Engine Pressurization

[stephenz]

Hi guys,

New here. I have a question about engine operation under "greater than atmospheric" pressure.

Let me clarify that I am currently not building anything, but still I'd like to understand practicalities and realities of an engine running pressurized.


1. First, assuming say an "alpha" configuration with the crank open to the ambient air, pressurizing the engine would mean pressurizing the volume of gas between the 2 pistons. And considering pistons can't be hermetically sealed, it means that ultimately the engine would lose its pressurization. Correct?

2. Now, let's assume that the same engine is now hypothetically fully enclosed in an hermetically sealed enclosure. In other words, now we have the volume in front of the pistons (effective engine volume), and the volume in the crank, behind the pistons. Both of these volumes are assumed sealed from the environment, but still not perfectly sealed relative to one another, since it's impossible for piston to create a perfectly air tight seal.

Here's my first set of questions:
- which of the 2 scenarios above is the desirable scenario? When you read about Stirling Engine optimization, they say pressurizing the engine is desirable. Is it desirable from the better physical properties of the compressed gas? or is desirable from the Pressure delta across the front and back of the piston?
- In the scenario 1 above, only pressurizing the engine volume, automatically create a force constantly "pushing back" on the pistons, right? is that desirable?


Finally, pushing this further:
- Ideally we want to minimize friction between piston and cylinder.
- Ideally we would also want an air-tight seal between piston and cylinder.

Say, you manage to achieve a pretty good seal between piston and cylinder. It means that pressurizing the engine volume would not immediately pressurize the volume behind the piston/crank. In this case would it make sense to have some kind of valve to allow the volumes in front and back of the piston to equalize quickly before operating the engine?


I love the simplicity of the Stirling cycle, but I also find the complexity of the practicalities of optimizing these engines absolutely fascinating.

If anyone has any recommendation for document/books talking about pressurization and hermetically sealed engines please let me know!

Thanks

[Bumpkin]

Well, everything about a Stirling engine is an art in compromise, so nobody’s an expert, but I’ll submit my two cents. Unless you’re running a focused high temperature, the relative size of the burner/exchangers will dwarf the actual engine, so I have to wonder at the importance of specific displacement performance. But certainly pressurization has benefits. I know you’re talking about Alphas, but here’s a way to pressurize a Gamma without a pressurized crankcase, the forum’s linked here to this for years. http://www.starspin.com/stirlings/jimd6.html Even if this isn’t what you’re interested in, understanding this engine might answer some of your questions.

Bumpkin

[stephenz]

Thanks for your response.

From a thermodynamics standpoint the higher gas pressure is obvious.
However, if the crankcase isn't pressurized, pressurizing cylinders isn't as an obvious benefit. The piston(s) will have to deal with greater pressure gradients, thus having to worker much harder in direction (and much less in the other). Since pressurizing the crankcase isn't that much of a challenge IMO, I haven't really given this much thought. But conceptually if I am to ever proceed with building something I would probably try hard to get a good piston to cylinder seal, and as such I was wondering how I would go about pressurizing the crankcase; since I heard of those snifter valves I thought I'd bring this up.

I've considered the gamma configuration which I really like in terms of layout.

[VincentG]

Stephenz, maybe I can help you understand this a bit as I try to learn more myself. To understand the benefits of pressurizing the working gas, reference the PVT plot of ideal gasses and play around with a guy-lussacs law calculator. There is more to it than the pressure gradient from either side of the piston. (edit, I missed you first statement so forgive me if this is obvious to you).

While pressurizing the crank case seems like a step in the right direction, and it may help with keeping the working gas pressure up in the cylinders, all it really does is increase pumping losses. As one piston rises the other falls so any pressure gain is counteracted. Ideally you would have a vacuum in the crankcase to eliminate pumping losses all together.

There may well be gains to be had here but at a large expense of complexity in an already complicated system.

[stephenz]

Yes, this part is really difficult for me to visualize and it's even more difficult to think objectively with a lot of people claiming one over the other, with example of working systems: fully pressurized systems, partially pressurized or not pressurized at all.

To me, the ultimate realities are practical ones:
- there are better gas than air
- fully sealing a moving piston is impossible without having to recharge
- sealing the crankcase is entirely possible

So, if someone wants to create a commercial product out of it:
- either they use air in a non pressurized system
- or they pressurize the entire system, including the crank

Going back to effect of a pressurized engine but non pressurized crank, I think the engine configuration matters a lot. In my original post/question I talk about an "alpha" configuration which technically has 2 pistons with backs exposed to environment, and since work is done only one of them, intellectually I don't see how that could work with more than just a bit more than atmospheric pressure. Maybe in a "beta" configuration things are different since only piston's back is exposed to atmospheric pressure, and while it may be "harder" in part of the cycle, it would "help" in the other.



It's in french about 3:30 he explains how the engine doesn't like more than 0.8 bar above atm.
https://www.youtube.com/watch?v=a5aket-HSYI

[VincentG]

Hmm..if you mean that only one alpha piston does work I would argue against that. For me the alpha cycle is the hardest to fully understand.

With that said, I'm in over my pay grade here. Maybe Matt Brown has something to add.

[stephenz]

I'm in the same boat, I don't know if my understanding is correct.

In an alpha configuration both pistons have a tight fit with their respective cylinders, so during expansion, both of them do work. But considering they're 90 degrees apart, there is still much more work done by the cooler piston than the heater piston (the heater piston is close to TDC). Which to me is why gamma/beta configurations are superior since the pressure on either side of the displacer piston is roughly equal thanks to the loose fit between the displacer and its cylinder.

To me it's much easier to visualize what's going on during expansion (pressure in front of the piston is greater than back due gas being heated just before that) than during compression. The word "compression" systematically makes me thing the gas is getting compressed externally, when in fact it's only getting compressed due to the gas being cooled and the pressure in front of the piston is now lower than the pressure behind the piston effectively driving the piston up, effectively compressing the gas.

And that to me is the reason, "just pressurizing the inside of the engine" has obvious limitations. I.E. it directly affects the forces seen by the pistons. And to me that limitation will be greater on an "alpha" configuration than the other 2.

[VincentG]

I tend to think of it as more of a shuttling of the gas from hot side to cold side, with compression and expansion being secondary but complimentary to the cycle. The forces that drive the engine come from the transfer of gas from one side to the other. Compression and expansion alone will not run the engine.

If you take the pressures to the extreme for the sake of visual aid, let's assume 100 bar in cylinder and 1 bar in the crank case.

An alpha with a 90 degree crank will spin by hand nearly the same as if it had 1 bar in the cylinder.

The dwell of the pistons near tdc and bdc is what allows the gas to do work on the other descending or ascending piston.

Now I would argue that the gas does more work when there is a restriction of flow from one cylinder to the other. This restriction slows the shuttling of gas so more time is spent gaining or losing temperature.

To understand this, imagine a hypothetical transfer port with unlimited flow rate (but assume no additional dead volume). I think this engine would make less power as the gas moves unrestricted towards the dwelling piston.

The point I'm trying to make, and I hope I'm not adding confusion, is that unlike a beta type engine, pressure on the back side of the piston is not required for the cold power stroke.

I might be off here but this is how I see the alpha cycle.

I will also add that there are some studies available that show the alpha type is quite a bit less efficient without advanced heat exchangers.

[stephenz]

Yeah, I don't see it that way. Taking your example of 100 bar inside the engine and 1 bar inside the crank, I think it would nearly impossible to move the flywheel by hand for the engine pressurized to 100 bar. The position of the crank is easy to identify and that is the position where both pistons are at the lowest position they can be, in other words the position that maximize the volume inside the engine, see attachment.

lowest-position.png

The reason why I think it would be nearly impossible to move by hand is the compression ratio.
Whether you pressurize at 1 bar or 100 bar, the compression ration is the same, 2:1 or so.

The pressure inside the engine makes the pistons want to be in the position shown above. Assuming you can rotate the flywheel close to half a turn you'll be nearing 200 bar in one case, and 2 bar in the other.

[VincentG]

My fault I never added to assume the engine had an insignificant compression ratio(similar to an ltd engine). I was just trying to visualize the gains from raising in cylinder pressure alone.

But im glad you called me out. Because, now let's pressurize the crankcase to the same 100bar and assume the standard 2:1 compression ratio. Is that going to counter act the added pressure from in cylinder compression?

I would think no, unless the crankcase side of the system has the same 2:1 compression ratio, and in that case wouldn't we be better off making a double acting piston engine?

To be clear I'm not being sarcastic here. I usually find bouncing around thoughts like this is the fastest way to understand some things.

[matt brown]

An alpha with a 90 degree crank will spin by hand nearly the same as if it had 1 bar in the cylinder.

No Vincent, the greater the charge pressure the harder an alpha is to turn through 360 deg, so stephenz is correct. The volumetric phasing diagram that stephenz posted shows hot & cold space volumes per relative crank angle, but lacks pressure data. An alpha will have a greater pressure swing than a beta or gamma (everything else equal) but any alpha will have a 'short' period when it is very easy to turn by hand (regardless of charge pressure) due to 2 pistons. This length of this 'short' period will depend upon phasing, but is greatest (per any given phasing) when both pistons have the same geometry and the Schmidt assumption that pressure is uniform throughout the cycle.

[matt brown]

Pressurizing the crankcase has to be carefully considered, and I tend towards Vincent's idea/l with crankcase vacuum.

The most practical crankcase pressurization involves multiple cylinders. Example: consider typical 2 cylinder alpha with single-acting pistons, and crankcase pressurization is not practical due to phasing. However, if engine is doubled with 2 single-acting compression pistons phased 180 deg in one crankcase, and 2 single-acting expansion pistons phased 180 deg in another crankcase, then each crankcase will have a fairly constant volume (ok, some 'flutter') and design can proceed with whatever phasing you want within each piston pair AND whatever piston geometry you want between each piston pair...both compression pistons are identical, and both expansion pistons are identical, but compression AND expansion pistons need not be identical. Overall, very similar to 2 double-acting pistons, just single-acting here.

[VincentG]

Now thats a neat idea Matt but one hell of a crank case.

[stephenz]

The crankcase has a much greater volume than the engine, so the pressure in the crank should remain close the nominal pressure the system was pressurized to. I haven't given it a lot of thoughts but I'm pressure if the crank volume was close to that of the engine volume it would affect the engine negatively. I.E. engine pressure rising would result in crank pressure falling, so as gas is expanding, the pressure in the crank would be rising opposing to the work done in the engine.


Most stirling engine are designed around that 2:1 compression ratio, from what I read. The higher the dead volume, the lower the compression ratio obviously. In the case of the Alpha configuration, and assuming identical pistons (diameter and stroke) and a 90 degree phase shift, I remember calculating a maximum theoretical compression ratio of close to 3:1. In practice the heater, cooler and regenerator volume aren't zero.


edit: just saw Matt's posts. thanks

[matt brown]

forced work PV plot.png (31.14 KiB) Viewed 8312 times

I found this PV recently and it reminded me of the Hall Engine discussion months ago (stephenz, search this site and you should find the link). Over the years, I've seen this PV for common partial vacuum flame licker and note that this PV could also represent pressurized crankcase. The text with this PV was eluding to the 'fact' that by adjusting the back pressure (raising the ambient pressure '0' on this PV) that the red backwork was less than typical PV (with ambient pressure at bottom edge of work area). This might excite some as a free lunch, but if we take this further by raising ambient pressure ('0' here) higher to the top of the work area, we find that the red backwork area is solely on the upper right (and larger) but that the blue positive work area remains.

There's definitely something work with this 'logic' for if we consider the following PV with ambient pressure at pts 2-4, then there's no similar red backwork area. I've been working on a response to this PV for weeks, but short on time these days.

sq cycle.jpg (58.59 KiB) Viewed 8312 times