MakerBot Print is our newest print-prepration software, which supports native CAD files and STL assemblies,
allows you to interact with all your printers via the Cloud, and many other exciting new features.

Download Now


OpenSCAD Turbine Propeller Generator

by CorrugatorSupercilii Mar 19, 2017
Download All Files

Thing Apps Enabled

Please Login to Comment

Is it possible to make the space bellow the propeller blades solid?
Like projecting the blade area down until it touches the support surface or like making the Cura supports solid.

It will make some cool one way rotation propeller which will print nice as well.

My other comment got flagged for moderation (because thingiverse flags comments every time you edit them sigh) so here are the pics to compare the scad output with a real prop. Explanation is in my other comment.

Yes I understood what you meant, and have read the other comments, short answer, it is not really a choice, the implementation is done with a rotation, defined by the number of layers, which might correspond to your printing layers, and the pitch at the edge of propeller sets the division of rotation per layer.

I would very much like others to improved upon what is done here, as the saying goes, necessity is the mother of all invention...

Attached is an e.g. of what I use the script for, the STL is imported to FreeCAD converted solid exported to STEP and reworked from there.

That's very nice, are those fans decorative or functional?

Have you tried using Fusion 360? You can import a mesh (as .stl), turn off design history (to enter direct modelling mode), convert the prop to a solid (using Mesh to BRep conversion) and work on it (or convert the solid to STEP) - all in one program. In fact in the prop for my second attempt the raw STL from openscad had some issues (the hole was covered over by a layer in the center) so I reworked it in fusion.

Thanks for the tip, I've tried it, it aborted, too many facet.

Surprisingly FreeCAD is like the Arnold Schwarzenegger of conversion of complex mesh to solid, my exported STEP was 120MB

There's a particular sequence of steps to follow. I first reduce the mesh to 9000 or so faces before doing the Brep conversion. Fusion 360 does four types of conversion, you have to use exactly the right one.

Rather than explain here, I've made you a screencast showing how I do it. It's about a minute and a half long and I demonstrate how to do the conversion and add the ring on the outside as in the example you posted.

Link here: https://autode.sk/2ASfmm7

Thanks! Very useful processing step indeed!

One last thing, FreeCAD can do 'refine shape' which simplify planar surfaces, which is better for further rework e.g. to make design look nicer with chamfer or fillet... Fusion360 does that?

You're welcome :)

To simplify planar surfaces, you need to:

  1. disable design history (like I did in the screencast)
  2. Switch to the patch environment (click "model" in the top left corner, patch should be the second item in the menu)
  3. In the modify menu, use the merge feature (the icon is a pic of a big square with a little square inside it)
  4. Tick "chain selection"
  5. Click one of the faces on the plane. The rest of the faces on the same plane should be selected automatically since chain selection is enabled
  6. Click ok

It also works on curved surfaces but you have to manually paint the surfaces to merge, it takes a long time, and doesn't always give a good result .

Hey, it's me again. I've started my testing and I believe I've found a bug. I tried to imitate an 8x4 RC prop since that's what I actually have (to compare real-world performance with). I uncommented the necessary lines at the top, and the console reports Pitch_Angle is 45, which seems to be correct.

However, your script appears to be applying that 45 degrees to the tips, then adding a positive twist (making blades more vertical) up until the hub. I believe it should actually it should be the other way round - the angle at the hub of the prop should be 45 degrees, and the twist should subtract from that angle up until the tip.

The twist direction is correct, it's just that you're taking the tips as Pitch_Angle reference when you should be taking the hub as Pitch_Angle reference.

I have attached the openscad output and a real 8x4 prop for comparison. Notice the angle at the hub of the real prop is around 45 degrees and the angle at the tip is around 20 degrees. Your script makes the hub angle almost vertical (90 degrees) and the tip angle about 45 degrees. That very high blade angles of attack means that the first half of your prop blade is statically stalled and so would not generate much useful thrust or airflow.

This is a very neat and useful thing, thank you for posting it! I'm trying to make fan blades to convert a spare brushless motor I have into a personal fan to help with the summer heat. I've tried designing props and fan blades on Fusion 360 (just by guessing and estimating the shape), but when printed and mounted to the motor their performance has been poor. I get more wind and less noise simply using a spare 8x4 wooden prop.

I'm hoping your thing will work a lot better than my attempts to design props freehand (I mean, I'm a mechanical engineer but my knowledge of aerodynamics is very limited). I will generate a design with your script and run a test this weekend and report back.

Do you have any suggestions on settings that would make a good fan? I've gotten familiar with the settings in your openscad script, but I'm not that familiar with propeller aerodynamics so I'm a bit lost on what the effect of changes will be. I'm looking for something around 140 to 200mm diameter that moves a lot of air at low revs and that makes predominantly low-frequency noise. Any help in picking good parameters for my first attempt would be much appreciated. Thanks!


I am not much better in propeller design, however what I done is to go look at application specific design, like desk fan, turbines, impellers and tune the design to fit the angle, number of blades, shape of blade.

After printing many design, the balance of the propeller tends have a great impact, especially printing with support for low blade angles is problematic. In general, i still believe that one should design the propeller to be as light as possible, and that is the key to better performance (for a 3D printed one anyway). It would suggest that larger fan blades, due to imperfection may perform worst than smaller fan blade at higher RPM, even though the air flow increases by the cube of the diameter...

My attempts have had 7, 3 and 2 blades. My last attempt was the worst print, I had a blade thickness of 1.0mm and a pitch of 40 degrees (so a 50-degree overhang angle). It printed quite badly (even got stuck to the raft) and it's too flimsy - in fact when spinning it up I can actually see the blades flex backwards due to aerodynamic forces.

My wooden prop has 2 blades, and I think I can save significant material by not making more blades and instead increasing the diameter of that. I understand what you mean about larger diameter having more chance of being unbalanced, but I can always balance the blades after printing using a prop balancer ;).

Thanks for the help, I will try some designs and see whether 2 big blades perform better than say 5-7 small blades.

Absolutely brilliant work! Thank you for building and providing this.

I keep getting this error. :(

Parser error in line 87: syntax error

ERROR: Compilation failed!

You are a Crack !!!!


Very Very USEFUL

scad file not working - always craches opensad if open it

I get syntax errors on line 87.
Good after I comment out the section.