What do you think of this simulation?

http://www.abc.net.au/news/2014-09-19/simulation-show-andromeda-and-milky-way-colliding/5757048

ps
we only have roughly 5 billion years
get your affairs in order peeps…

according the managers there’s no problem, everything will be fine, just keep doing what you are doing whilst they clear their desks out.

Aquila said:

ps
we only have roughly 5 billion years
get your affairs in order peeps…

which is fine because in a bit over 1.0 billion years the Earth will be likely frazzled by the sun

Aquila said:

What do you think of this simulation?

http://www.abc.net.au/news/2014-09-19/simulation-show-andromeda-and-milky-way-colliding/5757048

I ought to be qualified to comment on this because I know quite a bit about fluid dynamics, and know people who do similar sorts of simulation – some doing discrete element hydrodynamics (at CSIRO) and some doing massive multiparticle simulations of globular clusters (at Swinburne). I hadn’t known there were people doing this at University of Western Australia.

As for the hydrodynamics, I can’t fault it. I particularly like how the galaxy rotation causes pseudo-waves to appear spiral-like.

As for the initial conditions, I haven’t kept track of what the relative lateral motion of the two galaxies is, but the shown miss distance on first pass is plausible and so is that maximum distance after first pass. According to Wikipedia the time to collision is 4 billion years rather than 5 billion.

As for overall timescale, the Milky Way rotates about once in every 225 million years, ie. the Sun orbits the galaxy at that frequency. But on this scale of simulation the shown density waves wouldn’t be expected to match the rotation period. The galaxy rotation rate for the density waves shown is about 4 rotations before the collision giving a rotation period of about once in 1,250 million years, which seems far too slow to me.

I find it very worrying that the second pass is a head-on collision. This actually makes good sense from the lack of observations of merging distorted galaxies in the relatively close universe, eg. in the Virgo cluster. Also worrying is that the galaxy state after head-on collision is more compact than either of the initial unmerged galaxies.I would have expected less compact, but then I could easily be wrong about this one, the larger elliptical core of the merged galaxy compared to the elliptical core sizes of the parent galaxies is qualitatively consistent with the Tully-Fisher relation, but I don’[t know if its quantitatively consistent.

There’s an earlier simulation of the collision by NASA here . The similarities between the simulations outweigh the differences. In both simulations the second pass is a head-on collision and the result is more compact and more elliptical-like than either of the parent galaxies.The NASA version has a Milky Way rotation period of about 160 million years, much closer to the correct rotation rate.

> The galaxy rotation rate for the density waves shown is about 4 rotations before the collision giving a rotation period of about once in 1,250 million years, which seems far too slow to me.

Or perhaps they’re not starting the simulation at the present time, but at a time much closer to the collision.

The simulation has both the galaxies on the same horizontal plane. Is this actually the case or is it a simplification? I am sure it would affect the outcome.