hep-th0304237/y6.tex
1: \documentclass[12pt]{article}
2: \topmargin -1.1cm
3: \oddsidemargin 0cm \evensidemargin 0cm
4: \textwidth 16cm
5: \textheight 23cm
6: \usepackage[dvips]{graphicx,psfrag}
7: %\renewcommand{\thesection}{\arabic{section}\setcounter{equation}{0}}
8: \renewcommand{\theequation}{\thesection.\arabic{equation}}
9: \renewcommand{\thefootnote}{\fnsymbol{footnote}}
10: \def\beqa{\begin{eqnarray}}
11: \def\eeqa{\end{eqnarray}}
12: \def\papertitlepage{\baselineskip 3.5ex \thispagestyle{empty}}
13: \def\Title#1{\vspace{1.5cm}\begin{center}
14:  {\Large\bf #1} \end{center}
15: \vspace{1cm}}
16: \def\Authors#1{\begin{center} {\large #1} \end{center}}
17: \def\Abstract{\vspace{0.3cm}\begin{center} {\large\bf Abstract}
18:            \end{center} %\parbigskip
19: }
20: %\def\ICRRnumber#1#2#3{\hfill \begin{minipage}{3cm} #1
21: %              \parn #2 \parn #3 \end{minipage}}
22: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
23: \begin{document}
24: \papertitlepage
25: \vspace*{-1 cm}
26: %\ICRRnumber{UT-Komaba/03-8  }{April 2003}{hep-th/0304237 }
27: \hfill 
28: \begin{minipage}{4cm} {UT-Komaba/03-8}
29:           \par\noindent  {April  2003}\par\noindent {hep-th/0304237}  
30: \end{minipage}
31: %%%%%%%%%%%%%%%%%%%%%%%%%%%%
32: \Title{D-brane Dynamics and Creations of \\
33: Open and Closed Strings after Recombination}
34: 
35: \Authors{{   Takeshi Sato
36: \footnote{tsato@hep1.c.u-tokyo.ac.jp}} \\
37:  \vskip 3ex
38: {\itshape  Institute of Physics, University of Tokyo, \\
39: 3-8-1 Komaba, Meguro-ku, Tokyo 153-8902 Japan} \\
40: }
41: 
42: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
43: 
44: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
45: \Abstract
46: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
47: 
48: A quantum-mechanical technique is used within the framework of U(2) 
49: super-Yang-Mills theory to investigate what happens in the process
50: after recombination of two D-p-branes at one angle.   
51: Two types of initial conditions are considered, 
52: one of which with $p=4$ is a candidate of inflation mechanism.
53: It is observed that the branes' shapes come to have three extremes 
54: due to localization of tachyon condensation. 
55: Furthermore, open string pairs connecting 
56: the decaying D-branes are shown to be created;
57: most part of the released energy is used to create them.
58: It also strongly suggests 
59: that creation of closed strings happens afterward.
60: Closed strings as gravitational radiation from the D-branes are also 
61: shown to be created. 
62: A few speculations are also given on implications of the above phenomena 
63: for an inflation model.
64: 
65: \newpage
66: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
67: 
68: 
69: \section{Introduction}
70: \setcounter{equation}{0}
71: 
72: Recombination of D-branes at angles are processes 
73: increasing its value recently 
74: from both phenomenological 
75: and theoretical point of view, related to the presence of
76: tachyonic modes which appear as modes of open strings
77: between the two D-branes\cite{angle}.  
78: In the brane inflation scenarios,
79: these are some of the promising candidates 
80: to explain the mechanism of inflation\cite{halyo}\cite{hirano1}%
81: \cite{ky}\cite{oneangle}\cite{int2}\cite{hirano2}\cite{tye1}%
82: \cite{2angles}\cite{bro}\cite{br1}.
83: %and recombination of two D-p-branes is one of them
84: %\cite{oneangle}\cite{int2}\cite{tye1}\cite{2angles}.
85: In addition, among various string constructions of Standard Model,
86: intersecting brane models are one class of hopeful
87: candidates\cite{bl1}\cite{chiral}\cite{bl2}\cite{getjust}\cite{higgs}%
88: \cite{koko1}\footnote{
89: For a recent review and other references, see e.g.
90: ref.\cite{ur1}.},
91: in some of which it has been proposed that
92: the recombination process occurs as Higgs mechanism% 
93: \cite{chiral}\cite{higgs}\cite{koko1} (see also \cite{bachas}\cite{koko2}). 
94: %although for this case  
95: %multiple angle cases are rather favored. 
96: On the other hand, this system 
97: and process can be regarded as a generalized 
98: setting of $D\bar{D}$ system and its annihilation process, 
99: which has been studied
100: thoroughly\cite{sen1}\cite{y1}\cite{y2}\cite{sen5}%
101: \cite{lar}\cite{tera}\cite{tye2}. 
102: In this way, this is one of the most important phenomena 
103: to explore at present in string theory.
104: 
105: The behavior of the system after  recombination, however, 
106: had been almost unexamined until recently.
107: Suppose one consider a recombination of two D-branes at one angle, 
108: wrapping around some cycles in a compact space.
109: One might expect that after the recombination, 
110: each brane would begin to take a ``shorter cut'' and then make 
111: a damped oscillation around some stable configuration of branes,
112: while radiating RR gauge and gravitational waves (and some others),
113: leading to the stable configuration.
114: However, the case was that only the shape of the recombined branes 
115: at the initial stage was merely inferred\cite{pol}, 
116: though the final (infinitely late time) 
117: configuration can be determined in some cases (e.g. see ref.\cite{2angles}).
118: 
119: Recently, in ref.\cite{hn}, K. Hashimoto and Nagaoka
120: proposed that a T-dual of super Yang-Mills theory (SYM)
121: can be an adequate framework
122: to describe the process, and in the case of two D-strings at one angle and via 
123: classical analysis, they presumed how the shapes of the branes develop 
124: at the very initial stage\footnote{Just after ref.\cite{hn}, the process 
125: is also 
126: analyzed via tachyon effective field theory in ref.\cite{wh} and very recently,
127: the subsequent work was done by K. Hashimoto and W. Taylor in ref.\cite{khw}.}
128: 
129: Main purpose of this paper is to investigate 
130: the process after
131: the recombination via SYM
132: ``more rigorously'' and to reveal what happens in 
133: the process.
134: To be concrete, we will focus on the two points: 
135: time-evolution of the D-branes' shape,
136: and behavior or creations of fundamental strings.
137: The former can be of value in that 
138: we describes time-dependent, i.e. 
139: dynamical behavior of curved D-branes though for 
140: a rather short time-scale (which enables us to discuss
141: e.g. gravitational waves radiated by the branes).  
142: The former might also be regarded as a first step toward the
143: understanding of branes' shapes after the recombination 
144: in intersecting brane models as Standard Model,
145: which are responsible for physical quantities after the condensation
146: of Higgs fields. 
147: The latter point is what happens in the process itself.   
148: What we mean by ``more rigorously'' is that 
149: we set concrete initial conditions and make a quantum-mechanical 
150: analysis:
151: These are necessary steps in order to understand precise behavior of 
152: the system, because the 
153: system, its tachyon sector in particular, 
154: is essentially a collection of inverse harmonic oscillators,
155: and its behavior depends crucially on
156: the initial fluctuations of the system,
157: which can only be evaluated appropriately via a quantum analysis,
158: but has not been done at least in 
159: this setting.\footnote{
160: In the case of inflation and other systems, see 
161: ref.\cite{guth}\cite{boya1}, and for a non-perturbative analysis,
162: see ref.\cite{linde1}.}
163: 
164: The second purpose of this paper is to discuss
165: implications of the above two points (phenomena),
166: for the inflation scenario of the setting that two D-4-branes are
167: approaching each other at one angle as in
168: ref.\cite{oneangle}\cite{tye1},
169: aiming at the understanding of the whole process of the scenario.
170: 
171: For the above two purposes,
172: we consider two types of initial conditions for  
173: branes. As we will discuss later,
174: in order to make a quantum-mechanical analysis,
175: we need the fact that all the modes (except U(1) gauge fields on each of
176: the branes) 
177: have positive frequency-squareds at
178: $t=0$. So,
179: the first one we consider (denoted as case (I)) is that 
180: two D-p-branes have been parallel until $t=0$,
181: but are put
182: intersected at one angle $\theta$ at the instant $t=0$.
183: %\footnote{
184: %In the case of D-strings (and D-2-branes) 
185: %we consider the initial condition that
186: %the branes have been parallel to each other until $t=0$ instead,
187: %to regularize a typical value of fluctuations, as we will discuss later.} 
188: The second one (the case (II)) is that one of parallel two
189: D-p-branes were rotated 
190: by a small angle $\theta$ and are approaching each other very slowly.
191: The case (I) is one of the simplest conditions, rather easier 
192: to study the process itself.
193: The case (II) is more practical;
194: the case with $p=4$ is one of the hopeful setting for 
195: the brane inflation scenarios\cite{oneangle}\cite{tye1}.
196: From a practical viewpoint, the case (I) may seem less realistic,
197: but it may also be able to be regarded as some local or rough
198: approximation to more complicated setting of D-branes.
199: 
200: The outline of this paper is summarized as follows:
201: 
202: We consider two D-p-branes for each initial condition and for each 
203: $p\ge 1$, within the framework of a T-dual of SYM, 
204: a low energy effective theory of open strings around the D-branes
205: when distances between the branes are small, as in ref.\cite{hw}\cite{hn}.
206: Detailed set-ups are given in section 2.
207: 
208: In section 2 and 3 we evaluate time-evolution of typical amplitudes of 
209: tachyonic fluctuations in the following way; 
210: we consider fluctuations around the D-branes 
211: and make mode-expansions of them to examine their spectra.
212: Then, defining typical amplitudes (absolute values) of the fluctuations
213: in terms of VEV's of their absolute square,
214: we obtain them by using a time-dependent wave function of
215: each mode and focusing on contribution of tachyonic modes.
216: We note that though we use WKB approximation (second order one 
217: in fluctuations), 
218: it is enough good in the cases until the tachyons blow up,
219: because higher order terms are suppressed by a factor of
220: $O(g_{s})$, as we will discuss later.
221: 
222: In section 4 we discuss time-evolution of  D-brane's shape after 
223: recombination; we show that a seemingly queer behavior of the shape 
224: appears and explain why it happens. 
225: The information of the shape is obtained from a  
226: transverse U(2) scalar field 
227: by diagonalizing  its VEV's (i.e. choosing a certain gauge)
228: and looking at its diagonal elements\cite{hn}.\footnote{
229: For case (II), there appears a non-commutativity
230: between VEV's of scalar fields, which implies their uncertainty
231: relation. We will discuss this point in section 8.}
232: %So we will proceed with our discussion by concentrating on a VEV of one field
233: %while leaving alone another.}
234: We denote this as ``physical gauge'' because in the gauge direct 
235: correspondence holds between d.o.f. of each open string starting from and 
236: ending on certain branes
237: and d.o.f. of each U(2) matrix element.
238: 
239: 
240: Then, in section 5 we argue that open string pairs (composed of strings
241: with opposite orientations) connecting the two decaying branes are
242: created. 
243: Our argument is based on the behavior of the U(2) gauge and scalar fields 
244: {\it in the physical gauge};
245: investigation of
246: its non-vanishing energy density, the correspondence of d.o.f. stated
247: above, and charge densities of two U(1) symmetries on the D-branes
248: lead to such conclusion.
249: We also evaluate number density of the open strings and 
250: distribution of energy to the open strings, and show that 
251: most part of the released energy are used to create the pairs.
252: 
253: In section 6, we discuss creation of closed strings, going out of the
254: framework of SYM: We show that the decaying branes radiate  
255: closed strings as gravitational wave, regarding the branes as a
256: collection of massive small objects; its energy flux is evaluated
257: quantitatively. 
258: We also discuss the system after some relatively long time, i.e. 
259: beyond the reach of SYM;
260: creation of closed strings 
261: are strongly suggested after the two branes are separated further.
262: 
263: In section 7, we discuss implications of the above results for 
264: a brane inflation scenario, on the efficiency of the reheating and 
265: mechanism to create ``reheated'' gauge fields and fermions.
266: 
267: Finally in section 8 we present summary and discussion;
268: we discuss some points on 
269: non-commutative scalar fields appearing in the case (II).
270: In the appendix we give a short review of giving
271: a width of a Gaussian wave function for each mode at arbitrary time.
272: 
273: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
274: 
275: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
276: \section{Set-ups and preliminaries}
277: \setcounter{equation}{0}
278: \setcounter{footnote}{0}
279: 
280: In this section we present set-ups and preliminaries for a
281: quantum-mechanical analysis for the case (I) and (II).
282: 
283: First we present our notation and set-ups;
284: in this paper we set $l_{s}= 1 $ for convenience 
285: (and revive it when needed), 
286: and set the string coupling $g_{s}$ very small 
287: and consider D-p-branes in a 10-dimensional
288: flat target space with a metric $g_{\mu\nu}={\rm diag}(-1,1,..,1)$
289: and coordinates $x_{M}$ ($M=0,..,9$). The world-volumes are parametrized by 
290: $x_{0},x_{1},\cdots x_{p}$.
291: We compactify space dimensions $x_{p},\cdots,x_{9}$ 
292: on a (10-p)-torus with periods $L_{a}$ for $a=p,\cdots,9$.
293: We denote $x_{0}$ as $t$ and $x_{p}$ as $x$ below. 
294: The initial condition for the case (I) is the configuration that 
295: two D-p-branes have been parallel to each other with a displacement  
296: $y_{0}$ until $t=0$ but are put
297: intersected at one angle $\theta$ at the instant $t=0$.
298: The setting is depicted in Fig.1.
299: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
300: \begin{figure}[h] 
301: \setlength{\unitlength}{1mm}
302: \begin{picture}(160,40)
303: \put(4,17.5){\vector(1,0){56}}
304: \put(30.5,8){\vector(0,1){19}}
305: \put(14,17.5){\vector(0,1){5.5}}
306: \put(14,17.5){\vector(0,-1){5.5}}
307: \put(4,8){\line(1,0){56}}
308: \put(4,27){\line(1,0){56}}
309: \put(4,8){\line(0,1){19}}
310: \put(60,8){\line(0,1){19}}
311: \put(4,12){\thicklines\vector(1,0){56}}
312: \put(4,23){\thicklines\vector(1,0){56}}
313: %\put(120,18){\makebox(5,5){$v/2$}}
314: %\put(120,12){\makebox(5,5){$v/2$}}
315: \put(29,27){\makebox(5,5){$x_{p+1}$}}
316: \put(54,17){\makebox(5,5){$x$}}
317: \put(44,28){\makebox(5,5){$Dp$}}
318: \put(44,2){\makebox(5,5){$Dp$}}
319: \put(16,17){\makebox(5,5){$y_{0}$}}
320: \put(31,2){\makebox(5,5){$L_{p}$}}
321: \put(63,15){\makebox(5,5){$L_{p+1}$}}
322: \put(26,33){\makebox(5,5){(i) $t<0$}}
323: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
324: \put(71,16.8){\thicklines\vector(1,0){5}}
325: \put(71,18.2){\thicklines\vector(1,0){5}}
326: \put(80,17.5){\vector(1,0){56}}
327: \put(108,8){\vector(0,1){19}}
328: \put(80,8){\thicklines\vector(3,1){56}}
329: \put(80,27){\thicklines\vector(3,-1){56}}
330: \put(80,8){\line(1,0){56}}
331: \put(80,27){\line(1,0){56}}
332: \put(80,8){\line(0,1){19}}
333: \put(136,8){\line(0,1){19}}
334: \put(125,18){\makebox(5,5){$\theta /2$}}
335: \put(125,12){\makebox(5,5){$\theta /2$}}
336: \put(105,27){\makebox(5,5){$x_{p+1}$}}
337: %\put(111,22){\makebox(5,5){$x_{p+1}$}}
338: \put(131,17){\makebox(5,5){$x$}}
339: \put(126,28){\makebox(5,5){$Dp$}}
340: \put(126,2){\makebox(5,5){$Dp$}}
341: \put(107,2){\makebox(5,5){$L_{p}$}}
342: \put(139,15){\makebox(5,5){$L_{p+1}$}}
343: \put(102,33){\makebox(5,5){(ii) $t= 0$}}
344: \end{picture}
345: \caption{The initial configuration for the case (I)}
346: \end{figure}
347: 
348: \noindent
349: The initial condition for the case (II) 
350: is the one that one of parallel two D-p-branes
351: were rotated by a small angle $\theta$ and are approaching each other 
352: with a small relative velocity $v$.
353: The simplest setting for each case is represented in Fig.2.
354: We note that $x$ is parallel to neither of the branes,
355: but can parametrize a world-space
356: of either branes, though the energy density
357: is $\sqrt{1+\beta^2}$ times larger then the parallel embedding
358: (where $\beta\equiv \tan(\theta/2)$).
359: Although there are multiple intersecting points in 
360: cases of compact dimensions, in this paper we assume that the
361: recombination happens only at the point $x=0$ for simplicity.
362: \begin{figure}[h] 
363: \setlength{\unitlength}{1mm}
364: \begin{picture}(160,40)
365: \put(4,17.5){\vector(1,0){56}}
366: \put(32,8){\vector(0,1){19}}
367: \put(4,8){\thicklines\vector(3,1){56}}
368: \put(4,27){\thicklines\vector(3,-1){56}}
369: \put(4,8){\line(1,0){56}}
370: \put(4,27){\line(1,0){56}}
371: \put(4,8){\line(0,1){19}}
372: \put(60,8){\line(0,1){19}}
373: \put(49,18){\makebox(5,5){$\theta /2$}}
374: \put(49,12){\makebox(5,5){$\theta /2$}}
375: \put(29,27){\makebox(5,5){$x_{p+1}$}}
376: \put(55,17){\makebox(5,5){$x$}}
377: \put(50,28){\makebox(5,5){$Dp$}}
378: \put(50,2){\makebox(5,5){$Dp$}}
379: \put(31,2){\makebox(5,5){$L_{p}$}}
380: \put(63,15){\makebox(5,5){$L_{p+1}$}}
381: \put(26,33){\makebox(5,5){$x_{p}x_{p+1}$-plane}}
382: %%%%%%%%%%%%%%%%%%%
383: \put(80,17.5){\vector(1,0){56}}
384: \put(106.5,8){\vector(0,1){19}}
385: \put(88,17.5){\vector(0,1){5.5}}
386: \put(88,17.5){\vector(0,-1){5.5}}
387: \put(80,8){\line(1,0){56}}
388: \put(80,27){\line(1,0){56}}
389: \put(80,8){\line(0,1){19}}
390: \put(136,8){\line(0,1){19}}
391: \put(115,22){\vector(0,-1){3}}
392: \put(115,13){\vector(0,1){3}}
393: \put(80,12){\thicklines\vector(1,0){56}}
394: \put(80,23){\thicklines\vector(1,0){56}}
395: \put(120,18){\makebox(5,5){$v/2$}}
396: \put(120,12){\makebox(5,5){$v/2$}}
397: \put(105,27){\makebox(5,5){$x_{9}$}}
398: \put(130,17){\makebox(5,5){$x$}}
399: \put(120,28){\makebox(5,5){$Dp$}}
400: \put(120,2){\makebox(5,5){$Dp$}}
401: \put(92,17){\makebox(8,5){$z(t=0)$}}
402: \put(107,2){\makebox(5,5){$L_{p}$}}
403: \put(139,15){\makebox(5,5){$L_{p+1}$}}
404: \put(102,33){\makebox(5,5){$x_{p}x_{9}$-plane}}
405: \end{picture}
406: \caption{The initial configuration for case (II) (t=0)}
407: \end{figure}
408: 
409: Let us give a brief review of the process.
410: The configuration represented as (ii) in Fig.1 cannot be distinguished from 
411: that of the two branes in which both are 
412: bent and touched at $x_{p}=x_{p+1}=0$
413: since they have the same energy density and flux.
414: The latter is unstable and each brane begins to take a shorter cut
415: to save the energy. This is the recombination process.
416: Their shapes just after that are expected to be like a hyperbola.
417: This instability leads to the existence of a complex tachyon field 
418: with a mass-squared 
419: \beqa
420: m^{2}=-\theta,\label{mass1}
421: \eeqa 
422: appearing as modes of a string connecting the two D-branes,
423: as shown in ref.\cite{angle}.
424: We will investigate the behavior of the system using 
425: T-dualized versions of U(2) SYM as in ref.\cite{hw}\cite{hn}. 
426: 
427: For each $p$, the SYM action with $x_{p+1},\cdots,x_{9}$ T-dualized 
428: takes the form
429: \beqa
430: S_{2Dp}=-T_{p} \int d^{p+1} x \ 
431: {\rm Tr} \{\ (1+)\ \frac{1}{4}F_{\mu\nu}F^{\mu\nu} +\frac{1}{2 
432: }(D_{\mu}X_{i})^2
433: - [X_{i},X_{j}]^{2}\}
434: \eeqa
435: where $T_{p}=1/g_{s}$ is the tension of a D-p-brane. 
436: $F_{\mu\nu}$ is the field
437: strength of the world-volume U(2) gauge field $A_{\mu}$ for
438: $\mu=0,1,..,p$. $X_{i}$ for $i=p+1,..,9$ are 
439: U(2) adjoint scalar fields corresponding to 
440: coordinates $x_{i}$ transverse to the branes. 
441: We note that to apply SYM, the relation 
442: $L_{p+1}\simeq L_{p}\theta/2 < l_{s}(=1)$ should be satisfied, 
443: leading to the constraint
444: \beqa
445: L_{p}<\frac{2}{\theta}.\label{lcond1}
446: \eeqa
447: 
448: We discuss dynamical behavior of the system for $t\ge 0$.
449: The ``background'' D-p-branes for each case for $t\ge 0$ are
450: represented by the configuration (VEV's):
451: \beqa
452: \begin{array}{ccc}
453: X_{p+1}= \left(
454: \begin{array}{cc}
455: \beta x & 0 \\
456: 0 & -\beta x
457: \end{array}
458: \right),
459:  &
460: X_{9}= \left(
461: \begin{array}{cc}
462: z/2 & 0 \\
463: 0 & -z/2
464: \end{array}
465: \right) ,
466: & X_{p+2}=\cdots =X_{8}=A_{\mu}=0.
467: \end{array}
468: \eeqa
469: where $\beta\equiv \tan(\theta/2)$. 
470: For the case (I) ($z(t)=0$),
471: this is T-dual 
472: to the configuration of
473: two D-(p+1)-brane with a constant field strength
474: $F_{p,p+1} = \beta$, as discussed in ref.\cite{hw}. 
475: %Since tachyonic fluctuations are modes of a string
476: %connecting the two branes, 
477: %they are expected (and will be shown) to arise in off-diagonal 
478: %elements of $A_{\mu}$ and $X_{p+1}$.
479: Actually, in ref.\cite{baal}, 
480: tachyonic modes around the T-dualized ($F_{p,p+1} = \beta$)
481: background were shown to appear in off-diagonal ones of
482: $A_{p}$ and $A_{p+1}$ in a certain gauge,
483: which is T-dual to $X_{p+1}$ in the present case.
484: However, the connection via T-duality does not of course means
485: that the two systems are physically equivalent. 
486: (Actually, as we will show later,  
487: T-duality has a non-trivial effect for the system to change
488: of the width of a wave function by a  $\sqrt{2}$.)
489: Thus, we examine the spectrum from the beginning.
490: For case (II) ($z(t)\ne 0$),
491: $z/2$ is a VEV of an adjoint Higgs field
492: which decreases slowly.
493: U(2) gauge symmetry is spontaneously broken to
494: U(1) $\times$ U(1) by the non-trivial VEV's. 
495: Since we set $g_{s}$ and $\theta$
496: very small, the force between the two branes is so
497: weak\cite{jabbari}\cite{pol}(and for short distance, see
498: \cite{oneangle})
499: that we can regard the velocity as a constant, leading to
500: \beqa
501: z=z_{0} -v t.
502: \eeqa
503: In this case  
504: potentially tachyonic modes should appear 
505: in the off-diagonal elements of $A_{p}$ and $X_{p+1}$ again.
506: In addition, there might be  a possibility
507: that tachyonic modes appear in off-diagonal elements of 
508: $A_{0}$ and $X_{9}$ because the two D-p-branes
509: approaching with a small velocity is 
510: T-dual to a two D-(p+1)-brane system
511: with a constant field strength (though it is electric),
512: and is also 
513: considered to be a Wick-rotated version of 
514: branes at one angle $v$\cite{pol}.
515: So, we denote the fluctuations, 
516: including those of $A_{0}$ and $X_{9}$, as
517: \begin{equation}
518: \begin{array}{ccc}
519: A_{\mu}= \left(
520: \begin{array}{cc}
521: 0 & c_{\mu}^{*} \\
522: c_{\mu} & 0
523: \end{array}
524: \right), &
525: X_{p+1}= \left(
526: \begin{array}{cc}
527: 0 & d^{*} \\
528: d & 0 
529: \end{array}
530: \right) 
531: , &
532: X_{9}= \left(
533: \begin{array}{cc}
534: 0 & w^{*} \\
535: w & 0 
536: \end{array}
537: \right)
538: \end{array}\label{ansatz2}
539: \end{equation}
540: (and denote in particular, $c_{p}$ as $c$), and
541: discuss their spectra below.
542: 
543: The part of the action second order in the fluctuations is
544: \beqa
545: S_{2Dp}|_{{\rm 2nd}}
546: &=&T_{p} \int d^{p+1}x [ 
547: -\partial_{\mu}c^{*}_{\nu}\partial^{\mu}c^{\nu} 
548: +|\partial_{\mu}c^{\mu}|^2 
549: -\partial_{\mu}d^{*}\partial^{\mu}d 
550: -\partial_{\mu}w^{*}\partial^{\mu}w\nonumber\\ 
551: & &+2 i \beta (c^{*}d -c d^{*})
552: -2 i \beta x(c^{*}_{\mu}\partial^{\mu}d 
553: -c_{\mu} \partial^{\mu}d^{*}) 
554: +iv(c_{0}^{*}w-c_{0}w^{*})\nonumber\\
555: & &-iz(c^{*}_{\mu}\partial^{\mu}w 
556: -c_{\mu} \partial^{\mu}w^{*})
557: -(4(\beta x)^{2}+z^{2})|c_{\mu}|^{2}-|2\beta x w-zd|^{2}%\nonumber\\ 
558: %& &-\frac{1}{2}|c_{\mu}^{*}c_{\nu}-c_{\nu}^{*}c_{\mu}|^{2}
559: %-|c^{*} d-c d^{*}|^{2}]
560: ].\label{secondaction22}
561: \eeqa 
562: %(For later use, we left the higher order terms of $c_{\mu}$ and $d$ 
563: %in the action.)
564: We consider such a gauge fixing condition 
565: that it is written as $D_{\mu}^{(0)}A^{\mu}\equiv\partial_{\mu}
566: A^{\mu}-i[A_{\mu}^{(0)},A^{\mu}]=0$ in a T-dualized pure 
567: Yang-Mills system, which gives
568: \beqa
569: \partial_{\mu}c^{\mu}=i(2\beta x d+zw).\label{gf3}
570: \eeqa
571: This is a generalized version of the condition
572: chosen in ref.\cite{baal}\cite{hw}.
573: Integrating by part some terms and  using (\ref{gf3}),
574: we have 
575: \beqa
576: S_{2Dp}|_{{\rm 2nd}}&=&  T_{p} \int d^{p+1}x [
577: -c^{*}_{\nu} \{ -(\partial_{\mu})^{2}
578: +4(\beta x)^{2} +z^{2}\} c_{\nu} 
579: - d^{*}  \{ -(\partial_{\mu})^{2}
580: +4(\beta x)^{2} +z^{2} \}d \nonumber\\
581: & &-w^{*}\{ -(\partial_{\mu})^{2}
582: +4(\beta x)^{2}+z^{2} \} w
583: -4 i \beta (c^{*}d -c d^{*})-2iv(c_{0}^{*}w-c_{0}w^{*})
584: ].\label{secondaction23}
585: \eeqa
586: With the redefinition of the fields as
587: \beqa
588: \tilde{c} \equiv \frac{c+id}{\sqrt{2}} ,\ 
589: \tilde{d} \equiv \frac{d+ic}{\sqrt{2}}, 
590: \label{defctilder} 
591: \eeqa
592: the term $-4 i \beta (c^{*}d -c d^{*})$ in (\ref{secondaction23})
593: give rise to their mass terms
594: $-(-4\beta)|\tilde{c}|^{2}-(4\beta)|\tilde{d}|^{2}$.
595: 
596: Here we set $z(t)=v=0$ and discuss the case (I) for a while. 
597: Let us consider mode expansions of the fields $\tilde{c}$, 
598: $\tilde{d}$ and $c_{\hat{\mu}}$ ($\hat{\mu}=0,1,\cdots,p-1 $)  
599: with respect to $x$ and $x_{\bar{\mu}}$ 
600: ($\bar{\mu}=1,\cdots,p-1$) (the latter for D-p-branes with $p\ge 2$) 
601: to examine their spectra.
602: As for $x_{\bar{\mu}}$, we use $u_{k}(x_{\bar{\mu}})\equiv 
603: e^{ik_{\bar{\mu}}x^{\bar{\mu}}}$,
604: free wave functions with a wave vector $ k_{\bar{\mu}}$.
605: As for $x$, the operator for each field is written as
606: \beqa
607: [-(\partial_{x})^{2}+4(\beta x)^{2} -4 a \beta]\label{ho}
608: \eeqa 
609: where $a=1,-1,0$ for $\tilde{c}$, $\tilde{d}$ and 
610: $c_{\hat{\nu}}$, respectively.
611: If it were not for the periodicity of $x$,
612: the eigenvalues and eigenfunctions could be given by
613: $\tilde{m}^{2}_{n}$ and $f_{n}$ as
614: \beqa 
615: \tilde{m}^{2}_{n}&=&2\beta(2 n -a)\label{massform}\label{eigenv}\\
616: f_{n}&=&\frac{1}{\sqrt{2^{n}n!}}(\frac{2\beta}{\pi})^{1/4}
617: e^{-\beta x^{2}}H_{n}(\sqrt{2\beta}x) \label{eigenset}
618: \eeqa
619: for each non-negative integer $n$ where $H_{n}$ are Hermite polynomials;
620: since $x$ direction is compactified,   
621: exact eigenfunctions are expected to be mathematically complicated, 
622: e.g. related to theta functions
623: as in ref.\cite{baal}\cite{hw}.
624: However, we can see that $f_{n}$ in (\ref{eigenset}) is localized around 
625: $x=0$ with a width 
626: $\delta\equiv 1/\sqrt{2\beta}\simeq 1/\sqrt{\theta}$.
627: So, if one consider 
628: $L_{p}$ such that
629: \beqa
630: \frac{1}{\sqrt{\theta}} \ll L_{p},\label{lcond2}
631: \eeqa 
632: one can use (\ref{eigenv}) and (\ref{eigenset})  
633: as an approximate set of eigenvalues and eigenfunctions.
634: %\footnote{The width of $f_{n}$ is
635: %the same as the case of D-stings in ref.\cite{hn}.}
636: Then, we can see that
637: the lowest mode of $\tilde{c}$ (we denote as $\tilde{c}_{0}$)
638: is a complex tachyon field with a mass-squared
639: \beqa
640: (m_{0})^{2}=-2 \beta,
641: \eeqa 
642: which agrees with (\ref{mass1}) obtained in ref.\cite{angle} 
643: for a small $\theta$, while
644: the lowest modes of $c_{\bar{\nu}}$ are massless and the others are
645: massive. 
646: We  note that the condition (\ref{lcond2}) is compatible with
647: the one (\ref{lcond1}) if $\theta \ll 1$. 
648: So, we consider such a range of  $L_{p}$ throughout this paper.
649: (We set $\theta\simeq 2\beta=10^{-2}$ and $L_{p}=100$ later.)
650: We also note that the approximate width of $f_{n}$
651: is larger than the eigenfunction in the T-dualized theory
652: in ref.\cite{hw}\cite{baal} by  a factor $\sqrt{2}$, 
653: which may be one of non-trivial effects of the T-duality. 
654: 
655: Expanding $\tilde{c}$ and the other fields such as 
656: \beqa
657: \tilde{c}(x_{\mu}) =\sum_{n}\int\frac{d^{p-1}{\bf k}}{(2\pi)^{p-1}} 
658: \tilde{c}_{n,k_{\bar{\mu}}} (t) 
659: u_{k}(x_{\bar{\mu}})f_{n}(x), 
660: \label{expansion2}
661: \eeqa
662: we obtain the action 
663: \beqa
664: S_{2Dp}|_{{\rm 2nd}}=T_{p}\int d t
665: \sum_{n,k_{\hat{\mu}}} \{ |\frac{d\tilde{c}_{n,k}}{dt}|^{2}
666: -\omega_{n,k}^{\ \ \ 2}|\tilde{c}_{n,k}|^{2}
667: +(\tilde{c}_{n,k}\to \tilde{d}_{n, k}, \ c_{\hat{\nu},n,k})\}
668: \label{secondaction3}.
669: \eeqa
670: where $\omega_{n,k}^{\ \ \ 2}$ for $t\ge 0$ is given as
671: \beqa
672: \omega_{n,k}^{\ \ \ 2}= k^{2}+2 \beta (2n-a)\label{omega1}
673: \eeqa
674: where $k\equiv \sqrt{(k_{\bar{\mu}})^{2}}$.
675: (Note that $a=1,-1,0$ for $\tilde{c}_{n,k}, \tilde{d}_{n,k},  
676: c_{\hat{\nu},n,k}$, 
677: respectively.)
678: This looks much like the action of a collection of harmonic 
679: oscillators
680: with a mass $T_{p}$, although it includes modes with negative 
681: frequency-squareds; we can easily see that
682: the modes $\tilde{c}_{0,k}$ with 
683: $0 \le k^{2} \le 2\beta$
684: have negative $\omega^{2}$'s (while all the other modes 
685: have non-negative $\omega^{2}$'s). 
686: Thus, they are the tachyonic modes in the case (I) with $p\ge 2$.
687: For the case of D-strings, there is no d.o.f. of $k_{\bar{\mu}}$ and
688: $\tilde{c}_{0}$ is the only (complex) tachyonic mode with  
689: $(\omega_{0})^{2}=-2\beta $ for $t\ge 0$.
690: 
691: For $t<0$, one should set
692: $2\beta =\theta =0$ and instead turn on a displacement of the branes
693: $y_{0}$ in $x_{p+1}$ direction. $y_{0}$ contribute to the
694: second order action as a mass-squared for each of the off-diagonal
695: fluctuations. Since the direction of $x_{p}$ is compactified,
696: the momentum component $k_{p}$ is discretized as 
697: $k_{p}=2\pi n_{p}/L_{p}$ 
698: for an integer $n_{p}$, and the mode functions is
699: \beqa
700: g_{n_{p}}(x)=e^{ik_{p}x}/\sqrt{L_{p}}.
701: \eeqa
702: The modes  
703: $\tilde{c}_{n_{p},k}^{(0)}\equiv
704: (c_{n_{p},k}^{(0)}+id_{n_{p},k}^{(0)})/\sqrt{2}$,
705: corresponding to $g_{n_{p}}(x)$,
706: have such $(\omega_{n_{p},k}^{(0)})^{2}$  as
707: \beqa
708: \omega_{n,k}^{(0)\ \ 2}=(y_{0})^{2}+(k_{\bar{\mu}})^{2}
709: +(2\pi n_{p}/L_{p})^{2}\label{omega01}
710: \eeqa
711: (where we attach the index $(0)$ to the qualities defined for $t<0$).
712: 
713: Let us move on to the case (II) where $z\ne 0$.
714: As seen in (\ref{secondaction23}), $z^{2}$ merely plays a role of 
715: a mass-squared for each field.
716: Making mode-expansions of $\tilde{c},\tilde{d}$, and
717: $c_{\hat{\mu}}$ in terms of $u_{k}$ and $f_{n}$,
718: we have the action (\ref{secondaction3}) with frequencies
719: \beqa
720: \omega_{n,k}^{'\ \ \ 2}= k^{2}+2 \beta (2n-a)+(z_{0}-v t)^{2}
721: \label{omega2}
722: \eeqa
723: where $a=1,-1,0$ for $\tilde{c}_{n,k} \tilde{d}_{n,k},
724: c_{\bar{\nu},n,k}$, respectively again.
725: As we will discuss later,
726: we need the fact that all the modes have positive frequency-squareds 
727: at $t=0$, so we set $z_{0}=0.5 l_{s}$, relatively a large value
728: but not to exceed the applicable range of SYM.
729: Since $z(t)$ decreases gradually to zero,
730: again (potentially) tachyonic modes are 
731: the modes $\tilde{c}_{0,k}$ with $k^{2}\le 2\beta$ for the case with
732: $p\ge 2$, and $\tilde{c}_{0}$ is the only (complex) tachyonic mode
733: for the case of D-strings.
734: 
735: Qualitative behavior of the system in the case (II) is as follows:
736: at $t=0$, all the modes have positive $\omega^{2}$'s, hence have
737: some small fluctuations.
738: As $z(t)$ decreases to $z=\sqrt{2\beta}$ (which gives
739: $\omega_{0,0}^{'\ \ 2}=0$), $\tilde{c}_{0,0}$ starts to blow up, 
740: and
741: so do the modes $\tilde{c}_{0,k}$'s with a small $|k|$ one after
742: another when   
743: $z$ becomes a critical displacement $z^{*}_{k}$ 
744: which gives $\omega_{0,k}^{'\ \ 2}=0$.
745: The blow-ups together give rise to the tachyon condensation
746: (to cause recombination).
747: 
748: In the next section we will focus on
749: the potentially tachyonic modes $\tilde{c}_{0,k}$ with $2\beta \gg v$,
750: and make a quantum-mechanical analysis
751: using the action (\ref{secondaction3}) with the obtained $\omega^{2}$'s.
752: %(\ref{omega1}) or (\ref{omega2}).
753: 
754: 
755: We note that one can regard $c_{0}$ and $w$ as non-tachyonic fields for
756: the parameter $2\beta \gg v$.
757: As for the fields $c_{0}$ and $w$,
758: the last term in (\ref{secondaction23})
759: might also seem to produce mass terms of linear combination of 
760: $c_{0}$ and $w$, 
761: in the same way as the case of $c$ and $d$,
762: with $2\beta$, $c$,$d$ replaced by $v$,$c_{0}$, $w$, respectively, 
763: resulting in a complex tachyonic field.
764: However, the sign of the kinetic term of $c_{0}$
765: is different from that of $w$, and, if one carry out 
766: the diagonalization, it results in a non-local action including 
767: square roots of differential operators;
768: it is unclear whether $c_{0}$ and $w$ include tachyonic modes.
769: Let us consider here the case with $2\beta \gg v$.
770: Then, even if there would be tachyon modes composed of
771: $c_{0}$ and $w$, 
772: the critical distances $z'_{*}$'s of them are much shorter than
773: those of $\tilde{c}$, and
774: it is much before the postulated tachyons of 
775: $c_{0}$ and $w$ get ``awake"
776: that the tachyon modes of $\tilde{c}$ condensate and
777: the recombination proceeds.
778: Thus, we consider the parameter region $2\beta \gg v$ 
779: and deal with $c_{0}$ and $w$ as non-tachyonic fields.  
780: 
781: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
782: 
783: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
784: \section{A quantum-mechanical analysis on tachyon condensation}
785: \setcounter{equation}{0}
786: \setcounter{footnote}{0}
787: 
788: In this section we analyze in detail how the tachyon fields condense,
789: applying quantum mechanics to the action (\ref{secondaction3}). 
790: We assume that the system is initially 
791: at zero temperature, i.e. at the ground state, for simplicity,
792: though the essential point of our argument will not be changed  
793: as long as the temperature is low.
794: 
795: One of the two basic ingredients (we will use here) is
796: that in a harmonic oscillator with a positive constant $\omega^{2}$, 
797: the ground state wave function is Gaussian, and 
798: its variable $q$ has a typical amplitude (absolute value) of fluctuation
799: $q_{{\rm typical}}\equiv \sqrt{<q^{2}>}=1/\sqrt{2m
800: \omega}=\Delta/\sqrt{2}$ 
801: where $<A>$ denotes a VEV of an operator $A$,
802: $m$ is a mass of $q$, and $\Delta $ is a width of $q$'s wave function.
803: Another basic one is that even if $\omega^{2}$ is time-dependent and 
804: becomes negative afterward, time-evolution of its wave function 
805: and hence that of $q_{{\rm typical}}$ can be deduced within WKB 
806: (second order) approximation,
807: by using its propagator or kernel, as in e.g. ref.\cite{fey}.
808: Taking these two into account, time-evolution behavior of the 
809: typical amplitude of each mode can be evaluated using quantum mechanics,
810: if an initial condition is such that 
811: $\omega^{2}$ of the mode is positive at $t=0$ and 
812: change into a negative one later:
813: the initial typical value is $\sqrt{g_{s}/2\omega(t=0)}$ since
814: in this case the tension $T_{p}=1/g_{s}$ corresponds to the mass.
815: Since the wave function at $t=0$ is Gaussian,
816: the absolute value of the wave function at any $t$ 
817: remains to be Gaussian 
818: within WKB (second order) approximation, and 
819: the system is characterized only by its width $\Delta(t)$,
820: which is equal to 
821: a typical value of the variable $q(t)_{{\rm typical}}$
822: times $\sqrt{2}$.
823: (These are reviewed in the appendix A. See ref.\cite{fey} 
824: for more detail.)
825: We note that typical amplitudes of the modes with non-negative 
826: $\omega^{2}$'s are 
827: suppressed by $\sqrt{g_{s}}$ 
828: in their numerators as 
829: $|\tilde{c}_{n, k}(t)|_{{\rm typical}} =\sqrt{g_{s}/2 \omega(t) }
830: \ \ (n\ge 1)$.
831: We also note that the WKB approximation is quite good 
832: until tachyonic modes blow up because
833: higher order terms are also suppressed by an extra factor of $O(g_{s})$.
834: 
835: If we define a typical amplitude of fields as 
836: $|\tilde{c}(x_{\mu})|_{{\rm typical}}\equiv \sqrt{<|\tilde{c}(x_{\mu})|^{2}>}$,
837: we have to sum
838: $|\tilde{c}_{n, k}|_{{\rm typical}}^{2} (f_{n})^{2}|u_{n,k}|^{2}$ 
839: with respect to $n$ and integrate it with respect to 
840: $k_{\bar{\mu}}$ except for the case of D-strings,
841: but the integral is divergent.
842: However, this is a kind of divergence that appears 
843: in a usual quantum field theory
844: and has nothing to do with the blow-up of tachyons.
845: Thus, we ignore the contribution of all the non-tachyonic modes 
846: and define the typical amplitude of
847: $\tilde{c}(x_{\mu})$ in the case of D-p-branes with $p\ge 2$ as
848: \beqa
849: |\tilde{c}(x_{\mu})|_{{\rm typical}} \equiv \{ \int_{k^{2}\le 2\beta} 
850: \frac{d^{p-1}{\bf k}}{(2\pi)^{p-1}}<|\tilde{c}_{0, k}(t)|^{2}> 
851: (f_{0})^{2}\}^{1/2}\equiv A(t)f_{0}\label{typfldef}
852: \eeqa
853: where $A(t)$ is a time-dependent part of $|\tilde{c}(x)|$ and $f_{0}(x)=
854: (2\beta/\pi)^{1/4}e^{-\beta x^{2}}$.
855: In the case of D-strings 
856: we define 
857: \beqa
858: |\tilde{c}(x_{\mu})|_{{\rm typical}}
859: \equiv\sqrt{<|\tilde{c}_{0}(t)|^{2}>}\equiv  A(t)f_{0}.
860: \label{typfldef2}
861: \eeqa
862: In this paper we regard the quantity (\ref{typfldef}) 
863: or (\ref{typfldef2}) as a VEV of $|\tilde{c}(x)|$ due 
864: to tachyon condensation.
865: As in ref.\cite{fey} and the appendix A, 
866: the time-dependent width $\Delta_{k}(t)$ of the wave function of 
867: $\tilde{c}_{0, k}(t)$, which is equal to 
868: $\sqrt{2}|\tilde{c}_{0,k}(t)|_{{\rm typical}}$,
869: can be written in terms of 
870: the initial width $\Delta_{k}(0)=\sqrt{g_{s}/\omega_{k}(t=0)}$ 
871: and two independent solutions 
872: to the equations of motion 
873: \beqa
874: \frac{d^{2} \tilde{c}_{0,k}}{dt^{2}}= 
875: -\omega_{0,k}^{2} \tilde{c}_{0,k}.
876: \label{eom1}
877: \eeqa
878: We denote the solutions as ``basis functions'' .
879: Below we solve the equation of motion for each case
880: and use the basis functions and $\Delta_{k}(0)$
881: to evaluate $|\tilde{c}(x_{\mu})|_{{\rm typical}}$.
882: 
883: For the case (I) with $p\ge 2$,
884: we can easily see that 
885: the basis functions for $t\ge 0$ are
886: $\cosh (Kt)$ and $\sinh (Kt)$
887: where $K\equiv \sqrt{-\omega^{2}_{0,k}}=\sqrt{2\beta-k^{2}}$ 
888: ($\ge 0$ for tachyonic modes),
889: and the initial width for $\tilde{c}_{n_{p},k}^{(0)}$ ($t\le 0$)
890: is $\Delta_{n_{p},k}^{(0)}(0)
891: =\sqrt{2}|\tilde{c}_{n_{p},k}^{(0)}(t=0)|_{{\rm typical}}=
892: \sqrt{g_{s}/\omega_{n_{p},k}^{(0)}}$
893: where $\omega_{n_{p},k}^{(0)}$ is given in (\ref{omega01}). 
894: However, the mode (eigen-) functions (of the operator 
895: with respect to $x$ arising in the action) 
896: change at $t=0$ from $g_{n_{p}}(x)$ to $f_{n}(x)$, so we have to ``translate'' 
897: $|\tilde{c}_{0,k}^{(0)}(t=0)|_{{\rm typical}}$ 
898: into $|\tilde{c}_{0,k}(t=0)|_{{\rm typical}}$.
899: (We note that $\tilde{c}_{0,k}$ is related only to  
900: $\tilde{c}_{n_{p},k}^{(0)}$ with the 
901: same $k_{\bar{\mu}}$
902: since the system keeps translation symmetry in $x_{\bar{\mu}}$
903: directions
904: even at $t=0$.)
905: If we define $N_{n,n_{p}}\equiv \int dx f_{n}^{*} g_{n_{p}}$,
906: it holds $g_{n_{p}}=\sum_{n}N_{n,n_{p}}f_{n}$ and we have 
907: $\tilde{c}_{n,k}=\sum_{n_{p}}N_{n,n_{p}}c_{n_{p},k}^{(0)}$,
908: and hence 
909: \beqa
910: \Delta_{k}(t=0)^{2}=2|\tilde{c}_{0,k}(t=0)|^{2}_{{\rm typical}}
911: =\sum_{n_{p}}|N_{0,n_{p}}|^{2}\frac{g_{s}}{\omega_{n_{p},k}^{(0)}}.
912: \label{delta01}
913: %|c_{n_{p},k}^{(0)}(t=0)|^{2}_{{\rm typical}}.
914: \eeqa
915: With the $\Delta_{k}(t=0)^{2}$, the typical value 
916: is given through (\ref{formaldelta}) as
917: \beqa
918: |\tilde{c}_{0,k}|_{{\rm typical}}^{2}=\frac{\Delta_{k}(t)^{2}}{2}=
919: \frac{\Delta_{k}(0)^{2}}{2}\cosh^{2}Kt
920: +\frac{g_{s}^{2}}{2\Delta_{k}(0)^{2}K^{2}}\sinh^{2}Kt, 
921: \label{deltak1}
922: \eeqa
923: and we can obtain the typical amplitude $|\tilde{c}(x_{\mu})|$ as
924: \beqa
925: |\tilde{c}(x_{\mu})|_{{\rm typical}}^{2}=(A(t)f_{0})^{2}=
926: \frac{\Omega_{p-2}}{(2\pi)^{p-1}}
927: \int_{0}^{\sqrt{2\beta}} k^{p-2}dk 
928: \frac{\Delta_{k}(t)^{2}}{2}(f_{0}(x))^{2}
929: \label{staticcfinal1}
930: \eeqa
931: where $\Omega_{p-2}$ is the volume of (p-2)-sphere.
932: This is the formula for time-evolution of the typical amplitude of
933: the tachyonic fluctuation for the case (I) with $p\ge 2$.
934: For $p=1$ (D-strings) case, we have
935: \beqa
936: |\tilde{c}_{0}(x_{\mu})|_{{\rm typical}}^{2}\equiv (A(t)f_{0})^{2} 
937: =\frac{\Delta_{k=0}(t)^{2}}{2}(f_{0})^{2}\label{staticstring1}
938: \eeqa
939: where $\Delta_{k}(t)$ is given in (\ref{deltak1}).
940: 
941: For all value of $p$,
942: order of $|\tilde{c}(x^{\mu})|_{{\rm typical}}$ in terms of $g_{s}$
943: is ${\cal O}(g_{s})$.  In addition, $|\tilde{c}(x^{\mu})|_{{\rm typical}}$
944: has x-dependence through the Gaussian function $f_{0}(x)$.
945: This means that the excitation of the tachyon is localized around $x=0$
946: as in ref.\cite{baal}\cite{hw}\cite{hn}
947: though the width is relatively broad such as $\delta=1/\sqrt{\beta}$.
948: 
949: A rough behavior of $A(t)$ can be read from the factor
950: $e^{\sqrt{2\beta-k^{2}}t} \sim  e^{\sqrt{2\beta}t}$ in
951: (\ref{staticcfinal1}) or (\ref{deltak1}).
952: This can be inferred from the classical solution to the equation of
953: motion, and a typical time scale is defined as $T^{(I)}\equiv
954: 1/\sqrt{2\beta}$.
955: The formula for $A(t)$ is a bit complicated, but 
956: if we set $L_{p}=100l_{s}$, we have
957: $\sum_{n_{p}=-2}^{2}|N_{n,n_{p}}|^{2} \simeq 0.9987$, so
958: it suffices to have the terms with $n_{p}=0,\pm 1, \pm 2$ in
959: (\ref{delta01}).
960: If we set $y_{0}=0.5 l_{s}$, relatively a large value,
961: we can approximate $\omega_{n_{p},k}\simeq y_{0}$ for the $n_{p}$'s, 
962: and hence $\Delta_{k}(t=0)^{2}\simeq g_{s}/y_{0}$.
963: So, we have
964: \beqa
965: |\tilde{c}_{0,k}|^{2}_{{\rm typical}}\simeq
966: \frac{g_{s}y_{0}}{2}\{
967: \frac{1}{2\beta-k^{2}}\sinh^{2}\sqrt{2\beta-k^{2}}t
968: +\frac{1}{(y_{0})^{2}}\cosh^{2}\sqrt{2\beta-k^{2}}t
969: \}.\label{staticcfinal}
970: \eeqa
971: Since $y_{0}\gg 2\beta$ now, the first term in
972: (\ref{staticcfinal}) is dominant.
973: The time-dependence of $A(t)$'s for $p=1,2,4$ is plotted in Fig.3, for
974: a specific value
975: of the angle $\theta\simeq 2 \beta =0.01$ and $g_{s}=0.01$, which we set
976: on the basis of the brane phenomenology
977: (e.g. refs.\cite{oneangle}\cite{tye1}).
978: \begin{figure}[h]
979: \begin{center}
980: \includegraphics[width=8cm]{fig3}
981: \caption{
982: %
983: %Figure 3: 
984: $\log_{10}A(t)$ vs. $t/T^{(I)}$.The topmost
985: graph is 
986: for D-strings, the middle is for D-2-branes and the lowest is for
987: D-4-branes, with 
988: $\theta\simeq 2\beta=0.01$, $g_{s}=0.01$ and $T^{(I)}=10 l_{s}$.
989: }
990: \end{center}
991: \end{figure}
992: We can read a tendency from (\ref{staticcfinal1}) or Fig.3
993: that the blow-up is more
994: delayed for the case with a larger value of $p$.
995: This fact can be explained as follows: 
996: the tachyonic modes with less momentum tend to blow up more rapidly,
997: but there are larger number of tachyonic modes 
998: for larger momentum.
999: That is, except for the case of D-strings,
1000: for larger value of $p$,
1001: the most ``active'' tachyonic mode which has $k=0$ is 
1002: more smeared in some region of $k$  
1003: and the start of the blow up becomes slower.
1004: 
1005: 
1006: Next we derive $A(t)$ for the case (II). 
1007: The frequency-squared for $\tilde{c}_{0,k} $ is
1008: %in the equations of motion (\ref{eom1}) is
1009: \beqa
1010: \omega_{0,k}(t)^{2}= (z_{0}-v t)^{2} +k_{\tilde{\mu}}^2-2 \beta.
1011: \label{omega3}
1012: \eeqa
1013: We note that each $\omega_{0,k}^{\ \ 2}$ is positive at $t=0$ for 
1014: $z_{0}=0.5 l_{s}$.
1015: The behavior of the system in terms of its wave function is 
1016: as follows: 
1017: As $z(t)$ decreases gradually,
1018: so does each $\omega_{0,k}^{\ \ 2}$ and
1019: the wave function $\Psi(t,\tilde{c}_{0,k})$ of 
1020: $\tilde{c}_{0,k}$ spreads little by little.
1021: At $t= t_{0}\equiv (z_{0}-\sqrt{2\beta})/v$, 
1022: $\omega^{\ \ 2}_{0,0}$ 
1023: (with the lowest momentum $k=0$) becomes zero and 
1024: then negative, i.e.  tachyonic, and the corresponding
1025: wave function begins to spread radically, resulting in
1026: the blow-up of the mode.
1027: After that, the modes with lower $k$ are also getting tachyonic  
1028: and their wave functions also start to spread one after another, 
1029: leading to the 
1030: tachyon condensation.  
1031: 
1032: To obtain $A(t)$ in this case, we have only to find two solutions to
1033: (\ref{eom1}) with (\ref{omega3}) and substitute it for (\ref{formaldelta}).
1034: However, its exact solutions is a bit difficult to deal with,
1035: so we find some approximate solutions:
1036: Let us define
1037: $t_{*}(k)\equiv (z_{0}-\sqrt{2\beta-k^{2} })/v$
1038: which gives
1039: $\omega_{0,k}(t_{*})=0$,
1040: and rewrite the $\omega_{0,k}^{2}$ using $t_{*}$ as
1041: \beqa
1042: \omega_{0,k}(t)^{2}= -\alpha_{k} (t-t_{*})+v^{2}(t-t_{*})^{2}
1043: \label{omega22}
1044: \eeqa
1045: where we denote the coefficient $\alpha_{k} \equiv \sqrt{8 v^{2} \beta'}$
1046: and $\beta'\equiv \beta-k^{2}/2$.
1047: If $t$ approaches $t_{*}$, the second term quadratic in $t-t_{*}$ of
1048: (\ref{omega22}) becomes so small as to be neglected compared to
1049: the first term. (We denote the time as $t=t_{1}$.) 
1050: Thus, for $t>t_{1}$, two Airy functions can be used as basis functions.
1051: On the other hand, taking a closer look will let one notice that 
1052: (\ref{eom1}) with (\ref{omega22}) has the same form  
1053: as the Schrodinger equation 
1054: in one dimension with relatively slowly-changing potential
1055: (if we replace $t$ by $x$). 
1056: So, when $\omega_{0,k}(t)^{\ \ 2}$ is relatively large 
1057: and changing slowly, 
1058: we can apply ``WKB approximation''
1059: to get two approximate basis functions. 
1060: Thus, if this approximation can be applied until the time
1061: $t=t_{1}$, we can follow the time-evolution of 
1062: $\Psi(t,\tilde{c}_{0,k})$ and  $\Delta_{k}(t)$
1063: by using the ``WKB-approximated'' solutions until $t=t_{1}$, and then
1064: letting the Airy functions take over the role of basis functions. 
1065: We take this prescription to evaluate their 
1066: time-evolution.
1067: 
1068: The ``WKB-approximated'' basis functions are
1069: \beqa
1070: \phi_{1}(t)=\omega_{0,k}^{-1/2}\cos(\int^{t}_{t_{1}}dt'
1071: \omega_{0,k}(t')),\nonumber\\
1072: \phi_{2}(t)=\omega_{0,k}^{-1/2}\sin(\int^{t}_{t_{1}}dt'
1073: \omega_{0,k}(t')).\label{basis1}
1074: \eeqa
1075: This approximation is valid  as far as
1076: \beqa
1077: v\ll |\omega_{0,k}|^{2}{\rm \ \ and \ \ } 
1078: |\frac{\partial\omega_{0,k}}{\partial t}|\ll |\omega_{0,k}|^{2}.
1079: \label{cond1} 
1080: \eeqa
1081: On the other hand, the condition
1082: for the Airy functions to be used is
1083: %($2v\sqrt{2 \beta'} |t-t_{*}| \gg v^{2}(t-t_{*})^{2}$) is
1084: \beqa
1085: |t-t_{*}|\ll \frac{\sqrt{8\beta'}}{v}.\label{cond2}
1086: \eeqa
1087: So, if we choose $t_{1}\equiv t_{*}-\sqrt{8\beta'}/10 v$ 
1088: to satisfy (\ref{cond2}) (please note the sign),
1089: the condition to satisfy (\ref{cond1}) is written as
1090: \beqa
1091: v\ll 2\beta'.\label{cond3}
1092: \eeqa
1093: which is compatible with the condition to regard $c_{0}$ and $w$
1094: as only massive fields.
1095: So, in this paper we consider such a system.
1096: One may worry that tachyonic modes with large $k$
1097: do not satisfy (\ref{cond3}) for any set of $\beta$ and $v$.
1098: In fact, in a set-up with $v\ll 2\beta$,
1099: $v$ is so slow that the tachyonic fluctuation $\tilde{c}$
1100: blows up enough much before the tachyon modes with large $k$  
1101: ``wake up''. So, the above point 
1102: is not practically a problem.
1103: 
1104: Using the Airy functions $Ai(\tau)$ and $Bi(\tau)$,
1105: we obtain the width $\Delta_{k}(t)^{2}$ of 
1106: $|\Psi(\tilde{c}_{0,k},t)|^{2}$ as
1107: \beqa
1108: \Delta_{k}(t)^{2}&=&g_{s}[  \frac{\pi^{2}}{\omega_{1}}
1109: \{Bi'(\tau_{1})Ai(\tau)-Ai'(\tau_{1})Bi(\tau)\}^{2}\nonumber\\
1110: & &+\frac{\omega_{1}\pi^{2}}{(8v^{2}\beta')^{1/3}}
1111: \{Bi(\tau_{1})Ai(\tau)-Ai(\tau_{1})Bi(\tau)\}^{2}
1112: ],\label{movingck}
1113: \eeqa
1114: where $Ai'(\tau)=\frac{dAi(\tau)}{d\tau}$, 
1115: $\tau\equiv \alpha^{1/3}_{k} (t-t_{*})$,  
1116: $\omega_{1}\equiv
1117: \omega_{0,k}(t=t_{1})(=\frac{\sqrt{11}}{10}\sqrt{8\beta'})$ 
1118: and 
1119: $\tau_{1}\equiv \alpha^{1/3}_{k} (t_{1}-t_{*})$.
1120: And we can obtain $|\tilde{c}(x^{\mu})|_{{\rm typical}}$
1121: for $p\ge 2$ for case (II) by substituting
1122: (\ref{movingck}) for (\ref{staticcfinal1}).
1123: %\beqa
1124: %|\tilde{c}(x^{\mu})|_{{\rm typical}}^{2}\equiv A(t)^{2} f_{0}(x)^{2} 
1125: %=\frac{\Omega_{p-2} }{(2\pi)^{p-1}}
1126: %\int_{0}^{\sqrt{2\beta}} k^{p-2}dk \frac{\Delta_{k}(t)^{2}}{2}
1127: %\sqrt{\frac{2\beta}{\pi}}e^{-2\beta x^{2}}.\label{movingcfinal}
1128: %\eeqa
1129: For the case of D-strings, $|\tilde{c}(x^{\mu})|=(\Delta_{k=0}(t)f_{0})^{2}/2$
1130: where $\Delta_{k=0}(t)$ is given in (\ref{movingck}).
1131: 
1132: As in the case (I), the blow-up of the tachyonic fluctuation
1133: is localized around $x=0$ with the  width $\delta=1/\sqrt{\beta}$.
1134: A rough behavior of $A(t)$ is $\sim 
1135: \exp [ \frac{2}{3} \{(8v^{2}\beta)^{1/6}(t-t_{*})\}^{3/2} ]$, 
1136: so we can see that a typical time scale of the decay is
1137: $T^{(II)}\equiv 1/(8v^{2}\beta)^{1/6}\cong 1/(4v^{2}\theta)$.
1138: That is, the typical time scale depends more strongly on the relative
1139: velocity of the branes than on the angle $\theta$.
1140: The reason for this is that
1141: after reaching the critical distance, 
1142: the larger the velocity is, the earlier potentially tachyonic modes
1143: ``wake up'' to become truly tachyonic and blow up.
1144: Time-dependent behavior of
1145: $A(t)$ in the specific case of D-4-branes with 
1146: $2\beta=10^{-2}$ and $v=10^{-4}$ is plotted in Fig.4
1147: where $T^{(II)}\simeq 36.84 l_{s}$. 
1148: \begin{figure}[h]
1149: \begin{center}
1150: \includegraphics[width=8cm]{fig4}
1151: \caption{
1152: %
1153: %Figure 4: 
1154: $\log_{10}A(t)$ vs. 
1155: $(t-t_{0})/T^{(II)}$ for D-4-branes; $2\beta=10^{-2},\ v=10^{-4}$ and
1156:  $g_{s}=0.01$.
1157: }
1158: \end{center}
1159: \end{figure}
1160: 
1161: In this way, we have obtained, for both cases, explicit functions
1162: of typical amplitudes of tachyonic fluctuations
1163: $|\tilde{c}(x_{\mu})|_{{\rm typical}}=A(t)f_{0}$.
1164: 
1165: For convenience,
1166: let us define the following quantity for a function of $k$, $g(k)$:
1167: \beqa
1168: <<g(k)>>\equiv \int \frac{d^{p-1}{\bf k}}{(2\pi)^{p-1}}
1169: \frac{\Delta_{k}(t)^{2}}{2} g(k)\label{gk1}
1170: \eeqa
1171: where it is normalized as $<<1>>=A(t)^{2}$.
1172: For a later use, we also define a ``typical amplitude of 
1173: tachyon's wave vector'' of tachyonic fluctuations
1174: as $\sqrt{\overline{k^{2}}}
1175: \equiv (<<k^{2}>>/<<1>>)^{1/2}$.
1176: By $\sqrt{\overline{k^{2}}}$ we can estimate 
1177: which momentum-possessing tachyon modes 
1178: contribute the most to the blow-up, 
1179: since the weight for a tachyon mode with a momentum ${\bf k}$ 
1180: in $|\tilde{c}(x^{\mu})|^{2}$ 
1181: is $(\Delta_{k})^{2}/2$, as in (\ref{typfldef}) with
1182: $|\tilde{c}_{0,k}|_{{\rm typical}}=\Delta_{k}(t)^{2}/2$.
1183: 
1184: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1185: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1186: \section{Time-evolution of the D-p-branes' shapes}
1187: \setcounter{equation}{0}
1188: \setcounter{footnote}{0}
1189: 
1190: In this section we discuss time-evolution  
1191: of the D-p-branes' shapes and show seemingly queer behavior 
1192: of  the brane's shape occurs. 
1193: Then, we  give a physical interpretation of the behavior 
1194: to conclude that it is not actually queer.
1195: 
1196: With (\ref{defctilder}) and (\ref{expansion2}), 
1197: the typical {\it value} of $d(x_{\mu})$ is  
1198: \beqa
1199: d(x_{\mu})_{{\rm typical}} \cong 
1200: -i\frac{\tilde{c}(x_{\mu})_{{\rm typical}} }{\sqrt{2}}=
1201: \frac{-i}{\sqrt{2}}
1202: \int_{k^{2}\le 2\beta}\frac{d^{p-1}{\bf k}}{(2\pi)^{p-1}}
1203: \tilde{c}_{0,k}(t)_{{\rm typical}}
1204: u_{k}(x_{\bar{\mu}})f_{0}(x)
1205: \label{dbehavior}
1206: \eeqa
1207: since $\tilde{d}_{{\rm typical}}$ is negligible.
1208: We note that each $\tilde{c}_{0,k}(t)_{{\rm typical}}$
1209: has its phase as $\tilde{c}_{0,k}(t)_{{\rm typical}}=
1210: |\tilde{c}_{0,k}(t)|_{{\rm typical}}e^{i\delta_{k}}$,
1211: and so does $d(x_{\mu})_{{\rm typical}}$.
1212: We regard this quality as the VEV of the field $d(x_{\mu})$.
1213: Then, the transverse scalar U(2) field $X_{p+1}$
1214: has the VEV of the form: 
1215: \beqa
1216: X_{p+1}= \left(
1217: \begin{array}{cc}
1218: \beta x & d^{*}_{{\rm typical}} \\
1219: d_{{\rm typical}} & -\beta x
1220: \end{array}
1221: \right).\label{xp1}
1222: \eeqa
1223: 
1224: Let us get geometrical information on the branes 
1225: from (\ref{xp1}), first for the case (I). 
1226: In this case there is only one transverse scaler field 
1227: which has non-trivial VEV elements, and, all the scalar fields are 
1228: commutative with each other. 
1229: Thus, as done in such conventional cases and in ref.\cite{hn},
1230: it is a logical step 
1231: to choose the gauge to diagonalize the VEV's of 
1232: the scalar field and interpret the diagonal parts as
1233: the positions of the two branes.
1234: $X_{p+1}$ is diagonalized as
1235: \begin{equation}
1236: \begin{array}{cc}
1237: X^{'}_{p+1}=U_{0} X_{p+1} U_{0}^{-1} 
1238: &=
1239: \left(
1240: \begin{array}{cc}
1241: y(x,t) & 0 \\
1242: 0 & -y(x,t)
1243: \end{array}
1244: \right) 
1245: \end{array}\label{diagonal1}
1246: \end{equation}
1247: where $y=\sqrt{(\beta x)^{2}+|d_{{\rm typical}}|^{2}}$.
1248: $U_{0}$ is the transformation matrix to diagonalize $X_{p+1}$,
1249: whose explicit form we present here for a later use, as
1250: \beqa
1251: \begin{array}{cc}
1252: U_{0} &=\frac{1}{\sqrt{2 y }|d|_{{\rm typical}}}
1253: \left(
1254: \begin{array}{cc}
1255: \sqrt{y+\beta x}|d|_{{\rm typical}}\ & 
1256: \sqrt{y-\beta x}d^{*}_{{\rm typical}} \\
1257: -\sqrt{y-\beta x}d_{{\rm typical}} & 
1258: \sqrt{y+\beta x}|d|_{{\rm typical}}
1259: \end{array}
1260: \right). 
1261: \end{array}\label{u0}
1262: \eeqa  
1263: Since it holds  $|d_{{\rm typical}}|^{2}=
1264: |\tilde{c}(x_{\mu})_{{\rm typical}}|^{2}/2
1265: \cong |\tilde{c}(x_{\mu})|^{2}_{{\rm typical}}/2$\footnote{
1266: In the last equality we use the approximation
1267: $|\tilde{c}_{0k}^{*}(t)_{{\rm typical}}
1268: \tilde{c}_{0k'}(t)_{{\rm typical}}
1269: u_{k}^{*}u_{k'}|\cong
1270: (2\pi)^{p-1} \delta^{(p-1)}(k-k') (\Delta_{k})^{2}/2$
1271: where $u_{k}=e^{ik_{\bar{\mu}}x^{\bar{\mu}}}$ and 
1272: $\Delta_{k}$ is given in (\ref{deltak1}). 
1273: This holds good because only the contributions of
1274: $c_{k}$ and $c_{k'}$ with $k=k'$ has a coherent  phase factor 
1275: while the others do not.} 
1276: which we have obtained in the previous section,
1277: the formula for the shape of one of the recombined branes is
1278: \beqa
1279: y(x,t) =\sqrt{(\beta x)^{2} 
1280: + \frac{A(t)^{2}}{2}\sqrt{\frac{2\beta}{\pi}}e^{-2\beta x^{2}} }.
1281: \label{braneshape1}
1282: \eeqa
1283: This is the same form as obtained for D-strings in ref.\cite{hn}.
1284: However, we can make a precise analysis on the shape of the branes at 
1285: arbitrary $t$ based on (\ref{braneshape1}), 
1286: since we have the explicit function of $A(t)$ for each case
1287: including the overall factor and fine coefficients; 
1288: %(in (\ref{staticcfinal}) or (\ref{sstring})); %or (\ref{movingcfinal}));
1289: We can know explicitly 
1290: when and which applicable condition of the approximations 
1291: (SYM, WKB and some others) breaks.
1292: We will investigate the behavior of the system in detail using
1293: (\ref{braneshape1}) with $A(t)$'s obtained in the previous section.
1294: %(\ref{staticcfinal1}) or (\ref{staticstring1}). 
1295: %or (\ref{movingcfinal}).
1296: 
1297: For the case (II),
1298: its essential difference from the case (I) is that 
1299: there are two scalar fields which have non-trivial 
1300: VEV's of  $X_{p+1}$ and $X_{9}$.
1301: Since the two do not commute each other
1302: after tachyon condensation,
1303: it is probable that one cannot determine 
1304: definite positions of the branes in both of the
1305: dimensions $x_{p+1}$ and $x_{9}$ simultaneously (i.e. uncertainty relation).
1306: Here we assume that we can interpret each of $X_{p+1}$'s 
1307: diagonal elements 
1308: as the position of the two branes in the $x_{p+1}$  direction
1309: if we focus only on the positions in $x_{p+1}$ direction
1310: regardless of  $x_{9}^{(0)}$ direction.
1311: We will return to this problem in the final section.
1312: 
1313: At the initial stage, the shape of each brane is approximately 
1314: one part of hyperbola as in Fig. 5, 
1315: which is consistent with our intuition; 
1316: though $|d|_{{\rm typical}}$ is dependent on $x$ 
1317: through the Gaussian $f_{0}(x)$,
1318: its width $1/\sqrt{\beta}$ is broad and 
1319: its differential coefficient is so small.
1320: \begin{figure}[h]
1321: \begin{center}
1322: \includegraphics[width=8cm]{fig5}
1323: %
1324: \caption{
1325: %Figure 5: 
1326: The shape of D-4-branes;
1327: $\theta\simeq 2\beta=0.01$ and $A(t)=0.01$
1328: }
1329: \end{center}
1330: \end{figure}
1331: The larger $|d|_{{\rm typical}}$ becomes, 
1332: the further the distance of the two D-brane becomes.
1333: That is, $|d|_{{\rm typical}}$
1334: %the VEV of (the absolute value of) 
1335: %the blowing-up tachyon field $d$ 
1336: corresponds directly to such a  
1337: relative motion between the recombined branes as discussed 
1338: in ref.\cite{hn}. 
1339: 
1340: So far, the shape of the p-branes is within the reach of our
1341: intuition. 
1342: After more time passes, however, a seemingly queer behavior
1343: occurs; the shape of each brane deviates from the approximate
1344: hyperbola and comes to have three extremes 
1345: due to localization of tachyon condensation. 
1346: This happens for the cases of D-p-brane for $p\ge 2$,
1347: and is also expected to happen for the case of D-strings,
1348: as we will show below.\footnote{
1349: The appearance of such a shape was already discussed,  
1350: but was {\it denied}
1351: in ref.\cite{hn} in the case of D-strings.}
1352: 
1353: The values of $x$ giving extremes of $y(x,t)$ 
1354: are formally given by
1355: \beqa
1356: x=0,\pm\sqrt{\frac{1}{4\beta}\ln \frac{2 A^4}{\pi\beta}},
1357: \eeqa 
1358: which means that each brane's shape has three extremes
1359: if it holds
1360: \beqa
1361: A>(\pi\beta/2)^{1/4}\equiv A_{{\rm critical}}.\label{multi1}
1362: \eeqa
1363: Thus, all we have to do is to show that the condition 
1364: (\ref{multi1}) is compatible with applicable conditions of 
1365: the approximations we have used.
1366: %The formula (\ref{braneshape1}) is deduced from the action
1367: %second order in the fluctuations,
1368: One  of the important conditions is  
1369: that of the WKB approximation that
1370: higher order terms of the fluctuations in the action 
1371: do not disturb the behavior.
1372: Since only $c$ and $d$ include the tachyonic field,
1373: we estimate the orders of their second order terms $S_{2}$ and 
1374: fourth order terms $S_{4}$ given respectively as
1375: \beqa
1376: S_{2}&\equiv& -T_{p}\int d^{p+1}x 2 i \beta (c^{*}d - c d^{*})\nonumber\\
1377: S_{4}&\equiv& -T_{p}\int d^{p+1}x |c^{*}d-cd^{*}|^{2}\label{ss4}
1378: \eeqa
1379: and show $|S_{4}|\ll |S_{2}|$ below.
1380: 
1381: For D-p-branes with $p\ge 2$,
1382: if we include the non-linear terms in the equation of motion, 
1383: (\ref{expansion2}) (with $\tilde{c}_{0,k}$ replaced 
1384: by its time-dependent typical value)
1385: are not the solutions to the equation of motion any more. However, 
1386: since each of $f_{n}$ and $u_{k}$
1387: forms a complete set,
1388: the expansion is still available
1389: as a type of decomposition of d.o.f..
1390: So, we substitute (\ref{expansion2}) for $S_{2}$ and $S_{4}$
1391: and discuss their amplitudes. 
1392: Neglecting the contribution of non-tachyonic 
1393: (i.e. non-blowing-up) modes,
1394: we have 
1395: \beqa
1396: S_{2}&=&-T_{p}\int dt (-2\beta)A(t)^{2}
1397: \nonumber\\
1398: S_{4}&=&-T_{p}\int dt \sqrt{\frac{\beta}{\pi}} 
1399: \int\frac{d^{p-1}{\bf k_{1}}}{(2\pi)^{p-1}}
1400: \int\frac{d^{p-1}{\bf k_{2}}}{(2\pi)^{p-1}}
1401: \int\frac{d^{p-1}{\bf k_{3}}}{(2\pi)^{p-1}}
1402: \tilde{c}_{0,k_{1}}^{*}\tilde{c}_{0,k_{2}}\tilde{c}_{0,k_{3}}^{*}
1403: \tilde{c}_{0,-k_{1}-k_{2}-k_{3}}.\label{s4}
1404: \eeqa
1405: since $A(t)^{2}\equiv \int d^{p-1}{\bf k}/(2\pi)^{p-1} 
1406: <|\tilde{c}_{0,k}|^{2}>$.
1407: The factor $\sqrt{\beta/\pi}$ comes from extra 
1408: normalization constants of $f_{0}$'s. 
1409: To carry out the integrations with respect to momenta in (\ref{s4}) 
1410: is difficult, so we make order estimation of $S_{4}$,
1411: which will later prove to suffice here;
1412: first we estimate a typical value of $k=\sqrt{k_{\bar{\mu}}^{2}}$
1413: in condensation of $\tilde{c}(x_{\mu})$ using 
1414: $k_{{\rm av}}=\sqrt{\bar{k^{2}}}$ defined at the last part of section 2;
1415: it is about $30 \sim 5$ percent of $\sqrt{k^{2}_{{\rm
1416: max}}}=\sqrt{2\beta}$ and
1417: monotonically decreasing during $t\ge 10 T^{(I)}$ in the case (I)
1418: or $t\ge 10 T^{(II)}$ in the case (II).
1419: Next, we choose some appropriate value of $k_{{\rm av}}$ and 
1420: approximate $\tilde{c}_{0,k_{i}}$ for $i=1,2,3 $ in (\ref{s4})  
1421: to $|\tilde{c}_{0,k_{av}}|$,
1422: defining $A_{{\rm av}}(t)^{2}\equiv \int d^{p-1}{\bf k}/(2\pi)^{p-1} 
1423: <|\tilde{c}_{0,k_{{\rm av}}}|^{2}>$. 
1424: Then, we can show that the behavior of $A(t)_{{\rm av}}$ is 
1425: not so different from that of $A(t)$, and 
1426: we estimate  $S_{4}$ as
1427: \beqa
1428: S_{4}\sim-T_{p}\int dt  
1429: \frac{\Omega_{p-2}(2\beta)^{p/2}}{(2\pi)^{p-1/2}}
1430: (A_{{\rm av}}(t))^{4}.
1431: \eeqa 
1432: Thus, the condition for $|S_{4}|\ll |S_{2}|$ is
1433: \beqa
1434: A(t)\ll \sqrt{\frac{(2\pi)^{p-1/2}}
1435: {\Omega_{p-2}(2\beta)^{(p-2)/2} }}.
1436: \label{wkbdp}
1437: \eeqa
1438: We can easily see that for a small angle $\theta\simeq 2\beta$ and $p\ge 2$,  
1439: the inequality (\ref{wkbdp}) is sufficiently
1440: compatible with (\ref{multi1}).
1441: The other conditions, especially that of SYM that the displacement of
1442: branes should be 
1443: smaller than $l_{s}$ ($A(t)< l_{s}$) and D-branes should have 
1444: some low velocity ($\dot{y}\ll 1(=c)$ at all the value of $x$),
1445: are also shown easily to be satisfied.  
1446: Therefore, the shape of the recombined D-p-brane for $p\ge 2$ 
1447: surely come to have three extremes.
1448: An example is drown in Fig.6.
1449: This happens essentially due to the localization of tachyons 
1450: around $x$,
1451: since the factor $e^{-2\beta x^{2}}$ in (\ref{braneshape1}) is directly
1452: responsible for the multiple extremes.
1453: \begin{figure}[h] 
1454: \begin{center}
1455: \includegraphics[width=8cm]{fig6}
1456: %
1457: \caption{
1458: %Figure 6: 
1459: The shape of a D-brane with multiple extremes;
1460: $2\beta=0.01$ and $A(t)=0.13$
1461: }
1462: \end{center}
1463: \end{figure}
1464: 
1465: For the case of D-strings,
1466: the case is a bit more subtle, but we argue that 
1467: the same phenomenon is expected to happen. 
1468: In this case 
1469: $S_{2}$  and $S_{4}$ are written as
1470: \beqa
1471: S_{2}&=&-T_{D1}\int dt (-2\beta)|\tilde{c}_{0}(t)|^{2}_{{\rm typical}}
1472: \nonumber\\
1473: S_{4}&=&-T_{D1}\int dt \sqrt{\frac{\beta}{\pi}}
1474: |\tilde{c}_{0}(t)|^{4}_{{\rm typical}}.
1475: \label{s41}
1476: \eeqa
1477: If we substitute $\tilde{c}_{0}(t)=A(t)$
1478: for the condition $|S_{4}|< |S_{2}|$, we have
1479: \beqa
1480: A(t) < (4\pi \beta)^{1/4}.\label{wkbcond2}
1481: \eeqa
1482: If one consider the equation of motion for $\tilde{c}_{0}(t)$,
1483: \beqa
1484: \frac{d^{2} \tilde{c}_{0}(t)}{dt^{2}}= 2\beta \tilde{c}_{0}(t) -2
1485: |\tilde{c}_{0}(t)|^{2}\tilde{c}_{0}(t),\label{eomhi}
1486: \eeqa
1487: (\ref{wkbcond2}) is replaced by $A(t) < (\pi \beta)^{1/4}$.
1488: Anyway, there is, a little though , a range 
1489: for $A(t)$ to be allowed to have three extremes
1490: within second order approximation: 
1491: $(\pi \beta/2)^{1/4}< A(t) < (\pi \beta)^{1/4}$.
1492: This may seem subtle, but if we solve numerically the 
1493: equation of motion for $\tilde{c}_{0}(t)$
1494: we can see that $|\tilde{c}_{0}(t)|$ really develop to exceed 
1495: $A_{{\rm critical}}$ in (\ref{multi1}).
1496: The other conditions of the approximations are again
1497: shown to be satisfied. 
1498: Thus, it is probable that
1499: the shape of the D-string also comes to have three 
1500: extremes. 
1501: 
1502: 
1503: The physical interpretation of this seemingly queer behavior  is 
1504: as follows;
1505: the energy released via tachyon condensation 
1506: pushes the recombined branes away from each other, but
1507: it is given only to the local part of the branes around $x=0$
1508: due to localization of tachyon condensation, so, only 
1509: the part is much accelerated.
1510: Though the D-branes have a large tension, 
1511: they also have a large inertia, and 
1512: when the given energy of the local part is large enough,
1513: the branes extend, surpassing the tension, to form three extremes.
1514: That is, {\it localization of tachyon condensation or that of  
1515: the released energy, and the (large) inertia causes the shapes of branes 
1516: to have three extremes.}
1517: 
1518: The recombined picture has been obtained, and that the seemingly
1519: queer behavior can also be explained from 
1520: physical point of view. Thus, 
1521: we can say that 
1522: each diagonal element of $X_{p+1}$ describes the shape of each D-brane,
1523: and that
1524: in the gauge where 
1525: $X_{p+1}$ is diagonalized (which we denote as ``physical gauge''),
1526: the correspondence of degrees of freedom 
1527: is direct between open strings and
1528: the matrices $X_{i}$ and $A_{\mu}$;
1529: for example, (1,1) component of $A_{\mu}$ corresponds to excitations of 
1530: open strings whose both ends are on a first D-brane, etc..  
1531: 
1532: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1533: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1534: \section{Creation of open string pairs} 
1535: \setcounter{equation}{0}
1536: \setcounter{footnote}{0}
1537: 
1538: Let us next discuss creation of open string pairs.
1539: We mainly concentrate on case (I) in this section:
1540: We first discuss the case of D-strings, and then
1541: that of D-p-branes with $p \ge 2$.
1542: 
1543: The blow-up of the gauge field appears in the off-diagonal elements of 
1544: $A_{p}$ (in this case of course $p=1$) and a typical value of 
1545: its electric flux is
1546: \beqa
1547: \begin{array}{ccc}
1548: F_{0p}(x_{\mu})_{{\rm typical}}&
1549: =\left(
1550: \begin{array}{cc}
1551: 0 &  \partial_{0} c^{*}_{{\rm typical}}\\
1552: \partial_{0} c_{{\rm typical}} & 0
1553: \end{array}
1554: \right) 
1555: &\cong \frac{1}{\sqrt{2}}
1556: \left(
1557: \begin{array}{cc}
1558: 0 &  \dot{A}(t)f_{0}(x)e^{-i\delta}\\
1559: \dot{A}(t)f_{0}(x)e^{i\delta} & 0
1560: \end{array}
1561: \right) 
1562: \end{array}
1563: \eeqa
1564: where a dot represents a time-derivative and $\delta$ is a constant
1565: phase of $c_{{\rm typical}}$ (and $\tilde{c}_{{\rm typical}}$).
1566: This flux has a non-vanishing contribution to the energy density 
1567: ${\cal H}_{{\rm ele}}(x)$ as
1568: \beqa
1569: {\cal H}_{{\rm ele}}(x)=\frac{1}{2g_{s}}(\dot{A})^{2}(f_{0})^{2},
1570: \eeqa
1571: so this excitation should have some physical meaning.
1572: What is the physical meaning of the blow-up? 
1573: We argue that it means that {\it open string pairs (each composed of 
1574: two strings with opposite orientations) connecting the 
1575: two recombined D-branes  are created}.
1576: We will explain the reason in the following some paragraphs.
1577: 
1578: First, we show that this blow-up correspond to 
1579: excitations of open strings connecting the two branes.
1580: Here we would like to confirm some basic facts about SYM as low energy
1581: effective theory (LEET) of D-branes: 
1582: As discussed in \cite{witten}, each D-brane  (${\rm Dp}_{I}$ for I=1,2) 
1583: has U(1) symmetry
1584: (${\rm U(1)}_{I}$), and open strings starting at ${\rm Dp}_{1}$ and 
1585: ending at ${\rm Dp}_{2}$ (which we denote as 1$\to$2 strings) 
1586: have ${\rm U(1)}_{1}\times {\rm U(1)}_{2}$ charges (-1,1), and 
1587: 2$\to$1 strings have charges (1,-1). 
1588: In the U(2) SYM plus adjoint Higgs theory as LEET of branes, 
1589: effective fields corresponding to the strings also have such charges, and
1590: if degrees of freedom (d.o.f.) of ${\rm Dp}_{I}$ 
1591: correspond to (I,I)-components of
1592: the U(2) matrices for I=1,2 (as in the case of the physical gauge),
1593: one can represent
1594: ${\rm U(1)}_{1}$ and ${\rm U(1)}_{2}$ transformation as 
1595: $U_{1}={\rm diag}(e^{i\alpha_{1}},1)$ and $U_{2}={\rm diag.}
1596: (1,e^{i\alpha_{2}})$ 
1597: for parameters $\alpha_{1}$ and $\alpha_{2}$.
1598: Suppose we use a unified symbol of the U(2) field $\hat{A}_{M}$, 
1599: which is equal to 
1600: $A_{\mu}$ for $M=\mu$ (=$0,1,\cdots,p$) and is equal to  
1601: $X_{i}$ for $M=i$ (=$p+1,\cdots,9$). 
1602: Then, $\hat{A}_{M}$ transforms under  
1603: the ${\rm U(1)}_{1}$ and ${\rm U(1)}_{2}$ with constant 
1604: $\alpha_{I}$'s as
1605: \beqa
1606: \begin{array}{cc}
1607: \hat{A}_{M}
1608: =\left(
1609: \begin{array}{cc}
1610: \hat{a}_{M} &  \hat{b}_{M} \\
1611: \hat{c}_{M}  & \hat{d}_{M} 
1612: \end{array}
1613: \right) 
1614: &\to  \hat{A}_{M}^{'}=
1615: \left(
1616: \begin{array}{cc}
1617: \hat{a}_{M} &  \hat{b}_{M}e^{i(\alpha_{1}-\alpha_{2})} \\
1618: \hat{c}_{M} e^{-i(\alpha_{1}-\alpha_{2})} & \hat{d}_{M}
1619: \end{array}
1620: \right),
1621: \end{array}\label{am1}
1622: \eeqa
1623: so, d.o.f. of $\hat{c}_{M}$ correspond to those of $1\to 2$ strings and 
1624: d.o.f. of $\hat{b}_{M}$ to those of  $2\to 1$ strings.
1625: 
1626: In our case it is when $X_{p+1}$ is diagonalized 
1627: (i.e. in the physical gauge) that 
1628: d.o.f. of ${\rm Dp}_{I}$ correspond to (I,I)-components of
1629: the U(2), so we have to see the gauge field in the physical gauge.
1630: Using $U_{0}$ in (\ref{u0}) (and abbreviating $c_{{\rm typical}}$ to $c$
1631: here),  we have the field in the gauge 
1632: $\hat{A}_{M}^{'}=U_{0}\hat{A}_{M}U_{0}^{-1}
1633: +i U_{0}\partial_{M}U_{0}^{-1}$ as
1634: \beqa
1635: \begin{array}{ccc}
1636: A_{0}^{'} & =&\frac{\beta x}{4y^{2}}
1637: \frac{\partial_{0}c^{*} %_{{\rm typical}}
1638:  c %_{{\rm typical}}
1639: +c^{*} %_{{\rm typical}}
1640: \partial_{0}c} %_{{\rm typical}}}
1641:  {|c%_{{\rm typical}}
1642: |^{2}}
1643: \left(
1644: \begin{array}{cc}
1645: 0 &c^{*}%_{{\rm typical}}
1646: \\
1647:  c%_{{\rm typical}}
1648: & 0
1649: \end{array} 
1650: \right)\\
1651: A_{p}^{'} &
1652: =&\{1-\frac{\beta+2(\beta x)^{2}}{2y^{2} }\}
1653: \left(
1654: \begin{array}{cc}
1655: 0 &c^{*}%_{{\rm typical}}
1656: \\
1657: c%_{{\rm typical}}
1658: & 0
1659: \end{array} 
1660: \right)\\
1661: X_{p+1}^{'}& =&
1662: \left(
1663: \begin{array}{cc}
1664: y & 0\\
1665: 0 & -y
1666: \end{array}
1667: \right),
1668: \end{array}
1669: \label{newgaugeff}
1670: \eeqa
1671: %in addition to $X_{p+1}^{'}={\rm diag.}(y,-y)$,
1672: and the electric flux as
1673: \beqa
1674: \begin{array}{ccc}
1675: F_{0p}^{'}(x_{\mu})%_{{\rm typical}}
1676: &
1677: =U_{0}F_{0p}(x_{\mu})%_{{\rm typical}}
1678: U_{0}^{-1}=
1679: &\left(
1680: \begin{array}{cc}
1681: 0 & \partial_{0} c^{*}\\
1682: \partial_{0} c & 0
1683: \end{array}
1684: \right).\label{eflux1} 
1685: \end{array}
1686: \eeqa
1687: That is, the flux is invariant under the transformation via $U_{0}$; 
1688: the blow-up arises only in its off-diagonal parts and hence, 
1689: according to the above basic facts, 
1690: the blow-up of the electric flux is due to excitations
1691: of 1$\to$2 and 2$\to$1 strings.   
1692: 
1693: Then, how is the ratio of 1$\to$2 and 2$\to$1 strings?
1694: We can see that via charge densities $j_{0}^{(I)}$ (for I=1,2)
1695: of ${\rm U(1)}_{I}$ current $j_{\mu}^{(I)}$;
1696: the above basic facts means that 
1697: $j_{0}^{(1)}$ measures 
1698: the number density of endpoints of open strings incoming to 
1699: ${\rm Dp}_{1}$
1700: minus that of strings outgoing from ${\rm Dp}_{1}$
1701: (if it is appropriately normalized), and $j_{0}^{(2)}$ is equal to
1702: $-j_{0}^{(1)}$ in this setting. 
1703: In the notation of $\hat{A}_{M}$ in (\ref{am1}), 
1704: $j_{0}^{(1)}$ is written via Noether's theorem as
1705: \beqa
1706: j_{0}^{(1)}(=-j_{0}^{(2)})=&i\{ 
1707: (\hat{b}_{M}\partial_{0}\hat{c}_{M}-\hat{c}_{M}\partial_{0}\hat{b}_{M})
1708: -(\hat{b}_{M}\partial_{M}\hat{c}_{0}-\hat{c}_{M}\partial_{M}\hat{b}_{0})\}\nonumber\\
1709: &-2(\hat{a}_{0}-\hat{d}_{0})\hat{b}_{M}\hat{c}_{M}
1710: +(\hat{a}_{M}-\hat{d}_{M})(\hat{b}_{0}\hat{c}_{M}
1711: +\hat{b}_{M}\hat{c}_{0}).\label{j0}
1712: \eeqa
1713: Substituting the solution (\ref{newgaugeff}) for (\ref{j0}),
1714: $j_{0}^{(1)}$ is shown to vanish, which means that
1715: the number density of endpoints of open strings
1716: incoming to ${\rm Dp}_{1}$ at the point $x$ is equal to that
1717: of strings outgoing from ${\rm Dp}_{1}$ at $x$.
1718: This is also the case with $j_{0}^{(2)}$ and ${\rm Dp}_{2}$.
1719: That is, open string pairs connecting the two decaying D-p-branes 
1720: are created, each pair of which is composed of 1$\to$2 and 2$\to $1
1721: strings! 
1722: \footnote{We 
1723: can also look into the configuration of fundamental strings in our solution 
1724: (\ref{newgaugeff}) via electric source current $J_{0M}(x_{N})$ of
1725: spacetime NSNS field $B_{MN}$ by
1726: using the results of ref.\cite{raams}, as done in ref.\cite{khw}.  
1727: Then, we can show that $J_{0M}$ vanishes in this case, which is 
1728: consistent with $j_{0}^{(1)}=j_{0}^{(2)}=0$ and our interpretation 
1729: of it.}
1730: 
1731: Furthermore, excitations of the scalar and gauge fields in
1732: (\ref{newgaugeff}) also
1733: includes other 
1734: components which correspond to such open string pairs.\footnote{
1735: I thank K.Hashimoto for letting me notice these extra contributions.} 
1736: We can extract
1737: the components by 
1738: writing all the excitations
1739: in terms of U(2) covariant forms in the physical gauge, as
1740: \beqa
1741: D'_{0}X_{p+1}'&=& U_{0} D_{0}X_{p+1}U_{0}^{-1}
1742: %=\frac{1}{2 y}\left(
1743: %\begin{array}{cc}
1744: %\partial_{0}|d|^2 \& 2 \beta x \partial_{0}d^{*} \\
1745: %2 \beta x \partial_{0} d \& - \partial_{0}|d|^2
1746: %\end{array}
1747: %\right)
1748: =\left(
1749: \begin{array}{cc}
1750: \partial_{0}y & \beta x \partial_{0}d^{*}/y \\
1751: \beta x \partial_{0} d/y & - \partial_{0}y
1752: \end{array}\right),%\label{dx0}
1753: \nonumber\\
1754: D'_{p}X_{p+1}'&=&U_{0} D_{p}X_{p+1}U_{0}^{-1}=
1755: %\frac{\beta -2|d|^{2}}{ y}\left(
1756: %\begin{array}{cc}
1757: %\beta x \& -d^{*} \\
1758: %-d \& -\beta x 
1759: %\end{array}
1760: %\right)=
1761: \left(
1762: \begin{array}{cc}
1763: \partial_{x}y & -(\beta -2|d|^{2})d^{*}/y \\
1764: -(\beta -2|d|^{2})d/y & -\partial_{x}y
1765: \end{array}
1766: \right).\label{dx1}
1767: \eeqa
1768: By T-dualizing with respect to $x_{p+1}$ direction, 
1769: $D'_{0}X_{p+1}'$ and $D'_{p}X_{p+1}'$ correspond respectively to
1770: the electric and magnetic fluxes $F'_{0,p+1}$ and $F'_{p,p+1}$. 
1771: Thus, based on the same reason as 
1772: the case of the electric flux $F'_{0p}$,
1773: we argue that only diagonal elements in (\ref{dx1}) correspond
1774: to d.o.f. of D-1-branes and the remaining
1775: off-diagonal components in (\ref{dx1}) correspond
1776: to d.o.f. of open strings connecting the branes. 
1777: Consistency check of this argument is given in the following way:
1778: Non-vanishing part of the total energy density of th system is 
1779: \beqa
1780: {\cal H}_{{\rm total}}(x)=\frac{T_{D1}}{2}Tr\{1_{2\times 
1781: 2}+\frac{1}{2}(D'_{p}X_{p+1}')^{2}+\frac{1}{2}(D'_{0}X_{p+1}')^{2}
1782: +\frac{1}{2}(F'_{0p})^{2}\},\label{etotal}
1783: \eeqa
1784: and contribution of the diagonal elements of (\ref{dx1}) to 
1785: ${\cal H}_{{\rm total}}(x)$ is 
1786: \beqa
1787: {\cal H}_{{\rm diag}}(x)=T_{{\rm D1}}\{2+2\cdot
1788: \frac{1}{2}(\partial_{x}y)^{2}
1789: %\frac{(\beta-2|d|^{2})^{2}}{y^{2}}(\beta x)^{2} 
1790: +2\cdot \frac{1}{2} (\partial_{0}y)^{2}
1791: %\frac{|d|^{2}|\partial_{0}d|^{2}}{y^{2}} 
1792: +0\}.\label{ediag}
1793: \eeqa
1794: while that of the off-diagonal elements of (\ref{dx1}) and $F'_{0p}$
1795: is
1796: \beqa
1797: {\cal H}_{{\rm off-diag}}=T_{{\rm D1}}\{
1798: (\beta-2|d|^{2})^{2}|d|^{2}/y^{2}
1799: +(\beta x)^{2}|\partial_{0}d|^{2}/y^{2}+|\partial_{0}c|^{2}\}.
1800: \label{eoffdiag}
1801: \eeqa
1802: %(We note that diagonal elements of $F'_{0p}$ are zero in this case.)
1803: The first and second term in (\ref{ediag}) are rewritten within second
1804: order approximation 
1805: as $2\sqrt{1+(\partial_{x}y)^{2}}$, the energy density due to the
1806: D-strings' existence 
1807: (tension), and the third term
1808: is of course the density of the D-strings' kinetic energy. That is, only
1809: ${\cal H}_{{\rm 
1810: diag}}(x)$ 
1811: just gives the energy density that the D-strings themselves have.    
1812: Thus, it is plausible that the remaining energy 
1813: ${\cal H}_{{\rm off-diag}}$ is the energy due to 
1814: interaction of D-branes, which is consistent with our argument that
1815: open string pairs connecting the D-branes have the rest of 
1816: the energy.\footnote{Further confirmation of our argument
1817: that open string pair creation are created between the D-branes
1818: will be given in ref.\cite{hashi}.}
1819: 
1820: 
1821: We can also estimate roughly 
1822: the number density of the created open strings;
1823: since SYM describes the behavior of low energy modes or almost zero
1824: modes of open strings, the strings connecting the two D-branes
1825: should be nearly straight. 
1826: Since the tension, or 
1827: the energy density of fundamental strings 
1828: is constant (=1 in our notation), the number density of 
1829: the open strings (per unit length) 
1830: $n(x,t)$ is equivalent to
1831: the energy density divided by the displacement of the two D-p-branes 
1832: in $x_{p+1}$ direction.\footnote{In fact, open strings have momenta
1833: since the off-diagonal elements to momentum density is non-vanishing. 
1834: Here, we ignore the contribution of momenta as an approximation
1835: and and estimate the number density.} 
1836: Thus, it is obtained as 
1837: \beqa
1838: n(x,t)=\frac{{\cal H}_{{\rm off-diag}}(x,t)}{2y(x,t)}
1839: =\frac{|d|^{2}}{2 y^{3}}\{(\beta-2|d|^{2})^{2}
1840: +2\cdot 2\beta(\beta x)^{2}+2\beta |d|^{2}\}
1841: %=\frac{{\cal H}_{{\rm total}}-{\cal H}_{diag}}{2y(x,t)}
1842: %=\frac{(\dot{A}f_{0})^{2}}{4g_{s}
1843: %\sqrt{(\beta x)^{2}+(A(t)f_{0}(x))^{2}/2}}
1844: \label{numb1}
1845: \eeqa
1846: where we have used $\partial_{0}d =\sqrt{2\beta}d$
1847: and  $\partial_{0}c =\sqrt{2\beta}c$.
1848: A distribution of $n(x,t)$ for $g_{s}=10^{-4}$ and 
1849: $A(t)=0.1$ is given in Fig.7,
1850: and the situation is roughly depicted in Fig.8.
1851: \begin{figure}[h]
1852: \begin{center}
1853: \includegraphics[width=8cm]{fig7}\\
1854: %Figure 7
1855: \caption{Number density of open strings for $A(t)=0.1$ with $g_{s}=10^{-4}$}
1856: \end{center}
1857: \end {figure}
1858: \begin{figure}[h]
1859: \begin{center}
1860: \includegraphics[width=12cm]{fig8}\\
1861: %Figure 8
1862: \caption{A rough picture how open string pairs are created}
1863: \end{center}
1864: \end{figure}
1865: 
1866: In fact, a total number of the fundamental string in Fig.7 is
1867: $N\equiv \int dx n\cong 50.14$, i.e. not an even integer. 
1868: This is due to our partially incomplete analysis,
1869: in which we do not consider the effect of quantization of charges 
1870: or fluxes. Actually, $n(x,t)$ is an
1871: expectation value 
1872: or a probability of distribution of number density 
1873: for open string to be created, and the real process
1874: should happen in a quantum theoretic way; each pair will be created
1875: rather ``suddenly''  with the above probability.
1876: 
1877: 
1878: Before moving on to the case of D-p-branes with $p\ge 2$,   
1879: we discuss distribution of the released energy.
1880: The energy released by the shortening of D-branes is 
1881: \beqa
1882: E_{{\rm released}}&=&T_{D1}\int dx \  2
1883: \sqrt{1+(\partial_{x}y)^{2}}|_{t=0}
1884: -T_{D1}\int dx \ 2 \sqrt{1+(\partial_{x}y)^{2}}|_{t}\nonumber\\
1885: &\simeq & T_{D1}\int dx \{\beta^{2}-(\beta-2|d|^{2})^{2}(\beta x)^{2}
1886: /y^{2} \},
1887: \label{ereleased}
1888: \eeqa
1889: and it is used partly to accelerate
1890: D-branes and partly to create the open string pairs.
1891: Then, how much is the ratio of the energy distribution?
1892: D-branes' kinetic energy $E_{D1\ {\rm kinetic}}$ and the energy used 
1893: to create open string pairs $E_{{\rm open} }$ are given respectively 
1894: as
1895: \beqa
1896: E_{D1\ {\rm kinetic}}=T_{D1}\int dx
1897: (\partial_{0}y)^{2}\label{ekinetic}, \ \ 
1898: E_{{\rm open} } =T_{D1} \int dx 
1899: {\cal H}_{{\rm off-diag}}
1900: %\{
1901: %(\beta-2|d|^{2})^{2}|d|^{2}/y^{2}
1902: %+(\beta x)^{2}|\partial_{0}d|^{2}/y^{2}+|\partial_{0}c|^{2}\}.
1903: \label{eopen}
1904: \eeqa
1905: (Of course it holds $E_{{\rm released}}=E_{D1\ {\rm kinetic}}
1906: +E_{{\rm open}}$.)
1907: Since we know all the quantities in the formulas of the energies,
1908: we can calculate the ratios    
1909: $E_{D1\ {\rm kinetic}}/E_{{\rm released}}$ and  
1910: $E_{{\rm open} }/E_{{\rm released}}$ at the time $t$ numerically.
1911: The result is 
1912: \beqa
1913: \left\{
1914: \begin{array}{ccc}
1915: E_{D1\ {\rm kinetic}}/E_{{\rm released}}&< 10 \% &\nonumber\\
1916: E_{{\rm open} }/E_{{\rm released}}&> 90 \% & {\rm for}\  A(t)<0.1 
1917: \end{array}\right.
1918: \eeqa
1919: as in Fig.9. and  Fig.10.
1920: %\begin{minipage}{7cm}
1921: \begin{figure}[h]
1922: \begin{center}
1923: \includegraphics[width=8cm]{fig9}
1924: \caption{$E_{D1\ {\rm kinetic}}/E_{{\rm released}}$ vs. $t(\times l_{s})$}
1925: \end{center}
1926: \end{figure}
1927: %\end{minipage}
1928: %\begin{minipage}{7cm}
1929: \begin{figure}[h]
1930: \begin{center}
1931: \includegraphics[width=8cm]{fig10}
1932: \caption{$E_{{\rm open} }/E_{{\rm released}}$ vs. $t(\times l_{s})$}
1933: \end{center}
1934: \end{figure}
1935: %\end{minipage}
1936: 
1937: \noindent
1938: That is, {\it most part of the released energy is used to create the open
1939: string pairs connecting 
1940: the branes}; only a little part of it is used to accelerate the D-branes.
1941: Thus, from the viewpoint of D-branes' motion,
1942: dissipation of the energy of the branes is very large at least
1943: at the initial stage of decay. We note that the order of  $E_{{\rm open}}$
1944: in terms of $g_{s}$ is ${\cal O} ((g_{s})^{0})$, and asymptotic behavior
1945: of $E_{{\rm open}}$ is $E_{{\rm open}}\propto 
1946: e^{2\sqrt{2 \beta}t}$ or $e^{\sqrt{2 \beta}t}$ (its exponent is
1947:  decreasing along with time) for $A(t)<0.1$.
1948: 
1949: 
1950: 
1951: Next we study the case of D-p-branes with $p\ge 2$.
1952: The main difference is that they 
1953: have extra world-volume coordinates $x_{\bar{\mu}}$, and hence 
1954: the condensing tachyon field $\tilde{c}$
1955: (and $U_{0}$) can have dependence of $x_{\bar{\mu}}$.
1956: Then, in addition to $A_{0}^{'}$, $A_{p}^{'}$ and $X'_{p+1}$  
1957: in (\ref{newgaugeff}), 
1958: $A_{\bar{\mu}}$ ($\bar{\mu}=1,\cdots,p-1$) arises through 
1959: the transformation 
1960: via $U_{0}$, and both U(2) electric and  
1961: magnetic flux have non-vanishing contribution to the energy.
1962: A naive form of $A_{\bar{\mu}}$ is a bit complicated, but 
1963: using the approximation in the footnote under (\ref{u0}),
1964: we have 
1965: \beqa
1966: A^{'}_{\bar{\mu}}\cong\frac{1}{2y}
1967: \left(
1968: \begin{array}{cc}
1969: 0 & \partial_{\bar{\mu}}c^{*}\\%_{{\rm typical}}\\
1970: \partial_{\bar{\mu}}c
1971: %_{{\rm typical}} 
1972: & 0
1973: \end{array}
1974: \right).\label{amubar}
1975: \eeqa
1976: Non-vanishing fluxes and covariant derivatives of $X$'s are
1977: \beqa
1978: F_{\bar{\mu}p}^{'}(x_{\mu}) %_{{\rm typical}}
1979: &=&\left(
1980: \begin{array}{cc}
1981: 0 &  \partial_{\bar{\mu}}c^{*} %_{{\rm typical}}
1982: \\
1983: \partial_{\bar{\mu}}c %_{{\rm typical}} 
1984: & 0
1985: \end{array}
1986: \right)\\
1987: D'_{\bar{\mu}}X'_{p+1}&=&
1988: \left(\begin{array}{cc}
1989: \partial_{\bar{\mu}}y & \beta x \partial_{\bar{\mu}}d^{*}/y\\
1990: \beta x \partial_{\bar{\mu}}d/y & -\partial_{\bar{\mu}}y
1991: \end{array}
1992: \right),
1993: \label{mflux1} 
1994: \eeqa
1995: in addition to (\ref{eflux1}) and (\ref{dx1}). 
1996: It should be stressed  that the above is U(2) magnetic flux but 
1997: not U(1)'s one on each of the branes.
1998: Thus, even if we consider the case of D-3-branes, its existence does not 
1999: mean that D-strings are created. 
2000: Instead, according to the basic facts stated above, 
2001: these d.o.f. again correspond to the d.o.f. of fundamental strings,
2002: and $j_{0}$ is again shown to vanish by substituting (\ref{newgaugeff}) 
2003: and (\ref{amubar}) (with $x_{\bar{\mu}}$ dependence 
2004: taken into account). 
2005: Therefore, we conclude that  
2006: in the case of recombination of D-p-branes with $p\ge 2$,
2007: open string pairs connecting the two D-branes are created 
2008: while they are decaying.
2009: 
2010: Let us consider the number density of the string.
2011: In the case with $p\ge 2$,
2012: $|\partial_{0}d|^{2}$ and $|\partial_{0}c|^{2}$ in ${\cal H}_{{\rm off-diag}}$
2013: (in (\ref{eoffdiag})) in the case of D-strings are replaced by 
2014: $|\partial_{\hat{\mu}}d|$ and $|\partial_{\hat{\mu}}c|$ 
2015: where $\hat{\mu}=0,\cdots,p-1$. But then, if we use $<<g(k)>>$ defined in
2016: (\ref{gk1}), it holds 
2017: $|\partial_{0}d|^{2}\simeq<<2\beta-k^{2}>>(f_{0})^{2}/2$ and 
2018: $|\partial_{\bar{\mu}}d|^{2}\simeq <<k^{2}>>(f_{0})^{2}/2$, so
2019: $|\partial_{\hat{\mu}}d|$ gives  $2\beta |d|^{2}$ and
2020: the number density formula (\ref{numb1}) holds again,
2021: though in this case it is a density in a p-dimensional space. 
2022: 
2023: The released energy $E_{r}(t)$ is
2024: \beqa
2025: E_{r}(t)=-T_{Dp}\int d^{p}x\ [2\sqrt{1+(\partial_{x}y)^{2}
2026: +(\partial_{\bar{\mu}}y)^{2}}\ ]^{t}_{t=0}.
2027: \eeqa
2028: Since an extra minus term $-(\partial_{\bar{\mu}}y)^{2}|_{t}
2029: =-<<k^{2}>>(f_{0})^{2}|d|^{2}/y^{2}$ is added
2030: to $E_{r}(t)$, the density part of the released energy is a bit reduced
2031: from that in the formula (\ref{ereleased}), but the density of energy
2032: used to create open string pairs has the same form as that in
2033: (\ref{numb1}). (The density of D-p-branes' kinetic energy is reduced in
2034:  terms of $A(t)$.) Thus, in the case of D-p-branes with 
2035: $p \ge 2$, {\it the ratio of the energy used to create open
2036:  string pairs to the released energy is more than that in the case of
2037:  D-strings}, and we can say again that most part of the released energy 
2038:   is used to create the open string pairs.
2039:  
2040:  
2041: 
2042: 
2043: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2044: \section{Creation of closed strings}
2045: \setcounter{equation}{0}
2046: \setcounter{footnote}{0}
2047: 
2048: %In this section we go out of the framework of SYM and
2049: %show that closed strings as gravitational wave are also 
2050: %created from the D-branes after recombination.
2051: %In addition, we speculate about the 
2052: %behavior of the system beyond SYM.
2053: 
2054: Let us assume $g_{s}\ll 1$. Then, the branes can be regarded as 
2055: heavy objects, and after recombination, 
2056: they start to move away from each other, though  very slowly,
2057: but {\it with some acceleration} in opposite directions. Thus, 
2058: one can easily expect that the system has non-trivial time-evolution
2059: behavior of mass quadrupole moments, which causes gravitational 
2060: radiation. Below, going out of the framework of SYM,
2061: we show that the above expectation is true, and 
2062: evaluate its energy flux, based on the time-evolution of 
2063: the D-branes' shapes. In addition, we discuss creation of
2064: closed strings for configurations of D-branes beyond the reach of SYM. 
2065: In this section we mainly discuss the case (I) again, 
2066: but we expect that essential part of the 
2067: discussion also holds in the case of the case (II).
2068: 
2069: Since the coupling of a D-brane to the spacetime metric is 
2070: $G_{10}M\simeq {\cal O}(g_{s})$ (where $G_{10}=g_{s}^{2}/(M_{10})^{8}$ is 
2071: the gravitational constant, $M_{10}$ is a 10-dimensional mass parameter
2072: and $M$ is the brane's mass),   
2073: the background spacetime can be regarded as a flat one and
2074: the wave we will consider is a deviation around the flat spacetime. 
2075: Since we are dealing with  a motion of D-branes with a relatively slow
2076: velocity in order to apply SYM, a non-relativistic approximation 
2077: is applicable. In addition, the motion of the D-branes are 
2078: localized around $x=0$ at least at the initial stage, 
2079: so the following formulae can be applied to this case 
2080: (see e.g. \cite{gravi});
2081: the lowest order radiation of gravitational wave is 
2082: electric quadrupole radiation, and
2083: the mass quadrupole moment of a collection of particles with 
2084: masses $\{ m_{I} \}$ in 10 spacetime dimensions is 
2085: \beqa
2086: I_{ij}=\sum_{I}m_{A}[x_{i}^{I}x_{j}^{I}
2087: -\frac{1}{9}\delta_{ij}(x^{I})^{2}]
2088: \eeqa
2089: where $x_{i}^{I}$'s ($i=1,\cdots,9 $) represent (space) positions of 
2090: a particle labeled by I.
2091: The angular distribution of the energy flux due to the radiation
2092: is given by 
2093: \beqa
2094: \frac{d^{2}E_{{\rm gravi}}}{dtd\Omega_{8}}=\frac{G_{10}}{8\pi} 
2095: <(\frac{d^{3}I_{ij}}{dt^{3}})^{2}
2096: -2(n_{i}\frac{d^{3}I_{ij}}{dt^{3}})^{2}
2097: +\frac{7}{8}(n_{i}n_{j}\frac{d^{3}I_{ij}}{dt^{3}})^{2}>
2098: \label{angra}
2099: \eeqa
2100: where $n_{i}\equiv (\bar{x}_{i}-x_{i})/|\bar{x}_{i}-x_{i}|$ where
2101: $\bar{x} $ is the position of the observer and $<\cdots >$ denotes an
2102: average over several wavelength here. 
2103: We note that in (\ref{angra}) $x_{i}$ in $I_{ij}$ 
2104: is to be evaluated at the retarded time.
2105: To apply the above formulae, we regard the branes as an assembly
2106: of infinitesimal hyper-planes.  
2107: Then, 
2108: $I_{ij}$ in the case of branes is translated into 
2109: \beqa
2110: I_{ij}=T_{p}\sum_{I=1}^{2}\int d^{p}x\
2111: \sqrt{1+(\partial_{x}x_{i}^{I})^{2}}
2112: [x^{I}_{i}x^{I}_{j}-\frac{1}{9}\delta_{ij}(x^{I}_{k})^{2}]
2113: \eeqa
2114: where the index $I(=1,2)$ denotes each of the branes.
2115: 
2116: Let us first discuss the case of D-strings in the case (I), 
2117: the simplest case,
2118: where D-strings moves (decays) toward positive and negative 
2119: directions of $x_{2}$.
2120: The $I_{ij}$ which have non-trivial components are 
2121: \beqa
2122: I_{11}&=&\frac{2}{9g_{s}}\int dx \sqrt{1+(\partial_{x}y)^{2}}
2123: (8 x^{2}-y^{2})
2124: \nonumber\\
2125: I_{22}&=&\frac{2}{9g_{s}}\int dx \sqrt{1+(\partial_{x}y)^{2}}
2126: (8 y^{2}-x^{2})\\
2127: I_{33}&=&I_{44}=\cdots =I_{99}=-\frac{2}{9g_{s}}\int dx
2128: \sqrt{1+(\partial_{x}y)^{2}}(y^{2}+x^{2})\nonumber
2129: \eeqa 
2130: Here we also assume that the branes has no velocity in $x$ direction.
2131: If we use 9-dimensional polar coordinate system as
2132: $x_{2}=\cos\theta_{1},x_{1}=\sin\theta_{1}\cos\theta_{2},
2133: x_{3}=\sin\theta_{1}\sin\theta_{2}\cos\theta_{3},\cdots,
2134: x_{8}=\sin\theta_{1}\cdots \sin\theta_{7}\cos\theta_{8},
2135: x_{9}=\sin\theta_{1}\cdots \sin\theta_{8}$,
2136: we have 
2137: \beqa
2138: \frac{d^{2}E_{{\rm gravi}}}{dtd\Omega_{8}}&=&\frac{G_{10}}{8\pi}
2139: \{\frac{7}{2}\sin^{4}\theta_{1}(\frac{d^{3}I_{1}}{dt^{3}})^{2}
2140: +\frac{7}{2}(1-\sin^{2}\theta_{1}\cos^{2}\theta_{1})
2141: (\frac{d^{3}I_{2}}{dt^{3}})^{2}\nonumber\\
2142: & &+\sin^{2}\theta_{1}
2143: (\cos^{2}\theta_{1}\cos^{2}\theta_{2}-\sin^{2}\theta_{2})
2144: \frac{d^{3}I_{1}}{dt^{3}}\frac{d^{3}I_{2}}{dt^{3}}
2145: \}\label{eneflux1}\\ 
2146: \frac{d E_{{\rm gravi}}}{dt}&=&\frac{10 \pi G_{10}}{99}
2147: ((\frac{d^{3}I_{1}}{dt^{3}})^{2}+(\frac{d^{3}I_{2}}{dt^{3}})^{2}
2148: -\frac{1}{4}\frac{d^{3}I_{1}}{dt^{3}}\frac{d^{3}I_{2}}{dt^{3}})
2149: \label{eneflux2}
2150: \eeqa
2151: where $I_{1}\equiv \int dx \sqrt{1+(\partial_{x}y)^{2}}y^{2}/g_{s}$ and 
2152: $I_{2}\equiv \int dx \sqrt{1+(\partial_{x}y)^{2}}x^{2}/g_{s}$.
2153: Since $|\frac{d^{3}I_{1}}{dt^{3}}|\gg |\frac{d^{3}I_{2}}{dt^{3}}|$
2154: (because accelerated most are the part around $x=0$ of the D-branes but
2155: $\partial_{x}y(x,t) $ vanishes at $x=0$.),
2156: the first terms in (\ref{eneflux1}) and  (\ref{eneflux2}) are dominant.
2157: Using the concrete formula for $y(x,t)$ in (\ref{braneshape1}),
2158: we have the energy flux as
2159: \beqa
2160: \frac{d E_{{\rm gravi}}}{dt}&\simeq&\frac{10 \pi G_{10}}{99g_{s}^{2}}
2161: 16 (2\beta)^{3} A(t)^{4}
2162: \eeqa
2163: where $A(t)\propto e^{\sqrt{2\beta} t}$ is used.
2164: It is of course non-vanishing.  
2165: Thus, the decaying D-strings certainly radiate closed strings as
2166: gravitational wave.
2167: 
2168: In the cases of D-p-branes with $p\ge 2$ in the case (I), 
2169: only the diagonal elements
2170: of $I_{ii}$ have non-trivial values again 
2171: due to the integration with respect to $x_{\mu}$. 
2172: To carry out the computation of the energy flux of the radiation, 
2173: it is necessary to regularize the 
2174: world volume of the D-branes, and if we do it, 
2175: we can easily show that the energy flux
2176: is shown to be non-vanishing, though the final result depends on
2177: the regularization volume.
2178: Thus, the decaying D-p-branes with $p\ge 2$ also 
2179: radiate gravitational wave.  
2180: 
2181: In addition, radiation of RR gauge field should also occur after 
2182: recombination of D-branes; each of the D-branes in the initial
2183: settings is a BPS brane, so each part of the branes should be 
2184: a source of the flux of RR (p+1)-form gauge field. 
2185: Thus, RR gauge wave should also be radiated if the branes are 
2186: accelerated and have non-vanishing multi-derivatives of 
2187: their quadrupole moment with respect to time,\footnote{
2188: We note that the lowest order RR gauge radiation is also their 
2189: electric quadrupole one; 
2190: the two D-branes are regarded to have RR charges of the
2191: {\it same} signature because if one set $\theta=0$, the force due to RR
2192: gauge field is repulsive.}  
2193: which is the case we are considering.
2194: We expect that the energy flux of the RR gauge wave 
2195: has about the same order as that of the gravitational wave.
2196:   
2197: Let us discuss the order of the energy given to the fundamental strings
2198: in the case of D-strings.
2199: In this paper we do not discuss the scale of $M_{10}$, and set
2200: $M_{10}=M_{s}(=1/l_{s})$ for simplicity.
2201: Then, since $G_{10}=g_{s}^{2}$ and $A(t)^{2}\sim {\cal O}(g_{s})$,
2202: the flux $d E_{{\rm gravi}}/dt\sim {\cal O}(g_{s}^{2})$.
2203: The energy flux of the RR gauge radiation is expected to be the same
2204: order as  $d E_{{\rm gravi}}/dt$.
2205: Taking into account that $E_{{\rm gravi}}(t)\propto A(t)^{4}$ and 
2206: the asymptotic behavior of $E_{{\rm open}}$, we have
2207: the ratio of the energy given to the massless closed strings to that
2208: of open strings is 
2209: \beqa
2210: \frac{dE_{{\rm massless \ closed}}/dt}{dE_{{\rm open}}/dt}
2211: \propto 
2212: g_{s}^{2}e^{2\sqrt{2\beta} t} {\rm or\ } g_{s}^{2}e^{3\sqrt{2\beta} t}.
2213: \label{ratio1}
2214: \eeqa
2215: Thus, the energy given to the massless closed string is very small
2216: at the initial stage compared with that of open strings, but
2217: the ratio increases exponentially with a typical time-scale
2218: $T^{(I)}=1/\sqrt{2\beta}$
2219: 
2220: As for the case (II), the shape of the D-branes in all direction
2221: may not be defined due to the non-commutativity of the scalar fields,
2222: but it is not plausible that cancellation
2223: of the radiation would happen.
2224: So, gravitational radiation and that of RR gauge field are strongly 
2225: expected to occur as creation of closed strings.
2226: 
2227: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2228: 
2229: Next, we want to discuss the behavior of the system beyond SYM.
2230: If the distance between the D-branes becomes larger than $l_{s}$,
2231: it is difficult to imagine that the created open strings extend 
2232: unlimitedly, because the string has its tension.
2233: It also seems unnatural if one take into account increase of entropy.  
2234: Then, what will happen? Since open strings are created {\it in pairs} 
2235: it is expected that {\it each pair of open strings connecting the branes
2236: are cut to pieces to form some closed strings and two open strings},
2237: each of the latter of which has its both ends on each of the two D-branes,
2238: as Fig.11. 
2239: \begin{figure}[h]
2240: \begin{center}
2241: \includegraphics[width=12cm]{fig11}\\
2242: %Figure 9
2243: \caption{}
2244: \end{center}
2245: \end{figure}
2246: That is, such a picture seems to arise that the decaying (annihilating) 
2247: D-branes leave many  closed strings 
2248: behind (and radiate some of them) while producing many open strings 
2249: which start from and end on each brane.
2250: If this is true, the further the two branes go from each other,
2251: the more possibility there arises to create closed strings. 
2252: This expectation is consistent with tendency of the ratio of 
2253: $dE_{{\rm closed}}/dt$ to $dE_{{\rm open}}/dt$.
2254: The dominant mechanism to create closed strings in this setting might be  
2255: the above way.
2256: (Please note that if there was a single open string connecting 
2257: the decaying D-branes, the energy of the system was not saved by creating 
2258: a closed string.)
2259: 
2260: 
2261: It is also to be noted that 
2262: the above picture may be regarded as a generalization of 
2263: Sen's conjecture for $D\bar{D}$ annihilation\cite{sen1}; 
2264: in the case of  $D\bar{D}$ with no topological defects, 
2265: the sum of the tensions of a D-brane and an anti-D-brane are 
2266: precisely the height of the potential of the unstable vacuum and 
2267: they vanish to be nothing.  
2268: On the other hand, the  setting we are considering is a ``middle'' 
2269: configuration of  
2270: $D\bar{D}$  and parallel two BPS branes, the latter of which 
2271: has definitely positive energy. Thus, our setting also has positive 
2272: energy and hence there should remain something after the decay.
2273: According to our speculation, it may be the closed and open strings 
2274: stated above.   
2275:  
2276: 
2277: 
2278: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2279: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2280: \section{Implications for an inflation model}
2281: \setcounter{equation}{0}
2282: \setcounter{footnote}{0}
2283: 
2284: In this section we consider implications of the above results
2285: of the system  for the inflation model of the setting of two D-4-branes
2286: at one angle in ref.\cite{oneangle}\cite{tye1}. 
2287: 
2288: First, we would like to speculate about the efficiency of the reheating
2289: of the model.
2290: According to the analysis in section 5, most part of the released energy via
2291: tachyon condensation is used to create open string pairs connecting the
2292: branes and only a little part is used to accelerate the branes,
2293: at least at the initial stage of decay.
2294: Thus, from the viewpoint of motion of the branes,
2295: dissipation of the energy 
2296: is rather large initially, and might be large afterward, so
2297: the oscillation of the branes around some stable configuration 
2298: might occur only a few times. 
2299: That is, if this setting is applied to the inflation
2300: scenario, it is possible that the reheating might be efficient.
2301: 
2302: Second, we would like to discuss the mechanism to create 
2303: ``reheated'' gauge fields and fermions;
2304: from the viewpoint of cosmology they should be produced
2305: at the end of the inflation, 
2306: but the mechanism how are they produced in this setting of the inflation
2307: model has not been clarified.
2308: We speculate about a scenario of the mechanism based on our analysis:
2309: Although we are considering a simpler setting model,
2310: if a more complete model of brane-world is considered, 
2311: the models should be the one in which at the end of inflation,
2312: the configuration of branes are put so oriented
2313: that they reproduce excitations of Standard Model
2314: as given in e.g. ref.\cite{getjust}\cite{higgs} or some others.  
2315:  (A prototype of such models are discussed  e.g. in ref.\cite{2angles}.) 
2316: In the model, fields like fermions and gauge fields
2317: arise as open strings connecting the branes or on each of the branes.
2318: Our suggestion is that
2319: the open string pairs connecting the branes like those in Fig.8
2320: and the open strings on each of the branes like those in Fig.11
2321: might directly correspond to the ``reheated'' fermions or gauge fields
2322: created at the end of the inflation.
2323: It would be interesting to find a framework describing the process in
2324: the region of distance other than SYM
2325: or beyond SYM, and discuss the above possibilities.
2326: 
2327: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2328: \section{Summary and discussion}
2329: \setcounter{equation}{0}
2330: \setcounter{footnote}{0}
2331: 
2332: We have investigated what happens in the process after the recombination 
2333: of two D-p-branes at one angle for the two types of concrete 
2334: initial conditions;
2335: case (I):two D-p-branes have been parallel until $t=0$ but are put
2336: intersected at one angle $\theta$ at the instant $t=0$, and 
2337: case (II):one of parallel two
2338: D-p-branes were rotated
2339: by a small angle $\theta$ and 
2340: are approaching each other very slowly.
2341: In summary, the results are the following:
2342: \begin{itemize} 
2343: \item Explicit functions of typical amplitudes of condensing tachyonic
2344:       fields (dependent on time and location) have been obtained. 
2345: \item Time-evolution of the D-branes' shape after recombination
2346: has been obtained, and seemingly queer behavior of the D-branes' shape that
2347: each brane comes to have three extremes has been observed. Physical 
2348: interpretation of it has also been given.
2349: \item Pair-creation of (many) open strings with opposite orientations 
2350:       connecting the two decaying branes has been shown to occur.
2351:       Expectation value of number density of the created open strings
2352:       has
2353: been estimated,
2354:       and energy distribution into the open strings has also been
2355:       evaluated; it has been shown that most part of the energy released 
2356:       via
2357:       tachyon condensation is used to create the open string pairs, and
2358:       dissipation of the D-branes' energy is large at the initial stage.
2359: \item Creation of closed strings 
2360:       as gravitational radiation from the branes
2361:       has also been shown to occur; its energy flux has been evaluated
2362:       quantitatively,
2363:       In addition, a strong suggestion has been obtained that 
2364:       closed strings are created out of open string pairs after the
2365:       two branes get further from each other.
2366: \item Speculations on implications of the above results for the  inflation
2367:       model of the D-4-branes at one angle have been given: 
2368:       dissipation of the branes' energy (due to the creation of the
2369:       strings)
2370:       might be large
2371:       and reheating might be efficient. In addition, 
2372: ``reheated'' gauge fields and fermion which should appear at the end of the
2373:       inflation, might correspond as the open strings created
2374:       accompanying the decay of the D-branes.
2375: \end{itemize}
2376: 
2377: Among the above, the most important result is 
2378: the pair-creation of open strings connecting the decaying D-branes,
2379: which strongly suggests that closed strings are created afterward.
2380: These creations in the case of decay of D-p-branes at one angle
2381: are in some senses like the case of decay of an unstable
2382: D-brane (or s-branes\cite{stro1}) where rolling tachyons \cite{sen8}  
2383: are used to discuss
2384: production of open string pairs\cite{stro2}\cite{stro3}\cite{stro4} 
2385: and closed strings 
2386: \cite{closed1}\cite{closed11}\cite{closed2}\cite{closed3}
2387: (see also \cite{st1}\cite{sen9} 
2388: and other references therein).
2389: %The point that open string creation at tree level has strongly 
2390: %connected to the creation of closed strings
2391: %might be related to Sen's argument of open -closed duality in ref.{sen9}.
2392: Compared with the latter cases, the analysis used in this paper is
2393: only perturbative and can follow the behavior of the system 
2394: for only a rather short time-scale.
2395: For further study of recombination of D-branes,
2396: the approaches analogous to the case of an unstable D-brane
2397: may be useful and is to be explored.
2398: 
2399: 
2400:     
2401: Finally, we discuss the problem of non-commutativity, 
2402: uncertainty and diagonalization procedure (gauge choice) 
2403: of VEV's of adjoint scalar fields 
2404: $X_{p+1}$ and $X_{9}$, which appear in the solution of case (II).
2405: The commutation relation of the two is
2406: \beqa
2407: [X_{p+1},X_{9}] = z(t) \left(
2408: \begin{array}{cc}
2409: 0 & -d^{*} \\
2410: d & 0 
2411: \end{array}
2412: \right). 
2413: \label{comm}
2414: \eeqa
2415: It is plausible for the relation to give an uncertainty relation
2416: \beqa
2417: \Delta x_{p+1} \Delta x_{9} \ge |z(t)||dx,t|\propto 
2418: |z_{0}-vt|A(t)e^{-\beta x^{2}}.
2419: \label{uncer}
2420: \eeqa
2421: That is, this is a solution in which
2422: the scale of uncertainty changes depending on location
2423: and time.
2424: 
2425: Let us assume this uncertainty relation.
2426: Then, the situation is as follows:
2427: If one focuses on the positions in $x_{p+1}$ direction
2428: and do not determine those in $x_{9}$ direction,
2429: it seems possible to discuss the shapes of branes projected on 
2430: $x_{p}x_{p+1}$-plane with some high accuracy. 
2431: so, we have argued that the discussion in section 4 holds.
2432: As for the ``shape'' (or profile) in  $x_{9}$ direction, 
2433: suppose one diagonalizes $X_{p+1}$ and describes their positions
2434: in the direction with the accuracy of the order $|d(x=0)|$,
2435: i.e. the accuracy one can determine whether the branes are 
2436: combined or not at $x=0$. 
2437: Then, one is able to determine the positions in $x_{9}$ direction
2438: only with the accuracy of order $|z|e^{-\beta x^{2}}$, 
2439: which means that we cannot
2440: tell whether the branes are separated or not at $x=0$, and
2441: at the point a little away from $x=0$, one can tell the branes are 
2442: separated but cannot determine the shape. And at the point more away from
2443: $x=0$, one can tell the shape of the branes projected on 
2444: $x_{p}x_{p+1}$-plane.
2445: 
2446: One may notice that there seems a bit problematic thing: 
2447: When one takes the gauge which diagonalizes  $X_{9}$ instead,  
2448: no evidence of recombination process can be observed.
2449: It may be possible to attribute this problem to
2450: the  gauge choice: On the one hand there is a gauge where 
2451: geometric information of the branes is easy to extract 
2452: (``physical gauge'' where $X_{p+1}$ is diagonalized).
2453: On the other hand, 
2454: there is a gauge where
2455: the behavior of fluctuations like tachyon condensation
2456: is easy to describe (the gauge $X_{9}$ is diagonalized).
2457:  
2458: In addition, we would like to speculate about the meaning of
2459: diagonal elements of the scalar field which is not diagonalized; 
2460: it might have some geometric meaning.
2461: In the physical gauge, the VEV of $X_{9}$ is written as
2462: \beqa
2463: X_{9}^{'} = U_{0}X_{9}U_{0}^{-1} =
2464: \frac{z(t)}{2\sqrt{(\beta x)^{2}+|d|^{2}}}
2465: \left(
2466: \begin{array}{cc}
2467:  \beta x & -d^{*} \\
2468: -d & -\beta x
2469: \end{array}
2470: \right). 
2471: \label{x9}
2472: \eeqa
2473: The diagonal elements represent two kink-like shapes (profiles)
2474: along x-direction with a width $|d|/\beta$ with 
2475: both ends approaching $\pm z(t)/2$, which fits our expectation
2476: for the branes' shapes projected on $x_{p}x_{9}$-plane.  
2477: It may be a merely coincidence, 
2478: but there might be a possibility that it has some geometric meaning.
2479: Let us remember the solution of dielectric (spherical) 
2480: D2/D0 bound states
2481: given by Myers in ref.\cite{myers}. We concentrate
2482: the N=2 (two D0) case. The solution is
2483: \beqa
2484: \begin{array}{ccc}
2485: X_{1}= \frac{R}{\sqrt{3}}
2486: \left(
2487: \begin{array}{cc}
2488: 0 & 1 \\
2489: 1 & 0
2490: \end{array}
2491: \right),
2492:  &
2493: X_{2}= \frac{R}{\sqrt{3}}
2494: \left(
2495: \begin{array}{cc}
2496: 0 & -i \\
2497: i & 0
2498: \end{array}
2499: \right),&
2500: X_{3}= \frac{R}{\sqrt{3}}
2501: \left(
2502: \begin{array}{cc}
2503: 1 & 0 \\
2504: 0 & -1
2505: \end{array}
2506: \right)
2507: \end{array}
2508: \eeqa
2509: where $ R$ is the extent of the spherical 
2510: D-2-brane to which two D-0-branes are bound.
2511: If one focuses on $x_{3}$ direction, 
2512: the two D-0-branes are considered to locate in the north 
2513: and south pole of the sphere in the $x_{3}$ direction. 
2514: Then, one cannot determine the positions of two D0's
2515: in $x_{1}$ and $x_{2}$ direction, but the zeros 
2516: in diagonal elements of them may be interpreted as 
2517: center of mass positions or expectation values of their positions.
2518: Analogous to this case, 
2519: it might be possible that 
2520: the diagonal elements of $X_{9}^{'}$ describe center-of mass positions of
2521: the branes in $x_{9}$ direction though the definite position is uncertain. 
2522: The above point, related with non-commutativity, gauge
2523: choice and uncertainty, 
2524: would be interesting and to be explored in the future. 
2525: 
2526: \vskip 3ex
2527: \noindent
2528: {\large\bf Acknowledgement}
2529: \vskip 3ex
2530: 
2531: %
2532: I would especially like to thank Hidehiko Shimada for many fruitful
2533: discussions, encouragement and much adequate advice 
2534: in proceeding this work.
2535: I also thank to Prof. K. Hashimoto and S. Nagaoka for useful 
2536: communication and discussion.
2537: I also like to thank Dr. Taro Tani for many useful discussion 
2538: and encouragement, Prof. M. Kato for useful comments (questions),
2539: directing my attention to some aspects I was not considering, 
2540: and Y. Aisaka for teaching me how to use many computer tools.
2541: I would like to thank Prof. H. Tye for giving me helpful
2542: comments at the conference String Phenomenology 2003. 
2543: Finally, I would like to thank members of Theoretical Physics Laboratory
2544: at RIKEN for useful discussion.
2545: This work is supported by Japanese Society for the Promotion of 
2546: Science under Post-doctorial Research Program (No.13-05035). 
2547: 
2548: 
2549: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2550: \appendix
2551: 
2552: \vskip 3ex
2553: \noindent
2554: {\Large {\bf Appendix}}
2555: 
2556: \section{The width of a Gaussian wave function
2557: for a mode with a time-dependent (negative) frequency-squared}
2558: \setcounter{equation}{0}
2559: 
2560: %\setcounter{section}{5}
2561: 
2562: In this appendix we first review derivation of the propagator (kernel) 
2563: $K$ of the harmonic oscillator with a time-dependent frequency,
2564: and then the fact that
2565: an initially Gaussian wave function keeps to be
2566: Gaussian within WKB approximation,
2567: and finally present the formula for
2568: the time-dependent width of the Gaussian in terms of
2569: basis functions. See for more detail ref.\cite{fey}.
2570: 
2571: If one find two independent solutions $A(t)$ and $B(t)$ (which 
2572: we denote as basis functions) to the ``WKB-approximated 
2573: equation of motion
2574: \beqa
2575: \frac{d^{2} x(t)}{dt^{2}}= -\omega(t)^{2} x(t),
2576: \label{eoma}
2577: \eeqa
2578: we can immediately write essential part of the propagator\cite{fey}
2579: in terms of $A(t)$ and $B(t)$ up to WKB approximation.
2580: The solution representing the  classical path of a motion 
2581: from $(x_{0},t_{0})$ to $(x_{1},t_{1})$ is
2582: \beqa
2583: x(t)&=&\frac{1}{B_{1}A_{0}-A_{1}B_{0}}[\{ B_{1}A(t)-A_{1}B(t)\} x_{0}
2584: +\{ -B_{0}A(t)+A_{0}B(t)\} x_{1}]\\
2585: &\equiv & h_{0}(t) x_{0}+ h_{1}(t) x_{1}\label{classicalsol}
2586: \eeqa
2587: where we abbreviate $A(t_{1})$ as $A_{1}$, etc.
2588: Substituting (\ref{classicalsol}) for the action 
2589: $S=\int_{t_{0}}^{t_{1}} dt \ m(\dot{x}^{2}-\omega^{2}x^{2})/2$, 
2590: and exponentiating 
2591: the one times $i$, we have the propagator $K(x_{1},t_{1};x_{0},t_{0})=
2592: F e^{iS_{{\rm cl}}}$ where $S_{{\rm cl}}$ is quadratic in $x$ as
2593: \beqa
2594: S_{{\rm cl}}=\frac{m}{2}\sum_{i=0,1}x_{i}x_{j} \alpha_{ij}(t_{1};t_{0})
2595: \eeqa
2596: where   
2597: \beqa
2598: \alpha_{ij}(t_{1};t_{0})\equiv \int_{t_{0}}^{t_{1}} dt 
2599: [\frac{dh_{i}}{dt} \frac{dh_{j}}{dt} -\omega(t)^{2} h_{i}h_{j} ].
2600: \eeqa
2601: $F$ is a time-dependent normalization but not dependent on $x_{i}$,
2602: which can be determined 
2603: in several ways. 
2604: (In this paper we determine the factor $F$ so that the propagator 
2605: in each case $K$
2606: satisfies the Schrodinger equation.)
2607: Since the wave function for $t=t_{1}$ is given by
2608: $\Psi(x_{1},t_{1})=\int dx_{0} K(x_{1},t_{1};x_{0},t_{0})
2609: \Psi(x_{0},t_{0})$,
2610: the absolute value of
2611: an initially Gaussian wave function remains to be Gaussian,
2612: since the above is only a Gaussian integral.
2613: The behavior of the system  is represented by the 
2614: time-dependent width of the Gaussian
2615: obtained  in terms of the basis functions as
2616: \beqa
2617: \Delta(t)^{2}=\Delta(t_{0})^{2}
2618: \frac{\alpha_{00}(t;t_{0})^{2}}{\alpha_{01}(t;t_{0})^{2}}
2619: +\frac{1}{\Delta^{2}(t_{0})^{2}(\alpha_{01})^{2}}.\label{formaldelta}
2620: \eeqa 
2621: Thus, finding the basis functions leads directly to the 
2622: time-evolution behavior of the width and the system up to WKB
2623: approximation as far as the the initial wave function is Gaussian.
2624: We note that normalizations of
2625: $A(t)$ and $B(t)$ do not appear in
2626: (\ref{classicalsol}) and hence also not in the final result.
2627: 
2628: 
2629: 
2630: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2631: 
2632: \begin{thebibliography}{99}
2633: 
2634: \bibitem{angle} M. Berkooz, M. R. Douglas and R. G. Leigh,
2635: {\it ``Branes Intersecting at Angles''},  Nucl.Phys. {\bf B480} (1996) 265,
2636: hep-th/9606139.
2637: 
2638: \bibitem{halyo}  E. Halyo, {\it ``Inflation from Rotation''}, 
2639: hep-ph/0105341.
2640: 
2641: \bibitem{hirano1}  C. Herdeiro, S. Hirano and R. Kallosh, 
2642: {\it ``String Theory and Hybrid Inflation/Acceleration''},
2643: JHEP {\bf 0112} (2001) 027, hep-th/0110271. 
2644: 
2645: \bibitem{ky} B. Kyae and Q. Shafi, {\it ``Branes and Inflationary Cosmology''},
2646: Phys. Lett. {\bf B526} (2002) 379, hep-ph/0111101.
2647: 
2648: \bibitem{oneangle}  J. Garcia-Bellido, R. Rabadan and F. Zamora, 
2649: {\it ``Inflationary Scenarios from Branes at Angles''},
2650: JHEP {\bf 0201} (2002) 036, hep-th/0112147. 
2651: 
2652: \bibitem{int2} R. Blumenhagen, B. Kors, D. Lust and T. Ott,
2653: {\it ``Hybrid Inflation in Intersecting Brane Worlds''},
2654: Nucl.Phys. B641 (2002) 235, hep-th/0202124.
2655: 
2656: \bibitem{hirano2} K. Dasgupta, C. Herdeiro, 
2657: S. Hirano and R. Kallosh,
2658: {\it ``D3/D7 Inflationary Model and M-theory''}, 
2659: Phys.Rev. {\bf D65} (2002) 126002, hep-th/0203019.
2660: 
2661: \bibitem{tye1}  N. Jones, H. Stoica and S.-H. H. Tye,
2662: {\it ``Brane Interaction as the Origin of Inflation''}, 
2663: JHEP {\bf 0207} (2002) 051, hep-th/0203163.
2664: 
2665: \bibitem{2angles} M. Gomez-Reino and I. Zavala,
2666: {\it ``Recombination of Intersecting D-Branes and Cosmological Inflation''},
2667:  JHEP {\bf 209} (2002) 020, hep-th/0207278. 
2668: 
2669: \bibitem{bro}  J. H. Brodie and D. A. Easson,
2670: {\it ``Brane inflation and reheating''}, hep-th/0301138.
2671: 
2672: 
2673: \bibitem{br1}  R. Brandenberger, G. Geshnizjani and S. Watson,
2674: {\it `` On the Initial Conditions for Brane Inflation''},
2675: hep-th/0302222.
2676: 
2677: \bibitem{bl1}  R. Blumenhagen, L. Goerlich, B. Kors and D. Lust, 
2678: {\it ``Noncommutative Compactifications of Type I Strings 
2679: on Tori with Magnetic Background Flux ''}, JHEP {\bf 0010} (2000) 006,
2680: hep-th/0007024.
2681: 
2682: \bibitem{chiral} 
2683: G. Aldazabal, S. Franco, L. E. Ibanez, R. Rabadan, A. M. Uranga,
2684: {\it ``D=4 Chiral String Compactifications from Intersecting Branes''
2685: },  J.Math.Phys. {\bf 42} (2001) 3103-3126, hep-th/0011073;
2686: {\it ``Intersecting Brane Worlds''
2687: } JHEP {\bf 0102} (2001) 047, hep-ph/0011132.
2688: 
2689: \bibitem{bl2}  R. Blumenhagen, B. Kors and D. Lust, 
2690: {\it ``Type I Strings with F- and B-Flux''}, JHEP {\bf 0102} (2001) 030,
2691: hep-th/0012156.
2692: 
2693: \bibitem{getjust}  L.E. Ibanez, F. Marchesano and R. Rabadan,
2694: {\it ``Getting just the Standard Model at Intersecting Branes''},
2695: JHEP {\bf 0111} (2001) 002, hep-th/0105155.
2696: 
2697: \bibitem{higgs}  D. Cremades, L.E. Ibanez and F. Marchesano,
2698: {\it ``Intersecting Brane Models of Particle Physics and 
2699: the Higgs Mechanism''},
2700: JHEP {\bf 0207} (2002) 022, hep-th/0203160.
2701: 
2702: \bibitem{koko1}  C. Kokorelis, 
2703: {\it ``Exact Standard model Structures from Intersecting D5-Branes''}, 
2704: hep-th/0207234.
2705: 
2706: \bibitem{ur1} A. M. Uranga, 
2707: {\it``Chiral four-dimensional string compactifications 
2708: with intersecting D-branes''}, hep-th/0301032,
2709: and references therein.
2710: 
2711: \bibitem{bachas} C. Bachas, {\it ``A WAY TO BREAK SUPERSYMMETRY''}, 
2712: hep-th/9503030.
2713: 
2714: \bibitem{koko2}  C. Kokorelis,
2715: {\it ``New Standard Model Vacua from Intersecting Branes''}, 
2716: JHEP {\bf 0209} (2002) 029,
2717: hep-th/0205147;
2718: {\it ``Exact Standard Model Compactifications from Intersecting Branes''},
2719: JHEP {\bf 0208} (2002) 036, hep-th/0206108.
2720: 
2721: \bibitem{sen1} A. Sen, 
2722: {\it ``Tachyon Condensation on the Brane Antibrane System''}, 
2723: JHEP {\bf 9808} (1998) 012,
2724: hep-th/9805170; {\it ``Descent Relations Among Bosonic D-branes''},
2725: Int. J. Mod. Phys. {\bf A14 (1999) 4061}, hep-th/9902105;
2726: {\it ``Universality of the Tachyon Potential''}, 
2727: JHEP {\bf 9912} (1999) 027, hep-th/9911116.
2728: 
2729: \bibitem{y1} P. Yi,
2730: {\it ``Membranes from Five-Branes and Fundamental Strings
2731: from Dp Branes''},
2732: Nucl. Phys. {\bf B550} (1999) 214, hep-th/9901159.
2733: 
2734: \bibitem{y2} O. Bergman, K. Hori and P. Yi,
2735: {\it ``Confinement on the Brane''},
2736: Nucl. Phys. {\bf B580} (2000) 289, hep-th/0002223.
2737: 
2738: \bibitem{sen5} A. Sen,
2739: {\it ``Fundamental Strings in Open String Theory at the Tachyonic
2740: Vacuum''},
2741: J. Math. Phys. {\bf 42} (2001) 2844, hep-th/0010240.
2742: 
2743: 
2744: \bibitem{lar}  P. Kraus, F. Larsen, {\it ``
2745: Boundary String Field Theory of the DDbar System
2746: ''}, Phys. Rev. {\bf D63} (2001) 106004, hep-th/0012198.
2747: 
2748: \bibitem{tera}  T. Takayanagi, S. Terashima, T. Uesugi, {\it ``
2749: Brane-Antibrane Action from Boundary String Field Theory
2750: ''}, JHEP {\bf 0103} (2001) 019, hep-th/0012210.
2751: 
2752: \bibitem{tye2}  N. T. Jones and S.-H. H. Tye,
2753: {``An Improved Brane Anti-Brane Action from Boundary Superstring Field
2754: Theory and Multi-Vortex Solutions''},  JHEP {\bf 0301} (2003) 012,
2755: hep-th/0211180.
2756: 
2757: \bibitem{pol} J. Polchinski, {\it ``String Theory vols. 1 \& 2''},  
2758: Cambridge University Press,
2759: Cambridge 1998.
2760: 
2761: \bibitem{hn} K. Hashimoto and S Nagaoka, 
2762: {\it ``Recombination of Intersecting D-branes
2763: by Local Tachyon Condensation''},
2764: hep-th/0303204.
2765: 
2766: \bibitem{wh} W. H. Huang, 
2767: {\it ``Recombination of Intersecting D-branes in Tachyon Field Theory''},
2768: hep-th/0304171.
2769: 
2770: \bibitem{khw}  K. Hashimoto and W. Taylor,
2771: {\it ``Strings between branes''},
2772: hep-th/0307297.
2773: 
2774: \bibitem{guth} A. H. Guth and S. -Y. Pi, {\it ``Quantum mechanics of 
2775: the scalar field in the new inflationary universe''},
2776: Phys. Rev. {\bf D32} (1985) 1899.
2777: 
2778: \bibitem{boya1} : D. Boyanovsky, M. D'attanasio, H.J. de Vega, 
2779: R. Holman and D.-S.Lee, {\it`` Reheating and Thermalization, 
2780: Linear Vs. Non-Linear Relaxation''}, Phys. Rev. {\bf D52} (1995) 6805,
2781: hep-ph/9507414.
2782: 
2783: \bibitem{linde1} G. Felder, J. Garcia-Bellido, P. B. Greene, 
2784: L. Kofman, A. Linde and I. Tkachev, {\it 
2785: ``Dynamics of Symmetry Breaking and Tachyonic Preheating''},
2786: Phys. Rev. Lett. {\bf 87} (2001) 011601, hep-ph/0012142.
2787: 
2788: \bibitem{baal} P. van Baal, {\it ``SU(N) Yang-Mills Solutions with 
2789: Constant Field Strength on $T^{4}$''}, Commun. Math. Phys. {\bf 94} 
2790: (1984) 397.
2791: 
2792: \bibitem{hw} A. Hashimoto and W. Taylor IV, 
2793: {\it ``Fluctuation Spectra of Tilted and Intersecting D-branes 
2794: from the Born-Infeld Action''},
2795: Nucl. Phys. {\bf B503} (1997) 193,
2796: hep-th/9703217.
2797: 
2798: \bibitem{fey} R.P.Feynmann and A. R. Hibbs, 
2799: {\it ``Quantum Mechanics 
2800: and Path Integrals''},  McGraw-Hill Inc., (1965).
2801: 
2802: \bibitem{jabbari}  H. Arfaei, and M. M. Sheikh Jabbari
2803: {\it ``Different D-brane Interactions''}, Phys. Lett. {\bf B394} (1997) 288, 
2804: hep-th/9608167.
2805: 
2806: \bibitem{witten} E. Witten,
2807: {\it ``Bound States Of Strings And $p$-Branes''},
2808: Nucl. Phys. {\bf B460} (1996) 335, hep-th/9510135.
2809: 
2810: \bibitem{raams} W. Taylor and M. Van Raamsdonk,
2811: {\it ``Multiple Dp-branes in Weak Background Fields''},
2812: Nucl. Phys. {\bf B573} (2000) 703, hep-th/9910052; 
2813: {\it ``Multiple D0-branes in Weakly Curved Backgrounds''},
2814: hep-th/9904095. 
2815: 
2816: \bibitem{gravi} S. L. Shapiro and S. A. Teukolsky,
2817: {\it ``Black holes, white dwarfs and Neutron Stars''},
2818: sJohn WIley \& Sons, Inc. (1983).
2819: 
2820: \bibitem{stro1}  M. Gutperle, A. Strominger,
2821: {\it ``Spacelike Branes''}, JHEP {\bf 0204} (2002) 018,
2822: hep-th/0202210.
2823: 
2824: \bibitem{sen8} A. Sen,
2825: {\it ``Rolling Tachyon''}, JHEP {\bf 0204} (2002) 048, hep-th/0203211; 
2826: {\it ``Tachyon Matter''}, JHEP {\bf 0207} (2002) 065, hep-th/0203265;
2827: {\it ``Field Theory of Tachyon Matter''}, 
2828: Mod. Phys. Lett. {\bf A17} (2002) 1797, hep-th/0204143;
2829: {\it ``Time Evolution in Open String Theory''},
2830: JHEP {\bf 0210} (2002) 003, hep-th/0207105.
2831: 
2832: 
2833: 
2834: \bibitem{stro2} A. Strominger,
2835: {\it ``Open String Creation by S-Branes''},
2836: hep-th/0209090.
2837:  
2838: \bibitem{stro3} M. Gutperle, A. Strominger,
2839: {\it ``Timelike Boundary Liouville Theory''},
2840: Phys. Rev. {\bf D67} (2003) 126002, hep-th/0301038.
2841: 
2842: \bibitem{stro4} A. Maloney, A. Strominger, X. Yin, 
2843: {\it ``S-Brane Thermodynamics''},
2844: JHEP {\bf 0310} (2003) 048, hep-th/0302146. 
2845: 
2846: 
2847: \bibitem{closed1} B. Chen, M. Li, F.-L. Lin,
2848: {\it ``Gravitational Radiation of Rolling Tachyon''},
2849: JHEP {\bf 0211} (2002) 050, hep-th/0209222.
2850: 
2851: \bibitem{closed11} S.-J. Rey, S. Sugimoto,
2852: {\it ``Rolling Tachyon with Electric and Magnetic Fields 
2853: -- T-duality approach -----''}, Phys.Rev. {\bf D67} (2003) 086008,
2854: hep-th/0301049.
2855: 
2856: \bibitem{closed2}  N. Lambert, H. Liu, J. Maldacena,
2857: {\it ``Closed strings from decaying D-branes''},
2858: hep-th/0303139. 
2859: 
2860: \bibitem{closed3} D. Gaiotto, N. Itzhaki, L. Rastelli,
2861: {\it ``Closed Strings as Imaginary D-branes''},
2862: hep-th/0304192. 
2863: 
2864: \bibitem{st1}  S. Sugimoto, S. Terashima, {\it ``Tachyon Matter in
2865: Boundary String Field Theory''}, JHEP {\bf 0207} (2002) 025,
2866: hep-th/0205085;   J. A. Minahan,
2867: {\it ``Rolling the tachyon in super BSFT''}, JHEP {\bf 0207} (2002) 030,
2868: hep-th/0205098; P. Mukhopadhyay, A. Sen,
2869: {\it ``Decay of Unstable D-branes with Electric Field''},  JHEP 0211
2870: (2002) 047, hep-th/0208142;  T. Okuda, S. Sugimoto, 
2871: {\it ``Coupling of Rolling Tachyon to Closed Strings''}, Nucl.Phys. B647
2872: (2002) 101, hep-th/0208196;  F. Larsen, A. Naqvi, S. Terashima,
2873: {\it ``Rolling Tachyons and Decaying Branes''}, JHEP 0302 (2003) 039, 
2874: hep-th/0212248; A. Ishida, S. Uehara,
2875: {\it ``Rolling down to D-brane and tachyon matter''},
2876: JHEP 0302 (2003) 050, hep-th/0301179;  
2877: J. L. Karczmarek, H. Liu, J. Maldacena, A. Strominger,
2878: {\it``UV Finite Brane Decay''}, hep-th/0306132.
2879: 
2880: \bibitem{sen9} A. Sen, 
2881: {\it ``Open and Closed Strings from Unstable D-branes''}, hep-th/0305011;
2882: {\it ``Open-Closed Duality at Tree Level''}, hep-th/0306137;
2883: {\it ``Open-Closed Duality: Lessons from Matrix Model''}, hep-th/0308068. 
2884: 
2885: 
2886: \bibitem{myers} R.C. Myers,  {\it ``Dielectric-Branes''},
2887: JHEP 9912 (1999) 022, hep-th/9910053.
2888: 
2889: \bibitem{hashi} K. Hashimoto, to appear.
2890: 
2891: 
2892: \end{thebibliography}
2893: 
2894: \end {document}
2895: 
2896: ----
2897: Takeshi Sato  ts-sato@lily.ocn.ne.jp
2898: 
2899: