Max Lift Coefficient Comparison

[Howard Rush]

Just for fun, I ran the symmetrical airfoil data from NACA TR 586 in XFoil using the same airfoils and Reynolds numbers (I added a missing 0012 point).  I used the XFoil defaults in Profili, except for the max alpha, which I bumped up to 25 degrees.  Looks like the Voodoo was better than we gave it credit for.

[Serge_Krauss]

Thanks for doing that, Howard. I'll be thinking about whether these data can be used to run corrections to XFOIL values - if I trust the NACA data. I think they were pretty smart and thorough by the time they made #586; so it looks as though XFOIL is a bit optimistic all the way through. Do surface roughness differences factor in here? Making a "baseline" would be useful, and there are troubling qualitative differences here.

SK

[Dennis Adamisin]

Howard
My thanks also.  I have seen something like this in a text somewhere - nacherly I can't find it.  You mentioned the VooDoo, what kind of airfloil percentages were you using in your combat designs?

You have probably seen Dr. Micahel Selig's wind tunnel work on model sailplane airfoils.  One of his big challenges was getting good data at low RN - had to overcome turbulent flow in the wind tunnel (among other things).  The NACA data probably had some issues at low RN too.

[Tim Wescott]

I was wondering how much the wingwazzles in the NACA data were measurement artifacts, and how much were 'real'.

I also noticed in what little of the text that I read the phrase that read more or less "these should provide nice conservative estimates of maximum lift" -- in other words, the crew at NACA may have felt that their biggest imperative at the moment wasn't to find out the most that the airfoils in question could do, but to find out the least that they would do under all conditions.  If I were more interested in flight safety than absolute top performance -- particularly if I were investigating things below what was considered "useful" Reynold's numbers, I might choose to fudge things to the bottom, instead of trying to pull out all the stops to get top performance.

Howard, now that you're retired you have time to get another degree.  Maybe you should move to Urbana, and see if you can get Dr. Selig interested in control line flying.

Clearly what the AMA needs is a wind tunnel facility, preferably one on the west coast.

[Serge_Krauss]

I was wondering how much the wingwazzles in the NACA data were measurement artifacts, and how much were 'real'...

...If I were more interested in flight safety than absolute top performance -- particularly if I were investigating things below what was considered "useful" Reynold's numbers, I might choose to fudge things to the bottom, instead of trying to pull out all the stops to get top performance.

I believe these NACA pioneers did their best, most objective work. During the 1930's they discovered that mounting devices (struts, balances,...) as well as tunnel wall interference affected their Reynolds numbers more than they had originally believed and took precautions to fix that problem, yielding what they called "effective Reynolds Numbers". They published this information, and Report 586 was written after these corrective procedures were applied. I've posted several pieces noting this and affected data on this and SSW's forums.

SK

[Tim Wescott]

Good point.  I'm continually amazed at the quality of work that I see in those old NACA reports.

Still, getting some 'foils into some moving air would certainly shed some light on something.

[Howard Rush]

Thanks for doing that, Howard. I'll be thinking about whether these data can be used to run corrections to XFOIL values - if I trust the NACA data. I think they were pretty smart and thorough by the time they made #586; so it looks as though XFOIL is a bit optimistic all the way through. Do surface roughness differences factor in here? Making a "baseline" would be useful, and there are troubling qualitative differences here.

Gary James observed that the NACA Variable Density Tunnel had a lot of turbulence.  I saw that graph in the stacks of the PU library when I was in high school.  It convinced me to use a fat airfoil (16.5%) for combat.  I think it worked better than the 13% or so Voodoo airfoil, so I'm inclined to favor the NACA data to this particular XFoil run.  The reason I doubt the XFoil stuff is that I pretty much don't know what I'm doing running XFoil.  I got some tips from Gary James, Igor, and another guy who do know what they're doing, so I might take another stab at it.  I don't know about roughness.   Igor also sent me some historical stuff, including a picture of the Variable Density Tunnel.  It's tiny: http://crgis.ndc.nasa.gov/historic/Variable_Density_Tunnel.   For all I know, the guys at the U of Illinois or the guys in Stuttgart have tested these airfoils in their low-turbulence tunnels.  It would make a swell comparison.  I haven't looked.

[Howard Rush]

I have seen something like this in a text somewhere - nacherly I can't find it.

I stumbled across it today.  I'll bet it's the same book.  Hints: it's green, and one of the authors is Chuck Rudner's uncle.

16.5% until 1990, when I switched to a 15% Eppler airfoil based on a wee bit of CFD data.  Gary and XFoil say it was a bad move.

[Howard Rush]

I was wondering how much the wingwazzles in the NACA data were measurement artifacts, and how much were 'real'.

Me, too.  

Maybe you should move to Urbana, and see if you can get Dr. Selig interested in control line flying.

I had a little correspondence and looked into making some stunt airfoil sections for his folks to test, but I'm too lazy to actually do it.  I would like to test some flaps like those on the aircraft held by the clown at left.  CFD barfs on those airfoils, so it would have to be in flight or a wind tunnel.

[Tim Wescott]

Gary James observed that the NACA Variable Density Tunnel had a lot of turbulence.

It occurs to me that just getting measurements in perfect, non-turbulent air, might be doing yourself a disservice vis-a-vis making Model Aviation's Best Airfoil.  Worse, it occurs to me that you'd want to get your measurements in the kind of turbulence that you'd be experiencing in flight (big whooshy leftover maneuver turbulence, choppy little prop turbulence, sideways-blowing turbulence from the wind going over the judges, etc.).

[phil c]

The one thing I noticed in the graph is that the best airfoil changes greatly with Reynolds number.  Most of our stuff runs under 1,000,000, so thicker airfoils would be better.

The other thing is wing loading plays a roll.  Thin wings definitely run out of lift sooner than thicker wings.  Ain't nothing worse than an overweight VooDoo.  Nice thing about a Nemesis is it is hard to build one as heavy as a light VooDoo.  It is VERY hard to build a VooDoo as light as a run of the mill Nemesis.

It would be interesting to build the VooDoo leading out of very light cross-grained balsa with a carbon fiber cover.  Maybe with some extra CF toward the middle, so it doesn't break there.

[phil c]

I dug out a graph Gary J sent me a few years ago.  He compared his airfoil, PTGrandersons, Nemesis II, Gotcha 460, and the Eppler 475.  It was done in XFoil using Fast combat speeds with Rn of 1,100,000.

Gary's airfoil turned in the highest L/d ratio around 18 deg AoA.  PTG's had the best performance between 5-10 deg AoG.

If you look at the top of the Cl(alpha) graph the airfoils from the top are James, PTG, Nemesi, G460, and Eppler.  Of interest is the slight upward kink in the PTG graph between 5-10 deg.  Roughly 10-20% more lift than the others.

Interesting side note, years ago I built a copy of Ira Keeler's plane.  He used a very similar airfoil to Phil Granderson's.  Very far forward high point.  flat back from just behind the high point.  He noted that "the faster it goes the better it flies".

Sorry for the fuzzy graph, it's the biggest I could get to fit.

[riley wooten]

The one thing I noticed in the graph is that the best airfoil changes greatly with Reynolds number.  Most of our stuff runs under 1,000,000, so thicker airfoils would be better.

The other thing is wing loading plays a roll.  Thin wings definitely run out of lift sooner than thicker wings.  Ain't nothing worse than an overweight VooDoo.  Nice thing about a Nemesis is it is hard to build one as heavy as a light VooDoo.  It is VERY hard to build a VooDoo as light as a run of the mill Nemesis.

It would be interesting to build the VooDoo leading out of very light cross-grained balsa with a carbon fiber cover.  Maybe with some extra CF toward the middle, so it doesn't break there.

Phil, I just ran across this.  All the Voodoos I built for personal use had the leading edges hollowed starting just outside the ply doubler getting thinner until it had a wall thickness of appx. 1/8" from the next rib to the tip...
I made a router cutter to do this so it was quick and easy but not something you would want to do for a kit.
Remember, that wing was designed 57 years ago and there were many airfoils and versions of it... No instant glue,
plastic covering, epoxy or carbon fiber.  Also, the kits came out heavy because Carl Goldberg wanted a plane that
just about anyone could fly for sport or competition and be tough.  As he said to me, "we will sell at least 10 to  sport flyers to 1 for competition".  I will have to admit he was right on and sales proved it.........
RW

[phil c]

Carl was an excellent businessman.  I think that also had something to do with the fact that most of the VooDoo kits came through with 25 lb. wood for the leading edge and trailing, 16-18 lb for the ribs and sheeting.  But to tell the truth, most folks didn't care.  It still went faster than anything most people built and everyone else was flying one too.

Just for kicks I put the first TWA engine I got one one.  It went 135 mph using a pylon prop and 10% fuel.  That was a thrill, at least until I realized I couldn't react that fast, much less think that fast.

[riley wooten]

Yeah, Carl was sharp but really didn't think a combat kit would sell in the numbers he wanted.  Fooled him big time when sales were over 50k the first year. You are pretty close on wood. The first run had about 10-12# leading edges. Too many complaints of too light so from then on they were 16-22# and some were heavier than that. As I remember the rest was about 10-14# which is still pretty hard.
When the super engines came out it was obvious we needed bigger planes to help hold down the speed....
The first Nelson was so fast we converted it to draw and put it on a slow-still 100+
RW

[Tim Wescott]

Too many complaints of too light so from then on they were 16-22# and some were heavier than that. As I remember the rest was about 10-14# which is still pretty hard.

Now this is a revelation to me.  I always assumed that kit wood is rock-hard because of penny-pinching, not that it's been complained in that direction by the model-building public.

This is going to put a whole new spin on what I think about Comet kits.

[Chuck_Smith]

Gary James observed that the NACA Variable Density Tunnel had a lot of turbulence.  I saw that graph in the stacks of the PU library when I was in high school.  It convinced me to use a fat airfoil (16.5%) for combat.  ...

You're too modest, the Nemesis changed the way we looked at combat ships. Fat wing, long tail ( with that beautiful elliptical stab!)
First time I saw one my Sneekers looked like Model T's. I had to give up my pacifiers and find a source for pen bladders.

I was still in high school, and so smitten by the thinking that went into that plane,  I actually started to study to get into college and went on to become an aerospace engineer.

Chuck