Stirlingengineforum.com

In theory, given that hot and cold are really just artifacts of human perception, or how things "feel" relative to our own body temperature, what hot and cold really consist of is more or less kinetic energy.

So instead of a "hot" and a "cold" area of the engine, a partial vacuum might serve as an effective low kinetic energy "sink" or buffer space

At any rate, a partial vacuum might lower the boiling point of water, so that the engine could operate at a much lower temperature.

Something like this maybe?

@Tom Booth Interesting findings around the Leidenfrost effect! In a weird coincidence, I was actually saved from a burn a couple of days ago by the Leidenfrost effect when working on this engine. My hand was dripping wet from the sink and the back of my hand lightly touched a very hot surface, but all I felt was a sizzle for a split second and my hand was completely unharmed.

By the way, it looks like the Leidenfrost engine video that went private was successfully archived:

https://web.archive.org/web/20170816145 ... e=youtu.be

So cool! I love how they created what looks like an axial flux alternator on a piece of ice.

I would definitely like to know whether the Leidenfrost effect plays any role in our engine. The only true way to find out I think is to build a transparent engine.

It has been my contention for years, that the cooling (or heat removal) is not needed. The heat is converted to mechanical "work" output, which seems obvious in a way but very weird and hard to think about or accept in another way.

Yes, I definitely think there's something to the theory that heat is converted to mechanical energy and thus there's no catastrophic pressure buildup. The only time that can happen is if more heat is pushed through the system than the engine can handle. When the engine is being started, it can be easy to blow off the diaphragm because of the pressure buildup even with modest heat if that pressure isn't released. When the engine is running, and even in the moments before it actually takes off, way more heat can be applied and the pressure mostly becomes a non-issue.

As said by others, external cooling of the cans does not seem to be a major factor in energy conversion if it's a factor at all.

@chaorro Yep, it seems your experience is reflecting mine. Heat removal doesn't appear to be a requirement, as the engine seems to have a point of maximal conversion where the temperature of the cylinder remains stable. Cooling may be useful for preventing heat sensitive parts from breaking down, but it's not clear to me whether cooling improves performance. Makes sense given that most other systems that have heat sinks use them to handle an overabundance of heat, but the presence of a heat sink isn't required for fundamental operation.

So far I haven't noticed the temperature of the resonator section having a noticeable impact on engine performance. In the video of my engine I posted, the top section was actually fairly hot even with the towel. I also haven't witnessed the presence of a wet towel having any impact either. That doesn't mean it *can't* do anything, but it doesn't seem to in the case of the resonator. A heat sink probably still is a good idea for mitigating overload when too much heat is applied (which may only be important if there are heat sensitive parts on the end).

The only evidence that I've seen that a heat sink is needed is in NightHawkInLight's thermoacoustics video.

https://youtu.be/abswNCqnMRQ?t=895

Then again, we don't actually know whether he tested that engine without any cooling.

*AND* we don't necessarily know whether the wet thermoacoustic engine is operating on the same principles as the ones we usually see, or if any of these engines are truly "thermoacoustic". Having played with the wet engine a lot this week, I'm leaning more towards thermoacoustics than I was before. There's definitely a specific frequency that the engine wants to operate on, and any increase in heat input results in greater amplitude but not greater frequency.

When I increase the temperature, I have to release internal pressure. When I decrease the temperature, I have to bleed air back into the cans.

So not only do you need to regulate the applied temperature, but also the internal pressure of the system to maintain maximum performance.

I'm curious how you've discovered this. When it comes to my engine, I haven't needed to worry about the internal temperature at all. I'm sure it's possible that a well regulated internal temperature could improve or stabilize performance, but I've not found it at all critical to do so. I've been able to apply a considerable amount of heat to my engine and only had to worry about not overapplying heat.

At any rate, a partial vacuum might lower the boiling point of water, so that the engine could operate at a much lower temperature.

@Tom Booth Very interesting idea! I think there's a lot of potential there. The entire system should be at the same pressure, right? With the diagram you drew (awesome you did that btw), I could get the impression that only the area above the diaphragm would be depressurized.

Dunno if I linked to this already, but I came across another paper that describes using a rice engine. The researchers tested a water-ethanol solution as the working fluid (not sure how that's the right term here) and they observed a faster time-to-start.

https://sci-hubtw.hkvisa.net/https://do ... /1.4766940

From the conclusion:

We estimate and measure the temperature difference DTo for the onset of thermoacoustic oscillation under three different conditions. Under the first condition, dry air is used as the working fluid; under the second condition, air and water are used; and under the third condition, air and diluted ethanol are used. The experimental results show that the values of DTo for these experiments are 290 C, 56 C, and 47 C, respectively. Furthermore, the smallest measured temperature difference while maintaining thermoacoustic oscillations (DT) was 42 C and 21 C under the second and third conditions, respectively. Hence, we can conclude that the introduction of water or ethanol into a thermoacoustic oscillator has a large effect on the reduction of the temperature difference required to drive the oscillator.

56 C down to 47 C is pretty significant.

the diluted ethanol had a density of 0.83g / cm3 before the start of the experiment

Eh, I gotta get ready for work and don't have time to figure out how that can be converted to a percentage. I have no idea if the solution they used is one that can be used safely.

Let's put it this way.

Boiling water causes all the "air" proper (oxygen, nitrogen etc. gas) to be replaced by water vapor. (Once there is a lot of steam from the pin hole)

When the water vapor condenses then, due to the cyclic drop in temperature/pressure (power/work output) what is left? no air / vacuum.

I'm thinking of the video when the rice engine goes crazy. Lots of steam out the pin hole, which is then closed by sticking an actual pin in the hole. Then "AHHHhhh!!!!"

Reminds me of this:

https://youtu.be/p3b9pK-O6cE

There is another way to regulate pressure/vacuum.

By increasing the efficiency of the condensation medium. More material or coarser or finer, more or less surface area. Also, work output. Heavier weight, bigger magnet/generator, more or less electrical output.

Of course I'm theorizing wildly with no real expectation that it (whatever "IT" might be) actually working, or exactly how something like that might work.

A total vacuum, in the upper space could not work, as there would be no pressure to push down the diaphragm.

The can would probably explode or collapse, like in the video.

Getting the "right" balance might take a lot of experimenting, if it is possible at all.

But, with that earlier video of the rice engine going wild, I think the effect is that a vacuum has been developed by displacing the air with steam.

The expansion and contraction of the water vapor (in a partial vacuum) is much stronger than expansion and contraction of air.

Creating an artificial atmosphere of partial vacuum (or perhaps higher than atmosphere pressure ?) might make things more controllable

But I think, higher than atmospheric pressure would make boiling more difficult (but condensation easier ?) while a partial vacuum would make boiling at a lower temperature easier.

I had plans a long time ago to do some such similar experiments with a vacuum engine, but it seems my vacuum pump was misplaced (or stolen ?).

Anyway my intention with using a vacuum would mostly be in hopes of avoiding use of potentially volatile/flammable alcohol or something similar with a lower boiling point

A water/alcohol mixture might be (IS ? in light of the info just posted by tenbitcomb) a more practical way to reduce the necessity for so much heat.

A water/ammonia mixture crossed my mind as well, but hot ammonia gas could be lethal I think, but my curiosity often gets the better over my caution or common sense.

If the space above the diaphragm power piston were sealed off, maybe with a pinhole in the diaphragm ?

As the pressure from the steam increased, there would be a generalized pressure increase, above and below the diaphragm, perhaps without any necessity for equalization with the outer atmosphere.

Of course, that may be asking for more trouble as far as potential build up of pressure / explosion.

A sturdier engine along with some pressure relief valve would be needed as a safety measure.

If completely hermetically sealed, there should not be any need to add additional water, it would be a completely enclosed system But when not in use, there would be a permanent vacuum. Unless, of course, some sort of valve could be open and closed as necessary.

One thing I think is important to keep in mind in regard to potential "overheating".

Over the years, since at least ten years ago, I posted a reference to a patent application relating to a solar Stirling engine, that read:

When powering a Stirling engine with solar energy from a mirror array, the solar energy supplied to the engine is substantially constant, and so the load must be maintained at a sufficient level to use all the solar energy supplied by the mirror array. If the load drops, the engine very quickly overheats and is damaged.

To provide a level of control, the flat mirror segments on the mirror array can be mounted such that they can be moved by an actuator. Controllers activate the actuators and pivot the mirrors to produce the focused cone-shaped solar beam. The amount of energy received by the receptor can thus be varied. Thus when overheating is detected the mirror segments are moved out of focus to reduce the amount of energy received, such as when the load on a Stirling engine drops.

To me, that suggests that the power, in this case electrical output acts to remove energy/heat. Similar to an actual "heat sink".

Normally, we don't think of heat as being equivalent to electric output, but apparently the addition of an electric generator for power output helps keep a heat engine cool, which seems counter intuitive.

As stated above, the removal of the "load" results in a heat bottleneck.

For a higher power heat engine, in other words, the heat energy input must be utilized in some way. Just lifting a nut may not be enough "work" output to keep the engine from overheating.

Yep Tom, you're right on the money.

There's a kind of tradeoff and balancing act going on here between efficiency and how long the engine can operate at a given rate of heat transfer. The engine doesn't necessarily need to be cooled, nor is that cooling necessarily a good thing for fundamental operation. It seems capable of running continuously if the input isn't so much that the heat can't be eliminated either through work or loss. Because the water vapor both transfers and stores heat at a greater amount than air alone, the engine is prone to hitting a wall. I believe this is why I've been able to sustain the longest running time by carefully adjusting the heat and even pulsing it on and off. A sustained input is obviously good for smooth operation but can easily be overdone if the heat source is adding energy to the system at too great a rate. Once the engine is running, it seems to run indefinitely as a totally closed pressure system if I use my mini butane torch (really a glorified lighter). Yet, when I run it on my camping stove, even at the lowest possible setting the engine seems to eventually hit a wall. This is a problem I really want to overcome because part of the appeal of external combustion engines is the ability to burn a variety of fuels, and that means losing some control over the rate of input. Maybe that can be somewhat addressed by giving the engine enough work to do (i.e. greater weight or pressure to push against), so that is also something worth testing.

Cooling reduces efficiency because that's not where you want the energy going towards. What *would* make cooling more useful is if it could somehow be engaged only if there's more heat than there is work. How that can be accomplished in a way that doesn't require a lot of tech and an external power source, I have no idea.

The "wall" that I refer to appears to happen at the same time the diaphragm begins to push too far outward, which I think means that not enough work is being done. This changes the resonance of the diaphragm and either reduces performance or stops it all together. At worst, the diaphragm has a blowout. In putting aside the possibility of adding more work, *maybe* a way to address this is to add a pressure release valve to the top cylinder. It would maintain a level of pressure but not allow it to exceed that point. In homebrewing beer, people who ferment their beer directly inside of a keg use something called a Spunding valve to keep the keg pressurized at a specific level while permitting CO2 to pass out. I'm thinking a relatively low-pressure version of this type of valve may help the engine avoid hitting the wall or at least encountering a diaphragm blowout.

What does make me unsure about my theory of the wall is that the opposite doesn't seem to be true. Tom, you're right that there's a vacuum being formed, and when the engine is able to perform more work than there is available heat the diaphragm pulls inward. The only way that could happen is if the pressure inside is lower than the external environment. Yet the engine doesn't appear to struggle in this state. I'll have to test this again, but I remember witnessing this happen several times and being surprised that the engine didn't slow to a crawl or stop like it does when heat is oversupplied. In which case maybe it's *not* the change in resonance that matters (so much?) as is the rate at which heat can be converted to work.

But I think, higher than atmospheric pressure would make boiling more difficult (but condensation easier ?) while a partial vacuum would make boiling at a lower temperature easier.

My gut tells me that better condensation might be preferable. Lower pressure would result in a faster start and perhaps a lower temperature, but that might matter less with sufficient heat input anyway. Dunno. Definitely would be worth testing both. Next time I'm at Harbor Freight or whatever, I want to see if there are any valves etc. that I can buy that would apply to this engine both in general and for trying out these ideas.

If I can get some Everclear today, I'll try using the highest safe concentration of water and alcohol. I don't recommend that just anyone reading this try it, though. I've had enough experience exposing alcohol to the presence of flames, and lighting things on fire in general, that I have a pretty good idea of what's an acceptable level of risk *for me*.

Yeah, it's too bad ammonia probably isn't a good idea. I didn't know this, but apparently ammonia was actually used in refrigeration before Freon was invented to replace it.

I start by heating the stack, once the stack reaches a threshold temperature, the membrane will start vibrating all by itself. As this happens, internal pressure builds in the stack and needs to be released, by keeping the single opening slightly open, the stack can be further heated until slightly less than the desired stroke is reached. Then when the hole is sealed, the full stroke of the stack will engage. Then the heat source must be regulated to maintain the optimum position of the diaphragm.

Very interesting, chaorro! Perhaps this is more evidence that cooling is a fundamental requirement for operation. If the stack, which I'm assuming you mean to be the rice/glass/pebbles, works when heat is applied to it directly, then that may mean that one of the water vapor's roles is to transfer heat to the stack for the precise reason of heating up that stack. It can explain why the version of the engine that's just one can with steel wool to hold up the rice is able to operate.

And yes, that's more or less how I've been operating my engine. Most of my heat sources simply provide way too much heat even at a low setting, hence I have to regulate the heat or briefly turn it off entirely if I notice the engine is too far into overdrive.

It's starting to feel like a heat battery that leaks mechanical waves.

Sure seems that way. I at least like the idea that's how it works.

I just gotta say I didn't expect this engine to be this fascinating and nuanced. For the time being I've put aside any plan I had to build a Stirling because I'm absolutely enchanted by the simplicity of what we have here. The materials are just stuff you can find in any random garbage can. No glass test tubes, no steel wool, no heat sink, no graphite. Tell me this wouldn't be an awesome thing to demonstrate in a high school classroom.

Oh yeah, so there's a few things I want to hopefully get to testing out this weekend.

Building a clear version of this engine might be helpful in visualizing what's going on. I'm thinking I want to give my hand a try at cutting some glass cylinders out of wine bottles and see how that goes. If I can do that and stick a mesh between two of them and film it in slow-mo with my GoPro 9, maybe it will provide some insight. I regret not getting a thermal camera, and really don't want to spend that money at the moment, but I've got a regular infrared thermometer from Harbor Freight and that might be good enough. Can't guarantee that I'll be able to get to it, but I don't think we have existing data to work off of, so it's a worth experiment.

I'm also interested in seeing if a piston system can work. I'm thinking of replacing the diaphragm with a small piston connected to a flexure spring. I can 3D print the flexure spring. Though I'd like to 3D print the piston, the first version can be made of epoxy since that can be easily made to be super airtight and heat resistant. As much as I like the diaphragm, I've found the balloons eventually stretch a bit and actually do wear out over time as the cylinder gets hot. If the piston doesn't work, an appropriate diaphragm for long-term use should probably be either bicycle inner tube or a truck air brake diaphragm.

As far as heat transfer, there is heat of evaporation, or perhaps it would be more accurate to say cooling, but at any rate, water, and liquids in general take in heat when evaporating/boiling and release that same heat during condensation, so is heat transfer to the condensation medium inevitable ?

I'm wondering if, or how much the conversion of heat to work intercepts or replaces such a heat transfer process, and if so, does the water condense on the beads or whatever, releasing heat to the medium which causes cooling/contraction of the air, or does the air cool as a result of expanding and doing work which then results in condensation as a consequence ?

I would say both, but ideally, the goal would be to have the gas condense as a consequence of doing work rather than the other way around.

I've been wondering about getting the water to boil as a result of a vacuum, at room temperature. As in this demonstration'

https://youtu.be/QQ3o3_sqfdo

Doesn't look all that difficult.

There are some others experimenting with trying to actually get power from this "cold steam" using various somewhat complicated setups to run Tesla turbines and such.

https://youtu.be/W1K2CMRyDrw

If the whole process could take place in a bottle that would be interesting.

I was thinking I might try fastening the lids of two canning jars together with the diaphragm sandwiched in between.

I was wondering about the crazy video of "shaking the building", it seemed that the engine really went nuts after taking it off the burner.

I thought that possibly the water that was added was doing that Leidenfrost effect thing so not really boiling effectively and needed to cool down a bit.

As far as the canning jars, the possibility of the glass breaking is certainly a concern. Canning jars are made to be pretty strong, heat and pressure resistant to a degree.

I'll be sure to take appropriate precautions, probably.

But, for what I have in mind, an implosion might be more likely than an explosion.

After boiling some water inside and sealing the jar, I thought I might try getting the water to boil by applying some ice.

https://youtu.be/eciqeeFpX0w

There were some papers I was reading related to enhancing condensation, though I can't give any specific references at the moment, I'd have to try and backtrack my browser history, but the gist of it is in regard using one sort of hydrophobic coating or another on the condensation medium.

Combined with some of the observations posted above, like the engine running best when some water drips back down to the bottom.

Condensation tends to stick to a surface, which, presumably tends to inhibit further condensation. The surface becomes saturated.

A hydrophobic coating allows the water to be shed so that it does not continue to take up and stick to the surface area but instead rolls off, dripping down to the bottom so it continues to cycle.

This is important in things like water desalination. The hydrophobic coating accelerates the rate of condensation, basically by getting the already condensed water out of the way, freeing up more surface area so condensation can continue.

Rice eventually soaks up it's fill of moisture until it can't hold any more. Any close matrix of glass beads or whatever can eventually become saturated by capillary adhesion or surface tension or whatever, the water clings to the glass becoming trapped.

So, it may be worth experimenting with silicone or whatever coatings on the condensation medium as a means of extending run time and avoiding saturation.

I don't really have any specific recommendation other than I recall something about silicone lubricant mentioned, some soap-like substances to reduce surface tension

Maybe PTFE coated glass beads, used for balancing tires.
https://www.napaonline.com/en/p/BK_7041440

There has been a couple interesting new videos posted to the "High Voltage Engineering" YouTube channel.

https://youtu.be/IEPiTa9XJAw

https://youtu.be/j0zxwase0Jc

Something I would like to point out though, this engine produces a lot of power with very little heat when you have everything tuned in correctly. At one point I started the engine with 3 tea lights, which overpowered it realy fast, so I had to run it with 2 tea lights and even then I had to be careful of how close I was holding the cans to the fire.

@chaorro, is your engine totally enclosed when it's running? I'm not sure how it would be "overpowered" that fast unless it could not handle the pressure, but maybe I'm misunderstanding. Relatedly, I've realized that keeping the engine completely sealed is a mistake.

Earlier I was talking about a Spunding valve to prevent an overload of pressure. I omitted having a pinhole in the side of the upper can because I thought it wasn't needed (technically it's not), but now I realize its importance. It mitigates catastrophic pressure buildup and it also allows the amplitude of the diaphragm movement to increase according to the heat input. When there's no hole, the heat intensity needs to be closely managed and the engine hits a limit of amplitude (or work it can perform). Counterintuitively, a teeny tiny pinhole allows for better performance.

I added the hole again to my engine and, lo and behold, the engine started running much more proportionally to the rate of heat input. Enough that some other pots and pans on my stove were shaking! It's gotta be a teeny tiny hole, though. The only reason I've had to remove it from the heat was to prevent my hand from getting burned from the ambient heat of whatever the source was. I picked up a smokeless fire pit the other day so I actually got the engine to run on it for about 20 minutes; it might have gone longer than that, but my diaphragm failed. This is a problem I keep having that forces me to abort these tests.

Speaking of diaphragms, I must emphasize how important the tautness is. One thing I've discovered is that if my engine hits a wall, I can retighten the diaphragm and then it immediately starts running again for quite a while. Whether this is because the diaphragm gets stretched, or the resonant frequency is shifted, or because I've let cool air into the top section, I am not exactly sure. In general, the tighter the diaphgram, the better the performance. At least that's what I've found. The problem with these balloons is they are very stretchy and can become looser because of the heat. The failure I had when using my fire pit was a result of me stretching the balloon diaphragm to such a point that it escaped the washers I have it bolted between. Almost every time I've hit a hard wall, it's because either something like that happens or it's because some other hole developed in the diaphragm.

I'm still not sure just how much condensation actually matters. Water condensing and dripping into the boiler does seem beneficial, but it might be secondary. The engine seems to get to a wall at the same time that the upper cylinder gets really hot, but that may be coincidental. Retightening the diaphragm seems to help get past some of the stalling. I've also re-evaluated cooling the upper cylinder, and now I think it probably does help. There seems to be a difference between evaporative cooling and air cooling. Like in my first video, I wrapped a moist paper towel around the cylinder and kept it moist with a spray bottle. When I've tried using a fan, that seems to easily take away too much work from the diaphragm. With the evaporative cooling from the moist towel, it could be that the rate of cooling is supply-driven, hence it's excess heat that's more likely to be carried away by that means. It doesn't necessarily make the engine run better, but just keeps it running. If I can figure out how to not get my diaphragm to encounter a problem, I want to see how long I can get it to run with cooling vs not cooling.

Oh, and another things about hitting the wall; I've noticed that really pumping up the heat can push the engine back into running again. This may have to do with the diaphragm loosening. If the diaphragm becomes too loose, it may take more heat input to put upward pressure on the diaphragm and get it to a tautness that it will resonate properly.

Based on this, I now think that the ideal way to set this up would be to have a metal base attached to a water resistance heat insulating tube to form a boiling reservoir. Then there should be a mesh over top of that, attached in a low thermal transfer way, and hydrophobic glass beads or something else to support condensation on top of that contained in a heat conducting metal tube.

That definitely seems more optimal.

I really need to get a needle-syringe and try the flash boil approach. It makes me wonder if the engine could benefit from a "carburetor" of sorts that sprays small amounts of water into the boiler as a result of the fluctuating pressure.

I had some modest success with a diaphragm made of parchment paper. Sounds just like a two-stroke!

https://www.youtube.com/watch?v=T_-j91E6CCQ