Air engine 2


You can buy plans
for this air engine


I had previously built a reciprocating air engine out of wood back in 1985, and posted it here in 2007.

I received a number of enquiries for plans for that engine, but the design for that air engine wasn't really one that lent itself to building from plans. So I came up with a better design for my air engine 2.


My original air engine was made mostly from solid maple. Over the course of many years, some of the pieces in the valve assembly and cylinder warped just slightly, and I had to sand them down a bit to get the engine to run freely again. For this engine, I made the cylinder and valve assembly out of baltic birch plywood - the kind of plywood that is made of layers of birch throughout.

One of the trickiest bits of the engine is making the crankshaft. The main crank for the engine is actually directly off the flywheel, but a secondary crank is needed to actuate the sliding valve assembly. This secondary crank only has a 6 mm throw to it, so I could make it by gluing another piece of dowel to the main shaft. The second piece of dowel is carved to a crescent shaped cross section to allow it to fit neatly against the shaft. After that, I cut away part of the main shaft.

The initial cut away was made with a band saw, but the rest I carefully carved out by hand.


I made a guide to check how much to cut away. The guide is made by drilling a hole on the edge of a piece of plywood, then cutting away half of the hole. I used this to check how much I still needed to cut away as I was whittling the main part of the shaft down.


By pressing my guide firmly against the carved away section, and turning back and forth, I could always see the glossy sections where my guide had rubbed against the shaft, and used that as a guide for where to remove material.


Once I was satisfied that the middle part of my crankshaft was sufficiently round, I made two reinforcing plates to glue on either side of it. I made these by drilling two 5/8" holes with centers 6 mm apart. After drilling the holes, I cut out a small rectangle around the holes, and glued it on to the crank. Gluing the pieces on was a simple matter of sliding it on from the ends of the crank.


The finished crankshaft (after varnishing)


The crankshaft bearings blocks are made of two pieces. To make sure the holes were all lined up perfectly, I clamped the two halves of the bearing together, and then drilled the screw holes through them.


After screwing the top part of the bearing block on, I drilled the shaft hole through both parts. I used my biggest 5/8" drill. That is, the one of my 5/8" drills that seems to drill 5/8" holes that fits 5/8" dowels the loosest, so it's a tiny bit over 5/8". After drilling the holes, I finally cut out the whole bearing block with a band saw, and rounded the corners on it.


I used the same approach for making the holes in the connecting rod for the valve slider. First screw the pieces together, then drill a hole in the assembled connecting rod.


I did end up tweaking the bearings just a little bit, by carving out a very thin layer from the insides with a carving knife. I had to do this again after I varnished all the parts, as the varnish added a little bit of thickness everywhere. But I wanted to varnish even the bearing surfaces, so that when I'd put a few drops of oil on them, hopefully the oil would not soak into the wood too much.


The "cylinder" and piston are simply made square. To make them round would perhaps look more realistic, but I doubt I would have been able to make it very accurately. Also, I would have had to use a piece of solid wood to cut it out of, which would then be subject to slight warping over the years.

There are no piston rings or seals around the piston, so there is quite a bit of "blow by". But this engine is not designed to be very powerful or efficient, so it's OK. In fact, ideally, there would be a slight gap around the piston to reduce friction, on the order of about 0.1 mm. I cut the piston to have no clearance at all, and then sanded it to fit. This was an iterative process.

There are no gaskets in the assembly. Just screwing the pieces together closes the gap sufficiently to reduce leakage to acceptable levels - certainly, there is much less leakage around the cover than there is around the piston.


In the previous photo, you can see holes in the back of the cylinder, which are for the air inlets. The air inlets for the piston need to be towards the ends of the piston, but the valve assembly needs the inlets together, so an internal channel is formed between two plywood parts, by carving a cavity out of the plywood. I just roughed these out with a big forstner bit - the cavities are not visible with the engine assembled, so it's not critical.


This photos shows all the pieces of the piston and valve assembly. The two holes in the front-most piece of plywood are the air inlet and exhaust. By changing which inlet one blows (or sucks) on, the engine will run in the opposite direction.

All the pieces of the valve assembly are varnished. To keep the varnish smooth and level, I scraped the varnish between coats. A bit of light sanding was required after all was done as well to get the valves sliding easily again.

The whole assembly is put together with 3/4" #4 wood screws, 38 screws in all.


For the bearing on the crank, I used a 1.5" long screw with a shank with no thread on it. I had to cut the end of the screw off so it would not stick out of the other side of the flywheel by too far. I actual found an old style wood screw with a thicker shank in my collection (screw on the bottom). Newer screws have a shank that is just thinner than the thread, which would mean that the connecting rod would have a little bit of play on the shank.


The piston end of the connecting rod is joined to the piston shaft with a simple steel pin, which is just a cut off nail. The hole in the piston shaft is drilled slightly undersize so that the pin sits firmly in the piston shaft. The holes to the connecting rod are slightly oversized, allowing the connecting rod to pivot freely on the pin.


The whole engine mounts on a piece of plywood.

I made the flywheel as large as I could for this engine, which required cutting a slot out of the mounting plate for it to protrude into. I was about to start chiseling the slot for the flywheel in the base plate, when I realized that my flywheel was just a little smaller than a skillsaw blade, so I just made multiple cuts with a skillsaw blade in the plywood to carve out the cavity. I made these cuts by clamping a block to the fence to keep the plywood from sliding backwards, and just cranking the spinning blade up into the plywood.


I built the whole engine, and made sure it ran smoothly before varnishing all the pieces. The photo at left shows the pieces drying after I brushed on the final coat.

Varnishing the engine necessitated further tweaking to get the engine to run smoothly again. But with this engine essentially being a toy, there is an expectation that it would get handled a fair bit, and so if it's varnished, it's much easier to get it clean again. Also, the varnish I used is fairly slippery, so it should make the engine run more easily.


The varnish itself however was not quite slippery enough, and I ended up oiling the crankshaft a little bit, to keep it from squeaking. In my previous air engine, I used axle grease on the bearings and no varnish. This made them run really smoothly, but the whole thing is a bit messy. The 3 in one household oil is much cleaner.


After Donald Zorn built his engine and was able to spin it up to 1050 RPM, I experimented to see how fast I might be able to spin my air engine up to.

I have drawn up some very detailed plans for this engine, which should make it much easier to build such an engine if you wish to do so.


Air engine
plans for sale


Donald Zorn's air engine
Rodger's air engine
Gábor Németh's
gas engine model

Back to my Woodworking website