NACA 23015 or equivalent

Discussion in 'Aircraft Design / Aerodynamics / New Technology' started by BDD, Sep 11, 2009.

Help Support HomeBuiltAirplanes Forum by donating:

  1. Sep 11, 2009 #1

    BDD

    BDD

    BDD

    Well-Known Member

    Joined:
    Apr 2, 2005
    Messages:
    388
    Likes Received:
    3
    Location:
    WI
    I like some of the characteristics of this airfoil but not the sharp stall behavior at the usual Reynolds numbers.

    The lift coefficient is excellent for such a low pitching moment airfoil.

    Does anyone have lift-curve, drag, pitching moment, etc, data for this airfoil at low Reynolds numbers? I am interested in R numbers corresponding to 30 to 100 mph with a wing chord of 4 ft. or 5 ft.

    Is the stall behavior appreciably better at low R numbers?

    Does Harry Riblett, or any reputable source, have a development of this airfoil series with high lift, very low pitching moment and with a gentle stall?
     
  2. Sep 11, 2009 #2

    Dana

    Dana

    Dana

    Super Moderator

    Joined:
    Apr 4, 2007
    Messages:
    8,475
    Likes Received:
    2,881
    Location:
    CT, USA
    All in all a good airfoil. The curves show a relatively sharp stall, but the 23- series airfoils have been used very successfully. I'd like to see low Re curves too... I intend to use the 23012 for my ultralight biplane (actually I want the clean stall brake for aerobatics, snaps and spins).

    I suspect that if anything, the stall break will be a little sharper (and Clmax a bit lower) at low Reynolds numbers.

    The 23015 doesn't seem to offer any significant aerodynamic advantage over the 23012, but it would give greater structural strength... which may be why the strut braced Taylorcraft used the 23012, while the cantilever wing Bonanze used the 23015.

    -Dana

    "You sure it's broken? Let me make sure..."
     
  3. Sep 11, 2009 #3

    addaon

    addaon

    addaon

    Well-Known Member

    Joined:
    Feb 24, 2008
    Messages:
    1,686
    Likes Received:
    100
    Location:
    San Jose, CA
    Also, the low moment is a direct result of having the camber forward, where it influences CLmax about as much as aft camber, but influences moment a heck of a lot less. Unfortunately, this has the direct influence on stall that you're seeing -- the fore camber causes high curvature in front, and causes a bubble to form and then detach way forward. The two effects are linked; it's not going to be easy to get one (low moment high CL_max) without the other (sharp 2D stall).
     
  4. Sep 11, 2009 #4

    Richard Schubert

    Richard Schubert

    Richard Schubert

    Well-Known Member

    Joined:
    Feb 5, 2009
    Messages:
    146
    Likes Received:
    33
    Location:
    Pittsburgh, PA
    For what its worth, here is the data as found in the "Airfoil Optimizer" program from Davinci Technologies.

    page1.jpg page2.jpg page3.jpg page4.jpg page5.jpg
     
  5. Sep 11, 2009 #5

    BDD

    BDD

    BDD

    Well-Known Member

    Joined:
    Apr 2, 2005
    Messages:
    388
    Likes Received:
    3
    Location:
    WI
    I'm thinking of 15% or maybe more at the root for a cantilever wing.
    I'm also considering this for a plank type "flying wing" with stabilators or with a short coupled tail for use as an ultralight motor-glider.

    I prefer the tail idea to prevent the possibility of tumbling, for improved control authority, the ability to have some down elevator, control and recovery in a stall/ spin, and for a bit more possible C.G. range.

    "Airplane Design" by K.D. Wood 1941, 5th Edition, indicates a very gentle stall at low R numbers but gives no actual data for the lift cuve, etc. at those R numbers.
     
  6. Sep 13, 2009 #6

    addaon

    addaon

    addaon

    Well-Known Member

    Joined:
    Feb 24, 2008
    Messages:
    1,686
    Likes Received:
    100
    Location:
    San Jose, CA
    For a plank-type flying wing, you'll want positive Cm, which the 230xx family won't give you (they're only slightly negative, but still negative). The 231xx family gives you "theoretical zero" moment (very slightly positive from what I can tell), but isn't really well regarded.
     
  7. Sep 13, 2009 #7

    Norman

    Norman

    Norman

    Well-Known Member

    Joined:
    Nov 29, 2003
    Messages:
    2,877
    Likes Received:
    913
    Location:
    Grand Junction, Colorado
    The stabilators he mentioned are basicaly Junkers flaps that will provide the extra moment. The Mitchell B-10 is a tapered plank that uses 23015 and stabilators and a bit of twist.
     
  8. Sep 19, 2009 #8

    BDD

    BDD

    BDD

    Well-Known Member

    Joined:
    Apr 2, 2005
    Messages:
    388
    Likes Received:
    3
    Location:
    WI
    Richard:

    Will Airfoil Optimizer give data for lower Reynolds numbers?
     
  9. Sep 19, 2009 #9

    Richard Schubert

    Richard Schubert

    Richard Schubert

    Well-Known Member

    Joined:
    Feb 5, 2009
    Messages:
    146
    Likes Received:
    33
    Location:
    Pittsburgh, PA
    No, it gives 3 choices 1 million, 3 million and 6 million. 1 Million is getting pretty low. I set it up for your conditions, I used an wing area and weight that gave a stall lift coefficient of 1.5. If you look at the first page it shows a stall reynolds number of 1.3 million for a 5 foot chord at 25 kts. 4 foot would be around 1 million.
     
  10. Sep 19, 2009 #10

    BDD

    BDD

    BDD

    Well-Known Member

    Joined:
    Apr 2, 2005
    Messages:
    388
    Likes Received:
    3
    Location:
    WI
    The lift break off at the stall does appear to be less sharp than the NACA data shows.
     
  11. Sep 20, 2009 #11

    Birdmanzak

    Birdmanzak

    Birdmanzak

    Well-Known Member

    Joined:
    Apr 20, 2006
    Messages:
    54
    Likes Received:
    0
    Location:
    Sydney, Australia
    Xfoil (here) and XFLR5 (here - also does 3D wings) will give results for whatever Re you want. The accuracy will drop off at ridiculous values, I'm sure, but it should help.
     
  12. Sep 20, 2009 #12

    Norman

    Norman

    Norman

    Well-Known Member

    Joined:
    Nov 29, 2003
    Messages:
    2,877
    Likes Received:
    913
    Location:
    Grand Junction, Colorado
    Be cautious about trusting software results in the non linear range. The theory may be spot on in the linear range but separation makes the problem a whole lot harder and until recently there just wasn't a realistic model and the desk top computers couldn't do it in a reasonable amount of time. Don't trust Xfoil to model a stall very well. Same for XFLR5 because the 2D analysis is Xfoil.
     
  13. Sep 21, 2009 #13

    Topaz

    Topaz

    Topaz

    Super Moderator

    Joined:
    Jul 30, 2005
    Messages:
    13,434
    Likes Received:
    5,141
    Location:
    Orange County, California
    +1 on Norman: I've seen the same thing myself.

    Also watch the quality of your airfoil coordinates when working with XFoil at low Reynolds numbers. And/or be very careful about how you smooth the 'foil shape in XFoil before you make the data run. I've seen XFoil go pretty nuts when given a model that hasn't been properly smoothed, especially at low Re. All sorts of weird things happen - changes in the results over multiple runs with the same input parameters, weird fluctuations in the Cm and Cl curves, etc.

    Most sets of coordinates out there are fairly well smoothed already, but some of the older ones (pre-computer) are surprisingly "in the ballpark" of accurate curves. Or maybe "not so in the ballpark."
     
  14. Sep 21, 2009 #14

    Norman

    Norman

    Norman

    Well-Known Member

    Joined:
    Nov 29, 2003
    Messages:
    2,877
    Likes Received:
    913
    Location:
    Grand Junction, Colorado
    That shouldn't be a problem for BDD. The NACA 4, 5 and 6 digit airfoils are all derived from formulae and there is an app to generate them with a user definable number of panels under the <Foils> menu in XFLR5 so you can get very accurate shapes for those sections.

    The trailing edges of NACA airfoil sections often have finite trailing edges. Xfoil (and AFAIK many other panel methods) hate any sharp corner. You should run <set TE gap> and make sure it's zero.

    Even when the surfaces are smoothed analyzing airfoils below a critical Re often results in spiky polars. I've heard that this is at least partially because the spline algorithm in Xfoil isn't as good as some others. I've even seen it suggested to smooth the finicky airfoil with the Epplar code then bring it back into Xfoil.


    Yeah! Eiffel was particularly sloppy.
     
  15. Sep 21, 2009 #15

    Topaz

    Topaz

    Topaz

    Super Moderator

    Joined:
    Jul 30, 2005
    Messages:
    13,434
    Likes Received:
    5,141
    Location:
    Orange County, California
    I actually got better, more stable results for one of my early reflexed sections by giving it a blunt trailing edge thickness. However, that was probably a special case and in general I agree with you.

    Interesting, and yes, that follows with what I've seen. Nice to know it isn't something I was doing wrong. When I was getting these strange results initially, I ran some NACA 'foils through under the same conditions, and got mixed results - some were fine and others were 'twitchy' again. So the 'threshold' seems to be somewhat geometry-dependant, too. In most of the case I've seen the trouble, the Re has been down around 500,000 or a bit above.

    Good to know. I've also seen some coordinates with what seem to be "planted" errors. For example, the coordinate file on the UIUC website for the Roncz airfoil used on the Genesis sailplane has some sort of problem that gives XFoil fits. I've never been able to get a really good polar out of it. And that one should be very smooth and computer-friendly, considering Roncz develops all his airfoils that way.
     
    Last edited: Sep 21, 2009
  16. Sep 22, 2009 #16

    BDD

    BDD

    BDD

    Well-Known Member

    Joined:
    Apr 2, 2005
    Messages:
    388
    Likes Received:
    3
    Location:
    WI
    The source I quoted earlier claimed that at lower Reynolds numbers the 2300 series airfoils have a more gentle stall. Does this sound correct? I am interested in R numbers near the stall/ min. flying speed of an ultralight.

    I am interested in a plank type wing with either no taper in planform (rectangular) or very little. Maybe this would be enough to produce gentle stall behavior. Thoughts? I could provide some washout if necessary.

    I see that the Mitchell Wing could loose about 300 feet in recovering from a stall. I've also read that they enter spins easily. Maybe the wing sweep and taper aggravate this.

    My interest in this airfoil is partly because it seems to offer a far higher lift coefficient with minimum negative pitching moment. It would seem to me that it could therefore be used to make a more efficient flying wing than you could with an airfoil with more reflex or with swept wings with a lot of washout. I also don't like the idea of tip stalls and some of the other aerodynamic and structural issues involved with swept wings. I like the simplicity of a plank.
     
  17. Sep 22, 2009 #17

    Norman

    Norman

    Norman

    Well-Known Member

    Joined:
    Nov 29, 2003
    Messages:
    2,877
    Likes Received:
    913
    Location:
    Grand Junction, Colorado
    I've heard that too but can't verifie it.



    That seems like a lot. People trying to soar with it often move the CG to far back with predictably bad results. With the CG forward the control response isn't as fast and the span efficiency is low but that's what it takes to get a docile stall in a flying wing.

    I keep hearing that and it's just not true. The 5 digit sections still have some negative Cm that you will have to trim out with your trailing edge surfaces which will reduce CL at all speeds and the pitch control system of planks reduces CLmax. I'm not saying that it's a bad choice but its L/D is not exactly stellar in the first place then you increase drag and reduce lift with the controls. Lots of reflexed airfoils have better L/D even though they have lower CLmax. Here's a comparison of the 23015 with several reflexed sections. I've thickened most of them to 15% (didn't get around to the FX). The real advantage of the 230xx are that they get that performance without having to be as clean as the laminar reflexed sections. So I'm not saying "don't use it" just don't kid yourself that it's going to be much better.
     

    Attached Files:

  18. Sep 22, 2009 #18

    Topaz

    Topaz

    Topaz

    Super Moderator

    Joined:
    Jul 30, 2005
    Messages:
    13,434
    Likes Received:
    5,141
    Location:
    Orange County, California
    Well, planks are easier to design aerodynamically than a swept-wing flying wing (see my profile and the album therein for my project), so that's a point on the plus side. On the minus side, they are prone to a couple of behaviors that, while not necessarily unsafe, do demand more vigilance on the part of the pilot. And they end up with insanely narrow CG ranges, which translate (usually) into very, very narrow ranges of possible pilot weights. There's no way around that with a plank.

    And let's put to rest once and for all the idea that flying wings (or tailless) aircraft of any variety are less prone to stalls and spins than those with a conventional tail. They're not. I'm not quite sure why that myth has been popping up so much lately, unless there's something spreading on some R/C forums somewhere.

    As for which planform to choose for your flying plank, that's pretty much driven by the same considerations as with a conventionally-tailed aircraft. A rectangular-planform wing will have more benign stall characteristics, generally, than something with really significant taper, and washout can help. Won't help any more, being on a tailless airplane, but it won't help any less, either.

    A 300 foot stall recovery is a bit much if the pilot knows what he's doing, and entered the stall intentionally. If the pilot was caught by surprise, I suppose an airplane could take somewhat more, but it sounds a bit much to me, too.

    As for the part about airfoils, heed Norman. He knows what he's talking about. I'll just add that you can't use any airfoil with a negative pitching moment in on a flying plank. You can't use a neutral-moment (symmetrical) 'foil, either. It'll just force you to create the reflex by pulling back on the stick, raising the pitch control surfaces. If they occupy a very significant portion of the center section of the span, you've just effectively reversed any washout in the wing, and that can lead to the very tip stall you're worried about. Personally, I like the Eppler 300 series for a flying plank. Not terrifically-performing airfoils, but thoroughly decent performance and a nice stall.
     
  19. Sep 22, 2009 #19

    BDD

    BDD

    BDD

    Well-Known Member

    Joined:
    Apr 2, 2005
    Messages:
    388
    Likes Received:
    3
    Location:
    WI
    As a single passenger ultralight with little or nothing else being carried, the low c.g. range is less of an issue. I have always assumed that I would align the pilot's c.g. with that of the plane for this reason.

    Did I ever say anything about flying wings not stalling or spinning? Nope. I've been saying that all wings can be stalled. Maybe people are reading Marske limited elevator travel info. into the subject of flying wings and stalls.

    I've read about a 300 ft. stall recovery for the Mitchell Wing and being easy to spin on the internet.

    In the past I've compared the 2300 series airfoils with older reflexed sections. I had assumed with the reflex, that lift coefficient and lift/drag ratio would always be reduced from a very good lift coefficient that you would have for a cambered 4415 section (for a good lift coeff. and gentle stall), for example. This is the main reason for this post....are there better airfoils with good lift and low drag at low R numbers which are not tempermental laminar sections with very good stall characteristic that could be used for a plank type flying wing. Some of the sections that Norman posted look promising. The Eppler sounds interesting too.

    Does anyone have lift, moment, and drag curves for these at low R numbers? I think this would be very useful.

    As I said at the beginning, If I used a 2300 series airfoil it would either be with stabilators or with a "hybrid" short coupled tail which would provide the additional positive pitching moment for stability.

    Thanks for the input.
     
  20. Sep 22, 2009 #20

    Norman

    Norman

    Norman

    Well-Known Member

    Joined:
    Nov 29, 2003
    Messages:
    2,877
    Likes Received:
    913
    Location:
    Grand Junction, Colorado
    Generally I agree with that but a wing with a properly designed external airfoil flap should be thought of as having a reflexed airfoil with a slot at the elevon hinge. This is because the two elements have constructive interference and the minimum drag position of the flap airfoil is ~3 degrees nose down relative to the main wing's chord. Reflex dose sacrifice lift to reduce Cm and this slotted airfoil is similar to regular reflexed sections in that respect but the interaction of the two elements also has some positive effects like slightly better control into the stall. If you don't have them yet I have links to all the NACA reports on their external airfoil flap studies here
     

Share This Page

arrow_white