F4U turn performance, flaps, the real plane, etc (discussion)

[Noir]

Which one of them is Nuke?

Im SirNuke, not NUKE

[GScholz]

lol, I think I asked you in the MA too. Sorry about that.

[nrshida]

I mean aircraft types not people types.

Also convincing, not convinced. Apologies.

[Badboy]

Guys

Sorry to come back to this discussion so late, but I think I can offer an explanation that doesn't depend entirely on mathematics. While that may be a little less rigorous, it will be more accessible for those who don't have the patience or interest to do the math.

Firstly though I'd like to pick up on a few points:

I thought I understood that Fowler flaps increased the turn rate and given they are reasonably large on the Ki-84 I'm rather sceptical about the comparable rates.
How can the DPS be the same with a good deal more lifting surface available, longer chord, a better camber and a slightly higher thrust to drag ratio?

Fowler flaps will always increase the turn rate, but it is very important to be specific about what kind. The high lift coefficient and additional area for Fowler flaps will have the most dramatic influence on instantaneous turn rate and I think that anecdotal sources that describe that increase in maneuverability are referring to instantaneous turns, because the influence on sustained turns is far less dramatic and influenced by several other factors that can't be so easily attributed to the flaps alone, I'll explain that in more detail later.

I gathered from other discussions that the efficiency of the prop increased as the airspeed decreased.... Have I understood those aspects incorrectly?

That's the wrong way around, the efficiency of the prop does not increase as the airspeed decreases, if you check out an efficiency curve you will see that the efficiency decreases with airspeed. The amount is different for every engine/prop combination and while that is not important in a discussion about instantaneous turn rate, which Fowler flaps will always provide, it is important with regard to sustained turn rates because the best sustained turn rate occurs when the thrust and drag are in equilibrium, so the more thrust the higher the sustained turn rate.

It seems to me Brooke found an error, and tweaked his report to coincide, but I don't recall.  It may be worth looking for.

I checked the math when Brooke first posted and pointed out a few errors that were quickly corrected. There were updates since then but other than rather clumsy notation, the final draft is fine and the conclusions are valid.

For anyone who doesn't want to do the math the following written explanation may be helpful.

If you want to predict the sustained turn rate for an aircraft the factors you would focus on are the thrust, the drag and the lift. When flaps are employed, all three change. The lift increases due to the change in curvature of the wing and in the case of Fowler flaps due to the increase in area. The parasite drag increases, but how much depends on several factors that makes comparison between flap type difficult without all the data. Most notably, the size of the flaps as a fraction of the chord length and span, the section profile, the deflection and if they are slotted, to name a few. Depending on those factors the parasite drag can swing either way, but parasite drag has less influence during a sustained turn where it is the induced drag that has the greatest influence on the resulting sustained turn rate. Obviously the thrust is also an important consideration but if we set the parasite drag and thrust to one side for a moment and consider the influence of the flaps on induced drag we can find an explanation that will help to clarify much of what we see in Aces High.

Induced drag is caused by the wing tip vortices resulting from the span-wise lift distribution. It has long been understood that an elliptical lift distribution is most efficient and gives the lowest induced drag and thus the best sustained turn rates, so we should consider the impact that flaps have on the spanwise lift distribution. Take an untapered, untwisted rectangular wing planform. Disregarding interference at the root and the finite length the lift will be similar at each station and so the span-wise lift distribution will be clearly not elliptical. Now if you lower the inboard trailing edge flap, increasing the lift over the inboard part of the wing the span-wise lift distribution will be closer to the elliptical than before, so under otherwise similar conditions the induced drag should actually decrease. This can be verified using data from NACA reports which shows the improved efficiency. On the other hand, if you do the same thing in the Spitfire, you move away from the ideal lift distribution, which it already had due to its planform, so the efficiency factor is reduced and induced drag increases. The Spitfire achieves an elliptical spanwise lift distribution by having an elliptical planform, but you can achieve it by more subtle means, such as wing taper ratio, change of thickness and change of wing section. For those reasons it is difficult to predict the effect that flaps will have on induced drag based on the flaps alone. Some configurations integrated with the wing will move closer to or farther from the optimal distribution when flaps increase the lift, so some will reduce the imposed drag and some will increase it, with a corresponding effect on the sustained turn rate. It is difficult to be more specific than that, other than to say that in almost every case where information can be found so that the performance can be checked properly, Aces High appears to be spot on, and that is probably so because we are all using the same data.  

The evidence for the F4U having reduced induced drag and thus increased sustained turn rate is fairly convincing but only because I can run the calculations using more sophisticated thrust and lift models based on data from various NACA reports that produce an even closer match with Aces High than those carried out by Brooke. Unfortunately, I don't have the same wealth of information for the Ki84. Despite spending some time checking my sources today, I failed to find key data for the Ki84 that would be necessary to run calculations that would give further insight.

Meanwhile I think it is worth pointing out that the Ki84 does see a significant increase in sustained turn rate when flaps are employed. My own tests show an increase in sustained turn rate for the Ki84 at both flap settings with the maximum increase for one notch of around 1dps. While that is a smaller increase than the best that can be achieved with the F4U, it is worth pointing out that despite that, the Ki84 still has a superior maximum sustained turn rate than the F4U-4, and that is still very impressive.

Hope that helps

Badboy

[mtnman]

Just to play Devil's advocate a little further, what about the FW190?

I'd like all of the planes to fly as realistically as possible, and that includes the 190's.  

When it comes to my " too easy" comments, I feel that if my then-7 year old son with no aero-knowledge and no experience can easily figure out how to take off, fly, and land the planes in AH without me even being home I'd call them "too easy" to be realistic.  Heck, I had it set up for rudder pedals too, and he couldn't even reach them, lol.  Now, he wasn't landing on the runway, I'll give you that...  And it took him a little bit to figure out how to raise and lower the gear, slow down to deploy it, and change to a different plane than the F4U in the hanger (which is why he didn't immediately snap the gear off after he took off).

Fun to fly, and modeled pretty dang realistically, yes.  Hard to fly?  No.

And I'm not talking about flying/fighting effectively in them.

The F4U is about the only plane I'm interested in from WWII.  Some of the others hold a passive interest for me, like the P51, P40, B17, and B25.  Not enough interest for me to fly them much though.  The rest of the planes in WWII I find just plain aesthetically ugly.  My only interest in them is as targets...

That doesn't mean I think the F4U should out-perform or be "better" than the rest.  I'd prefer it to be one of the more difficult actually.  

But when it comes to realism, I want measurable facts and data to be the basis.  If the F4U is wrong, fix it!  But honestly, I think the pilot model is skewing things so much that we can't see the reality of the modeling.  If the pilot were modeled more completely, I think we'd see the F4U doing what we hear and read about in the history books...  

You can't have an accurate model representation of what was in reality a combination of man and machine if only the machine is accurately modeled.

My contention is that a real WWII pilot would never be able to coax full "down and dirty" performance out of a real F4U.  Even off by himself somewhere, with no threats apart from the ground, he's still going to be limited by what his body can do.  

Of course, HTC may not feel like devoting that much time and effort to modeling something that wouldn't be seen as an exciting addition by many if not most players, and that would limit the fancy (i.e. exciting) flying that we see now to something that many would see as "less" exciting.  Similar to the argument that removing the engine management requirements makes the game more fun for the majority...

[mtnman]

A very clear explanation...
Badboy

Thank you sir! 

I appreciate the information presented in a way that I can easily understand and visualize.

 

[mtnman]

I found a few things in there which makes me think you didn't look very hard. Within the first few pages of the report is the statement that aileron efficiency diminishes rapidly with flap deployment. Now, I have not flow the F4U much, but I would expect that to be readily tested.

As Saxman mentioned, those weren't production-type flaps.

Also, the tests described were run March/April of '42, the first production F4U-1 was test flown in June of '42, and that report was issued in October 1942.  The first -1's saw combat in February of '43.  Those early F4U's went through quite a few changes, including aileron improvements and various drag-reduction improvements.

In AH, we don't see the F4U-1 all that much either.  We normally see the -1A and -1D, which had been modified enough from the -1's to warrant a different designation.

All that led me to conclude that that report wasn't all that applicable to our discussion. 

I did find it particularly interesting though for a different reason...  I've been building a Radio Controlled F4U, and since I'm not real happy with the kit I've been modifying it heavily to make it much more scale.  I found the diagrams helpful and saved quite a few of them into my research folder.

[nrshida]

I'd like all of the planes to fly as realistically as possible, and that includes the 190's. 

I didn't say you didn't, just that the departure characteristics of the AH Fw190s could be fairly described as viscous and without warning.

The F4U is about the only plane I'm interested in from WWII.  Some of the others hold a passive interest for me, like the P51, P40, B17, and B25.  Not enough interest for me to fly them much though.  The rest of the planes in WWII I find just plain aesthetically ugly.  My only interest in them is as targets...

Corsair pilots wear Sailor suits and sing 'In the Navy' 



Sorry some more questions for you Badboy, if you've got time.

Fowler flaps will always increase the turn rate, but it is very important to be specific about what kind. The high lift coefficient and additional area for Fowler flaps will have the most dramatic influence on instantaneous turn rate...

Let me just clarify this point, in this context you mean instantaneous because they produce so much lift, and hence drag that there is insufficient thrust to overcome the latter and hence the speed must drop which in turn reduces airflow hence lift, hence turn, which is why it is instantaneous and not sustainable?

That's the wrong way around, the efficiency of the prop does not increase as the airspeed decreases, if you check out an efficiency curve you will see that the efficiency decreases with airspeed.

But whichever way round you state it there is more thrust available at low speed than high, the opposite of a jet, correct?

If you want to predict the sustained turn rate for an aircraft the factors you would focus on are the thrust, the drag and the lift.

Isn't the weight of the aircraft also significant, obviously it only changes slightly and slowly in the same airframe, but in comparing different designs you must also consider the weight, the Corsair being rather 'big boned' compared to a lot of the other fighters.

The evidence for the F4U having reduced induced drag and thus increased sustained turn rate is fairly convincing

Do you mean the flaps were essentially making lift (and hence drag) just about up to the level of thrust? Again isn't the thrust related to the weight, a heavy aircraft needing an awful lot more power in the first place to ensure there is excess power left over to provide for the extra induced drag? So when you say reduced drag you mean compared to other aircraft rather than a Corsair with its flaps retracted, obviously?

Unfortunately, I don't have the same wealth of information for the Ki84. Despite spending some time checking my sources today, I failed to find key data for the Ki84 that would be necessary to run calculations that would give further insight.

What exact information do you require? If the information isn't available then what did HTC use to determine the flight model?

[mtnman]

I didn't say you didn't, just that the departure characteristics of the AH Fw190s could be fairly described as viscous and without warning.

Sure, I guess they could...  But since I know that, aren't they also predictable?

The problem with "viscous and without warning" is it means different things to different people at different times, with different levels of experience, in different situations, and different opinions.  Way too variable and subjective.  Interesting, but not useful.  Also, too easy to "cherry-pick" the terms that best fit a preconceived notion or agenda. 

Just because objective evidence is difficult to find, doesn't justify reverting to subjective evidence, IMO.  In my world I'd lose my job if I fell into the trap of using subjective rather than objective evidence.

I'd refer you back to post 106...

Corsair pilots wear Sailor suits and sing 'In the Navy' 

[nrshida]

I don't summarily dismiss evidence from a reputable source just because there is no statistical quantification.

Even in Formula 1 drivers and engineers discuss characteristics in qualitative terms before the engineers make alterations. I don't imagine them dismissing the impressions of the drivers because they don't have measurable data to back their impression. That would be unwise and wasteful. Of course they are motivated.

The official Corsair training film states: "Stalls are abrupt, and preceded by very little warning in the nature of buffeting".

I've just had a quick flap about offline to see for myself. The second Corsair model, 25% fuel, gear down just as configured in the training film. I've probably flown the Corsair less than a dozen times and then usually only as required in duelling matches.

My impressions:

"Stalls are predictable and docile with plenty of communication from the aircraft, the left wing wants to drop, you can feel it announcing after the buffet is well established, the departure it is not abrupt but progressive, you can even push it over the edge and recover with a little fettling of the rudder, throttle and medium stick inputs. I had the EB6 computer open the speed fluctuated to as low as 64 m.p.h. indicated in a nose up attitude in full control".

Quite a contrast.

[Badboy]

Let me just clarify this point, in this context you mean instantaneous because they produce so much lift, and hence drag that there is insufficient thrust to overcome the latter and hence the speed must drop which in turn reduces airflow hence lift, hence turn, which is why it is instantaneous and not sustainable?

Yep, I think you've got it. Anytime you pull enough g that you can't sustain your airspeed and altitude your turn rate is instantaneous and not sustained, you are flying in the area of the envelope above the Ps=0 line, in the region of negative excess power, or negative Ps. 

But whichever way round you state it there is more thrust available at low speed than high, the opposite of a jet, correct?

Well it matters which way around you state it, you had the efficiency v speed relationship the wrong way around and since you were asking if you understood it correctly I simply clarified that point for you. You have the trust v speed relationship the right way around.

Isn't the weight of the aircraft also significant, obviously it only changes slightly and slowly in the same airframe, but in comparing different designs you must also consider the weight, the Corsair being rather 'big boned' compared to a lot of the other fighters.

Not if you are only considering the effect of employing flaps on a particular aircraft, because as you lower the flaps the weight of the aircraft doesn't change so you can treat it as a constant rather than a variable, which is what I was doing. Of course if you want to know a value for sustained turn rate or do a comparison with another aircraft then of course the weight becomes a factor. The full range of data and aerodynamic ratios needed are as follows:

• The air density for the altitude which can be considered constant in a sustained turn.
• The gravitational constant, which actually varies slightly depending on location.
• The prop efficiency, which will change as the aircraft speed changes.
• The zero lift drag coefficient, which increases as flaps are lowered.
• The induced drag coefficient which may increase or decrease depending on the final lift distribution.
• The weight of the aircraft.
• The wing area which may change depending on flap type.
• The engine power at that altitude.
• The maximum lift coefficient for the given configuration.

The aerodynamic ratios needed are:

• The wing loading.
• The power loading.
• The maximum lift to drag ratio.
• The coefficient of lift at the maximum lift to drag ratio.

From that data you can calculate the maximum sustained load factor, the maximum sustained turn rate, the radius at that value, and the speed at that value. If you want me to run through an example calculation just ask.

Do you mean the flaps were essentially making lift (and hence drag) just about up to the level of thrust?

Not quite sure what you are asking here, but yes the flaps increase the lift and that can enable you to produce higher instantaneous turn rates.

Again isn't the thrust related to the weight, a heavy aircraft needing an awful lot more power in the first place to ensure there is excess power left over to provide for the extra induced drag?

No, thrust isn't related to the weight. They have an opposite effect on the sustained turn rate, additional weight being detrimental and additional thrust being beneficial, but they are independent of each other.

What exact information do you require? If the information isn't available then what did HTC use to determine the flight model?

I'm sure the information is available somewhere, I just couldn't find it when I wanted it the other day when I had the urge to do the calculations. I didn't get very far into it before I got stuck trying to find the reduction gear ratio for the Ki84 engine/prop combination for my thrust model. Unfortunately I was distracted before I had exhausted my search and haven't had an opportunity to resume. I always enjoy researching the aircraft, but as I browse documents and books I like to allow myself to wander in whatever direction my reading or interest takes me. It's an inefficient way to research and gather information, but each and every time is like an adventure that I never grow tired of.   

Anyway I hope something here helps...

Regards

Badboy

[nrshida]

Yep, I think you've got it.

 

Not quite sure what you are asking here, but yes the flaps increase the lift and that can enable you to produce higher instantaneous turn rates.

I meant is it the case that the designers of the Corsair just hit the 'sweet spot' of a lift to drag ratio which the engine was still able to overcome with spare thrust?

No, thrust isn't related to the weight. They have an opposite effect on the sustained turn rate, additional weight being detrimental and additional thrust being beneficial, but they are independent of each other.

I meant net thrust remaining once all of the drag (more weight, more drag) has been dealt with. Does momentum play a part by the way?

I'm sure the information is available somewhere, I just couldn't find it when I wanted it the other day when I had the urge to do the calculations. I didn't get very far into it before I got stuck trying to find the reduction gear ratio for the Ki84 engine/prop combination for my thrust model. Unfortunately I was distracted before I had exhausted my search and haven't had an opportunity to resume. I always enjoy researching the aircraft, but as I browse documents and
books I like to allow myself to wander in whatever direction my reading or interest takes me. It's an inefficient way to research and gather information, but each and every time is like an adventure that I never grow tired of.   

Maximum power is at 2900 r.p.m. for Ha-45-11 and Ha-45-12 with 1800 and 1825 hp respectively. The Ha-45-21 makes 1990 hp at 3000 r.p.m. The propeller reduction ratio is 0.5 on all three with an electric pitch control but it is acknowledged that the propeller diameter was too small for the engine this restriction being caused by the length of the main undercarriage. I told them to use a five blade prop 

Anyway I hope something here helps...

Yes thanks for your time Badboy, I actually now feel I have at least a basic understanding of all the elements and I think other people will benefit too. Once I can find the handbrake I'll be able to takeoff and try it all out 

[jeep00]

Nothing substantial to add to this barring thanks (nearly) all for some interesting information on the most beautiful aircraft in the game. I nearly always up F4U's, and most frequently I use the -1 because it is just so dang much fun to fly. contrary to one particular comment, I NEVER drop gear in dogfights, nor do I see it often, and never from other better sticks than me. I agree, it is a clear sign of an upcoming kill, and I'm not that good. The -1 can absolutely hang in a kow and slow 1v1 with many aircraft, even with me at the stick. But it is ready to flop and fall from the sky, it has taken a lot to learn that sweet spot. It is handy knowledge that is more feel that gauged though. And since 1v1 becomes 5v1 when on the deck like I am, invariably I meet an unpleasant end. So thanks again for this wealth of knowledge, food for thought. fwiw, this version is harder than it was in 2007 or so. It sure seems it to me, again just a feel thing.

[mtnman]

I don't summarily dismiss evidence from a reputable source just because there is no statistical quantification.

Even in Formula 1 drivers and engineers discuss characteristics in qualitative terms before the engineers make alterations. I don't imagine them dismissing the impressions of the drivers because they don't have measurable data to back their impression. That would be unwise and wasteful. Of course they are motivated.

The official Corsair training film states: "Stalls are abrupt, and preceded by very little warning in the nature of buffeting".

I've just had a quick flap about offline to see for myself. The second Corsair model, 25% fuel, gear down just as configured in the training film. I've probably flown the Corsair less than a dozen times and then usually only as required in duelling matches.

My impressions:

"Stalls are predictable and docile with plenty of communication from the aircraft, the left wing wants to drop, you can feel it announcing after the buffet is well established, the departure it is not abrupt but progressive, you can even push it over the edge and recover with a little fettling of the rudder, throttle and medium stick inputs. I had the EB6 computer open the speed fluctuated to as low as 64 m.p.h. indicated in a nose up attitude in full control".

Quite a contrast.

The second Corsair model, 25% fuel, gear down just as configured in the training film.

On my way out of town, but had some time to review the training film, search for some test documents, and do some tests.

http://www.youtube.com/watch?v=kpxyyLQ7u7g

My findings were that the AH F4U-1 stalls pretty dang close to the way it does in the film, but is more difficult to recover than it should be.  The F4U-1A may arguably stall too harshly in AH, and is almost definitely too difficult to recover.  But I think I'm ok with that  

The key is to do the test in as close to a realistic manner as possible, and in order to get the results we all think we should see we need to do the tests similarly to what we see in the training film...

First, it's important to start with the correct plane, right?  The plane shown is an early F4U-1, without the spoiler on the right wing.  That spoiler was added later because the left wing had a tendency to stall before the right wing, especially under deceleration.  The F4U-1A, and all later models had the spoiler (mounted on the leading edge of the right wing, just outboard of the gun ports).  

This addition was said to effectively cause both wings to stall at the same time.  We should expect the later model corsairs to have a somewhat "gentler" departure as a result...

We also know it's a -1 due to the canopy, the functional top three cowl flaps (training film at :52) which were sealed in later models, and the instructions to burn the fuel from the left wing first.  The "heavy controls" is also an early F4U trait (that was later improved).

Can you imagine that they would have test flown the plane, and made that training video with 25% fuel?  Not me...  I loaded 100%, but am confident you could load somewhat less and see the same result (I also fly with 100% in the arenas).  Adding the fuel adds weight, increasing the wing loading, which effects the stall...

Next, air density also effects the stall...  So, how high were they in the training film?  We should be in similar air density to expect similar results.  I launched at 15K, but also performed the tests down to 12K.  I'm fairly certain they were higher than that in the film.

I also trimmed for level at @325, and flew with Combat Trim off, to get that out of the equation.

So, get up there, drop your gear, and drop 3 notches of flaps (the training film stall you're referring to had 30 degrees of flaps).  This first stall is POWER OFF, per the training film.  Slow it down, and bring the stick back, and...

It'll drop it's left wing, and put you almost on your back.  Very similar to the training film!  I didn't "crank" the stick back hard, but also didn't go super-gentle (like I would normally fly).  What did he do in the film?  They don't say, but he was intentionally trying to show a stall, so would he be super gentle with it?  I doubt it.

However, the training shows an easy recovery after each stall!  The AH F4U stall is more problematic to recover from.  I purposely just centered my controls to see what would happen, but the nose doesn't come down as nicely as it does in the training film...

So, the departure I found to be convincing; the recovery less so.  Too difficult in AH compared to the film.  Just my subjective opinion of course.

Next, they show the POWER ON, CLEAN stall, which is very similar.  Again, AH does a good job of replicating the film.  If anything recovery again seems easier in the training film though.  If I use the advice given in the film "prompt positive action results in a normal recovery", I can live with the AH stall.

The POWER ON, LANDING CONFIG stall at 4:20 looks pretty dang gentle too!  Wow!

Now, part of that may have been the different position of the cowl flaps.  According to the test documents I found, opening them adversely effects the stall characteristics.  They're closed in the POWER ON stall, which is noticeably more gentle, and open in the POWER OFF stall, which is more harsh.  

The flaps are closed in AH; should we have a gentler stall? (See the test documents I'll link at end of post).

http://www.4shared.com/file/6CQ669gT/F4U-1_Stalls_0000.html

I also did the tests with the F4U-1A, with very similar (identical?) results.  Where does the stall strip benefit come into play?  I don't see it, for some reason?  Oh well.

http://www.4shared.com/file/mV-JAob1/F4U-1A_Stalls.html

I found a few other things VERY interesting in the training film.  

At around 16 minutes in they say the maneuvering flaps (25 degrees vs. the often-stated 20 degrees) are for use to increase maneuverability at slow speeds.  The warning they give is to not dogfight with more maneuverable planes!  Why?! They call it a bad tactic, but give no warning AT ALL about flight characteristics due to being slow and flaps out.  Hmmm, food for thought.  He actually says the flaps are designed for increased maneuvering at slow speeds AND to assist takeoff and landings.  Not the other way around...  More food for thought.

If we put emphasis on the "Stalls are abrupt, and preceded by very little warning in the nature of buffeting" quote, shouldn't we give equal weight to this one??

I like these ones too...

"There's nothing about the corsair that good pilot technique can't handle" and "There's plenty of sting in her guns".

Want to see a "floating corsair" outside of AH?  Check out the training film at 19 minutes in.  Low, slow, flaps out, and low throttle.  I bet if he gave her full throttle he could sure make 'er dance!

Here's a link to some "Confidential" test reports I found.  Pretty neat.  http://www.wwiiaircraftperformance.org/f4u/f4u.html

Definitely some info on stall characteristics, and some direct comparisons with the 190A-4, F6F, P51B, P47.  I haven't had time to go through them all, but they're definitely interesting reads.

Another interesting tidbit I found was the undesirable cockpit layout mentioned several times.  It gives credence to the RL vs. AH pilot theory I mentioned earlier.

[nrshida]

Yes you are right it was the wrong model. I took the early one up with 100% fuel, three notches, gear out and power off, as you suggested. The stall speed was higher as expected and this one it felt like either wing wanted to go. In all honesty, it still felt very communicative to me. Trouble started way after the buffet was fully set in. I was abusing it pretty hard and was able to work it to the point were I could maintain full back pressure on the stick (yes to the stop) and used the ailerons and rudder to keep the wings roughly level.

I then repeated the test with the second model, this one seems a little less docile, but conversely responded quicker to inputs.

Recovery in both models even when it snapped right over I found easy, had it back in under control and upright in around two seconds (estimated).

I didn't recheck the film to look at the cowl flaps but both the first and second model in AH have those modelled exactly the same graphically. Both the skins I have have that green paint inside. The cowl flap in line with the antiglare strip is closed on both, the other ones open (top ones at least).


<Shrug> I guess we'll just have to agree to disagree, clearly we are having a very different qualitative experience in AH. I'm not sure why that is.