Is there anywhere where I can find a database of experimental or CFD-obtained NACA series four or series 6 airfoil data?

I'd like to have access to a relatively large set of high-fidelity drag polars/$C_L \times \alpha$ curves in order to train a neural network for a personal project, and I don't have the time to digitalize the plots in NACA's original contractor reports.

Any help is appreciated. =)

  • 1
    $\begingroup$ Resource location questions are explicitly off-topic. Please check the help center for more details. $\endgroup$
    – FreeMan
    Nov 3, 2020 at 19:26

1 Answer 1

  1. The technical report NACA-TR-824 has been digitized. The digitized version includes Cl and Cd for 118 airfoils at 3 Reynolds numbers. To get it, create an account on x-plane.org. (It's been online sporadically elsewhere over the decades, e.g. here. People associated with the digitizing include Gregory Peter, Gregory Siemens, and James Sonnenmeier.)

  2. Convert plots to tables of numbers with Data Thief or WebPlotDigitizer.

  3. At lower Reynolds numbers, http://airfoiltools.com/polar/index lists XFOIL-calculated polars for many of those airfoils, as text as well as plots, e.g. http://airfoiltools.com/polar/details?polar=xf-n0009sm-il-500000:

alpha CL CD CDp CM Top_Xtr Bot_Xtr

-13.500 -0.9606 0.06449 0.06155 -0.0264 1.0000 0.0073
-13.250 -0.9745 0.05737 0.05426 -0.0319 1.0000 0.0074
-13.000 -1.0011 0.04929 0.04597 -0.0377 1.0000 0.0072
-12.750 -1.0177 0.04406 0.04052 -0.0404 1.0000 0.0073

You may need some clever webscraping script to collect them all in a NN-friendly format. Or you could run XFOIL yourself: they describe exactly how they generated these tables.

  • $\begingroup$ Those polars are at very low Reynolds numbers which are often unsuitable for the particular airfoil. Only model airplane airfoils are well represented on airfoiltools. But maybe the classic by Abbot and Doenhoff is a better choice? $\endgroup$ Nov 3, 2020 at 21:53
  • $\begingroup$ Aha! I didn't know that anyone had digitized NACA-TR-824. Yes, better. Propose that as an answer. $\endgroup$ Nov 3, 2020 at 22:45
  • $\begingroup$ I have enough reputation already. Also, it is better to keep all relevant stuff in one answer. So please feel free to add the link to your answer. $\endgroup$ Nov 3, 2020 at 22:46
  • $\begingroup$ What, no upvote so far? This is an excellent answer! $\endgroup$ Nov 4, 2020 at 18:12
  • $\begingroup$ Fear not, google will direct people here in due course! $\endgroup$ Nov 4, 2020 at 19:45

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