1: \documentclass[amsmath,amssymb,aps,eqsecnum,nofootinbib,preprintnumbers]{revtex4}
2: %\documentclass[showpacs,preprintnumbers,amsmath,amssymb,aps,eqsecnum]{revtex4}
3: \usepackage[dvips]{graphics,color}
4: \usepackage{latexsym}
5: \usepackage{graphicx}
6:
7:
8: \begin{document}
9:
10: \preprint{KUNS-2113}
11:
12: \title{Orthogonal black di-ring solution}
13:
14: \author{
15: Keisuke Izumi\footnote{e-mail:
16: ksuke@tap.scphys.kyoto-u.ac.jp}
17: \\~}
18:
19: ~\\
20:
21: \address{Department of Physics, Kyoto University, Kyoto 606-8502, Japan}
22:
23:
24: \begin{abstract}
25: We construct a five dimensional exact solution of the orthogonal black di-ring
26: which has two black rings
27: whose $S^1$-rotating planes are orthogonal.
28: This solution has four free parameters
29: which represent
30: %distances between the center and each black ring,
31: radii of
32: and speeds of $S^1$-rotation of the black rings.
33: We use the inverse scattering method.
34: This method needs the seed metric.
35: We also present a systematic method how to construct a seed metric.
36: Using this method, we can probably construct other solutions having
37: many black rings on the two orthogonal planes with or
38: without a black hole at the center.
39: \end{abstract}
40: \maketitle
41:
42: \section{Introduction}
43:
44: In recent years, higher dimensional black objects
45: have been actively studied.
46: This is partly because string theory tells us that
47: the spacetime we live in is higher dimensional.
48: In order for us to feel that the number of spacetime dimensions is four on large scales,
49: the extra dimensions must be compactified.
50: However, when we observe a phenomenon on a small scale,
51: the effect of higher dimensions may appear.
52: If the compactification scale is sufficiently large,
53: this effect can be detected in future collider experiments~\cite{collider}.
54:
55: The only asymptotically flat static solution of the vacuum Einstein equations
56: in higher dimensions is Schwarzschild-Tangherlini solution~\cite{Tan,static},
57: which is stable against perturbation~\cite{Ishibashi}.
58: These properties are the same as in the four dimensional case.
59: However, in the asymptotically flat,
60: stationary and axisymmetric case,
61: % which implies spacetime is ,
62: the uniqueness theorem of the black hole
63: does not exist in higher dimensions unlike the four dimensional case.
64: Myers and Perry discovered the higher dimensional black hole~\cite{Myers}
65: whose topology is $S^{D-1}$,
66: which is an extension of the Kerr black hole to higher dimensions.
67: This solution was obtained also by the solitonic solution-generating methods~\cite{Iguchi2}
68: and the inverse scattering method \cite{Tomizawa3}.
69: A five dimensional Myers-Perry black hole is unique~\cite{Morisawa}
70: if black hole topology is $S^3$ in the asymptotically flat spacetime
71: and if the spacetime has three commuting Killing vectors.
72: The black ring solution with horizon topology $S^1 \times S^2$
73: was discovered by Emparan and Reall~\cite{Emparan}.
74: The black ring rotating on the $S^1$ plane,
75: which helps the balance against its attractive self-gravity force.
76: The $S^2$-rotating black ring solution was
77: discovered by Mishima and Iguchi~\cite{Mishima,Tomizawa2,Figueras},
78: in which a plane supporting black ring from falling due to the self-gravity
79: is needed.
80: The solution of black ring with $S^1$ and $S^2$ rotations was also found
81: by Pomeransky and Sen'kov~\cite{Pomeransky},
82: which first had been found by numerical method~\cite{Kudoh}.
83: %as Myers-Perry black hole,
84: %Uniqueness theorem of the black ring solution was proved
85: It was proved that the black ring solution with horizon topology $S^1 \times S^2$
86: is only Pomeransky and Sen'kov solution~\cite{Yazadjiev,Tomizawa}.
87: In the case with Maxwell fields, a generalzed analysis has done~\cite{Yazadjiev6}.
88: Moreover, the solution which has a number of black objects,
89: such as black di-ring~\cite{diring,diring2} and black saturn~\cite{saturn}
90: were also discovered.
91: Rotating dipole black ring solution~\cite{Yazadjiev2,Yazadjiev3,Yazadjiev5}
92: and black saturn with dipole black ring solution~\cite{Yazadjiev4} have been generated in five-dimensional
93: Einstein-Maxwell-dilaton gravity.
94:
95:
96: In this paper, we construct the solution having
97: two black rings which are orthogonal to each other.
98: We call this solution ``orthogonal black di-ring".
99: We use the inverse scattering method~\cite{ISM}.
100: In inverse scattering method,
101: we need a seed solution.
102: % which has already been known
103: %or which is trivial.
104: We present a method of constructing a diagonal seed metric.
105: %From the seed metric which is obtained by our method,
106: %we can probably construct the solution with the many black objects
107: %which are black rings on the two orthogonal planes
108: %with or without black hole at the center.
109:
110: This paper is organized as follows.
111: In Sec.~\ref{inverse}, we will review the inverse scattering method.
112: In Sec.~\ref{seed}, we will present a method of constructing a seed metric,
113: giving a seed metric for the orthogonal black di-ring solution.
114: In Sec.~\ref{trans}, we will show how to obtain the
115: orthogonal black di-ring solution using the inverse scattering method
116: and we will write the obtained metric explicitly.
117: In Sec.~\ref{analysis}, we will analyse the regularity of the obtained solution.
118: The solution with several parameters generally has singularities.
119: However, we can remove all of these singularities
120: if we choose the parameters appropriately,
121: leaving four free parameters.
122: We show the conditions that the parameters must satisfy for regularity.
123: In Sec.\ref{summary}, we will summarize our results.
124:
125: \section{Inverse scattering method} \label{inverse}
126:
127: In this section, we briefly explain the inverse scattering method~\cite{ISM},
128: by which a new metric can be obtained from a known seed metric.
129: This method can be used when the metric which we want has
130: $D-2$ commuting killing vector fields, one of which is timelike,
131: in $D$-dimensional spacetime.
132: In this paper, we only present the procedure for generating a new solution
133: without giving its deviation.
134: %If you want to know particular logical explanations,
135: For detailed deviation, see Belinsky's Paper~\cite{ISM}.
136:
137: Thanks to the symmetries, the metric can be written as
138: \begin{eqnarray}
139: ds^2 = f (d\rho^2+ dz^2) + g_{ab} dx^a dx^b, \label{metric}
140: \end{eqnarray}
141: where $f$ and $g_{ab}$ depend only on $\rho$ and $z$.
142: Here without loss of generality,
143: we can set the determinant of $g_{ab}$ as
144: \begin{eqnarray}
145: \det g_{ab} =-\rho^2.
146: \end{eqnarray}
147: Then the Einstein equations become
148: \begin{eqnarray}
149: &&\partial_\rho U + \partial_z V =0, \label{diffg}\\
150: &&U=\rho(\partial _\rho {\bf g}) {\bf g}^{-1},
151: \qquad
152: V=\rho(\partial _z {\bf g}) {\bf g}^{-1},
153: \label{UV}\\
154: &&\partial_\rho \ln f =
155: -\frac{1}{\rho} +\frac{1}{4\rho} \mbox{Tr}(U^2-V^2),
156: \label{rhof}\\
157: &&\partial_z \ln f = \frac{1}{2\rho} \mbox{Tr}(UV),\label{zf}
158: \end{eqnarray}
159: where ${\bf g}$ is $g_{ab}$ in matrix notation.
160: These equations can be classified.
161: The first three are the differential equations of ${\bf g}$.
162: The others are the equations from which $f$ is obtained for a given ${\bf g}$.
163: From eq.(\ref{diffg}),
164: we find that integrability condition for $\ln f$,
165: $\partial_\rho\partial_z \ln f = \partial_z\partial_\rho \ln f$,
166: is satisfied.
167:
168: Suppose that a seed metric ${\bf g_0}$
169: which satisfies the Einstein equations (\ref{diffg}-\ref{zf}) is prepared.
170: Then we consider linear differential equations
171: \begin{eqnarray}
172: &&\left(\partial_z
173: -\frac{2 \lambda^2}{\lambda^2+\rho^2}\partial_\lambda\right) \Psi
174: =\frac{\rho V_0-\lambda U_0}{\lambda^2 +\rho^2} \Psi,
175: \label{psi1}
176: \\
177: &&\left( \partial_\rho +
178: \frac{2 \lambda \rho}{\lambda^2+\rho^2}\partial_\lambda \right) \Psi
179: =\frac{\rho U_0-\lambda V_0}{\lambda^2 +\rho^2} \Psi,
180: \label{psi2}
181: \end{eqnarray}
182: where $U_0$ and $V_0$ are $U$ and $V$ with ${\bf g}={\bf g_0}$,
183: $\lambda$ is the complex spectral parameter
184: independent of $\rho$ and $z$,
185: and
186: $\Psi=\Psi(\lambda,\rho,z)$ is a $(D-2)\times(D-2)$ matrix.
187: Solving these equations, we get the matrix $\Psi_{g_0}$.
188:
189: We prepare functions $\mu_k(\rho,z)$ and $\bar \mu_k(\rho,z)$,
190: which we call solitons and antisolitons, respectively,
191: defined by
192: \begin{eqnarray}
193: \mu_k = \sqrt{\rho^2 + (z-a_k)^2} -(z-a_k) , \qquad
194: \bar \mu_k = - \sqrt{\rho^2 + (z-a_k)^2} -(z-a_k) \label{soliton},
195: \end{eqnarray}
196: where $a_k$ is a real constant.
197: %In this paper, $\tilde\mu_k$ with $+$ ($-$) sign is called soliton (antisoliton)
198: %and it is denoted by $\mu_k$ ($\bar \mu_k$).
199: We choose ${\mu'}_{\!\!k}$ $(i=1,\cdots,n)$ from either $\mu_k$ or $\bar\mu_k$.
200: We introduce $n$ 3-vectors $m^{(k)}$
201: associated with ${\mu'}_{\!\!k}$.
202: %which correspond to $\tilde\mu_k$.
203: $m^{(k)}$ is called BZ vector.
204: We make a $n\times n$ matrix as
205: \begin{eqnarray}
206: &&\Gamma_{kl} =\frac{ m^{(k)}_a (\Psi_{g_0}^{-1} ({\mu'}_{\!\!k},\rho,z) )^{ab}
207: g_{0bc} (\Psi_{g_0}^{-1} ({\mu'}_{\!\!l},\rho,z) )^{cd} m^{(l)}_d}
208: {\rho^2 +{\mu'}_{\!\!k} {\mu'}_{\!\!l}}.\label{Gamma}
209: \end{eqnarray}
210: %where ${\mu'}_{\!\!i}$ is $\mu_i$ or $\bar\mu_i$.
211: Then a metric
212: \begin{eqnarray}
213: && {g'}_{\!\!1ab}=
214: \left( g_{0ab} - \sum_{kl} (\Gamma^{-1})_{kl}{{\mu'}_{\!\!k}}^{-1}{{\mu'}_{\!\!l}}^{-1}
215: N^{(k)}_a N^{(l)}_b
216: \right) , \label{hatg}
217: \\
218: &&N^{(k)}_a\equiv m^{(k)}_b (\Psi_0^{-1} ({\mu'}_{\!\!k},\rho,z) )^{bc} g_{0ca}.\label{N}
219: \end{eqnarray}
220: satisfies the Einstein equations.
221: In general, ${\bf{ g'}_{\!\!1}}$ does not satisfies $\det {\bf g}=-\rho^2$.
222: However, ${\bf {g'}_{\!\!1}}$ multiplied by $\rho$ and ${\mu'}_{\!\!i}$
223: also satisfies the Einstein equations.
224: With the help of this property,
225: we can get the metric satisfying both the Einstein equations
226: and $\det {\bf g_1} =-\rho^2$ as
227: \begin{eqnarray}
228: g_{1ab}= \rho^{-\frac{2n}{D}} \prod _{k=1}^n {{\mu'}_{\!\!k}}^{\frac{2}{D}} {g'}_{\!\!1ab}.
229: \label{newg}
230: \end{eqnarray}
231:
232: Moreover, if we choose $f$ as
233: \begin{eqnarray}
234: f=C f_0 \rho^{\frac{-(n^2 +2n -Dn)}{D}}
235: \left(\prod_{k=1}^n {{\mu'}_{\!\!k}}^{\frac{2(n+D-1)}{D}}\right)
236: \left(\prod_{k=1}^n ({{\mu'}_{\!\!k}}^2 \rho^2)^{\frac{2-D}{D}}\right)
237: \left(\prod _{k,l=1,k>l}^n ({\mu'}_{\!\!k} -{\mu'}_{\!\!l})^{\frac{4}{D}}\right)^{-1}
238: \det \Gamma_{kl}, \label{newf}
239: \end{eqnarray}
240: with a constant $C$,
241: %where $C$ is some constant,
242: $f$ satisfies eq.(\ref{rhof}) and eq.(\ref{zf}).
243: %In the left hand side of eq.(\ref{newf})
244: %the transformation from $g'$ to $g$ causes
245: %the product except for $C$, $f_0$ and $\det \Gamma_{kl}$ to appear.
246: The above set of operations is called soliton transformation with \{${\mu'}_{\!\!i}$\}.
247:
248: In this paper
249: we use the strategy taken in Ref.\cite{Pomeransky2}.
250: Using this method,
251: a new metric ${\bf g}$ automatically satisfies $\det {\bf g}=-\rho^2$
252: without the operation of eq.(\ref{newg}).
253: We give a diagonal seed metric ${\bf g_0}$ and $f_0$ which is constructed using $\mu_k$ and $\rho$
254: as explained in the next section.
255: First, we apply the soliton transformation
256: with \{${\mu'}_{\!\!i}$\}
257: % the solitons (or antisolitons)
258: and simple BZ vectors to the seed metric ${\bf g_0}$,
259: to obtain the metric $ {\bf g'_1}$
260: %which is corresponding to $g'$
261: (see eq.(\ref{hatg})).
262: %This removal can be done by the operation of antisolitons (solitons).
263: If we operate the soliton transformation to the metric ${\bf g'_1}$
264: with \{${{\bar\mu}'}_i$\}
265: % the antisolitons (solitons) corresponding to the solitons (antisolitons)
266: %used for obtaining the metric $g'_0$
267: and the same BZ vectors,
268: the metric is transformed back into the seed metric ${\bf g_0}$
269: where we refer to $\bar\mu_i$ ($\mu_i$) as ${{\bar\mu}'}_{i}$
270: when ${\mu'}_{\!\!i}$ is $\mu_i$ ($\bar\mu_i$).
271: If we operate the soliton transformation to the metric ${\bf g'_0}$
272: with \{${{\bar\mu}'}_i$\}
273: % the same antisolitons (solitons)
274: and general BZ vectors,
275: we get a new metric ${\bf g}$.
276: Since the change of $\det {\bf g}$ is independent of BZ vectors,
277: change of $\det {\bf g}$ by soliton transformation from ${\bf g'_1}$ to ${\bf g_0}$
278: is the same as that from ${\bf g'_1}$ to ${\bf g}$.
279: This means $\det {\bf g}=\det {\bf g_0}$ and
280: we don't need to rescale the new metric.
281: %When we consider that the seed metric $g_0$ and the new metric $g$ are
282: %obtained from the metric $g'_0$,
283: %the each complicate coefficient of eq.(\ref{newg}) is the same.
284: Moreover, we don't care the complicated factors composed of $\rho$ and $\mu'_{k}$ in eq.(\ref{newf}).
285: As a result, the metric ${\bf g}$ and $f$
286: % can be easily obtained as
287: take a simple form as
288: \begin{eqnarray}
289: &&\!\!\!\!\!\!
290: g_{ab}= {g'}_{\!\!1ab} - \sum_{kl} ({\Gamma'}^{-1})_{kl}{\mu'}_{\!\!k}^{-1}{\mu'}_{\!\!l}^{-1}
291: N^{(k)}_a N^{(l)}_b
292: ,\\
293: &&
294: f=C f_0 \frac{\det {\Gamma'}_{kl}}{\det {\Gamma'}^{(0)}_{kl}} ,\label{f}
295: \end{eqnarray}
296: where ${\Gamma'}_{kl}$ and ${\Gamma'}_{kl}^{(0)}$ are constructed
297: from the metric $g'_1$, and $\Psi_{g'_1}$
298: %where $\Psi'$ is $\Psi$ $\Gamma_{kl}$ constructed from $g'_0$, and
299: with general BZ vectors
300: % corresponding to $\det {\Gamma'}_{kl}$ and $\det {\Gamma'}_{kl}^{(0)}$ are
301: the simple ones, respectively.
302:
303: In the inverse scattering method
304: regular points where $\det {\bf g} \neq 0$ are transformed to regular points.
305: Some points where $\det {\bf g} =0$ become physical singularities.
306: However, we don't need to care about physical singularities at the level of the seed metric
307: because physical singularities can be transformed to coordinate singularities.
308: %However, singularities including coordinate artifacts
309: %are transformed to coordinate or physical singularities.
310: In many cases, the seed metric from which a new metric
311: without a physical singularity is obtained has
312: physical singularities on the axis.
313:
314:
315:
316:
317: \section{Seed metric}\label{seed}
318:
319: %Following the prescription of Ref.\cite{Emparan2},
320: %which discusses the generalized higher-dimensional
321: %Weyl solutions,
322: In this section,
323: we will give a seed solution
324: which is one of the Weyl solutions~\cite{Emparan2}.
325: We have introduced solitons $\mu_k$ in eq.(\ref{soliton}).
326: Using these solitons, we can construct a solution of the Einstein equations as follows.
327: We prepare the $3\times3$ diagonal metric like
328: \begin{eqnarray}
329: {\bf g}=\mbox{diag} \left\{
330: -\frac{\mu_{i_1^1} \mu_{j^1_2} \cdots }{\mu_{{i'}^1_1}\cdots} ,
331: \frac{\rho^2 \mu_{i^2_1} \cdots}{ \mu_{{i'}^2_1} \cdots},
332: \frac{\mu_{i^3_1} \cdots}{\mu_{{i'}^3_1} \cdots}
333: \right\}, \label{gg}
334: \end{eqnarray}
335: where the total number of each $\mu_k$ in the numerator
336: of all components is equal to that in the denominator.
337: Though in the above example $\rho^2$ appears in the second component,
338: it can be in any component but it appears only once.
339: You can check that the metric ${\bf g}$ satisfies Einstein equations (\ref{diffg}) and
340: $\det{\bf g} = -\rho^2$ (See Appendix \ref{apg}).
341: Next, we give $f$ as
342: \begin{eqnarray}
343: &&f=\kappa^2 \left( \prod_{k} \mu_{i^2_k}\right)
344: \left(\prod_{k} \mu_{{i'}^2_k}^{-1} \right)
345: \prod_{a=1}^3\Biggl\{
346: \left(\prod_{k,l}\frac{ \rho^2+\mu_{i^a_k}\mu_{{i'}^a_l}}{\mu_{i^a_k}\mu_{{i'}^a_l}}\right)
347: \left(\prod_{k\neq l}\frac{ \rho^2+\mu_{i^a_k}\mu_{i^a_l}}{\mu_{i^a_k}\mu_{i^a_l}}\right)^{-1}
348: \left(\prod_{k\neq l}\frac{ \rho^2+\mu_{{i'}^a_k}\mu_{{i'}^a_l}}{\mu_{{i'}^a_k}\mu_{{j'}^a_l}}\right)^{-1}
349: \nonumber\\
350: &&\qquad\qquad\quad\qquad\qquad\qquad\qquad\qquad\qquad\qquad\qquad\qquad\times
351: \left(\prod_{k}\frac{ \rho^2+\mu_{i^a_k}^2}{\mu_{i^a_k}^2}\right)^{(-1/2)}
352: \left(\prod_{k}\frac{ \rho^2+\mu_{{i'}^a_k}^2}{\mu_{{i'}^a_k}^2}\right)^{(-1/2)}
353: \Biggr\}, \label{F}
354: \end{eqnarray}
355: where $\kappa$ is a constant
356: and $\mu_{i^a_k}$ ($\mu_{{i'}^a_k}$) are
357: the solitons that appear in the numerator (denominator) of the $a$-th component.
358: %$\prod_l$ ($\prod_m$) is the product with regard to $\mu_l$ in
359: %the numerator (denominator) in the component including $\rho^2$ in $g$,
360: %$\prod_{ik}$ is the product with regard to $\mu_i$ in the numerator
361: %and $\mu_k$ in the denominator in the same component,
362: %$\prod_{ij}$ is the product with regard to $\mu_i$
363: %and $\mu_j$ ($i\neq j$) in the numerator
364: %or in the denominator in the same component
365: %and
366: %$\prod_{i}$ is the product with regard to all $\mu_i$ in $g$.
367: Although, in the above example, $f$ has the product
368: $\prod_{k} \mu_{i^2_k}\prod_{k} \mu_{{i'}^2_k}^{-1}$,
369: we must replace it with the product of $\mu_{i^a_k}$
370: corresponding to the component having $\rho^2$ in the general case.
371: Then, $f$ satisfies Einstein equations (\ref{rhof}) and (\ref{zf}) (See Appendix \ref{apg}).
372: Since $\mu_i>0$ except for the rod,
373: the seed metric is regular in the region satisfying $\rho>0$.
374: Although this seed metric generally has singularities on the rod,
375: %as we explained above,
376: it is not a problem at all
377: as we explained above.
378: %if the obtained solution has no physical singularity.
379:
380: Using this prescription, we will show a method
381: % by which
382: %if a rod structure of the seed metric is given,
383: %we can
384: for constructing a seed metric corresponding to a given rod structure.
385: Before that, we give a brief explanation about a rod structure.
386: %which shows that the direction of a zero-norm vector.
387: Since $\det {\bf g} =-\rho^2$,
388: at least one eigenvalue of ${\bf g}$ becomes zero at $\rho = 0$.
389: An eigenvector with a zero eigenvalue is not always the same but depend on $z$.
390: A rod structure represents how the eigenvector changes depending on $z$.
391: The positive density rod indicates the direction of the eigenvector
392: with a zero eigenvalue.
393: In this paper, we consider only the case in which
394: the eigenvalues corresponding to positive density rods
395: become $O(\rho^2)$ in the limit $\rho\to 0$.
396: Then, if there are two positive density rods,
397: one of the other eigenvalue of ${\bf g}$ must become infinite
398: as $O(\rho^{-2})$ in the limit $\rho \to0$.
399: We call it a negative density rod.
400: %A rod structure shows that,
401: %if the positive (negative) density rod exists,
402: %its metric component is $0$ (infinity).
403:
404: A method for constructing a seed metric corresponding to a given rod structure
405: is given as follows.
406: First, we put a minus sign to $t$-$t$ component and
407: $\rho^2$ to the numerator of
408: % the diagonal component
409: %which corresponds to
410: the component having rod at the left-end ($z=-\infty$).
411: Next, starting with the left-end,
412: we add $\mu_i$ to the numerator of the corresponding diagonal component
413: if the positive density rod appears at $a_i$ or
414: if the negative density rod disappears at $a_i$.
415: Similarly, we add $\mu_i$ to the denominator of
416: the corresponding diagonal component
417: if the positive density rod disappears at $a_i$ or
418: if the negative density rod appears at $a_i$.
419: %The metric $g$ which is obtained by the above operation
420: %corresponds to the rod structure which we want.
421: Constructing $f$ by eq.(\ref{F}),
422: we complete the construction of a seed metric.
423: %which is the one of metrics corresponding to
424: %the rod structure we want.
425: %Now, we check that
426: The metric ${\bf g}$ obtained by the above operations
427: has to the rod structure.
428: Suppose a positive (or negative) rod exists on the left of $a_i$.
429: When the value of $z$ crosses $a_i$ from the left,
430: the leading power in $\rho$ of $\mu_i$ changes from $O(1)$ to $O(\rho^2)$ near the rod.
431: Since the component corresponding to the positive (negative) density rod
432: should behave as $O(\rho^2)$ ($O(\rho^{-2})$),
433: %the rod disappears over $z=a_i$ if the positive (negative) density rod exists before $a_i$
434: %and if its component has $\mu_i$ in the denominator (numerator).
435: we can make the positive (negative) density rod
436: appearing from $z=a_i$
437: by adding $\mu_i$ to the numerator (denominator)
438: of the metric component.
439: The disappearance of rod can be explained similarly.
440:
441: % 図の挿入
442: \begin{figure}[tbp]
443: \begin{center}
444: \includegraphics[keepaspectratio=true,height=25mm]{BR4.eps}
445: \end{center}
446: \caption{the rod structure of the orthogonal black di-ring solution we want.}
447: %{}内にタイトルを記入してください
448: \label{fig:BR4.eps}
449: \end{figure}
450:
451: Next, we construct the seed metric of the orthogonal black di-ring.
452: The rod structure of the orthogonal black di-ring solution is as shown in fig.\ref{fig:BR4.eps}.
453: The thick solid lines correspond to the positive density rods,
454: while the dashed lines correspond to the negative density rods.
455: Naively one may think that we need to prepare the seed metric as the one corresponding
456: to this rod structure directly.
457: However, referring to the construction
458: of the black saturn solution~\cite{saturn},
459: we expect that we cannot add the angular momentum of $S^1$-plane of black ring
460: starting from the seed metric of this rod structure.
461: Following the case of black saturn,
462: we introduce a negative density rod as shown in fig.\ref{fig:BR1.eps}.
463: Then the metric corresponding to this rod structure can be written as
464: \begin{eqnarray}
465: {\bf g_0}=\mbox{diag} \left\{
466: -\frac{\mu_1 \mu_5}{\mu_3\mu_7} ,
467: \frac{\rho^2 \mu_3 \mu_7}{ \mu_2 \mu_4 \mu_6},
468: \frac{\mu_2\mu_4\mu_6}{\mu_1\mu_5}
469: \right\},\label{seedg}
470: \end{eqnarray}
471: where the constants $a_i$ contained in $\mu_i$ are ordered such that
472: $a_i \ge a_j$ for $i>j$.
473: Then $f_0$ becomes
474: \begin{eqnarray}
475: &&f_0=k^2 \frac{\mu_2\mu_4\mu_6}{\mu_1\mu_5} (\rho^2 +\mu_1\mu_2)(\rho^2 +\mu_1\mu_3)
476: (\rho^2 +\mu_1\mu_4)
477: (\rho^2 +\mu_1\mu_6)(\rho^2 +\mu_1\mu_7)(\rho^2 +\mu_2\mu_3)
478: (\rho^2 +\mu_2\mu_5)
479: \nonumber\\
480: &&\qquad\times
481: (\rho^2 +\mu_2\mu_7)(\rho^2 +\mu_3\mu_4)(\rho^2 +\mu_3\mu_5)
482: (\rho^2 +\mu_3\mu_6)
483: (\rho^2 +\mu_4\mu_5)(\rho^2 +\mu_4\mu_7)(\rho^2 +\mu_5\mu_6)
484: (\rho^2 +\mu_5\mu_7)
485: \nonumber\\
486: &&\qquad\times
487: (\rho^2 +\mu_6\mu_7)
488: (\rho^2 +\mu_1\mu_5)^{-2}(\rho^2 +\mu_2\mu_4)^{-2}
489: (\rho^2 +\mu_2\mu_6)^{-2}
490: (\rho^2 +\mu_3\mu_7)^{-2}(\rho^2 +\mu_4\mu_6)^{-2}
491: \prod_{i=1}^7 (\rho^2 +\mu_i^2)^{-1}
492: .\label{seedf}
493: \end{eqnarray}
494: %The rod structure of this metric is drawn as fig.(\ref{fig:BR1.eps}).
495: This solution has singularities on the rods $z\in[a_1,a_2]$ and $z\in[a_6,a_7]$ ,and
496: is of no physical interest by itself.
497: However, once soliton transformation is applied to this solution appropriately,
498: the negative density rod moves to the $t$-direction
499: and cancels the positive density rod
500: on the segments $z \in [a_1 ,a_2]$ and $z\in [a_6 , a_7]$.
501: Although in general it leaves the singularities at $z=a_1$ and $z=a_7$,
502: we will find that we can remove these singularities
503: by choosing the appropriate BZ vectors.
504: % 図の挿入
505: \begin{figure}[tbp]
506: \begin{center}
507: \includegraphics[keepaspectratio=true,height=30mm]{BR1.eps}
508: \end{center}
509: \caption{the rod structure of the seed metric of the orthogonal black di-ring
510: the solid (dashed) lines means positive (negative) density rods.}
511: %{}内にタイトルを記入してください
512: \label{fig:BR1.eps}
513: \end{figure}
514:
515: We must construct the matrix $\Psi_{g_0}$
516: solving eq.(\ref{psi1}) and eq. (\ref{psi2}).
517: Although it seems at first sight difficult
518: to solve these equations,
519: it is easy to obtain one of the solutions in fact.
520: The method is as follows.
521: %In the component of $g_0$ which has $\rho^2$,
522: We eliminate $\rho^2$ using $\mu_i \bar \mu_i=-\rho^2$.
523: Here we can use any $\mu_i$ other than those used in soliton transformation\footnote{
524: You might worry that, if we replace $-\rho^2$ with $\mu_i \bar \mu_i$ of the different $\mu_i$,
525: the different $\Psi$ is obtained.
526: Then the difference of the component in $\Psi^{-1}(\lambda=\mu_k)$ is any factor.
527: In the new metric, $\Psi$ appears in the only form of $m_a^{(k)}\Psi^{-1}_{ab}(\mu_k)$.
528: These facts mean the difference of $\Psi$ can be finally absorbed in the parameters in BZ vectors.}.
529: Then basically we only have to change $ \mu_i$ in ${\bf g_0}$ into
530: $( \mu_i- \lambda)$ to get $\Psi_{g_0}$.
531: However, if we perform a soliton transformation with $ \mu_i$ when ${\bf g_0}$ has $ \mu_i$,
532: $\Psi_{g_0}( \mu_i,\rho,z)$ or $\Psi_{g_0}^{-1}( \mu_i,\rho,z)$
533: becomes infinity.
534: In order to avoid this pathology,
535: we replace $\mu_i$ contained in ${\bf g_0}$
536: with $-\mu_k \bar\mu_k / \bar\mu_i$ .
537: Then,
538: we obtain the matrix $\Psi_{g_0}(\lambda,\rho,z)$ which has no pathology
539: at $ \lambda= \mu_i$.
540: %where $ \mu_i$ are solitons or antisolitons used in
541: %succeeding transformation.
542:
543:
544:
545: \section{soliton transformation}\label{trans}
546:
547: In this section, we explain the soliton transformation
548: to obtain the orthogonal black di-ring solution
549: following the construction of the black saturn solution~\cite{saturn}.
550: The orthogonal black di-ring solution can be obtained by
551: performing the following soliton transformation on the
552: seed metric (\ref{seedg}).
553:
554: We perform a 2-soliton transformation\footnote{
555: The metric obtained by 2-soliton transformation is the same as that
556: obtained by two 1-soliton transformations with the same solitons
557: and the same BZ vectors.
558: The calculation of two 1-soliton transformations is easier than that of
559: 2-soliton transformation.}
560: with
561: $\mu'_{i}=\{\mu_1 ,\bar\mu_7\}$ and both of BZ vectors being
562: $(1,0,0)$.
563: The metric obtained by this transformation is
564: \begin{eqnarray}
565: {\bf g'_{1}}= \mbox{diag} \left\{
566: -\frac{\mu_5 \mu_7}{\mu_3\mu_1} ,
567: \frac{\rho^2 \mu_3 \mu_7}{ \mu_2 \mu_4 \mu_6},
568: \frac{\mu_2\mu_4\mu_6}{\mu_1\mu_5}
569: \right\}.
570: \end{eqnarray}
571: We rescale this metric as
572: \begin{eqnarray}
573: {\bf {\tilde g'}_{1}} = \frac{\mu_1}{\mu_7} {\bf g'_{1}}=
574: \mbox{diag} \left\{
575: -\frac{\mu_5}{\mu_3} ,
576: \frac{\rho^2 \mu_1 \mu_3}{ \mu_2 \mu_4 \mu_6},
577: \frac{\mu_2\mu_4\mu_6}{\mu_5\mu_7}
578: \right\}.\label{tildeg}
579: \end{eqnarray}
580: This rescaling makes the calculation easier,
581: although it does not change the result\footnote{
582: Due to the rescaling,
583: $\Psi$ is also rescaled as
584: $\Psi'_0= (\mu_1-\lambda)/(\mu_7-\lambda)\tilde\Psi_0$.
585: This affect $\Gamma_{kl}$ of eq.(\ref{Gamma}) and $N^{(k)}$ of eq.(\ref{N}) as
586: $\Gamma_{kl} \to (\mu_1-\mu_k)(\mu_1-\mu_l)/((\mu_7-\mu_k)(\mu_7-\mu_l))\Gamma_{kl}$
587: and $N^{(k)} \to (\mu_1-\mu_k)/(\mu_7-\mu_k) N^{(k)}$.
588: We can find from eq.(\ref{hatg}) that in the new metric the effects cancel each other.
589: }.
590: We make $\Psi$ corresponding to ${\bf \tilde g'_1}$ as
591: %From this matrix,
592: %we made $\Psi$, which becomes
593: \begin{eqnarray}
594: &&
595: \Psi_{\tilde g'_1} =
596: \mbox{diag}\biggl\{
597: -\frac{(\mu_5-\lambda)}{(\mu_3-\lambda)} ,
598: -\frac{ (\mu_1-\lambda) (\mu_3-\lambda)(\bar\mu_4-\lambda)}
599: { (\mu_2-\lambda) (\mu_6-\lambda)}
600: ,\frac{(\mu_2-\lambda)(\mu_6-\lambda)(\bar\mu_7-\lambda)}
601: {(\mu_5-\lambda)(\bar\mu_4-\lambda)}
602: \biggr\}.
603: \end{eqnarray}
604: Next, we apply the 2-soliton transformation to ${\bf \tilde g'}$ with $\bar \mu'_i$.
605: %in which one of the solitons is $\bar\mu_1$ with BZ vector
606: The BZ vectors associated with $\bar\mu_1$ and $\mu_7$ are chosen to be
607: (1,0,c) and
608: %and the other is $\mu_7$ with BZ vector
609: (1,b,0), respectively.
610: We denote the resulting metric by ${\bf \tilde g}$.
611: Finally, to undo the rescaling in eq.(\ref{tildeg})
612: %from $\tilde g_0$ to $g'_0$,
613: we rescale ${\bf \tilde g}$ as
614: \begin{eqnarray}
615: {\bf g}= \frac{\mu_7}{\mu_1}{\bf \tilde g}.
616: \end{eqnarray}
617:
618: We must also compute $f$ given in eq.(\ref{f}).
619: %transform $f_0$ into $f$.
620: %Although this transformation need $\Gamma_{kl}$,
621: $\Gamma_{kl}$, which is necessary to compute eq.(\ref{f}),
622: is obtained in the process of constructing ${\bf \tilde g}$.
623: Moreover, $\Gamma_{kl}^{(0)}$ is given by
624: $\Gamma_{kl}^{(0)} =\Gamma_{kl}|_{b=0,c=0}$.
625: %As a result, from eq.(\ref{f})
626: %we can construct $f$.
627:
628: %\section{orthogonal black di-ring solution}\label{solution}
629:
630: The obtained metric
631: %of the orthogonal black di-ring solution
632: is given by eq.(\ref{metric}) with
633: %written as
634: %\begin{eqnarray}
635: %ds ^2 = f(d\rho^2 + dz^2) + g_{ab} dx^a dx^b ,
636: %\end{eqnarray}
637: $a,b=t,\psi,\phi$ and
638:
639: \begin{eqnarray}
640: &&
641: f=Cf_0 H/F,
642: \\
643: &&
644: H=F +b^2 F_{(b)} +c^2 F_{(c)} + b^2 c^2 F_{{(bc)}},
645: \\
646: &&
647: F = -\frac{\mu_5^2(\mu_3-\mu_7)^2(\mu_1 \mu_3 +\rho^2)^2
648: (\mu_1 \mu_7 +\rho^2)^2}{\mu_3^2 (\mu_5-\mu_7)^2
649: (\mu_1\mu_5 +\rho^2)^2 (\mu_1^2 +\rho^2)(\mu_7^2+\rho^2)\rho^4},
650: \\
651: &&
652: F_{(b)}=-\frac{\mu_1\mu_4\mu_5(\mu_2-\mu_7)^2
653: (\mu_6-\mu_7)^2(\mu_1\mu_3+\rho^2)^2}
654: {\mu_2\mu_6(\mu_3-\mu_7)^2
655: (\mu_1\mu_5 +\rho^2)^2(\mu_4\mu_7+\rho^2)^2
656: (\mu_1^2+\rho^2)(\mu_7^2+\rho^2)},
657: \\
658: &&
659: F_{(c)}=-\frac{\mu_1^2 \mu_2 \mu_6 \mu_7
660: (\mu_1-\mu_4)^2(\mu_3-\mu_7)^2 (\mu_1\mu_5+\rho^2)^2}
661: {\mu_3\mu_4(\mu_5-\mu_7)^2
662: (\mu_1\mu_2 +\rho^2)^2(\mu_1\mu_6+\rho^2)^2
663: (\mu_1^2+\rho^2)(\mu_7^2+\rho^2)\rho^2},
664: \\
665: &&
666: F_{(bc)}=\frac{\mu_1^3 \mu_3 \mu_7
667: (\mu_1-\mu_4)^2(\mu_2-\mu_7)^2
668: (\mu_6-\mu_7)^2 (\mu_1\mu_5+\rho^2)^2}
669: {\mu_5(\mu_1-\mu_7)^2(\mu_3-\mu_7)^2
670: (\mu_1\mu_2 +\rho^2)^2(\mu_1\mu_6+\rho^2)^2(\mu_4\mu_7+\rho^2)^2
671: (\mu_1^2+\rho^2)(\mu_7^2+\rho^2)},
672: \end{eqnarray}
673: \begin{eqnarray}
674: &&
675: g_{tt}=H^{-1}(A +b^2 A_{(b)} +c^2 A_{(c)} +b^2 c^2 A_{(bc)}),
676: \\
677: &&
678: A=\frac{\mu_1 \mu_5^3(\mu_3-\mu_7)^2(\mu_1\mu_3+\rho^2)^2
679: (\mu_1\mu_7+\rho^2)^2}
680: {\mu_3^3\mu_7(\mu_5-\mu_7)^2
681: (\mu_1\mu_5+\rho^2)^2(\mu_1^2+\rho^2)(\mu_7^2+\rho^2)\rho^4},
682: \\
683: &&
684: A_{(b)}=-\frac{\mu_1^2 \mu_4 \mu_5^2\mu_7(\mu_2-\mu_7)^2(\mu_6-\mu_7)^2
685: (\mu_1\mu_3+\rho^2)^2}
686: {\mu_2\mu_3\mu_6(\mu_3-\mu_7)^2
687: (\mu_1\mu_5+\rho^2)^2(\mu_4\mu_7+\rho^2)^2
688: (\mu_1^2+\rho^2)(\mu_7^2+\rho^2)\rho^2},
689: \\
690: &&
691: A_{(c)}=-\frac{\mu_1 \mu_2 \mu_5 \mu_6(\mu_1-\mu_4)^2(\mu_3-\mu_7)^2
692: (\mu_1\mu_5+\rho^2)^2}
693: {\mu_3^2\mu_4(\mu_5-\mu_7)^2
694: (\mu_1\mu_2+\rho^2)^2(\mu_1\mu_6+\rho^2)^2
695: (\mu_1^2+\rho^2)(\mu_7^2+\rho^2)},
696: \\
697: &&
698: A_{(bc)}=-\frac{\mu_1^2 \mu_7^2(\mu_1-\mu_4)^2(\mu_2-\mu_7)^2
699: (\mu_6-\mu_7)^2(\mu_1\mu_5+\rho^2)^2}
700: {(\mu_1-\mu_7)^2(\mu_3-\mu_7)^2
701: (\mu_1\mu_2+\rho^2)^2(\mu_1\mu_6+\rho^2)^2(\mu_4\mu_7+\rho^2)^2
702: (\mu_1^2+\rho^2)(\mu_7^2+\rho^2)},
703: \end{eqnarray}
704: \begin{eqnarray}
705: &&
706: g_{\psi\psi}= B+H^{-1}(b^2 B_{(b)} +b^2 c^2 B_{(bc)}),
707: \\
708: &&
709: B=\frac{\mu_3 \mu_7 \rho^2}
710: {\mu_2 \mu_4 \mu_6},
711: \\
712: &&
713: B_{(b)}=\frac{\mu_1 \mu_3\mu_5(\mu_2-\mu_7)^2(\mu_6-\mu_7)^2
714: (\mu_1\mu_3+\rho^2)^2 \rho^2}
715: {\mu_2^2\mu_6^2\mu_7(\mu_3-\mu_7)^2
716: (\mu_1\mu_5+\rho^2)^2(\mu_4\mu_7+\rho^2)^2(\mu_1^2+\rho^2)},
717: \\
718: &&
719: B_{(bc)}=-\frac{\mu_1^3 \mu_3^2 (\mu_1-\mu_4)^2(\mu_2-\mu_7)^2
720: (\mu_6-\mu_7)^2 (\mu_1\mu_5+\rho^2)^2 \rho^2}
721: {\mu_2\mu_4\mu_5\mu_6(\mu_1-\mu_7)^2(\mu_3-\mu_7)^2
722: (\mu_1\mu_2+\rho^2)^2(\mu_1\mu_6+\rho^2)^2
723: (\mu_4\mu_7+\rho^2)^2(\mu_1^2+\rho^2)},
724: \end{eqnarray}
725:
726: \begin{eqnarray}
727: &&
728: g_{\phi\phi}=C+ H^{-1}(c^2 C_{(c)} +b^2 c^2 C_{(bc)}),
729: \\
730: &&
731: C=\frac{\mu_2 \mu_4 \mu_6}
732: {\mu_1 \mu_5},
733: \\
734: &&
735: C_{(c)}=\frac{\mu_1 \mu_2^2\mu_6^2\mu_7(\mu_1-\mu_4)^2(\mu_3-\mu_7)^2
736: (\mu_1\mu_5+\rho^2)^2 }
737: {\mu_3\mu_5(\mu_5-\mu_7)^2(\mu_1\mu_2+\rho^2)^2
738: (\mu_1\mu_6+\rho^2)^2(\mu_7^2+\rho^2) \rho^4},
739: \\
740: &&
741: C_{(bc)}=-\frac{\mu_1^2 \mu_2\mu_3\mu_4\mu_6\mu_7(\mu_1-\mu_4)^2(\mu_2-\mu_7)^2
742: (\mu_6-\mu_7)^2(\mu_1\mu_5+\rho^2)^2 }
743: {\mu_5^2(\mu_1-\mu_7)^2(\mu_3-\mu_7)^2(\mu_1\mu_2+\rho^2)^2
744: (\mu_1\mu_6+\rho^2)^2(\mu_4\mu_7+\rho^2)^2(\mu_7^2+\rho^2) \rho^2},
745: \end{eqnarray}
746:
747: \begin{eqnarray}
748: &&
749: g_{t\psi} = H^{-1} b(D +c^2 D_{(c)}),
750: \\
751: &&
752: D=\frac{\mu_1 \mu_5^2 (\mu_2-\mu_7)(\mu_6-\mu_7)
753: (\mu_1\mu_3+\rho^2)^2 (\mu_1\mu_7+\rho^2)}
754: {\mu_2\mu_3\mu_6\mu_7(\mu_5-\mu_7)
755: (\mu_1\mu_5+\rho^2)^2(\mu_4\mu_7+\rho^2) (\mu_1^2+\rho^2) \rho^2},
756: \\
757: &&
758: D_{(c)}=-\frac{\mu_1^2 (\mu_1-\mu_4)^2 (\mu_2-\mu_7)(\mu_6-\mu_7)
759: (\mu_1\mu_5+\rho^2)^2 }
760: {\mu_4(\mu_1-\mu_7)(\mu_5-\mu_7)
761: (\mu_1\mu_2+\rho^2)^2(\mu_1\mu_6+\rho^2)^2
762: (\mu_4\mu_7+\rho^2)(\mu_1^2+\rho^2)},
763: \end{eqnarray}
764:
765: \begin{eqnarray}
766: &&
767: g_{t\phi} = H^{-1} c(E +b^2 E_{(b)}),
768: \\
769: &&
770: E=\frac{ \mu_2\mu_5\mu_6 (\mu_1-\mu_4)(\mu_3-\mu_7)^2
771: (\mu_1\mu_3+\rho^2) (\mu_1\mu_7+\rho^2)}
772: {\mu_3^2(\mu_5-\mu_7)^2(\mu_1\mu_2+\rho^2)
773: (\mu_1\mu_6+\rho^2)(\mu_7^2+ \rho^2)\rho^4},
774: \\
775: &&
776: E_{(b)}=-\frac{\mu_1 \mu_4\mu_7 (\mu_1-\mu_4) (\mu_2-\mu_7)^2
777: (\mu_6-\mu_7)^2 (\mu_1\mu_3+\rho^2) }
778: {(\mu_1-\mu_7)(\mu_3-\mu_7)^2(\mu_1\mu_2+\rho^2)
779: (\mu_1\mu_6+\rho^2)(\mu_4\mu_7+\rho^2)^2(\mu_7^2+\rho^2)\rho^2 },
780: \end{eqnarray}
781:
782: \begin{eqnarray}
783: g_{\psi\phi}=H^{-1} bc \frac{\mu_1 (\mu_1-\mu_4) (\mu_2-\mu_7)
784: (\mu_6-\mu_7) (\mu_1\mu_3+\rho^2) }
785: {(\mu_1-\mu_7)(\mu_5-\mu_7)(\mu_1\mu_2+\rho^2)
786: (\mu_1\mu_6+\rho^2)(\mu_4\mu_7+\rho^2)\rho^2 }.
787: \end{eqnarray}
788:
789: This solution has generally point singularities at
790: $(\rho,z)=(0,a_1)$ and $(0,a_7)$
791: and conical singularities on the rods $z \in[-\infty,a_2]$,
792: $z \in[a_3,a_4]$, $z \in[a_4,a_5]$ and $z \in[a_6,\infty]$.
793: However, in the next section, we show
794: these singularities can be removed if we choose the parameters appropriately.
795: When we remove the singularities,
796: this solution becomes the orthonormal black di-ring solution.
797:
798:
799:
800:
801: \section{analysis}\label{analysis}
802:
803: First, we analyse the rod structure of the orthogonal black di-ring solution.
804: Although
805: generally singularities appear at $z=a_1$ and $a_7$,
806: we can remove these singularities by choosing parameters $c$ and $b$
807: appropriately.
808: After setting the parameters $b$ and $c$ to the particular values that
809: eliminate the singularities at $z=a_1$ and $a_7$,
810: we next analyse the asymptotic behavior of this solution.
811: If we choose $C \kappa^2=1$, we will find that this solution becomes asymptotically flat
812: %when the solution has no singularity at $z=a_1$ and $z=a_7$.
813: Finally, we analyse the conical structure around the axis.
814: We will find that all conical singularities can be removed by choosing
815: the parameters appropriately.
816: As a result, four parameters are left in the end.
817:
818:
819:
820: \subsection{rod structure}
821:
822: % 図の挿入
823: \begin{figure}[tbp]
824: \begin{center}
825: \includegraphics[keepaspectratio=true,height=30mm]{BR2.eps}
826: \end{center}
827: \caption{General rod structure after the transformation.
828: Although this solution has the singularities at $z=a_1$ and $z=a_7$,
829: we can remove these singularities
830: if we choose the appropriate parameters.}
831: %{}内にタイトルを記入してください
832: \label{fig:BR2.eps}
833: \end{figure}
834:
835: The rod structure of the obtained metric with general parameters
836: is illustrated in fig.\ref{fig:BR2.eps}.
837: % the representation of \cite{Harmark},
838: We calculate the direction of the rod
839: which is defined in Ref.\cite{Harmark}.
840: The direction of the rod represents the vector whose norm
841: becomes zero at $\rho=0$.
842: The semi-infinite rod $z\in [-\infty,a_2]$ and
843: the finite rod $z[a_3,a_4]$ have directions $(0,1,0)$
844: which correspond to the $\psi$-axis.
845: The semi-infinite rod $z\in [a_6,\infty]$ and
846: the finite rod $z[a_4,a_5]$ have directions $(0,0,1)$
847: which correspond to the $\phi$-axis.
848: The finite rods $z[a_2,a_3]$ and $z[a_5,a_6]$, which
849: correspond to the location of the black ring horizons,
850: have directions
851: \begin{eqnarray}
852: v=(1,\Omega_\Psi^{(1)},\Omega_\Phi^{(1)}) \qquad \mbox{and}
853: \qquad u=(1,\Omega_\Psi^{(2)},\Omega_\Phi^{(2)}), \label{v}
854: \end{eqnarray}
855: respectively,
856: where
857: \begin{eqnarray}
858: &&\Omega_\psi^{(1)}=-\frac{b\, a_{75}\, a_{76}}{2\, a_{73}^{\ 2}\, a_{71}},\qquad
859: \Omega_\phi^{(1)}=-\frac{c\, a_{51}^{\ 2}}{2\, a_{31}\, a_{61}\, a_{71}},
860: \\
861: &&\Omega_\psi^{(2)}=-\frac{b\, a_{76}}{2\, a_{71}\, a_{74}}, \qquad
862: \Omega_\phi^{(2)}=-\frac{c\, a_{41}}{2\, a_{61}\, a_{71}},
863: \\
864: &&
865: a_{ij}\equiv a_i-a_i.
866: \end{eqnarray}
867: This solution has no negative density rod as anticipated.
868: The negative density rod in the seed metric was cancelled by the positive density rod in $t$-direction.
869: However, the soliton transformation leaves the singularities at
870: $z=a_1$ and $a_7$.
871: We find that the metric component $g_{\phi\phi}$ becomes the singular at $z=a_1$.
872: %On the rod $z\in [-\infty,a_1]$,
873: %$g_{\phi\phi}$ is written as
874: %\begin{eqnarray}
875: %g_{\phi\phi}=\frac{-c^2 (\mu_1-\mu_4)^2 \mu_5^2 +\mu_2\mu_3\mu_4\mu_6\mu_7}
876: %{\mu_1\mu_3\mu_5\mu_7}.
877: %\end{eqnarray}
878: %If $z\to a_1$, this goes to infinity
879: %because $\mu_1 = a_1-z$ on the rod $z\in [-\infty,a_1]$.
880: In order to see this,
881: we perform the coordinate transformation,
882: \begin{eqnarray}
883: \rho= r \sin \theta, \qquad z = r \cos \theta + a_1.
884: \end{eqnarray}
885: In the limit $r \to 0$ (which means $\rho \to 0$ and
886: $z\to a_1$), the leading term of $g_{\phi\phi}$ becomes
887: \begin{eqnarray}
888: g_{\phi\phi} \to 4
889: \frac{a_{21}\, a_{41}\, a_{61}\, (-c^2\, a_{41}\, a_{51}^{\ 2}\,
890: +2\, a_{21}\, a_{31}\, a_{61}\, a_{71})\sin\theta(1-\cos\theta)}
891: {c^2\, a_{41}\, a_{51}^{\ 3}\, + 2\, a_{12}\, a_{13}\,
892: a_{51}\, a_{61}\, a_{71} \sin^2\theta (1-\cos\theta)^2} r^{-1}.
893: \end{eqnarray}
894: Obviously, $g_{\phi\phi}\to\infty$ for $r\to 0$.
895: However, this singularity can be removed
896: if we choose the parameter as
897: \begin{eqnarray}
898: c^2 = \frac{2\, a_{21}\, a_{31}\, a_{61}\, a_{71}}
899: {a_{41}\, a_{51}^{\ 2}}. \label{c}
900: \end{eqnarray}
901: Then the term proportion to $r^{-1}$ in $g_{\phi\phi}$ becomes $0$,
902: and the leading term of $g_{\phi\phi}$ becomes
903: \begin{eqnarray}
904: g_{\phi\phi} \to
905: 2\frac{a_{31}\, a_{41}\, a_{51}\, a_{61}\, a_{71}
906: +a_{21}\, a_{41}\, a_{61}\, a_{71}\, a_{53}
907: +a_{21}\, a_{31}\, a_{61}\, a_{71}\, a_{54}
908: +a_{21}\, a_{31}\, a_{41}\, a_{51}\, a_{61}
909: +a_{21}\, a_{31}\, a_{41}\, a_{51}\, a_{71}}
910: {a_{31}\, a_{51}^{\ 2}\, a_{71}}.
911: \end{eqnarray}
912: In this limit, the other components of $g$ and $f$ become
913: \begin{eqnarray}
914: &&g_{tt}=O(r), \qquad
915: g_{\psi\psi}\to \frac{a_{31}\, a_{71}}{2\, a_{21}\, a_{41}\, a_{61}} r^2 \sin^2\theta,
916: \qquad
917: g_{t\psi}=-\frac{b\, a_{31}\, a_{72}\, a_{75}\, a_{76}}
918: {4\, a_{21}\, a_{41}\, a_{61}\, a_{71}\, a_{73}^{\ 2}}
919: r^2 \sin^2 \theta,
920: \nonumber\\
921: &&
922: g_{t\phi}\to c\frac{a_{41}\, a_{51}}{a_{31}\, a_{71}},
923: \qquad
924: g_{\psi\phi}=\frac{c\, a_{41}\, a_{51}}{a_{31}\, a_{71}} r^2 \sin^2 \theta \qquad
925: \mbox{and} \qquad f\to \frac{a_{31}\, a_{71}}{2\, a_{21}\, a_{41}\, a_{61}}.
926: \end{eqnarray}
927: We find that, in this limit,
928: % the $\psi$ is orthogonal to $t$-$\phi$ plane
929: %and
930: $g_{t\phi}$ and $g_{\phi\phi}$ are constant.
931: Moreover, when we introduce new variable
932: \begin{eqnarray}
933: \psi_1= \psi-\frac{b\, a_{72}\, a_{75}\, a_{76}}
934: {2\, a_{71}^{\ 2}\, a_{73}^{\ 2}}t
935: +\frac{2\, c\, a_{21}\, a_{41}^{\ 2}\, a_{51}\, a_{61}}
936: {a_{31}^{\ 2}\, a_{71}^{\ 2}}\phi.,
937: \end{eqnarray}
938: the metric becomes
939: \begin{eqnarray}
940: ds^2 \simeq f(dr^2 +r^2 d\theta^2 + r^2 \sin^2\theta d{\psi_1}^2)
941: +2 g_{t\phi} dt d\phi +g_{\phi\phi} d\phi^2,
942: \end{eqnarray}
943: Since this metric is the Minkowski metric,
944: we find that there is no singularity at this point.
945:
946: In a similar way, we can remove the singularity at $z=a_7$
947: if we choose the parameter as\footnote{
948: If the seed metric is exchanged using $\mu_i\bar\mu_i=-\rho^2$,
949: the result by a soliton transformation does not change
950: as we show the footnote-1.
951: When we write the seed metric with only $\bar\mu_i$ using $\mu_i\bar\mu_i=-\rho^2$,
952: the original seed metric and the transformed seed metric is symmetric as
953: $\mu_i \to \bar \mu_{7-i}$ and $\psi \leftrightarrow \phi$.
954: From this, we find that, when we can remove a singularity at $z=a_1$,
955: we can have singularity appear at $z=a_7$.}
956: \begin{eqnarray}
957: &&b^2 = \frac{2\, a_{71}\, a_{73}^{\ 2}\, a_{74}}
958: {a_{72}\, a_{75}\, a_{76}}.\label{b}
959: \end{eqnarray}
960: With these conditions (\ref{c}) and (\ref{b}),
961: the metric is smooth across $z=a_1$ and $a_7$.
962: Then the rod structure is as illustrated in fig.\ref{fig:BR3.eps}.
963:
964: % 図の挿入
965: \begin{figure}[tbp]
966: \begin{center}
967: \includegraphics[keepaspectratio=true,height=33mm]{BR3.eps}
968: \end{center}
969: \caption{The rod structure of the orthogonal black di-ring,
970: which is obtained by removing the singularities by fixing $c$ and $b$.
971: The vectors accompanied with rods are such defined in Ref.~\cite{Harmark}.}
972: %{}内にタイトルを記入してください
973: \label{fig:BR3.eps}
974: \end{figure}
975:
976:
977: \subsection{asymptotic structure}
978:
979: To analyse the asymptotic structure,
980: we take other coordinates defined by
981: \begin{eqnarray}
982: \rho= \frac{1}{2} r^2 \sin{2\theta},
983: \qquad
984: z= \frac{1}{2} r^2 \cos{2\theta},
985: \end{eqnarray}
986: where $0< \theta< \pi/2$.
987: Then in the asymptotic limit $r \to \infty$, we have
988: $g_{t\psi}\to 0$, $g_{t\phi}\to 0$ and $g_{\psi\phi}\to 0$.
989: Since
990: \begin{eqnarray}
991: d\rho^2 +dz^2 =r^2 (dr^2 +r^2 d\theta^2),
992: \end{eqnarray}
993: and the asymptotically value of $f$ becomes
994: \begin{eqnarray}
995: f=C\kappa^2\frac{1}{r^2} +\cdots,
996: \end{eqnarray}
997: we find we must choose $C\kappa^2=1$
998: in order to get the asymptotic flat solution .
999: Setting $C\kappa^2=1$ the asymptotic metric becomes
1000: \begin{eqnarray}
1001: ds^2 = -dt^2 +dr^2 + r^2 d\theta^2
1002: +r^2 \cos^2\theta d\psi^2
1003: +r^2 \sin^2\theta d\phi^2.
1004: \end{eqnarray}
1005: The asymptotic flatness also requires
1006: that the angles $\psi$ and $\phi$ should have periodicities
1007: \begin{eqnarray}
1008: \Delta\psi =\Delta\phi =2\pi.
1009: \end{eqnarray}
1010: %Therefore, the solution is asymptotically flat.
1011:
1012: \subsection{regularities on the axis}
1013:
1014: In order to avoid conical singularities,
1015: the period $\Delta\eta$ of the spacelike coordinate $\eta$($=\psi$ or $\phi$)
1016: corresponding to the angle around each positive density rod
1017: must satisfy~\cite{Harmark}
1018: \begin{eqnarray}
1019: \Delta\eta= 2\pi \lim_{\rho\to 0}
1020: \sqrt{\frac{\rho^2 f}{g_{\eta\eta}}}.
1021: \end{eqnarray}
1022: %where $\eta$ is $\psi$ or $\phi$.
1023: When we impose the conditions (\ref{c}) and (\ref{b}),
1024: the regularity conditions on the rods $z\in [-\infty,a_2]$
1025: and $z\in [a_6,\infty]$
1026: are automatically satisfied for $\Delta\psi=\Delta\phi=2\pi$
1027: %fixes the period of $\psi$ to be $\Delta\psi=2\pi$ and
1028: %the regularity condition on the rod $z\in [a_6,\infty]$
1029: %fixes the period of $\phi$ to be $\Delta\phi=2\pi$.
1030:
1031: Next we consider the regularity condition on the rod
1032: $z\in [a_3,a_4]$.
1033: The regularity condition is written as
1034: \begin{eqnarray}
1035: \Delta\psi = 2\pi \sqrt{\frac{
1036: a_{14}\, a_{16}\, a_{17}\, a_{25}\, a_{27}\, a_{34}\, a_{35}\, a_{36}}
1037: {a_{15}^{\ 2}\, a_{24}^{\ 2}\, a_{26}^{\ 2}\, a_{37}^{\ 2}}}.
1038: \end{eqnarray}
1039: Since $\Delta\psi$ must be $2\pi$,
1040: this implies
1041: \begin{eqnarray}
1042: 1 = \frac{
1043: a_{14}\, a_{16}\, a_{17}\, a_{25}\, a_{27}\, a_{34}\, a_{35}\, a_{36}}
1044: {a_{15}^{\ 2}\, a_{24}^{\ 2}\, a_{26}^{\ 2}\, a_{37}^{\ 2}}.
1045: \label{regpsi}
1046: \end{eqnarray}
1047: Similarly, the regularity condition on the
1048: rod $z\in [a_4,a_5]$
1049: implies
1050: \begin{eqnarray}
1051: 1 = \frac{a_{74}\, a_{72}\, a_{71}\, a_{63}\, a_{61}\, a_{54}\, a_{53}\, a_{52}}
1052: {a_{73}^{\ 2}\, a_{64}^{\ 2}\, a_{62}^{\ 2}\, a_{51}^{\ 2}}.
1053: \label{regphi}
1054: \end{eqnarray}
1055: Under the transformation,
1056: $a_i \to a_{7-i}$,
1057: eq.(\ref{regpsi}) becomes eq.(\ref{regphi}).
1058: This is expected because
1059: the solution we want has a symmetry
1060: corresponding to the exchange of $\psi$ and $\phi$.
1061:
1062: We must check the existence of the parameters
1063: which satisfy eq.(\ref{regpsi}) and eq.(\ref{regphi})
1064: with $a_i>a_j$ for $i>j$.
1065: Without loss of generality,
1066: we can set $a_4 =0$.
1067: For the sake of simplicity,
1068: we consider the symmetric case where
1069: \begin{eqnarray}
1070: a_1 = -a_7,
1071: \qquad
1072: a_2 =-a_6,
1073: \qquad
1074: a_3=-a_5.
1075: \end{eqnarray}
1076: In this case, eq.(\ref{regpsi}) and eq.(\ref{regphi})
1077: become the same equation.
1078: Moreover, we set $a_7 =1$,
1079: which corresponds to fixing the scale.
1080: Then, eq.(\ref{regpsi}) and eq.(\ref{regphi})
1081: are written as
1082: \begin{eqnarray}
1083: a_5^2(1+a_6)^2(a_5+a_6)^2 -(1+a_5)^4 a_6^4=0.
1084: \label{reg}
1085: \end{eqnarray}
1086: Here, we denote the left hand side of this equation
1087: by $f(a_6)$ as a function of $a_6$.
1088: Then
1089: \begin{eqnarray}
1090: &&f(a_5)=a_5^4 (1+a_5)^2(1-a_5)(3+a_5),
1091: \label{a5}\\
1092: &&f(1)= (1+a_5)^2(3a_5+1)(a_5-1).\label{1}
1093: \end{eqnarray}
1094: Since $a_5$ satisfies the condition
1095: $0<a_5<1$,
1096: $f(a_5)$ is always positive.
1097: On the other hand,
1098: $f(1)$ is always negative.
1099: Therefore the parameter $a_6$ which satisfies eq.(\ref{reg}) exists
1100: between $a_5$ and $a_7(=1)$.
1101: This means that the orthogonal black di-ring solution
1102: which has no singularity outside the horizon exists.
1103:
1104: \subsection{regularities at end points of rods}
1105:
1106: In this subsection, we show there is no singularity at the end points of rods.
1107: From the symmetry of this solution (see footnote-3),
1108: we only have to check the points $z=a_2$, $a_3$ and $a_4$.
1109: Near each point $(\rho,z)=(0,a_i)$, we transform $\rho$ and $z$ as
1110: \begin{eqnarray}
1111: \rho=\frac{1}{2}r^2 \sin2\theta, \qquad z=\frac{1}{2}r^2 \cos2\theta -a_i.
1112: \end{eqnarray}
1113:
1114: Near the point $(\rho,z)=(0,a_2)$,
1115: we take basis as $v$ (see eq.(\ref{v})), $q=(0,1,0)$ and $s=(0,0,1)$.
1116: In the limit $r \to 0$, inner products and $f$ become
1117: \begin{eqnarray}
1118: &&g(v,v)\to -4\frac{c^2\, a_{42}\, a_{62}\, a_{51}^{\ 4}}
1119: {a_{31}\, a_{52}\, a_{61}\, a_{71}}r^2 \sin^2\theta,
1120: \qquad
1121: g(q,q) \to \frac{a_{32}\, a_{72}}{a_{42}\, a_{62}}r^2 \cos^2 \theta, \qquad
1122: g(s,s) \to 4\frac{a_{21}\, a_{31}^{\ 2}\, a_{61}^{\ 2}\, a_{71}^{\ 2}\, a_{52}}
1123: {c^2\, a_{51}^{\ 4}\, a_{32}\, a_{72}},
1124: \nonumber\\
1125: && \!\!\!\!
1126: g(v,q)=O(r^4), \qquad
1127: g(v,s)=\alpha_2 r^2 \sin^2\theta, \qquad
1128: g(q,s)=\beta_2 r^2 \cos^2\theta, \qquad
1129: f\to \frac{c^2\, a_{41}\, a_{51}^{\ 2}\, a_{32}\, a_{72}}
1130: {2\, a_{21}\, a_{31}\, a_{61}\, a_{71}\, a_{42}\, a_{62}}r^{-2},
1131: \end{eqnarray}
1132: where $\alpha_2$ and $\beta_2$ are some constants.
1133: We find that $g(s,s)$ is constant.
1134: % and $v$, $q$ and $s$ are orthogonal basis in this limit.
1135: When bases are changed like
1136: \begin{eqnarray}
1137: \psi_2=\psi+\frac{\beta_2}{f}\phi
1138: \qquad \mbox{and} \qquad
1139: \eta_2= \frac{1}{f}
1140: \left(
1141: 4\frac{c^2 \, a_{42}\, a_{62}\, a_{51}^{\ 4}}
1142: {a_{31}\, a_{52}\, a_{61}\, a_{71}}\eta - \alpha_2\phi
1143: \right) ,
1144: \end{eqnarray}
1145: where $\eta$ is defined as $(\partial/\partial \eta)^a\equiv v^a$ ,
1146: the metric becomes
1147: \begin{eqnarray}
1148: ds^2 \simeq f(dr^2 +r^2 d\theta^2 + r^2 \cos^2\theta \psi_2^2-r^2 \sin^2\theta d\eta_2^2)
1149: + 4\frac{a_{21}\, a_{31}^{\ 2}\, a_{61}^{\ 2}\, a_{71}^{\ 2}\, a_{52}}
1150: {c^2\, a_{51}^{\ 4}\, a_{32}\, a_{72}} d\phi^2,\label{a2}
1151: \end{eqnarray}
1152: where we used eq.(\ref{c}).
1153: Since eq.(\ref{a2}) is locally the Minkowski metric,
1154: % by a coordinate transform,
1155: there is no singularity at the point $(\rho,z)=(0,a_2)$.
1156:
1157: Similarly, we can obtain the metric near the point $(\rho,z)=(0,a_3)$.
1158: The metric around this point becomes
1159: \begin{eqnarray}
1160: ds^2 \simeq \frac{a_{32}\, a_{73}}{a_{43}\, a_{63}}
1161: (dr^2 +r^2 d\theta^2 + r^2 \sin^2\theta d\psi^2-r^2 \cos^2\theta d\eta_3^2)
1162: + 2\frac{a_{31}\, a_{43}\, a_{63}}{a_{32}\, a_{53}} d\phi^2,
1163: %\\
1164: %&&g(v,q)=O(r^4), \qquad
1165: %g(v,s)=O(r^4),
1166: %\qquad
1167: %g(q,s)=O(r^4),
1168: \end{eqnarray}
1169: where
1170: \begin{eqnarray}
1171: \eta_3= \frac{2\, a_{31}\, a_{43}\, a_{53}\, a_{63}}
1172: {a_{32}\, a_{73}^{\ 2}}\eta.
1173: \end{eqnarray}
1174: %where $\alpha_3$ and $\beta_3$ are some constants defined as in similar way to
1175: %$\alpha_2$ and $\beta_2$.
1176: From this, we find there is no singularity at the point $(\rho,z)=(0,a_3)$.
1177:
1178: Finally we check the point at $(\rho,z)=(0,a_4)$.
1179: The metric around this point becomes
1180: \begin{eqnarray}
1181: &&ds^2 \simeq \frac{a_{24}\, a_{74}}{a_{43}\, a_{63}}
1182: (dr^2 +r^2 d\theta^2 + r^2 \sin^2\theta d\psi_4^2
1183: +r^2 \cos^2\theta d\phi_4^2)
1184: -\frac{a_{34}\, a_{54}}{a_{41}\, a_{74}} dt^2,
1185: \nonumber\\
1186: &&
1187: \psi_4= \psi+ \frac{b\, a_{41}\, a_{75}\, a_{76}}
1188: {2a_{71}\, a_{64}\, a_{74}^{\ 2}}t
1189: \qquad\mbox{and}\qquad
1190: \phi_4=\phi+\frac{c\, a_{51}^{\ 2}}{2\, a_{41}\, a_{61}\, a_{71}}t,
1191: %&&g_{t\psi}=O(r^2), \qquad
1192: %g_{t\phi}=O(r^2)\qquad \mbox{and} \qquad
1193: %g_{\psi\phi}=O(r^4),
1194: \end{eqnarray}
1195: where we used eq.(\ref{regpsi}) and eq.(\ref{regphi}).
1196: We find that
1197: %this metric is the Minkowski metric by a coordinate transform,
1198: there is no singularity at the point $\rho=0$ and $z=a_4$.
1199:
1200:
1201: \section{summary and discussion}\label{summary}
1202:
1203: We present a method for constructing a seed metric.
1204: Using this method, we can obtain a seed metric
1205: corresponding to any rod structure that we want.
1206: In this paper, we constructed a solution which has two mutually
1207: orthogonal black rings.
1208: We call this solution the orthogonal black di-ring.
1209: Although the solution that we obtained by using the inverse scattering method
1210: generally has singularities,
1211: we have shown that we can remove these singularities
1212: by choosing the parameters appropriately.
1213: This solution has four free parameters
1214: which represent
1215: %distances between the center and
1216: radii of black rings,
1217: and speeds of $S^1$-rotation of black rings.
1218:
1219: % 図の挿入
1220: \begin{figure}[tbp]
1221: \begin{center}
1222: \includegraphics[keepaspectratio=true,height=25mm]{BR5.eps}
1223: \end{center}
1224: \caption{The expected rod structure of the seed metric of the many
1225: black rings solution.}%{}内にタイトルを記入してください
1226: \label{fig:BR5.eps}
1227: \end{figure}
1228:
1229: If we want to construct many black rings solution,
1230: probably we should prepare the seed metric corresponding to
1231: the rod structure as shown in fig.\ref{fig:BR5.eps},
1232: which can be easily obtained by our method.
1233: %In this solution,
1234: The soliton transformation applied to this seed metric
1235: leaves naked singularities
1236: with the general parameters.
1237: However, in generating the orthogonal black di-ring solution,
1238: the naked singularities can be removed
1239: when we choose the parameters by imposing the continuity of the periodicities at this point.
1240: This seems to tell us that
1241: these singularities are related to the discontinuity of the periodicities.
1242: The physical reason why the discontinuity of the periodicities exists in the seed metric is
1243: that a plane which helps the balance against its attractive self-gravity force is needed.
1244: If the rotation of the black ring balances against its attractive self-gravity force,
1245: such a plane is not needed.
1246: Since the rotation of the black ring is introduce through the BZ parameters,
1247: it seems that we can choose the parameters such that no discontinuity of the periodicities appears.
1248: Therefore, the naked singularities can be probably removed.
1249: %Performing the inverse scattering method to the seed metric made by this method,
1250: %we can probably obtain the solution having many black objects which
1251: %are the black rings on the orthogonal planes with or
1252: %without black hole at the center.
1253:
1254: Similarly, it seems that
1255: the solution, which has black rings on the two orthogonal planes and a black hole at the center,
1256: can be probably obtained.
1257: We will report such a solution in the forthcoming paper.
1258:
1259: \acknowledgements
1260: The author is grateful to Takahiro Tanaka for careful reading the manuscript and
1261: useful comments.
1262: He also thanks Takashi Nakamura for his valuable comments and continuous
1263: encouragement.
1264: This work is supported in part by the 21st Century COE ``Center for
1265: Diversity
1266: and Universality in Physics'' at Kyoto university, from the Ministry of
1267: Education,
1268: Culture, Sports, Science and Technology of Japan.\\
1269:
1270: \appendix
1271:
1272: \section{}\label{apg}
1273:
1274: In this section, we show that eq.(\ref{gg}) and eq.(\ref{F})
1275: satisfy the Einstein equations (\ref{diffg}-\ref{zf}) and $\det {\bf g}=-\rho^2$.
1276: Thanks to the diagonalization, we can divide eq.(\ref{diffg})
1277: into that for each component of ${\bf g}$.
1278: The differential equation of $g_{tt}$ becomes
1279: \begin{eqnarray}
1280: \partial_\rho \{\rho \partial_\rho (\ln g_{tt}) \}+
1281: \partial_z \{\rho \partial_z (\ln g_{tt}) \}=0. \label{gtt}
1282: \end{eqnarray}
1283: Suppose that $g_{tt}=g_{i}\cdots /(g_{j}\cdots)$.
1284: Since $g_{tt}$ appears only in the form of $\ln g_{tt}$,
1285: $g_{tt}$ becomes the solution of eq.(\ref{gtt})
1286: if each $g_j$ satisfies
1287: \begin{eqnarray}
1288: \partial_\rho \{\rho \partial_\rho (\ln g_i) \}+
1289: \partial_z \{\rho \partial_z (\ln g_i) \}=0.\label{mudiff}
1290: \end{eqnarray}
1291: In fact, if $g_i=\mu_i$ or $-\rho^2$,
1292: %\begin{eqnarray}
1293: %g_i=\mu_i=\pm\sqrt{\rho^2+(z-a_i)}-(z-a_i),\label{soli}
1294: %\end{eqnarray}
1295: %where $a_i$ is a arbitrary constant
1296: %and which is the same equation as eq.(\ref{soliton}),
1297: eq.(\ref{mudiff}) holds.
1298: Therefore, $g_{tt}$ which is a product of $\rho^2$ and $\mu_i$
1299: satisfies eq.(\ref{gtt}).
1300: We can construct $g_{\psi\psi}$ and $g_{\phi\phi}$ similarly.
1301: Moreover, $\det {\bf g}= -\rho^2$ is achieved
1302: provided that the total number of each $\mu_k$ in the numerator
1303: in all components is equal to that in the denominator
1304: and that $\rho^2$ appears once in the numerator among
1305: all components.
1306: This means that ${\bf g}$ given in (\ref{gg}) satisfies
1307: eq.(\ref{diffg}) and $\det {\bf g}=-\rho^2$
1308:
1309: We must also solve eq.(\ref{rhof}) and eq.(\ref{zf}).
1310: Thanks to the diagonal form of the metric ${\bf g}$, the traces
1311: $\mbox{Tr}(U^2-V^2)$ and $\mbox{Tr}(UV)$ in these equations
1312: become the summations of the contributions from each component of ${\bf g}$.
1313: In addition, since $f$ appears only in the form of $\ln f$ in the equations (\ref{rhof}) and (\ref{zf}),
1314: the left hand sides of these equations can be written as a summation of $\ln f_m$
1315: with $f=f_1f_2 \cdots f_n$.
1316: By this fact, we have only to solve for $f_m$ corresponding
1317: to each term of the right hand side of eq.(\ref{rhof}) and eq.(\ref{zf}).
1318: First, we consider a component without $\rho^2$,
1319: which is written as
1320: \begin{eqnarray}
1321: \frac{\mu_{i_1}\cdots \mu_{i_m}}{\mu_{i'_1}\cdots \mu_{i'_n}}.
1322: \end{eqnarray}
1323: Then $(1/4 \rho) \mbox{Tr}(U^2-V^2)$ and $(1/2 \rho)\mbox{Tr}(UV)$ becomes
1324: \begin{eqnarray}
1325: &&\frac{1}{4\rho}\mbox{Tr}(U^2-V^2) \supset
1326: \frac{\rho}{4}\sum_{k}\left(\frac{(\partial_\rho \mu_{i_k})^2-(\partial_z \mu_{i_k})^2}{ \mu_{i_k}^2} \right)
1327: +\frac{\rho}{4}\sum_{k}\left(\frac{(\partial_\rho \mu_{i'_k})^2-(\partial_z \mu_{i'_k})^2}{ \mu_{i'_k}^2} \right)
1328: \nonumber\\
1329: &&\qquad\qquad\qquad\qquad\qquad
1330: +\frac{\rho}{2}\sum_{k\neq l}\left(
1331: \frac{(\partial_\rho \mu_{i_k})(\partial_\rho \mu_{i_l})-(\partial_z \mu_{i_k})(\partial_z \mu_{i_l})}
1332: { \mu_{i_k} \mu_{i_l}} \right)
1333: +\frac{\rho}{2}\sum_{k \neq l}\left(
1334: \frac{(\partial_\rho \mu_{i'_k})(\partial_\rho \mu_{i'_l})-(\partial_z \mu_{i'_k})(\partial_z \mu_{i'_l})}
1335: { \mu_{i'_k} \mu_{i'_l}} \right)
1336: \nonumber\\
1337: &&\qquad\qquad\qquad\qquad\qquad\qquad\qquad\qquad\qquad\qquad\qquad\qquad\qquad
1338: -\frac{\rho}{2}\sum_{k,l}\left(
1339: \frac{(\partial_\rho \mu_{i_k})(\partial_\rho \mu_{i'_l})-(\partial_z \mu_{i_k})(\partial_z \mu_{i'_l})}
1340: { \mu_{i_k} \mu_{i'_l}} \right),
1341: \end{eqnarray}
1342: and
1343: \begin{eqnarray}
1344: &&\frac{1}{2\rho}\mbox{Tr}(UV)\supset
1345: \frac{\rho}{2}\sum_{k}\left(\frac{\partial_\rho \mu_{i_k}\partial_z \mu_{i_k}}{ \mu_{i_k}^2} \right)
1346: +\frac{\rho}{2}\sum_{k}\left(\frac{\partial_\rho \mu_{i'_k}\partial_z \mu_{i'_k}}{ \mu_{i'_k}^2} \right)
1347: \nonumber\\
1348: &&\qquad\qquad\qquad\qquad\qquad
1349: +\frac{\rho}{2}\sum_{k\neq l}\left(
1350: \frac{(\partial_\rho \mu_{i_k})(\partial_z \mu_{i_l})+(\partial_z \mu_{i_k})(\partial_\rho \mu_{i_l})}
1351: { \mu_{i_k} \mu_{i_l}} \right)
1352: +\frac{\rho}{2}\sum_{k\neq l}\left(
1353: \frac{(\partial_\rho \mu_{i'_k})(\partial_z \mu_{i'_l})+(\partial_z \mu_{i'_k})(\partial_\rho \mu_{i'_l})}
1354: { \mu_{i'_k} \mu_{i'_l}} \right)
1355: \nonumber\\
1356: &&\qquad\qquad\qquad\qquad\qquad\qquad\qquad\qquad\qquad\qquad\qquad\qquad\qquad
1357: -\frac{\rho}{2}\sum_{k,l}\left(
1358: \frac{(\partial_\rho \mu_{i_k})(\partial_z \mu_{i'_l})+(\partial_z \mu_{i_k})(\partial_\rho \mu_{i'_l})}
1359: { \mu_{i_k} \mu_{i'_l}} \right).
1360: \end{eqnarray}
1361: As we explained above, we only have to solve the equations
1362: \begin{eqnarray}
1363: &&\partial_\rho \ln f_m =
1364: \frac{1}{2\rho}\left(
1365: \frac{(\partial_\rho \mu_k)(\partial_\rho \mu_l)-(\partial_z \mu_k)(\partial_z \mu_l)}
1366: { \mu_k \mu_l} \right),
1367: \\
1368: &&\partial_z \ln f_m = \frac{1}{2\rho} \left(
1369: \frac{(\partial_\rho \mu_k)(\partial_z \mu_l)+(\partial_z \mu_k)(\partial_\rho \mu_l)}
1370: { \mu_k \mu_l} \right).
1371: \end{eqnarray}
1372: A solution of these equations is given by
1373: \begin{eqnarray}
1374: f_m = \frac{\mu_k\mu_l}{\rho^2+\mu_k\mu_l}.
1375: \end{eqnarray}
1376: Next, we consider the component with $\rho^2$.
1377: The difference from the case of the component without $\rho^2$ is that
1378: $(1/4 \rho) \mbox{Tr}(U^2-V^2)$ and $(1/2 \rho) \mbox{Tr}(UV)$, respectively, have extra terms
1379: \begin{eqnarray}
1380: \frac{1}{\rho}
1381: + \sum_{k}\frac{\partial_\rho \mu_{i_k}}{\mu_{i_k}}
1382: -\sum_{k}\frac{\partial_\rho \mu_{i'_k}}{\mu_{i'_k}},\label{mumu}
1383: \end{eqnarray}
1384: and
1385: \begin{eqnarray}
1386: \sum_{k}\frac{\partial_z \mu_{i_k}}{\mu_{i_k}}
1387: -\sum_{k}\frac{\partial_z \mu_{i'_k}}{\mu_{i'_k}}.\label{mumumu}
1388: \end{eqnarray}
1389:
1390: The first term of (\ref{mumu}) cancels $-1/\rho$ on the right hand side of eq.(\ref{rhof}).
1391: $f_m$ corresponding to the the second and the third of (\ref{mumu}) and (\ref{mumumu}) is $\mu_k$.
1392: As a result, remembering that $f$ can be written with the product of $f_m$,
1393: $f$ is obtained as eq.(\ref{F}).
1394:
1395: \begin{thebibliography}{}
1396:
1397: \bibitem{collider}
1398: T.~Banks and W.~Fischler,
1399: hep-th/9906038;
1400:
1401: S.~B.~Giddings and S.~D.~Thomas,
1402: Phys.\ Rev.\ D {\bf 65}, 056010 (2002);
1403:
1404: S.~Dimopoulos and G.~Landsberg,
1405: Phys.\ Rev.\ Lett\ {\bf 87}, 161602 (2001);
1406:
1407: P.~Kanti,
1408: Int.\ J.\ Mod.\ Phys.\ A {\bf 19}, 4899 (2004).
1409:
1410: \bibitem{Tan}
1411: F.~R.~Tangherlini,
1412: Nuovo\ Cimento {\bf 27}, 636 (1963).
1413:
1414: \bibitem{static}
1415: G.~W.~Gibbons D.~Ida and T.~Shiromizu,
1416: Prog.\ Theor.\ Phys.\ Soppl. {\bf 148}, 284 (2003);
1417: Phys.\ Rev.\ Lett\ {\bf 89}, 041101 (2002).
1418:
1419: \bibitem{Ishibashi}
1420: A.~Ishibashi and H.~Kodama,
1421: Prog.\ Theor.\ Phys. {\bf 110}, 901 (2003).
1422:
1423:
1424: \bibitem{Myers}
1425: R.~C.~Myers and M.~J.~Perry,
1426: Ann.\ Rev. {\bf 172}, 304 (1986).
1427:
1428: \bibitem{Iguchi2}
1429: H.~Iguchi and T.~Mishima,
1430: Phys.\ Rev.\ D {\bf 73}, 121501 (2006).
1431:
1432: \bibitem{Tomizawa3}
1433: S.~Tomizawa and M.~Nozawa,
1434: Phys.\ Rev.\ D {\bf 73}, 124034 (2006).
1435:
1436: \bibitem{Morisawa}
1437: Y.~Morisawa and D.~Ida,
1438: Phys.\ Rev.\ D {\bf 69}, 124005 (2004).
1439:
1440: \bibitem{Emparan}
1441: R.~Emparan and H.~S.~Reall,
1442: Phys.\ Rev.\ Lett. {\bf 88}, 101101 (2003).
1443:
1444: \bibitem{Mishima}
1445: T.~Mishima and H.~Iguchi,
1446: Phys.\ Rev.\ D {\bf 73}, 044030 (2006).
1447:
1448: \bibitem{Tomizawa2}
1449: S.~Tomizawa, Y.~Morisawa and Y.~Yasui,
1450: Phys.\ Rev.\ D {\bf 73}, 064009 (2006).
1451:
1452: \bibitem{Figueras}
1453: P.~Figueras,
1454: J.\ High\ Energy\ Phys. {\bf 07}, 039 (2005).
1455:
1456: \bibitem{Pomeransky}
1457: A.~A.~Pomeransky and R.~A.~Sen'kov,
1458: hep-th/0612005.
1459:
1460: \bibitem{Kudoh}
1461: H.~Kudoh,
1462: % gr-qc/0611136.
1463: Phys.\ Rev.\ D {\bf 75}, 064006 (2007).
1464:
1465: \bibitem{Yazadjiev}
1466: S.~Hollands, S.~S.~Yazadjiev,
1467: %Uniqueness theorem for 5-dimensional black holes with two axial Killing fields.. Jul 2007. 18pp.
1468: arXiv:0707.2775
1469:
1470:
1471: \bibitem{Tomizawa}
1472: Y.~Morisawa, S.~Tomizawa and Y.~Yasui
1473: arXiv:0710.4600
1474:
1475: \bibitem{Yazadjiev6}
1476: % A Uniqueness theorem for 5-dimensional Einstein-Maxwell black holes.
1477: S.~Hollands, S.~S.~Yazadjiev,
1478: arXiv:0711.1722 [gr-qc]
1479:
1480:
1481: \bibitem{diring}
1482: H.~Iguchi and T.~Mishima,
1483: Phys.\ Rev.\ D {\bf 75}, 064018 (2007).
1484:
1485: \bibitem{diring2}
1486: %The Black Di-Ring: An Inverse Scattering Construction.
1487: J.~Evslin, C.~Krishnan
1488: arXiv:0706.1231
1489:
1490:
1491: \bibitem{saturn}
1492: H.~ Evlvang and P.~ Figueras,
1493: JHEP {\bf 05}, 050 (2007).
1494:
1495:
1496: \bibitem{Yazadjiev2}
1497: %Completely integrable sector in 5-D Einstein-Maxwell
1498: %gravity and derivation of the dipole black ring solutions.
1499: S.~S. Yazadjiev,
1500: Phys.\ Rev.\ D {\bf 73}, 104007 (2006).
1501: %e-Print: hep-th/0602116
1502:
1503:
1504: \bibitem{Yazadjiev3}
1505: %Solution generating in 5D Einstein-Maxwell-dilaton
1506: %gravity and derivation of dipole black ring solutions.
1507: S.~S.~Yazadjiev,
1508: JHEP {\bf 0607}, 036 (2006).
1509: %e-Print: hep-th/0604140
1510:
1511: \bibitem{Yazadjiev5}
1512: %Rotating dyonic dipole black rings: Exact solutions and thermodynamics.
1513: S.~S.~Yazadjiev,
1514: Gen.\ Rel.\ Grav.\' {\bf 39}, 601 (2007).
1515: %e-Print: hep-th/0607101
1516:
1517: \bibitem{Yazadjiev4}
1518: %Black Saturn with dipole ring.
1519: S.~S.~Yazadjiev,
1520: Phys.\ Rev.\ D {\bf 76}, 064011 (2007).
1521: %e-Print: arXiv:0705.1840 [hep-th]
1522:
1523: \bibitem{ISM}
1524: V.~A.~Belinsky and V.~E.~Zakharov,
1525: Sov.\ Phys.\ JETP {\bf 48}, 985 (1978);
1526:
1527: V.~A.~Belinsky and V.~E.~Sakharov,
1528: Sov.\ Phys.\ JETP {\bf 50}, 1 (1979).
1529:
1530: \bibitem{Pomeransky2}
1531: A.~A.~Pomeransky,
1532: Phys.\ Rev.\ D {\bf 73} 044004 (2006)
1533:
1534: \bibitem{Emparan2}
1535: R.~Emparan and H.~S.~Reall,
1536: Phys.\ Rev.\ D {\bf 65}, 084025 (2002).
1537:
1538: \bibitem{Harmark}
1539: T.~Harmark,
1540: Phys.\ Rev.\ D {\bf 70}, 124002 (2004).
1541:
1542: \end{thebibliography}
1543:
1544:
1545:
1546: \end{document}
1547: