I'm planning to build a Stirling generator to work off heat from our kitchen stove.

[Tom Booth]

Right now though, I'm leaning more towards a dual rotor system on the same axle.
One main reason is that heating one side of a cylinder will warp it and that will affect the sealing by the vanes.
Another is separating a cold and hot rotor increases the delta t.

It allows for a little more design freedom to increase efficiency.

Well, it's an intriguing possibility. I like the directness and simplicity. If nobody else tries it I might, if I ever get the opportunity.

About the potential for warping due to temperature changes in the metal, I recently ran into information about some iron-nickel "Invar" alloys that don't change size with temperature. Used in such precision applications.

It would be interesting to see if the concept worked, if so it might be worth the investment of using such (more expensive) alloys, even if a prototype quickly warped and quit working.

I was thinking too, the idea of heating working fluid directly through a transparent engine has been discussed here.

Also looking forward to seeing your design. You seem to have dived right in to the deep side of the pool. When it comes to prototyping, or building larger power producing engines, I think I'll probably start out with something more conventional, like a scaled up version of one of these:

Resize_20231127_062215_5672.jpg (143.2 KiB) Viewed 27568 times

[Jack]

I like to jump into the deep end, it forces me to swim hehe.
I'm somewhat of a perfectionist (still don't know if that's a blessing or a curse though), which doesn't allow me to build something I know I could improve. But it also keeps me going in circles about this design.

Anyway, I went back to the drawing board and imposed the pv diagram onto a rotor design.
I used arbitrary sizes, I'm just looking for the needed shape of the cylinder to make the ideal volume change happen.

First the single rotor. It reveals a non-circle shape for a cylinder. (granted, this is a quick sketch that could use some smoothing)
There's still the matter of heat transfer. In a single rotor this is only the cylinder wall. Ideally the rotor and vanes would be as insulating as possible.

I also drew the double rotor shapes.
I'm more excited for the possibilities this opens up. For one the rotors and vanes can be part of the heat transfer, because they each have their own side of the engine.
The other possibilities are timing change and regenerator design.
In general it's much more flexible.

For both more vanes/chambers would mean a volume change closer to the ideal one.
I assumed clockwise rotation.

PV diagram rotors.jpg (52.46 KiB) Viewed 27560 times

[Jack]

More advantages of two rotors.

I can play with a difference in size between the hot and cold cylinder to get more surface area where needed.

Passing the air from one cylinder to the other creates some turbulence, which is good for mixing and heat transfer.
Haven't decided if I'm going for a port between the two cylinders or let the air move through the rotors with canals.

[Tom Booth]

So far I'm having some difficulty understanding how the two rotor design works. That is, what makes it go around?

[Jack]

The same way as in the single cylinder, but I see I've misrepresented the sizes a bit. It looks more apparent in the single rotor.

Between 1 and 2 is where most power is made.
Between 2 and 3 there's a constant volume, but it doesn't look like it because the fluid is moving from hot to cold.

[Jack]

Another thought.
If I eject the hot air from the hot rotor after 2, through a regenerator (sort of like a radiator type, with separated fluids) and into a bigger pool of cool fluid with more cooling capacity I gain efficiency. I can pick up cold fluid with the cold rotor on the other side of it.
1. I don't have to shuttle the fluid back to the cold cylinder, which takes some energy.
2. I don't heat up the cold rotor and cylinder with the hot fluid.

I'm not really sure about the physical effect of the air ejection on the vanes. My gut tells me I gain work output because of it, but I'm not very sure about it.

[Jack]

Triple rotor.jpg (138.67 KiB) Viewed 26823 times

I've come up with this. It's a triple rotor system. I'll try to explain hehe.
I don't like how a regenerator takes heat from the fluid right when it is creating the work and putting heat in when the work needs to be put in. Basically taking a bit of power out when we want most of it and adding a bit of "difficulty" when we want our energy input as low as possible.

What I thought of here is a hot, cold and regenerator rotor. They all move clockwise and are place in line of each other. All the same size. The colored areas are ports to the other rotor.
There are 8 chambers in each rotor, so the ports are always separated by one chamber to prevent cross flow when we don't want it.
So following the fluid from the biggest chamber in the cold rotor (bottom one) it gets compressed while being cooled. Eventually it combines with the fluid that's in the regenerator, basically mixing the temperature before being pushed out of the cold rotor completely and being divided in the regenerator and hot rotor.
Part of the fluid goes back into the cold rotor, putting out some work there and being cooled and expanded in the process.
The other part goes into the hot rotor and gets heated, putting out work, before going back into the regenerator where it repeats the whole process.

So in stead of trying to transfer heat into a metal regenerator, I use the hot air directly here. And divide it into two cycles.
The volume evolution is very close to that of a carnot cycle. I've plotted them in an excel file with 5 degree increments.

Triple rotor volume evolution.jpg (210 KiB) Viewed 26823 times

The overlapping part shows a bit strange in the graph.

I think this is worth trying to build and experiment with.
The biggest concern is friction as always. I'm hoping it will stay manageable with graphite vanes and properly honed cylinders.

Onto trying to solve the mechanical side of this.
Because the heating area is about 25cm wide, the hot rotor should be that long. This in turn will make the other two a similar length, which makes the whole assembly a bit long.
I'm trying to see if there's a way to fit the regenerator cylinder inside the hot rotor or any other solution to try and shorten the thing, but keeping it on one axle (I'd like to avoid gears for noise and power loss concerns). But then again I also like the idea of the fluid only having to move in two directions, in stead of a system of tubes or passages (which also adds dead space).

Did I explain this in a way you can understand my thoughts?
Any thoughts on this?

[Bumpkin]

“Did I explain this in a way you can understand my thoughts? Any thoughts on this?”

The first, sorta. The second, maybe; as in I hope my participation in the thread didn’t contribute to your delinquency towards other responsibilities, whilst you pursue the sick madness of hot-air engines. I don’t get and/or am not sure I agree with your regenerative middle of the three rotors process, but I sorta get where you’re going. If you get enough torque to beat friction, rpm is the only limit. Anyway, that’s a chickenshi% answer I’ll try to improve on later, I’m cheering you on.

Bumpkin

[Jack]

It's an addictive hobby I'm finding out. The Mrs has to sometimes remind me to take care of my other work as well haha.

[Tom Booth]

...
Any thoughts on this?

If I understand your diagrams and explanations;

The only thing I notice that IMO might present a problem is that in my general experimentation with Stirling engines, after the expansion/power output, the working fluid has lost energy and tends to "contract", regardless of being externally cooled or not.

It appears, though, that you make a transfer from the hot to the cold side and then continue to expand the working fluid

Resize_20231206_164932_2176.jpg (53.75 KiB) Viewed 26711 times

The air passes into the cold side at the top right through the port that looks like a pink sliver?

The gas then expands as the cold chamber grows until reaching the bottom, where the chamber is as big as it gets, then starts shrinking again

So, if that is the case, between the port and the full enlargement of the chamber the engine is attempting to expand working fluid that has already done its work and should already be cooling and contracting as a result of the energy loss from work output.

Now, maybe there is still some energy in the gas, so that it will continue to expand and output work though it has already been transfered over to the cold side?

On the other hand, it is true that in general, the cooling of the working fluid takes place 90° before BDC while the gas is still expanding in the power cylinder, still, this appears to be more than 90° in advance of full expansion at BDC (not that there is a "bottom dead center in a rotary engine, but the functional equivalent).

I'm thinking that the dynamics of a rotary engine are different than a reciprocating engine so that likely the timing should be less advanced rather than more.

At any rate, if the gas is already cooling from work output and the engine is still expanding the gas, it could be working against a vacuum condition being created by the contracting gas.

I'm undecided as to the possibility of any benefit to be derived from this apparent excessive(?) Expansion.

Some means of adjusting the timing on the fly to find the best advance position would of course be ideal

[Jack]

My theory about this is a bit different.
The system is thought out to be pressurized. Let's assume a random 5 bar.
The biggest chamber at the bottom of the cold rotor would be where that base pressure is. Moving up the fluid gets compressed, while cooled, to about double, so 10 bar. Ideally here it would go into the hot cylinder directly where the heat brings it further up, let's use 15bar peak as a hopeful number. This is where one part of the work output is done.
From there it goes into the "regenerator" cylinder which brings the temperature down a bit while maintaining an equal volume for some degrees rotation before it shrinks down in volume. I'm thinking there might even be a slight work output for the first bit of the transfer to the regenerator cylinder, as it's still being heated up in the hot cylinder.
Only then is it transferred into the cold cylinder, still hotter and with higher pressure, ideally a little over 10 bar. The cold cylinder decompresses and cools the fluid back down while hopefully getting some work out of that too.

Timing in the normal sense in this concept is of less importance as there are 8 cycles evolving in each rotation, and two separate work output "locations".
It's a very different way of looking at a regenerator. In stead of trying to heat up some metal and picking that up again with the next reversal of air movement, this uses the heat to power a different cylinder.
The timing of when to move the air from one cylinder to the next is more important. And I do agree that it would be good if that could be made adjustable for the prototype.

[Tom Booth]

Another thing that concerns me, again, if I follow what's going on:

Looking at it from the cold side again,

Resize_20231206_164932_2176.jpg (53.75 KiB) Viewed 26599 times

The "wedge" of air at the top of the rotor appears to be between two ports.

At the upper left greenish port, air is pushed out through the regenerator over to the hot side and at the upper right pinkish colored port the gas is brought back through the regenerator to the cold side. Is that correct?

The segment or wedge at the top center location is apparently touching, or nearly touching both ports simultaneously, or almost simultaneously.

Or, the instant the cold to hot port (top left) fully closes the hot to cold (top right) starts to open.

From 1 to 2 then, on the PV diagram, the effective power/expansion leg, there is virtually no time where there is not an open port between the hot and cold sides.

Is there a separate power piston (not shown) attached somewhere in some way, so that what is shown is only a kind of rotary displacer system?

[Jack]

No all ports are separated this way by at least one segment. But this particular segment has a negligible volume as the rotor here is very close to the cylinder wall. I didn't explain this, but the small circle inside the bigger one is the rotor. The chambers are the spaces between the small and big circle.

There is no other piston attached to this. This is it. Three rotors.

The green port is basically a port to equalize the hot and regenerator rotor and push the last bit of fluid from the cold rotor in. Then is closes for a split second after which only a port between the regenerator and cold side is open. This allows the regenerator rotor to empty into the cold rotor.

[Tom Booth]

What would the "regenerator" consist of?

It appears (upper right) to reduce to zero volume.

[Fool]

Jack, are you striving for a first ever pseudo-Carnot cycle?That's what it looks like. Excellent.