Singularities in colliding black holes

Jolb

When two black holes collide, do their singularities merge the instant their event horizons touch? Or is there a lag between the event horizons touching and the singularities merging when there are actually two singularities inside a single black hole?

Dmitry67

Yes, there is a lag. Note that during the collision their horizons become distorted.
So yes, it takes additional time for them to collide but I doubt anybody can provide more information without numerical simulations.

bcrowell

My intuition says pretty much the same thing as Dmitry67's. There has to be a lag.

However, I think the answer may be even more complicated. General relativity has a tendency to develop singularities spontaneously any time a sufficiently violent process occurs. For example, there's a broad research program being carried out by many different groups around the world to try to define cosmic censorship rigorously, and determine under what conditions it holds. A collision of two black holes is exactly the kind of violent, highly nonlinear process in GR that they have to worry about. So for instance, I could imagine that when two black holes collide you get more singularities than you started with, maybe some of them naked singularities -- I don't know. Someone who's an expert on this kind of thing could probably say more. I'm sure people have tried hard to model black hole mergers, since they're one of the best candidates for producing detectable gravitational waves. I would guess that collisions with a high degree of symmetry are the easiest to model, whereas highly asymmetric ones are probably much more difficult to say anything about.

Nabeshin

This is an interesting question, actually... I think next time I get the chance I'll ask some of the folks at work (who numerically solve the E.Eqns and do these kind of black hole mergers) what happens to the actual singularities during a merger.

bcrowell

That would be great. If they can point us to a review article on arxiv or something, that would be very cool.

atyy

I haven't studied this, but here is a reference which maybe someone can interpret:

Fig 60, p322 of http://books.google.com/books?id=Qag...gbs_navlinks_s