"...But how am I to know a good flight model from the bad?"
"You will know... when you are calm, at peace, passive. A pilot uses the flight model for positioning and defense, NEVER to attack."
"But tell my why I can't..."
"No, no! There is no why."
Seriously though, I read through this whole conversation and I noticed two trends: Cherry-picking data, and magical thinking. There have been more than one person to do this in this conversation.
Cherry-picking data means that you look at accounts, pick the ones that support your opinion, and analyze them with the exclusion of other, conflicting reports such as actual, physical hard data about the aircraft themselves - their mass, wing area, engine power, thrust, control forces, critical angle of attack, stall characteristics etc. etc.
Magical thinking is a bit more complicated and is a continuation of the cherry-picking process. Since pilot accounts describe plane A in a way that contradicts our understanding of flight dynamics, and what its performance should be, the pilot accounts must be right and our understanding of flight dynamics OR the parametres of the aircraft A must be wrong.
This is magical thinking: "To match pilot accounts, magical properties must be added to the aircraft so that no conflict occurs."
However, when it comes time to explain what physical phenomena facilitate these properties, things become a bit difficult because mechanical physics is, on the most part, quite well known collection of knowledge and adding "unknown flight characteristics" smells incredibly fishy.
It's like trying to explain Grand Canyon if the preconception is that the world is only eight thousand years old: Since the world can't be billions of years old for these sediment layers to slowly form, there must have been a global flood that accumulated all these sediments at one go! This has, actually, been used as
evidence for a global flooding... along with the fossil record... but I digress.
While I wouldn't wish to accuse anyone of dogmatic faith in scripture or pilot accounds, I can't help but notice certain tendencies in the argumentation on this thread.
In reality, I would not automatically consider any pilot accounts - either combat- or test pilot - "reliable" accounts of the characteristics of the aircraft itself, but rather specific accounts of what this pilot did in this particular situation against that pilot in that plane and how it happened to work out. Additionally, you're assuming that the undoubtedly highly skilled, experienced and intuitive pilots had the ability to put their experiences, feelings, and fly-by-the-seat-of-their-pants gauging of different planes into objective format.
It's absolutely right that pitting aircraft against one another in evaluation test flight might not reveal what plane is the "best one" at some specific flight regime. But comparing the accounts of combat pilots is not exactly reliable either, because - you know -
the accounts are written mostly by surviving veteran pilots who were probably both highly experienced
and biased toward their own particular aircraft. This type of bias is perfectly normal and expected from humans put into situation where their life hangs on the performance of a machine and how well they can handle said machine. It's easy for the most rational person to ascribe almost mythical qualities to such a machine, especially if it happened to bring them home time after time.
Regardless. Ignoring the possibility of hidden variables, I'm sure we can all agree that the laws of physics apply on aircraft regardless of their type, manufacturer or pilot. The pilot can bring them closer to the edge of their performance, and possibly do tricks that other pilot can't, but as far as raw performance goes, the capabilities of an aircraft are
fairly straightforward. I say fairly because aerodynamics is a really complicated science and there are often surprises even from quite simple designs, but nevertheless some simplifications still hold true.
If we look at a situation where an aircraft is turning, there are two key variables that describe its performance. Ones is turning radius, and the other is turn rate. Transient turn rate (and turn radius) mainly depend on the g-loading of the airframe at critical angle of attack, but transient turns bleed energy - the aircraft's energy state is not at equilibrium.
In sustained turn, the aircraft is banked and has elevated angle of attack, produces lift toward the direction of the turn, and maintains constant airspeed (producing constant lift), constant turn rate and constant turn radius.
The first requirement is constant airspeed. The aircraft typically produces quite high amount of drag with high angle of attack, so the aircraft's engine must produce the thrust to offset drag so that the airspeed does not further reduce. Since thrust is (mostly) a function of how much air the propeller can move, this is pretty much a function of propeller efficiency and engine power: The
work done by the drag forces must be equal to the
work done by the propeller's thrust.
Therefore: An aircraft with more thrust can maintain higher angle of attack and therefore its sustainable rate of turn is better. Inversely, however, an aircraft with less drag at optimal angle of attack might well be able to sustain higher turn rate at lower engine power, so this is not at all clear-cut parametre. A simplified expectation would be that an aircraft with more engine power should be able to sustain higher turn rate assuming other variables are identical.
The constant turn rate and constant turn radius mean that the aircraft can produce a
constant centripetal acceleration, ie. a force accelerating it toward the centre of the turn. While vast majority of this force is basically the lift of the aircraft's wings and control surfaces, at very high angles of attack the propeller's thrust is also partially directed "inward" of the turning circle. In fact, if you're dancing at the edge of stall - critical angle of attack being let's say 15 degrees - as much as
25% of the aircraft's thrust is directed "upward" relative to the local airflow around the aircraft. While the main function of the propeller is to, should we say, propel the aircraft through air, its effect on "hanging on the prop" in high AoA turns should not be neglected.
That said: Most of the centripetal force is accounted by the lift produced by the aircraft's wing at that specific angle of attack.
Centripetal
acceleration, then, is the
sum of aerodynamic pressure forces divided by
mass of the aircraft (a = F/m).
Note that most talk about wing loading is, at best, a gross simplification that assumes the wing's characteristics are very similar between two aircraft.
Wing loading, as a parametre, is merely the mass of the aircraft divided to the total surface area of the wing. While somewhat indicative of the general characteristics of the aircraft - especially within a specific class of aircraft such as WW2 fighter aircraft - there can still be
radical differences in performance. Wing loading does not determine turn performance. Total lift produced by the aircraft does, and wing area is only one part of that equation.
The other part of the equation is the wing's airfoil profile. The two most influential factors are the wing's camber and chord thickness. Those are the ones that affect the wing's lift coefficient most. Other variables tend to affect the wing's critical angle of attack and lift-to-drag ratio.
I found an interesting site which includes references to the approximate airfoil shapes of quite a few aircraft:
http://www.ae.illinois.edu/m-selig/ads/aircraft.html
For example:
Focke Wulf Fw 190A-8
Wing root: NACA 23015.3
Wing tip: NACA 23009
Chord thickness ratio (root/tip): 15.3% - 9%
Wing loading: 241 kg/m² / 48.4 lb/ft²
Power/mass ratio: 0.29-0.33 kW/kg
(I couldn't find A-5 wing loading but if the A-5 variant's loaded mass is known it would be trivial exercise to find out).
Focke Wulf Fw 190D-9
Wing root: NACA 23015.3
Wing tip: NACA 23009
Chord thickness ratio (root/tip): 15.3% - 9%
Wing loading: 238 kg/m² / 48.7 lb/ft²
Power/mass ratio: 0.30-0.35 kW/kg
Supermarine Spitfire Mk. V
Wing root: NACA 2213
Wing tip: NACA 2209.4
Chord thickness ratio (root/tip): 13% - 9.4%
Wing loading: 133.5 kg/m² / 27.35 lb/ft²
Power/mass: 0.36 kW/kg
Supermarine Spitfire Mk. IX
Wing root: NACA 2213
Wing tip: NACA 2209.4
Chord thickness ratio (root/tip): 13% - 9.4%
Wing loading: 159.8 kg/m² / 32.72 lb/ft²
Power/mass ratio: 0.42 kW/kg
For some reference, here are some chord thickness ratios at wing root and wing tip from some other prominent fighters:
F4F Wildcat: 15.3% - 9%
F6F Hellcat: 15.3% - 9%
F4U Corsair: 15.3% - 9%
(Identical airfoil profile with FW-190, tremendously lighter wing loading...)
La-5/F/FN/7: 16% - 10%
MC.205: 18% - 9%
Bf-109 G-6: 15% - 9% (NACA 2315 mod - NACA 2309 mod)
Hurricane: 19% - 12.2%
Typhoon: 19% - 13%
Tempest: 14% - 10%
MiG-3: 14% - 8%
Yak-1/7/9/3: 14% - 10%
Now what has this got to do with anything on this thread?
Well, aside from
wing area, the chord thickness affects the amount of lift that the wing produces.
That means that if you have same nominal wing loading on two planes, but the other one has thicker wing, the one with thicker wing is producing more lift. For example, the Hawker Hurricane has wing loading of 121.9 kg/m² which is only 11.6 kg/m² lower than Spitfire Mk.V's wing loading - but the thicker wing would produce more lift, which pretty much explains why the Hurricane turns so much better than Spitfire both in-game and by the pilot accounts:
More lift means more centripetal acceleration.
More centripetal acceleration means higher sustained turn rate.
But wait, that's not all! Higher chord ratio means that the wing also produces more drag. So that means the aircraft will need more engine power to offset the work done by increased drag force, or it will travel slower through the sustained turn - and reduced airspeed reduces available lift which reduces centripetal acceleration which reduces the sustained turn rate.
True to this assumption, the Hurricane really does lose its energy quite fast in hard turns and while its sustained turn is still better than Spitfire's, it is also really slow at that point.
Spitfire's wing is slightly thinner than the FW-190 wing at root, but slightly thicker at the wing tip; however, as the FW-190 wing was trapezoid and Spitfire wing elliptical, the root chord of Spitfire has much bigger significance and it can be said that Spitfire's wing is overall thinner than FW-190 wing.
What this means is that while FW-190 A-8's wing loading is as high as 241 kg per square metre and Spitfire Mk.IX's wing loading is 160 kg per square metre, the Spitfire doesn't in fact turn 33% better than the FW-190 A-8. Instead the difference would be somewhere between 0%-33% in unpowered turns. In sustained powered turns, the thrust of the propeller will also affect things as it is directed "outward" from the turning circle, and the thrust/weight ratio also comes to play - and Spitfire IX has a lot more power.
So how can we make any sense of any of this?
The answer is: It's really hard, really complicated, and
the vast odds are against armchair pilots trying to think how each of these parametres affects each other.
As far as simulation accuracy goes... if you have a good flight dynamics model, and you have the correct values for relevant terms for the aircraft, they should have
close to historical performance characteristics.
Will they behave like the planes historically did? Hell no. First of all most pilots in IL-2 tend to operate very differently from the paranoid survivors who checked their six every twelve seconds, kept track of everything that happened around them, and aspired to never put themselves in a position to get shot at, while putting themselves in a position where they could shoot at the enemy.
As has been said in this discussion, bullets flying tends to motivate men and mice. In a real combat situation, a lot more factors affect the outcome than just the performance of the aircraft. For example, FW-190 offers much better all-round visibility than the Spitfire (or Bf-109 for that matter). Additionally, when Bf-109's and FW-190's were operating together, it would make perfect sense for the FW-190's to fly at lower altitude and Bf-109's at higher altitude
because the BMW engine was inferior at higher altitudes compared to the DB engine of the 109.
I am reasonably certain that no FW-190 pilot would have wanted to enter into a sustained turn fight with any allied fighter aircraft if they had any other choice.
As the facts may be, they often may not have had any other choice as the quality and amount of material and pilots on the Allied side grew and Luftwaffe was run over by P-51's, P-47's, Spitfires, Tempests etc. etc.
Whether or not any of these designs were objectively "better" at turning than FW-190 or Bf-109 didn't really matter much at this point. The Luftwaffe fighters' main task was to go after the bomber fleets, and they would have tried to avoid combat with Allied fighters as much as possible.
And even so: As the war progressed, many Allied fighter pilots flew all their sorties with no enemy combat, while Luftwaffe pilots engaged in combat almost every sortie. Is it a big surprise that pilot accounts of the capabilities of individual aircraft may have been skewed by the other factors affecting the set-ups of the individual fights? I think not.
If there's something I've learned while studying physics it is that complicated interactions of a veritable horde of parametres is not always quite exact science and often the only reliable data comes from experimentation.
Sadly, the majority of war-time fighter aircraft have been destroyed or otherwise rendered flightless. The best solution to the question of evaluating flight performances would be to construct new production planes of each fighter, and then test their performance.
Needless to say this may prove somewhat expensive, so in lieu of that, the best alternative is to look at the sources of data, form some sort of opinion on what data to use, and then use it. In the end, this is first and foremost a game. As much as I would love to know that the aircraft we fly on bit sky are accurate representations of their real world counterparts, I'm willing to accept that sometimes we can't get what we want quite as much as we want.
With that in light, as long as I'm having fun and I can find viable ways to use different aircraft in the game, I can live with possible historical inaccuracies.
Finally, an anecdote.
I have encountered a couple FW-190 pilots in this game that would outmaneuver a Spitfire flown by me. I have, on occasion, done so myself. But whether "outmaneuver" is the same as "out-turn" is anyone's guess...
End of story. Take of it what you will, ignore the rest if you wish.