Suppose we know the 3D CAD model of a commercial propeller. Starting from that, we do the slicing to get the profile of airfoil being used to generate the complete shape of the propeller. How do we know the kinds of the airfoil from those section? Knowing there are thousands of airfoil profiles out of there, and we don't want to compare it one by one and mathcing the shape

  • $\begingroup$ Good reason to dig in and learn about how certain airfoil shapes give better L/D efficiency for a given range of AoA and rotation velocity. $\endgroup$ Mar 18 at 1:56
  • $\begingroup$ Right! Airfoils distribution, twist distribution, and chord distribution all play their role in a propeller generation $\endgroup$
    – IFX21
    Mar 18 at 2:28

1 Answer 1


You don't.

If you had a guess as to the family of airfoils -- and that family was parametrically defined -- you could write a program to do a nonlinear least squares fit of that family to your points. However, if the airfoil was in-house or custom designed, then that would only find the closest airfoil in the family to the one that was used. It probably isn't worth the trouble.

In the old days, most propellers used a Clark Y airfoil (along with everything else). The NACA 16-series airfoil series was specifically developed for propellers. The NACA 65-series airfoils are also frequently used on propellers.

If you have a CAD model that represents the foils accurately (often CAD models of propellers don't have real foil properties accurately represented) -- such that you can extract points and believe them, you should just use those points.

I assume you're after sectional aerodynamic data (lift, drag, moment) of the airfoils. If you have high quality points, just perform a 2D analysis of the points and use your own computation. That will work no matter the airfoil source.

  • $\begingroup$ Good point! I just feel curious whether we could somehow "decode" the propeller's airfoil being used and remake it with our own manufacturing capability $\endgroup$
    – IFX21
    Mar 17 at 10:53
  • $\begingroup$ You could -- take the points from your slice and fit a smoothing spline to them. You can use that curve without knowing where it came from (if the points are accurate). If you really want to know where they came from, you can do the least squares fit to the airfoil family. But what does knowing that the airfoil at mid-span was a NACA 16-408 get you vs. already having the curve and points? If you really want to go down that path, I can help you more. But you must decide if it is worth the hassle. $\endgroup$ Mar 17 at 16:39
  • $\begingroup$ Oh I see.....I think the least square fit just enough for this case $\endgroup$
    – IFX21
    Mar 18 at 2:31

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.