[NO BUG] FM bug due to CG being initially too far forward then shifting aft...

 

Ultimately I think this thread has probably reached the end of its utility by now.

No no no. I wanna know it all ends. :)

No no no. I wanna know it all ends.

Spoiler:

The cockroaches inherit the earth, but then they discover capitalism and wipe out all life on the planet.

Look at other flight simulators which do actually cover a lot more in aerodynamics (all aero coefficients versus Mach and Beta) than DCS and it's a fact!

All the BS aside I'd like to know the answer to the accusation above just to know if we're really missing something in the DCS FMs compared to "other flight sims" (consumer products I assume, not professional).

I really don't want to brag about it, but I have manged to convince the Developers to correct their flight models TWICE already with rather short and clear posts consisting more or less accurate data and documents. They do care about this, even though not everything is perfect.

Yes? Did you convince them to correct aerodynamic issues? Certainly not! I could also point out the problems using far less sentences, but I'm afraid that I might leave a gap of misunderstanding therefore the conversations would go on and on, so by trying to detail the problem more and give a more accurate understanding of it I may inadvertently complicate my phrases and sentences. I'm not a good narrator, but I want to prove my points with what I know and interpret!

Back to the subject, as you've seen in those tracks, there's not a smooth transition of CG, thus maximum constant achievable AoA, when the fuel constantly drops from a quantity to another and it's logically at doubt that the real system has the capability to "instantly" transfer fuel from one side of the plane to another to make the CG (thus AoA) shift as quick as seen in the tracks, besides that this shift doesn't even have a logic, because now it's fore, now it's aft.

I don't get it why you people try to find alternate explanations to something that won't make any sense.

LOL

Yes, and that makes a lot of sense now! This is sad, but not for me, I've made my point and I'm greatly disappointed!

Make your own sim please.

Stop wasting your "talent" here.

Yes? Did you convince them to correct aerodynamic issues? Certainly not!

More than once. And I've also been wrong more than once, after going over my own testing and comparing to the aircraft's performance charts.

Back to the subject, as you've seen in those tracks, there's not a smooth transition of CG, thus maximum constant achievable AoA, when the fuel constantly drops from a quantity to another and it's logically at doubt that the real system has the capability to "instantly" transfer fuel from one side of the plane to another to make the CG (thus AoA) shift as quick as seen in the tracks, besides that this shift doesn't even have a logic, because now it's fore, now it's aft.

 

I don't get it why you people try to find alternate explanations to something that won't make any sense.

I don't get why instead of just saying what's in the above paragraph, you had to dive into conspiracy theories and accusations. You could could have asked for/waited until Yo-Yo responded, and then simply presented what you believe is wrong.

Instead you accused him of not knowing what he's doing, and now you're a joke :)

We have checked the probable issue. In the realistic mode fuel management and CoG travel works AS IT IS INTENDED. But, actually, there was a bug in unlimited fuel mode as the plane goes at negative g. Unlimited fuel management sometimes is harder to implement over the realistic fuel system...

Yo-yo, in my own testing, I’m seeing a consistent spike in the AoA upward centered on 7400 kg fuel remaining. It lasts for about 300 kg of fuel, then drops back down. Also as the last 1200 kg of fuel begins to be used, the AoA suddenly decreases. Is this as it should be?

Yes. There are initial balance changes at full load end and at low fuel end. And there is a long flat range between. Almost neutral.

Excellent! That's what I'm seeing. Thank you. I'll post what I have when it's ready simply because it's interesting to see it in the data.

So it does exist! Thanks for the confirmation YoYo.

So here's what I found, when I fly a high-AoA profile under direct control. While the CG remains fairly steady as fuel is consumed, there are a few shifts toward the rear (as indicated by the higher AoA values) before it moves forward again depending on which tanks fuel is being drawn from. The real-world manual states the fuel consumption sequence which Esac_mirmidon kindly posted in a different thread: Su-27 Fuel Tanks and Draw Sequence. EDIT: The sequence referenced in the real-world manual is for mil power, not afterburner. So all of my flights were made under mil power--full mil for the high-AoA runs, about 83-84% for the long trimmed "level" flight. ::EDIT ENDS

My initial flight was with 90% fuel because Maverick Su-35S's complaints began at a lesser fuel quantity and I had not been planning to get this involved. I could see that there was some shifting but his tracks made it difficult to follow.

At any rate, the track of my initial flight is attached as is the spreadsheet itself. There's a 2nd page to the spreadsheet for a flight that started with the full 9400 kg and ran down to 6000 kg as a double check to make sure nothing strange was happening at the full fuel end.

The following image is my first spreadsheet. The flight starts somewhat "nose heavy" and the AoA slowly increases as the CG shifts rearward during the first drawdown from Tank 1. There's a short-lived spike in AoA at around 7400 kg fuel remaining that appeared consistently. Then, during the drawdown of Tank 4, the CG shifts forward again. As the remainder of Tank 1 is emptied, the CG again starts moving toward the rear. And so on.

Here's that flight graphed in Tacview. I need to mention that this flight was flown with full ammo, chaff and flares. I also accepted the sim's nose down trim for the airstart. My initial airspeed set in the ME was 350 kph.

Here's another flight, this time with full fuel, no internal stores, and the aircraft trimmed to neutral nose down trim. A similar pattern of CG shifts follows, though the flight itself was noticeably rougher throughout and the AoA values are a bit higher. The track for this flight is attached as well. (This track is just as boring to view as the first. Download at your own risk.)

This last flight is a "level" trimmed flight made under direct control. What I found fascinating is that the same CG shift pattern is readily observable. The flight lasted 3+ hours.

The nice thing about trimmed flight is that you don't have to sit in front of the computer for it. I got bored after 45 minutes and left the sim to run it's course while I went outside and did more productive things. I came back, happily, just as the fuel was being depleted and deadsticked a landing somewhere in the interior of Ukraine. That last part was the only fun part.

CG vs Fuel Quantity.zip

Ironhand CG vs Fuel Quantity-90pc.trk

Ironhand Direct Control 9400kg--Neutral Trim_No Internals.trk

Amazing research Ironhand.

The BEST results always came from boring, long observations not everybody are willing to do. Thats how science work.

Thanks a lot for this

Amazing research Ironhand.

 

The BEST results always came from boring, long observations not everybody are willing to do. Thats how science work.

 

Thanks a lot for this

You're quite welcome. I knew Maverick was seeing something but with the chaos his methodology introduced, it was hard to see the signal for the noise. I was hoping to quiet the noise a bit so the "signal" could come through.

With the data and tracks, anyone so inclined can take the time to verify the results. For the set with the spreadsheet, you have the sample times at which to stop the sim and doublecheck. I imagine some data might be a bit off from time to time if for no other reason than sampling error (a polite way of saying I screwed up the observation). But all of my spot checks were right on. That plus the same basic pattern revealed itself across multiple flights.

Outstanding work Ironhand, wish I still had some rep to give :notworthy:

And let this be a lesson in how to user-test the DCS flight models: recording verifiable, repeatable, observable phenomena is the essence of the scientific method. Everything else is mere opinion and / or conjecture.

So, theory predicts the changes in CG with fuel depletion. Computational analysis verifies this. The £1K question then is: can this be verified by experiment? Does this happen in the real Su-27? It'd be great to have an answer to that.

Outstanding work Ironhand, wish I still had some rep to give :notworthy:

 

And let this be a lesson in how to user-test the DCS flight models: recording verifiable, repeatable, observable phenomena is the essence of the scientific method. Everything else is mere opinion and / or conjecture.

 

So, theory predicts the changes in CG with fuel depletion. Computational analysis verifies this. The £1K question then is: can this be verified by experiment? Does this happen in the real Su-27? It'd be great to have an answer to that.

Thanks. Though it's not something I'd expect everyone to do. But the more verifiable and easy to interpret you make things, the easier it is for everyone involved.

At any rate, I hope Yo-Yo comes through again. I'm curious about the rapid spike followed by a dip followed by an increase that seems to occur consistently toward the full end of the fuel spectrum. It doesn't make sense to me given the drawdown sequence in the manual. But, then again, the "tank empty" illuminations don't match what I'd expect from what I've seen in the manual either. I haven't completely translated it, though. So I could very well be missing something.

Thanks. Though it's not something I'd expect everyone to do. But the more verifiable and easy to interpret you make things, the easier it is for everyone involved.

 

At any rate, I hope Yo-Yo comes through again. I'm curious about the rapid spike followed by a dip followed by an increase that seems to occur consistently toward the full end of the fuel spectrum. It doesn't make sense to me given the drawdown sequence in the manual. But, then again, the "tank empty" illuminations don't match what I'd expect from what I've seen in the manual either. I haven't completely translated it, though. So I could very well be missing something.

My guess would be that the depletion sequence was designed to be as flat as possible whilst avoiding the need for a complex & heavy automatic fuel re-distribution system. It may also be significant that the main spikes at the start occur at fuel loads that are above the "nominal" load of 5,160 that Esac_mirmidon's translation suggested (tank 2 & tank(s) 3).

Might also be relevant that the drop off at the end of the fuel curve IIRC happens at a fuel load that's safely below the fuel required for a nominal landing weight, so in practise the drop off at the end should almost never be experienced unless a pilot goes in to emergency fuel.

Finally, I think the available manual is for the 27SK variant rather than the 27S that we have in DCS World. Wonder if that makes any difference.

My guess would be that the depletion sequence was designed to be as flat as possible whilst avoiding the need for a complex & heavy automatic fuel re-distribution system. It may also be significant that the main spikes at the start occur at fuel loads that are above the "nominal" load of 5,160 that Esac_mirmidon's translation suggested (tank 2 & tank(s) 3).

 

Might also be relevant that the drop off at the end of the fuel curve IIRC happens at a fuel load that's safely below the fuel required for a nominal landing weight, so in practise the drop off at the end should almost never be experienced unless a pilot goes in to emergency fuel.

 

Finally, I think the available manual is for the 27SK variant rather than the 27S that we have in DCS World. Wonder if that makes any difference.

:) The only one I'm really wondering about is the one at around 7400 kg remaining. The dropoff toward the end makes sense if the CG with empty tanks is behind the feeder tank. But this stuff is most noticeable only when flying in direct control, something that isn't the norm. Otherwise the FBW system is there to smoothe it out, anyway. So, in that sense, this is a tremendous trifle.

Don't really know the differences between the two models. If the -SK has slightly larger tanks, that could account for the differences.

:) The only one I'm really wondering about is the one at around 7400 kg remaining. The dropoff toward the end makes sense if the CG with empty tanks is behind the feeder tank. But this stuff is most noticeable only when flying in direct control, something that isn't the norm. Otherwise the FBW system is there to smoothe it out, anyway. So, in that sense, this is a tremendous trifle.

 

Don't really know the differences between the two models. If the -SK has slightly larger tanks, that could account for the differences.

Tremendous trifle may be, but it's nice to know it's there & frankly after a couple of K hours in the DCS Su-27, exploring the very edges of the envelope are where fun & interest are still to be had. Test pilot style investigations are much more interesting than going A-->B on automatic control :thumbup:

Everything I've been able to find indicates that the -SK is simply the export version of the baseline Su-27S. Max fuel loads are quoted as being the same as is the radar and EOS. As far as I can tell the Su-27SK is simply the designation given to the original Russian manufactured tranche of aircraft that would become the J-11A.

In Soviet days export models were inferior to domestic versions (possibly with heavier, older equipment) but there's nothing to suggest that was the case with the -27SK so I'm not sure. I guess the final dip in the AOA graph may have been an anomaly that Sukhoi were simply prepared to accept given that it shouldn't ever be experienced by line squadron pilots if they keep to SOP.

Tremendous trifle may be, but it's nice to know it's there & frankly after a couple of K hours in the DCS Su-27, exploring the very edges of the envelope are where fun & interest are still to be had. Test pilot style investigations are much more interesting than going A-->B on automatic control :thumbup:

:) True enough...

Everything I've been able to find indicates that the -SK is simply the export version of the baseline Su-27S. Max fuel loads are quoted as being the same as is the radar and EOS. As far as I can tell the Su-27SK is simply the designation given to the original Russian manufactured branche of aircraft that would become the J-11A.

OK. Thanks.

...I guess the final dip in the AOA graph may have been an anomaly that Sukhoi were simply prepared to accept given that it shouldn't ever be experienced by line squadron pilots if they keep to SOP.

Unless you're in direct control at that point rather then using the FBW system, it's not anything you'd even notice.

I don't have the patience to conduct these kind of indepth tests, but I do enjoy reading them :) Good work and thanks for the effort

I don't have the patience to conduct these kind of indepth tests, but I do enjoy reading them :) Good work and thanks for the effort

Not sure it was patience. More a combination of curiosity and stubbornness. And, besides, it’s fun to go through the spreadsheet.

...

 

The following image is my first spreadsheet. The flight starts somewhat "nose heavy" and the AoA slowly increases as the CG shifts rearward during the first drawdown from Tank 1. There's a short-lived spike in AoA at around 7400 kg fuel remaining that appeared consistently. Then, during the drawdown of Tank 4, the CG shifts forward again. As the remainder of Tank 1 is emptied, the CG again starts moving toward the rear. And so on.

...

Hello "Ironhand",

At least they believe you for the exact things I also pointed out in repetitive ways! Who knows, maybe they have a difficulty understanding what I write (although if I read my own posts everything should be well understood).

I've only said (from what I was able to observe without any special software and just by eyes) that as the fuel depletes (besides the case of unlimited fuel), at full aft stick direct pitch control and constant elevator position, the AoA jumps some 10 degrees higher at an instant when the fuel status becomes X, then at some lower fuel status Y the AoA reverts, then again at some other fuel status it jumps back to 10 degrees higher. I don't get it why it was so hard to understand this, but maybe the word "AoA spike" finally convinced them.

Thanks "Ironhand"! Seriously..., I truly appreciate that you made an abstraction from how the conversations evolved (because I have my limits of patience and I sometimes feel that I'm talking to who I should) and separately done a verification and finally "smashed" all the ignorance and false claims that everything is "OK" when they're not OK in plain sight! I kind of like the hypocrisy that goes around by trying to deny what's true and I became so used to it that I no longer feel like needing to put it to the heart.

In general, I'm kind of done with DCS! I've made my impression about DCS's level, talking about flight models only (what I only care about) and I really wouldn't want to share my words for it, especially for how it's creators and/or defenders treated me here here! I've made my points clear and fair almost every time and never sought to get into conflict with anyone, insult or misbehave, but the responses I get, especially by re-naming the thread's title as a "NO BUG" when I know for sure and can clearly see what's wrong, is too much for me giving what I know. I moved to simulators which I personally find on the right track and on which the flight model problems can be more quickly spotted and resolved.

Regards!