Stephenz's work

[stephenz]

My design was initially a pure Alpha and now is pure Gamma, but it is still uses the typical alpha crank design, where both connecting rods connecting at the same point on the crankshaft (which also means my design has the same stroke on both cylinders) and the 90 degree angle between the 2 cylinders is what provides the phase angle.

the transfer tube connects the head of the power piston to an annular copper liquid cooled heatsink. The regenerator is composed of 18 tubes (100mm long and 5mm in diameter) which can be fitted with various materials for testing purposes. I've tested a lot of things, from various stainless steel wools, fine mesh/matrices and even ceramic balls of various diameters. The heater is composed 18 stainless steel tubes which are heated by a propane torch fitted to a stainless steel enclosure acting as furnace. The torch is also supplied secondary air (5psi regulator) to adjust the mixture as the enclosure has a minimal amount of opening of exhaust and as a result it's difficult to have a good combustion with a naturally aspirated burner. I can reach 500-550 C at the base of the heater tubes with this setup..

[stephenz]

Yes, I can machine a lot of things, but not a crankshaft. So I bought this cheap compressor from HF.
Initially I was even using their crankcase, which I eventually replaced by a version I designed that be more easily serviced, so orings and screws, and also added ports for my lubrication system, various sensors, etc.

[Tom Booth]

What direction of rotation are you using?

For example, if both piston and displacer are""up" together near top dead center, which direction will the crankshaft turn? Toward the displacer or toward the power piston?

[stephenz]

direction of rotation is counter clockwise when facing the flywheel

[Tom Booth]

Did you make some connecting rod replacement?

Does the displacer extend up above the copper cooling area into the heater head?

Is the bottom of the displacer cylinder closed off?

Z-Mk-I-001.jpg (239.65 KiB) Viewed 1630 times

Sorry if you've already covered all this but it seems you've been doing a lot since you were previously posting about this.

[stephenz]

Yes the bottom of the displacer cylinder is closed off, if it wasn't it would be a displacer but a piston. What makes a displacer not a piston is having (roughly) equal pressure under and above.

I am using stock connecting rods. But I made a part that mounts to the displacer connecting rod (with a shorter pin) to which I mount a 8mm rod going through a linear bearing mounted at the bottom of the liquid cooled heatsink. After the linear bearing, that rod goes through a seal to avoid pressure losses with the crankcase. Even though both engine and crankcase are pressurized, the delta of pressure can be pretty significant (pressure ratio of about 2), the higher the pressurization the greater that delta is.

[Tom Booth]

So the crankshaft is still the same and the swept volume of displacer and piston I assume? Did you also maintain the original crank angle, offset between power piston and displacer?

Somehow when looking at mine, I don't quite know or understand exactly why because I haven't even looked inside the crankcase yet, but the advance or whatever between the two pistons reaching TDC at the top of each compression stroke does not seem like 90° to me.

I guess it could be if I rotated the crank all the way around, but it looks more like about maybe 45°


https://youtu.be/TV5cCFceAc4?si=Te-WqN14Jyvpq1id


Was that the same on yours?

[stephenz]

Yes, no change to crankshaft, like I said above the stroke is fixed and is the same for both cylinders.

The phase angle exactly 90 degrees. If you're not sure, mark the flywheel with tape when you exactly reach tdc on one of the piston and then do the same for the other piston.

if you have a carpenter square you can also verify you have a geometric 90 degrees between the 2 cylinders.

Edit: just watched your video, it's hard to see but seems a bit odd to me. Mine is a true 90 degree. Either your crankcase is different (doubtful) or the 2 cylinders are not at 90 degrees. Do you have a construction/carpenter square?

[Tom Booth]

... Do you have a construction/carpenter square?

Yes, several. But I can't get over to the shop (where the compressor is) for a while. Maybe tomorrow.

It looks as though the two cylinders are at a weird angle so they are closer together at the bottom maybe?

If so that would put them at a skewed angle so the offset was less.

It is also not a two stage compressor but a one stage.

Resize_20221213_173024_4529.jpg (291.04 KiB) Viewed 1620 times

Anyway, it looks like you do amazing machine work. That engine looks like it should be 5kw or something.

I thought I might be able to get some ideas why you could only get 30 watts, but so far, I don't know.

I would say, way too effective cooling is sucking out the power.

My thoughts at the moment is water cooling would mitigate the cooling by the copper.

Copper draws heat away instantaneously.

[Tom Booth]

Do like the crazy YouTube ROCKNTV1 guy and heat your cooling water up to boiling.

To my mind "real science" is about experimenting not following the letter of whatever it says in the textbooks. That ROCKNTV1 guy might have been on to something.

I know if I try to heat up a copper pipe doing plubing work, with a map gas torch and there is even a bit of water in there that can boil to draw off the heat, within about five feet of the joint I'm trying to solder you could heat that little copper pipe all day long and it will never get hot enough to melt the solder. Not until every drop of that water boils off.

Copper conducts heat just like it does electricity. The copper cooling jacket and copper pipe is like a heat pipe sending heat straight out and over to the other cylinder most likely.

[stephenz]

The main reason for the poor power is really poor design on my part. I did everything based on empirical figured found online backed up by nothing but ear-say.

But once I got the engine working I figured the best course of action for me was to simulate this engine to see if the simulation would agree with actual data. And aside from some possible variance within estimations for friction for example, I was able to validate the numerical to a surprising degree. And it turns out, in case you didn't know high temperature engine are really sensitive to a lot of parameters. I spent a good amount of time fine improving the design which means remaking a lot of parts, cooler, heater, regenerator. I am pretty confident the Mk-II will output over 300 W or more.

My advise:
- wall thickness for cylinder, displacer, burner/heater: as thin as possible. Stainless Steel over any other metal
- friction vs seal: the sweet spot is hard to reach, aim for the best seal you can get at the lowest friction you can get. There is no substitute for one or the other.
- and the most important of all: if you want to build something remotely efficient, forget anything you think you know and use software to do the optimizations for you. There is absolutely nothing trivial about those optimizations, and stirling engines are not intuitive machines. Oversizing any of the components of the engine will result in performance degradation, because all characteristics of all components of the engine are working against each other, meaning you improve one: you will damage another. It's all about balance. There are not many ways of doing this but iteratively which is insanely difficult to do without proper software.

This is not meant to be discouraging but my personal experience.

And yes, this probably doesn't apply to LTD's which are fundamentally easier to design and build.




just found my receipt, this is the one I am using. I had 90 degree phase right off the box. Perhaps your smaller compressor is built different.
https://www.harborfreight.com/37-hp-v-s ... 4bd71e5753

[stephenz]

Do like the crazy YouTube ROCKNTV1 guy and heat your cooling water up to boiling.

To my mind "real science" is about experimenting not following the letter of whatever it says in the textbooks. That ROCKNTV1 guy might have been on to something.

I know if I try to heat up a copper pipe doing plubing work, with a map gas torch and there is even a bit of water in there that can boil to draw off the heat, within about five feet of the joint I'm trying to solder you could heat that little copper pipe all day long and it will never get hot enough to melt the solder. Not until every drop of that water boils off.

Copper conducts heat just like it does electricity. The copper cooling jacket and copper pipe is like a heat pipe sending heat straight out and over to the other cylinder most likely.

This is why you would never build a regenerator, cylinder walls, displacer out of a conductive material. I only have copper on about 1/3 of the length of the displacer cylinder. The burner only applies heat to the heater tubes that start roughly 100 mm away from the cooler. The regenerator separate the cooler from the heater. The role of the regenerator is create a temperature gradient across its length. You want the temperature at the top of the regenerator to be as close as possible to the heater temperature, i.e. as high as possible and you want the temperature at the bottom of the regenerator to be as close as possible to the cooler temperature, i.e. as low as possible.
For for a cooler, copper is the best you can do. You want to be able to reduce the temperature gradient between the gas and the water, because your goal is to get the working gas as cold as possible.


I am asking you kindly, and only once, not to start talking about how "not cooling the cooler will make the engine run better". I don't want any of this non-sense in this thread. How you think this makes any sense is just beyond me. I have tried to point out to you multiple plausible explanations that would explain your observations, and proposed to you how to correct your experiments. The main issue is that all your experiments were done on toy LTD's where the quantities of heat you are dealing with are ridiculously low, LTD's are designed to run off ridiculously low temperature delta's, in other words your signal to noise ratio is not adequate, meaning your conclusions are meaningless. Build a bigger engine running at 1000 F, measure the rpm and then cut off liquid cooling and call the local news when the rpm go up . I'm kidding obviously, but I don't want to discuss this here, I'm happy to try to explain things again in one of your threads. (and again: I mentioned this recently, but I forgot to turn on the pump/fans for the liquid cooling in one of my tests, the cooler hit 100 C and only realized that because the RPM seemed a bit lower than usual. As soon as I turned on the pump/fans the cooler temperature dropped from 100 C to 40 C and the RPM increased from 300 RPM to 350 RPM).
[Subject Closed]

[Tom Booth]

Well I wish you luck with your project. Your machining work is really quite beautiful.

[stephenz]

thank you