P-47 Maneuverability or Lack Thereof

[Holtzauge]

2 minutes ago, ACG_Cass said:

My point was these are very extreme tests. You would never find yourself pulling 100% stick at 214 IAS as you'd very quickly stall, as we see. I can test at higher speed but my guess would be that the rate of onset differences will be more pronounced there and a difference of a few tenths or half a second head start to a turning rate can be the difference between life and death. Especially when approaching 7g cornering speeds.  

 

My question would be, outside of maximum performance turns would profile drag be a factor? You aren't always doing tight turns and the P47 bleeds speed in even the lightest of manoeuvres. 

 

 

If the the added wing profile drag is an issue at lower Cl turns? Good question: Intuitively I would say no and it's still the induced drag that adds the most but I would have to calculate it or maybe someone else can answer? However, my wife is competing for attention now and I don't want to sleep on the couch so this means I'm AFK for the rest of the evening!

[Denum]

35 minutes ago, Holtzauge said:

What you describe (the Me-109 maintaining speed better) is due to the P-47 having a higher span loading than the Me-109 so the P-47 should slow down faster. However, note I'm not saying that it's correct the way it is now: Could be there is something off and it IS slowing to fast in-game but this would be very difficult to prove even if it was so I think. That it out-tuns you is again due to wing loading and Clmax.

I fully expect to be out turned, it's just how little the 109s slow down pulling similar G's or in some cases higher is frustrating. 

 

If I push and continue the turn I'm fighting the aircraft more then the enemy, if I straighten out and go straight to try and extend the 109s often catch you as they maintain so much speed. 

 

What's frustrating is if you commit to the fight with the 109 there's no running or leaving the engagement. 

You either win or die. 

 

The P47 is supposed to be quite fast but when your opponent can maintain speed so easily it really puts you into a enormous disadvantage. Even more so when even slight damage completely cripples your aircraft. 

 

 

 

Edited December 21, 2021 by Denum

[Bert_Foster]

1 hour ago, Legioneod said:

This is another thing I'm curious about. Was the drag really so bad on the P-47 that it loses speed at the drop of a hat? I've read or watched a video somewhere stating the drag on the 47 was on par (or not much worse) in comparison to other aircraft so I'm not sure why it losses it's speed so quickly. Also its inability to maintain energy is very frustrating.

"Americas 100,000" Chap 5 contains some interesting relative comparison data. Accuracy of the info I guess needs to be verified however it implies that in the 1G acceleration department the P47 was no slouch:

 

 

Edited December 21, 2021 by Bert_Foster

[ACG_Cass]

29 minutes ago, Holtzauge said:

Good question: Intuitively I would say no and it's still the induced drag that adds the most but I would have to calculate it or maybe someone else can answer? However, my wife is competing for attention now and I don't want to sleep on the couch so this means I'm AFK for the rest of the evening!

Even if it's something fairly small like 5%, that could still be significant.

 

That was me last night and I had to watch a crappy Sarah Jessica Parker Christmas movie, good luck!

 

Anyway, for tomorrow:

 

So here is the same test conducted at 6500 feet with enough engine power to get a constant 300mph Indicated airspeed (added in the curve lines because they look prettier).

 

Now again, this is a fairly short period. But we are talking about a very unrealistic (idiotic) maneuver here. This is a cruise setting pull up where stall will be reached very quickly. 

 

If the same pull was done at full power either without altitude loss or diving then these times would likely longer. 

 

Even still the tactical significance of being 1-1.5g behind a plane that's at exactly the same speed and altitude and has pulled the stick at the same time is absolutely massive, even if for these fractions of a second. Either bouncing someone or defending against a bounce, these are huge differences. Sure the other plane will reach it's Clmax and be able to out-turn you, but getting to your Clmax much faster and at high speed, burning through your speed quickly to tighten the turn rate is again hugely significant. 

 

If this scenario played out at 15,000 feet diving down with the A5 on the P47s tail, the P47 has no defense as it doesn't have the ability to pull inside the A5. Also the added control ability that would factor in if all of the elevator had an effect rather than a bit at full deflection can't be understated.  

 

I think we're starting to see that even outside of any Clmax changes, the P47s elevators were an enormous asset IRL and the significant under modeling of their ability in IL2 are a big attributing factor to it's sluggishness. 

[354thFG_Rails]

1 hour ago, Denum said:

What's frustrating is if you commit to the fight with the 109 there's no running or leaving the engagement. 

You either win or die. 

 

 

 

Try to git gud. 

[Denum]

13 minutes ago, QB.Rails said:

Try to git gud. 

 

I'm trying!

[Bert_Foster]

Best P47D28 sustained Turn performance I can achieve at Sea Level 200MPH IAS 65"/2700 Water injection on 50% internal fuel clean is a Turn time of 23seconds averaged out over 3 turns riding the buzz.

 

That equates to Turn rate 15.65 deg sec with Sust G at 2.6. .... That fits in well with Greg's numbers in his P47 manv video:

 

and the support Fan plot generated by an aerodynamics engineer ASJ27:

 

As to Acceleration rates. Using Mil pwr 52"/2750 acceleration from 250MPH to 350 MPH IAS took 49 seconds so equates to 1.496 Feet per sec ... significantly slower than the quoted Americas 100,000 value of 2.13Feet per sec. However 300MPH would seem to be close to limit IAS in this condition. Going from 250mph IAS to 270mph IAS takes 23seconds so equates to 1.91 Feet per sec ... still down on Americas 100,000 value.

 

Edit: Test with Water injection on 65"/2700RPM I get Accel rates of 4.4 feet per sec from 250-270mph and 3.49 feet per sec from 250-300mph.

Edited December 22, 2021 by Bert_Foster

[Holtzauge]

14 hours ago, ACG_Cass said:

My point was these are very extreme tests. You would never find yourself pulling 100% stick at 214 IAS as you'd very quickly stall, as we see. I can test at higher speed but my guess would be that the rate of onset differences will be more pronounced there and a difference of a few tenths or half a second head start to a turning rate can be the difference between life and death. Especially when approaching 7g cornering speeds.  

 

My question would be, outside of maximum performance turns would profile drag be a factor? You aren't always doing tight turns and the P47 bleeds speed in even the lightest of manoeuvres. 

 

 

Made a ballpark estimate about the influence of profile drag for the P-47:

 

As a first assumption, I’m assuming a drag area of 6.36 sqrft, so a zero lift drag coefficient of around 0.0213.

 

Second assumption is that 30% of this is due to wing (usually this holds true for WW2 type aircraft within reason)

 

Since I don’t have a polar for the S-3 I’m going with NACA 230 data and that profile has a low Cl drag coefficient of around 0.006. Looking at the polar, there is very little difference in profile drag until you go as high as Cl=1 at which point it has risen to 0.008. This is 33% so not a small amount on profile level but then we need to remember that the wing only stands for 30% of the total so a modest total zero lift drag increase of say around 10%. In addition this is a rather tight turn given the Clmax is circa 1.2 (at least for now unless until new compelling data is discovered!).

 

Calculating the induced drag at Cl=1 for the P-47 this yields a result that is in more than three times as high as the zero lift drag so at least by my ballpark estimate the wing profile drag is insignificant.

 

About the later post you did after this one about the onset of the g-load: I still don’t get it why that slight difference between the planes in the g-load onset from 0.25-1.25 s is worth chasing if you after that time is still limited by a lower Clmax for multiples of 10 s but for sure, just like in the Yak-3 and Me-109 videos I posted: All moves in the sim towards a more “crisp” handling and less lag are welcome!

 

11 hours ago, Bert_Foster said:

Best P47D28 sustained Turn performance I can achieve at Sea Level 200MPH IAS 65"/2700 Water injection on 50% internal fuel clean is a Turn time of 23seconds averaged out over 3 turns riding the buzz.

 

That equates to Turn rate 15.65 deg sec with Sust G at 2.6. .... That fits in well with Greg's numbers in his P47 manv video:

 

and the support Fan plot generated by an aerodynamics engineer ASJ27:

 

As to Acceleration rates. Using Mil pwr 52"/2750 acceleration from 250MPH to 350 MPH IAS took 49 seconds so equates to 1.496 Feet per sec ... significantly slower than the quoted Americas 100,000 value of 2.13Feet per sec. However 300MPH would seem to be close to limit IAS in this condition. Going from 250mph IAS to 270mph IAS takes 23seconds so equates to 1.91 Feet per sec ... still down on Americas 100,000 value.

 

Edit: Test with Water injection on 65"/2700RPM I get Accel rates of 4.4 feet per sec from 250-270mph and 3.49 feet per sec from 250-300mph.

 

I updated my C++ simulation code by lowering my previously assumed Clmax for the P-47 so I now have a low Mach power off Clmax of 1.2 like in the game and the results I get in my simulations are basically the same:

 

At 200 mph IAS SL altitude standard atmosphere I get a turn time of slightly below 23 s. However, in my simulation, I get the peak turn rate going a bit slower at 182 mph IAS in which case the turn time goes down to a bit above 22 s. So I seem to be a tad more optimistic than Il-2 in this case!

[ACG_Cass]

2 hours ago, Holtzauge said:

Since I don’t have a polar for the S-3 I’m going with NACA 230 data and that profile has a low Cl drag coefficient of around 0.006. Looking at the polar, there is very little difference in profile drag until you go as high as Cl=1 at which point it has risen to 0.008. This is 33% so not a small amount on profile level but then we need to remember that the wing only stands for 30% of the total so a modest total zero lift drag increase of say around 10%. In addition this is a rather tight turn given the Clmax is circa 1.2 (at least for now unless until new compelling data is discovered!).

How does this compare with other wings? That would seem like in low clmax turns the speed loss from the wing should be fairly minimal. 

 

I was more talking about the profile drag from the increased elevator angle needed at the moment, but I'm not sure how that works in a turn. 

2 hours ago, Holtzauge said:

About the later post you did after this one about the onset of the g-load: I still don’t get it why that slight difference between the planes in the g-load onset from 0.25-1.25 s is worth chasing if you after that time is still limited by a lower Clmax for multiples of 10 s but for sure, just like in the Yak-3 and Me-109 videos I posted: All moves in the sim towards a more “crisp” handling and less lag are welcome!

 

Sure it's not going to help you in prolonged fights but most of the time you live or die on your instantaneous turn and being able to get to your max turn rate quicker is almost certainly going to be an advantage. 

 

If you spot someone behind you with a 100 mph overspeed that is 100m from their firing point, you have about 2 seconds before they are there. Having the ability to quickly pull to your Clmax is absolutely crucial. Not even spotting them, but timing your break turn and having the ability to pull quickly is enormously important. 

 

We could do plenty of calculations around speed and turn rate. But you can see there that even an A5 in that scenario could have overspeed and still be able to pull with the P47. 

 

Also this issue is going to get worse with speed so in high speed fights trying to evade and pull lead shots, it's hugely valuable. The same scenario as above but reversed with the current P47 elevators means that the A5 will easily be able to evade if timed right.

 

Looking that the data, the P47 should be incredibly difficult to escape as it can pull to it's Clmax very quickly, burn speed and get a lead shot. 

 

 

 

[ACG_Cass]

So I configured the P51 and P47 to have the same stall speed. P47 at it's minimum weight with 4 guns, no ammo and 10% fuel. P51 with 75% fuel. 

 

As they have the same Vmin, they can pull to the same G limit. I understand there will be MAC differences between the 2, but even still this is a significant difference, especially when compared to the results in DCS. 

 

If you look at the details in this scenario it doesn't make a huge amount of sense. 

 

	Elevator Area sqft	Wing Area sqft	Weight kg
P51	13	240	4400
P47	22	300	5033

 

 

	Wing Area to move per sqft of Elevator	Weight to move per sqft of Elevator
P51	18.46153846	338.4615385
P47	13.63636364	228.7727273

 

The P47 shouldn't just be matching the P51, it should be beating it. I know the P51 has superior controls at higher speeds, but at 310 IAS at 2000m that won't have taken effect and this is purely a the P47 missing a huge chunk of it's elevator effectiveness. 

 

[JtD]

1 hour ago, ACG_Cass said:

Wing Area to move per sqft of Elevator // Weight to move per sqft of Elevator

 

I'm not trying to take away the point of the sluggishness of the P-47 response, but these two parameters way oversimplify the issue of aerodynamic control and pilot input. Referring to NACA test results makes much more sense.

 

For instance stick forces per g, which depending on centre of gravity could be between 8 and 3 lb/g for both aircraft at around 250mph - so fairly similar.

(And what does this have to do with response to pilot input? Well, low force (and small stick travel) both make for easy quick pilot input, which is as essential to a quick reaction of the aircraft as anything in the aerodynamic layout of the aircraft as such.)

[ACG_Cass]

 

1 hour ago, JtD said:

I'm not trying to take away the point of the sluggishness of the P-47 response, but these two parameters way oversimplify the issue of aerodynamic control and pilot input. Referring to NACA test results makes much more sense.

 

For instance stick forces per g, which depending on centre of gravity could be between 8 and 3 lb/g for both aircraft at around 250mph - so fairly similar.

(And what does this have to do with response to pilot input? Well, low force (and small stick travel) both make for easy quick pilot input, which is as essential to a quick reaction of the aircraft as anything in the aerodynamic layout of the aircraft as such.)

 

Correct yes. I'd made a point about that comparison earlier in the thread that it's an oversimplification but probably should have added it again. There's actually a lot of data about the planes stick forces that I realised haven't been included yet. I was making the point having known the P47 stick forces aren't that different and even factoring in the differences in position, design etc for the elevators, we're talking about a 25-33% difference. For there to be that performance gap between the designs seemed a little far fetched.

 

To be honest if we just matched the response and time to G of the P51 I'd be happy with that.

 

P-47

http://www.wwiiaircraftperformance.org/p-47/Naca_TN_2899__F-47D-30_Flight_Test.pdf

 

 

 

Here is a comparative 180 degree left turn at similar speeds:

 

P51

 

 

 

P47

P47s Mac would have been between 28.3 and 30.2

Edited December 22, 2021 by ACG_Cass

[Legioneod]

14 hours ago, Holtzauge said:

...

I updated my C++ simulation code by lowering my previously assumed Clmax for the P-47 so I now have a low Mach power off Clmax of 1.2 like in the game and the results I get in my simulations are basically the same:

 

At 200 mph IAS SL altitude standard atmosphere I get a turn time of slightly below 23 s. However, in my simulation, I get the peak turn rate going a bit slower at 182 mph IAS in which case the turn time goes down to a bit above 22 s. So I seem to be a tad more optimistic than Il-2 in this case!

This sustained turn? If so it's vastly superior to what we currently see in-game. 

[Bert_Foster]

18 minutes ago, Legioneod said:

This sustained turn? If so it's vastly superior to what we currently see in-game.

I get 23 seconds sustained turn Sea level as posted earlier

 

"Best P47D28 sustained Turn performance I can achieve at Sea Level 200MPH IAS 65"/2700 Water injection on 50% internal fuel clean is a Turn time of 23seconds averaged out over 3 turns riding the buzz. "

 

 

 

Edited December 23, 2021 by Bert_Foster

[Legioneod]

34 minutes ago, Bert_Foster said:

I get 23 seconds sustained turn Sea level as posted earlier

 

"Best P47D28 sustained Turn performance I can achieve at Sea Level 200MPH IAS 65"/2700 Water injection on 50% internal fuel clean is a Turn time of 23seconds averaged out over 3 turns riding the buzz. "

 

 

 

Ah ok, I was reading from the stat card. Says 27sec max performance turn.

[Cpt_Siddy]

On 12/23/2021 at 4:18 AM, Legioneod said:

27sec max performance

That's what she said.

 

But seriously, are we looking at the wrong thing here, you all have been mulling over the low speed stuff. that have been more or less fixed and honed in.

 

The main gripe of people flying p-47 is it high to mid altitude performance, lackluster dive acceleration and poor instantaneous turn with huge energy bleed.. 

All that what make it be bad at its intended high altitude role. 

 

[ACG_Cass]

Ran a close enough test to this at the same altitude, speed and weight loading.

 

 

I took stills of maximum and zero deflection as a comparison. 

 

The elevators should be deflecting up to 30 degrees which looks about right. 

 

 

According to the IRL test we should be seeing a max of 2-2.5 degrees of a deflection to initiate a maneuver like this. 

 

 

In IL2 it looks more like 10-12 degrees. 

 

[ZachariasX]

1 hour ago, ACG_Cass said:

In IL2 it looks more like 10-12 degrees. 

You will find this in many other sims as well. And it is what I am repeating over and over. I don't know how much this is intended to maybe simulate stick forces on your desktop joystik or not, but it is a phenomenon at the base of what is the wobble. That one has thank God been damped out in most cases, but the control authority at low deflections is still a far cry of what it should be on aircraft that are at least reasonably neutral or stable to neutral in pitch. In the Spitfire, the nose (and wings) follows the stick like the mouse pointer follows your mouse movement. It is that direct. It is also shown very clearly in the hugely excessive stick travel required to compensate for trim changes in the Spit IX, where the elevator trim position is far off anyway.

 

The frustrating thing about how modern sims are representing stck input is that they tend to come with lag (as it has been shown above in this thread) and hence you cannot really cancel insufficient control response with skewing a response curve without creating new artifacts.

[ACG_Cass]

2 hours ago, ZachariasX said:

You will find this in many other sims as well. And it is what I am repeating over and over. I don't know how much this is intended to maybe simulate stick forces on your desktop joystik or not, but it is a phenomenon at the base of what is the wobble. That one has thank God been damped out in most cases, but the control authority at low deflections is still a far cry of what it should be on aircraft that are at least reasonably neutral or stable to neutral in pitch. In the Spitfire, the nose (and wings) follows the stick like the mouse pointer follows your mouse movement. It is that direct. It is also shown very clearly in the hugely excessive stick travel required to compensate for trim changes in the Spit IX, where the elevator trim position is far off anyway.

You're completely right, the other planes appear to be the same. I'd wager that the actual elevator perhaps isn't indicative of the total travel. That certainly seems to be the case when looking at the sticks. 

 

Stick forces appear to be much higher than they should on the 47. I don't have a way to track the input unfortunately, but you can see from an incredibly unscientific look at the percentage of travel, that the P51 needs far less.  

Top: Neutral

Middle: 3g pull

Bottom: Full stick deflection

 

edit: quite frankly I don't think this shows us anything so I've no idea why I posted this. 

Edited December 24, 2021 by ACG_Cass

[[CPT]Crunch]

Then old training wheels conundrum, easy to see in a Tempest or Typhoon, get about 400+ Mph going, pull max stick, and as you bleed speed all of a sudden the training wheel decides to quit and a sudden maximum stick pull is auto induced, and not by you or your input.  I absolutely hate that crap, it ruins many a plane, nothing realistic about a protection circuit, many like 109's really reap the benefit, you absolutely can't high speed stall one out or commit any error.  I'd prefer to make it so the pilot has direct control and it stalls, stick shakes violent inducing undesirable characteristics, the tail rips right off or wings snap, or the plane departs.  Anything but training wheels, its a natural ace maker, no talent, no finesse even required.  The uproar would be epic among the aces.

[Mtnbiker1998]

On 12/24/2021 at 12:05 PM, [CPT]Crunch said:

Then old training wheels conundrum, easy to see in a Tempest or Typhoon, get about 400+ Mph going, pull max stick, and as you bleed speed all of a sudden the training wheel decides to quit and a sudden maximum stick pull is auto induced, and not by you or your input.  I absolutely hate that crap, it ruins many a plane, nothing realistic about a protection circuit, many like 109's really reap the benefit, you absolutely can't high speed stall one out or commit any error.  I'd prefer to make it so the pilot has direct control and it stalls, stick shakes violent inducing undesirable characteristics, the tail rips right off or wings snap, or the plane departs.  Anything but training wheels, its a natural ace maker, no talent, no finesse even required.  The uproar would be epic among the aces.

Thats all fine and dandy, except that theres nothing realistic about a pilot ripping off his wings because he instantly applied a billion pounds of force on his stick. The current system we have even with its flaws is far better than unrealistic difficulty just for the sake of making it harder.

 

If the system isn't perfect, refine it, don't just remove it entirely.

[[CPT]Crunch]

No, the point is its silly talking about realistic flight modeling and behaviors when you can't even depart the published envelope any way.  The charts were posted as a warning, they're not a safety rail, wall or limit to keep an airframe contained from exceeding as a game does.  So enjoy the 'game', it is what it is.

[69th_Mobile_BBQ]

2 hours ago, [CPT]Crunch said:

No, the point is its silly talking about realistic flight modeling and behaviors when you can't even depart the published envelope any way.

 

Just a friendly reminder that Chuck Yeager wasn't the first pilot to break the sound barrier.  He was the first to live to tell about it (with recorded proof to back him up).

[II/JG17_HerrMurf]

On 12/24/2021 at 12:05 PM, [CPT]Crunch said:

The uproar would be epic among the aces.

 

I think both real pilots and the virtual aces would adapt in a matter of days. The rest of the community would follow not too far behind. It would only wreck those new to simming and BFM.

Edited December 27, 2021 by II/JG17_HerrMurf

[[CPT]Crunch]

Until the day we get one to one controls with direct haptic feedback, there's no point in getting to excited about all this, and that doesn't exist in home PC games.  None of us on the user end can really know where the limits are, we get a vote on input, the game actually decides what force is actually used, when, and how much.  Hence we're stuck with training wheels, and not a damn thing anyone can do about that, we can only argue the diameter of the training wheel, but its still going to be there.

[Mtnbiker1998]

5 hours ago, [CPT]Crunch said:

Until the day we get one to one controls with direct haptic feedback, there's no point in getting to excited about all this, and that doesn't exist in home PC games.  None of us on the user end can really know where the limits are, we get a vote on input, the game actually decides what force is actually used, when, and how much.  Hence we're stuck with training wheels, and not a damn thing anyone can do about that, we can only argue the diameter of the training wheel, but its still going to be there.

So then why did you come here in the first place with your post complaining about "training wheels" and why you think they need to be removed? Way to totally 360 and add nothing to the thread.

[ACG_Cass]

We can't have our cake and eat it too. You either get proper elevator sensitivity and the player can do some crazy stuff, or you get a more muted and dampened approach like in IL2. It certainly has its faults but is fit for purpose and works most of the time.

 

Anyway, within the current framework that we have it looks like the P47 doesn't have the elevator authority or stick forces it has IRL. I think a big factor people are forgetting about is how effective the flaps are on it. Even ~5 degrees of flaps can knock the clmax up to 1.3. Some extra sensitivity could go a long way.

 

I'll see what the devs need from us in order to potentially look into it.

Edited December 27, 2021 by ACG_Cass

[[CPT]Crunch]

Because we don't need the wheel expanded any more than necessary, it's already too encompassing.  A 4 year old kid can fly damn near anything in the game, max perform it, and won't get in sort of trouble.

[Cpt_Siddy]

9 hours ago, [CPT]Crunch said:

Because we don't need the wheel expanded any more than necessary, it's already too encompassing.  A 4 year old kid can fly damn near anything in the game, max perform it, and won't get in sort of trouble.

 

Flying is not that hard. 

 

It is like riding a bike. 

[ACG_Cass]

I found an RAF report on Take Off Trials for the P47 D-22 (pretty sure it's the exact model based on other parts of the test).

 

 

This was conducted at full fighter load (6409kg). I replicated the windspeed (perfectly blowing head on), MAP and weight and ran some tests on Kuban Autumn (so ISA) - I can replicate the temperature and pressure of the test if needed but I don't think it would have that significant an impact looking at them. 

 

 

Our P47 can most certainly not take off at those speeds. We are looking at 6-8mph higher than 115mph. 

 

I added some flaps and found that you would need between 15-20% flaps to match the take off speed seen in the IRL test - this points to the P47s lift being too low. 

 

Edited January 9, 2022 by ACG_Cass

[K3FKefuddle]

Cass, What a great piece of work you have done there. Fascinating!

FWIW, I am very new to IL2GB, the very few aircraft I have messed around with in IL2 seem to struggle to get off the ground with what I would presume to be combat weights typical for the aircraft, so your finding does not surprise me. I could be very wrong there.

In reality, CofG location would make a difference, the further aft (within limits) would reduce the amount of lift needed by the main wings. I have no idea how much difference this would be though, without a flight manual. I noticed in another thread, the IL2 Kuban P47 has the ground defence kit which is additional armour and weight, and I would guess that armour would be up front...

Saying all that, I think it would be very difficult to make a sim that would be mph accurate in all phases of flight, I am sure compromises have to be made in the FM, but a 5mph difference is quite a lot. The only sim I have ever seen nail book numbers is Elite Pilot 8, which is not, but any measure, an entertainment product.

Edited January 10, 2022 by Kefuddle

[JtD]

On 1/9/2022 at 1:19 PM, ACG_Cass said:

I can replicate the temperature and pressure of the test if needed but I don't think it would have that significant an impact looking at them.

 

It would make 115mph TAS to 115.3mph IAS. Indeed insignificant.

 

I can manage to pull the plane off the ground at about 115mph, but that's certainly not in a way I'd do it in real life. My biggest issue with replicating take off run tests is that it is difficult to identify the issue, if there is one. It can be airfoil lift, it can be propwash, it can be attitude on the ground it can be lift over angle of attack it can be ground effect related and there's more. In case of the P-47 I managed to keep it airborne (gear down) at about 113mph IAS flaps up military power full fuel extra ammo low altitude. With a stall speed of ~113mph I'd certainly not take off at 115mph in real life (and also not at 110 like the other pilot managed). Since being airborne eliminates some unknown variables from the actual take off scenario, I think you're right with your conclusion about lack of (maximum) lift in power on config.

[ZachariasX]

On 1/9/2022 at 1:19 PM, ACG_Cass said:

this points to the P47s lift being too low. 

I would be careful taking such a test as a mean to determine lift. The aircraft we have in the sim do not always fly themselves of the ground as the real aircraft. that is not something all too straightforward to bring in a sim.

[ACG_Cass]

19 hours ago, JtD said:

I can manage to pull the plane off the ground at about 115mph, but that's certainly not in a way I'd do it in real life. My biggest issue with replicating take off run tests is that it is difficult to identify the issue, if there is one. It can be airfoil lift, it can be propwash, it can be attitude on the ground it can be lift over angle of attack it can be ground effect related and there's more. In case of the P-47 I managed to keep it airborne (gear down) at about 113mph IAS flaps up military power full fuel extra ammo low altitude. With a stall speed of ~113mph I'd certainly not take off at 115mph in real life (and also not at 110 like the other pilot managed). Since being airborne eliminates some unknown variables from the actual take off scenario, I think you're right with your conclusion about lack of (maximum) lift in power on config.

 

Thanks @JtD

Can I ask the technique you used? In that configuration I cannot get it off the ground at 115mph. Even full elevator defection with the tail on the ground only lifts the the front wheels up at 119mph for me. It's a good point about ground effect, I'm not sure whether that is modelled in IL2.

 

15 hours ago, ZachariasX said:

I would be careful taking such a test as a mean to determine lift. The aircraft we have in the sim do not always fly themselves of the ground as the real aircraft. that is not something all too straightforward to bring in a sim.

Absolutely. I was careful to word it as "points". This is by no means an definitive evidence that its the lift that is the issue. To @JtD's point, there are a lot other variables that could be the issue here. But, providing everything else is right, lift would be the factor that is incorrect.

[ZachariasX]

53 minutes ago, ACG_Cass said:

lift would be the factor that is incorrect.

But it would be lift in ground effect, not the lift you have up there. And I have no idea how ground effect is realized in this game. I mean, I see what you are doing and it is an interesting datum that you brought to the discussion. Still, personally, I am hesitant to go as far as conclusions based on that.

[ACG_Cass]

4 hours ago, ZachariasX said:

But it would be lift in ground effect, not the lift you have up there. And I have no idea how ground effect is realized in this game. I mean, I see what you are doing and it is an interesting datum that you brought to the discussion. Still, personally, I am hesitant to go as far as conclusions based on that.

As far as I can find, ground effect does not increase Clmax it just shifts the lift-slope back and decreases the AoA needed to achieve it. By most standards it appears to be ~2-4 degrees between airfoils (although could be different for the P47).

 

In these examples the technique described is lifting the tail once there is enough airspeed and then "pulling" the plane off the ground. The P47 sits at a 10-11 degree angle while the tail is on the ground and has a normal max AoA of ~16 degrees (from the stats page). We can safely assume that the pilot most definitely did not want to pull to stall on take off and couldn't due to the tail hitting the ground before that was even achievable. Even factoring in ground effect that would change the Clmax AOA to 10 degrees, providing everything else is modelled correctly, the test indicates that maybe, although not entirely certainly, the P47 has a power on stall speed lower than what we currently have in IL2. In my tests, even with with the tail down and full elevator deflection, the P47 front wheels only lift at 119mph

 

And if prop wash is modelled properly, a lower power-off stall speed. 

 

[JtD]

9 hours ago, ACG_Cass said:

Can I ask the technique you used? In that configuration I cannot get it off the ground at 115mph. Even full elevator defection with the tail on the ground only lifts the the front wheels up at 119mph for me.

 

I raise the tail during the take off run and pull back fully at about 110-112mph, the front wheels come up which increases AoA which increases lift and I can get all wheels off the ground at 115.

But, reading the "technique" bit in the report carefully, it says that when the mentioned speed was reached, the aircraft was "pulled off the ground". It doesn't really say that all wheels were free at that point. So I suppose if you pull back at 115 and are clear at 119mph, it would still qualify as 115 in the sense the report uses the figure. However, I don't feel that the safety margin at these speeds makes the lift off recommended.

[ACG_Cass]

2 hours ago, JtD said:

I raise the tail during the take off run and pull back fully at about 110-112mph, the front wheels come up which increases AoA which increases lift and I can get all wheels off the ground at 115.

Ahhh I see now. If you pull up aggressively from tail wheel up you can get it to pivot on the tail wheel which gives you the AoA to get in the air. I got about 116-117mph but can see it would be possible to get 115. I also got roughly the same power on stall at ~113mph, although it's mega tricky to get that.

 

2 hours ago, JtD said:

But, reading the "technique" bit in the report carefully, it says that when the mentioned speed was reached, the aircraft was "pulled off the ground". It doesn't really say that all wheels were free at that point. So I suppose if you pull back at 115 and are clear at 119mph, it would still qualify as 115 in the sense the report uses the figure. However, I don't feel that the safety margin at these speeds makes the lift off recommended.

I originally wondered this as well, but the fact that they've measure the TAS with an external camera, the easiest reference point would have been wheels up. The elevator authority was clearly very good on the P47 + ground effect I'd imagine the pull up was quite rapid unlike what we see.

[ACG_Cass]

So this is the model specification for the P-47, from the 2nd August, 1944. 

 

 

 

 

It contains performance information from the P47D-1 to the D-30 (Design useful load). 

D22:

D:28

 

and also runs into detail for each plane on the design useful load

 

D22:

D30:

It also states a stall speed:

 

 

Big question: What weight are these for?

 

The Design Gross Weight is shown the page before, but it would be very odd to use this as this is an estimation and doesn't include the pilot and some equipment. Also the speed performance was given at Design Useful Load. 

 

At the Design Gross Weight, we are looking at a low mach, power off Clmax of 1.25, higher than we currently have in IL2.

 

The Useful loads don't fluctuate for 25 & Up, but do for the 1 Up to 25-RE. 10 Up to 25-RE is fairly static. At Useful loads, you are looking at a Clmax of 1.28. 

 

It would be odd for the model specification to give a stall speed for a plane without a pilot, so I'm almost certain it would be the latter that the stall speeds to relate. 

 

Also to factor in is the fact stalling speeds are specified with gear down. I'm unsure as to whether this increases or decreases the stall speed as IRL tests seem to be inconsistent on this. I think tweaking the IL2 P47 so it corresponds to those stalling speeds with the gear down and letting the current modelling do the rest would probably be best. 

Edited January 22, 2022 by ACG_Cass

[ACG_Cass]

Another issue that's a bit more severe appears to be the stick forces on the P47. 

 

Despite having a lower Clmax than the P51, the elevators appear to have been powerful enough to pitch the P47 so that the stick forces per G were comparable. The test conditions are very similar in terms of altitude and speed and are comparable all the way up to 350mph. This is the forward center of gravity position. There are tests for high altitude and rear center of gravity as well that compare that I can post later. 

 

Understand it's a 51H and there were considerable differences, but the elevators remained the same and I would imagine stick forces were very similar.