P-47 Maneuverability or Lack Thereof

[354thFG_Panda_]

36 minutes ago, TacticalOni said:

Still watching the changelogs with bated breath. 

It boggles me to go from the P-51, to the P-38 and have them react in a believable manner only for the P-47 to fall so short. With half stick deflection it feels like I'm trying to hang the plane on its prop. 

Hopefully we will see some information soon!
 

It's now spring, a large update and a patch has dropped. Hopefully they have the time and people to look into it before moving to the next deadline. ?

[ACG_Cass]

10 hours ago, ZachariasX said:

On the whole, I feel the plane should handle more direct, but the penalty from flying slow should be higher. There should be no „helicoptering“ around, but you go the way of the brick instead if you did try to torque the aircraft. No amount of power can save you past coffin corner.

I wish this principal was applied to all planes in IL2 to a more extensive degree with the P47 certainly to a more exaggerated degree.

[357th_KW]

5 hours ago, 86th_Rails said:

I feel like this is probably true or that everything else is over performing to where it seems like it’s under performing. I’m still hoping the devs are looking at again and we’ll get an update soon. 

 

This likely plays a big role - the 109s and 190s (among others) can WILDLY exceed their real world critical mach numbers in a dive with no repurcussions, retaining full control while doing so.  In the case of the 109, the test aircraft had to be modified to remove half the aileron travel, and add an additional trim tab, and even then the trim had to be locked at a very specific setting to even be able to perform the dive and then recover.  And even with that, they added an ejection seat as well, because they were so close to the edge in their testing.  And all that to achieve Mach .8 (the unmodified aircraft was in an aileron reversal around .75-76 and basically uncontrollable and in danger of shedding a wing) - in game with none of those mods, and paying no attention to trim (you have so much elevator authority, it just doesn't matter) you can blast up to .87-89 easily and recover.  So the overperformance of the P-47's competitors are a big factor with regard to dives.

Edited April 6, 2023 by 357th_KW

[354thFG_Panda_]

On 4/6/2023 at 3:53 PM, theRedPanda said:

It's now spring, a large update and a patch has dropped. Hopefully they have the time and people to look into it before moving to the next deadline. ?

 

The topic was brought up on the official discord and the Devs responded. There isn't enough time or manpower to look into it

 

https://discord.com/channels/1015218984313946173/1030463581625528411/1093566677658120263

 

Edited April 7, 2023 by theRedPanda

[Dr_Molem]

"And mind that any changes there, although necessary, won't make the Jug into a nimble fighter, it simply wasn't one."

 

This was not needed at all, actually.

 

I don't think anybody here is waiting to use the P-47 in slow speed turnfights against 109s.

[LukeFF]

Stay on topic, or the discussion gets locked. One comment has already been edited due to snide comments toward the developers; let's not make it more than that.

[TacticalOni]

It seems like there's a disconnect between what we are asking for and what the developers are interpreting as that question. 

We all know the P-47 wasn't a nimble fighter, thats not what we are asking about. 
It's about the way it flies. As I said before, "almost there." 

None of the responses in here are asking for the P-47 to be a nimble fighter, it's all (or at least mostly) about the pitch response, engine power, and energy retention. This thread has just about every test a guy can do on a fighter to determine the break between how the real thing acted and how the one in IL-2 deviates from that. 

I have no experience in engineering a flight model, but I feel that once you find a couple guys to do it, it feels like it shouldn't take long thanks to all the info dumped into this thread. Lets hope that someone finds the time to look into it soon! 

[332FG-CoolHand]

Any word on this per chance?

[354thFG_Panda_]

6 minutes ago, Coolhand3011 said:

Any word on this per chance?

 

I asked again on the official discord but it looks like they are really really hard at work with the upcoming planes announced for WWI, WWII and the "new project". They have the desire to do it but no time in the schedule.

Edited May 9, 2023 by theRedPanda

[Letka_13/Arrow_]

P-47 has been considerably improved in the last patch in terms of its maneuverability, however now it seems that it got some strange landing characteristics. With full flaps and landing gear down the becomes very tail heavy and during powered approach requires a lot of down trim or stick push. I am basically touching down at 85-90 mph with stick in neutral or slightly pushed for a three point attitude. I have flown P-47 extensively before and at touchdown it required nearly full stick aft for a three pointer, now I am landing with centered stick (trim neutral) and elevator at neutral which is very far off the real data, which show that the elevator should be nearly fully deflected up between 20-30 degrees. I am using FFB stick. The graph is from NACA report (https://ntrs.nasa.gov/api/citations/20090022749/downloads/20090022749.pdf). 

 

[69th_Mobile_BBQ]

23 minutes ago, Letka_13/Arrow_ said:

P-47 has been considerably improved in the last patch in terms of its maneuverability, however now it seems that it got some strange landing characteristics. With full flaps and landing gear down the becomes very tail heavy and during powered approach requires a lot of down trim or stick push. I am basically touching down at 85-90 mph with stick in neutral or slightly pushed for a three point attitude. I have flown P-47 extensively before and at touchdown it required nearly full stick aft for a three pointer, now I am landing with centered stick (trim neutral) and elevator at neutral which is very far off the real data, which show that the elevator should be nearly fully deflected up between 20-30 degrees. I am using FFB stick. The graph is from NACA report (https://ntrs.nasa.gov/api/citations/20090022749/downloads/20090022749.pdf). 

 

 

That's odd. I've been finding that I've had to add a rather large amount of tail-heavy trim in most flight conditions to keep the updated model from lawn darting.  

But, once I get it "balanced" it flies alot better than the previous model. I suspect the trimmer tabs weren't reviewed also along with the new changes.

[354thFG_Rails]

Same, I usually have to put in some nose up trim to keep it nice. Otherwise the nose will still fall. The ballooning though seems more pronounced than before, not sure if this is correct. I would also like the flaps to be a bit more draggy and produce a bit less lift, but again I have yet to find any data for the flaps. I used to be able to maintain 20” and 2550 rpm for my approach settings all the way in to the threshold. Maintaining 140-150 IAS the whole time, I’ve read you do not want to turn the 47 below 130 when you’re on approach so I try to keep the speed up. I’m finding the plane a bit to controllable at those low speeds especially below 130, it feels like you have a lot of authority still. 

Edited March 5, 2024 by 86th_Rails

[JeanStravinsky]

On Special Missions: The Luftwaffe's Research and Experimental Squadrons 1923-1945   John Richard Smith, Eddie J. Creek, Peter Petrick Classic, 2003 - History - 128 pages This is the remarkable story of the Verschuchsverband, the Trials and Research Unit of the Luftwaffe High Command, one of the most intriguing, clandestine and rarely-covered elements of the Luftwaffe before and during World War 2.

 

  Quote (early captured Razorback without full power available, and with needle tip prop): "The P-47D out-turns our Bf-109G."  (I think it is likely later Bubbletop versions did not do quite as well as the earlier Razorback did.)

 

  Bf-109G vs P-51B comment (in the same book): "The P-51 has a dangerous stall which killed two of our pilots."

 

 

  Contrast with P-51D:

 

 Osprey, "RAF Mustang and Thunderbolt Aces", P.42:

Sq. Lt. Hearner (No 19 Sq) commenting 11 April 1945 battle over Lister airfield (P-51 Mk IV vs late Me-109Gs or Ks )

  "The 109s we encountered were obviously an experienced bunch of boys. Their turning circle is decidedly better than ours at low speed. The lowering of 20 degrees of flaps may just enable us to hold them in the turn, although I feel they could outclimb us."

 

 

 Quote, 1989 SETP (Society of Experimental and Test Pilots): "Heading Change Time (180 deg at METO, 220 KIAS at 10,000 ft.)

FG-1--8.5 sec / P-47--9.7 sec / F6F--9.9 sec / P-51--10.0 sec

 

  This is not sustained speed turning. Note this is a Bubbletop, so (in my opinion) not quite similar to a needle tip Razorback (at least in left sustained turns).


  Quote 1989 SETP test: "AIR-TO-AIR TRACKING 210 KIAS at 10,000 ft. (straight & level into a 3g turn to the left building to 4g followed by a hard reversal into a 4g right turn.)


  FG-1 best, followed by P-47, F6F and, trailing badly, the P-51." (In this test, the roll stick forces were so high on the P-51D that in one instance the stick was wrenched from the pilot's grasp...)

 

  The real instantaneous roll rates of the P-47D and P-51 in those tests were both at 75 degrees per sec at 1G, but under 3Gs of turning effort the P-47 was still at 66/s while the P-51 dropped to 55/s, and with much higher stick forces...

 

  P-47D vs Me-109G [probably gondola equipped]: "We got to the deck. After 3-4 climbing turns I managed to get in a position to fire a deflection shot... We continued in a climbing lufberry indicating 140 mph. We continued climbing for another full turn, when he suddenly snapped out and spun in."


http://www.wwiiaircraftperformance.org/p-47/er/78-mcdermott-...

 

 

  (P-47Ds) "We started turning with several 109s and were having no difficulty doing it at 23 500 ft., with full tanks" "The E/A (109s) started to turn [12 000 ft.], and we out-turned them immediately."


http://www.wwiiaircraftperformance.org/p-47/er/78-covelle-19...

 

 

 But my favourite of them all...:

 

  Osprey, "P-47 Thunderbolt units of the 12th Air Force".

  P.32: 15th May 1944, 87th Fighter squadron operational report (Paddle-blade propellers only started to be delivered to the group in late May 44, and only with new aircrafts, so all these are needle-tip props, which does explain in part their turning performance).

  :"That afternoon, the 87th FS took off (16 aircrafts) with 32 X 1000 lbs bombs underwing to add to the destruction in Acquapedente. Target: Acquapedente bridges."

  "A flight of 15 Me-109s and 5 FW-190s was encountered. One section kept the fighters occupied while the remainder attacked the bridges. Three enemy fighters were destroyed for one of ours damaged."

  "A gratifying result of this engagement was that a P-47, not considered a low-altitude aircraft, can maneuver advantageously with Me-109s almost on the deck, even though under the handicap of being on a bomb run." [Meaning 2 X 1000 lbs of bombs underwing...]

 

  Now I know some will inevitably say the P-47s "occupying" the 109s had already dropped their bombs, but that is not what is implied by the text. If they had dropped their bombs, then they would not be "under the handicap", would they? 

 

  I actually do not think the Razorback needle tip prop P-47 was better in low speed sustained turns than a clean Me-109G-6. I think they were roughly equal, if the Me-109G reduced the throttle, which it could afford to do more than the essentially no-flaps P-47 (flaps limited to 140 mph on the Razorback, and automatic "blown flaps" on late production runs of the Bubbletop), and both were better to the left at low speed. The P-51D was slightly worse than both, but similarly preferred left turns at low speed. (Not all types were noticeably asymmetrical in sustained turns: The Ki-43, FW-190A, and especially the Spitfire were all fairly symmetrical, although the FW-190A clearly hated hard un-sustained right turns...) 

 

  In higher speed (230 mph +) right turns that did not spiral down, the Me-109G was better than the P-47 (and also better than the P-51 if spiral climbing, but likely less so in level right turns or spiralling down), as explained by Johannes Steinhoff: "I turned into a right climbing chandelle. The P-51 and P-47 could not duplicate this maneuver. The P-38 however could match us.".

 

  High speed (above 230 mph) sustained altitude left turns were poor on the Me-109, because of the boot full of very heavy LEFT rudder deflection needed (even for just 2 degrees of rudder deflection) to keep the nose from drifting right on its own... This applied not only to the E, but also to the G, and was commented on by British testers: "The lack of rudder trimmer is severely felt, and seriously reduces the ability to perform left turns."

 

  I do not know to what extent the later tall tail alleviated this. My impression from a Karhila interview is that it might have made things worse, but it is not clear.

 

  The well-known joke among Me-109G pilots went: "You can tell a 109 pilot by his over-developed left leg." The effect was worse with speed, so later MW-50 versions might have contributed to the saying. It should be noted that the Me-109 was neutral below 230 mph, and even required some right rudder to keep the nose straight when climbing, so the "heavy" left foot effect only existed above that median speed of about 230 mph, but it is important to note it rapidly became heavy quite soon after that speed was passed.

 

  This rudder effect did not affect a 90 degrees maximum rate turn that spiralled down, since, in that maximum bank attitude, the rudder is not as important. However such turns would drop you far below your opponent, which is why most wanted to keep the altitude as much as possible, or at least keep the spiral as flat as possible... The P-51 was so slick that it maintained speed, in downward left spirals, better than any other type, so allowing a downward left spiral to the P-51 was probably a serious error.

 

  It should be noted that the P-47 was generally light on the elevator, like the Spitfire, and so was generally "mushy" in hard turns (mushier than the Spitfire, which also quickly rumbled its wings in hard turns), giving a false impression of poor overall P-47 turning ability before the speed had stabilised to a lower value. The Spitfire could engage the turn much harder, but had to keep the radius wide once the speed had dropped (possibly in part because all Marks had no partial flap position).

 

  The mushing "deceleration phase" of a P-47 could still occasionally cause a pursuer to overshoot. This did not happen as easily with the FW-190A, because its mushing was far worse, and continued indefinitely in a very broad curve, inflicting very high deceleration Gs on the pilot while generating no angles (Red Fleet: "There is a moment where the FW-190 "hangs": It is then convenient to fire.". In turns, one of its wing could drop violently.

 

  The P-51D was much crisper at high speed than all of these, but had very heavy controls, shockingly so in the case of rolls. In low speed turns, using 10-20 degrees of flaps was essential on the Merlin P-51. A FW-190A using flaps would easily out-turn the Merlin P-51 in low altitude low speed right turns, but the contest was closer in left turns, simply because the P-51 performed better to the left when slow (but seemed more symmetrical at higher speeds). 

 

  800 P-47 Encounter Reports which I have read (at least twice each), available here:  http://www.wwiiaircraftperformance.org/p-47/p-47-encounter-reports.html

 

  Read those and you will see that the "Hit and Run" use of the P-47D ranges from rare on the Bubbletop to near non-existent on the Razorback, and the Razorback in particular proves shocking adept (and obsessive) in using prolonged turn fighting to the left (especially against Me-109s), at the exclusion of virtually everything else, except slowly sneaking up under the tail...

 

  As a conclusion, I would say the misleading aspect of the P-47's turning ability is that its initial response to a hard pull was mushy and thus seemed poor, but that this did not reflect how it could perform once the speed had gone down and the turn speed was stabilised to a sustained value. The Fw-190A was broadly similar, yet much more extreme in its contrasts, being better at low speed and much inferior at high speed, the transition point being around 250 mph, as this test demonstrated:

  

Stability and control committee, "S.C. 1718", 24 April 1944:

P-47D versus FW-190A at Low Altitudes:

Turning above 250 mph:

"The P-47 easily out-turned the FW-190 at 10 000 ft., and had to throttle back to avoid overrunning the FW-190. The P-47's turning superiority increased with altitude. The FW-190 vibrated excessively (at higher power) and had a tendency to black out its pilot. [Again, pitch up tail-sinking deceleration]

Turning below 250 mph:

"The turns were made so rapidly it was impossible for the airplanes to accelerate, and the ability of the FW-190 to hang in its propeller and turn inside the P-47 was very evident."

 

  This below is a link to the Air Force board game variant I made, which compiles the results of my research. It has just now been the object of a 2 year long update, based on some of the quotes above:

 

  Air Force | Advanced Air Force | File | BoardGameGeek

Edited March 9, 2024 by JeanStravinsky
added details

[BillsPlane]

I appreciate the last update of the P-47, but I still question was the plane really this sluggish in real life?  Also, it doesn't seem to perform as well as I thought it would at high altitude, both in turning ability but also just speed.  I find it seems to do better speed wise at 20-25k ft than 30k, yet it was rated at best max speed at 30k.  I find it is slower than the P-51D at 30k.  With 150 octane, about 250mph IAS is about all I can get out of it at WEP, cowl flaps closed, oil/intercooler tried at everything from closed to neutral to full open, etc.  In comparison, I get 270mph IAS out of the P-51D at 30k ft. Not complaining, just wondering if the specs are off that are used in the flight model?  I assume, with the level of detail of this simulation, that it is a pretty full flight model?  I.e., to give an analogy with something I am quite familiar with, for designing a car acceleration simulation one would use the input parameters of the vehicle speed, vehicle weight, engine torque curve, gear ratios, frontal area, drag Cd, weather conditions (barometer, humidity, temp), tire coefficient of friction to model the acceleration of the car rather than using actual acceleration values as a function of speed and gear (the empirical method) to model it.  Related to aircraft, I would expect similarly that to model an aircraft acceleration you would have input parameters for engine power output (which itself would have a sort of mini-model for calculating the engine output based on manifold pressure and rpm, modified by temp and mixture), aircraft velocity (to calculate air drag), aircraft weight, etc.  As opposed to just have acceleration numbers for an aircraft hard coded in (i.e. empirical rather than modeled).  Anyway I am hoping it is modeled as mentioned.  If that is the case, if the "correct" values for things such as weight, engine power and drag Cd are used for a given aircraft, the performance should be accurate.  With the assumption that the model is correct.  That brings me to my next point.  Sometimes an inconsistency in the performance actual vs the performance expected for a given vehicle/plane, while the rest are matching up is a reason to take another look at the performance model.  Though, this must then be rechecked on all of the planes to make sure there aren't any other "gotchas".  I say this humbly and based on my own experiences in designing/evolving and selling a car acceleration simulation back many years ago called PC Hotrod.  It was decidedly simpler than what you guys have built, but it illustrates the concept of coming up with a physics model and then adjusting it over time as "exceptions" pop up.  When an exception would pop up... i.e., the 1/4 mile ET/mph was not matching up for a given car or range of cars, first question I would ask is "why?", followed by "what did I fail to model?"  or "what did I perhaps not model quite accurately enough"?  For example, in the early days I found that cars with larger engines were accelerating faster than the magazine articles were showing that model, yet for the smaller engined cars it was spot on.  I had already correctly modelled the engine torque curve and tracing it through the drivetrain (torque multiplication by gear ratios, drivetrain efficiency), along with aerodynamic and rolling drag, weight, etc.  What I determined was I incorrectly modeled the rotational inertia of the engine crank and flywheel.  I had a hardcoded value used but that worked for smaller engined cars, which I had originally designed the program model around.  Once I revised the model to take in account the engine displacement and type of flywheel used to better estimate the engine rotational moi... basically going into a little more detail of that part of the model, the results then became accurate when I tested it with a much larger range of cars (small 4 cylinder VWs, big block V8s and everything in between).  Anyway, just some food for thought regarding the physics modelling.  Sometimes, all one has to do is make slight revisions to the model to take something into account that was missed.  Doesn't mean the whole model has to be thrown away or necessarily a lot of work.

Edited March 19, 2024 by Spitfire_Enthusiast1

[1PL-Husar-1Esk]

 

14 minutes ago, Spitfire_Enthusiast1 said:

I appreciate the last update of the P-47, but I still question was the plane really this sluggish in real life?  Also, it doesn't seem to perform as well as I thought it would at high altitude, both in turning ability but also just speed.  I find it seems to do better speed wise at 20-25k ft than 30k, yet it was rated at best max speed at 30k.  I find it is slower than the P-51D at 30k.  With 150 octane, about 250mph IAS is about all I can get out of it at WEP, cowl flaps closed, oil/intercooler tried at everything from closed to neutral to full open, etc.  In comparison, I get 270mph IAS out of the P-51D at 30k ft. Not complaining, just wondering if the specs are off that are used in the flight model?  I assume, with the level of detail of this simulation, that it is a pretty full flight model?  I.e., to give an analogy with something I am quite familiar with, for designing a car acceleration simulation one would use the input parameters of the vehicle speed, vehicle weight, engine torque curve, gear ratios, frontal area, drag Cd, weather conditions (barometer, humidity, temp), tire coefficient of friction to model the acceleration of the car rather than using actual acceleration values as a function of speed and gear (the empirical method) to model it.  Related to aircraft, I would expect similarly that to model an aircraft acceleration you would have input parameters for engine power output (which itself would have a sort of mini-model for calculating the engine output based on manifold pressure and rpm, modified by temp and mixture), aircraft velocity (to calculate air drag), aircraft weight, etc.  As opposed to just have acceleration numbers for an aircraft hard coded in (i.e. empirical rather than modeled).  Anyway I am hoping it is modeled as mentioned.  If that is the case, if the "correct" values for things such as weight, engine power and drag Cd are used for a given aircraft, the performance should be accurate.  With the assumption that the model is correct.  That brings me to my next point.  Sometimes an inconsistency in the performance actual vs the performance expected for a given vehicle/plane, while the rest are matching up is a reason to take another look at the performance model.  Though, this must then be rechecked on all of the planes to make sure there aren't any other "gotchas".  I say this humbly and based on my own experiences in designing and selling a simulation back many years ago called PC Hotrod.  It was decidedly simpler that what you guys have built, but it illustrates the concept of coming up with a physics model and then adjusting it over time as "exceptions" pop up.  The first question I would ask is "why?", followed by "what did I fail to model?"  or "what did I perhaps not model quite accurately enough"?  For example, in the early days I found that cars with larger engines were accelerating faster than the magazine articles were showing that model, yet for the smaller engined cars it was spot on.  I had already correctly modelled the engine torque curve and tracing it through the drivetrain (torque multiplication by gear ratios, drivetrain efficiency), along with aerodynamic and rolling drag, weight, etc.  What I determined was I incorrectly modeled the rotational inertia of the engine crank and flywheel.  I had a hardcoded value used but that worked for smaller engined cars, which I had originally designed the program model around.  Once I revised the model to take in account the engine displacement and type of flywheel used... basically going into a little more detail of that part of the model, the results then became accurate when I tested it with a much larger range of cars (small 4 cylinder VWs, big block V8s and everything in between).  Anyway, just some food for thought regarding the physics modelling.  Sometimes, all one has to do is make slight revisions to the model to take something into account that was missed.  Doesn't mean the whole model has to be thrown away or necessarily a lot of work.

They high altitude model is definitely wrong as proven by a book  regards WW1 FC FM . About P47 and high altitude handling I have the same observation as you have.

 

About the book - 

 

Edited March 19, 2024 by 1PL-Husar-1Esk