Newton's third law to explain lift

[Stan Butchart]

Re:downflow ( the following is contingent upon Cyrus not shooting me down!)
Newton only requires a ballance of forces for lift. The only directionality in fluid pressure is in the orientation of surfaces exposed to it.
We look at flow from the standpoint of the wing for purposes of calculation. In almost all cases what is happening must be seen from the remote still air. The basic displacement flow is essentually circular. The circular path of circulation directs most of the displaced air over the top. The flow following the top surface travels a 360deg path with a forward displacement.
The Bernoulli energy equation is from relative tangential acceleration between particles. Weltner and Graig prefer centrifugal pressure from the particles in normal acceleration. (Personnaly I could not get big enough numbers.) Both of these accelerations occur within the curving flow.
In the Bernoulli case all of the pressure reduction would be created where the flow has an upward component. Where the flow has a downward component the reduced pressure is being "destroyed", if you will. In the centrifugal case, half of the pressure reduction would occur with upwards and half with the downward components.
High pressure air cannot flow to low pressure areas unless the low pressure air has somewhere to go. If separation is not present, the air at the trailing edge will have a forward vector.
This represents my quandry.

 

[Cyrus]

Re:downflow ( the following is contingent upon Cyrus not shooting me down!)
Newton only requires a ballance of forces for lift. The only directionality in fluid pressure is in the orientation of surfaces exposed to it.
We look at flow from the standpoint of the wing for purposes of calculation. In almost all cases what is happening must be seen from the remote still air.

I don't understand this, and I don't agree with it if your saying what I think your saying.

The basic displacement flow is essentually circular. The circular path of circulation directs most of the displaced air over the top. The flow following the top surface travels a 360deg path with a forward displacement.

Again, I don't follow you. What do you mean circular?

The Bernoulli energy equation is from relative tangential acceleration between particles. Weltner and Graig prefer centrifugal pressure from the particles in normal acceleration. (Personnaly I could not get big enough numbers.) Both of these accelerations occur within the curving flow.

Again, I don't follow.

In the Bernoulli case all of the pressure reduction would be created where the flow has an upward component. Where the flow has a downward component the reduced pressure is being "destroyed", if you will. In the centrifugal case, half of the pressure reduction would occur with upwards and half with the downward components.
High pressure air cannot flow to low pressure areas unless the low pressure air has somewhere to go. If separation is not present, the air at the trailing edge will have a forward vector.
This represents my quandry.

Still not following you...

 

[Stan Butchart]

"In almost all cases, <to understand> what is happening it must be seen from the remote still air." Poor sentence. Pressure change comes from the relationship of air to air, not air to surface.

"What do you mean circular? " Let's take the 2d cylider as a model. We look at it from the remote still air in an ideal fluid so that separation is not present. The instantainious source/sink picture is made up of pure circles. As the cylinder moves forward thr air in front is accelerated forward. The air following the surface traces a cursive "e" path through space ,top and bottom. This is a full 360deg path with a cosiderable forward displacement.

"Both of these accelerations occur within the curving flow." Pressure changes occur within accelerations. Tangential relative accelerations occur because each air particle has its own individual flowpath. The normal accelerations are the "centrifical force" in the curving flow. I stick with tangential but you can pick your own poison.

"In the Bernoulli case all of the pressure reduction would be created where the flow has an upward co..." One half of wing lift occurs on the front half to third of the wing. Except for under the L.E. allof the flow here has an upwards component. This extends out for many,many chord lengths. Aft of here, where the flow all has a downward component, pressure is constantly increasing.

 

[Cyrus]

Im sorry, I still can't understand what your trying to say. Can you make it clearer please.

If your talking about flow circulation, that's a mathematical construct. The air is NOT doing circles in the actual flow field. All the air is deflected downstream and down to creat lift on the airfoil.

 

[Stan Butchart]

What I am trying to say is that I cannot find a dedicated downflow that is a reaction to lift force. This is not argument, I simply do not see it. The required pressures are produced within curving flows where the verticle components of acceleration are equally up and down and the horizontal components contribute as much as the verticle. Unless one doubts lifttig line the downflow of the vortex ribbon is a charactoristic of parallel vortexes and not a reaction to any lifting force.

If I find downflow, a specific momentum, it will be the product of a specific pressure force and I am back to ballancing weight with static pressures.

Like I say, this is only a search. For hundreds of statements about deflected air no one has ever offered the application to the forces against the wing.
No circulation is not a rotating flow but its effect applies equally around the whole wing.

Then there is always the balloon.

 

[rcgldr]

What I am trying to say is that I cannot find a dedicated downflow that is a reaction to lift force.

A wing (producing lift) accelerates air downwards (and forwards), and the air reacts to this acceleration with an upwards (and backwards) force. These accelerations coincide with pressure differentials, lower above and behind a wing, higher below and in front of a wing.

Above a wing, Coanda and what some call "void" effect, cause the air to (mostly) follow the convex upper surface of a wing (during the transition from laminar to turbulent flow, which occurs on almost all wings, there is a "bubble" of separation of flow from the wing). The curvature of the air above the wing results in a lower pressure above the wing and a net downwards acceleration of air, and the cross-section of the affected air extends well above the wing, it's not just a near surface effect. Below a wing, the air is simply deflected downwards (assuming a typical air foil), increasing the pressure and more net downwards acceleration of air. The drag related effects cause pressure differentials in front and behind a wing with a net forwards acceleration of air.

 

[Count Iblis]

So, given the Navier Stokes equations, what is the optimal shape of a glider if we fix the mass and minimize the descent angle?

 

[rcgldr]

So, given the Navier Stokes equations, what is the optimal shape of a glider if we fix the mass and minimize the descent angle?

Don't know about the equations, but the optimal shape is simply a huge wing span with narrow chord, in otherwords a huge aspect ratio (as pointed out by Fred Garvin in another thread), combined with good airfoils (as I pointed out in the same thread). The result is 60 to 1 glide ratios at around 60mph in open class cross country gliders. The wingspans are huge, 80 feet or more. The 15 meter (just under 50 feet) class gliders get up to 50 to 1 glide ratios. Aerobatic gliders are around 35 to 1.

Gliders with 60 to 1 glide ratios.

http://www.sailplanedirectory.com/PlaneDetails.cfm?planeID=28

http://www.sailplanedirectory.com/PlaneDetails.cfm?next=118

http://www.sailplanedirectory.com/PlaneDetails.cfm?next=274

http://www.sailplanedirectory.com/PlaneDetails.cfm?next=277

Wiki link to Nimbus 4
Schempp-Hirth Nimbus 4 Wiki.htm

Official site for Nimbus 4, with photos:
http://www.schempp-hirth.com/index.php?id=nimbus-4dm0&L=1

The ETA is a motorized prototype, with a 101 foot wingspan, quite a few photos here:
http://www.eta-aircraft.de

For radio control gliders the shape is basically a skinny pole with a wing and a tail:

jrartms.wmv

 

[Cyrus]

What I am trying to say is that I cannot find a dedicated downflow that is a reaction to lift force. This is not argument, I simply do not see it. The required pressures are produced within curving flows where the verticle components of acceleration are equally up and down and the horizontal components contribute as much as the verticle. Unless one doubts lifttig line the downflow of the vortex ribbon is a charactoristic of parallel vortexes and not a reaction to any lifting force.

If I find downflow, a specific momentum, it will be the product of a specific pressure force and I am back to ballancing weight with static pressures.

Like I say, this is only a search. For hundreds of statements about deflected air no one has ever offered the application to the forces against the wing.
No circulation is not a rotating flow but its effect applies equally around the whole wing.

Then there is always the balloon.

Im sorry, I still don't understand a word your saying. Its as if your throwing in physics words into a sentence and hoping something sensible will come out. Is english your first language?

I mentioned the raleigh transport theorem on directed flow and force analysis.

I don't have a clue what your talking about. Is your background aerospace engineering?

 

[Stan Butchart]

Yes, after rockets and spacecraft I spent twenty years configurating tactical combat aircraft.
Guys, I guess I should bail on this. I am looking for help IDENTIFYING A NET downflow (other than induced alpha) and am unable to get that across.

Is there a sight that might provide the gist of the raleigh transport theorem?

 

[Cyrus]

I would just grab any old fluid mechanics textbook.


Sorry, its Reynolds not raleigh.

 

[rcgldr]

One half of wing lift occurs on the front half to third of the wing. Except for under the L.E. allof the flow here has an upwards component. This extends out for many,many chord lengths.

This conflicts with every source of information about wings that I have found. In all the articles I've seen the air flow is downwards for the rear 1/2 to 2/3rds of an wing, both above and/or below the wing, depending on the airfoil.

Aft of here, where the flow all has a downward component, pressure is constantly increasing.

This transition occurs above a wing, not behind the wing, and produces a small bubble during the transition from laminar to turbulent flow. I've already posted links to web sites that mention this.

"All airfoils must have adverse pressure gradients on their aft end. The usual definition of a laminar flow airfoil is that the favorable pressure gradient ends somewhere between 30% and 75% of chord."

http://www.aviation-history.com/theory/lam-flow.htm

For non-laminar airfoils, this transition from acceleration+decreasing pressure to deceleration+increasing pressure occurs even sooner along the chord.

Here's a picture and link to another web site:

http://www.dreesecode.com/primer/p4_f003.jpg
http://www.dreesecode.com/primer/airfoil4.html

I cannot find a dedicated downflow that is a reaction to lift force.

I prefer to think of lift force as an inertial reaction to the downwards acceleratoin of air. Perhaps the links I just posted will help point out the fact that downward air flow occurs while still flowing along the chord of a wing, and is not delayed for "several" chord lengths as you suggested.

If the frame of reference is the air itself, then as a wing passes through a volume of air, there is downwards and forwards acceleration of air, corresponding to reactive lift and drag forces.

A model flying or gliding inside a sealed box is a good means to "prove" that wings are air pumps. As long as the center of mass of the sealed box, air, and model are not accelerating vertically, then the weight of this system remains constant, regardless if the model is resting at the bottom of the box, or if flying or gliding within the box. In the case that the model is resting in the box, then the weight of the air creates a pressure differential that decreases with height so that the net downforce of the pressure differential exactly equals the weight of the air, while the model exerts it's weight directly on the bottom of the box. In the case that the model is flying or gliding, then the model increases the pressure differential within the box so that the net downforce due to the pressure differential exactly equals the weight of the model and air inside the box; therefore the models lifting surfaces are effectively air pumps.

 

[Stan Butchart]

Cyrus - My Shames makes no reference.Do you have a more specific reference? Is this our old Reynolds frome Reynolds Number?.

Jeff - Note that Laminar/Terbulent flow is a different subject than a Laminar/Turbulent boundry layer. To describe the basic way in which lifting pressures are derived they consider the flow outside of the boundry layer. The downflow along the wing is not greater than the upflow at the front part of the wing. The upper pressure reduction is CREATED in the upflow.
I'll stuggle with the box.

 

[rcgldr]

The downflow along the wing is not greater than the upflow at the front part of the wing. The upper pressure reduction is CREATED in the upflow.

The downflow doesn't occur just along the wing, it occurs well above and/or below the wing as well (depending on the airfoil). The upper pressure reduction occurs because of Coanda and "void" effect, which cause pressure differentials and acceleration of air. Also, it's not the direction of the flow that counts, it's the direction of the acceleration, which is downwards, the flow curves downwards.

There's no way around the "model flying in a sealed box proof" that wings are "air pumps", as it's a closed system. The classic anecdotes for this are a closed truck full of birds and a guy banging on the truck to get the birds to fly so the truck "weighs" less, or the question about the weight of a plane if birds are flying inside the plane.

 

[Cyrus]

Cyrus - My Shames makes no reference.Do you have a more specific reference? Is this our old Reynolds frome Reynolds Number?.

Jeff - Note that Laminar/Terbulent flow is a different subject than a Laminar/Turbulent boundry layer. To describe the basic way in which lifting pressures are derived they consider the flow outside of the boundry layer. The downflow along the wing is not greater than the upflow at the front part of the wing. The upper pressure reduction is CREATED in the upflow.
I'll stuggle with the box.

Try fluid mechanics by Munson Young Okishi.

 

[Phrak]

"All airfoils must have adverse pressure gradients on their aft end. The usual definition of a laminar flow airfoil is that the favorable pressure gradient ends somewhere between 30% and 75% of chord."

http://www.aviation-history.com/theory/lam-flow.htm

For non-laminar airfoils, this transition from acceleration+decreasing pressure to deceleration+increasing pressure occurs even sooner along the chord.

I think you may have inverted what you intended to say. As I recall a turbulent boundry layer detaches further into an adverse pressure gradient than a laminar boundry layer due to greater mixing with the general airsteam, thus the interest in tripping the boundry layer toward the leading edge into turbulence with devices such as turbulators.

 

[rcgldr]

"All airfoils must have adverse pressure gradients on their aft end. The usual definition of a laminar flow airfoil is that the favorable pressure gradient ends somewhere between 30% and 75% of chord."

http://www.aviation-history.com/theory/lam-flow.htm

For non-laminar airfoils, this transition from acceleration+decreasing pressure to deceleration+increasing pressure occurs even sooner along the chord.

I think you may have inverted what you intended to say. As I recall a turbulent boundry layer detaches further into an adverse pressure gradient than a laminar boundry layer due to greater mixing with the general airsteam, thus the interest in tripping the boundry layer toward the leading edge into turbulence with devices such as turbulators.

I was just quoting the article in that link. Did I get my post link description of the transition wrong? In the case of laminar airfoils, the ideal is to move the transition point further back on the airfoil. In the case of gliders, the low speeds result in the laminar "bubble" causing more drag than the turbulent flow, so a rough surface, or tubulators below and/or above a wing are used. Turbulators are also used on some powered aircraft, but I'm not sure why (maybe at higher air speeds like above mach .6 or mach .8?).

 

[djeitnstine]

This website does a descent job of explaining it, with a lot of emphasis on Coanda effect, but towards the end of this web site, there's a diagram of a wind blowing over a roof, and although the air downwind of the roof is turbulent, it's also at lower pressure, due to what some call "void" effect: when a solid object passes through a fluid, or when a fluid passes around a solid object, low pressure "voids" are created because the solid object blocks or diverts the fluid flow away from these low pressure areas.

After visiting a large number of web sites, my conclusion is that lift is a combination of Coanda and "void" effects.

Thank you sir, this clears up quite a bit although my current knowledge only carries my understanding half way through that lecture lol.

 

[Phrak]

I was just quoting the article in that link. Did I get my post link description of the transition wrong? In the case of laminar airfoils, the ideal is to move the transition point further back on the airfoil. In the case of gliders, the low speeds result in the laminar "bubble" causing more drag than the turbulent flow, so a rough surface, or tubulators below and/or above a wing are used. Turbulators are also used on some powered aircraft, but I'm not sure why (maybe at higher air speeds like above mach .6 or mach .8?).

Something doesn't add up with that link. Really, I haven't looked at either fluid dynamics or theory of flight in a number of years, but I recall the laminar flow idea is a matter of historical note in so far as the belief that maintaining a laminar boundry layer was either advantages or practically achievable.

A model aircraft with a 4 inch cord and maybe 30 mph flies in the laminar regime. An aircraft wing any much larger or faster is in the turbulent regime except under the most tedious conditions: a polished surface (no rivets of course) and any variations in the surface such as wavyness could trip the boundry layer to turbulent flow.

Like so much of fluid dynamics, it's counter-intuitive, but a turbulent boundry layer is advantageous because of the improved L/D. Any of the upper wing surface past the point at which the boundry layer detatches, generates, effectively no lift. The further along the upper surface you can delay the onset of boundry separation, the better.

The best of both worlds would be to have a laminar boundry layer as far past the leading edge as possible, then trip to turbulent flow somewhere immediately before the point where of laminar separation would occur, but I recall the efforts to obtain this are hampered by practical limitations from varying angle of attach and enviromental conditions like bugs, rain, air pressure.

I couldn't find any links worthy of posting, but one of them places the critical Rynyolds number for aircraft wings at 100,000-500,000 with small aircraft wings having values from 2,000,000 to 20,000,000 -- well within the turbulant regime.

 

[rcgldr]

After visiting a large number of web sites, my conclusion is that lift is a combination of Coanda and "void" effects.
http://user.uni-frankfurt.de/~weltner/Mis6/mis6.html

I should also include direct deflection of the air, which is what happens under a wing (assuming that significant lift is being produced from below a wing). So it's Coanda effect, "void" effect, and direct deflection of air.

A model aircraft with a 4 inch cord and maybe 30 mph flies in the laminar regime.

Actually it's more like 6 inch chord but 10mph airspeed. In the case of gliders, the laminar bubble creates more drag than turbulent airflow, so these small hand launch (now called discus launch because of the launching method) gliders use turbulator strips.

The best of both worlds would be to have a laminar boundry layer as far along the leading edge as possible, then trip to turbulent flow somewhere immediately before the point where of laminar separation would occur, but I recall the efforts to obtain this are hampered by practical limitations from varying angle of attach, and enviromental conditions like bugs, rain and air pressure.

This is done for model and full scale gliders. Note that the laminar detachement isn't "permanent" except for very low Reynolds numbers. Typically there's a reattachment of the flow once it's gone turbulent, because of what I call "void" effect, a wing produces a moving void of low pressure on it's aft portion, and this detachment is called a "seperation bubble". To control this transition bubble, sometimes just sanding the wing surfaces with 600 grit sandpaper is enough, in other cases. turbulator strips are used below and/or above on a wing.

In the case of gliders, oil flow tests are used to determine detachment of air flow. As noted at this web site, "Partially developed separation bubbles can actually have beneficial effects."

http://www.standardcirrus.org/OilFlows.html

 

[Phrak]

I should also include direct deflection of the air, which is what happens under a wing (assuming that significant lift is being produced from below a wing). So it's Coanda effect, "void" effect, and direct deflection of air.

Actually it's more like 6 inch chord but 10mph airspeed. In the case of gliders, the laminar bubble creates more drag than turbulent airflow, so these small hand launch (now called discus launch because of the launching method) gliders use turbulator strips.

This is done for model and full scale gliders. Note that the laminar detachement isn't "permanent" except for very low Reynolds numbers. Typically there's a reattachment of the flow once it's gone turbulent, because of what I call "void" effect, a wing produces a moving void of low pressure on it's aft portion. Sometimes just sanding the surfaces with 600 grit sandpaper is enough, in other cases. turbulator strips are used below and/or above on a wing.

In the case of gliders, oil flow tests are used to determine detachment of air flow. As noted at this web site, "Partially developed separation bubbles can actually have beneficial effects."

http://www.standardcirrus.org/OilFlows.html

Whoah, you've sure done your homework! I'd completely forgotten about the detachment, reattachment and all that.

From what I get out of that link, it's best to have a turbulator stip somewhere at the quarter cord point rather than a bubble. I think I'll go back to lurking.

 

[Phrak]

Originally Posted by Jeff Reid

For non-laminar airfoils, this transition from acceleration+decreasing pressure to deceleration+increasing pressure occurs even sooner along the chord.

Originally Posted by Phrak

I think you may have inverted what you intended to say...

Originally Posted by Jeff Reid

I was just quoting the article in that link. Did I get my post link description of the transition wrong?

Nope. Sorry. I misunderstood.

Originally Posted by Stan Butchart

I am looking for help IDENTIFYING A NET downflow ...

Nobody has answered this yet?? The upward force on the lifting surfaces is equal to the downward force on the surrounding air where the airplane has zero verticle accelleration. The verticle force on the lifting surfaces would be the integral of the pressure times area normals doted projected in the vertical direction.

$$F_{airplane}=\int P \hat{k}\cdot d\bar{n}$$

The force on the airsteam is harder. The downward velocity of each air particle decreases the further back in the airstream you find it as it shares it's momentum with surrounding particles. Is this Navier-Stokes?

 

[Stan Butchart]

I am not bright enough to figure out getting the quotes.
"The upward force"... Well put, but some interpretation (personal?) goes with the words.
"The downward velocity of each"...Having worked with classical aerodynamics I fail to find a downward flow which is the result of a force that contributes to, or results in, the actual creation of lifting force.

 

[Phrak]

I am not bright enough to figure out getting the quotes.

Basically, I was appologizing to JR, as it seems I may have insulted his intelligence.

"The upward force"... Well put, but some interpretation (personal?) goes with the words. "The downward velocity of each"...Having worked with classical aerodynamics I fail to find a downward flow which is the result of a force that contributes to, or results in, the actual creation of lifting force.

I've been searching for an easily understood explanation of what's going on, with as little extra clutter as possible. I think the best way to go about it is to start with the flow of air over an outwardly curved surface.

If you're willing to accept that the airflow doesn't break-away from the surface, leaving stagnant air or a vortex underneath, I think it's an acceptable explanation. As the air flows over the top surface of the wing the overall flow above the wing follows an arc. This means that the air is subjected to an acceleration normal to it's direction of flow--that is, toward the wing surface. The acceleration time the mass of the air exerts a counter-force on the wing--the lift. This is just F=ma. The arcing flow redirects the air slightly downward as it leaves the trailing edge.

The overall downward amount of flow as it gets as far as the tailsection is only a couple degrees or so.

What I would like to know is this: "Is stall a result of the stagnation point migrating toward the leading edge?

 

[rcgldr]

"Is stall a result of the stagnation point migrating toward the leading edge?

As I pointed out in previous posts, it's normal for the air stream to separate and reattach while it transitions from laminar to turbulent flow. As the angle of attack increases, the detachment zone gets larger, but the lift still continues to increase until you reach a crictical angle angle of attack where the lift is at it's maximum. Go beyond this, and the lift decreases, but it doesn't vanish completely, although there may be quite a drop in the amount of lift.

In the case of slow speed flight, the problem is excessive angle of attack reduces lift, causing the aircraft to sink, which increases the effective angle of attack, since the direction is now more downwards than before, this decreases lift more, which causes the aircraft to sink even more so; a viscous cycle, where control is negatively stable.

In the case of high speed flight, the wings aren't identical, so one wing goes past crictical before the other. Since the wing past critical has less lift, there's a viscous roll reaction, downwards on the wing past critical, so it makes even less lift, while upwards on the other wing which reduces it's effective angle of attack but doesn't reduce the lift as much as the wing past critical. The result is a snap roll.

On aerobatic radio control models, with excessive elevator throw, pulling back hard on the elevator results in a fast roll response with no hint of the expected pitch response, and without any aileron control input. It spooked me the first time it happened to me, which was with a friends small aerobatic glider, which I had in a dive out over a tall slope well above the ground below, so I had plenty of time to recover by easing off the elevator input. After that I thought it was cool that a pitch control input would result in a roll response. The owner of the model had set it up that way. Contest aerobatic powered models are setup similar to this, to produce a true snap roll for competion.

Snap rolls are bad during a speed contest, where the snap roll results from pulling too many g's in a turn, while low to the ground. It's a 50/50 chance that the model will roll downwards into the ground and crash, or upwards with no harm done.

For glider being launched via a line drawn by a winch or strong latex tubing, the high loading combined with too aggressive pitch input (excessive elevator trim) can result in a snap roll. The instinct is to try to recover with aierlons but this make the situation worse because the alieron input increases camber and effective angle of attack on the downwards moving wing, reducing it's lift further still. The general rule for gliders is that if something goes wrong, down elevator should be the first control input, to make sure that the glider isn't experiencing a stall or snap roll situation.

If you're willing to accept that the airflow doesn't break-away from the surface, leaving stagnant air or a vortex underneath.

Actually for almost all situations, there's always some break-away "bubble", but it's very small, in the mm range in some cases, during the transition from laminar to turbulent flow. Even if the turbulent flow is composed of small oval eddies, it's still not an issue as long as the pressure is still below ambient in that turbulent flow, which it usually is. In the case of delta wings, they can handle huge angles of attack (over 20 degrees) without stalling, because the shape of the wing (triangular front, flat back) allows it to take advantage of turbulent eddies that flow across it.

 

[Stan Butchart]

Again Phrak states well.
I have been inmeshed in the fuzzy world of logic and symantics. Stuff like -I could not get the arithmetic to work for "centrifical" effects. The force for following the surface came from ambient pressure. Fluid "static" pressure has no directionality, only the receiving surface. This clouds the word "downward". The balloon was pure differential static pressure so did achieving the reduced static pressure by dynamics require "down"?
The tangential accelerations that exist within the curving flow produce the right answers.

Anyway, thanks I can't argue with that but my mind will have to do some smoothing yet.

 

[Phrak]

As I pointed out in previous posts, it's normal for the air stream to separate and reattach while it transitions from laminar to turbulent flow. As the angle of attack increases, the detachment zone gets larger, but the lift still continues to increase until you reach a crictical angle angle of attack where the lift is at it's maximum. Go beyond this, and the lift decreases, but it doesn't vanish completely, although there may be quite a drop in the amount of lift.

Ok. It had occurred to me that I haven't seen any satisfying graphs of pictures of flow fields as a wing section progresses into stall. I've presumed it to be a progressive separation of the boundry layer advancing toward the leading edge with increasing angle of attack. I'm speaking of the permanent separation of the boundry layer, rather than the laminar to turbulent transition. And of course the position of the line of separation isn't necessarily stable but could be oscillatory or even chaotic for all I know.

But I'm given thought to an interesting alternative mechanism (that you may have alluded to--I can't tell.) whereby the boundry transition bubble fails to reattach. Perhaps a reflex wing section, or a high lift system could exhibit this.

 

[Stan Butchart]

Regaurding normal acceleration: When I look at a pressure gradient using Rhov^2/r = dP/dr it comes up way short of the gradient useing Bernoulli. Am I missing a basic princple here?

 

[Phrak]

Regaurding normal acceleration: When I look at a pressure gradient using Rhov^2/r = dP/dr it comes up way short of the gradient useing Bernoulli. Am I missing a basic princple here?

What geometry are you using to with the equation Rhov^2/r = dP/dr?

 

[rcgldr]

I haven't seen any satisfying graphs of pictures of flow fields as a wing section progresses into stall.

How about a video? It appears to be a narrow wind tunnel, so it's considered a "2d" airflow (equivalent to a 3d wing with infinite wingspan). It's also apparently a small airfoil so the Reynolds number would be quite low, and the air flow is much more laminar and the angle of attack is much higher than it would be if everything were scaled up to a larger size. The transition into the stalled condition is very abrupt. In the segment annotated as "stall", there's virtually no lift, but near the end of the video, that starts off "flow attached", then "stall", there's still significant lift although there is a stall.

http://www.youtube.com/watch?v=6UlsArvbTeo&fmt=18

Assuming this next video isn't a GGI video, it appears to be a flame aimed at various angles over an glowing (from the heat) airfoil at a fixed angle of about 45 degrees. As the flame angle is made more horizontal, the effective angle of attack becomes higher. What I call "void" effect is more evident here, as the flame flow is detaches from the aft end of the airfoil at low effective angle of attack. At higher effective angle of attack, the flame flow detaches from the "upper" surface of the airfoil, but it's stil accelerated (curved) "downwards", while below the airfoil there is significant direct deflection.

http://www.youtube.com/watch?v=hkJaTTIiXSc&fmt=18

The second video looks much different than the first video. I can think of 3 reasons for this. First the behavior of the heated gas is similar to a wind tunnel with a much higher air speed than the wind tunnel video. Second, it's a heated gas instead of normal air. Third, it's an open environment, whereas the wind tunnel is sealed above and below, preventing much downwards flow of the air (resulting in more pressure effects and less flow effects).

I'll keep searching for a more open (larger) and higher air speed wind tunnel.

 

[rcgldr]

More videos:

For this model, the stalling angle of attack is fairly small:
http://www.youtube.com/watch?v=5wIq75_BzOQ&fmt=18

A small wing and slow air speed:
http://www.youtube.com/watch?v=TGUSmdFmXDg&fmt=18

This is so wrong, but it's the classic and incorrect equal transit time "hump" theory from an old film:
http://www.youtube.com/watch?v=mTW4Fdcoyqs&fmt=18

 

[Stan Butchart]

What geometry are you using to with the equation Rhov^2/r = dP/dr?

For the circular cylinder, the min radius at the top of the cursve "e" for the flow relative to the remote still air.

 

[Phrak]

It's a pity that detailed videos are so hard to find. With luck, in a few years, someone will come up with a quality video of a section undergoing stall, complete with the laminar/turbulant transition bubble included. In fact, it could be done by combining both smoke and the oil film you showed us, if need be. This does bring up a question. Do you have a source I could look at that talks about stall proceeding from the transition point?

This Cambride video http://www.youtube.com/watch?v=6UlsArvbTeo&fmt=18 is the best, overall, I think. If you look at the pulsed smoke part of it, over the top of the section the pulses remain in a nearly verticle row; the v_x velocity remains nearly constant. The overall velocity increases. How much of this effect is due to the top of the wind tunnel interferring with the wing is hard to tell. I recall, the top of the box is only about 3/4 cord from the wing.

In this, http://www.youtube.com/watch?v=5wIq75_BzOQ&fmt=18,
video the abruptness of stall is frightening!

In both videos the boundry layer separates at the trailing edge. The greater the angle of attack the sooner the separation, It makes sense of course; the sooner the flow reaches stagnation, the sooner it separates.

In the second video the sudden separation at the leading edge is the most suspicious. Why should the stagnation point transite so suddenly to the leading edge? Do you think it not a result of the stagnation point moving forward, but a failure of the boundry layer to reattach at turbulent/laminar transission; that is, failure to tranite to an attached turbulent boundry layer?

 

[Phrak]

By the way, you both might find this interesting. A high lift foil set quoted at an incredible CL_max=4.81.

http://adg.stanford.edu/aa241/highlift/highliftintro.html"

 

[Stan Butchart]

Three years later! I have to close by saying that these conversations do produce added insite.
Previously I tried to calculate the centripedal acceleration of curved flow from its inertial path. In reality it works just fine when useing the velocity relative to the surface and surface radious. Interestingly, when we multiply v^2/R *Rho times R/2 (which integrates the entire normal column) we wind up with the Bernoulli equation even though "Bernoulli flow" is not present.

The place of vertical Newtonian acceleration is still not straight forward. Normal acceleration produces pressure change against the local surface element. Lift contribution is the vertical component of that surface pressure. The vertical component of that normal mass acceleration was indeed equal to the lift contribution. It satisfies Newtons laws of force ballance. However, to explane lift, the pressure change against the suface element is created by the normal acceleration. The surface element dose not recognize
the direction from which it created. Lift contribution is determined by surface element orientation.