[NO BUG] flight model

[Quid]

i decided to go with a number the charts do list, that is 50% fule, 4x4 at sea level. Changed the mission accordingly and flew again.

 

Just for your SA, none of the F-14B's charts are for sea level. The lowest chart is 5000 feet, so you want to test there, not SL.

 

I ran a few tests myself with a 4-4, 54% fuel (based on the weights listed in the 1.1, it's right around there), and while I'm not great at just holding a turn, I was stable enough to get some data out of it. At around 330-340 KTAS (.51M), I am sustaining 5-5.1g, which is right about where it's supposed to be on the charts. At .56M, max STR, and 4857 feet (because i started a little low) I'm at 5.5g, which is pretty close to where it should be; TACVIEW lists a fluctuating 16.6-16.7 STR, it should be around 16.1, or about .5-.6deg/sec off, possibly slightly less due to my suck-tastic entering conditions. Seems pretty close to me at 5,000 feet; then again, running it with a script that makes no mistakes could yield different results.

[IronMike]

1.5 seconds limit is ok difference if we speaking about 30 seconds turn. Like 30.7 instead 29.3. Not awesome, not fail. Good enough for digital combat simulator. But here it is 9% of the turn rate.

And my personal opinion 9% gap is too much to be fully satisfied as user or development. For example I hope you woudn't ignore if awg-9 doppler filter has 9% lower filter settings. Here is the same. This is not major problem, but it still require minor polish.

As DCS F-14 pilot I hope you take care about it.

And even if another module have bigger gap, that doesn't mean you shouldn't try make yours better.

 

 

 

 

It is not 9% off. Delan's test was wrong, for reasons explained in the post above. Or let me phrase it differently, it is pretty much where it should be if you reduce your fuel and/or go to sea level. Quid answers it very nicely in the post above.

 

 

Please guys: if you wanna test the FM, test it as supposed to in the EM charts. Same altitude, same fuel, same loadout. If you decrease on of these three, of course you will get even better performance, that is to be expected.

 

And I will add also this:in order for this type of testing to be valid, you need to:

 

• hold constant speed +- 2 knots
  
• roughly hold +-100 ft/min on rate climb (hence the Ps=0 curve)
  
• have the correct loadout/weight
  
• have unlimited fuel
  
• be at the correct altitude MSL
  
• have standard day conditions (15°C at sea level)
  
• have the correct flap setting
  
• stabilize this flight envelope for 10 seconds before you take any G or turn rate measurement
  
• have it stabilized at least 4 (or more) times like that (for 10 sec minimum) and take an average
  
  

Captain Delan, as much as the test is appreciated, did not even do one of those things - this is just to illustrate why data taken like this does not say much beyond that it is well within what it is supposed to be. But if you want to know for sure, you need to fly it accordingly. (And personally I know only 1 person who can do that consistantly, which is why such performance charts are theoretical in the first place and not tested by humans in general, but by programs.)

 

We do appreciate your input guys, but I want to re-assure you once more that there is nothing wrong or off with the FM beyond anything that needs any kind of fixing atm performance wise. We do always appreciate your input, but please understand that flight model tests cannot be done "just in any kind of way". So far Totmacher's tests were the ones done closest as supposed to (because a program did it), and they show it to be off once by 1.2 seconds and once by 1.3 seconds. Which in a sim, is nothing to worry about. If you do, fine. But we don't. Neither do our SMEs. Neither would either the FAA or EASA. You're nitpicking on something, which hardly will find a more precise equivalent in consumer simming. If you do find one, please show us. That is: an FM in any consumer sim, that tested as described above according to its EM charts is not off by more than 1.1 seconds and still feels natural and not scripted or forced. And all that said, we will still look into it, if we cannot get it even closer. In our opinion there is always room left for improvement, and we'll likely make it even closer in the future. But let us focus first on where it is needed more. Thank you.

 

Edited June 30, 2020 by IronMike

[captain_dalan]

Just for your SA, none of the F-14B's charts are for sea level. The lowest chart is 5000 feet, so you want to test there, not SL.

 

I ran a few tests myself with a 4-4, 54% fuel (based on the weights listed in the 1.1, it's right around there)

 

It is not 9% off. Delan's test was wrong, for reasons explained in the post above.

 

Gentlemen, i would point you to chapter IX, Combat Performance. More specifically pages XI-9-1 to XI-9-6. Therein you'll find the specific excess power diagrams for the tested configuration (55620 lbs gross weight, 4X4), for 4 acceleration (g) values, 1g (level flight), 3g, 5g and 6.5g. The charts run from sea level to aircraft ceiling (and slightly above). Just pick your desired excess power curve for the desired g-load, find where it intersects with the desired altitude on the y-axis and you have the required mach number on the x-axis. In this case, the sea level values, are naturally at the bottom of the y-axis.

 

Connect the dots for each value and you get the entire sea level envelope. Or don't and just test the 5g chart. It should be enough. You'll see that the 1.3s increase in turn performance over desired value, becomes a 1.5-1.6s increase, a rise from 6% to 8+% (8.2-8.6 depending on how fine you can read the chart) and what should be a 5g sustained turn at 18.5 deg/s, becomes 20-20.1 deg/s at 5.5g.

 

Is this detrimental to the overall ACM performance for the bandits? No. Right now there is no plane in DCS that contests this part of the envelope from the F-14. Is it really that big of an issue? Debatable..... i would argue not something to rob one self from sleeping comely each night. Is it there and is it statistically significant? Yes it is. Half a g difference at 450 knots and 8g, would hardly be felt in the terms of turn performance. At 300 knots and 5g though, it will. And it is felt even by a clumsy stick like myself. Try to maintain the turn at the said load, and you will accelerate. A 0.1-0.2, maybe even 0.3 g is difficult to sense. Or can be, depends on how good you are. But a 0.5g? Especially this slow. Yeah, you can tell.

 

Now if testing is considered not so relevant, i bet some of the new (and much smarter) kids around here can script and create an automated mission where the F-14 will do its best sustained turn at 300 knots. Let's see what gives.

 

If however this is not considered significant and is withing "tolerable" deviations from the expected values for ANY reason, then so be it. I resign. :thumbup:

[totmacher]

It is amazing to see that those who try here to change the FM of the Tomcat, do not even have it. However, the FM of an F18 which holds in the air like a helicopter it does not matter or a Su27 which gear down in full dogfight.

 

The others developpers changed anything ? so why would Heatblur do it without evidence and for people who don't have it ?

 

Can anyone give me link to documents with turn data of F/A-18c please? Thanks!

[Freakerr]

In our opinion there is always room left for improvement, and we'll likely make it even closer in the future. But let us focus first on where it is needed more. Thank you.

 

Ok. This is the answer we need. Let take closer look on FM after "A" release.

Based on Totmacher's test it seems like "B" have too mach power or lift or less drag until 1.15 mach.

Hes measurements show right shape but always too high.

And anyway it's nice to know dev's keep communicate with customers.

See you:pilotfly:

[Kaiju]

Can anyone give me link to documents with turn data of F/A-18c please? Thanks!

 

 

 

 

Ask here :

 

 

https://forums.eagle.ru/forumdisplay.php?f=594

[IronMike]

Gentlemen, i would point you to chapter IX, Combat Performance. More specifically pages XI-9-1 to XI-9-6. Therein you'll find the specific excess power diagrams for the tested configuration (55620 lbs gross weight, 4X4), for 4 acceleration (g) values, 1g (level flight), 3g, 5g and 6.5g. The charts run from sea level to aircraft ceiling (and slightly above). Just pick your desired excess power curve for the desired g-load, find where it intersects with the desired altitude on the y-axis and you have the required mach number on the x-axis. In this case, the sea level values, are naturally at the bottom of the y-axis.

 

Connect the dots for each value and you get the entire sea level envelope. Or don't and just test the 5g chart. It should be enough. You'll see that the 1.3s increase in turn performance over desired value, becomes a 1.5-1.6s increase, a rise from 6% to 8+% (8.2-8.6 depending on how fine you can read the chart) and what should be a 5g sustained turn at 18.5 deg/s, becomes 20-20.1 deg/s at 5.5g.

 

Is this detrimental to the overall ACM performance for the bandits? No. Right now there is no plane in DCS that contests this part of the envelope from the F-14. Is it really that big of an issue? Debatable..... i would argue not something to rob one self from sleeping comely each night. Is it there and is it statistically significant? Yes it is. Half a g difference at 450 knots and 8g, would hardly be felt in the terms of turn performance. At 300 knots and 5g though, it will. And it is felt even by a clumsy stick like myself. Try to maintain the turn at the said load, and you will accelerate. A 0.1-0.2, maybe even 0.3 g is difficult to sense. Or can be, depends on how good you are. But a 0.5g? Especially this slow. Yeah, you can tell.

 

Now if testing is considered not so relevant, i bet some of the new (and much smarter) kids around here can script and create an automated mission where the F-14 will do its best sustained turn at 300 knots. Let's see what gives.

 

If however this is not considered significant and is withing "tolerable" deviations from the expected values for ANY reason, then so be it. I resign. :thumbup:

 

 

 

 

Ok, I see which chart you went for, however you are not flying the test stabilized enough. You got very large deviations in climb reate, deviations in G and deviations in speed. They all need to be very stable, practically no change on any parameter over a 10 second period before you take the measurement.

 

You would have to fly it stabilized right there, at the red dot in the chart attached below. You can then either close on speed or on G with the stick while keeping the rate of climb very low. If you choose to close on G, check to see if the speed is correct. If you choose to close on speed, you'd check for the constant G load.

 

Flown like in your video, your numbers are anything but reliable. It btw is only really possible to hand fly points on the Ps=0 curve, all other tests require an automated testing system with a perfect "automatic pilot"... Your tests are appreciated still, because they give us insight in how you guys fly, how she performs in your hands and so forth, which is all valuable to us, too. As is your input. But data wise, please try to stick to the Ps=0 curve and even then, you need to take your own results with a huge grain of salt for above mentioned reasons.

Edited June 30, 2020 by IronMike

[Quid]

Well, this brings up an idea: could Totmacher run a test at 5,000 feet ASL, 4x AIM-9, 4x AIM-7, full gun ammo, no tanks, 50% fuel and see how well it matches given an AI script which can fly the circle "perfectly"? I didn't see a loadout on the last page, only a weight. If it was done with a clean aircraft, there is no impact from stores, a lower DI, and it won't follow the chart as closely.

[HDpilot]

As for turn rate: show us a track or video and we will take a look. Else please stop with your unsubstantiated accusations. The FM is not wrong, even if you believe it is. There is a margin of error, and as I said, it is both acceptable and negligable.

 

 

 

If you spike once into around 15G that is possible. But I doubt you can reproduce pulling 14-15G or pulling 18G 10/10 on the initial pull. If you can, then this is a bug. Instead of throwing tantrums, please show us the proof and we will fix it. that simple.

 

But please take your aggressive tone and attitude elsewhere, or this discussion will be over. Thank you.

 

Also, if you do not own the tomcat, you are obviously complaining because you think that we made it "OP" in order to give it an advantage. This is yet another accusation, we most certainly will not stand for. Nothing could be further from the truth. So if your intention is to help us improve something we could, then please be our guest. If you just come here to unload your frustrations, then please let it be. It will not get you anywhere.

 

Sorry, i am only responsible for what I say, not for your interpretation. I´m not aggressive, have no tantrums, insulted anyone and do not violate forum rules. I would be frustrated if i owned the module, but i have no reason to be it. It is my right not to like your FM interpretation and to share it here publicly. But enough "polemics"

 

You said 1,5sec are acceptable, then i try to clarify what 1.5 seconds are comparable to:

~1,2sec represent a 1850kg gross weight difference, 59695pounds vs 55620puonds (5000feet, devices on auto), means + 0,4G load factor at M0,55 on Ps=0 line, means ~7,6% more thrust, means ~ 1850 more pounds force per engine if you estimate the required thrust according to the known formula

Parameter checking should be part of your development process, so i don't understand why you ask for evidence videos or tracks.

In the short free trial period I didn't have much time to master the handling. Keeping flight parameters precise is very difficult, but I'm sure you can reproduce in your test environment a sustained 5.6-5,7g turn at 334kts IAS at 5000 feet with 55620lbs and devices on auto. That would be ~ 20 seconds per turn. Decide for yourself whether this is acceptable.

Edited June 30, 2020 by HDpilot

[TLTeo]

Parameter checking should be part of your development process, so i don't understand why you ask for evidence videos or tracks.

 

 

Sweet baby Jesus. BECAUSE WITHOUT A TRACK THEY CAN'T CHECK EXACTLY WHAT'S GOING IN WITH WHAT YOU ARE DOING AND/OR WITHIN THE CODE.

 

 

How complicated can it be? Just show us your proof or go away. And then you go on to say

 

 

In the short free trial period I didn't have much time to master the handling. Keeping flight parameters precise is very difficult, but I'm sure you can reproduce in your test environment a sustained 5.6-5,7g turn at 334kts IAS at 5000 feet with 55620lbs and devices on auto. That would be ~ 20 seconds per turn. Decide for yourself whether this is acceptable.

 

 

So your claim is based on you not knowing how to test the FM properly, yet people are supposed to take your word for the absolute truth. THIS is why devs need proof. "Demonstrating this thing is easy but I didn't fly it well enough to demonstrate it but you should be able to demonstrate what I'm saying because it's easy" is not a valid argument.

[IronMike]

But enough "polemics"

 

 

 

 

Quite right, thank you. Show us proof, or stop "claiming" things please. Totmacher showed you the exact numbers. And that is that, and yes, again, it is acceptable and most likely closer than any other module you are flying.

 

 

 

And yes, you are starting to break forum rules, from this post on, keep in mind that you have been hereby warned:

 

 

1.10 Product feedback and constructive criticism is encouraged when provided in a mature and courteous manner. However, feedback that is abusive, insulting or condescending is not welcome. Additionally, to bring up a particular issue repeatedly after it has already been acknowledged will be considered "trolling" - in such cases a warning will be issued to the author and the post will be removed.

 

 

Your tone was aggressive and condescending, you did suggest that we falsified the flight model to be better to paraphrase you "such a coincidence that it turns faster", calling my replies "nonsense" etc etc.. So do not act as if you did not and are the vitcim here, because you are not.

 

Furthermore I've been warned by several members of the Russian community that you are known as a troll on the russian forums, where you throw around the same unsubstantiated accusations against us, in the same kind of manner and the same kind of tone. Again, don't act as if you do not.

 

On top of that you don't seem to come here to help, but with a hostile attitude towards us, out of the suspicion we would have intentionally falsified the module to give it an advantage. I will tell you again, that we will not stand for such preposterous accusations.

 

And lastly, you are time and time bringing up an issue that has been acknowledged, and you give such unsubstantiated input that it does not let us act upon, even if you were genuinely interested to help, which you are not.

 

 

 

You do have the right to dislike our flight model, you also have the right to go around and call it the worst thing in the world, if you like. And we have the right to absolutely not care about your unfounded, unproven and quite honestly, unqualified opinion, in all regards concerning the FM: understanding the charts properly, knowing it's history and employment (and flight) data and the ability to even fly remotely close to what you should, in order to get valuable data.

 

So here is my warning to you: Bring proof, come with a genuine attitude that you wanna help, or with a legitimate complaint, but any more unsubstantiated posts of yours like this, and you will get a strike under rule 1.10 and your post will get deleted. You are costing us valuable development time with your trolling.

[fat creason]

a sustained 5.6-5,7g turn at 334kts IAS at 5000 feet with 55620lbs and devices on auto. That would be ~ 20 seconds per turn. Decide for yourself whether this is acceptable.

 

If you make this claim, the burden of proof is on YOU to prove it, not on us to disprove it. Is it not our job to investigate every wild claim that someone makes. We have better things to do with our time.

 

Is the F-14 FM 100% perfect, with zero error across the entire flight envelope? No, and no flight model on earth can claim this. Is it within an acceptable margin of error in the critical areas? Yes. Even Level D flight trainers have acceptable margins or error (I would know, I used to work on them). I would invite you legitimately test every other module in the game across their entire flight envelope and report your results with proof. I'd bet the F-14 FM is closer than most. Totmacher's testing methodology is by far the most valid way of testing EM chart in-game. We appreciate his efforts and candor regarding his test results, we plan to make the FM even better in the future.

 

Another aspect beyond pure FM performance is subjective handling qualities which I have been spending far more time on, working closely with our SME pilots, (in addition to the A/TF30) over FM performance aspects. This is far more important to the feel of the FM than a ~7% 1.1 deg/sec turn rate error. I will probably spend more time on FM performance in the future, but it is in a state that does not require immediate attention.

 

Totmacher's test demonstrates that the actual Ps=0 peak is somewhere in the area of the blue dot. These two claims are wildly different, and only one has validity.

 

On top of all of this, I would like to point out that these EM charts are purely based on estimations of available energy. These charts are estimated from flight test. It even says it right there, at the top left. There is no way to verify these charts in any real-world scenario, since it would require constant weight (aka no fuel burn) which would be impossible at full after burning power. Once you stray from the Ps=0 line, this testing becomes even more implausible since it would require an instantaneously measured stabilized turn right at 5000 ft while having a significant climb rate at a constant speed. I have seen takeoff climb charts from FAA approved performance manuals that under-report the actual climb rate from real flight test data by 50-100%! The flight test data showed that the aircraft was capable of climbing twice as fast as the manual stated. In this context, I consider a 7% error to be a very minor concern.

 

I hate to break it to you, but if you are losing dogfights to Tomcats it is not because of this issue. :music_whistling:

Edited July 2, 2020 by fat creason
Edited for misreading HDPilot's post. (20 sec turn vs 20 deg/sec turn). Regardless, HDPilot's claim is still false as a 20 sec turn would imply an 18 deg/sec turn which he has no proof of.

[Freakerr]

Mike, please, calm down. No one here call f-14 FM bad. Maybe HDpilot a little aggressive, but it all because in our (Russian) community we too sensitive when western aircraft flight better when a log. Do not blame us!

And I hope after F-14A release we will find where are this gap is hiding. In the engines, on the wings or whatever.

 

Well, this brings up an idea: could Totmacher run a test at 5,000 feet ASL, 4x AIM-9, 4x AIM-7, full gun ammo, no tanks, 50% fuel and see how well it matches given an AI script which can fly the circle "perfectly"? I didn't see a loadout on the last page, only a weight. If it was done with a clean aircraft, there is no impact from stores, a lower DI, and it won't follow the chart as closely.

 

Titmacher's test made with same conditions. It's show ~0.48 mach for 5g turn, instead ~0.52, if I understand charts right.

[Quid]

Titmacher's test made with same conditions. It's show ~0.48 mach for 5g turn, instead ~0.52, if I understand charts right.

 

 

I see a weight, but not a loadout. It doesn't tell me if the jet is clean and fuel was added or if there were stores added to make the weight. They'll have a different impact since the aircraft's drag index will be different as a result, and now I'm genuinely interested since it was pretty damn close when I flew it (~.5-.6 degrees off, not 1.2-1.3), but I am not a script, and had to derive my numbers from TACVIEW recordings in sections of the circle where the altitude and airspeed didn't vary.

[Freakerr]

4 aim-9, 4 aim-7

https://forums.eagle.ru/attachment.php?attachmentid=236060&d=1589796419

[IronMike]

Mike, please, calm down. No one here call f-14 FM bad. Maybe HDpilot a little aggressive, but it all because in our (Russian) community we too sensitive when western aircraft flight better when a log. Do not blame us!

 

:)

 

I am happy to discuss any discrepancies that come with substantiated proof or even hand flown, if at least attempted correctly and/or deviating slightly, all that is fine, and we are happy to receive any kind of criticism. I also never take things personally, so no worries.

 

But, accusations like us intentionally tweaking the FM to be "OP" is not something I take lightly. We put a lot of effort in getting everything as right as we can and we listen to everyone, also HD pilot. But if his conclusions are wrong more than once and without proof, toned aggressively and as if we didnt knew what we are doing, but he does, what do you expect? We're not here to stand for that kind of stuff, and at this point we are losing time with this thread. I'm not angry about it at all, but I call it as I see it. Because this is where it becomes trolling. It is ok to let anger go, I don't mind. But one or two ranting posts should be enough. I totally understand the passion behind it. But we have a certain level, too, where we say "ok, that's enough now". It does not lead anywhere else than us re-explaining things which have been explained and losing time, while meanwhile other issues get pushed back. Such accusations steal our attention away from "more important" things, and I say more important, because as explained, perfomance does not require immediate attention with the margin as is. We're talking about negligable margins. It should be enough if we say this once, I've repeated it at least 5 times now.

 

I don't like to be "forced" by a "loud minority" to focus on something which has a small priority, because it is unfair to others who have (more) legitimate concerns and do not get the attention then which they require from us more immediately than this.

 

I know you just want it fair. And trust me, we want that, too. We don't care if you are direct customers or not, if you have it or not, we want everyone in DCS to have a great sim experience. But the Tomcat is what it is, and once we will have brought the margins down even further, it will not change in that aspect to any level where you would notice a difference without looking very, very closely.

 

Mind you we started with the MiG-21 back when we were still with LNS, and the same kind of approach went into both. We would approach any russian aircraft in the same way, with the same pride. Within the limits of DCS we always strive for maximum realism, if only possible. If we cannot agree on that, it is quite pointless to discuss, because then the accusations start to fly around left and right and it leads nowhere. Rest assured: we want the same things. And I've always had a special place in my heart for REDFOR and so does everyone on the team. But you guys also need to trust us, when we say the margin is acceptable. And that we will improve it even further. This is where this thread could have stopped, because we do hear you and we decided almost immediately that we want to get the margins even better than they are. There is atm just no point in going on more about it, unless of course you find something new, and can show us, we will always be happy to look into it. Or if you have a question, etc, always happy to answer. But at this point going on about how off it is or not, without proof next to already provided proof (by Totmacher), we will start to turn in circles. I hope that makes sense and thank you for your kind understanding.

 

 

 

 

:smilewink:

[captain_dalan]

Ok, I see which chart you went for, however you are not flying the test stabilized enough. You got very large deviations in climb reate, deviations in G and deviations in speed. They all need to be very stable, practically no change on any parameter over a 10 second period before you take the measurement.

 

You would have to fly it stabilized right there, at the red dot in the chart attached below. You can then either close on speed or on G with the stick while keeping the rate of climb very low. If you choose to close on G, check to see if the speed is correct. If you choose to close on speed, you'd check for the constant G load.

 

Flown like in your video, your numbers are anything but reliable. It btw is only really possible to hand fly points on the Ps=0 curve, all other tests require an automated testing system with a perfect "automatic pilot"... Your tests are appreciated still, because they give us insight in how you guys fly, how she performs in your hands and so forth, which is all valuable to us, too. As is your input. But data wise, please try to stick to the Ps=0 curve and even then, you need to take your own results with a huge grain of salt for above mentioned reasons.

 

Copy that. The video was done in a hurry yes. As soon as i'm able i'll do a sea skimming one (hopefully tomorrow) . But a script would work even better i think. Eliminating the human factor from the equation. :thumbup:

 

 

Titmacher's test made with same conditions. It's show ~0.48 mach for 5g turn, instead ~0.52, if I understand charts right.

 

Hey guys, can you make the test for a best sustained turn at mach 0.45-0.46, 55620 lbs gross weight in a 4X4? Sea level? It's the best break point we can use to find how much deviation is there right now. :thumbup:

[IronMike]

Hey guys, can you make the test for a best sustained turn at mach 0.45-0.46, 55620 lbs gross weight in a 4X4? Sea level? It's the best break point we can use to find how much deviation is there right now. :thumbup:

 

 

 

 

I'm sorry, but no, it is not. Best is the Ps=0 curve (from the other chart, Delan)... And the deviation is already known, see Totmacher's test. It's 1.2 seconds in the first and 1.3 seconds fast in the second test. This was the scripted test.

 

Please guys, stop doctoring with stuff you do not understand fully. You are not the experts on this. I do not mean this in any kind of condescending way, but that's how it is, and when we say you need to test the Ps=0 curve for best results, then we mean that. I really don't know why we have to repeat ourselves here so often.

 

 

 

"Once you stray from the Ps=0 line, this testing becomes even more implausible since it would require an instantaneously measured stabilized turn right at 5000 ft while having a significant climb rate at a constant speed."

 

You have our promise that we will make the FM even more precise than it is and that atm nothing is off outside of a margin that is not acceptable. If you do not want to take our word for it (including the many repeated explanations), fine, that is up to you. But please understand that we will not be repeating ourselves any further. Thank you for your kind patience and understanding in the meantime. :)

 

This thread will be closed until further notice. [We will reopen the thread for further discussion once we have made further tweaks on FM performance.]

Edited July 1, 2020 by IronMike