Hacker News Reader

Comment by lowkey | Hacker News Reader

Can confirm. I worked at Pratt & Whitney testing jet engines early in my career. At the time I read a similar article and spread it amongst my colleagues - the cognitive dissonance was palpable.

As engineers we had been taught that lift was due to air above the wing traveling faster than air below the wing and thus creating lift by way of a pressure differential. The more accurate answer as seen in the article is that the effect is better explained through Newton’s laws as a re-vectoring of horizontal thrust in a downward direction. Literally the engine pushes horizontally accelerated air downward and the action-reaction mechanic causes an equal but opposite upward force lifting the plane.

Amazing how so many experts could be so wrong in their understanding while the planes continue to fly.

It reminds me of how hummingbirds don’t know that they violate the known laws of physics when they fly.

View on HN

As a child I used to stick a school ruler out of the back window of the car and rotate it slightly to make it move upwards, like a plane's wing. Intuitively I felt that this happened because it was pushing some of the horizontal airflow downwards and the air was pushing back up on the ruler. Yet the books I read about aeroplanes referred to something called Bernoulli's principle which was pretty demoralising because I couldn't understand it.

I suspect, like many other things that didn’t make sense - the reason was that it wasn’t actually true.

The Bernoulli effect explains that lift is due to the design of the wing such that the path above the wing is longer than the path below the wing.

This coupled with the fact that due to the Bernoulli effect an air particle just above the wing would reach the back of the wing at the same time as an air particle just below, and that since the upper particle would therefore have to travel faster than the lower particle the pressure differential would cause lift.

The problem is the theory doesn’t hold up under testing because it isn’t true.

Isn't Bernoulli's principle only applicable when talking about the same flow? I've always found the "above path is longer than the lower path" explanation to be unintuitive because we're not talking about the same flow. They're separate flows.

That may be the correct answer here. The important point to note is that there is no physical reason why the two separate upper and lower streamlines would collude to arrive at the back of the wing at the same time and in fact they do not.

Is it possible to think of.. the roundness on the front disrupting the airflow over the top causing air to become turbulent and less dense on the top. Where as the air flow under the wing high higher relative density and the wing will rise to the less dense position?

Bernoulli's principle, the actual thing, has very strict criteria* to be applicable. People usually neglect this entirely in casually throwing the term around.

- points 1 and 2 lie on a streamline,

- the fluid has constant density (note effects of height difference > gravitational potential energy between point 1 and 2),

- the flow is steady, and

- there is no friction.

I usually just answer Socratically: "So how can (some) planes fly upside-down?" whenever I encounter the Bernoulli-adherents.

Has someone taken an rc airplane and turned the wing upside down and flown it? It should be pretty easy to demonstrate.

RC airplanes have insane thrust to weight ratios, with enough thrust a brick can fly.

That said with my historic experience with single piston RCs if you didn’t compensate for the upside down flight with your ailerons you would nose dive.

Because there is not up or down for the wing when its cutting through a fluid. It is not that we have seen planes flying intercontinental flights upside down.

And those upside down events do not happen at 10 feet above ground. There is plenty of fluid (air) above and below the aircraft and power (fighters jet engines are the most powerful ones on aircrafts) to be able to correct any up-downward force with flaps (basically walls to air)

This is nonsensical. Glider pilots (no engine!) fly aerobatics programs all the time, in planes with curved wings. If the pilot is a bit of a masochist, they could fly upside down for half an hour in still air given enough altitude. The various forms of air resistance will be greater and energy/altitude loss hence higher, but the wing will still be generating lift equal to the weight of the airplane.

> Because there is not up or down for the wing when its cutting through a fluid.

Except that there obviously is. Unless you ignore gravity.

But the "true" explanation given above is that the engine pushes the horizontally accelerated air downwards (with respect to its own orientation). Wouldn't that also lead to the conclusion that upside down flight is impossible?

Changing the angle of attack effectively changes the shape of the airfoil. No gotcha here either.

I'm not immediately convinced that "effectively changing the shape" is a coherent idea. The lift effect either crucially depends on the actual, unchanging shape of the aerofoil or it doesn't. Flying upside-down proves that it doesn't. Maybe all we're disproving is a straw-man of a "Bernoulli-ist" position, but we're disproving it all right.

EDIT: trying to think what you might mean by "effectively changing the shape". Do you just mean that an upside-down aerofoil is a reflection of the aerofoil the right way up? Because that's the entire point of the argument you seem to be trying to rebut.

If you are asserting that the angle of attack does not affect the lift, you are wrong.

A plane flying upside down is most certainly not using the same angle of attack as it does right side up. The real difference in performance is efficiency, the upside down plane is burning more fuel due to the increased drag from sub-optimal operation (a high angle of attack to overcome the optimization for right-side-up flying).

Note that a right-side-up wing can easily plummet by dropping its angle of attack. That is what it's doing while upside down to generate lift.

> If you are asserting that the angle of attack does not affect the lift, you are wrong.

I am certainly not asserting that, and I'm baffled how you could have formed the impression that I was.

You appeared to be attempting to rebut an argument in favour of the significance of angle of attack. We have another pointless internet misunderstanding on our hands.

I am replying to this:

"I usually just answer Socratically: "So how can (some) planes fly upside-down?" whenever I encounter the Bernoulli-adherents."

Which is a lazy and garbled gotcha attempt.

It is a lazy gotcha attempt. It is a lazy attempt at gotcha-ing someone who believes angle of attack ISN'T important. Unless you think a plane flying upside-down is somehow evidence AGAINST the importance of angle-of-attack?

Again: this entire pointless misunderstanding has arisen because you didn't see - apparently STILL HAVEN'T SEEN - which side of the debate the comment you replied to is arguing for.

Oh I see. You think the Bernoulli approach is independent of angle of attack? It's not.

I'm afraid it is most evident that you do not see. Forget trying to guess what I might or might not think about aerodynamics. Just see if you can follow the following recap of the conversation:

1) anvandare says aeroplanes can fly upside down. This is an argument AGAINST a putative person who argues that lift is entirely a function of aerofoil shape, ignoring angle of attack. In advancing this argument, anvandare implies that he DOES understand and contend that angle of attack is significant.

2) You say something unclear about "effective change of shape", apparently attempting to rebut anvandare, who, remember, contends that angle of attack is significant.

3) I say that what you said about "effective change of shape" is unclear, meaning I am rebutting you, meaning I agree with anvandare that angle of attack is significant.

4) You form the impression that I believe angle of attack is not significant, and tell me that if I believe angle of attack is not significant, then I am wrong.

Can you see where you have gone wrong there?

Having written all this, I'm come to the point of actually becoming quite concerned about your neurological state. If you've had a recent head injury or you're old enough that Alzheimers is a possibility, you need medical advice - you've failed to follow the simple thread of a conversation.

As a layman this explanation makes more sense than the "fast air over the top of the wing" one.

Causing a pressure difference is the same as causing a force vector. There's no gotcha here, it's just two ways of looking at the same thing.

My point about the Bernoulli effect is that there is no physical reason why the upper flow should move faster than the lower flow and in fact testing shows that they do not.

The upper flow does move faster, and must for the sake of vortex production.

Please explain what you mean here in simple language. I don’t understand how you conclude that the upper flow must move faster and I don’t understand how it relates to vortex production.

To get a force acting upwards on the wing, there must be a downward reaction on the airstream.

This downward force on the airstream bust change its direction.

This change of direction is a rotation about the airfoil. Specifically a downwards rotation.

The airstream moving above the centerline of this rotation is moving in the same direction, and thus will be accelerated faster.

The airstream moving below the centerline of this rotation is moving in the opposite direction, and so will be decelerated to a slower speed.

The center of this rotation happens to be along the camber line of the airfoil, so all the air above the camber line (ie over the top of the airfoil) must move faster, while all the air below the airfoil must move slower.

The vortex is just the bulk rotational movement of the airflow. In fact, the airfoil can be replaced by anything that will generate the same vortex, like a rotating cylinder.

See my sister comment. The Bernoulli effect explains the pressure differential by way of an above-wing streamline reaching the back of the wing at the same time as the below wing streamline. Since the upper wing is curved and therefore a longer path the theory claims the pressure differential is caused by the upper streamline traveling faster than the lower streamline.

The problem with this theory is that there is no physical reason why both streamlines must arrive at the back of the wing at the same time - and per experimental verification, in fact they don’t.

> The Bernoulli effect explains the pressure differential by way of an above-wing streamline reaching the back of the wing at the same time as the below wing streamline. Since the upper wing is curved and therefore a longer path the theory claims the pressure differential is caused by the upper streamline traveling faster than the lower streamline.

That's not how the Bernoulli effect explains the pressure differential. The bernoulli explanation is that air builds up in front of the airfoil, creating a high pressure region, while there is a low pressure region created in the wake of the wing. This pressure differential forces accelerates air over the wing. For an asymmetric airfoil, more of this flow is over the top than the bottom, so the airflow over the top is faster, and thus lower pressure, than the airflow under the wing.

The "equal time" thing is a pop-science misunderstanding.

That is sadly literally how it is explained both in classes and in textbooks. It’s incorrect but it’s how it’s taught which is the root of a lot of misconceptions and why this “explanation” which is not correct keeps living on.

> a re-vectoring of horizontal thrust in a downward direction. Literally the engine pushes horizontally accelerated air downward and the action-reaction mechanic causes an equal but opposite upward force lifting the plane.

You mean the wing, not the engine.

But even then that doesn't answer the question. It's just another way of looking at the effect, but it doesn't explain the cause. The question is then why/how does the wing pushes that air downward?

> You mean the wing, not the engine.

Yes, clarifying I meant the wing not the engine.

> The question is then why/how does the wing pushes that air downward?

Fair point. I honestly didn’t expect this quality of analysis on the topic.

I believe that another phenomenon is required to complete the explanation in addition to Newton. Couette flow explains why streamlines closest to the wing tend to follow the shape of the wing. Hence the vectoring effect. I’m sure a better article exists but Wikipedia is a bit lacking unfortunately.

[1]Couette Flow - https://en.m.wikipedia.org/wiki/Couette_flow

The experts weren't _wrong_ in their understanding. Bernoulli (creating lift by way of pressure differential) and Newton (reaction to redirection of the flow downwards) are different ways of describing the same thing; integrating either the pressure or velocity vector of the airflow around the wing will give you the correct results for lift.[1]

Whenever people argue about which interpretation of lift is correct I think back to this (https://xkcd.com/895/) comic about teaching how gravity works in general relativity. Only in the case of lift the explanations are actually _correct_, albeit somewhat circular. ("So the air above the wing sticks to the surface, which redirect it downwards. But _why_ does the air stick to the wing?!")

Also in no way do hummingbirds violate any known laws of physics, although they do have a pretty impressive way of harnessing them.[2]

[1] https://www.grc.nasa.gov/www/k-12/airplane/bernnew.html

[2] https://phys.org/news/2005-06-hummingbird-flight-evolutionar...

I commented elsewhere that there is no physical reason why the Bernoulli effect would cause the upper streamline and the lower streamline to reach the back of the wing at the same time - and to my knowledge there is no experimental evidence that it does. I may be wrong about that but I have never seen an adequate rebuttal.

You're right- there isn't, and it doesn't. In fact a parcel of air moving over the top of the wing will beat its counterpart moving below it to the trailing edge. This[1] video has a very good demonstration of this. (Relevant part begins around 0:25.)

So saying that air on top of the wing moves faster, creating a pressure differential and thus lift, is absolutely correct. The problem begins when some people try to come up with an intuitive explanation for _why_ the air would need to speed up. "Because this is the lowest energy state that conserves energy, momentum, and mass" isn't a very satisfying answer; neither is "because this system of PDEs say so"; so they came up with the "equal transit time" explanation, which is simple, intuitive, and completely wrong.

Hopefully aeronautical engineers at P&W didn't actually believe that last bit?

[1] https://www.youtube.com/watch?v=UqBmdZ-BNig

The streamlines don't need to meet for Bernoulli to apply.

That statement, that the upper airstream flows faster because it has to meet up with the lower airstream, is wrong and easy to disprove experimentally.

So, putting aside that specific statement, re-read the post you're replying to.

The airstream above the aerofoil does travel faster, and there is a lower pressure region there (commensurate with the effect described by Bernoulli), and it turns out that this is a valid way to model the forces involved just as it is valid to model them as a redirection of the airstream.

Same as sailing, does the airplane have an equivalent of the keel?

Yes, the engine.

Speaking of sailing, the angle and shape of the sails are both important to maintaining velocity. Racing boats adjust (trim) the shape of the sails all the time.

Merely deflecting the wind obeys conservation of momentum, but conservation of energy in an unpowered sailboat (overcoming losses due to friction) also means extracting energy from the airflow.

Reply on news.ycombinator.com