What happens to entropy in a cyclical universe?

[sci-guy]

I have only a basic understanding of science, but don't the Second Law of Thermodynamics and the concept of entropy preclude a cyclical universe (i.e. one that will eventually contract to a singularity)?

 

[sci-guy]

For those interested, I just found this, which explores the topic:

http://en.wikipedia.org/wiki/Heat_death_of_the_universe

There doesn't seem to be clear consensus, so I'm still interested in discussion.

 

[Dmitry67]

This is why scientists don't believe in that model.
Also, based on the observations, our Universe will never ever contract again. So if it was cyclical, then current cycle is the very last one. (But I suggested a version of cyclical Universe where there is no problem with entropy: https://www.physicsforums.com/showthread.php?t=476735 )

 

[bcrowell]

There is no well established answer to this question, mainly because there is no well established way of defining the entropy of a gravitating system in general relativity. Roger Penrose has a model (CCC) whose main selling point is that he thinks it makes sense out of the entropy issues -- more than non-cyclical models. The idea is that in a standard cosmology, exceptional fine-tuning is required to explain why we didn't have a maximum-entropy big bang (in which case we would have a universe that experienced heat death right when it was born).

 

[skydivephil]

There is no well established answer to this question, mainly because there is no well established way of defining the entropy of a gravitating system in general relativity. Roger Penrose has a model (CCC) whose main selling point is that he thinks it makes sense out of the entropy issues -- more than non-cyclical models. The idea is that in a standard cosmology, exceptional fine-tuning is required to explain why we didn't have a maximum-entropy big bang (in which case we would have a universe that experienced heat death right when it was born).

Can you elaborate on what you mean by "no well established way of defining the entropy of a gravitating system in general relativity."

 

[Chronos]

Depends on how you define 'entropy'. It is ... complicated. Is gravity a form of 'negative' energy? Many physicists think so. Shannon entropy provides an interesting way of viewing the big picture.

 

[bcrowell]

Can you elaborate on what you mean by "no well established way of defining the entropy of a gravitating system in general relativity."

Here's a paper that discusses some of the issues: http://arxiv.org/abs/hep-th/0410270

 

[bapowell]

This is why scientists don't believe in that model.
Also, based on the observations, our Universe will never ever contract again. So if it was cyclical, then current cycle is the very last one.

This is really not true. Observations don't tell us this at all. They tell us that the current epoch is undergoing an accelerated rate of expansion. Sure, if you extrapolate this behavior into the future, then there will never be a contracting phase again. But that's a big if, given that we don't currently have a good understanding of what's causing the accelerated expansion.

 

[bcrowell]

Also, based on the observations, our Universe will never ever contract again. So if it was cyclical, then current cycle is the very last one.

This is not completely established. Penrose's CCC is cyclical and is consistent with the evidence that our universe will never contract again.

 

[Chalnoth]

I have only a basic understanding of science, but don't the Second Law of Thermodynamics and the concept of entropy preclude a cyclical universe (i.e. one that will eventually contract to a singularity)?

Well, it certainly disproves any overly-naive cyclical universe model. But there are a number of ways around it. For instance, if you make it so that the post-bounce universe is different in such a way that entropy increases the whole time, then there is no problem (e.g. if the volume of each subsequent bounced universe is vastly larger).

 

[Chalnoth]

This is not completely established. Penrose's CCC is cyclical and is consistent with the evidence that our universe will never contract again.

Yeah, but it's also not a full model (there is no explanation whatsoever for the remapping of the large universe onto a small early one...it just happens).

 

[bapowell]

Well, it certainly disproves any overly-naive cyclical universe model. But there are a number of ways around it. For instance, if you make it so that the post-bounce universe is different in such a way that entropy increases the whole time, then there is no problem (e.g. if the volume of each subsequent bounced universe is vastly larger).

But it's important to point out that this kind of universe is not truly cyclic, since it is not past-eternal.

 

[Chronos]

I think a cyclical universe should leave an imprint in the CMB. Cosmologists have been looking for this for years ... and still looking. Anisotropies in the CMB are apparent, but, no reasonable explanation has yet been offered, IMO.

 

[Chalnoth]

I think a cyclical universe should leave an imprint in the CMB. Cosmologists have been looking for this for years ... and still looking. Anisotropies in the CMB are apparent, but, no reasonable explanation has yet been offered, IMO.

Perhaps, but there are no guarantees this is the case. Unfortunately, the properties of inflation are such that is is all too easy to completely and utterly wipe away any trace of what happened before.

 

[skydivephil]

and what about the BAum Frampton Model?
is that still on the table?

 

[yenchin]

The question of what happens to entropy when the universe evolves towards a crunch seems largely depends on who you ask. At least for now...

1. There is a model called Gold Universe in which the arrow of time turns the other way around and entropy starts decreasing as the universe reaches maximum size and evolve towards a crunch. [Thomas Gold, The Arrow of Time, American Journal of Physics 30, pp. 403-410, doi:10.1119/1.1942052, 1962.]

2. As pointed out by a few posters already, the fact that the expansion is currently accelerating doesn't tell us much about the *far future*. It is possible to cook up a model in which the universe has *negative* cosmological constant (so that it eventually re-collapses) yet has accelerating phase [Brett McInnes, Quintessential Maldacena-Maoz Cosmologies, JHEP 0404 (2004) 036, hep-th/0403104]

3. In Penrose's Weyl Curvature Hypothesis, he basically assumes that near the Big Bang, the Weyl Curvature vanishes, but this constraint is not imposed in the far future, so that even in the case of a crunch, the geometry will be "wild" and not as smooth as that at the bang. That means the crunch has high entropy unlike the bang. There have been a few ways to make sense of this. See for example, the very readable http://arxiv.org/abs/0711.1656v2" in which the author argued that entropy should continue to increase even if the universe re-contracts.