Power Producing Stirling Engine # Team Comments

Moderator: stan.hornbaker

Forum rules
Be nice!

Power Producing Stirling Engine # Team Comments

Postby info74 » Wed Oct 16, 2002 11:45 am

I have been asked by the university where my father attended medical
school to build them a low cost, low tech, reliable Stirling engine.
Naturally I had to say yes.

After much discussion we have decided to post much of the information
about this design on our web site as the design develops. The initial
web site is located here.

Everyone who is interested in Stirling engines (whether you are a part
of this team project or not) is welcome to start their own free
Stirling engine web site using our server. To do that you simply
register as a user, then click on Homepage Maintenance to get your
site going. The page cited above was built very quickly using
Netscape Composer and the site editing tools built into our server.
You need to upload a file called index.html for your home page.

This thread is for use by team members only. If you are not sure if
you are part of the team or not, your comments should probably go on
the - General Comments page instead of this one.

Posts: 70
Joined: Thu Aug 02, 2001 7:41 pm
First Name: Brent
Last Name: Van Arsdell

High Temp Gaskets

Postby info74 » Mon Oct 21, 2002 12:49 pm

Flex Tech Seals has some high temp gaskets that may work well for this application.
Posts: 70
Joined: Thu Aug 02, 2001 7:41 pm
First Name: Brent
Last Name: Van Arsdell

Hot Head Stress

Postby info74 » Mon Oct 21, 2002 10:21 pm

I ran a quick NASTRAN model (188 k) of your proposed structure, and the results are, shall we say, sobering. I kept all of the thicknesses of the entire structure 0.06" thick just to see what would happen. There is about 300 ksi generated in the membrane web, and the stiffeners are seeing up to 700 ksi in the outer edges. Attached is the contour plot of the major principal stresses. The stiffeners are 2" tall, and the cylinder is 3" deep as we discussed on the phone. The load is 450 psi applied normal to the membrane. I did not apply 450 psi to the cylinder, so hoop tension is being ignored for this case.

Obviously, 700 ksi is ridiculous, especially since your operating at temperatures around 1200 degrees. I found the typical yield point of stainless steel at that temperature to be around 18 ksi. There is a long way to go to get from 700 to 18. I haven't increased thicknesses yet to see what it will take to make this work; however, by the results of this particular model, I doubt 0.06" is a reasonable gage. Just guessing, the membrane will end up in the 0.2 - .3" thick range, depending on how the stiffeners are designed.

I just wanted to send you this preliminary while I had the chance. I will continue to play with the model to see what works.


Re-posted by Brent

Posts: 70
Joined: Thu Aug 02, 2001 7:41 pm
First Name: Brent
Last Name: Van Arsdell


Postby info74 » Wed Oct 30, 2002 4:07 pm

I re-modeled the Stirling engine cylinder using a "fan fair" corrugated membrane, and used the same stiffener arrangement that I had previously. This model shows much more reasonable results. The thicknesses are as follows:

Corrugated membrane: 0.06 in.
Stiffeners: 0.3 in.
Cylinder Wall: 0.2 in.

The maximum stress shown in the outer edge of the stiffeners is 24 ksi, and the maximum stress in the membrane is around 18 ksi. This hits my target of getting the membrane under 20 ksi, and the stiffener stress somewhere close to 20 ksi. The model is shown with 150 psi of pressure. Also remember, there is NO factor of safety used for this analysis. I'm not sure exactly how the engine will fatigue a the high temperatures, so I am just providing stresses for actual load.

I have attached a .jpg file of the stress contour plot, (184k) and I have also done some hand analysis of the stiffener attachment which is included in the .PDF file. (394 k) I would recommend that you attach the stiffeners with a pinned joint of some kind, so there is no moment reaction by the beam. This will create more deflection, but it will save you some fatigue problems at the stiffener attachments. One thing I have learned about structural design is that larger structure creates more stiffness, which in turn creates more internal stress. Sometimes making a part less rigid actually reduces the stress. In this case, making the end fixity a pinned joint will eliminate unwanted stress.

Take a look at the attachments, and email or call with questions.


re-posted by Brent

Posts: 70
Joined: Thu Aug 02, 2001 7:41 pm
First Name: Brent
Last Name: Van Arsdell

Hot Head Stress

Postby info74 » Wed Oct 30, 2002 10:05 pm

We have been trying to design a Stirling engine that does not have the pressure vessle do double duty as the hot and cold heat transfer membrane. The initial reports from our stress analyst have not been encouraging.

We have considered rounding the hot head of the engine with an approach along these lines. Humm...
Posts: 70
Joined: Thu Aug 02, 2001 7:41 pm
First Name: Brent
Last Name: Van Arsdell

Our approach has changed

Postby info74 » Fri Nov 29, 2002 12:35 pm

While I belive that the previously discussed concept could work, I have decided to change our approach. We are now planning on building an engine that is very similar to the Sunpower Rice Husk engine. Merrick Lockwood has graciously offered to help us with this project.

The Sunpower Rice Husk engine is discussed in the book Principles and Applications of Stirling Engines by Colin West. An authorized re-print of this book is available from Stirling Machine World. When our engine is complete we will be publishing a book by Merrick Lockwood that will include photos of the engine and a complete set of plans.

One of our first cad drawings of part of the engine is located here. As always, team members comments should be posted on this thread and general public comments should be posted on the General Public comments thread.
Posts: 70
Joined: Thu Aug 02, 2001 7:41 pm
First Name: Brent
Last Name: Van Arsdell

Comments from Merrick

Postby info74 » Fri Nov 29, 2002 12:44 pm

I have re-posted some comments from Merrick Lockwood about this project. These will be more meaningful to others after they have had a chance to learn more about the original project, but for now I think they stand on their own.


As I said before I feel a rhombic drive for the RHEP
(or ST-5) would have been a good arrangement Using a
dry crankcase and teflon alloy rings the engine could
have been in either horozontal or vertical
orientation. By being vertical the friction losses
and wear due to piston/displscer weight would be less.
The two crankshafts, six connecting rods and timing
gears add to the cost, complexity and weight of the
engine but would provide vibration free durability.
The Philips patents on the rhombic drive were not
renewed in the UK in 197?. I presume that they have
expired in other countries, like the USA, but have
never checked.

The RHEP engine avoided oil for fear that vaporized
oil would migrate into the hot end and and with
pressurized hot air produce an explosive mixture. The
one fatality that Philips experienced in Eindhoven was
due to a hot air engine (with oil lubrication)
exploding. After that all the Philips engines used
either hydrogen or helium. I feel (but am not
positive) that an engine operating with air near
atmospheric pressure and not very high temperatures
would be safe (like the 1800's versions). You should
check this out before suggesting ai high temperature,
high pressure, air engine. Even if oil droplets don't
migrate into the regenerator and hot end, oil might
vaporize from the cylinder wall(??).

I have doodled a bit with the idea of a rhombic drive
for the ST-5 but no detailed drawings. This, along
with my micro rice mill, is something that I keep
hoping to get some source of funding to work on here
in India. Time will tell.

That's it for now.



The original Sunpower design for our engine used the
articulated connecting rod for the piston that
incorporates a swing link. The purpose of this was to
avoid the large side forces that a single connecting
rod would put on the piston and the resulting power
loss and wear. The bell crank minimized the side
forces on the displacer rod. These problems would be
pronounced if the connecting rods are short but less
significant with long con rods. A double throw
crankshaft would be an promising solution if the con
rods were long. A single central throw could drive
the displacer with the tail of the displacer rod
supported by a spider (as in the RHEP Ericsson
engine). The same piston design could be used with
two (longer) piston connecting rods extended to a pair
of throws on the crankshaft on either side of the
displacer throw.

The big advantage of this arrangement will be that the
displacer drive can be almost as strong as the piston
drive and thus avoid that (fatal) flaw in the original
Sunpower arrangement. The drawback of long connecting

rods is that the crankcase is longer. If you
pressurize it would probably be best to stay with a
crankcase built around a stock mild steel pipe with
welded bits for the crankshaft bearings. The cost and
complexety for adding a foot or so of pipe to make the
crank case longer would be minimal. I wholeheartedly
endorse having the crankshaft supported on both ends
with dual counterweights. I used that arrangement
when we switched to a Ross linkage in 1986 and in 3 of
the 4 engines that I have designed since then.

Be sure to keep the phase between the displacer and
the piston the same, I suspect Sunpower had this
right. As you can see it is NOT 90 degrees. You can
work it out from the geometry of the crankshaft and
bellcrank. It looks like something in the vicinity of
60 degrees.

The plain hot end that I made had the same dimensions
as the finned one except that the heater can was 50%
longer. Of course the displacer has to be
corrispondingly longer also. I built up the hot end
by bolting the flange to the lathe face plate and
successively welding/aligning/trimming the regenerator
can, regenerator shoulder piece, heater can and heater

Next I welded a bunch (? my guess would be 12
longitudinal rows) of tabs to the inside of the heater
can. These were around 16 mm long, 3 mm (1/8") thick
with a height greater than the required gap. If I
were to do this again I would opt for thicker
stainless steel, say 3/16", welded on both sides to
facilitate the next operation.

The hot end was finished by mounting it, flange
outward, in a jig on the lathe face plate. The jig
can be made of two mild steel flanges welded to a
section of mild steel pipe big enough to hold the hot
end. The jig was clamped or bolted to the face plate
and the outside flange faced.

Now I'm sort of guessing. The hot end flange can be
drilled with one set of loose holes for assembly to
the engine and a second set of tapped holes for
mounting on the jig. The jig would be drilled with
loose holes and the hot end mounted by bolts through
the back of the flange of the jig. With this one
set-up the heater flange can be faced, have the
outside diameter turned to its finish size. Finally
the inside tabs are bored out to fit the liner (make
the liner first so it can be fitted at this time).
The reason for the thicker tabs is that this machining
operation should be easier, the 1/8" material tended
to bend under the force of the boring tool. The last
circumferential row of tabs at the top should only be
bored part of their length so that when the liner is
pushed in and seated you have the correct end
clearance for air to flow from the hot space into the
annular heater space. Use a angle grinder or die
gringer to bevel the leading edge of the tabs so the
liner doesn't catch as you push it in.

Hope I haven't completely lost you. Let me know if
something isn't clear.

Good luck,



Comments on the power output of a plain hot end that 50% longer than the original engine hot end, but was fabricated without corrugations:


Back to the plain hot end. I gave you the details of
fabrication but didn't mention performance. I used
the plain hot end on our Ross engine but we only had a
couple of weeks to work the bugs out of the engine.
In this time we went from zero horsepower (mainly due
to leaks) to 3 horsepower. I think we could have
increased that with further de-bugging but some of
this may have been due to limitations of the hot end.
I figure that for a plain heater one could expect 1
watt/sq cm. I don't have the measurements of the hot
end here but you can figure that out.


Posts: 70
Joined: Thu Aug 02, 2001 7:41 pm
First Name: Brent
Last Name: Van Arsdell

Progress so far...

Postby info74 » Fri Dec 06, 2002 6:28 pm

Here's a picture of the current version of our engine. We've already started purchasing materials, so this is getting exciting.
Posts: 70
Joined: Thu Aug 02, 2001 7:41 pm
First Name: Brent
Last Name: Van Arsdell

Regenerator Etc.

Postby info74 » Sun Dec 08, 2002 2:33 pm

The regenerator material used in our engine was Metex
which was formed from knitted stainless steel wire.
It was made from a knitted tube, perhaps 4" or 5" in
diameter which was folded up and sheathed in another
knitted tube about 1" in diameter this came in any
length one wanted. To install it in the regenerator
space I machined a cast iron mandrel that seated on
the inside liner of the hot end so that the outside of
the mandrel was flush with the outside of the inner
liner (about 302 mm). One layer of metex was wrapped
tightly around this mandrel to the point that it
filled the annular space in the regenerator housing
and then pressed into the housing using a cast iron
ring that just fitted into the regenerator space.
This was repeated however many times (4?) to tightly
fill the regenerator space. This left only two loose
ends at each end of each layer so there was no worry
about loose bits getting loose in the engine. I was
told that Metex was expensive but as we were supplied
enough for our engines I don't know just how much.
The regenerator space is open at the bottom of the hot

Operating the engine on propane, biogas, etc. will be
relatively simple. I haven't made this sort of burner
but Andy Ross has described in detail how he has made
ring burners for his engines in which gas is mixed
with air and then distributed around the cylindrical
part of the hot end. Check his book on making
Stirling engines, the description was in that and/or
one of his papers (most of my references are in the US
so I can't say for sure). You will have to cover or
insulate the end of the hot end.

Biogas is a mix of methane, CO2 and other things so it
has a lot less calorific value and will burn
differently than propane. You might dilute the
propane with CO2 or something to try to emulate biogas
in order to get your jet size and flame holes right.

I have checked the drawings and find that the diameter
of the flange on the displacer base was 292 mm and the
displacer can was 0.8 mm thick so the OD of the
displacer can was (roughly) 294 mm. The diameter of
the heater liner was 302 mm so the clearance between
the displacer and liner was about 4 mm.


Posts: 70
Joined: Thu Aug 02, 2001 7:41 pm
First Name: Brent
Last Name: Van Arsdell

Return to Power-producing Stirling Engines

Who is online

Users browsing this forum: No registered users and 1 guest