Let's consider a massive sphere (with an approximate constant density, not charged, not rotating, no anti-matter for example an Earth simplified model).

On the outside of this sphere, $g(r) = \alpha/{r^2}$, or I could say the space-time shrinks as $1/r^2$.

On the inside of this sphere, $g(r) = \beta.r$, or I could say the space-time expands as $r$.

And $g(0) = 0$ at the center, or I could say the space-time is plain flat.

Let's now consider a simplified non-rotating, non-charged, matter (and not anti-matter) black hole of Schwarzschild radius $r_s$, and let's imagine the geometry of this black hole is a sphere.

Is there any possible way to compute the variation of the gravitational field below this $r_s$ as $r \rightarrow g(r)$? ($g(r)$ being here a way to interpret the space-time expansion inside of a black hole, because as you progress inside a black hole the quantity of matter behind you increases and the quantity of matter forward you decreases).

I guess this is the solution of a differential equation where $g$, $d$ (density) and $r$ are related to themselves. But before attacking this level of math tools I am in search of an intuitive image of where the solution would be.

For symmetry reasons I imagine that at $r=0$, $g=0$ and there is no singularity. But I didn't find any proof of this or the contrary.

Where are the best publication on this computation of $g$ and models (simplifications) involved?

  • 1
    $\begingroup$ What do you mean by $g$? The acceleration of gravity? This is going to be tough to define inside the horizon, since it's not possible to stand still. Not to mention there definitely is a singularity at $r=0$. $\endgroup$
    – Javier
    Jun 18, 2017 at 16:39
  • $\begingroup$ [My Q & A are in the standard reading order.] I mean $g = || { \vec {g} } ||$. Yes. Nothing stands still within this neighbour of the universe, for example on earth, nonetheless, an acceleration of gravity (or movment) is possible to compute. I can't see the difficulty you see here. Could you explain it in simpler terms? $\endgroup$
    – dan
    Jun 18, 2017 at 22:03
  • 2
    $\begingroup$ To amplify on what Javier said, you might want to read up on the equivalence principle. The equivalence principle tells us that g has no objectively defined value unless we pick a frame of reference. To measure g on the earth, we pick a frame of reference relative to the earth's surface, and in GR this is considered a noninertial (accelerating) frame. (The definition is different than the one in Newtonian mechanics.) In a good inertial (free-falling) frame, g=0. Inside the horizon, there is no frame that is at rest relative to the black hole. $\endgroup$
    – user4552
    Jun 18, 2017 at 23:09

1 Answer 1


Given a metric one can try to understand what the Newtonian potential is like by expanding the metric about flat space $\eta_{\mu\nu}$. If you can write $g_{\mu\nu} = \eta_{\mu\nu} + h_{\mu\nu} + O(h^2)$ where $h_{\mu\nu}$ is small in some sense then the gravitational potential is given by $h_{00} = - 2 \Phi$ (as you can derive from the Newtonian limit of Einstein's equations).

For the Schwarzschild metric inside the horizon, as $0\leq r \leq 2GM$ and the metric functions are like $1-2GM/r$, it will be very hard to define a sensible perturbation about flat space.

What you can look at is how the gravitational forces act non-perturbatively. The way to do this is to look at the so called "tidal forces" where you find that you will get pulled apart in a processes called "spaghettification." https://en.wikipedia.org/wiki/Spaghettification


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.