A new airfoil mean line.

Discussion in 'Aircraft Design / Aerodynamics / New Technology' started by bifft, Aug 27, 2011.

Help Support HomeBuiltAirplanes Forum by donating:

  1. Aug 27, 2011 #1

    bifft

    bifft

    bifft

    Well-Known Member

    Joined:
    Apr 17, 2011
    Messages:
    208
    Likes Received:
    75
    Location:
    Utah
    Been studying up on airfoils, reading the books, etc. Thinking about simple airfoils for low performance aircraft. Most of my designs (none of which have progressed beyond the X-plane model) just use the old NACA 4 digit series, but I have been thinking about moving a little beyond.

    The other day, had an idea on changing the mean line. Because the NACA 4-digit mean line is two parabolic sections, the slope is increasing all the way to the trailing edge. This means you end up with fairly high moments. The NACA 5-digit uses a straight line from the max camber point to get low moments, but at the cost of having a sharp stall due to the kink (discontinous 1st derivative) in the mean line. So, my idea is to use a hyperbolic mean line after the max camber point. For the front, use the same parabola as the standard 4-digit, then for the back use a curve of the form a/x+bx+c where a,b,c are chosen to have the peak at the max camber point, then back to 0 at the x=1 position.

    Worked this up in the attached spreadsheet. Just using the y+-thickness for now, instead of perpendicular to the mean line as the real NACA method, so that is also a little diffrerent, but using the NACA 4-digit thickness profile. Testing with javafoil (as it is free and easy to use) shows my foils as having slightly lower drag, the same Clmax, and about 25% lower moment than the same parameters for the NACA 4 digit.

    I know that what I'm doing is basically 1930s tech, and there are many better airfoils available now. But, I want something that is less critical to exact shape (working with wood or Al instead of composite) at low Reynolds numbers (1-3x10^6). Just thought I'd throw this out here for the many more knowledgeable posters to take a look at. Thanks for your comments.
     

    Attached Files:

  2. Aug 28, 2011 #2

    Autodidact

    Autodidact

    Autodidact

    Well-Known Member

    Joined:
    Oct 21, 2009
    Messages:
    4,531
    Likes Received:
    799
    Location:
    Oklahoma
    The way you're measuring the thickness sounds similar to Ribblet's method. I'm not really an airfoil geek but using a hyperbolic in place of the parabolic seems brilliant to me especially if the numbers are similar in real world application.
     
  3. Aug 29, 2011 #3

    pwood66889

    pwood66889

    pwood66889

    Well-Known Member

    Joined:
    Feb 10, 2007
    Messages:
    1,324
    Likes Received:
    120
    Location:
    Sopchoppy, Florida, USA
    Not a geek either. Had good luck with 43013 so far (see avatar). Do get Ribblet's texts - from the EAA or where ever. Thank you for a very informative post.
    Percy in SE Bama
     
  4. Aug 31, 2011 #4

    bifft

    bifft

    bifft

    Well-Known Member

    Joined:
    Apr 17, 2011
    Messages:
    208
    Likes Received:
    75
    Location:
    Utah
    Yes, I am doing the thickness the same as Ribblet. Reading his work along with Theory of Wing Sections was the main inspiration for this.

    No "real world" testing, just simulation so far. Further playing around leads me to think that the Clmax might be a little lower than the same designation on the NACA main line, but you can slightly increase the camber to get around that, still with lower drag, moment than the standard line.

    Would be nice to get some real world testing, but that is harder. I doubt I could build a wind tunnel to get the reynolds numbers that would be reasonable for a full size plane, smaller scale means you have to go faster. Measuring the drag accurately, when it is only about 1% of the lift would be tricky.

    I tend to trust the simulations for low-medium angles of attack, but not near or at the stall. This area is also the part most sensitive to Reynold's number differences. The goal for this foil is to give slightly better L/D, including trim drag, but keep the gentle stall of the NACA 4-digit series. So the stall behavior is important.

    I suppose the best test would be a large scale model on a pylon attached to my truck, Rutan style. Maybe after I finish my RV. (Should be in the garage working on it now).
     
  5. Aug 31, 2011 #5

    ClippedCub

    ClippedCub

    ClippedCub

    Well-Known Member

    Joined:
    Jul 15, 2011
    Messages:
    105
    Likes Received:
    5
    Location:
    aa
    eom
     
    Last edited: Apr 16, 2012
  6. Sep 1, 2011 #6

    bifft

    bifft

    bifft

    Well-Known Member

    Joined:
    Apr 17, 2011
    Messages:
    208
    Likes Received:
    75
    Location:
    Utah
    Thanks for that, that was an interesting read. Based on thier forumula, as I am pretty close to the standard NACA lines for the forward section, I can expect the same type of stall behavior. I did notice that the 230xx foils stalled differently than this formula would predict tho.
     
  7. Sep 1, 2011 #7

    ClippedCub

    ClippedCub

    ClippedCub

    Well-Known Member

    Joined:
    Jul 15, 2011
    Messages:
    105
    Likes Received:
    5
    Location:
    aa
    eom
     
    Last edited: Apr 16, 2012

Share This Page

arrow_white