Ten Cylinder "Elbow Engine"
USA - Manufacture date: 2007
Bore = 1/4", Stroke 1 1/8", Flywheel diameter = 3 1/2"
Style: Multi Cylinder, Single Acting,  Hobson's Coupling

Dimensions: 71/2" length x 6" width x 5 1/2" height
(includes wooden base)

The Completed Elbow Engine

This little engine's basic design has been a personal favorite for years. I once had the opportunity to repair one for a collector and I've been in search of my own ever since. They seldom show up on Ebay and hardly ever leave a collection, so the only alternative is to build one. It is known as an Elbow Engine and is based on a mechanism known as a Hobson's Coupling. It is used to transfer rotation across 90 degrees via a mechanical coupling. The action produced in the process is known as "simple harmonic motion".  I just call it hypnotic and fun to watch. The design you see here is a "10 cylinder version, requiring only 7 moving parts to operate. You read that correctly... a 10 cylinder engine with only 7 moving parts. A 6 cylinder version is much more common and uses only 5 moving parts.

After years of lamenting my lack of machining knowledge and skills, I finally acquired a small lathe and a small milling machine  and began teaching myself to work metals. The exciting part of this engine, at least for me,  is that it is my very first engine building project. A  few weeks before these photos were taken, this engines was only a small stack of scrap aluminum, steel and bronze bar stock, which I was finally able to turn into a working engine. I'm just a little bit chuffed about that...LOL. The basic engine design was published by Popular Mechanics, sometime in the 1960's and quite a few engine builder have taken up its deceptive challenge. I was brilliantly ignorant enough to toss out the plans and choose to go with the design in my head.  Thusly, the saga begins. ....LOL

There were a number of considerations that went into deciding to build this engine, and to build it "my way". The first was the desire for something one of a kind that would fit nicely into this collection. The second was to make it a challenging project and the third was to try out a couple of ideas that had been fermenting in my head for a long time. The appearance I finally decided to shoot for was something close to a Victorian style preferably something that would look at home on a gentleman's desk or residing in his library. This lead to the idea of using bronze accents and gun blued steel for various components of the engine. This was modeled out in 3D and proved to work well together within the overall idea.

3D Proof of Theory Rendering

This project began with more than enough challenge for any novice engine builder. I decided to make a classic flywheel with 10 spokes, rather than use the normal flat hub style with a few holes drilled in it. Having fitted my lathe with an indexing plate, drilling evenly spaced holes in a radial pattern was not a huge undertaking. This part of the job went quickly, as did turning the tapered and shaped spokes from 1/4 inch hot rolled steel rod. Things got a bit more interesting when it came time to assemble all the parts. How exactly does one install 10 individual spokes into 2  separate rims and make them run true?

It was a puzzle that took a or two day to sort out, but the answer woke me from a sound sleep at about 4:00 in the morning. I got up and went to the shop where I placed the outer bronze rim on the lathe's chuck and then put the center hub on a piece of 1/4 inch drill rod and chucked it in the tailstock. When the two rows of holes were aligned, and measured with a dial test indicator, the hubs were obviously concentric. Ahhhh.... the smell of success is sweet.

By carefully inserting opposing spokes and hand turning the lathe chuck to check things, as I went along, the rim and the hub remained concentric and held their position quite nicely as I used red loctite to secure them in place. A press fitted outer rim completed the flywheel by locking the spokes in place and hiding the exposed holes in what was now the interior rim of the flywheel. A quick spin test on the lathe proved the flywheel was concentric and ran true, with no wobble. Beginner's luck?... maybe, but the idea worked out well and the project progressed another step.

Building the flywheel was surprisingly easy and is something I'll be willing to tackle again in future engine projects. I did mention that this little engine has hooked me on building my own... I didn't?  A few dollars worth of scrap metal and a few enjoyable hours of work beats paying the rediculously high prices that handmade engines have risen to, by a very long shot.

The Finished Flywheel Installed on Drum

The Revolving Drums and Bearing Plates

The next step in the journey was to make the pistons. The engine uses 5 piston sets bent to 90°. Each end of a set acts as the piston in one of the cylinders of one of the revolving drums. Each set acts in unison as air is applied or exhausted to extend or retract the pistons. This action causes the drums to turn in unison and run the engine. The piston sets have to be perfectly set at 90° or they lock in the cylinders and nothing will turn freely. This process took several attempts before things were completely squared and turning "relatively free" within the 10 cylinders. 

One also has to take care that the pistons are made to the proper length or they will strike the bearing plates and stop the motor. By adding the decorative spheres, I had managed to make things even harder, since I now had to avoid having them strike the top of the drums too. Some experimentation ensued and the proper middle ground was soon achieved by a bit of easy math and some trial and error testing. There is very little margin for error in any Elbow engine, but this one has next to no wiggle room at all. A slight misalignment is all it takes to stop the motor dead in its tracks.

Test Fitting the Pistons

My training as a real world artist and experience in the world of 3D design has taught me that sculpture must be interesting from any angle from which it is viewed. You can't ask the viewer to stand in one place, just so you can hide your errors or lack of imagination. Any viewing angle without some point of interest means your design has missed the mark. The whole engine needed to be interesting and it took a bit of work to come up with a workable design that had the right sort of total visual impact while retaining the intended style..

Rear Structure and Partial Air Line Test Fit

The original plans for an Elbow engine used a rather plain looking piece of aluminum plate as a rear support for this engine. To achieve the Victorian look that I hoped to have when finished, something a bit more visually complex would have to be created. A number of ideas were modeled in 3D until a pair of classic columns were chosen. They worked well with the spoked flywheel and a flying bridge soon followed them into the design. This meant the supply and exhaust lines could no longer be hidden within the engine base itself.  Exposed lines were anticipated, but their final design would wait until the main structure was in place and fairly close to properly adjusted.

Air Lines Installed for Test Run

As things progressed, the need for fitting the air lines grew. The drums needed to be adjusted to their final positions and the engine was nearing a place where it could be test run, but not without an air supply. The bronze plumbing "T" design was chosen for appearance as were the flanged fittings. These were turned on the lathe in three pieces and soldered together. The spherical section was drilled for a slip fit, making it adjustable. Once the proper location was found, the pipe was drilled using the T fitting as a drill guide. The pipe was then locked in position with red loctite, in order  to avoid having to solder the anodized steel pipe surfaces. Red loctite is an excellent substitute, in certain instances, where soldering is not possible or is considered undesirable. It forms quite a strong joint and acts as an airtight seal to prevent leaks.

One would think the hard part of the project was behind. The machining work is all done, much of the heavy polishing is complete and all that remains is the test run.  Nope... it ain’t that easy there Sparky...(grin). These little fellows are very finicky in their simplest form. This one was even more so with its additional cylinders and somewhat altered stance. No... this is where things got difficult and rather frustrating. Proper alignment is actually difficult to achieve and the sweet spot is quite elusive. Until everything is exactly where it needs to be, friction is the enemy and binding in the extended pistons is the norm. Consider that each time the engine has to be disassembled, it also has to be readjusted and you can soon begin to quietly "converse" with yourself as you work.


This engine has been apart uncounted times, the pistons were tried in combination after combination until finally an undetected problem was spotted. Using Play-Doh to make an impression for comparing the pistons revealed a slight, but identical deviation in the bend of each one. This tiny problem was causing a .020 inch mismatch when all the pistons weren't "facing the same way". Once matched in proper fashion, the friction problem was soon conquered.

Ok... time to test run it again... sort of . My air compressor is of the small bench top 5 gallon variety and does well to keep up with a large bore steam engine, having a sealed system. An Elbow engine is anything but a sealed system, so the poor compressor can't keep up with the demand. The engine will only run for a few revolutions before the compressor runs out of air, but at least it runs. Excitement enough for a first build ,but a fair disappointment in that I can't easily play with it. Sadly, that fun will have to await the arrival of a larger air compressor, hopefully really soon...LOL

The engine was finished up with a detailed hand polishing and the addition of a hand rubbed wooden plinth.  I'll offer a couple of tips for those looking to build this engine. It is a deceptive build. The machining part is pretty straight forward and well within the abilities of most intermediate level machinists. That is where the deception begins. You'll need the patience of a biblical  Job and stubborn perseverance bordering on obsession, when it comes time to adjust and test the little dickens. This is not the perfect choice for a first build engine project and modifying it, as I did, certainly tested my rather limited experience levels to the maximum. I got through it with some serious head scratching and a lot of encourqagement from the crew over at the Home Model Engine Machinist Forum. I learned a tremendous number of new tricks, but sometimes failure was feeling uncomfortably near.

All in all, I'm pleased with the outcome and I've totally enjoyed making the discovery that I too can successfully build my own engines. You can bank on there being more to come.


Stuart Beam  -  Reeves Tangye  -   Williamson   -   Stuart H9   -   Stuart D-10   -   Jensen 51
"Minnie"   -   Cretors D  -  Lewiston High   -  Radial  -  Elbow   -  Rectilinear   -  Curved Cylinder
Böhm Stirling  -  Water Pressure  -  Coomber Rotary
Toy  Engines   -   Turbines
Steam Links
Jensen Steam Engine Store
Antique Steam Engine store

Do You Have an Unusual Steam Engine to Sell?
Email Information to