On Airplane lift

One comment on the discussion of airfoil lift: the discussion speaks of a region of high pressure “pushing” the wing up and a region of low pressure “pulling” the wing up. After working for years with vacuum systems, I can assure you that “Vacuum Does Not Suck”. There is no “pulling” action associated with the low pressure region. A net force exists only because the pressure (number of molecular collisions per unit area per second) is greater on the higher pressure side than on the low pressure side. A pressure differential only pushes - it does not pull.
This is the same trick that vacuum cleaner sales associates use to show how your lousy, old vacuum cannot pick up a steel ballm while their slick new machine picks it up easily. Note that the nozzle of your vaccum cleaner over the ball on the floor first, and then turn on the machine. The vacuum is formed uniformly around the ball and it does not go into the vacuum. But when they demonstrate their model, they turn it on first and then approach the ball. A pressure differential is created between the front face and the back face of the ball, and it is pushed towards the low pressure side: the vacuum cleaner.

Welcome to the Straight Dope Message Boards, Tunneller, we’re glad to have you with us.

When you start a thread, it’s helpful to other readers if you provide a link to the Column (or, in this case, Staff Report) that you’re commenting on. The link saves search time and helps keep us all the same page (avoids people repeating what’s in the Staff Report, for example.) In this case, it’s the Staff Report by guest aerodave, which will appear this coming Tuesday (July 12) on the website: How do airplanes fly, really?

No biggie, you’ll know for next time… and, as I say, welcome!

And while I’m here, congrats to aerodave for his first guest report!

I very much enjoyed Aerodave’s article on Airplane Lift.

I would like to point out, however, that symmetrical airfoils are used all the time on things that really fly… and they seem to do just fine. A symmetrical airfoil is a wing cross-section that has the same shape on the top and bottom.

I fly model Radio Controlled aicraft and symmetrical airfoils are used for aerobatic models. One popular aircraft that is modelled is the Extra-300, which (I believe) uses a symmetrical airfoil on the real aircraft.

In this case, there is nothing but AOA and the ability of the engine to pull the aircraft through the sky. A symmetrical airfoil truly does “plane” through the sky just like waterskis plane across the water.


the leading edge of a symmetric airfoil is round.
there’s a point on the leading edge that separates the air that goes over the top and the air that goes underneath.
when the angle of attack is positive, this point moves towards the underside of the airfoil.
so from the point of view of the air molecules that generate the lift the airfoil is no longer symmetric.

While I’m no aeronautical engineer, this portion of aerodave’s report apparently isn’t totally correct:

In fact, flat plate wings do work, at least at the Reynolds Numbers encountered in small model airplanes.

Here is a site that specializes in flat plate winged planes, with links to a few videos that appear to show them actually flying.

Not sure how this all figures into the original report, just thought it might be worth consideration.

Considering all the threads we’ve shredded over the Bernoulli effect in the past, I was a little concerned to see a staff report that not only spoke positively of it, but also centrifugal force and magic. :slight_smile:

As near as I can tell, the centrifugal force idea is in relation to a stationary wing, is that correct? Sort of a wind tunnel description?

Given all that was said in the article, especially regarding the shape and contour of the upper surface of a wing, why don’t wings have ridges, like Ruffles. Seems to me that would greate a greater surface area and lots of mini-wing front edges. Or perhaps surface dimples, like on a golf ball…?

I see you are capitalizing Ruffles. Have you trademarked it? :slight_smile:

Besides, wings already have Rivets.

Thanks, Dex, for the congrats and to everyone for the quick interest. I should first point out that I’ve just seen the article in its current for for the first time this afternoon. Some things didn’t make it past Ed’s edit, and I’m not sure if any of those revisions changed the message at all.

I will address a couple of the comments so far, but expect that some minor things may get tweaked before it hits the street on Tuesday.

  1. Flat plate wings are real, and do make some things fly. Balsa gliders, small R/C planes, etc. But for large scale things at reasonable speeds, they’re horribly inefficient. I did not mean to imply they don’t work…just that they’re not as good for most applications as a shaped airfoil.

  2. Symmetric (uncambered) airfoils also make perfectly good wings. And I don’t believe anything in my article says otherwise (although I’d appreciate it being pointed out if it does). I think Skadar is confusing my mention of flate-plate airfoils with symmetric ones. The Extra 300 does, in fact, have symmetric airfoils…but they also have thickness–that is, aren’t just flat plates. The only thing worth mentioning about symmetric foils is that they produce no lift at zero AOA. Cambered airfoils produce some lift even at zero AOA.

  3. RM Mentock, the centrifugal force analogy does make the most sense if you picture a stationary wing in moving air. Aerodynamicists always think of the body as being stationary, and the air as moving. So picture yourself moving over the wing the way you might drive your car really fast over a crest in the road…the negative G’s you feel on the way over the top (way fun, I know) are analogous to how the air is pulled away from the surface.

  4. And I realize that vacuums don’t suck, and that high pressure pushes. But raising the pressure on one sode of something, and lowering it on the other are functionally the same things. The fact remains that if you integrate the pressures over the whole wing, you’ll came up with the right number for the lift. The question really becomes one of, “Why does that pressure distribution take the form is does?”

One overarching point I tried to make, but didn’t make it to the final print, was that it’s hard to separate the different phenomena going on around a wing. Describing the flow deflection is just one way that lift presents itself. Describing the varied pressure distribution around the wing is equally valid, and is just another manifestation of the same thing. They are, in fact, necessary consequences of each other. In other words, the flow won’t turn down around the wing unless you also find a low-pressure area above. All these things are interrelated, and codependant, and that’s what makes describing lift so hard.

I’m one of the people who had his arse handed to him on a platter for supporting Bernoulli in a thread a few years ago, so I don’t want to get into a debate over Newton vs. Bernoulli.

I’m coming into this thread to ask about two things: Lifting bodies, and supercritical wings.

As can be seen in this photo, the M2-F1 Lifting Body is curved on the bottom and flat on the top. I’ve downloaded Robert G. Hoey’s Testing The Lifting Bodies At Edwards. Actually, I downloaded it years ago, but never printed it out. I don’t see the .pdf file at the NASA Dryden site, but I did find it on my computer and wrote a copy to CD. Since I don’t like reading long .pdf files (this one is 218 pages), I still haven’t read it. I’m sure it contains information on the aerodynamics of Lifting Bodies, but I wonder if someone could encapsulate them.

The supercritical wing is flatter on the upper surface than on the bottom – the reverse of a conventional airfoil. How does this work?

Great report. Now for some questions.

  1. Ok, back to [del]the explanation that I once read in the World Book Encyclopedia[/del] the Bernoulli principle. If an equal volume of air approaches a wing, and the top half of it is spread out over a larger volume than the lower half, I would think that the upper volume would acquire less density. Less density=lower air pressure, etc. etc.

Is this wrong? Confused? I understand (now) that there are 2 other effects such as AOA and downwash.

  1. “That’s why we have a hard time predicting how real bodies will behave, and why experimentation is so important in aeronautics.”

Do wind tunnel experiments tell you all you need to know? How much extra do you learn when the full-scale mockup is done?

  1. Air is a fluid? I thought it was a gas. (Compare and contrast the properties of air and water, in this context.)

  2. What about that Coanda effect? Is it driven by air pressure or atomic-scale effects? (Or is this a dubious question?)

In science and engineering, “fluid” means liquid or gas.


Airplanes fly by the mass they accelerate.

The action of a plane on the air causes an equal and opposite re-action on the plane.
Please consider the following:

Mass behaves paradoxically only at extremes. The transformation of solid to fluid is not extreme. Mass behaves as mass even when it is fluid.

Bernoulli showed that fluid pressure is inversely proportional to fluid velocity. In other words, even the airflow accelerated parallel to the wing provides lift. IAMA mathematician, but I understand this force to be the inverse square of parallel acceleration. This is compared to force in the multiplication of mass and velocity vector perpendicular to the wing.

Airplanes fly by the mass they accelerate.
Only when we understand this, can we pursue the details.

Only through Liberty


You’re including jet engines and propellors in this, right?

aerodave, your answer is so much better than most. Since I speak ADD, I gave an answer for the people that have yet to read your whole answer. Like you, I also have found that “Nearly all of the common ‘theories’ are misleading at best, and usually flat-out wrong”. The sad thing is that these are from “authorities”.

Now that I have had time to look for nits, please allow some gentle picks.

This statement contradicts my understanding. Although you seem to recognize that Coanda and Magnus forces are Newton; it is my experience that few authorities do. You may wish to review the source behind your statement.

The angle of attack relates to drag. It is the angle of departure that determines lift. A wing with a negative angle of attack can have a positive angle of departure. Those pesky authorities with their misleading theories!

I look forward to further discussion.

Only through Liberty


  1. I don’t know what you mean by “spreading over a larger area.” Rephrase if if you can, and I’ll be happy to help try to clear it up.

  2. Development of an aircraft, or part of an aircraft, is a long process, and is an exhaustive mix of wind tunnel experiments, CFD (computaional fluid dynamics, or computer simulations) and flight tests. Used to be some paper analysis and lots of windtunnel work. CFD has lessened the load of WT-based development somewhat in more recent years. We’re finally getting to the point where they play well together, and complement each other. Any real system nowadays will have a ton of both in its development.

  3. Air is a fluid…but this was already gotten by someone else.

  4. It’s all related to turbulence and viscosity. If there were no viscosity, air wouldn’t follow a curve, it would just “jump the ramp.” Imgaine air cresting the hump at the top of the wing, but there’s no friction, no viscosity. The air won’t stay over the convex curve, it just flies off in a straight line. More fluid is left stagnant on the back side of the hill, just sticking to the wing.

Now, let’s switch on the viscosity like magic. Now, the air that wants to fly off the top of the hill rubs against the still air below. Before, there was no rubbing because there was no viscosity (the fluid could slide past itself without friction). As the ski-jumping air rubs against the still air, the top portion of the still air starts moving, too. And it rubs the air below it, which moves and starts rubbing the air below it…and so on…until you’ve reached metal. We’d say the fluid below is “entrained.” What it really means is that before ling, you can’t tell the ski-jump air from the entrained air. And since there’s only so much energy to go around, and so much of the energy from the ski-jumping air is dumped into accelerating the air below, there’s hardly any ski-jumping air left.

The net result? Air that would have jumped off the top of the curve was transformed into air that hugs the curve. This should give you a vague idea of why air would rather stick to a curved surface than fly off. Or not…it’s really hard to get this stuff across without doodling.

A salute of Saint Phonzie to RM Mentock

Its all Newton.

The physics are the same whether in the perspective of the wing or the airflow. I prefer to use the word inertia. As the air passes the height of contour, inertia creates an area of lowered pressure above the wing.


At high speeds, shaped airfoils are horribly inefficient; I do not mean to imply they don’t work… just that they’re not as good as a flattened airfoil. Shaped airfoils create excess drag at high speed.

The ideal wing shape is the one that most efficiently deflects mass from any given angle of attack through any given angle of departure. At low speeds, these angles diverge, at high speeds, they converge. A flatter and thinner wing is more efficient at high speeds.

Hear, hear. I would even say it is impossible to separate the phenomena. It’s all Newton.


It is so wonderful to see Real Scientists™ admitting that we really don’t have all the answers; we just use what works!

Yeah, but they have rivits on BOTH sides! No? Negative gain.

But, a wonderful idea, if the principle is valid, let’s invent wings with rivits only on the top side! :slight_smile: