Aerodynamics - why wings create lift - current vs historical discussions

[sophiecentaur]

Then there would be no net angular momentum, so what was your point in bringing angular momentum up?

Whether or not I have understood the whole aerodynamics picture I can be pretty sure that the zero "net" angular momentum doesn't mean that the two counter-rotating hamster wheels and their individual angular momenta cannot provide the hamster with some lift. The fact that the net angular momentum is zero is not relevant to the two forces that give the lift to the hamster.

 

[A.T.]

The fact that the net angular momentum is zero is not relevant to the two forces that give the lift to the hamster.

Then how is angular momentum relevant at all here?

 

[sophiecentaur]

Then how is angular momentum relevant at all here?

Did you see the video?

 

[A.T.]

Did you see the video?

Yes

 

[FactChecker]

This video is an interesting and informative summary of the problems with simple explanations. I liked it enough to make a list of topics and times in the video so that people can look up their favorite erronious explanation:

    1. Outline: 4:23
    2. Basic physics: 6:00
    3. Lift explanations: 13:55
        a) Bernoulli: 15:25
            (1) Longer path, equal time: 15:40
            (2) Streamline pinching and conservation of mass: 17:45
        b) Downward-turning based (Newton): 19:50
            (1) Newton's second and third: 20:34
            (2) Coanda effect: 21:20
        c) Bernoulli-versus-Newton "controversy": 23:00
        d) Getting it right: 24:53
    4. Vorticity and Biot-Savart: 28:30
        a) Tip device drag reduction errors 33:28
    5. Lift and momentum in 3D: 41:00
        a) Trefftz-plane slice: 41:30
        b) Momentum accumulation in the atmosphere: 44:49
    6. Boundary-layer separation: 47:03
    7. Base drag and Hoerner's jet pump: Ran out of time

 

[FactChecker]

OMG. So they're all wrong? Which is the right alternative?

I think it's better to say that all the simple explanations that people like are wrong or at best grossly incomplete. People do not like to accept the differential equations of fluid dynamics as an explanation. Even if people accepted DEQ, they would require infinite integrals and/or complicated boundary value assumptions.

Question is whether the pressure at the ground is high enough to measure for a plane at normal height.

I don't think that is an important question. Nobody would use that to estimate lift anyway. A model in a wind tunnel is the best way to directly measure lift. The validity that the total force on the ground equals the lift is not in question. The problem with that as a theoretical explanation is that it is like a platitude -- it seems profound, but it adds no insight into how lift works.

 

[Swamp Thing]

This one is pretty interesting and may be of interest to anyone following this thread

 

[cjl]

wingtip vortices cause induced drag

Ehh.. wingtip vortices are definitely related to induced drag. Certainly it's a complex interaction, but I would hesitate to try to separate the two too much.

He says that "there is no downward momentum accumulated in the atmosphere" (at 45:10). This should not be surprising. How should the atmosphere as a whole accumulate downward momentum, with the ground in the way?

There certainly would be accumulated downward momentum if the ground didn't exist though. Behind the aircraft, there absolutely is accumulated downward momentum. The only reason that doesn't happen is because the ground is in the way.

 

[Swamp Thing]

the ground is in the way

It doesn't necessarily have to be the ground, per se. If we think of a mini drone that is hovering several kilometers up, then it is creating a small column of downward air that accounts for the lift. But this column widens and slows down as it moves down. At some point, the undisturbed air around the column and at its base will have a mass much greater than all the air that the drone is causing to move. The mass (inertia) of all that stationary air could play the role of the earth, in that it is heavy enough to account for momentum exchange with a negligible velocity change of its own.

In other words, if the aircraft is high enough, we can reasonably draw a large closed surface that encloses it and say that FAPP, everything that's happening can be accounted for by the air within this surface without considering the existence of the earth. You need the Earth's surface only to account for why the air wasn't falling under gravity in the first place.

 

[cjl]

Negligible, but nonzero. Without something in the way like the ground, that can support a force without gaining momentum of its own, there will be accumulated downward momentum. As the drone hovers for longer and longer (or more and more aircraft fly overhead), the air will gain more and more downward momentum, and eventually you'll end up with non-negligible downward velocity unless you have a hard surface to absorb that downward force.

 

[Swamp Thing]

you'll end up with non-negligible downward velocity unless you have a hard surface to absorb that downward force.

That's hard to argue with. But perhaps that would only be moving the problem along one step, because then the Earth would have to acquire a negligible but non-zero velocity to balance the books. And maybe that is indeed the case, with the Earth and the air having their own little movements but maintaining a common CG that orbits the sun as per usual.

 

[cjl]

Sure, but the downward force eventually placed on the Earth by the downwash is counteracted by the gravitational attraction between the Earth and the plane in the first place.

 

[boneh3ad]

I don't think that is an important question. Nobody would use that to estimate lift anyway. A model in a wind tunnel is the best way to directly measure lift. The validity that the total force on the ground equals the lift is not in question. The problem with that as a theoretical explanation is that it is like a platitude -- it seems profound, but it adds no insight into how lift works.

*wind tunnel engineer slowly nods, grins*

 

[K41]

*wind tunnel engineer slowly nods, grins*

I agree with your point but I would like to add that people must always be aware that experiments are also difficult as well. I feel sometimes experimentalists just aren't appreciated in some parts of engineering.

In general, for example, PIV isn't great for near surface measurements due to the reduction in signal with window size. Multiple point measurements with Laser Doppler is possible but will take time and are complicated by the necessity of needing a kinetic stage to move the measurement volume accurately. In both cases you also need a method to introduce particles or droplets homogeneously in the air flow.

Using particles means that some particles will deposit on the surface of your model gradually over time, so you have to continually replace your model. You can use droplets but risk the droplets evaporating, particularly in high speed flows, so you may need to add things like glycol to the water to reduce evaporation - again more complexity.

In some cases you will need to ensure the upstream flow is uniform - which means you probably need a very long test section or a very small model. I also imagine small aerofoil models have problems with specifying the surface roughness in some cases, which I imagine makes studying transitional behaviour particularly challenging.

3D measurements are possible using PIV and Laser Doppler, but I imagine applying them to a wind tunnel, whilst possible, is not easy. This makes it difficult to study lift formation, particularly the "starting vortex" at the rear end of the aerofoil.

Big up to people like yourself who have to overcome these challenges and I am sure, many more!

 

[boneh3ad]

@K41

This is why I:

• build big tunnels,
• don't use particle seeding methods like PIV/LDV,
• work with supersonic flows, and
• get my graduate students to do all the tedious parts.

 

[FactChecker]

I agree with your point but I would like to add that people must always be aware that experiments are also difficult as well.

There certainly are great limitations on any aerodynamic estimation method, be it wind tunnel, CFD, or anything else. And the various results can take a lot of work to merge into a single, consistant aerodynamic model. That is why the first "flight" of a new design is so often a high speed run down the runway, with no planned flight at all. At every step, the models are adjusted to match the results measured on the real plane before the test envelope is expanded a small amount.

 

[Aeronautic Freek]

@K41

This is why I:

• build big tunnels,
• don't use particle seeding methods like PIV/LDV,
• work with supersonic flows, and
• get my graduate students to do all the tedious parts.

How do you mean you bulid big wind tunnel ?
What are you doing,what is your job?

 

[boneh3ad]

How do you mean you bulid big wind tunnel ?
What are you doing,what is your job?

The post by @K41 mentioned issues with small models. Solution: build a larger wind tunnel so you can use larger models.

What do I do? If it wasn't obvious from my mention of graduate students, I am a professor who studies/teaches fluid mechanics/aerodynamics, especially high-speed aerodynamics.

 

[Aeronautic Freek]

The post by @K41 mentioned issues with small models. Solution: build a larger wind tunnel so you can use larger models.

What do I do? If it wasn't obvious from my mention of graduate students, I am a professor who studies/teaches fluid mechanics/aerodynamics, especially high-speed aerodynamics.

Nice job!
Did you maybe read Understading Aerodynamics-book,do you agree with Doug Mcleans explanations/obseravations?

 

[boneh3ad]

I have not read it and do not intend to.

 

[Aeronautic Freek]

I have not read it and do not intend to.

You sound like you have something against this book, what is "problem" with this book?

 

[Aeronautic Freek]

This one is pretty interesting and may be of interest to anyone following this thread

current aero engineers are very sharp at what they are doing but they don't understand at all what Al Bowers talking about...but it is not their fault,schooling is "wrong"..they just fallow what they have learned..

 

[FactChecker]

You sound like you have something against this book, what is "problem" with this book?

A better question might be what book @boneh3ad recommends as an introductory book. I would also be interested in that.

PS. Who am I kidding? I don't have what it takes to read a serious technical book any more.

 

[boneh3ad]

You sound like you have something against this book, what is "problem" with this book?

Not at all. I own 123 textbooks assuming my spreadsheet keeping track of them is correct. Just looking at the shelf behind me, at least 55 of them are related to fluid mechanics/aerodynamics. I haven't even read all of them, just bits and pieces that I need for a given project. How much time do I have to read another book? Nothing I've seen anyone post of his is particularly Earth-shattering. It all fits with everything else I have read and teach.

If someone wants to give me a free copy I will add it to my library. Otherwise, I probably won't go out of my way.

current aero engineers are very sharp at what they are doing but they don't understand at all what Al Bowers talking about...but it is not their fault,schooling is "wrong"..they just fallow what they have learned..

There was an entire thread on this recently. Maybe you were involved, I don't know. The bottom line is that aerospace engineers aren't doing it "wrong" or "right." It's purely a question of what your optimization parameters are when determining an optimal wing shape as well is what is actually manufacturable. Until relatively recently, the shape described by Al Bowers was not something that was readily manufacturable.

A better question might be what book @boneh3ad recommends as an introductory book. I would also be interested in that.

PS. Who am I kidding? I don't have what it takes to read a serious technical book any more.

At this point I still stick with Anderson's Fundamentals of Aerodynamics for a standard introduction to aerodynamics. I am less familiar with it but Bertin and Cummings's Aerodynamics for Engineers seems pretty good as well. For more of a mechanical engineering flavor, I think I prefer White's Fluid Mechanics to the various other options (can't stand Fox and McDonald, for example). Chemical engineers seemingly gravitate toward Transport Phenomena by Bird, Stewart and Lightfoot.

The point here is that fluid mechanics is a complicated topic and is approached differently by different disciplines due to the desired end application being different. Once you get to the graduate level, the books converge a bit because they are "allowed" to get a bit more mathematical and fundamental in nature (after all, it's the same physics and mathematics that underpin all of the various undergraduate books). If you are looking more at that level, then something like Incompressible Flow by Panton or An Introduction to Fluid Mechanics by Batchelor are classics and comprehensive.

 

[FactChecker]

At this point I still stick with Anderson's Fundamentals of Aerodynamics for a standard introduction to aerodynamics.

That is not surprising. When I retired, I gave my copy away. It was wasted in my hands anyway.

 

[Aeronautic Freek]

There was an entire thread on this recently. Maybe you were involved, I don't know. The bottom line is that aerospace engineers aren't doing it "wrong" or "right." It's purely a question of what your optimization parameters are when determining an optimal wing shape as well is what is actually manufacturable. Until relatively recently, the shape described by Al Bowers was not something that was readily manufacturable.

this is Al Bowers words,not mine..i just write what he told about "bell spanload distribution" when he was asked, why today aircraft industry don't use this "new" much more efficeint concept...

 

[boneh3ad]

this is Al Bowers words,not mine..i just write what he told about "bell spanload distribution" when he was asked, why today aircraft industry don't use this "new" much more efficeint concept...

He wouldn't be saying that modern engineers are doing it wrong so much as choosing the wrong criteria for optimization if the goal is pure efficiency. The problem is that the standard criteria were selected with more than just efficiency in mind, and given the decades of infrastructural development related to supporting aircraft of the current form factor, doing something with considerably larger wingspan likely won't work with that infrastructure. In other words, moving to a new design like this is more than just "doing it right."

 

[K41]

boneh3ad is correct when he recommends Anderson's book. It is probably one of the best books for graduates studying aerodynamics. It covers everything quite well. If you want a good book on turbulent flow, people often say to use Lumley's book but I'd recommend David Ting's as a short introduction. If you are interested in multiphase flows, then Sommerfeld's book is the one you should get. The seminal book for CFD in turbulent flows is Pope's book but it is quite mathematical in content. I think Anderson has a book which focuses on CFD in aerodynamics.

 

[boneh3ad]

Turbulence textbook talk with no mention of Monin and Yaglom?

 

[K41]

Turbulence textbook talk with no mention of Monin and Yaglom?

I'd say that is more like a very detailed encyclopaedia than a textbook! In my view start with the basics and introductory textbooks, build a solid foundation and if you find yourself fortunate to continue research in that area after your masters/PhD, then you can read those very long seminal books like Bachelor, Monin and Yaglom etc if you have the time. Though in the case of Monin ad Yaglom, they should probably only be used as reference guide unless you study specifically turbulent flows. I studied experimental multiphase flows so I didn't have much time for it.

 

[Swamp Thing]

Hi, there is a good explanation by Professor Holger Babinsky in this paper he wrote here.

From fluids, we know that the pressure increases radially outwards from the centre of curvature (an intuitive explanation is that the pressure needs to be greater on the outside to cause the net pressure acting radially inwards required for centripetal acceleration). A proof is shown on page 503 of the paper above.

If we look at the deflection of streamlines around a wing (images on pages 501 and 498 in paper above) and using the fact that pressure increases with radius, we can first look at the top and recongnise that it must be at a higher pressure than atmospheric pressure (which the pressure assumed for the air above the wing) as the top of the wing is closer to the centre of curvature. Thus $p_{top} < p_{atm}$. If we now look at the bottom and compare it to the streamlines below the wing, we can see that it must be at a higher pressure than atmospheric pressure $p_{bottom} > p_{atm}$. Thus, we have $p_{bottom} < p_{top}$, which provides the pressure difference which causes lift.

I hope this made some sort of sense.

I just read the Babinsky paper, then searched on this thread for the terms "centripetal" and "centrifugal". Apparently there has been no follow-up comment on this idea that lift has to do with the centripetal force required to keep an element of fluid curving downwards. This idea sounds convincing, although one must keep in mind that the exact shape of the streamlines has to be known somehow before we can apply this concept, and those streamlines have to be calculated using diff. eqns or whatever.

But as far as intuitive appeal goes, I find it elegant and simple to think of the air molecules above the wing trying to stay in a straight path, and thus trying to fly away from the wing -- hence low pressure near the top. And air molecules below the wing flying into it, creating high pressure below. What can be simpler?

Having said that, this picture is perhaps not mutually exclusive or contradictory to the "weak" Bernoulli picture where you focus on accelerations along the streamlines rather than across them. Maybe both pictures are valid, i.e.

(1) Start from a point infinitely above/below the wing and integrate pressure differentials along the normal to the flow (based on centripetal acceleration along local radius of curvature) until you get to the wing surface.

(2) Start from a point infinitely upstream and integrate pressure differentials along the flow, based on acceleration along the flow.

The fact that the two methods would give the same answer might be somehow baked into the differential equation, Navier Stokes or what have you.

 

[FactChecker]

Clinging to the hope that there is a single, intuitive, reason for the shape of the streamlines and their associated flow properties (other than Navier-Stokes) is, IMHO, wishful thinking.

 

[Aeronautic Freek]

So every time when fluid goes in curved path ,we have pressure gradient (low p. out,high p. in) which act like centipetal force which force fluid to go in curved path..

But not when water come from pipe at freefall parabola.Now pressure gradient don't exist in curved path ,now gravity act as centripetal force.
am i correct?

 

[FactChecker]

So every time when fluid goes in curved path ,we have pressure gradient (low p. out,high p. in) which act like centipetal force which force fluid to go in curved path..

But not when water come from pipe at freefall parabola.Now pressure gradient don't exist in curved path ,now gravity act as centripetal force.
am i correct?

View attachment 265070

In the Navier-Stokes equations, the external force (usually gravity) is one term and the negative internal pressure gradient is another term. the third force term is the internal viscosity force.

density * flow field time derivative = -pressure gradient + friction (viscosity) + external force (often just gravity)

 

[Aeronautic Freek]

In the Navier-Stokes equations, the external force (usually gravity) is one term and the negative internal pressure gradient is another term. the third force term is the internal viscosity force.

density * flow field time derivative = -pressure gradient + friction (viscosity) + external force (often just gravity)
View attachment 265083

even flow is curving,there is no pressure gradient in that part of water flow,on picture below..
isn it?