Can rotational speed increase gas density in a cylinder?

[Gh778]

Hi,

I would like to know if it's possible to turn a cylinder full of gas (air for example) at W rd/s for have more density up ? The first drawing show the basic system turn around the center. The second drawing show the density of gas like I think it can be, black=more density, white=less density. Is it correct ? The air is at 20°C, I'm interesting about the effect of temperature with centripetal forces.

 

[mfb]

To get a significant effect, the cylinder needs a velocity comparable to the speed of the molecules inside, this is ~500m/s for air. If the height of the cylinder is small compared to the radius, you need even more.
If you can reach this, or if very small effects are fine, it might be interesting to look at actual numbers.

 

[Gh778]

Ok, the speed must be high. In this case (theory study), is it possible to recover energy from temperature if the cylinder move up a little (green parts turn only) ? I think we recover energy from rotational speed in the same time but it's the energy we have gave.

 

[mfb]

The rotation needs a lot of energy. And I do not see why/how the pressure gradient should help in any way.

 

[Gh778]

I'm not sure, but I see the centripetal forces like a separator, the cylinder (black/grey color) has one surface with gas not the other (or less gas) and the force apply from temperature is more at one surface than other. It's not the pressure gradient that is interesting but the fact that the gas is more at one surface than other. Maybe the efficiency is not good due to the high speed but theorically is it possible to recover energy ?

 

[mfb]

Recover which energy?
Thermal energy? No.
Energy from the rotation of the whole setup? Maybe, if you have some connection to the environment, but I would do it in a different way.

Your pressure gradients just move mass away from the axis of rotation - you can get the same effect if you shift your whole cylinder a bit.

 

[Gh778]

Why is not possible to recover thermal energy, could you explain ?

 

[mfb]

Without an external cold reservoir, this would reduce the total entropy and violate one of the fundamental laws of physics. While it can be interesting to study thermodynamic devices to see why they do not work, they cannot reduce entropy. This can be proven from deeper theories in statistical physics. If your device could reduce entropy, there would be some fundamentally new thing at the microscopic level, doing this. While it is not possible to completely exclude that possibility: In the framework of ordinary thermodynamics, your device will not reduce entropy.

You could calculate all the thermodynamic processes here, and get the same result, but as you can see there is a nice shortcut. It does not work.
In addition, I do not even see how you get the idea that any thermal energy would be used in any way.

 

[russ_watters]

Yes, you can centrifuge a gas. Yes, you can recover heat in the process. No, it cannot be recovered continuously, nor can you recover more energy than you put in.

 

[boneh3ad]

You should look up the gas centrifuge, particularly of the Zippe-type.

 

[Gh778]

No, it cannot be recovered continuously

I think with a screw and the help from gravity this system can give a torque (for 180° after reverse the direction of screw around its axis). Gravity give more atoms at bottom to the gas, so more chocs from heating. I don't say it's the difference of pressure that give torque but the number of chocs is different due to the difference of pressure. Imagine a screw with 50 m of height, the difference of pressure due to gravity is about 25 Pa. When the screw turn around high circle of 180°, it turn around its axis of 180°.

 

[Gh778]

Only with gravity. I draw pressure curves from gravity with attraction in 1/d². With a screw, green points can't compensate themselves and magenta points can't compensate themselves too. I don't know how all these small torques can compensate themselves.

 

[mfb]

With a screw, green points can't compensate themselves and magenta points can't compensate themselves too.

There is no reason to expect that.

I don't know how all these small torques can compensate themselves.

Integrate, or use energy momentum conservation.

 

[Gh778]

There is no reason to expect that.

Red lines on my last drawing are lines where pressure are equal. Screw exist only at right on the drawing. I exaggerate the vector gravity 1/d².