Building a Hero's Engine (Aeolipile)




This instructable will show steps for creating a Hero's Engine. Also known as an Aeolipile, which translates to the Ball of Aeolus (the Greek god of wind), the device consists of a chamber that rotates due to thrust produced by tipjets. In my version of the Engine, I used a rounded steel chamber with two curved copper tubes as the tipjets. When the chamber is heated, the substance in the chamber (usually water) is converted to vapor, which then exits the vessel via the copper tubes, producing thrust and therefore the trademark rotational motion of the engine.

Please let me know if there are any ways you think I could improve this instructable, and if you have any suggestions for design or constructions modifications.

Shameless Self-Promotion: If you like it, vote for it in the Make it Move contest! Please?

Disclaimer: Multiple instruments are used within the construction of this project that are dangerous if used improperly. Furthermore, if the project is successful and the end result functions correctly, then it will spew hot steam, boiling water, flame, or any number of potentially hazardous substances. Don't hurt yourself, and its not my fault if you do.
Another Disclaimer (New videos=more uses of the engine=more minor injuries=more safety advice): Because of condensation, the method of filling the chamber, overfilling the chamber, the nature of the apparatus, and more, it is possible (read: likely [read: pretty much inevitable]) that you will have some sort of liquid (rather than vapor) discharge from the tubes. In my Engine, the drips from the tubes land right on the hand holding the torch. Although in my experience, this is significantly more likely with the water-filled vs the alcohol-filled Engine, PLEASE take necessary precautions to keep yourself safe. Furthermore, even though the alcohol-filled Engine is less likely to drip, this is by no means a reason to take fewer precautions. The risk from fire not only exists from the rather conspicuous jets of flame from the tubes, but also from the drips which primarily happen at the beginning (before the flame heats the tubes enough to vaporize them before they reach the ends). In short, make sure the area around the Engine is safe, and make sure you protect yourself adequately. One last word of wisdom: don't try to "pre-heat" the tubes with the torch. It's difficult, unnecessary, and if there is any liquid in the tube at all, the tiny amount that you vaporize will eject the rest of the volume of liquid, most likely onto some place where it will cause you pain. Trust me.

More information on Hero's Engine:

A video of my Hero's Engine filled with water:

Alcohol-Filled Engine (the real reason you should make one of these)

Step 1: Materials and Equipment

These are the materials and equipment I used. I will try to indicate when other tools/materials would work, but feel free to ask or look at the rest of the instructable if you think something will work and I haven't listed it.
1 Sheet 22 gauge steel (12in x 24in)
5 ft of copper tubing (1/8 in) I used cooling tubing originally intended for a refrigerator/air conditioner/something else from Home Depot
String (I used old kite string I had lying around)

Dremel tool with cutting wheel and small grinder
Ball-peen hammer
Vise (both to hold material and as place to work the metal)
Blowtorch (for sweating [brute force soldering] the connections and for operation of the final product)
Aviation Snips (for cutting the steel: very heavy duty, and ideally not ones you mind getting a little dull)
Bag of something impact absorbent and moveable for shaping the metal on (I used charcoal, but anything like sand, gravel, dirt, etc     should work. This will become clearer later)

Step 2: The Chamber Hemispheres

The shaping of the two halves of the chamber was, for me, the most difficult part of the entire process. It requires a lot of patience, and there is a significant learning curve. After I saw that my second one was so much better than my first one, I went back and tried to fix up the first one. It kind of worked. What you are seeking to achieve is two equally sized (especially in diameter, but also, for aesthetics, in depth) hemispheres that protrude from the FLAT metal. This is important because the two sides are soldered together later, and the smaller your gaps are the less irritating that step will be. Anyhow, on to the step by step part:

First, if necessary, cut your steel until you have two 12x12in squares (this is a little large, but personally, I needed the room to work with).
On your bag of charcoal (or sand, or dirt, or gravel, or whatever), place on metal sheet. Find (roughly) the center, and hit it with the ball end of the hammer.
Work in generally concentric circles, continuing hitting the metal until you have an indention that is slightly smaller than how large you want your chamber to be. IMPORTANT: your chamber can only be as large as the largest distance your vise can open.
Open your vise to the diameter that you want your chamber to be, place the indention within the vise, and begin to hit it with the hammer again. Work in one area, slowly rotating the piece, so that you end up with a perfect circle.
Repeat for the second hemisphere
Optional (but HIGHLY recommended): at this point, trim the excess metal from your hemispheres. I used old, super heavy duty aviation snips, and left about 3/4 of an inch of metal. However, I did this AFTER soldering the tubes into it, which made it more difficult than it needed to be. Another note about aviation snips: they exert torque when they cut, which, in mine, led to warping of the metal. I had to go back to the vise and hammer it flat again.

Step 3: The Tubes and Their Placement(or Nozzles, or Tipjets, or Thrusters, Etc)

By the very nature of the engine, there must be tipjets to provide thrust. Unfortunately, this means cutting two (albeit small) holes in one of the chamber hemispheres that is so precious to you. Something to note: you could theoretically finagle the design so that the two tubes came out from the seam between the hemispheres, but that seemed way too hard, and you could also put one hole at the apex of each hemisphere, and have the thrust be perpendicular to the way I did it, but I did not trust that the chamber would be balanced and didn't want to deal with the accompanying problems that that caused. Moving on,

First, choose which hemisphere you like more (which is prettier). Then, drive a nail through it twice, (ironic, I know) making two holes that are close to the rest of the sheet (depthwise, see picture) and across the diameter of the hemisphere from each other.
Using the small grinder on the Dremel, widen each of these punctures until they are the outside diameter of your copper tubing. 
Using the cutting wheel on the Dremel, cut two sections of copper tubing of equal length.
These sections will be soldered into the holes you just made. As I said before, I recommend trimming the excess steel off before you put the tubes in. For trimming instructions, see step 2.
Bend the tubes into whatever shape you wish, but be careful about two things:
Do NOT bend the tubes so much that you kink the tube
Make sure that the tubes will provide thrust (torque) around the same axis of rotation, in the same direction

Design Idea from djokimaki: To avoid/attenuate expulsion of hot, dangerous liquids from the tubes, it may be a good idea to have the tubes extend further into the center and then bend toward what will be the top of the chamber (further up into the hemisphere), so that they have a greater propensity to collect vapor rather than vapor and water. However, due to condensation, this will probably not completely solve the problem, so please take steps to protect yourself. Also, don't try to preheat the tubes, because its difficult, fairly dangerous, not worth it, and chances are that you'll melt the solder connections between the tubes and the chamber and just feel stupid (like I did)

Step 4: Sweating (soldering) the Engine Together

First, sweat the tubes into the holes in the chamber you made holes in (fairly obvious there)
I will give a quick explanation, but for more in depth analysis of the process, I encourage you to look at this wonderful instructable . The general idea of sweating is to use something (a torch) to heat the living daylights out of a piece (or two) of metal, and then squish solder into it, which immediately melts and seals the joint. So:
Stick the tubes into the holes in the chamber. It should be a snug fit, both because it cuts down on unnecessary solder, and because if you're like me, you got really sick of grinding the holes and made them as small as they could be and still work.
Using the torch, begin to heat the area where the two metals join. Two tips here: while possible to do with a pair of pliers, I much preferred using the vise to hold the work at this point. Also, I soldered the connections from the inside, mainly for reasons of vanity and wanting it to look pretty at the end.
After the area is screaming hot, squish solder into it until there is a fully sealed connection.
Repeat for other tube.

Sweating the two chamber hemispheres together was a pain. Two main things to shoot for are a flat, gapless interface between the two sides. Also, keep the work level at this point, so the solder doesn't just leak out the bottom when it melts. 
Same procedure as above, but with a much larger area. Another tip: torch the work from the side that does NOT have your tubes in it, so that you don't melt the solder holding your tubes and create the need for a repair.  

Step 5: Design Recommendations and Other Musings

Design recommendations:
Now, although you theoretically have a functional engine, it works much better suspended from a string. I took some of my leftover copper tubing, cut it into about a 16inch section, cut a notch in each end, and squished it around the edges of my engine, then tied the string to that.

One thing I did after the novelty of steam wore of was I filled mine with alcohol, just for fun. Although great fun, it requires constant application of the torch to continue to vaporize the alcohol. My as yet untested idea is to bend the jets to the point at which they blow fire onto the bottom of the chamber, allowing the Engine to perpetuate itself. 

Please let me know if you have any other modifications, questions, ideas, comments, concerns, worries, cares, or really anything relating to this project. Or relating to a related project. Et cetera.

Also, I will try to get a video of the flaming, alcohol filled Hero's Engine up as soon as possible.

Make It Move Challenge

Second Prize in the
Make It Move Challenge



    • Build a Tool Contest

      Build a Tool Contest
    • Organization Contest

      Organization Contest
    • Pocket Sized Contest

      Pocket Sized Contest

    55 Discussions

    Awesome!! I have that problem too. It can definitely be solved by using very little water, or using rubbing alcohol. It might be able to be solved by curving the ends of the copper tubing inside the chamber upwards, but I haven't tried that yet.

    Thanks for the comment, please let me know if you find a solution!


    7 years ago on Introduction

    Looking at your second musing: It's a neat idea, but I think the perpetuation of the heat source would lose a lot of the propelling energy by bringing the moving source to the center and decreasing the centrifugal force. Maybe if you had something with six to eight legs, two or four reflexively heating your pod and the other four or six propelling the engine.

    6 replies

    I think as long as the tubes still exerted some torque on the engine, it would be ok. One problem that I ran into with the alcohol-filled Engine (which, admittedly, I didn't mention) was that if I mindlessly heated it with the torch, it started spinning fast enough that it actually put itself out at the ends of the tubes. Obviously, the reflexive heating system would impart anywhere near the same amount of thermal energy to the mass of the engine and whatever it was filled with, but it still seems like there would be enough energy transfer to make it viable.

    On a tangentially related note, I would prefer a slower, reflexively heated apparatus as opposed to one that I had to A, get near with a torch while its spewing fire, and B, constantly regulate the temperature to avoid it extinguishing itself.

    One other option for a reflexive heating system would be a small hole in the bottom of the engine, with a very large, sturdy wick. The pure alcohol/alcohol vapor would keep the fluid inside the chamber from igniting (lack of oxygen) , and the outer part would reflexively heat the chamber. However, that configuration has multiple design challenges and risks.

    As always, comments/feedback/constructive criticism much appreciated. Thank you for your comment!

    Have an air inlet along the way, like on a torch head. So the flame hovers in the protected part of the tube while still having air. Or hook it up to an oxygen tank and put some nozzles on the tubes =).

    I tried housing the end of the small diameter tube in a short section of a larger diameter tube, but the effect was just a much lower velocity (if that make sense) flame... more like an oversized cigarette lighter than a propane torch, especially when compared to the original, small diameter resultant flames. This is still an aspect of the project that I'm very interested in, so any ideas on design are much appreciated and will probably be field tested, so long as they are possible for me to make as attachments/non-permanent mods to the engine.

    Thanks for the idea!

    (Also, as entertaining as charging the entire assembly with O2 would be, I don't have any blast shields, or and gimballed nozzles =/ )

    Wow. Didn't even think about it blowing itself out. I bet that was annoying to deal with!

    And no prob. Glad to be of service.

    You wouldn't necessarily need more volume within the chamber. There would be more fuel consumption, clearly, but that could easily be rectified by using smaller openings for your tipjets (aka, tapping the ends of your copper tubing with your trusty hammer). One word of caution on reducing the size of the "nozzles"; as I mentioned above, the engine does have the potential to spin fast enough to put itself out, as the alcohol vapor has to mix with airborne oxygen, and if it spins too fast, the flame cannot keep up with the volume containing a viable alcohol/air mixture. Reducing the size of the nozzles reduces the speed as which the engine will self extinguish.

    Thank you for your comment! Any other ideas you have are much appreciated!


    7 years ago on Introduction

    What if you added an extra 2 tubes off the top down the sides and aimed up at the bottom so it would sustain itself once the alcohol is lit??

    1 reply

    Again, I really like this idea, and have already discussed it at length with dionysis.god and others in the posts below. Any new insights, of course, are always appreciated. Especially if they aren't too labor intensive =D

    Thanks for the comment!

    Yeah, I'm still pondering designs for one like that. With the one I made, the chamber to hold the water is the same as the chamber to be heated, and thus the chamber to eject the steam from. The other version would be really cool to make, but would require a design that would allow the chamber to spin on a stand, and the stand would have to have hollow supports that allowed the steam to flow into the rotating chamber. While this is by no means impossible, at this point its a little bit beyond my skill and design level. But I'm working on it.

    Thanks for the comment and the compliment!


    7 years ago on Step 3

    I think your aeolipile is cool! I also think that if you insert your thruster tubes fully to the center, and bend them so the inside ends are at he very top/center of the chamber; you might be able to more effectively vaporize the liquid thus reducing the amount of spitting dangerously hot liquid! How many RPM can it generate?

    1 reply

    Reply 7 years ago on Step 3

    I do think that would reduce the irritating problem of boiling liquid spewage, but by the time I figured out that it would be a problem, the chamber was already soldered together, and I did not want to crack it apart for more experimental design changes. I don't think it fully solves the problem, because you still have condensation that forms before the tubes are fully heated, but it would definitely attenuate the problem. I will make a notation in the instructable for this advice.
    The RPM is variable on the fluid being used and the heat level. I felt like I had more RPM with the alcohol, but that was probably because it was exciting and on fire. I would say that the alcohol-filled engine gets to about 60 (or a little under) RPM before it blows itself out. In my relatively few tests (I was more interested in the flaming version) of the water-filled engine, I definitely got it above 60 RPM, and probably could have gotten it faster, but I still hadn't figured out a good way to keep heating on the vessel without burning my hand (if you look at the water-filled video on the introduction page, each time the torch flinches/moves away is me getting hit with very hot water).

    Thanks for the ideas!


    7 years ago on Introduction

    Paint the chamber black, put a reflector dish under it and put it in the sun. Put a check valve in the top and mount a water bottle over the top with a hose running into the chamber.

    5 replies

    I like the ideas here, especially the black chamber and reflector dish, but I do see a few problems with the water bottle feed.
    First of all, barring sturdy attachment to a really nice ball-bearing rotational stand, the assembly would have to be suspended, ideally exactly through the centers (for rotational symmetry and its attendant stability), making the water bottle, and more importantly the hose and check valve assembly, load bearing.
    Furthermore, you would have to be able to somehow regulate the flow rate of the water, so that you didn't flood the chamber initially (one thing to note would be that if you had enough water flow, the water out of the tubes would provide the same thrust, but I doubt that's what you're looking for), and so that you got enough water flow through it to matter.
    However, I do think that the black chamber and reflector would work great with just the liquid in the chamber.
    If you get some water bottle feed apparatus working, I would love to see pics/descriptions.
    Thanks for the ideas!


    Reply 7 years ago on Introduction

    No offense intended . But there is no need for a check valve on the water bottle. If you mount the water bottle upside down on top of the engine, simply run a dip tube from the bottle down to where you want the top of the water pool in the engine. Think of a office water cooler, as water is removed air is allowed into the bottle and water out as needed.

    Have a good (Safe) weekend everybody ;-)


    Reply 7 years ago on Introduction

    Unless you increase the heat and restrict the output valves
    and experience
    which is what the check valve is intended to stop, not regulating water coming out into the expansion tank. so you'd orient the check valve that way,
    sorry if that was unclear.
    You don't want hot gas feeding back into the supply and pressurizing it.
    Of course, I'm talking about the ridiculous notion of making it powerful instead of what the OP may intend. Pay no attention at your pleasure.


    Reply 7 years ago on Introduction

    If there is no regulation, it seems to me that its possible for the water to just run out of the tubes, which would provide propulsion, but not in the manner I was looking for. The check valve assembly would also be able to be a stepping stone to the pulsejet idea mentioned in another post.
    Thanks for the comment!