Hey! This thing is still a Work in Progress. Files, instructions, and other stuff might change!

Radial Engine ( Compressed Air Engine ) Experimental

by ZekeAsakura Apr 24, 2018
Download All Files

Thing Apps Enabled

Please Login to Comment

I see an idea, main is make it larger to become an engine for a bicycle, prop for a sprocket

I wonder if it was possible that you share the STEP files to this project?

I'm having an issue with air leaking out of the 5 vent holes in the cylinders, and can't get the motor to run. Did anybody else have this issue? Closing the vents isn't an option since air needs to flow into the engine, and then vent. The only way that air could leak out of the cylinders is when the crankshaft is leaking. Removing the back cover causes air to leak out of the back instead of out of the vent holes. The crankshaft has a cut out notch to allow air to go into one cylinder at a time. This shows that the crankshaft is still leaking even with close tolerance. Too close tolerance causes a binding condition. Adding an O-ring causes excessive friction along with a binding condition.

Another question, or two actually; in what position did you print the case and what supports did you use if you did at all? And how did you manage to get the crankshaft fit properly in the case?

The case prints upright, and gets filled with support material. It takes a little time to clean up. Slicers have an option for the amount of support material being used just like they have an option for infill density. I kept it at the default setting for a normal amount of support material.

I have several remixes of the crankshaft. It had to be redesigned & resized to fit correctly. I've also experimented with various types of materials such as PLA, ABS, PETG, eALFILL, etc. The crankshaft has a gap all around which varies as it turns. One side is very close while the other sides have more space. Non smooth materials will cause it to bind so only use PLA. I didn't have good results with Inland PLA for the crankshaft since it was a little soft. It was fine for the case as well as other parts. The hex head bolt indentation kept getting stretched. I used some cheap brand of PLA from EBay that was harder & smoother.

The original design for the piston isn't round. One side is a different size. This is why everybody has issues with it fitting correctly, and they have to sand it. I didn't remix the original shape of the piston but I resized it to fit into a 3/8" copper bore. It works a lot smoother but is more work since the copper requires drilling air relief holes, and sanding the rough edges.

I had to remix the size of the crankshaft rods, and the crankshaft master to work with the smaller remixed pistons.

Where are these crankshaft remixes? i cannot fit it in the body

Click the word remixes on the upper right hand corner of this page.

What is the diameter of the case? In Repetier host it looks like it is 40-50 mm in horizontal axis and the scale is 1 (default). Is this size correct comparing to your original model?

The diameter varies at different points. The average size is 45.50 mm more or less. It will be more at the mounting tabs & the air inlet.

I finished printing and assembling the air engine, but I had to do some post processing and made some modifications that are listed below.

-The pistons had to be sanded for a good fit. I would lower the diameter of the pistons by 0.2 mm.

-The case cracked a bit when I screwed in the cylinders, I drilled beforehand to increase the size of the holes, but that did not seem to do it. I would change the holes on the case for the cylinder screws to be 3.0 mm, because the holes will end up being slightly smaller due to volume expansion of the plastic during cooling.

-I used a couple layers of masking tape as a gasket instead of buying flexible filament. I have done this before and it works well for pressures up to 20 psi.

-I used cooking oil to lubricate the pistons. It worked very well, I also tested with oil for rotary vein air motors, and the cooking oil seemed better.

-I resized the crankshaft to fit in the case.

-The holes for the ball bearing were a little small and I use a rotary tool to increase the diameter.

-I did not use tubing to line the holes of the connecting rod, I did use washer on both sides of the connector and did not notice much friction.

I tried this weekend to run the engine, but I do not have much experience with radial engines and did not get it run. Is the back supposed to be sealed for the motor to move? When supplied air to it, most of the air came out of the back. If this is the issue, I will make an attachment that screws onto the back of the case, so it does not need to be mounted. Has anyone else had problems getting the engine to move? how did you fix those problems? I will post my make once I get it to run well enough.


The thin part of the crankshaft is porous since it's so thin so I think that this might be the source of the air leaking out of the rear. This is where there's support material. I don't have software to edit this part of the stl for the crankshaft. I was experimenting with resizing it to add additional layers but this messed up the height.

The reason that everybody has issues with the size of the pistons is because they are not round. One side of the hemisphere is different than the opposite side. It needs to be resized to correct this issue to make it round.

I have not been able to get the motor moving. The back is currently open and I did not mount it to a block or anything. Do I need to mount it to get it moving? Also, does the crankshaft need to be sealed in the case, or can there be a gap?

I've tried several sizes of the modified crankshaft that you made, and the one that I made. It seems like there's always air leaking around the side of it. If the diameter of the crankshaft is too close to the case it will bind. If it has a small gap it will not bind but the air leakage issue gets worse. I think that the rear of the case could use a flat flexible gasket with a hard background to seal it. Could you post an stl for the gasket, and another stl for a hard background? So far two people posted videos of it working, but it was mounted on a slab of wood. I think that this sealed the back to get it running.

As for both versions of the modified crankshaft, I don't know why the original top design is only thick on one side, and has a small piece extended for a screw hole. I printed it upside down, and left support material in place to have a full diameter top, and it didn't interfere with anything. Maybe you could modify another version of the crankshaft to have a full round top?

I'm experimenting with modifying the inner part of the case where the bearing is. I've used a rotary tool to lower it, and tried adding a thrust bearing. It seems like it turns easier, but there's the same issue with air leaking. Check your messages from last week, and let me know what you think.

The piece inside needs a bearing. The housing that I printed didn't have a recess for the bearing. I also noticed that the bolt going through that area slips without adding a nut. Maybe a lock washer or a carriage style bolt might be better.

I added a hexagon recess in the crankshaft, so that it could turn without slipping. I used a M4 hex head bolt (50 mm long). You may need to modify the size of the hexagon depending on the hex head bolt you are using.

Thanks for taking the time to modify the opening for a hex head bolt. Which program did you use? I didn't print your design yet, but I'm wondering if you made the lower area which is very thin a little thicker? I think that this might be the source of the air leak coming out of the rear of the motor. Could you add your crankshaft to the remix section?

You got it! To design the crankshaft, I used SolidWorks. I first used calipers to find the dimensions of the current crankshaft to make my part, and I modified it. I did make a slice of the crankshaft 2 mm instead of 1 mm, so that the part would be stronger, but I agree with your point and changed it back. See the picture for details. I will add this to the remix section.

There are a couple of issues with the design. I printed it upright, and it didn't print some areas thick enough. The size is also slighty off so I resized it, flipped it over, and made a test print. I have to clean up the support material, and see how it fits. I could add it to the remix files if it's good. The hex head cap bolt fits in the top, so that was a great improvement over the round design which has no grip. The old design with the round head could fit a hex head bolt if a nut is used to tighten the bolt. The pressure of the bolt tightening will make a hex shape in the round head since it's plastic.

Hi, I am in the process of making this air engine. I having trouble getting the crankshaft to fit in the case. How can post-process the parts for the best fit?

I printed using PLA and a Monoprive Maker Select Printer (prusa i3 knockoff).


The original crankshaft has an issue with the size. I had to resize it to get it to fit without binding. It's now in the remix section along with a photo of the resized part.

Make sure that you remove all of the support material with a narrow needle nose pliers. The photo that you posted looks better than the way mine came out. What kind of PLA filament did you use? I'm going to print another one.

I also resized the crankshaft to get it to fit. I am using generic PLA manufactured by Inland.

does anyone have a list of the quantities of certain parts that need to be printed

Everything is a single quantity except for the parts on each cylinder. You have 5 of everything since there are 5 cylinders, rods, gaskets, pistons, pins, etc.

I wanted to post an update on the types of bearings which could be used to improve on this design. You could use a larger inner diameter bearing for the output bearing if you want to use a thicker shaft such as an 8/32" shaft instead of an M4 shaft. The bearing number is MR125ZZ. The size is 5MM X 12MM X 4MM. The inner bearing is a 3 piece set, and is called a needle thrust bearing. Two of them have to be stacked then placed between the rotary crank section, and the case. The bearing number is AXK0619. The size is 6MM X 19MM X 2MM. A 4MM thick bearing will work if you could find one like this. The inner diameter could be smaller than 6MM, but make sure that you use a 19MM outer diameter.

Here's an update. I got the engine running today. The problem wasn't a lack of compression, it was end play in the prop shaft which would bind the crankshaft when air pressure was applied. I will post a video as soon as I get a better air line fitting.



Hello Zeke. After many hours I have completed the engine, run it in with a drill to seat the pistons but cannot get it to run on it's own.
I even bored the cylinders and matched pistons to them. All I'm getting is the sound of escaping air. In order to get it to run freely I may have made too loose a fit between pistons and bores. I don't seem to have enough compression.
Any advise you can provide would be greatly appreciated. After all this work I would really like to see it run.


Roger Flagg

Comments deleted.

Hey mate can u send me the updated crank stl as i dont no how to modify the one u uploaded thanks mitch

Hi. Is it possible to add a pulley (either with a groove or teeth for a belt) instead of a propeller? Are the upper pistons (not the lower part which is needed for support) able to be made out of a flexible material like TPE or TPU to make a better seal? Could the pistons have a groove for an O ring to make a better seal?

Hi Brooklyn. I'm going through the prosses to build this myself. Your question about the pulley, yes it is possible to add on but that is something you would have to design yourself. Also, I'm not sure if this will help but TPE is more elastic and TPU is more flexible. Personally, I'm going to try TPU since that's what I have.

I was having difficulty trying to find some of the hardware such as the bushings, metric bolts, bearings, and the hydraulic/pneumatic fitting. The fitting appears to be made for PTFE tubing for hotends. There are several types available online but nobody sells them locally. Home Depot has 3/8" OD X 1/4" ID polyethylene tubing which fits into the bore without any coupling. It's a tight fit. The M3 screws interchange with 6/32" screws, and the M4 screws interchange with 8/32" screws. The printable screws didn't seem to come out correct on my printer since they are so small.

There was one thing which I've noticed in the photos. The STL which is included has no inner bearing in the case. The STL has a built in bushing which is part of the case. The photos show one design with the bushing, and another photo with an inner bearing. That STL file wasn't included with the other STL files. Maybe the original poster will see this, and add it to the listing.

As for the brass bushings, a plain 6/32" screw will fit without a brass insert. There's a tube style compression sleeve with a top hat flange which works if you increase the inner bore diameter in the crank arm, and grind down the length of that tube. I saw two different sizes of small bearings online which I've ordered to see if they fit. The inner diameter of those bearings is very small so it might need a smaller screw or some sort of pin.

Please let me know how the piston works with TPU. I saw somebody mention in another listing for a similar engine that O rings on pistons add resistance to the movement so an actual TPE or TPU piston might cause resistance too but I don't know since I haven't tried it. Please add photos when you make it.

I'm still waiting on bearings and brass tubing but just from messing around with it the TPU doesn't seem to affect the movement but I think that's more of a design thing since the piston can only come in contact with the upper seal and even then I don't think the piston has the travel to do so. Since I have first started this project I have had to print smaller pistons and thinner connecting rods to get a smooth and relatively effortless movement.

The smaller piston might cause air to leak if there's too much clearance. How much did you scale down the piston? Could you post updated STL files of the scaled down piston & connecting rods in a remix once you get it to work properly? Will the connecting rods still have the same diameter openings? I understand that you scaled down the length to make them shorter. Maybe you could add brass sleeves in the cylinder bores since the pistons were scaled down?

I think everything should still work fine because the pistons are only 4-5% smaller and with the connecting rods I didn't make them any shorter I just made them skinnier so they would fit in the smaller slot in the piston. You should be able to do this in most slicing software by changing the scaling factor to 95-96% for the piston heads (for x, y, and z) and around 95% z height for the connecting rod.

I think I found the sweet spot at 97% on the pistons (x, y, and z) and 93% on the connecting rods (z-axis only)

I'm still waiting for parts to come in the mail. Did you get your model working? The inner movable part rubs, and doesn't turn freely. The photos show one design with an inner bearing, but the STL doesn't have a place for an inner bearing. I wanted to try a disc shaped flat bearing. I hope that it could fit.

The piston is too tight, and doesn't move in the cylinder unless I do a lot of sanding. So far it looks like the piston is sanded unevenly with a wider gap in some areas, and a little rubbing in other areas. Could you post remixed STLs for the modified parts? I think a scaled down piston with a groove for an O ring would be better than a plain piston. Maybe a brass or copper inner sleeve for the cylinder bore with a scaled down piston would work better?

Hello Zeke. I am printing your excellent design now and I thank you for all the work you put into this engine .
Although it's not really a complex assembly it would be nice to see an exploded view to see all the part relationships. I'm not sure what CAD program you used to model the engine but for example Fusion 360 will create an exploded view of an assembly. Right now I'm not sure where the bearings go and how the prop spinner is attached. Also where are the brass collars used and what is the purpose of the recently altered piston and crankshaft?
I hope you can help me with these questions and an exploded view would be the easiest answer.
Thanks again for all your work. I will post a picture of the finished engine.


Roger Flagg


J'ai modifié le diamètre du piston afin qu'il fasse 14.5mm et j'ai aussi écarté les parois intérieure du piston qui font maintenant 5.7mm afin que la bielle ne soit plus gênée .

Je suis entrain de terminer d'imprimer ton moteur et je le trouve vraiment magnifique.


oui, effectivement j'ai modélisé quelques pièces du moteur plutôt ajuster pour éviter les fuite d'air et aussi qu'il y est trop de jeux , mais s'est vrai quel selon le réglage de l'imprimante 3D, les pièces sont surement plus dur a assembler!
Donc je vais publier les fichier STEP du piston et aussi celui du vilebrequin/admission pour que vous puissiez ajuster les côtes et aussi peut être mieux régler l'admission ( enfin sa j'y travaille aussi en ce moment , pour augmenté sont rendement et sa vitesse )
Mais toute les idées d'amélioration sont les bienvenue

Et merci j'ai essayer de faire un truc un peu jolie , mais surtout le plus petit possible ( on sait jamais , si j'arrive a avoir asser de rendement , il pourrait peut être servir en modélisme )

Great design, would it be possible to have the case without the writing on the back? To make it easier to print.

Thank you!
I update the case

Looks like there is alot of support material, on the housing, actually everywhere there are tons of support material. Did you use water soluble support material?

i am not using water soluble support material!

but the support this removed quite easily

Good morning dude!

Nice work! You and I seem to be on the same skill level. I am a mechanical engineering inter at a robotics firm. I know that sounds profound but I ensure you I am just a beginner in CAD and I've learned a lot since I've been working here that past 8 months. I've been looking for someone that I can collaborate with on a project, whatever we decide. Just for fun as a hobby.


I'm happy you appreciate my work. I'm not a professional but I try to improve myself every time!

I'm really interested about your project but I want to know more about it. Can you contact me in private to tell me what your project is about?

Thank you for your interest

Hello. This is awesome!

As soon as I can buy some Nylon filament I'll print it and I'll let you know how it went.

Thank you for your comment!

Let me know when you did it, I'm really interested by know how it's done with this material

How are you timing the cylinders?

Tow Time cylinders

what material did you use?

I Used PLA filament but I think if you use a better plastic who is high hot resistant is better.

Amazing, have you had it running? worth a print even if it does not produce useful power!

Yes I had it running but not with any object.

I didn't put the gasket so there were a lot of air leakage. I have to redo the air inlet to gain power.

I'll check it theses weeks