Ivan's practical tips on building a gamma type Stirling Motor
(or, things that the other sites don't talk about)

My twin cylinder (2 x 8cc) Stirling Motor (56K).
Do you want to see it run? ( 490K) animation from my colleague Hanspeter Holzhauser
I have constructed this site because of the difficulties I had in scraping together enough solid facts to be able to design and build my own Stirling Motor. The amount of information about these Motors on the web is vast but try to find anything actually related to building them! One can order commercial kits or even full working models, Koichi Hirata has some interesting free plans but I wanted to build one to my own design.

So, if you can't find the answers to some of the infuriating little problems that nobody ever mentions on their homepages and, like me, you are too embarrassed to trouble the "big guns" (profi's) with seemingly stupid questions then perhaps you can learn from the experience of a "little gun".

Table of Contents:

Things that the others don't talk about:

I wanted to build a working model of a Stirling Motor incorporating the concept of using waste heat from a paraffin lamp (light) to generate power and warm water. I didn't quite fulfil all my aims but on the whole I am satisfied. I gave up with paraffin, I couldn't get the heaters to burn evenly and cleanly I always ended up with soot over everything - now I use alcohol instead (for the burners - not for me!). The design and finish had to be pleasing enough that I could display it in the living room without disturbing anyone - judge for yourself if I succeeded. This is the first and only stirling motor that I have built, perhaps it will be a one-time wonder - I am still getting over the shock of success and the five years needed to reach it. It has, up until now run for a total of about 2.5 hours (divided into 15 minute sessions). Its speed has been measured twice (with a Hioki 3402 Tachometer), first run from cold - it ran independently after 1min 50sec of heating, reached a peak speed of 1,020 rpm after 4 min., slowed down to 580 rpm over the next 10min. Second run (after cooling down) was similar but with a peak speed was 930 rpm. Output water temperature after 10 min. (into the tank) was 50 ºC and obviously had a major influence on the rpm. External cooling would be much more efficient but it wouldn't look nice on the dining-room table!

I am very lucky, I have at work a very small workshop (just for me although I have no formal training as a machinist) it contains a small lathe and a milling machine. I wish to thank our glassblower (Mr. Daniel Schnarwiler) who made the chimney with its two special holes (very tricky).

As you read on you will see that I am not an engineer or scientist or trained metal worker (typical lab. tech. - knows a little about everything and nothing about anything specific). I haven't made any thermodynamic calculations, I followed my instincts and vaguely remembered physics lessons from school as well as logic and common-sense (hopefully). The web was very useful for basic concepts but was not really very useful when it came to designing and building.

Schematics of an alpha type Stirling Motor

Schematics of the beta and gamma type Stirling Motors

Schematic of my gamma type Stirling Motor.

Back to the text
As I learnt to my cost you should make sure that you understand the three main classifications properly. I confused a beta animation from K. Hirata (laid out in the form of a gamma) with a gamma and tried to build something in between - which did not work! The model that you see in the main picture incorporates 80% of the parts from my first failed attempt. This is why I have shown more than one variation of a type, avoid the mistake that I made. There are many types and variations of Stirling Motors plus various kinds of heat motors (Ringbom, vacuum, atmospheric (flame lickers), etc., etc.). I would say if you are good enough to build an alpha or a beta then you probably won't need my advice.

Any ideas you glean from this page are my solutions to various problems encountered, they are not necessarily the best way of doing things. A large number of heads are better than one - someone out there is bound to have found better solutions. I will assume that you have seen the major web-sites (links) and the various brilliant animations from Koichi Hirata and Michael Abendschön demonstrating the principles involved.

Here are some of the "little" details that gave me trouble as I started out:

1) I decided to build a gamma because the displacer does not need to seal to the cylinder which could have a temperature around 500°C. The only practical way (for amateurs) to build an alpha / beta would be to use a well-fitting graphite piston for the displacer (with its resultant mass) and all the associated infrastructure to support it.

2) I wanted to avoid too much mass for the displacer piston, I could imagine that my two rather frail brass levers (but they look nice) would start to flex badly should the motor ever reach a reasonable RPM. I decided to use aluminium, the piston was turned out of the solid leaving the wall as thin as possible (0.4mm). An end cap, containing the thread for the push rod, was glued onto the "cool" end with high temperature araldite (XD 4510 -180ºC). Displacer Ø 22mm, length 65mm, total weight about 8g.

The open motor featuring the displacers (139K)
3) I used a 3mm (internal) Ø "O"-ring and oil to seal the push rod of the displacer piston. A long bronze bush would have sealed well and given little friction but would have made the tolerances for rod straightness and component alignment too small for my limited engineering skills.

4) I probably have not left a big enough gap between the displacing piston (ext. Ø 22 mm) and the cylinder (int. Ø 23.4 mm). It is difficult to decide the optimum between having too much "dead" volume and the extra air resistance caused by the air having to force its way through the narrow gap around the displacer. I do not know of an ideal displacing piston diameter to gap ratio, maybe someone else does?

5) My original power cylinder was a simple construction made out of three pieces of aluminium, an "O"-ring and a threaded rod. It proved unreliable, it would seize with each speck of dirt even though well lubricated with silicon grease dried out with graphite powder. I solved the problem by making an Aluminium/Teflon hybrid. It was much more complicated and difficult to make but it has turned out to be worth it - see photo's. Tip: the Teflon lips were left oversize until the cylinder had been assembled, only then were they turned down to the size of the cylinder (it was surprising how much out of true it could get due to the very soft Teflon). The retaining groove in the piston is such that the "O"-ring touches the cylinder but does not press on it (the seal is not perfect - but not bad) it has very little resistance (too much would stop the engine from running).

Power pistons:
Hybrid Aluminium/Teflon
Original in Aluminium
Power piston - open (39K)
Power piston - partially assembled (49K)
Cylinder & power piston (fully assembled) covered with silicon grease and graphite (69K)
6) My first brass flywheel had a diameter of 5cm and a weight of 190g, it wasn't enough to overcome the compression of the beta/gamma motor. When I decided to design a new motor I also decided to increase the size of the flywheel to 10.5 cm giving it a weight of about 440g. It is probably bigger than it needs to be (cuts down the RPM) but it works.

7) One of the most important factors in getting my motor to run was the reduction of the thermal conduction between the hot and cold areas of the "hot" cylinder. Yes, I know it is obvious! Now! Who would have thought that a wall thickness of about 0.6mm would have transferred so much energy. I turned down the walls of the "hot" cylinders until they became mechanically unstable in the lathe. I achieved a wall thickness of around 0.3-0.4 mm. Unfortunately my cutting tool was not the best for stainless steel especially when cutting at a depth of up to 9.5 cm. After this improvement the cooling water didn't get quite so hot and if I positioned the flames of the lamps to burn right at the ends of the cylinders instead of in the middle the motor would run - and well! Another alternative (or additional) measure would have been to increase the length of the cylinders and displacers. Thermal separation would have been improved at the cost of increased "dead" volume, air resistance and displacer mass. The type of material used for the cylinder could also have been exchanged for one with a lower thermal conductivity, latheable ceramic for instance. One could use high temperature Araldite for glueing the cooling tank on - good luck to any one that tries it.

8) The compression ratio is governed by the ratio of power volume (swept) to displacement volume (see graphic). My motor is roughly 1:1. The lower the compression ratio (displacer >> power) the easier it is to get the motor to run (but slower). This is a great help if the thermal separation is poor. The diameter of the displacer could be increased until it enters the league of the so-called LTD Stirling Motors (low temperature difference). These can be powered from as little as a hot cup of coffee (tea, water..). Eventually the displacer becomes so large that it has to be made out of polystyrene or something similar to keep its mass down and its stroke has to be reduced to avoid too much air resistance. These materials are only suitable for use at relatively low temperatures and hence, can't be used to increase power output. Basically one should be able to increase the size of the displacer a little with almost no negative side-effects - although I haven't tried it for myself.

9) There are lots and lots of pictures of all kinds of Stirling Motors on the web. I don't remember seeing a single one with some sort of starting device. As we all know, except for some very rare exceptions, Stirling Motors are not self-starting. I can only assume that one has to try and flip the flywheel or one of the cams on those other motors (very crude). I solved the problem by splitting the flywheel into two and putting a toothed cog in between the two halves. A toothed belt connects the flywheel to the starting knob - you can see it on the animation.

A couple more photo's for good measure:

Power crank (37K)

Milled aluminium with an inset bronze bush and a 10mm ball race (secured with Loctite).

Displacer lever (45K)

Made out of sheet, a block and tubular brass all silver soldered together.

If you have any questions I will do my best to answer them - the correctness of the answers is not guaranteed although I will do my best. Suggestions for further "little details" would also be welcome as would improvements or corrections to anything that I have all ready suggested. Questions in English or German please, answers will be mainly in English - my German grammar is horrible.

Here are a few of the many available links concerning Stirling Motors.

One of the best ones for beginners is: http://www.bekkoame.ne.jp/~khirata/indexe.htm

Some really good animations of the various motor types: http://michael.abendschoen.bei.t-online.de/hauptmenue.html

A small selection of some of the main sites: http://stirling.sites.solnet.ch/index.htm , http://www.geocities.com/~rrice2/ , http://www.stirlingsouth.com/ , http://www.jerry-howell.com/ , http://stirlingengines.org.uk/ , http://www.stirlingengine.com/ , http://sesusa.hypermart.net/

Why , How , Me , Alpha , Beta , Schematic of my gamma motor , General / Motor type? , Displacer material? , Displacer seal? Displacer gap / diameter ratio? , Power cylinder construction , Flywheel mass , Improvement of thermal separation , Compression ratio (Power swept vol.:displacer vol.) , Starting

Ivan Woodhatch April 2003