A question about reducing gas entropy

[Edgard Neuman]

hi,

that may be a stupid question, but I'm not sure about the answer..  is it really impossible to reduce the entropy of a gas ? 
Picture a cavity with a special shape (probably parabolas) where molecule would preferentially bounce in carefully selected directions.. it could also be some sort of tube, where when you put some gas at one end, molecules end up going all parallel at the other end..  I suppose what would be difficult is to make the speed of the particles uniform, but maybe if you use some material with a specific bouncing properties (i mean that molecules would bounce to different direction varying with their speed, and so you can filter them using only geometry)

is it really impossible ? 

[studiot]

35 minutes ago, Edgard Neuman said:

is it really impossible to reduce the entropy of a gas ? 

Of course not.

 

You don't need  fancy curved demon apparatus. That would not work anyway.

 

Simply cool the gas.

You can do this until it becomes liquid or solid, but again it is not necessary.

 

Change of entriopy is defined as

[math]dS = \frac{q}{T}[/math]

So if q is negative (cooling) then dS is negative.

This is the basic reason the Entropy function was introduced in the first place.

 

However, a caveat.

That refers (as requested) to the entropy of the gas.

There will be counterbalancing changes in the rest of the Universe.

[swansont]

It's possible, but you would have to raise the entropy somewhere else. 

edit: xpost with studiot

 

[Edgard Neuman]

As always, you seem to not understand my very complicated question, and give me very basic answers, that I understood a thousand times.
 I'm challenging the theory. I try to understand it fully, and I thought of a specific case that seems (at least to me) to contradict it. I may be wrong, but I want to understand HOW.  
Imagine a gas. In a perfect closed (or opened ?) chamber. All molecules of the gas are bouncing on the walls of the perfect chamber according to its shape. I'm a well aware of the laws of thermodynamics. That's not what I'm asking. My "vivid visual imagination" tells me (and some article I read about specific shapes that modify the density of things when they bounce in it), that, maybe for some specific shapes of the chamber, ordinary bouncing of particles could be altered in a certain way so that, from a state where all molecule are going in a random directions, they end up all going in the same direction and at the same speed. It's a box that SORT molecule and ORGANIZE them because of it's specific shape. It's a kind of maxwell's demon, doing so only with its shape. Because of the shape and the bounces. 
You know that parabolas for instance, have specific properties.. I know you can create some chamber with very specific shapes where some place are only accessible after a certain serie of specific bounces (at least in 2D), or with forbidden area where a bouncing never goes. I'm thinking of very specific asymmetric shaped billiard that affect trajectories is a particular way.
Are you aware of that ? (the problem is I read a lot of articles and it's very hard to find this one back)
I understand, that in that very particular case, if in the end, (after the bounces in the very specific shapes) molecules effectively all go in the same direction, at the same speed, one could then argue that the entropy of the gas diminished. (because in their own shared frame of movement, they are not moving relative to each others, therefore their temperature is zero (may I remind you that "temperature" is not a mathematical object that exists by its own or is carried by space, but a statistical property of the molecules  of a gas) ).

There's two possibility :
- such a shape does not exist (I don't see why but maybe for mathematical reasons). That would solve the paradox. In that case WHY NOT ? 
- in that thought experiment, if the shape exist, did something gained entropy that I missed ? (we have to suppose the walls are perfect)..So in that case WHAT DID ? 

I think I have the answer : the box necessarily gained some momentum in the opposite direction (the momentum of molecule that were redirected by the box), so for the law of thermodynamic to work, and the whole thing should have globally the same entropy : the box and the gas moving in opposite directions have globally the same entropy, because of their relative movement. That give me some insight (believe it or not )... so the entropy of a system of objects is proportional to the mass of the component of the system and to the disparity of speed around the barycenter.. (because speed is relative to the frame, but the disparity of speed is the invariant)... it's like the "n object gas" and a zero entropy box turned into "2 object" gas with high entropy (because the high mass).. 
Don't bother answering, I have my answer.. 

Edited December 7, 2019 by Edgard Neuman

[Enthalpy]

Hi EN,

What you describe with oriented molecule speed is called a nozzle. If the pressure ratio is big enough, a nozzle converts well over 90% of the enthalpy into work, here kinetic energy of the ejected material, let's say gas.

The entropy of the nozzle is usually taken as zero because it's a mechanical part. Some people consider the heat stored in the nozzle's material, but I see no use for the entropy, and anyway the heat amount in the nozzle is normally negligible as compared with the gas.

The entropy of the ejected gas diminishes a lot, sure, as does its temperature. The oriented speed does not count in the entropy because it's ordered.

==========

Whether a closed box can do the same? General notions of thermodynamics tell "no". But I doubt this can be proven generally, whatever the attempted shape.

Usually, one has to propose a specific shape, and then it can be proven to keep the gas' entropy. But this must be done for each individual shape.

And, yes, the second principle can be proven, if leaving the 19th century thermodynamics and entering statistical physics. You can find it in Landau and Lifschitz for instance. But that won't prove mechanically that no box shape achieves ordering, only that the attempts are vain.

==========

You mentioned molecules "bouncing" at the walls, more or less like elastic balls. A more common description is that every impinging molecule sticks at the wall for some time, then is ejected by the wall's temperature. Though, such descriptions are imperfect too.