1: \documentclass[floatfix,showpacs]{revtex4}
2: %\documentclass[12pt,a4wide,epsf]{article}
3: \usepackage{epsf}
4: \usepackage{epsfig, latexsym}
5: \usepackage{pictex}
6: \usepackage{amsmath}
7: \usepackage{amssymb}
8: \usepackage{amsfonts}
9: %\usepackage{exscale}
10: \usepackage[latin1]{inputenc}
11:
12:
13: \textwidth=5.5truein
14: \oddsidemargin=1.cm
15: %\textheight= 8.125truein
16: \usepackage{graphicx}
17:
18: \newcommand{\insertplot}[5]{\begin{figure}
19: \hfill\hbox to 0.05in{\vbox to #5in{\vfill
20: \inputplot{#1}{#4}{#5}}\hfill}
21: \hfill\vspace{-.1in}
22: \caption{#2}\label{#3}
23: \end{figure}}
24: \newcommand{\inputplot}[3]{\special{ps:: gsave #2 -36 mul 0 rmoveto
25: currentpoint translate #2 7.0 div #3 5.0 div scale}
26: \special{ps: plotfile #1}\special{ps:: grestore}}
27: \newcounter{fig} \newcommand{\lbfig}[1]{\refstepcounter{fig}
28: \label{#1} }
29:
30: \newcommand{\vphi}{\varphi}
31: \newcommand{\DS}{\displaystyle}
32: \newcommand{\pih}{\frac{\pi}{2}}
33: \newcommand{\sqdetg}{\sqrt{-g}}
34: \newcommand{\sqdetgi}{\frac{1}{\sqrt{-g}}}
35: \newcommand{\edil}{e^{2\kappa \phi}}
36: \newcommand{\bA}{\bar{A}}
37: \newcommand{\bF}{\bar{F}}
38: \newcommand{\bD}{\bar{D}}
39:
40: \renewcommand{\a}{\alpha}
41: \renewcommand{\b}{\beta}
42: \renewcommand{\c}{\gamma}
43: \renewcommand{\d}{\delta}
44: \newcommand{\f}{\phi}
45: \newcommand{\e}{\mu}
46: \newcommand{\g}{\nu}
47: \renewcommand{\l}{\lambda}
48: \renewcommand{\t}{\theta}
49: \newcommand{\rot}{\mbox{rot}}
50: \renewcommand{\div}{\mbox{div}}
51: \newcommand{\cosec}{\mbox{cosec}}
52:
53:
54: \begin{document}
55:
56: \title{
57: Rotating Boson Stars and $Q$-Balls II:\\
58: Negative Parity and Ergoregions
59: }
60: % \vspace{1.5truecm}
61: \author{Burkhard Kleihaus, Jutta Kunz,}
62: \affiliation{ Institut f\"ur Physik, Universit\"at Oldenburg,
63: D-26111 Oldenburg, Germany}
64: \author{Meike List, Isabell Schaffer}
65: \affiliation{ZARM, Universit\"at Bremen,
66: Am Fallturm, D-28359 Bremen, Germany}
67:
68: \vspace{0.3cm}
69:
70: \pacs{04.40.-b, 11.27.+d}
71:
72: \begin{abstract}
73: We construct axially symmetric, rotating boson stars
74: with positive and negative parity.
75: Their flat space limits represent spinning $Q$-balls.
76: $Q$-balls and boson stars
77: exist only in a limited frequency range.
78: The coupling to gravity gives rise to a spiral-like
79: frequency dependence of the mass and charge of boson stars.
80: We analyze the properties of these solutions.
81: In particular, we discuss the presence of ergoregions in boson stars,
82: and determine their domains of existence.
83: \end{abstract}
84:
85: \maketitle
86:
87: \section{Introduction}
88:
89: $Q$-balls represent stationary
90: localized solutions in flat space.
91: They arise, when a complex scalar field
92: has a suitable self-interaction
93: \cite{lee-s,coleman}.
94: The global phase invariance of the scalar field theory
95: is associated with a conserved charge $Q$,
96: corresponding for instance to particle number \cite{lee-s}.
97:
98: The time-dependence of the $Q$-ball solutions
99: resides in the phase of the scalar field
100: and is associated with a frequency $\omega_s$.
101: $Q$-balls exist only in a certain frequency range,
102: $\omega_{\rm min} < \omega_s < \omega_{\rm max}$,
103: determined by the properties of the potential
104: \cite{lee-s,coleman,lee-rev,volkov,list}.
105: At a critical value of the frequency,
106: both mass and charge of the $Q$-balls assume their minimal value,
107: from where they rise monotonically towards both limiting values
108: of the frequency.
109: Considering the mass of the $Q$-balls as a function of the charge,
110: there are thus two branches of $Q$-balls, merging and ending
111: at the minimal charge and mass.
112:
113: The simplest type of $Q$-balls is spherically symmetric.
114: These possess finite mass and charge, but carry no angular momentum.
115: Besides the fundamental spherically symmetric $Q$-balls
116: there are also radially excited $Q$-balls \cite{volkov}.
117: The fundamental spherically symmetric $Q$-balls
118: are stable along their lower mass branch,
119: as long as their mass is smaller than the mass of $Q$ free bosons
120: \cite{lee-s}.
121: %The stress-energy tensor of the spherically symmetric $Q$-balls is diagonal.
122:
123: Recently, the existence of rotating $Q$-balls was demonstrated
124: \cite{volkov,list}.
125: The rotation is achieved by including an additional dependence
126: of the phase factor of the scalar field
127: on the azimuthal angle $\varphi$,
128: where the proportionality constant $n$ must be integer.
129: The resulting stationary localized $Q$-ball solutions
130: then possess finite mass and finite angular momentum.
131: Interestingly, their angular momentum $J$ is quantized
132: in terms of their charge $Q$,
133: $J=nQ$ \cite{volkov,schunck}.
134: Rotating $Q$-balls with $n=1$ thus have the smallest angular momentum $J$
135: for a given charge $Q$.
136: The energy density and angular momentum density of
137: rotating $Q$-balls possess axial symmetry.
138: There are no (infinitesimally) slowly rotating $Q$-balls \cite{volkov}.
139:
140: The field equations allow for $Q$-balls
141: with positive and negative parity.
142: The scalar field is then symmetric resp.~antisymmetric
143: w.r.t.~reflections \cite{volkov}.
144: For a given charge $Q$ one finds thus two sequences of
145: rotating $Q$-balls, whose angular momentum increases with $n$:
146: the positive parity sequence $n^+$ and the negative
147: parity sequence $n^-$.
148: While the energy density of rotating $Q$-balls with positive
149: parity corresponds to a torus, the energy density of rotating $Q$-balls
150: with negative parity corresponds to a double torus.
151:
152: %----------------------------------------------------------------------
153:
154: When the scalar field is coupled to gravity, boson stars arise
155: \cite{lee-bs,lee-rev,jetzer,ms-review}.
156: The presence of gravity has a crucial influence
157: on the domain of existence of the classical solutions.
158: Stationary spherically symmetric boson stars ($0^+$)
159: also exist only in a limited frequency range,
160: ${\omega}_{0}(\kappa) < \omega_s < \omega_{\rm max}$,
161: where $\kappa$ denotes the strength of the gravitational coupling.
162: They show, however, a rather different type of frequency dependence.
163: In particular, in the lower frequency range, the boson star
164: solutions are not uniquely determined by the frequency.
165: Instead a spiral-like frequency dependence of the charge and the mass
166: is observed, where the charge and mass approach finite limiting values
167: at the center of the spiral.
168: Furthermore,
169: the charge and the mass of these boson stars tend to zero,
170: when the maximal value of the frequency is approached
171: \cite{lee-bs,lee-rev}.
172:
173: Rotating boson stars with positive parity have been
174: obtained before \cite{schunck,schunck2,schunck3,japan,list}.
175: A study of the frequency dependence
176: of the $1^+$ boson stars showed,
177: that their frequency dependence is analogous to the one of
178: the non-rotating $0^+$ boson stars \cite{list}.
179: In particular,
180: their charge and mass also exhibit a spiral structure.
181: Like rotating $Q$-balls,
182: rotating boson stars exhibit the angular momentum quantization,
183: $J=nQ$ \cite{schunck}.
184:
185: Here we reconsider
186: rotating boson stars with positive parity and, for the first time,
187: present rotating boson stars with negative parity.
188: We construct families of boson stars of both parities numerically
189: for constant values of the gravitational coupling strength
190: and $n=1$ and 2.
191: We investigate the properties of these solutions.
192: In particular, we analyze their frequency dependence,
193: and we demonstrate the appearence and development of ergoregions
194: for these families of boson star solutions \cite{Cardoso}.
195:
196: In section II we recall the action, the general equations of motion
197: and the global charges.
198: In section III we present the stationary axially symmetric Ansatz
199: for the metric and the scalar field, we evaluate
200: the global charges within this Ansatz, and present the
201: boundary conditions for the metric and scalar field function.
202: We discuss $Q$-ball solutions and their properties in section IV
203: and boson star solutions and their properties in section V.
204: Section VI gives our conclusions.
205:
206: \section{Action, Equations and Global Charges}\label{c1}
207:
208: \subsection{Action}\label{c1s1}
209:
210: We consider the action of a self-interacting complex scalar field
211: $\Phi$ coupled to Einstein gravity
212: \begin{equation}
213: S=\int \left[ \frac{R}{16\pi G}
214: -\frac{1}{2} g^{\mu\nu}\left( \Phi_{, \, \mu}^* \Phi_{, \, \nu} + \Phi _
215: {, \, \nu}^* \Phi _{, \, \mu} \right) - U( \left| \Phi \right|)
216: \right] \sqrt{-g} d^4x
217: \ , \label{action}
218: \end{equation}
219: where $R$ is the curvature scalar,
220: $G$ is Newton's constant,
221: the asterisk denotes complex conjugation,
222: \begin{equation}
223: \Phi_{,\, {\mu}} = \frac{\partial \Phi}{ \partial x^{\mu}}
224: \ ,
225: \end{equation}
226: and $U$ denotes the potential
227: \begin{equation}
228: U(|\Phi|) = \l |\Phi|^2 \left( |\Phi|^4 -a |\Phi|^2 +b \right)
229: =\lambda \left( \f^6- a\f^4 + b \f^2 \right) \ ,
230: \label{U} \end{equation}
231: with $|\Phi|=\f$.
232: The potential is chosen such that nontopological soliton solutions
233: \cite{lee-s}, also referred to as $Q$-balls \cite{coleman},
234: exist in the absence of gravity.
235: As seen in Fig.~\ref{potentiale},
236: the self-interaction of the scalar field has an attractive component,
237: and the potential has a minimum, $U(0)=0$, at $\Phi =0$
238: and a second minimum at some finite value of $|\Phi|$.
239: The boson mass is thus given by $m_{\rm B}=\sqrt{\lambda b}$.
240: %
241: In the numerical calculations of the solutions presented,
242: we choose the potential parameters
243: \cite{volkov,list}
244: \begin{equation}
245: \lambda = 1 \ , \ \ \ a=2 \ , \ \ \ b=1.1 \ .
246: \label{param}
247: \end{equation}
248:
249: \begin{figure}[h!]
250: %\centering
251: {\centerline{
252: \mbox{
253: \includegraphics[width=70mm,angle=0,keepaspectratio]{F1a.eps}
254: \includegraphics[width=70mm,angle=0,keepaspectratio]{F1b.eps}
255: }}}
256: %1
257: \caption{Left: The potential $U(\f)$ versus $\f$
258: for $\lambda=1$, $a=2$ and $b=1.1$ resp.~$b=1$.
259: Right: The effective potential
260: $V(\phi) = \frac{1}{2} \, \omega_s^2 \, \phi^2
261: - \frac{1}{2} U(\phi)$
262: versus $\f$
263: for several values of the frequency $\omega_s$.}
264: \label{potentiale}
265: \end{figure}
266:
267: \subsection{Equations}\label{c1s2}
268:
269: Variation of the action with respect to the metric
270: leads to the Einstein equations
271: \begin{equation}
272: G_{\mu\nu}= R_{\mu\nu}-\frac{1}{2}g_{\mu\nu}R = \kappa T_{\mu\nu}
273: \ , \label{Einstein}
274: \end{equation}
275: with
276: $\kappa = 8\pi G$
277: and stress-energy tensor $T_{\mu\nu}$
278: \begin{eqnarray}
279: T_{\mu \nu} &=& \phantom{-} g_{\mu \nu} L_M
280: -2 \frac{ \partial L}{\partial g^{\mu\nu}}
281: \\
282: &=&-g_{\mu\nu} \left[ \frac{1}{2} g^{\alpha\beta}
283: \left( \Phi_{, \, \alpha}^*\Phi_{, \, \beta}+
284: \Phi_{, \, \beta}^*\Phi_{, \, \alpha} \right)+U(\f)\right]
285: + \left( \Phi_{, \, \mu}^*\Phi_{, \, \nu}
286: +\Phi_{, \, \nu}^*\Phi_{, \, \mu} \right
287: ) \ .
288: \label{tmunu} \end{eqnarray}
289: Variation with respect to the scalar field
290: leads to the matter equation,
291: \begin{equation}
292: \left(\Box+\frac{\partial U}{\partial\left|\Phi\right|^2}\right)\Phi=0 \ ,
293: \label{field_phi}
294: \end{equation}
295: where $\Box$ represents the covariant d'Alembert operator.
296: Equations (\ref{Einstein}) and (\ref{field_phi}) represent
297: the general set of non-linear Einstein--Klein--Gordon equations.
298:
299: \subsection{Global Charges}\label{c1s3}
300:
301: The mass $M$ and the angular momentum $J$
302: of stationary asymptotically flat space-times
303: can be obtained from their respective Komar expressions \cite{wald},
304: \begin{equation}
305: {M} =
306: \frac{1}{{4\pi G}} \int_{\Sigma}
307: R_{\mu\nu}n^\mu\xi^\nu dV
308: \ , \label{komarM1}
309: \end{equation}
310: and
311: \begin{equation}
312: {\cal J} = -
313: \frac{1}{{8\pi G}} \int_{\Sigma}
314: R_{\mu\nu}n^\mu\eta^\nu dV
315: \ . \label{komarJ1}
316: \end{equation}
317: Here $\Sigma$ denotes an asymptotically flat spacelike hypersurface,
318: $n^\mu$ is normal to $\Sigma$ with $n_\mu n^\mu = -1$,
319: $dV$ is the natural volume element on $\Sigma$,
320: $\xi$ denotes an asymptotically timelike Killing vector field
321: and $\eta$ an asymptotically spacelike Killing vector field
322: \cite{wald}.
323: Replacing the Ricci tensor via the Einstein equations by the
324: stress-energy tensor yields
325: \begin{equation}
326: M
327: = \, 2 \int_{\Sigma} \left( T_{\mu \nu}
328: -\frac{1}{2} \, g_{\mu\nu} \, T_{\gamma}^{\ \gamma}
329: \right) n^{\mu }\xi^{\nu} dV \ ,
330: \label{komarM2}
331: \end{equation}
332: and
333: \begin{equation}
334: {\cal J} = -
335: \int_{\Sigma} \left( T_{\mu \nu}
336: -\frac{1}{2} \, g_{\mu\nu} \, T_{\gamma}^{\ \gamma}
337: \right) n^{\mu }\eta^{\nu} dV \ .
338: \label{komarJ2}
339: \end{equation}
340:
341: A conserved charge $Q$ is associated
342: with the complex scalar field $\Phi$,
343: since the Lagrange density is invariant under the global phase transformation
344: \begin{equation}
345: \displaystyle
346: \Phi \rightarrow \Phi e^{i\alpha} \ ,
347: \end{equation}
348: leading to the conserved current
349: \begin{eqnarray}
350: j^{\mu} & = & - i \left( \Phi^* \partial^{\mu} \Phi
351: - \Phi \partial^{\mu}\Phi ^* \right) \ , \ \ \
352: j^{\mu} _{\ ; \, \mu} = 0 \ .
353: \end{eqnarray}
354:
355: \section{Ansatz and Boundary Conditions}\label{c2}
356:
357: \subsection{Ansatz}\label{c2s1}
358:
359: To obtain stationary axially symmetric solutions,
360: we impose on the space-time the presence of
361: two commuting Killing vector fields,
362: $\xi$ and $\eta$, where
363: \begin{equation}
364: \xi=\partial_t \ , \ \ \ \eta=\partial_{\varphi}
365: \ \label{xieta} \end{equation}
366: in a system of adapted coordinates $\{t, r, \theta, \varphi\}$.
367: In these coordinates the metric is independent of $t$ and $\varphi$,
368: and can be expressed in isotropic coordinates
369: in the Lewis--Papapetrou form \cite{kkrot1}
370: \begin{eqnarray}
371: ds^2 &=&- f dt^2
372: + \frac{l}{f} \, \biggl[ h \left( dr^2 + r^2 \, d\t^2 \right)
373: + r^2 \, \sin^2 \t \, \, \left( d \varphi
374: - \frac{\omega}{r} \, dt \right)^2 \biggr] \ . \label{ansatzg}
375: \end{eqnarray}
376: The four metric functions $f$, $l$, $h$ and $\omega$
377: are functions of the variables $r$ and $\theta$ only.
378:
379: The symmetry axis of the spacetime, where $\eta=0$,
380: corresponds to the $z$-axis.
381: The elementary flatness condition \cite{book}
382: \begin{equation}
383: \frac{X,_\mu X^{, \, \mu}}{4X} = 1 \ , \ \ \
384: X=\eta^\mu \eta_\mu \
385: \ \label{regcond} \end{equation}
386: then imposes on the symmetry axis the condition \cite{kk}
387: \begin{equation}
388: h|_{\theta=0}=h|_{\theta=\pi}=1 \ .
389: \end{equation}
390:
391: For the scalar field $\Phi$ we adopt the stationary Ansatz \cite{schunck}
392: \begin{eqnarray}
393: \Phi (t,r,\t, \varphi)& = & \f (r, \t)
394: e^{ i\omega_s t +i n \varphi} \ , \label{ansatzp}
395: \end{eqnarray}
396: where $\f (r, \t)$ is a real function,
397: and $\omega_s$ and $n$ are real constants.
398: Single-valuedness of the scalar field requires
399: \begin{equation}
400: \Phi(\varphi)=\Phi(2\pi + \varphi) \ ,
401: \end{equation}
402: thus the constant $n$ must be an integer,
403: i.e., $n \, = \, 0, \, \pm 1, \, \pm 2, \, \dots \ $.
404: We refer to $n$ as the rotational quantum number.
405: When $n \not=0$, the phase factor $\exp{(i n \varphi)}$
406: prevents spherical symmetry of the scalar field $\Phi$.
407:
408: Solutions with positive and negative parity satisfy, respectively,
409: \begin{eqnarray}
410: \f (r, \pi-\t) & = & {\phantom{-}} \f (r, \t) \\
411: \f (r, \pi-\t) & = & - \f (r, \t)
412: \ , \label{parity}
413: \end{eqnarray}
414:
415: To construct stationary axially symmetric boson star solutions
416: a system of five coupled partial differential equations
417: needs to be solved.
418: In contrast, for $Q$-balls %existing in flat space
419: the metric is the Minkowski metric, i.e.,$f=l=h=1$, $\omega=0$.
420: Here, at least in principle,
421: only a single partial differential equation for the scalar field
422: function needs to be solved.
423:
424: \subsection{Mass, angular momentum and charge}\label{c2s2}
425:
426: The mass $M$ and the angular momentum $J$
427: can be read off the asymptotic expansion of the metric functions $f$
428: and $\omega$, respectively,
429: \cite{kkrot1}
430: \begin{equation}
431: f = 1- \frac{2MG}{r} + O\left( \frac{1}{r^2} \right) \ , \ \ \
432: \omega = \frac{2JG}{r^2} + O\left( \frac{1}{r^3} \right)\ ,
433: \label{MQasym1} \end{equation}
434: i.e.,
435: \begin{eqnarray}
436: M=\frac{1}{2G} \lim_{r \rightarrow \infty} r^2\partial_r \, f
437: \ , \ \ \ J=\frac{1}{2G} \lim _{r \rightarrow \infty} r^2\omega \ .
438: \label{MJ2}
439: \end{eqnarray}
440: This is seen by considering the Komar expressions
441: Eqs.~(\ref{komarM1}) and (\ref{komarJ1}),
442: with unit vector $n^\mu = (1, 0, 0, \omega/r)/\sqrt{f}$,
443: and volume element
444: $dV =1/ \sqrt{f} \, |g|^{1/2} \, dr \, d\t \, d\varphi$,
445: leading to \cite{kkrot2}
446: \begin{eqnarray}
447: M &=&
448: -\frac{1}{ 8 \pi G} \int_\Sigma R_t^t \sqrt{-g} dr d\theta d\vphi
449: \nonumber \\
450: &=& \lim_{r\to\infty}
451: \frac{2\pi }{8\pi G} \int_0^{\pi}
452: \left. \left[\frac{\sqrt{l}}{f} r^2 \sin\theta
453: \left( \frac{\partial f}{\partial r} -
454: \frac{l}{f} \sin^2\theta \omega
455: \left(\frac{\partial \omega}{\partial r} - \frac{\omega}{r} \right)
456: \right)
457: \right] \right|_{r} d\theta
458: \ , \phantom{\frac{2\pi }{4\pi G}}
459: \end{eqnarray}
460: and similarly
461: \begin{equation}
462: J=\lim_{r\to\infty}
463: \frac{2\pi }{16\pi G} \int_0^{\pi}
464: \left. \left[
465: \frac{l^{3/2}}{f^2} r^2 \sin^3\theta
466: \left( \omega - r \frac{\partial\omega}{\partial r} \right)
467: \right] \right|_{r} d\theta \ .
468: \label{localJ} \end{equation}
469: Insertion of the asymptotic expansions of the metric functions
470: then yields expressions (\ref{MJ2}).
471:
472: Alternatively, the mass $M$ and the angular momentum $J$
473: can be obtained by direct integration of the expressions
474: (\ref{komarM2}) and (\ref{komarJ2}), where
475: \begin{eqnarray}
476: M
477: &=& \phantom{2}\int_{\Sigma} \left( 2 T_{\e}^{\g}
478: - \d _{\e}^{\g} \, T_{\c}^{\c} \right) \, n_{\g} \, \xi^{\e} dV \ ,
479: \nonumber\\
480: &=&\int \left(2 \, T_t^t -T_{\e}^{\e} \right) \, |g|^{1/2}
481: \, dr \, d\t \, d \varphi \ , \label{tolman}
482: \end{eqnarray}
483: corresponds to the Tolman mass, and
484: \begin{eqnarray}
485: J&=& -
486: \int T_{\, \varphi}^{\, t} \, |g|^{1/2} \, dr \, d\t \, d \varphi \ .
487: \label{ang2}
488: \end{eqnarray}
489:
490: The conserved scalar charge $Q$ is obtained from the time-component
491: of the current,
492: \begin{eqnarray}
493: Q &=- & \int j^t \left| g \right|^{1/2} dr d\t d\varphi
494: \nonumber \\
495: %& = & -i\int \left( \Phi^* \dot{\Phi} - \Phi \dot{\Phi}^* \right)
496: %\left|g \right|^{1/2} dr d\t d\varphi \ ,
497: %\label{ladung} \end{eqnarray}
498: %thus $Q$ vanishes, unless $\Phi$ is time-dependent.
499: %Evaluation yields for the scalar charge
500: %\begin{eqnarray}
501: %Q&=& 4 \pi \omega_s \int_0^{\infty}\int _0^{\pi}
502: &=& 4 \pi \omega_s \int_0^{\infty}\int _0^{\pi}
503: |g| ^{1/2} \frac{1}{f} \left( 1 +
504: \frac{n}{\omega_s}\frac{\omega}{r} \right) \f^2 \,
505: dr \, d\t
506: \ . \label{Qc}
507: \end{eqnarray}
508:
509: From Eq.~(\ref{ang2}) for the angular momentum $J$
510: and Eq.~(\ref{Qc}) for the scalar charge $Q$,
511: one obtains the important quantization relation for the angular momentum,
512: \begin{equation}
513: J=n \, Q \ , \label{JnQ}
514: \end{equation}
515: first derived by Schunck and Mielke \cite{schunck},
516: by taking into account that
517: $T_{\, \varphi}^{\, t} = n j^t$, since
518: $\partial_\varphi \Phi = i n \Phi$.
519: %
520: Thus a boson star with $n=0$ carries no angular momentum, $J =0$.
521:
522: \subsection{Boundary conditions}\label{c2s3}
523:
524: The choice of appropriate boundary conditions must guarantee
525: that the boson star solutions
526: are globally regular and asymptotically flat,
527: and that they possesses finite energy and finite energy density.
528:
529: For rotating axially symmetric boson stars
530: appropriate boundary conditions must be specified
531: for the metric functions $f(r,\theta)$, $l(r,\theta)$, $h(r,\theta)$
532: and $\omega(r,\theta)$,
533: and the scalar field function $\f(r,\theta)$
534: at infinity, at the origin, on the positive $z$-axis ($\theta=0$),
535: and, exploiting the reflection properties w.r.t.~$\theta \rightarrow
536: \pi - \theta$, in the $xy$-plane ($\theta=\pi/2$).
537:
538: For $r \rightarrow \infty$
539: the metric approaches the Minkowski metric $\eta_{\a\b}$
540: and the scalar field assumes its vacuum value $\Phi=0$.
541: Accordingly, we impose at infinity the boundary conditions
542: \begin{equation}
543: f|_{r \rightarrow \infty} =1 \ , \ \ \
544: l|_{r \rightarrow \infty} =1 \ , \ \ \
545: h|_{r \rightarrow \infty} =1 \ , \ \ \
546: \omega|_{r \rightarrow \infty} =0 \ , \ \ \
547: \f| _{r \rightarrow \infty}=0 \ .
548: \label{bc4} \end{equation}
549:
550: At the origin we require
551: \begin{equation}
552: \partial_r f|_{r=0}=0 \ , \ \ \
553: \partial_r l|_{r=0}=0 \ , \ \ \
554: h|_{r=0}=1 \ , \ \ \
555: \omega|_{r=0}=0 \ , \ \ \
556: \f| _{r =0}=0 \ .
557: \label{bc3} \end{equation}
558: Note, that for spherically symmetric boson stars the scalar field
559: has a finite value $\f_0$ at the origin, though.
560:
561: For $\t=0$ and $\t=\pi/2$, respectively,
562: we require the boundary conditions
563: \begin{equation}
564: \partial_{\t} f|_{\t=0}=0 \ , \ \ \
565: \partial_{\t} l|_{\t=0}=0 \ , \ \ \
566: h|_{\t=0}=1 \ , \ \ \
567: \partial_{\t} \omega |_{\t=0}=0 \ , \ \ \
568: \f |_{\t=0}=0 \ ,
569: \label{bc5} \end{equation}
570: and
571: \begin{equation}
572: \partial_{\t} f|_{\t=\pi/2}=0 \ , \ \ \
573: \partial_{\t} l|_{\t=\pi/2}=0 \ , \ \ \
574: \partial_{\t} h|_{\t=\pi/2}=0 \ , \ \ \
575: \partial_{\t} \omega |_{\t=\pi/2}=0 \ , \ \ \
576: \partial_{\t} \f |_{\t=\pi/2}=0 \ ,
577: \label{bc6} \end{equation}
578: for even parity solutions,
579: while for odd parity solutions
580: $\f |_{\t=\pi/2}=0$.
581:
582: In non-rotating solutions, the metric function $\omega$
583: is identically zero.
584: In spherically symmetric solutions
585: the functions depend on the radial coordinate only,
586: with metric function $h=1$.
587: For the flat space solutions the metric corresponds
588: to the Minkowski metric everywhere.
589:
590: \subsection{Numerical method}
591:
592: Rotating solutions are obtained when $n \ne 0$.
593: The resulting set of coupled non-linear partial differential equations
594: is solved numerically \cite{FIDISOL}
595: subject to the above boundary conditions, Eqs.~(\ref{bc4})-(\ref{bc6}).
596: Because of the power law fall-off of the metric functions,
597: we compactify space by introducing the compactified radial coordinate
598: \begin{equation}
599: \bar r = \frac{r}{1+r} \ .
600: \label{rcomp} \end{equation}
601: The numerical calculations are based on the Newton-Raphson method.
602: The equations are discretized on a non-equidistant grid in
603: $\bar r$ and $\theta$.
604: Typical grids used have sizes $90 \times 70$,
605: covering the integration region
606: $0\leq \bar r\leq 1$ and $0\leq\theta\leq\pi/2$.
607:
608: \section{$Q$-balls}\label{c3}
609:
610: \subsection{Spherically symmetric Q-balls}
611:
612: We first briefly review the main features
613: of spherically symmetric $Q$-balls
614: (see e.g.~\cite{lee-s,coleman,volkov}, or \cite{lee-rev} for a review).
615: %to compare their properties with those of rotating axially symmetric $Q$-balls.
616: %
617: The equation of motion for
618: spherically symmetric $Q$-balls is given by
619: \begin{eqnarray}
620: \label{radeom}
621: 0 & = & \phi'' + \frac{2} {r} \, \phi' -
622: \frac{1}{2} \frac{d U(\phi)} {d\phi} +
623: \omega_s^2 \, \phi \ .
624: \end{eqnarray}
625: It may be interpreted to
626: effectively describe a particle moving with friction in the
627: effective potential
628: \begin{eqnarray}
629: V(\phi) = \frac{1}{2} \, \omega_s^2 \, \phi^2
630: - \frac{1}{2} U(\phi) \ ,
631: \label{Vpot}
632: \end{eqnarray}
633: exhibited in Fig.~\ref{potentiale}.
634:
635: $Q$-balls exist in the frequency range
636: \begin{equation}
637: \omega_{\rm min}^2 < \omega_s^2 < \omega_{\rm max}^2 \ ,
638: \label{olimits}
639: \end{equation}
640: where the maximal frequency $\omega_{\rm max}$
641: and the minimal frequency $\omega_{\rm min}$ are determined
642: by the properties of the potential $U$ \cite{volkov},
643: \begin{eqnarray}
644: \label{cond1}
645: \omega_s^2 & < & \omega^{2}_{\rm max} \; \equiv \;
646: \frac{1}{2} U''(0) = \lambda \, b \ ,
647: \end{eqnarray}
648: \begin{eqnarray}
649: \label{cond2}
650: \omega_s^2 & > & \omega^{2}_{\rm min} \; \equiv \;
651: \min_{\f} \left[{U(\phi)}/{\phi^2} \right] \; = \;
652: \lambda \left(b- \frac{a^2}{4} \right) \ .
653: \end{eqnarray}
654:
655: The mass $M$ of the $Q$-balls
656: can be obtained from
657: \begin{eqnarray}
658: \label{radenergy}
659: M
660: %& = & 4 \pi \int_0^\infty T_{tt} r^2 \, d r \
661: =
662: 4 \pi \int_0^\infty \left[\omega_s^2 \, \phi^2 +
663: \phi^{\, \prime \, 2} + U(\phi)\right] r^2 \, d r \ ,
664: \end{eqnarray}
665: where the prime denotes differentiation with respect to $r$, and
666: and their charge $Q$ from
667: \begin{eqnarray}
668: \label{charge}
669: Q(\omega_s) & = & 8 \pi \, \omega_s \int_0^\infty \phi^2 \, r^2 \, d r \ .
670: \end{eqnarray}
671:
672: Fixing the value of $\omega_s$ in the allowed range,
673: one then obtains a sequence of $Q$-ball solutions,
674: consisting of the fundamental $Q$-ball and its
675: radial excitations \cite{volkov}.
676: The boson function $\f$ of the fundamental $Q$-ball has no nodes,
677: while it has $k$ nodes for the $k$-th radial excitation.
678:
679: We exhibit the fundamental $Q$-ball solution ($k=0$) and the first
680: radial excitation ($k=1$) in Fig.~\ref{Qvsomega0}.
681: Here the charge $Q$ is shown versus the frequency $\omega_s$.
682: As seen in the figure, at a critical value
683: $\omega_{\rm cr}$, which depends on $k$,
684: the charge assumes its respective minimal value $Q_{\rm cr}$.
685: Thus two branches of solutions exist for each $k$,
686: the lower branch $\omega_s < \omega_{\rm cr}$
687: and the upper branch $\omega_s > \omega_{\rm cr}$,
688: which merge and end at $\omega_{\rm cr}$.
689: When $\omega_s \rightarrow \omega_{\rm min}$
690: as well as when $\omega_s \rightarrow \omega_{\rm max}$
691: the charge diverges \cite{lee-s}.
692:
693: In Fig.~\ref{Qvsomega0} also the mass $M$ of these two sets of solutions
694: is shown, but it is now exhibited versus the charge $Q$.
695: The lower mass branches correspond to the lower values of the frequency,
696: $\omega_s < \omega_{\rm cr}$, while the upper mass branches
697: correspond to higher values of the frequency, $\omega_s > \omega_{\rm cr}$.
698: As long as the mass is smaller than the mass of $Q$ free bosons,
699: $M<m_{\rm B}Q$, the fundamental $Q$-ball solutions are stable,
700: both classically and quantum mechanically \cite{lee-s}.
701: Classically the solutions remain stable all along the lower branch.
702: The classical stability changes only at the critical point, $Q_{\rm cr}$,
703: where the solutions acquire an unstable mode \cite{lee-s}.
704: In contrast, from a quantum perspective the stability changes
705: when the lower mass branch intersects the line corresponding to
706: the mass of $Q$ free bosons.
707: On the unstable upper branch
708: the mass approaches the mass $M=m_{\rm B}Q$ of $Q$ free bosons from above,
709: as $\omega_s \rightarrow \omega_{\rm max}$ \cite{lee-s}.
710:
711: \begin{figure}[h!]
712: \parbox{\textwidth}
713: {\centerline{
714: \mbox{
715: \epsfysize=10.0cm
716: \includegraphics[width=70mm,angle=0,keepaspectratio]{F2a.eps}
717: \includegraphics[width=75mm,angle=0,keepaspectratio]{F2b.eps}
718: }}}
719: \caption{
720: %2
721: Left: The charge $Q$ versus the frequency $\omega_s$
722: for spherically symmetric fundamental $Q$-balls ($k=0$)
723: and their first radial excitations ($k=1$).
724: Also shown are the limiting values of the frequency, $\omega_{\rm min}$
725: and $\omega_{\rm max}$.
726: %
727: Right: The mass $M$ versus the charge $Q$
728: for spherically symmetric fundamental $Q$-balls ($k=0$)
729: and their first radial excitations ($k=1$).
730: Also shown is the mass for $Q$ free bosons, $M=m_{\rm B}Q$.
731: The mass on the upper branches is very close this mass.
732: The $k=0$ branches close to the critical
733: charge $Q_{\rm cr}$ are seen in the inset.
734: }
735: \label{Qvsomega0}
736: \end{figure}
737:
738: We remark, that also angularly excited
739: non-rotating $Q$-balls have been obtained recently \cite{Brihaye:2007tn}.
740: These new solutions are axially symmetric,
741: and their angular dependence is given by spherical harmonics.
742: They have higher energy and charge than the spherically
743: symmetric $Q$-balls, but lower energy and charge than the rotating $Q$-balls.
744: In addition, interacting Q-balls have been studied \cite{Brihaye:2007tn}.
745:
746: \boldmath
747: \subsection{Rotating $Q$-balls}
748: \unboldmath
749:
750: The existence of rotating $Q$-balls was shown by
751: Volkov and W\"ohnert \cite{volkov}.
752: Based on the Ansatz (\ref{ansatzp}) for the scalar field $\Phi$ \cite{schunck},
753: rotating $Q$-balls are solutions of the field equation
754: \begin{eqnarray}
755: \label{eom3d}
756: \left( \frac{\partial^2} {\partial r^2} + \frac{2} {r} \frac{\partial}
757: {\partial r} + \frac{1} {r^2} \frac{\partial^2} {\partial \theta^2}
758: + \frac{\cos\theta} {r^2 \sin\theta} \frac{\partial} {\partial
759: \theta} - \frac{n^2} {r^2 \sin^2\theta} + \omega_s^2 \right) \phi &
760: = & \frac{1}{2} \frac{d U(\phi)}{d \phi} \ , \phantom{abcd}
761: \end{eqnarray}
762: with mass \cite{volkov}
763: \begin{equation}
764: \label{erot}
765: M = 2 \pi \int_0^\infty \int_0^\pi
766: \left(\omega_s^2 \phi^2 + (\partial_r \phi)^2 +
767: \frac{1} {r^2} (\partial_\theta \phi)^2 + \frac{n^2\phi^2} {r^2
768: \sin^2 \theta} + U(\phi) \right)
769: r^2 \, d r \, \sin\theta\, d \theta
770: \ ,
771: \end{equation}
772: and charge
773: \begin{equation}
774: Q= 4 \pi \omega_s \int_0^{\infty}\int _0^{\pi}
775: \f^2 \, r^2 dr \, \sin \theta \, d\t \
776: \ . \label{3} \end{equation}
777: Their angular momentum satisfies the quantization relation
778: $J=nQ$.
779:
780: In their pioneering study \cite{volkov}
781: Volkov and W\"ohnert
782: showed, that for a given value of $n$ there are two types of solutions,
783: positive parity $Q$-balls $n^+$ and negative parity $Q$-balls $n^-$.
784: In the following we restrict our considerations
785: to fundamental rotating $Q$-balls.
786:
787: \subsubsection{Positive parity $Q$-balls}
788:
789: \begin{figure}[t!]
790: \parbox{\textwidth}
791: {\centerline{
792: \mbox{
793: \epsfysize=10.0cm
794: \includegraphics[width=70mm,angle=0,keepaspectratio]{Fig3phi_n1.eps}
795: \includegraphics[width=70mm,angle=0,keepaspectratio]{Fig3eps_n1.eps}
796: }}}
797: {\centerline{
798: \mbox{
799: \epsfysize=10.0cm
800: \includegraphics[width=70mm,angle=0,keepaspectratio]{Fig3phi_n2.eps}
801: \includegraphics[width=70mm,angle=0,keepaspectratio]{Fig3eps_n2.eps}
802: }}}
803: {\centerline{
804: \mbox{
805: \epsfysize=10.0cm
806: \includegraphics[width=70mm,angle=0,keepaspectratio]{Fig3phi_n3.eps}
807: \includegraphics[width=70mm,angle=0,keepaspectratio]{Fig3eps_n3.eps}
808: }}}
809: \caption{
810: %3
811: The scalar field $\f$ (left column)
812: and energy density $T_{tt}$ (right column)
813: for rotating positive parity $Q$-balls
814: with charge $Q=410$ and %quantum numbers
815: $n^P=1^+$ (upper row), $2^+$ (middle row) and $3^+$ (lower row)
816: versus the coordinates $\rho=r \sin \theta$
817: and $z= r \cos\theta$ for $z\ge 0$.
818: }
819: \label{QT410_n2}
820: \end{figure}
821:
822: We first recall the properties of
823: rotating $Q$-balls with positive parity
824: and, at the same time, extend the previous studies \cite{volkov,list}
825: to higher rotational quantum number $n$.
826: In Fig.~\ref{QT410_n2}
827: we illustrate the scalar field $\f$ and the energy density
828: $T_{tt}$ of rotating $Q$-balls with charge $Q=410$,
829: %quantum numbers
830: $n^P=1^+$, $2^+$ and $3^+$.
831: For rotating $Q$-balls the scalar field $\f$ vanishes on the
832: $z$-axis.
833: The energy density $T_{tt}$ of even parity rotating $Q$-balls
834: typically exhibits a torus-like structure.
835: This torus-like shape becomes apparent by
836: considering surfaces of constant energy density
837: of e.g.~half the respective maximal value of the energy density.
838: As seen in the figure, the maximum of the energy density
839: increases and shifts towards
840: larger values of the coordinate $\rho=r \sin \theta$,
841: when $n$ and thus the angular momentum increases.
842: In terms of the energy density tori this manifests as an
843: increase of their radii in the equatorial plane with increasing $n$.
844: (Note, that in \cite{list} a weighted energy density was shown.)
845: %i.e., the torus increases its size with increasing $n$.
846: %The mass $M$, the angular momentum $J$ and the frequency $\omega_s$ of these ro tating $Q$-balls are exhibited in Table \ref{table1}.
847:
848: We next address the frequency dependence for these
849: rotating $Q$-ball solutions with positive parity.
850: In Fig.~\ref{rot-flat} we exhibit the charge $Q$
851: versus the frequency $\omega_s$
852: for $Q$-balls with $n=0-3$.
853: We observe the same upper limiting value $\omega_{\rm max}$, Eq.~(\ref{cond1}),
854: for the frequency $\omega_s$, as for non-rotating $Q$-balls,
855: ensuring asymptotically an exponential fall-off of the scalar field $\f$.
856: For a given frequency $\omega_s$ the charge of a
857: $Q$-ball increases with $n$.
858: We thus conclude, that the frequency of rotating $Q$-balls
859: is limited by the minimal frequency
860: $\omega_{\rm min}$, Eq.~(\ref{cond2}), as well,
861: independent of $n$.
862:
863: \begin{figure}[h!]
864: \parbox{\textwidth}
865: {\centerline{
866: \mbox{
867: \epsfysize=10.0cm
868: \includegraphics[width=70mm,angle=0,keepaspectratio]{F4a.eps}
869: \includegraphics[width=70mm,angle=0,keepaspectratio]{F4b.eps}
870: }}}
871: \caption{
872: %4
873: Left: The charge $Q$ versus the frequency $\omega_s$
874: for $Q$-balls with %quantum numbers
875: $n^P=0^+$, $1^+$, $2^+$, $3^+$.
876: The asterisks mark the critical values of the charge.
877: Right: The mass $M$ versus the charge $Q$
878: for $Q$-balls with %quantum numbers
879: $n^P=0^+$, $1^+$, $2^+$, $3^+$.
880: The upper branches of the mass $M$ are hardly discernible (on this scale)
881: from the mass of $Q$ free bosons, $M=m_{\rm B}Q$.
882: }
883: \label{rot-flat}
884: \end{figure}
885:
886: The mass of these rotating $Q$-balls with positive parity
887: shows the same critical behaviour and thus cusp structure
888: as the mass of the non-rotating $Q$-balls.
889: This is also illustrated in Fig.~\ref{rot-flat}.
890: Thus we conclude, that these rotating solutions are classically stable
891: along their lower mass branches.
892: Overall we note, that the sets of rotating $Q$-balls with positive parity
893: exhibit the same general pattern as the sets of non-rotating $Q$-balls.
894:
895:
896: \subsubsection{Negative parity $Q$-balls}
897:
898: Turning to rotating $Q$-balls with negative parity,
899: we first illustrate in Fig.~\ref{QT410_n3}
900: the scalar field $\f$ and the energy density
901: $T_{tt}$ of $Q$-balls with
902: fixed charge $Q=720$
903: and %quantum numbers
904: $n^P=1^-$ and $2^-$.
905: For these $Q$-balls the scalar field $\f$ vanishes not only
906: on the $z$-axis but everywhere in the $xy$-plane, as well.
907: The energy density $T_{tt}$ now typically
908: exhibits a double torus-like structure,
909: where the two tori are located symmetrically
910: w.r.t.~the equatorial plane
911: \cite{volkov}.
912: When $n$ and thus the angular momentum increases,
913: the maxima of the energy density again shift towards
914: larger values of the coordinate $\rho=r \sin \theta$.
915: For the double tori of the energy density this shift
916: with increasing $n$ manifests as an
917: increase of the radii of the tori in the equatorial plane.
918:
919: \begin{figure}[h!]
920: \parbox{\textwidth}
921: {\centerline{
922: \mbox{
923: \epsfysize=10.0cm
924: \includegraphics[width=70mm,angle=0,keepaspectratio]{Fig5phi_n1.eps}
925: \includegraphics[width=70mm,angle=0,keepaspectratio]{Fig5eps_n1.eps}
926: }}}
927: {\centerline{
928: \mbox{
929: \epsfysize=10.0cm
930: \includegraphics[width=70mm,angle=0,keepaspectratio]{Fig5phi_n2.eps}
931: \includegraphics[width=70mm,angle=0,keepaspectratio]{Fig5eps_n2.eps}
932: }}}
933: \caption{
934: %5
935: The scalar field $\f$ (left column)
936: and energy density $T_{tt}$ (right column)
937: for rotating negative parity $Q$-balls
938: with charge $Q=720$ and %quantum numbers
939: $n^P=1^-$ (upper row) and $2^-$ (lower row)
940: versus the coordinates $\rho=r \sin \theta$
941: and $z= r \cos\theta$ for $z\ge 0$.
942: }
943: \label{QT410_n3}
944: \end{figure}
945:
946: In Fig.~\ref{rot-flat2} we exhibit the charge $Q$
947: %and the mass $M$
948: versus the frequency $\omega_s$
949: for negative parity $Q$-balls with %quantum numbers
950: $n^P=1^-$ and $2^-$, and compare to
951: the positive parity $Q$-balls.
952: While we conclude, that
953: for the negative parity $Q$-balls
954: the limiting values $\omega_{\rm min}$ and $\omega_{\rm max}$
955: are retained,
956: numerical difficulties have unfortunately prevented a more complete study.
957: %As before, the condition $\omega_s < \omega_{\rm max}$
958: %ensures an asymptotically exponential fall-off of the scalar field $\f$.
959:
960: \begin{figure}[h!]
961: \parbox{\textwidth}
962: {\centerline{
963: \mbox{
964: \epsfysize=10.0cm
965: \includegraphics[width=70mm,angle=0,keepaspectratio]{F6a.eps}
966: \includegraphics[width=70mm,angle=0,keepaspectratio]{F6b.eps}
967: }}}
968: \caption{
969: %6
970: Left: The charge $Q$ versus the frequency $\omega_s$
971: for negative parity $Q$-balls with $n^P=1^-$ and $2^-$.
972: For comparison the positive parity $Q$-balls
973: with $n^P=0^+$, $1^+$, $2^+$, $3^+$ are also shown.
974: Right: The mass $M$ versus the charge $Q$
975: for the same set of solutions together with
976: the mass of $Q$ free bosons, $M=m_{\rm B}Q$.
977: }
978: \label{rot-flat2}
979: \end{figure}
980:
981: Comparing the sets of negative and positive parity $Q$-balls,
982: we conclude,
983: %observe,
984: that they exhibit the same general pattern.
985: However,
986: concerning the magnitudes of their charges and masses,
987: the positive parity $Q$-balls and the negative parity $Q$-balls
988: do not quite alternate.
989: Instead, they are ordered according to
990: $n^P=0^+,\, 1^+,\, 2^+,\, 1^-,\, 3^+,\, 2^-$, as seen in
991: Fig.~\ref{rot-flat2}.
992: Thus for $Q$-balls of a given charge $Q$,
993: the double torus-like structure of the negative parity solutions
994: leads to a considerably higher mass
995: than the single torus-like structure of the positive parity solutions.
996:
997: \section{Boson stars}
998:
999: \subsection{Spherically symmetric boson stars}
1000:
1001: When the scalar field is coupled to gravity, boson stars arise
1002: (see e.g.~\cite{jetzer,schunck2} for reviews).
1003: %
1004: We first demonstrate the effects of gravity for
1005: spherically symmetric boson stars in Fig.~\ref{qvsos_col},
1006: where we exhibit the charge $Q$ versus the frequency $\omega_s$
1007: for several values of the gravitational coupling constant $\kappa$.
1008:
1009: \begin{figure}[h!]
1010: \parbox{\textwidth}
1011: {\centerline{
1012: \mbox{
1013: \epsfysize=10.0cm
1014: \includegraphics[width=80mm,angle=0,keepaspectratio]{F7a.eps}
1015: \includegraphics[width=68mm,angle=0,keepaspectratio]{F7b.eps}
1016: }}}
1017: \caption{
1018: %7
1019: Left: The charge $Q$ versus the frequency $\omega_s$
1020: for fundamental spherically symmetric boson stars ($k=0$)
1021: in the full range of existence
1022: for a sequence of values of the gravitational coupling constant $\kappa$.
1023: Right:
1024: The charge $Q$ versus the frequency $\omega_s$
1025: in the frequency range of the spiral
1026: for the gravitational coupling constant $\kappa=0.2$.
1027: }
1028: \label{qvsos_col}
1029: \end{figure}
1030:
1031: For solutions in curved space the frequency $\omega_s$
1032: is also bounded from above by
1033: $\omega_{\rm max}$, Eq.~(\ref{cond1}),
1034: ensuring an asymptotically exponential fall-off of the scalar field.
1035: But, unlike for flat space solutions,
1036: the charge $Q$ does not diverge for boson stars,
1037: when $\omega_s$ approaches $\omega_{\rm max}$.
1038: Instead the charge tends to zero
1039: in the limit $\omega_s \rightarrow \omega_{\rm max}$.
1040:
1041: Also, for the smaller values of $\omega_s$
1042: the solutions approach the limiting lower value
1043: $\omega_{\rm min}$, Eq.~(\ref{cond2}),
1044: no longer monotonically.
1045: Instead a spiral-like behaviour arises in the presence of gravity,
1046: where the boson star solutions exhibit an inspiralling
1047: towards a limiting solution \cite{lee-bs}.
1048: The location and size of the spiral depend on the gravitational
1049: coupling strength $\kappa$,
1050: and can be well below $\omega_{\rm min}$
1051: for small values of $\kappa$
1052: (see Fig.~\ref{qvsos_col})
1053: \cite{list}.
1054:
1055: \begin{figure}[h!]
1056: \parbox{\textwidth}
1057: {\centerline{
1058: \mbox{
1059: \epsfysize=10.0cm
1060: \includegraphics[width=70mm,angle=0,keepaspectratio]{F8a.eps}
1061: \includegraphics[width=75mm,angle=0,keepaspectratio]{F8b.eps}
1062: }}}
1063: \caption{
1064: %8
1065: Left: The scaled charge $\kappa Q$
1066: versus the frequency $\omega_s$
1067: for fundamental spherically symmetric boson stars
1068: in the limit $\kappa \rightarrow \infty$.
1069: For comparison the scaled charge $\kappa Q$
1070: is also shown for $\kappa = 0.2$.
1071: Right: The mass $M$ versus the charge $Q$
1072: for fundamental spherically symmetric boson stars
1073: for the value of the gravitational coupling constant
1074: $\kappa=0.2$
1075: }
1076: \label{mvsq_col}
1077: \end{figure}
1078:
1079: The limit $\kappa \rightarrow \infty$
1080: is obtained by introducing the scaled
1081: scalar field $\hat{\phi}(r)=\sqrt{\kappa} \phi(r)$ \cite{list}.
1082: Substituting $\phi(r)$ in the field equation and taking the
1083: limit $\kappa \rightarrow \infty$, the terms non-linear
1084: in $\hat{\phi}(r)$ vanish. Similarly, in the Einstein equations
1085: the terms of higher than second order in $\hat{\phi}(r)$ vanish and
1086: the dependence on $\kappa$ cancels.
1087: The resulting set of differential equations is then identical to
1088: the original one, except that $\kappa=1$ and
1089: $U(\hat{\phi})= \lambda b \hat{\phi}^2$,
1090: reducing the potential to a mass term.
1091: This also yields a spiral pattern for the
1092: (scaled) charge $\hat{Q} = \kappa Q$, as seen in Fig.~\ref{mvsq_col}.
1093:
1094: The mass $M$ has an analogous dependence on the frequency $\omega_s$
1095: as the charge $Q$.
1096: When the mass $M$ is considered as a function of the charge $Q$, however,
1097: one now observes a new intricate cusp structure \cite{lee-bs,list}.
1098: As gravity is weekly coupled,
1099: % (e.g.,~$\kappa=0.2$)
1100: the single cusp present in flat space
1101: (associated with the minimal value $Q_{\rm min}$ of the charge)
1102: is retained. But it is supplemented by a second cusp,
1103: since in the presence of gravity the charge and the mass
1104: do not diverge for $\omega_s \rightarrow \omega_{\rm max}$,
1105: but tend to zero in this limit.
1106: %Along the branch, leading from the second cusp zero, then
1107: %boson stars are bound again, $M<m_{\rm B}Q$.
1108: For larger values of the gravitational coupling $\kappa$ %(e.g.,~$\kappa=1$)
1109: these two cusps merge and disappear.
1110: However, an additional set of cusps arises in the presence of gravity,
1111: which is due to the appearance of the spirals,
1112: and therefore only present in curved space \cite{lee-bs,list}.
1113: The cusp structure is illustrated in Fig.~\ref{mvsq_col}.
1114:
1115: The classical stability of boson stars can be analyzed according
1116: to catastrophe theory,
1117: implying a change of classical stability at each cusp encountered
1118: \cite{catastrophe}.
1119: In the following we refer to the main branch of solutions
1120: as the branch between
1121: the maximum of the charge and the mass
1122: (at a relatively small value of the frequency $\omega_s$)
1123: and the
1124: %(for small gravitational coupling only local)
1125: local minimum of the charge and the mass
1126: (at a relatively large value of the frequency $\omega_s$).
1127: Note, that for the flat space solutions the minimum is a global minimum,
1128: the charge and the mass assume their critical values here,
1129: while for large gravitational coupling $\kappa$ the local minimum
1130: has disappeared, and the relevant minimum corresponds to the global
1131: minimum at $\omega_{\rm max}$.
1132:
1133: Along this main branch the boson star solutions are classically stable.
1134: We note though, that a small part of this main branch
1135: may become quantum mechanically unstable,
1136: when $M>m_{\rm B}Q$ close to the first cusp.
1137: Since at each cusp terminating the main branch a negative mode is acquired,
1138: the solutions on the adjacent branches become classically unstable.
1139: In the spiral, the boson stars are expected to acquire
1140: at each further cusp an additional negative mode,
1141: making the solutions increasingly unstable.
1142: The physically most relevant set of solutions
1143: should thus correspond
1144: to the stable boson star solutions on the main branch.
1145:
1146: \boldmath
1147: \subsection{Rotating Boson Stars}\label{c4}
1148: \unboldmath
1149:
1150: We now turn to rotating boson stars,
1151: which emerge from rotating $Q$-balls,
1152: when the gravitational coupling constant is increased from zero.
1153: We first recall the previous results
1154: on rotating positive parity boson stars with $n^P=1^+$
1155: \cite{schunck,japan,list}
1156: and present new results for $n^P=2^+$.
1157: We then turn to rotating negative parity boson stars.
1158: For both cases we demonstrate and analyze the occurrence of ergoregions.
1159: We here restrict our analysis to fundamental rotating boson stars.
1160:
1161: \subsubsection{Positive parity boson stars}
1162:
1163: We begin the discussion by recalling the main results on
1164: boson stars with $n^P=1^+$, obtained previously \cite{list}.
1165: For that purpose we exhibit in Fig.~\ref{Qlimrot}
1166: the charge $Q$ of $1^+$ boson stars versus the frequency $\omega_s$
1167: for several values of the gravitational coupling constant $\kappa$,
1168: including the flat space limit.
1169:
1170: \begin{figure}[h!]
1171: \parbox{\textwidth}
1172: {\centerline{
1173: \mbox{
1174: \epsfysize=10.0cm
1175: \includegraphics[width=70mm,angle=0,keepaspectratio]{F9a.eps}
1176: \includegraphics[width=70mm,angle=0,keepaspectratio]{F9b.eps}
1177: }}}
1178: \caption{
1179: %9
1180: Left:
1181: The charge $Q$ versus the frequency $\omega_s$
1182: for rotating $1^+$ boson stars
1183: for several values of the gravitational coupling constant
1184: $\kappa$. Also shown are the limiting flat space values.
1185: The dots indicate the onset of ergoregions.
1186: Right:
1187: The mass $M$ versus the frequency $\omega_s$
1188: for the same set of solutions.
1189: }
1190: \label{Qlimrot}
1191: \end{figure}
1192:
1193: Again, the frequency $\omega_s$ of the solutions
1194: is bounded from above by
1195: $\omega_{\rm max}$, Eq.~(\ref{cond1}),
1196: ensuring an asymptotically exponential fall-off
1197: of the scalar field.
1198: Also, the charge of the rotating boson stars tends to zero,
1199: when $\omega_{\rm max}$ is approached.
1200: Likewise, $1^+$ boson stars
1201: exist only in a finite interval of the frequency $\omega_s$,
1202: where the lower bound increases with increasing $\kappa$.
1203:
1204: Furthermore, we observe the presence of spirals also
1205: for these rotating boson stars,
1206: i.e., the $1^+$ boson stars exhibit an inspiralling
1207: towards a limiting solution just like the non-rotating boson stars.
1208: For small values of the coupling constant $\kappa$ the spirals are
1209: located in the region of small frequency $\omega_s$,
1210: for larger coupling $\kappa$ the spirals shift towards larger frequencies.
1211: A good and rather complete determination of the spirals is, however,
1212: numerically often extremely difficult \cite{list}.
1213:
1214: The mass $M$ of the $1^+$ boson stars
1215: has an analogous dependence on the frequency $\omega_s$
1216: as the charge $Q$, as seen in Fig.~\ref{Qrotbs}.
1217: When the mass $M$ of these rotating boson stars
1218: is considered as a function of the charge $Q$,
1219: we observe an analogous cusp structure as for the non-rotating boson stars
1220: \cite{list}.
1221: In particular, for a given gravitational coupling $\kappa$,
1222: a main branch of rotating boson star solutions
1223: is present, and located between
1224: the maximum of the charge and the mass
1225: (at a relatively small value of the frequency $\omega_s$)
1226: and the (for small gravitational coupling only local)
1227: minimum of the charge and the mass
1228: (which is not part of the spiral).
1229: Following arguments from catastrophe theory,
1230: the boson star solutions are again
1231: expected to be classically stable along this main branch.
1232:
1233: Let us now turn to positive parity boson stars with higher
1234: quantum number $n$ and thus greater angular momentum.
1235: We exhibit in Fig.~\ref{Qlimrot2}
1236: the charge $Q$ and the mass $M$
1237: of $2^+$ boson stars versus the frequency $\omega_s$
1238: for several values of the gravitational coupling constant $\kappa$,
1239: including the flat space limit.
1240: Because these calculations are numerically very involved
1241: and time-consuming, we have concentrated on the main branches
1242: of these $2^+$ boson stars, omitting the numerically even more involved
1243: spirals.
1244: Clearly, the $2^+$ solutions show a completely analogous
1245: $\kappa$- and frequency dependence (in the regions studied)
1246: as the $1^+$ solutions.
1247: %and are expected to possess also an analogous spiral structure.
1248:
1249: \begin{figure}[h!]
1250: \parbox{\textwidth}
1251: {\centerline{
1252: \mbox{
1253: \epsfysize=10.0cm
1254: \includegraphics[width=70mm,angle=0,keepaspectratio]{F10a.eps}
1255: \includegraphics[width=70mm,angle=0,keepaspectratio]{F10b.eps}
1256: }}}
1257: \caption{
1258: %10
1259: Left:
1260: The charge $Q$ versus the frequency $\omega_s$
1261: for rotating $2^+$ boson stars
1262: for several values of the gravitational coupling constant
1263: $\kappa$. Also shown are the limiting flat space values.
1264: The dots indicate the onset of ergoregions.
1265: Right:
1266: The mass $M$ versus the frequency $\omega_s$
1267: for the same set of solutions.
1268: }
1269: \label{Qlimrot2}
1270: \end{figure}
1271:
1272: To further demonstrate the dependence of the positive parity boson stars
1273: on the rotational quantum number $n$, we exhibit in Fig.~\ref{Qrotbs}
1274: the charge $Q$ and the mass $M$ as functions of the
1275: frequency $\omega_s$ for boson stars with
1276: quantum numbers $n^P=0^+$, $1^+$, $2^+$
1277: %and gravitational coupling $\kappa=0.01$ and $0.1$.
1278: and gravitational coupling $\kappa=0.2$.
1279:
1280: \begin{figure}[h!]
1281: \parbox{\textwidth}
1282: {\centerline{
1283: \mbox{
1284: \epsfysize=10.0cm
1285: \includegraphics[width=70mm,angle=0,keepaspectratio]{F11a.eps}
1286: \includegraphics[width=70mm,angle=0,keepaspectratio]{F11b.eps}
1287: }}}
1288: \caption{
1289: %11
1290: The charge $Q$ (left column) and the mass $M$ (right column)
1291: versus the frequency $\omega_s$
1292: for boson stars with $n^P=0^+$, $1^+$, $2^+$
1293: for the gravitational coupling $\kappa=0.2$.% (upper row)
1294: %and $\kappa=0.2$ (lower row).
1295: %The fat lines indicate the presence of ergoregions.
1296: }
1297: \label{Qrotbs}
1298: \end{figure}
1299:
1300: Considering the spatial structure of the solutions,
1301: the scalar field $\f$ typically gives rise to
1302: a torus-like energy density $T_{tt}$
1303: for the rotating $1^+$ and $2^+$ boson stars,
1304: just as for the rotating $n^+$ $Q$-balls.
1305: Again, the torus-like shape becomes apparent by
1306: considering surfaces of constant energy density
1307: of e.g.~half the respective maximal value of the energy density.
1308: As in the flat space solutions, the maximum of the energy density
1309: of the boson stars increases and shifts towards
1310: larger values of the coordinate $\rho=r \sin \theta$,
1311: when $n$ and thus the angular momentum increases.
1312: Thus in terms of the energy density tori this again manifests as an
1313: increase of their radii in the equatorial plane with increasing $n$.
1314:
1315: Rotating objects may possess ergoregions.
1316: But unlike the ergoregions of black holes, the ergoregions of
1317: regular objects like boson stars
1318: would signal the presence of an instability
1319: \cite{Cardoso,ergo-paper1,ergo-paper2,ergo-paper3}.
1320: This instability can be traced back to superradiant scattering,
1321: and its possible presence was put forward by Cardoso et al.~\cite{Cardoso}
1322: as an argument against various black hole doubles
1323: as potential horizonless candidates for compact dark astrophysical objects.
1324:
1325: Therefore we now analyze the occurrence of ergoregions for
1326: rotating positive parity boson stars.
1327: The ergosurface is defined by the condition
1328: \begin{equation}
1329: g_{tt}= 0 = -f + \frac{l}{f} \sin^2 \theta \omega^2 \
1330: \label{ergo-c}
1331: \end{equation}
1332: in the metric parametrization Eq.~(\ref{ansatzg}).
1333: The ergoregion resides inside this ergosurface.
1334: The presence of an ergosurface thus implies instability
1335: of the respective boson stars \cite{Cardoso}.
1336:
1337: Examining the condition $g_{tt}= 0$ for the sets of boson star solutions,
1338: we indeed observe that rotating boson stars possess ergoregions
1339: in a large part of their domain of existence.
1340: To discuss these ergoregions, let us recall Figs.~\ref{Qlimrot} and
1341: \ref{Qlimrot2}, where the charge $Q$ is exhibited versus the
1342: frequency $\omega_s$ for several sets of boson stars solutions with
1343: $n^P=1^+$ and $2^+$, respectively.
1344: The dots in these figures indicate the onset of ergoregions
1345: and thus ergoregion related instability
1346: at a corresponding critical frequency $\omega_s$.
1347: The solutions to the right of the dots do not possess an ergoregion,
1348: whereas the solutions to the left of the dots as well as in the spirals
1349: do possess an ergoregion.
1350:
1351: The onset of ergoregions thus almost always occurs on the main branch
1352: of boson star solutions, supposed to be classically stable.
1353: The presence of ergoregions then tends to diminish the physically
1354: relevant - since stable - set of solutions.
1355: As seen from the figures, the diminishment is greater
1356: for larger gravitational coupling and larger rotational quantum number $n$,
1357: and thus greater angular momentum.
1358: However, there always remains a part of the main branch
1359: of classically stable boson star solutions,
1360: not suffering from an ergoregion instability.
1361:
1362: Let us now discuss the emergence and structure of ergoregions
1363: in more detail for these boson stars.
1364: For that purpose we exhibit the ergoregions
1365: of $1^+$ boson stars at $\kappa=0.1$
1366: and $2^+$ boson stars at $\kappa=0.1$ and $0.2$
1367: in Fig.~\ref{ergorot}.
1368:
1369: \begin{figure}[h!]
1370: \parbox{\textwidth}
1371: {\centerline{
1372: \mbox{
1373: \epsfysize=10.0cm
1374: \includegraphics[width=75mm,angle=0,keepaspectratio]{F12a.eps}
1375: \epsfysize=10.0cm
1376: \includegraphics[width=75mm,angle=0,keepaspectratio]{F12b.eps}
1377: }}}
1378: {\centerline{
1379: \mbox{
1380: \epsfysize=10.0cm
1381: \includegraphics[width=75mm,angle=0,keepaspectratio]{F12c.eps}
1382: \epsfysize=10.0cm
1383: \includegraphics[width=75mm,angle=0,keepaspectratio]{F12d.eps}
1384: }}}
1385: \caption{
1386: %12
1387: Location and shape of the axially symmetric ergoregions
1388: of boson stars exhibited via cross-sections with the $xz$-plane:
1389: $1^+$ boson stars at $\kappa=0.1$ (upper row)
1390: and $2^+$ boson stars (lower row)
1391: at $\kappa=0.1$ (left) and at $\kappa=0.2$ (right).
1392: The numbers in the legend indicate the respective values of the charge $Q$.
1393: }
1394: \label{ergorot}
1395: \end{figure}
1396:
1397: Beginning with the $1^+$ boson stars, we observe, that
1398: at the critical frequency (where $Q\approx 986$ for $\kappa=0.1$)
1399: the condition (\ref{ergo-c})
1400: is satisfied precisely on a circle in the equatorial plane.
1401: %centered at the origin.
1402: In the (upper left) figure, only the two points of intersection of this
1403: circle with the $xz$-plane are seen.
1404: As $Q$ is increased (i.e., $\omega_s$ is decreased)
1405: the circle turns into a torus-like surface.
1406: In the figure,
1407: its intersection with the $xz$-plane is then seen as two small circles
1408: ($Q=911$).
1409: From this approximately maximal value of the charge,
1410: the set of solutions now enters the first part of the spiral,
1411: where stability is lost anyway.
1412:
1413: As the solutions evolve into the spiral the charge decreases.
1414: But the ergoregion continues to grow in size, while it remains
1415: (approximately) centered at the location, where it first arose.
1416: The ergoregion then tends towards its largest size
1417: ($Q \approx 800$), while the center of the torus of the ergoregion
1418: starts to shift inwards.
1419: Interestingly, the ergoregion is not only topologically a torus,
1420: but also geometrically almost a torus.
1421:
1422: As seen in the (upper right) figure,
1423: with further decreasing charge $Q$,
1424: the ergoregion then starts to decrease in size again,
1425: when the solutions evolve deeper into the spiral.
1426: At the same time, the center of the torus of the ergoregion
1427: continues to move further inwards.
1428: Thus the innermost part of the ergosurface gets increasingly
1429: closer to the origin, but never touches it.
1430:
1431: The ergoregions of the set of boson stars at different values
1432: of the gravitational coupling constant $\kappa$
1433: exhibit an analogous behaviour.
1434: Likewise,
1435: the ergoregions of the sets of boson stars at higher $n$
1436: exhibit an analogous behaviour in the regions studied.
1437: As seen in the (lower part of the) figure, the shift of the centers of the
1438: tori of the ergoregion arises faster for $n=2$
1439: than for $n=1$, yielding
1440: almost touching ergosurfaces already in the vicinity
1441: of the maximal size of the ergosurfaces.
1442:
1443: \subsubsection{Negative parity boson stars}
1444:
1445: For boson stars with negative parity
1446: we observe analogous features as for positive parity
1447: boson stars. This is illustrated in the following.
1448: %
1449: We first consider $1^-$ boson stars
1450: for gravitational coupling $\kappa=0.1$ and $0.2$.
1451: The dependence of their charge $Q$ and mass $M$
1452: on the frequency $\omega_s$ is exhibited
1453: in Fig.~\ref{Qrotbs1-}. For $\kappa=0.1$ a large part of
1454: the spiral could be obtained numerically.
1455:
1456: \begin{figure}[h!]
1457: \parbox{\textwidth}
1458: {\centerline{
1459: \mbox{
1460: \epsfysize=10.0cm
1461: \includegraphics[width=70mm,angle=0,keepaspectratio]{F13a.eps}
1462: \includegraphics[width=70mm,angle=0,keepaspectratio]{F13b.eps}
1463: }}}
1464: \caption{
1465: %13
1466: The charge $Q$ (left) and the mass $M$ (right)
1467: versus the frequency $\omega_s$
1468: for $1^-$ boson stars
1469: for gravitational coupling $\kappa=0.1$ and $0.2$.
1470: The dots indicate the onset of ergoregions.
1471: }
1472: \label{Qrotbs1-}
1473: \end{figure}
1474:
1475: The mass $M$ of negative parity boson stars
1476: exhibits the same characteristic cusp structure
1477: as the mass of positive parity boson stars,
1478: when considered as a function of the charge $Q$ \cite{list}.
1479: This is seen in Fig.~\ref{m_rot-}
1480: for $1^-$ boson stars
1481: at gravitational coupling $\kappa=0.1$ and $0.2$.
1482:
1483: \begin{figure}[h!]
1484: \parbox{\textwidth}
1485: {\centerline{
1486: \mbox{
1487: \epsfysize=10.0cm
1488: \includegraphics[width=70mm,angle=0,keepaspectratio]{F14a.eps}
1489: \includegraphics[width=70mm,angle=0,keepaspectratio]{F14b.eps}
1490: }}}
1491: \caption{
1492: %14
1493: The mass $M$ versus the charge $Q$
1494: for $1^-$ boson stars
1495: for gravitational coupling $\kappa=0.1$ (left) and $0.2$ (right),
1496: together with the mass for $Q$ free bosons, $M=m_{\rm B}Q$.
1497: }
1498: \label{m_rot-}
1499: \end{figure}
1500:
1501: In Fig.~\ref{QrotPM} we compare the $1^-$ boson stars
1502: to the $1^+$ boson stars.
1503: In particular, we exhibit the charge $Q$ versus the frequency
1504: $\omega_s$
1505: for gravitational coupling $\kappa=0.1$ and $0.2$.
1506:
1507: \begin{figure}[h!]
1508: \parbox{\textwidth}
1509: {\centerline{
1510: \mbox{
1511: \epsfysize=10.0cm
1512: \includegraphics[width=70mm,angle=0,keepaspectratio]{F15a.eps}
1513: \includegraphics[width=70mm,angle=0,keepaspectratio]{F15b.eps}
1514: }}}
1515: \caption{
1516: %15
1517: The charge $Q$
1518: versus the frequency $\omega_s$
1519: for $1^-$ boson stars and $1^+$ boson stars
1520: for gravitational coupling $\kappa=0.1$ (left) and $0.2$ (right).
1521: }
1522: \label{QrotPM}
1523: \end{figure}
1524:
1525: The $n$-dependence of the negative parity boson stars
1526: is demonstrated in Fig.~\ref{Qrotbs-}.
1527: Here the charge $Q$ is exhibited versus the frequency $\omega_s$
1528: for coupling constant $\kappa=0.2$.
1529: The figure also shows the mass $M$ versus the charge $Q$
1530: for $2^-$ boson stars for $\kappa=0.2$, exhibiting
1531: part of the cusp structure.
1532: %As expected the characteristic cusp structure
1533: %of boson stars is also present for the higher value of $n$.
1534:
1535: \begin{figure}[h!]
1536: \parbox{\textwidth}
1537: {\centerline{
1538: \mbox{
1539: \epsfysize=10.0cm
1540: \includegraphics[width=70mm,angle=0,keepaspectratio]{F16a.eps}
1541: \includegraphics[width=70mm,angle=0,keepaspectratio]{F16b.eps}
1542: }}}
1543: \caption{
1544: %16
1545: Left: The charge $Q$ versus the frequency $\omega_s$
1546: for $1^-$ and $2^-$ boson stars
1547: for gravitational coupling $\kappa=0.2$.
1548: The dots indicate the onset of ergoregions.
1549: Right: The mass $M$ versus the charge $Q$
1550: for $2^-$ boson stars
1551: for gravitational coupling $\kappa=0.2$
1552: together with the mass for $Q$ free bosons, $M=m_{\rm B}Q$.
1553: }
1554: \label{Qrotbs-}
1555: \end{figure}
1556:
1557: Considering the spatial structure of the boson stars
1558: with negative parity,
1559: the scalar field $\f$ here typically gives rise to
1560: a double torus-like energy density $T_{tt}$
1561: for the $1^-$ and $2^-$ boson stars on the main branch,
1562: seen already in the negative parity $Q$-balls.
1563: As in flat space, the radii of the double tori of the energy density
1564: increase when $n$ and thus the angular momentum increases.
1565: In the spiral the spatial structure of solutions
1566: can become more complicated, exhibiting for instance a quadruple
1567: torus-like structure.
1568:
1569: Addressing the occurrence of ergoregions
1570: for the above sets of boson star solutions,
1571: we observe that rotating boson stars with negative parity also
1572: possess ergoregions in a large part of their domain of existence.
1573: The onset of the ergoregions for the $1^-$ and $2^-$
1574: boson star solutions %shown in Figs.~\ref{Qrotbs1-} and \ref{Qrotbs-}
1575: is again indicated by dots in the respective figures.
1576: Again, if the onset occurs on the main branch,
1577: then the solutions
1578: to the left of the dot and in the spiral do possess an ergoregion,
1579: while the solutions to the right of the dot do not possess an ergoregion.
1580: Likewise, if the onset occurs only in the spiral, then the solutions further
1581: down the spiral all possess ergoregions, while the other solutions do not.
1582: %
1583: We demonstrate the emergence and structure of these ergoregions
1584: in more detail in Fig.~\ref{ergorot-},
1585: where we exhibit the ergoregions
1586: of $1^-$ boson stars at $\kappa=0.1$
1587: and $2^-$ boson stars at $\kappa=0.2$.
1588:
1589: \begin{figure}[h!]
1590: \parbox{\textwidth}
1591: {\centerline{
1592: \mbox{
1593: \epsfysize=10.0cm
1594: \includegraphics[width=75mm,angle=0,keepaspectratio]{F17a.eps}
1595: \epsfysize=10.0cm
1596: \includegraphics[width=75mm,angle=0,keepaspectratio]{F17b.eps}
1597: }}}
1598: {\centerline{
1599: \mbox{
1600: \epsfysize=10.0cm
1601: \includegraphics[width=75mm,angle=0,keepaspectratio]{F17c.eps}
1602: }}}
1603: \caption{
1604: %17
1605: Location and shape of the axially symmetric ergoregions
1606: of boson stars exhibited in the $xz$-plane:
1607: $1^-$ boson stars at $\kappa=0.1$ (upper row)
1608: and $2^-$ boson stars (lower row)
1609: at $\kappa=0.2$.
1610: The numbers in the legend indicate the respective values of the charge $Q$.
1611: }
1612: \label{ergorot-}
1613: \end{figure}
1614:
1615: Addressing first the $1^-$ boson stars, we observe, that
1616: at a critical value of the frequency (where $Q\approx 540$ for $\kappa=0.1$)
1617: the condition (\ref{ergo-c})
1618: is satisfied precisely on two circles,
1619: %centered on the symmetry axis and
1620: located symmetrically w.r.t.~the equatorial plane.
1621: As seen in the (upper left) figure,
1622: the circles turn into two torus-like surfaces,
1623: when $Q$ is decreased in the first part of the spiral.
1624: The ergoregion thus consists of two disconnected tori,
1625: after it first arises.
1626:
1627: Following the set of solutions further into the spiral,
1628: the two parts of the ergoregion increase in size
1629: until they touch at a critical value of the charge
1630: ($Q\approx 447$ for $\kappa=0.1$)
1631: in the equatorial plane, while their centers
1632: shift only little. Beyond this critical value,
1633: the ergoregion then consists of only one connected piece,
1634: a topological torus with an $8$-shaped cross-section.
1635:
1636: As seen in the (upper right) figure,
1637: the ergoregion then rapidly changes shape
1638: and assumes a more ellipsoidal cross-section,
1639: when %the charge $Q$ decreases further and
1640: the solutions evolve deeper into the spiral.
1641: At the same time, the center of the torus
1642: moves further inwards.
1643: Again the innermost part of the ergosurface gets increasingly
1644: closer to the origin, but never touches it.
1645:
1646: The ergoregions of the set of boson stars at different values
1647: of the gravitational coupling constant $\kappa$
1648: exhibit an analogous behaviour.
1649: Likewise, the ergoregions of boson stars at higher $n$ also
1650: exhibit an analogous behaviour %in the regions studied,
1651: as seen in the (lower part of the) figure.
1652:
1653: \section{Conclusions}
1654:
1655: We have addressed boson stars and their flat space limit, $Q$-balls,
1656: recalling and extending known results for positive parity solutions
1657: and presenting new negative parity solutions.
1658: Our main emphasis has been to study the general pattern
1659: displayed by these regular extended objects,
1660: to determine their domain of existence, and to investigate their
1661: physical properties.
1662:
1663: $Q$-balls and boson stars
1664: exist only in a limited frequency range.
1665: Whereas both mass and charge of $Q$-balls assume a minimal value
1666: at a critical frequency,
1667: from where they rise monotonically towards both smaller and larger
1668: frequencies,
1669: boson stars show a different type of behaviour.
1670: Their mass and charge tend to zero when the maximal frequency is approached,
1671: while for smaller values of the frequency
1672: the charge and mass of boson stars exhibit a spiral-like frequency dependence.
1673: The spirals end in limiting solutions with finite values of
1674: the mass and charge,
1675: which depend on the gravitational coupling constant $\kappa$.
1676:
1677: $Q$-balls and boson stars possess radial excitations,
1678: which so far have only been studied for spherically symmetric solutions,
1679: and even there a systematic study has not yet been performed.
1680: Preliminary results indicate, that
1681: radially excited $Q$-balls and boson stars
1682: possess an analogous frequency dependence %and critical behaviour
1683: as the corresponding fundamental solutions.
1684:
1685: Rotating $Q$-balls and boson stars possess an angular momentum $J$
1686: which is quantized in terms of the charge $Q$, $J=nQ$,
1687: with integer $n$.
1688: The positive parity solutions $n^+$ exhibit a torus-like energy density,
1689: while the negative parity solutions $n^-$
1690: exhibit a double torus-like energy density,
1691: with the two tori located symmetrically w.r.t.~the equatorial plane.
1692: %
1693: %We have investigated solutions with $n=1-3$ and both parities.
1694: %For a given charge and parity,
1695: %the maximum of the energy density increases with increasing $n$
1696: %and moves to larger radii. At the same time the mass increases.
1697: Clearly, the double torus-like structure of the negative parity solutions
1698: is energetically unfavourable as compared to the single torus-like structure
1699: of the positive parity solutions.
1700: %making positive parity solutions with quantum number $n+1$
1701: %lighter than negative parity solutions with $n$.
1702: %%yielding for the masses the relation
1703: %%$M((n+1)^+)<M(n^-)$.
1704:
1705: The classical stability of $Q$-balls and boson stars
1706: can be analyzed according to catastrophe theory,
1707: implying a change of classical stability at each cusp encountered,
1708: when the mass is considered as a function of the charge.
1709: Both type of solutions then possess classically stable branches,
1710: representing the physically most relevant sets of solutions,
1711: which in the case of boson stars
1712: have potential applications to astrophysics.
1713:
1714: Rotating objects may possess an ergoregion.
1715: But for globally regular objects such as boson stars
1716: the presence of an ergoregion would imply an instability,
1717: associated with superradiant scattering
1718: \cite{Cardoso,ergo-paper1,ergo-paper2,ergo-paper3}.
1719: Thus rotating boson stars should become unstable,
1720: when they develop an ergoregion.
1721: Therefore the possible presence of ergoregions %in boson stars
1722: was put forward by Cardoso et al.~\cite{Cardoso}
1723: as an argument to exclude boson stars and various other black hole doubles
1724: as potential horizonless candidates for compact dark astrophysical objects.
1725:
1726: Our analysis has shown that ergoregions can indeed be present
1727: for boson star solutions on the classically stable branch.
1728: But their presence only diminishes the set of boson star solutions
1729: with possible physical relevance
1730: (where the diminishment is greater
1731: for larger gravitational coupling and larger angular momentum),
1732: while there always remains a part of the branch
1733: of classically stable boson star solutions,
1734: not suffering from an ergoregion instability.
1735: Therefore, rotating boson stars cannot yet be excluded
1736: as potential candidates for compact astrophysical objects.
1737: It remains to be seen, whether boson stars with
1738: appropriate values of the physical parameters
1739: to fit observational data
1740: will or will not suffer from such an ergoregion instability.
1741:
1742:
1743: \begin{acknowledgments}
1744: We gratefully acknowledge V.~Cardoso for drawing
1745: our attention to the ergoregions of boson stars.
1746: B.K. gratefully acknowledges support by the German Aerospace Center
1747: and by the DFG.
1748: \end{acknowledgments}
1749:
1750:
1751: %\clearpage
1752:
1753: \begin{thebibliography}{99}
1754:
1755: \bibitem{lee-s}
1756: R. Friedberg, T.~D. Lee and A. Sirlin,
1757: Phys. Rev. D13, 2739 (1976)
1758: %Class of scalar-field soliton solutions in three space dimensions
1759:
1760: \bibitem{coleman}
1761: S. Coleman
1762: Nucl. Phys. B262, 263 (1985) (E: B269, 744 (1986))
1763: % Q BALLS
1764:
1765: \bibitem{lee-rev}
1766: %NONTOPOLOGICAL SOLITONS.
1767: T.~D. Lee, and Y. Pang,
1768: Phys. Rept. 221, 251 (1992)
1769:
1770: \bibitem{volkov}
1771: M.~S. Volkov and E. W\"ohnert,
1772: Phys. Rev. D66, 085003 (2002)
1773: %Spinning Q--Balls
1774:
1775: \bibitem{list}
1776: B. Kleihaus, J. Kunz, and M. List,
1777: Phys. Rev. D72, 064002 (2005)
1778:
1779: \bibitem{schunck}
1780: F.~E. Schunck and E.~W. Mielke,
1781: in \textit{Relativity and Scientific Computing},
1782: edited by F.~W. Hehl, R.~A. Puntigam and H. Ruder
1783: (Springer, Berlin, 1996), 138
1784: %Rotating Boson stars
1785:
1786: \bibitem{lee-bs}
1787: R. Friedberg, T.~D. Lee and Y. Pang,
1788: Phys. Rev. D35, 3658 (1987)
1789: %Scalar soliton stars and black holes
1790:
1791: \bibitem{jetzer}
1792: %BOSON STARS.
1793: P. Jetzer,
1794: Phys. Rept. 220, 163 (1992)
1795:
1796: \bibitem{ms-review}
1797: E. Mielke and F.~E. Schunck,
1798: Proc. 8th Marcel Grossmann Meeting, Jerusalem, Israel, 22-27 Jun 1997,
1799: World Scientific, 1607 (1999)
1800: %gr--qc/9801063
1801: %Boson Stars: Early History and recent prospects
1802:
1803: \bibitem{schunck2}
1804: F.~E.Schunck and E.~W. Mielke,
1805: Nucl. Phys. B564, 185 (2000)
1806: %Boson stars: alternativesto primordial black holes?
1807:
1808: \bibitem{schunck3}
1809: E.~W. Mielke, and F.~E. Schunck,
1810: Proc. 8th Marcel Grossmann Meeting, Jerusalem, Israel, 22-27 Jun 1997,
1811: World Scientific, 1633 (1999)
1812:
1813: \bibitem{japan}
1814: S. Yoshida and Y. Eriguchi,
1815: Phys. Rev. D56, 762 (1997)
1816: %Rotating boson stars in general relativity
1817:
1818: \bibitem{Cardoso}
1819: %\cite{Cardoso:2007az}
1820: %\bibitem{Cardoso:2007az}
1821: V.~Cardoso, P.~Pani, M.~Cadoni and M.~Cavaglia,
1822: %``Ergoregion instability rules out black hole doubles,''
1823: arXiv:0709.0532 [gr-qc].
1824: %%CITATION = ARXIV:0709.0532;%%
1825:
1826: %\bibitem{japan2}
1827: %S. Yoshida and Y. Eriguchi,
1828: %Phys. Rev. D55, 1994 (1997)
1829: %New static axisymmetric and nonvacuum solutions in general relativity:
1830: %Equilibrium solutions of boson stars.
1831: %B. Schupp, J.J. van der Bij,
1832: %An axially symmetric Newtonian boson star,
1833: %Phys. Lett. B366, 85-88 (1996)
1834:
1835: \bibitem{wald}
1836: R.~M. Wald,
1837: General Relativity
1838: (University of Chicago Press, Chicago, 1984)
1839:
1840: \bibitem{kkrot1}
1841: B. Kleihaus and J. Kunz,
1842: Phys. Rev. Lett. {86}, 3704 (2001)
1843:
1844: \bibitem{book}
1845: D. Kramer, H. Stephani, E. Herlt, and M. MacCallum,
1846: Exact Solutions of Einstein's Field Equations,
1847: (Cambridge University Press, Cambridge, 1980)
1848:
1849: \bibitem{kk}
1850: B. Kleihaus and J. Kunz,
1851: Phys. Rev. Lett. {78} (1997) 2527;
1852: Phys. Rev. D57, 834 (1998)
1853: %Static axially symmetric Einstein--Yang--Mills--dilaton solutions: regular solutions
1854:
1855: \bibitem{kkrot2}
1856: B. Kleihaus, J. Kunz, and F. Navarro-L\'erida,
1857: Phys. Rev. {D66}, 104001 (2002)
1858:
1859: \bibitem{FIDISOL}
1860: W. Sch\"onauer and R. Wei\ss ,
1861: J. Comput. Appl. Math. 27, 279 (1989)\\
1862: M. Schauder, R. Wei\ss\ and W. Sch\"onauer,
1863: The CADSOL Program Package,
1864: Universit\"at Karlsruhe, Interner Bericht Nr. 46/92 (1992).
1865:
1866: %\cite{Brihaye:2007tn}
1867: \bibitem{Brihaye:2007tn}
1868: Y.~Brihaye and B.~Hartmann,
1869: %``Interacting Q-balls,''
1870: arXiv:0711.1969 [hep-th].
1871: %%CITATION = ARXIV:0711.1969;%%
1872:
1873: \bibitem{catastrophe}
1874: F.~V.~Kusmartsev, E.~W.~Mielke, F.~E.~Schunck,
1875: Phys. Rev. {D43}, 3895 (1991)
1876:
1877: \bibitem{ergo-paper1}
1878: J.~L. Friedman, Commun. Math. Phys. 63, 243 (1978)
1879:
1880: \bibitem{ergo-paper2}
1881: N. Comins and B.~F. Schutz, Proc. R. Soc. Lond. A364, 211 (1978)
1882:
1883: \bibitem{ergo-paper3}
1884: S. Yoshida and Y. Eriguchi, MNRAS 282, 580 (1996)
1885:
1886:
1887: \end{thebibliography}
1888:
1889: \end{document}
1890:
1891:
1892:
1893: