Bound Systems in an Expanding Universe: How Can Galaxies Merge?

[Simon Peach]

As the universe expands all objects are moving away from us, so how is it posible that the Andromeda galaxy is on a path to merge with our galaxy?

 

[Vanadium 50]

How is it possible that you get closer to your grocery store when you go shopping?

 

[PeterDonis]

As the universe expands all objects are moving away from us

This is not correct. Only objects that are not part of the same local gravitationally bound system as our galaxy are moving away from us because of the expansion of the universe. The Andromeda galaxy is part of such a local gravitationally bound system.

 

[Ibix]

As the universe expands all objects are moving away from us, so how is it posible that the Andromeda galaxy is on a path to merge with our galaxy?

It's only in idealised model universes that literally everything is moving away from everything else. In the real world, there was a tiny bit of jitter in the early universe which meant that some regions had slightly over the average density of stuff (I would say "density of matter" but it wasn't matter yet), and those regions eventually collapsed to form planets, stars, galaxies, and galaxy clusters. Those objects pull on each other with their gravity, so they don't move in the perfectly regimented Hubble expansion way.

All distant galaxies are moving away from us, faster the further away they are. On average nearby galaxies are moving away from us, also faster the further away they are, but at (relatively) short distances the random velocity from gravitational interaction may be larger than that average "away" motion so individual galaxies may be moving towards us. An extreme example is our planet, held together by its own gravity, which is not expanding at all.

 

[PeterDonis]

On average nearby galaxies are moving away from us, also faster the further away they are

This is not correct as you state it. Galaxies that are in the same gravitationally bound system as ours are not moving away from us on average.

the random velocity from gravitational interaction

I don't think it's useful to view velocities due to gravitational interaction in gravitationally bound systems as random.

 

[Ibix]

This is not correct as you state it. Galaxies that are in the same gravitationally bound system as ours are not moving away from us on average.

Fair enough - I should have excluded the local group (which includes Andromeda). But once you get outside the local group I think it's correct.

I don't think it's useful to view velocities due to gravitational interaction in gravitationally bound systems as random.

Deterministically chaotic, then (unless I'm missing your point). I'm not sure the distinction is particularly important in a B level thread.

 

[PeterDonis]

once you get outside the local group I think it's correct.

Not quite, because the local group is itself part of a larger gravitationally bound system, the Virgo Supercluster:

https://en.wikipedia.org/wiki/Virgo_Supercluster

You would have to look at galaxies outside that system (and possibly even more than that, if the speculations about a larger supercluster, mentioned in the Wikipedia article, are correct) to be sure you were looking at relative motion due to the overall expansion of the universe.

 

[PeterDonis]

Deterministically chaotic, then (unless I'm missing your point)

I agree the motions are deterministically chaotic, but my point wasn't about chaos, but about the fact that the system is gravitationally bound. The gravitational binding means that there are correlations between the individual components of the system, so their relative motions are not random, or at least not completely random.

 

[Bandersnatch]

Not quite, because the local group is itself part of a larger gravitationally bound system, the Virgo Supercluster:

https://en.wikipedia.org/wiki/Virgo_Supercluster

You would have to look at galaxies outside that system

They definitely do move away from us, even in that system. All of them, in fact - other than the members of the Local Group. Hubble's original graph, with all its ailments, included objects just within this super cluster. Only four of his data points are blueshifted, and the value of the Hubble constant that pops up from averaging all twenty four (using modern data, of course) is not that far from what can be gleaned from large-scale or CMB observations.

That a system is gravitationally bound means only that its components won't eventually join the Hubble flow, however long it may take. It doesn't necessarily mean the recession is not observable. The way I understand this, it can take an enormous amount of time for distant components to bleed off the initial recessive behaviour inherited from the primordial gas, and start coming back on the return leg of the orbit. Even after 14 billion years all but the most tightly bound systems are still relaxing.

 

[PeterDonis]

Even after 14 billion years all but the most tightly bound systems are still relaxing.

Hm, yes, this is a fair point--if the orbital time scale for the bound system is longer than the time since the Big Bang, we can't expect the bound system to have reached a state in which the orbits are approximately stationary.