1: \documentclass{article}
2: \usepackage{amssymb,epsfig}
3: \begin{document}
4: \centerline{\large\bf SELF--SIMILAR AND CHARGED RADIATING SPHERES:}
5: \vspace*{0.035truein}
6: \centerline{\large\bf AN ANISOTROPIC APPROACH}
7: \vspace*{0.245truein}
8: \centerline{\footnotesize{\large W. Barreto\footnote{Departamento de F\'\i isica,
9: Centro de F\'\i sica Fundamental, Facultad de Ciencias, Universidad de Los Andes}
10: B. Rodr\'\i guez\footnote{Computational Science Research Center,
11: College of Sciences,
12: San Diego State University, San Diego, California.}
13: L. Rosales\footnote{Laboratorio de F\'\i sica Computacional,
14: Universidad Experimental Polit\'ecnica ``Antonio Jos\'e de
15: Sucre'', Puerto Ordaz, Venezuela.}
16: and O. Serrano\footnote{Departamento de Ciencia y Tecnolog\'\i a,
17: Universidad Experimental de Guayana,
18: Puerto Ordaz, Venezuela.}}}
19: \baselineskip=12pt
20: \vspace*{0.21truein}
21: \date{\today}
22:
23: \begin{abstract}
24: Considering charged fluid spheres as anisotropic sources
25: and the diffusion limit as the transport mechanism, we
26: suppose that the inner space--time admits self--similarity.
27: Matching the interior solution with the Reissner--Nordstr\"om--Vaidya exterior one,
28: we find an extremely compact and oscillatory final state with a redistribution of
29: the electric charge function and non zero pressure profiles.
30:
31: \vspace*{0.21truein}
32: Key words: Charged spheres; anisotropy; compact objects
33:
34: \end{abstract}
35: \section{Introduction}
36: It is well known that astrophysical objects are not significantly charged,
37: For this reason charged bodies have limited interest
38: (for an historical overview see Section I in Ref. \cite{remlz03}).
39: Observed stars, like the Sun or a neutron star, cannot support a great
40: amount of charge. This is so basically because the particles that compose
41: a star have a huge charge to mass ratio, as is the case for a proton or an
42: electron. For cold stars the electric charge is about $100$ Coulombs per solar mass \cite{g00}.
43: Nevertheless, in compact stars the equilibrium conditions are different,
44: allowing some more net charge \cite{b71}.
45: For highly relativistic stars near to form a black hole, the huge gravitational pull
46: can be balanced by huge amounts of net charge. On this kind of configurations
47: we are concerned in this paper and have been considered by other authors
48: \cite{remlz03},\cite{b71}--\cite{ar01}.
49:
50: To include the electric charge a number of authors
51: make additional assumptions such as an equation of state or a
52: relationship between metric variables \cite{hs97}--\cite{hm84}.
53: Bonnor and Wickramasuriya \cite{bw75} have studied
54: electrically charged matter, with electrostatic
55: repulsion balancing the gravitational attraction.
56: Most works have been done under static conditions,
57: including the ones by Ivanov \cite{i02}, who exhaustively surveyed
58: static charged perfect fluid spheres in general relativity, and Ray {\it et
59: al.} \cite{remlz03}, who studied the effect of electric charge in compact stars
60: and its consequences on the gravitational collapse.
61: Bekenstein \cite{b71}
62: found that for highly relativistic stars, whose radius is on the verge
63: of forming an event horizon, the large gravitational pull can be
64: balanced by large amounts of net charge.
65:
66: In this paper we study an electrically charged matter distribution
67: as an anisotropic fluid.
68: It is well known that different energy--momentum tensors can lead to
69: the same space--time \cite{carotplus}. For instance, under spherical symmetry
70: viscosity can be considered as a special case of anisotropy \cite{brt}.
71: To illustrate our approach to obtain
72: dynamical models we consider the diffusion approximation as the transport mechanism, and a
73: self--similar space--time for the inner region. Our treatment allows
74: different scenarios for the gravitational collapse, including the
75: reported some years ago \cite{bd99}. In particular we are now interested in
76: non stationary initial conditions to explore the fate of the gravitational collapse.
77: The set problem has some physical and geometrical features that we review briefly, these are: anisotropy,
78: dissipation and self--similarity.
79:
80: For certain density ranges, local
81: anisotropy in pressure can be physically justified in self--gravitating
82: systems, since different kinds of physical phenomena may take place,
83: giving rise to local anisotropy and in turn relaxing the upper limits
84: imposed on the maximum value of the surface gravitational potential \cite{l33}.
85: The influence of local anisotropy in general relativity has been
86: studied mostly under static conditions (see \cite{hs97} and references therein).
87: Herrera {\it et al.} \cite{hetal04} have reported a
88: general study for spherically symmetric dissipative anisotropic fluids
89: with emphasis on the relationship among the Weyl tensor, the shear
90: tensor, the anisotropy of the pressure and the density inhomogeneity.
91: In another context, a scalar field with non--zero spatial gradient is
92: an example of a physical system where the pressure is clearly
93: anisotropic.
94: Boson stars, hypothetical
95: self-gravitating compact objects resulting from the coupling of a
96: complex scalar field to gravity, are systems where anisotropic
97: pressure occurs naturally. Similarly, the energy--momentum tensor of
98: both electromagnetic and fermionic fields are anisotropic \cite{dg02}.
99: We want to stress that in order to have isotropy we need an extra
100: assumption on the behavior of the fields or on the fluid describing
101: interiors.
102:
103: The proposed dissipative process
104: is suggested by the fact that they provide the only plausible
105: mechanism to carry away the bulk of binding energy, leading to a
106: neutron star or black hole \cite{ks79}. Furthermore, in cores of densities
107: close to $10^{12}\, g/cm^3$ the mean free path for neutrinos becomes small
108: enough to justify the use of the diffusion approximation \cite{a77,k78}.
109: Here we do not discuss the temperature distribution during diffusion; we consider
110: this transport mechanism because it provides an efficient way to
111: radiate energy. Additional studies are necessary to explore the
112: consequences of incorporating a hyperbolic theory of dissipation to
113: overcome the difficulties inherent to parabolic theories.
114:
115: Few exact solutions to the Einstein--Maxwell equations are relevant
116: to gravitational collapse. For this reason, new collapse solutions
117: are very useful, even if they are simplified ones \cite{op90}. It
118: is well known that the field equations admit homothetic motion
119: \cite{ct71}--\cite{lz90}. Applications of self--similarity
120: range from modeling black holes to producing counterexamples
121: to the cosmic censorship conjecture \cite{ch74}--\cite{cc98}.
122: It is well established that in the critical
123: gravitational collapse of an scalar field the space--time can be self--similar
124: \cite{c93}--\cite{g99}. We can expect on physical grounds that this scenario
125: remains similar for charged matter \cite{bcgn02}.
126:
127: In this work the Darmois--Lichnerowicz junction conditions
128: at the surface of the distribution are satisfied. We match the
129: self--similar interior solution with the Reissner--Nordstr\"om--Vaidya exterior one.
130: We obtain a model that resembles the Seidel--Suen
131: solitonic star formed by a real and massive scalar field \cite{ss91}.
132:
133: Section \ref{main_eq} contains the field equations, written in such form that the
134: electrically charged fluid can be considered anisotropic.
135: In this section we also discuss the junction
136: conditions to match the dissipative interior space--time with the
137: exterior space--time, a Reissner--Nordstr\"om--Vaidya one, and write the
138: generalized Tolman--Oppenheimer--Volkoff equation. Section \ref{selfsimilar} contains a
139: review of self--similarity for the sake of completeness. Then we propose
140: a simple solution to include electric charge by means of the active
141: mass. We show two example solutions in section \ref{modeling}
142: and finally discuss our results
143: in section \ref{conclusion}.
144:
145: \section{The field equations and Junction conditions}\label{main_eq}
146: Let us consider a non static distribution of matter which is spherically
147: symmetric and consists of charged fluid of energy density $\rho$, pressure $p$,
148: electric energy density $\epsilon$ and radiation energy flux $q$ diffusing in
149: the radial direction, as measured by a local Minkowskian observer comoving
150: with the fluid (with velocity $-\omega$).
151: In radiation coordinates \cite{b64} the metric takes the form
152: \begin{equation}
153: ds^2=e^{2\beta}\bigg(\frac Vrdu^2+2du\,dr\bigg)
154: -r^2\bigg(d\theta ^2+\sin^2\theta\,d\phi ^2\bigg),
155: \end{equation}
156: where $\beta$ and $V$ are functions of $u$ and $r$.
157: As it is well known for the spherical symmetry $F^{ur}=-F^{ru}$ are the only
158: non vanishing electromagnetic field tensor components. Now, defining the
159: function $C(u,r)$ by the relation
160: \begin{equation}
161: F^{ur}=\frac{C}{r^2}e^{-2\beta}
162: \end{equation}
163: the inhomogeneous Maxwell equations become
164: \begin{equation}
165: C_{,r}=4\pi r^2J^ue^{2\beta},
166: \end{equation}
167: and
168: \begin{equation}
169: C_{,u}=-4\pi r^2J^re^{2\beta},
170: \end{equation}
171: where the comma subscript represents partial derivative with respect to the
172: indicated coordinate; $J^u$ and $J^r$ are the temporal and radial components of
173: the electric current four vector, respectively. Thus, the function $C(u,r)$ is
174: naturally interpreted as the charge within the radius $r$ at the time $u$.
175: Thus, the inhomogeneous Maxwell field equations entail the conservation of charge
176: \begin{equation}
177: U^\mu C_{,\mu}=0,
178: \end{equation}
179: where the four--velocity is given by
180: \begin{equation}
181: U^\mu=e^{-\beta}\left[\sqrt{\frac{r}{V}}\left(\frac{1-\omega}{1+\omega}\right)\delta^\mu_u+
182: \sqrt{\frac{V}{r}}\frac{\omega}{(1-\omega^2)^{1/2}}\delta^\mu_r\right].
183: \end{equation}
184: We can write the conservation equation in a more suitable
185: form, that is
186: \begin{equation}
187: C_{,u}+\frac{dr}{du} C_{,r}=0,
188: \label{ce}
189: \end{equation}
190: where
191: \begin{equation}
192: \frac{dr}{du}=e^{2\beta}\frac{V}{r}\frac{\omega}{1-\omega}\label{eq:drdu}
193: \end{equation}
194: is the matter velocity.
195: It is clear that inside a sphere comoving with the fluid the
196: charge is conserved.
197:
198: Introducing the mass function by means of
199: \begin{equation}
200: \mu=\frac{1}{2}\left(r-Ve^{-2\beta}\right),
201: \end{equation}
202: we can write the Einstein field equations as follows
203: \begin{equation}
204: \frac{\hat\rho + p_r\omega^2}{1-\omega^2}+\frac{2\omega q}{1-\omega^2}=
205: -\frac{e^{-2\beta}\mu_{,u}}{4\pi r(r-2\mu)} +\frac{\mu_{,r}}{4\pi r^2}
206: \label{eq:one}
207: \end{equation}
208: \begin{equation}
209: \bar \rho=\frac{\mu_{,r}}{4\pi r^2}
210: \end{equation}
211: \begin{equation}
212: \bar\rho + \bar p=\frac{\beta_{,r}}{2\pi r^2}(r-2\mu)
213: \end{equation}
214: \begin{eqnarray}
215: p_t=-\frac{1}{4\pi}\beta_{,ur}e^{-2\beta}+\frac{1}{8\pi}(1-2\mu/r)
216: (2\beta_{,rr}+4\beta_{,r}^2-\beta_{,r}/r)\nonumber \\
217: +\frac{1}{8\pi r}[3\beta_{,r}
218: (1-2\mu_{,r})-\mu_{,rr}]
219: \end{eqnarray}
220: where
221: $$
222: \bar p=\frac{p_r-\omega\hat\rho}{1+\omega}-\frac{1-\omega}{1+\omega}q,
223: $$
224: and
225: $$
226: \bar \rho=\frac{\hat\rho-\omega p_r}{1+\omega}-\frac{1-\omega}{1+\omega}q,
227: $$
228: with $\hat \rho=\rho + \epsilon$, $p_r=p-\epsilon$, $p_t=p+\epsilon$
229: and the electric energy density
230: $\epsilon=E^2/8\pi$, where $E=C/r^2$ is the local electric field intensity.
231: Observe that we have introduced the mass function $\mu$ instead the usual
232: total mass $\tilde m=(r-Ve^{-2\beta})/2+C^2/2r$.
233:
234: If we define the degree of local anisotropy induced by charge as
235: $\Delta=p_t-p_r=2\epsilon$, the electric charge determines such a degree at any point.
236:
237: It can be shown that the Tolman--Oppenheimer--Volkoff equation reduces to
238: \begin{eqnarray}
239: \frac{\partial\bar p}{\partial r}+\frac{\bar\rho+\bar p}{1-2\mu/r}
240: \bigg[4\pi r \bar p +\mu/r^2\bigg]
241: -e^{-2\beta}\bigg(\frac{\bar\rho+\bar p}{1-2\mu/r}\bigg)_{,u}=\nonumber \\
242: \frac{2}{r}
243: \bigg(p_t-\bar p\bigg).
244: \label{eq:tov}
245: \end{eqnarray}
246: This generalized equation is the same as for an anisotropic fluid \cite{cosenzaetal}.
247:
248: The most general junction conditions for our
249: system can be written
250: as \cite{bd99}
251: \begin{equation}
252: \mu_a= \tilde m_a-\frac{C_T^2}{2a},
253: \end{equation}
254: \begin{equation}
255: \beta_a=0,
256: \end{equation}
257: and
258: \begin{equation}
259: -\beta _{,u}|_a+(1-2\mu _{a}/a)\beta _{,r}|_a-\frac{\mu _{,r}|_a}{2a}+
260: \frac{C_{T}^{2}}{4a^{3}}=0,
261: \end{equation}
262: where a subscript $a$ indicates that the quantity is evaluated at the surface, $\tilde m_a$ and
263: $C_T$ are the total mass and the total charge, respectively.
264: Remarkably, the last equation is equivalent to
265: \begin{equation}
266: p_{a}=q_{a}.
267: \end{equation}
268:
269: We have obtained a system of equations which are equivalent to the anisotropic
270: matter case (see \cite{hs97} and references therein). In the next
271: section we will see how this approach lead us to an extra
272: simplification.
273:
274: Up to this point all the written equations are general. We can try to solve the system using
275: numerical techniques or a seminumerical approach to gain
276: some insight. The last method is worked out in the next section.
277:
278: \section{The self--similar space--time: A simple solution}\label{selfsimilar}
279: We shall assume that the spherical
280: distribution admits a one--parameter group of homothetic motion.
281: A homothetic vector field on the manifold is one that satisfies
282: $\pounds_\xi {\bf g}=$2$n{\bf g}$ on a local chart,
283: where $n$ is a constant on the
284: manifold and $\pounds$ denotes the Lie derivative operator. If $n \ne 0$ we
285: have a proper homothetic vector field and it can always be scaled so as to have
286: $n = 1$; if $n = 0$ then $\xi$ is a Killing vector on the manifold.
287: So, for a constant rescaling, $\xi$ satisfies
288: \begin{equation}
289: \pounds_\xi{\bf g}=2{\bf g}
290: \end{equation}
291: and has the form
292: \begin{equation}
293: \xi =\Lambda (u,r)\partial_u +\lambda (u,r)\partial_r.
294: \end{equation}
295: The homothetic
296: equations reduce to
297: \begin{equation}
298: \xi(\chi)=0,\label{se1}
299: \end{equation}
300: \begin{equation}
301: \xi(\upsilon)=0,\label{se2}
302: \end{equation}
303: where $\lambda=r$, $\Lambda=\Lambda(u)$,
304: $\chi=\mu/r$ and $\upsilon=\Lambda e^{2\beta}/r$.
305: Therefore, $\chi=\chi(\zeta)$ and $\upsilon=\upsilon(\zeta)$ are solutions
306: if the self--similar
307: variable is defined as
308: \begin{equation}
309: \zeta\equiv r\, e^{- \int du/\Lambda}.
310: \end{equation}
311: We have reported \cite{bd99,bpr99} a
312: simple homothetic solution
313: which can be written
314: for the present case as
315: \begin{equation}
316: \mu=\mu_a(u) (r/a)^{k+1}
317: \end{equation}
318: and
319: \begin{equation}
320: e^{2\beta}=(r/a)^{l+1}
321: \end{equation}
322: with $k$ and $l$ constants. It can be shown that for a given
323: $k$, $l$ must be a root of a complicated seventh degree
324: polynomial in order to
325: satisfy the additional symmetry equations.
326: It is
327: obvious that these solutions satisfy the continuity of
328: the first fundamental form, and to satisfy the continuity
329: of the second fundamental form, the radial velocity at the
330: surface must be
331: \begin{equation}
332: \omega_a=1-\frac{2(l+1)(1-2\mu_a/a)}{2(k+1)\mu_a/a-C_T^2/a^2}.
333: \label{omega}
334: \end{equation}
335: \subsection{Surface equations}
336: Now, to completely determine the metric we have two
337: differential equations at the surface, the equation
338: (\ref{eq:drdu}) and the field equation (\ref{eq:one}) which
339: evaluated at the surface become respectively
340: \begin{equation}
341: \dot a=(1-2\mu_a/a)\frac{\omega_a}{1-\omega_a}\label{ap}
342: \end{equation}
343: and
344: \begin{equation}
345: \dot\mu_a=-Q(1-2\mu_a/a)+\dot a \frac{C_T^2}{2a^2}, \label{mup}
346: \end{equation}
347: where a dot over a variable denotes the derivative with respect to time.
348: The quantity $Q$ is defined as
349: \begin{equation}
350: Q\equiv 4\pi a^2 q_a \left(\frac{1+\omega_a}{1-\omega_a}\right),
351: \label{Q}
352: \end{equation}
353: and can be explicitly determined using equation (\ref{eq:tov})
354: evaluated at the surface, resulting in
355: \begin{eqnarray}
356: Q=\{(1-2\mu_a/a)[2(l+1)^2+k(k+1+3(l+1))]-2(C_T/a)^2 \nonumber \\
357: -k(k+1+3(l+1))\}
358: \{k+1-(1-2\mu_a/a)(k+l+2)-(C_T/a)^2\}/
359: \nonumber \\
360: \left[4(1-2\mu_a/a)(l+1)a^2\right].\label{Q+}
361: \end{eqnarray}
362:
363: The surface equations (\ref{ap}) and (\ref{mup}) together with
364: equations (\ref{omega}) and (\ref{Q+}) can be integrated numerically
365: (using Runge--Kutta for instance) for initial conditions
366: $a(0)$ and $\mu_a(0)$ and some $k$ and $C_T$.
367: The only restrictions on these conditions and parameters come from the need to
368: get a physically acceptable model.
369:
370: For $k=0$ (which implies $l=0$) we obtain the same results of reference \cite{bd99}.
371: As we have mention before, we gain some simplification with the anisotropic approach
372: of this paper, that is, one of the similarity equations
373: (Eq. (23) in Ref. \cite{bd99}) was decoupled,
374: leading us to less restrictive models.
375:
376: \subsection{Integrating the conservation equation}
377: Once the surface equations are integrated we have to integrate
378: the conservation equation (\ref{ce}) to obtain all the physical
379: variables inside the source. To do that we use comoving spatial markers
380: $x=r/a$. Thus the conservation equation can be written as
381: \begin{equation}
382: C_{,u}=-\frac{dx}{du} C_{,x}
383: \end{equation}
384: which is a wave--like equation that can be integrated numerically
385: using the Lax method (with the appropriate Courant--Friedrichs--Levy (CFL)
386: condition). The evolution of the conservation equation is restricted
387: by the surface evolution and was implemented as follows
388: \begin{equation}
389: C^{n+1}_j=\frac{1}{2}\left(C^n_{j+1}+C^n_{j-1}\right)+
390: \frac{\delta u}{2\delta x} \left(\frac{dx}{du}\right)^n_j
391: \left(C^n_{j+1}-C^n_{j-1}\right).
392: \end{equation}
393: The superscript $n$ indicates the hypersurface $u=n\delta u$
394: and the subscript $j$ the spatial position for a comoving
395: observer at $x=j\delta x$. We typically used $\delta u=10^{-2}$
396: with a CFL condition $\delta u=2\delta x$. In order to integrate
397: the conservation equation we need to specify a boundary
398: condition and an initial condition.
399: For this work we have used the boundary condition
400: \begin{equation}
401: C(x=0,u) = 0,
402: \end{equation}
403: with an initial condition
404: \begin{equation}
405: C(x,u=0) = C_T\, x^{\mathcal P},
406: \end{equation}
407: where the power ${\mathcal P}$ allows us
408: to test the sensitivity to the initial conditions of our results.
409:
410: \section{Modeling}\label{modeling}
411: \subsection{Evolving towards dust: $k=0$}\label{A}
412: As a test of our approach to dealing with electric charge as anisotropy is to
413: reproduce the results reported in \cite{bd99}. For this scenario
414: $k=l=0$, $a(0)=3.5$, $\mu_a(0)=1$, ${\mathcal P}=1$ and $C_T=0.09$.
415: Even when the initial mass $m_a (0)$ is not equal to one, we obtain the same
416: results, that is, the collapse is halted and the boundary oscillates
417: when the total electric charge asymptotically approximates the total
418: mass $m_a$ and the radius equals asymptotically to twice the
419: total electric charge. The ratio $P/\rho$ as a function of time is the same for any region.
420: Also, as the distribution evolves,
421: becomes dust asymptotically with damped oscillations and the pressure
422: at the surface coincides with the heat flow. The matter velocity profiles
423: at all regions have damped oscillations that evolve to the static regime.
424: A striking result not reported before is the ability to obtain any interior profile for
425: matter velocity or electric charge function using the simple rule
426: \begin{equation}
427: {\mathcal F}(u,x)=x{\mathcal F}(u,1),
428: \end{equation}
429: where ${\mathcal F}$ could be either $dr/du$ or $C$, that is,
430: once the surface profile is determined we can go without additional work inside the
431: distribution. Another interesting result is that we can
432: get the interior profiles for pressure and energy density by
433: \begin{equation}
434: {\mathcal G}(u,x)={\mathcal G}(u,1)/x^2,
435: \end{equation}
436: where ${\mathcal G}$ could be $\rho$ or $p$.
437:
438: \subsection{Redistribution of the electric charge: $k=0.1$}\label{B}
439: Now that we have successfully tested our new approach, we
440: explore another physical scenario, one with redistribution
441: of the electric charge. For this case we choose $k=0.1$,
442: $a(0)=3.5$, $\mu_a(0)=1$ and $C_T=0.09$
443: to obtain (among other roots) $l=0.741632$.
444: For ${\mathcal P=1}$ (power of the
445: initial condition to integrate the conservation equation) the
446: results are shown in figures \ref{one}--\ref{six}. The system
447: collapses and losses mass until it reaches a stable oscillatory
448: state. When the distribution
449: is in oscillatory regime the mass behaves in the same way, that is, oscillates between
450: two extremal values.
451: This oscillatory behavior
452: extends to all physical variables at all pieces of the material.
453: The interior redistribution of electric
454: charge responds to the surface dynamics.
455: Electric charge (repulsively) let the distribution expand with absorption
456: of energy up to some maximum point in which gravitation compensates
457: electric repulsion. Therefore the collapse undergoes with emission of
458: energy, up to some minimum point in which electric repulsion is again dominant
459: and so on indefinitely. This model seems to be independent
460: of the initial conditions, as shown in figure \ref{seven}.
461: The dependence of the radius on the total electric charge is
462: shown in figure \ref{eight}.
463: Analysis of this last result
464: indicates scale invariance with respect to the total electric charge.
465: In fact, once the stationary state (oscillatory) is reached, the
466: amplitude and period depend on the total electric charge.
467: Thus, if we know the oscillatory regime for
468: some specific value of electric charge, we can reproduce the
469: oscillatory regime for any other, non zero, value of electric charge.
470: In the limit of zero charge the distribution shrink completely,
471: forming a very small region with very high energy density and pressure.
472: This last result was found in another context (see section IV. B in Ref. \cite{bpr99}).
473:
474: The energetic at the boundary surface
475: in conjunction with the pull of gravity and push of electric charge,
476: explain the dependence between the amplitude and period with
477: the total electric charge. These features are analogous to the reported
478: by Brady {\it et al.} \cite{bcg97} for a real and massive scalar field.
479: In this context the mass destroys the scale invariance
480: of the Einstein--scalar field equations, but after the nearly critical phase
481: the oscillation period depends on the inverse of scalar field mass,
482: i.e. the Compton wavelength. In our
483: case the situation is analogous respect to the electric total charge.
484: Parenthetically, the Seidel--Suen oscillaton \cite{ss91} is the stable solution
485: far from the critical behavior of a massive and real scalar field.
486: Our solution behaves like that stable configuration from quite simple
487: considerations.
488:
489: \section{Discussion}\label{conclusion}
490:
491: In this paper we have shown that electrically charged matter
492: can be considered as anisotropic matter.
493: We exemplify the approach obtaining
494: a dynamical model under the diffusion approximation and
495: self--similarity within the source. The showed example
496: is representative of many others varying only the initial conditions,
497: $k$ and $C_T$, and imposing the following physical conditions:
498: $-1<\omega<1$, $p<\rho$ and $\rho> 0$.
499:
500: Some general considerations concerning our results are listed next,
501: \begin{itemize}
502: \item {\sc Extremely compact distribution:}
503: The collapse is halted when the total electric charge is about
504: the total mass (in geometrical units).
505: If the initial radius of the electrically charged
506: distribution is $5.2\,\, km$,
507: the radius of the final stable source
508: is $\approx 266\,\, m$. Also the final pressure (if not zero)
509: is about $ 1.5 \times 10^{39} \,\,Pascal$ and the energy
510: density $ 1.8\times 10^{19}\,\, g/cm^3$.
511: \item {\sc High electric field:}
512: The electric charge used in our models
513: can be as high as $1.5\times 10^{19} \,\,Coulomb $.
514: Thus, the electric field for the stable configuration
515: is about $2.2\times 10^{19} \,\,V/m$ \cite{b71, remlz03}.
516: The sign of net charge does not affect our results.
517: \item {\sc High powered:} If the initial total mass is
518: $1 M_{\odot}$ the final is
519: $\approx 0.1 M_{\odot}$, that is, almost $90 \%$ of the
520: the mass is radiated in $0.15\,\, ms$.
521: \end{itemize}
522:
523:
524: A more general solutions to the homothetic
525: equations (\ref{se1}) and (\ref{se2}) could be necessary to explore
526: how general our results are. However, the results found using
527: the anisotropic approach are consistent with previously reported results
528: when the static regime is the final state \cite{l95}--\cite{hp85}.
529: If the solution reaches stationarity it oscillates. For this case
530: we did not investigate the stability of the system,
531: but independence of initial conditions is apparent.
532: It is clear that the presence of some finite electric charge
533: (or anisotropy) avoids singularity formation.
534: Finally, we would like to conclude by posing a question.
535: If the spherically symmetric
536: source has an upper limit for the total net electric charge
537: that carries \cite{remlz03,fst03}, is it relaxed by the transport mechanism or
538: by the additional homothetic symmetry? Work in this direction is in
539: progress.
540:
541: \section*{Acknowledgements}
542: {This work was supported partially by FONACIT
543: under grants S1--98003270 and F2002000426;
544: by CDCHT--ULA under grant C--1267--04--05--A.
545: Computer time was provided by the Centro Nacional de
546: C\'alculo Cient\'\i fico, Universidad de los Andes (CeCalcULA).}
547:
548: \begin{thebibliography}{99}
549: \bibitem{remlz03} Ray S, Espindola A L, Malheiro M, Lemos J P S and Zanchin V T
550: 2003 {\it Phys. Rev. D} {\bf 68} 084004.
551: \bibitem{g00} Glendenning N K 2000, {\it Compact Stars: Nuclear Physics,
552: Particle Physics, and General Relativity}, 2nd edition, Springer Verlag.
553: \bibitem{b71} Bekenstein J 1971 {\it Phys. Rev} {\bf 4} 2185.
554: \bibitem{dyf95} de Felice F, Yu Y and Fang Z 1995 {\it Mon. Not. R. Astron. Soc.} {\bf 277} L17.
555: \bibitem{dly99} de Felice F, Liu S M and Yu Y 1999 {\it Class. Quantum Grav.} {\bf 16} 2669.
556: \bibitem{yl00} Yu Y and Liu S M 2000 {\it Comm. Teor. Phys.} {\bf 33} 571.
557: \bibitem{ar01} Anninos P and Rothman T 2001 {\it Phys. Rev. D} {\bf 65} 024003.
558: \bibitem{hs97} Herrera L and Santos N O 1997 {\it Phys. Rep.} {\bf 286} 53.
559: \bibitem{b92} Bondi H 1992 {\it Mon. Not. R. Astr. Soc.} {\bf
560: 259} 365.
561: \bibitem{t84} Tikekar R 1984 {\it J. Math. Phys.} {\bf 25} 1481.
562: \bibitem{hm84} Humi M and Mansour J 1984 {\it Phys. Rev. D} {\bf 29} 1076.
563: \bibitem{bw75} Bonnor W B and Wickramasuriya S B P 1975 {\it Mon. Not. R. Astr.
564: Soc.} {\bf 170} 643.
565: \bibitem{i02} Ivanov B V 2002 {\it Phys. Rev. D} {\bf 65} 104001.
566: \bibitem{carotplus} Tabensky R and Taub A 1973 {\it Comm. Math. Phys.}
567: {\bf 29} 61; Tupper B 1981 {\it J. Math. Phys.} {\bf 22} 2666;
568: Tupper B 1983 {\it Gen. Rel. Grav.} {\bf 15} 849;
569: Raychaudhuri A and Saha S 1981 {\it J. Math. Phys.} {\bf 22} 2237;
570: Raychaudhuri A and Saha S 1982 {\it J. Math. Phys.} {\bf 23}
571: 2554; Carot J and Ib\'a\~nez J 1985 {\it J. Math. Phys.}
572: {\bf 26} 2282.
573: \bibitem{brt} Barreto W and Rojas S 1992 {\it Ap. Sp. Sc.} {\bf 193}
574: 201; Barreto W 1993 {\it Ap. Sp. Sc.} {\bf 201} 191;
575: Coley A A and Tupper B O J 1994 {\it Class. \& Quantum Grav.} {\bf 11} 2553.
576: \bibitem{bd99} Barreto W and Da Silva A 1999 {\it Class. \& Quantum Grav.} {\bf 16} 1783.
577: \bibitem{l33} Lema\^\i tre G 1933 {\it Ann. Soc. Sci. Bruxelles} {\bf
578: A53} 51.
579: \bibitem{hetal04} Herrera L, Di Prisco A, Martin J, Ospino J, Santos N O and
580: Troconis O 2004 {\it Phys. Rev. D} {\bf 69} 084026.
581: \bibitem{dg02} Dev K and Gleiser M 2002 {\it Gen. Rel. Grav.} {\bf 34} 1793.
582: \bibitem{ks79} Kazanas D and Schramm D 1979 {\it Sources of Gravitational Radiation},
583: L. Smarr ed. (Cambridge University Press, Cambridge).
584: \bibitem{a77} Arnett W D 1977 {\it Astrophys. J.} {\bf 218} 815.
585: \bibitem{k78} Kazanas D 1978 {\it Astrophys. J.} {\bf 222} L109.
586: \bibitem{l88} Lattimer J 1988 {\it Nucl. Phys.} {\bf A478} 199.
587: \bibitem{op90} Ori A and Piran T 1990 {\it Phys. Rev. D} {\bf 42}
588: 1068.
589: \bibitem{ct71} Cahill M E and Taub A H 1971 {\it Comm. Math. Phys.}
590: {\bf 21} 1.
591: \bibitem{hp91} Henriksen R N and Patel K 1991 {\it Gen. Rel. Grav.}
592: {\bf 23} 527.
593: \bibitem{lz90} Lake K and Zannias T 1990 {\it Phys. Rev. D}
594: {\bf 41}, R3866.
595: \bibitem{ch74} Carr B J and Hawking S 1974 {\it Mon. Not. R. Astr. Soc.}
596: {\bf 168}, 399.
597: \bibitem{bh78a} Bicknell G V and Henriksen R N 1978 {\it Astrophys. J.}
598: {\bf 219} 1043.
599: \bibitem{bh78b} Bicknell G V and Henriksen R N 1978 {\it Astrophys. J.},
600: {\bf 225}, 237.
601: \bibitem{cy90} Carr B J and Yahil A 1990 {\it Astrophys. J.}
602: {\bf 360}, 330.
603: \bibitem{eimm86} Eardley D M, Isenberg J, Marsden J and Moncrief V 1986
604: {\it Comm. Math. Phys.} {\bf 106} 137.
605: \bibitem{l92} Lake K 1992 {\it Phys. Rev. Lett.} {\bf 68} 3129.
606: \bibitem{b95} Brady P R 1995 {\it Phys. Rev. D} {\bf 51} 4168.
607: \bibitem{cc98} Carr B J and Coley A A 1999 {\it Class.
608: \& Quantum Grav.} {\bf 16} R31;
609: Carr B J and Coley A A 2000 {\it Phys. Rev. D} {\bf 62} 044023.
610: Coley A A and Taylor T D 2001 {\it Class. \& Quantum Grav.} {\bf 18} 4213;
611: Carr B J, Coley A A, Goliath M, Nilsson U S and Uggla C 2001 {\it Class. \& Quantum Grav.} {\bf 18} 303;
612: Carr B J and Coley A A 2005 {\it Gen. Rel. Grav.} {\bf 37} 2165;
613: \bibitem{c93} Choptuik M W 1993 {\it Phys. Rev. Lett.} {\bf 70} 9.
614: \bibitem{hs96} Hamad\'e R and Stewart J M 1996 {\it Class. \&
615: Quantum Grav.} {\bf 13} 497.
616: \bibitem{g99} Gundlach C 1999 {\it Living Reviews in Relativity},
617: {\bf 2}, 4.
618: \bibitem{bcgn02} Brady P R, Choptuik M W, Gundlach C and
619: Nielsen D 2002 {\it Class. \& Quant. Grav.} {\bf 19} 6359.
620: \bibitem{ss91} Seidel E and Suen W M 1991 {\it Phys. Rev. Lett.} {\bf 66} 1659.
621: \bibitem{b64} Bondi H 1964 {\it Proc. R. Soc. A} {\bf 281} 39.
622: \bibitem{bcg97} Brady P, Chambers C and Goncalves S 1997 {\it Phys.Rev. D} {\bf 56} 6057.
623: \bibitem{cosenzaetal} Cosenza M, Herrera L, Esculpi M and
624: Witten L 1982 {\it Phys. Rev. D} {\bf 25} 2527.
625: \bibitem{bpr99} Barreto W, Peralta C
626: and Rosales L 1999 {\it Phys. Rev. D} 59 024008.
627: \bibitem{l95} L\'opez C 1995 {\it Gen. Rel. Grav.} {\bf 27} 85.
628: \bibitem{b65} Bonnor W B 1965 {\it Mon. Not. R. Astron. Soc.} {\bf 129}
629: 443.
630: \bibitem{f69} Faulkes M 1969 {\it Can. J. Phys.} {\bf 47} 1989.
631: \bibitem{cd78} Cooperstock F and de la Cruz V 1978 {\it Gen. Rel. Grav.}
632: {\bf 9} 835.
633: \bibitem{hp85} Herrera L and Ponce de Le\'on J 1985 {\it J. Math. Phys.}
634: {\bf 26} 2302.
635: \bibitem{fst03} Fayos F, Senovilla J and Torres R 2003 {\it Class. \&
636: Quantum Grav.} {\bf 20} 2579.
637: \end{thebibliography}
638: \newpage
639:
640: \begin{figure}
641: \centerline{\epsfxsize=4.in\epsfbox{figure1.ps}}
642: \caption{Total electric charge (short dashed line) and the evolution
643: of radius $a$ (solid line) and the total mass $m_a$ (dashed line).}
644: \label{one}
645: \end{figure}
646:
647: \begin{figure}
648: \centerline{\epsfxsize=4.in\epsfbox{figure2.ps}}
649: \caption{Energy density
650: as a function of time for different pieces
651: of the distribution: $x=0.25$ (short dashed line); $x=0.5$ (large
652: dashed line); $x=0.75$ (solid line) and $x=1$ (dotted line).}
653: \label{two}
654: \end{figure}
655: \begin{figure}
656: \centerline{\epsfxsize=4.in\epsfbox{figure3.ps}}
657: \caption{Pressure as a function of time for different pieces
658: of the distribution: $x=0.25$ (short dashed line); $x=0.5$ (large
659: dashed line); $x=0.75$ (solid line) and $x=1$ (dotted line).}
660: \label{three}
661: \end{figure}
662: \begin{figure}
663: \centerline{\epsfxsize=4.in\epsfbox{figure4.ps}}
664: \caption{Heat flux as a function of time for different pieces
665: of the distribution: $x=0.25$ (short dashed line); $x=0.5$ (large
666: dashed line); $x=0.75$ (solid line) and $x=1$ (dotted line).}
667: \label{four}
668: \end{figure}
669: \begin{figure}
670: \centerline{\epsfxsize=4.in\epsfbox{figure5.ps}}
671: \caption{Matter Velocity as a function of time for different pieces
672: of the distribution: $x=0.25$ (short dashed line); $x=0.5$ (large
673: dashed line); $x=0.75$ (solid line) and $x=1$ (dotted line).}
674: \label{five}
675: \end{figure}
676:
677: \begin{figure}
678: \centerline{\epsfxsize=4.in\epsfbox{figure6.ps}}
679: \caption{Electric charge function versus time for different
680: pieces of the material: $x=0.25$ (solid line); $x=0.5$ (large
681: dashed line); $x=0.75$ (short dashed line).}
682: \label{six}
683: \end{figure}
684:
685: \begin{figure}
686: \centerline{\epsfxsize=5.0in\epsfbox{figure7.ps}}
687: \caption{Electric charge function versus time for $x=0.75$ and
688: different values of power ${\mathcal P}$
689: in the initial value of $C$ to integrate
690: the conservation equation. Curves correspond to: ${\mathcal P}=1$
691: (continuous line); ${\mathcal P}=2$ (large dashed line);
692: ${\mathcal P}=3$ (short dashed line) and ${\mathcal P}=4$ (dotted line).
693: Many values of $a(0)$ were used with essentially the
694: same results.}
695: \label{seven}
696: \end{figure}
697:
698: \begin{figure}
699: \centerline{\epsfxsize=4.0in\epsfbox{figure8.ps}}
700: \caption{Radius as a function time for $C_T=0.00$ (solid line);
701: $C_T=0.09$ (large dashed line);
702: $C_T=0.18$ (short dashed line); $C_T=0.36$ (dotted line).}
703: \label{eight}
704: \end{figure}
705:
706: \end{document}
707: