4

I have seen a question here on PSE asking the same thing, but it has no answers. To reformulate the question slightly differently, suppose you have a general spherically symmetric metric in $(1+D)$ dimensions, $$g_{\mu\nu}dx^\mu dx^\nu = -A(r,t)dt^2 + B(r,t)dr^2 + r^2 d\Omega_{D-1}^2$$

The standard proof of Birkhoff's Theorem for $D=4$, follows by calculating the Ricci tensor and setting its components to zero. For higher dimensions, I imagine the Ricci tensor will become increasingly convoluted in higher dimensions and will not be as easy to solve to the resulting differential equations.

Question: Can we prove/disprove the validity of Brikhoff's Theorem, namely spherically symmetric vacuum solutions to the Einstein Equations are static, without any other assumptions?

I would imagine disproving it would not be as difficult, if we can find an example, but I could not find one nevertheless. Any help would be appreciated.

Qmechanic
  • 201,751
Lelouch
  • 3,616
  • 2
    You can calculate this on your own, surely? Work out the Ricci tensor and solve Einstein's equations! – Prahar Jul 08 '21 at 08:52
  • That is doable in (1+3) D. I did try for general dimensions and the calculation is too exhausting, albeit proceeding towards a seemingly definite resolution (which I don't know yet). I wanted to ask if there is a (simple ?) counter-example to show that Birkhoff's Theorem fails in the given form for (say) 5-D gravity. – Lelouch Jul 08 '21 at 10:43
  • The calculation is doable in any dimension. In any case, something being "doable" or not should stop you from actually doing it. It's the best way to learn. Surely if someone knew the answer, it means they actually did the calculation. If they can, why can't you? It's not like you don't know what needs to be done - what's stopping you is the amount of effort that you would need to put in. – Prahar Jul 08 '21 at 11:02
  • 1
    I ended up doing the calculation for you, only because I thought it was an interesting thing to work out. However, these are things you should be doing yourself! There is really no excuse not to. – Prahar Jul 08 '21 at 13:08

1 Answers1

3

We assume that a spherically symmetric metric can be written in the form \begin{equation} \begin{split} ds^2 = - e^{2A(t,r)} dt^2 + e^{2B(t,r)} dr^2 + r^2 d\Omega_{D-2}^2 , \qquad d\Omega_{D-2}^2 = \gamma_{ab} d\theta^a d\theta^b \end{split} \end{equation} where $\gamma_{AB}$ is the metric of the unit $S^{D-2}$. This satisfies $$ R_{\mu\nu} = 0 . $$ Our goal is to show that the unique solution to this is the Schwarzschild black hole.

The Christoffel symbols are \begin{equation} \begin{split} \Gamma^t_{tt} &= \partial_t A , \qquad \Gamma^t_{tr} = \partial_r A , \qquad \Gamma^t_{ta} = 0 , \qquad \Gamma^t_{rr} = e^{-2(A-B)} \partial_t B , \qquad \Gamma^t_{ra} = \Gamma^t_{ab} = 0 , \\ \Gamma^r_{tt} &= e^{2(A-B)} \partial_r A , \qquad \Gamma^r_{tr} = \partial_t B , \qquad \Gamma^r_{ta} = \Gamma^r_{ra} = 0 , \qquad \Gamma^r_{rr} = \partial_r B , \qquad \Gamma^r_{ab} = - r e^{-2B} \gamma_{ab} , \\ \Gamma^a_{tt} &= \Gamma^a_{tr} = \Gamma^a_{tb} = \Gamma^a_{rr} = 0 , \qquad \Gamma^a_{rb} = \frac{1}{r} \delta^a_b , \qquad \Gamma^a_{bc} = \Gamma^a_{bc}[\gamma] , \\ \Gamma^\rho_{t\rho} &= \partial_t ( A + B ) , \qquad \Gamma^\rho_{r\rho} = \partial_r ( A + B ) + \frac{D-2}{r} , \qquad \Gamma^\rho_{a\rho} = \Gamma^b_{ab}[\gamma] . \end{split} \end{equation} The Ricci tensor components are \begin{equation} \begin{split} R_{tt} &= - \partial_t^2 B + \partial_t B \partial_t ( A - B ) + e^{2(A-B)} \left[ \partial_r A \partial_r ( A - B ) + \partial_r^2 A + \frac{D-2}{r} \partial_r A \right] , \\ R_{tr} &= \frac{D-2}{r} \partial_t B , \\ R_{rr} &= - \partial^2_r A - ( \partial_r A )^2 + \left[ \partial_r A + \frac{D-2}{r} \right] \partial_r B - e^{-2(A-B) } \left[ \partial_t ( A - B ) \partial_t B - \partial_t^2 B \right] , \\ R_{ab} &=e^{-2B} [ - r \partial_r ( A - B ) + ( D - 3 ) ( e^{2B} - 1 ) ] \gamma_{ab} . \end{split} \end{equation} where in the last line, we used that on $S^{D-2}$, we have $R_{ab}[\gamma] = (D-3) \gamma_{ab}$. We now solve Einstein's equations in a vacuum $R_{\mu\nu} = 0$. The second equation implies that $\partial_t B = 0 \implies B = B(r) $. Then, $\partial_t R_{ab} = 0$ implies $\partial_t \partial_r A = 0 \implies A = A(r) + C(t)$. We can set $C(t)=0$ by redefining the coordinate $t$. Next, we look at \begin{equation} \begin{split} R_{rr} + e^{-2(A-B)} R_{tt} = \frac{D-2}{r} \partial_r ( A + B ) \implies A(r) = - B(r) - c . \end{split} \end{equation} We can set $c=0$ by redefining $t$. The remaining equations can all be simplified to \begin{equation} \begin{split} 2 r \partial_r B + ( D - 3 ) ( e^{2B} - 1 ) = 0 \implies e^{-2B} = 1 - \frac{c}{r^{D-3}} . \end{split} \end{equation} Putting all this together, we find that the metric takes the form \begin{equation} \begin{split} ds^2 = - f(r) dt^2 + \frac{dr^2}{f(r)} + r^2 d\Omega_{D-2}^2, \qquad f(r) = 1 - \frac{c}{r^{D-3}} . \end{split} \end{equation} This is the Schwarzschild solution in $D$ dimensions.

Prahar
  • 25,924
  • Concerning your first assumption about the form of the metric: a fairly rigorous argument about when such coordinates exist can be found in Box 23.3 of Misner, Thorne, and Wheeler. On a cursory examination, I don't see any impediments to extending it to higher dimensions (under the same implicit assumptions), so I think this is a valid starting point for the derivation. – Michael Seifert Jul 08 '21 at 13:40
  • @MichaelSeifert - I would tend to agree. However, I wasn't sure if there were some issues with topology or cobordisms (which do arise in other similar contexts) so I made sure to clearly present my starting assumption. – Prahar Jul 08 '21 at 13:51
  • The proof does rely on the statement that there are no non-vanishing rotationally symmetric vector fields on a 2-sphere. I think this is still true in higher dimensions, but vector fields do behave differently on higher-D spheres and I wouldn't be shocked if this step doesn't hold. – Michael Seifert Jul 08 '21 at 14:00