Hello guys,
I have a question regarding the behaviours of the different airfoil section type regarding a near-stall (or post stall) situation.

In particular, I ask that question because I read a post of Bruce in a different thread who said that it is not true that after a stall the wing does not produce lift anymore: if I understand correctly, it is possible (for some airfoils I imagine) that after a stall the C_l don't drop but it remains almost constant - hence continuing to provide lift, albeit I understand that the C_d will rise sharply.

So, if the above is correct the question is the following. Which kind of airfoils exhibit that behaviour, and conversely which airfoil will drop its lift after stall?

I guess that a bottom flat airfoil (i.e. a trainer airfoil) would behave like I described, whereas a fully symmetrical (i.e. acro airfoil) would behave like the second one.

Thanks for any reply!

Actually, none of them stop lifting when they stall.

If you look at any plot of lift and drag at angle of attack for any airfoil, you'll see that. As the AOA increases, we see the lift increase until at some AOA the lift starts to decrease. The line that shows drag does much the same except at that same AOA, it doesn't start to decrease. It keeps rising and usually does so at an increased rate.

What we see our airplane do, we often think of as a sudden loss of lift, but what it really is, is simply a loss of ENOUGH lift. That's coupled with steadily increasing drag that keeps increasing. And our airplane, which we were slowing and slowing, suddenly doesn't have enough lift to keep that speed, and it drops it's nose or a wing.

I suspect that what you're looking for is an airfoil that has as broad a stall speed range as is possible. Some have quite a sharp peak at the point where separation takes place and the wing is considered to be stalled. Often these types of shapes are ones with fairly rearward max thickness points. The 6 digit NACA laminar flow airfoils are examples of this.

In the model world there's a few rules of thumb related to soft stalling airfoils. Well rounded leading edges, thick and max thickness points fairly close to the 25 to 30% marks. The NACA 0015 or 0018 and any of the cambered airfoils that use that shape are examples of very soft stalling airfoils. Something like a NACA 4415 could be pretty much guaranteed to have a very broad speed range from the initial mushiness to the final soft sort of stall.

On the other hand with thickness also comes some extra drag. But on a trainer where you don't want the model to suddenly pick up a lot of speed when the pilot puts the model into an inadvertent dive this is a good thing.

If this is what you're looking for then I suspect it would make a pretty nice trainer airfoil.

The link below is to the first definitive 1945 NACA report on airfoil section characteristics... should be able to find a section that meets your criteria...

http://naca.larc.nasa.gov/reports/1945/naca-report-824/

Anyone care to bet money that the section used --other than thickness --has no bearing on post stall/near stall flying .
remember we are talking about models here --
My own experience is that the wing loading is far , far, more relevant .
Why anything other than a simple ,strong symmetrical airfoil is used on any of our powered models -is a puzzle to me
or for that matter , any airfoil which is easy to build and strong enough. Thickness affects stall angle on most shapes other than that -I see no real differences.
Heaven knows I have tried enough different combinations.

I would have to disagree. You are correct in that wing loading is extremely important, but there are two sources that create increased wing loading. 1) heavy airframe/payload and 2)high g turns. Obviously for something like a pattern plane airfoil section is going to be more critical than that of a trainer because they are trying to optimize their lift vs. drag in a high alpha turn. So simply adding wing area might help with the turns, but kill them on the strait out runs. Models such as trainers and such are not flown to their design limits, but pattern, jets, and similar aircraft are routinely flown in an environment where airfoil choice is very important. I guess the differences might be minute, but at 100+ MPH it could make or break the run.

Dick, you went out on the limb and then sawed the tree away on this one...

Camber significantly raises the max lift coefficient that can be reached before a stall occurs. This means that a cambered airfoil can slow down to a lower speed than a symetrical airfoil before the stall occurs for otherwise equal models.

respectfully-- no
let's take a wing -- say - 12" chord 60"span - trainer
now the airfoil -- let's make it 12% thick-- one a Clark Y type/ 2412/ maybe give it a raised entry -make it what you like
make an identical wing - SAME le radius but symmetrical foil- both airfoils same thickness
same weight same size engine etc
The difference being one is cambered -one is not

pretend these otherwise identical models start down the same runway to take off and they both weigh- say 5 lbs
When either setup reaches 5 lbs lift - the model is flying .
My premise is that the speed for take off (min flying speed) will be the same - tho the physical angle (apparant AOA)may be different.
There is a possible difference in efficiency - but I don't believe -in a model of this size --you will ever find it
I see the two wings as really the same - one with the aft section even. the other with it drooped

Ok guys,
I see the discussion is starting!!! Good...

I think that, for my modest experience, airfoil has an effect (camber, actually). Let me explain a simple experience I have had recently.

Let's imagine a classic trainer, with a "correct" CG and tipical wing loading. We all know that the we can easily land that kind of plane in a classic 3-point landing - touching with the rear wheels in a slow and highly AoA situation - without an apparent brutal stall - at least no brutal decrease of lift. Ok.

Now, my last model is an aerobatic plane made with depron and foam, with a symmetrical wing of about 12 - 14 % wing thinckness, similar to a Funtana. Its wing loading is FAR lower that the trainer one (mine was about 16g/dm^2, a trainer could be about 50g/dm^2).

The fact: if you try to do a 3-point landing with such an airplane, what you get is a brutal stall and an hard touchdown, even if the wing loading is almost NULL!


So... what can be different, other than the airfoil?

easy
your far lower wing loading requires far lower speed to land
the wing flys to a very low speed then the loss of lift is quick as you increase AOA for th 3 pointer.
example
I had a electric Cub total weight was 3 lbs very light
wheel landings were always a breeze
but
if I tried to slowly rotae to a 3 pointer - it would "dump at the 3 point attitude
My Funtana (90) does exactly the same thing
your old trainer would rotate and land before speed was very low.
Typical wing loading (?) on some trainers is pretty high -compared against slow speed 3D models
On My many years old Seniorita tho ,it is extremely low -it weighs under 4 lbs and is 800+ sq inches
it will also settle quickly if over rotated.
I keep it around for old times sake -to let new guys try their hand at it etc.. Further -- my flat airfoil foamies with wing loadings of 3 ounces to the foot -- same thing - they fly up to almost a dead stop -then just --drop
If you bend the foil to an undercambered foil (curved flat plate) -- same thing
maybe the speed is different -but is too slow to see any difference.
Trainer airfoils are simply a carry over from full scale appearances -in my bok. One of the best trainers I have flown--is the large Diablotin -with 40 ZDZ - it simply will crawl along and land like a feather - the airfoil is symmetrical and extremely thick- so lift wize it is quite compromised -- it can be rotated to higher AOA with no sudden break out.

I'm going to take Dick's side in this. But I'm going to hedge my bets. More on that later.

First of I believe he agreed that there is a difference but slight. At least that's what I've seen. I built Esquires and TriSquires to sell back when I was young. And I sorta had a production line setup. And they've got rubber band on wings. And I had a bunch of different airfoil templates I wanted to try. So I'd make wings with different airfoils and cut the saddle to match on the next plane I built. I always got to fly them so could directly check out how the airfoil flew. And a lot of them lasted awhile so I got to see how they worked over time (could recheck how they felt compared to a new airfoil I'd just sold).

I really don't think many of my customers had a clue that their airfoil wasn't the "factory" one, which if I remember correctly was just a ClarkY. I was never perfectly sure that any airfoil handled differently than any other. Granted, I really didn't try any wild things, but I did more than a couple of symmetrical ones and a number of different LEs on flatbottoms that had different thickness. I had lift/drag/AOA plots for everyone that I tried so I knew what incidence to set the wing into the fuselage, so wouldn't have been confused that what bad rigging would cause was what the airfoil performance was.

OK.... what all that told me back then was that airfoils as different as ClarkY versus symmetrical really flew about the same. I also understood that in our imperfect laboratory (the flying field) it's darn near impossible to accurately measure performance. And that we can't hear speed or see AOA worth spit.

Now, all that said..... Yeah, no lie, when I fly a trainer upside down that has a cambered airfoil wing, it usually takes more elevator to fly level than a similar trainer that has a symmetrical wing. But those two different wings carry those very similar airplanes around nearly identically.

And to throw a small monkey wrench into the ideas stated.......
I designed my own thermal gliders for 10-15 years and did a bunch of different wings for a couple of my 2m fuselages that held the wings with rubber bands. Those wings did do differently. But truth is, nobody but me ever noticed any performance difference. I THINK I felt a difference but there was no way to measure it. And I'm convinced that I could pick out the right wing for the conditions, but never could tell for sure.

BTW....... I've got a symmetrical wing on a trainer and have taught a number of people to fly with it. Some of my students have flown it and then their run-of-the-mill "lifting" wing trainers and commented how mine "lifts" better. That was years ago and there were only a few trainers and most had flatbottom wings. Mine flew like it had more lift from the wing. It couldn't have, right?

just to add-
I designed and built hundreds of pattern planes of many different layouts and wing loadings and airfoils etc,etc..
the structural part of wing design is hardly ever mentioned here --but is far more important than the actual airfoil
a flat bottom with a curved top is easy to do and is -- a decent layout for strength -so --it is often thought of as a good airfoil
and it is . But
it really has little else going for it on a model
On full scale I have doubts of any advantage too- except it is easy to build.
Some full scale foils are intended to maximize a certian load and speed.
others for broad loads and speeds - which brings us to the airoils for the full scale EXTRA aircraft -- an extremely large blunt LE ,which simply transfers to straight lines to a thick TE .
this is literally a flat plate but thick enough not to break-and flies extreeeemly well over broad speed ranges

I never see it mentioned on this forum.

IMO, Dick and Bruce are arguing opposite ends of the same issue... Further, both are correct within the limitations of how they define the argument...

IMO, in practice the effects of thickness and camber are similar, and likely to be indistinguishable without a much more detailed analysis in a controlled environment... but it does make for an interesting discussion!!

I hoped it would be just that -not just a wild shot at the importance of airfoil design
My new project of the Cassutt racing plane - will use a wing shape that is highly tapered and the airfoil will change from a moderate 12% section to an extremely thin tip of 8-9%
The sections are chosen as a best guess of strength as the taper of each panel is 2-1.
I expect very low wing loading to cover the other results desired
they are
1 strength
2 stiffness
3 rapid shift from lifting to stalled to lifting again
The desired AOA for stall is low (quick flick setups)
The low wing loading on the setup should keep accidental snapping at bay

Dick,

I thought the Cassutt was an extra wide chord 'Hershey Bar' wing and that the area with less span was one of the attractions.

See cassutt.lornet.com/

Could you explain the taper---- a plan form you haven't been too high on in the past???

There are a bunch of wings for the Cassutt
16 ft 18 ft rectangular --then some long highly tapered composite stuf of recent decelopement for longer course at Reno
A highly tapered wing for areobats is fine --as long as the span is not excessive - basically--they are very responsive as the mass is really inboard made light enough - they work fine
I tried this already on my Petrol Petrel
roll rate is a blur.
The rectangular is still the best at very low speed work - here are some examples - the 3view of the tapered is not the taper I will use . it will be like the red and white example.

Dick,

Thanks.

The Cassutt has areas and lines which seem 'made to order' for model aerobatic purposes.
Now with a variety of wing planforms it is even more appealing

There will no doubt be those who point out its very sparse use a full-scale aerobat-----to which I answer so what.
The current crop includes low wingers modeled as mid-wings and the reverse ---- not to mention other dimensions which seem to be well outside the 10% rule.

I'll be interested to hear of the results you obtain.

How an airfoil behaves at the stall point depends on a lot of things. Often it's hard to apply "rules of thumb" to them. If they stall with a sudden and large separation bubble that produces a sudden very high drag then the model can often appear to suddenly fall out of the sky. This can often be compounded by a loss of lift if the airfoil is of the type that has a sharp peak rather than a nice rounded lift curve at the stall.

Stek, as an experiment you could try adding a turbulator strip to the top surface of your aerobatic wing to see if you can soften the stall and help avoid that sudden drop. Often a turbulator can be used to reduce the severity of the separation bubble size. But be sure to try it out 3 mistakes high before commiting to a landing....

That link to the NACA report has some great info in it to study. You'll see how some airfoils have a nice rounded or even "flat" lift curve at the stall point while others look like Mt Everest in their sharpness at the stall.

I'm sorry if I had given the impression that airfoils NEVER suffer a loss of lift at the stall. They idea was that when stalled that they don't just stop lifting all together. The point was supposed to be that when our models "stall" that they are suffering from a loss of lift due to a reduction, but not total failure, in lift along with a sudden increase in drag which slows the model down and further reduces the lift.

Thanks Bruce for the advice!

Too bad my plane is gone some months ago!

I've read the NASA doc and found it very interesting! But I din't find a clear example like the one you're saying, can you please tell me the page number or the section where you saw those lift curves?

Thanks to all!

Dick,

Hi, first of all, I have to tell you that I really respect your work on models and I think the model airplane community owes you a debt of gratitude for making this hobby/obsession what it is today. I still have one of your Zlins that I'm currently molding because I didn't know who made one of the truest planes I've flown. But I would like to add to your post as to what you're theoretically going to see when you compare the wings.

The first one is the 2412 at that wing size. To lift 5lbs, you will be going about 39 MPH at 0deg AOA.

You said that we will only change the AOA for the symmetrical wing, in this case, a 0012. In order for the foil to lift 5lbs at 38MPH, it would have to reach an AOA of around 2.2 degrees, which is basically very near the start of turbulent flow. Whereas you still have about 1.5 degrees to play with until the 2412 begins to get turbulent. BTW, the drag at this configuration is about the same as the 2412.

Now, to put in another foil in the mix, a 63-A012, a laminar flow, 12% foil. To get 5lbs at 38MPH, the foil would have to be at 2.6 AOA. Because it is a laminar flow airfoil, the oncoming air tends to stick to the surface further than the 0012 and thus giving it a bigger AOA margin than the 0012. As an added benefit, it actually has less drag than both previous airfoils.

Here's the thing. We're talking about small foils that have small consequences on flying...really. I have hooked up wings on trainers backwards and the plane still flew well. When we get to full scale planes, that's when drag becomes a major factor and fuel consumption is critical.

All in all, the airfoil, though important, is not as important to us when selecting for our projects. Here's the caveat before anyone flames me. The airfoils should have the same dimension and, incidence, dihedral, maximum thickness as well as LE roundnes and that the TE comes to a point. Then the shape of the sirfoil should not bother the plane much. No cheating with a flat airfoil that happens to have a 12% thickness ratio...that doesn't count.

B

That ice cream cone airfoil Dick mentioned is used primarily for three reasons, not necessarily in order:

It has stall characteristics that are desireable for an aerobatic airplane: the stall progresses very rapidly (almost instantly) from the TE forward. This is good for clean snaps. NACA airfoils with their further aft high points and smoother curves hang on to the airfow longer, giving a more gradual stall, which is good for other than aerobatic airplanes.

It has a similar lift curve to an equal thickness NACA section, except that it drops off much more rapidly at stall (there is little or no pre-stall lift penalty).

It is easy to build.

I have done comparisons in XFoil and have flown a few ice cream cones and the theory agrees with experience.

Thanks for mentioning those, Dick. Ice cream cones have gotten a bad rap in modeling for the very reason they are so popular with full scale acro: sudden, clean stalls with no pre-stall buffet.

Of course there are many airplanes with other types of airfoils that snap well - the ice cream cone is just one way to do it and is truly a matter of personal preference. I like the ones I have flown.