Thread Closed

Flawed Flying Models
#16

I have a feeling oleg uses some kind of aerodynamics software program to get a lot of estimations on his fm's. He has mentioned this a few times over the years. Only trouble is a lot of planes don't perform like they were estimated too even by the original designers. Sad

He does get a few planes spot on. These are a joy to fly. 8)
#17

When I can fly a sim aircraft and have it perform like it was said to real life and have it's opposing oposition perform as they did I am one happy camper.

In the old ww2 sim Janes WW2 fighters I found the flight models quite good. How did I know this? Because I could perform the very same combat manauvers I read about in real life and my opposing numbers accurately reflected what I had read they did. It wasn't perfect but it was a very good reflection of what I had read. 8)
#18

that's it though, a lot of it comes down to feelings or with reference to what has been read. not meaning any offence, but the manouvers described could have been beyond a lot of wartime pilots abilities. so pulling them off, for us average pilots, could be something we simply aren't capable of. i'm just recounting a lot of what has been said on this, and remembering a lot of the lengthier discussions relating to performance. one that comes to mind is kurfursts insistence that the k4 climbs slower than it should, and that spits are overmodelled with regard to overheating. the findings were that the in game k4 outperformed it's most optimistic test results, the spits overheating was about right, and the k4 was in fact overmodelled, despite kurfys insistence it was undermodelled and an example of red bias. the fm debates are long, painful and ultimately down to feelings (tagart should tm that particular phrase) as opposed to reference to detailed, hard to come by testing results. there are also the test results based on clapped out, incorrectly reassembled captured examples that can skew expectations, as well as the problem that some of these dodgier test results could have been used for the fm data (for the simple reason that no other extensive testing data was available). also there is the issue that we have certain models of aircraft that aren't particularly representative of the actual models used in wartime - i'm particularly thinking of the tempest we have in game, 9lbs boost model and the most common wartime version was 11lbs or 13lbs boost. which again could cause some of the "this isn't performing as described" issues.
one thing worth noting is that the undermodelled US planes issue is apparently a western only gripe. a lot of russian players complain of a US bias, almost as much as people here complain of a perceived russian bias! if all sides of the issue are complaining of inaccuracies or bias in equal measure then the middle ground of reality can't be far off.
#19

FM debates really are painful since most people aren't qualified or unbiased to make a civilized discussion possible.

I think maddox did a good job regarding FM parameters which are easy to measure and for which data exists. For instance, level speeds are VERY close to test data.

Climb rate comparative difference between aircraft are mostly OK, but there are many which are significantly wrong (most planes climb too well, some are way above even the most optimistic tests - VVS stuff , especially early)

But how do you model acceleration? Turn rate (the single most important characteristic in Il-2)? Damage modeling (should a zero really fall apart from 5 .50 cal hits?)? General handling? For these we have only descriptive or anecdotal evidence without much scientific data.

Here the devs went for "the feeling" more then anything else.
For example:
* In Soviet version of history, FW-190 was overweight, underpowered monstrosity with acceleration of an overloaded flying boat whose only quality was superior level speed to earlier soviet design.
* Someone in the dev team read something about P-38 early compressibility problems and they modelled it for the sake of the game, however not at the speed range or alt where it was supposed to happen. Also the dive recovery flaps were not a bat turn on a push button device.
* Someone wrote that P-51 was superior to Bf-109 in high speed maneuvers, so they created an aircraft in which can pull 15G in a 750km/h dive with two fingers. Hence the disintegrating 51 in the game.
* Il-2 was renowned as a very tough aircraft so they made it ridiculously strong. Il-10 must have been better so they made it almost invulnerable (although Il-10s were massacred by US fighters armed mostly with .50 MG in Korea)
* Based on the British observations of a single 109, someone came to the conclusion that 109 had no controls authority over ~470km/h and that it had a seriously limited maneuverability
* Someone thought that if an aircraft has elliptical wings, it should turn like a Spitfire and thats how we got our Tempfire.
* P-47 was supposedly a good diver, so they made it supersonic (actually it had a pretty low critical Mach and P-51 was notably superior). P-47 actually had lower critical Mach number then Bf-109 Smile

I'm sure I could find more examples, but you get the point, much of the aircraft performance in the game is based on devs perception or "feeling" as well as historical performance data. Also, one must remember that PCs do not have and will not have sufficient "horsepower" to run real aerodynamic simulations for quite a while.
#20

15/JG52_Riddler Wrote:But how do you model acceleration?

Accel is function of excess power, if plane have correct climb rates at different speeds it's accel is correct too. It's relatively easy to make accel calculation when you know aerodynamic properties of plane.

Hard part in FM modeling is to give "character" to a plane, some things are very hard to express with numbers.

FC
#21

FC Wrote:
15/JG52_Riddler Wrote:But how do you model acceleration?

Accel is function of excess power, if plane have correct climb rates at different speeds it's accel is correct too. It's relatively easy to make accel calculation when you know aerodynamic properties of plane.

Climb rates are more often incorrect then correct (most being too high)
#22

It often appears to me that when an obvious problem is identified in IL2, be it FM load out or whatever. It more often than not leans towards the game side than history. By this I mean, more power, more bombs, shorter distance to target etc. etc.

After all it's main function was to sell and therefore it had to appeal to a wide audience. The majority of fighter pilots in WW2 never shot anything down and even the top experten never got as many kills as the top jollies on hyper lobby over the same period of combat. This says allot to me about how realistic it is.

So where do we go from here. It's been made a taboo subject, and even if it wasn't there are few who could sort it properly, ie. give the numbers character, to paraphrase FC. Even then if it was handed to most people on a plate they probably wouldn't like it.

IL2 flight and weapons modelling is, I feel, one of life's little ironies :mrgreen:
#23

I had the intention to come back and write about which models I believe are wrong, but after reading some posts I think that I am misunderstood somewhat. I never thought that IL-2 offer real aerodynamic simulations.I did CFD in the uni and I know very well what it means to perform a calculation about a certain part e.g an airfoil over just a simple set of parameters.Even calculating the static forces that are acting on this complexity of beams and spars inside a fuselage through a dedicated serious engineering program is a time consuming process, even if we assume that all the revelant engineering data were obtained.
Still I think that given the restrictions that were taken place,especially in the capability of home systems, it was an extremely good effort and the impression that was given to me was that the developers tried really hard to make it seem as real as possible.I think that most people don't respect enough Oleg and his collegues for this.
It is too long to explain from here (unfortunately), but in most planes that are represented in IL-2 what we experience is justified by the engineering and aerodynamic specs of the REAL planes.This is what I always tend to look at before I make a judgement about one or the other flying model.
Question No1:Are they real FMs up the level of military simulators?
Of course not!
Question No2:Are all of them near perfection?Are there any faults on them?
Yes, definetely they are faults in some models.
Question No3:Can we just -at last- trust Oleg a bit when he was approving the one or the other model, and put aside our fetish and our surprise when some flying legends do not appear the way we always imagined?
I 'll let people to answer that question to themselves.
#24

High Plains Drifter Wrote:I have a feeling oleg uses some kind of aerodynamics software program to get a lot of estimations on his fm's. He has mentioned this a few times over the years. Only trouble is a lot of planes don't perform like they were estimated too even by the original designers. Sad

He does get a few planes spot on. These are a joy to fly. 8)

The flight behaviour is calculated on the fly. It is much more sophisticated than just producing turnrate vs. speed vs. altitude tables.

One should never jump to conclusions

Remember the complaints about the P-47 climb rate a couple of years back? After several long threads someone posted a document that showed people were using references that stated the instantaneous rate of climb figures and comparing them to continuous climbs to altitude in game. If you used actual figures for extended climbs with an open radiator Oleg's sim matched perfectly.

Regarding Veteran's accounts: one Soviet ace is convinced to this day that an I-16 was about as fast (or faster) than a Bf-109. He not only engaged in combat but scored victories in that aircraft...

So, no Oleg didn't make massive changes to the FW-190 (just tweaked stall effects at one point), we simply got better at flying her.
#25

VT-51_Razor Wrote:The TBF/TBM for sure, as well as the Kate and Jill (B5N & B6N) The TBF is at least flyable. You can trim most of the nastiness out of her, but the two IJN attack planes don't even have any trim controls!! If this group ever decides to do something with the FMs of some of these planes that they have brought to life, those three ought to be at the top of the list. At the very least, you might be able to turn the two IJN planes into decent flyers by just adding trim controls in all three axes.

AMEN!!!!!!!! I would not support in any way modifying the FM of standard flyable aircraft because although they might not be completely correct I like to believe Oleg & co made an honest effort and their "best guess" is just as good or better than anybody elses. The problem is with the FM of some of the new AI flyable planes, especially the ones listed above. I think the speed, acceleration,etc... are all pretty good but the torque effects are insane..especially on the TBM which has a humongous wing which is also very long (giving it a ton of leverage on the airframe). There is no way in hell that the torque effect on this plane is even close to being correct.

The TBM had a 1700hp R2600 engine while the F4U had an 1850 hp R2800 engine. The TBM was heavier with more wing surface (ie: a lot more resistant to torque rotation) yet it has at least 10 times the torque effect of the F4U. This is the type of thing I would support changing.
#26

15/JG52_Riddler Wrote:FM debates really are painful since most people aren't qualified or unbiased to make a civilized discussion possible.

I think maddox did a good job regarding FM parameters which are easy to measure and for which data exists. For instance, level speeds are VERY close to test data.

Climb rate comparative difference between aircraft are mostly OK, but there are many which are significantly wrong (most planes climb too well, some are way above even the most optimistic tests - VVS stuff , especially early)

But how do you model acceleration? Turn rate (the single most important characteristic in Il-2)? Damage modeling (should a zero really fall apart from 5 .50 cal hits?)? General handling? For these we have only descriptive or anecdotal evidence without much scientific data.

Here the devs went for "the feeling" more then anything else.
For example:
* In Soviet version of history, FW-190 was overweight, underpowered monstrosity with acceleration of an overloaded flying boat whose only quality was superior level speed to earlier soviet design.
* Someone in the dev team read something about P-38 early compressibility problems and they modelled it for the sake of the game, however not at the speed range or alt where it was supposed to happen. Also the dive recovery flaps were not a bat turn on a push button device.
* Someone wrote that P-51 was superior to Bf-109 in high speed maneuvers, so they created an aircraft in which can pull 15G in a 750km/h dive with two fingers. Hence the disintegrating 51 in the game.
* Il-2 was renowned as a very tough aircraft so they made it ridiculously strong. Il-10 must have been better so they made it almost invulnerable (although Il-10s were massacred by US fighters armed mostly with .50 MG in Korea)
* Based on the British observations of a single 109, someone came to the conclusion that 109 had no controls authority over ~470km/h and that it had a seriously limited maneuverability
* Someone thought that if an aircraft has elliptical wings, it should turn like a Spitfire and thats how we got our Tempfire.
* P-47 was supposedly a good diver, so they made it supersonic (actually it had a pretty low critical Mach and P-51 was notably superior). P-47 actually had lower critical Mach number then Bf-109 Smile

I'm sure I could find more examples, but you get the point, much of the aircraft performance in the game is based on devs perception or "feeling" as well as historical performance data. Also, one must remember that PCs do not have and will not have sufficient "horsepower" to run real aerodynamic simulations for quite a while.

I agree, except... :wink:

Some notes:
- The FW-190 has a tremendous rate of roll (they certainly got that right)
- P-38s did use combat flaps (so at least its modeled - even if not accurately)
- I can shoot down the Il-2 in a Mig-3 (and it only takes three runs! A Bf-109 takes four)
- I think the Bf-109 elevator is modeled in game and the effects are a result of calculations - not history, although I could be wrong).
- There are two claims of P-47 pilots who believe they broke the sound barrier (although I don't have the reference on hand).

No comment on the others (especially the Tempest which is a wierd bird in many ways).
#27

FM's have nothing to do with weapons, whole different story. I can agree the Avenger has bad FM. To much Torque + Trims Screwy, the Avenger was never a flyable aircraft. Unless added into the .mis file useing bornplace, now it has a cockpit. Yes i agree its off but theres not much you people can do about that at this time.

FM's contain info on reaction time on all control surfaces. How Heavy the aircraft is, How many engines it has, speed at altitude. Speed related to pitch. Choke points, and a bunch more things.
#28

Have some sort of checks and balances been considered re: FM changes?

I ask because my fear is that "new FM" might become "correct FM" if that new FM is popular enough
#29

Former_Older Wrote:Have some sort of checks and balances been considered re: FM changes?

I ask because my fear is that "new FM" might become "correct FM" if that new FM is popular enough

No... because FM's wont be changed for the current standard version of the game. At least not for a while. Not here anyway.
#30

Hellzone Wrote:FM's have nothing to do with weapons, whole different story. I can agree the Avenger has bad FM. To much Torque + Trims Screwy, the Avenger was never a flyable aircraft. Unless added into the .mis file useing bornplace, now it has a cockpit. Yes i agree its off but theres not much you people can do about that at this time.

FM's contain info on reaction time on all control surfaces. How Heavy the aircraft is, How many engines it has, speed at altitude. Speed related to pitch. Choke points, and a bunch more things.

Where is the AI, BTW? Is the min alt they will fly a plane by plane thing? Is the AI they use load dependent (ie: they know they have a fish, so they act like a torpedo plane)?
Thread Closed


Forum Jump:


Users browsing this thread: 5 Guest(s)