HELP! Alpha Stirling Engine Not Running

[NKTL9414]

Hi All,

I built an alpha-type Stirling engine in a V-shape cylinder configure as a class project with the following specs:
Bore: 1.125''
Stroke: 2''
Compression: around 2.3
Crank case pressure: 1 bar
Working fluid: air
Hot side temperature: 450F
Cold side temperature: 125F
Crank offset: 90 degree

When the hot piston is at the top dead center, there is a roughly 0.5'' gap between the top surface of the piston and the ceiling of the cylinder. Same is true for the cold side.

These parameter are pretty much set in place already since I have already machined all the part and assembled the engine. Some of the parameters are not as ideal, leading to the problem I have now:

I'm having trouble starting the engine.

From checking each moving part, everything seems to run smoothly and with very low friction. Piston and cylinder create good seal - when they crank rods are not connected, sliding one piston into the cylinder will cause the other piston to move out of the piston. However, the engine is very difficult to turn when fully assembled. The only reason for this, from my view, is the high compression that the engine needs to overcome. The current flywheel-free assembly cannot push the crank over the compression, as some of the posts here on the forum suggest.

I therefore plan to add this:
https://www.grainger.com/product/SPEEDA ... 3?opr=ILOF
huge (10''OD, ~9lb) air compressor flywheel to the crank shaft.

I wonder if anyone has any suggestion on what to do. Or if my decision of adding this flywheel seems reasonable. Also let me know if there are any questions!

Many thanks!

[Tom Booth]

Why is your "cold side" 125°F ?

Various types of Stirling engines will (sometimes) be able to run "free piston" or without a flywheel. This may not be true of Alpha type engines.

I've never known any Alpha type Stirling to run without a flywheel, but I don't know if anyone ever tried before.

Including or adding a flywheel seems like a reasonable thing to do, or try, though that monstrous 10" flywheel seems like overkill for an engine with a 1.125" bore.

IMO generally, for a Stirling engine to run without a flywheel it needs to be VERY efficient. Opinions differ, but for various reasons I've never considered Alpha Stirlings to be highly efficient. Inherently, it seems to me that the pistons tend to work against each other, at least through part of the cycle, which would likely necessitate the use of a flywheel.

[VincentG]

Pictures of the engine would help. If your compression ratio is really 2.3:1, then your thermal ratio is far too low considering conventional thinking. Larger diy Alphas are notoriously hard to get running.

Did you model an existing engine? What is your cylinder angle? Do the rods share a common crank pin?

[NKTL9414]

Pictures of the engine would help. If your compression ratio is really 2.3:1, then your thermal ratio is far too low considering conventional thinking. Larger diy Alphas are notoriously hard to get running.

Did you model an existing engine? What is your cylinder angle? Do the rods share a common crank pin?

Thanks for the reply!

The two cylinders are positioned 90 degrees apart, so cylinder angle is 90 degrees. Crank rod share a common crank pin. The first attached image is the engine configuration, and the second is the assembled engine. I'm not sure if I calculated the compression ratio correctly. I calculated it by:

(smallest volume enclosed by the two pistons)/(largest volume enclosed by the two pistons)

Note that the two volume shown above do not occur when pistons are at their top end/bottom center. In CAD, the ratio is 2.3:1. In reality I think the ratio is probably lower, but should be larger than 1.5.

Let's say if the compression is really that high, what kind of temperature ratio should I achieve for it to run? The cold side is air cooled currently so it's not as cold as it could be. I can also heat up the hot side by switching to a much more powerful torch.

[NKTL9414]

Why is your "cold side" 125°F ?

Various types of Stirling engines will (sometimes) be able to run "free piston" or without a flywheel. This may not be true of Alpha type engines.

I've never known any Alpha type Stirling to run without a flywheel, but I don't know if anyone ever tried before.

Including or adding a flywheel seems like a reasonable thing to do, or try, though that monstrous 10" flywheel seems like overkill for an engine with a 1.125" bore.

IMO generally, for a Stirling engine to run without a flywheel it needs to be VERY efficient. Opinions differ, but for various reasons I've never considered Alpha Stirlings to be highly efficient. Inherently, it seems to me that the pistons tend to work against each other, at least through part of the cycle, which would likely necessitate the use of a flywheel.

Thank you for the reply!

It's mainly because thermal conduction from the hot side. The cold cylinder is air-cooled. I can water cool it and reduce the temperature. What would be a reasonably flywheel size that you think is appropriate?

[Tom Booth]

Smart looking design, very cool looking. I like it a lot. Can you post an image with something for size comparison?

Just a "looks about right" guess, but Maybe a 3 1/2 inch or 4 inch diameter flywheel.

To big a diameter would tend to slow down the RPM.

In this animation:

ezgif.com-speed.gif (279.09 KiB) Viewed 6920 times

At the start of 3 there is still some hot air in the cylinder being pushed out towards the cold side, while the cold side piston is also pushing the air in the opposite direction.

Especially if the temperature difference is not great there will be resistance for maybe about 1/4 turn. The added momentum of the flywheel can help carry the engine through such points, but you don't want the flywheel too big.

[VincentG]

I'll second what Tom said, really cool design, and a flywheel is definitely needed.

Maybe Matt Brown or Stephenz could chime in here but I seem to remember your assumed compression ratio (2.3:1) to be around the max achievable with a 90 degree Alpha. That said, your .5" of cylinder head clearance is probably keeping you from seeing that ratio.

Generally, a 2:1 compression ratio should be matched with a 2:1 thermal ratio, so a cold side of 300k(80 F) and a hot side of 600k(620 F).

My own theory on these type Alphas is that the transfer tube should be much smaller to prevent backflow into the cold cylinder.

Just some reasoning, using the engine valve calculator here https://www.gregraven.org/hotwater/calc ... izing.html with your bore and stroke size and a target of 1000rpm, a 4.2mm intake valve is suggested.

Now, a Stirling engine is transferring even less air mass than an internal combustion engine, so the passage could (and should IMO) be even smaller.

[matt brown]

A lot of good suggestions, but I'll add some more...

The long time trick to testing Stirlings (especially alphas) is to run them 'backwards' with a small motor and test for chiller effect. The theory here is that as an engine, work output during expansion must exceed work input during compression, and this requires enough heat input to make it into gas (think temp, time, etc). However, running as 'reefer', you can use relatively massive motor input to check for any chiller effect.

Beautiful engine !!! If you chose to sell it after class, let me know, even if it never runs.

Yes, alphas need a flywheel and likely bigger than 4" that Tom suggested. I'm thinking closer to 6" and around 5 lbs and ideally iron/steel with most mass in 'rim' (the slower the rpm, the greater the flywheel weight required). I see that you used brass for cylinders with copper conduits connecting regenerator. Well, brass is a poor conductor, so I guess you're using the copper conduits as heater and cooler. Vincent is right about how small conduits can be, but here's my pitch...

(1) the 1/2" clearance volume is way to much "dead volume" and will rob max pressure when power piston is descending. I'd suggest fitting "plug" in each head, about 3/8" of something (thick gasket material) glued to cylinder head with high temp silicone.

(2) use the copper tubes as heater and cooler but lightly packed with copper 'steel wool' to decrease volume and aide surface area.

(3) good call on actual compression ratio which is not Vmax/Vmin but depends upon sinusoidal piston motion. Typical alphas require thermal ratio greater than volume ratio (aka Tr>Vr) and converting Fahrenheit to Kelvin your 125-425F thermal cycle is 325-505K or about 1.5 thermal ratio (always best to use Kelvin scale with this stuff). Wise decision to use 1 bar air at this scale.

(4) I assume you mount this to a baseboard, since this is waaaay too big for simple stand, and more so when a flywheel is added.

(5) the regenerator appears way to long, what's the matrix material inside ???

[matt brown]

Hold on...reviewing engine, is that a plastic tube in lieu of regenerator ???

[Tom Booth]

I hadn't noticed the copper tubing until Matt pointed it out.

There have been people in here from time to time asking "why won't my engine run"?

Previously I've pointed out engines made of copper likely are loosing heat/energy due to the copper parts.

Copper draws/conducts heat like a sponge, it is a heat conductor rather than a heat insulator

So in an Alpha, one cylinder is heated with the piston acting more or less like a displacer. The expanding hot air then goes into and through the transfer tube.

As the gas expands through the tube it will eventually reach the other piston, which is more the actual power piston.

Now if the transfer tube itself is absorbing and conducting heat away to the outside air prematurely, before the expanding hot gas can reach the "cold" power piston then you have an effective short circuit to "ground" or the cold ambient air outside the copper tube.

In other words, IMO the copper tubing should be replaced with something non-heat conducting, or at least LESS heat conducting.

Copper is about 400 w/m K

Aluminum would actually be a little better at about 230 w/m K

Carbon steel is not bad at about 50 w/m K

Stainless steel is better at about 15 w/m K

Can't get much better using metal but acrylic is 0.1 w/m K but might not take the heat depending on how hot you run the engine, but just insulating the copper pipe with anything non-heat conducting should help prevent some of the heat loss.

[matt brown]

Now if the transfer tube itself is absorbing and conducting heat away to the outside air prematurely, before the expanding hot gas can reach the "cold" power piston then you have an effective short circuit to "ground" or the cold ambient air outside the copper tube.

In other words, IMO the copper tubing should be replaced with something non-heat conducting, or at least LESS heat conducting.

Tom, we don't know where he planned to add heating & cooling.

I'm assuming the finned cylinder is the compression cylinder and the smooth cylinder is the expansion cylinder. I don't see any soot anywhere, but the idea of directly heating and cooling alpha cylinders is a stretch, regardless of material. The only guy who managed this successfully was Rider. Instead, typical alpha has heating between regenerator and expansion cylinder with cooling between regenerator and compression cylinder. So, those copper tubes are typical alpha heater and cooler locations, where the main issue is maximizing surface area and minimizing dead volume (maybe add external fins and inner matrix to each tube). We need more info, especially regarding regenerator.

[Tom Booth]

... the idea of directly heating and cooling alpha cylinders is a stretch, ...

Is it?

Admittedly, I'm not a big fan of Alpha engines generally, and don't have any hands on experience building one, but I think every illustration, animation (such as above) show heating and cooling of the cylinders.

I think stephenz had a large LTD type engine with heating and cooling on either side of the regenerator, but I don't recall ever seeing an Alpha where heat was applied to the transfer pipe rather than the hot cylinder. If that is a "standard", it's certainly news to me. I'd like to see some citation or reference in support of that because I've never seen it.

If Wikipedia is anything to go by, the article on the Alpha reads:

Alpha-type Stirling engine.

There are two cylinders. The expansion cylinder (red) is maintained at a high temperature while the compression cylinder (blue) is cooled. The passage between the two cylinders contains the regenerator
An alpha Stirling contains two power pistons in separate cylinders, one hot and one cold. The hot cylinder is situated inside the high-temperature heat exchanger and the cold cylinder is situated inside the low-temperature heat exchanger.

At any rate, if the copper pipe is cooled, cooling the air in the transfer pipe BEFORE it reaches the power (cold) cylinder, that's even a worse situation than just passive, incidental loss. Though admittedly NASA seems to think it's a good idea

[NKTL9414]

Thank you guys so much for all the great comments and suggestions! My next steps would be 1.add a fly wheel; 2.insulate the connecting copper conduit; and 3.add some material to displace the top dead volume in the cylinder.

I'll update when I'm done with those and how it turns out.

In the meantime, I did a test where I power the shaft with a drill and reverse the Stirling cycle. The thermal image of that is attached. There is a clear temperature differential on the horizontal connecting tube. This is roughly after 10 seconds of running at around 300-500 rpm by the drill. The thermal camera does not measure temperature well for shine metal surfaces due to their low emissivity and thus the lack of temperature reading on the metal parts. Temperature reading on the top left is in degree C.

To answer the questions regarding the regenerator. I packed very fine steel wool in the horizontal plastic connecting tube. I unfortunately don't have a thermal mass measurement of how much I packed into it, but I don't think it's a lot since the limited volume and my intention to not have big pressure drop across the connecting tube.

[Nut]

I'm just watching and learning with interest. Congratulations on an impressive looking creation. Beautifully built. I look forward to the day you "crack the code" and we see it spinning.