1: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2: %% ws-procs9x6.tex : 27 January 2003
3: %% Text file to use with ws-procs9x6.cls written in Latex2E.
4: %% The content, structure, format and layout of this style file is the
5: %% property of World Scientific Publishing Co. Pte. Ltd.
6: %% Copyright 1995, 2002 by World Scientific Publishing Co.
7: %% All rights are reserved.
8: %%
9: %% Proceedings Trim Size: 9in x 6in
10: %% Text Area: 7.35in (include runningheads) x 4.5in
11: %% Main Text is 10/13pt
12: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
13:
14: %\documentclass[draft]{ws-procs9x6}
15: \documentclass[final]{ws-procs9x6}
16: \def\lb{\left(}
17: \def\rb{\right)}
18: \def\ls{\left[}
19: \def\rs{\right]}
20: \def\lc{\left\{}
21: \def\rc{\right\}}
22: \def\del{\delta}
23: \newcommand{\be}{\begin{equation}}
24: \newcommand{\ee}{\end{equation}}
25: \newcommand{\bea}{\begin{eqnarray}}
26: \newcommand{\eea}{\end{eqnarray}}
27:
28: \begin{document}
29:
30: \title{Dilaton Stabilization in Brane Gas
31: Cosmology\footnote{\uppercase{T}his work is supported by the \uppercase{C}entre
32: for \uppercase{H}igh \uppercase{E}nergy \uppercase{P}hysics at \uppercase{M}cGill, and \uppercase{NSERC}.}}
33:
34: \author{A.~Berndsen\footnote{\uppercase{E}mail: aberndsen@physics.mcgill.ca}~ and~ J.~Cline\footnote{\uppercase{E}mail: jcline@physics.mcgill.ca}}
35: \address{Centre for High Energy Physics\\
36: Ernest Rutherford Physics Building\\
37: McGill University\\
38: 3600 rue University\\
39: Montréal, QC\\
40: Canada H3A 2T8}
41:
42:
43: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
44: % You may repeat \author \address as often as necessary %
45: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
46:
47: \maketitle
48:
49: \abstracts{
50: Brane Gas Cosmology is an M-theory motivated attempt to reconcile
51: aspects of the standard cosmology based on Einstein's theory of general
52: relativity. Dilaton gravity, when incorporating winding $p$-brane
53: states, has verified the Brandenberger--Vafa mechanism ---a
54: string-motivated conjecture which explains why only three of the nine
55: spatial dimensions predicted by string theory grow large. Further
56: investigation of this mechanism has argued for a hierarchy of
57: subspaces, and has shown the internal directions to be stable to initial
58: perturbations. These results, however, are dependent on a rolling
59: dilaton, or varying strength of Newton's gravitational constant
60: $G_N$. In these proceedings we show that it is not possible to
61: stabilize the dilaton and maintain the stability of the internal
62: directions within the standard Brane Gas Cosmology setup.}
63:
64: \section{Introduction}
65: \subsection{Dilaton Gravity}\label{sec:dil}
66: Dilaton Gravity comes from the low-energy effective action of Type II-A
67: string theory, which is the result of M-theory compactified on $S^1$. Ignoring contributions from the bulk anti-symmetric two-form and
68: including a potential $V(\phi)$ for the dilaton, the
69: action of this system is described as
70: \be
71: S=\int d^{10}x\sqrt{-G}e^{-2\phi}\lb
72: R+4G^{MN}\nabla_M\phi\nabla_N\phi-V(\phi)\rb\ ,\label{eq:iia}
73: \ee
74: where $G$ is the determinant of the background metric $G_{MN}$, $\phi$
75: is the dilaton, $R$ is the Ricci-scalar.
76:
77: Assuming a spatially-homogeneous dilaton and an FRW-type
78: metric of the
79: form~(\ref{eq:frw})\footnote{The space is described by two subspaces
80: with the intention of one subspace describing the large directions ($\lambda$), and the
81: other describing the remaining internal dimensions ($\nu$).}, the action reduces to
82: \bea
83: ds^2&=&-dt^2+e^{2\lambda(t)}\sum_{i=1}^3(dx_i)^2+e^{2\nu(t)}\sum_{j=1}^3(dy_j)^2\label{eq:frw}\\
84: S&=&\int d^{10}x\sqrt{-G_{00}}e^{-\varphi}G^{00}\lb
85: -3\dot\lambda^2-6\dot\nu^2+\dot\varphi^2-V(\phi)\rb\ ,
86: \eea
87: where the shifted dilaton $\varphi=2\phi-3\lambda-6\nu$ is introduced
88: for notational convenience.\cite{tv} Variation of the action with respect to
89: $G_{00}$, $\phi,\lambda$, and $\nu$ yields the system of equations
90: \bea
91: -3\dot\lambda^2-6\dot\nu^2+\dot\varphi^2&=&e^\varphi E + V(\phi)\label{eq:bgc1}\\
92: \ddot\varphi-3\dot\lambda^2-6\dot\nu^2&=&\frac{1}{2}e^\varphi
93: E+\frac{1}{4}V^\prime\nonumber\label{eq:bgc2}\\
94: \ddot\lambda-\dot\varphi\dot\lambda&=&\frac{1}{2}e^\varphi P_\lambda-\frac{1}{4}V^\prime\nonumber\label{eq:bgc3}\\
95: \ddot\nu-\dot\varphi\dot\nu&=&\frac{1}{2}e^\varphi
96: P_\nu-\frac{1}{4}V^\prime\label{eq:bgc4}\ .
97: \eea
98: The scale-factor equations of motion in~(\ref{eq:bgc4}) indicate one of the
99: substantial departures from general relativity: negative pressure
100: terms will cause deceleration\footnote{Recall that in general
101: relativity negative pressure terms are normally associated with
102: inflation, not deceleration.}. Indeed, it is precisely this last
103: property that is exploited in Brane Gas Cosmology (BGC) to stabilize the internal
104: dimensions; BGC provides the negative pressure source\cite{bw}.
105: \par
106: In the standard approach to BGC\cite{abe,bek,bw}, energy $E$ and pressure $P_\lambda,\ P_\nu$ contributions come from $1$-branes, which are
107: described by the Dirac-Born-Infeld (DBI) action, while
108: $V(\phi)=0$. The DBI action admits winding
109: and momentum states, whose energy and pressure contributions are
110: (respectively)
111: \bea
112: E_w=&\mu N e^\lambda,\hspace{2cm}P_w=&-\mu Ne^{\lambda}\nonumber\\
113: E_m=&\mu M e^{-\lambda},\hspace{2cm}P_m=&\mu Me^{-\lambda},
114: \eea
115: where $\mu$ is the brane tension, $N$ is the number of windings, and $M$
116: is the number of momentum modes. The winding modes contribute the
117: negative pressure $P_w$ necessary to inhibit expansion of the internal
118: directions.
119:
120: \subsection{The Rolling Dilaton}
121: We have remarked on one feature of dilaton-gravity which differs from
122: that of general relativity; if, however, this theory is to be consistent
123: with our current understanding of gravity, these differences must
124: disappear and we must recover general
125: relativity at some point in the early universe.
126: The transition occurs in the
127: case of a static dilaton: the action of dilaton-gravity, eq.~(\ref{eq:iia}), reduces to the
128: Einstein-Hilbert action, where the vacuum value of the dilaton
129: determines the strength of gravity as $G_N=e^{2\phi}/(16\pi)$. Thus, a
130: rolling dilaton can be interpreted as a changing Newton's constant, which is
131: strongly constrained on experimental grounds.\cite{cfos}.
132: If BGC is
133: to make contact with our understanding of today's universe, the
134: dilaton must stabilize at some point of its evolution --- it is with this
135: motivation that we have introduced a dilaton potential into the
136: original action~(\ref{eq:iia}).
137:
138: \section{Dilaton Stabilization}
139: \subsection{Perturbation Analysis}
140: To understand the effects of a stabilized dilaton we linearly perturb about a
141: static solution by $\phi=\phi_0+\del\phi,\
142: \lambda=\lambda_0+\del\lambda$, and $\nu=\nu_0+\del\nu$. Such an
143: expansion should draw similar conclusions to a perturbation about a
144: slowly expanding universe since both the
145: radion and dilaton must have masses much greater than the eventual
146: Hubble rate. Due to our
147: choice to expand about a static solution, and choosing
148: $\phi_0=\lambda_0=\nu_0=0$, the zeroth-order equations of the
149: system~(\ref{eq:bgc4}) imply the relations
150: \bea
151: E&=&-V(\phi)\nonumber\\
152: E+3P_\lambda+6P_\nu&=&8V^\prime\nonumber\\
153: P_\lambda&=&\frac{1}{2}V^\prime\nonumber\\
154: P_\nu&=&\frac{1}{2}V^\prime\label{cc}
155: \eea
156: at the stationary point.
157:
158: To understand the effect of small fluctuations about the stationary point, we expand the
159: system~(\ref{eq:bgc4}) to first order. Keeping in mind that
160: $E,\ P_\lambda$, and $P_\nu$ depend on $\lambda$ and $\nu$, we find
161: that
162: \be
163: \left(\begin{array}{c} \ddot\phi\\ \ddot\lambda\\ \ddot\nu
164: \end{array} \right) =
165: {\bf S}
166: \left(\begin{array}{c} \phi\\ \lambda\\ \nu\\
167: \end{array} \right)
168: \ee
169: where the stability matrix $\bf S$ is given by
170: \be
171: {\bf S} =
172: \left(\begin{array}{ccc} 2(V'-\frac12V'') & \ -\frac{25}{8} V'
173: + \frac34 {\partial P_\lambda\over\partial\lambda} +
174: \frac32 {\partial P_\nu\over\partial\lambda}& \ -\frac{49}{8} V'
175: + \frac34 {\partial P_\lambda\over\partial\nu} +
176: \frac32 {\partial P_\nu\over\partial\nu}\ \ \\
177: \frac12 V'-\frac14V''& -\frac34 V' + \frac12{\partial
178: P_\lambda\over\partial\lambda}
179: & -\frac32 V' + \frac12{\partial P_\lambda\over\partial\nu} \\
180: \frac12 V'-\frac14V'' & -\frac34 V' + \frac12{\partial
181: P_\nu\over\partial\lambda}
182: & -\frac32 V' + \frac12{\partial P_\nu\over\partial\nu}\\
183: \end{array} \right)
184: \ee
185:
186: This matrix is diagonalized by a similarity transformation, $\bf P^{-1}SP$, so that
187: the general solution takes the form
188: \be
189: \left(\begin{array}{c} \phi\\ \lambda\\ \nu\\
190: \end{array} \right) = {\bf P} \left(\begin{array}{c}
191: A_1 e^{i\omega_1 t}\\ A_2 e^{i\omega_2 t}\\ A_3 e^{i\omega_3 t}\\
192: \end{array} \right) + {(\omega_i\to -\omega_i)}
193: \ee
194: We insist that the eigenvalues $\omega_1$ and $\omega_3$ are real, giving a
195: stable dilaton and radion, while $\omega_2$ must be tuned to vanish (and the
196: solution degenerates to $A_2 t + B$), so that the large dimensions are free to
197: expand when the tuning (\ref{cc}) is relaxed.
198:
199: However, the tuning of $\omega_2=0$ is not enough. The matrix elements
200: $P_{\phi,2}$ and $P_{\nu,2}$ must be tuned to also vanish; otherwise $\phi$ and
201: $\nu$ will mix with the unstable mode and stability of the dilaton and radion
202: will be lost. Certain linear combinations of $\phi$, $\nu$ and $\lambda$ will
203: oscillate, but this is not adequate: we really need $\phi$ and $\nu$ separately
204: to settle to some fixed values. This can be accomplished by adjusting
205: parameters so that $S_{\phi,\lambda}=S_{\nu,\lambda}=0$. Unfortunately, this
206: requires an exact cancellation between the first derivative of the dilaton
207: potential and quantities ($dP_\nu/d\lambda$ and $dP_\lambda/d\lambda$) characterizing the
208: brane gas, entities which have no reason to be related to each other. If the
209: tuning fails by even a small amount, the contamination of the dilaton and
210: radion by the mode which expands like the scale factor of the large dimensions
211: will eventually lead to unacceptably large evolution of these fields.
212:
213: Let us contrast this to the usual BGC scenario where the dilaton is rolling and has no
214: potential. In the radion-stability analysis\cite{bw}, the dangerous off-diagonal term
215: $S_{\nu,\lambda}=\frac12 dP_\nu/d\lambda$ is taken to be zero, and there is no
216: need to cancel it against $V'$. The pressure in the extra dimensions is
217: assumed to be due to winding and momentum modes which exactly cancel each other
218: at $\nu=0$,
219: \be
220: P_\nu = -\mu N (e^\nu - e^{-\nu}).
221: \ee
222: Here $\mu$ is the tension of the brane, and therefore positive,
223: and $N$ is the number of winding modes.
224: However $dP_\nu/d\nu$ is not zero, and has the right sign to stabilize the
225: extra dimensions, since $S_{\nu,\nu} < 0$. If we try to use the same
226: approach but also stabilize the dilaton, then $V=V'=P_\lambda=P_\nu=0$ at the stationary
227: point, while $V''$ can be nonzero. We then have one vanishing eigenvalue,
228: as needed for the growth of the large dimensions, while the other two eigenvalues are given
229: by
230: \be
231: -\omega^2_\pm = \frac12\left( \frac12 P' - V'' \pm
232: \sqrt{\left(\frac12 P' - V''\right)^2 + \frac12 V''P'} \right)\label{eq:bgcev}
233: \ee
234: where $P'=dP_\nu/d\nu$. It can be shown that one of $\omega^2_\pm < 0$ for any value of
235: $P'$ or $V''$, meaning that there is always one unstable mode in
236: addition to the unstable $\omega=0$ mode. This result does not
237: contradict previous studies of the stability of the extra
238: dimensions in BGC because these works were only
239: concerned with the stability of the
240: radion\cite{bw,BWlinstab}. Instead, this result
241: shows that stabilization of the dilaton and the radion is
242: not possible without the introduction of some new potential for the
243: radion, and it demonstrates the importance of a rolling dilaton to the
244: stabilization mechanism in BGC.
245:
246:
247: \section{Conclusions}
248: The original proposal of Brandenberger and Vafa argues why, within a
249: string theory context, only three dimensions will be able to grow
250: large.\cite{bv} This idea has been extended to argue for a hierarchy of
251: subspaces, has been numerically verified, and issues such
252: the stability of the internal dimensions have been explored\cite{abe,bek,bw}. A
253: remaining aspect of this scenario is to understand the evolution of
254: the dilaton, an evolution which must stabilize at some later phase,
255: thus marking the transition to general relativity. To this end, we have introduced a potential for the dilaton
256: into the original action and tracked the evolution of perturbations
257: about a static configuration.
258: \par
259: The analysis indicates the importance of a rolling dilaton in order
260: for the negative-pressure winding modes to inhibit growth of the extra
261: dimensions. When the dilaton stabilizes, the original action reduces to
262: the usual Einstein-Hilbert action of general relativity, where all
263: sources act to accelerate expansion, causing the radion to grow. The eigenvalues of eq.~(\ref{eq:bgcev})
264: quantify this previous statement, indicating that any static
265: solutions are unstable to perturbations, so that two unstable modes
266: will always exist. Thus, solutions with one
267: growing mode (corresponding to three directions growing large), and
268: two stable modes (corresponding to a stable dilaton and radion) are
269: physically implausible within the normal BGC setup.
270: \par
271: Although BGC can stabilize the radion during the dilaton-gravity
272: epoch, some new mechanism must stabilize the internal dimensions once
273: the dilaton has stopped rolling.\cite{bc} Presumably the onset of the
274: dilaton potential will coincide with the appearance of a potential for
275: the radion as well: one possibility is from compactifications with
276: fluxes\cite{gkp}. This possibility provides an interesting link
277: between BGC and the work of Giddings, Kachru, and
278: Polchinski and subsequent investigations.
279:
280:
281: %\begin{figure}[ht]
282: %\epsfxsize=10cm %width of figure - will enlarge/reduce the figures
283: %\epsfbox{fig3.eps}
284: %\figurebox{2cm}{3cm}{} %to have a box alone
285: %aar\centerline{\epsfxsize=3.9in\epsfbox{procs-fig1.eps}}
286: %\caption{First 3 normalized frequencies versus release location for
287: %clamped simply supported beam with internal slide
288: %release. \label{inter}}
289: %\end{figure}
290:
291:
292:
293: \section*{Acknowledgments}
294: We would like to thank Horace Stoica for helpful discussions.
295: A.~B. would like to acknowledge NSERC for their support in this
296: research via grant ES~B~267789, as well as support from the Physics Department at
297: McGill University.
298:
299: %\appendix
300:
301: %\section{First Appendix}
302:
303: %\renewcommand{\theequation}{A.\arabic{equation}}
304:
305: %\section{Footnotes}
306: %Footnotes should be numbered sequentially in superscript lowercase
307: %roman letters.\footnote{Footnotes should be typeset in 8~pt Times
308: %roman at the bottom of the page.}
309:
310:
311: \begin{thebibliography}{0}
312: %\bibitem{BV}
313:
314: %\cite{bv}
315: \bibitem{bv}
316: R.~H.~Brandenberger and C.~Vafa,
317: ``Superstrings In The Early Universe,''
318: Nucl.\ Phys.\ B {\bf 316}, 391 (1989).
319: %%CITATION = NUPHA,B316,391;%%
320:
321: \bibitem{tv}
322: A.~A.~Tseytlin and C.~Vafa,
323: ``Elements of string cosmology,''
324: Nucl.\ Phys.\ B {\bf 372}, 443 (1992)
325: [arXiv:hep-th/9109048].
326: %%CITATION = HEP-TH 9109048;%%
327:
328: %\cite{abe}
329: \bibitem{abe}
330: S.~Alexander, R.~H.~Brandenberger and D.~Easson,
331: ``Brane gases in the early universe,''
332: Phys.\ Rev.\ D {\bf 62}, 103509 (2000)
333: [arXiv:hep-th/0005212].
334: %%CITATION = HEP-TH 0005212;%%
335:
336: %\cite{gkp}
337: \bibitem{gkp}
338: S.~B.~Giddings, S.~Kachru and J.~Polchinski,
339: ``Hierarchies from fluxes in string compactifications,''
340: Phys.\ Rev.\ D {\bf 66}, 106006 (2002)
341: [arXiv:hep-th/0105097].
342: %%CITATION = HEP-TH 0105097;%%
343:
344: %\cite{bek}
345: \bibitem{bek}
346: R.~Brandenberger, D.~A.~Easson and D.~Kimberly,
347: ``Loitering phase in brane gas cosmology,''
348: Nucl.\ Phys.\ B {\bf 623}, 421 (2002)
349: [arXiv:hep-th/0109165].
350: %%CITATION = HEP-TH 0109165;%%
351:
352: \bibitem{bw}
353: S.~Watson and R.~Brandenberger,
354: ``Stabilization of extra dimensions at tree level,''
355: JCAP {\bf 0311}, 008 (2003)
356: [arXiv:hep-th/0307044].
357:
358: %\cite{BWlinstab}
359: \bibitem{BWlinstab}
360: S.~Watson and R.~Brandenberger,
361: ``Linear perturbations in brane gas cosmology,''
362: JHEP {\bf 0403}, 045 (2004)
363: [arXiv:hep-th/0312097].
364: %%CITATION = HEP-TH 0312097;%%
365:
366: %\cite{cfos}
367: \bibitem{cfos}
368: R.~H.~Cyburt, B.~D.~Fields, K.~A.~Olive and E.~Skillman,
369: ``New BBN limits on physics beyond the standard model from He-4,''
370: [arXiv:astro-ph/0408033].
371: %%CITATION = ASTRO-PH 0408033;%%
372:
373: \bibitem{bc}
374: A.~Berndsen and J.~Cline, In Progress, (2004).
375: \end{thebibliography}
376:
377: \end{document}
378: