P-40 turn rate/Flight model check

[Venturi]

There are four weights - the minimum 3264kg, full maximum 4414kg - "standard" is 3819kg. 3846kg appears nowhere: in your calculation you are taking very nearly the minimum of the range of speeds with a weight about 18% higher than their "standard". Of course this gives a high CLmax. I wanted to be nice.

 

I covered this in post 320:

 

      I take the minimum stall speed in DD 123  153 kph as being at the minimum weight: 3264 kg, > CLmax of 1.32

      At maximum stall speed - 176 kph I assume "maximum takeoff weight"  4414 kg. > CLmax 1.35

      Somewhere in the middle, "standard weight"  3819kg, speed (153+176)/2 = 165kph > CLmax  1.34

 

Well if you watch Kai Lae's videos, he stalls at 96mph with full fuel and ammunition. Which would be about 3850 per Boscombe trials. But I recognize that you do not give those documents any credence. 

 

If you like, the pilot's operating manual states stall at 90mph (does not specify weight, but let's say they decided maybe it was safest to be conservative here, what do you think?)

It also states best glide speed at 110-115mph, again strange that such a low airspeed is given for a power off glide considering that stall happens, according to you (and devs? apparently) at only a few mph lower. Must be a strange sort of glide with the power off you know, if the engine quit a few miles from the runway and full fuel/ammo load, just a few mph above stall.)

Actual wings have manufacturing imperfections, are not completely smooth and have lift interfering gun-ports etc. Planes have interference between wing and fuselage plus +/- lift effects from non - wing surfaces.

Consequently Clmax is always less for wing than airfoil, and for whole plane than for wing. Stalls at 90mph, per handbook "plus or minus a few mph depending on aircraft"

 

Whatever your view is here, the developers clearly disagree with Crump and their game choices reflect this, I do not expect the P-40 to be treated any differently. Low expectations never fixed anything. I expect more from BoX.

Edited January 1, 2017 by Venturi

[unreasonable]

The real issue both players and the AI have with the P-40 - as with the Fw190 - is actually the low critical AoA.  Most of the complaints are not about stalling out at too high a Vmin: they are about unexpected accelerated stalls.

 

Checking the 109F4 using Hans figures and the same methodology gives an in game Clmax of 1.40 -1.42, ie only 5% higher than the P-40.

 

But the critical AoA is 19.9% compared to the 14% for the P-40.  That is a huge difference: 42% if expressed as an advantage for the F-4.

 

However, the graph for the 22-- wing in the report you posted shows a AoA at Clmax of about 20-22%. I suspect that if everyone could fly at 20% AoA they would be much happier, whether the Clmax is set at 1.5 or 1.35

 

What I do not know if the critical AoA for a plane relates to that for a wing in the same way as the Clmax: ie that the wing represents a maximum but the reality could be somewhat worse. Also I have no idea if the FM determines one from the other or are they separate variables.  Increasing Clmax might not increase critical AoA - perhaps it would just make the curve steeper?

 

But I cannot help think that this might be a more productive line of enquiry. 

[Silavite]

The real issue both players and the AI have with the P-40 - as with the Fw190 - is actually the low critical AoA.  Most of the complaints are not about stalling out at too high a Vmin: they are about unexpected accelerated stalls.

 

Checking the 109F4 using Hans figures and the same methodology gives an in game Clmax of 1.40 -1.42, ie only 5% higher than the P-40.

 

But the critical AoA is 19.9% compared to the 14% for the P-40.  That is a huge difference: 42% if expressed as an advantage for the F-4.

 

However, the graph for the 22-- wing in the report you posted shows a AoA at Clmax of about 20-22%. I suspect that if everyone could fly at 20% AoA they would be much happier, whether the Clmax is set at 1.5 or 1.35

 

What I do not know if the critical AoA for a plane relates to that for a wing in the same way as the Clmax: ie that the wing represents a maximum but the reality could be somewhat worse. Also I have no idea if the FM determines one from the other or are they separate variables.  Increasing Clmax might not increase critical AoA - perhaps it would just make the curve steeper?

 

But I cannot help think that this might be a more productive line of enquiry. 

Erm, when you say 19.9% in respect to AoA, do you mean 19.9 degrees? I was confused reading that.

[ZachariasX]

Erm, when you say 19.9% in respect to AoA, do you mean 19.9 degrees? I was confused reading that.

It should be. He's not that unreasonable

 

I think unreasonable has a point here by mentioning the lack of AoA imposed by a reduced Clmax. This extra alpha is just what gives you a gun solution or not while turning after the other guy. Not having the ability to pull up feels very much like having a wider turn radius.

 

As constant circling has not much of a place in a multi-ship engagement, I'd say the extra alpha is even more important than the extra lift. At a "good" combat speed you are power limited anyway, so it is a game of power to weight, not a game oft lift to drag.

 

It seems to me that this is also what makes the F6F as well as the Fw190 great gun platforms: you can follow your mark at higher speeds and have this extra alpha when you need to pull alpha for the gun solution. And this in combination with a great view over the nose.

 

I see the importance of published polars for wing profiles mostly in this light. It gives me an angle, at wich my wing root will stall. Lift however, in total, will be slightly less in the real world as unreasonable summarized above.

 

While "whole aircaft total lift" lift (defining the circle diameter and substained turn time) gets affected by wing shape, washout, build quality, etc., the stalling AoA of the wing root will not for the profile given. And that stalling angle is at the Clmax of the published wingroot profile.

 

1. Airspeed measurement was not accurate enough in the 1940's to calculate CLmax from a flight test. That is exactly why Grumman used acceleration measurements and the normal force coefficient in their inflight measurements. That data gave good agreement with their measured results for their wing design. They did not use airspeed and weight estimations in a simply lift formula analysis. It is simply NOT accurate enough.

After using unreasonable's spreadsheet, this is my feeling as well. But I really find that one useful to check if figures are plausible. And that it does very well.

[Dakpilot]

It also states best glide speed at 110-115mph, again strange that such a low airspeed is given for a power off glide considering that stall happens, according to you (and devs? apparently) at only a few mph lower. Must be a strange sort of glide with the power off you know, if the engine quit a few miles from the runway and full fuel/ammo load, just a few mph above stall.)

 

In the manual and training video it talks about best speed for landing glide approach as 110-115mph, down to throttle cut at 10 feet above runway, this is full flaps at the last 500ft,

 

At a landing configuration stall of 84 this correlates to 1.3 X  stall for landing  config, which is pretty much a universal aviation generalism, and makes sense

 

In which state of flight could the engine 'quit' just a few MPH above stall?? especially a few miles from the runway??  min climb out speed is 120mph, 30mph above stall in clean config and the slowest you are likely to ever be at any stage except on approach

 

It does not state in the manual that best power off glide speed clean is 110-115 (which you highlighted and quoted from) It states

 

"when gliding in for a landing use your trim tabs to establish an indicated airspeed of about 110-115mph"

 

Cheers Dakpilot

[unreasonable]

Erm, when you say 19.9% in respect to AoA, do you mean 19.9 degrees? I was confused reading that.

 

Yes, sorry! 

Edited January 1, 2017 by unreasonable

[Crump]

As for Crump's comments, this has all been covered ad-nauseam in Fw190 threads, so I will just summarize, I hope not misrepresenting anyone.

 

The maximum CLmax for the plane cannot exceed that for the wing, and for the wing cannot exceed that of the airfoil. Everyone agrees.

 

Actual wings have manufacturing imperfections, are not completely smooth and have lift interfering gun-ports etc. Planes have interference between wing and fuselage plus +/- lift effects from non - wing surfaces.

Consequently Clmax is always less for wing than airfoil, and for whole plane than for wing.

 

I think everyone agrees with that, the disagreement is about the size of the differences, especially between wing model and whole plane tests.

 

Crump argues that the wing Clmax represents the goal towards which the designers aspire, and that differences represent problems that are solved, so that the measured plane Clmax should be at most a few % lower than the wing.

 

Others agree that this is the goal, but point out that problems are not in practice always solved. This view is backed up by NACA empirical tests that show the 0.10-0.20 differences from wing - plane I previously mentioned.

 

Whatever your view is here, the developers clearly disagree with Crump and their game choices reflect this, I do not expect the P-40 to be treated any differently.

 

 

Good post except the NACA does not agree that Clmax gets lowered and designer just accept it.  The NACA publishes guidelines and consolidates knowledge for engineers to reduce the amount of time spent troubleshooting their designs.  Design Teams simply will not accept a wing design that does not meet the design aerodynamic coefficients.  They will fix it.

 

As for turning performance calculations, it is the WING CLmax that is used:

 

 

– nmax is a function of CLmax ; – at low speeds it will be limited by the aerodynamic lifting capability (stall) of the lifting surfaces

 

 

http://people.clarkson.edu/~pmarzocc/AE429/AE-429-12.pdf

 

You can get a practical demonstration of the fact only the wing(aerodynamic surfaces) is considered in the pre-contamination and pre take off checks for icing conditions that every pilot performs (or at least should perform)  in pre-flight checks.

 

 

 

 

 

The current regulations in parts 121, 125, and 135 prohibit a takeoff when frost, ice, or snow (contamination) is adhering to the wings, control surfaces, or propellers of an airplane (see § 121.629(b), part 125, § 125.221(a), and part 135, § 135.227(a)).

 

 

 

 

Section 121.629(b) prohibits takeoff when contamination is adhering to critical surfaces of an airplane or when takeoff would not be in compliance with § 121.629©. The exception to that general rule is that the Administrator may approve takeoff with “frost under the wing in the area of the fuel tanks.”

 

 

 

A pretakeoff check is a check of the aircraft’s wings

 

http://fsims.faa.gov/WDocs/8900.1/V03%20Tech%20Admin/Chapter%2027/03_027_001.htm

 

 

 

After using unreasonable's spreadsheet, this is my feeling as well. But I really find that one useful to check if figures are plausible. And that it does very well.

 

I agree.  I think unreasonable provided the community a great tool for gaining insight into how the science works and why the argument over the FW-190's CLmax was so drawn out without it.  Folks simply did not understand the factors and the relationships being discussed until they were able see for themselves on that sheet.

 

 

What we really need is the designers wing CLmax.

[303_Kwiatek]

Erm, when you say 19.9% in respect to AoA, do you mean 19.9 degrees? I was confused reading that.

In BOS P40 got 14,5 degree critical angle of attack which is suprisly very low value comparing to other planes which are in range 16-18 degree without slats

Edited January 1, 2017 by 303_Kwiatek

[Crump]

 

 

In BOS P40 got 14,5 degree critical angle of attack which is suprisly very low value comparing to other planes which are in range 16-18 degree without slats

 

And having a NACA 4 digit series in the 20% Cmax..... 

[ACG_KaiLae]

 

What we really need is the designers wing CLmax.

 

 

So, how do we get it? What tree do we need to shake? While it's been 70 years I don't see why all the curtiss documents would have simply vanished.

[Crump]

 

 

Actual wings have manufacturing imperfections, are not completely smooth and have lift interfering gun-ports etc. Planes have interference between wing and fuselage plus +/- lift effects from non - wing surfaces. Consequently Clmax is always less for wing than airfoil, and for whole plane than for wing.

 

Here is where you are misinterpreting experimental results that detail the problem with the not being able to fix the problem.

 

 

The general conclusion of the NACA airfoil data reflects aeronautical engineering norms in that designers can, will, and do attain the aircraft design CLmax.  If they cannot for some reason, then it is back to the drawing board until they DO attain their design coefficients.

 

Even such things as gun installation are corrected, unreasonable.  

 

You can clearly see the line of experimentation used by Focke Wulf in the gun installation design.

 

 

 

General check of the flight characteristics and determination of level speeds after installing an MG 151 in each wing instead of the MG FF.

 

 

 

With respect to flight characteristics, no differences were noticed due to the slightly modified external form (different hatch under the wing) compared to A-5 production aircraft.

 

This is a retest of a previous hatch design which did not give good agreement with the design coefficients of lift their wing design should produce.  Hence, they went back to the drawing board, found the problem, redesigned the cover, and tested the results.

 

 

 

The values obtained lie within the usual limits for the 190 A-5 without ETC 501, so that the acceptance flight report the betr. Serie can be accepted without change for the 190 A-6.

 

http://www.wwiiaircraftperformance.org/fw190/fw190-861-nr1.html

 

Every design team in existence conducts this kind of testing to eliminate the possible errors that can be introduced.  Grumman did it, Curtiss did it, Supermarine, Focke Wulf, Lavochkin, etc....

 

Taking the NACA's determination of the scope of that possible error and determining that error is normal or acceptable is not the correct conclusion.

 

 

o, how do we get it? What tree do we need to shake? While it's been 70 years I don't see why all the curtiss documents would have simply vanished.

 

No they have not vanished. I am waiting on work to give me a schedule for the next month.

 

I will have some time off in the last two weeks of January and now that I live very close to the Smithsonian, I will hit the archives and find it.  It has to be in there somewhere.  I also plan to get the report on the Longitudinal Flying Qualities of the P-40 too.

[Farky]

So, how do we get it? What tree do we need to shake?

 

Without a doubt Smithsonian National Air and Space Museum.

 

 

I will have some time off in the last two weeks of January and now that I live very close to the Smithsonian, I will hit the archives and find it.  It has to be in there somewhere.

 

I think this could be useful for you Crump (Curtiss-Wright Corporation Records Finding Aid for Smithsonian) - http://sirismm.si.edu/EADpdfs/NASM.XXXX.0067.pdf

[Crump]

That really helps and compiling a list of documents I want to examine.  

[Crump]

Farkey, the information is there and should be contained in numerous reports.  So far I have six pages of likely areas to find the design coefficients starting with "Report of Engineering Data" submitted by the Engineering Dept., 2/11/1942 Models covered: Tomahawks & Kittyhawks.

 

There are numerous wing torsional test and structural testing that will give us the Design Clmax too.

[Venturi]

Current Facts:

• 1. the published stall speed of the P40E in the Operating Manual is 90mph ASI ("plus or minus a few mph depending on aircraft"),

• 2. the 25+ page report from the RAF Boscombe Down P40E trials measured its stall speed at 90mph ASI,

• 3. the CLmax of a generic 2218-09 airfoil is not out of the ballpark of the CLmax value which is empircally derived from the Boscombe Trial data,

• 4. the plane does not currently perform at the level historically expected by period reports and pilot interviews

• 5. with a fully loaded stall at 96mph TAS as measured by Kai Lae, the P40E's current CLmax is significantly below what it should be, using data from the above 4 points

I await the published design CLmax.

[Venturi]

As for PEC graphs...

 

 

[ACG_KaiLae]

No they have not vanished. I am waiting on work to give me a schedule for the next month.

 

I will have some time off in the last two weeks of January and now that I live very close to the Smithsonian, I will hit the archives and find it.  It has to be in there somewhere.  I also plan to get the report on the Longitudinal Flying Qualities of the P-40 too.

 

[BlitzPig_EL]

Good news Crump.

 

I am not going to hold my breath on getting a reply from the Curtiss Museum.  

 

Perhaps they will wake up after the Holidays, but I'm kind of doubting it.

 

 

As an aside for my own knowledge, when we talk about a wing having a NACA X profile at the root, and a NACA Z profile at the tip, is it not true that the ONLY place that wing will have those profiles is at those two precise places? (Unless it is a totally straight wing with no chord change at all over the entire span).

 

It's always been one of the things that I have thought about over the years when us simmers discuss turn capabilities and the focus always seems to be on wing and power load, yet no one takes into consideration the effect of what wing type is used.

[Venturi]

More stall data.

 

 

...86mph at sea level, flaps up.

...80-81mph, flaps down.

more... 88mph flaps and gear up.

[Crump]

 

 

As an aside for my own knowledge, when we talk about a wing having a NACA X profile at the root, and a NACA Z profile at the tip, is it not true that the ONLY place that wing will have those profiles is at those two precise places? (Unless it is a totally straight wing with no chord change at all over the entire span).

 

Generally the wing design will blend the two airfoils at some point and you would have to examine the details of the specific wing design.  Outside of "neat to know" it has little bearing on aircraft performance.  The root airfoil determines the stall point in terms of aircraft performance math for an unswept high aspect ratio wing without or moderate taper.   That means the P-40 fits and will use the root airfoil characteristics.

 

The wingtip airfoil is mainly there to mitigate stall behavior and maintain aileron control at the stall point.  That we can keep our wings level with our ailerons has not bearing on fact the airplane is not turning anywhere or performing anything except a stall with the root airfoil sections no longer flying.

 

Airfoil selection makes a large difference and is a very important part of the design.  Think of it like the tires on a race car.

[BlitzPig_EL]

Thanks.

[Crump]

You are most welcome, Blitz_Pig.

[unreasonable]

Current Facts:

• 1. the published stall speed of the P40E in the Operating Manual is 90mph ASI ("plus or minus a few mph depending on aircraft"),

• 2. the 25+ page report from the RAF Boscombe Down P40E trials measured its stall speed at 90mph ASI,

• 3. the CLmax of a generic 2218-09 airfoil is not out of the ballpark of the CLmax value which is empircally derived from the Boscombe Trial data,

• 4. the plane does not currently perform at the level historically expected by period reports and pilot interviews

• 5. with a fully loaded stall at 96mph TAS as measured by Kai Lae, the P40E's current CLmax is significantly below what it should be, using data from the above 4 points

I await the published design CLmax.

 

The only points here which are facts are 1 and 2, 3 is simply false, the rest is interpretation, contested and yet to be decided.

 

Take 3:

 

 If you use the Boscombe figures of 90mph and 8480lbs you get Clmax of 1.73   (assuming zero PEC, even higher if PEC is negative at stall as two out of your 4 examples show - your "blue dots" example does not )

 

Assuming that the plane Clmax (flaps up) cannot be higher than the airfoil Clmax, which is 1.60 according to the NACA doc, this is physically impossible.

 

To get it at or below 1.60 you either have to knock off 300kg (660lbs ie about 3/4 of the whole fuel weight IIRC) or add 3.7 mph to the stall speed or some combination of the two.

 

Kae Lae's test at 96mph fully loaded does indeed get you into the zone of possibility at Clmax 1.53, so sensitive is the calculation to the speed change, whether this is below what it should be is still very much in doubt.

 

There is still the open issue of why - if you accept results calculated from flight tests  in the 1.50-1.60 range - the loss of Clmax from airfoil to plane results would be so much less for the P-40 than for other RL planes measured, according to some posters.   This will still be an issue even if we get a manufacturer's figure for the wing that is close to the 1.60 for the airfoil, just as in the Fw190 discussion. 

 

I would add to your list of facts that the in-game critical AoA of 14 degrees is much lower than the documented figure for the airfoil of about 20 degrees and at the moment no-one seems sure why, or how this and the Clmax are linked in the game FM.

[Venturi]

Here's another fact, we now have four period sources giving flaps up stall speed at 86-90mph, and another giving flaps down at 80mph from Curtiss, which agrees with the other sources' flaps down stall speeds.

 

You can keep on making up explanations, go ahead.

Edited January 2, 2017 by Venturi

[unreasonable]

Here's another fact, we now have four period sources giving flaps up stall speed at 86-90mph, and another giving flaps down at 80mph from Curtiss, which agrees with the other sources' flaps down stall speeds.

 

You can keep on making up explanations, go ahead.

 

I am not "making up" any explanations,  I agree that there is a bit of a mystery here.  But how can the Clmax measured from flight tests possibly be higher than that of the airfoil?

 

Either:

 

1) The airfoil/wing number we have is wrong,

2) The equations do not apply to the P-40 or

3) The Boscombe and Manual data is incomplete or we are interpreting it incorrectly by plugging it directly into the equations.

 

I am obviously going with (3) given current information. Which is your choice?

[Crump]

 

 

But how can the Clmax measured from flight tests possibly be higher than that of the airfoil?

 

It is not without the use of high lift devices like flaps or slats.  However you must read the 2D data correctly and use the correct airfoil data under the same RE number in order to get the Clmax.  

 

If the polar is not at the same conditions of flight and it is not the same airfoil then there is nothing to be learned in terms of any specific value of Clmax.

 

Either:   1) The airfoil/wing number we have is wrong,

 

 

 

The airfoil data is wrong.  The P-40 used the NACA 2215 airfoil at the root and NOT the NACA 2218.09.  

We will get the correct data.

[Venturi]

There are five sources giving relevant data:

 

1. USAAF pilot's operating handbook - 88mph clean stall, 79mph gear and flaps down, 78mph stall flaps down

2. USAAF pilot's training manual - 90mph clean stall, 84mph gear and flaps down

3. RAF Boscombe Down trials - 90mph clean stall, 80mph gear and flaps down 

4. USAAF HQ Technical Order for P40E - 86mph clean stall, 81mph gear and flaps down

5. Curtiss-Wright Corp. Hawk series comparison sheet - 80mph stall flaps down

 

So I have a really hard time believing that these are all wrong.

[ACG_KaiLae]

Let's see what comes out of the national archives on what the design specification of the plane is, and we can then go from there.

[Venturi]

It is not without the use of high lift devices like flaps or slats.  However you must read the 2D data correctly and use the correct airfoil data under the same RE number in order to get the Clmax.  

 

If the polar is not at the same conditions of flight and it is not the same airfoil then there is nothing to be learned in terms of any specific value of Clmax.

 

The airfoil data is wrong.  The P-40 used the NACA 2215 airfoil at the root and NOT the NACA 2218.09.  

 

For example, here is a source giving 2212 airfoil (same family, slight differences as 2215) a CLmax of 1.72...

From Aerodynamic characteristics of a large number of airfoils tested in the variable-density wind tunnel, pub 1938.

[unreasonable]

It is not without the use of high lift devices like flaps or slats.  However you must read the 2D data correctly and use the correct airfoil data under the same RE number in order to get the Clmax.  

 

If the polar is not at the same conditions of flight and it is not the same airfoil then there is nothing to be learned in terms of any specific value of Clmax.

 

The airfoil data is wrong.  The P-40 used the NACA 2215 airfoil at the root and NOT the NACA 2218.09.  

We will get the correct data.

 

Fair enough.

 

Re for 90mph is about 5.2 million at standard conditions (chord 1.9m).   The 2218-09 model was at Re 3.1 million, (so underestimating Clmax at stall)  wheras Venturi's source above for 2212 is at 8.4 million (so overestimating it).

 

Although I do not know how much the difference is in either case, I would be inclined to take the difference 1.72-1.60= 0.12, then add prorata to Re difference, ie 1.60 + (0.12 * (5.2-3.1)/(8.4-3.1)) = 0.05 to arrive at a new guess at Re 5.2 = 1.65    

 

 

I hope you can find good data - especially if it also shows the critical AoA.

Edited January 2, 2017 by unreasonable

[Venturi]

If you want another 2212 CLmax value with different Re (3.2mil, CLmax 1.60) try:

 

National advisory Committee for aeronautics, Report 460: The Characteristics of 78 related airfoil sections from tests in the variable-density wind tunnel, pub 1935

 

[graph attached]

 

 

Additionally,refer to:

 

NACA report 824, Summary of Airfoil data, pub 1945

 

Which among much other data, gives whole-airplane CLmax for different airplane configurations... interestingly their models are exactly as for real aircraft. Including a P40 along with Liberator, Corsair, etc. I am having some trouble understanding the airfoils listed for these models. NACA 66,2-118 is listed for root and NACA 66(2x15)-116 is listed for tip.

 

But, regardless, it is interesting because one of the P40 model variants does not have a horizontal stabilizer... and its CLmax is reduced compared to the other variants of the P40 with normal horizontal stabilizer.

 

So if you were asking before if only the wing matters for airplane CLmax. There's your answer.

Edited January 2, 2017 by Venturi

[JtD]

It's a laminar flow airfoil, not the 22xx series. The model also features washout, which the P-40 did not have, and has a higher aspect ratio than the P-40. It's a model used for experiments, it has not much to do with the actual aircraft.

 

They did check a P-40 in the wind tunnel, but that was for drag reduction in the nose section, not for clmax determination.

 

It's great you're looking at all of this. I hope you find something I missed...after all, I still consider my best guess of propwash as an explanation as pretty lame.

[Venturi]

It still demonstrates the horizontal stabilizer can add to CLmax and emphasizes the need to look at whole aircraft CLmax.

I figured the CLmax value itself was not applicable since the airfoils are not the same.

...even though the airfoil position is. 

[unreasonable]

Having a read of NACA 824 throws up a few pieces of the jigsaw.

 

1) P 34. In discussing NACA 24,44 and 230 series with thickness from 12 to 24%, "Increasing the Reynalds number from 3*10^6 to 9*10^6 results in an increase in [CLmax] of approximately 

0.15 to 0.20

 

To redo my half joking calculation, 1.60 + (5-3)/(9-3)* 0.175 = 1.66    

 

2) P 37. "The predicted maximum lift coefficient for the wing will be somewhat lower than [..] the sections used becasue of the nonuniformity of the spanwise distribution of lift coefficient. The difference amounts to about 4 to 7 percent for a rectangular wing with an aspect ratio of 6."

 

1.66 * (1 - ((4+7)/2)/100) = 1.57

 

3) P.38 "Tests of several airplanes in the Langley full-scale tunnel show that many factors besides the airfoil sections affect the maximum lift coefficient of airplanes. Such factors as roughness, leakage, leading edge air intakes, armament installations, nacelles and fuselages make it difficult to correlate the airplane maximum lift with the airfoils used."

 

"...at approximately the same Reynalds number. The results show that the airplane in service condition had a maximum lift coefficient more than 0.2 lower than that of the model."

 

The examples you show for illustrating Clmax at R2 of 4.9 are of models.

 

This has all been posted before, but I repeat it here: it is not me saying that RL airplane Clmax is significantly less than the airfoil, wing or model plane number, it is NACA.

 

 

The fact that the tail-less one shows a lower Clmax is indeed interesting. The fuselage takes away and the tail gives back, in this case perhaps, or is there something weird going on? After all it does not have a vertical stabilizer either. 

 

Anyway, not sure that XXI and XXII are really a P-40 as such, although it is the right rough shape, the wing section data says NACA 66, a different numbering system, are you sure this is the same wing? The last two digits suggest not.

 

edit - composed offline, I see you already addressed the model wing issue.

Edited January 2, 2017 by unreasonable

[unreasonable]

It's great you're looking at all of this. I hope you find something I missed...after all, I still consider my best guess of propwash as an explanation as pretty lame.

 

I also wondered if it could be to do with the point made (P.34 IIRC) that in some circumstances where AoA is increasing, the Clmax can be higher than when AoA is constant? This presumably applies in a power off level stall test and maybe applies more for some planes than others.

 

The trouble is the source document NACA 618 is difficult to interpret, for me at any rate.  But if this effect is a) significant and b) not modeled it could account for some of the difficulties.  

[JtD]

It still demonstrates the horizontal stabilizer can add to CLmax and emphasizes the need to look at whole aircraft CLmax.

Yes, but then the reference area is bigger than the wing area already, because it includes for instance the fuselage section, which itself does not produce significant lift. This typically compensates quite well for added lift of the tail plane. Additionally, you'll have to consider that in a wind tunnel the tail assembly is typically fixed, while in flight the elevator will be angled upwards and induce a downwards force or at least less lift than in the wind tunnel.

 

Speaking of the reference area, since it is one of the figures in the equation I considered it possible that on the P-40 some odd figure was used, but judging from the drawings it in exactly chord at half span multiplied by span (minus the rounded tips). So it appears to be calculated as usual. Additionally, looking at the tail, the tail surface area is fairly large compared to some other aircraft, but then it isn't particularly outstanding or noteworthy. To sum it up, nothing special about the P-40 reference area (as was to be expected).

[JtD]

I also wondered if it could be to do with the point made (P.34 IIRC) that in some circumstances where AoA is increasing, the Clmax can be higher than when AoA is constant? This presumably applies in a power off level stall test and maybe applies more for some planes than others.

Absolutely possible, it was universally found that rate of change of angle of attack effects maximum angle of attack as well as maximum lift coefficient. Changing it quickly will result in higher angles of attack and clmax's than slow changes or steady condition. 'The airflow keeps sticking to the wing for a moment past the steady critical AoA' was the technical explanation. I recall having read a ballpark, but can't remember. It was big enough to be relevant, but wouldn't increase coefficients from say 1.5 to 2.0. It probably was also in one of the NACA reports you're currently looking into.

[unreasonable]

The source quoted in 824 is note 41 which is NACA 618, which turns out to be an analysis of a prewar highwing aircraft. Trouble is the pdf is not great quality, I have difficulty reading the graphs, and the "standard"  formula given (which the writers say underestimated their observed effects) looks horribly like calculus, which I have not used for about 40 years.

 

I have learned and remembered an awful lot from taking part in these discussions, but I fear that relearning calculus is probably a bridge too far for me.

[Venturi]

The actual value of the model CLmax is irrelevant other than the change in the value illustrated by the removal of the tail plane, since the airfoils are not the same as the actual P40. As I said.

 

We do not know the P40 design CLmax, but we DO know the empiric CLmax from reality as given by five different documents.

 

But the model itself is also in fact reality, just smaller scale and not quite the same as the real plane. It shows you that you can get a 0.1 increase in total CLmax or more from having the lifting surface of the tail present vs absent. It is possible this is a factor in this "mystery".

 

I additionally think that your argument that the elevator would induce a negative or minimal lift on the real plane in flight is not especially convincing sans example or math, in the presence of much evidence that says CLmax was much higher than 1.5, for which reason we do not know why. I think we need to see a whole aircraft design CLmax to account for this.

 

And I'm sure there are more documents to be found, too.

 

To add to the additional point of loss of CLmax in real aircraft vs models due to surface imperfections etc. I did some research on this. NACA airfoils have differing responses to imperfections based on their series characteristics. The NACA four digit series is said to be relatively resistant to any surface imperfections as might be expected in a airfoil with its leading edge characteristics (large leading edge camber and thickness).

 

By the way, critical angle of attack on the 2212 airfoil was 16 deg, not 14 deg as the devs state. The 2212 is not a 2215 as on the P40. Still.

 

Anyways, the stall speeds for the P40 are as listed in Multiple Sources. Those speeds are what the airplanes stalled at in real life, I am convinced of it.

 

Whether that is because of extra wing lift in the form of tail lift, or because of rearward cg inducing higher CLmax (yes even Boeing states this is a reality, aft cg increases CLmax - another reason aircraft CLmax is necessary and not airfoil), or for some other reason we might not completely understand some at a distance of more than 70 years since its design. It is moot because the plane stalls at 90mph.

 

This is a phenomenon often encountered in the almost infinitely complex world of human medicine. It is called agnostic science. It ultimately does not matter if we do not always understand completely WHY, if we can predict accurately what a outcome of a particular set of circumstances will be. And we have multiple historical and detailed sources all corroborating each other in this case, stating 90mph or even less for stall speed.

[Venturi]

When we have 96 mph in game. And I'm convinced now too to look at critical angle of attack.