NACA Four-Digit Airfoil Stories

[Howard Rush]

I'll take Serge's bait.

I read an old NACA report that essentially said that there was something magic about the NACA airfoils that made them better than airfoils that came close to to them in shape, but weren't exactly the same.  I believed it religiously until Gary James straightened me out some years later. 

The Nemesis I combat plane had an NACA 0016.5 airfoil with the max thickness squished forward linearly to 20% chord.  The Nemesis II had an NACA 0016.5 airfoil with the max thickness squished forward linearly to 25% chord. Based on scant testing, I thought it was better than the 20% version, so I built a lot of the 25% version. Gary used XFOIL to show that the Nemesis I's airfoil was better.  Gary also showed me how to derive airfoil shapes using the NACA four-digit formula.  I used that to draw the airfoils for the Rushpuppy, my favorite combat plane.  It was a Nemesis II airfoil with continuous derivatives.

I have used the NACA formula and fiddled with the inputs to match given airfoils.  That gives formulas that can go to a laser cutter and make very accurate ribs.  It's a math version of Bob Hunt's Lost Foam process.

At least one world champion stunt plane was built using ribs derived and plotted from the NACA formula before laser cutting was available.

"Laminar flow" airfoils were all the rage after WW2.  They probably influenced the sharp-LE airfoils popular in old stunters.  They made awful airfoils for CL stunt and combat airplanes.  The designer of the Swoop and Super Swoop combat planes told me that he used a "laminar flow" airfoil on the Swoop on the advice of an aeronautical engineer.  It was no good.  He switched to a good airfoil for his Super Swoop.  When Sterling kitted it, they reverted to the earlier, substandard airfoil. 

[Dennis Adamisin]

Hi Howard
Thanks for sharing - combat model design is foreign to me but not underappreciated.  Couple questions: did you try the 30% high point NACA's?  As you moved the high point forward you/Gary determined it was better - how so?  Tighter turning? Faster?   Holding energy thru manuvers?

Thanks in advance...

[Howard Rush]

We tried an NACA 0018 and NACA 0015, but didn't compare airfoils very scientifically.  I must have got the notion to move the max thickness forward from an old NACA report, but I forgot which one.  All I remember at our Reynolds number is the one I compared with XFoil here: http://stunthanger.com/smf/index.php?topic=24769.0 .  Airplanes using the good airfoils turned tight and didn't slow down much in maneuvers.

[phil c]

This goes back a bit, but my son did a science fair project on combat wing airfoils.  He built 4 planes.  They all had the exact same layout and stab size, with the tip 75% of the root chord and the spars straight at 25% chord.  The control used what we were using at the time on competition planes- an elliptical section leading edge transitioning at the spar to a straight to the trailing edge.  The second plane used a pollywog Eppler airfoil(designed for extra lift on glider tails, recommended by Steve Sacco), the third had the highpoint pushed back with a pronounced curve at about 50% chord going back to the TE(similar to many F2D airfoils at the time) and the fourth was the same as the first, but with a much more streamlined shape(it still had a decent LE radius) but more streamlined than an ellipse.

Testing included level flight speed, time to do 4 figure eights, and a measurement of the width of the figure eights.  The results were interesting.

All the planes had virtually identical speed.  Any differences were due to slight variations in needle valve setting and wind.

The Eppler airfoil seemed to get the figure eights about 10% narrower.

The biggest surprise was in the eights.  The Eppler foil was a few tenth's of a second faster.  But the biggest surprise was that neither one of us could do a consistent, low level figure eight to measure with the flatback airfoil.  We both ended up crashing several times because the plane would not turn consistently at the interesection.  You'd start to change directions and it wouldn't do anything for a moment and then suddenly go to full control the other way.  The stock wing seemed a little steadier and more pointable, with the other the Eppler and the sharp leading edge a little edgier.

Gary James did test 4 combat airfoils with X foil.  They all were virtually identical results.  Phil Granderson's had a small up kink in the curve at around 7deg AOA which gave it a 5% or so advantage there with a fairly sharp peak lift.  Gary's pet airfoil had the the highest Cl at a couple degrees higher AOA.  My stock airfoil was right in the middle of the pack, but a little lower peak lift.  The Eppler 475 pooped out a couple degrees lower.

The bottom line was that in a useable range up to maybe 10 deg AOA there was only a slight difference, with PG's being just a bit better.  Nobody ever flies a combat plane anywhere near the peak lift, so peak lift and stall characteristics never come into play.  So now everybody in F2D is using the baseball bat leading edge, which I think Rich Brasher came up with on foam planes.  The high point goes as far forward as 17-18% of chord.

Phil C.

[Brett Buck]

But the biggest surprise was that neither one of us could do a consistent, low level figure eight to measure with the flatback airfoil.  We both ended up crashing several times because the plane would not turn consistently at the interesection.  You'd start to change directions and it wouldn't do anything for a moment and then suddenly go to full control the other way. 

    That's interesting. After evaluating many types of airfoils on stunt planes, the only thing that was consistently bad was having a straight-line section i the aft part. In fact, they were almost always evil to fly, even if otherwise good. For a long time I thought it was simply due to the fact that they usually also had corner where it transitions from curved to straight. But even when it fairs in nicely, it still doesn't work. Air just doesn't want to stick to the flat part. In almost every case we had to use wires taped to the surface to try to get even moderate corners with any sort of consistency. But even with no other problems and obviously sufficient lift,  they also tended towards very high hinge moments for reasons I can't even guess.

    Near as I can tell, as long as it's not a perfectly straight line, it doesn't make much difference what shape the aft part of the airfoil is.

    Brett

[Tim Wescott]

    That's interesting. After evaluating many types of airfoils on stunt planes, the only thing that was consistently bad was having a straight-line section i the aft part.

So, just to clarify: even a "pollywog" airfoil with a convex section behind the high point is better than straight?  Or did you stop at straight?

[Howard Rush]

I wouldn't think air would notice anything special about a straight piece, except the discontinuity in curvature, if there is one.  There's no aerodynamic significance of the "high point" either.  I'll sell you vortex generators to fix your flat-in-the-back airfoils.  PJ says to put 'em at the high point.

[Tim Wescott]

I wouldn't think air would notice anything special about a straight piece, except the discontinuity in curvature, if there is one.

I think the perfect title to a book on applied aerodynamics would be Aerodynamics: It's Stranger than You Think.  So I'm gonna believe Brett and Phil's experimental results on this one.  Particularly because Brett said something about trying to deal with said discontinuity.

I keep trying to figure out if this matters for cracked-rib rubber power thingies, and the answer I keep getting is "probably not", because you're not running those at every possible angle of attack.  (And 'cause they're teeny and you want them light and simple anyway).

[Brett Buck]

I keep trying to figure out if this matters for cracked-rib rubber power thingies, and the answer I keep getting is "probably not", because you're not running those at every possible angle of attack.  (And 'cause they're teeny and you want them light and simple anyway).

   And you operate then at the Reynolds number of a flea. It's like flying through mineral oil.

   Brett

[Brett Buck]

I wouldn't think air would notice anything special about a straight piece, except the discontinuity in curvature, if there is one.  There's no aerodynamic significance of the "high point" either.

   I didn't mention "high point" in my response. There is some relevance to it, in the sense that that is where any hope of developing a normal pressure from mere forward motion ends.

    I didn't think there was any relevance to the fact that it was straight instead of curved, either. But in fact it was pretty consistently bad. Even in cases when there was no discontinuity where it transitioned.

   To answer Tim's question, yes, a pollywog seems to work better than a flat section. Al's car-hood experiment seem to be adequate to explain why you don't want to use a polywog on a flapped airfoil, but that's a different effect than we are talking about here.

  I'll sell you vortex generators to fix your flat-in-the-back airfoils.  PJ says to put 'em at the high point.

      When we "fixed" them with turbulators made of old flying lines taped to the surface, we put them at the high point, too. Trying to drag the air around the corner. You may scoff but it took several airplanes from unflyable to merely unpleasant. If you don't believe me, I will bring Uncle Jimby's "Frankenstunt" to the NATs and let you give it a try at a 2500 foot density altitude with and without the wire. You will believe me afterwards.

   Brett

[Chuck_Smith]

   I didn't mention "high point" in my response. There is some relevance to it, in the sense that that is where any hope of developing a normal pressure from mere forward motion ends.

    I didn't think there was any relevance to the fact that it was straight instead of curved, either. But in fact it was pretty consistently bad. Even in cases when there was no discontinuity where it transitioned.

   To answer Tim's question, yes, a pollywog seems to work better than a flat section. Al's car-hood experiment seem to be adequate to explain why you don't want to use a polywog on a flapped airfoil, but that's a different effect than we are talking about here.

      When we "fixed" them with turbulators made of old flying lines taped to the surface, we put them at the high point, too. Trying to drag the air around the corner. You may scoff but it took several airplanes from unflyable to merely unpleasant. If you don't believe me, I will bring Uncle Jimby's "Frankenstunt" to the NATs and let you give it a try at a 2500 foot density altitude with and without the wire. You will believe me afterwards.

   Brett

I very much believe it. At the ridiculously low Re fly at there is insufficient energy in the flow to keep it attached. The vorticity added to the flow by the turbulator adds energy to the flow to help keep it attached. The cost is drag, but (and YMMV) for me drag ain't too bad a thing on a stunt ship.

[Howard Rush]

That thing about wire taped to the wing has too many credible witnesses to elicit much scoffing.  Bewilderment, maybe, but not scoffing.  According to XFoil, natural transition happens farther forward than the wire--at least for an Impact--so I still can't imagine what the wire does.  

Nevertheless, I recommend a couple strings of artisanal vortex generators for everyone on your Christmas list.  

  I didn't mention "high point" in my response.

Nor stagnation point.  It's me going off-topic to grind axes again.

[Brett Buck]

That thing about wire taped to the wing has too many credible witnesses to elicit much scoffing.  Bewilderment, maybe, but not scoffing.  According to XFoil, natural transition happens farther forward than the wire--at least for an Impact--so I still can't imagine what the wire does.  

Nevertheless, I recommend a couple strings of artisanal vortex generators for everyone on your Christmas list.  

Nor stagnation point.  It's me going off-topic to grind axes again.

     It's important to remember that "functionally adequate" explanations, whether or not they are accurate on close engineering scrutiny, are frequently sufficient for any practical purpose.

     Brett

[Howard Rush]

     It's important to remember that "functionally adequate" explanations, whether or not they are accurate on close engineering scrutiny, are frequently sufficient for any practical purpose.

other than giving guidance on where to go from where you are.  "It just works" is probably the most useful explanation for a lot of these stunt phenomena.

[Tim Wescott]

other than giving guidance on where to go from where you are.  "It just works" is probably the most useful explanation for a lot of these stunt phenomena.

In my experience "it just works" helps to keep you from extrapolating yourself right off of a cliff, where an inaccurate but convincing engineering explanation may encourage you to do so.

Not that I've had the experience with anything aerodynamical, but certainly with things electro-mechanical.

[Brett Buck]

In my experience "it just works" helps to keep you from extrapolating yourself right off of a cliff, where an inaccurate but convincing engineering explanation may encourage you to do so.

   People are going to do that anyway.

      Brett