Talk:Lift (force)/Archive 5

Talk page archive

Archive 4 has been created with a link at above right. It is an exact copy of the talk page as it was before this edit (besides the last four sections which I left here). Archive 5, when needed in the future, should be a new subpage (same as creating an article) titled "Talk:Lift (force)/Archive 5" and the link added to the template on this page's code. For further information on archiving see Wikipedia:How to archive a talk page. See also User:5Q5 for the used archiving procedure. Thank you. -- Crowsnest (talk) 01:23, 15 May 2010 (UTC)

Re-casting Deflection Theory section in plain prose

Per the discussion above, I've taken a stab at recasting the deflection theory in plain language, removing the quotes from the article and placing them in footnotes. I think this approach is more readable to the general public than the 'cut and paste' set of quotations that is there now. The proposed new language is in my sandbox: http://en.wikipedia.org/wiki/User:Mr_swordfish/Lift#Deflection Please take a look and offer comments or edits.

One thing that would help the article is a nice picture or diagram for this section, but I haven't found a public domain one yet. Suggestions cheerfully appreciated. Mr swordfish (talk) 21:20, 3 May 2010 (UTC)

Thanks for the work you have done here, and for the invitation to comment. I have made my comments at User talk:Mr swordfish/Lift#May 2010. Dolphin (t) 00:09, 4 May 2010 (UTC)

As of tomorrow it will be ten days since I posted the edits in my sandbox. Dolphin, Crowsnest and I have made some minor tweaks in that time, and while it is still a work in progress (and probably always will be) I think it's time to merge the changes into the main article. Pending any objections, I'll make the changes live tomorrow. Mr swordfish (talk) 17:25, 13 May 2010 (UTC)

Bernoulli's Principle

Today I reverted Sliceofmiami's good faith edit to the "equal transit time" section of the article. If we want to present a description of lift based on Bernoulli's principle (and perhaps we should) it deserves it's own section, not a rear-guard action tacked on to the end of equal-transit-time.

The difficulty in presenting a simple explanation based on Bernoulli's principle is explaining why the air changes velocity as it flows past the airfoil. I don't necessarily disagree with Sliceofmiami's statement "lift is still generated by Bernoulli's principle", but without some supporting arguments explaining why the velocity changes the assertion is out of context.

The next section (a more rigorous physical description) explains it pretty well, but is not exactly simple. My take is that it's not possible to explain lift simply via Bernoulli's principle, but I'm all ears if someone wants to take a crack at it.

BTW, to answer his question "Bernoulli's principle is discussed negatively in this section, why?", I don't think Bernoulli's principle is discussed negatively, only the mis-application of it via equal transit time. Mr swordfish (talk) 13:37, 14 June 2010 (UTC)

I agree that Bernoulli's principle on its own is not sufficient to explain lift because it doesn't explain why the air travels faster past the upper surface of the airfoil than the lower surface. In the past I have grappled with this predicament by incorporating the Kutta condition. When a reader is comfortable with the Kutta condition, the extra information provided by Bernoulli's principle should complete the picture. In addition, the vortex necessary to meet the Kutta condition completes the equation at Kutta-Joukowski theorem so that the amount of lift can be considered to be quantified. Dolphin (t) 04:08, 15 June 2010 (UTC)

Bernoulli's principle gives the pressure (dynamics) provided the velocity field (kinematics) is known (in incompressible flow). As already said by Dolphin51, the problem is in the determination of the flow kinematics. The equal transit time assumption is known to be false. Also the descriptions purely based on blockage and mass conservation (Lift (force)#In terms of a difference in areas) are inadequate. As pointed out by Dolphin51, the Kutta condition in combination with irrotational and incompressible flow are sufficient to provide the flow kinematics (as a potential flow). However, this is not a simple explanation: the Kutta condition has to be justified through Prandtl's description of boundary layers, and the flow fields from potential flow theory require quite some math (but have -- according to Lord Kelvin -- minimum kinetic energy).

However, we could rephrase the last sentence in the section from a double negation ("...does not imply that Bernoulli's principle is incorrect...") to a more positive one, e.g.: "Note that while this theory depends on Bernoulli's principle, the fact that this theory has been discredited is independent from the validity of Bernoulli's principle". -- Crowsnest (talk) 09:00, 15 June 2010 (UTC)

Weltner & Ingelman quote

I removed the Weltner & Ingelman-Sundberg quote:

The cause of the aerodynamic lifting force is the downward acceleration of air by the airfoil

— Klaus Weltner and Martin Ingelman-Sundberg

since it does not seem to be written by someone who has published much on the subject of lift in renowned journals in the field, or is often cited. Nor does it seem to have been published in a peer-reviewed article, or being a reliable secondary source. So, why should this particular quote go into the article? If such a quote is needed, are there no better phrased alternatives? Also note that the wording is quite opposite to the common wording of Newton's 2nd law: this quote says that force is caused by acceleration, while the common description is that net force causes acceleration. So, instead of an explanation, this quote may confuse. (Apart from wordings, as in the original phrasing by Newton, of the 2nd law as a balance between net force and momentum change rate without reference to cause and effect). -- Crowsnest (talk) 23:42, 23 April 2010 (UTC)

I have no objection to Lift (force) supplying a number of different explanations for the phenomenon of lift, and I have no objection to Lift and the deflection of the flow being one of those explanations. It has a legitimate place in this article.

I see no benefit in having one or more of the different explanations supported by a quotation, regardless of the significance of the author of that quotation. If one of the different explanations is to be supported by a quotation, they should all be supported by such a quotation.

The quotation from Weltner and Ingelman-Sundberg is probably technically correct when surrounded by the complete text by these authors. However, this quotation, isolated from its surrounding text, is without merit. Essentially, all it says is the cause of the ... force is the ... acceleration of air ...

There is nothing in physics, and certainly nothing in Newton's 2nd or 3rd Law of Motion, that says the cause of a force is an acceleration. So taking these words out of their intended context does Weltner and Ingelman-Sundberg a great disservice. Without the support of the surrounding text provided by these two authors, this quotation is technically incorrect.

If each of these different explanations for lift is to be supported by a quotation that quotation must be eloquent and scrupulously correct. The quotation by Weltner and Ingelman-Sundberg is neither of these, so I am not in favour of it being used in Lift (force). In fact, I think it is sufficient to state each of these different explanations and provide suitable in-line citations. Quotations are not warranted. Dolphin (t) 05:21, 24 April 2010 (UTC)

Dolphin,

I do not think that the quote is taking Weltner out of context. Here is the entirety of the Abstract:

Abstract. The conventional or standard explanation of aerodynamic lift states the higher streaming velocity at the upper side of the airfoil as cause of the lower pressure, due to Bernoulli’s law. But a higher streaming velocity is the effect of a lower pressure and never its cause. The cause of the aerodynamic lifting force is the downward acceleration of air by the airfoil - which depends on the angle of attack and its velocity.

In relation to the airfoil the normal acceleration of the air in case of curved streamlines must be regarded which results in pressure gradients perpendicular to the streamlines and reaction forces acting perpendicular on the deflecting surfaces.

And here is what he says in the summary, in it's entirety:

The conventional explanation of aerodynamical lift based on Bernoulli’s law and velocity differences mixes up cause and effect. The faster flow at the upper side of the wing is the consequence of low pressure and not its cause.

The generation of lift by an airfoil can be explained correctly and simply taking the downward acceleration of air into consideration. This approach allows to derive the dependency of lift from angle of attack, flow velocity and the air’s density in a streightforward and coherent way.

A detailed explanation of the generation of pressure differences is possible if the normal acceleration of streaming air is taken into consideration.

I don't think it could be much clearer what he is saying: deflection of the air causes lift. See the quotation from the American Journal of Physics in my reply to Crowsnest below for further corroboration.

As to the question of whether it merits a quotation as opposed to an inline citation, that's a question of style. I don't claim to be an expert on Wiki style, so I'll defer to those who are.

My actual preference would be to dispense with the cut-and-paste style of a quote from Weltner followed by a quote from Anderson followed by a quote by Langewiesche and replace it with plain prose (with citations, of course). Stitched-together quotes are less readable than plain statements and the article would be improved by simply describing the theory. Unfortunately, the subject is enough of hot-potato that we (the various editors) are ultra-careful about the language and have resorted to quotations in order to protect ourselves from criticism by other editors. The result is that the article evolves to be written for the editors rather than for the general public. Unfortunately, I don't know that it's possible under the current circumstances to come to agreement on language.

Mr swordfish (talk) 20:33, 26 April 2010 (UTC)

Hello Mr swordfish. Thanks for your comprehensive and thoughtful response.

I fully agree with your final paragraph. (I particularly agree with your comment that pasting of quotations in this section of Lift (force) is related to the contentious nature of the subject. I think this Talk page is the place for the quotations so that all interested editors can see that in-line citations accurately support the text in the article.) Let’s move the quotations from the article to this Talk page, and rely on plain prose that will best serve the general public.

The first sentence at Lift (force)#Description of lift on an airfoil is There are several ways to explain how an airfoil generates lift. I fully agree with this statement and I am glad it is the opening sentence in this potentially-contentious section. There are too many people who seem to be saying My favourite explanation of lift is correct and therefore all other explanations must be incorrect. The quotation by Weltner and Ingelman-Sundberg is dogmatic in that it asserts a single cause of aerodynamic lift. It fails to acknowledge that this is one explanation of lift, but not the only one. This quotation is incompatible with the first sentence which establishes Wikipedia’s position that there are several ways to explain lift. The quotation belongs on this Talk page so we can all see exactly what the in-line citation is pointing to. Regards. Dolphin (t) 23:18, 26 April 2010 (UTC)

Crowsnest,

Google Scholar, while a useful tool, is not comprehensive. In particular, your search missed Weltner's seminal paper in the American Journal of Physics 1987 vol 55 No. 1 A comparison of explanations of the aerodynamic lifting force. This paper is in a peer reviewed journal, but unfortunately is behind a paywall. It is similar to the online paper cited by the wiki article, and contains the following quote:

"Textbooks stating that the higher streaming velocity is the reason for the low pressure are wrong. It is the other way around. The low pressure is the reason for the higher velocity of the streaming air." (emphasis in the original)

Although the paper cited is self-published, Weltner certainly qualifies as a reliable secondary source. The online paper says basically the same thing as the AJP paper; I could have quoted the AJP paper instead, but the online article said it more succinctly, and it seems valuable to have an easily accessible reference rather than one that's behind a paywall.

Newton's 2nd law is usually presented as F=ma. Whether force causes the acceleration or the other way around is open to interpretation. For example, when you are sitting in the passenger seat on the bus and the driver presses the accelerator you feel a force from the back of the seat. The force you feel is caused by the acceleration of bus; it would be absurd say it was the other way around, i.e. to insist that you are making the bus speed up by pressing on the back of your seat. So, in some situations, it makes complete sense to say that acceleration causes a force.

As for the quote "confusing" the reader, there is nothing confusing about the quote at all: it is the simplest, most direct and straightforward statement of the deflection theory. That's why it was chosen. I understand that you disagree with it. That's your prerogative. Mr swordfish (talk) 19:45, 26 April 2010 (UTC)

The quote is from a self-published source, so in general not acceptable. A possible reason to still include it in the article would be, when Weltner and Ingelman-Sundberg are renowned experts in the field. And the Google search was just intended to get an indication about that (how much have they published on the subject), as well as an indication about the amount of citations to their work.

Nor is it logical to take a quote from a self-published source because it is "basically the same thing" as what was said in another, peer-reviewed paper by the same author. Wikipedia is full of refences to articles in peer-reviewed papers which are in general not freely accessible for everyone, and references to for many difficult to obtain books. There is no policy stating that quotes should be from easily accessible references.

The example given about the bus is off-topic, while the representation F=ma is not a suitable depiction of Newton's 2nd law for use in fluid-dynamical flow modeling. Nor do I believe that in fluid dynamics, in general, flow forces cause acceleration or the other way around. On the contrary, in fluid dynamics most flows are balances between pressure (and viscous stresses), and momentum changes; in complex interactions.

People like John D. Anderson (note 9 at this moment in the article), and Landau & Lifshitz (note 26) are very careful in their wording of the relationship between lift force and momentum changes, preventing to end up in endless discussions about cause and effect (which may be useful for the philosophy of science).

Your opinion or my opinion on what is "confusing" or "the simplest, most direct and straightforward statement of the deflection theory" is of limited importance. The point is that a stand-alone quote in that location in the article has to represent scientific consensus (or the consensus within the notable scientific faction it stands for). The cause-and-effect formulation of Weltner & Ingelman-Sundberg gives an opinion, not a scientific fact that is widely accepted (notable in the WP sense). -- Crowsnest (talk) 07:45, 27 April 2010 (UTC)

Mr. Swordfish, you wrote "it would be absurd say it was the other way around, i.e. to insist that you are making the bus speed up by pressing on the back of your seat."

Did you mean to say, "it would be absurd say it was the other way around, i.e. to insist that you are creating a force on your back by accelerating." ?

I think so. You were making the (good) point that simply reversing F and a in the cause and effect statement seems intuitively absurd in this case. But the F in statement 1 is the forward force exerted on the back, not the backward force by the back, and the a is the acceleration of the person, not of the bus.

If we make the reversal one of switching equal and opposite forces carefully, we come up with "The backward force of the back on the bus causes the acceleration of the bus to be lower than it otherwise would be" which is different than the way you wrote it. In this case there is no intuitive absurdity at all. Both cause and effect statements are equally intuitive.

Crowsfoot is right that F=ma, or its fluid dynamics equivalent, doesn't by itself indicate an asymmetric cause and effect relationship. If causes must precede effects, then we could even argue that F=ma forbids a either causal interpretation, since it refers to instantaneous values of both F and a (and m).

But its helpful to the reader, I think, to go with the intuitively appealing bias that net forces (pressure differences) cause accelerations (of bodies or particles, or for us, of differential volumes of fluid). The real difficulty comes when the reader wants to know if the pressure field is caused by the flow field, or vice versa. We must get him/r to abandon this truly meaningless question, I believe, and understand that Newton's differential equation, and the continuity and mass conservation constraints all have to hold at the same time. The infinite set of 3d velocity vectors and the infinite set of local pressures have to be consistent with each other, and there is no intuitively meaningful way to say that one causes the other.

91.103.41.50 (talk) 21:48, 27 October 2010 (UTC)

The animation contradicts the text?

The animation that is supposed to show the flow on top is faster than the flow on the bottom appears to show the top flow slower than the bottom, visually it doesn't look faster, and after the air leaves the airfoil the top part is behind the bottom part; is it an optical ilusion or is the animation really showing the top flow being slower? --TiagoTiago (talk) 00:48, 27 August 2010 (UTC)

It is neither. You are not looking at the animation carefully enough. The flow on top is faster, particularly between the leading edge of the airfoil and about the quarter-chord position. Above the airfoil the spacing between the vertical black dotted lines is about twice the spacing of the lines below the airfoil. That shows the average speed above the airfoil is about twice the speed below the airfoil.

The vertical black dotted lines in the fluid are all one color, but if they were alternating colors (say alternating red and blue) you would see clearly that upstream of the airfoil each vertical line was all one color but downstream one color above the wake would almost line up with the other color below the wake. It is a brilliant piece of animation and there is no error or optical illusion. Dolphin (t) 03:02, 27 August 2010 (UTC)

The animation is misleading. Using the alternating red and blue lines for example, after having passed over the aerofoil red lines are almost in-line with blue lines in the current animation. But in an ideal flow, as is illustrated in the animation, the red and blue lines would remain in-line and unnaffected by the aerofoil at distances sufficiently above and below the aerofoil. The animation does not make this clear and it looks as if the presence of the aerofoil affects the entire column of air above and below it. It's hard to describe but I hope you understand. 86.159.128.111 (talk) 23:05, 5 January 2011 (UTC)

I do not see alternating red and blue lines. I suppose you mean the vertical lines of black dots (time lines). You are right that in ideal flow the influence of the airfoil becomes smaller and smaller, far away from the airfoil. But the region shown in the animation is by far not large enough to show this. The scale of the animation is limited due to restrictions on a workable file size (band width) and resolution. The main feature to show is the fact that the flow above the airfoil is faster than below, and that air parcels just above and below the dividing streamline do not catch up behind the trailing edge. A second animation of a much larger region (say five times in diameter) would be needed to show that the time lines far above and below the airfoil are unaffected by the presence of the airfoil. -- Crowsnest (talk) 08:21, 6 January 2011 (UTC)

Very basic summary

So, as a very, very basic summary that glosses over it, would it be fair to say that the airfoil's angle of attack or camber "deflects" the upper stream towards other layers, compressing it and causing it to speed up, while it allows more room for the lower stream, expanding it and causing it to slow down?208.249.136.187 (talk) 14:24, 2 November 2010 (UTC)

It's fair to say that the airfoil deflects the air, but not that it compresses it. Almost every mathematical model of lift assumes the fluid is non-compressibile, and in the few models that do model compressibility it is a minor effect at best. IOW, while there are pressure changes, there are no density changes.

(This is wrong. Gasses can't change "pressure" like that (independent of other measures). The air traveling under the wing of a jumbo jet would be hot enough to ignite the aluminum if it maintained constant density. The mathematical models assume incompressible fluids because they are created by engineers who are (generally) lazy and do shoddy work. Engineers work only to approximations; you tarnish mathematicians by calling these models "Mathematical", they are merely arithmetic. I also don't see how a windtunnel could be modeled at all without adjusting for compression.76.126.215.43 (talk) 15:54, 5 January 2011 (UTC))

If you want a very, very basic summary that glosses over the details, air goes down, plane goes up works as well as any. Or for a sailboat,air goes backwards, boat goes forward. - Mr swordfish (talk) 14:44, 2 November 2010 (UTC)

208.249.136.187 has written compressing it and causing it to speed up. This is incorrect. Bernoulli's principle asserts correctly that where the pressure of a fluid increases its speed reduces. If 208.249.136.187 had written compressing it and causing it to slow down it would have been correct.

Similarly, 208.249.136.187 has written expanding it and causing it to slow down. Again, this is incorrect. If a fluid expands it is because its pressure has reduced. Bernoulli's principle asserts that where the pressure on a fluid reduces, the speed must increase. Dolphin (t) 21:40, 2 November 2010 (UTC)

There seems to be two meanings of the word "compress" in play here. If by "compressing" you mean increasing the pressure then what you (Dolphin) says is correct. But the usual meaning of compressibility in the context of fluid mechanics relates to changes in density not pressure. See http://www.centennialofflight.gov/essay/Dictionary/Compressibility/DI136.htm

In laymans terms compression can mean either an increase in pressure or an increase in density. In fact many people don't even realize that there is a difference; in many circumstances the two occur simultaneously (e.g. pumping up a tire). But I think it is important for the article to stick to the conventional scientific meaning (i.e. increasing the density) and not use "compress" to refer to an increase in pressure. If nothing else, it will confuse people who read further into the article and encounter the discussion of incompressible flow. Mr swordfish (talk) 15:17, 3 November 2010 (UTC)

In any very basic summary of lift, or even a basic summary, we must assume air to be of constant density. Once we begin talking about air as a compressible fluid, and speeds greater than about 30% of the speed of sound, we are no longer in the realm of a basic summary of lift

A basic summary is appropriate to understanding the flight of birds, gliders and most propeller-driven airplanes. As you know, for an understanding of these vehicles the phenomenon of compressibility is unnecessary. Dolphin (t) 22:12, 3 November 2010 (UTC)

Page directly contradicts diagram

There is a motion image in "A more rigorous physical description" which appears to show a column of dots representing equi-sized air parcels as a wing moves through them. This image clearly shows the parcels above the wing being displaced forward (in the wing's FOR) from the parcels below the wing.

However the text in at least two places directly contradicts this observation:

Same section:
The upper stream tube constricts as it flows up and around the airfoil, a part of the so-called upwash. From the conservation of mass, the flow speed must increase as the stream tube area decreases. The area of the lower stream tube increases, causing the flow inside the tube to slow down. It is typically the case that the air parcels traveling over the upper surface will reach the trailing edge before those traveling over the bottom.

Section "Popular" explanation based on equal transit-time:
However, equal transit time is not accurate and the fact that this is not generally the case can be readily observed. Although it is true that the air moving over the top of a wing generating lift does move faster, there is no requirement for equal transit time. In fact the air moving over the top of an airfoil generating lift is always moving much faster than the equal transit theory would imply. —Preceding unsigned comment added by 76.126.215.43 (talk) 15:46, 5 January 2011 (UTC)

I don't see a contradiction. You have written the parcels above the wing being displaced forward. What do you mean by forward? I see the parcels above the wing being displaced downstream, in the direction of motion of the air, relative to the parcels below the wing. The parcels above the wing acquire greater speed than the parcels below the wing. Dolphin (t) 03:34, 6 January 2011 (UTC)

Can you further clarify where you see contradictions? -- Crowsnest (talk) 08:28, 6 January 2011 (UTC)

Request for real world flow diagram (as opposed to idealized flow)

The current animated diagram doesn't show the "big picture" effect of the flows further away from a wing. This avweb article includes an image showing the flow and one animated diagram with a vertical column of dots and a wing passing through, that gives a more realistic picture of how the air is affected by a wing passing through it:

avweb_article.html

Since the flow speed and size of the wing in the current animated diagram (the one with the colored and black dots) isn't known, I don't know how realistic it is, but the upwash effect is exaggerated compared to the flow you'd see with a real airplane. In the real world, starting at a sufficient distance above and in front of an approaching wing, there's virtually no upwash, just an acceleration backwards and downwards towards the low pressure area above the wing which transitions into a mostly downwards path as the wing passes by. I'd like to see an animated diagram showing this "macroscopic" view of the air flow caused by a wing passing through the air.

Jeffareid (talk) 10:00, 22 February 2010 (UTC)

You can take a look at this video and see the upwash and downwash (as well as the faster flow above the wing; non-equal transit time) in a "real flow". -- Crowsnest (talk) 19:45, 12 May 2010 (UTC)

I have serious issues with the current animated diagram:

• If you look at the air far above and far below the wing, the effects of the wing should become negligible and thus the timelines must move at the same speed. As it is, it looks like all air below the wing moves one black line slower than above the wing. That can't be true.
• Given the arguments laid out below "The upper streamtube is squashed the most in the nose region ahead of the maximum thickness of the airfoil, causing the maximum velocity to occur ahead of the maximum thickness." That squashing effect is more pronounced directly above the wing and fades higher above the wing. Thus the air moves faster directly above the wing than higher up. Thus the black lines must bend to the right and not to the left.
• Or put another way: the air is supposed to speed up above and slow down below the wing: thus the black lines cannot bend in the same direction. To the right of the wing, the black lines of the diagram should look something like this:

  |
  |
   \
\
 \
  |

-- ol@bofh.priv.at 83.136.33.3 (talk) 09:42, 20 January 2011 (UTC)

See also above animation: avweb_article.html, giving the same result as the one in the Wikipedia lift article. The maximum Darwin drift occurs further above the airfoil due to the flow circulation around the airfoil. See further the discussions further on with respect to the animation. -- Crowsnest (talk) 10:07, 20 January 2011 (UTC)

I think the confusion ties to what "far above and far below the wing" means. The animation only shows air about one chord-length away from the airfoil -- not nearly far enough away to see relatively "undisturbed" air. 214.4.238.180 (talk) 18:12, 24 January 2011 (UTC)

YC-15 and Coandă effect

The article states "Two aircraft, the Antonov An-72 and An-74 "Coaler", use the exhaust from top-mounted jet engines flowing over the wing to enhance lift,[35] as did the Boeing YC-14 and the McDonnell Douglas YC-15. [36][37]". This seems to imply that the YC-15 has top-mounted jet engines, with exhausts flowing over the wings, which is contradicted by the references and pictures of the YC-15. Reference [36] http://www.theaviationzone.com/factsheets/amst.asp described the mechanism used in the YC-15 as "under-surface blowing" and reference [37] http://www.globalsecurity.org/military/systems/aircraft/c-14.htm described it as "Externally Blown Flap", as opposed to the "Upper Surface Blown Flap" of the YC-14. —Preceding unsigned comment added by 169.143.244.115 (talk) 13:21, 6 January 2011 (UTC)

I agree that the article creates a misleading impression concerning the YC-15 which does not have top-mounted engines. It has underwing engines and when wing flaps deploy they move into the engine exhaust stream, causing the flaps to be externally-blown flaps and therefore take advantage of the Coandă effect.

The offending sentence is primarily about the AN-72 and AN-74 and the fact that they have top-mounted engines. The simplest solutions to the problem would be simply to delete references to the YC-15, or create a new sentence in which externally-blown flaps are cited as another application of the Coandă effect. Dolphin (t) 21:36, 6 January 2011 (UTC)

The flow image

Just my two cents, the image could include pressure information as well as a force vectors. You can see that the air is more dense in some areas and less dense in others by looking at the dots, and you can follow the dots to see how the forces act in the air, but the actual reason why the aerofoil is "lifted" is missing from the image. —Preceding unsigned comment added by 80.75.107.170 (talk) 12:22, 19 January 2011 (UTC)

What you "see" actually *is* pressure information. If you can imagine it: think of the lower "dots" as tall (high pressure) skinny (slow) patches of air and the upper "dots" as short (low pressure) fat (fast) patches of air. The density is actually constant across the animation.214.4.238.180 (talk) 19:47, 24 January 2011 (UTC)

Is viscosity necessary for lift?

I am new to this, so apologies if I step on any toes.

Under "Flowfield formation" it is suggested that viscosity is responsible for the establishment of the Kutta condition, and is thus a root cause of lift. I don't believe that viscosity is necessary, or even sufficient in itself, to cause lift. Any thoughts?

Zapletal 117.120.18.132 (talk) 02:17, 4 February 2011 (UTC)

Hello, welcome. This talk page is for discussing improvements to the article (see the top of this page). For questions about the subject of an article, Wikipedia has a reference desk.

Regarding your question: laboratory experiments show, that when the flow around an airfoil is suddenly started from rest, the initial flow field looks like the potential flow field without rotation. So the flow separates from the upper side of the airfoil and there is no lift. Only the build up of the boundary layers by viscosity forces the separation of the flow at the sharp trailing edge (around which the flow cannot stay attached). This leads to the use of the Kutta condition as a very handy approximation to calculate lift with potential flow models.

As far as I remember, there also have been experiments at very low temperatures with superfluids – with zero viscosity – leading to the disappearance of the lift force (and the applicability of the Kutta condition). -- Crowsnest (talk) 09:58, 4 February 2011 (UTC)

Extended content

Thanks for prompt reply. I am trying to improve the article by pointing out an unjustified assertion, namely that “viscosity generates the starting vortex, which then enforces the Kutta condition, and is therefore the root cause of lift’. I’m sorry, but I believe that this is false! I first studied aerofoil lift theory 30+ years ago. Back then I saw an old black and white movie (possibly made by Prandtl in early 1900s) showing the development of the starting vortex during initial aerofoil movement. This may be the "laboratory experiments" you refer to. But since then I have never heard or seen any adequate justification, either theoretical or empirical, for the assertion that “viscosity is the cause of lift”. I would be interested to hear more about the superfluid experiments, but these would only be negative results, and would not affirm the “viscosity causes lift” claim. Contrary to what is often said these days, there are many real everyday flows around aerofoils at +ve AoAs that look very similar to the classic mathematical potential flows with zero circulation/lift. Interestingly, the more that viscosity dominates the flow, the less likely is lift. That old movie showed some of these flows, and I can give you other simple examples that you can see for yourselves, if you want. However, I understand that I am not to present “original research” (or opinions?) here, so I won’t go into details just now of what I consider to be the origin of lift. So for now, could somebody please provide their best justification/evidence for the claim that “viscosity causes lift”? I am confident I can refute all arguments supporting viscosity, but first I have to hear them. Again, I am trying to improve the article. Zapletal—Preceding unsigned comment added by 117.120.18.132 (talk) 01:41, 7 February 2011 I agree with Crowsnest. Your original question was about whether viscosity is necessary for lift. Your latest request is about whether viscosity causes lift. Empirically, lift on an airfoil is reliant on numerous things - relative speed, density of the fluid, angle of attack on the airfoil. I think we should proceed on the assumption that the topic, and the question, is simply about whether viscosity is necessary for lift to be generated. Dolphin (t) 02:20, 7 February 2011 (UTC) The lifting characteristics of airfoil sections, and also three-dimensional airfoils, vary with Reynolds number. The Reynolds number can be changed by altering the size of the airfoil, the relative speed of airfoil and air, the density of the air, or the dynamic viscosity of the air. Hence it is possible to alter the lifting characteristics of an airfoil section or a three-dimensional airfoil by altering the viscosity but holding all the other parameters constant. This suggests to me that dynamic viscosity is significant in the generation of lift. In the Kutta-Joukowski model of lift-generation, the strength of the lift is partly determined by the strength of the bound vortex. A vortex is generated by viscous forces. In the absence of viscosity there could be no viscous forces and so no vortices. Without a bound vortex there could be no lift. Dolphin (t) 02:34, 7 February 2011 (UTC) I am happy to proceed on the basis of "Is viscosity necessary for lift?" Dolphin's second to last paragraph above suggests "dynamic viscosity is significant in the generation of lift". I agree viscosity has an influence, but it is generally deleterious to lift. Many empirical tests of aerofoils show generally worse lifting performance at lower Reynolds number (higher viscous/inertial forces), and better at higher Re (lower vis/init). This is over "typical" ranges of Re for real wings in air. At extremely low Re (Stokes flow) lift is almost impossible to generate. Roughly speaking, viscosity and lift is a bit like friction in an engine - more viscosity/friction just makes things worse. Dolphin's last paragraph states "a vortex is generated by viscous forces". Sorry, but here that is absolutely wrong! The starting and bound (and trailing) vortices of lifting flows are accurately modeled in theory (ie. in close agreement with empirical measurements) by "free" vortices. Free vortices have V~1/R and are by definition irrotational, which means they have zero vorticity. Free vortices are a concept intended to describe "freely" swirling flows that are uninfluenced by viscosity (ie. they are not "forced" vortices). If viscosity really does generate the lifting vortices, then all the flow surrounding the aerofoil, to a long(?) way out, would be full of "vorticity", which is the "footprints" of viscosity. That is, all fluid particles would be seen to be individually rotating. The vortices would not look like the "free" ones seen in real lifting flows. Zapletal—Preceding unsigned comment added by 117.120.18.132 (talk) 01:23, 8 February 2011 (UTC) An elementary math model of the bound and trailing vortices is the free vortex with zero vorticity everywhere except at the singular point at the center. A much more realistic and accurate model of these vortices is the Rankine vortex in which there is a full complement of vorticity distributed across the core of the vortex where the flow is rotational.Dolphin (t) 02:14, 8 February 2011 (UTC) In a free vortex the velocities increase towards the centre, so across the centre the velocity gradient is extremely high (large velocities in opposite directions). In real viscous fluids this high velocity gradient generates a "forced" vortex at the core of the otherwise irrotational free vortex. In effect the core is a boundary layer of the oppositely directed flows around it. The simplified model of a "Rankine" vortex has "solid body rotation (V~R)" at the core, and irrotational motion (V~1/R) outside the core (a more realistic velocity distribution has a smooth curve blending the two regions). The key point is that in a Rankine vortex the outer irrotational flow is the "cause", and the inner forced vortex is the "effect". If you see it the other way around, then by what mechanism does the inner core get the outer flow moving, without imparting vorticity to it? Action at a distance? Note that a typical aeroplane wing in flight has a boundary layer thickness at the TE measured in centimetres, this being the extent of its "viscous influence". Yet in the first few chord lengths of the wing's movement it has created two large swirls of air (starting and bound free vortices) that extend outward many chord lengths. How can these swirls be the result of viscosity? Zapletal—Preceding unsigned comment added by 117.120.18.132 (talk) 00:46, 9 February 2011 (UTC) Zapletal, I see you are still not managing to sign your edits with the necessary four tildes. Please see my message to you HERE. You have written that the forced vortex at the core is effect, and the irrotational flow outside the core is the cause. Do you have some objective reasoning for saying that or is it your personal view? Helmholtz's third theorem states: In the absence of rotational external forces, a fluid that is initially irrotational remains irrotational. The airstream approaches an airfoil in irrotational flow, but as it departs from the airfoil there are regions of rotational flow, such as in the cores of the trailing vortices and in the starting vortex. Clearly, the flow has not remained irrotational, so according to Helmholtz's third theorem some rotational external force has been at work on the air. My view is that the rotational external forces on the air are the viscous forces constantly at work in the boundary layer surrounding the airfoil. The viscous forces impart a spin to the boundary layer, initially causing a starting vortex, and then causing a vortex sheet in the wake behind the airfoil. This vortex sheet quickly consolidates into trailing vortices. Dolphin (t) 02:05, 9 February 2011 (UTC) Dolphin, apologies for bad manners. Four tildes, indenting(??)... working on it... Only have one hour per day (max) for this, so corrections welcome. Free vortices are irrotational "in compliance with" Helmholtz's Third. That is, if an inviscid fluid (which by definition is too "slippery" to have a torque applied to it) is to move in a generally circular ("swirling") motion, then it must do so irrotationally. A truly inviscid flow around an aerofoil has lots of irrotational swirling (as per Helmholtz's 3rd), but it also has regions of infinite velocity gradient, namely next to the aerofoil's surface, and at the centres of any shed free vortices. With viscous fluids these high velocity gradients, together with the fluid's viscous resistance to shearing, generate shear stresses that are the "rotational external forces" that generate "vorticity" (what you called "rotational flow" above) in the fluid. Going backwards in the cause-effect chain, the initial cause of the vorticity that appears in the fluid is a force acting on the aerofoil pushing it forwards. So vorticity in the BL and wake is caused by the force that overcomes "skin friction and profile drag". Vorticity in the cores of the trailing vortices is due to the force that overcomes "induced" or "vortex drag". And, interestingly, the vorticity that results from the decay of the starting vortex is due to the force that overcomes an increased drag on the aerofoil during its initial acceleration. I'll cover this last one under a different heading, where it is more relevant (when I get time). The force required to overcome skin friction drag is easy to understand, being simply the sum of the viscous shear forces at the surface. The vortex drags, however, are initially felt as pressure forces normal to the surface of the aerofoil. These pressures have gradients out into the fluid, and these gradients accelerate the fluid particles in various directions (more accurately "change their momenta") causing large scale swirling motions. If the fluid is viscous, then wherever there are high velocity gradients in these swirls (such as at the cores of free vortices) the viscosity converts the momentum and kinetic energy of the large scale and orderly swirl into, ultimately, increased heat energy in the fluid (ie. faster but more randomly moving molecules). This also applies in turbulent BLs and wakes. Entropy increases, etc... Zapletal 117.120.18.132 (talk) 00:29, 10 February 2011 (UTC) Zapletal, your use of the four tildes after your latest edit went well! With a little practice you will find it is very easy. Keep it up. In physics, there are many phenomena in which two or more parameters can change simultaneously. (eg Bernoulli's principle explains simultaneous changes in fluid pressure and speed.) I sometimes see evidence of people debating which of these parameters is the cause, and which is the effect. (eg some people believe pressure changes as the result of a change in speed, and others believe speed changes as the result of a change in pressure.) These people are incorrect on this point. The changes in these parameters are caused by other forces – they are not cause and effect for each other. You have commented on the cause-effect chain of vorticity in the core and the irrotational flow in the free vortex outside the core. You also wrote "The key point is that in a Rankine vortex the outer irrotational flow is the "cause", and the inner forced vortex is the "effect"." Both these flow patterns are caused by the shape of the airfoil and the relative motion of the fluid. They are not cause and effect. Your thesis is that viscosity is not necessary for the generation of lift. If your thesis is dependent on the idea that vorticity in the core of a Rankine vortex and the irrotational flow outside the core have a cause and effect relationship then I think your thesis is unsound. Everything on Wikipedia must be based on information already existing in reliable published sources. These sources must be identified using References and In-line citations. I seriously doubt you will find a reliable published source to support the idea that the above two flows have a cause and effect relationship. Your time would be better spent researching what reliable published sources actually say on this subject rather than pursuing your personal beliefs if those personal beliefs are not supported by reliable published sources. You will find some useful background information on this subject at No original research.Dolphin (t) 01:51, 10 February 2011 (UTC) Being able to distingish cause and effect is of utmost importance to understanding. Imagine someone 150 years ago watching the horse-and-carts go by. After some time they come to the conclusion that “a horse can only move forward if it is caused to do so by a cart behind it, pushing it along.” If they then suggest that they can build a faster and more efficient cart by simply removing the horse, well, they are plain wrong. Now consider a streamlined body moving through viscous but stationary fluid. Viscous shear drags some fluid along with the body, generating a moving boundary layer (BL) and wake. The “effect” is the moving fluid, and its “cause” (some way back in the chain) is a force causing the body to move forward. To suggest, say, that the moving body is the effect, and the BL the cause, is plain wrong. My main reason for attribution of cause and effect in a Rankine vortex is Occam’s Razor. As explained above, I reckon the outer free vortex is generated by pressure forces caused by the moving aerofoil, and the inner forced vortex grows in the manner of a BL. No new assumptions, theories, etc., are required to support this view. But if it is suggested that the inner vortex (my BL) “causes” the outer fluid to move in a V~1/R manner, then some new “fast action at a long distance” mechanism must also be postulated in addition to the usual viscous diffusion behaviour seen in BLs, which is slower and shorter in range. Occam favours the simpler explanation. Statements such as “trailing vortices induce [cause] a downwash at the wing”, seen in many textbooks, are just sloppy language used while borrowing EM theory’s Biot-Savart equation. These are compounded by an education system obsessed with teaching only the “steady-state”, with negligible mention of cause and effect. Ultimately, the misinformation has spread like an urban myth, with few people bothering to check the details. Here are some references. Most of my fluid/aerodynamics books claim, as in this article, that “viscosity is necessary for lift”. However, one of them, “Aerodynamics for Engineering Students”, Houghton and Brock, 2nd Ed. 1970, has the following footnote at p96 (possibly added at the suggestion of a reviewer?)- “There is no fully convincing explanation for the production of the starting vortex... [the usual viscosity explanations are described as “incomplete”]... It may be, however, that local flow acceleration is equally important and that this is sufficiently high to account for the failure of the flow to follow round the sharp trailing edge, without invoking viscosity. As a corollary, it is possible that circulation and aerodynamic lift can be generated in an inviscid fluid. This opens up interesting possibilities for further research...”. Another book you might look at is “Introduction to Aerodynamics”, Craig, 2002. This may be self-published (??) so don’t know how Wiki rates its “reliability”. This book is mostly in accordance with my views, although some of the figures could be improved (I’m a hard marker). Finally, I have recently been talking to some of the teachers at the local university. Z “Many very poor explanations of aero lift on the web... mostly "flat earth" theory...”. Drs “Well, those are for the great unwashed...” (not their exact words, but that gist). Z “Most explanations claim viscosity is necessary for lift, the rest don’t say either way.”. Drs “No, no, inviscid flows will generate lift... great unwashed...”. Z “But most of the explanations have university level maths...”. Drs “Err, well, the people who really need to know it, they do understand it”. Z "I'm not so sure...". Anyway, I’m here trying to establish the level of understanding of this subject. I hope to convince the local university to present a better explanation, possibly via their website (unfortunately they seem reluctant to spend their resources on such trivial matters - they have their pet projects). This, of course, might lead to improvements to this article. So, does anyone have some more arguments supporting the assertion that “viscosity is necessary for lift”. Zapletal117.120.18.132 (talk) 23:50, 10 February 2011 (UTC) Sorry, but as pointed out by Dolphin51, this is an encyclopaedia striving for a neutral point of view, with no room for original research. For material to be included, reliable secondary sources are required, which have substantial support in the scientific community. So e.g. explanations on a university website will not suffice. Wikipedia works on the basis of consensus. For this well-established subject of lift, probably only highly-cited scientific papers – and books by renowned experts in the field – will suffice to warrant inclusion of such a theory in the article (besides the common explanations on the origin lift). If you want to pursue inclusion of your ideas here about the inviscid origins of lift on airfoils, you have to provide evidence that there is support for them in (a substantial sub-group of) the scientific community. -- Crowsnest (talk) 01:56, 11 February 2011 (UTC) @Zapletal: You are going well with use of four tildes to sign off. It isn't appropriate to add four tildes to something you wrote a day or two ago because that leaves an incorrect date and time. Only apply the four tildes today to what you wrote today. Your statement about the horse and cart is clearly correct, but as an analogy directed at fluid dynamics it is unconvincing. If you still believe it, try answering this question. In Bernoulli's principle, does a change in pressure cause the change in fluid speed, or does the change in speed cause the change in pressure? If there is a clear majority of reliable published sources that support the notion that lift cannot be generated in an inviscid fluid, then this is what Wikipedia must publish. There may be a minority of sources that don't support that notion, that minority may be very attractive, and it may be your preferred position, but Wikipedia doesn't arbitrate where there are differences of opinion. Wikipedia must publish the majority position. If the contrary position is supported by some reliable published sources then Wikipedia might mention that there is an alternative position on the matter, and state what that alternative position is, but Wikipedia certainly doesn't arbitrate on the two different views and publish only the winner. There are other documents, including on-line documents, that are entirely suitable for experts such as yourself to argue the case for and against the different theories and hypotheses that abound in the field of fluid dynamics. I know it sounds boring, but Wikipedia aims to state what has already been published as the majority view in reliable sources. Sometimes we editors find ourselves writing, and looking for citable sources for, things that we don't entirely believe in. That is the situation in which we place ourselves when we volunteer to contribute to Wikipedia. Dolphin (t) 05:03, 11 February 2011 (UTC) As stated before, I am hoping to improve general understanding of this subject, and possibly also this article. Really! I thank you both (Dolphin & Crowsnest) for your input (in this section - more to come!), and for your time and efforts here at Wiki. Hopefully I can use your above comments to convince the local university to provide the resources that may lead to a "better explanation". This will take time. Meanwhile, some changes are in the pipeline with NASA's GRC FoilSim program, and Marco Colombini's very colourful aerofoil graphics program at the University of Genoa site, although these will take time too... Zapletal117.120.18.132 (talk) 23:57, 13 February 2011 (UTC) I have just noticed the reference to the superfluid experiments at last entry "Further Reading". Typically I am expected to pay for this paper, even though it was written in 1957. (Used to be scientific papers were freely available in libraries for anyone to read. I am in a library now, but such is progress...) Anyway, the key phrase in the abstract is "[lift vanishes] at sufficiently low velocity". Presumably, above this velocity the aerofoil generates lift. If (??) the flow is still inviscid at the higher velocity, then the "viscosity is necessary for lift" claim is refuted. All sorts of weird things happen at the quantum level, so more details would be good. Has anyone read this paper? Is it worth the effort/money to buy? The "sufficiently low velocity" seems to support my (non-viscous) ideas for the origin of lift... Zapletal—Preceding unsigned comment added by 117.120.18.132 (talk) 01:51, 8 February 2011

Correction Required in "Flowfield Formation".

About half way down this section is the statement "At the instant when the flow is 'turned on'...". This is followed by a description of a flow resembling potential-flow-with-no-circulation, and Bernoulli is then applied to determine pressures based on the velocities. (I note that potential flow is a purely kinematic model. It takes no account of the flow's mass, or any forces acting on it. It considers only "continuity" or "conservation of volume". So it can give velocities, but forces such as "pressure" must be derived separately, such as by tacking Bernoulli on to the end of PF.) The paragraph concludes with "The net pressure difference between upper and lower surfaces is zero".

I assume that the phrase "when the flow is 'turned on'..." means the same as "when the aerofoil accelerates from rest, relative to the bulk stationary fluid", or "when the fluid starts to move, relative to a stationary aerofoil". In any of these cases the description is one of an accelerating flow and Bernoulli does not apply. Perhaps the easiest way to see this is in the moving-aerofoil-stationary-fluid reference frame. Here all the fluid has initially V=0, but later some parts of the fluid have acquired velocity, and hence also momentum and kinetic energy. Bernoulli only applies when no energy is put into, or taken out of, the flow.

During the aerofoil's (or even a streamlined non-lifting body's) initial acceleration it will have positive pressure in front, and negative pressure behind it, even in inviscid flow. The body experiences a drag force. This is a direct consequence of Newton's laws. For an aerofoil at +ve AoA the positive pressure is also under the aerofoil, and negative pressure above.

I hope this is clear? I await any comments before continuing (ie. Dolphin's earlier question regarding cause and effect in Bernoulli can be answered...). Zapletal117.120.18.132 (talk) 00:30, 14 February 2011 (UTC)

The flow is turned on in these experiments (near) instantaneously, i.e. the initial flow field described is hardly accelerating any more. It is this near-stationary situation which is treated in this section. The flow field resembles the potential flow field, and only thereafter the starting vortex is formed.

Furthermore, notice that Bernoulli's principle is also valid in accelerating potential flows (as at the very rapid "turning on" of the flow, not covered in the flow field description of the article), see here. -- Crowsnest (talk) 01:21, 14 February 2011 (UTC)

@Zapletal: I don't agree that Bernoulli's principle does not apply to an accelerating flow. Whenever fluid flows into a nozzle its speed increases, and into a diffuser its speed decreases, and Bernoulli applies to both these situations. Similarly, where streamlines are curved Bernoulli applies even though the fluid is experiencing a centripetal acceleration. Dolphin (t) 06:23, 14 February 2011 (UTC)

If I interpreted Zapletal correctly, he/she refers to the applicability of Bernoulli's principle to a flow accelerating in time (dv/dt ≠ 0). That was the basis of my reaction. -- Crowsnest (talk) 09:05, 14 February 2011 (UTC)

That is a good point. I think your interpretation is correct. Dolphin (t) 10:18, 14 February 2011 (UTC)

Extended content

(Is this indenting ok?). Many issues to cover here... 1) Yes I am talking about non-steady flows (dV/dt non-zero), which are different to "pseudo-steady" flows such as Dolphin's nozzle/diffuser/curved flows where the "picture" is essentially the same at any time T. Note that stationary-aerofoil-moving-fluid (the usual picture) is pseudo-steady, while essentially the same physical situation of moving-aerofoil-stationary-fluid is non-steady, but easily modelled with a simple coordinate transformation (subtract Vo). 2) I'm not sure what Crowsnest is saying? A flow that is "turned on near instantaneously" will have extremely high accelerations, implying extremely high pressure in front and low pressure behind the body (see below). And then you say "near stationary situation"? If this is before the aerofoil or flow picks up significant speed (ie. while the aerofoil is hardly moving relative to the fluid) then we have "Stokes flow" (very small Re), which typically resembles PF-with-no-circulation (move a spoon through honey to see this). The Prandtl movie shows this no-circ. flow only during the first (very small) increments of aerofoil movement. 3) The reference to Bernoulli in Unsteady Potential Flows points out that the equations are "momentum...integrated", namely Newton integrated. Also they are "pressure + kinetic + gravity + f(t) = 0 (energy)". Or (omitting gravity) "pressure + kinetic = variable-energy". This is different to the more usual version of Bernoulli, such as that in the article where zero velocity at front and rear stagnation points implies the same pressure, because of constant-energy. 4) Another way to look at it. Take any body, say a sphere, and accelerate it from rest relative to the surrounding stationary fluid. Clearly it must push the fluid particles in front of it out of the way. According to Newton's First (Law of Inertia) these particles would rather "remain at rest" unless "compelled" to do otherwise by forces "impressed" upon them (words from Motte's translation of Principia). So a positive pressure gradient must develop in front of the body to get the particles moving. Similarly the particles behind the body wish to remain at rest, so a negative pressure gradient develops there, while the pressure of the particles further back gets things moving. Considering an aerofoil at typical AoA, it must push the particles in front of and under it away, and it creates a "vacant space" (ie. lower pressure) above and behind it as it accelerates forward. The next step (particles moving from high to low pressure zones) explains the development of the starting and bound vortices... Zapletal117.120.18.132 (talk) 01:37, 15 February 2011 (UTC) Sorry, but again, as stated at the top of this talk page, this page is for discussing improvements to the article. This is not a forum to request explanations of the physics involved (see for that e.g. Batchelor, Landau & Lifschitz, etc.; references in the article). What is expected here are argumented discussions (including reliable sources) on which material is missing, or needs improvement, in the article. See also here Also this discussion is getting more-and-more off-topic. Ad your point #2: there are several phases 1st the fluid is at rest, then 2nd there is a rapid acceleration of the flow towards a new situation with constant mean flow (during this acceleration phase the flow is well described as an unsteady potential flow), 3rd there is a new situation with constant mean flow; initially the flow strongly resembles potential flow, 4th the starting vortex is formed, separates and there is circulation and lift. Only step 3 & 4 are relevant here, and discussed in the article. -- Crowsnest (talk) 09:27, 15 February 2011 (UTC) And again, I have been trying to improve this article by pointing out what I see as obvious "falsehoods". I don't want to waste any more of your time (as I said, I appreciate your voluntary efforts here), other than to make this last request. Could you please give me a list of the references at the bottom of the article that you consider to be the most "reliable/trusted" sources that support the issues discussed above (ie. "viscosity is necessary for lift", and "no extra drag force at start-up"). I guess the first two I will check are Batchelor, and Landau & Lifschitz, but are there any others that you strongly recommend? Zapletal117.120.18.132 (talk) 02:22, 16 February 2011 (UTC) Copied from User talk:117.120.18.132: Hi Zapletal, from flying through turbulence you know what a rough ride unsteady flow can give. Besides the variations in the instantaneous drag and lift by the flow variations, there are also inertia forces (see added mass) due to the temporal accelerations of the flow (as seen from the frame of reference moving with the plane). Even according to potential flow theory (without circulation), there are forces in unsteady flow (d'Alembert's paradox is for steady flow). What to call drag & lift in the unsteady case is up to convention, but for an asymmetric airfoil both force components -- in the direction of the flow acceleration vector and perpendicular -- are present. Unsteady flow and resulting forces, including added-mass effects, is treated in e.g. Joseph Katz & Allen Plotkin (2001) Low-speed aerodynamics, ISBN 0521665523. See also Küssner effect and the references there. I think Wikipedia may be expanded and improved with respect to these unsteady effects on aerodynamic forces. The flow field formation in a starting flow around an airfoil is described in e.g. Batchelor (1967) An introduction to fluid dynamics ISBN 0521663962. -- Crowsnest (talk) 07:59, 16 February 2011 (UTC) Crowsnest, I was going to stop annoying you people for a while (until I checked Batchelor, et al, which will take some time) but your above reference to "added mass" (presumably a reliable source, being in Wiki) seems to confirm my comments here. I note "added/virtual mass" is a metaphor for "greater fluid pressure in front, than behind". It applies to symmetric bodies where the extra pressure force is in the opposite direction to acceleration (so F=(Mbody+Mextra)A), such as the air bubbles mentioned in the Wiki article. Anyway, for an aerofoil at typical AoA the extra pressure force is roughly perpendicular to the chord line, so this Lift article could be corrected to reflect this. Also, I've been looking more closely at the Prandtl pictures (can't find the original movie anywhere on web) and, 1) the development of the starting vortex and change from non-circ. to circulatory flow takes place within less than a chord length of movement, 2) the velocity seems to be increasing throughout this process (ie. it is accelerating or unsteady flow). I'll leave you with the following suggestion. Fill the kitchen sink with water, place a large spoon in it at a typical aerofoil-like AoA, and move the spoon quickly sideways. Water level (and hence pressure) rises in front/under the spoon, drops behind/above spoon, and a very obvious starting vortex develops (put drops of food colouring in water to see better). The vortex is of the "free" type, with "V" shaped bottom - steeper slope closer to middle. Next take a large tub of honey, a small spoon, and repeat above but more slowly (to keep Re down). The flow closely resembles potential flow with no circulation, the type of flow often described as "physically impossible" in textbooks, although I see it every morning (sweet tooth)! You can draw your own conclusions from the above experiments, but I see 1) viscosity is not good for lift, and 2) pressure and inertial forces are great at generating free vortices. I know it's not a "published" source, but I reckon you can't beat nature as a "reliable" source! —Preceding unsigned comment added by 117.120.18.132 (talk) 01:06, 18 February 2011 (UTC) Honey is very viscous, so the flow you describe is Stokes flow, not inviscid potential flow. Discussion closed, as far as I am concerned. -- Crowsnest (talk) 08:06, 28 February 2011 (UTC)

Force vectors and the definition of airfoil lift

This article describes airfoil lift which is a vector in a direction perpendicular to the chord of the airfoil. That vector separates into propulsion (forward force counteracting drag) and lift (upward force counteracting weight.) The article does mention that down force is the opposite of lift which is the case if the vector only separates into drag and a downward vector adding to the weight. In the case of a propeller, it produces propulsion since the airfoil lift force is "forward" and the drag force resists the rotation of the propeller.

76.105.60.21 (talk) 21:59, 27 February 2011 (UTC)

The article does not say lift is a vector perpendicular to the chord of the airfoil. The article correctly states that lift is the component of the surface force that is perpendicular to the direction of the oncoming flow. There is a significant angle between the chord and the oncoming flow - it is called the angle of attack. Dolphin (t) 11:36, 28 February 2011 (UTC)

Bernoulli's and Newton's premise of motion

I removed:
"The academically accepted theories include Bernoulli's and Newton's premise of motion."
Because I cannot grasp it. Can somebody explain what is meant? -- Crowsnest (talk) 15:45, 9 March 2011 (UTC)

This sentence was added by an IP address - see diff. Perhaps the intention was to highlight that both Bernoulli's principle and Newton's laws of motion provide explanations of the phenomenon of lift that are accepted in academic circles; implying that some alternative theories such as the equal-transit time model are not accepted in academic circles. In the absence of an in-line citation or a clear understanding of what is intended it is reasonable that the sentence should be removed. Dolphin (t) 21:28, 9 March 2011 (UTC)

I think the contributor was trying to say that both Bernoulli's Law and Newton's Laws can be used to explain lift, but the phrase premise of motion has no meaning that I can discern, and the statement is out of place in the overview which explains what lift is rather than how it comes about. I'd say just delete it (which you have already) and move on. Mr swordfish (talk) 21:40, 9 March 2011 (UTC)

Hoffman and Johnson

Take it up with Hoffman and Johnson who have published their work in a peer reviewed journal: J. Hoffman and C. Johnson, Resolution of d´Alembert's paradox, Online First, Dec 10, 2008, Journal of Mathematical Fluid Mechanics. Journal of Mathematical Fluid Mechanics Sept 12 2008. http://knol.google.com/k/why-it-is-possible-to-fly#New_Mathematical_Theory_of_Lift Bcebul (talk) 05:32, 31 March 2011 (UTC)

That paper is about d'Alembert's paradox, not about lift. This paper is referenced sideways once in another paper, not from the group of Hoffman and Johnson. Their lift theories have not been published in renowned peer-reviewed scientific journals. So I reverted the addition of this theory, not being notable at the moment. -- Crowsnest (talk) 07:15, 31 March 2011 (UTC)

Ok. How would you like to introduce their critique of the current controversial, ill-posed K-Z model? see also http://www.nada.kth.se/~cgjoh/flightnormat-5.pdf Bcebul (talk) 09:14, 31 March 2011 (UTC)

As long as there is no significant discussion (or use) of their approach in the scientific literature, the need to include a discussion of their material in Wikipedia is lacking. See also the past discussions on the inclusion of their views: Talk:D'Alembert's paradox/Archive 1 and Talk:Navier–Stokes existence and smoothness/Archive 1. -- Crowsnest (talk) 12:44, 1 April 2011 (UTC)

Stick and Rudder

Langewiesche's Stick and Rudder is referenced by the article. Yesterday, someone marked it as an "unreliable source". I undid that categorization, pending resolution or consensus here. So, does Stick and Rudder meet the requirements of a reliable source?

My take is that it does. It's a standard textbook that's been around for over 60 years and is still in print. Interested to hear other opinions.... Mr swordfish (talk) 14:16, 24 May 2011 (UTC)

I acknowledge that Langewiesche's book is still available. It is an introductory book for student pilots and newcomers to aviation. Introductory books on a subject are often not suitable as sources of information for an encyclopedia. For example, Langewiesche says That’s what keeps an airplane up. Newton’s Law says that if the wing pushes the air down, the air must push the wing up. This explanation will be satisfactory to many student pilots and newcomers to aviation and science, but most people with a sound understanding of science will find it to be entirely unsatisfactory, for the following reason.

Newton's third law of motion can be considered to say that whenever an object A exerts a force F on another object B, then object B simultaneously exerts the same force F on object A but it is in the opposite direction. Every force that has ever existed has been accompanied by an equal and opposite force. In considering the lift force exerted on an airfoil by the passing air, Langewiesche seeks to explain it by observing that the airfoil exerts a force on the passing air and therefore, by Newton's third law, the air must be exerting an equal but opposite force on the airfoil. This is not an explanation of either of these forces; it is just dumbing-down the description to the point where the majority of student pilots and newcomers will be willing to accept it, so they are then able to move on to the next topic.

Science students study the effect of an electric current in deflecting the needle of a magnetic compass. They learn about the subject by trying to explain why the presence of an electric current can exert a force on the needle of the compass. Very few science students, and none of their teachers, would be impressed by an explanation that says The compass needle exerts a force on the electric conductor and therefore, by Newton's third law of motion, the electric conductor must exert a force on the compass needle. That is not an explanation of either of these electro-magnetic forces, but it is what Langewiesche is doing in Stick and Rudder. See Talk:Bernoulli's principle/Archive 2#Stick and Rudder.

Langewiesche's book is still in print but it is merely an introductory book for newcomers. It is not an authoritative source of information suitable for an encyclopedia, and it is certainly not an authoritative source for an encyclopedia article about the lift on an airfoil. Dolphin (t) 23:20, 24 May 2011 (UTC)

> This explanation will be satisfactory to many student pilots and newcomers to aviation and science, but most people with a sound understanding of science will find it to be entirely unsatisfactory...

Who are we writing this article for? Newcomers, or people who already have a "sound understanding" ? My take is that it's mostly for lay people, or at least a layperson should be able to read the article and come away with something. If all they get out of it is "air goes down, plane goes up" at least that's something. If they wish to read on and find out why the air goes down, good for them.

But I'm getting off topic at hand. The question is whether Stick and Rudder is worth citing in the footnotes. My opinion is that it is, because 1) it's a respected introductory textbook on aviation and 2) it states the deflection theory in a way that's easily understood.

Instead of Langewiesche, we could cite Halliday & Resnick:

"...the effect of the wing is to give the air stream a downward velocity component. The reaction force of the deflected air mass must then act on the wing to give it an equal and opposite upward component."

Halliday and Resnick Fundamentals of Physics 3rd Edition John Wiley & Sons page 378

Although both are saying the same thing, I think Langewiesche states it more succinctly and simply. That said, it's in the footnotes, not the main body of the article, so I don't think it matters much. My take is that we should either leave the cite as-is, or remove it - if it's truly an "unreliable source" we should just remove it and replace it with the Halliday & Resnick quote. Mr swordfish (talk) 20:19, 26 May 2011 (UTC)

@Mr swordfish. I don't disagree with anything you have written. It is often attractive to contemplate that Wikipedia should be a simple, how-to guide for lay people. WP:NOTGUIDE explains that Wikipedia is not an instruction manual, guide book or textbook. (Conversely, WP:MTAA explains that technical articles should be accessible, so it isn't a simple matter of using encyclopedic language.)

My objection to Langewiesche is not that he advocates the momentum approach to explaining lift (or deflection model) but that he deprecates Bernoulli's principle. In Stick and Rudder he says Bernoulli's Theorem doesn’t help you the least bit in flying. While it is no doubt true, it usually merely serves to obscure to the pilot certain simpler, much more important, much more helpful facts. This indicates to me that his book is not one dedicated to a scientific explanation of the subject, but is a specialised book dedicated to student pilots. It makes non-scientific statements in order to provide a simple, narrow path to the end goal.

I have the greatest respect for the books by Halliday and Resnick. At Bernoulli's principle, Note No. 20 is a citation that I added from H&R's 1960 book Physics. Please go ahead and add your H&R reference in place of the one by Langewiesche.

Thanks for your assistance and co-operation on this point. Dolphin (t) 05:42, 27 May 2011 (UTC)

I'm not sure I follow your reasoning here. You say that "Wikipedia is not an instruction manual, guide book or textbook.", and then seem to argue that since Stick and Rudder is one of those things we cannot cite it. But Wikipedia is not lots of things - it's not a newspaper, a magazine, a scientific journal, or a place to publish original research. But that doesn't prohibit us from citing magazines, newspapers, scientific journals, etc. I fail to see how a prohibition on citing instruction manuals follows from the fact that Wikipedia isn't one itself. By that logic, Wikipedia could only cite other encyclopedias.

Regarding Langewiesche's "deprecation" of Bernoulli, I don't see that sentence saying anything depreciatory towards Bernoulli's law, merely that you don't need to understand it for the practical act of flying a plane. The analogy I'd make here is a textbook for one of those "physics without calculus" courses. Of course such a book has serious limitations, but there are many phenomenon that can be described without using calculus and if the book described some physical phenomenon succinctly and correctly (if not completely) I'd have no issue with citing it. Langewiesche does not claim Bernoulli's Law is wrong, any more than the authors of our physics-without-calculus textbook claim that calculus is wrong. So I don't understand your objection.

I'll add the H & R cite, and await some other opinions before deleting the Stick and Rudder cite. Mr swordfish (talk) 15:18, 31 May 2011 (UTC)

Thanks for adding the H & R citation. My comment that "Wikipedia is not an instruction manual etc" was aimed at your comment Who are we writing this article for? Newcomers, or people who already have a "sound understanding"? My take is that it's mostly for lay people ...

I agree that Langewiesche does not claim Bernoulli's Law is wrong. He stops short of that. He diminishes its significance by dismissing it as unhelpful to student pilots. He has written Forget Bernoulli's Theorem. Bernoulli’s Theorem doesn’t help you the least bit in flying. While it is no doubt true, it usually merely serves to obscure to the pilot certain simpler, much more important, much more helpful facts. Langewiesche had his reason for dismissing Bernoulli's Theorem. He was writing a book for student pilots, presenting the essential information in the briefest and most simple way, and he was successful considering his book is still readily available. The article Lift (force) in Wikipedia is not written for student pilots and it doesn't seek to present the essential information in the briefest and most simple way. Consequently we need to be careful before citing Langewiesche, just as we would be careful before citing the physics-without-calculus book if it says Forget calculus. Calculus doesn’t help you the least bit in physics. While calculus is no doubt true, it usually merely serves to obscure to the student certain simpler, much more important, much more helpful facts. Conversely, I would not have a problem with the physics-without-calculus book if it said Calculus is a powerful tool in physics, but it is still possible for people who haven't studied calculus to gain a broad understanding of physics. This book is written for all those people who are keen to study physics without the extra challenge of calculus.

Langewiesche spawned a genre of writing that repudiated Bernoulli's Theorem, and was much more aggressively dismissive of the relevance of Bernoulli to the phenomenon of fluid dynamic lift. For example, in Understanding Flight, Anderson and Eberhardt pay tribute to Langewiesche and say Bernoulli's equation has mistakenly become linked to the concept of flight. Dolphin (t) 23:02, 31 May 2011 (UTC)

What interactions cause the increase of energy of the air when producing lift (and drag)?

Using the air as a frame of reference, after a wing passes through a volume of air, the affected air ends up with an increase in energy, a downwards (lift related) and somewhat forwards (drag related) non-zero "exit" velocity where the affected flow's pressure returns to ambient.

Using a 1500 lb glider with 60:1 glide ratio at 60 mph as an example, (a Nimbus 4T with an 80 foot wingspan does this at 68 mph), in a steady 1 mph descent, the gravitational power applied to the glider is 1500lb x 1 mph = 4 horsepower. The glider in turn applies this power to the air, adding 2200 lb ft of energy to the air every second.

Most of the energy is in the form of an impluse wave that transmits the 1500 lb weight of the glider through the air and eventually onto the earth's surface.

Clearly Bernoulli is violated by this increase of energy of air (work is peformed). What are the interactions between a wing and air that are responsible for this increase in energy? In real world situation, such as assuming a finite wingspan of 100 feet or less, is there a lower limit to the minimal amount of energy that must be added to the air in order to produce lift for a 1500 lb glider at some range of speed, perhaps 35 mph to 70 mph?

Jeffareid (talk) 20:18, 23 February 2010 (UTC)

Viscosity? When anything moves through the air you get molecules hitting the surface and they inevitably bounce off faster, they then hit each other and similar things happen there. The net upshot is that the thermal energy of the air around the object increases. If you think about it, drag has to involve heating because energy is force times distance, in any second the object has moved a distance against the drag force, and that energy has to go somewhere.- Wolfkeeper 01:56, 24 February 2010 (UTC)

In this case of a glider, the temperature increase of the affected air is small, and the amount of change in velocity of the air in the direction of flight, related to drag, is also small. Most of the change in velocity (from zero relative to the air, to the "exit velocity" speed) is in the direction of lift, and most of the increase in total mechanical energy is related to the increase of "exit velocity" in the direction of lift. Since the wing is at an effective angle of attack, it could be possible to estimate work done in the direction of lift as the average force of the wing times the average component of distance the wing surfaces move perpendicular to the direction of flight, as a wing moves through a cross-sectional plane perpendicular to the direction of flight, but I don't know how to take the pressure differential effects into account. Jeffareid (talk) 03:31, 24 February 2010 (UTC)

There's a handy formula for relating fluid velocity to pressure - it's called Bernoulli's principle :-) If the temperature increase of the affected air is small, it will provide good answers. But, there's no work done in the directin of lift for straight and level flight. (No distance travelled in the dirction of the force = no work.) The propulsion system does the work in overcoming drag.69.1.23.134 (talk) 02:54, 14 March 2010 (UTC)

I agree that no work is done by the lift vector, but it isn't only in straight and level flight. Providing the frame of reference is the one in which the atmosphere has zero velocity, lift is the component of aerodynamic force that is perpendicular to the vector representing the velocity of the airfoil (or wing or aircraft) and so zero work is done, regardless of whether the aircraft is in straight flight or turning, level flight, climbing or descending. Thrust, drag and weight are all capable of doing work on the airfoil (or wing or aircraft) but lift does no work when using this particular reference frame. Dolphin51 (talk) 11:32, 14 March 2010 (UTC)

True for an airfoil, but not for a wing or aircraft. Turning, climbing and descending all involve work in the direction of lift on the wing/body resulting in energy storage/release by the wing/body. 214.4.238.180 (talk) 18:11, 12 May 2010 (UTC)

True for an airfoil, but not for a wing ...? What is a wing if it isn't an airfoil? Climbing and descending involve work, but it is work done by the weight and thrust. Turning flight doesn't involve work. If we choose the reference frame attached to the atmosphere, lift never does work because lift is defined to be the component of aerodynamic force that is perpendicular to the relative velocity, and any force that acts perpendicular to the displacement or velocity vector does zero work. See the definition of Work (physics). Dolphin (t) 21:56, 27 October 2010 (UTC)

For a two-dimensional airfoil or an "infinitely rigid" body there would be no work. But in the case of a real wing/aircraft the wing is displaced in the direction of lift (relative to the center of gravity) during maneuvers. This does real mechanical work on the wing. This is easily visible on the wings of large body aircraft during takeoffs and significant maneuvers. 69.1.23.134 (talk) 01:31, 14 November 2010 (UTC)

69.1.23.134 has written the wing is displaced in the direction of lift ... This is incorrect. Lift is defined to be the component of the aerodynamic force that is perpendicular to the direction of motion, therefore any displacement is perpendicular to the lift. Wherever force and displacement are perpendicular, no work is done - for example a centripetal force is perpendicular to the direction of motion and so centripetal force does no work and the body continues in circular motion with no change in kinetic energy. Dolphin (t) 07:04, 14 November 2010 (UTC)

You are correct that lift is the component of the aerodynamic force that is perpendicular to the direction of motion of the airflow. The work done by the force is the dot product of the force vector and the motion of the wing, so in the case of an airplane flying at a steady altitude, the motion of the wing is perpendicular to the force vector and thus the lift force does no work. By way of contrast, look at a sailboat - the motion of the sail is not perpendicular to the lift force and the lift force does do work (ie transfer energy to the foil). So, sometimes the lift force does work on the foil, sometimes it does not. In particular, when an airplane is climbing, the wings move in the direction of the lift force and the lift force does work on the wings. Mr swordfish (talk) 02:38, 15 November 2010 (UTC)

Much of what Mr swordfish has written is correct. However, his statement So, sometimes the lift force does work on the foil, sometimes it does not. In particular, when an airplane is climbing, the wings move in the direction of the lift force and the lift force does work on the wings is incorrect. Yes, work is the dot product of the force vector and the displacement vector; lift is defined to be the component of aerodynamic force that is perpendicular to the vector that represents the relative velocity of the airfoil and the freestream, therefore the dot product of these two vectors is always zero. Whether an airplane is climbing, descending or flying level, and we are using the reference frame attached to the atmosphere, the wing does not move in the direction of the lift force so the lift force does no work. (When an airplane is flying level or climbing, and its kinetic energy increases, it is the thrust that does work, not lift. When an airplane is descending and its kinetic energy increases it is the weight that does work.) If we use a reference frame other than the frame attached to the atmosphere, lift will usually not be perpendicular to the displacement so lift will do work. For example, when using the reference frame attached to the Earth's surface the airplane's velocity will be its ground speed and that will sometimes change due to work done by lift, such as when an airplane accelerates while turning downwind, or when it decelerates when turning upwind, and the horizontal component of lift during the turn does work on the airplane. Dolphin (t) 10:45, 15 November 2010 (UTC)

But if you define a single velocity vector (forward) and a single lift vector (up) for a non-rigid body, you will have wing displacements along the "up" axis resulting from the internal torque of the "real" lift at each wing. You cannot establish infinitely many inertial reference frames and discount each contribution to torque individually. If you have one frame for a flexible body, there is real work occuring in that frame. 214.4.238.180 (talk) 16:49, 18 November 2010 (UTC)

This discussion is about airfoils, wings and airplanes. They are being considered as rigid bodies. The comment above by 214.4.238.180 is unclear but in referring to a non-rigid body and a flexible body he appears to be describing the dynamics of wing bending. If so, it is off the current topic. 214.4.238.180 has written You cannot establish infinitely many inertial reference frames ... True, but nobody has done that. Dolphin (t) 22:08, 18 November 2010 (UTC)

Perhaps a rename to "Lift (Thin airfoil theory)" or "Lift (Two-dimensional flow)" is in order. The work performed by lift on any real air vehicle (or even any rotating rigid body) cannot be analyzed from a single 2-D inertial reference frame. 214.4.238.180 (talk) 17:58, 20 January 2011 (UTC)

So may I presume that Dolphin now agrees that wing bending in the case of a real wing/aircraft is a displacement in the direction of lift (relative to the center of gravity) which does real mechanical work on the wing? And that what 69.1.23.134 has written is actually correct? (albeit outside the two-dimensional flow model used in the article - I failed to grasp that the article covered only thin-airfoil theory) 69.1.23.134 (talk) 03:13, 22 March 2011 (UTC)

You asked what interactions cause the increase in energy involved in lift and drag. I don't know if this really answers your question, or just leads to another question, but here is a stab.

(Hope we are indirectly supporting the article here, since Wikipedia policy forbids discussions for other purposes.)

Each tiny patch of wing collides with air molecules. Each collision imparts a change in momentum of the particle (and the glider). For some of those patches, for example at the forward stagnation point, the collisions increase the magnitude of the momentum of the particles, increasing their kinetic energy (by one half the square of the increase, divided by the mass). For others, they decrease it, as on the top surface aft. A patch there is receding from the particle, which dulls the rebounds like a snowball hitting a moving car from behind. The power transfer from a given patch is the sum of those energy changes per unit time, which smooths out to a constant value if all else is steady. For some patches that power is positive, for others negative, But the sum of the powers from all the patches is positive. These are the interactions that cause the increase in energy, I think.

Mark.camp (talk) 16:05, 27 October 2010 (UTC)

@Mark.camp:A body experiences drag and leaves behind a wake where previously there was no movement. The kinetic energy in the wake is the result of work done by the drag force. Lift does no work because it is perpendicular to the direction of motion. Dolphin (t) 11:33, 22 March 2011 (UTC)

• there's no work done in the direction of lift for straight and level flight. (No distance travelled in the direction of the force = no work.) - For straight and level flight, the wing isn't moving in the direction of lift, but the air is being accelerated in the negative direction of lift, and as mentioned in the original post, the wing applies a downwards force on the air, and the surfaces of a wing are angled downwards, so there is some vertical component of the displacement of the surfaces of a wing as it passes through any cross sectional area of air. Rcgldr (talk) 15:44, 26 September 2011 (UTC)

@Rcgldr: Lift is defined to be the component of the aerodynamic force that is perpendicular to the vector representing the relative motion between an airfoil and the atmosphere. So lift is always perpendicular to the direction of motion, regardless of whether the airfoil is moving straight and level or performing aerobatic maneuvers. Consequently, when we are using a reference frame attached to the atmosphere, the lift on an airfoil is always perpendicular to the direction of motion and the lift does no work on the airfoil. This explains why lift can't be used to increase the kinetic energy of an airfoil - that requires either thrust or loss of gravitational potential energy. Dolphin (t) 22:36, 26 September 2011 (UTC)

lift can't be used to increase the kinetic energy of an airfoil - that requires either thrust or loss of gravitational potential energy.

Then how do you explain how a sail accelerates a sailboat? It's not gravity, and it's not some external thrust as from an engine. Mr. Swordfish (talk) 14:56, 27 September 2011 (UTC)

@Mr swordfish: An excellent question! I haven’t grappled with that one before so it has exercised my brain. The answer focuses on the choice of reference frame.

Lift is defined to be the component of the aerodynamic force that is perpendicular to the vector representing the relative motion between the atmosphere and the airfoil. If we choose an inertial reference frame attached to the atmosphere (or even one attached to the airfoil) lift is incapable of doing work. But if we choose any reference frame other than these two, lift is no longer perpendicular to the motion between atmosphere and airfoil, and it is capable of doing work and altering the kinetic energy of the airfoil.

When thinking of a sailboat I instinctively use a reference frame attached to the ocean. I see the wind (the atmosphere) coming from one direction with speed W, and the sailboat initially stationary and then accelerating in a different direction to the wind. However, in doing so I am not using an inertial reference frame attached to the atmosphere or the airfoil so I must expect that the lift on the sail will be responsible for increasing the kinetic energy of the sailboat.

If I use a reference frame attached to the atmosphere, the entire ocean is moving in a straight line with speed W. Similarly, the sailboat is initially moving at speed W. (All counter-intuitive!) The sailboat of mass M initially has kinetic energy half*M*W squared. The lift on the sail is perpendicular to the apparent direction of movement of the sail so it acts as a centripetal force and the sailboat appears to follow a circular path, still with speed W and kinetic energy half*M*W squared. Using this reference frame, the sailboat began with speed W and ended up with speed W, but in a different direction, so the lift on the airfoil has changed the direction of movement of the sailboat but hasn’t changed its kinetic energy.

I suspect some high-speed sailboats can actually move faster than the wind speed. That is a complicating consideration! I will have to think about it. Dolphin (t) 22:56, 27 September 2011 (UTC)

Ummmm... I don't know where you got that from, but the idea that lift can't do work is completely false. That's precisely how windmills work; the air flows over an aerofoil, this generates lift force perpendicular to the axis and it turns the rotor, doing work (force times distance). Also, when an aircraft lowers its nose, it loses altitude... and gains horizontal speed. That horizontal speed increase is entirely due to the lift vector being tilted forwards to the horizontal, literally pulling the aircraft forwards and gaining it kinetic energy.- Sheer Incompetence (talk) Now with added dubiosity! 22:59, 16 January 2012 (UTC)

Hi SI. Thanks for taking an interest in this subject. You are correct in saying the lift force on a windmill does work. (There are actually many examples where the lift force does work, but it all depends on choice of reference frame.)

In my edit dated 27 Sept (immediately above yours) I wrote:

If we choose an inertial reference frame attached to the atmosphere (or even one attached to the airfoil) lift is incapable of doing work.

Notice the emphasis on choice of reference frame! When contemplating a windmill it is challenging to use any reference frame other than one attached to the Earth. The frame attached to the Earth is clearly not one attached to the atmosphere (unless the atmosphere is stationary, in which case the windmill won't be turning.) As an exercise in basic aerodynamics, try drawing the blade of a windmill using the reference frame attached to the atmosphere. Be sure to include all the forces acting on the blade. You will see what force is responsible for the work on the mill, and you will probably be surprised. (Let me know if this one stumps you and I will be happy to explain.)

The purpose behind this topic of discussion is to reinforce the idea that lift is not a stand-alone force, as many people think. The stand-alone force is the aerodynamic force. Mankind chooses to resolve the aerodynamic force into two components - lift and drag. Mankind defines these two to be the components perpendicular to, and parallel to, the vector representing the relative velocity between the airfoil and the atmosphere.

When a pilot lowers the nose of his aircraft in order to accelerate, it is the weight of the aircraft that does work on the aircraft and causes its kinetic energy to increase; not the lift. Dolphin (t) 01:58, 17 January 2012 (UTC)

Um. Careful here. Although energy is to a degree frame dependent, changes in energy occur in all frames in a pretty similar way. The lift force does work in all inertial reference frames on the windmill blades. Lift force is not an ultimate source of energy of course, in the case of a windmill that's the sun, but that doesn't stop the lift force doing real work. In the case of the aircraft, the potential energy is the ultimate cause of the acceleration, but the lift force does do work by accelerating the aircraft when the lift force is tilted forwards. If you think about it gravity can't accelerate a vehicle forwards, since it acts always downwards, and $work={\vec {f}}orce\cdot {\vec {d}}istance$ giving always zero work done for any horizontal distance when the force is purely vertical, as with gravity, and yet the aircraft gains horizontal speed, work has been done. It's the lift force that actually does the work to accelerate the vehicle laterally.- Sheer Incompetence (talk) Now with added dubiosity! 02:31, 17 January 2012 (UTC)

What I'm saying is, both the aerodynamic force as well as the lift force can and do do work.- Sheer Incompetence (talk) Now with added dubiosity! 02:31, 17 January 2012 (UTC)

You have written If you think about it gravity can't accelerate a vehicle forwards, since it acts always downwards. If a person is riding a bicycle down a hill, he or she can take their feet off the pedals and accelerate forwards, at least until arriving at the bottom of the hill. How can that happen? Dolphin (t) 04:52, 17 January 2012 (UTC)

Gravity doesn't do that either, and for the same reason. It's the reaction to the gravity that accelerates the bicycle down the hill. The gravity pushes the bicycle downwards, but the hill pushes back, perpendicular to the surface, which is at an angle. The component of that reaction force that faces forwards down the hill accelerates the bicycle laterally. Gravity itself CANNOT accelerate a bicycle sideways, it points 90 degrees- it's in the wrong direction.- Sheer Incompetence (talk) Now with added dubiosity! 05:55, 17 January 2012 (UTC)

Let me see if I have got it right. The reaction from the hill pushes upwards, perpendicular to the surface. The component that faces down the hill accelerates the bicycle. So the force that is perpendicular to the hill has a component that faces down the hill. Have I got it right? Dolphin (t) 07:11, 17 January 2012 (UTC)

Yes, that's pretty much right, the reaction isn't entirely upwards, it's perpendicular to the surface, which is tilted and that does the work.- Sheer Incompetence (talk) Now with added dubiosity! 17:05, 17 January 2012 (UTC)

I agree that the reaction isn't upwards. It is perpendicular to the surface, and the surface is the side of the hill, which is sloping. You wrote:

$work={\vec {f}}orce\cdot {\vec {d}}istance$

The reaction is perpendicular to the side of the hill, and perpendicular means ninety degrees. The dot product of two vectors at ninety degrees is zero, suggesting the work done by the reaction force would be zero. Where are we going wrong? Dolphin (t) 21:49, 17 January 2012 (UTC)

You can pick the distance vector in any direction you want. A horizontal vector is parallel to the horizon. The lift force in the case of the aircraft and for the reaction to the weight of the bicycle both do work in that direction. In a sailing boat, the lift vector also does work horizontally and in a windmill it does work as a torque. For a kite, the lift can be have a vertical component and if the kite rises then it has done work vertically (note that for a kite the drag vector cannot do work if the wind is lateral). The idea that the lift vector cannot do work is false. Of course if you deliberately pick a vector that is at 90 degrees to the lift vector, then it can do no work, but otherwise it will do work.GliderMaven (talk) 01:16, 18 January 2012 (UTC)

GliderMaven is absolutely correct when he writes Of course if you deliberately pick a vector that is at 90 degrees to the lift vector, then it can do no work

Lift is just a component of aerodynamic force. It is defined to be the component of the aerodynamic force that is at 90 degrees to the vector representing the relative motion between the airfoil and the atmosphere. That is the same as saying lift is deliberately picked to be at 90 degrees to the motion.

That is why it is true to say that in the reference frame attached to the atmosphere, lift is incapable of doing work, but in almost every other reference frame lift is capable of doing work. Dolphin (t) 02:05, 18 January 2012 (UTC)

"in the reference frame attached to the atmosphere, lift is incapable of doing work" NO, that is WRONG, because the reference frame attached to the atmosphere is NOT the relative wind reference frame, the lift vector is perpendicular to the relative wind NOT the reference frame attached to the atmosphere!!!!- Sheer Incompetence (talk) Now with added dubiosity! 07:40, 19 January 2012 (UTC)

The relative wind is a vector, not a reference frame. You are writing about the relative wind reference frame.

You have also written the lift vector is perpendicular to the relative wind ... Yes!, we both agree on that point

You have also written the lift vector is ... ... NOT the reference frame attached to the atmosphere. Correct - the lift vector is not a reference frame. In fact, no vector is a reference frame.Dolphin (t) 12:27, 19 January 2012 (UTC)

Basically, the relative wind reference frame and the atmosphere's reference frame are only the same when you're in level, constant speed flight, and yeah, in that case there's no work being done. But as soon as the wing is accelerating, they're different reference frames, and in general, the relative wind reference frame is a non inertial one, and can generate power. That's how windmills work, the blades' relative wind is constantly changing direction and the lift force is very definitely generating power!!!- Sheer Incompetence (talk) Now with added dubiosity! 07:40, 19 January 2012 (UTC)

There it is again - the relative wind reference frame. Relative wind is a vector, not a reference frame. Dolphin (t) 12:27, 19 January 2012 (UTC)

Reference frames have an origin and a speed, which is a vector. You can align a reference frame with the relative wind, this is commonly done to analyse a jet engine in flight for example. In that frame of reference the air will approach the engine along the x-axis.- Sheer Incompetence (talk) Now with added dubiosity! 17:53, 19 January 2012 (UTC)

When we talk about aligning the relative wind with the axis of a jet engine we are constructing a Cartesian coordinate system. For an explanation of what I mean by a frame of reference have a look at Frame of reference#Examples of inertial frames of reference. Dolphin (t) 22:04, 19 January 2012 (UTC)

Look you said: " If we choose the reference frame attached to the atmosphere, lift never does work because lift is defined to be the component of aerodynamic force that is perpendicular to the relative velocity". But I'm not sure which reference frame you mean by "reference frame attached to the atmosphere". Do you mean the reference frame in which the atmosphere is (upstream) stationary and the wing moves or the one in which the wing is momentarily stationary with the wind approaching (say) aligned along one of the reference frame's axes?- Sheer Incompetence (talk) Now with added dubiosity! 17:53, 19 January 2012 (UTC)

Imagine the wind is blowing at 10 knots and an aircraft is flying upwind at 200 knots TAS (190 knots GS). Using the frame of reference attached to the atmosphere, the air is stationary (except in the immediate vicinity of the aircraft), the aircraft is moving at 200 knots and the Earth's surface is moving at 10 knots. Another way of describing the frame of reference attached to the atmosphere is to say As seen by an observer in a hot air balloon ... Dolphin (t) 22:04, 19 January 2012 (UTC)

If the former, then the relative wind points in different directions depending on what the wing does and the force can certainly do work, as in a helicopter or a windmill. If the later, then it's correct to say that at the moment the wing is stationary that the wind is generating no power (since work=force.distance, so power=force.speed, but speed is zero since it's momentarily stationary. BUT, and it's a big but that acceleration can still be non zero, as in a helicopter or a windmill and so it will not be zero a moment later unless the wing is moving at completely constant speed, which in general it WON'T be.)- Sheer Incompetence (talk) Now with added dubiosity! 17:53, 19 January 2012 (UTC)

If that's what you're talking about, then the wing can't be guaranteed to be stationary unless you're using an accelerated reference frame, but those get even more complicated and don't change the underlying physics. And if it's not stationary then the wing can generate power.- Sheer Incompetence (talk) Now with added dubiosity! 17:53, 19 January 2012 (UTC)

What you're saying is very obviously wrong in general, although is true in restricted situations like level flight. The physics is that lift can and does do work, if it didn't no aircraft would be able to takeoff. I mean what other force operates to lift a helicopter?- Sheer Incompetence (talk) Now with added dubiosity! 17:53, 19 January 2012 (UTC)

This is a very interesting discussion but it has drifted away from improving the article Lift (force). I have transferred the latest discussion to my Sandbox2 and I will post my reply there. Dolphin (t) 02:09, 20 January 2012 (UTC)