Talk:Lift (force)/Archive 3

“Obstruction” explanation of lift

The new section on “Obstruction” theory reads like original research. There are a few problems with this section as written: 1) Calling it “obstruction” theory is something unique to this talk page: Anderson does not use that label and I haven't found a source who does use the term. 2) It needs a citation that says it's an "alternative" explanation. Anderson says it's the simplest explanation of lift. Yet, here it's said to be an alternative explanation. 3) JDA is cited numerous times, but the interpretation is a synthesized analysis. I don't think that the summary appropriately interprets the source.

Since this appears to advance a position in the debate we had previously, I moved the section here for further discussion before putting it back in the article.

• • • Begin new section added 22:19, 8 Februaryh 2009 by Dolphin51

Obstruction

When a fluid flows relative to a solid body, the body obstructs the flow, causing some of the fluid to change its speed and direction in order to flow around the body. The obstructive nature of the solid body causes the streamlines to move closer together in some places, and further apart in others. Different areas on the surface of the body present different levels of obstruction to the fluid flow.

When an airfoil is moving relative to a fluid and generating lift as the result, the upper surface of the airfoil presents a greater obstruction to the fluid than the lower surface. As the result, the streamlines in the fluid flowing over the upper surface move closer together than the streamlines over the lower surface. The consequence of the streamlines being closer together is that the speed of the fluid is faster over the upper surface, and slower over the lower surface. The streamlines near the nose of the airfoil, ahead of the region of maximum thickness of the airfoil, are squashed together the most, causing the speed of the fluid to be greatest near the nose of the airfoil.

In accordance with Bernoulli's principle, where the fluid is moving faster the pressure is lower, and where the fluid is moving slower the pressure is greater. The fluid is moving faster over the upper surface, particularly near the leading edge, than over the lower surface so the pressure on the upper surface is lower than the pressure on the lower surface. The difference in pressure between the upper and lower surfaces results in lift.

• • • End new section

Michael Belisle (talk) 20:56, 9 February 2009 (UTC)

Hi Michael,

1) I agree that Anderson does not call his alternative explanation obstruction theory. Anderson uses the word obstruction numerous times, and it is the key word in his alternative explanation. If someone can find a more appropriate word than obstruction I would be happy for it to be used, but that is the key word used by Anderson so I doubt a more appropriate word can be found without looking outside section 5.19 of Anderson’s Introduction to Flight.

2) It needs a citation that says it’s an alternative explanation. The citation given three times is section 5.19. The title of that section contains the expression alternative explanation. Why is that not a sufficient citation?

Are you satisfied that there are citations saying Equal transit-time and Coanda effect are alternative explanations?

3) WP:SYN says Synthesis occurs when an editor puts together multiple sources to reach a novel conclusion that is not in any of the sources. The paragraphs on obstruction are based solely on the words of Anderson. There is no second source involved. Clearly, this cannot be a case of multiple sources being amalgamated to reach a novel conclusion.

You have written I don't think that the summary appropriately interprets the source.

Are you able to explain where you think the summary differs from the source? Dolphin51 (talk) 23:00, 9 February 2009 (UTC)

1) Giving "obstruction" its own section elevates the word beyond the manner in which I think JD Anderson used it. When he invokes the word obstruction, he does so by the dictionary definition of obstruction as just one part of a complete explanation. But it is not the "obstruction theory“ of lift. This is primarily why I say I don't agree with the interpretation: it defines Anderson's explanation of lift by one of its parts. There is no “obstruction theory“ of lift just as there is no “Kutta condition theory” or “void theory” of lift. These are all concepts frequently encountered in explaining lift, but no reliable sources use them to distinguish between different explanations of lift. Anderson's treatment might be called the “mass conservation and Newton's second law“ explanation of lift (which he says on p. 355).

2) After thinking about it more, this complaint might be moot. I may have misinterpreted the import of the word "alternative" here, probably because this used to be the “Common Misconceptions“ section. When you say "alternative explanation", I read that in the same sense as "alternative lifestyle", i.e. something apart from the mainstream. Also note that on p. 356, Anderson begins explaining what he considers “alternate” explanations of lift “that are in reality not the fundamental explanation, but rather are more an effect of lift being produced, not the cause.” We need to be clear about which sense of alternative we mean.

3) You're correct. Scratch this one too.

My main point is that if we're going to present a bunch of explanations of lift, it doesn't make sense (and isn't NPOV) to have a favored explanation (which is implicit since we present one first and separate from the others), and then present "alternative explanations" unless we do it in the same manner as a verifiable source. Since there is dispute about which explanation is the explanation of lift, we may do well to present a bunch of explanations. But if we include technically-incorrect explanations like ETT and Coanda mixed in with technically correct explanations, we stand the risk of leaving the reader confused, e.g. "I came here to find out how lift is generated and all I got was an argument about how lift is generated."

The technically correct explanations are not mutually exclusive, so I've tried to write the main text of the article to be inclusive. If we agree that JD Anderson's explanation is technically correct, there's no reason it can't be integrated above. Coanda and ETT were once in a separate section because they are not generally accepted by the scientific community as correct explanations of lift. But of course, if we integrate them together, we run the risk of committing an act of WP:SYN.

I'm not sure what to do, but I think that having the main text and then adding a bunch of "alternative explanations" is not the best approach. Even JD Anderson prefaces his explanation of lift by saying that it's "what this author advocates as the most fundamental explanation of lift". A substantial reorganization is probably in order. Maybe we should write an introduction that describes common features present in different explanations of lift, and then present more detailed treatments of the various explanations. Michael Belisle (talk) 00:43, 10 February 2009 (UTC)

Thanks for the prompt response. Anderson's alternative explanation is not the only one to rely on continuity and Bernoulli's principle. ETT is a simple, easy-to-understand explanation of the kinematics of the flow around an airfoil. Continuity and Bernoulli are then invoked to make the transition from kinematics to a resultant force on the airfoil. Anderson's explanation is also a simple, easy-to-understand explanation of the kinematics.

I suggest you restore the summary of Anderson's alternative explanation to Lift (force)#Alternative explanations and then start a new topic on this Talk page to start a debate about the substantial reorganization. (If you don't like Obstruction as the heading, you could try Anderson or Anderson's alternative or Anderson's obstruction.) Dolphin51 (talk) 01:37, 10 February 2009 (UTC)

I have called it "Cambered Airfoils", since that's the heading of the equivalent explanation in The Illustrated Guide to Aerodynamics. The differential area argument as presented needs to be tweaked if the airfoil isn't cambered. The argument is expanded to symmetrical and inverted airfoils on p. 23. Michael Belisle (talk) 01:51, 10 February 2009 (UTC)

Eh, that's not quite correct, since it's not the only way to explain lift on a cambered airfoil. I'll try "In terms of a difference in areas", which is factual without labeling it. Michael Belisle (talk) 02:56, 10 February 2009 (UTC)

1900 lift equation

In the section "Mathematical approximations", the article gives the lift equation used by the Wright brothers, including the Smeaton coefficient, in the early 20th century. Which equation is not used any more in modern aircraft design. However, a treatise is missing on the lift equation and lift coefficient as used nowadays (see e.g. Anderson's Introduction to flight, 2004, 5th ed., pp. 257-261).
Further, the Smeaton coefficient cannot be dimensionless, but dimensions are missing both in the article and referred NASA website. -- Crowsnest (talk) 23:44, 9 February 2009 (UTC)

I agree but haven't gotten around to addressing it. It might be good to incorporate some info from the lift coefficient article. I'm not even sure we need to include the Smeaton coefficient at all. Michael Belisle (talk) 23:59, 9 February 2009 (UTC)

In the article lead is said:

The mathematical equations describing the generation of lift forces have been well established since the Wright Brothers experimentally determined a reasonably precise value for Smeaton's Smeaton coefficient more than 100 years ago ...

This more or less suggests that the modern form of the lift equation is not a big improvement, compared with the equation used by the Wright brothers. However, the 1900 lift equation contains two empirical coefficients: a lift coefficient and the Smeaton coefficient. While the modern lift equation only depends on the lift coefficient, which is a reduction by a factor of two (and now thoroughly based on dimensional analysis).

For a layperson on the subject, this sentence may also easily give the impression that a math calculation of lift is a straightforward thing to do since the early 1900's, while in practice laboratory tests and complex CFD models are most often needed. -- Crowsnest (talk) 00:32, 10 February 2009 (UTC)

I think the usefulness of Smeaton's equation (and its connection to modern approaches involving the lift coefficient) is that it establishes the dependence of lift on the square of the velocity. This dependence, at least, has been recognized since 1900. But you're right that the mathematical understanding of lift has not been around since 1900. That sentence has annoyed me for a while and I don't think it contributes anything to understanding, so I'm moving it here until we come up with something better:

• • • snip from article introduction

The mathematical equations describing the generation of lift forces have been well established since the Wright Brothers experimentally determined a reasonably precise value for Smeaton's Smeaton coefficient more than 100 years ago, but the practical explanation of what those equations mean is still controversial, with persistent misinformation and pervasive misunderstanding.

• • • /snip from article introduction

Michael Belisle (talk) 00:52, 10 February 2009 (UTC)

OK. Perhaps the Smeaton coefficient can have its own (stub) article, or be put somewhere else in a history of lift force or lift equation section. Note that Smeaton Coefficient redirects to this article. -- Crowsnest (talk) 01:12, 10 February 2009 (UTC)

Moved to John Smeaton#Smeaton coefficient. It's all history, so now it's with other history related to him. Michael Belisle (talk) 01:32, 10 February 2009 (UTC)

Thanks! Looks a lot better to me this way. -- Crowsnest (talk) 14:43, 10 February 2009 (UTC)

Determination of the level of obstruction

Am trying to understand this quote from Anderson:

"Different areas on the surface of the body present different levels of obstruction to the fluid flow."

For a given area "X" on the surface of the body, how does one determine the level of obstruction "L"?

Mark.camp (talk) 22:43, 10 February 2009 (UTC)

Anderson’s alternate explanation of lift, which we are calling the obstruction theory, is not a rigorous, math-based, method of quantifying lift. It is a simple, easy-to-understand explanation of lift, suitable for student pilots, aviation enthusiasts and other newcomers to aviation. It appears to serve a similar purpose to the Equal Transit-time theory.

Anderson’s obstruction theory appears to work as follows. You can identify a couple of points on the surface of an airfoil — one subjected to higher air pressure than atmospheric, and the other subjected to lower air pressure than atmospheric. Using the obstruction theory you can then say the level of obstruction of the airfoil surface in the vicinity of the first point is lower than the level of obstruction in the vicinity of the second point. In the vicinity of the first point the streamlines are not squashed together as much as they are in the vicinity of the second point. Consequently the air is flowing slower in the vicinity of the first point than it is in the vicinity of the second point. Applying Bernoulli's principle, you can then conclude that the air pressure is higher in the vicinity of the first point than in the vicinity of the second point. (But this is where you started from, so you are assured of being correct.)

The obstruction theory does not have a scientific definition of obstruction, or obstruction per unit area, or anything else that might put it on a rigorous base. That is why I think it is just a simple, easy-to-understand explanation that will assist newcomers for whom the Kutta condition is still too advanced a concept. Dolphin51 (talk) 23:17, 10 February 2009 (UTC)

It suffices to say that invoking the notion of an obstruction is just a way to skip the formidable task of explaining rigorously why the flow field looks the way it does by giving the reader an intuitive feel for how the flow deforms in the presence of an airfoil. That's it. But unlike ETT, it's not incorrect. It's just the one-sentence, high-level summary of the total picture. Follow these steps to determine the "level of obstruction" (which is not a term used in any reference) and lift:

1. Calculate the streamlines (using for example potential-flow theory as in the NACA 0012 graphic)
2. Apply conservation of mass between the streamlines (${\dot {m}}=\rho UA=$ constant within a streamtube). Since density is assumed constant, the change in area of the streamtube gives you the resulting change in velocity.
3. Use Bernoulli's equation ($p_{t}=p+\rho U^{2}/2=$ constant along a streamline) to get the pressures
4. Integrate the pressure over the surface to get the aerodynamic forces

That's it. Using the notion of an obstruction is just a way to give the reader an intuitive feel for step 1 without getting bogged down in any level of detail (and notably without making up laws of physics like ETT does). This is why I argue against including a detailed section on it: it's just a simple, high-level, qualitative concept. It's not the key to understanding lift in terms of pressures; steps 2–4 are the key. If you want to understand where the obstruction comes from, look into ways of calculating the flowfield. (Actually, there is Bernoulli Obstruction Theory, but that's a topic for another article.) Michael Belisle (talk) 06:50, 11 February 2009 (UTC)

Or try imagining a wind tunnel with a cambered airfoil like in the picture at right. (This will not give a very accurate value for lift, but it illustrates the obstruction concept.) There are two stream tubes: the upper one bounded by the upper wall and the upper surface of the airfoil and the lower one bounded by the lower wall and the lower surface of the airfoil. Upstream of the airfoil, the streamtubes have equal areas $A_{0}$. The streamtubes divide at the leading edge. At the point shown here near the maximum thickness of the airfoil, the upper streamtube has an area $A_{u}$ and the lower one $A_{l}$. Since $A_{u}<A_{l}$, conservation of mass requires the average velocity at this point in the upper streamtube be greater than that in the lower streamtube. Add your favorite numbers to the variables, integrate, and you have the "level of obstruction" and can estimate the lift. Michael Belisle (talk) 07:27, 11 February 2009 (UTC)

How would "obstruction theory" explain these reverse cambered (flat top, curved bottom) airfoils? M2_F2_glider.jpg M2_F3_rocket_powered.jpg Jeffareid (talk) 11:44, 16 February 2009 (UTC)

Hello Jeffareid. It can not explain them. It only works for the cambered airfoil at zero angle of attack shown in the figure. This wind tunnel analogy is an attempt to make it plausible to a layperson that lift is generated. But it faintly resembles the real flow around an airfoil, and quickly becomes invalid in other cases (like the 3D reverse-chambered flying objects pointed out by you). -- Crowsnest (talk) 12:08, 16 February 2009 (UTC)

You could also rotate that cambered airfoil in the diagram 180 degrees, with flat side on top, cambered side on bottom (with "hump" towards the aft end of the wing), and it would also produce lift, although with a lot more drag. Obstruction theory seems too much like equal transit time to me. I still prefer what I term "void theory" (better stated as "void abhorence theory") to explain the Coanda like effect of air following a convex surface (because if it didn't a void would be created). Given this, then it could be explained that an airfoil separates an air flow then diverts (accelerates) the flow downwards via interaction with upper and/or lower surfaces of the airfoil. Jeffareid (talk) 00:08, 17 February 2009 (UTC)

Thanks, both of you. Since it isn't intuitively obvious, please include in the article a one-sentence version of the above explanation of how the reader would determine the relative level of obstruction of two arbitrary surfaces on a wing. Please make the explanation simple, intuitive, and non-mathematical?

Mark.camp (talk) 11:44, 11 February 2009 (UTC)

"When fluid flows past a cambered airfoil at zero angle of attack, the upper surface has a greater area than the lower surface and hence presents a greater obstruction to the fluid than the lower surface." is a one-sentence version of the idea (a picture like figure 2-11 here would help). If there's something here that makes it clearer to you, please improve the article. But I'm not sure it's possible to explain how to determine the amount of obstruction without math and I think that's going into too much detail with this idea. Michael Belisle (talk) 16:35, 11 February 2009 (UTC)

Put the hump at the back of the airfoil, (turn the wing backwards), and it's the same surface areas, but it's generates downforce and more drag. I don't see the validty of the greater obstruction theory. Jeffareid (talk) 08:07, 2 March 2009 (UTC)

For zero angle-of-attack and a chambered airfoil: moving forward, most of the lift is near the leading edge of the airfoil (see J.D Anderson, p. 352-361). Missing in his depiction of the flow is the influence of the sharp trailing edge, enforcing the flow to separate there instead of somewhere on top of the wing. When using the Kutta condition and potential flow, this backward movement of the separation point (to enforce it to be at the trailing edge) will further increase the velocity on top of the wing, and decreases it below. And by Bernoulli's principle increases the pressure on the lower side, while decreasing it on the upper side.

And now to the situation you sketch: with the wing moving backwards through the air, the flow separates on the round side (now at the back). There is no fixed separation point (so no a Kutta condition enforcing circulation), and just as for other bluff bodies like circular cylinders there will be a wide turbulent wake and a large amount of drag. The flow near the round trailing edge (where most of the lift was in forward flight) will be strongly changed by the presence of the wake. The result will be hardly any lift and a lot of drag, so perhaps that is the reason you do not see this type of wing.

And yet the m2-f2 and m2-f3 mentioned above with links to pictures glided and flew just fine, probably a lot of drag, but quite a bit of lift. Jeffareid (talk) 06:44, 6 March 2009 (UTC)

In fact, they are quite streamlined. As correctly pointed out by you, they cannot be explained using J.D. Anderson's intuitive approach. Also note that these are 3D lifting shapes. Lift on them can however be calculated using "panel methods" (boundary element methods), computer models for potential flows including vortex sheets (lifting-line theory) as introduced by Hess & Smith in the early 60's. -- Crowsnest (talk) 13:34, 6 March 2009 (UTC)

So, the flow field will change drastically when reversing the direction in which the wing moves: with a wide region of separated flow (the wake) and probably — as for a circular cylinder — multiple (and possibly unsteady) separation points at the rounded rearward side. Potential flow theory cannot predict such a turbulent wake flow. The argument of J.D. Anderson, based on his expectation of how a potential flow around a wing looks like, will no longer be useful. As said before, all these simple explanations contain additional assumptions, limiting the range of conditions for which they can qualitatively make it plausible that lift is generated by a wing. -- Crowsnest (talk) 19:55, 2 March 2009 (UTC)

My point here is that it's the shape of the airfoil, not relative surface areas. A thin cambered airfoil has the same surface areas above and below, yet because of it's shape, it generates lift with a reasonable amount of drag. Jeffareid (talk) 06:44, 6 March 2009 (UTC)

"A thin cambered airfoil has the same surface areas above and below"? Can you explain this. A cambered airfoil has more surface above than below its chord line. -- Crowsnest (talk) 13:34, 6 March 2009 (UTC)

Well virtually the same. A very thin cambered airfoil, such as a kite style hang glider, or the early gliders and planes that used a single "sheet" of cloth for the wing surfaces. I also mean real thin airfoils, not the zero thickness airfoils use in thin airfoil theory, however some stuff here: Airfoil#Thin_airfoil_theory Jeffareid (talk) 20:40, 6 March 2009 (UTC)

You can attach a plus or minus sign to the area's, depending on their side of the chord line: in this case the area associated with the upper surface of the airfoil is positive, while that of the lower surface is negative (because it is above the chord line, instead of below for a thicker airfoil). So then it is in line with the expectations, when using the surface-area argument to make lift plausible. -- Crowsnest (talk) 21:15, 6 March 2009 (UTC)

You are right that it is in the end the shape that determines the lift. But the area is dependent on shape, so that is why it is used in this surface-area approach to make the occurrence of lift, using a cambered airfoil as example, plausible. -- Crowsnest (talk) 21:22, 6 March 2009 (UTC)

Signed surface areas? What about symmetrical airfoils? Jeffareid (talk) 16:40, 7 March 2009 (UTC)

If the upper surface is fully above the chord line, all its contributions to the upper area are positive. If the lower surface is fully below the chord line, it also has only positive contributions to the lower area. For the rest you can forget about this, because it is original research. -- 16:47, 7 March 2009 (UTC)

Sorry, you are right--I misinterpreted the sentence. I thought it was restricting the definition to the specific case of those two surfaces.

But as I re-read it, I see it didn't say necessarily say that. So, it is clear how the reader can determine the relative amount of obstruction, at least for the case of a typical doubly convex positive-cambered airfoil, at zero angle of attack as conventionally defined.

To put it simply, whichever surface has the greater area, presents the greater obstruction.

This is exactly what I was looking for, a simple definition, not a complex mathematical one.

Mark.camp (talk) 19:12, 11 February 2009 (UTC)

Jeffareid wrote (08:07, 2 March 2009):

Put the hump at the back of the airfoil, (turn the wing backwards), and it's the same surface areas, but it's generates downforce and more drag. I don't see the validty of the greater obstruction theory.

Although I agree that the theory isn't valid, that isn't relevant. Its inclusion is advocated by one contributor, who has included a citation of a book which mentions the assertion, though only vaguely, indirectly, and in passing. This is apparently all that is required by Wikipedia policy. Attempts to debate its objective validity, such as yours above, or to find out what the theory is actually saying, have been unproductive.

I think I came across something about an official Wikipedia initiative to eliminate junk science from its pages. I didn't stop to read about it, but I hope that such an effort is underway, and that it will raise the academic standards for submissions in the future, and we will be able to eliminate scientifically unacceptable speculations such as ETT, the celestial dome theory, and differential obstruction Mark.camp (talk) 18:45, 2 March 2009 (UTC)

Mark, as said before, if it gets undue weight in the article (since his line of arguing is not/hardly repeated in articles or books by others), we may leave it out (if we can reach consensus on that). Otherwise, the phrasing of it may be changed to increase the neutral point of view of the article. Something like: "John D. Anderson, in his book ..., gives the following explanation of lift occurring for a chambered airfoil at near-zero angle of attack: ...". -- Crowsnest (talk) 20:08, 2 March 2009 (UTC)

ETT needs to be mentioned, since it is such a widespread theory which has been proven wrong. It would be strange if it was not mentioned in this encyclopedia. -- Crowsnest (talk) 20:13, 2 March 2009 (UTC)

I like that phrasing. I disagree with leaving it out entirely. Michael Belisle (talk) 21:40, 2 March 2009 (UTC)

Since he is an authority on aerodynamics, this has also my preference. -- Crowsnest (talk) 21:49, 2 March 2009 (UTC)

Flowfield generation and other explanations

I moved "Other explanations" up to the section about the process through which the flowfield around an airfoil is formed. These other explanations are not so much alternative explanations of lift, but alternative ways to simplify Step 1 above. For the most part, they don't affect steps 2-4. Additional sections on at least the Kutta Condition, Curvature, and maybe Circulation should be added.

However, these really have more generally applicability beyond lift. I think that perhaps the detailed explanation should be moved to the airfoil article and a terse summary presented here. The thing about lift is that if you understand all the principles related to an airfoil in a flow, then lift just comes "naturally" from the principles. It's a bit of wasted effort to focus so much time on explanations of the principles specific to lift when small changes make them more general. Michael Belisle (talk) 20:10, 11 February 2009 (UTC)

Hello Michael, it is not clear to me what your main idea is with the present resectioning. The "Flowfield formation" subsection is on the transitional stage from a near-potential flow (without circulation) just after the start from rest, to the steady state in flight as described in subsection "Lift in an established flow".

But the "Equal transit-time" fallacy, the "Coandă Effect" explanation, and the "In terms of a difference in areas" alternate explanation are about the steady state in established flow again. Side remark: what is the Coandă effect anyway, for the case of a single-density fluid at low Mach numbers? Is it well defined, other than just saying (not explaining why) the flow sticks -- for low enough angles of attack, before stalling -- to the wing. Does it verifiably provide a real physical mechanisms distinguishable from other mechanisms?

Further the whole approach of step 1 to 4 is limited to low Mach numbers, using an incompressible potential flow approximation. Only in that case the streamlines can be computed from flow kinematics, separate from the flow dynamics. At transonic and supersonic speeds such a separate solution of flow kinematics (streamlines) and dynamics (density, pressures and lift) is impossible. -- Crowsnest (talk) 23:50, 12 February 2009 (UTC)

I would like that the three alternative descriptions (equal transit time; Coanda effect; area approach) go back to a separate section, as they were before. I do not object including Anderson's and/or Smith's alternative explanations, if not given overdue weight (and text length). If these renowned experts in the field are not capable of providing a "simple explanation" for lift for general flow conditions, we probably cannot either. -- Crowsnest (talk) 02:28, 13 February 2009 (UTC)

My point with the reorganization is that ETT, Coanda, Areas, Curvature, and Kutta+Circulation are all ways to address the question "Why does the flow field around an airfoil look the way it does?" They aren't alternative explanations of lift; they're alternative explanations of the first step. The first step, though, has little to do with lift and I think it should be deemphasized in this article.

See Coanda effect for an explanation of the Coanda effect. It's a real effect, just not applicable to airfoils in general. Some employ it to as a way of hand waving and say that the flow follows the surface of the airfoil because of the Coanda effect.

As for your comment on supersonic and transonic flow, sure. Those should be another section in the article, along with hypersonic flow and 3D flow and the consequences of lift generation. The current article explicitly limits the discussion to incompressible, 2-D flow. We're spending so much time arguing about a throwaway line about obstructions that it's tough to move on to more complicated topics. Michael Belisle (talk) 03:07, 13 February 2009 (UTC)

Unfortunately, the first step is causing all the trouble in explaining lift. The other three, as you showed, are straightforward consequences of mass conservation, and using Newton's 2nd law c.q. the Bernoulli equation.

The Coandă effect is historically associated with jets (of different density) flowing over a convex surface. As you said, it is often used in a loose sense, remarking that flow attachment on a convex surface is "...due to the Coandă effect...": just labelling without explaining and identifying physical mechanisms. I think the description in the article, as it is, is quite adequate.

In my opinion, "Flowfield formation" (formerly "Stages of lift production"), with its description closely linked to "Lift in an established flow", is more appropriate in this article than it is in the airfoil article. The same is true for the three alternative explanations of lift, which are specific on lift (more than on the general subject of an airfoil). -- Crowsnest (talk) 03:45, 13 February 2009 (UTC)

Scope of Differential Obstruction Theory

Does the differential obstruction explanation apply to the potential flow model? I mean the model which assumes 2D, steady, inviscid, incompressible flow. Mark.camp (talk) 12:20, 12 February 2009 (UTC)

It's a natural result in potential flow, yes. I'm not sure what you mean by "applies" since it's a simplified description of the actual physics. Potential flow (with sources, sinks, etc) is a more quantitative way of simplifying the physics. If you put a source in a free stream, yes, it'll act like an obstruction. Michael Belisle (talk) 19:13, 12 February 2009 (UTC)

Thanks, you interpreted my question correctly. I am working to understand the differential obstruction theory, and I did not know what model it starts with (inviscid or no, steady-state or no, etc.,) Mark.camp (talk) 22:01, 12 February 2009 (UTC)

I believe the theory is applied to two-dimensional, steady-state, inviscid flow. Its value is in helping newcomers to feel comfortable with the kinematics of the primary flow around an airfoil when it is generating lift. The situation with a three-dimensional airfoil, and transient effects, and the effects of viscosity are all more advanced concepts that are not needed by newcomers. Dolphin51 (talk) 22:12, 12 February 2009 (UTC)

Thanks, both. Does it apply only to asymmetric sections? Mark.camp (talk) 23:00, 12 February 2009 (UTC)

Anderson only mentions cambered (asymmetric) airfoils, but I see no reason to imagine it is not relevant to symmetric sections. We know the pressure on the upper surface of a symmetric airfoil generating lift is lower than on the lower surface, so we can say the upper surface must be presenting a greater obstruction, causing the streamlines to be squashed closer together near the upper surface than near the lower surface. (Yes, Anderson actually uses the word squash. This is a very simple explanation.)

Michael Belisle has written about the greater surface area on the upper surface of a cambered airfoil than on the lower surface, and equated the greater area to greater obstruction. This is an attractive variation, but newcomers might imagine there is a linear relationship — if the path length of the upper surface is 10% greater than the path length of the lower surface, newcomers to aeronautics might imagine the airspeed is 10% faster past the upper surface than past the lower surface. This would be as much a fallacy as the Equal transit-time model which is sometimes called a fallacy. The difference in path length does not change with angle of attack; it is a characteristic of the geometry of the airfoil section. However, we know that the difference in speed of the air passing the upper and lower surfaces is strongly dependent on angle of attack, and can be much, much greater than 10%.

I think the best thing to do with Anderson's obstruction theory is to use it to gain an acceptance that airfoils generate lift by causing the air to flow faster past the upper surface than the lower, and then move on to something closer to a rigorous scientific explanation of lift, such as the Kutta condition, circulation and the Kutta-Joukowski theorem. Dolphin51 (talk) 23:19, 12 February 2009 (UTC)

When deviating from J.D. Anderson's description in his book, within the article, that is effectively original research. Anderson claims on p. 355 (5th ed.) that his explanation is valid for flat plates at an angle of attack as well. The weak point in Anderson's alternate explanation is that it does not explain why the streamlines are more dense on the upper side of the (chambered) wing. It inherently contains all his knowledge about the streamline patterns around such an airfoil. If you ask somebody unknown to flow around airfoils, he might even find arguments why he thinks the flow should be more obstructed by the lower side. And consequently that the streamlines should be denser there. Although introduced by a renowned authority, this alternate explanation does not seem to have become popular over the past 30 years (since the 1st print in 1978). So is it notable enough to be included in the article? The "equal transit time" fallacy is, since it is very popular. As probably is the "Coanda effect" (although it does not explain anything, in my opinion: it is just a label without meaning for flows without density differences). -- Crowsnest (talk) 00:25, 13 February 2009 (UTC)

There are at least two sources: the Illustrated Guide to Aerodynamics also includes it. It's a pretty basic idea, so I wouldn't be suprised if there are more.

I did not say greater surface area. In 2-D flow that would be equivalent to saying that the surface has a longer path, which is wrong. The difference in interior area comes from the Illustrated Guide to Aerodynamics, linked previously and referenced in the article. If you will continue reading from page 20 to page 23, you'll see that the idea is extended to symmetric airfoils at angle of attack and on p. 24 to upside-down cambered airfoils.

One might be able to generate arguments why the lower side should have a greater obstruction, but one would be wrong and should perhaps consider reviewing elementary geometry. I don't see how you say that it doesn't explain why the streamlines are more dense on the upper surface: to me that exactly what it does do. That's the reason why he brings up the obstruction and it's only thing bringing it up attempts to do. See my figure above for another way of putting it. The streamlines must be closer together because there is no other way. Michael Belisle (talk) 00:55, 13 February 2009 (UTC)

Michael, thanks for your clarification in Lift (force) that the area pertinent to the upper surface is the interior area between the chord line and the upper surface of the airfoil. What would be more meaningful would be the area of the section, divided not by the chord line, but by the line between the trailing edge and the leading edge stagnation point. That explanation would account for symmetrical airfoils because, as angle of attack increases, the stagnation point moves towards the lower surface. It would also account for a progressively increasing area obstructing the flow as angle of attack increases, and a progressively decreasing area as angle of attack decreases and becomes negative.

However, it is likely this enhancement is not to be found in the Illustrated Guide to Aerodynamics so it is only my original research and cannot be added to Wikipedia. Dolphin51 (talk) 01:40, 13 February 2009 (UTC)

No, that enhancement is in the Illustrated Guide. With pictures, even. Michael Belisle (talk) 01:53, 13 February 2009 (UTC)

Regarding Michael's response at 00:55, 13 February 2009 (UTC):

Yes, there are no doubts on verifiability. But if others also consider Anderson's alternate explanation an appealing one, one would expect quite some references.

You are certainly right that it is erroneous to consider the streamlines denser on the lower side of a chambered airfoil at zero or positive angle of attack. But my main point it that J.D. Anderson's explanation is only more or less intuitive for a chambered airfoil at near zero angles of attack. For large angles of attack a layman may argue that the obstruction posed by the vertical distance from leading to trailing edge is much larger than from leading edge to top of the chambered airfoil. And erroneously sketch streamlines more dense at the lower side. Anderson's explanation is strongly dependent on his fast experience of what the flow field should look like (see e.g. p. 356 for the flat plate case). Within your low Mach-number approach of four steps: it is still difficult to explain how to obtain the correct streamlines in step 1, if you have no expectations based on experience what a (potential) flow around an airfoil looks like.

With regard to your figure above: it has has strong blockage effects, different from an airfoil in usual flight conditions (see also objections against the Venturi explanation at the NASA web site). So that makes it not very applicable for a general explanation to be used in the article (apart from being different from the alternate approach of J.D. Anderson to explaining lift). The same is true for H.C. Smith's explanation, which is also inconsistent with Anderson (p. 353-355), who states that the lowest pressure is close to the leading edge and not where the wing is thickest (in the blockage approach of Smith). Note that for (symmetric) airfoils at non-zero angles of attack also Smith heavily relies on his knowledge of streamlines (stagnation points) in his line of argument. -- Crowsnest (talk) 02:02, 13 February 2009 (UTC)

A straight application of H.C. Smith's approach of a finite-height windtunnel to situations of e.g. symmetric airfoils at angles of attack would lead to ridiculous results. He uses his knowledge on streamlines to sketch the effects of non-zero angles-of-attack on the upper and lower area's. My main objections against Anderson's and Smith's alternative explanations is that they do not explain much if you do not know what the flow around an airfoil should look like. So for a lay person do not provide more insight than what is given in the section "Lift in an established flow", where the streamlines are just given. -- Crowsnest (talk) 02:18, 13 February 2009 (UTC)

Thanks Michael, for placing my view on this also in the previous section, where it also applies. The last lines of it also applies here, so I copied it here as well. Crowsnest (talk) 03:09, 13 February 2009 (UTC)

I would like that the three alternative descriptions (equal transit time; Coanda effect; area approach) go back to a separate section, as they were before. I do not object including Anderson's and/or Smith's alternative explanations, if not given overdue weight (and text length). If these renowned experts in the field are not capable of providing a "simple explanation" for lift for general flow conditions, we probably cannot either. -- Crowsnest (talk) 02:28, 13 February 2009 (UTC)

You pointed out that this exceedingly simple explanation doesn't apply to extreme cases. Of course not. It's a simple concept that applies to simple situations. If it doesn't make intuitive sense to you, then it's not the explanation for you. There are others, hence the presentation of a bunch of explanations (with a few yet to come beyond what is there now).

The diagram is different from an airfoil in flight. But it's how things work in a wind tunnel. It's a related situation.

We are not to provide our own "simple explanation" of lift for general flow conditions. The best we can do is present a summary of what is written by the experts in a NPOV way. Michael Belisle (talk) 03:15, 13 February 2009 (UTC)

It does not only not apply to extreme cases, it does not apply to moderate angles of attack for the chambered airfoil, and not at any angle in case of a symmetric airfoil. For moderate angles of attack: only by bringing in experience on flow patterns around airfoils -- i.e. only if you know beforehand what the flow field should approximately look like -- you can predict it.

No, the diagram is not what happens in a windtunnel, it is a strong simplification assuming a uniform velocity from the airfoil to the windtunnel wall. With that assumption you can apply mass conservation to determine that uniform velocity. But the real flow near the airfoil is very different from that (as are the pressures, as also pointed out by Anderson): velocities near the airfoil are very different from the mean velocity between the airfoil surface and the windtunnel wall. Quantitatively that will largely underpredict lift.

I fully agree, so I do not object against their inclusion in the article. But we must be careful not to mix J.D. Anderson's and H.C. Smith's approaches, which are different in many respects. -- Crowsnest (talk) 04:10, 13 February 2009 (UTC)

HC Smith considers symmetric airfoils at moderate angle of attack, so I'm not sure what you mean. It's also not a means of prediction: it gives you no quantitative answers in any way. It's like "First, let's start by giving you a high-level, intuitive feel for what's going on by talking about areas and obstruction. Then we'll move to specifics like viscosity or potential flow and the Kutta condition."

The diagram is what happens in a wind tunnel, as far as the average velocity is concerned. It's a standard control-volume analysis, but I neglected to draw the control volume. If we consider the upper $A_{0}$ and $A_{u}$ lines as the inlet and outlet of the upper control volume (and likewise for the lower CV), $U_{0}A_{0}=U_{u}A_{u}=U_{l}A_{l}$, where $U$ refers to average velocity across the boundary. That's it. It's basic continuum mechanics. But you are right that the velocity isn't necessarily uniform, which is why I said it's not a very accurate calculation of lift.

Related to this analysis is the fact that it's fairly standard to measure lift by means of static pressure taps on the upper and lower walls of a wind tunnel test section. You can also measure drag using a pressure rake in the wake of the airfoil. The validity of doing so can be shown by applying a momentum balance to the control volume. Michael Belisle (talk) 07:25, 13 February 2009 (UTC)

Take the symmetric airfoil at the right. H.C. Smith's windtunnel approximation, using mean velocities, will result in even the wrong sign of lift: downward instead of upward. Near cross section B the mean flow velocity — averaged from above the foil to the windtunnel ceiling — is much smaller than below. From Bernoulli's principle it follows that the pressure above is higher than below the airfoil. Resulting in a negative lift, concentrated near the trailing edge. -- Crowsnest (talk) 22:22, 15 February 2009 (UTC)

Sure you can measure lift by static pressure taps in ceiling and floor of the wind tunnel, as a result of a force balance (momentum fluxes through windtunnel walls and wing surface being zero). But that has nothing to do with the mean-velocity approach of H.C. Smith.

Further, lift can also be estimated by measuring vertical velocities in the wake of the airfoil and using a momentum balance (see Landau & Lifshitz, note 5 in the article). -- Crowsnest (talk) 22:34, 15 February 2009 (UTC)

Oooh, good point with the symmetric airfoil at angle of attack. Note anyway that it's my explanation, not H.C. Smiths. I didn't see him use it. Michael Belisle (talk) 06:08, 18 February 2009 (UTC)

Calculating the streamlines

On 11 Feb, Michael wrote:

Follow these steps to determine the "level of obstruction" (which is not a term used in any reference) and lift:

1. Calculate the streamlines (using for example potential-flow theory as in the NACA 0012 graphic)

Must the Kutta condition be imposed in this step? Mark.camp (talk) 21:54, 13 February 2009 (UTC)

Yes, the Kutta condition is needed to determine the streamlines. Although the Kutta condition is applied to fix the circulation in the potential flow, the rationale behind why the Kutta condition is a good approximation requires much more than just potential flow theory. -- Crowsnest (talk) 00:09, 16 February 2009 (UTC)

I agree with everything you said. So here is the point of my question: if the theory starts by assuming potential flow conditions, plus the Kutta condition, then we have the classical explanation of lift--the streamlines follow by applying Newton's second law and Bernoulli's principle (which itself follows from Newton's laws) and so does the lift.

Then, what purpose is served by introducing the concept of "differential obstruction"? It seems superfluous.

I think the answer is that Michael does not agree with what you said. In particular, when I said the same thing you just did, he said I was exaggerating the importance of the Kutta condition. Not sure what his thinking is in this area--asked in what way this was an exaggeration, and don't recall ever getting an answer.

In fact, I think that the differential obstruction theory is the opposite of assuming the Kutta condition holds. (And, as you point out, explaining why it holds in a separate discussion, which must transcend potential flow and take viscosity into account). Differential obstruction theory, as presented by Michael (Anderson says almost nothing about it) seems to be is an attempt to explain lift in potential flow theory without assuming that the Kutta condition holds.

I think that's mathematically impossible, in a sense. One could of course impose the forward stagnation point, or the circulation, by pulling them out of thin air. But the only way to impose a circulation in a physically justifiable way is with the Kutta condition. Mark.camp (talk) 22:29, 16 February 2009 (UTC)

The Kutta condition is necessary (but not sufficient) to get lift in potential flow. But potential flow theory is not the only way to get the streamlines. Other ways include experimental measurements and solving the Navier-Stokes equations.

I said you were overemphasizing the Kutta condition because, in the physical world, the Kutta condition is neither necessary nor sufficient for lift. The real world is viscous and hence one does not need to impose a circulation. It arises naturally. Michael Belisle (talk) 07:20, 18 February 2009 (UTC)

Illustrated Guide to Aerodynamics as an original source

The explanation of lift given in this popular science book, which is not peer-reviewed by scientists, and written not for scientists but for aircraft pilots, is the foundation of the interpretation of differential obstruction theory.

The theory is what I think of as the "Celestial Dome" theory. The theory starts, like all popular theories of its kind including ETT, with the special case of an asymmetric section at zero angle of attack, with potential flow, and no mention of the Kutta condition. The chord of the wing is treated as a physically significant parameter--another common thread in these theories, including differential obstruction and ETT--and it is observed that there is a high spot in the wing relative to the chord. The sky is treated as an impermeable surface--a celestial dome--so close to the upper surface of the wing that it and the high spot form, in effect, the narrow part of a Venturi tube. Pressure is reduced and lift results. The earth is regarded as so close that it and the relatively flat lower surface form another passage, this one with "more room" for the air to flow (in the language of Illustrated Guide, ie, Celestial Dome theory) or "less of an obstruction" (in the language of differential obstruction theory).

I think that the celestial dome theory is very easy to understand. But I also think that it has no scientific validity and the incorrect assumpion on which it is based is easily demonstrated. For that reason, I don't thing that the Illustrated Guide is a good first source for this article.

Anderson never explains differential obstruction theory, which has itself been raised here as a reason not to include it in the article even as an alternative theory. The theory is mainly introduced only on this page--original research. I'm more concerned that the original research presented here isn't scientifically valid.Mark.camp (talk) 03:01, 15 February 2009 (UTC)

The verifiability and status of obstruction theory appears to be the same as equal transit-time theory. Both are simple, easy-to-understand attempts to de-mystify the kinematics of flow around an airfoil. There are references which can be cited to support Wikipedia's coverage of both, but the titles obstruction theory and equal transit-time theory appear not to be used in any of those citable references. Wikipedia's coverage of equal transit-time was originally called a fallacy until Michael Belisle correctly changed it to the more appropriate theory. As far as I am aware, use of the work fallacy was pure original research by the original editor.

I originally added some information on obstruction theory because this theory was being used in the body of the text, not as an alternative explanation of lift, but as part of the mainstream explanation. Anderson's Introduction to Flight, p.353 was the citation to support this element of mainstream explanation. See Lift (force) immediately prior to Michael Belisle's first edit on 15 January 2009. Having focussed attention on Anderson's alternate obstruction explanation there now appears to be much wider acceptance that this is not a rigorous explanation of lift and therefore not appropriate as part of the mainstream explanation.

Obstruction theory and equal transit-time have the same status and verifiability. If one goes from Wikipedia they should both go. Coanda effect may also share the same status and verifiability as the other two. Perhaps there is no longer a place in Lift (force) for alternative explanations. Dolphin51 (talk) 05:21, 15 February 2009 (UTC)

Do you agree that "Illustrated Guide to Aerodynamics" lacks scientific rigor, and for that reason shouldn't be relied on as an authority about what is accepted scientific theory? I am concerned that if this sort of popular science book is treated as if it were a scientific source document, that the floodgates will be opened for every book full of invalid theories, which a single well-meaning individual invented in order to explain the science. The scientific community has rigorous standards that must be met before an idea is accepted as valid. The above theory has not met those standards. A high school physics student could compare the accepted potential flow theory and the celestial dome theory side by side and see that they are mutually exclusive.

Think of the simple example of a circular section with circulation 6 * PI and free flow speed in the positive x direction of 1. The accepted potential flow theory, sans Kutta condition, gives a simple explanation based solely on Newton's second law, for

the position of the stagnation point, which is somewhat above the wing.

the nature of the flow near the wing, which is a series of concentric rings in the counterclockwise direction.

the direction and amount of lift. It is negative in this case.

The theory in the above book would have to give some explanation completely incompatible with the above. The lift has reversed from the normal case, so perhaps the advocate of the celestial dome theory would explain that the sky must have gotten higher, or the circle has gotten closer to the earth.Mark.camp (talk) 18:26, 15 February 2009 (UTC)

Professor Hubert C. Smith appears to be a renowned scientist in the field. His attempt to explain lift in a simple way works to some extend for a chambered airfoil at zero angle of attack. -- Crowsnest (talk) 23:39, 15 February 2009 (UTC)

We agree to disagree then. You believe the theory is scientifically valid and I think that it is not.Mark.camp (talk) 00:26, 16 February 2009 (UTC)

Personally I do not think the theory is scientifically valid (for instance not verified by experiments or against numerical results obtained from established methods of lift calculation). It is one of the attempts to give a simple explanation for something that does have an explanation, but not a simple one. But what counts on WP is whether it is verifiable (by reliable sources), relevant and notable. Prof. Smith's book is published by McGraw-Hill and prof. Smith has received several awards. Both H.C. Smith's and J.D. Anderson's alternative explanations do not seem to be contested, nor supported, in the scientific literature. Both appear to be reliable sources in the WP sense, see WP:RS. So at the moment only notability (see WP:UNDUE) and relevance for the article are important regarding whether — and to which extend — to include these explanations in the article. -- Crowsnest (talk) 00:41, 16 February 2009 (UTC)

The above makes me think of (see Occam's razor#Science and the scientific method):

Albert Einstein probably had this in mind when he wrote in 1933 that "The supreme goal of all theory is to make the irreducible basic elements as simple and as few as possible without having to surrender the adequate representation of a single datum of experience" often paraphrased as"Theories should be as simple as possible, but no simpler."

The explanations suffer from being too simple to be able to give an adequate description of what is really happening regarding the streamline pattern (2D potential flow, Kutta condition) -- Crowsnest (talk) 12:44, 16 February 2009 (UTC)

In this case, the problem is in Wikipedia's policies, and not one which we can solve here. As currently stated, they permit unreviewed, unscientific content in ostensibly scientific articles.

Here is the problem with that. If the Wikipedia user opens a scientific article to learn what science says about a scientific subject, and comes away with not only with rubbish, but with rubbish uncritically presented as scientific fact, then the potential value of this online encyclopdedia can never be realized. Mark.camp (talk) 03:16, 16 February 2009 (UTC)

Well, Smith and Anderson both give it a try, to make lift a plausible phenomenon for a layperson. They should be credited for that.

But with regard to inclusion, and (if included) how it is presented, we can decide by trying to obtain consensus. The WP rules and guidelines are aimed at preventing the horror scenario you sketch. With regards to the Smith and Anderson explanations (which are different) I see the following reasons why they should not be included:

1. These two mechanisms for explaining lift do not seem to be backed-up or preceded by others in the scientific literature. As stated in WP:RS and WP:PSTS, that makes them primary sources, while WP articles should be based on secondary sources.
2. In WP:UNDUE it is said: "If a viewpoint is held by an extremely small (or vastly limited) minority, it does not belong in Wikipedia regardless of whether it is true or not and regardless of whether you can prove it or not, except perhaps in some ancillary article." And: "Keep in mind that in determining proper weight we consider a viewpoint's prevalence in reliable sources, not its prevalence among Wikipedia editors." So, both explanations only being based on one source, and not being supported by others, are not significant enough to be included.

On the other hand, if we do decide to include them, we have limited possibilities to express our view, since that would easily become original research. Rreliable sources for our critique seem to be lacking, although Smith's reasoning can be considered similar to the Venturi explanation criticized at the NASA web site. -- Crowsnest (talk) 10:27, 16 February 2009 (UTC)

Thanks for these comments.

I fear that all of this will lead to great confusion in Wikipedia's scientific articles.

In science, the burden of proof is on the proponent of any new scientific theory.

In Wikipedia's policies, its seems that a new theory contradicting the accepted theory need not have been critically reviewed by scientists to see if it is valid. The burden of proof is on anyone who wished to remove it to find a book claiming to prove the new theory false.

This makes me very concerned. I think its a case where the old rules worked better.

A man I worked with years ago sought investment money from his co-workers for a get-rich quick scheme based on a perpetual motion device. He could argue for hours on why the well-accepted scientific theory denying perpetual motion was wrong and his theory was correct.

The most influential encyclopeda at the time was Brittanica. If that encyclopedia's policies were such that, merely by finding a non-peer-reviewed book on the new theory, this fellow could get his theory published in the Encyclopedia as accepted science, he'd have been delighted to get this stamp of authenticity. But the public's knowledge of true science would have suffered.

It seems that for Wikipedia, science is in effect guilty until proven innocent, when an unreviewed theory is entered which contradicts the well-accepted scientific theory. Mark.camp (talk) 13:14, 16 February 2009 (UTC)

The burden of proof is also in Wikipedia with the editor who includes material in the article, see WP:PROVEIT. With regards to controversial theories: see also WP:REDFLAG. So there are rules to reduce the risks you identify. -- Crowsnest (talk) 13:27, 16 February 2009 (UTC)

(Indent reset to zero)

Thanks, Crowsnest. I've now read PROVEIT. Good stuff as far as it goes. It requires that sources be cited. But the proponents of the differential obstruction theory, and the celestial dome theory on which it is built, have done that.

Concerning the content of the referenced info, this Wikipedia policy imposes no burden of proof of the scientific validity of the new theory, either on the author of the source, or on those who edit the theory into Wikipedia.

Firstly, the book itself is acceptable even though it has already bypassed the burden of proof required by the scientific tradition for new scientific assertions. Secondly, the editors who include the text in Wikipedia haven't violated the policy even if they fail to provide satisfactory answers to serious questions about the validity of the theory.

Suffice it to say that I object to either of these theories being presented in this article, unless it is to make it clear that they are fallacies. To include them as fallacies may or may not be warranted; I don't much care whether they are not. Note that when I use the term "fallacy" to describe a theory, it is with the generally accepted definition of the term in math and philosophy, including logic and science. It is an objective description of a theory as written, and implies absolutely nothing about the proponents of the theory.Mark.camp (talk) 18:53, 16 February 2009 (UTC)

All we have done is to present properly cited information without passing our own judgement. I have now added a third citation. Two books are university-level textbooks (Introduction to Flight and Introduction to Aeronautics), and one is a book for general audiences (Illustrated Guide to Aerodynamics). All authors are respected in their field (most notably JD Anderson, Curator of Aerodynamics at the Smithsonian). The presentation meets all the standard criteria and hence is suitable for inclusion in WP. I'll respond to two specific complaints:

1. Even if these books are primary sources, WP:PSTS says “Primary sources that have been reliably published (for example, by a university press or mainstream newspaper) may be used in Wikipedia, but only with care, because it is easy to misuse them. Any interpretation of primary source material requires a reliable secondary source for that interpretation.” AIAA and McGraw-Hill are reputable university-level publishers. Both books are used in university-level Aerospace Engineering courses.
2. WP:UNDUE says “If a viewpoint is held by a significant minority, then it should be easy to name prominent adherents.” It is easy: the article now names four prominent adherents. The mention is brief and concise. Undue weight was previously a concern when it was included in the mainstream text, but now it is on equal footing with other explanations of its class. The weight will be even less when other alternative explanations are added.

Although you (mark.camp) personally have decided that the theory is a fallacy, are there any reliable sources for your claim? If not, then you have not met the standard imposed by WP:PROVEIT and your conviction, as logical as you may believe it is, is original research. Since you claim they are primary sources, you would do well to reread the last sentence above: “Any interpretation requires a reliable secondary source.”

Finally, as Dolphin51 pointed out above, the first sentence is WP:V is “The threshold for inclusion in Wikipedia is verifiability, not truth.” Some comments above have digressed from arguing about lift and instead are addressing the way you think an encyclopedia should be written. While the complaints about Wikipedia's approach may well be valid, this talk page on lift is not the appropriate forum. That is all I will say here. Michael Belisle (talk) 07:03, 18 February 2009 (UTC)

I very much like the text as it is now in the article. It is along the lines of J.D. Anderson. So, is it not confusing to refer to Smith, since his wind tunnel analogy is different (using blockage/Venturi effects and mean velocities)? While Anderson points towards the real streamlines and flow field? I do not have the Brandt & Stiles book available, so I cannot comment on that.

Michael, is it possible (and are you willing) to create an image of the flow field around a cambered airfoil (e.g. NACA 4412 or another one) at near-zero angle of attack (mean flow more or less parallel to the flatter parts of the lower surface). Like "Streamlines_around_a_NACA_0012.svg", but now more similar to the streamline figure accompanying Anderson's explanation in his book. -- Crowsnest (talk) 09:01, 18 February 2009 (UTC)

Smith doesn't use the wind tunnel analogy in the same way I did. Attempting to use it to explain the obstruction concept was my own creation and hadn't intended for it to go into the article. Your point about angle of attack is a valid one.

Smith's and Anderson's treatments of this concept are very similar in my evaluation, except that Smith's is a bit simpler for the layman to understand and includes some more general examples.

I can make that figure later this week. Michael Belisle (talk) 19:50, 18 February 2009 (UTC)

= = = =

Michael, I've now read the policy, and realize that you are right. To include this sort of theory is absolutely within the guidelines.

I'm disappointed to see our search for common scientific understanding here end in a resort to policy, and my optimism for a really excellent article coming out of our discussion has soured a little.

Here is why I think that the policy has been shown in this case to be inadequate.

No-one is accountable to respond to questions about a novel theory's scientific or logical validity--least of all, the author of the pop science book (by definition, a book formally unquestioned by scientists), who may well have passed on decades ago.

No-one is obliged by the policy to remove such theories until such time as someone is able to respond to fundamental arguments questioning their validity, such as the observation that the definitions in it are circular, or to questions about exactly what the author might have been trying to articulate in a couple of offhand, somewhat incomprehensible sentences which form the sole citation for the theory.

According to policy, to object on logical grounds to a Talk page interpretation of the theory that says "the flow is determined by first determining the amount of X, and the amount of X is determined by first determining the flow" is unacceptable. The person questioning the theory has merely "personally decided that the theory is a fallacy", but his point has no merit because he has given no "reliable sources for his claim".

Since the circular argument was made in a Talk page, not the originally cited reference, there will never be any "reliable sources for the claim" that it is circular. Talk page comments don't motivate authors to write books simply to refute the comment.

Along the same lines, if the theory itself lacks sufficient gravitas to attract any scientific attention, much less a book countering it, then it is permanently immune from attack; by policy, the burden of literature research is on anyone who questions the theory, and no-one with scientific credentials will likely ever bother to waste time writing a paper refuting an undefined theory in a pop science book, whose author didn't see fit to write a scientific paper about it. (My surmise is that Anderson was probably sleepy when he wrote it, and had a 9:00 AM deadline to hand in the manuscript!).

Instead, I'm now holding my breath, waiting to see the next scientific theory to pop up here.Mark.camp (talk) 19:39, 18 February 2009 (UTC)

You now understand one of the many flaws of the Wikipedia approach:

"If I told you I had a bucket and showed you the bucket, would you ask for a newspaper article to prove I had a bucket?" 210.9.136.63 (talk) 12:53, 23 May 2008 (UTC)

"Yes. Yes, we would. That's the way it works on Wikipedia." FisherQueen (talk)12:57, 23 May 2008 (UTC)

Michael Belisle (talk) 20:05, 18 February 2009 (UTC)

Mark, I sympathise but Wikipedia is not a scientific journal; it is an encyclopedia. In a scientific journal there can be argument and counter-argument in favour, and against, a proposition, and out of that there may be a consensus as to whether the argument or the counter-argument is correct. An encyclopedia does not attempt to present the views of individuals, or to determine which of competing arguments is correct. An encyclopedia impartially presents information that exists and can be verified. If conflicting information exists on a topic, an encyclopedia should present both sides of the conflict.

For example, if there are quotable sources that point to a belief that the earth is flat, Wikipedia should not dismiss that belief. Wikipedia should say something like “the earth is widely accepted to be spherical (suitable citations) although there are published works advocating that the earth is flat (suitable citations).”

My personal view of Equal transit-time theory and Obstruction theory is that they are simple, easy-to-understand models to help newcomers, not rigorous scientific theories, but my personal views have no place in Wikipedia. These theories deserve to be exposed for what they are, but the place for that to happen is a scientific journal or similar medium. Providing these theories are presented in Wikipedia in a dispassionate and non-judgemental way, and are supported by suitable citations, I see no problem with that. That is not to say Lift (force) is the appropriate article. Perhaps there should be an article on Introduction to lift (force) where the various simple, easy-to-understand models can be listed. Dolphin51 (talk) 23:19, 18 February 2009 (UTC)

Suppose your kid says

"Daddy, the Encyclopedia Brittanica says the earth's round, but Wikipedia says it may be flat or round. It explains how the photos from the moon may have been faked by NASA. Which encyclopedia should I believe?"

What would you respond?

Crowsnest: yeah, me too.

Michael: yeah, good answer, you get an "A" in Non-judgmentalism. But wait till you have kids. Flat-earthers tend to get SAT scores way below the medical school cut-off.  :-)

Mark.camp (talk) 22:37, 19 February 2009 (UTC)

“You're asking the wrong question. You shouldn't ‘believe’ either encyclopedia. You should skeptically read the articles, then read the sources and decide for yourself whether or not you want to believe the incontrovertible scientific consensus or the lunatic fringe.“

I don't need an anonymous Wikipedia editor (who may or may not be qualified or properly informed) to judge the quality of a source for me and decide whether or not it's correct. I can do that myself, eliminating one degree of separation between me and the primary source. In this current battle between the eminently qualified and respected JD Anderson and Wikipedian Mark.camp over how best to explain lift to a general audience, I'm siding with JD Anderson. Sorry Mark. It's nothing personal. Michael Belisle (talk) 23:13, 19 February 2009 (UTC)

No offense taken, Michael. It turns out our disagreement was not about how to improve a specific encyclopedia article; it was about what an encyclopedia is. Mark.camp (talk) 12:11, 20 February 2009 (UTC)

Logical relationship between the physical principles

Current text:

"Lift is generated in accordance with the fundamental principles of physics such as Newton's laws of motion, Bernoulli's principle, conservation of mass and the balance of momentum (where the last is the fluid dynamics version of Newton's second law).[2] Each of these principles can be used to explain lift on an airfoil.[3] As a result, there are numerous different explanations with different levels of rigour and complexity. For example, there is an explanation based on Newton’s laws of motion; and an explanation based on Bernoulli’s principle. Neither of these explanations is incorrect, but each appeals to a different audience.[4]"

The sentence

"Each of these principles can be used to explain lift on an airfoil."

implies that these principles are a partial list of alternative principles, each of which can be used independently to explain lift.

The remainder of the paragraph amplifies this impression.

I think it can be reworded to show how these principles relate logically to each other in explaining lift.

I will give it some thought. Mark.camp (talk) 12:56, 21 February 2009 (UTC)

A very good point! Some remarks (just my opinion):

The fundamental principles of physics required are: Newton's laws of motion, conservation of mass and a force (pressure) model. The last one is missing at the moment. Note that Newton's 2nd law relates the net force on a fluid parcel to the rate of momentum change. But it does not describe the force itself (e.g. in gravitation you also need Newton's law of gravity or general relativity theory to describe the force). This force/pressure model can have several forms:

1. At low Mach numbers the incompressibility constraint (isochoric process) is a good approximation, i.e. the pressure constraints the flow velocity field to be divergence free.
2. More generally: an equation of state relating pressure, mass density and eventually energy; if also dependent on energy than an energy evolution/conservation equation is needed. And in chemically reacting flows the chemical processes need to be modelled.
3. For inviscid flows only the pressure has to be related to the flow. In case of a viscous flow also the (turbuhlence) shear stresses have to be modelled (e.g as in the Navier–Stokes equations; turbulence models).

Further, the balance of momentum is just Newton's second law. While Bernoulli's principle results from integration of Newton's 2nd law along a streamline; under the additional assumption (which needs to be justified) that irreversible processes (e.g. dissipation through shocks and viscosity, acoustic radiation, heat radiation) can be neglected.

Under the assumptions of irrotational and incompressible flow, potential flow theory (Laplace's equation) applies. Then dynamics only enter the model through the Kutta condition and Bernoulli's equation (to compute the lift force).

Good luck with the rephrasing/improvements! -- Crowsnest (talk) 16:26, 21 February 2009 (UTC)

It is just my opinion, but I agree with everything you said about how these principles are logically related. It remains to write the improved text.

If the text agrees with what you've written, then I am confident that it will be (a) verifiably correct, which is critical to one faction in this debate, and permissible according to the other.

It remains to make it (b) easy to understand and (c) to include all required citations, which conditions are critical to both factions.

(b) is a challenge. One cannot reduce the explanation to one or two sentences and still include all of the relevant facts, and all the logical development of the explanation.

Even simplifying the model drastically doesn't make it a three sentence explanation. Leaving out time-dependent flow (and thus turbulence, and the true rationale for cambered foils), the third dimension (and thus, understanding of induced drag and trailing and starting vortices); friction (Navier-Stokes, laminar flow, boundary layer, separation of boundary layer, and the theoretical justification for applying Kutta condition to potential flow model), and compressibility (supersonic flow), it still takes a progression of several simple explanatory stages, each one building on the previous.

(Einstein, on his American tour, was pestered by every reporter at every stop to explain, in one sentence, his general theory of relativity. He complained privately, 'I have spent years trying to explain it in a single book, and they want me to explain it in one sentence!') (Citation needed. The source book is sitting less than ten meters away from me, but I am very lazy about citations, as Michael B. has noted above more than once.) Mark.camp (talk) 03:52, 22 February 2009 (UTC) Superfluous sentences removed. Mark.camp (talk) 13:26, 22 February 2009 (UTC)

Hello Mark. The nice thing is, that you do not need to explain the whole story in four sentences. The four sentences, as they are, are confusing and incorrect; as pointed out by you. They need to be replaced by some other introductory remarks. But which impossibly can contain the whole story, or even point to all aspects involved. Fortunately there is plenty of room for expanding elsewhere at Wikipedia, either in this article and in others (existing ones or not yet created; often a wikilink may suffice). -- Crowsnest (talk) 08:24, 22 February 2009 (UTC)

Good point. I've started working on it. Mark.camp (talk) 13:26, 22 February 2009 (UTC)

A summary could start something like the following

Still needs a lot of work. But the critical break with the non-scientific explanations is already made. The premise of the article will be that (a) every reader understands Newton's laws--'if you push on something its speed in that direction increases'--to be common sense, and (b) the reader will need nothing more than that scientifically proven, concrete mental picture to understand everything that follows in the summary.

"By using a simplified model, and three reasonable physical assumptions mentioned below, one can partially understand lift by applying Newton's second law

F=mA

to find the flow "solution": the velocity and static pressure at every point. It's discovered that the wing disturbs the flow in such a way that the pressure on the top surface is reduced significantly below ambient pressure. The pressure on the bottom is generally increased slightly. By summing (or integrating) the vertical component of these pressures, one finds that the net downward force of the air on the upper surface is less than the upward force from below. The difference in these two forces is the lift.

"Newton's second law requires a force law. This is supplied by the Bernoulli principle, which is itself derivable directly from Newton's second law and conservation of mass.

"Unfortunately, just as in the case of a particle subjected to a force field, Newton's equation allows an infinite number of solutions for the velocity and pressure fields, and an infinite range of values for the lift. An explanation of lift must provide a unique value.

"For a particle, one finds the unique solution for velocity by specifying an initial value, giving a physical explanation for the value chosen. Similarly, for the simplified lift model, one does it by making the three assumptions mentioned above, and providing a physical explanation of each.

"First, one assumes that the confluence of the upper and lower flowstreams is at the sharp trailing edge. The physical explanation of why this condition, the Kutta condition, occurs is explained later in the article.

"The other two assumptions may be more intuitively obvious.

(a) Since the wing can't create or absorb air, the flow at the wing's surface must be zero or tangent everywhere.

(b) At an infinite distance in front of the wing, the speed of the air is constant and horizontal.

Mark.camp (talk) 13:26, 22 February 2009 (UTC)

If you have references for the claims in this introduction, please provide them. I think the explanation would benefit from citations: your phrasing is unfamiliar to me and so I can't tell if it's wrong, misinterpreted, or correct but written in an manner unfamiliar to me. My initial comments are

1. Newton's equation (here $F=ma$), which for a fluid continuum becomes the Navier-Stokes equations, doesn't admit an infinite number of solutions. The potential flow model where viscosity is neglected allows an infinite number of solutions, which is why the Kutta condition is necessary to simulate the effects of viscosity. If you include viscosity, there is only one solution and it's not necessary to impose the Kutta condition explicitly.
2. $F=ma$ can't be directly applied to a fluid continuum, as it's the statement of Newton's second law for a single particle. You do need more than $F=ma$ to find the velocity and pressure at every point. A lot more.
3. No-penetration says that the flow must be tangent. No-slip says the flow must be zero at the surface. Whether or not either applies depends on the assumptions, so you should state these as assumptions, not as inherent limitations (i.e. say something like "The wing can't create or absorb air, so an initial assumption is that the flow direction at the wing's surface must be tangent to the surface. Additionally, since the flow is viscous and the no-slip condition applies, the flow velocity must be zero.")
4. "Newton's second law requires a force law." I'm not sure what this sentence means.

Michael Belisle (talk) 18:30, 23 February 2009 (UTC)

Thanks, Michael I will have a try at cleaning up the things you mention. I have no citations yet, still just working on the content. Mark.camp (talk) 19:29, 23 February 2009 (UTC)

On 23 Feb MB wrote:

1. Newton's equation (here $F=ma$), which for a fluid continuum becomes the Navier-Stokes equations, doesn't admit an infinite number of solutions.The potential flow model where viscosity is neglected allows an infinite number of solutions, which is why the Kutta condition is necessary to simulate the effects of viscosity.

Absolutely agree. I will address this issue by making it clearer what the 'simplified model' is meant.

"By using a simplified model called "potential flow theory", and...

Mark.camp (talk) 21:30, 23 February 2009 (UTC)

On 23 Feb MB wrote:

F = ma can't be directly applied to a fluid continuum, as it's the statement of Newton's second law for a single particle. You do need more than F = ma to find the velocity and pressure at every point. A lot more.

Absolutely agree. Some text which makes this transition--in an intuitive, non-mathematical way--from the familiar form of Newton's second law, for particles, to the continuum, is desparately needed. Have been thinking of how to do this succinctly. For the time being, it is an open issue.

To me, this connection was the single most difficult thing to grasp about lift; I worked on it in my mind for forty years before I understood it even a little bit. Much of this is my own fault, I admit. I would never accept a mathematical explanation without an intuitive understanding of the physics, nor an easy-to-follow explanation without being convinced it was backed by rigorous Newtonian math.

One reason for my stubbornness is that I have never seen anything in mathematical physics or engineering which could not be explained in an intuitive way to a curious layman. (With the obvious exception of quantum physics, where if you think you understand it, you really need to go back and read it again.) Nor a common sense physical fact--"ain't no water coming out of that pipe except what already gone in"--which cannot be obscured by the language of math ("the divergence of the velocity field of the water is zero")

We've all seen lots of formal mathematical proofs of this bit or that bit--conformal mapping, the law of constant circulation, the vortex line must form a closed loop, etc. And lots of easy-to-follow explanations. We just need to get both in one place and then find some citations to wrap it up in, so's won't nobody erase it. Mark.camp (talk) 22:45, 23 February 2009 (UTC)

Foil (fluid mechanics)

Just when I thought it was safe to go back in the water, because we have eliminated from Lift (force) most of the pseudoscience, original research and whacky alternative explanations, I discovered the existence of Foil (fluid mechanics)! It contains a section called Physics of foils, and it is just about all pseudoscience. I have flagged this section with a Dispute tag, and dissected its contents on the Talk page. I would appreciate other Lifters having a look at it. The best thing might be to merge Foil (fluid mechanics) and Airfoil. — Dolphin51 (talk) 02:48, 25 February 2009 (UTC)

I added merge tags to both articles, the discussion on the merge is at Talk:Airfoil#Merger of Foil (fluid mechanics) into Airfoil. -- Crowsnest (talk) 07:18, 25 February 2009 (UTC)

Upwash

The wording of the current text gives impression that upwash is specifically related to the upper streamtube adjacent to the stagnation line:

"The upper stream tube constricts as it flows up and around the airfoil, the so-called upwash

My understanding has been that upwash refers to the upward acceleration in front of the air in general, and not specifically to the region mentioned. If this is correct, could someone with expertise and access to references please re-word it? Thanks. Mark.camp (talk) 19:22, 25 February 2009 (UTC)

I changed it to ""The upper stream tube constricts as it flows up and around the airfoil, a part of the so-called upwash". See image at [1]. Michael Belisle (talk) 19:53, 25 February 2009 (UTC)

Seems like it might be nice to include the fact that the upper stream tube constricts because it's speed increases as it acclerates towards the lower pressure zone above the wing and since mass flow with in a stream tube is constant, the tube has to constrict as speed increases. Currently there is no explanation for why it constricts. Jeffareid (talk) 12:38, 26 February 2009 (UTC)

On a related issue, it would be nice to see a diagram of streamlines (tubes) that use the air as a frame of reference as opposed to the wing. These would be nearly vertical lines. Jeffareid (talk) 12:38, 26 February 2009 (UTC)

Velocity vectors in a frame of reference moving with the airfoil		Velocity vectors in a frame of reference moving with the undisturbed air in front of the airfoil

-- Crowsnest (talk) 16:52, 17 March 2009 (UTC)

I disagree that there would be any near-vertical lines. If we use the reference frame in which the velocity of the atmosphere is zero, any particle of air that is influenced by the passage of a wing will execute a closed circuit that is elliptical in shape. As the wing approaches a particle from the right, the particle first moves a little to the left, then up or down as the wing is passing, then down or up, then a little to the right, causing the particle to finish where it started. Particles that enter the boundary layer of the wing will end up with a residual velocity so they don't finish where they started.

The near-vertical lines Jeffareid has in mind may be those often shown around a 2-D airfoil to show how the flow cleaves in two so that the upper lines and lower lines never meet again after the airfoil has passed. See Airfoils and airflow. These lines aren't streamlines, and they are shown in a reference frame in which the velocity of the airfoil is zero. Dolphin51 (talk) 23:27, 26 February 2009 (UTC)

causing the particle to finish where it started. - This conflicts with the fact that the particles that the wing passes under end up displaced from the particles that the wing passes over. The interaction between a wing and air results in the net downwards and somewhat forwards acceleration and flow of air. The downwards flow from a wing would be similar to what is called the exit velocity from prop wash propeller_analysis.htm. The affected particles never return to their former position, circulation will result in other particles filling in the positions of the original particles. Jeffareid (talk) 13:42, 1 March 2009 (UTC)

The particles do not end at the same position as where they start, after the passage of the wing. This effect is called Darwin drift, after Charles Galton Darwin. -- Crowsnest (talk) 08:09, 2 April 2009 (UTC)

Velocity vectors in a frame of reference moving with the undisturbed air in front of the airfoil. Take a look at the first photo (Cessna) and adjacent diagram from here: airplane_wings.htm. A series of pitot static tubes were mounted on a pole about 1/2 foot in front of the wing, vertically spaced 1 foot apart. There is significant difference in horizontal velocity above (faster than aircraft) and below (slower than aircraft) due to the lower pressure zone above and higher pressure zone below a wing. Going back to the air as a frame of reference, it would seem that air would accelerate towards the lower pressure zone above a wing and away from the high pressure zone, in all directions (except through the solid wing), and I would expect a significant downwards flow of air at some reasonable distance away from the wing, both above and below, which is what I meant by near vertical lines of flow. Jeffareid (talk) 05:50, 25 March 2009 (UTC)

See the pictures above. Streamlines are everywhere tangent (parallel) to the velocity vectors. -- Crowsnest (talk) 08:09, 2 April 2009 (UTC)

I recall seeing a diagram of streamlines relative to the air, from about 2/3'rd wingpan away from the wing. In that diagram, the air flow was downwards and backwards in front of the wing, and mostly downwards and a bit forwards behind the wing. This diagram didn't really show the circulation near the wing, but it gave a better idea of air flow further away from a wing, which is what I was looking for. Unfortunately, I've looked but can't find that diagram. Jeffareid (talk) 18:25, 7 April 2009 (UTC)

Jeffareid, the diagram just above, which Crowsnest is referring to, shows you kind of what the path of each parcel of air would be. You have to mentally sketch the path in, based on the fact that the path is everywhere tangent (parallel) to the velocity vectors shown in his diagram. This is what you mean by "streamlines", I think.

But you must realize that there really are no "streamlines" in the frame of reference you are talking about. In the wing's frame, every parcel follows the same path as its upstream and downstream neighbors, and these paths, which NEVER CROSS each other, are streamlines. In the frame of reference you want, which is the one Crowsnest graciously provided a diagram of, (that of the still air far forward of the wing) the paths of neighboring parcels cross each other, so they aren't really streamlines.

Mark.camp (talk) 21:56, 7 April 2009 (UTC)

I agree with J. that a diagram of the directions of motion relative to earth at a point in time, in addition to the kind shown, would be helpful for understanding.

J., are you proposing a diagram based on

a. the simplified model of potential flow theory,

-- irrotational

-- inviscid

-- 2D (=infinite 3D wing of constant section)

-- steady

-- Kutta condition imposed, or

b. 3D, viscous, steady flow (no need to impose Kutta condition), or

c. 3D, viscous, unsteady flow

?

A diagram based on the real 3d flow of air at some reasonable sub-sonic speed, not some theoretical model. Jeffareid (talk) 13:42, 1 March 2009 (UTC)

Note that the ubiquitous streamlines diagram, which is given in the current article, is a time-independent diagram, because in wing frame the flow is time-independent (if the simplifying assumption of steady flow is made, of course). In earth frame, the flow is time-dependent, so the less commonly presented diagram J. proposes would be a snapshot of the flow at an arbitrary point in time. Mark.camp (talk) 15:05, 28 February 2009 (UTC)

See for instance Anderson & Eberhardt "A Physical Description of Flight; Revisited". p. 5. -- Crowsnest (talk) 16:43, 5 March 2009 (UTC)

Interesting apparent contradiction in this book.

"For straight and level flight this momentum eventually strikes the earth."

A few pages later is a picture of the trailing vortex, which shows that, in fact, most of the air aft of the wing which is moving down, at one point, will soon be moving horizontally, and then upward, without ever striking earth. This change from downward to upward motion would happen even if there were no earth below, because it isn't caused by the earth. It is caused by the circulation generated by the wing. This spinning motion of the air the air behind the wing is actually its "natural state of motion", in the sense that it is the motion which would occur if there were no friction or externally applied forces, only the radial pressure gradient and the resulting centripetal acceleration, which is given by F=mA applied to each differential volume of air. Mark.camp (talk) 19:55, 5 March 2009 (UTC)

Sentence still needs tweaking. It says that something is "..a part of the so-called upwash." Exactly what is that "something"? Mark.camp (talk) 16:31, 27 February 2009 (UTC)

Jeffareid, above, you wrote (emphasis mine):

"There is significant difference in horizontal velocity above (faster than aircraft) and below (slower than aircraft) due to the lower pressure zone above and higher pressure zone below a wing. Going back to the air as a frame of reference, it would seem that air would accelerate towards the lower pressure zone above a wing and away from the high pressure zone, in all directions (except through the solid wing)...25 March 2009 (UTC)

I don't this idea should be added to the article, because I don't think it is factual.

In fact, the acceleration vector of a given parcel of air is simply the net force vector on the parcel*. This force vector is the frictional force vector at that point (determined by the velocity 'shear' at that point) plus the negative of the pressure gradient vector at that point*

(*Suitable units of mass, force, and distance being chosen, to allow a simpler sentence and to avoid the need for equations and a diagram with the size of the parcel shown. I use the more familiar form of Newton's law, with forces, objects, and masses, rather than the customary pressures, volumes, and densities, for the same reason.)

Thus, the acceleration at a point is entirely determined by the local velocities (for frictional force) and local pressure gradient at that point. The pressure in a distant region, say above the wing, is meaningless, and the average pressure in that region is irrelevant.

When you say that "air would accelerate towards the lower pressure zone above a wing and away from the high pressure zone, in all directions (except through the solid wing)" you are making a statement that simply isn't true or even meaningful in general.

If you look at a pressure contour map, the direction of the pressure gradient at a point is perpendicular to the contour line at that point. The magnitude is higher the closer the lines are spaced--just like the grade of a hill is determined by how close the elevation lines are on a topo map. Now apply what I said about the acceleration being the essentially the negative of the pressure gradient, you will see that some air is accelerating toward the low pressure zone above the wing, but some air is not. Some air is even accelerating away from that zone.

Mark.camp (talk) 11:58, 27 March 2009 (UTC)

"Most often"

The edit summary for reinserting "most often" reads

re-insert "most often": when the streamline passes through a shock discontinuity, there is dissipation in the shock (i.e. the Bernoulli eq. constant makes a jump

while the text in the article reads

Most often, the Euler equations for inviscid flow can be integrated along a streamline, resulting in Bernoulli's equation.

In this context I read "most often" as though it's saying that the most common thing done with the Euler equations is to integrate them along a streamline. That's not necessarily true, which is why I took it out. The edit summary suggests it's meant to say something else, but I don't understand what it's trying to say or how "most often" has anything to do with shocks. Crowsnest, would you please clarify? Michael Belisle (talk) 04:48, 26 February 2009 (UTC)

Michael, I was confused with hydraulic jumps, where only mass and momentum are conserved through the jump (shock). In the compressible Euler equations mass, momentum and energy are conserved through the Rankine-Hugoniot conditions. So there is no dissipation in the shock, and Bernoulli's equation still applies (see e.g. Roskam, Jan; Lan, Chuan-Tau Edward (1997), Airplane Aerodynamics and Performance (3th illustrated ed.), DARcorporation, ISBN 1884885446, p. 20). Entropy however has a jump across the shock. -- Crowsnest (talk) 07:50, 26 February 2009 (UTC)

Lift on circular cylinder in a free stream

Although it's an interesting topic, I think the mention of the lift generated by flow past a circular cylinder is a bit specific or esoteric for the introduction. I think the intro should have a more general description of lift on bodies other than airfoil without going into that much detail. There should be another section added on lift on bodies other than airfoils (which would include a summary of related topics like the Magnus effect). This would do well as part of that section. For now, we can leave it where it is. Michael Belisle (talk) 19:59, 3 March 2009 (UTC)

I agree. The following text does not belong in the introduction.

Also the flow around bluff bodies may generate lift, besides a strong drag force. For instance, the flow around a circular cylinder generates a Kármán vortex street: vortices being shed in an alternating fashion from each side of the cylinder. The oscillatory nature of the flow is reflected in the fluctuating lift force on the cylinder, whereas the mean lift force is negligible. The lift force frequency is characterised by the dimensionless Strouhal number, which depends (among others) on the Reynolds number of the flow. For a flexible structure, this oscillatory lift force may induce vortex-induced vibrations — under resonance conditions the resulting motions of the structure due to the lift fluctuations may be strongly enhanced.

Dolphin51 (talk) 21:54, 3 March 2009 (UTC)

I moved the text to a separate section. And made a shorter intro in the lead. Still a bit longer than I would like. So feel free to improve. -- Crowsnest (talk) 22:04, 3 March 2009 (UTC)

What do people think of the following as a development of Crowsnest's new paragraph?
Non-streamlined objects such as bluff bodies and flat plates may also generate lift when moving through a fluid. This lift may be steady or its magnitude and direction may alternate, causing the object to vibrate. Non-streamlined objects are inefficient generators of lift because they usually experience significantly more drag than lift. Dolphin51 (talk) 01:19, 4 March 2009 (UTC)

Michael Belisle (talk) 03:28, 4 March 2009 (UTC)

Lift and the deflection of the flow

I added a sub-section of this name. I am not sure whether my interpretation of Langewiesche is correct. Input and comment from others is appreciated. -- Crowsnest (talk) 23:58, 5 March 2009 (UTC)

Current text

However, some go as far as stating that one can explain lift along this line of reasoning. For instance Wolfgang Langewiesche states: "… the wing keeps the airplane up by pushing the air down."[31] This, however, is a misinterpretation of Newton's laws.

I think that the statement of Langewiesche (with my emphasis) is fine as an interpretation of Newtons' laws. As it goes through the air, the wing pushes the air down. Therefore, the air pushes the wing up. Mark.camp (talk) 00:25, 6 March 2009 (UTC)

Thanks Mark. I changed the text accordingly. -- Crowsnest (talk) 00:28, 6 March 2009 (UTC)

I want to focus on water waves, and some other aspects of fluid dynamics, so I removed this page from my watchlist. Good luck with improving and extending the article. -- Crowsnest (talk) 11:06, 10 March 2009 (UTC)

Momentum changes - a difficult explanation to express

For a long time I found the explanation of lift in terms of momentum change to be very easy to understand. I could explain it with confidence to anybody, with statements that sounded intelligible. But if anyone had stopped me the first time I said "the momentum of the air" and asked me what that term means, I would have stared and stammered and then finally turned away muttering to myself about how rude people have gotten.

An intelligible explanation could be given, but I've never seen it done.

From one perspective, the problem is this. Newton's law not withstanding, it is not true that

"the force on a parcel of air equals the time derivative of its momentum"

if the parcel is in contact with the wing (which does result in a mutual repulsive force) and the "force" being considered is the force of the wing on the parcel. Mark.camp (talk) 19:40, 6 March 2009 (UTC)

Where do you read this statement in the article? Besides, it is true. But the force (net force actually) is the sum of the force of the wing on the parcel and the force by the surrounding fluid on the parcel. -- Crowsnest (talk) 19:53, 6 March 2009 (UTC)

Momentum is perhaps easier to grasp using a real-life example: someone throws two two balls towards you, the first one has a mass of one kilo, and the second is five kilo. Both balls are thrown at the same velocity towards you, and you have to catch them. Which one is more difficult to stop? -- Crowsnest (talk) 20:08, 6 March 2009 (UTC)

Ah, sorry, it wasn't meant to be a quote from the article per se.

You said "it [the statement] is true". I disagree.

You said "the net force is the sum of the force of the wing on the parcel and the force by the surrounding fluid."

I agree. In fact, I believe that this proves that the first statement is false.

If "the air" were a rigid body, then one could speak of "the momentum of the air" as if it were a 5 kilo ball, and it would be an understandable phrase. But the air is a continuum. Mark.camp (talk) 02:03, 7 March 2009 (UTC)

Because it is a continuum, one normally uses momentum density, which is momentum per unit volume. Like for air one uses mass density, instead of mass itself.

If you talk about "the momentum of the air", the natural question arises: which air? In general, it obviously cannot be all air in the whole world. But if you take at some moment a certain amount of air in a finite volume, mark the air parcels within, and follow those air parcels (while the position and shape of that volume changes due to the flow), it is true again.

It is important to note this is all from a Lagrangian perpective, because the time derivative you are talking about is in that frame of reference. -- Crowsnest (talk) 07:50, 7 March 2009 (UTC)

So I think you are saying that the lift force on the wing is the time derivative of the downward momentum of "a certain amount of air in a finite volume".

In that case, my question is essentially the same: "which finite volume?" I think that's a fair question.

I think maybe you do mean "all the air in the whole world". That's fine, I would understand that. All I am saying is that, if this is what one means, he should state it explicitly, because otherwise it is not clear. Mark.camp (talk) 21:59, 13 March 2009 (UTC)

Re-work of introduction

The introductory paragraph of Lift (force) is less than ideal so I am proposing a re-worked version. You can see it on my personal sandbox. Changes relative to the current version can be determined by using the History page to see the last set of changes. (I am yet to find an ideal citation for my comment about non-streamlined objects are inefficient generators of lift so I have inserted a {{fact}} tag instead.) Please leave any comments on the sandbox talk page. All comments will be appreciated. Dolphin51 (talk) 02:31, 20 March 2009 (UTC)

Excellent work, Dolphin51. Tweaked one sentence.

Mark.camp (talk) 00:10, 28 March 2009 (UTC)

Most of the text on my personal sandbox was transferred to the introductory paragraphs of Lift (force) on 31 March 2009. There was a minor problem with the Wiktionary link obscuring some of the text so I re-positioned the link above the image of the A380. I will shortly erase my Sandbox and create text for my next project. Dolphin51 (talk) 02:02, 1 April 2009 (UTC)

Funny. Now Dolphin51 interchanged the wiktionary link and the image, some text is partly obscured in my browser (Firefox 3.0.8, 1400x1050 display). While it was OK before. -- Crowsnest (talk) 00:09, 1 April 2009 (UTC)

Opening sentence

Hi Crowsnest. The opening sentence of Lift (force) begins In the context of … Use of the word context appears to be rare in Wikipedia. (I am not actually aware of any other article that uses the word in this way.) It is not a word in common useage so I am concerned that young readers, and readers who are new to the English language, may be deterred from proceeding further because they are unfamiliar with the meaning of context.

In Wikipedia, it is much more common for an article to begin by specifying the field or subject in which the article is relevant. That is why my proposed text on my sandbox begins In aeronautics and fluid dynamics … …

I am keen to repair the opening sentence to eliminate the word context. What do you think of the following?

In aeronautics and fluid dynamics the lift force is the component of the aerodynamic force that is perpendicular to the oncoming flow direction.

Dolphin51 (talk) 05:39, 2 April 2009 (UTC)

The phrase was introduced by Michael Belisle in this edit. To me it does not sound odd, but I am not a native English-speaking person. Agitated depression also starts this way, but I agree it is rarely used at the start of the first sentence in a WP article. The important elements in "In the context of a fluid flow relative to a body, ..." are: "fluid flow" (which is a wikilink redirecting to the "Fluid dynamics" article) stating the force is by a fluid flow and occurs in fluid dynamics, and "body" stating the subject of the aerodynamic force mentioned later on. The "Aerodynamic force" article has a very strong depiction of the phenomenon right from the start (due to your edits there).

Some of these strong elements are lost in your present proposition.

As said in the talk page of your sandbox, both using "aeronautics" and "fluid dynamics" sounds double, since they are so closely related: aerodynamics is the application of fluid dynamics in aeronautics. If necessary, personally I would only use one, and then my preference would be "fluid dynamics" (since lift also occurs in other fields than "aeronautics"). Further the main subject here is lift force, so primarily it is dynamics (either fluid dynamics or aerodynamics).

Why not simply replace "In the context ..." with just "For ..." or "In..."? So: "For fluid flow relative to a body, ...". Then the article has an good, interesting and not-so-standard start, retaining the strong elements introduced by Michael. While the terms and wikilinks "fluid flow" and "aerodynamic force" set the context. -- Crowsnest (talk) 07:41, 2 April 2009 (UTC)

Hi guys, I could have sworn I wrote "context". I do talk and write like that. I checked my "contributions" though and it looks like I haven't. I even wrote a bit on this page taking "credit" which I had to "undo" (Doh!). Anyway, with regard to "context", maybe the writer was trying to get away from the idea of necessarily assigning strict subject areas. Or, maybe a fancy "blue" subject is as daunting to a timid reader as an unusual word like "context". I don't know. I do think "In aerodynamics and fluid mechanics" is excellent though.

-- Gummer85 (talk) 09:14, 3 April 2009 (UTC)

Agree that the sentence works just as well without "...context". I removed it, and reworked the sentence accordingly.

Mark.camp (talk) 16:28, 3 April 2009 (UTC)

Actually, the more I think about it, the more I like the idea of not being so tied down to a rigid form. I think it is good communication to first give a big picture (the "context" or subject area), then refine it (sort of like the inverted pyramid in newspaper writing). However, I'm not so sure it always has to have the form:

"In some Wikipedia article about a subject area, X is...".

It makes me want to go look in some printed encyclopedias and see how they start up things and what their range of acceptable forms are.

I learned about "context clues" in 3rd grade reading class. Maybe I'm exceptional in that I've embraced the word, but I know that every American has learned it. "Context" is indeed colloquial where I come from, slightly high colloquial, but colloquial. It's not too unusual to use it to refer to a "situation" or "subject area" even if a few might need to use "context clues" to remind themselves of the meaning.

--Gummer85 (talk) 18:51, 3 April 2009 (UTC)

The article now begins

"This article is about lift force in fluid dynamics. For other uses, see lift."

which sets the context.

The next sentence is

"A fluid flowing relative to a body exerts a force on it."

which gives the context for the next sentence

"Lift is the component of this force which is perpendicular to the oncoming flow direction."

which defines Lift in the context of fluid dynamics.

So, the context pyramid of which you speak is present, even though the term "context" is no longer there.

Mark.camp (talk) 21:13, 3 April 2009 (UTC)

I like it!  :-)

--Gummer85 (talk) 21:27, 3 April 2009 (UTC)

Lift and drag are the two components of aerodynamic force. In Mark's edit, the words aerodynamic force have disappeared. This is not really satisfactory — if the article is talking about a component of X, the article must identify what X is! Dolphin51 (talk) 01:06, 4 April 2009 (UTC)

The "this force" is pretty clear to me to be the fluid-dynamic force from the previous sentence. Although, maybe it would be an improvement to demonstrate (by way of context(!)) that the force is called a fluid-dynamic force. For example:

"A fluid flowing relative to a body exerts a fluid-dynamic force on it." ...

or

"A fluid flowing relative to a body exerts a force on it. This force is a fluid-dynamic force." ...

In the first example, knowledge (that it is a fluid-dynamic force, etc.) is imparted by use of a term in context. In the second example, some of that knowledge is stated explicitly. Rules of simple and clear writing (taught to me by the Air Force) say to break up things into smaller explicit bites like in the second example. But, I like the first way in this case because it is shorter while still being clear.

The problem of "fluid-dynamic" vs. "aerodynamic" is ignored in both examples, permitting the reader to infer their relationship by the fact that the link leads to aerodynamic force. To explain this relationship explicitly, especially so early, would be digression. The aerodynamic force article does a better and more direct job of talking about fluid-dynamic forces than the fluid dynamics article. Its problem is that it is not about general fluid-dynamic forces. Waddayagonna do? We can't fix that problem easily! I've noticed that kind of muddling crops up everywhere in "Aviation-Aero-Fluids". I guess we just have to believe that people are smart enough to know that air is a gas, and that a gas is a fluid, etc. Although, I like to be correct when I can by using "fluid-dynamic" in this case.  :-)

--Gummer85 (talk) 17:16, 4 April 2009 (UTC)

Added terms "fluid-dynamic" and "aerodynamic" to opening paragraph. Would someone do me a favor, and convert them to references?

But Gummer, I should note in passing that you are right, the referent of "this force" was already perfectly clear.

BTW, are we not indenting responses here? It's ok with me, just so we are consistent with the style guide. I admit that indenting responses can get to be messy when we have extended chats.

--Mark.camp (talk) 18:02, 5 April 2009 (UTC)

I don't know, or rather didn't know about style guidance for discussion. I know now to look it up. I've generally been using visual judgment on indenting in discussions and copying previous patterns. Too deep can be messy I agree. Also, and I know this is silly, but now that I think about it, indenting seems subservient! (Bah! Ha! Ha!) Walk tall! Don't indent! (Ha! Ha! Ha! Ha! Ha!) Although seriously, I think indenting is good for a fairly short answer that contains only a response to only the previous "thing". Otherwise, it's okay to promote it to the top IMHO. A different indentation from the previous (and possibly following) is useful too to help the eye distinguish it from other entries. Then people can scan pretty easily to see how the subject(s) fits in I think.

Hey, I noticed a mangled sentence at the end of the first paragraph. I couldn't figure out its intent. I'll look in the histories and see if I can fix it. If I can't, I think you (Mark) might be able to.

--Gummer85 (talk) 22:44, 5 April 2009 (UTC)

I see Dolphin51 took care of the stray sentence a few minutes ago --Gummer85 (talk) 22:57, 5 April 2009 (UTC)

The second paragraph of Drag (physics) contains a sentence saying:

• For a solid object moving through a fluid, the drag is the component of the net aerodynamic or hydrodynamic force acting opposite to the direction of the movement.

Taking this as our guide, the opening sentence of Lift (force) should say:

• For a solid object moving through a fluid, the lift is the component of the net aerodynamic or hydrodynamic force acting perpendicular to the direction of the movement.

What do people think of that? Dolphin51 (talk) 05:24, 7 April 2009 (UTC)

Crowsnest has refined the opening sentence to say:
A fluid flowing relative to a body exerts a force on it, called an aerodynamic force if the fluid is air.
I think this is satisfactory and we can turn our attention to other parts of the article that are in need of improvement. Dolphin51 (talk) 02:17, 9 April 2009 (UTC)

Agree.

Mark.camp (talk) 16:14, 9 April 2009 (UTC)

---