: 

4681: 

4682: We assume a matter content described by an energy-momentum tensor

4683: of the form

4684: \begin{eqnarray}

4685: \tilde \kappa^2\, \tilde T_{AB} = \rho u_{A}u_{B} + p_Lh_{AB} + p_Tn_{A}n_{B}\,,

4686: \end{eqnarray}

4687: where

4688: \begin{eqnarray}

4689: u_A=(-n(t,y),\mb{0},0)\,,~~~~

4690: h_{AB} = \tilde g_{AB} + u_Au_B - n_An_B \,,~~~~

4691: n_A=(0,\mb{0},b(t,y))\,,

4692: \end{eqnarray}

4693: where $\rho\,,$ $p_L\,,$ and $p_T$ denote, respectively, the energy density and

4694: the longitudinal and transverse pressures with respect to the

4695: observers $u^A$.  They are functions of $t$ and $y$.  The energy-momentum

4696: conservation equations, $\nabla_A \tilde T^{AB}=0$, reduce to:

4697: \begin{eqnarray}

4698: \dot{\rho} = -\frac{\dot{b}}{b}(\rho+p_T) -3\frac{\dot{a}}{a}(\rho+p_L)\,,

4699: \label{5ceq}

4700: \end{eqnarray}

4701: \begin{eqnarray}

4702: p_T' = -3\frac{a'}{a}(p_T-p_L)-\frac{n'}{n}(\rho+p_T) \,.

4703: \end{eqnarray}

4704: 

4705: The $\{ty\}$-component of the field equations for the metric

4706: (\ref{themetric}) in Lanczos gravity [eq.~(\ref{fieldeq})] has the

4707: form

4708: \begin{eqnarray}

4709: \left(1+4\alpha\Phi\right)\left(\frac{\dot{a}}{a}\frac{n'}{n}+

4710: \frac{a'}{a}\frac{\dot{b}}{b}-\frac{\dot{a}'}{a}\right) = 0\,. \label{tycom}

4711: \end{eqnarray}

4712: If we discard the possibility $1+4\alpha\Phi=0$ by restricting ourselves

4713: to models with a well-defined limit in Einstein gravity ($\alpha\rightarrow

4714: 0$), we see that the metric functions must satisfy

4715: \begin{eqnarray}

4716: \dot{a}' = \frac{n'}{n}\dot{a} + \frac{\dot{b}}{b}a' \,. \label{keyrel}

4717: \end{eqnarray}

4718: Using this consequence of the $\{ty\}$-component, we can rewrite the

4719: remains components of $\tilde G_{AB}$ and $\tilde H_{AB}$ as

4720: \begin{eqnarray}

4721: \tilde G_{tt} = \frac{3n^2}{2a^3a'}\left(a^4\Phi\right)'\,,~~~~

4722: \tilde G_{yy} = - \frac{3b^2}{2a^3\dot{a}}\left(a^4\Phi\right)^{\displaystyle{\cdot}}   \,,

4723: \end{eqnarray}

4724: \begin{eqnarray}

4725: \tilde G_{ij} = \frac{1}{2\dot{a}a'}h_{ij}\left\{\frac{\dot{b}}{b}\frac{a'}{\dot{a}}

4726: \left(a^4\Phi\right)^{\displaystyle{\cdot}} + \frac{n'}{n}\frac{\dot{a}}{a'}

4727: \left(a^4\Phi\right)' - \left(a^4\Phi\right)^{\displaystyle{\cdot}}{}'  \right\}   \,,

4728: \end{eqnarray}

4729: \begin{eqnarray}

4730: \tilde H_{tt} = \frac{3n^2}{2a^3a'}\left(a^4\Phi^2\right)'\,,~~~~

4731: \tilde H_{yy} = - \frac{3b^2}{2a^3\dot{a}}\left(a^4\Phi^2\right)^{\displaystyle{\cdot}}   \,,

4732: \end{eqnarray}

4733: \begin{eqnarray}

4734: \tilde H_{ij} = \frac{1}{2\dot{a}a'}h_{ij}\left\{\frac{\dot{b}}{b}\frac{a'}{\dot{a}}

4735: \left(a^4\Phi^2\right)^{\displaystyle{\cdot}} + \frac{n'}{n}\frac{\dot{a}}{a'}

4736: \left(a^4\Phi^2\right)' - \left(a^4\Phi^2\right)^{\displaystyle{\cdot}}{}'  \right\}   \,.

4737: \end{eqnarray}

4738: 

4739: Then the field equations~(\ref{fieldeq}) for the metric~(\ref{themetric})

4740: are equivalent to eq. (\ref{keyrel}) and

4741: \begin{eqnarray}

4742: \left[a^4\left(\Phi+2\alpha\Phi^2+\frac{1}{l^2}\right)\right]' =

4743: \frac{1}{6}(a^4)'\rho  \,, \label{ttcom}

4744: \end{eqnarray}

4745: \begin{eqnarray}

4746: \frac{\dot{b}}{b}\frac{a'}{\dot{a}}

4747: \left[a^4\left(\Phi+2\alpha\Phi^2+\frac{1}{l^2}\right)\right]^{\displaystyle{\cdot}}

4748: + \frac{n'}{n}\frac{\dot{a}}{a'}

4749: \left[a^4\left(\Phi+2\alpha\Phi^2+\frac{1}{l^2}\right)\right]'-\cr

4750: - {\left[a^4\left(\Phi+2\alpha\Phi^2+\frac{1}{l^2}\right)\right]^{\displaystyle{\cdot}}{}}' =

4751: 2\dot{a}a'a^2p_L\,, \label{ijcom}

4752: \end{eqnarray}

4753: \begin{eqnarray}

4754: \left[a^4\left(\Phi+2\alpha\Phi^2+\frac{1}{l^2}\right)\right]^{\displaystyle{\cdot}} =

4755: -\frac{1}{6}(a^4)^{\displaystyle{\cdot}} p_T \,. \label{yycom}

4756: \end{eqnarray}

4757: Introducing (\ref{ttcom}) and (\ref{yycom}) into (\ref{ijcom}), we obtain

4758: the conservation equation~(\ref{5ceq}).

4759: 

4760: \section{Bulk geometry in static coordinates} \label{B2}

4761: 

4762: In this section we show how to derive the modified Friedmann equation for any gravitational theory\footnote{See also \cite{Barcelo-edge} for

4763: the closed Friedmann  brane case ($k=+1$).}.

4764: A Friedmann brane at $y=0$ in a bulk metric

4765: \be \label{matric1}

4766: ds^2=-n^2(\tau,y)d\tau^2+a^2(\tau,y)d\vec x\cdot d\vec x+ b(\tau,y) dy^2

4767: \ee

4768: is locally equivalent to a Friedmann brane moving geodesically in a black-hole-type metric

4769: \be\label{metricBH}

4770: ds^2=-f(R)dT^2+R^2d\vec x\cdot d\vec x + \frac{dR^2}{f(R)}\ .

4771: \ee

4772: The form of $f(R)$ is determined by solving the field equations of the theory.

4773: 

4774: In this picture we imagine the brane as a hypersurface that is moving towards or away

4775: from a black-hole with an expansion factor $a(\tau,y(\tau))=R(T(\tau))$. At

4776: each fixed radial distance from the black-hole, $dR=0=da$, one has

4777: \be

4778: da=\dot ad\tau+a'dy=0\ \Rightarrow\ dy^2=\left(\frac{\dot a}{a'}\right)^2 d\tau^2\ .

4779: \ee

4780: Substituting into (\ref{matric1}), we obtain

4781: \be

4782: ds^2=-\left(n^2-\frac{\dot a^2}{a'^2}b^2\right)d\tau^2\ .

4783: \ee

4784: We suppose that the brane is a hypersurface in the black-hole background in geodesic

4785: motion. Therefore its four-velocity in static coordinates is

4786: \be

4787: u_Adx^A=-\sqrt{f(R(T))+\dot R(T)^2}dT+\frac{\dot R(T)}{f(R(T))}dR=-d\tau\ ,

4788: \ee

4789: where $\tau$ is the proper time on the brane. For $dR=0$ we have

4790: \be\label{tau}

4791: -\sqrt{f(R)+\dot R^2}dT=d\tau\ .

4792: \ee

4793: Substituting into (\ref{metricBH}), we have

4794: \be\label{BH}

4795: ds^2=-\frac{f(R(T))}{f(R(T))+\dot R(T)^2}d\tau^2\ .

4796: \ee

4797: Equating now eq. (\ref{tau}) and (\ref{BH}), using $R(T(\tau))=a(\tau,y(\tau))$, we obtain

4798: \be

4799: n^2-\left(\frac{\dot a b}{a'}\right)^2=\frac{f(a)}{f(a)+\dot a^2}\ .

4800: \ee

4801: Since $\tau$ is the proper time, $n(\tau,0)=1$, and

4802: \be

4803: f(a)=\left(\frac{a'}{b}\right)^2-\dot a^2\ .

4804: \ee

4805: Therefore defining the Hubble rate $H=\dot a/a$ and the function $H_y=a'/ba$, we obtain

4806: \be

4807: \frac{f(a)}{a^2}=H_y^2-H^2\ ,

4808: \ee

4809: where $H_y$ must be found by junction conditions that are dependent on the gravity theory

4810: chosen. In particular, for the Randall-Sundrum-type model, we have

4811: \be

4812: f(a)=k-\frac{M}{a^2}\ ,

4813: \ee

4814: and we find \cite{Kr:99,Id:00}

4815: \be

4816: H^2+\frac{k}{a^2}=H_y^2+\frac{M}{a^4}\ ,

4817: \ee

4818: where $M/a^4$ is the dark radiation term. Instead for Lanczos gravity, we have

4819: \be

4820: f(a) = k +

4821: \frac{a^2}{2\alpha}\left(1-\sqrt{1+\frac{4\alpha M}

4822: {3a^4}+\frac{2}{3}\alpha\Lambda}\right)\ ,

4823: \ee

4824: leading to \cite{GS1}

4825: \be

4826: H^2+\frac{k}{a^2}=H_y^2+\frac{1}{2\alpha}\left(1-\sqrt{1+\frac{4\alpha M}

4827: {3a^4}+\frac{2}{3}\alpha\Lambda}\right)\ .

4828: \ee

4829: Local equivalence of the metrics (\ref{matric1}) and (\ref{metricBH}) has been proved in the

4830: Randall-Sundrum model \cite{BH} and in the Lanczos model \cite{Charmousis}.

4831: 

4832: \newpage

4833: \begin{thebibliography}{99}

4834: 

4835: 

4836: 

4837: \bibitem{AbMo:01} B. Abdesselam and N. Mohammedi, Phys. Rev. {\bf D65}, 084018  (2002).

4838: 

4839: \bibitem{bhads} S.~O. Alexeyev, Gravitation \& Cosmology {\bf 3},  161  (1997);

4840:                 A. Chakrabarti and D.~H. Tchrakian, Phys. Rev. {\bf D65},

4841:                 024029  (2002).

4842: 

4843: \bibitem{KK}

4844: T. Appelquist, A. Chodos and P. G. O. Freund, {\it Modern Kaluza-Klein theories}

4845: (Addison-Wesley, Reading, USA, 1987).

4846: 

4847: 

4848: \bibitem{ADM}

4849: R. Arnowitt, S. Deser and C. W. Misner, in {\it Gravitation: An Introduction to Current

4850: Research}, edited by L. Witten (Wiley, New York,

4851: 1962).

4852: 

4853: \bibitem{BK}

4854: V. Balasubramanian and P. Kraus, Commun. Math. Phys. {\bf 208}, 413 (1999).

4855: 

4856: \bibitem{barcelo}

4857: C. Barcel\'o, C. Germani and C. F. Sopuerta, Phys. Rev. {\bf D68}, 104007 (2003).

4858: 

4859: \bibitem{Barcelo-edge}

4860: C.~Barcel\'o and M.~Visser,

4861: %``Living on the edge: Cosmology on the boundary of anti-de Sitter space,''

4862: Phys.\ Lett.\ B {\bf 482}, 183 (2000).

4863: 

4864: \bibitem{bdl}

4865: P. Binetruy, C. Deffayet, U. Ellwanger and D. Langlois, Phys. Lett. {\bf B477}, 285 (2000).

4866: 

4867: \bibitem{bida}

4868: N. D. Birrell and P. C. W. Davies, {\it Quantum fields in curved

4869: space} (Cambridge University Press, Cambridge, 1994).

4870: 

4871: \bibitem{BH}

4872: P. Bowcock, C. Charmousis and R. Gregory, Class. Quant. Grav. {\bf 17}, 4745 (2000);

4873: S. Mukohyama, T. Shiromizu and K. Maeda,

4874: Phys. Rev. {\bf D62}, 024028 (2000); Erratum: ibid., {\bf D63}, 029901 (2001).

4875: 

4876: \bibitem{Bousso}

4877: R. Bousso, JHEP {\bf 06}, 028 (1999).

4878: 

4879: \bibitem{bcg}

4880: P.\ Bowcock, C.\ Charmousis, and R.\ Gregory, Class.\ Quantum

4881: Grav.\ {\bf 17}, 4745 (2000).

4882: 

4883: \bibitem{holo1}

4884: J. D. Bekenstein, Nuovo Cim. Lett. {\bf 4}, 737 (1972).

4885: 

4886: \bibitem{holo2}

4887: J. D. Bekenstein, Phys. Rev. {\bf D7}, 2333 (1973).

4888: 

4889: \bibitem{holo3}

4890: J. D. Bekenstein, Phys. Rev. {\bf D9}, 3292 (1974).

4891: 

4892: \bibitem{Bruck}

4893: P.~Brax and C.~van de Bruck,

4894: %``Cosmology and brane worlds: A review,''

4895: Class.\ Quant.\ Grav.\  {\bf 20}, R201 (2003).

4896: 

4897: \bibitem{Brohe}

4898: J.D. Brown and M. Henneaux, Commum. Math. Phys. {\bf 104}, 207 (1986).

4899: 

4900: \bibitem{BY}

4901: J. D. Brown and J. W. York, Jr., Phys. Rev. {\bf D47}, 1407 (1993).

4902: 

4903: \bibitem{collapse}

4904: M. Bruni, C. Germani and R. Maartens, Phys. Rev. Lett. {\bf 87}, 231302 (2001).

4905: 

4906: \bibitem{C1}

4907: C. G. Callan, D. Friedan, E. J. Martinec and M. J. Perry, Nucl. Phys. {\bf B262}, 593 (1985).

4908: 

4909: \bibitem{CamSop}

4910: A.~Campos and C.~F.~Sopuerta,

4911: %``Evolution of cosmological models in the brane-world scenario,''

4912: Phys.\ Rev.\ {\bf D63}, 104012 (2001).

4913: 

4914: \bibitem{245Duff}

4915: D. M. Capper and M. J. Duff, Nucl. Phys. {\bf B82}, 147 (1974); D. M. Capper, Nuovo Cim.

4916: {\bf A25}, 29 (1975).

4917: 

4918: \bibitem{CLR}

4919: D. M. Capper, G. Leibbrand and M. Ramon Medrano, Phys. Rev. {\bf D8}, 4320 (1973).

4920: 

4921: \bibitem{SWi}

4922: D. M. Capper and M. Ramon Medrano, Phys. Rev. {\bf D9}, 1641 (1974).

4923: 

4924: \bibitem{Cha:89} A.~H.~Chamseddine, Phys. Lett. {\bf B233}, 291 (1989);

4925:                  F.~M\"uller-Hoissen, Nucl. Phys. {\bf B346}, 235 (1990).

4926: 

4927: \bibitem{blackstring}

4928: A. Chamblin, S. W. Hawking and H. S. Reall, Phys. Rev. {\bf D61}, 065007 (2000).

4929: 

4930: \bibitem{Charmousis}

4931: C.~Charmousis and J.~F.~Dufaux,

4932: %``General Gauss-Bonnet brane cosmology,''

4933: Class.\ Quant.\ Grav.\  {\bf 19}, 4671 (2002).

4934: 

4935: \bibitem{ChNeWe:01} Y.~M. Cho and I.~P. Neupane, arXiv: hep-th/0112227 (2001);

4936:             Y.~M. Cho, I.~P. Neupane, and P.~S. Wesson, Nucl. Phys. {\bf B621},  388 (2002).

4937: 

4938: \bibitem{chrifu}

4939: S. M. Christensen and S. A. Fulling, Phys. Rev. {\bf D15}, 2088

4940: (1977).

4941: 

4942: \bibitem{chridu}

4943: S. M. Christensen and M. J. Duff, Nucl. Phys. {\bf B154}, 301

4944: (1977); see also M. J. Duff, Class. Quant. Grav. {\bf 11}, 1387 (1994) and references therein.

4945: 

4946: \bibitem{Corradini}

4947: O.~Corradini and Z.~Kakushadze,

4948: %``Localized gravity and higher curvature terms,''

4949: Phys.\ Lett.\ {\bf B494}, 302 (2000).

4950: 

4951: %\bibitem{distribution}

4952: %T. M. Cover and J. A. Thomas.

4953: %{\it Elements of Information Theory},

4954: %(John Wiley and Sons, 1991).

4955: 

4956: \bibitem{Friedmann} C.~Cs$\acute{\mbox{a}}$ki, M.~Graesser, C.~Kolda

4957: and J.~Terning,

4958: Phys.\ Lett.\ {\bf B462}, 34 (1999);

4959: J.~M.~Cline, C.~Grojean and G.~Servant,

4960: %``Cosmological expansion in the presence of an extra dimension,''

4961: Phys.\ Rev.\ Lett.\/ {\bf 83}, 4245 (1999).

4962: %[arXiv:hep-th/];

4963: 

4964: \bibitem{CGKT}

4965: C. Csaki, M. Graesser, C. Kolda and J. Terning, Phys. Lett. {\bf B462}, 34 (1999).

4966: 

4967: \bibitem{C2}

4968: T. Curtright and C. Zachos, Phys. Rev. Lett. {\bf 53}, 1799 (1984).

4969: 

4970: \bibitem{dmpr}

4971: N. K. Dadhich, R. Maartens, P. Papadopoulos and V. Rezania,

4972: Phys. Lett. {\bf B487}, 1 (2000).

4973: 

4974: \bibitem{Davis}

4975: S.~C.~Davis,

4976: %``Generalised Israel junction conditions for a Gauss-Bonnet brane world,''

4977: Phys.\ Rev.\ {\bf D67}, 024030 (2003).

4978: 

4979: \bibitem{DD}

4980: N. Deruelle and T. Dolezel, Phys.\ Rev.\ {\bf D64}, 103506 (2001).

4981: 

4982: \bibitem{DeruDole}

4983: N. Deruelle and T. Dolezel, Phys. Rev. {\bf D62}, 103502 (2000).

4984: 

4985: \bibitem{5}

4986: N. Deruelle, T. Dolezel and J. Katz, Phys. Rev. {\bf D63}, 083513 (2001).

4987: 

4988: \bibitem{lovelock} N. Deruelle and L. Farina-Busto, Phys. Rev. {\bf D41},  3696  (1990);

4989:                    G.~A.~M. Marugan, Phys. Rev. {\bf D46},  4320  (1992);

4990:                    4340  (1992).

4991: 

4992: \bibitem{DG}

4993: N. Deruelle and C. Germani, arXiv: gr-qc/0306116.

4994: 

4995: \bibitem{Deruelle-Madore}

4996: N.~Deruelle and J.~Madore,

4997: %``On the quasi-linearity of the Einstein- Gauss-Bonnet gravity field

4998: %equations,''

4999: arXiv: gr-qc/0305004.

5000: 

5001: \bibitem{BdW}

5002: B. DeWitt, Phys. Rev. {\bf 160}, 113 (1967).

5003: 

5004: \bibitem{DeWitt}

5005: B. S. DeWitt, Phys. Rep. {\bf 19C}, 295 (1975).

5006: 

5007: \bibitem{DGP}

5008: G. Dvali, G. Gabadadze and M. Porrati, Phys. Lett. {\bf B485}, 208 (2000).

5009: 

5010: \bibitem{MT1}

5011: M. J. Duff, Int. J. Mod. Phys. {\bf A11}, 5623 (1996).

5012: 

5013: \bibitem{DL}

5014: M. J. Duff and J. T. Liu, Phys. Rev. Lett. {\bf 85}, 2052 (2000); Class. Quant. Grav. {\bf 18},

5015: 3207 (2001); see also

5016: M. J. Duff, Phys. Rev. {\bf D9}, 1837 (1974).

5017: 

5018: \bibitem{duff}

5019: M. J. Duff, Nucl. Phys. {\bf B125}, 334 (1977).

5020: 

5021: \bibitem{E1}

5022: A. Einstein, Preussiche Akademie der Wissenschaften (Berlin)

5023: Sitzungsberichte, pg. 688 (1916).

5024: 

5025: \bibitem{Emparan}

5026: R. Emparan, A. Fabbri and N. Kaloper, JHEP {\bf 0208}, 043 (2002).

5027: 

5028: \bibitem{F1}

5029: E. S. Fradkin and A. A. Tseytlin,

5030: %``Effective Field Theory From Quantized Strings,''

5031: Phys.\ Lett.\ {\bf B158}, 316 (1985).

5032: 

5033: \bibitem{F2}

5034: D. Friedan, Phys. Rev. Lett. {\bf 45}, 1057 (1980); L. Alvarez-Gaum\'e and D. Z. Freedman,

5035: Phys. Lett. {\bf 94B}, 171 (1980);

5036: L. Alvarez-Gaum\'e, D. Z. Freedman and S. Mukhi, Ann. Phys. {\bf 134}, 85 (1981).

5037: 

5038: \bibitem{V2}

5039: S. Fubini and G. Veneziano, Ann. Phys. {\bf 63}, 12 (1971).

5040: 

5041: \bibitem{bunchNew}

5042: J. Garriga and T. Tanaka, Phys. Rev. Lett. {\bf 84}, 2778 (2000);

5043:  C. Csaki, J. Erlich, T. J. Hollowood and Y. Shirman, Nucl. Phys. {\bf B581}, 309 (2000);

5044:  S. B. Giddings, E. Katz and L. Randall, JHEP

5045: {\bf 3}, 023 (2000); C. Csaki, J. Erlich and T. J. Hollowood, Phys. Rev. Lett. {\bf 84},

5046: 5932 (2000); C. Csaki, J. Erlich and

5047: T. J. Hollowood,

5048: Phys. Lett. {\bf B481}, 107 (2000); I. Ya. Aref'eva, M. G. Ivanov,

5049: W. Muck, K. S. Viswanathan and I. V. Volovich, Nucl. Phys. {\bf B590}

5050: 273 (2000);  Z. Kakushadze, Phys. Lett. {\bf B497}, 125 (2000).

5051: 

5052: \bibitem{static}

5053: C. Germani and R. Maartens, Phys. Rev. {\bf D64}, 124010 (2001).

5054: 

5055: \bibitem{GS1}

5056: C. Germani and C. F. Sopuerta, Phys. Rev. Lett. {\bf 88}, 231101 (2002).

5057: 

5058: \bibitem{GS2}

5059: C. Germani and C. F. Sopuerta, Astrophys. Space Sci. {\bf 283}, 487 (2003).

5060: 

5061: \bibitem{Germani}

5062: C.~Germani and C.~F.~Sopuerta,

5063: %``String inspired braneworld cosmology,''

5064: Phys.\ Rev.\ Lett.\  {\bf 88}, 231101 (2002); ibid.,

5065: %[arXiv:hep-th/0202060];

5066: %%CITATION = HEP-TH 0202060;%%

5067: %----------------------------------------------------------------

5068: %\cite{Germani:2002vs}

5069: %\bibitem{Germani}

5070: %C.~Germani and C.~F.~Sopuerta,

5071: %``Varying fundamental constants from a string-inspired brane world model,''

5072: Astrophys.\ Space Sci.\  {\bf 283}, 487 (2003);

5073: %[arXiv:hep-th/0210086].

5074: %%CITATION = HEP-TH 0210086;%%

5075: %----------------------------------------------------------------

5076: %----------------------------------------------------------------

5077: %\cite{Kim:2000pz}

5078: %\bibitem{Kim:2000pz}

5079: J.~E.~Kim, B.~Kyae and H.~M.~Lee,

5080: %``Various modified solutions of the Randall-Sundrum model with the  Gauss-Bonnet interaction,''

5081: Nucl.\ Phys.\ {\bf B582}, 296  (2000),

5082: Erratum: ibid.,\ {\bf B591}, 587 (2000);

5083: %[arXiv:hep-th/0004005].

5084: %%CITATION = HEP-TH 0004005;%%

5085: %----------------------------------------------------------------

5086: %\cite{Abdesselam:2001ff}

5087: %\bibitem{Abdesselam:2001ff}

5088: B.~Abdesselam and N.~Mohammedi,

5089: %``Brane world cosmology with Gauss-Bonnet interaction,''

5090: Phys.\ Rev.\ {\bf D65}, 084018 (2002).

5091: %[arXiv:hep-th/0110143].

5092: %%CITATION = HEP-TH 0110143;%%

5093: 

5094: \bibitem{GW}

5095: G. W. Gibbons and S. W. Hawking, Phys. Rev. {\bf D15}, 2752 (1977).

5096: 

5097: \bibitem{Giovannini}

5098: M.~Giovannini,

5099: %``Thick branes and Gauss-Bonnet self-interactions,''

5100: Phys.\ Rev.\ {\bf D64}, 124004 (2001).

5101: 

5102: \bibitem{Ginsparg2}

5103: P. Ginsparg and G. Moore, arXiv: hep-th/9304011.

5104: 

5105: \bibitem{ginsparg}

5106: P. Ginsparg, arXiv: hep-th/9108028.

5107: 

5108: \bibitem{Gravanis}

5109: E.~Gravanis and S.~Willison,

5110: %``Israel conditions for the Gauss-Bonnet theory and the Friedmann  equation on the brane

5111: %universe,''

5112: Phys.\ Lett.\ {\bf B562}, 118 (2003).

5113: 

5114: \bibitem{GSW}

5115: M. B. Green, J. H. Schwartz and E. Witten, {\it Superstring theory}, volumes I \& II,

5116: (Cambridge University Press, Cambridge, 1987).

5117: 

5118: \bibitem{Gregory-Padilla}

5119: J.~P.~Gregory and A.~Padilla,

5120: %``Braneworld holography in Gauss-Bonnet gravity,''

5121: arXiv: hep-th/0304250.

5122: 

5123: \bibitem{GRS}

5124: R. Gregory, V. A. Rubakov and S. M. Sibiryakov, Phys. Rev. Lett. {\bf 84}, 5928 (2000).

5125: 

5126: \bibitem{G1}

5127: D. Gross, J. Harvey, E. Martinec and R. Rohn, Phys. Rev. Lett. {\bf 54}, 502 (1985).

5128: 

5129: \bibitem{GBE}

5130: D. J. Gross and J. H. Sloan, Nucl. Phys. {\bf B291}, 41 (1987).

5131: 

5132: \bibitem{H2}

5133: S. W. Hawking, Nature {\bf 248}, 30 (1974); Comm. Math. Phys. {\bf 43}, 199 (1975).

5134: 

5135: \bibitem{predict}

5136: S. W. Hawking, Phys. Rev. {\bf D14}, 2460 (1976).

5137: 

5138: \bibitem {H1}

5139: W. Heisenberg, Z. Physics {\bf 110}, 251 (1938).

5140: 

5141: \bibitem{tH1}

5142: G. t'Hooft, Nucl. Phys. {\bf B62}, 444 (1973);

5143: G. t'Hooft and M. Veltman, Ann. Inst. H. Poincar\'e, {\bf 20}, 69 (1974);

5144: S. Deser and P. V. Nieuwenhuizen, Phys. Rev. {\bf D10}, 401 (1974); ibid.,

5145: Phys. Rev. {\bf D10}, 411 (1974).

5146: 

5147: \bibitem{tH2}

5148: G. t'Hooft, arXiv: gr-qc/9310026.

5149: 

5150: \bibitem{HW}

5151: P. Horava and E. Witten, Nucl. Phys. {\bf B460}, 506 (1996).

5152: 

5153: \bibitem{positive}

5154: G. T. Horowitz and R. C. Myers, Phys. Rev. {\bf D59}, 026005 (1999).

5155: 

5156: \bibitem{Id:00} D. Ida, JHEP {\bf 09},  14  (2000).

5157: 

5158: \bibitem{israel}

5159: W. Israel, Nuovo Cim. {\bf 44B}, 1 (1966).

5160: 

5161: \bibitem{IZ}

5162: C. Itzykson and J.-B. Zuber, {\it Quantum Field Theory} (McGraw-Hill, Singapore, 1980).

5163: 

5164: \bibitem{others}

5165:                  J.~E.~Kim, B.~Kyae and H.~M.~Lee, Phys.\ Rev.\ {\bf D62}, 045013 (2000);

5166:                  N.~E. Mavromatos and J. Rizos, Phys. Rev.  {\bf D62},  124004  (2000);

5167:                  I. Low and A. Zee, Nucl. Phys. {\bf B585},  395  (2000);

5168:                  S. Nojiri and S.~D. Odintsov, Phys. Lett. {\bf B494},  135  (2000);

5169:                  S. Nojiri, S.~D. Odintsov, and S. Ogushi, Phys. Rev.  {\bf D65},  023521

5170:                  (2002);

5171:                  M. Giovannini, Phys. Rev.  {\bf D64},  124004  (2001);

5172:                  I.~P. Neupane, Class. Quant. Grav. {\bf 19}, 1167 (2002).

5173: 

5174: \bibitem{exact}

5175: D. Kramer, H. Stephani, M. MacCallum and E. Herlt, {\em Exact

5176: Solutions of Einstein's Field Equations} (Cambridge University Press, Cambridge, 1980).

5177: 

5178: \bibitem{Kr:99} P. Kraus, JHEP {\bf 12},  011  (1999).

5179: 

5180: \bibitem{kre}

5181: E. Kreysig, {\it Introduction to differential geometry and

5182: Riemanian geometry} (University of Toronto Press, Toronto, 1968).

5183: 

5184: \bibitem{Lanczos}

5185: C. Lanczos,

5186: Z. Phys. {\bf 73} 147  (1932); ibid., Ann. Math. {\bf 39}, 842 (1938).

5187: 

5188: \bibitem{Lidsey}

5189: J.~E.~Lidsey,

5190: %``Inflation and Braneworlds,''

5191: arXiv: astro-ph/0305528.

5192: %%CITATION = ASTRO-PH 0305528;%%

5193: %----------------------------------------------------------------

5194: %\cite{Lidsey:2003sj}

5195: \bibitem{Lidsey2}

5196: J.~E.~Lidsey and N.~J.~Nunes,

5197: %``Inflation in Gauss-Bonnet brane cosmology,''

5198: Phys.\ Rev.\ {\bf D67}, 103510 (2003).

5199: %[arXiv:astro-ph/0303168].

5200: 

5201: \bibitem{Lov}

5202: D. Lovelock, J. Math. Phys. {\bf 12}, 498 (1971).

5203: 

5204: \bibitem{Low:2000pq}

5205: 

5206: I.~Low and A.~Zee,

5207: %``Naked singularity and Gauss-Bonnet term in brane world scenarios,''

5208: Nucl.\ Phys.\ {\bf B585}, 395  (2000);

5209: N.~E.~Mavromatos and J.~Rizos,

5210: %``String inspired higher-curvature terms and the Randall-Sundrum  scenario,''

5211: Phys.\ Rev.\ {\bf D62}, 124004  (2000);

5212: N.~E.~Mavromatos and J.~Rizos,

5213: %``Exact solutions and the cosmological constant problem in dilatonic  domain wall

5214: %higher-curvature string gravity,''

5215: Int.\ J.\ Mod.\ Phys.\ {\bf A18}, 57 (2003);

5216: I.~P.~Neupane,

5217: %``Consistency of higher derivative gravity in the brane background,''

5218: JHEP {\bf 0009}, 040 (2000);

5219: P.~Binetruy, C.~Charmousis, S.~C.~Davis and J.~F.~Dufaux,

5220: %``Avoidance of naked singularities in dilatonic brane world scenarios  with a

5221: %Gauss-Bonnet term,''

5222: Phys.\ Lett.\ {\bf B544}, 183 (2002);

5223: C.~Charmousis and J.~F.~Dufaux,

5224: %``General Gauss-Bonnet brane cosmology,''

5225: Class.\ Quant.\ Grav.\  {\bf 19}, 4671 (2002);

5226:  K.~A. Meissner and M. Olechowski, Phys. Rev. Lett. {\bf 86},  3708  (2001).

5227: 

5228: \bibitem{m}

5229: R. Maartens, Phys. Rev. {\bf D62}, 084023 (2000).

5230: 

5231: \bibitem{mwbh}

5232: R. Maartens, D. Wands, B.A. Bassett and I. P. C. Heard, Phys. Rev.

5233: {\bf D62}, 041301 (2000).

5234: 

5235: \bibitem{maacov}

5236: R. Maartens, Prog. Theor. Phys. Suppl. {\bf 148}, 213 (2003).

5237: 

5238: \bibitem{M1}

5239: J. Maldacena, Adv. Theor. Math. Phys. {\bf 2}, 231 (1998); Int. J. Theor. Phys.

5240: {\bf 38}, 113 (1999).

5241: 

5242: \bibitem{MeOl:00} K.~A. Meissner and M. Olechowski, Phys. Rev. Lett. {\bf 86},  3708  (2001).

5243: 

5244: \bibitem{local} K.~A. Meissner and M. Olechowski, Phys. Rev. {\bf D65}, 064017  (2002);

5245:                 J.~E.~Kim and H.~M.~Lee, Nucl. Phys. {\bf B602}, 346 (2001);

5246:                 ibid., {\bf B619} 763 (2001).

5247: 

5248: \bibitem{M3}

5249: R. R. Metsaev and A. A. Tseytlin, Nucl. Phys. {\bf B293}, 385 (1987).

5250: 

5251: \bibitem{gravitation}

5252: C. W. Misner, K. S. Thorne and J. A. Wheeler, {\it Gravitation} (W. H. Freeman,

5253: New

5254: York, 1998).

5255: 

5256: \bibitem{Mounaix}

5257: P.~Mounaix and D.~Langlois,

5258: %``Cosmological equations for a thick brane,''

5259: Phys.\ Rev.\ {\bf D65}, 103523 (2002).

5260: 

5261: \bibitem{msm}

5262: S.\ Mukohyama, T.\ Shiromizu and K.\ I.\ Maeda, Phys. Rev. {\bf

5263: D62}, 024028 (2000).

5264: 

5265: \bibitem{M2}

5266: C. Myers, Phys. Lett. {\bf B199}, 371 (1987).

5267: 

5268: \bibitem{Myers}

5269: R.~C.~Myers,

5270: %``Higher Derivative Gravity, Surface Terms And String Theory,''

5271: Phys.\ Rev.\ {\bf D36}, 392 (1987).

5272: 

5273: \bibitem{Ne:01a} I.~P. Neupane, Phys. Lett. {\bf B512},  137  (2001).

5274: 

5275: \bibitem{Ca:01} S. Nojiri and S.~D. Odintsov, Phys. Lett. {\bf B521}, 87 (2001); R.-G.

5276: Cai, Phys. Rev. {\bf D65},

5277:                 084014  (2002).

5278: 

5279: \bibitem{pet}

5280: A. Z. Petrov, {\it Einstein spaces} (Pergamon, Oxford, 1969).

5281: 

5282: \bibitem{padilla}

5283: A. Padilla, Ph.D. thesis (Durham University, 2002), arXiv: hep-th/0210217.

5284: 

5285: 

5286: \bibitem{Padilla}

5287: A.~Padilla,

5288: %``Surface terms and the Gauss-Bonnet Hamiltonian,''

5289: arXiv: gr-qc/0303082.

5290: %%CITATION = GR-QC 0303082;%%

5291: 

5292: \bibitem{P1}

5293: A.M. Polyakov, Phys. Lett. {\bf 103B}, 207 (1981); ibid., Phys. Lett. {\bf 103B}, 211 (1981).

5294: 

5295: \bibitem{PO}

5296: J. Polchinski, {\it String Theory} volumes I \& II, (Cambridge University Press,

5297: Cambridge, 1998).

5298: 

5299: \bibitem{tesi}

5300: M.\ M.\ Prats, PhD thesis (University of Barcelona, 1995).

5301: 

5302: \bibitem{RS}

5303: L. Randall and R. Sundrum, Phys. Rev. Lett. {\bf 83}, 3370 (1999).

5304: 

5305: \bibitem{R1}

5306: C. Rovelli, arXiv: gr-qc/0006061.

5307: 

5308: \bibitem{SV}

5309: I. Savonie and E. Verlinde, Phys. Lett. {\bf B507}, 305 (2001).

5310: 

5311: \bibitem{ssm}

5312: M. Sasaki, T. Shiromizu and K. Maeda, Phys. Rev.  {\bf D62},

5313: 024008 (2000).

5314: 

5315: \bibitem{schwinger}

5316: J. Schwinger, Proc. Nat. Acad. Sci. (USA) {\bf 37}, 452 (1951).

5317: 

5318: \bibitem{SMS}

5319: T. Shiromizu, K. Maeda and M. Sasaki, Phys. Rev. {\bf D62}, 024012 (2000).

5320: 

5321: \bibitem{MT3}

5322: J.H. Schwarz, Nucl. Phys. Proc. Suppl. {\bf B55}, 1 (1997).

5323: 

5324: \bibitem{MT4}

5325: A. Sen, arXiv: hep-th/9802051.

5326: 

5327: \bibitem{Synge}

5328: J. L. Synge, {\it Relativity: the general theory} (North Holland, Amsterdam,

5329: 1971).

5330: 

5331: \bibitem{s}

5332: H.\ Stephani, {\it General Relativity} (Cambridge University

5333: Press, Cambridge, 1990).

5334: 

5335: \bibitem{S1}

5336: L. Susskind, J. Math. Phys. {\bf 36}, 6377 (1995).

5337: 

5338: \bibitem{hawholo}

5339: T. Tanaka, Prog. Theor. Phys. Suppl. {\bf 148}, 307 (2003);

5340: R. Emparan, A. Fabbri and N. Kaloper, JHEP {\bf 0208}, 043 (2002);

5341: R. Casadio, arXiv: hep-th/0302171.

5342: 

5343: \bibitem{MT2}

5344: P. K. Townsend, arXiv: hep-th/9612121.

5345: 

5346: \bibitem{U1}

5347: W. G. Unruh, Phys. Rev. {\bf D14}, 870 (1976).

5348: 

5349: \bibitem{V1}

5350: G. Veneziano, Nuovo Cimento {\bf 57A}, 190 (1968).

5351: 

5352: \bibitem{Visser}

5353: M. Visser, Phys. Lett. {\bf B159}, 22 (1985).

5354: 

5355: \bibitem{Wald}

5356: R. M. Wald, {\it General Relativity} (University of Chicago Press, Chicago, 1984).

5357: 

5358: \bibitem{wiseman}

5359: T. Wiseman, Phys. Rev. {\bf D65}, 124007 (2002).

5360: 

5361: \bibitem{W1}

5362: E. Witten, Comm. Math. Phys. {\bf 92}, 455 (1984).

5363: 

5364: \bibitem{W2}

5365: E. Witten, Adv. Theor. Math. Phys. {\bf 2}, 253 (1998).

5366: 

5367: \bibitem{SU}

5368: E. Witten, Nucl. Phys. {\bf B443}, 85 (1995).

5369: 

5370: \bibitem{York}

5371: J. W. York, Found. Phys. {\bf 16}, 249 (1986).

5372: 

5373: \bibitem{BoDe:85} B.~Zwiebach, Phys. Lett.  {\bf B156}, 315 (1985);

5374:                   D.~G. Boulware and S. Deser, Phys. Rev. Lett. {\bf55},  2656  (1985).

5375: 

5376: 

5377: 

5378: %----------------------------------------------------------------

5379: \end{thebibliography}

5380: %----------------------------------------------------------------

5381: 

5382: 

5383: 

5384: 

5385: %----------------------------------------------------------------

5386: \end{document}

5387: %----------------------------------------------------------------

5388: 

5389: 

5390: 

5391: 

5392: \end{document}

5393: