The Carnot efficiency problem

[Fool]

Matt I'd like to hear more about the double PV diagram for a gamma engine. Please. Always, many ways to model a system.

"Yes. All of them." You got my meaning of the first sentence even though it should have been written:

"I'd like to know, for the engines discussed here, how many of them could have work-expansion without a load?"

Sheesh... I even used a period where a question mark goes. Total drunk talk... LOL

To clarify and expand for those who are interested:

Yes. My point is all of them. Thanks. It's the old elastic vs inelastic collision theory. The only thing molecules care about in a collision is velocity and mass. (Impact angle, shape, rotational inertia, spin, chocolate pudding.. etc.). What is on the other side of the molecules doesn't matter, unless touching, so the mass adds. Molecules don't touch. Close but initial response is to move independently, for solids or gases.

It, borrowing a Tom analogy, is like a baseball and bat. The ball and bat react to the collision solely on the velocity between the two (speed direction angle offset spin). That isn't effected by what acceleration happened before or after the collision. Nor what the pitcher or batter ate that morning.

In other words, the gas's adiabatic cooling depends on the speed and direction of the piston (turbine, Tesla rotor, etc.). A bunt happens when there is a specific velocity differential. A difference is that the gas molecules are moving significantly faster than the piston. To bunt, the bat has to move with the ball to slow it down. To cool a gas the piston needs to move away from the gas molecules. Gas molecules are moving, on average, around the speed of sound.

Load controls the piston speed. Or over speed. For our engines here, rpms go from 500 to 50. Is that a big speed difference when compared to molecular speed of a gas? Probably not, only a very small percentage. Probably very close to the same amount of work for either engine speed.

Now for a cryocooler running at 1000 2000 psi. Serious rpms can be obtained, or over obtained. A load is important there just to keep the speed within a safe and effective range. The gas molecules don't see the load, just the wall speed. Work output is not what does the cooling, wall/piston speed is.

For the little micro-horsepower engines discussed here, an external load is moot.


Sorry. I ramble on.

[Tom Booth]

(...)
When a piston compresses a gas the piston does work to the gas. This means there is a force, or momentum, driving the piston in. This is adiabatic temperature rise with work. The energy/work, driving the piston in, is absorbed by the gas.

If both strokes are identical and adiabatic, the temperature volume and pressure will be the same as the initial point. Zero net work will be available to output. Area enclosed will be zero. One adiabatic path/line. Work gained minus work applied equals zero.

What you describe here, generally speaking, is certainly true. In an "ideal" engine. but I think there are a few things being overlooked.

During compression,in a real engine, (being heated at one end, the end that the working fluid is compressed into) the gas gets hot. This is certainly true.

Infact it gets very hot. Hotter even than the heat source.

Which seems strange but if you think about it, in a Stirling engine the hot end is heated continually maintaining a more or less constant minimum temperature.

Adding compression on top of that results in additional heat above that minimum. The minimum being the temperature of the heat source or "hot reservoir"

Remember the gas is being compressed by atmospheric pressure. Yes, this is just acting like a spring returning some work done during expansion that displaced the outside atmospheric air in the cylinder. I'm not advocating free energy from atmospheric pressure, just trying to get a clearer picture of what actually goes on in a Stirling engine.

If the gas is in contact with the heat source but as a result of being compressed the gas is hotter than the heat source, then which way is heat going to "flow" at TDC?

OK I think Matt or "nobody" or whomever has tried to tell me many times that heat "rejection" to the sink takes place during compression and my argument has been that this is impossible because the cold "sink" is thermally isolated from the working fluid at that time, covered by the displacer. So if there is "heat rejection" during compression it is putting heat BACK into the heat source and not into the "sink".

The compressed gas is at least partly COOLED down by the relatively LOWER temperature of the heat source.

It is possible that some heat of compression is "lost" at the hot side back into the heat source before the "working fluid" re-expands.

What matters for heat flow is the relative temperature and during compression the working gas is hotter than the heat source and thermally isolated from the "cold source".

This is already such a departure from conventional theory about a heat engine operating I think we need to just forget everything and start over with a completely blank slate as all this old theory just clouds up thinking and actual objective observation.

The tendency is just to reject all this "new theory" as irrelevant or inconsequential and stick with what is already known. But let's continue.

Let's just say that theoretically, the compressed gas lingers for some time during the brief "Dwell" period at TDC and the compressed gas COOLS by rejecting heat back into the supposed heat source or hot "reservoir".

Although the heat source is relatively hot, the compressed gas is hotter.

This is exactly how a heat pump works. The refrigerant is compressed so that it becomes hotter than the "hot reservoir" and heat is ",pumped" UPHILL so to speak. When the refrigerant is then expanded it gets very cold and can absorb heat from the cold outdoors in winter because after being compressed and loosing heat, when expanded it gets colder than the "cold reservoir" and the refrigerant then absorbs heat from the relatively HOT winter air.

So after lingering at TDC and loosing heat the working gas expands.

If it just expanded it would get cool because 1) gas cools when expanding 2) some heat was "lost' to the hot reservoir at TDC, but 3) it is also doing work displacing atmosphere and possibly driving a load which causes additional cooling.

And no, the load on an air cycle refrigerator or expansion engine is not just to "prevent over revving". It is to take out additional "internal energy" which is what results in extreme cooling to potentially cryogenic temperatures.

So on expansion the hot gas SNAPS almost instantly to cold and contracts but is now colder than the "cold reservoir" so absorbs some heat at BDC and during "compression".

The piston is accelerated during "compression" by atmospheric pressure so that atmospheric pressure is converted to velocity/momentum and back into heat at TDC so that at TDC there is a sudden burst of heat that drives the piston back out, then at BDC there is a sudden SNAP back to cold that results in a "vacuum".

This is very rapid.

If there is a heat "flow" it is a flow out of the colder heat source into the hotter heat source like a heat pump.

This is why, contrary to expectations, attempts at using Stirling engines as heat sinks to cool internet/Google data centers didn't work. The Stirling engines acted like heat pumps putting heat back into the data center or at best were like insulation, preventing cooling.

https://youtu.be/pVZVnxrcosc?t=912

Now, of course I'm "exaggerating" this probable "heat pump" type action in order to describe it. If there is any actual cooling effect it is so slight as to have gone almost completely unnoticed for over a century.

Experimentally, though, I've been finding it very difficult to detect any real "heat rejection" to the "cold reservoir" in any Stirling engine.

After most heat "leaks" are plugged, by eliminating avenues for heat to be conducted through the engine in ways other than through the working gas, like conduction through bolts or convection or metal engine body etc. It seems the working fluid is carrying heat the opposite way.

Maybe I'm wrong, but that is the basic "Tom Booth" Stirling engine theory of operation, after ten years or researching and experimenting just trying to understand how a Stirling engine REALLY operates.

I'm not trying to disprove Carnot or prove Tesla. Ultimately I think they were both wrong.

The basic assumption that heat would "automatically" FLOW "downhill" into a "cold hole" was wrong. The entire theory of heat engine operation: that the engine derives energy by intercepting heat as it tries to "flow" from a hot to a "lower" cold "reservoir" is wrong.

Tesla was at least half right by recognizing that "heat" is a form of energy that can be converted to other forms of energy not an indestructible fluid.

You can't get 100% efficiency at absolute zero.

First of all, the gas would all liquify. A hot air engine needs air to expand as well as atmospheric pressure so the gas can "contract".

A "cold hole" engine like Tesla envisioned is not feasible because, for one thing, apparently the "cold hole" is actually the effective heat source for the "heat pump" so the cold hole cannot be too cold.

Using "atmosphere" or ambient as a primary heat or power source by allowing heat to flow into a "cold hole" is, at best, a misconception. Heat is not going to just flow through or into the engine due to the presence of a "cold hole".

If you start out with a cold engine with cold working fluid and then add ambient heat to expand the gas. As the gas expands it displaces atmosphere until pressure equalizes. The gas has to get cold enough in the process of expansion for atmosphere/ambient to push the piston back. That is not going to happen by passively "allowing" heat to flow in. This is not the same as being able to apply unlimited amounts of heat from an above ambient heat source.

A heat pump/Stirling engine needs an actual energy source to displace atmosphere and using atmospheric heat to displace atmosphere itself is a much more difficult proposition than using a high heat source to displace atmosphere.

Since a Stirling engine IS a heat pump, that makes the "cold hole" the actual energy/heat source, which is a rather huge disadvantage when trying to displace the vast atmosphere each cycle.

It might just be barely possible, I don't know. All I do know (I think) is that a Stirling engine appears to work almost directly opposite to what is generally supposed.

[Fool]

You seem to have the same understanding for a heat pump as I. For a C.O.P. of 3. One unit of work is input and 3 units of heat are supplied to the hot mass/side/plate/source/sink/room/etc.

One unit is supplied by converting, work to heat, and two units are supplied by transfer of heat from the cold mass/side/plate/room/etc.

That makes the hot mass hotter and the cold mass colder. If the masses are small enough to measure that.

What would happen if we tried to pull work out of that pump, at the same temperatures? Wouldn't we get one unit of work out? Wouldn't three units of heat have to enter the hot side? Wouldn't two units of heat have to be rejected to the cold side? Isn't that the law of reversibility? The engine/pump works the same only reversed?

Wouldn't a heat pump C.O.P. of three dictate an efficiency factor of 1/3 for that engine?

Wouldn't the two, connected together, produce zero net work and zero net heat transfer? 100% perfect recycle of work and heat? We can't have one without the other? Mathematical proof? Maybe?

All that was assuming an ideal perfect reversible engine/pump.

Also, wouldn't all real pump's be worse than those two?

Lastly: I have to fight dyslexia. I need to be very careful and double, triple, quadruple, check what I write, and still get it wrong sometimes.

Isn't the displacer tight against the hot side with the bulk of the gas in the cold side during compression? And tight against the cold side during expansion? Assuming an ideal Stirling Engine cycle too. Or do I have the concept backwards?

Doesn't a real Stirling Engine try to mimic that with a harmonic motion corner cutting 90° phase processes?

Wouldn't that mean that the bulk of the gas is cooled during compression and heated during expansion?

Sorry I ask too many questions.

[Fool]

Matt,

I think I have gotten it. Breaking up the system into two or more subsystems and having a PV chart for each. Right?

[VincentG]

I think it may help to consider the neutral state of a real Stirling engine. So considering an engine with a 300k to 600k thermal cycle, the neutral state of the engine would be 450k with the power piston at middle dead center.

A heat pump can have this neutral state biased to one side or the other depending on what the desired outcome is.

[Tom Booth]

...

Isn't the displacer tight against the hot side with the bulk of the gas in the cold side during compression? And tight against the cold side during expansion? Assuming an ideal Stirling Engine cycle too. Or do I have the concept backwards?

Doesn't a real Stirling Engine try to mimic that with a harmonic motion corner cutting 90° phase processes?

Wouldn't that mean that the bulk of the gas is cooled during compression and heated during expansion?
...

Lots of interesting questions.

A lot depends on the engine and what type, how it's built etc. There is a lot of room for sloppiness as far as advance angle, how close the displacer comes to the heat exchanger plates etc. And an engine will still operate.

An LTD with a magnetic displacer seems to be "better" in that the displacer snaps up to cover the cold (top) side very quickly and it stays tight up against the cold side until it suddenly drops.

Attached to a crank that kind of "ideal" displacer movement is nearly impossible.

The angle or "advance" (timing) can also be adjusted to any angle in some engines.

With a standard 90° "advance" on the displacer attached to a crank the piston is 1/2 through the compression stroke when the displacer lifts from the hot side to cover the cold side. The displacer is 1/2 way up at full compression.

The cold air is being pushed down to be heated while the piston is nearing full compression and the cold air is still being pushed down away from the cold side during and after full compression. So IMO not much chance for heat rejection to the cold side during compression.


With a magnetic displacer the cold side is covered completely through something like 45° more or less, depending on how the magnet is adjusted, before during and after full compression.

In many of my experiments where I saw little if any heat leaving the cold side It was using a magnetic displacer.

At high RPM though, the magnet (on the piston) moves so fast the displacer barely has time to lift at all, just making a little "hop" at TDC.

The top of this engine is acrylic, a very poor heat conductor, and that is covered with a silica aerogel blanket. Supposed to have insulating properties close to vacuum.

https://youtu.be/i9nz0vt7eQA

[Tom Booth]

BTW, in case it isn't obvious. What I think is peculiar about this is that without the insulation the engine normally runs more slowly and the displacer goes all the way up.

I had assumed when first trying this kind of "insulating the sink" experiment that with "heat rejection" blocked, the engine would quickly overheat, slow down and stop.

Instead it runs faster and the displacer lifts so little, I'd have to assume heat input is reduced to a bare minimum.

For whatever reason, preventing heat loss or retaining heat at the "sink"... (Or is it blocking heat intake from ambient by the "heat pump" at the sink?) apparently increases efficiency; less heat out, less heat in but engine runs at higher RPM.

Originally I thought insulating the sink might block ambient heating INTAKE on the cold side of the "heat pump" which would increase efficiency by increasing the ∆T.

A greater ∆T, makes possible a higher or stronger "adiabatic bounce".

Whatever is going on it seems contrary to conventional wisdom, unless the insulation is actually improving CONDUCTIVE heat loss making "heat rejection" more effective, but to me that seems contrary to reason.

Years ago, I just thought I would do a quick experiment, put some insulation over the sink, the engine would overheat and stop and that would be the end of that.

Now years later I'm still puzzling over why the engine runs faster with the cold side insulated.

[matt brown]

Matt,

I think I have gotten it. Breaking up the system into two or more subsystems and having a PV chart for each. Right?

Correct. In the paper Tom linked to (where the regen graph came from) scroll down to fig. 13 which gives another idea of how to plot values. The original paper I saw this composite PV mentioned was simpler, but had a better description. I'll look for it this weekend (I have a good idea on time stamp).

[Tom Booth]

Don't really know what you mean by double PV plot but it reminds me of the Andrew Hall demonstration video, towards the end where he says (about a Stirling engine being able to run without a flywheel)

Something like (paraphrasing) " I call that a one stroke engine. The expansion stroke is a power stroke, and the contraction stroke is ALSO a power stroke.".

I've been thinking along the lines that the expansion stroke is a power stroke and the return stroke is a heat pump.

The power stroke drives the engine which, due to the change in position of the displacer, in effect converts the engine into a heat pump on the compression stroke.

Between the regenerator and this apparent heat pump action the engine "resists" heat flow enough to divert the energy into power production.

I've been thinking that perhaps a greater effect could be produced in an "ambient heat engine" if the engine were pushing against an artificial atmosphere. Sealed, compressed air buffer space.

[Tom Booth]

The reason for the later conclusion (or speculation) is I've been thinking hard on what might be the actual difference between a normal heat engine and an "ambient" heat engine.

I think I've more or less narrowed it down to the relative pressure differential between the internal engine pressure and the external atmospheric pressure.

In experiments running an engine "on ice' (really ambient heat) the engines always seemed to run disappointingly slow and without much vigor. Almost as if running in slow motion.

When running an engine with a greater than ambient (actually hot) heat source atmosphere is displaced. The gas ('working fluid") is "lean" (rarified) and contracts easily when cooled.(or is compressed easily by the relatively dense atmosphere) during compression.

So in the ordinary hot air engine you have a relatively"thin" but hot inner working fluid and "thick" dense outer atmosphere.

In a COLD engine the working fluid is cold and dense relative to atmosphere.

Expanding the cold dense working fluid with ambient heat just tends to equalize the density by expanding the working fluid to match that of the atmosphere. What seems to be lacking is the immediate "push back" or air spring effect from the displaced atmosphere. The expansion work is lost. The dense working fluid can hardly be made thin enough when expanded by ambient heat alone to "contract" or be pushed back in by the more dense atmosphere, because, well, it is not more dense than the relatively cold working fluid.

Maybe I'm barking up the wrong tree but I haven't been able to come up with much of anything else.

Enclosing the space above the piston might create an "artificial atmosphere" with higher density to act as an "air spring". Maybe a snift valve in the piston itself could gradually increase the pressure and density of this "artificial atmosphere", but are pressure and density the same thing? Should the buffer space be chilled as well?

Somehow this seems rather dubious to me because it makes expansion more difficult.

Thinking it through though, this "buffer pressure" would be returned for use during the compression (heat pump) phase.

If what cools the working fluid in a "normal" heat engine is expansion work against a higher density atmosphere then would having an artificially high density buffer space in an "ambient heat engine" provide a stronger air spring to push against to act as a temporary energy store for the compression work to pump heat back, increasing the temperature differential?

Maybe just a literal metal spring might do the trick.

Intuitively this somehow seems like a loosing battle and I have a nagging feeling I must be missing something but I can't quite put my finger on it.

Resize_20230802_060107_7424.jpg (154.15 KiB) Viewed 4765 times

As a theory though it is at least testable. Possibly just adding a small spring or weighting the flywheel so the engine (working fluid) does more work on expansion that would be returned (by gravity) during compression. On the theory that expansion work output that returns during compression contributes to the "heat pump" phase.

Somehow, pushing heat back into the heat source seems a waste though if it is just allowed to dissipate.

Maybe some kind of heat diode on the hot (ambient) side could retain heat if the heat exchanger plate exceeds ambient temperature but allow ambient heat through if the temperature drops below ambient. Or keeping it simple, maybe just a thicker plate with more heat storage capacity would serve as a temporary heat battery. It would only need to store heat for a millisecond or in essence "reflect" the heat of compression back into the working fluid.

[Tom Booth]

You seem to have the same understanding for a heat pump as I. For a C.O.P. of 3. One unit of work is input and 3 units of heat are supplied to the hot mass/side/plate/source/sink/room/etc.

One unit is supplied by converting, work to heat, and two units are supplied by transfer of heat from the cold mass/side/plate/room/etc.

That makes the hot mass hotter and the cold mass colder. If the masses are small enough to measure that.

What would happen if we tried to pull work out of that pump, at the same temperatures? Wouldn't we get one unit of work out? Wouldn't three units of heat have to enter the hot side? Wouldn't two units of heat have to be rejected to the cold side? Isn't that the law of reversibility? The engine/pump works the same only reversed?

Wouldn't a heat pump C.O.P. of three dictate an efficiency factor of 1/3 for that engine?

Wouldn't the two, connected together, produce zero net work and zero net heat transfer? 100% perfect recycle of work and heat? We can't have one without the other? Mathematical proof? Maybe?

All that was assuming an ideal perfect reversible engine/pump.

Also, wouldn't all real pump's be worse than those two?

(...)

Regarding the above, a lot more good questions.

IMO a Stirling "heat pump" is not a Stirling engine "running in reverse", or driven in reverse.

A Stirling heat pump IS identical to a Stirling engine running in the same direction.

This gets ambiguous with a thermal lag or Ringbom that can run in either direction.

It's just that by driving a Stirling engine with a motor the load is lessened, the "heat pump" no longer has to waste heat driving itself so it works more efficiently as JUST a heat pump.

In other words, if you spin a Stirling engine manually in the SAME direction it usually runs it will act as a heat pump.

So, if the engine runs as an engine 1/2 cycle during expansion and is DRIVEN by stored momentum/atmospheric pressure 1/2 cycle during compression you don't have all the mechanical loses of a heat engine running a heat pump.

A Stirling "heat pump" will also work as a cooler if driven in reverse, just like a Stirling engine will run in reverse on ice but the heat "flow" direction is reversed.

It gets complicated and can be difficult to sort out and keep track of.

I think, then, running as an engine, a high heat source is used to drive the Stirling machine as both an engine and heat pump simultaneously.

The heat pump is "resisting" heat flow by trying to push it back, acting kind of like a dam holding back a river so the heat can only leave as "work".

In that sense the water wheel analogy works but the engine is not letting heat just pass through freely like a water wheel. It's holding the energy back forcing it to change or be converted.

So I THINK, (could be wrong) that the heat input on the hot side does the expansion work that does external work output and stores some heat for use (in various forms) for driving the engine during compression.

A Stirling engine that is driven IS a Stirling heat pump.

So the ENGINE drives during expansion and the "heat pump" is driven during compression.

That is something of a simplification. I think there is actually more functional overlap than a simple clean division consisting of expansion(engine) compression(heat pump).

Work output during expansion leaves the working fluid cold which is what draws heat in at BDC to one degree or another, so the distinction between heat engine and heat pump is not so clear cut.

This is a complete revision of heat engine theory and I'm putting it forward as nothing more than that: THEORY to be tested or as a basis for experimentation, not any kind of fact or "claim".

The theory is not entirely baseless though. IMO the whole Caloric based concept that has largely dominated heat engine theory, in a somewhat disguised form, cannot be true.

Based mostly on experiment and observation, this seems at least plausible.

[Goofy]

But as I remember, Stirling constructed the engine where the power-piston is on the "cold-side".
So if you are running it with the ice on the bottom side, you are actually somehow running it "reverse" of what it´s mend too ?

Perhaps that why it seams to run slower that way. Try to put the ice on top and let the ambient heat "go in" on the bottom,
just as if it was on hot water.

There must be a difference either way or ?

[Tom Booth]

But as I remember, Stirling constructed the engine where the power-piston is on the "cold-side".
So if you are running it with the ice on the bottom side, you are actually somehow running it "reverse" of what it´s mend too ?

Perhaps that why it seams to run slower that way. Try to put the ice on top and let the ambient heat "go in" on the bottom,
just as if it was on hot water.

There must be a difference either way or ?

How it seemed to me is the sluggish response running "cold" (on ice) , aside from the temperature difference being smaller (on the kelvin scale), probably about 1/2 what it is running on hot water. It's just my subjective impression I guess.

I haven't tried dry ice or liquid nitrogen.

Other factors could be responsible, like the lower temperature causing binding of the piston in the cylinder.

I did put ice on top of an LTD covered in a little insulated chamber once and it seemed to run quite well for a while then somewhat mysteriously "seized up" and refused to run. I couldn't get it to start up again until I thoroughly warmed up the engine and ran it on hot water.

I can only speculate. Maybe the cylinder shrunk from the cold and caused binding. Maybe the cold "sunk down" into the engine making it too uniformly cold. Not enough data.

Here is a problem I had generally.


https://youtu.be/DmkVR7hF14Y


More detail in the video description.

I found that unlike an engine running on hot water, I could not get an engine to start up on top of a cup of ice. The vacuum insulated cup had been kept in the ice box. Then I dropped in an ice cube.

The engine would not run at all that way. The cold air just pooled in the bottom of the cup.

I kept adding ice until the cup was full, one ice cube at a time. But I still could not get the engine to even start up at all.

The Cold in the cup just would not "rise up" to cool the bottom of the engine the way steam rises. And/or the heat was making no effort to dive down through the engine to get to the cold.

Finally to do these controlled ice cube experiments I had to fill the cup with ice water until it touched the bottom surface of the engine so that the floating ice cube and cold water came into direct contact with the bottom of the engine. Then the engine would start and run s. l o w l y.

I had this frustrating difficulty whenever the cold was under the engine, so perhaps you are right. The ambient heat needs to be let in from underneath.

The same tendency for cold to pool on the bottom even after INSIDE the engine seems to be a problem.

Anyway, this is what I mean by the heat does not automatically "flow" towards cold to power the engine which is what I was originally expecting. That heat would automatically "flow" towards the cold.

Getting heat into the engine has to be "engineered" in some way, by using direct contact or convection or moving air or radiant heat or something.

A cup of cold ice will just sit under the engine indefinitely and the engine will stay too uniformly warm to operate, as long as there is air around or above the ice.

Maybe the air is just too good an insulator and cold air will just pool under the engine.

As I recall, having the ice on top of the engine did seem to work better, for a while, but for a rather short while, until the engine got "filled up" with cold or something. That was so frustrating I haven't done much experimenting with the ice on top, I should try to get to the bottom of it though.

My general impression though is "heat" itself is completely indifferent and not at all anxious or willing to "flow" towards or into a cold sink even millimeters away on the opposite side of the plate.

[Tom Booth]

Re-reading my last post and thinking about the time I ran an engine with ice on top, a thought occured to me.

I had the ice on top of the engine in a small cup I made out of aluminum foil in direct contact with the cold (top) plate of the little LTD.

I had a styrofoam disk on top of the engine with a hole to receive the "cup" of ice. The hole was then covered with another layer of styrofoam.

If my "heat pump" theory is correct, then the engine would need to be able to draw heat from the cold side to "pump" over (or down in this case) to the hot side on the bottom.

Ice itself is a very good insulator. So I've heard.

So if heat were being "pumped" out of the ice, there would be less and less heat available as time went on and no ambient heat available to replace it as the ice was buried in insulation. Some heat taken from the ice on the outer surface would leave the rest of the ice to serve as "insulation"

My theory was that the ice would get colder and the ∆T would increase so the engine would run better and better, and this seemed to be the case, when I went to show my wife.

But after a while the piston seemed to freeze up. The power piston was also on top next to the ice surrounded by insulation.

It seems too far fetched to imagine that the piston literally FROZE due to getting ice cold, but if it started out nearly cold as ice by being in contact with ice under the insulation, it might only take a very minute drop in temperature, or a very slight "heat pump" cooling action to cause some moisture or condensation in the cylinder to freeze.

At the time I was showing the running engine to my wife and the way it looked like the engine abruptly jammed or something, she thought, and said out loud that "maybe it froze..." Meaning that literally.

I was rather excited about how well the engine was running, then very frustrated when it quit running and wouldn't start again just as I was trying to demonstrate my "successful" experiment.

I can relate that I've had several instances where an engine left to sit on top of ice would refuse to start up, though in contact with ambient air on top. Hitting the top with a hair dryer can help, but not as much as you would think. If the engine gets "saturated" with cold, it can be quite impossible to get running again.

But could that engine have become "saturated with cold" with ambient heat going in through the bottom?

The ice would not have much heat to "pump out" so, when the engine finally quit running or froze...

Anyway, I have more experimenting to do once I get some new thermocouples.

[Goofy]

I have build several expansion engines (not exactly Stirling engines), when running on compressed air at perhaps 6 bar, they
easily go down to -50 Celsius on the exhaust, in one stage. At least that's what my (cheap) digital thermometer can go down to.
They also seems to suddenly "get stocked" or freeze up, but I think that because of moisture in the compressed air, as I don´t
do anything special to remove it after compressing.
Btw, this will only happens with load on, because PRESSURE is converted to work during expansion.

We have learnt over and over again that it is HEAT that is converted to work, but there shouldn't be any thing wrong with at gas at 200 bar being at 10 kelvin, and still do "work" on a piston ?
We are also told, that all the energy in a compression process, turns into heat. So when put it in at storage tank to go down to ambient temperature, what energy are you left with ? Still a lot in my opinion.
OK, we then tell our self, that the stored energy comes from the ambient heat energy, and then the equation comes to an equal.

If we imagine an engine/turbine expand a gas at 200 bar/10 kelvin out in freezing/vacuum space will it run at all ? ? ?
If yes, then by what heat ?
Carnot tells us there is still "a lot" energy down to 0 kelvin, so . . .