A New Type of Hot Air Engine

[matt brown]

I always thought 'steel' wool from copper was a good place to start due to high conductivity. Porosity of matrix will (may) figure into dead volume, but can be easily verified by testing samples in container with vs without water. I favor a 'perfect' seal to separate heat short, and a moving matrix which forces the gas thru it mechanically rather then drawing it thru via pressure differential. There's many ways to game this basic scheme and 2 stroke ICE is ripe for conversion. Over the decades, this has become my favorite ECE chase, but you'll discover that there's a lot more going on within this 'simple' scheme than is casually apparent. A typical thermal lag scheme solves the mass volume problem by being unregulated which partially drains down the matrix and its hot chamber during expansion, but a moving matrix reduces the hot chamber volume more. As you get into this further, you'll discover all types of possibilities for gaming this, and I remain clueless why most SE guys (especially DIY) are chasing this.

All common compression cycles will require 'cooling' (and adhere to Carnot) but cooling can be open cycle ambient or closed cycle to cold reservoir. The key thing with this scheme is to be wary of hot chamber reservoir issues (Pmax at the wrong place & time). Also, Tom's gif is counterflow and I favor a uniflow design (conduit w valves between chambers). I usually scheme in conv'l pistons, cylinders, and slider-cranks, so no biggie thinking valves, cams and conduits. What's good enough for ICE is good enough for ECE.

[VincentG]

With such a simple operating theory, are there no running examples out there?

[matt brown]

I've never seen any and always got dissed when suggesting such. This doesn't make any sense, since even the gurus drive around in simple/similar ICE. So, I mainly kept it to myself while perfecting the stuff everyone expects. Just mention valve to any SE guy...

[matt brown]

...and I remain clueless why most SE guys (especially DIY) are chasing this.

Oops, correction...I remain clueless why most SE guys are NOT chasing this.

[VincentG]

Glad you made that correction, it had me a bit confused. I think a valve solves the stirling's main issue of temp mixing. Shouldn't even be a question after any analysis at all.

[Tom Booth]

I should be able to prototype that hot potato fairly quickly. What was you idea for the regenerator material? And I imagine the valve was meant to be a perfect seal? Or was it just a flow restriction? Also how much importance do you place on the regenerator moving with the valve vs being stationary in the hot chamber?

Obviously, (I hope) I gag when Matt Brown says things like "All common compression cycles will require 'cooling' (and adhere to Carnot)" and yet latches on to this "Hot Potato" engine idea as if it were his own now. Very strange IMO.

Anyway, the basic concept is on that thread. "Cooling" as described there is a consequence of expansion/work output, as covered exhaustively on the other thermodynamics thread also mentioned there.

The "Carnot efficiency limit" is complete fiction. I've been on several science/physics forums and have also asked Matt here, how, where, when and by whom was this supposed "limit" equation empirically established. It seems nobody has an answer. Fact is, it never was experimentally established in any way shape or form. Experimentally, it doesn't hold water.

3D printing (plastic) would not handle the heat, I don't think. I'd probably use ceramic for most of the engine body to retain heat as much as possible.

A ring of conductive material around the "regenerator" to transfer heat into the engine rapidly.

Traditionally stainless steel wool for the regenerator holds heat just well enough but also gives it up readily to a flow of air.

Probably an air spring would do as well as an actual spring.

Also this has evolved into the Ringbom version as air pressure serves as well as a bash rod to move the regenerator/displacer/"potato"/heat accumulator whatever you want to call the wad of steel wool.

To reduce dead air space it would likely help to recess the spring so that when compressed there is no air (or as little air as possible) left in the hot bulb.

Maybe something like this:

Polish_20230319_005719525.jpg (564.43 KiB) Viewed 2470 times

The port, IMO would serve no purpose other than keeping the "hot potato" centered on the metal ring. The piston could just as well push it directly I think.

The main idea is just to accumulate heat in a metal matrix until pushed by the piston causing air (or whatever working fluid) to flow through the matrix releasing the stored heat. This will expand the air and drive the piston back out cooling the air by expansion and power output.

There is only heat input at the "hot potato" matrix. No heat output to any "sink" anywhere.

The piston should likely be non-heat conducting as well.

I've shown the end as a separate "plug" but it could all just as well be one piece.

The matrix, or "hot potato" to function as intended should probably move, if only slightly, in order to force rapid air flow to release the heat.

Feel free to experiment with different materials or whatever.

Probably the whole thing could be made from a glass tube or if it is not heated too hot, some PVC pipe, except for the metal ring section .

I'm thinking, if possible, the "ring" should only be exposed where it will be in contact with a flame or heating source.

Of course there needs to be contact between the ring and matrix, but it also needs to be free to slide inside the tube.

A valve or port IMO is not necessary because heat travels from hot to cold. At expansion the metal matrix will be cold and absorb heat from the surrounding metal ring. In theory it should mostly hold onto the heat until jolted by the piston and forced to release it due to the sudden air flow through it.

It's a little hard to read but that should say NON- heat conducting ceramic, for the engine body. The "non" is not showing up well in the photo due to the lighting.

Basically ALL the available air would be forced into the heat matrix. So porting the air through valves and such seems unnecessary, IMO.

Simplify rather than complicate, as Matt seems to love to do: viewtopic.php?f=1&t=5425&p=17544#p17544

[Tom Booth]

In that link above I posted one of this guy's videos. Here is another:

https://youtu.be/cgz3pRk_ZHE

About as simple as you can get.

I'm not in agreement however as far as the explanation provided for how this simple engine actually works:

This engine is a kind of free piston Stirling engine which consists of glass tube, injector, and stainless wool. The mechanism is very simple. As the stainless wool move to the hot side, working gas is moved to cold side, then rejects heat to heat sink and compression. Similarly, as the stainless wool move to the cold side, working gas is moved to hot side, then absorbs heat from heat source and expansion and repeats this cycle.

Maybe, to a degree, but I'm betting that a cold side is actually unnecessary, cooling being the result of expansion + work output rather than "As the stainless wool move to the hot side, working gas is moved to cold side, then rejects heat to heat sink and compression". I'm not sure that even makes sense, but perhaps there is a language/translation issue.

[matt brown]

The only downside to ECE Otto (and Atkinson) vs Stirling is that they can't use as low of deltaT as Stirling, but with wood or solar, who cares. I've read soooo many parametric studies that it's as if the grantmeisters don't want any solution, just endless funding. ALL the major engines in history came about without any academic help AND despite a lot of academic naysayers. The thing to remember with ECE Otto is that it's basically a gas spring with some heat added between compression & expansion. I don't worry about friction, cause it never kept the ICE camp up at night, and friction increases with speed...so, as an ICE runs faster & faster, each stroke has more friction...more Wneg per stroke from a paltry Wpos per stroke. The only reason ICE dominant the world is due to crazy high cycle rate (rpm) but I'll trade that any day for the ability to pressurize.

I also tend towards argon as a working gas in prototypes. It may be 1/3 less conductive than 'air' but it doesn't have any moisture issues (water & oxygen have a thing for each other or the fish would die). Furthermore, argon is monatomic with a lower thermal capacity for adiabatic processes, so it's P&T rise & fall faster than diatomic air for same deltaV. Per that adi index I posted, a diatomic (ideal) gas with 6:1 compression will rise T about 2x and P about 12x; whereas a monatomic gas with 3:1 compression will rise T about 2x and P about 6x. This is definitely where basic thermo is handy, since none of this is linear like isothermal stuff (these examples just happen to look linear). The point here is comparing PVT values between diatomic & monatomic gases for the same compression ratio, diatomic are more powerful (due to higher P) but monatomic are more efficient (due to higher T). So...a monatomic gas allows more efficiency from a lower Pmax within the same deltaV constraint.

Another prototype gimmick is using a belt driven dyno-motor input to spin up engine whereupon engine outputs to dyna-motor once producing power. Yep, lots of different ways to game this and where is everyone ???

[matt brown]

The "Carnot efficiency limit" is complete fiction. I've been on several science/physics forums and have also asked Matt here, how, where, when and by whom was this supposed "limit" equation empirically established. It seems nobody has an answer. Fact is, it never was experimentally established in any way shape or form. Experimentally, it doesn't hold water.

There is no Carnot whatever for a single process...like taking energy from the wind or from moving or falling water. This Old World linear experience of energy is what led to the caloric theory.

The Carnot buzz only becomes an issue when you have a cycle, when you're adding energy (work and/or heat) to a gas BEFORE removing energy (work and/or heat). This is easiest seen with the common compression cycles, but Carnot wins with most irregular cycles. Take the famous 3 legged Lenoir with isochoric heating, adiabatic expansion, and ambient exhaust. The ambient exhaust will count as isobaric 'cooling' but even without any engine compression (no Wneg) this cycle looses big time to Carnot. Why ??? Simply take the massive isochoric heat input and divide by the paltry adiabatic engine output, and Carnot keeps his crown. PV plots are a good reference for work, but there's no trip odometer showing heat & work, in & out, per process as you trace the perimeter of any PV plot. Nope, calcs are req'd, and why engineers go to school. Heck, you wouldn't want a brain surgeon learning on the job, would ya ???

So, riddle me this...why does a 10:1 compression ICE get better 'mileage' than a 6:1 compression ICE ???

[Tom Booth]

The nearly completed hot end, save for the bash valve bumper and possibly the reed valve modification to the bash valve face.
assembled hot end.jpg

I think at larger scale, all of this could be cast from 5000psi concrete, or perhaps a more insulative aircrete mixture.

That's what I call rapid prototyping VincentG. You have an idea today, tomorrow it's already done. I wish I had your skills. I could debate theory with Matt here for months,... Years... and not get as far as you do in a day. Nice work. Your contributions here are greatly appreciated.

[Tom Booth]

I'll propose one more example. A piece of unburnt wood is like a gallon of diesel fuel. Just a store of potential energy. The diesel can power a 1500hp pulling tractor but it is nothing without first putting energy into starting the process.

In this way, although the thermal efficiency of the diesel engine is only 35%, in effect its total efficiency is nearly infinite based on the input energy required to start the process.

Perhaps the hot air engine can behave in a similar manner with a well designed system.

If I understand what your suggesting here, as well as the previous post, the "dippy bird" will theoretically run forever, as long as it has water available, but it still requires someone holding its head down in the water to get the process started.

There is a long list of inventions/inventors who seem to have at least had some kind of apparently working prototype of some such thing or other here:

https://www.aircaraccess.com/achf-intro.html

I've actually encountered people on the science forums who apparently do not believe the drinking bird, (or a heat engine running on evaporative cooling) are real or possible, because they would be operating within a single "heat reservoir", which would be "a violation of the second law of thermodynamics".

[Tom Booth]

I should be able to prototype that hot potato fairly quickly. What was you idea for the regenerator material? And I imagine the valve was meant to be a perfect seal? Or was it just a flow restriction? Also how much importance do you place on the regenerator moving with the valve vs being stationary in the hot chamber?

On second thought, possibly this COULD operate as a low temperature engine (relatively low) That is, 3D printed with some relatively high temperature plastic.

The heat holding matrix, if it were all flattened out, would, or could have as much actual surface area as an LTD engine.

Originally, back in, when was it? 2010? I think I imagined the regenerator as somewhat more rigid than simple steel wool, but that would probably do for testing

And I imagine the valve was meant to be a perfect seal?

Matt has provided his response, mine however would be Absolutely NOT!. IMO there should be more than sufficient "thermal lag".

The gas expands and cools due to work output, which also cools the heat matrix, allowing the matrix to rapidly absorb heat for the next compression stroke.

The gas is "spread out" or expanded when cold. The matrix will take time to reheat, by the time the matrix absorbs some heat the gas has already been re-compressed.

I see no necessity whatsoever for any "perfect seal" ports or valves of any kind. Not due to any prejudice against such, I just don't consider it necessary. The port was not intended as any flow restriction, the drawing or animation was just based on existing laminar flow or thermoacoustic type engine designs.

There was a kind of hard plate on the steel wool to provide a hard surface for the rod to push against, but that was not intended as a valve or seal. IMO air needs to be able to flow in and out of the matrix as freely as possible.

The engine in that video (with what looks like a simple string plunger in a test tube) works, if a little jittery, without any seal or restrictions of any kind. A spring of some sort and the metal ring would just give better control over the timing and heat input for a smoother running and higher compression/higher power engine.

Could the matrix be stationary?

My original idea was the bump from the piston, moving the "hot potato" would effect the rapid release of heat.

Possibly just compressing gas into the matrix and the gas then expanding out through the matrix again would be sufficient air flow. A stationary matrix might be well worth experimenting with.

Maybe a compressible matrix would eliminate even more dead air space and give a higher compression ratio., Like a matrix made of spring steel wool or some Nitinol type super-elastic mesh. A normal stainless steel wool pot scrubber is probably springy enough for testing though

That is, instead of moving the whole wad of mesh it could just be squeezed a bit in place, serving as its own "spring".

[Tom Booth]

With such a simple operating theory, are there no running examples out there?

Personally, I think it just takes some time and trouble, as well as money to prototype anything, so without a theory of operation that anyone considers viable, nobody has bothered.

Even now, Matt, though earlier stating that perhaps I'm right and Carnot efficiency is bogus, still argues for changes and modification that include "cooling".

Your ability and willingness to dash off prototypes can certainly make for more progress than endless theorizing and debate.

I think previously, though, even considering any hot air engine design that does not incorporate some active "heat sink" cooling mechanism for removing the inevitable "excess" heat would be thrown out as a priori "impossible". "Don't waste your time" is about all I've heard for the past decade.

[VincentG]

I could write an essay in response to everything here, but I'll keep it brief. Thanks for the kind words Tom, it certainly keeps me enthusiastic about the project. I am "making" time to work on this engine which has become a bit of an obsession at this point. I took the day to think about a new plan of attack.

Matt seems to constantly contradict himself back and forth, which is frustrating, but it keeps me thinking about new ideas. I now realize that having a perfect seal, in the hot potato or the valve in my design is not ideal and would cause the mass volume issue mentioned above. It seems just blocking most of the flow is most important, and the leak by will keep the chambers balanced.

I now believe fully that a cooling system is unnecessary and detrimental to the greater cooling offered by load and expanding the gas at BDC.

Tom's post in his perpetual ideas thread about the heating vanes being exposed seems the best way forward. I had abandoned the idea due to complexity and previously thinking that "cooling" vanes need to be introduced as well. With only heat input now needed I think there is a simple solution available. This eliminates the "reservoir scheme" issues too, which I now realize could really pose a problem. I think it can also be easily combined with a secondary compression cycle and a dense but porous matrix like an extruded brass filter element, or a simple OD lathe turned heat exchanger.

Finally, I am attached to keeping the Stirling cycle with Otto like heat application and a hybrid open/closed cycle. It seems silly to throw out a whole power stroke, when I don't see the ability to double the hot stroke power with a pure Otto cycle.

[KristofB1982]

Hello all, I am new on this forum and I like al these interesting info. I believe also that the Stirling Engine can be a solution for the future.