Jump to content
JtD

Another look at turn times

Recommended Posts

Well, the A-20 does slightly in excess of 500km/h on the deck at 3200hp. So it is worse with thrust/drag and modelled as such than a Bf109G or a Bf110G. The latter of which, while similar overall, has a slightly worse span and wing loading than an A-20 without bombs. It has a slightly better power loading, though. All in all there'd only be small differences between the two, still in favour of the A-20 assuming similar efficiencies. In game, of course, there appears to be quite a large difference, owing to the worse efficiency of the Bf110.

Share this post


Link to post
Share on other sites

Hey @Holtzauge

 

I've been reading down on your comment history trying to understand what you are up too, but you have hundreds of forum comments and I'm having a hard time getting a grip on what your C++ program actually does. Is there anywhere where you have summarized what you are doing? Namely, what is your mathematical model, assumptions, simplifications, and validation cases?

 

Cheers

Share this post


Link to post
Share on other sites
28 minutes ago, Pupo said:

Hey @Holtzauge

 

I've been reading down on your comment history trying to understand what you are up too, but you have hundreds of forum comments and I'm having a hard time getting a grip on what your C++ program actually does. Is there anywhere where you have summarized what you are doing? Namely, what is your mathematical model, assumptions, simplifications, and validation cases?

 

Cheers

 

Funny you should ask right now since I just replied to basically the same question here. But to provide some input on the model: The backbone in the simulator is Simula code I wrote in the 80's as part of my Masters Thesis to assess the effects of external stores like missiles, drop tanks etc on jet fighters which I more than 10 years ago modified and converted to C++ to be able to model prop planes. But basically it works by having an atmospheric model so I can do simulations at any altitude and also climbs and dives that take changing atmospheric conditions into account. I calculate power available, exhaust thrust, propeller efficiency based on advance ratio, tip speeds, disc loading and lift and drag depending on aircraft geometry and Mach effects on Clmax, Oswald factors and drag etc etc. I also take compressibility into account which is absolutely needed to do any meaningful dive analysis so I would say it's a pretty advanced simulator. Also, I have the luxury of being able to wait while the numbers are crunched while in BoX everthing is done in real time for multiple aircraft which makes this sim all the more impressive since in most cases the performance we see in game is pretty close to IRL performance.

  • Like 1
  • Upvote 2

Share this post


Link to post
Share on other sites

Ok @Holtzauge. You clearly know what you are talking about. And as such it must come as no surprise that no one should trust your results before you publish your book. Unless the actual mathematical model is made public (not just what is considered, but the actual formulas being used) it is impossible to reproduce your results and confirm they are in accordance with the mathematical model you are proposing.

 

I understand you may want to keep your code private, for future commercial use. It is your work and you should use it for whatever purpose you see fit. But you can't keep the cake and eat at the same time. If you want to use your results to suggest improvements to the sim, or to make a critique to the current state of the FMs, your results must be reproducable.

 

I wish you good luck with your book. 

Edited by Pupo
spell check
  • Upvote 1

Share this post


Link to post
Share on other sites
53 minutes ago, Pupo said:

Ok @Holtzauge. You clearly know what you are talking about. And as such it must come as no surprise that no one should trust your results before you publish your book. Unless the actual mathematical model is made public (not just what is considered, but the actual formulas being used) it is impossible to reproduce your results and confirm they are in accordance with the mathematical model you are proposing.

 

I understand you may want to keep your code private, for future commercial use. It is your work and you should use it for whatever purpose you see fit. But you can't keep the cake and eat at the same time. If you want to use your results to suggest improvements to the sim, or to make a critique to the current state of the FMs, your results must be reproducable.

 

I wish you good luck with your book. 

 

That is of course one way to look at it. Another would be to look at it as a black box systems model in which case you judge the validity of the model by how good or bad the results coming out of it are based on the input.  As I said before, I have posted a lot of simulation results over the years and in general I would say that the results tab pretty well with IRL numbers but if you think something looks off I’m certainly willing to take a closer look at it.

Share this post


Link to post
Share on other sites
6 hours ago, Holtzauge said:

 

That is of course one way to look at it. Another would be to look at it as a black box systems model in which case you judge the validity of the model by how good or bad the results coming out of it are based on the input.  As I said before, I have posted a lot of simulation results over the years and in general I would say that the results tab pretty well with IRL numbers but if you think something looks off I’m certainly willing to take a closer look at it.

 

FWIW I agree with the above sentiment expressed by Pupo. You of all people should know that reproducibility is a key tenet of science (including engineering, which is simply applied science). You and I have discussed this privately in the past and I respect your decision not to release your source code even if I categorically disagree with it.

 

That said, you should not expect to change any minds (or more importantly, game behaviours) with results from an unvetted black box.

Edited by JG13_opcode
  • Upvote 3

Share this post


Link to post
Share on other sites
8 hours ago, JtD said:

Well, the A-20 does slightly in excess of 500km/h on the deck at 3200hp. So it is worse with thrust/drag and modelled as such than a Bf109G or a Bf110G. The latter of which, while similar overall, has a slightly worse span and wing loading than an A-20 without bombs. It has a slightly better power loading, though. All in all there'd only be small differences between the two, still in favour of the A-20 assuming similar efficiencies. In game, of course, there appears to be quite a large difference, owing to the worse efficiency of the Bf110.

 

Well if we pick the standard weights provided, 7500 & 8366 kg  respectively, we get (lower is better):

 

Bf-110G2

Wing loading = 193.3 kg/m^2

Span loading = 460 kg/m

Power loading = 2.57 kg/hp

 

A-20B

Wing loading = 193.7 kg/m^2

Span loading = 447 kg/m

Power loading = 2.61 kg/hp

 

Note: I expected a bigger difference based on the respective empty weights (~5600 vs ~6700 kg), but I guess the 110's amount of ammo for the 5x MG17 +2x MG151/20's weighs more than the extra crew member + 4x .50 cal ammo in the A20. 

 

So based on the above the only advantage the A20 has seems to be span loading, whilst the 110 features a smaller wetted and frontal area, plus the inline engines should reduce form drag over that of the A20, and as such the 110 should suffer less parasitic drag than the A20. In addition to this ofcourse the 110 features slats for improved lift.

 

In terms of induced drag both aircraft need to haul well over twice the weight round the turn than most of the single seaters, which means over twice the lift and therefore induced drag will be produced. Add this to the well over double amount of parasitic drag and you should start to see why these two big birds simply shouldn't be able to match the single seaters round a turn, esp. sustained.

 

In the end I don't really see the A20 beating the 110 round a sustained turn (they should be about the same in that regard based on the rough numbers, whilst the 110 probably boasts a better instantaneous rate), let alone a single seater. Infact neither of two should come close to the 109, and if talking sustained not a 190 either. Holtzauge can plug the numbers into his C++ program and he will without a doubt come to the same conclusion. 

 

 

1 hour ago, JG13_opcode said:

FWIW I agree with the above sentiment expressed by Pupil. You of all people should know that reproducibility is a key tenet of science (including engineering, which is simply applied science). You and I have discussed this privately in the past and I respect your decision not to release your source code even if I categorically disagree with it.

 

That said, you should not expect to change any minds (or more importantly, game behaviours) with results from an unvetted black box.

 

Whilst I most certainly would also like to have Holtzauge's program, mostly because it dramatically cuts down the time it takes to make each calculated prediction, I can respect his decision to keep it private considering:

1. It no doubt took a long time to make

2. How closely it predicts performance in comparison to actual real life figures 

 

Also in Holtzauge's defence he DID provide a detailed list of all the aerodynamic parameters used in his program, and I didn't notice anything important missing when it comes to predicting the speed, climb & turn rate of an aircraft. Thus the only question becomes wether his program can combine all of these figures into an accurate end result, which I honestly don't see why it shouldn't be capable of, however I did double check by doing a few doghouse plot calculations by hand on the 109F4 and arrived at basically identical results to those he's shown from his program. Thus as far as Im concerned it's simply a custom pocket calculator that, providing you plug in the proper basic values, will also show you an accurate end result. The true benchmark is how closely it matches the real life results though, and here I've been mighty impressed so far.

 

 

 

 

Edited by Panthera
  • Thanks 2
  • Upvote 1

Share this post


Link to post
Share on other sites

People use and trust simulators daily without having access to the source code. There are many different CFD and FEM programs out there whose trust is built on the results they produce and not by every user auditing the source code. As I said earlier on, I have posted results from my simulations on-line for years so there is plenty of output to judge the simulator on. Those who still don’t trust my work can simply ignore it like you can with any other posts in the forum. Anyway, I see from earlier replies here and in the now closed acceleration thread that at least some people I have come to respect by what they post here think my results have added value and that’s good enough for me.

  • Upvote 4

Share this post


Link to post
Share on other sites
16 hours ago, Panthera said:

 

Well if we pick the standard weights provided, 7500 & 8366 kg  respectively, we get (lower is better):

 

Bf-110G2

Wing loading = 193.3 kg/m^2

Span loading = 460 kg/m

Power loading = 2.57 kg/hp

 

A-20B

Wing loading = 193.7 kg/m^2

Span loading = 447 kg/m

Power loading = 2.61 kg/hp

 

Note: I expected a bigger difference based on the respective empty weights (~5600 vs ~6700 kg), but I guess the 110's amount of ammo for the 5x MG17 +2x MG151/20's weighs more than the extra crew member + 4x .50 cal ammo in the A20.

 

So based on the above the only advantage the A20 has seems to be span loading, whilst the 110 features a smaller wetted and frontal area, plus the inline engines should reduce form drag over that of the A20, and as such the 110 should suffer less parasitic drag than the A20. In addition to this ofcourse the 110 features slats for improved lift.

 

In terms of induced drag both aircraft need to haul well over twice the weight round the turn than most of the single seaters, which means over twice the lift and therefore induced drag will be produced. Add this to the well over double amount of parasitic drag and you should start to see why these two big birds simply shouldn't be able to match the single seaters round a turn, esp. sustained.

 

In the end I don't really see the A20 beating the 110 round a sustained turn (they should be about the same in that regard based on the rough numbers, whilst the 110 probably boasts a better instantaneous rate), let alone a single seater. Infact neither of two should come close to the 109, and if talking sustained not a 190 either. Holtzauge can plug the numbers into his C++ program and he will without a doubt come to the same conclusion.

 

Whilst I most certainly would also like to have Holtzauge's program, mostly because it dramatically cuts down the time it takes to make each calculated prediction, I can respect his decision to keep it private considering:

1. It no doubt took a long time to make

2. How closely it predicts performance in comparison to actual real life figures 

 

Also in Holtzauge's defence he DID provide a detailed list of all the aerodynamic parameters used in his program, and I didn't notice anything important missing when it comes to predicting the speed, climb & turn rate of an aircraft. Thus the only question becomes wether his program can combine all of these figures into an accurate end result, which I honestly don't see why it shouldn't be capable of, however I did double check by doing a few doghouse plot calculations by hand on the 109F4 and arrived at basically identical results to those he's shown from his program. Thus as far as Im concerned it's simply a custom pocket calculator that, providing you plug in the proper basic values, will also show you an accurate end result. The true benchmark is how closely it matches the real life results though, and here I've been mighty impressed so far.

 

 

Thanks for the vote of confidence Panthera but modelling the A-20 in C++ would take some time and frankly right now my enthusiasm for posting results here has somewhat waned if you get my drift. :dry:

 

Anyway, I know you listed the span loading above which is fine but I think to get an even better feel for how they compare (when it comes to the stationary turn rate) then looking at the aspect ratio is even better and here the difference is quite large: The A-20 has an impressive 8.11 while the Me-110 comes in at a modest 4.6 which would have a large impact on the stationary turn rate. But of course there are other factors like the slats you mentioned and since both have similar profiles (230-series and 2R1) and root profile thickness (18.5%) while the Havoc has a thinner tip profile (9%), maybe the slats and thicker 11% tip profile gives the Me-110 an overall higher Clmax to work with........

 

Edited by Holtzauge

Share this post


Link to post
Share on other sites
Quote

Thanks for the vote of confidence Panthera but modelling the A-20 in C++ would take some time and frankly right now my enthusiasm for posting results here has somewhat waned if you get my drift. :dry:

 

I get it, but don't let the words of a few discourage you mate, afterall physics don't change at the whim of people ;)

 

3 hours ago, Holtzauge said:

Anyway, I know you listed the span loading above which is fine but I think to get an even better feel for how they compare (when it comes to the stationary turn rate) then looking at the aspect ratio is even better and here the difference is quite large: The A-20 has an impressive 8.11 while the Me-110 comes in at a modest 4.6 which would have a large impact on the stationary turn rate. But of course there are other factors like the slats you mentioned and since both have similar profiles (230-series and 2R1) and root profile thickness (18.5%) while the Havoc has a thinner tip profile (9%), maybe the slats and thicker 11% tip profile gives the Me-110 an overall higher Clmax to work with........

 

 

Yes I agree AR is more relevant, I listed span loading because JtD refered to it. That said I get an aspect ratio of 6.84 for the Bf-110 and 8.08 for the A-20, so not a huge difference there :)

 

Bf110 = 16.3^2/38.8 = 6.84

A20 = 18.69^2/43.18 = 8.08

 

 

Edited by Panthera

Share this post


Link to post
Share on other sites

My, bad: The aspect ratio of the Me-110 is like you say 6.84 and not 4.6. So with this, the power loading and the (at least to me) unknown Clmax in mind it's not totally clear who would be king of the hill it seems. ;)

Share this post


Link to post
Share on other sites
Just now, Holtzauge said:

My, bad: The aspect ratio of the Me-110 is like you say 6.84 and not 4.6. So with this, the power loading and the (at least to me) unknown Clmax in mind it's not totally clear who would be king of the hill it seems. ;)

 

Indeed, by my so far rough calculations they should be very similar in the sustained realm, perhaps with a slight advantage to the A20, whilst the 110 more than likely has the advantage in the instantaneous. So it would be a rate vs radius fight which I predict would be quite close with either one having to rely on a mistake by the other.

 

Share this post


Link to post
Share on other sites

Some good news: Both Han and Jason have now replied to the Me-109 turn rate FM report I did earlier on and I will give an update as soon as the discussion has concluded.

  • Like 1
  • Thanks 1
  • Upvote 11

Share this post


Link to post
Share on other sites
53 minutes ago, Panthera said:

Yes I agree AR is more relevant, I listed span loading because JtD refered to it.

 

Lift induced drag equals

Lift²/Span² * A

A being a factor for the rest of (unknown) aircraft parameters and flight condition. Instead of Span² you may also find wing area * aspect ratio, which is the same thing ((span*chord)*(span/chord) = span²). Bottom line is the plane with the better span loading produces less induced lift, other parameters being equal.

So with the numbers given by the devs, the heavier A-20B comes with the 448kg/m, the lighter Bf110G-2 with 461kg/m. So if all other things were equal, which they aren't, the heavier A-20 could turn with the lighter Bf110 using slightly less power than the Bf110, because it produces less induced drag. That's a simple take away from looking at span loading, which you don't get from any other as simple parameter. Hence I refer to it as important.

Keep in mind the slightly lower wing loading (194kg/m² vs. 196kg/m²) and the higher power output (3200hp vs. 2960hp) of the A-20, you can see the A-20 has quite a few advantages over the Bf110 in terms of sustained turning. The Bf110 has, as you mentioned, the cleaner/smaller airframe and leading edge slats, so also a few advantages on its side. On the bottom line, I'd expect the same thing you do:

 

40 minutes ago, Panthera said:

Indeed, by my so far rough calculations they should be very similar in the sustained realm, perhaps with a slight advantage to the A20, whilst the 110 more than likely has the advantage in the instantaneous. So it would be a rate vs radius fight which I predict would be quite close with either one having to rely on a mistake by the other.

 

 

  • Like 1
  • Upvote 2

Share this post


Link to post
Share on other sites
46 minutes ago, Holtzauge said:

Some good news: Both Han and Jason have now replied to the Me-109 turn rate FM report I did earlier on and I will give an update as soon as the discussion has concluded.

 

Where did they post Holtzauge ? Looking fwd for their updates - Thanks for the excellent contributions!

Share this post


Link to post
Share on other sites
4 hours ago, JtD said:

 

Lift induced drag equals

Lift²/Span² * A

A being a factor for the rest of (unknown) aircraft parameters and flight condition. Instead of Span² you may also find wing area * aspect ratio, which is the same thing ((span*chord)*(span/chord) = span²). Bottom line is the plane with the better span loading produces less induced lift, other parameters being equal.

So with the numbers given by the devs, the heavier A-20B comes with the 448kg/m, the lighter Bf110G-2 with 461kg/m. So if all other things were equal, which they aren't, the heavier A-20 could turn with the lighter Bf110 using slightly less power than the Bf110, because it produces less induced drag. That's a simple take away from looking at span loading, which you don't get from any other as simple parameter. Hence I refer to it as important.

Keep in mind the slightly lower wing loading (194kg/m² vs. 196kg/m²) and the higher power output (3200hp vs. 2960hp) of the A-20, you can see the A-20 has quite a few advantages over the Bf110 in terms of sustained turning. The Bf110 has, as you mentioned, the cleaner/smaller airframe and leading edge slats, so also a few advantages on its side. On the bottom line, I'd expect the same thing you do:

 

No worries JtD, I understand why you used it seeing that we were using other ballpark parameters such as wing & power loading too. AR is more relevant when doing the more in depth calculations.

  • Upvote 1

Share this post


Link to post
Share on other sites
19 hours ago, jcomm-in-combat said:

 

Where did they post Holtzauge ? Looking fwd for their updates - Thanks for the excellent contributions!

 

I PM:ed them so that's why I want the discussion to converge before I do an update. :)

  • Upvote 4

Share this post


Link to post
Share on other sites

Seems like the developers have said what they want in reply to the Me-109 FM report I did earlier in February. I’m posting what I have on this in a separate thread instead of here to avoid the risk of derailing this thread.

  • Like 1
  • Upvote 1

Share this post


Link to post
Share on other sites

Can someone generate empiric turn times for all these aircraft?

 

So we can see actual in-game form drag. :)

I think that if someone were to do testing of all aircraft on JtDs chart, and provided the turn times, someone would be happy to generate an empiric set of efficiencies. :)

 

Share this post


Link to post
Share on other sites

I think that's an excellent idea, which is why I'm already working on it. It isn't even a lot of work, it's just terribly boring. :)

  • Thanks 1
  • Upvote 1

Share this post


Link to post
Share on other sites

Well, slow progress, as it really is boring. Anyway, I have checked 11 fighters so far, after the systematic start with all 109's I decided to just get a couple of samples from a mix of other planes. I'm trying to turn near the stall, which makes me take other turning speeds than the ones listed by the devs. I push it as far as I'm comfortable with, so aircraft handling matters a bit. Oddly enough, I found the G-2 more forgiving than any other 109 and therefore pushed it a bit further. All turns done were to the right. Near sea level (50m), autumn map. Standard loadout and full fuel.

 

Two clear losers: Bf109F-2 and La-5S8 - both of them were already losers according to dev data. In particular the F-2 sucks for no reason, not so sure about the La-5, but I don't think it makes sense.

One clear winner: P-40E1, it's much better than is listed in the specs. I only used 42" boost, flying full load and it already turns with the 109's.

 

The Spitfire, flown at more suitable speeds than given by the devs, is a good turner, you can chop off about another second if you enable 16lb boost.

The Yak-7b, already magic by dev data, is even more magic in game. 1.4 times as efficient as the average opposition, it's just unreal.

 

I will get back with when I can motivate me enough to continue.

bla005.jpg

Edited by JtD
Standard loadout and full fuel.
  • Thanks 6
  • Upvote 5

Share this post


Link to post
Share on other sites

Your load-outs were all full fuel and weapons?

 

You would expect the form drag of the 109E to be higher than the F-G series with their rounded wingtips and generally cleaner airframes. On the other hand, it is lighter than they are so less lift-induced drag (as less weight). This may explain why the efficiency is overall similar.

 

You would expect the CLmax to be the same between the F4, G2, G6 as no wing nor tailplane nor fuselage lifting area changes.

 

The efficiency of the P40 is a surprise. However, it is two seconds slower in a turn at combat power than the 109F4, so obviously it can't compete in a turn except when using power greater than combat and of course the F4 pilot then can too. I would expect its form drag to be in the same region as the F-G series 109s or worse as it is generally a larger plane. But then it would absolutely be a brick, which we know isn't correct by historical and contemporaneous accounts. To my mind, the major issue here is the extra weight it has been given, which requires some other black box factor to be tweaked to make it perform plausibly. More plausible would be to increase its form drag or such, and reduce the weight...

 

It's strange that the CLmax is different for the two Spit MkV with different engines.

 

I'm sure more strange things will appear as this continues. But, need to know what weights each aircraft was tested at, yes?

Edited by Venturi

Share this post


Link to post
Share on other sites

Yes, sorry. All planes at standard loadout, full fuel.

 

The efficiency takes into account weight, so this is not an explanation for the Bf109E. It's just odd to have this unclean airframe with rectangular wingtips (bad for efficiency) be as efficient as the later series.

 

As you know from other discussions, I consider the P-40E weight in game correct. What you can do to improve turn time without upping the power, is to take less guns and less fuel. Compared to the Bf109F4, the P-40 carries a lot of both in the standard configuration. It turns good enough to make me stay away from sustained turning contests against it while flying 109, which previously it didn't.

 

Both Spitfires probably have the same clmax, the difference is likely down to testing. I tested the M46 first, and was probably more confident and pushed it a bit harder when I tested the M45 thereafter. Plus I was aiming for 140 / 145mph, which more accurately is 233 / 240km/h instead of 235 / 240 as listed, which would already put the cl of the M46 at 1.28. I just don't keep things constant within 3km/h, so I felt 5km/h steps in the table are sufficiently accurate.

Share this post


Link to post
Share on other sites

Well, the Spitfire by design is a great low speed turner. You can see that it doesn't need to trick physics to get to this performance. Turning at 270IAS is just nowhere near best performance turn for the SpitfireV, 240ish more so. That partially explains the large difference.

 

If StarWars, then Yak-7B, from what it looks like so far.

 

You're welcome. Since the weather outside is too good for outdoor activities, I guess I can post some more figures this afternoon.

Share this post


Link to post
Share on other sites

So, in game test continued, same procedure - standard loadout, sustained right turns near sea level (50m), standard conditions, close to stalling, but far enough away not be stressful. I think for the most part, the table below speaks for itself. Today, with one really unexpected result:

 

I never thought about turnfighting in a Fw190 at sea level, but in game, it's one of the best there is at that. Please note - this is with stab trimmed all the way up. I checked also with stab all the way down, and the time got worse by about half a second. So it has some effect, but overall, the Fw is kind of the X-Wing to the Yak-7B TIE fighter. They are clearly following other FM's than most of the rest does. Not sure why I could push the A-3 further. I tested the A-5 first, so maybe it was a better feeling for the aircraft in the later test.

I never really bothered with stabilizer trim with the other aircraft, I mostly tested with default settings. So it is quite possible that one can chop half a second off the Bf109 times as well, if one bothers to trim a lot. For me, it's just too slow to be useful.

 

The Yak-1S69 clearly handles the worst of the three, I could not push it as much as I could push the other two, so I flew it at a lower Cl.

 

The La-5FNS2 is a lot better than the early La-5, but in turns it's still just mediocre for a fighter of that size, weight and power.

 

The Mc202 looks much better in game than devs stats suggest. Performance wise, that is, looks as such are great anyway.

 

The I-16 is a real disappointment, not only does it handle poorly / stall early, you also feel as if it lacked several hundred of its 1100hp. Gotta say though, flying open cockpit made up for it. It's always a blast.

 

I also had a look at two ground attack aircraft, noteworthy is the high lift coefficient achieved in the Hs129, which imho comes from strong propwash at these low speeds. Both of them are capable of flying really tight circles, but very slowly so. Might still be enough to dodge a 109 or a Yak.

 

Anyway, outside of a look at in game efficiencies, we now have some performance numbers for comparison. Personally I still can't believe the 190, but I have since then tried to fly it more aggressively (against AI), and it worked out.

 

Sorted by name:

bla006.jpg.7480cd9fe142cd1af70f8281ffec3f42.jpg

Sorted by efficiency:

bla007.jpg.cc999a0ecb2ba99322a8e6c7c28f45bb.jpg

  • Thanks 12

Share this post


Link to post
Share on other sites

lol at 190 turn time, so its best at anything on axis side, best turn, roll, speed, firepower, durability, pilot view... and they call yak ufo whats 190 then... a mitical unicorn :)

Edited by 77.CountZero
  • Haha 3

Share this post


Link to post
Share on other sites

Thanks for all the effort you have put in this Jtd!

 

Im wondering the difference the dev data and tha game test have. Im not accusing You or the devs, i just would like to know why the numbers are so inconsistent?

How about all the other data? Data should be as accurate as the real thing what it should represent or its not correct. I dont undesrtand.

Edited by VesseL

Share this post


Link to post
Share on other sites
7 hours ago, 77.CountZero said:

lol at 190 turn time, so its best at anything on axis side, best turn, roll, speed, firepower, durability, pilot view... and they call yak ufo whats 190 then... a mitical unicorn :)

 

Not a helpful post.  Do you think it's kinda biased and disingenuous to pick one aircraft from his testing for your "special" brand of sarcasm?

 

Nothing to say about some of the other aircraft discrepancies at all?  Maybe the Spitfire getting a full 5-7 seconds over the dev listed turn times?  Now I did find it odd the Spitfire turn times are so high according to the devs, but a JtD's excellent testing (thanks JtD) shows a fair few discrepancies on most aircraft.

Edited by ICDP

Share this post


Link to post
Share on other sites
On 4/7/2018 at 1:48 PM, JtD said:

Well, slow progress, as it really is boring. Anyway, I have checked 11 fighters so far, after the systematic start with all 109's I decided to just get a couple of samples from a mix of other planes. I'm trying to turn near the stall, which makes me take other turning speeds than the ones listed by the devs. I push it as far as I'm comfortable with, so aircraft handling matters a bit. Oddly enough, I found the G-2 more forgiving than any other 109 and therefore pushed it a bit further. All turns done were to the right. Near sea level (50m), autumn map. Standard loadout and full fuel.

 

Two clear losers: Bf109F-2 and La-5S8 - both of them were already losers according to dev data. In particular the F-2 sucks for no reason, not so sure about the La-5, but I don't think it makes sense.

One clear winner: P-40E1, it's much better than is listed in the specs. I only used 42" boost, flying full load and it already turns with the 109's.

 

The Spitfire, flown at more suitable speeds than given by the devs, is a good turner, you can chop off about another second if you enable 16lb boost.

The Yak-7b, already magic by dev data, is even more magic in game. 1.4 times as efficient as the average opposition, it's just unreal.

 

I will get back with when I can motivate me enough to continue.

bla005.jpg

I dont want to be the contrarian, especially since I generally agree with your assertions in this thread, but I dont know how how you getting some of these numbers. Especially at full fuel weights. My own tests I have done in the past produced very different results. Different for the dev numbers too. In fact I long suspected the reason for the weird dev numbers was due to being at full fuel in the stat pages compared to my own tests. I do all mine at 200 liters for each plane. The F4 and G2 do not turn the same. The F-4 is quite a bit better. The P-40 is a cow(in game). 

 

Did you let the aircraft "settle" into the turns? It is sometimes easy to overlook, but the airplanes will decelerate slowly until they reach PS=0. Many of the planes you give higher marks for can turn with the 109 seemingly at first if you dont do enough turns to settle the turn. Once they bleed E, say from 180 down to like 130, they stop staying with the 109. 

 

As it stands right now in game the 109 can out turn both the 190 and the P40 in game, handily. Especially the F, and it is only slightly worse than the Yak1 (F4)

 

These are off the top of my head, ill redo my tests later, but the general gist of what happens in game is like this:

 

Spit (16lbs) 16sec.

109F4 (18sec) (1.3ata)

Yak1b(17.5sec)

G2(19)

A3(21)

P40(24) (43)

39(21ish)

 

 

Edited by Fumes

Share this post


Link to post
Share on other sites
8 hours ago, JtD said:

So, in game test continued, same procedure - standard loadout, sustained right turns near sea level (50m), standard conditions, close to stalling, but far enough away not be stressful. I think for the most part, the table below speaks for itself. Today, with one really unexpected result:

 

I never thought about turnfighting in a Fw190 at sea level, but in game, it's one of the best there is at that. Please note - this is with stab trimmed all the way up. I checked also with stab all the way down, and the time got worse by about half a second. So it has some effect, but overall, the Fw is kind of the X-Wing to the Yak-7B TIE fighter. They are clearly following other FM's than most of the rest does. Not sure why I could push the A-3 further. I tested the A-5 first, so maybe it was a better feeling for the aircraft in the later test.

I never really bothered with stabilizer trim with the other aircraft, I mostly tested with default settings. So it is quite possible that one can chop half a second off the Bf109 times as well, if one bothers to trim a lot. For me, it's just too slow to be useful.

 

The Yak-1S69 clearly handles the worst of the three, I could not push it as much as I could push the other two, so I flew it at a lower Cl.

 

The La-5FNS2 is a lot better than the early La-5, but in turns it's still just mediocre for a fighter of that size, weight and power.

 

The Mc202 looks much better in game than devs stats suggest. Performance wise, that is, looks as such are great anyway.

 

The I-16 is a real disappointment, not only does it handle poorly / stall early, you also feel as if it lacked several hundred of its 1100hp. Gotta say though, flying open cockpit made up for it. It's always a blast.

 

I also had a look at two ground attack aircraft, noteworthy is the high lift coefficient achieved in the Hs129, which imho comes from strong propwash at these low speeds. Both of them are capable of flying really tight circles, but very slowly so. Might still be enough to dodge a 109 or a Yak.

 

Anyway, outside of a look at in game efficiencies, we now have some performance numbers for comparison. Personally I still can't believe the 190, but I have since then tried to fly it more aggressively (against AI), and it worked out.

 

Sorted by name:

bla006.jpg.7480cd9fe142cd1af70f8281ffec3f42.jpg

Sorted by efficiency:

bla007.jpg.cc999a0ecb2ba99322a8e6c7c28f45bb.jpg

 

Great stuff JtD, really interesting read.

Share this post


Link to post
Share on other sites
1 hour ago, ICDP said:

 

Not a helpful post.  Do you think it's kinda biased and disingenuous to pick one aircraft from his testing for your "special" brand of sarcasm?

 

Nothing to say about some of the other aircraft discrepancies at all?  Maybe the Spitfire getting a full 5-7 seconds over the dev listed turn times?  Now I did find it odd the Spitfire turn times are so high according to the devs, but a JtD's excellent testing (thanks JtD) shows a fair few discrepancies on most aircraft.

 

Nah i just need to read few post abow my and see how things are going with this, planty of bias there

Also im not suprised with Spit or other but im suprised with 190 so why should i not coment on its majic abilitys :P

  • Upvote 1

Share this post


Link to post
Share on other sites

Turn Times. Sea Level. 200Liters of Fuel. Sustained. No gain or loss in altitude. Other conditions specified where necessary. I just redid all of these to make sure. 

 

Yak1b (full rads): 18.06 seconds. 

 

190A3 (1.42ata) 21seconds

 

P-40 (43inches, 4 guns): 24 seconds

 

G-2: 20.5

 

F4 (Full combat power): 19.5 seconds. 

 

 

Share this post


Link to post
Share on other sites
8 hours ago, VesseL said:

Thanks for all the effort you have put in this Jtd!

 

Im wondering the difference the dev data and tha game test have. Im not accusing You or the devs, i just would like to know why the numbers are so inconsistent?

How about all the other data? Data should be as accurate as the real thing what it should represent or its not correct. I dont undesrtand.

You're welcome. One thing is of course, that the conditions the devs give the data for and the conditions I tested at are not the same. I'm testing as close to the stall as I can, the devs usually give figures for 270km/h, and often at lower lift coefficients. Also, the power is sometimes different. That's why I gave power, speed and cl of the devs figures and of mine.

I also suspect that some of the figures from the devs are outdated, the P-40 figures for instance are still the ones we originally got, but the FM was reworked since then. The aircraft feels very different, it's possible some performance figures changed, and the devs did not update the numbers.

Then the devs once said their figures were flown with a 'robot pilot'. I don't know how this works, what this is exactly. Maybe that thing is not as good at riding the edge as a dedicated human. Different methods produce different results.

And finally, my figures are test figures. I've tested a couple of planes several times, to see how accurate I am and it was OK to about one tenth of a second. But I did not make repeated tests with all aircraft, and I probably even made simple mistakes. I also realized that getting used to a particular plane had an influence, second runs in the same plane typically producing better results if I didn't fly it often. So my numbers aren't the absolute truth, as you can see from from Fumes questions above. They are just test results.

 

4 hours ago, Fumes said:

Turn Times. Sea Level. 200Liters of Fuel. Sustained. No gain or loss in altitude. Other conditions specified where necessary. I just redid all of these to make sure.

Thanks for your efforts as well. Always good to have someone check results.

We don't do the tests exactly the same, but generally your figures should not be worse than mine, given the lower fuel load. Since I did 10 to 20 turns depending on how quick they were and how long the engine power was available (3 mins in the Fw190 for instance mean 10-11 turns until the engine blows), I'm sure little loss in speed or altitude did not matter much. I tried to keep it constant, and would estimate 5 km/h speed difference and 20m maximum between start and finish. When I started I wrote down the times for each turn in several 20 turn runs, and individual turns varied quite a bit (example: 18.5-22.5), but already the averages of 5 turns were fairly close (example: 19.7-20.2). I then went on and averaged at least 10 turns (example: 19.95 & 20.05, with the 20 turn average being 20s). So I'm positive that my figures are fairly reliable. I'll post a track later today, so you can take a look at what I did.

  • Like 2
  • Thanks 1

Share this post


Link to post
Share on other sites

Nice info JtD, if you want to test the most restricted engine limits, you can check the difficulty option "Unbreakable" and you will be able to use any engine setting for unlimited time. The option "unlimited fuel" also makes fuel consumption non-existent, so fuel consumption during prolongued flight doesn't affect the numbers because of lower weight, those are the settings I use for my speed tests :salute: They will still overheat though, the engine won't die but it will start to run rough in the RPMs decreasing performance so you still need to take care of the temps.

Edited by -=PHX=-SuperEtendard

Share this post


Link to post
Share on other sites

Yes, thanks for the reminder. I started with difficulty fairly full to make sure I don't accidentally switch off something important. I also thought I'd want it to be relevant for playing, so testing a 109 at a 1 minute Notleistung would be somewhat misleading, as you couldn't really use it in game. Same thing about fuel consumption, you do use fuel when you do sustained turns.

Well, the latter probably wasn't the best idea, because fuel consumption changes the condition of the aircraft somewhat. But now I wouldn't want to mix different ways of testing.

 

Attached a track of a Fw190A-3 in a six turn run. I didn't quite manage 19.0 with a cold start today, got it to 19.3 in the second try. Hope that's still OK. While speed varies a bit as does altitude, starting conditions are (if instruments can be believed) about 12m higher and 2km/h slower than ending conditions, which is an about 300kJ error (additional energy available for turning). That translates to about 0.25% of total engine power output in favour of the Fw190. I hope my partially not so great flying makes up for it.

Fw190.zip

  • Thanks 1
  • Upvote 1

Share this post


Link to post
Share on other sites
On 8.4.2018 at 4:57 PM, 77.CountZero said:

lol at 190 turn time, so its best at anything on axis side, best turn, roll, speed, firepower, durability, pilot view... and they call yak ufo whats 190 then... a mitical unicorn :)

 

What a BS post!

 

I invite you! You can have this uber 190, the best turn, roll, speed, firepower and durability plane as you call it and I take this mediocre Yak, ok. Should not be a problem for you to shoot me down.

 

Do we have a deal? :biggrin:

  • Haha 2

Share this post


Link to post
Share on other sites

Oh come on, we've had far more comments on Yak flaps or La roll rates. The Fw - in this regard - is clearly up in the same league.

 

But what I wanted to ask, has anyone of you online pilots maybe taken the Fw into a sea level turn fight in the last couple of days to see if the the test results can be utilized in practice? It's still not best at climb or acceleration, it can only maintain great turning for about 3 minutes, so I wonder how useful it really is. Anyone to mix it with a couple of Lavochkins at sea level and came out on top?

 

I only checked against AI, which went great, but it isn't that meaningful. Like I said, I never even thought about turn fighting in the Fw190. Now I've changed my style completely, and it works against pretty much anything except Spitfires, which turn too tightly.

Edited by JtD
  • Upvote 2

Share this post


Link to post
Share on other sites
On 7/4/2018 at 10:24 PM, Venturi said:

The efficiency of the P40 is a surprise. However, it is two seconds slower in a turn at combat power than the 109F4, so obviously it can't compete in a turn except when using power greater than combat and of course the F4 pilot then can too. I would expect its form drag to be in the same region as the F-G series 109s or worse as it is generally a larger plane. But then it would absolutely be a brick, which we know isn't correct by historical and contemporaneous accounts. To my mind, the major issue here is the extra weight it has been given, which requires some other black box factor to be tweaked to make it perform plausibly. More plausible would be to increase its form drag or such, and reduce the weight...

 

You have to be really careful with historical accounts, esp. ones involving comparisons done by pilots based on combat incidents, everything happens so fast in combat and we will never know what went on inside the cockpit of the pilot who was shot down.

 

That said according to all the real data available for both aircraft the P-40 will never match the 109 in a sustained turn and probably not an instantaneous one at low speed either. However AFAIK the P-40 featured a much lighter elevator at speed than the 109, right? In that case this could in many incidents during combat make it seem as though it was able to turn better, as it was simply more responsive in pitch above a certain speed.

 

 

22 minutes ago, JtD said:

Oh come on, we've had far more comments on Yak flaps or La roll rates. The Fw - in this regard - is clearly up in the same league.

 

But what I wanted to ask, has anyone of you online pilots maybe taken the Fw into a sea level turn fight in the last couple of days to see if the the test results can be utilized in practice? It's still not best at climb or acceleration, it can only maintain great turning for about 3 minutes, so I wonder how useful it really is. Anyone to mix it with a couple of Lavochkins at sea level and came out on top?

 

I only checked against AI, which went great, but it isn't that meaningful. Like I said, I never even thought about turn fighting in the Fw190. Now I've changed my style completely, and it works against pretty much anything except Spitfires, which turn too tightly.

 

I tried turn fighting a La-5FN in the 190A3 and vice versa, the 190 lost in both cases. 

 

Should be said though that 1.45 is exactly the CLmax I use for the Fw190 in calculations. The 109's figures are incredibly/oddly low according to your ingame tests though.

Edited by Panthera

Share this post


Link to post
Share on other sites

I think that combat involving continued turning were hardly ever occurring. More than two full turns maintaining altitude beeing unlikely. What I would consider likely is a defensive spirale. And here I would expect the P-40 to easily match the 109. This mostly due to the lighter elevator contols as Panthera said.

 

Still I am surprised at the abyssmal performance of the Spitfire in that spreadsheet. In MP even I can outturn any other plane with ease. Even in those slow speed turning contests right above the grass line.

 

The divergence of how the aircraft feels in combat and how it actually performs is almost comical at times. But I guess it also has to do with an aircrafts ability to pull alpha for the shot. In the Spit, even when you are slow in a tight turn, you can pull up the nose a bit further to get a snapshot. The Yak I find also great for that. The 109 I find terrible in that regard, to me it feels like you have to actually steer it 10 seconds ahead of where you are. 

 

Thus, I think there are some factors coloring anectotal evidence...

 

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

×
×
  • Create New...