Flying upside down

Agreed. I’ve run into this problem before… various exams my students write are from flying schools, government agencies, and the military. Each has its own set of little prejudices like that. I’ll have to be sure to prepare them for this. I plan to teach it the way bbeaty does, which includes both downwash and Bernoulli explanations, but make clear that the path difference is not what causes accelerated flow.

If you read my post above, you’ll see that I was already teaching that downwash is an important component, and that the Bernoulli effect due to differential path lenghts was another (in fact, likely not the major contributor). I’d been taught, when I was a student, that it was all about the unequal path length (with a bit of contribution from downwash), but discovered through my own reason and investigations that this didn’t quite jive, and adjusted accordingly.

I greatly prefer bbeaty’s model, which I find much more intuitive, and which answers many niggling little questions (“Sir, if path length is the deciding factor, why is the thickest part of the wing always close to 25% chord?”)

Agreed. I was just sort of neglecting concerns about flow separation and just focusing on what generated maximum downwash, though clearly the effects of those are why a flat plat won’t do (in many applications).

Also, I concur (by intuition if not by citation) with Ale’s analysis that the air would have a departure vector slightly angled to the bisector of the trailing edge angle, due to greater speed of the upper flow.

Ah, geez. Now you’ve got my head all full of airplanes and airflows again. How am I ever going to get any work done? :smiley:

Wow I wasn’t expecting such a large response.Thanks!! I especially like the “how things work” link!!

Nope, now that I think about it more, that can’t work. An upper/lower velocity difference at the trailing edge is forbidden because the flow pattern would immediatly change until the air was flowing smoothly off some other point on the wing, and NOT off the trailing edge. In other words, at the trailing edge of the wing, the upper and lower velocities are normally the same because of “Kutta Condition” which is caused by inertia. It’s not a law though (see the following.)

Here’s a diagram which shows what would happen if those two velocities were not the same. This is an exaggerated diagram (the zero-circulation condition), and this sort of effect would never be so large.

http://www.av8n.com/how/img48/nocirculation.png
And here’s the more usual situation:

http://www.av8n.com/how/img48/normal.png
Even simpler, here are diagrams of two possible flows around flat wings. The second one illustrates conventional flight. In that second diagram, just imagine the wing to be thick and streamlined, then imagine it to be cambered. These two changes will not alter the trailing-edge flow pattern. The air still flows parallel to the bisector.

http://www.av8n.com/how/img48/barn20x.png
http://www.av8n.com/how/img48/barn20z.png
Of course the flow doesn’t REMAIN going straight off the trailing edge. It curves, so it starts off straight but rapidly deviates from a straight line.

Weird tidbit: if the trailing edge of a wing is not sharp, then the lifting force vanishes. This is not just useless trivia. If you understand this strange fact, then you REALLY have a solid grasp of the mysteries hindden in airfoil physics.

In other words, a flat plate “stalls” as soon as you try to tilt it.
If you bend the leading edge of a flat plate downwards so its leading edge meets the oncoming air at zero angle, then the trailing edge can also be bent downwards so it deflects lots of air. And… the more you bend the trailing edge downwards, the more you have to bend the leading edge. Camber!

Heh. That quote contains no errors!

As long as they carefully avoid saying WHY the air above a wing flows faster, their explanation is perfectly correct.

Here’s something to think about: in the following diagram, the air above the aifoil is flowing MUCH faster than the air flowing below. Wing shape obviously has no effect. This illustrates something the books never mention: the fact that positive angles of attack cause lift ONLY because the positive angle of attack forces the air flow faster above the wing and slower below.

http://www.av8n.com/how/img48/barn20.png

If you mean that the wing wouldn’t work if the two edges were separated (ie, the trailing edge becmes a trailing wall), then:
Is this because, aside from the recirculation drag, you have two airflows roughly parallel but separated from each other, with a low pressure area in between, resulting in the lower airflow deflecting up towards the low-pressure area behind the wing?

If you mean that the wing is the same shape as normal, except that the upper and lower surfaces meet at a wide angle, then:
Is this just a question of flow separation becasue of the excessively steep angle, similar to what you get if you have a diffuser with too steep an angle? If the top surface of the wing is too steep near the back, the momentum of the flow would overcome the tendency to adhere to the surface. (Flow separation.) The drag would be large, and the flow wouldn’t wash downward much, if at all.

Which situation do you mean? Wing ending in a wall, or wing ending in wide angle? And am I right, at all?

Note that the following stuff is sort of subtle, and would only be taught after the simpler concepts were solidly learned. It’s about the reasons why wings fling air downwards, or more specifically, the reasons why ‘circulation’ appears.

I mean, use a blunted, non-sharp trailing edge. If the trailing edge was just as blunt as the leading edge, then even if the wing was cambered and even if it had a positive attack angle, the wing would still not deflect air downwards, and lift would vanish.

Certain kinds of wings will fail to deflect air downwards, since the air flowing beneath the wing would collide with any downwards-flowing stream and push it back upwards again.

In terms of those diagrams, below is what happens if either the trailing edge is too blunt, or if the viscosity of the flowing fluid is high enough to damp out all the inertia effects:

WING WITH NO ‘KUTTA CONDITION’

Notice that the flow pattern in the above diagram is symmetrical, where the flow approaching the leading edge has the same shape as the flow departing the trailing edge. The wing is unable to fling the air downwards and have it keep flowing downwards. The air divides and lets the wing pass by, but the wing doesn’t deflect the flow in a permanent way.

If instead the trailing edge is made sharp, then intertia effects become strong, and the pattern of flow changes to this one:

WING IN NORMAL FLIGHT

Are you aware that “sucking” is different than “blowing?” For example, if you have a pump with its inlet and outlet having identical shapes, then the flow pattern near the inlet is radial like a sunburst, while the flow pattern near the outlet looks like a narrow stream. Inertia is the cause. Air that was flowing as a narrow jet inside the pump will continue flowing as a narrow jet when it leaves the outlet. But air which approaches the inlet has no reason to form itself into a long narrow jet. (And you can blow out a candle from far away, but just try sucking out a candle from the same distance!)

This type of inertia stuff is what happens in normal flight as shown in the second diagram; the leading edge cannot form distant air into a narrow stream, but the trailing edge can do this easily because of inertia. The leading edge is surrounded by a pattern resembling “suction”, while the trailing edge is surrounded by a “blowing” pattern: a long narrow jet of air. The wing essentially “grabs” the air and flings it downwards, the air continues moving after the wing has gone past, and the “claw” that grabs the air is located at the sharp trailing edge.
(Those diagrams are from  3  Airfoils and Airflow)

PS

In three dimensions the air continues flowing downwards forever (or until it hits something, or until it slows down from friction.) But this topic is never mentioned except in the most advanced textbooks. But it’s simple! (Well, it’s simple if you have an animated diagram.) I thought these concepts should be part of introductory teaching, so I put the math-free version on my website here:

http://amasci.com/wing/rotbal.html

Note the animated GIF that’s a ways down on the page. When I first figured this stuff out (while fighting with J. Denker!), I was amazed to discover that wingtip vortices aren’t just an unimportant detail which should be ignored. Instead they’re right at the center of any full-blown explanation of flight. They don’t just trail behind the wing. Instead, the wingtip vortices are part of the air which remains moving downwards after the wing has passed by. Planes fly BECAUSE they create wingtip vortices. I think this whole part of wing explanations is ignored in introductory texts because they focus on 2D diagrams (which simplify things a bit TOO much.)

The book Elementary Mechanics of Fluids quite clearly states that the presence of viscous shear with circulation produces stream lines that leave the wing tangential to the bottom of the wing and thus produces a lifting force.

Our ground school instructor in Cadet training told us to just consider that the wing “throws the air downward behind it” and thus produces lift. The low pressure-high pressure is true but only serves to redirect the airflow around the wing.

The Stearman biplane that I learned to fly in had fabric covered wings. If low pressure above were lifting the plane then the fabric on the top of the wing would have bulged upward wouldn’t it? Or the fabric on the underside would have bulged upward. Or both. But that didn’t happen.

Actually, I believe the cloth surfaces of a Stearman are sewed to the ribs for precisely that reason - to prevent the fabric from pulling upward and distorting the wing shape. After all, there is a low pressure area there.

If there was any change in the form of the fabric between ribs it wasn’t evident to me. We used to speculate on the subject at the time, and didn’t really know enough to come to a firm conclusion. Of course, it was a long time ago.

Hmmm. A WAG.

I wonder if it has anything to do with the fact that the air in actual contact with the wing is stationary with respect to the wing and will be at stagnation pressure. And, because of viscous drag, the next thin layer will be almost stationary and so on. So there might be a near stagnation pressure boundry layer that keeps the fabric from bulging. And the air in that boundry layer doesn’t flow out into the nearby low pressure zone but the pressure is maintained by the forward motion of the aircraft.

bbeaty, that’s more or less what I’d envisaged.

To be clear, the wing shapes in those diagrams aren’t representative of what wing shapes would generate those flows, right? 'Cause both of those diagrams have wings that look like flat plates.

Right, those are computer simulations where “stall” and turbulence are forbidden. If stall couldn’t occur, then a flat plate would make an excellent wing.

I haven´t seen it myself, but I´ve heard accounts of crop-duster pilots that have seen the wing upper sking “ballooning” when pulling up (I don´t remember what plane though).
Probably, the Steraman being a biplane, distributes the lifting forces over a larger area, so the low pressure over the wings is not as great as it should be to pull the covering up.

By the way, last month I had the chance to take a ride on a Steraman, that´s the way I like my flyin´!
I´d take an open cockpit plane anytime, it´s a wonderful experience.

In A&P school, we were shown pictures taken in flight of the upper surface of a wing. It had a poorly done recovering job, and the fabric was ballooned up between the ribs.

The cloth on Stearman wings is pretty tight, precisely to help it maintain aerodynamic shape.

Back when I was flying ultralights with Dacron wings and no rib-stitching we did see some ballooning on the upper surface and “sucking” on the bottom. During stalls, you could also see small ripples in the fabric traveling from root outward.

Which is not to say we were flying on billowing bedsheets, just that the wing fabric was looser than on many larger airplanes and if you looked you could see these effects.

I have to admit that I didn’t look at the wing fabric during stalls. The instructor was there insisting that the nose be made to drops straight down through the horizon before recovery. And that habit carried through into solo’s.

However, I think it is quite possible there was some disturbance to the fabric during a full stall as there is a lot of separation and turbulence over the top of the wing.

Just the same, I would guess the wing loading on a Stearman is in the neighborhood of 15-20 lb/ft[sup]2[/sup] and in level flight that should cause quite a bulge that I’m sure I would notice. After all, in a climbing turn one of the ways to hold a constant bank angle is to put the inside, upper wing tip on the horizon and hold it there. I don’t recall any change in the configuration of the covering when looking at the wing in rolling in or out of such turns.

In a steep bank, say 60[sup]o[/sup] the lift is doubled and I never noticed any change in the “fit” of the covering on the wings then either.