Thermodynamics, gravity, entropy

[kernelpenguin]

The laws of thermodynamics (the second one, if I'm not mistaken) basically say that if matter is in clumps, it will become equally distributed over time and not the other way around.

But if we look into space, we see that matter was (probably) equally distributed in the beginning and has now condensed into clumps. (Known as planets, even.)

The thermodynamic model was originally made to model the behaviour of gases. It does not take into account gravity at all. (From what I can tell.)

Therefore, thermodynamics does not apply on the bigger scale as matter tends to collect into clumps, not become equally distributed.

This was the point argued by Ian Stewart, Jack Cohen and Terry Pratchett in The Science of Discworld II. (They've been wrong in that book before -- wouldn't surprise me if they were wrong again.)

So which is the case?

Oh and I've heard that black holes have a very high entropy. How is that possible? Or was this added just so that the laws of thermodynamics would hold?

 

[Clausius2]

The thermodynamic model was originally made to model the behaviour of gases. It does not take into account gravity at all. (From what I can tell.)

Therefore, thermodynamics does not apply on the bigger scale as matter tends to collect into clumps, not become equally distributed.

Why not?. You only have to make \rho to take variations in the space. Surely the thermodynamic field inside a star suffers of convection and diffusion phenomena. That effects are mirrored by thermodynamics equations, and the principal cause of such mass movements are the spatial density differences.

 

[kernelpenguin]

The way my physics lecturer explained it (I'm majoring in compsci so it wasn't a very technical explanation) was that the probability of having an even distribution of matter is higher than having it all gather into a clump or clumps.

For example, if you have 4 indistinguishable particles of gas in a two-part container, then the probability of finding the particles in different parts is higher than finding them both in the left part or both in the right part.

From that, it's supposed to follow that the universe is more likely smoother than clumpier as the probability of many particles gathering into the same place is very very small.

However, the above gas model does not take into account attractive forces and as such can't be applied to the universe at large. Or so the story went in The Science of Discworld.

 

[kuenmao]

This seems to have two sides:
1. The matter of the universe becomes one piece and clumps together.
2. The matter of the universe is evenly distributed out, so that for any particle, attractive forces from every side cancel out each other's effects.
Both seem reasonable, but personally I think the first is more probable.