I've been trying to calculate $d\phi/dt$ of photons orbiting a Kerr black hole (Kerr metric in Boyer-Lindquist coordinates) on the equatorial plane, both counter and along with its rotation. So I used that $u_\mu u^\mu=0$, which applies for the four-velocities of photons and, because I wanted a circular orbit, I said $r = \mathrm{const}$. The same goes for $\theta$ (we have one specific plane, $\theta=\pi/2$)

My problem is that I reach a result expressed in $$g_{ab}$$ which I can't seem to bring to a simpler form (for example with $\Delta$):

$$ \frac{d\phi }{dt}_\mathrm{with} = \frac{\sqrt{g_{t\phi }^{2}-g_{tt}g_{\phi \phi }} - g_{t\phi }}{g_{\phi \phi }} $$ $$ \frac{d\phi }{dt}_\mathrm{counter} = -\frac{\sqrt{g_{t\phi }^{2}-g_{tt}g_{\phi \phi }} - g_{t\phi }}{g_{\phi \phi }}$$

Is my approach completely wrong or am I missing something? Should I use l and e instead?

  • $\begingroup$ Have you actually tried substituting the BL metric elements into those equations? You seem to have stopped short! This is probably something you want to do in a computer algebra package, like wxMaxima wxmaxima-developers.github.io/wxmaxima/index.html $\endgroup$
    – m4r35n357
    Jun 7, 2021 at 9:04
  • $\begingroup$ yes, i've been trying i use mathematica but it just can't bring it to a simpler form $\endgroup$ Jun 7, 2021 at 9:16
  • $\begingroup$ OK, just checking, been there done that - that is pretty much what it is! If you keep delta unexpanded you might miss simplifications, but if you expand delta, you end up with a mess. The CAS will help you avoid errors, but rarely gives you an answer in the form you want without a lot of meddling. Good luck! $\endgroup$
    – m4r35n357
    Jun 7, 2021 at 9:38
  • $\begingroup$ thank you so much! $\endgroup$ Jun 7, 2021 at 11:03

1 Answer 1


You are right and you only need to develop the calculations with BL values and obtain the simplifications.

You have $$ g_{\phi\phi}(\frac{d\phi}{dt})^2+2g_{t\phi}\frac{d\phi}{dt}+g_{tt}=0\ \ \ \ \ \ [1] $$

With $ m=\frac{GM}{c^2} $, and:

$ g_{\phi\phi}=r^2+a^2+\frac{2ma^2}{r} $

$ g_{t\phi}=-\frac{2ma}{r}c $

$ g_{tt}=-\left (1-\frac{2m}{r}\right )c^2 $

By multiplying $ [1] $ by $ r^2 $, you have: $$ (r^4+a^2r^2+2ma^2r)(\frac{d\phi}{dt})^2-4marc\frac{d\phi}{dt}-(r^2-2mr)c^2 = 0\ \ \ \ \ \ [2] $$ After some calculations and simplifications, the discriminant of $ [2] $ is written as:

$ 4c^2r^2(r^2-2mr+a^2) $ or $ 4c^2r^2\Delta $ with $ \Delta=r^2-2mr+a^2 $.

Then the two solutions for $ \frac{d\phi}{dt} $ are:

$$ \frac{d\phi}{dt}_{with}=c\frac{2ma+r\sqrt{\Delta}}{r^3+a^2r+2ma^2} $$ and $$ \frac{d\phi}{dt}_{counter}=c\frac{2ma-r\sqrt{\Delta}}{r^3+a^2r+2ma^2} $$

Hoping to have answered the question,

Best regards.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.