(Mostly) Wooden "Steam" Engine
After thinking about it for years, I finally built a (mostly) wooden “Steam” Engine. It actually runs off air (either blown as from a hair drier, or sucked as from a shop vac), so it should more correctly be called a pneumatic engine, but it illustrates the same principles as a steam engine. Steam would be a bad idea here, as the moisture would make the wood swell, and bad things would happen.
It is a modified version of the one done by Matthias Wandel (http://woodgears.ca/air_engine/). Or perhaps one might say it was inspired by his engine. I kept the same general layout and scale as his, but I redesigned the valve assembly, and I used a rather different axle arrangement.
The main “non-woodenness” of it is that I used a bolt for the axle rather than a dowel, and a pair of bearings to allow the bolt to rotate. It was not feasible to offset the axle as in Matthias’s design, so I used a more normal eccentric.
As illustrated below, the “cylinder” is actually rectangular, and the valve body is formed by several layers of plywood screwed on, each with one or more openings cut into it. It would have been neater to use a router to make the channels, but I just used a drill press with a Forstner bit and a lot of chisel work.
The two pieces of plywood above the valve slider (one of which forms the side of the “cylinder”) contains a pair of air channels from the ends of the “cylinder” to the middles of the slider, and a central channel that forms the “outlet”. (Note that “outlet” and “inlet” are interchangeable.)
The two pieces below the slider contain a single channel that directs air from the inlet to the ends of the slider.
The slider consists of two openings. The operation is illustrated below:
When the valve slider moves to the left, air is directed from the inlet to the left side of the piston, pushing the piston to the right, and the air to the right of the piston is directed out the outlet. When the slider moves to the right, things are reversed:
The opening for the slider is just a hair bigger than the slider—only as much as will allow it to move easily. I trimmed the fixed side pieces to be just thicker than the slider (maybe 0.005 inches?). Then I glued and screwed one in place. I placed the slider and a piece of paper in place, and then glued and screwed the other side piece in. After removing the paper, and cleaning up any squeeze-out, I found that I had succeeded, and the slider was tight but would slide easily.
For the piston, I cut three squares at the same time, and drilled a hole through all three. The two end pieces in the stack became the ends of the cylinder, and the center one became the piston. I lightly sanded each of the pistons four edges, so that it could slide.
For the axle mechanism, I used a carriage bolt for the axle. The head engaged the crank. Then after the first bearing I put the eccentric. This is followed by the flywheel and the final bearing. I left the bolt extra-long so that if I ever wanted to put something on to be driven by the engine, I would have some axle to attach it to.
For the eccentric I wanted a bearing surface that was smooth and round, and which preferably would have reasonably low friction sliding against wood. Also, it had to be easily workable. I originally thought of using a short length of PVC pipe, but then I ended up using an end cut off of some sort of PVC coupling. It was cheaper and was a better diameter.
To make the eccentric, I laid out a circle on a wooden blank the same size as the inside of the PVC. I located the center of this circle and the offset circle for the axle. I drilled a small hole at the center and the appropriately-sized hole (3/8”) for the axle. I then drilled a quarter-inch hole radially from the outside into the axle hole (see the drawing below). I’ll say what this hole was for shortly.
Then I roughed out the outer circle with a jig saw. I put a small bolt through the center hole, chucked it in my drill press, and then sanded it. I first got it round, and then I gradually reduced the size until it just fit inside the PVC.
I needed a way to key it to the axle. Just snugging down a nut on either side could still allow it to shift.
What I ended up doing was to use a 4-penny brad. I had a drill bit just a tad larger than the brad. I figured out where on the bolt the eccentric would sit, and then I drilled a small hole through the bolt, being careful that the hole would be 90-degrees off of the crank angle. I also made a similar hole where the flywheel would go.
Then I assembled onto the axle the crank, bearing, and eccentric. I carefully oriented the eccentric to be 90-degrees from the crank, and then I tightened the nut down on it. Once it was held in the proper orientation, the hole in the axle was visible in the radial hole I had drilled into the wooden blank. I used that hole as a guide to drill a hole through the bolt and into the center of the blank. I could then slide a brad into that hole, which would keep the eccentric from twisting on the axle. The PVC cylinder would keep the nail from falling out.
I did something similar for the flywheel, except that I didn’t have access to the side, like for the eccentric. Here I just put a nail halfway through the hole in the axle, so that it stuck out both sides. I put the flywheel against it and traced out the nail. Then I used a jig saw to cut a slot on both sides of the central hole of the flywheel, to line up with the nail. This allowed me to slide the flywheel over the nail, which then prevented the flywheel from slipping with respect to the axle.
The follower for the eccentric wasn’t as easy. The size of the PVC was between two of my Forstner bits, so I couldn’t just drill it directly. I had to drill it one size smaller and then sand it to make it larger. I couldn’t spin it to make sure that it was staying round, so I just had to be careful and keep trying the fit.
Most of the junctions (e.g. at the end of the crank) are screws, where I used a long enough screw that the rotating part did not have any threads on them. After screwing the screw to the proper depth, I then removed the screw, cut off the excess length, and screwed it back (now that the threads had already been cut).
The exception was the piston rod to crank shaft. That was just a cut down 4-p nail. As I mentioned previously, one of my drill bits was just a tad larger than the nail, and the next smaller size was a tad small. If I had drilled all of the holes with the larger bit, then the nail would be loose, and I would afraid that it would fall out. So I made the outer holes (on the fork) with the larger bit, and then drilled the inner hole with the smaller bit. That hole was a bit too small to get the nail through, so I cut off the head of a nail, chucked it into my drill, and “drilled” the nail through the smaller hole. This enlarged it enough that I could push a nail through it, but with enough resistance that it wouldn’t fall out.
The rest of the construction was fairly straight-forward. I first attached the axle to the base, and then carefully positioned the piston/valve assembly so that the valve slider was centered, i.e. so that as the eccentric turned, the slider moved the same distance away from the center. Then I measured and drilled the piston’s connecting rod so that the piston would be roughly centered in the cylinder. It wasn’t as critical that the piston be centered as that the valve slider be centered.
I shouldn’t have been, for some reason I was surprised when I first put the shop vac nozzle in place, and the engine actually ran.
I found that it actually ran at a good clip. I don’t have an accurate way to measure its speed, but looking at video footage seems to show that it runs between 900 and 1000 RPM.
One issue is that the engine is totally unbalanced. So with the piston going back and forth 15 times a second, it vibrates rather a lot. I thought about adding some counterweights to the flywheel, but particularly as the flywheel is on the other end of the axle, I’m not sure how well that would work.
Given that this project was an engine, and the purpose of an engine is to move, I figured that still pictures would not do it justice. So here is a short video of it in operation.