Best running ever.
The trick is that it only starts up at room temperature, and then runs until it heats up too much. It will run about a minute and stop. After that, it tries to turn over, where it will run for a few seconds at a time. It might be that the best position for that alcohol lamp flame is moving, but knowing where to put the fire is not exactly documented. Nor is there anyway I know to calculate it. It's pure trial and error.
I frankly don't know how much more can be learned out of this. It's possible a slightly bigger piston would get it to the state I've seen other videos, where the cylinder has to be preheated with the lamp or a torch, and then it runs solidly. It's also possible the issue is something else (that I don't see or understand). It might be time to finish the hardwood base it's going to be mounted to and call it an engine.
If you had a high-frame "bullet-cam" I think you'd see it right off, but just a cursory look tells me that it's failing when the flame is more outside and up, and when the flame is being sucked into the engine, it runs better.
ReplyDeleteKinda like a revolver with too much cylinder gap, and all the bang leaks out the sides, and the bullet just kind of plops on the ground, instead of heading downrange.
Note also that after you started it, and then pivoted the base a bit, it altered airflow on the flame enough to blow the flame out of the draw gate, which then induced progressive failure.
DeleteThat's a finicky little beastie.
They are very finicky little beasties.
DeleteThey're a historical oddity and relegated to things hobbyists build because they're low power and just a milepost in the evolution of the engine into something useful. There's no real compression cycle. The expanding flame that gets sucked in helps push the piston down the chamber, but then the cylinder drops down to ambient pressure and the atmospheric pressure presses the piston back up.
OTOH, they have a neatness coefficient because they predate the internal combustion engine by lots. The patent for this sort of engine goes back to Henry Wood in 1758. Otto didn't develop the four stroke, internal combustion engine until 1876.
If I had to hazard a guess from up here at the 30,000 foot level: if it runs ok at room temperature and not when it's hot, then increase the area of the cooling fins.
ReplyDeleteThis is hard to do at this point, I realize, but if it's going to be a desk ornament anyway, you could turn very thin slots in the centers of the fins, down to about 1/2 of the depth between them.
Or you could get more completely insane and add liquid cooling...
Or you could get more completely insane and add liquid cooling...
DeleteSo guess which one I was thinking of.. ?
If you said, add liquid cooling, congratulations.
There are similar engines that use the flywheel's energy to turn a little fan. Seems to me you'd have to be careful to not blow out the lamp. So maybe liquid cooling isn't all that crazy. Maybe?
from Aesop:
DeleteWell, if you're going to get that silly (as opposed to building the thing to begin with), you'll first need to add a thermocouple, collect some data, and see what kind of temp range we're talking about, before you decide on how best to manage it...
I like the thermocouple idea, but it should feed a control loop after you've assessed the operating range. Hook it up to a Raspberry Pi and use it to control a robot arm with an ice cube, per Mark (an engine must run without human intervention, after all).
DeleteI have a FLIR infrared camera for my iPhone and can measure the temperature that way. Until the dynamic range gets too big for it.
DeleteBear in mind, my home field is electronics. One of my first thoughts is a CPU cooler. Those vary tremendously in geometry, but are designed to be mounted onto the metal cap of CPU package. It would be an easy mod to put the cylinder back on the mill and make a flat spot to epoxy down a heat sink - silver-filled, conductive epoxy is widely available for that. There are (mostly) passive ones that use heat pipes to conduct the CPU heat up onto a bigger surface area heat sink that a fan blows over, and there are more-active ones that use a Peltier effect device that electrically cool the chip. Peltier devices are specially doped diodes that get cooler on one end and hotter on the other when you run current through them. You attach the cold end to the CPU, put a heat sink and fan on the other end and they make electronic refrigerators.
Something that ran on the energy the engine produces (little that there is) would have a higher coolness factor. There's another guy's version of this engine that has a little plastic fan that blows across the cylinder from the back.
Another thought that goes through my messed up mind would be to drill a ring of holes through the fins of the heat sink, parallel to the cylinder, and then run copper tubing through the holes. The fins are 3/16 tall, .188, so the pipe would have to be around 1/8 OD. This makes the cylinder look more like a radiator, with fins and small tubes going through the fins. Then I'd pump water through the tubes. Might have to chill the water!
All of them are crazy ideas, but are they crazy enough?
The problem with putting copper pipes through holes is getting them to fit tightly enough that there is good heat transfer. You might slather them with heatsink compound, I suppose. On the other hand, even if it wasn't terribly efficient it would look cool as hell...
Deletefrom Aesop:
DeleteWhen in doubt, Steampunk is the cowbell of the mechanical world.
Always opt for more.
"My engine's got a fever, and the only prescription is more cowbell...!"
https://www.youtube.com/watch?v=Qa9Zg6yGlQ4
Would not the easier way be to apply an ice cube to the far side of the cylinder after about 10 seconds, pull it away after a few more seconds, rinse, and repeat?
ReplyDeleteAdd a water jacket which is open at the top and boils off? Silicone a sheet of something to the top of the fins on every side but the opening?
ReplyDeleteTry freezing the whole thing and seeing if it runs a bit longer, to prove this theory.
ReplyDeleteThat's how I got to where I am.
DeleteI started freezing the cylinder by using "redneck freeze mist" (take a "compressed-air-in-a-can" dust blower and turn it upside down). It ran for the first time. That got me to where I tried to run it at room temp instead of warming it up first. Every video I'd seen of a successful build, they warm it up first.
Some comment to a video of an engine said, "these things either run warm or cold; if it runs cold it only runs until it warms up too much". Seems to be the case.
Sounds reasonable. Why don't you make a model diesel engine next time? We know they run pretty well at all temperatures...
DeleteI'm trying to narrow down the next project over the next couple of weeks around some improvements to my mill and the shop in general. It will very likely be a four stroke internal combustion engine, something more "conventional" than the flame eater.
DeleteIf I could find a diesel that relative newbies can make, that would sure be considered.
RC airplane engines are essentially diesel. There are plenty of kits, but of course that doesn't give the thrill of making each part. On the other hand, you could buy an inexpensive one and make a duplicate by reverse engineering it. Your version could add some flare and pizzazz, like scrollwork and fancy decoration on the casings.
DeleteFor instance:
Deletehttps://www.amazon.com/Four-Stroke-Glow-Powered-Control-Aircraft-Muffler/dp/B000TXUBAW
If your flame pot was mounted on an x/y/z movement with readouts of some sort, combined with temp readings of various parts of the cylinder, you may be able to chart the changing flame requirements to keep it running. With motor drives a system could be designed to move the flame as required.
ReplyDeleteBTW, what has been done for lube of the connecting rod ends? Is it possible the small end is binding as it heats up?
BTW, what has been done for lube of the connecting rod ends?
DeleteThere is none. The instructions in the book (and everywhere else I find) says not to use oil in the cylinder. Dry graphite only. I use a very small amount of light oil on the crank, crank roller and flywheel, but let it dry out pretty fully. Those parts never go into the cylinder.