Fear of flying Bernoulli-Style

[bunyip]

I think your theories make more sense than most, Nimbus. At least I can understand it the way you present it.
The spoiler on top of the wing does its job by increasing drag as well as by disrupting the flow of air. Since the air is faster (?) on top, it is more affected by drag. Some airplanes do have spoilers on the bottom of the wing (gliders, Vampire etc).
The theory boys ignore the fact that without wingtip vortices there would be no lift. A wing of infinite span creates no lift (Kermode). So wind tunnel experiments do not explain lift unless the wing is complete (ie has wing tips in the flow).
A flat plate creates lift, even at a small angle of attack, when I would assume no flow separation would occur, so the difference in speed of the particles would not be great.
When moving your hand through water at an angle, the force against it is felt almost solely on the bottom side, which means that, to me, the majority of the lift is being produced by the angle of the wing to the relative wind, deflecting the airflow down, and thus the wing up. The rudder on a boat surely does its job by deflecting the water directly, not by creating a low pressure side which causes vortices to form? (water is incompressible). And an aileron must act directly on the flow of air? Wind tunnel theory only partly relfects the real world, and could very well be wrongly interpreted.
If someone kicks my butt, I think it will be the toe of his boot I feel, not the air flow from the boot vortice!

[Prof2MDA]

Nimbus-
You write:
_________________________________________
"Not sure about your statement that the downwash is caused by wingtip vortices…!
I would have thought that the wingtip vortices are an undesirable result of the whole up-/downwash action, not the
cause!"
__________________________________________

While I agree with the statement made previously that you are referencing, I should point out to you that I was QUOTING Dr. HC Smith, Director of undergrad studies in Aerospace Engineering, Penn State! Are you saying he is wrong? If so, then so is John Anderson, Chris Carpenter, et al. Go look at the NASA web site I posted earlier as well. All the PhDs in Aero agree with each other. You don't agree with them. Sorry, but I think I'll beg to differ with you and go along with the folks who write the textbooks on the subject!

As for your statement referencing bricks as "proof" that it's Newton and not Bernoulli, all that tells me is that you truly don't understand what Bernoulli is and how to apply it. You clearly lack the basic concept of what is generating the flow that brings about the differences in pressure. I suggest again, more strongly than ever, that you buy and read Flightwise: Principles of Aircraft Flight, by Chris Carpenter (who is, incidentally, the head of aerodynamics at the Royal Air Force college).

[This message has been edited by Prof2MDA (edited 22 April 2001).]

[This message has been edited by Prof2MDA (edited 22 April 2001).]

[bookworm]

<font face="Verdana, Arial, Helvetica" size="2">The theory boys ignore the fact that without wingtip vortices there would be no lift. A wing of infinite span creates no lift (Kermode). So wind tunnel experiments do not explain lift unless the wing is complete (ie has wing tips in the flow).</font>

Well that's one of the more absurd assertions of this thread! Reference from Kermode please, bunyip, so we can see the context?

<font face="Verdana, Arial, Helvetica" size="2">A flat plate creates lift, even at a small angle of attack, when I would assume no flow separation would occur, so the difference in speed of the particles would not be great.</font>

What do you mean by "flow separation" here? If you mean "some air flows over the plate, some under", I beg to differ. That's exactly what happens. You might want to look at http://www.monmouth.com/~jsd/how/htm...sec-thin-wings
to see the difference in speed.

[chicken6]

NIMBUS

The only way to get a brick to fly is to provide so much power that the vertical component overrides the weight, and then you might as well call it a helicopter. How exactly were you going to provide this power? Through a jet? A prop? It would only be the upward component of thrust that would get it off the ground unless you had a brick unlike one I've ever seen - one that deflected more air down with the undersurface than it deflected up with the frontal/upper surface.

And parking aircraft outside in horrendously strong winds does create enough lift to flip them over, that's why tiedowns were invented.

Bunyip

The common name for spoilers under the wing is flaps. Different types of flap exist, and can be hinged in different places along the chord, but in general a flap extending from under the aerofoil is called a flap, a flap on top of it is a spoiler, and on the fuselage they are called airbrakes.

Note they are all named after their function, and there is only one called a 'spoiler' because it spoils the lift.

And the boat rudder thing is more to do with the water being 1000kg/m3 whereas air is only about 1.2kg/m3 (at sea level). Being 1000 times more dense, deflecting a certain amount of water is bound to require a much much greater force (then equal and opposite reaction rah rah rah) than the same volume of air. If water was compressible, then we wouldn't even be talking about this because there wouldn't be any dry land on this fair planet.

When I learned all this stuff and could write pages of calculus on the stuff, I seem to remember Bernoulli being pretty well right and Newton being pretty well right, although memories fade with time!

------------------
Confident, cocky, lazy, dead.

[Prof2MDA]

A brick could fly with enough speed if it were inclined, just as a flat plate works, with enough q, of course.

Lift can be explained either with Bernoulli or via the Newton method (regardless of the fluid you are travelling through)

Bernoulli depends on Newton in various ways (such as continuity.

Newton's method depends on airflow being deflected by Bernoulli, so to get the amount of air deflected off the top of the wing you need you must have the pressure differences that are explained only by Bernoulli so the flow is forced around.

What is being missed here is that these are not conflicting theories at all. They are not even complimentary (meaning that you don't get "part of the lift" from one theory and "part" from another). They are just different ways of _quantifying_ the results. Either way you can come up with ALL the force you need. If you want to work harder you could do part and part I suppose, but that would be a bit silly.

[bunyip]

OK, the flat plate would only work until flow separation (ie stall), so it is restricted to a low angle of attack.
The spoiler or speed brake works because it causes drag. It has to be close to 90 degrees to the airflow to do this. It matters not where it is located; on top or bottom of the wing or the fuselage. If it is on top and moves only a little bit it is a spoiler, and works by disrupting the airflow, and causing, again, drag. The 747 spoiler is restricted in range in flight for example. In flight it operates as a spoiler and on the ground as a speed brake. By spoiling the flow it also reduces the lift produced, since a clean flow is needed to set up the whole lift thingy (technical term).
Kermode Chapter 3 has a lot of stuff in it and most is above my head, I admit. But I agree with the circulation theory and it says you need wingtip vortices to have lift. Without them you have just drag. If you have an infinite wing then the CL is reduced because the correction for aspect ratio is to divide the CL by Pi.A (A being aspect ratio). If A is infinite then CL is zero. The discussion is in the book to cover induced drag, but I am assuming the drop in CL would also affect lift. Could be wrong, so don't run out and buy any lottery tickets based on my advice!
However if this is right then there will be insufficient lift produced by a wing to fly, and it is my own theory that the wind tunnel results are not real world. I can see that the force on the bottom of the wing needs to be greater (impact lift to give it a name), and in order for it to give more lift than the theory supposes, I can see that it is affected by the approach of the wing and is diverted downward before the wing actually passes, and so the wing is riding on a rising flow of air, so it planes, much as does a water ski. A well designed wing will be efficient because it reduces drag, as well as producing lift by increasing the downward flow of ambient air. As someone said, the total of all effects gives the lift; nothing is solely responsible.

[Prof2MDA]

&lt;sigh&gt;

yes, could be wrong about your infitite wing not creating lift...

[bookworm]

<font face="Verdana, Arial, Helvetica" size="2">OK, the flat plate would only work until flow separation (ie stall), so it is restricted to a low angle of attack.</font>

Correct, just like a shaped aerofoil. The flat plate may typically stall at about 10 degrees with a maximum lift coefficient between 0.5 and 1. The shaped aerofoil is better at avoiding separation at higher AOAs, and can manage a lift coefficient of, say, 1.5. That's why it's shaped the way it is! But the principle by which the shaped aerofoil and flat plate produce lift are the same.

<font face="Verdana, Arial, Helvetica" size="2">Kermode Chapter 3 has a lot of stuff in it and most is above my head, I admit. But I agree with the circulation theory and it says you need wingtip vortices to have lift. Without them you have just drag. If you have an infinite wing then the CL is reduced because the correction for aspect ratio is to divide the CL by Pi.A (A being aspect ratio). If A is infinite then CL is zero. The discussion is in the book to cover induced drag, but I am assuming the drop in CL would also affect lift. Could be wrong, so don't run out and buy any lottery tickets based on my advice!</font>

You misunderstood the left hand side of the formula you repeated. Cl/(Pi.AR) is the "induced angle of attack". That's the correction to the angle from which the flow approaching the wing appears to be coming because of the wingtip vortices.

To estimate the lift of a finite wing you can subtract the induced AOA from the geometric AOA. For example, a finite wing at 10 degrees AOA with an induced AOA of 1 degree will have the lift per unit span that an infinite span aerofoil has at 9 degrees.

As AR approaches infinity, the correction, not the lift, tends to zero!

The induced AOA can also be used to find the induced drag. The lift vector is effectively tilted back by the induced AOA, so the induced drag is the lift multiplied by sin(induced AOA). Again, as AR goes infinite, the induced AOA and thus the induced drag tends to zero.

[This message has been edited by bookworm (edited 25 April 2001).]

[bunyip]

Ah, you caught my deliberate error! But without using the formula for Aspect Ratios (BTW, what is the correction (if any) on lift if the aspect ratio is changed?) the circulation theory says that lift is generated as a result of wing tip vortices. So if you have an infinite A you would not have wingtip vortices and so you would not have any lift. And that's using Kermode again.
I still think there must be more to it, and that the theories are just that. In the real world there could be things happening that do not show up in a wind tunnel.
Imagine that there are some mad scientists who build a round tank of water and put a large water ski on a radial drive and at a positive angle, which moves around the pool. The ski has a weight in the form of a concrete block that spoils any aerodynamic lift and makes the thing sink if it goes too slow. At some speed it will lift off the bottom and as speed increases it will plane on the surface. Now the mad scientists let alcohol or compressed air into the tank to reduce the SG/density. The ski would have to go faster to stay on top, right? And what if the water was so aerated it was close to the density of air? The ski would have to go pretty fast now, with the mad scientists' hampsters pedalling like mad. But it would still plane, right?
If the ski planes, why not a wing?

[Smurfjet]

bunyip I don't know what you on about, or what you're confusing in Kermode's book, but please please stop saying

<font face="Verdana, Arial, Helvetica" size="2">...if you have an infinite A you would not have wingtip vortices and so you would not have any lift. And that's using Kermode again.</font>

My head hurts and I'm sorry I can't help you out...

[bunyip]

Why? Afraid you might fall out of the sky?

[Evo7]

<font face="Verdana, Arial, Helvetica" size="2">
A wing of infinite span creates no lift (Kermode)
</font>

bunyip, I'm not sure I understand this quote. A wing produces circulation in proportion to its angle of attack, and the
Kutta-Zhukovsky (or Joukowski - not sure of the correct spelling, as both seem to be used) says that lift (per unit span) is proportional to circulation (and, FWIW, density and velocity). Now vortex-lines can't have lose ends, so in for a finite wing they spill off at the wingtips, and mark the boundary of the region of descending air behind the wing that, thanks to Newton's Third Law, keeps us in the air. For an infinite wing, no vortex shedding occurs (obviously, as there are no wingtips), but circulation and hence lift occur unless the wing is at a zero angle of attack. The lift per unit span should therefore be finite, not zero, while the total lift is, of course, infinte. The modifications to make 2-D calculations apply to 3-D are well documented.


Edited for tyops

[This message has been edited by Evo7 (edited 26 April 2001).]

[bunyip]

That sorta makes sense, thanks. I like your explanation in fact. I will put my grey cells to work. I don't doubt it is correct. I was taught aerodynamics a long time ago by some pretty intelligent guys, and the statement that a theoretical wing of infinite span, or in a wind tunnel when the wing extended from the sides of the tunnel, would produce no circulation and therefore, in theory, would produce no lift, stuck. It was explained at the time but I cannot remember how. I might have misunderstood (probably did) and I was hoping someone else would be able to explain it to me here.
But my real agenda is to have someone explain why impact lift is not considered.
When I hit turbulence on final in a Cessna 152 and lose 50 feet real quick, then slam out at the bottom of the air pocket, the little Cessna stops going down in a real hurry. I can feel the wing hitting solid air. No way is it just the circulation starting up again, it is solid air, and feels just the same as a water ski coming down off a wave.
Ignoring all the theory, I cannot see why a wing set at an angle of attack simply does not push the air out of its way and so generate lift. The circulation, Bernoullis theory etc simply explaining why a wing does it so well, and why drag can be reduced with a well-designed wing. A wing at 90 degrees angle of attack is still going to produce lift, yet there is little flow theory working there!

[This message has been edited by bunyip (edited 26 April 2001).]

[bookworm]

<font face="Verdana, Arial, Helvetica" size="2">But my real agenda is to have someone explain why impact lift is not considered.
...
Ignoring all the theory, I cannot see why a wing set at an angle of attack simply does not push the air out of its way and so generate lift. The circulation, Bernoullis theory etc simply explaining why a wing does it so well, and why drag can be reduced with a well-designed wing. A wing at 90 degrees angle of attack is still going to produce lift, yet there is little flow theory working there!</font>

Newton was the first person to consider what you call 'impact lift'. If you go to very high altitudes (space) where the mean free path of the air molecules is of the same magnitude as the wing chord, then impact lift is what you get. The particles behave as individual packets of momentum bouncing off the underside of the wing.

Let's do some sums.

The dynamic pressure of an airflow is essentially the force applied by bringing all the air molecules in the airflow to a halt and thus changing their momentum.

So if you put a flat plate at 45 degrees to the airflow, the air molecules would change direction by 90 degrees as they bounced off. You'd get a rate of change in momentum downwards of the dynamic pressure times the area of the plate. That would produce a force upwards on the plate, your 'impact lift'.

The lift coefficient for that 45 degree angle of attack is 1 (one). (Of course you'd also have a drag coefficient of 1, but we'll ignore that for now). If you reduce the AOA of the plate, you get less deflection of momentum and therefore less lift. For a 10 degree AOA, you'd get a deflection of 20 degrees, and therefore a lift coefficient of about sin(20 degrees) or 0.3.

In the wind tunnel that bunyip is so fond of, you can do the experiment, to find that the actual lift coefficients are much greater than that. It's about 1.0 per 10 degrees of AOA and is roughly linear. So one reason "why a wing set at an angle of attack simply does not push the air out of its way and so generate lift" is that, from the point of view of reconciling theory with experiment, a perfectly efficient deflector of that sort doesn't generate more than a fraction of the lift actually observed.

Before you say "but it's a part of it" and go all additive on me, bear the following in mind. In reality, in the density of air that we fly in, the airflow pattern doesn't look anything like the simple "deflector" picture outlined above. Because the air molecules interact with each other over short distance scales, the cross-section of air disturbed is much greater than the cross-sectional area of the wing. The easiest way to calculate the lift produced is to use the laws of physics that were designed to work with interacting molecules of fluid, namely Bernoulli's theorem and, more generally the momentum theorem. If you do that, with the airflow patterns predicted by circulation theory, you get a pretty good and complete picture of the lift coefficient and the way it varies with geometry (as well as avoiding the ludicrously high drag coefficients that the 'impact lift' theory also predicts).

[bunyip]

Thankyou, Mr Worm, good information. I am sure you are right, and will not argue with you. I don't think that it is necessary for the angle of attack to be so high as to make a 90 degree change of airflow; a small diversion is enough, which would not have a high drag penalty. But I guess any lift so generated would also be small.

[Prof2MDA]

Impact lift does have an application in hypersonic flight vehicles at the edge of the atmosphere, actually, where that is the source of most of the lift.

At lower altitudes its contribution is very small, as has been stated here.

[Speedbird252]

...so how do aircraft fly upside down??

[Evo7]

Very well, thanks....

http://www.monmouth.com/~jsd/how/htm...nverted-camber

explains it better than I can...

[pulse1]

A fascinating discussion and I have one question. If the major component of lift is due to Newton, why are differential spoilers such a useful alternative to ailerons? They seem to be effective on a wide range (speed) of aircraft from microlights (Snowbird) through the ATR to some fast jets, where a very small application of spoiler seems to have a rapid effect on roll rate.

------------------
"If you keep doing what you've always done, you will keep getting what you've always got"

[Prof2MDA]

It is incorrect to state "that a major portion of the lift is due to Newton". Newton and Bernoulli are inseparable, it is not a case of "part from one, part from another" at all. Rather, you need Newton to get Bernoulli and vice versa!