Large piston rings/sealing

[BTYork]

Hi all,
I'm new here so I'll give a brief intro, I got my masters in mechanical engineering working on Stirling engines, specifically doing flow analysis' through regenerators. I'm now working out of the same university lab prototyping an alpha engine. We have some funding for the prototyping, with the end goal being to bring the engine to market for rural, off grid communities, indigenous peoples, and adventurers. I'm making what I would consider a reasonably large engine; we're not looking to power full homes or anything, but the end goal is ~1kW powered off of wood burning at 600C.

The first version of the engine is specifically for testing. It is in no way economical or really robust quite frankly, but in the matter of a day I can have it completely disassembled, change the dead volume, regenerator mass, stroke length etc. It's also an ongoing experiment, so its loaded with sensing equipment.

Our Achilles heel thus far has been our piston rings. The engine was designed around a rebuild kit for a diesel tractor. The cylinders are 3.375" diameter, and we're usually running an oversquare stroke ratio. The iron split rings that are with the rebuild kit simply wont due, they really need oil and it just burns away at that temperature, not to mention foils the regen. Graphite may be an option, but I'm worried about its longevity. Our newest attempt is custom pistons with graphite infused PTFE rings, no split. These seal really well, but have a little more friction that I would like. On top of that, we now need to cool a portion of our hot cylinder to prevent them from melting which obviously hurts our efficiency.

Does anyone here have experience with larger pistons? I know looong pistons with no seals is relatively popular, is leakage not an issue with these? We dont have a pressurized crankcase right now, is this something that you think is necessary?

Im open to any feedback or advice of course, and happy to answer questions about the engine. This forum seems pretty awesome, its nice to see so many others that are as passionate about Stirlings as we are.

Kind regards,
BTYork

[staska]

Hello in our small forum. On st05g stirling engine piston rings are made from iglidur with slice - just as an ordinary piston rings + plus have some spring to energise them. Their bore is 75mm as far as i remember. But they are kind of precious one. Same are used in whispergen, but made from graphite ptfe.

Other way is w type cut out of in piston ring https://www.nmri.go.jp/oldpages/eng/khi ... /index.htm

Me, an lazy one being try to reuse cup seal ptfe from oil free air compressor. Bigger from ready made is for 69mm bore, so close to your need too.

But i am working on Alpha-gamma type stirling engine. Gamma like with thicker displacer rod - which gives correct geometry for our medium temperature. Like ~ 120-130 degree Alpha with 0.8 volume ratios. Google for it.. Similar task, just lower power in 250-400 watts range. But it have to be easy and nice to make :)

And drop me Pm if you want, will send some drawings. Always open to discuss this topic.

[BTYork]

Hey staska,
Those look great! It seems as though we're on the right track with the high temperature graphite/ptfe rings. Our experienced friction may just be the clearances involved. Our tolerances are super tight, and the rings like a 0.001" gap. I've designed the cylinder and piston undersized so they'll reach spec at operating temperature (200C), maybe I'll just need to heat it and see. Here's an image of a ring, for those that are curious.

IMG_6229.JPEG (174.79 KiB) Viewed 5458 times

[staska]

Could you share data of that graphite ptfe ? I think it is different than mine. So you have oring backed graphite ring ?

Tolerance or no - but if you do not give it any pressure - friction will eat all of you power. Or most of it. Learned that hard way. And for that 1 kw you need about 3-4 liters of swept volume. And truck air brake diaphragms as your cylinders. Now i am back to square One with my design. Was not able to make dual cylinder with acceptable tolerances and surface finish after anodising. But i do start all from case. Mine is here : https://stirlingengineforum.com/viewtop ... f=1&t=5223 But this kind of old data.

Have you geometric data ? I am running some matlab code for predictions of performance, as far as i did test on other stirlings +- on right side.

[BTYork]

Sorry, I don't have the material data sheet for the rings. They are Carbon finber filled teflon from https://www.hitechseals.com/index.asp?lang= and yes, they are oring backed. The oring is Viton, bringing the max temp for the seals down to 200C.

I do have all of the geometric data, I modeled the engine in excel to give rough initial estimates. The design is pretty modular, so the only real aspect we're locked into is the cylinder diameters, hot and cold, at 3.375"(85.725mm). Our stroke is currently set at 1.5" (38.1mm), and we have ~0.043L of dead volume. This wont bring us near our 1kW, more cylinders will need to be added in the future for that. We're just hoping to get something to turn to get us started.

For your matlab code, are you able to estimate the rpm of the engine?

[staska]

Sorry, I don't have the material data sheet for the rings. They are Carbon finber filled teflon from https://www.hitechseals.com/index.asp?lang= and yes, they are oring backed. The oring is Viton, bringing the max temp for the seals down to 200C.

I do have all of the geometric data, I modeled the engine in excel to give rough initial estimates. The design is pretty modular, so the only real aspect we're locked into is the cylinder diameters, hot and cold, at 3.375"(85.725mm). Our stroke is currently set at 1.5" (38.1mm), and we have ~0.043L of dead volume. This wont bring us near our 1kW, more cylinders will need to be added in the future for that. We're just hoping to get something to turn to get us started.

For your matlab code, are you able to estimate the rpm of the engine?

I could estimate rpm, power level. pumping loses etc.

[BTYork]

That's awesome, did you create this code? Is it available?

[staska]

That's awesome, did you create this code? Is it available?

I rewrite from old FORTRAN code found in some scholar papers. I can run numbers for you. I need following data :

Compression space, expansion space + dead volume of each of them.

On heater / cooler i need number of tubes / slots, length and cross dimesion.

For regenerator - housing id / od, length and filling type and density.

Working pressure, medium and heater cooler temperatures.

I am getting similar results as st05g / whispergen and some Andy Ross engines.

Or just buy Allan latest book
https://www.amazon.com/Stirling-Cycle-E ... 1118818431

- which leads to same criteria as found in real life working engines 120-130 alfa and 0.8 volume ratio. Or around 1.7 of compression ratio for our rural style stirling engine with heater at 500 degree celsius.

[BTYork]

I rewrite from old FORTRAN code found in some scholar papers. I can run numbers for you. I need following data :

Compression space, expansion space + dead volume of each of them.

On heater / cooler i need number of tubes / slots, length and cross dimesion.

For regenerator - housing id / od, length and filling type and density.

Working pressure, medium and heater cooler temperatures.

I am getting similar results as st05g / whispergen and some Andy Ross engines.

Or just buy Allan latest book
https://www.amazon.com/Stirling-Cycle-E ... 1118818431

- which leads to same criteria as found in real life working engines 120-130 alfa and 0.8 volume ratio. Or around 1.7 of compression ratio for our rural style stirling engine with heater at 500 degree celsius.

Unfortunately I have an NDA for some of the 'novel' design aspects of our engine so I cant share all of the necessary info. Thank you for pointing me towards some useful engines and data though. Its impressive to see the amount of work shared on this forum.

[staska]

Unfortunately I have an NDA for some of the 'novel' design aspects of our engine so I cant share all of the necessary info. Thank you for pointing me towards some useful engines and data though. Its impressive to see the amount of work shared on this forum.

Contribute yourself too on it. But was that NDA data tested or only idea ? There is a ton of data on stirling engine which could be classified as secret, but rarely shows any value. If it was not proven by more than three designers - it is too wrong to start with.

[BTYork]

Contribute yourself too on it. But was that NDA data tested or only idea ? There is a ton of data on stirling engine which could be classified as secret, but rarely shows any value. If it was not proven by more than three designers - it is too wrong to start with.

Its one of those things that has been proven from a thermodynamics perspective to improve efficiency, but validating it from a mechanical perspective is just easier done experimentally.

[Bumpkin]

Hey BTYork: For the overheating rings issue: (and you’ve probably already been there but if not,) look into heylandt ? (spelling) crowns - “hot caps” with which you add a thermally isolated extension to the hot cylinder to keep the heat away from the bearing parts.
Bumpkin

[BTYork]

Hey BTYork: For the overheating rings issue: (and you’ve probably already been there but if not,) look into heylandt ? (spelling) crowns - “hot caps” with which you add a thermally isolated extension to the hot cylinder to keep the heat away from the bearing parts.
Bumpkin

I've seen those on many engines, is there any particular reason for it being a dome? My initial guess what for easier manufacturing of thin walled pistons.

[Tom Booth]

.... no seals is relatively popular, is leakage not an issue with these? ...

I suppose, with an Alpha type engine, some type of flexure bearings would be out of the question?

There is also the old Clupet type steam engine piston rings. They are able to expand, but there is no through air gap.


https://youtu.be/80jzsNM6olE


I don't really see how a W type ring presents an advantage in terms of creating a better seal. It appears to create multiple effective "slits" where the ring contacts the actual cylinder wall, rather than just the one slit in a conventional piston ring gap (?)

I would try to engineer a "contactless" piston and cylinder using flexure bearings if at all possible in any way.

[Bumpkin]

About hot caps, I mistakenly said cylinder when I meant piston. Anyway it can be domed or flat as long as it matches the cylinder for least dead space, but if it is flat it would need to be heavier to handle the same pressure differences. It should be longer than the stroke and slightly less diameter than the bore so all wear and sealing is kept below the hottest zone. It should be as thin as possible for the best thermal break to the actual piston and it needs to be a contained hollow above the actual piston to keep direct heat away from the crankcase. To complete the effect, the cylinder should also have the best possible thermal break between the hot zone and the sealing/sliding surface. Taken all together I guess it can make quite a difference. Good luck, have fun.
Bumpkin