1:
2:
3:
4: \documentclass[]{ptptex}
5: \usepackage{graphicx}% Include figure files
6: \usepackage{bm}% bold math
7:
8: \def\ds{\displaystyle}
9: \def\etal{{\it et al.\ }}
10: \def\beq{\begin{equation}}
11: \def\eeq{\end{equation}}
12: \def\pa{\partial}
13: \def\ra{\rightarrow}
14: \def\tt{\tilde{t}}
15: \def\tr{\tilde{r}}
16: \def\tp{\tilde{\phi}}
17: \def\tv{\tilde{V}}
18: \def\tm{\tilde{m}}
19: \def\to{\tilde{\omega}}
20: \def\tE{\tilde{E}}
21: \def\tQ{\tilde{Q}}
22:
23: \begin{document}
24: %\preprint{YITP-07-84}
25:
26: \title{Stability of Q-balls and Catastrophe}
27: \author{Nobuyuki \textsc{Sakai}
28: \footnote{nsakai@e.yamagata-u.ac.jp}
29: and Misao \textsc{Sasaki}\footnote{misao@yukawa.kyoto-u.ac.jp}}
30: \inst{Department of Education, Yamagata University, Yamagata 990-8560, Japan\\
31: Yukawa Institute for Theoretical Physics, Kyoto University, Kyoto 990-8502, Japan}
32: %\recdate{December 26, 2007}
33: \recdate{April 15, 2007}
34:
35: \abst{
36: We propose a practical method for analyzing stability of Q-balls for
37: the whole parameter space, which includes the intermediate region between
38: the thin-wall limit and thick-wall limit as well as Q-bubbles
39: (Q-balls in false vacuum), using catastrophe theory.
40: We apply our method to the two concrete models,
41: $V_3=m^2\phi^2/2-\mu\phi^3+\lambda\phi^4$ and
42: $V_4=m^2\phi^2/2-\lambda\phi^4+\phi^6/M^2$.
43: We find that $V_3$ and $V_4$ Models fall into
44: {\it fold catastrophe\/} and {\it cusp catastrophe}, respectively,
45: and their stability structures are quite different from each other.
46: }
47:
48: %\pacs{03.75.Lm}
49:
50: \maketitle
51:
52: Q-balls \cite{Col85}, a kind of non-topological solitons \cite{LP92},
53: appear in a large family of
54: field theories with global U(1) (or more) symmetry,
55: and could play an important role in cosmology.
56: For example, the Minimal Supersymmetric Standard Model may contain
57: baryonic Q-balls,
58: which could be responsible for baryon asymmetry \cite{EM98} and dark
59: matter \cite{KS98}.
60:
61: The stability of Q-balls has been studied in the literature.
62: Coleman argued that Q-balls are absolutely stable if the charge $Q$ is
63: sufficiently large, using the thin-wall approximation \cite{Col85}.
64: Kusenko showed that Q-balls with small $Q$ are also stable for the
65: potential
66: \beq\label{V3}
67: V_3(\phi)={m^2\over2}\phi^2-\mu\phi^3+\lambda\phi^4 ~~~
68: {\rm with} ~~~ m^2,~\mu,~\lambda>0,
69: \eeq
70: using the thick-wall approximation \cite{Kus97a}. Here the thick-wall limit is
71: defined by the limit of $\omega^2\rightarrow m^2$, where $\omega$ is the
72: angular velocity of phase rotation.
73: Multamaki and Vilja found that in the thick-wall limit the stability
74: depends on the form of the potential \cite{MV00}. Paccetti Correia
75: and Schmidt showed a useful theorem which applies to any equilibrium
76: Q-balls \cite{PCS01}:
77: their stability is determined by the sign of $(\omega/Q)dQ/d\omega$.
78:
79: It is usually assumed that the potential has an absolute minimum at $\phi=0$.
80: If $V(0)$ is a local minimum and the absolute minimum is located at $\phi\ne0$,
81: true vacuum bubbles may appear \cite{Col77}.
82: If $Q=0$, vacuum bubbles are unstable: either expanding or contracting.
83: Kusenko \cite{Kus97b} and Paccetti Correia and Schmidt \cite{PCS01} showed,
84: however, that there are stable bubbles if $Q\ne0$.
85: They called those solutions ``Q-balls in the false vacuum".
86: Hereafter we simply call them ``Q-bubbles".
87:
88: The standard method for analyzing stability is to take the second variation
89: of the total energy (given by Eq.(\ref{E}) below) and evaluate its sign.
90: However, this calculation can be executed analytically only for some
91: limited cases; in general the eigenvalue problem should be solved numerically,
92: as Axenides \etal did \cite{AKPF00}.
93: In this paper, we propose an easy and practical method for analyzing
94: stability with the help of catastrophe theory. The basic idea of
95: catastrophe theory is described in Appendix.
96: As we shall show below, once we find {\it behavior variable}({\it s}),
97: {\it control parameter}({\it s}) and a {\it potential\/} in the Q-ball system,
98: it is easy to understand the stability structure of Q-balls for
99: the whole parameter space including the intermediate region
100: between the thin-wall limit and thick-wall limit as well as Q-bubbles.
101:
102: Consider an SO(2)-symmetric scalar field, whose action is given by
103: \beq\label{S}
104: {\cal S}=\int d^4x\left[
105: -\frac12\eta^{\mu\nu}\{\pa_{\mu}\phi_1\pa_{\nu}\phi_1
106: +\pa_{\mu}\phi_2\pa_{\nu}\phi_2\}
107: -V(\phi) \right],
108: ~~~{\rm with}~~~
109: \phi\equiv\sqrt{\phi_1^2+\phi_2^2}.
110: \eeq
111: We consider spherically symmetric configurations of the field.
112: Assuming homogeneous phase rotation,
113: \beq\label{phase}
114: (\phi_1,\phi_2)=\phi(r)(\cos\omega t,\sin\omega t),
115: \eeq
116: the field equation becomes
117: \beq\label{fe}
118: {d^2\phi\over dr^2}=-\frac2r{d\phi\over dr}-\omega^2\phi+{dV\over d\phi}\,.
119: \eeq
120: This is equivalent to the field equation for a single static scalar
121: field with the potential $V_{\omega}\equiv V-\omega^2\phi^2/2$.
122: Due to the symmetry there is a conserved charge,
123: \beq\label{Q}
124: Q\equiv\int d^3x(\phi_1\pa_t\phi_2-\phi_2\pa_t\phi_1)=\omega I,
125: ~~~{\rm where}~~~
126: I\equiv\int d^3x~\phi^2.
127: \eeq
128:
129: Monotonically decreasing solutions $\phi(r)$ with the boundary conditions,
130: \beq\label{bc}
131: {d\phi\over dr}(0)=0,~~~\phi(\infty)=0,
132: \eeq
133: exist if min$(V_{\omega})<V(0)$ and $d^2V_{\omega}/d\phi^2(0)>0$,
134: which is equivalent to
135: \beq\label{omega}
136: \omega_{\rm min}^2<\omega^2<m^2
137: ~~~{\rm with}~~~
138: \omega_{\rm min}^2\equiv{\rm min}\left({2V\over\phi^2}\right),~~~
139: m^2\equiv{d^2V\over d\phi^2}(0)\,,
140: \eeq
141: where we have put $V(0)=0$ without loss of generality.
142: The two limits $\omega^2\ra\omega_{\rm min}^2$ and $\omega^2\ra m^2$
143: correspond to the thin-wall limit and the thick-wall limit, respectively.
144: The condition $\omega_{\rm min}^2<m^2$ is not so restrictive because it is
145: satisfied if the potential has the form,
146: \beq\label{V0}
147: V={m^2\over2}\phi^2-\lambda\phi^n+O(\phi^{n+1})~~
148: {\rm with} ~~ m^2>0,~\lambda>0,~~n\ge3\,.
149: \eeq
150: The total energy of the system for equilibrium solutions is given by
151: \beq\label{E}
152: E={Q^2\over2I}+\int d^3x\left\{\frac12\left({d\phi\over dr}\right)^2+V\right\}.
153: \eeq
154: Note that the variation of $E$ under fixed $Q$, $\delta E/\delta\phi|_Q=0$,
155: reproduces the field equation (\ref{fe}).
156:
157: Let us discuss how we apply catastrophe theory to
158: the present Q-ball system. Catastrophe theory is briefly
159: described in Appendix.
160: An essential point is to choose {\it behavior variable}({\it s}),
161: {\it control parameter}({\it s}) and a {\it potential\/} in the Q-ball
162: system appropriately.
163: For a given potential $V(\phi)$ and charge $Q$,
164: we consider a one-parameter family of perturbed field configurations
165: $\phi_{\omega}(r)$ near the equilibrium solution $\phi(r)$.
166: The one-parameter family is chosen to satisfy $I[\phi_{\omega}]=Q/\omega$.
167: Then the energy is regarded as a function of $\omega$,
168: $E(\omega)\equiv E[\phi_{\omega}]$.
169:
170: Because $dE/d\omega=(\delta E/\delta\phi_{\omega})d\phi_{\omega}/d\omega=0$
171: when $\phi_{\omega}$ is an equilibrium solution, $\omega$
172: may be regarded as a {\it behavior variable\/} and
173: $E$ as the {\it potential}.
174: On the other hand, the charge $Q$ and the model parameter(s) of $V(\phi)$
175: can be given by hand, and therefore should be regarded as {\it control parameters}.
176: We denote the model parameter(s) by $P_i$ ($i=1,2,\cdots$).
177: Then we analyze the stability of Q-balls as follows.
178: \begin{itemize}
179: \item
180: Solve the field equation~(\ref{fe}) with the boundary
181: condition~(\ref{bc}) numerically to obtain equilibrium
182: solutions $\phi(r)$ for various values of $\omega$ and model parameter(s) $P_i$.
183: \item
184: Calculate $Q$ by (\ref{Q}) for each solution to obtain the
185: {\it equilibrium space} $M=\{(\omega,P_i,Q)\}$.
186: We denote the equation that determines $M$ by $f(\omega,P_i,Q)=0$.
187: \item
188: Find folding points where $\pa P_i/\pa\omega=0$ or $\pa Q/\pa\omega=0$
189: in $M$, which are identical to the stability-change points,
190: $\Sigma=\{(\omega,P_i,Q)\,|\,{\pa f/\pa\omega}=0, ~f=0\}$.
191: \item
192: Calculate the energy $E$ by (\ref{E}) for equilibrium solutions
193: around a certain point in $\Sigma$ to find whether the point is
194: a local maximum or a local minimum. Then we find the stability
195: structure for the whole $M$.
196: \end{itemize}
197:
198: Now, using the method devised above, we investigate the stability
199: of equilibrium Q-balls.
200: Because it was shown \cite{PCS01} that in the thick-wall limit Q-balls are stable
201: if $n<10/3$ for the potential (\ref{V0}) and unstable otherwise,
202: we consider two typical models. One is given by (\ref{V3}),
203: which we call $V_3$ Model, and the other is given by
204: \beq\label{V4}
205: V_4(\phi)={m^2\over2}\phi^2-\lambda\phi^4+{\phi^6\over M^2} ~~~
206: {\rm with} ~~~m^2,~\lambda,~M^2>0\,,
207: \eeq
208: which we call $V_4$ Model.
209: For $V_3$ Model, rescaling the quantities as
210: \beq\label{rescale3}
211: \tt\equiv{\mu\over\sqrt{\lambda}}t,~~ \tr\equiv{\mu\over\sqrt{\lambda}}r,~~
212: \tp\equiv{\lambda\over\mu}\phi,~~
213: \tv_3\equiv{\lambda^3\over\mu^4}V_3,~~
214: \tm\equiv{\sqrt{\lambda}\over\mu}m,~~
215: \to\equiv{\sqrt{\lambda}\over\mu}\omega,
216: \eeq
217: the field equation (\ref{fe}), the potential (\ref{V3}),
218: the charge (\ref{Q}) and the energy (\ref{E}) are rewritten as
219: \beq\label{fe3}
220: {d^2\tp\over d\tr^2}=-\frac2{\tr}{d\tp\over d\tr}-\to^2\tp+{d\tv_3\over d\tp},~~~
221: \tv_3={\tm^2\over2}\tp^2-\tp^3+\tp^4,~~~
222: \tE={\lambda^{\frac32}\over M}E,~~~\tQ=\lambda Q.
223: \eeq
224: Similarly, for $V_4$ Model, rescaling the quantities as
225: \beq\label{rescale4}
226: \tt\equiv\lambda Mt,~~ \tr\equiv\lambda Mr,~~
227: \tp\equiv{\phi\over\sqrt{\lambda}M},~~
228: \tv_4\equiv{V_4\over\lambda^3M^4},~~
229: \tm\equiv{m\over\lambda M},~~
230: \to\equiv{\omega\over\lambda M},
231: \eeq
232: the field equation (\ref{fe}), the potential (\ref{V4}),
233: the charge (\ref{Q}) and the energy (\ref{E}) are rewritten as
234: \beq\label{fe4}
235: {d^2\tp\over d\tr^2}=-\frac2{\tr}{d\tp\over d\tr}-\to^2\tp+{d\tv_4\over d\tp},~~~
236: \tv_4={\tm^2\over2}\tp^2-\tp^4+\tp^6,~~~
237: \tE={E\over M},~~~\tQ={Q\over\lambda}.
238: \eeq
239:
240: In both models the system is regarded as a mechanical system with
241: the {\it behavior variable} $\to$, the {\it control parameters} $\tm^2$ and $\tQ$,
242: and the {\it potential} $\tE(\to;\tm^2,\tQ)$.
243: Because $\to_{\rm min}^2=\tm^2-1/2$, the existing condition~(\ref{omega})
244: reduces to
245: \beq
246: 0<\tm^2-\to^2<\frac12\,.
247: \eeq
248: The thin-wall and thick-wall limits correspond to
249: $\tm^2-\to^2\ra1/2$ and $\tm^2-\to^2\ra0$, respectively.
250: The condition for ordinary Q-balls, $\to_{\rm min}^2\ge0$,
251: reduces to $\tm^2\ge1/2$, while that for Q-bubbles,
252: $\to_{\rm min}^2<0$, to $\tm^2<1/2$.
253:
254: \begin{figure}
255: \centerline{\includegraphics[scale=.57]{f1.eps}}
256: \caption{\label{f1}
257: Structures of the {\it equilibrium spaces},
258: $M=\{(\to,\tm^2,\tQ)\}$, and their catastrophe map, $\chi(M)$,
259: into the {\it control planes}, $C=\{(\tm^2,\tQ)\}$,
260: for $V_3$ Model.
261: The dash-dotted lines in $M$ denote stability-change points $\Sigma$,
262: and the dash-dotted lines in $C$ denote their catastrophe maps $\chi(M)$.
263: Solid lines in $M$ (on the light-cyan colored surface) and
264: dashed lines (on the light-magenta colored surface)
265: represent stable and unstable solutions, respectively.
266: The arrows indicated by ``thin" and ``thick" show
267: the thin-wall limit, $\to^2\ra\to_{\rm min}^2=\tm^2-1/2$,
268: and the thick-wall limit, $\to^2\ra \tm^2$, respectively.
269: In the regions denoted by S, SU and N on $C$,
270: there are one stable solution, one stable and one unstable solutions,
271: and no equilibrium solution, respectively, for fixed $(\tm^2,\tQ)$.}
272: \end{figure}
273:
274: Figures \ref{f1} and \ref{f2}
275: show the structures of the {\it equilibrium spaces},
276: $M=\{(\to,\tm^2,\tQ)\}$, and their catastrophe map, $\chi(M)$,
277: into the {\it control planes}, $C=\{(\tm^2,\tQ)\}$,
278: for $V_3$ and $V_4$ Models, respectively.
279: We only show the results for $\to>0$; the sign transformation
280: $\to\ra-\to$ changes nothing but $\tQ\ra-\tQ$.
281: The dash-dotted lines in $M$ denote stability-change points $\Sigma$.
282: Because the equilibrium space alone does not tell us which lines,
283: solid or dashed, represent stable solutions,
284: we evaluate the energy $\tE$ for several equilibrium solutions,
285: as shown in Figs.\ \ref{f3} and \ref{f4}.
286: When there are double or triple values of $\tE$ for a given set
287: of the control parameters $(\tm^2,\tQ)$, by energetics the solution
288: with the lowest value of $\tE$ should be stable and the others should be unstable.
289: In Figs.~\ref{f3} and \ref{f4}, we also give a sketch of the {\it potential}
290: $E(\omega; \tm^2,\tQ)$ near the equilibrium solutions.
291: Once the stability for a given set of the parameters $(\tm^2,\tQ)$ is found,
292: the stability for all the sets of parameters which may be reached continuously
293: from that set without crossing $\Sigma$ is the same.
294: We therefore conclude that, in Figs.\ \ref{f1} and \ref{f2}
295: as well as in Figs.\ \ref{f3} and \ref{f4},
296: solid and dashed lines correspond to stable and unstable solutions,
297: respectively.
298:
299: \begin{figure}
300: \hspace*{-10mm}\includegraphics[scale=.57]{f2.eps}
301: \caption{\label{f2}
302: The same as Fig.\ \ref{f1}, but for $V_4$ Model.
303: Because the structure of $M$ is complicated in this case,
304: we show two pictures of $M$: The left one shows the
305: upper (front) sheet of the equilibrium space, while the right one
306: the lower (back) sheet.
307: In the regions denoted by N, U, SU and SUU on $C$,
308: there are no equilibrium solution, one unstable solution,
309: one stable and one unstable solutions, and one stable and
310: two unstable solutions, respectively, for fixed $(\tm^2,\tQ)$.}
311: \end{figure}
312:
313: According to the configurations of $\chi(\Sigma)$ in the
314: {\it control planes\/} in Figs.\ \ref{f1} and \ref{f2},
315: we find that $V_3$ Model falls into {\it fold catastrophe}
316: while $V_4$ Model falls into {\it cusp catastrophe}.
317: In the {\it control planes}, the numbers of stable and unstable solutions
318: for each $(\tm^2,\tQ)$ are represented by N, S, U, SU and SUU
319: (see the figure captions for their definitions).
320: Thus we find the stability structures of the two models
321: are very different from each other. They are found as follows.
322:
323: \begin{figure}
324: \centerline{\includegraphics[scale=.67]{f3}}
325: \caption{\label{f3}
326: A schematic picture of the {\it potential} $E(\omega; \tm^2,\tQ)$
327: of $V_3$ Model with $\tm^2=0.2$ near the equilibrium solutions,
328: and the locus of equilibrium solutions on $(\tQ, \tE)$ plane.
329: The solid and dashed lines represent stable and unstable
330: solutions, respectively.}
331: \end{figure}
332:
333: \begin{list}{}{}
334: \item[$\bm V_3$ \bf Model]~
335: \begin{list}{$\bullet$}{}
336: \item $\tm^2\ge1/2$: All equilibrium solutions are stable.
337: \item $\tm^2<1/2$ (Q-bubbles): For each $\tm^2$ there is a maximum charge,
338: $\tQ_{\rm max}$, above which equilibrium solutions do not exist.
339: For $\tQ<\tQ_{\rm max}$, stable and unstable solutions coexists.
340: It is interesting to note that stable Q-bubbles exist no matter how small $\tQ$ is.
341: \end{list}
342:
343: \item[$\bm V_4$ \bf Model]~
344: \begin{list}{$\bullet$}{}
345: \item $\tm^2\ge1/2$: For each $\tm^2$ there is a
346: minimum charge, $\tQ_{\rm min}$, below which equilibrium solutions do not exist.
347: For $\tQ>\tQ_{\rm min}$, stable and unstable solutions coexists.
348: \item $\tm^2<1/2$ (Q-bubbles): For each $\tm^2$ there is a maximum charge,
349: $\tQ_{\rm max}$, as well as a minimum charge, $\tQ_{\rm min}$,
350: where stable solutions do not exist if $\tQ<\tQ_{\rm min}$
351: or $\tQ>\tQ_{\rm max}$.
352: For $\tQ_{\rm min}<\tQ<\tQ_{\rm max}$, there are one stable
353: and two unstable solutions.
354: \\
355: As $\tm^2$ becomes smaller, $\tQ_{\rm max}$ and $\tQ_{\rm min}$ come
356: close to each other, and finally merge at $\tm^2\approx0.26$,
357: below which there is no stable solution.
358: \end{list}
359: \end{list}
360: The above results for the two models are consistent with the previous
361: results for some special cases such as the thin-wall limit, the thick-wall
362: limit and bubbles with $Q=0$.
363:
364: \begin{figure}
365: \includegraphics[scale=.67]{f4}
366: \caption{\label{f4}
367: The same as Fig.\ \ref{f3}, but for $V_4$ Model with $\tm^2=0.3$.}
368: \end{figure}
369:
370: Although we have investigated only two concrete models,
371: taking account of the fact that the stability structure falls into
372: two classes in the thick-wall limit, that is, the fact that Q-balls are
373: stable if $n<10/3$ for the potential (\ref{V0}) and unstable otherwise,
374: one expects that there are essentially two distinct stability
375: structures in the general case. Then the two types of models investigated
376: here, $V_3$ and $V_4$, may be regarded as the representatives of
377: these two distinct stability structures.
378:
379: For example, in the gravity-mediated supersymmetry breaking model \cite{SUSY},
380: the lowest-order negative term of the potential is $\sim-\phi^2\log\phi$.
381: Because this term corresponds to $n<3$ in (\ref{V0}), the stability
382: structure of this model falls into $V_3$ Type.
383: Furthermore, because the potential is positive everywhere, which
384: corresponds to $\tm^2>1/2$ in Fig.\ 1, all equilibrium solutions are
385: stable in this model.
386:
387: In summary, we have proposed a new method for analyzing
388: the stability of Q-balls using catastrophe theory.
389: An essential point is that, although the Q-ball system (\ref{S}) includes
390: infinite degrees of freedom, practically it can be regarded as a mechanical
391: system with one variable, $\omega$, near equilibrium solutions.
392: Therefore, we have applied catastrophe theory, which was
393: established for mechanical systems with finite degrees of freedom,
394: to the Q-ball system.
395: A similar analysis but on the stability of exotic black holes
396: was done by Maeda \etal \cite{Maeda} some time ago, and catastrophe theory
397: was found to be very useful. Thus it seems worthwhile to consider
398: the application of catastrophe theory to other
399: cosmological (gravitating) solitons such as gravitating Q-balls \cite{GQ},
400: topological defects, and branes.
401: It may be also interesting to apply the catastrophe-theoretic approach to
402: non-relativistic atomic Bose-Einstein condensates \cite{BEC}, where
403: Q-ball-like solitons appear.
404:
405: We thank H. Kodama, K. Maeda, K. Nakao, V. Rubakov, H. Shinkai,
406: T. Tanaka and S. Yoshida for useful discussions.
407: A part of this work was done while NS was visiting at Yukawa
408: Institute for Theoretical Physics, which was supported by Center for
409: Diversity and Universality in Physics (21COE) in Kyoto University.
410: The numerical computations of this work were carried out at the Yukawa
411: Institute Computer Facility.
412: This work was supported in part by JSPS Grant-in-Aid for Scientific Research
413: (B) No.\ 17340075, (A) No.\ 18204024 and (C) No.\ 18540248.
414:
415: \appendix
416: \section{Basic Idea of Catastrophe Theory}
417:
418: To illustrate the basic idea of catastrophe theory \cite{PS78},
419: we consider a system with one {\it behavior variable} $x$, two
420: {\it control parameters} $p,q$ and a {\it potential} $F(x;p,q)$.
421: An equilibrium point of $x$ is determined by $dF/dx=0$ for
422: each pair of $(p,q)$.
423: The set of the control parameters, $C\equiv{(p,q)}$, spans a plane
424: called the {\it control plane}, and the set of equilibrium points,
425: \beq
426: M\equiv\left\{(x,p,q)|f(x,p,q)\equiv {dF\over dx}=0\right\},
427: \eeq
428: is called the {\it equilibrium space}.
429: Because equilibrium points are stable if $\pa f/\pa x>0$, the
430: boundary of stable and unstable equilibrium points are given by the curve,
431: \beq
432: \Sigma\equiv\left\{(x,p,q)|{\pa f\over\pa x}=0, ~f=0\right\}.
433: \eeq
434:
435: The {\it catastrophe map} is defined as
436: \beq
437: \chi:~M\ra C,~~(x,p,q)\ra(p,q).
438: \eeq
439: According to Thom's theorem, depending on the configurations of the image $\chi(\Sigma)$,
440: all mechanical systems with stability-change are classified into several catastrophe types.
441: If the number of control parameters is two, as is this example, possible catastrophe types are {\it fold catastrophe\/} and {\it cusp catastrophe}.
442: As we show in the text, Q-ball models are also classified into these two types.
443:
444: If the {\it potential} $F(x;p,q)$ is known, it is easy to find equilibrium
445: points and their stability. However, even if we do not know
446: the explicit form of $F(x;p,q)$,
447: we can still find $\Sigma$ by analyzing equilibrium points as follows.
448: The Taylor expansion of $f(x,p,q)$ in the vicinity of a certain point
449: P$(x_0,p_0,q_0)$ in $M$, where $f=0$, up to the first order yields
450: \beq\label{qex}
451: q=q(x,p)=q_0-\left({\pa f\over\pa q}\right)^{-1}\left\{{\pa f\over\pa x}(x-x_0)
452: +{\pa f\over\pa p}(p-p_0)\right\},
453: ~~~{\rm if} ~~~ {\pa f\over\pa q}\ne0.
454: \eeq
455: Because $\pa f/\pa x=0$ in $\Sigma$, it follows from (\ref{qex})
456: that $\pa q/\pa x=0$ in $\Sigma$.
457: Similarly, unless $\pa f/\pa x=0$, $\pa p/\pa x=0$ in $\Sigma$.
458: Therefore, surveying the points with $\pa p/\pa x=0$ or $\pa q/\pa x=0$
459: in the {\it equilibrium space} $M$, we can obtain the set of stability-change
460: points $\Sigma$.
461:
462: \begin{thebibliography}{99}
463: \bibitem{Col85}S. Coleman, Nucl. Phys. {\bf B262}, 263 (1985).
464: \bibitem{LP92}For a review of non-topological solitons, see,
465: T. Lee and Y. Pang, Phys. Rep. {\bf 221}, 251 (1985).
466: \bibitem{EM98}K. Enqvist and J. McDonald, Phys. Lett. B {\bf425}, 309 (1998).
467: \bibitem{KS98}A. Kusenko and M. Shaposhnikov, Phys. Lett. B {\bf418}, 46 (1998).
468: \bibitem{Kus97a}A. Kusenko, Phys. Lett. B {\bf404}, 285 (1997).
469: \bibitem{MV00}T. Multamaki and I. Vilja, Nucl. Phys. B {\bf 574}, 130 (2000).
470: \bibitem{PCS01}F. Paccetti Correia and M. G. Schmidt, Eur. Phys. J. {\bf
471: C21}, 181 (2001).
472: \bibitem{Col77}S. Coleman, Phys. Rev. D {\bf 15}, 2929 (2000).
473: \bibitem{Kus97b}A. Kusenko, Phys. Lett. B {\bf406}, 26 (1997).
474: \bibitem{AKPF00}M. Axenides, S. Komineas, L. Perivolaropoulos and M.
475: Floratos, Phys. Rev. D {\bf 61}, 085006 (2000).
476: \bibitem{SUSY}K. Enqvist and J. McDonald, Phys. Lett. B {\bf 425}, 309 (1998);
477: Nucl. Phys. B {\bf 538}, 3210 (1999);
478: S. Kasuya and M. Kawasaki, Phys. Rev. D {\bf62}, 023512 (2000).
479: \bibitem{Maeda}K. Maeda, T. Tachizawa, T. Torii and T. Maki, Phys. Rev. Lett. {\bf 72}, 450 (1997);
480: T. Torii, K. Maeda and T. Tachizawa, Phys. Rev. D {\bf 51}, 1510 (1995);
481: T. Tachizawa, K. Maeda and T. Torii, {\it ibid.} {\bf 51}, 4054 (1995);
482: T. Torii, K. Maeda and T. Tachizawa, {\it ibid.} {\bf 52}, R4272 (1995);
483: K. Maeda, J. Korean Phys. Soc. 28, S468 (1995).
484: \bibitem{GQ}B. W. Lynn, Nucl. Phys. {\bf B321}, 465 (1989);
485: W. Mielke and F. E. Schunck, Phys. Rev. D {\bf66}, 023503 (2002);
486: T. Matsuda, {\it ibid.} {\bf68}, 127302 (2003).
487: \bibitem{BEC}K. Enqvist and M. Laine, JCAP {\bf0308}, 003 (2003).
488: \bibitem{PS78}For a review of catastrophe theory, see, e.g.,
489: T. Poston and I.N. Stewart, {\it Catastrophe Theory and Its Application}, Pitman (1978).
490: \end{thebibliography}
491:
492: \end{document}