In the previous installments in this series, we built most of the major pieces of our little wobbler steam engine. We just have a couple of accessory parts to make, and then we’ll be ready for fitting things together.
The second-last part we need is the tension adjustment knob for the cylinder pivot. This allows us to achieve a balance between steam-seal at frame/cylinder, and still allowing smooth movement. As is tradition, let’s take a look at the drawing first.
The last step we want to do is knurl the outer diameter. Knurling is a machining technique to create a 3D pattern on a surface so that human meat sticks can grasp it easily. A special knurling tool is used, which generally consists of a pair of wheels, each with half the pattern on them. You apply pressure to the stock with these wheels, and they form the pattern as the stock turns between them. Knurling is a forming operation, not a cutting operation, which means it requires a lot of force. Cutting knurlers do exist, but they’re hard to find and expensive.
Getting good knurls is one of the trickier lathe skills, but there are basic things you can do to jump ahead on the learning curve:
The final (yes, final!) piece we need to make is a fitting to attach our air compressor to the steam inlet. We’re going to run this engine on compressed air, so we need a way to get the air in there. To the drawings!
As usual, we start with a piece of stock of generous length, and face the end.
This part has a lot of narrow ring-shaped features with sharp transitions. Making features this small can be tricky, but a cut-off tool is your friend here.
Okay, let’s do some fit-up! Even if you make all these parts perfectly, there’s still going to be some tweaking required. Let’s put it all together and check for binding.
As you assemble, apply a little bit of 3-in-1 machine oil to any surfaces that move against other surfaces.
Now we can do our fit-up. The whole assembly should spin freely a couple of times if you give the flywheel a twirl. The cylinder head is not installed, thus we have no compression, so there should be nothing impeding movement. Odds are something is binding though, because physics is hard and the real world laughs at 3D models. It can be difficult to figure what is binding though, because we have a lot of moving parts here. Let’s look at all the places you should check. Much like debugging software or electronics, you want to test moving part relationships in isolation first. Make sure the piston moves smoothly in the cylinder by itself, to eliminate that variable, for example.
Once everything is running smoothly, we can install the cylinder head.
We’re almost ready to run this little guy! We need a system to deliver air at an appropriate pressure. Garage air compressors generally don’t go much lower than 20-30 psi, but we only need 5-10 here. A low-pressure regulator and gauge are needed. The regulator is easy to find, but a low-pressure gauge is trickier. I’m using a tire-pressure gauge intended for garden tractors and other small soft-tired things. I found it at an auto parts store.
Does it run?
Huzzah! It may need a little spin to get going, since this isn’t a self-starting engine design. If it won’t run, make sure you’re spinning it in the correct direction, and when in doubt turn up the pressure. Many small flaws in the craftsmanship can be compensated for with more pressure. As the engine runs-in, it will get more efficient and require less pressure as well. Ideally, an engine like this shouldn’t need more than about 5psi, but if it runs on 10psi, that’s nothing to be ashamed of.
Speaking of shame, a final exercise that I thought would be fun is to compare this engine to the first version of it that I made. I feel like I made more mistakes on the second one, but I honestly don’t know which will run better. In order to be all sciencey about it, I went back to the lathe and whipped up a little tee pipe for my air system. I want to supply the same pressure to both engines and compare how they run in real time.
The final step in making a fitting like that is an optional one. After the silver-soldering is done, you then need to run the drill through it again to clear all the clumsy excess solder you used that plugged the passage. Then, break the drill bit in the process, and then also break a larger drill bit that you use to try and free the broken chunk of the first one. Now that you have two pieces of hardened tool steel blocking the passages in the tee, the part is thoroughly ruined and you can finish up by making the whole thing a second time. Again, I must stress those final steps are all optional. I opted to do them, but you should do what works best for you.
Okay, now that we can supply equal parts of Science Air™ to both engines, it’s moment of truth time. Let’s fire ’em up!
The results are quite interesting indeed! The newer engine (with the black flywheel) starts a whole lot easier. However, the older engine runs on much lower pressure. As I turn down the pressure, the new engine dies at about 7 psi, despite my attempts to give it a second chance. The older engine, amazingly, happily puffs away all the way down to 3 psi. The reason for the difference doesn’t show on video, but is clear in person- the piston fitment on the new engine is looser. You can feel leakage around it, which you cannot on the older engine. I got a better fit the first time around, despite having zero experience at the time. It just goes to show that learning is a non-linear process. The looser fitting engine starts much easier, because it has less resistance, but requires more pressure to run because it’s sealing poorly.
Well, that wraps up this engine project. We have more steam related things in the pipeline, as well as more electronics and other classic Blondihacks-style stuff coming. Don’t touch that dial! Thanks as always to my Patreon patrons, who make this blog possible.