FW-190A vs Spitfire, Me-109G, P-47D in dogfights

[K_Freddie]

Originally Posted by Gaston

I am amazed that you say that the Il-2 FW-190A is unlike what everyone says it is: Do you actually downthrottle to sustain tighter slow speed turns?
Gaston

Most planes, at certain speeds, can certainly turn better than the FW.
It all amounts to which tactics to use at which speed/situation.

From my experience the FW and ME109 have excellent slow speed characteristics, and on many occassions I have outturned other allied a/c by throttle variation.

At high speed the LH turn favours the FW, but at low speed the FW's niche is the RH turn. This is where the propwash and engine torque pulls the nose up and left, enabling one to use full pitch and rudder to bring it around a lot quicker than other aircraft. This is a fine balance, but very workable.

The FW is also excellent in the vertical, whether it's pulling out or up. It does wash a bit with lots of pitch, but this is very usefull as a deception. The 'washing' if used properly can enable the FW to change direction very quickly.

Using a combination of the above, can make the FW a very dangerous foe for any a/c...

Oh!.. forgot to mention the roll rate

[Gaston]

It is an interesting picture: The high speed turn preference to left is correct, as is the low-speed preference to right! (Though I had, perhaps wrongly, assumed that at low speed this reversal of preference was due to the deployment of flaps in low speed fighting: See Eric Brown's description of the landing configuration stall: The stall wing drop direction is reversed from a harsh left to a gentler right wing drop with flaps down: The side of the wing drop is the better turning side)

According to Closterman, FW-190As did not initially use their flaps to turn in combat, but later in the war they did, and it did tighten their turning ability...

At high-speed, use of flaps is very costly in speeds because the engine does not accelerate enough to compensate, at these speeds, for the extra drag, in addition to turning.

Combining downthrottling and flaps should, in real-life, allow out-turning at low speeds, in sustained turns, any major later war Western Front Allied fighter, with only the Spitfire being a question mark and maybe the P-38 at extremely low speeds. The FW-190A Western ace stated, despite his exclusive turning tactic: "I feared no other fighter in my FW-190A-8"

Turning, sustained or not, above 250 MPH MIGHT be about equal or better to the Me-109G if flying an early short-nose FW-190A-3/4, but DEFINITELY not for any later Anton variant (see A-5 test vs P-47:
http://img105.imageshack.us/img105/3950/pag20pl.jpg

). These later variants above 250 MPH IAS should be worse in turns and pull-outs than most other fighters.

So for late war Antons, the Il-2 picture at high speed is wrong: The Me-109G should be superior, as it should be superior to the FW-190A's vertical maneuvering except maybe for the first zoom from a very high speed. The Me-109G's absolute superiority in climb rate is what impressed the Soviets the most, and this made it an essential complement to the Anton on the Eastern Front at least.

Interesting note: Starting the turn fight at high speed, hoping to decelerate into the better lower speed while turning, is a dangerous idea in a Fw-190A... Maybe especially so for later longer-nose Antons: As the Fw-190A decelerates into its more favorable lower speed turn speed region (around 220 knots-250 MPH IAS) it abruptly changes pitch, which has to be compensated by the pilot instantly by pushing forward on the stick... Or it will stall: This is why the FW-190A Western ace described downthrottling long PRIOR to the merge: Decelerating from faster into lower speed while turning was risky... E. Brown also mentions this abrupt change in pitch, but did not find it dangerous on a short-nose Anton. It may have been worse on later Antons, as a few combat anecdotes seem to indicate...

The Me-109G was better in the vertical, but still inferior in zoom or dives to US fighters!

The Me-109G was better off downthrottling into very slow flat turns at 160 MPH against US fighters. The spiral climb might have helped, but it was very rarely used, so it must not have been convenient to use... Downward spirals are a bad tactic for all chased fighters...

The firepower and strenght of the Fw-190A made countering dive and zoom tactics by turning to face head-to-head into the attack worthwhile.

You say throttle variations, but once committed to lower speed turn fights, there is usually no upthrottling except maybe for catching up to a zoomer or a diver...

I don't think it is likely the FW-190A liked abrupt pitch transitions... You had to work the stick gently...

I will post later an English test of the FW-190D-9 that shows it to have far inferior maneuverability to the Anton, to the point of nullifying its climb and speed advantages over the Anton in the opinion of the tester...

Thanks for your picture of the FW-190A's handling in Il-2: It could be reasonably accurate for early shorter-nose FW-190A-3/4s...

Gaston

[David603]

I was intrigued by your description of the Anton as a low speed turner, so I decided to try some rough tests in Il2 to see if this supported the idea.

They are by no means perfect, but I did my best to keep conditions the same throughout the test and the results are averages of the good turns (ie if I stalled the aircraft or found myself more than 100m above or below my starting altitude I discarded the results).

I used the Fw109A-5 and the Spitfire LF MkIXc for the tests.

Technique used was a flat turn, using rudder where needed to keep the aircraft's nose up, and just trying to see the tightest turn that I could produce, regardless of speed.

Full throttle @ 500m (left turn)
MkIX 17.1 sec
A-5 22.4 sec

80% throttle @ 500m (left turn)
MkIX 16.0 sec
A-5 18.7sec

As you can see, while the Spitfire only gains 1.1 sec by downthrottling, the A-5 gains 3.7 sec, halving the gap from 5.3 sec to 2.7 sec, and the difference between a down-throttled A5 and a full throttle MkIX is only 1.6 sec.

The A-5 turns on a similar radius but at a lower speed, and is noticeably easier to stall, especially if you try to change direction once you have slowed down in the turn. Since both aircraft have props that rotate to the right, I didn't try repeating the tests with right turns.

[Ernst]

Originally Posted by David603

I was intrigued by your description of the Anton as a low speed turner, so I decided to try some rough tests in Il2 to see if this supported the idea.

They are by no means perfect, but I did my best to keep conditions the same throughout the test and the results are averages of the good turns (ie if I stalled the aircraft or found myself more than 100m above or below my starting altitude I discarded the results).

I used the Fw109A-5 and the Spitfire LF MkIXc for the tests.

Technique used was a flat turn, using rudder where needed to keep the aircraft's nose up, and just trying to see the tightest turn that I could produce, regardless of speed.

Full throttle @ 500m (left turn)
MkIX 17.1 sec
A-5 22.4 sec

80% throttle @ 500m (left turn)
MkIX 16.0 sec
A-5 18.7sec

As you can see, while the Spitfire only gains 1.1 sec by downthrottling, the A-5 gains 3.7 sec, halving the gap from 5.3 sec to 2.7 sec, and the difference between a down-throttled A5 and a full throttle MkIX is only 1.6 sec.

The A-5 turns on a similar radius but at a lower speed, and is noticeably easier to stall, especially if you try to change direction once you have slowed down in the turn. Since both aircraft have props that rotate to the right, I didn't try repeating the tests with right turns.

Nice. Do you have the track? How about using flaps combat in FW190? This is a nice test to do. May your FW turn was not sustained, but spitfire one did?

[AndyJWest]

Two questions, David603:

Were you trying to do a minimum-radius turn, or a best-rate turn? The results would be different. In combat, rate is usually more important than radius.

Are these sustained turn rates? Unless you can maintain speed, altitude and turn rate continuously, the results may be misleading. Even a loss of height of a few meters can make a noticable difference to results.

EDIT ---
I've been doing a bit of experimenting, using my prototype autopilot application (see http://forums.ubi.com/eve/forums/a/t...097#4121016097), and though I need to investigate further, I find it very difficult to believe a Fw 190 A5 will do a 360 degree sustained turn in 18.7 seconds, regardless of throttle settings. IL-2 compare suggests the best turn time will be around 24s, which is much more consistent with the results I'm getting at full throttle, and trying to turn at that sort of rate at 80 % throttle results in a rapid decay in airspeed. My autopilot is struggling to hold a smooth turn in these conditions (it was never designed to do this), but I'd be surprised if a human pilot could do much better - the plane is right on the edge of the stall.

As Ernst says, we need to see a track.

[Gaston]

The Spitfire Mk IX at full throttle is still 1.6 seconds faster than a FW-190A-5 at 80% throttle: This could be plausible if the FW-190A-5 does not use an optimal flap setting: The flap setting was critical to low-speed performance.

The Spitfire IX could not use a combat flap setting: 2 position only...

Soviet tests of a FW-190A-4 did show about 19 seconds in turn times, but it was displayed as 19-23 seconds, implying the difference in side of the turn(?).

Note the best Soviet Gustav time for that test, for a clean Me-109G-2, WAS 22 seconds, so right there you have the FW-190A out-turning the Me-109G to one side at least...

Me-109F was 20 seconds.

Note that the longer-nose A-5 was said by the Soviets to shave a second off these figures, which would give about 18-22 seconds.

I assume these real-life tests were all done at full power without flaps...

At partial power it could be all these aircrafts do not do much better in sustained turn TIME, but much better in RADIUS, which gives some advantage in sight lead also...

A FW-190A-8 in sustained low-speed level turns, flaps down, at 70% power(?), can gain nearly 180° per 360° on a P-51D at full power riding on the edge of a stall: If flaps up full power for the P-51D means 23-24 seconds, then the FW-190A-8 with the broad-blade prop could be as low as 16-17 seconds to one side (it was the right side with flaps down), to reverse a tail position in 2.? X 360° turns... The A-8 was said to be better than all previous FW-190As in low-speed maneuverability, especially with the broad-blade prop.

Of note is that the FW-190A riding on the edge of a stall requires the use of the ailerons to catch the stall's wing drop: At low speeds this favoured the choice of the longest chord of three different types of ailerons that could be used. The FW-190A Western ace in AH's forum described adding "spacers" to the hinges of his longest-chord aileron choice, to increase low-speed wing-drop "catch" performance further, this of course at the expense of aileron leverage and performance at high speed...

This choice of his was specifically described by him as being exclusively aimed at low-speed turn performance...

These aileron hinge modifications could explain the out-of-the-ordinary low-speed turn performance he mentions for his P-51D shootdown: 2.? X 360 to reverse a tail position on the deck, the P-51D almost stalling in front of him...

Not clear if the aileron hinge extensions were a field modification, or availabe as a kit...

Interesting tests for the in-game figures... Thanks!

Gaston

[AndyJWest]

Soviet tests of a FW-190A-4 did show about 19 seconds in turn times, but it was displayed as 19-23 seconds, implying the difference in side of the turn(?).

Nope 19-23 seconds means exactly what any engineer/test pilot would expect it to mean - somewhere between the upper and lower limit, but not accurately measured. Or if it doesn't, it is down to you to provide evidence why, not just assume it means what you want it to.

And as for 'catching the stall's wing drop' with aileron, this is nonsense if you are talking about a sustained turn (along with airspeed, turn rate and altitude, AoA must be constant so either the wing is stalled or it isn't), and dubious as a means to recover from a stall anyway. If a wing is stalled, down aileron is going to make it worse.

Even with the luxury of an autopilot, and no worries about structural/engine failure, fatigue from G forces, instrumentation errors and the rest, practical experience with the few tests I've run tells me that any measurements of turn rates need to be taken with some scepticism. Out of curiosity, does anyone actually know how turn rate was measured? The compass would be useless, and I'm not sure a gyro would be much better - they tended to tumble with extreme manouvering.

[JtD]

Originally Posted by AndyJWest

Nope 19-23 seconds means exactly what any engineer/test pilot would expect it to mean - somewhere between the upper and lower limit, but not accurately measured. Or if it doesn't, it is down to you to provide evidence why, not just assume it means what you want it to.

Actually the numbers are 22-23 seconds. So 19-23 means the author has manipulated original data to prove his point, nothing else. Don't blame the engineers / test pilots.

[David603]

Originally Posted by AndyJWest

Two questions, David603:

Were you trying to do a minimum-radius turn, or a best-rate turn? The results would be different. In combat, rate is usually more important than radius.

Are these sustained turn rates? Unless you can maintain speed, altitude and turn rate continuously, the results may be misleading. Even a loss of height of a few meters can make a noticable difference to results.

EDIT ---
I've been doing a bit of experimenting, using my prototype autopilot application (see http://forums.ubi.com/eve/forums/a/t...097#4121016097), and though I need to investigate further, I find it very difficult to believe a Fw 190 A5 will do a 360 degree sustained turn in 18.7 seconds, regardless of throttle settings. IL-2 compare suggests the best turn time will be around 24s, which is much more consistent with the results I'm getting at full throttle, and trying to turn at that sort of rate at 80 % throttle results in a rapid decay in airspeed. My autopilot is struggling to hold a smooth turn in these conditions (it was never designed to do this), but I'd be surprised if a human pilot could do much better - the plane is right on the edge of the stall.

As Ernst says, we need to see a track.

To be honest, I think I shouldn't have put 18.7 as the Fw time. Although the Fw can do between 2-3 360 degree turns with times that are very close to produce this data (18.7 is the average of many turns), it only maintains this turn rate at 80% throttle by losing speed and tightening up the turn, so after 2-3 turns at this rate you run out of speed and either stall out or have to leave the turn to regain speed.

As such it is not a true sustained turn, although 2-3 turns would still be quite useful in a combat situation. By contrast, the Fw @ full throttle and the Spitfire at either setting can maintain the times listed for as long as you wish, so those are real sustained turn times.

Bear in mind I said this was a rough test, and by no means perfect, if you can come up with a better technique I will be happy to accept the figures produced by it.

In the meanwhile I will try to get a good track of the Fw turning @ 80% throttle (will change altitude from 500m to just above the deck so you can more easily see that it is a horizontal turn).

[AndyJWest]

Thanks for clarifying, David603. That sounds more in line with what I was finding. If you are going to run tests, you really need to specify as much as you can about the conditions: map, fuel load, difficulty settings etc. As has been shown on the Ubi forums, even changes in the 'wind and turbulence' setting can have a measurable effect on performance, in supposedly calm conditions.

If you can provide a track, it is possible to analyse it later using DeviceLink, which is often more useful than trying to make measurements in real time.

I need to do more work with my autopilot to be able to test turn rates accurately, by the look of it, but I think I'm getting somewhere. I'll post a demo of what I've achieved up to now if anyone is interested, though the plane still lacks a little stability.