Saturday, 30 September 2017

Event horizons of approaching black holes

I think that the event horizons of approaching black holes should repel as they approach.

I was intrigued when I posted a comment on PBS space times' YouTube video and the response was "do the math".

It took me a while but now at least I have something to point at, and I think it was worth the effort given my lack of background in (astro)physics.

The source code is available in SourceForge for you to try out yourself.
I'm on Debian so you'll need make and g++/clang++, I tried to keep it as simple as possible.
The program generates an ".off" file which describes a point cloud.

The image above is what you get when you load the file into MeshLab and take a snapshot.
This was generated by:
./event-horizon  --d=1.6 --res=1e-4
 You can try other values by varying the parameters, the code isn't that complicated to understand.


Update 2017-10-03

I wrote a CGAL variant that generates full surfaces but as it turns out one sometimes gets its normals inverted - it looks kinda cool though!

Update 2017-10-03 part 2

After several hours mulling it over, I figured out how to write a libigl version with the NanoGUI.
With this version one can look around the model and directly modify viewing attributes as well as the distance (d) parameter.

These screenshots were taken from running
./event-horizon-igl -d 1.382012173 

The thing to note is the blob hiding in the middle of the combined event horizons - it's outside the event horizon but isolated from the rest of space-time.

It's also shrinking as the singularities approach one another.

I'm guessing its entropy is decreasing too, so time is going backwards.

It may or may not be being bombarded with leaking firewall from the two black holes.

Finally, it's becoming so small and so close to the event horizons that millions or billions of years will pass for it while for us it would be just a moment.

Sounds like a pocket universe to me, a time-reversed big bang!

Update 2017-10-13

It looks like the singularities meet inside that pocket of regular space-time!
I used the radii corrosponding to the 29 and 36 solar mass black holes whose gravity waves were detected.

If you run
./event-horizon --r1=.855484 --r2=1.06198 -d 5e-13 --res=1e-11
 here's a very zoomed-in view of what you get (the dot in the middle is the origin)

I wasn't expecting that!