POLL: BRING BACK SFM MISSILES UNTIL AFM GETS APPROPRIATE GUIDANCE

[danilop]

This poll is lacking "I don't care" option! :D

 

I gave up on fighters in DCS long, long time ago.

 

Flying helicopters and attack aircraft exclusively nowadays.

[GGTharos]

That is the one.

 

What paper? Care to give a link to that?

 

Is this the "thread" or "paper" you're talking about?

 

http://forums.eagle.ru/showthread.php?t=131806&highlight=Missiles+MOd

[zaelu]

Yes there is, and yes we do. If we don't know the real world performance of missiles, and I'm not talking about subjective things like pk, but facts such as motor burn time, thurst, drag/lift ratios, etc. then there is no way to compare the performance of the in-game missiles against them no matter how many precise tests you do.

 

In any case it has already been mentioned in this thread and others that IAGTSG and others have done significant testing, and comparisons going so far as to actually do costly CFDs on their models, to get some kind of baseline that we are talking about here, and still their results were dismissed by Eagle Dynamics.

 

All our arguing is academic at this point, and threads like these are pointless. Ive accepted that missiles is a topic that is not up for debate in the eyes of the developers, and after that I'm more happy in flying within these limitations.

 

 

Well, usually fallacies have a name can you give me the name of the fallacy I am guilty of? I would like to correct myself and I can't see where I committed a fallacy.

 

I agree that having real data would be important but we do know that real data is either secret/confidential or "infected" with false data. So all we have close to real tangible experience that would be "acceptable" to all is ED experience, maybe.

 

So all we could to do... is to make sure our findings or observations about incorrect are accurate rigorous observations. Maybe ED wanted Z missile to perform in x way but it doesn't, but if we give incorrect test data... ED will disregard them as you observed.

 

If community comes with not perfect designed experiments (tracks and some stories) about missiles performance on their own is normal for ED to disregard them because from missile launch to hit, other things happen beside simple missile performance. The most important imho is target knowing the missile was shot and taking actions that cannot be controlled. Secondly are the conditions of the launch that mostly are disregarded by the testers.

[GGTharos]

I agree that having real data would be important but we do know that real data is either secret/confidential or "infected" with false data. So all we have close to real tangible experience that would be "acceptable" to all is ED experience, maybe.

 

... seriously? How do you know that they are 'infected'?

 

There is plenty of real data around, you just might not be able to find it on the internet for free. There are research papers out there, but you have to know what you are looking for and you have to be willing to pay money for the papers. Sometimes they might contain something you're looking for.

 

If community comes with not perfect designed experiments (tracks and some stories) about missiles performance on their own is normal for ED to disregard them because from missile launch to hit, other things happen beside simple missile performance. The most important imho is target knowing the missile was shot and taking actions that cannot be controlled. Secondly are the conditions of the launch that mostly are disregarded by the testers.

In general we can control experiments quite well by having capable targets, and the AI can be used as a simple, non-maneuvering drone, too.

 

ED themselves have limited time to deal with missiles: They have an entire flight sim to deal with. Sometimes they also don't have all the facts either. They've done a pretty good job with missiles in general, but things can always be better.

 

So yes, with that in mind good work/proof is needed to make changes, and most of the community is not in a position to provide that. That's fine, they own a game and have an opinion about it. There's no reason to tell them to stop talking about it. It's not like people can't discriminate what they're going to read about.

Edited December 4, 2015 by GGTharos

[IASGATG]

Ahh, another one of these threads again, my favourite!

 

Okay uninitated, here we go. I'll break this down as simply as I can.

 

Firstly we have the hard coded flight model. In a nutshell a series of equations that have look-up data that creates outputted forces which ultimately govern how the missile behaves. The maths equations are an oversimplification of supersonic flow for idealized bodies which is causing large problems in edge cases. (Edge cases here are for AoA >1) As a result the equations need to be augmented, likely with the addition of new variables, in order to to resolve this problem.

 

Secondly we have the look-up data. This is your standard pressure, temperature stuff, but also your drag and lift curves which are in turn derived from look up data found in lua files etc. What this grants is the ability to have custom drag profiles for each missile and so each missile flies slightly differently. The problem here is that the drag and lift curves do not match what the literature suggests they should look like.

 

Thirdly we have guidance. This is the laws that the missile obeys in order to attempt to hit the target. As of at least 1.2 they are very basic which is fitting of early generation missiles and of heat seeking missiles, but not for new gen tech. This reduces the effective range of the missile against maneuvering targets but not against non-maneuvering targets. (Although when lofting is taken into consideration it effects this as well)

 

Fourth we have seeker head and countermeaure susceptibility. This is how easily the missile is able to track the target, how easily the missile can be spoofed by countermeaures and how easily it can reacquire after being spoofed. This appears to be what ED is putting the most effort into at the moment as the code changes every other update. It's an unfortunate grey area and is very divisive as we have literature that says how easily missiles can be defeated, but this suddenly makes an effective weapon useless. This is what I'm least qualified to talk on so I'll stop this point here.

 

What this ultimately means is we have a missile who is losing too much energy in level flight because the drag curves are incorrect, who is losing too much energy in turning because the maths is also incorrect and is turning when it shouldn't be because the guidance is incorrect.

 

Oh, and certain variables are entirely missed from the flight model like drag reduction with active motors and such.

 

I hope we all learnt something today and have gone away as more enriched and happy students.

 

Class dismissed.

[zaelu]

 

There is plenty of real data around, you just might not be able to find it on the internet for free. There are research papers out there, but you have to know what you are looking for and you have to be willing to pay money for the papers. Sometimes they might contain something you're looking for.

 

Is what I would call unreliable data. Rare and payed...

 

In general we can control experiments quite well by having capable targets, and the AI can be used as a simple, non-maneuvering drone, too.

 

Controlled for me is controlled continuously while the experiment is going and also has to be repeatable in same conditions. Is what I described would help.

AI means you don't know what exactly the target does thus it might negate your experiment. Also if you would be able to see what the missile "see" a lot of errors might be disregarded. Otherwise is just a smart bullet with little to do with a multiplayer "strange" event.

[GGTharos]

Is what I would call unreliable data. Rare and payed...

 

Ok, if you want to call papers published in scientific journals unreliable, then I don't think there's really any discussion to be had about realism, since there's no baseline to compare to.

[zaelu]

The problem I see with such sources/data GGTharos is that if you take it after the name "Scientific Journals" you might be tempted to say... good data. But... Scientific journals usually don't talk about: "Classified Military Commercial Products Characteristics". You see... it can't be a consensus on AIM120 characteristics. The Russians for example... can't test them to check it. And viceversa for R27R/R77 etc missiles.

 

For user/gamer of DCS is about deciding to believe a source and put 100% confidence in it. For ED it might be a mix of things. None matters much, what matters is when you want to present a problem in this aspect of the simulation you have to do it is such way that ED could actually think about it at least for a bit. You need to put yourself in their shoes sort of speak.

 

Why would they actually spend a second look into the problem if what is presented are some sort of blank charts (like on that thread) and all wrapped in a bit of arrogance like in the post of IASGATG?

 

Seriously, I can't blame them for not looking at it even for a second.

 

Regarding the base line. As I said. The data is classified/"murked" by false data because it is a military product after all. So declaring "I" have the base line just put you on the wrong foot.

 

So better is to say:

 

"This is how the missile Z is performing in this and this situations/conditions. Is this how it was intended? "

 

To do that you need your tests to be rigorous not just declare "I made my test very good sir" you actually need to provide all the info and control variables.

 

That's why I said AI won't do it. And probably ED knows it. When you put AI versus AI of the missile you basically in front of Schrodinger's box. And pulling a chart out of it is not helping but the contrary.

 

We don't need 7 charts/acmi's with 7 fire situations. We need "100" charts or acmi's with clear conditions with but all in the same situation. No AI, no maneuvering target, no maneuvering launching platform, no test on huge AI bomber then conclusion about some small fighter, etc.

 

That's why I said we need a sort of firing range sandbox for missiles.

 

P.S. just as an anecdote... same super scientific researches made the F4 Phantom without gun at the beginning. Solid data, I am sure.

Edited December 5, 2015 by zaelu

[GGTharos]

Right, I don't do research based on paranoia, so I don't think your advice is useful.

 

P.S. just as an anecdote ... the gun didn't do much for the F-4. Training did.

[OnlyforDCS]

Ok, if you want to call papers published in scientific journals unreliable, then I don't think there's really any discussion to be had about realism, since there's no baseline to compare to.

 

I tried using that line of thought, but it seems a lot of people don't know what "peer reviewed" means and aren't willing to put up the effort to try and understand the scientific process, especially if it disagrees with their line of thought and "reasoning". *sigh*

 

It's a shame that this wonderful sim is still so lacking in one of the most important areas it can simulate. BVR combat. Oh well, we'll always have the mod I guess.

[zaelu]

GGTharos, call it what you like but I take it as simple skepticism. If something might be unreliable because of people having interests and secrets... then probably it is.

 

OnlyforDCS, I thought also at peer review when GGTharos said about Scientific Journals and I find hard to believe USAAF for example would ask VVS to do a peer review on AIM120 characteristics. I am sure you meant something different but... it can mean more things :) .

[GGTharos]

What you did isn't skepticism. It is paranoia.

 

Further, the stuff you're typing out here just shows that you either did not read, or did not understand the work that was done, or what it was based on. You're not asking questions, you're making up some sort of weird scenario that may apply to a movie, but it isn't reflected in reality.

[zaelu]

It could be as you say.

 

But then you should ask why ED didn't embrace that mod or those claims? Try not to be paranoid also.

[IASGATG]

If you read the thread you'd understand why they didn't accept it? They believe that a DLZ curve from a 90's Su-27 handbook of estimated AIM-120 performance based off of Russian intelligence is the maximum aerodynamic performance of the missile and matched the code to that.

 

The dispute is that ALL other data which shows actual Cd/Cl/CF diagrams, velocity/time distance/time diagrams, and USAF/USN DLZ diagrams contradict the Russian DLZ curve.

 

But if you'd read the post you'd know that.

[OnlyforDCS]

It could be as you say.

 

But then you should ask why ED didn't embrace that mod or those claims? Try not to be paranoid also.

 

*disclaimer* This is my opinion and it is pure speculation, and in no way should be construed as negative criticism of Eagle Dynamics or their work.

 

There could be many reasons, but primarily I feel it's down to two things. First of all it would make the AIM120C the best missile in the game hands down. This is something that the Russian community would probably reject out of hand. Secondly (and I feel more importantly) due to the many incoming 3rd and 4th generation fighters, they want to level the playing field so to say. Shorter ranged, less accurate, and non CM resisting missiles (from both the Su27 and F15) allow fighters such as the Mig21 and the incoming Mirage2000C (and even the F14) a chance to get in the fight and contribute. Something that they would probably have a really hard time with in real life, considering their weapon systems. Recent improvements to the short ranged IR missiles gives even more credence to this "theory" of mine, as it encourages players to perhaps go into the merge with more confidence instead of disengaging for another BVR pass.

All of this can be shortened down into the most hated word in the simulation genre: Balance. Which is why I speculate we will never hear about it, or it would be outright denied.

Edited December 5, 2015 by OnlyforDCS

[GGTharos]

Actually that work was a proof-of-concept. Even though it was specific to the 120, it applies to other missiles also, meaning that other (radar guided) missiles should also gain range.

 

The changes to IR SRM maneuvering were made based upon our recommendation, supported by proof of maneuvering capability for this class of missiles.

[OnlyforDCS]

Actually that work was a proof-of-concept. Even though it was specific to the 120, it applies to other missiles also, meaning that other (radar guided) missiles should also gain range.

 

Yes I know that the range of most missiles was increased including the ERs. However as I understand it, in the mod the ER is limited by it's max operation of 60 seconds.

 

The changes to IR SRM maneuvering were made based upon our recommendation, supported by proof of maneuvering capability for this class of missiles.

 

Yes I know that the changes to the IR missiles is based on real data, didn't know it was your data though. I'm not disputing the changes, just speculating as to why those were accepted and others rejected.

[GGTharos]

The data is from a research paper, same data was used to validate the CFD AFTER the CFD was done. :)

 

As for ERs ... so what? 60 seconds is a looong of time and the ERs are pretty fast ... at those ranges, a crank will prevent a missile from reaching you :)

[zaelu]

If you read the thread you'd understand why they didn't accept it? They believe that a DLZ curve from a 90's Su-27 handbook of estimated AIM-120 performance based off of Russian intelligence is the maximum aerodynamic performance of the missile and matched the code to that.

 

The dispute is that ALL other data which shows actual Cd/Cl/CF diagrams, velocity/time distance/time diagrams, and USAF/USN DLZ diagrams contradict the Russian DLZ curve.

 

But if you'd read the post you'd know that.

 

And it seems to me they didn't accepted your data isn't it? So you think the way you presented the problem helped your cause or not?

 

You try to make it look like missiles are fired by universities around the globe for peer reviewing and only "VVS" is somehow stubborn to accept the fact that "that god exists". Which for me... as a noob is ridiculous. I don't need to know anything about those missiles. All I need to know is that they are a military product that is sold to USAF USN etc and the studies might (or not) need to fit a budget and an agenda and I would be skeptic on definitive data claims.

 

Seems to me that ED is doing the same. So you need to ask. If you want your findings to be taken in consideration... should you insist in your previous approach or should you try something else? If you would have something like my proposed sandbox missile testing range would you be able to break the problem maybe in smaller parts that would reveal better the problems if any?

[OnlyforDCS]

Physics is physics. All of those studies were peer reviewed. You can't fake something like that. Unless you believe that the peer review process can be faked. In that case there is nothing further really to discuss.

[zaelu]

I think a missile is a weapon system produced/advertised by a manufacturer for a govern branch and not simple physics. Other than that, physics is physics, yeah.

[IASGATG]

Zaelu, want to talk verbally about this? Skype/Teamspeak/Something? I think that doing it this way will take too long and wont convey things as clearly.

[zaelu]

It would not be of any use as I am not in the position to decide anything about how missiles behave in DCS. Also I couldn't make the mod or whatever would be necessary to create that sandbox testing range for missiles.

[IASGATG]

It would not be of any use as I am not in the position to decide anything about how missiles behave in DCS. Also I couldn't make the mod or whatever would be necessary to create that sandbox testing range for missiles.

 

Nevermind then.

[Angel101]

Last year I had the opportunity to talk with an spanish hornet pilot, and I was surprised to know that they train BVR to shoot the Aim120 not beyond 11 nm. and you know "train like you fight, fight like you train". (maybe there are other considerations for this, expensive missile, low supply of them, get an higher PK) but at the end of the conversation I get the mental image that the real missiles behave more near of actual DCS representation than the BMS one (range and missile PK between Sim´s are like Day and Night)