Jump to content

Re-Post: Updating the SE5a - 10 Sept , 2011 - An.Petrovich


Recommended Posts

Posted

I re-post this as I think it provides fascinating insight into the process An.Petrovich took while updating the SE5a back in 2011 for RoF. I haven't edited or touched the text but note there was one power curve chart unavailable by the time I made the copy. It doesn't impact the narrative.
It's purely meant to clarify his process and correct some false assumptions that have taken root since.  

Given that the FM is essentially the same in Flying Circus the post is still relevant and sheds light on how he tried to address community concerns, while maintaining certain target performance metrics and include new FM data that had just become available to him. It's clear he wasn't averse to workarounds like using the experimental ultra-high compression engine(5.68 : 1 ) in order to hit his targets. Ultimately though, you can sense his frustration and confusion at being unable to hit all the metrics and not understanding why.  

.......................................................................................................................................

 

 

 

An.Petrovich

 

Posts: 565

 

Posted 10 September 2011 - 23:12

https://riseofflight.com/forum/topic/22802-se5a-fm-review-fixes/?p=316642

 

Hi guys! :S!:

 

So, I have some good news for you: the fixed S.E.Va finally is already under beta test now, and at last I’d like to make a summary report here about its final performances.

 

As you know, the main S.E.Va FM problem was a wrong engine revolutions («glass engine» in the dive), and also it had a very fast climb, and a little high stall speed. For fix these issues I had to revise many performances, as they all depend on each other. It was really hard research (as usual for any airplane) because many sources contain conflicting data (as usual again). I think I’ve found the compromise which seems like a true, but many questions still not answered.

 

Now, some facts from the 777 hangar:

 

180 (200?) h.p. Wolsely W.4A Viper engine.

 

As I wrote before, we have 2 different sources with the engine power curve,

the first source:

 

v7zZByhwo5hi1uvjnBbra8HTL1EENCT_sjL-2KRs1-BKSULm_sscaKhvlj5NC4x7tZaHedWCJBzy1_tIa0y2vFQ8iFVvBPzz0fnDO81Q-uAhQSP_zAxe8idOufnvThW31tPesRTyBKbN9LCDDPWBgQ

And the second source:

 

Image (missing)

 

This difference is due to the different compression ratio. I've tested both engines in RoF, and chose the high compression engine, because it provides the best fit of all the characteristics of the aircraft. As a result, the maximum error of Viper power in RoF does not exceed 0.5 BHP (0.3%) against the source curve from 1200 to 2200 RPM.

 

BTW, the heat balance of engine has also retuned.

 

Airframe aerodynamics

 

Many thanks for piecost, who found the polar curve of S.E.Va! So, the aerodynamic performance have also retuned, and now our S.E.Va in RoF has Drag and Lift the same to the real airplane:

 

uPrByl5pEB-6PiIGnmokLgzaGudw33R8iN6cj33EOakQSlC97t0tggjcxQr3cmnwi72MS2dF8pGMWEZ_5iZ5B9SeBbFw1CAk2I8q0QIfaYB01P_Ia8s4dzEufsHVhKYzeTF9mKmV7XhtUuTk44bnyw

 

dWmoHSTAo08BZ4TxPeE_YTYef6yWeOm8xiOqdREOECYohjp9gbf7fBBwb9HO6K9dFOiXbUbgi2OZTeGTf16HyM11WhSX6jEfOnv8TwnXhD-7S6s6CRGIqi0T5n10A-oHbC3edepdx2lBVZdS1k3W-Q

 

(Note: the Drag curve on this picture has recalculated for 2050lb. weight, although the S.E.Va in RoF has 1940lb. gross weight; moreover, the stopped airscrew in RoF has 52.5lb. of drag at 100f.p.s, exactly according to this source, with error less than 0.9%).

 

The maximum CL, in accordance with our discussion (started here), has been refined using this source. Taking into account the stall speed 54-56mph for 2050lb. the average CL_max value is 1.09. Finally, the stall speed of S.E.Va in RoF is about 45-48mph on the airspeed indicator w/o thrust, and less than 40mph at full power.

 

Since aerodynamics of the wing have been modified (according to new CL and CD curves), then the roll rate (aileron performance) is also changed as a consequence. Now the roll rate is worse then before.


 

Airscrew

 

After airframe aerodynamics and engine performances have been modified, all I can do to ensure the right characteristics of the aircraft (I mean the climbrate and speed) - this is an airscrew tuning. All the more that «glass engine» was because wrong blade aerodynamic performances.

 

The general geometry of the airscrew I've taken from this source:

 

 


 

After analyzing the drawing I put into the RoF basic geometric parameters of T28118 – as an original airscrew for S.E.Va with Viper engine:

 

Attached File  Airscrew.zip   23.72KB   54 downloads

 

And then I’ve done many iterations to found aerodynamic performance of blade (CL and CD polar curve) that match general performance of typical British 2-blades fixed pitch wooden airscrew:

 

NrO1j1NWXYCuG027f3Brpc0ScCy7L06N0_swGc0B59v0kXdNP9G17PSb53VY2kA3X6GA-IT6YbXVq2NIZOPIXijvEKnzQF4_6HJ6-VNyVZAEymLM-1O2WCV7u27KmfhUwUhiGorvxu0Wa7e663-Jbw

6Tl9C0jOCqoW09TRohKxTPZYATx4NwJDc8RQA02hCKxl6LYMFj8-SwEv71Pkl4s66c9quzqjbW_m7DptQD07fsoVq1Ym82uCTBFnyvALXAtn0STtvSvEMbqxmjr4UH_SJJ3zcZ18obQkqaOrJ879Cg

Selected CL and CD curves for the blade provide a good agreement between the results of tests in the game and calculations of thrust and RPM using charts above:




 

FFWTkiXt87AM8UH_4ZfHWB4E5UtfXhM210Nx9kXLV8Ql0s-NU5ZB7s9hQp1yECqm9bpt4OK0MXqqsgEJh_E7DZR4xStAOKP7ItsUCqzDKNQrlufqEB6VRXMNasgZ1Dn8ndrWZd-Gm6OLGIOqG1UyNQb5XQblN175J7DS-lGESqC_wQivBosoZpahd19frlmpKczyfkqpD4WhJ1V63gze-QO9LVZ6BJgBHzkfBTVBv06xjkHzx98fZ--bxivy3EYkqRZsxNzVV6sT3r45k8q7fK0fyTpP-JibRe7NQ224c4jw

 

And most important, this airscrew gives needed performance of airplane in general…




 

Airspeed, climb rate and ceiling

 

Well, guys, as we have already discussed about the maximum speed, I just publish the test results here:

kuPEhgPyZ-PlYFyc9aoqNAl1i_xOPexuxEbG1Hc7boYzP9uE-TVE0vedYrCsglY5W3SF0vjlsaeHpSNDUYK5UKm6e2uc4D94xlBX59S7o7LEYY0wevbu2GTuhFBnLufrgX42DU0NsMWAjbOd4nc0Qg




 

The climb rate… yea, it’s a really most hard task, because as usual every source data is uneven, they often contradict each other, roughly approximated, and we very seldom know anything about the conditions (the weather, airplane weight, fuel, ammo, etc.) or when the timer started: in the moment when the pilot set full throttle, or when wheels were up from runway, or may be when the pilot started climb after level flight at some altitude? We don’t know anything. So, this is a always investigation, painstaking research and walking around strange puzzle, analysis and many trying to know the truth. The same words about the ceiling, but the climb rate, of course, is more important for gameplay.


 

For determine what data can be trusted and what are not I do always analysis the time data from each source: study how its derivative depend on altitude, calculate several theoretical climb paths, with taking into account the various vertical velocity at sea level, and various ceiling (as is known, the climb rate often very close to linear function of altitude). I combine some data from different sources each other, shift all of them on the same time for acceleration or deceleration the beginning of climb, to obtain a uniform vertical velocity dependence on altitude, w/o the time, which may be was spend for moving on the runway, or on the contrary add the time, which may be need to spend for takeoff and for climbing to some altitude. After this analysis sometimes it becomes clear, that airplane can't climb to some altitude so fast, because the time in the source is from level flight, not from start on the ground, f.e. Or sometimes it becomes clear, that climb time to some altitude can't be so much, because then the ceiling must be too low (may be this data for the airplane with bombs, or it's a mistake and value is for the other engine, etc.). I write this to you just for your understanding, that sometimes the source contains data that cannot exist together, because so many reprints, errors, and speculations have accumulated in the books.

 

h8SYfV7zwFm1ZrnjunxNg-tp0b1mXJO6_8J4X47aUVy61b25BtmpExj_sUL8mRrrRkPN40w0cDSU_xR_3wLZ9ncOgyIV8JkmSpLefMEKBmhafCHbNL0CIPJih3r4mS_7ng9oqWZJVpTv_14UkY7Rqw

 

-rSXJj3_ozxhTZdRPs6x0-T0_iT6GPp91OWoZgD_ZN80VhVsQAaZ0vNTvmU0TrN9gR423-CpYLit6xVj3XjXRrK0souFvShEFGkiF0849gXSfVAZ6n16ei5HgsWsMciI7NrDYLHKH7e5na_G08MJHQ

VtT0hiVaScGd4Cou89u4y-o62yQtGsHMXYBr9rHgrfS0phBg9Cs9E_fexDuFqLIJoKqUYPd7Pdfy_ECqMdQXw8YJ25CnoxvzQBEiq8G9JogGZOnb0LFajF5bb94m7o6PgY9oqHdCMrpikIRp9xfueA

As usual, if airplane FM is developing and tuning in accuracy, the climb rate doesn’t need for additional tuning, and it is quite close to some of sources. In this case I usual believe this source, and finish work. But sometimes I have to spend much time for understanding: what’s wrong? Is here our mistake, or maybe it’s a wrong source data? Nobody knows.

 

Anyway, the new S.E.Va climb rate seems quite close to the one of combinations of source data. It is a bit faster all the time, about 10 seconds for any altitude, except 10’000ft (there is a bit delay). I don’t know why this time shift is present, may be because in real life pilot is starting easier, and moving throttle smooth or something else. The most important is that vertical velocity is quite realistic at any altitude. However, the theoretical ceiling of S.E.Va in RoF (where is the level flight is possible yet) almost w/o fuel is 6700m (with accuracy less than 25m), and it seems very-very close to 22’000ft (error does not exceed 18ft, 0.1%). That’s why I don’t believe in 19:55 or 20:50 to 15’000ft with full fuel loading (I note, that we test climb rate in RoF from the start with the place with full fuel loading for all airplanes). If 19:55 would be possible, then ceiling must be much higher, but I’ve never read anywhere about ceiling more than 22’000ft for S.E.5. Moreover, the theoretical ceiling of S.E.Va in RoF with full fuel loading is 5900m (with accuracy less than 25m), and it again seems very-very close to 19’500ft (error does not exceed 143ft, 44m, 0.7%). Magic? I don’t know. So, the climb rate for S.E.Va in RoF is:

 

TuBYrs5SbjCsW-9v-FvEwQaS03vfukWeuln1EbMDHgb2YZVVmje1PCDuiN2ME3p80_8O4X8h8ofg55oOba-jeDlPUmVF2-2gm1L0b1n1W3CRddZlpLqx-Bm2uWZdLk3Y3kHE22Wo3iJhO6vjo1t-qw

 

Of course, I clearly understand that it’is climbing to 5000ft up to 14 seconds faster, and the error is 5.1% (this is the biggest error here). I have tried many ways to fix it and spend about 3 days over the plan for this job – but I have no good result. Each method corrupts other performances more than improve climb rate. And to the end, I don’t understand one more thing: how is it possible: if S.E.Va in RoF has a right engine power curve; has a right Drag and Lift with and w/out aircrew; has right performances of T28118 airscrew, with consider an its slipstream and additional drag for airplane parts which are in the slipstream; and at last, S.E.Va has a quite right horizontal speed at any altitude, has right ceiling and almost right climb but.. look at this yourself:

 

VrafFBAxzFSiMr9krJsWicIkmOigd123MPlRN0ly_eg1CeciB7n0XoWFgbFRayM3VVzm0EW35m8kCAL0v2G55k0vpWYdEoMV7pZblJaMq6tQZ89L5FspptWxlzbtIOGl5Hyg6Wd3Cb1HbCTa9J8vMQ

 

- is here some another S.E.Va with another Viper engine??? I don’t understand it all… No any weight loading or low engine compression can give the same result to the this source. On this picture the airplane is dramatically slow and it's a very pure climbing. May be somebody tell me - how is it possible?.. and… welcome to the club, who has read all until this line. :)

 

(and sorry for my English, maybe I wrote something not clear here…)

 

Sincerely,

…………………………………………………………………………………………………..


 

Source Climb Rate images added for clarity

17hOsXx4hc91TC3bCHlcUCriyTEMgN12k87Jq6QMEcQlXyUCMX8NbPfyiHfUgvktq5WlPxHGucINTATPLNUqYj_wUyjek_jFWVyqZPwkYBDZqOdC4UVtpV-z5ZOFjUGQpdwqojP6t5nXSscFx-pVhQ

 

6Nn9oeGGYzfVTsjefXsY1pPmwcgOEC-JnivbjoIaV5svB2SxoYXsdKx8vvKtg3Nv9T-cZLZ-35iIKPoCnzgxUc7BVAm61KxGUEauyePBIDNdYygVqsUC5OMM0pSvrywuGzA7vyuvOkmnFaITbUzU5Q

  • Thanks 4
Posted

Very interesting @US103_Baer, and thanks for posting!

 

First point: There is no mention of how the turn performance was modeled. And since speed and climb occur at low to intermediate Cl’s, one cannot use speed or climb data to tune the turn rate FM models since this occurs at much higher Cl’s. Consequently, it’s very difficult to do this unless you rely on modeling FM’s according to the physical properties of the aircraft.

 

But since Petrovich’s post is about speed and climb modeling, let’s focus on that. As I am reading it, the way he refers to how he is doing the tuning, hints that the models are being tuned to data for specific airplanes and that the FM modeling is not generic, as in models of the aircraft based on their physical characteristics. And earlier this year I had a brief exchange with some other Il-2 developers concerning another in-game aircraft and the approach was the same here: “This is the data we are tuning to and look how close it is to the curve”. When I suggested that that particular data set may not be representative, and that other data might be more appropriate, this was dismissed and there was no interest in any further discussion.

 

But for sure, this approach of picking a particular data set and tuning to it could work if there was a statistically significant amount of data for each plane and it’s totally clear under what conditions it applies. But there isn’t. So the problem here is you have no way of knowing if the data you have in hand is representative. And if you dogmatically tune to this specific set of data, the plane could end up being a high-, mean or low end performing model, or even worse, not representative at all. In addition, there may be measurement errors in the data, which can add yet another level of confusion.

 

As an example, I too find the climb speeds in the final two blueprint figures above very low. And in the same way as Petrovich, I have no ready explanation for them. OTOH, they could be the result of a position error in the measuring equipment. In fact, it’s not unusual with a position error of 10 mph IAS or more in some aircraft. But there is simply no way of knowing if this applies. But I too get the S.E.5a’s optimum climb speed at a much higher IAS, in the order of 75-80 mph just as Petrovich’s blue curve shows. So I understand his frustration.

 

But all of this just highlights the problems of tuning the aircraft models as separate entities: You get what you tune to, and with this approach it will be all too easy to get FM’s that are not connected to the physical characteristics of the real aircraft. True, it could (and should!) work, if what you tune to is data corrected for measurement errors done for all aircraft at 15 C standard atmospheric conditions on mean performing aircraft. If not, you could end up with FM models that make no sense and that may not even tab with pilot accounts (on meta level!), and people scratching their heads trying to make sense of it all……

 

Finally, too bad Petrovich has left the team. This old post about the S.E.5a and others he did before he left include questions and musings and indicate no fixed mindset. Rather to the contrary, his interactions with the forum community always hinted that he was open to new ideas and input, not dogmatically fixing on any particular data set. In DCS, I have had very good discussions with one of the developers there about the P-51 Mustang and Fw-190 turn modeling FM’s and more. We shared and discussed data, had back and forths, but usually ended up in some sort of agreement. And this I think is the key to developing any good product. Not just good flight simulators. Use the information and people available, and always be open to new ideas and input! ;)

  • Thanks 1
  • Upvote 2
Posted

Interesting insights, thanks @Holtzauge. I wasn't really aware that it was basically one FM per aircraft. But since this is stemming from the old days of RoF, maybe it was a reasonable path to follow at a stage where the "global FM" wouldn't have been ready to deal with monoplanes, biplanes and triplanes in the same way? Would it have been faster just tweaking an existing system for each aircraft rather than striving for a global solution, that, until perfect doesn't make the aircraft perform as they should across the bench? I would think there are a ton of legacy reasons stuck in these sims. I wonder if the planes FMs in GB are still made that way.

 

I see the "porting" of aircraft from RoF to FC in this sense, as you basically just make the aircraft follow the same chart lines for performances in the new sim engine. They clearly use a different base FM, and now devs have to come up with flight parameters that make each aircraft follow the chart lines they used for RoF. The differences we see is the evolution of the base FM.

 

Given the planes FMs had to made new from scratch, I really find it a shame that they didn't take the time of re-evaluating their index charts, and be it just for not having a discussion.

Posted

Thanks for taking the time to put this up @US103_Baer. There's a lot to digest here. I've just had a rudimentary squizz at it and what stands out immediately is Anp's charts/references to a "180hp Viper engine" which he possibly assumes is a 200hp Viper? And maybe influenced his "compromises"? There was a 180hp engine but it was a direct drive Hispano-Suiza not a 200hp Wolseley Viper as he have in-game. It's likely that early Se5a's used this engine in some versions. In fact, "The Vintage Aviator" Se5a replica uses the 180hp Hispano-Suiza. 

 

I'm no aeronautical engineer but could this have a bearing on why the in-game Se5a is a bit of a dog? Or am I barking up the wrong tree. 

 

 

 

 

 

 

 

 

Posted
53 minutes ago, ZachariasX said:

Interesting insights, thanks @Holtzauge. I wasn't really aware that it was basically one FM per aircraft. But since this is stemming from the old days of RoF, maybe it was a reasonable path to follow at a stage where the "global FM" wouldn't have been ready to deal with monoplanes, biplanes and triplanes in the same way? Would it have been faster just tweaking an existing system for each aircraft rather than striving for a global solution, that, until perfect doesn't make the aircraft perform as they should across the bench? I would think there are a ton of legacy reasons stuck in these sims. I wonder if the planes FMs in GB are still made that way.

 

I see the "porting" of aircraft from RoF to FC in this sense, as you basically just make the aircraft follow the same chart lines for performances in the new sim engine. They clearly use a different base FM, and now devs have to come up with flight parameters that make each aircraft follow the chart lines they used for RoF. The differences we see is the evolution of the base FM.

 

Given the planes FMs had to made new from scratch, I really find it a shame that they didn't take the time of re-evaluating their index charts, and be it just for not having a discussion.

 

Maybe I should clarify a bit better about my theory regarding the in-game FM’s: What I’m saying is that I believe the evidence points to the Il-2 FM’s being one per plane in the sense that you input values for Cl and Cd for example, and that this is how you model the wing, and which was why I think Petrovich showed how diligently he tuned his model towards the RAE data he had.

 

What I do on the other hand when I do a new aircraft in my simulation is this: I enter the wing area, wingspans of the wings, gap, e as a function of Cl, Clmax as a function of Mach, engine power as a function of altitude, propeller rpm, solidity, diameter etc. Then I simply cut the simulation “loose” so to speak. And the satisfying part here is that usually the results tab very well with historical data. No tweaking needed. And this goes both for WW2 and WW1 aircraft. And I have done close to 60 different aircraft to date and all of them conform quite well to historical data. Granted, I use some historical data as well to tune to, but since those data points in some sense are “valid” for all aircraft, I immediately spot outliers in data.

 

So in summary, my approach is to model the physics and let the chips fall where they may. The other approach is to model each individual aircraft dialing in the Cl/Cd polars etc. you believe are representative for each. Both methods work, but the latter is of course more susceptible to errors since the flexibility also allows you to dial in whatever you want.

 

And yes, my interpretation of the "porting" is much the same as yours. I first thought that the FM's would improve in the step from RoF to FC since it was a new game engine utilizing the physical properties of each plane and higher fidelity would follow automatically but apparently the FM tuning is done in a similar way still allowing things like the Albatros’ optimistic turn performance.

 

11 minutes ago, ST_Catchov said:

Thanks for taking the time to put this up @US103_Baer. There's a lot to digest here. I've just had a rudimentary squizz at it and what stands out immediately is Anp's charts/references to a "180hp Viper engine" which he possibly assumes is a 200hp Viper? And maybe influenced his "compromises"? There was a 180hp engine but it was a direct drive Hispano-Suiza not a 200hp Wolseley Viper as he have in-game. It's likely that early Se5a's used this engine in some versions. In fact, "The Vintage Aviator" Se5a replica uses the 180hp Hispano-Suiza. 

 

I'm no aeronautical engineer but could this have a bearing on why the in-game Se5a is a bit of a dog? Or am I barking up the wrong tree. 

 

 

 

 

 

 

 

 

 

The engine modelling could for sure have something to do with it, but my simulation model of the S.E.5a is not that far off from the one in FC in terms of speed and climb. Granted, I am a bit more optimistic regarding the climb performance in my book, but it’s the turn and energy retention where the FC S.E.5a really stands out.

 

  • Thanks 1
354thFG_Panda_
Posted

@Holtzauge I have some queries regarding aerodynamics. I tried to DM but no messaging available. How to get in touch?

Posted
9 minutes ago, Red_Panda said:

@Holtzauge I have some queries regarding aerodynamics. I tried to DM but no messaging available. How to get in touch?

 

My Bad: Apparently I'm at 102% of my mailbox capacity and I did not get any warning message. I'll clean it up later today and then you should be able to DM me. 

  • Thanks 1
Posted
1 hour ago, Holtzauge said:

So in summary, my approach is to model the physics and let the chips fall where they may.

Thanks for the explanation. In terms of getting a computational method for a simulator, this is rather ballsy, given the nitpicing clientele finding (of course, we do that instead of just playing the game) deviations in published lift values that scream for explaining. But you think that would work for real time comutation in a sim?

 

 

 

Posted
22 minutes ago, ZachariasX said:

Thanks for the explanation. In terms of getting a computational method for a simulator, this is rather ballsy, given the nitpicing clientele finding (of course, we do that instead of just playing the game) deviations in published lift values that scream for explaining. But you think that would work for real time comutation in a sim?

 

 

 

 

For real time applications like in game then my method will not work because it requires iterations. However, I could easily write out Cl Cd polars just like Petrovich used for tuning the S.E.5a and this would solve the problem I think.

 

I have no doubt it could be done and it only requires, as Dr. Merkwerdigliebe would have put it, the WILL to do it!

 

 

  • Like 1
  • Haha 1
Posted (edited)

@OP, many thanks for posting that info. from RoF's days of yore -- always interesting to see how FMs were tweaked and tuned in that particular sim. Makes sense then that the devs. are somewhat leery of devoting too many resources to FM tweaking in the (continued) FC series, considering, most likely, that aircraft FMs are still done on a per-aircraft basis (a very slow process; speaking from personal experience in First Eagles 2, even though there we use static tables, modular data initialization files, and whatnot -- none of this newfangled dynamic physics modeling ?).

 

At any rate, good luck to you gents' and here's hoping that you at least inspire more creative (open-ended) thinking about FM tweaking in the FC series. I suppose that a good compromise between an extensive FM overhaul across the board and no FM changes at all would be, if the devs. are interested of course, to add a few extra dials to the in-sim menu in FC, to have user-tweakable climb and dive parameters per aircraft, as well as turn rates, per aircraft - for the single-player enthusiast (but of course locked to tweaking for the MP crowd).

 

Cheers all & happy contemplating of FM tweak possibilities,

?

Edited by VonS
Edited post.
Posted
17 hours ago, VonS said:

@OP, many thanks for posting that info. from RoF's days of yore -- always interesting to see how FMs were tweaked and tuned in that particular sim. Makes sense then that the devs. are somewhat leery of devoting too many resources to FM tweaking in the (continued) FC series, considering, most likely, that aircraft FMs are still done on a per-aircraft basis (a very slow process; speaking from personal experience in First Eagles 2, even though there we use static tables, modular data initialization files, and whatnot -- none of this newfangled dynamic physics modeling ?).

 

At any rate, good luck to you gents' and here's hoping that you at least inspire more creative (open-ended) thinking about FM tweaking in the FC series. I suppose that a good compromise between an extensive FM overhaul across the board and no FM changes at all would be, if the devs. are interested of course, to add a few extra dials to the in-sim menu in FC, to have user-tweakable climb and dive parameters per aircraft, as well as turn rates, per aircraft - for the single-player enthusiast (but of course locked to tweaking for the MP crowd).

 

Cheers all & happy contemplating of FM tweak possibilities,

?

 

For sure, any update of the FM’s need to be done per aircraft in Flying Circus. But this is no different from the rest of Il-2.

 

And there have been many tweaks and adjustments on the various Bf 109, Spitfire, Thunderbolt and Mustang variants. And these changes have nearly always been driven by the forum community. Someone finds something off, reports it and usually it was fixed or there were long threads and complaints about it. Granted, some threads were not supported by facts, but in the cases they were, the water hollowed out the stone in the end. However, for some reason this does not seem to apply to the WW1 parts of the simulation. Neither the developers nor the community are rocking the boat.

 

As far as I know, in Il-2 is that the developers have a stated goal to be within 5% of the data. And for WW2 aircraft many are. But for some reason, the WW1 aircraft being introduced into FC from RoF should be accepted as they are. However, some of the the aircraft ported over from RoF deviate with 30% or more.

 

And now installments FC3 and FC4 are in the works. And the aircraft ported from RoF will probably look smashing in new graphics! However, while I do not wish to rain on the parade overly much, my interest in sims is not 4K graphics but rather to get a feel of how the different designs compared to each other. For example, can you effectively combat a tighter turning (angles fighter) Spitfire Mk5 in a Bf 109F (energy fighter)? Then try out different approaches to this and try to figure out how does what you experience in the sim tab with what happened IRL? Now imagine we had a Bf 109F which turned better than the Spitfire in-game? So now if I want to try out what Marseille did on North Africa to shoot down Hurricanes and Spitfires, I can skip all that and instead just outturn them. How much fun is that? Well it’s not for me anyway. And this analogy is not as crazy as it sounds because when matching up certain planes in FC in its current state, this is the situation we are looking at.

 

End of rant! ;)

 

  • Thanks 1
  • Upvote 4
Posted (edited)

Petrovich did a great job at times considering the dodgy, frustrating published data sets, the amount of research necessary for fuller understanding of each plane, and the time he had available for each FM.

But with the SE5a I can't get over the use of an experimental 5.6:1 CR Hisso of which only 19 were made, in order to try hit some very optimistic performance targets. Surely the need for an ultra high comp engine raises a red flag that something is wrong? Either the data, the model or the approach - or all 3!

Anyway, that was then.

Now we know better. User expectations have changed too.

We're fortunate to now have the models and processes to provide credible performance data. We just need a developer to realise the same :)

 

 

 

 

 

Hisso Type 34 S perf extract with red lines for 5.6 to1 .png

Hisso book - Wolsely Hisso perf table.png

Edited by US103_Baer
  • Thanks 2
  • Upvote 1
Posted
On 10/25/2023 at 7:04 PM, Holtzauge said:

there have been many tweaks and adjustments on the various Bf 109, Spitfire, Thunderbolt and Mustang variants. And these changes have nearly always been driven by the forum community. Someone finds something off, reports it and usually it was fixed or there were long threads and complaints about it. Granted, some threads were not supported by facts, but in the cases they were, the water hollowed out the stone in the end. However, for some reason this does not seem to apply to the WW1 parts of the simulation.

 

The weird turn/zoom energy loss issues of the RoF Se5a FM update (ported into FC) has been discussed ever since its inception in 2011. Over the ensuing years, there has been great effort, with the best data available (including your book), to persuade the devs to revisit and revise the FM. All to no avail. I don't understand it.

 

But then, there are things we don't understand, there are known knowns. There are things we know we know. We also know there are known unknowns. That is to say, we know there are some things we do not know. But there are also unknown unknowns, the ones we don't know we don't know.

 

That basically sums it up.

 

 

1PL-Husar-1Esk
Posted

I wander if the S5e would benefit from different propeller model udes for FM tuning, one more optimized for climb not for speed. Would that make significant difference in regainig energy in boom nad zoom flight? 

In general I think virtual Se5 build energii to slow and deplite it to quickly.

Posted
17 hours ago, 1PL-Husar-1Esk said:

In general I think virtual Se5 build energii to slow and deplite it to quickly.

 

Correct. Welcome aboard Husar. 

  • Haha 1
Posted (edited)
18 hours ago, ST_Catchov said:

 

The weird turn/zoom energy loss issues of the RoF Se5a FM update (ported into FC) has been discussed ever since its inception in 2011. Over the ensuing years, there has been great effort, with the best data available (including your book), to persuade the devs to revisit and revise the FM. All to no avail. I don't understand it.

 

But then, there are things we don't understand, there are known knowns. There are things we know we know. We also know there are known unknowns. That is to say, we know there are some things we do not know. But there are also unknown unknowns, the ones we don't know we don't know.

 

That basically sums it up.

 

 

 

18 hours ago, 1PL-Husar-1Esk said:

I wander if the S5e would benefit from different propeller model udes for FM tuning, one more optimized for climb not for speed. Would that make significant difference in regainig energy in boom nad zoom flight? 

In general I think virtual Se5 build energii to slow and deplite it to quickly.

 

For sure, the engine and propeller modeling could be a part of it. However, since the in-game S.E.5a’s climb and speed performance are within reason, I think the main culprit (when it comes to the poor turn and energy retention) is the modeling of induced drag at higher lift coefficients.

 

What points in this direction is that the speed for best climb rate and the speed for best sustained turn performance are about the same, and since this in turn means that the engines power output is about the same under these conditions, it has to be the induced drag at high Cl being that is the culprit. Otherwise the climb rate would be off.

 

What this means for the FM modeling, is that the lift/drag polar for the in-game S.E.5a could probably be unchanged for low (high speed) and intermediate Cl (climb), but then for high Cl and Clmax like in turns, it should be reduced.

 

Such a tweak would leave the high speed and climb performance unchanged, but improve the turn and energy retention performance, thus bringing the in-game S.E.5a’s FM closer to its IRL performance.

 

Edited by Holtzauge
  • Like 2
  • Upvote 3
Posted

An astute observation as always @Holtzauge. I can't pretend to understand your theory completely as that would be dishonest and unbecoming of me. But as one gentleman to another, I think you're onto something. Can we bring the devs onboard though? A tweak for them to consider rather than a complex FM overhaul, and one that may deliver the desired result we all pine for.

Posted

My dear chap: Always a please to converse with another gentleman. Now all we have to do is convince the developers to scrap this awful thingy with jets and whathavenyou and instead concentrate their efforts on the kind of aircraft that gentlemen fly. In addition, jets will attract those frightful Americans, and we can’t have that now, can we?

  • Like 1
Posted

It's too beastly to even contemplate. Jets, if there is such a ghastly thing, are an abomination, favoured only by the lower classes. Devs take note, and act accordingly.

  • Haha 1
Posted
35 minutes ago, ST_Catchov said:

, favoured only by the lower classes.

 

Ere, you toffee nosed chinless wonder, I resemble that remark!

Posted
13 minutes ago, Trooper117 said:

 

Ere, you toffee nosed chinless wonder, I resemble that remark!

 

Good God man! It's lucky I don't put you up on a charge for insolence! Get back to the shed and fix the Se's engine. She's losing revs. I want her ready for the dawn patrol. And there's a war on, in case you've forgotten!

  • Haha 1

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...