CLmax Calculator

[-TBC-AeroAce]

just a question unreasonable did u use pounds mass or pounds force? I hate Imperial !

[Crump]

 

 

I guess my years of studying Aerodynamics to masters level was a waste then.

 

I was taught this specifically about Prandtl's Lifting line theory..

 

 

 

Considering that many of the FW190 design team studied under Prandtl, I would say that is the theory they used.  In keeping with that theory, they measured everything. 

 

The fact that under the same Reynolds number, the NACA 23015 airfoil returns a section CLmax of 1.7 and Focke Wulf used a CLmax that falls within the normal range of losses for Prandtl's methods....

 

Kind of backs that up.

[unreasonable]

just a question unreasonable did u use pounds mass or pounds force? I hate Imperial !

 

Everything is SI, the units given in the spreadsheet. You can input kph and kg, the spreadsheet converts to m/s and Newtons automatically.

[JG13_opcode]

What proportion of the 190 wing was 23015?  Isn't part of the wing 23009?

[Crump]

PS Compressible effects only start to happen at a Mach number of 0.3 and only really have a big impact approaching the trans sonic range. Air below Mach 0.3 is considered incompressible i.e the stalling speed of a ww2 fighter should not have any real compressed flow.  

 

Good general assumption but simplification rule of thumbs that do not apply at the level Focke Wulf used to derive that CLmax.  

 

There is no real agreement on that mach .3 either.  The NACA says mach .14 and above.

 

Which is why if you run the compressibility formula it comes out with a ~~.5 knot correct at stall speed if the aircraft is at 10,000 feet.  

 

That is one reason why I use IAS converted to EAS with a universal application of compressibility rules but that is another topic.

 

That is also why the RAE includes a compressibility error correct at the very low IAS of best climb speed which is very close to the speeds under discussion.

As to your independant CFD analysis...Contact David Lednicer.  He has already done a CFD analysis and has the 3D models already built.  You can check his data if you would like.

 

He is a nice guy, introduce yourself and I am sure he will respond to polite requests for a colleague.  He and I have corresponded several times.

 

What proportion of the 190 wing was 23015?  Isn't part of the wing 23009?

 

 

IIRC, the last 20% of untwisted wing section is NACA 23009. 

Edited October 17, 2016 by Crump

[-TBC-AeroAce]

Well using some basic siz i arrive at CLmax=1.52 at sea level and max takeoff weight! I did not adjust IAS for TAS as they will be almost the same at sea level! This using the in game numbers! The only way this is wrong is if the published stall speed was not at sea level and or max takeoff weight!!!!!!!!

 

I may have been a bit wrong about Lifting Line theory Blah Blah but all u need to calculate this is L=W=1.2*Vs^2*rho*CLmax*S

 

Seriously no more is needed!!! Stop over complicating it

 

Actually what I forgot is that I used to use a 2D aerofoil program to get curves (that did have a CLmax) and I would feed that in to a Lifting line Algorithm that calculated all the 3D curves Sorry I forgot the first bit was not lifting line. Look for numerical Lifting line Algorithm!

Edited October 17, 2016 by AeroAce

[Crump]

http://forum.il2sturmovik.com/topic/25722-need-some-help-community/

 

 

 

Well using some basic siz i arrive at CLmax=1.52 at sea level and max takeoff weight! I did not adjust IAS for TAS as they will be almost the same at sea level!

 

 

 

I got a Clmax of ~1.17 which is what the devs say they used.

 

Can you check it?

 

The test is aligned with the RAE results.  That is why it is at 3000 meters altitude, full fuel, and wing MGFF's.

[JtD]

Any updates on how discussions with devs are going?

Nothing new, I'm not expecting any information / meaningful discussion before the release of the next patch. Having been on the other side of the discussion, I can imagine where the priorities are, I'm happy to remind the developers every now and them, but I don't want to cross the line into annoying them.

[=EXPEND=13SchwarzeHand]

Maybe AeroAce took the stall speed that was published, but not adjusted after the patch and crump took numbers from actual flight tests (which would represent numbers after the patch)?

[-TBC-AeroAce]

Maybe AeroAce took the stall speed that was published, but not adjusted after the patch and crump took numbers from actual flight tests (which would represent numbers after the patch)?

 

Like i said I just need the correct speed, height, temp and weight! With that I can get CLmax with no problem and 100% certainty with out any other complicated BS!!!!!

 

A CLmax calculation (the way I do it) does not care about any other variable if you have the above as it will account for what is actually happening as it is based on observation. i.e. all this complex BS will be inclusive 

Edited October 17, 2016 by AeroAce

[-TBC-AeroAce]

Ok it seems like we all share a common ground which is trying to work out the flight behavior of a 70 year old ship! 

 

What I suggest is we all seem quite academic, is going quiet for a bit an producing a paper!

 

Let use see what we can do and let real peer review answer it!!!!!

 

Goes with out saying real references ......

 

Lets do this with a mounth deadline! And try make it so our methods are transferable to other aircraft!

 

What I would aim to do is use modern methods and historical reference to try give a full performance index of ww2 aircraft!

 

God knows maybe the results of such things combined might bear some fruit

Edited October 17, 2016 by AeroAce

[JG13_opcode]

i.e. all this complex BS will be inclusive 

 

Some of it will.  For power-on stalls you need to account for the inclination of the thrust axis, among other things. 

 

For power off stalls you still want to temper your observed/tested coefficient to correct for deceleration effects.  The standard way is to multiply by the quantity (x+1)/(x+2), where x = (observed stall speed) / ( 0.5 * chord length * deceleration rate)

Edited October 17, 2016 by JG13_opcode

[Crump]

 

 

Maybe AeroAce took the stall speed that was published, but not adjusted after the patch and crump took numbers from actual flight tests (which would represent numbers after the patch)?

 

 

I think that is exactly what happened.  Yes, I am seeking to replicate the conditions of the RAE test because they are known, set conditions that agree with Focke Wulf data.  I wanted him to replicate the conditions and see his data.  More data is good.

 

 

 

Well using some basic siz i arrive at CLmax=1.52 at sea level and max takeoff weight!

 

IMHO, if it was this in the game, it is close enough for government work and gives good agreement.   That is within 4% of measured data which is pretty good for a PC flight sim that as I understand it, tries to replicate CFD by using a computer model flying thru a virtual atmosphere.  It is not as easy as plugging in a value for CLmax and the FM will respond but rather a virtual "physical" environment.  So a little slack needs to be cut for the devs.  As long as the FM gives good agreement with the real world performance, I could give a hoot what the actual value they use.

 

In this case, CLmax is important for both sustained turning ability and even more important to the Focke Wulf, instantaneous turn performance. 

 

As the NACA 23015 airfoil was selected by designers because of FACT it delivered a high CLmax, the relative CLmax lineup needs to reflect that.  In other words, we should not find a ClarkYH having a higher CLmax than a NACA23015 for example taking into account Re.

 

 

 

Ok it seems like we all share a common ground which is trying to work out the flight behavior of a 70 year old ship! 

 

Agreed.  I appreciate your immediately owning up to the mistake in lifting line theory and not turning this into a 20 page discussion.  I am pretty sure some good things can come of further discussion.

 

Sounds good.

[Crump]

 

 

Some of it will.  For power-on stalls you need to account for the inclination of the thrust axis, among other things. 

 

That makes a HUGE difference.  It is one of the reasons why turning performance math is about relative performance and not specific performance.  The relationship of aircraft advantages will not change but the specific numbers will.

 

Here is the F6F Hellcat in flight CLmax measurements power on:

 

 

A Clmax of 2.0.

 

But look at the stall warning.  The pilot gets ONE second of elevator vibration before the aircraft dumps.  That is not a lot of warning at all.  That is the downside to the NACA 23015.

[ZachariasX]

Interesting effort, unreasonable.

 

So far, I have not seen any flaws in the math of your spreadsheet (it does work for me in the recent iteration of MS' blessing).

 

However, I think it does not actually do what you mean it to do. Looking at what I can do with it by computing CL, it appears to me that you compute the lift coefficient required for a certain state of flight. This is not CLmax of an airfoil. You can however infer if the required lift coefficient is possible and consistent with “an assumed CLmax”. You can throw in loadouts and airspeeds and see whether the crate should still flying or not. It would be still flying if your CLmax is below assumed Clmax and it would be crashing if it was above your CLmax

 

In short, with this, we can see that a postulated CLmax of 1.17 for the Fw-190 is a tad low compared to what is required to keep the plane airborne at known speeds and loadouts.

 

As a side note and it has been mentioned, the lift formula doesn’t reflect the atmosphere and Mach effects the results have to be taken with a grain of salt if you cross check your formula with real world data.

 

I sense the difference of calculating the lift coefficient and obtaining CLmax as the central point of why Crump dismissed your proposal, even though you took "his formula".

 

My main concern is that your effort is made “in the wrong part” of aircraft design.

 

CLmax is something that you get along by using a specific, known profile for a wing. In the optimal case, you have some guys (like at NASA) with a wind tunnel that publish these data along with the respective profiles. Best case, they put it in a wind tunnel and measured lift and drag vs AoA. Getting this corrected data along with your desired profile, you know among other things:

• how much wing you need to keep your aircraft in an optimal state of flight at the desired flight speed
• the suitable angle of incident of cord vs span of the airframe
• stalling angle
• The height of your landing gear (tricycle gear) to prevent tail strike upon landing
• Landing gear length / angle to make your taildragger flare out at three point attitude.

 

Without knowing this, there is no point in even going to the drawing board. Even better, when they measure the obtainable lift with the full size aircraft in flight and correct for factors such as local weather and flight speed/Mach effects. Still it should be consistent with what the guys back with the wind tunnel had. If it’s not, someone “made and error”.

 

One thing that you have to be really clear about, and I feel that is the wrong part of your spreadsheet, is that CLmax does not change, no matter how fast you fly or how much you load the aircraft.

 

You simply cannot have CLmax changing without changing the wing!

 

Here’s an example test question that may illustrate this point:

 

Flap selection at constant IAS in straight and level flight will increase the :

a) Lift coefficient and the drag
b) Stall speed
c) Lift and drag
d) Maximum lift coefficient (Cl max) and the drag.

 

 

 

d)

Why?

 

Because you maintain level flight at the same IAS speed. This translates into the wing having absolutely the same lift. If it had more, you’d be climbing. As you increase drag as well you have to open throttle a bit to compensate.

This is also why flaps are high lift devices. They increase the possible max lift of the airfoil. But that doesn’t mean you actually have this lift in steady state flight.

 

 

 

Taken together, lift may vary depending on state of flight and aircraft. CLmax varies with you varying the airfoil.

 

It’s been many years since I built RC aircraft from scratch myself and had to do that kind of math. Thank God for Hewlett-Packard. And lots of educated guesses on what does work and what doesn’t.

 

People who have no idea about all that have no idea about how many ways there are to mess up an aircraft design.

 

Even in flight simulations it is a problem. In my dear RoF, the SE5a and the Nieuports happily fire through the propeller arc with their Foster guns, just because the 3D artists misaligned the span of the fuselage.

 

I whish I had your spreadsheet 25 years ago…

 

[unreasonable]

Interesting effort, unreasonable.

 

So far, I have not seen any flaws in the math of your spreadsheet (it does work for me in the recent iteration of MS' blessing).

 

However, I think it does not actually do what you mean it to do.

 

 

Thanks for constructive critical feedback, Z.  Let me make a few points of clarification.

 

The point was simply to use the equation W = L = CLmax * 1/2* ρ * Vmin^2 * S  to see what combinations for the variables are possible, and the sensitivity of any one of the variables to changes in others.

 

If the equation is a good approximation to Vmin, then by definition you can solve for any of the terms if you know (or assume) all of the others.

 

So if you have empirical data for W, rho, Vmin and S, all of which can be directly measured to some degree of accuracy, then you can determine the CLmax - not of the airfoil, but of the whole plane. CLmax, on the other hand, cannot be measured empirically - it is a ratio.

 

If this is incorrect please explain why - it seems logically watertight to me. Edit: I think you have misunderstood - the only state of flight being addressed is at Vmin - which is the level flight minimum stall speed. At this point the CL=CLmax. Or are you saying the formula is incorrect? 

 

The next step is changing some variables to see what happens. I understand perfectly well that the CLmax of an actual aeroplane does not (cannot) change just because you add a few percent to the weight, for example - but the formula is simply a consistency test.  You cannot keep a constant CLmax and only change one variable - you have to change at least two. I built the toy to get an easy way to see how eg a 10% change in an assumed CLmax affected Vmin, other things being equal, or visa versa, and so on.

 

Edit - for clarification as this is where i think you have misunderstood my spreadsheet - I am not saying that changing a physical variables in a RL aircraft can somehow change the actual CLmax. I am saying that, if the equation holds, changing a variable in the equation requires at least one other change - so if you change say W, and hold rho, S and Vmin constant, CLmax has to change.

 

In an experiment attempting to determine an unknown CLmax, allowing it to vary in the equation is perfectly OK, since it is the unknown for which we solve. 

 

As for the RL case of the RAE test. My contention is that given the sparse and approximate evidence we have from that test, we can fit a whole range of reasonable estimates for the variables that are consistent with CLmax of anywhere between 1.60 and 1.20.  A common sense test for probability would put the most likely outcome somewhere in the middle of the range, but with very low certainty.

 

My conclusion, therefore, is that the test is simply not a very good one for the purpose of discovering anything because the conditions are so poorly defined and the measurements so obviously approximations. If someone had all of the specification for that test I have not seen them btw, simply a lot of assumptions.

 

This is important because the measurement for speed in particular is so sensitive. As should be obvious in principle, because Vmin is squared, small changes in Vmin require larger changes in another variable(s). 

 

The IAS we are given was "approximately 110mph". The other numbers in the report strongly suggest that this was simply eyeballed to the nearest 5mph. Try solving for CLmax using standard assumptions and then change the Vmin +/- 2.5mph. From a base of CLmax of 1.58 the number would now be 1.50 or 1.65 - that is just the range of error from the IAS approximation.

 

We do not know the PEC - in Crumps worked example he simply used the difference number that made his equation work. The number he came up with is not unreasonable, but I am sure if he had an actual PEC table for a real Fw we would have seen it and not had to justify the number by extrapolating off a graph of a Spitfire PEC.  We just do not know.

 

Consequently the assertion that the RAE gives good agreement with the estimates of 1.58 is unwarranted - as soon as you look at the sensitivities involved you can see that this is an example of specious rigour.

 

I hope this clarifies what I was attempting to do. If you still think I am in error please tell me why - I am treating this as a learning exercise!

Edited October 18, 2016 by unreasonable

[JtD]

Let me try to summarize:

 

If you enter the correct stall speed, correct weight and correct wing area, you'll end up with the correct clmax. There's no way around it.

 

The purpose of the spreadsheet is to show how sensitive the determined clmax is towards the input properties, because that's much easier shown by providing people with a tool where they can play with the numbers themselves, than it would be by posting lengthy texts on a message board.

 

And if you do play with the numbers, you'll find you can't determine the second digit of the clmax if you only have a rough estimate for speed and no idea about the weight of the aircraft, in fact, you can't even reliably estimate the first digit.

[unreasonable]

Horray!

 

JtD says it all in admirable concise terms unlike my rambling.

 

Edit - since there is always one pedant lurking, (in this case me )  "the correct stall speed, correct weight and correct wing area and correct air density" since the spreadsheet uses TAS for calculations.

Edited October 18, 2016 by unreasonable

[ZachariasX]

Unreasonable, I very much stand by my statement saying that your math is correct, but you did something you should be very careful of. You implied a relation between wing loading and CLmax, when in effect, there is none except the CLmax required to keep the aircraft airborne at given minimal speed.

 

As this equation had to be fed from real world data in a real world atmosphere at real world airspeeds, it is an inaccurate and extensive way in getting to know something that you know already. It can be useful, for instance in the case of you wanting to design an aircraft and you a given wing area, and now you are looking for profiles giving you little enough lift “to make it fly properly”. But that is more than awkward for a design procedure.

 

The fact that enough of us here are resistant to fundamental published figures of a profile makes us doing the most elaborate assessment to get to the same result.

 

The RAE was much less interested in the CLmax (in case of testing the Fw-190), because they knew the aircraft flies very well and once you know your stall speed you are pretty much there. They finally could make test flights to confirm a CLmax but you should consider their results more like an indication of how much effort they put in. The more effort, the closer CLmax has to be with Focke Wulf’s figure. The fact that CLmax is merely a side note in all extensive testing with the Fw-190 tells us the same thing.

 

In this sense unreasonable, you have an awkward way of using the formula. You usually know your weight and your CLmax, and you are maybe interested in the minimal airspeed. And this is how your formula (in the source document you mentioned) is advertised: “Also, it is interesting to notice that the minimal speed an airplane can have, can be calculated, if the maximum lift coefficient is known:”

 

“…if the maximum lift coefficient is known.”, I should write that a thousand times. (No, I don’t.)

 

Thus, what we are doing here is really awkward regarding how aircraft are designed. As I stated above, CLmax is such a fundamental value to your aircraft design, you have start out with that. Before you take the pen in your hand and the ruler (or: turn on your computer) to start drawing your aircraft, you know your CLmax! You absolutely cannot start when not knowing this. And where do you get it? From the guys who are designing airfoils. Then you look up profiles AND THE PUBLISHED CLmax (1.58 in our case) and you start designing.

 

Build an aircraft yourself. The first two lines that you’ll be drawing are the lines of the chord and the span of the fuselage that have to intersect at a specific angle. You cannot draw them when not knowing the angle at which they intersect. You get that angle by calculating the lift your wing will produce at intended airspeeds. Where do you get that info from? NACA charts, and lo and behold CLmax is published there as well. Now use your spreadsheet plus add the formula matching AoA, lift with the polar and you can draw you first two lines. Your entire aircraft is built on these two lines.

 

From that perspective, you had to come up with a pretty elaborate way to come up with something that in the best case matches the published documentations. And in your case, it does pretty well. It’s just.. why would you do that when you have the result anyway?

 

In your case here, It serves indeed the purpose that you (in contrast to people how just market opinions) made the valid effort to understand what’s going on. Us being mostly laymen on the topic, this is cool!

 

Once you have your aircraft, you can send it up with a test pilot and put it through it’s paces. If the pilot comes back with values that significantly differ from your numbers that you based your design on, you should maybe step back from the drawing board and go empty bins for your community. Doing so, you most likely won’t endanger anyone with your way of doing your work.

 

The case of the Grumman F6F where they assessed CLmax shows how closely design and real world data matches if you do your homework correctly. You also get a very good aircraft.

 

Taken together, I stand by my opinion that

 

1. Your math is correct.
2. This is a dangerous way to assess fundamental fight characteristics as it implies relations that don’t exist beyond your line of arguments (but laymen probably will assume).

 

Maybe I should have another signature:

“Published manufacturers values for CLmax are more correct that you’ll ever be!”

[unreasonable]

Z, I am not trying to design an aircraft! Merely understanding some sensitivities, while accepting the wing CLmax as a theoretical maximum for the aeroplane CLmax.

 

My recollection on the full size operational condition F6F wind tunnel test document was that the tested CLmax came in at a considerably lower figure than the airfoil, but the difference was reduced if gun ports and gaps in the skin were taped over etc, but I do not have that doc downloaded, perhaps you do. Other posters with aeronautical experience have also posted that RL plane CLmaxes are often in the region of 0.2 less than that of their wing's theoretical maximum. After all, the wing =/= the plane model =/= actual factory built plane with the usual imperfections.

 

Crump's response - and I guess this would be yours too based on your comments above, is that these differences represent problems to be overcome. I agree, but that does not mean that they were indeed overcome. The F6F test I allude to suggest that many of them simply cannot be. So what is the right answer - I do not know, nor claim to know.

 

Again, please note that I am NOT saying that you can use the RAE test to determine the CLmax - the whole point is that, given the flakey data and the sensitivities, you cannot.  But if you cannot use it to determine the CLmax, you cannot use it to show "good agreement" with some pre-determined design number either!

 

“Also, it is interesting to notice that the minimal speed an airplane can have, can be calculated, if the maximum lift coefficient is known:”  I wondered if anyone was going to raise that point.  The thing is that the formula is written as an equation, not as a function, so it is a logical necessity that we can rearrange algebraically and solve for anything. Or do you think that the formula is wrong and should be replaced by a function where the(whoops edit) Vmin can be derived from the other variables but not visa versa?

 

 As for the RAE not being interested in the CLmax - upto a point I agree, in relation to stall speed - the differences in speed (Vmin) we are talking about are hardly likely to make a plane drop out of the sky in power off condition to the huge surprise of its pilot. But as per the Fw controversy here, it really is not about the clean-wing power-off stall speed anyway - who even does it? - it is about how easy it is to get into an accelerated stall by exceeding the critical AoA.

 

Presumably the difference between 1.17, 1.40 and 1.58 will make a noticeable difference to the critical AoA (although perhaps not in direct proportion?) and this really would have been of interest to real as well as virtual pilots.

 

So who to believe? So many people noisily claiming to be authorities. The developers with their 1.17, one group with about 1.40, another with 1.58

 

As to anything I say here being dangerous - I had to laugh, it is not as though anyone who will make any decision will be at all affected by anything they read here - or more likely, do not read. From my own point of view this is simply an intellectual exercise to make sense of some of the information and noise about the topic.

 

BTW I do appreciate your taking the time to see what I was up to and addressing it sensibly: even when I do not agree it helps to sharpen understanding.

Edited October 18, 2016 by unreasonable

[ZachariasX]

Z, I am not trying to design an aircraft!

 

Well, in trying to understand CLmax, you're a very close bystander then

 

I wanted to outline the decision process of aircraft design to somewhat illustrate where the values come from just to put your effort into perspective.

 

Of course you can invert equations to solve for any given variable, but no one gets their CLmax from there. Only us flat earthers do that, much to the grief to forum moderators and devs. And why do we do that? Because we don’t believe what’s written in the spec sheets from the manufacturer.

(Just imagine how on earth would you create a 20-page thread as it happened  before on the topic if we believed in common practice in aeronautics as well as having an idea about it! Probably Bearcat had to start ice fishing to make use of all the extra time he had at hand.)

 

Now, forum wisdom is, “Ah, this is in the manual, but in the air, it’s different, of course, and I am only interested in the truth!”

 

So how would we for instance assess the length of a particular plane amongst people like us?

Simple: Get exactly the aircraft (not just the type!) and have it fly past a boffin that is sitting on top of a hangar operating a stop watch. Now when he repeatedly measures the time it takes when the tip of the spinner passes by until the time the end of the tail passes by and normalizes that time, adjusts the speed (don’t forget PEC!), then he can calculate the length of the aircraft. (You can do this with your car too, just measure the time it takes between the front and the rear license plate passing you.)

 

Also then JtD could say that you end up with the correct length; “there’s no way around it”. But you could also open the glove compartment and just look it up.

 

The devs face a different predicament. They not only need CLmax, they need the entire polar. And they have one. Alas, Murphy at work, they ended up with the one that would deserve a closer inspection. Because people do make errors. Especially in reports that few people had a look at. But if you have a source where millions of eye pairs had their look at, chances are low that you have nonsense.

 

 

Measured differences between "planned values" and "measured values" should not be significant and (IIRC) in the case of the F6F they were very consistent. Power on configuration however significantly increases CLmax; in the case of the F6F IIRC it goes up from ~1,6 to about ~2,0.

 

After all (using your formula), the difference between CLmax of 1.17 and 1.58 is roughly 4,5 m^2 of wing area in case of the Fw-190! (From a total of 18.3m^2 in a Fw-190.) That is really a different blueprint then, something like the Ta-152H compared to the ta-152C…

 

 

So, you’re doing the noble thing and were calculating correctly, but you are not supposed to use the equation that way, because there’s better ways to get to CLmax. Alas, in this situation it happened to be the only way…

 

All we need are some more (German) polars, other than the “French” one (Résistance at work there??). The “truth” we need is in there. But whatever the Germans come up with, it can’t be different from what NACA measured for the same profile. As VASERO calculations come up with 1,58 for a whole aircraft (no propeller), you have a pretty good guess what your polar MUST look like if some Krauts put a full scale Fw-190 into a wind tunnel. If it looks different, then they worked for the allies.

 

 

As to anything I say here being dangerous - I had to laugh

 

 

 "It's all fun and games until someone loses an eye!"

 

[unreasonable]

Z you need to find the right F6F wind tunnel test document - it was quoted/pictured in one of the previous Fw threads. It is the one where an actual plane "in operational condition" IIRC, was tested and got a CLmax of 1.40 ?

 

Cannot remember the exact number. Then they covered the gun ports, filled obvious cracks etc and tested again and got 1.50 - compared to the wing value of 1.60  

 

I am not talking about the document with the range of different wing models etc. I do not think it was a Japanese document.....

 

Clearly they thought the test was worthwhile, even though you obviously do not, so I would be interested in how that fits with your approach that the manufacturer is always right. 

 

I will see if I can find it, but searching this forum is not so easy....

 

I got +6.4m^2 addition to the wing to get 1.17 BTW, using the calculator and starting assumptions delivering 1.58        Fortunately that is the only variable in the RAE test we can be reasonably sure of.

[unreasonable]

Found it - I think this is the one: hope you can read it.  Posted by Crump in the "NACA data indicates" thread.  (We may have had our disagreements, but he does post a lot of interesting stuff).

 

Edit - you can get the whole report by checking out the NACA link JtD made in the same thread. (I have not: too hard for me).

 

Read the whole paragraph, not just the red box.

 

 

 

Edited October 18, 2016 by unreasonable

[Crump]

Let me try to summarize:

 

If you enter the correct stall speed, correct weight and correct wing area, you'll end up with the correct clmax. There's no way around it.

 

The purpose of the spreadsheet is to show how sensitive the determined clmax is towards the input properties, because that's much easier shown by providing people with a tool where they can play with the numbers themselves, than it would be by posting lengthy texts on a message board.

 

And if you do play with the numbers, you'll find you can't determine the second digit of the clmax if you only have a rough estimate for speed and no idea about the weight of the aircraft, in fact, you can't even reliably estimate the first digit.

 

Which is all the more reason why the engineering design teams measured CLmax, the result of comparitive studies, is pure aerodynamic gold.

 

That value is not the result of a single calculation or simple working of the Lift formula.  It is the result of all aerodynamic engineering knowledge of multiple engineers working on the design.  Once more, by December 1944, this is a very well estabilished design with tens of thousand of flying airplanes that have given no reason for that design team to reevaluate their results.

 

It is a fact it represents the Clmax of the clean wing.  More importantly it represents the fixed relationship of Angle of Attack to Coefficent of Lift for the design.

 

Assuming a properly constructed airplane with a proper finish....

 

If the airspeed measurement is accurate and weight is accurate....the result will be at the design teams CLmax.  Weight is very easy to adjust but accurate speed measurement is very problematic.

 

unreasonable has done a very good job of identifying the difficulty in accurately measuring weight and airspeed.  Having the design teams Clmax eliminates those errors.

Edited October 18, 2016 by Crump

[Crump]

Found it - I think this is the one: hope you can read it.  Posted by Crump in the "NACA data indicates" thread.  (We may have had our disagreements, but he does post a lot of interesting stuff).

 

Edit - you can get the whole report by checking out the NACA link JtD made in the same thread. (I have not: too hard for me).

 

Read the whole paragraph, not just the red box.

 

 

 

Aircraft CLmax.jpg

 

 

What do you think it is talking about?

 

The first part identifies issues if your data does not align with measured results.  It is pointing out the problems and solutions so that it will align.

 

The second portion above the outlined box refers the NACA 6 series airfoils and is specific to that airfoil family.

 

Measured results always take precendence over calculations when it comes to airfoil theory. 

[JG13_opcode]

 

 

It is a fact it represents the Clmax of the clean wing.

 

Do you have design documents that clearly and unequivocably state this?  All I've ever seen is that one chart with various coefficients, and never in the context of the entire report.

 

Moreover, not having intimate knowledge of 1940's German aero bureau design practices, how did FW actually assure that their design achieved the design target value?  Surely they did their own testing at some point prior to Chalais-Meudon.  Peenemunde, perhaps?

[unreasonable]

What do you think it is talking about?

 

The first part identifies issues if your data does not align with measured results.  It is pointing out the problems and solutions so that it will align.

 

The second portion above the outlined box refers the NACA 6 series airfoils and is specific to that airfoil family.

 

Measured results always take precendence over calculations when it comes to airfoil theory. 

 

I am referring to the text from the paragraph heading downwards. It is all perfectly clear English.

 

It states that there are a variety of things that make the airfoil CLmax hard to correlate with that of the aeroplane.  Now some of the things it mentions - roughness, leakage - you could classify as problems which can be fixed, but armament installations, gun ports and so on are an inescapable part of being a real fighter aircraft rather than a model. That was the airfoil - airplane correlation. 

 

(The airfoil - model correlation is not discussed here - let us just assume that it is good).

 

The paragraph then quantifies the model - airplane difference as "more than 0.2" based on empirical testing: again same causes for the difference not all of which can be "solved".

 

The VSAERO model with a CLmax 1.58 (?) is a smooth model, so you might reasonably expect the same issues to apply in which case you would estimate a range somewhere below 1.58.  1.40 -1.50? 

 

I do not know why this would be controversial TBH, since you made a rather similar comment in an earlier thread - I can find it for you if you like.....

 

So if I suggest that the plane CLmax should be less than the wing or smooth model CLmax I am simply reflecting what the NACA report says.

 

However, this is a little OT, I reposted that because Zacharias seems to have missed it, judging from his comments above.

Edited October 18, 2016 by unreasonable

[=EXPEND=13SchwarzeHand]

Moreover, not having intimate knowledge of 1940's German aero bureau design practices, how did FW actually assure that their design achieved the design target value? Surely they did their own testing at some point prior to Chalais-Meudon. Peenemunde, perhaps?

As far as the discussion about the in game implementation goes, again making comparability between airplanes a central point, AFAIK devs used the calculated (design target value) CLmax of the wing for the Yak as a clmax flight model target value. So if you talk real clmax for the FW it's a valid point. The discussion about implemented or targeted Clmax is a different story then.

 

What I'm basically saying is that whatever value you take as face value is irrelevant as long as you don't know what was used as a target clmax for other planes in the game.

Edited October 18, 2016 by II/JG17_SchwarzeDreizehn

[JtD]

Also then JtD could say that you end up with the correct length; “there’s no way around it”.

 

That's of course true, but the analogy is faulty for several reasons. Primarily, aircraft dimensions are nominal values you will find on every technical drawing, prepared by the engineers, to be built after by the guys on the production line. Engineering sends a document that gives the dimensions, construction will built it accordingly. That's not the case for lift coefficients. Lift coefficients are a result of various dimensions and loads of other factors. They are not specified on construction drawings the way dimensions are. Therefore, your assertion that you start with a known clmax and build the plane around is, is plain wrong. You start with an airfoil that you think will give you certain characteristics, and then you alter these characteristics by not building an infinite span, perfectly smooth, constant dimensions wing. Every chosen dimension, every angle, every transition radius, every gap, every hatch, every panel, every rivet, the surface finish including the choice of paint effects the eventual clmax. These effects can be calculated with todays knowledge and computing power to a reasonably accurate degree, but that was far beyond the capabilities and capacities of WW2 aircraft manufacturers. They made rough estimates, and then validated their results in tests. In fact, it is still done with the sophisticated models used today, just to a higher accuracy.

 

So, if you think you can look into the manual of an aircraft and look up the clmax - feel free to do so. We have 20 aircraft in game - try finding their maximum lift coefficients in the manuals. The only nation I know who occasionally did that, were the Soviets, who occasionally in particular early on included wind tunnel test data in their documentation. Dimensions, however, are found in about every single manual.

One reason you won't find it - unlike the aircraft length, the maximum lift coefficient varies a lot, because it is extremely sensitive to minor changes on the wing. Even after it has been determined sufficiently accurate in tests under specific conditions, the clmax on individual aircraft will vary to a degree that it does not make sense to put it into a manual. What you insert are figures relevant for piloting the aircraft, calculated with a sufficient margin of safety, taking the typical range of conditions into account.

 

Bottom line, if you want to accurately know the clmax of your aircraft with certainty, you need sophisticated instrumentation and testing - measuring stalling speed, weight and density.

Edited October 18, 2016 by JtD

[JtD]

Moreover, not having intimate knowledge of 1940's German aero bureau design practices, how did FW actually assure that their design achieved the design target value?  Surely they did their own testing at some point prior to Chalais-Meudon.  Peenemunde, perhaps?

There were several wind tunnels in Germany, probably the most famous one was (in fact still is) at Adlerhof in Berlin (Nice picture.). There were also centralized facilities at other places, most famously Göttingen and Braunschweig, plus some manufactures had their own wind tunnels, such as Junkers in Dessau (another picture). However, only a few of these tunnels could be used for full scale aircraft and it would be unlikely for a company to use the competitors wind tunnel or vice versa for a company to allow the competitor to use the own wind tunnel.

 

I don't know where Fw eventually went.

Edited October 18, 2016 by JtD

[JG13_opcode]

It states that there are a variety of things that make the airfoil CLmax hard to correlate with that of the aeroplane.  Now some of the things it mentions - roughness, leakage - you could classify as problems which can be fixed, but armament installations, gun ports and so on are an inescapable part of being a real fighter aircraft rather than a model. That was the airfoil - airplane correlation. 

 

(The airfoil - model correlation is not discussed here - let us just assume that it is good).

 

The paragraph then quantifies the model - airplane difference as "more than 0.2" based on empirical testing: again same causes for the difference not all of which can be "solved".

 

The VSAERO model with a CLmax 1.58 (?) is a smooth model, so you might reasonably expect the same issues to apply in which case you would estimate a range somewhere below 1.58.  1.40 -1.50? 

 

I do not know why this would be controversial TBH, since you made a rather similar comment in an earlier thread - I can find it for you if you like.....

 

So if I suggest that the plane CLmax should be less than the wing or smooth model CLmax I am simply reflecting what the NACA report says.

 

However, this is a little OT, I reposted that because Zacharias seems to have missed it, judging from his comments above.

 

Be sure not to confuse the Clmax you get from the airfoil charts with those you get from flight tests.

 

The data on the curve for a given airfoil, such as the 23015, is presented as what's called the Section Lift Coefficient, also sometimes called the 2D lift coefficient.  It is the Cl that would be developed by an infinitely long wing.  Usually denoted by Cl with a lower-case L.

 

On a real aircraft you have the so-called "3D" coefficient, usually CL with an upper-case L, and it is always lower than the 2D data.  The primary reason for this is due to spanwise flow along the wing.  Air spills outward along the wing, causing tip vortices and reducing the lift coefficients from the idealized 2D case.

 

The report you're citing and the 0.2 difference in CL observed is between a model of an aircraft and the full-scale aircraft.  The model is not the same as the airfoil - I see you're conflating terms in this quoted post.

 

 

There were several wind tunnels in Germany, probably the most famous one was (in fact still is) at Adlerhof in Berlin (Nice picture.). There were also centralized facilities at other places, most famously Göttingen and Braunschweig, plus some manufactures had their own wind tunnels, such as Junkers in Dessau (another picture). However, only a few of these tunnels could be used for full scale aircraft and it would be unlikely for a company to use the competitors wind tunnel or vice versa for a company to allow the competitor to use the own wind tunnel.

 

I don't know where Fw eventually went.

 

Awesome pics.  Agree about allowing competitors to use your tunnel; it just seems strange to me that there'd be no design validation data since in the modern world, experimental validation is an integral part of any design.  I assume there's something out there that I just can't find.  The only thing I've seen giving the 1.58 is the "widerstandsdaten von fluzeugen" image that Crump posts now and again.

 

edit:  I'm encouraged by the American test of the Hellcat that showed a clean-wing CLmax of 1.5 or so.  I'm cautious about using these results too dogmatically since the F6F had a much larger wing and there doesn't seem to be any atmospheric conditions listed in the hellcat report.

Edited October 18, 2016 by JG13_opcode

[3./JG15_Kampf]

Why is it so difficult to get documents CLmax? Because Devs use the test Chalais meudon? The wind speed was less that the plane could fly? What is the influence of this error in the CLmax to fw190? Someone in the community has any answers? Do we have to get a shovel and dig some fw190 in Turkia to take a test in the wind tunnel? This last joke lol

Edited October 18, 2016 by JAGER_Kampf

[unreasonable]

 

The report you're citing and the 0.2 difference in CL observed is between a model of an aircraft and the full-scale aircraft.  The model is not the same as the airfoil - I see you're conflating terms in this quoted post.

 

 

Not sure that I see your point, I simply summarized the paragraph, concluding:

 

As I said - "The paragraph then quantifies the model - airplane difference as "more than 0.2" based on empirical testing: again same causes for the difference not all of which can be "solved"."

[JG13_opcode]

Why is it so difficult to get documents climax? Because Devs use the test Chalais meudon? The wind speed was less that the plane could fly? What is the influence of this error in the climax to fw190? Someone in the community has any answers? Do we have to get a shovel and dig some fw190 in Turkia to take a test in the wind tunnel? This last joke lol

 

AFAICT the difficulty is that FW didn't actually do stall testing.  There were several problems with the test at the Chalais-Meudon tunnel, and JtD or someone else is more able to tell you what those are than I am, however there's also another report kicking around that someone posted screenshots of, stating that the folks at the C-M tunnel measured the angle of attack incorrectly among other things.

Not sure that I see your point, I simply summarized the paragraph, concluding:

 

As I said - "The paragraph then quantifies the model - airplane difference as "more than 0.2" based on empirical testing: again same causes for the difference not all of which can be "solved"."

 

I was just trying to help you out.  See the parts I underlined.  I guess what I'm saying is the 0.2 you're quoting is not applicable to 2D data:  you can't just take the 1.7@6M from a 23015 polar, subtract 0.2, and expect that to be valid for the real aircraft.

 

The 0.2 difference is likely due to scaling effects between the tunnel-sized models and the real aircraft; I have not examined Lednicer's work but I would not expect a properly-done CFD run to be subject to the same difference.  CFD can be very accurate but it is also notoriously finicky - people spend their careers doing PhDs in applying CFD to a particular class of problems, etc.

Edited October 18, 2016 by JG13_opcode

[Holtzauge]

First of all thanks for doing the excel sheet unreasonable. Works perfect for me and is quite useful for the purpose you have outlined in the OP and explained in the follow up posts. I’m sure many forum members will find it useful.

 

However, looks like Crump is back with his Fw-190 Clmax 1.58 mantra and I was actually planning to stay out of this because I know from experience that talking some sense into Crump is an exercise in futility.

 

The problem with Crump’s 1.58 figure for Clmax for the Fw-190 is that it’s never going to happen: the wing lift Clmax in the order of the 2D values only acts on the exposed wing area, not on the fuselage. This is also why the attentive observer will see that Lednicer's figure of the Clmax distribution ends at a bit over 0.1 of the semispan fraction. In addition, if you look at Lednicer’s result of the average wing Clmax distribution you will see that it is in the order of 1.4-1.45 not 1.58 which is only attained at the wing root station.

 

So multiplying the wing Clmax of 1.58 with the wing area of 18.3 sqm is wrong on two counts: First of all the assumed Clmax 1.58 is too high as witnessed also by Lednicer's figure which since it is a CFD simulation also includes 3D effects and is more in the order of 1.4-1.45 as the figure shows. Secondly it is wrong to multiply this wing based Clmax figure with the reference wing area since this area also includes the fictitious wing area inside the fuselage which actually contributes little to the lift. You can actually see this in the CFD figure: the low pressure does extend up on the fuselage sides but that only results in lateral forces that cancel and don’t contribute to lift. The area on top of the fuselage will of course contribute since the area projects in the right direction but as the pressure distribution scale shows it is far below the wing component. In addition the fuselage will have a negative impact on the wing lift close to the fuselage, effectively reducing wing lift here due to the fuselage interaction.

 

IIRC then RAE stated the Fw-190 wing areas as 177 sqft exposed and 197 sqft reference area. Going metric this equates to 16.44 and 18.3 sqm. So if we assume the wing lift Clmax in the order of 1.45 as per Lednicer the wing will roughly contribute L=q*Clmax*Sexp, where Sexp=16.44 sqm. However, since we are using Sref=18.3 sqm as a base (as is the convention) the wing will only contribute 1.45*16.44/18.3=1.3. Now of course the fuselage will add a bit to this but then again you need to remove the trim drag so a Clmax figure in the order of 1.3 to 1.4 is more reasonable. But 1.58? Simply not going to happen. This is also why I keep saying pushing this number is simply damaging our case because the developers can see we don’t know what we are talking about. Remember also that the Spitfire and Me-109 had Clmax figures using the same type of reference wing area as above at 1.36 and 1.4 respectively.

 

In addition, as JtD already pointed out the 1.5 to 1.6 2D Clmax figure from the Fw-190 report SchwarzeDreizehn was kind enough to provide comes from a structural analysis by Focke-Wulf so should not be confused with an aerodynamic analysis. Why do I say so? Well because I have worked for an aeronautical company and I have done aircraft structural analysis and even been paid for it. In fact I have even done the type of strip torsional analysis the report describes. And if Focke-Wulf worked anywhere close to how the company I worked for did it then I can tell you that that it is most likely the 2D lift figures you see in the report because when you do structural analysis the aerodynamics department will want to be conservative cause if a wing falls of you don’t want it to land in your lap because you provided them with a too low lift figure. Anyway the lesson here is you simply can’t use structural loading data to reverse engineer a Clmax.

Edited October 18, 2016 by Holtzauge

[unreasonable]

@ JG13_ opcode:  I appreciate that, and perhaps I was not sufficiently clear, but I do not think I am making that mistake. Sorry if I seem defensive - I have taken a fair bit of flak.......

 

Just to summarize my views to date:

 

I have said all along that CLmax for airfoil =/= wing =/= model =/= perfect real plane =/= typical operation plane. They may give "reasonable approximations" but how much?

 

Crump et al are suggesting that 1.58 is the CLmax for the FW 190, based on

 

1) Having seen the number quoted somewhere on a Fw document, 

 

2) The VSAERO 3D model,  

 

3) The RAE test results.

 

In my view, based on the calculator's sensitivity, the RAE results could be used to support almost anything, so I ignore them.

 

The 2D number Crump came up with was IIRC 1.70 - the 7% drop to the 1.58 CLmax he concedes due to the probable effects of wing shape - I am happy to take that as a given since I have no idea.

 

He also reads the VSAERO data as showing a model CLmax at 1.58 (approximately). I have no reason to argue with that so take it for now and move on.

 

My reading of the NACA report quoted above is that a further drop from smooth model to real plane is probable. The question is how much. You may be right that part of the 0.2+ in the NACA report is due to scaling, although they were done at the same(ish) Reynolds number, but it cannot all be that since the report notes that part of the gap can be close by sealing leaks.

 

NACA says "the  aircraft lift characteristics are strongly affected by details not reproduced on large scale smooth models".  So does VSAERO incorporate these kinds of details? I do not know, the pictures certainly did not look like it, so if not then a cut consistent with the NACA findings would be reasonable. How much exactly? No idea - but it just means clinging to 1.58, the whole 1.58 and nothing but the 1.58 seems a bit, well, unreasonable.  

Edited October 18, 2016 by unreasonable

[unreasonable]

First of all thanks for doing the excel sheet unreasonable. Works perfect for me and is quite useful for the purpose you have outlined in the OP and explained in the follow up posts. I’m sure many forum members will find it useful.

 

 

Thanks for the kind words!  As for the rest of your post, I will read it again, very slowly.... but I do get the big picture you are painting even if the some of the details are above my pay grade!

[Holtzauge]

Thanks for the kind words!  As for the rest of your post, I will read it again, very slowly.... but I do get the big picture you are painting even if the some of the details are above my pay grade!

 

You are to humble Sir.

[ZachariasX]

Hi Holzauge

 

Thnx for stopping by! With your comment #76 I guess I finally understand your point, and I think in general we are very much on the same page (Who would have though that would be unreasonable's spreadsheet?)

 

I think the basic miscommunication derives from that fact that with the CLmax of 1,58 I am talking about the airfoil, not the entire aircraft "handled as a single airfoil".

 

In this sense, my assumption is that when you create a wing, not what is by convention also included between the outermost tips of a wing, then one should be able to reproduce measured lift qualities as published alson with the profile.

 

Knowing that, I (as the designer) must imply that the wingtips as well as regions departing areodynamically from the "clean profile" will have an impaired lift.

 

You pointed out that roughly 2 m^2 of "included wing area (by convention)" do not really participate in giving lift, thus "diluting" the nominal lift value for the profile. This is logic enough that I didn't even think of raising the argument in the first place, as no aircraft (maybe Horten's design choices get close to it) will reach as whole the lift coefficient of  the infinite clean wing.

 

My point was that if you take your airplane and assess stall speed, the lift of the participating airfoil(s) (we usually have more than one profile section or washout) should not diverge much from published lift values. This means that you cannot expect in this situation that a single lift section can be used as the one giving you the CLmax corresponding to the full aircraft. Your I see your subtractions form the 1.58 as the way to correct for that, which is the thing you have to to.

 

Myself, I always took the 1,58 as it is published by Focke Wulf and NACA as the value for the plain vanilla airfoil not the whole aircraft. Same goes with the 1,58 from VASERO, to me it always looked that this is the computational result of the wing profile used not the whole aircraft. But it would help having the whole article. So, this is guesswork from my side. But it is at least as I understood it and the base of my previous arguments.

 

Regarding the whole aircraft I said

 

The devs face a different predicament. They not only need CLmax, they need the entire polar.

And with „the entire polar" I meant the whole aircraft in the wind tunnel. There you cannot expect, as you say, 1.58 and therefore you cannot use the published lift drag values for the profile of the infinite wing.

 

Holtzauge, you gave a plausible estimate for the whole aircraft. Good wind tunnel data should point in that direction and tell us the exact story.

 

And I would just find that the wind tunnel is the more elegant (but extensive) solution that just take the airplane up and assess stall speeds and correct for all factors that arise “in the open world”. But, as JtD says, correct values WILL give you correct answers.

 

So, let’s find good wind tunnel data for the devs!

 

I have my difficulties with the F6F article that unreasonable was kind enough to repost (I didn’t find it) and there they lament being 0,2 below published values. Now, if they included the fuselage section and things like that that hardly contribute lift then I find a deviation of 0,2 not really surprising, however if they calculated that their wing alone I mean the lift giving airfoil is 0,2 below specs, then that would be bad indeed.

 

 

So let me thank unreasonable again for his effort, as I think the discussion as well as the math is very instructive. You didn’t take flak, you raised a good discussion. This is a forum. What more do you want?

 

 

 

 

[unreasonable]

So let me thank unreasonable again for his effort, as I think the discussion as well as the math is very instructive. You didn’t take flak, you raised a good discussion. This is a forum. What more do you want?

 

   Nothing, it has in the end worked out to be very educational for me at least, and I am now glad that the thread was not shut down as I requested, purely out of frustration with the tone of Crump's comments.

 

Thanks to all for lots of specific and clear input. I am pleased that the calculator served as a tool to stimulate discussion - that is what I had hoped would happen.