Interview with Anderson-Eberhardt: Flight Experts

Interview with David F. Anderson and Scott Eberhardt serving as scientific experts on Aeoronautics Learning Laboratory funded by NASA and on The Aviation History Online Museum with the article

How Airplanes Fly: A Physical Description of Lift.

CJ: In your book Understanding Flight and the related article The Newtonian Description of Lift-Revised from February 2009, you state that what is often taught as the physical explanation of lift of a wing, is incorrect; most explanations do not even make sense. You claim that a correct explanation is that lift comes from viscosity. Could you briefly explain in more detail how lift is generated by viscosity?

DFA: Lift doesn't come from viscosity. Viscosity is a characteristic of air that makes flight possible. But then, non -viscous air would be an oxymoron. I don't want to go there. It really is complicated, the concept of zero viscosity. To say lift comes from viscosity is like saying a race car accelerates because of friction with the road. It is part of the answer but not the most important part. Let me give you some quotes from the new book:

A jet engine and a propeller produce thrust by blowing air back. A helicopter’s rotor produces lift by blowing air down. In the same way, a wing produces lift by diverting air down. A jet engine, a propeller, a helicopter’s rotor, and a wing all work by the same physics: Air is accelerated in the direction opposite the desired force.So why does the water bend around the glass or air over a wing? 

First, consider low-speed flight (subsonic). In low-speed flight, the forces on the air and the associated pressures are so low that not only is the air considered a fluid but it is also considered an incompressible fluid. This means that the volume of a mass of air remains constant and that flows of air do not separate from each other to form voids (gaps). 

A second point to understand is that streamlines communicate with each other. A streamline, in steady-state flight, can be looked at as the path of a particle in the moving air. It is the path a small, light object would take in the airflow over the wing. The communication between streamlines is an expression of pressure and viscosity. Pressure is the force per area that the air exerts on the neighboring streamline. Viscosity in a gas or liquid corresponds to friction between solids.  Think of two adjacent streamlines with different speeds. Since these streamlines have different velocities, forces between them try to speed up the slower streamline and slow down the faster streamline. The speed of air at the surface of the wing is exactly zero with respect to the surface of the wing. This is an expression of viscosity. The speed of the air increases with distance from the wing. Now imagine that the first non-zero-velocity streamline just grazes the high point of the top of the wing. If it were to go straight back initially and not follow the wing, there would be a volume of zero-velocity air between it and the wing. Forces would strip this air away from the wing, and without a streamline to replace it, the pressure would lower. This lowering of the pressure would bend the streamline until it followed the surface of the wing. 

In brief, lift is created by the wing pulling air down from above producing an upward lift. It is not by the air striking the bottom of the wing. For a well designed wing, non of the lift is produced by air striking the bottom of the wing. In fact, the pressure is slightly lowered on the bottom of the wing to prevent air striking causing high drag. 

I find it better to understand lift by looking at it in the rest frame where the air is standing still and the wing flies by. A bit of air sees the slope of the top side of the wing as a receding surface. Not unlike a piston being pulled down. This lowers the pressure and causes the air to start moving down. It chases the wing until it gets out of the way and then continues straight down. This is what happens.  Hope this is of some help.

CJ: Thanks for response. It is interesting to follow your argument of the role of viscosity for lift. Would you say that the more viscous a fluid is, the better it is for flying? Is gliding flight possible for little insects?

DFA:  Lift is possible in the limit of zero viscosity. Air is just fine. The more viscous the medium has the greater the parasite drag. Can you imagine flying through honey? Viscosity is the communication of information in the medium. If there were zero viscosity the molecules would not know about the others and they would all end up on the ground. 

I've never thought about insects gliding. Induced power goes as load squared. That is why hummingbirds can hover but not eagles. I suspect that an insect takes so little energy to fly, and distances so short,  that it is not necessary for their wings to be efficient. On the other hand monarch butterflies fly thousands of miles and have large wings. And the space shuttle "glides" with a glide ratio of 4:1. 

CJ: Thanks for valuable information. If you are interested in my own thoughts, please take a look at Why It Is Possible to Fly and related material. Any comments are wellcome.

DFA: Claes,   I started to look at your web site. But it is 5 am and I am about to leave on a trip for a week. There is too much stuff for the moment. I can see that you've collected a lot of material. I'll give the site more time when I get back next week. 

A few key points on understanding flight are: The shape of the wing has nothing to do with lift. Any explanations that rely on the shape of the wing are just wrong. The shape only effects drag and stall characteristics.  All wings have the same lift as a function of effective angle of attack inverted.  Lift requires work. A jumbo jet is diverting about its own weight in air per second. Before the plane came by the air was standing still. Afterwards there is a great deal of air with energy heading down. Energy is not conserved in real flight and so air pressure and speed are not related by the Bernoulli principle. This is not so for the infinite wings that calculations are make on. Stick an Rudder got it right when it says to forget Bernoulli. In fact, until after the WWII only Newton was taught. Then the engineers got computers and forgot how to think. Any explanation the says Bernoulli is wrong. Likewise any explanation that start out with the acceleration of the air causing the lowering of the pressure is violating Newton's first law: A body at rest will remain at rest, and a body in motion will continue in straight-line motion unless acted on by an external applied force. Lift is not a surface effect. A great deal of air from above the wing is involved. Lift on a real wing is from lowering the pressure on the top of the wing with little or no (often negative) increase in pressure on the bottom. So forget the bottom of the wing. All the action is on the top. 

So much for your morning sermon from Reverend Anderson Ciao, David 

PS. Where in the world are you located? I forget where .se is.

CJ: Great that you will take a look. Sweden is the place, the country where the name of most people is Andersson.

SE: I'd be happy to give an interview.  We just completed the second edition of the book, which should be coming out in August.  In the first edition and the paper, we made the point that viscosity is a necessary evil for lift.  But, we didn't say what we meant correctly  and the interpretation many take home is that we say lift comes from viscosity.  Lift is not generated by viscosity, it is generated by pushing air down.  However, without intermolecular communication, which shows itself as pressure and viscosity, there would be no lift.

CJ: Can you comment on the fact that NASA, according to the Interview with NASA, presents three incorrect theories for lift but no theory claimed to be correct.

DFA: I looked at "How Airplanes Work" on this site. It is much better than I have ever seen before on a NASA web site. Their description is essentially correct.

CJ: Correct description? NASA presents three incorrect theories, but which is the correct

theory you indicate is presented? Circulation theory? 

DFA: When NASA says: These velocity variations are caused by the disruption and turning of the air flowing past the wing. They are right on. It is the first time I've seen this in print somewhere and I wish we'd used this line in the book. It neatly sumes up an important part of our book.

CJ: This is hand-waving intended to confuse the reader and the general public and give the impression that NASA understands.  If you know a correct answer to a certain question, there is no reason to present three incorrect answers; in fact there are infinitely many wrong answers, but only one correct. NASA does not claim to present a correct theory, unless handwaving is a theory. Do you really mean that to just say that the turning of the flow of air around a wing is what causes lift, is a theory?

DFA: The air curves because air has the characteristics given it by viscosity. The air is bent by the lowering of the pressure. If it didn't bend there would be a vacuum which would bend the air. This lowering of the pressure propagates out accelerating down much more air above the wing. In the rest frame where the air is moving and the wing is standing still, their explanation is a heck of a lot better than and Bernoulli explanation. It may not be complete but for a change it isn't wrong.

CJ: Maybe it is not wrong, but it does not explain anything either; it only says that what happens, happens. It is not a theory, according to any reasonable scientific standard. The real problem is not only to explain lift, but to explain why the lift/drag quotient is of the order 10-20.Right?

DFA: You've given me something to think about. The reason that this is a mystery to you is that you are thinking in drag not power. If you just look at the drag associated with producing lift, I'm sure that there are gliders with lift to parasite drag on the order of 100:1. When the airplane is sitting on the ground the lift due to the wheels is still the weight of the airplane but the drag is now zero. Lift to drag is just the glide ratio, nothing more profound. It is not like saying the work out divided by the work in is 20:1. You could lift the airplane with a 20:1 ratio with a block and tackle. 

CJ: I don't understand: If lift/drag was only = 1-2, then you would see no airplanes or birds in the sky. Why can lift/drag be as large as 10-20 or even up 70 for extreme glider wings? Have you acted as NASA's flight expert?

DFA: I have spent 35 years investigating flight and never saw anyone make a point that L/D is 20. L and D may have the same units but in reality they are apples and oranges. Flight is not fairly efficient. Yes, if L/D were 1 or 2 there would be no powered flight. The question is why should it be 1? As for NASA the answer is No. I can tell you that I'm sure that the statements that have about the bending of air is probable due to Scott and my work.

CJ:I am amazed: Have you not realized that L/D = 10-20 is what makes flying possible?

DFA: L/D (glide ratio) of the space shuttle is 4. That of a high performance glider is 60. It just relates to the size of the wing compared to the weight. You can make an aircraft with almost any glide ration you want. It's all in Newton's laws. Lift goes as mv and power goes as mv^2. Since lift is mv equals weight (a constant) if you could divert an infinite amount of air (VERY large wings) you would set v to zero and thus power to zero. That would be an infinite L/D. Saying l/D of 1 or 2 is what make flight impossible is just saying that an eagle can't fly with humming bird wings. L/D = 20 is not a miracle by any means. Just physics and engineering.

CJ: I am even more amazed: The space shuttle can land but not take off! A barn door at 45 degrees angle of attack has L/D = 1, and you can't fly on barn door unless you have a very powerful engine. Yes, you can design very slender large wing gliders with L/D = 70 flying very slowly at a small angle of attack, but birds and commerical airflight cannot afford that. To say that flying is just physics and engineering is an insult to the clever birds, right?

It is a common misconception that D scales like L*L, but that is not true for real birds and normal airplanes for which L/D = 10-20 for normal angles of attack say 4-12.

DFA: So insult the birds. A bullet follows a correct trajectory without having taken a physics class. But it is still physics. I'm getting a little tired of this conversation. Induced power does scale as load squared for fixed speed. That is a fact. Since drag is just power divided by speed and speed is a constant in this conversation, induced drag varies as load squared. Parasite drag only changes with load because the angle of attack of the whole airplane changes with load.

CJ: The flight of a a bullet is trivial physics, but the flight of a bird is not trivial.

DFA: Either is string theory but it's still physics.

CJ: Maybe. Anyway, many thanks for giving an interview concerning state-of-the-art. Your input is useful in a continued discussion.