Electricity from ambient heat

That would be perpetual motion machines, not perpetual motion itself.

The phrase "perpetual motion" is often used to refer to perpetual motion machines, but also used to refer to actual perpetual motion such as orbiting bodies and thermal vibration, eg Feynmann's "statement [that] would contain the most information in the fewest words":

All things are made of atoms-little particles that that move around in perpetual motion, attracting each other when they are a little distance apart, but repelling upon being squeezed into one another.​

 

Log in or Sign up to hide all adverts.

Re Maxwell's demon, iirc the problem is worse than described in this thread. The main sticking point is the _valve_. As 'free energy' is captured the valve has to work against a back pressure. Anyone care to guess the minimum energy the valve will consume?

Yep.

 

Log in or Sign up to hide all adverts.

Why? We're thinking about the case where the valve only opens when there is a random, transient reverse in the pressure differential, I think.

Is that what you're thinking, RJBeery?

 

Log in or Sign up to hide all adverts.

Yes this is right. The process of the demon repeatedly opening and closing the valve based upon the velocity of the incoming molecules very well could take energy greater than what is expected to be gained, I wouldn't know.

In the scenario I described, though, the higher-velocity molecules (actually almost all of the molecules) are captured "all at once" due to random chance. I don't think the same problem exists here, since the valve is only moving once for an arbitrarily large amount of gas molecules (which gives us as much available energy as we want for a given pressure).

This is only true if you used the work to keep the machine in motion. Even if we could extract work from ambient heat the heat level would be reduced, otherwise you are violating both the 2nd AND the 1st law of thermodynamics. Violating the 2nd is a bit of a misdemeanor, while violating the 1st is a felony.

 

By the way, this is right. I was thinking of Alfred Hitchcock Presents from 1960, and then later "Four Rooms" by Tarantino, but they are all apparently based upon Roald Dahl's original "Man from the South" short story.

If you have 20 minutes to kill, here's TOTU episode.

If you're a bit less patient, or if you want to contrast the subtlety of TOTU with the in-your-face Tarantino, watch these 2 minutes.

I love YouTube!

Please Register or Log in to view the hidden image!

 

For Maxwell's demon, the energy spent controlling the valve isn't the problem, it's the energy spent and entropy introduced in measuring the molecules.
I think I was a bit off the mark talking about energy spent earlier (since you could have a completely passive system like a simple pressure valve). It's actually measurement and entropy control that's the direct problem.

But the valve could also move for an arbitrarily small amount of gas that bounces off it, and that gas might lose the excess pressure in the act of opening the valve. I.e. there is no reason for the pressure increase that triggers the valve opening to be maintained long enough for gas transfer. And while the valve is open, pressure drops that transfer gas in the wrong direction are not necessarily going to close the valve, because the valve mechanism/sensor isn't in exactly the same place as the opening.

So no matter how large or small the setup, I think that there could (will?) always be random cases that work in the wrong direction.

 

This is a really interesting discussion. I'm not completely certain that the position I'm arguing is right, by the way (although I'd give long odds that it is, in no small part due to my bias for mainstream.)


But maybe we're at the stage now where we're beyond hand waving and need to do the hard quantitative analysis? I don't expect to be able to actually do it, but how would we approach it?

• Large scale dynamic fluid modeling (ie no molecules, just continuous fluid).
• An ideal gas with some stochastic variation in its pressure field.
  • How do we define the variation? Is there a Brownian motion model that can be applied?
• No gravity.
• Perfectly insulating, perfectly rigid, high mass container walls.
• Perfectly insulating, perfectly rigid, zero mass valve/gate/switch components
• Outcome measure of interest is system entropy
  • Long term trends
  • Short term fluctuation probabilities
  • Probabilities of measurements of short term fluctuations, so we can take advantage of them? (or is that becoming circular?)

The hard part is going to be the dynamic fluid modeling, and the stochastic pressure field model.
I'm also thinking vaguely of an incompressible liquid model, but I don't know if there is an incompressible liquid model that allows a variable pressure field.

Then there's the minor details of potential mechanisms:

• Active gate (easiest to model, I think?)
  • Pressure (and temp?) sensor/s on gate and around gate opening
  • Instantly opens or closes (at zero energy cost) to desired size on defined sensor conditions
• Passive valve
  • Flat poppet - A section of the wall that lifts perpendicularly to some defined limit. Movement defined by average pressure on each side.
  • Flap - as above but a hinged flap
  • Valves may be spring loaded, with constant or distance dependent closing force

 

Carefully chosen words for a concession.

Please Register or Log in to view the hidden image!

In the passive system the measurement comes from the essentially continuous pressure monitoring by the sensor. The entropy control is handled with nothing but a ludicrous amount of patience.

This comment makes me think I'm not explaining the setup clearly. The gas container is separated down the middle by a barrier almost completely, except for a small, spring-loaded valve that is in the open position. The pressure sensor located in one half triggers the valve when it reaches a certain threshold. In this setup, the valve can be thought of as being a small frictionless "plate trap" that runs parallel to the separating barrier. As such, pressure on either side should be inconsequential after it has shut.

(I suspect Pete understands the following point, but I'm writing for others' benefit) If there is any doubt that the sensor can be activated with the valve in the open position, consider what happens when we BEGIN with all gas in one half of the container and the valve initially closed, and then we open the valve...the gas presumably equalizes in both sides but not instantly. Now simply reverse this process, which is completely legitimate according to Physical laws (if unlikely) and you can see that the system should work.

 

OK... and what is the consequence?
How much does the pressure at the sensor tell you about the state of the gas in each container?

Could work.
But what other states could the system be in after the valve closes?
How many states are there with locally high pressure at the sensor?
Quantitative time?

 

Sitting here, not sure. I don't know much about PSI sensors. It seems that the more surface area of the to-be-enclosed half of the container it covered (or sampled), the more accurate it would be with regards to not giving a false positive. A false positive actually seems fairly likely with a small sensor since we could consider the area immediately surrounding the sensor to be much more likely to be "compressed by chance" than the entire half of the container. To me the exercise is really a proof of concept, though, so as long as the probability of extracting usable work from ambient heat is non-zero that's all that we really care about.

Oh god, I have no idea. Something like this would definitely be a learning project for me.

Actually, now that I think of it, this probably has a better chance of producing energy than many of the Green projects that Obama is, um, investing in. Shall we get serious, Pete??

Please Register or Log in to view the hidden image!

 

I have to say, Pete, my threads with you both tax and advance my understanding. Do you realize the last time we had a discussion (regarding whether truly incompressible fluids could flow in a sealed container) we basically deduced the right answer? (that being, only specific flowing motions are allowed, such as that in the form of spheres or portions thereof).

Anyway I think I can resolve the "false positives" problem, and that's by maximizing surface-area/volume of the pressured half of the container by making it "almost" flat. That way, presuming the PSI sensor can cover the entire surface area, then it may be possible to show that any positive signal whatsoever necessitates that there is a true higher pressure differential in that half of the container.

 

Right, but we're only looking at half and ideal picture right now, where we're considering the distribution of entropy (or something) between the chambers.
The other half is in extracting work from that difference... and I think we will then need to consider whether there's a non-zero probability that we'll spend work instead of extracting it.

Me too, and one I don't have time for. Like my tag says, I really should be studying!
Right now I'm between a Population Health lecture on smoking, and a physiology lecture on gas exchange in people with COPD. Tomorrow I'm off to an all day workshop on Men's and Women's Health, where I learn to do testicular, vaginal, rectal, and breast exams (all on real healthy people, who I can only assume get paid well for their sacrifice!)

No. As interesting as it is, as wonderful it would be if it were feasible and practical, and as shaky as my grounding in my thermodynamics grounding is, I think that it has less than a million to one chance of being even theoretically viable, and less of being practical.

Please Register or Log in to view the hidden image!

We're the best thing on SciForums

Please Register or Log in to view the hidden image!

Interesting. Or a long thin pipe, with a valve in the middle.
Could be worth modeling.

 

Love, look at the two of us...

http://www.youtube.com/watch?v=exhiNToY3eI

lol, sorry.

 

Ah, Neddy... you'll always be my [post=887999]first love[/post]
:cheers:

(5.5 years ago... good times!)

 

I wonder if you can extract work from them?

Again, I left off the sarcasm tags (but I put that big smiley face!). That voice in my head when I'm typing must not sound the same as the voice in other peoples' heads when they're reading my stuff. This would be a ridiculous practical pursuit (unless you could convince a Gov't bureaucrat to grant you some research money!).

Yes, I wonder if pressure sensors can be manufactured to cover the inside of a cylinder?

 

Depends on their major

Please Register or Log in to view the hidden image!

Sorry... I'm not very sensitive to social subtleties in text media, so I tend to play it straight unless it's as subtle as a baseball bat.

Imagine they can.
Ideal gas, ideal sensors, ideal instant gate, instant processing time and instant communication from sensors to gate.
Tube length L and radius r (r much less than L).
Initial pressure Po, initial temperature To.
Where do you place the gate, under what conditions does it close, and what is the probability of those conditions being met at any give time?

 

Zing!

I'd say this part depends on our goal. Since its a simple proof on concept then in order to wait as little time as possible we need to establish a minimum pressure differential that could be used to establish "successfully extracted work" from the system.

Another thing I was just wondering is, does the size of the valve act as a dampener to pressure change between the chambers? I'm almost certain it does so we'd want to make it as large as possible.

 

Sorry for not responding sooner. Rather bad computer problem. FYI "win.com not found" is a _bad_ thing.

I was thinking that in Maxwell's original analysis that the "free energy" produced by those random fluctuations was at most equal to the energy required by the valve to "rectify" the energy, but I'm going strictly by memory which may well be wrong.

 

That's what I remember as well. And as bad as that is, it's still not THE big kicker/killer of this idea.

What REALLY dooms it is the fact that even *if* you could extract a minute amount of energy from such a device, doing so would reduce the energy (motion) of the molecules inside. And even assuming you could continue to reduce said energy (which is still a BIG stretch)

Please Register or Log in to view the hidden image!

you would eventually reach a point - at zero K - where there was no energy left inside the system.

Just as there is no such thing as a free lunch, there's no free energy either. If you want to ride the train you *must* buy a ticket.

Please Register or Log in to view the hidden image!

 

Right, you can't create energy, you can at most recycle that which is there. With good insulation the heat could be recycled many times, but eventually the system would lose all of its heat unless you have perfect insulation.

Unless you want to be really cool and ride on top of the train, or go to the park in early summer and eat a crapload of delicious berries..

Please Register or Log in to view the hidden image!

Diffusion (molecular heat) can be used for transportation. It happens often at the cellular level. Random movement of particles is actually quite effective at getting metabolites from point A to point B as long as the cell isn't too large.