Air-to-Air Missile Discussion

[Falcon_S]

IASGATG, spyro23, Exorcet - Could this your science fix the missile? ... :book: ... :joystick:

[TAW_Blaze]

They are talking about kinematics, not guidance :)

[spyro23]

I work with automotive aerodynamics currently. We fix the area when it comes to reporting out results because we often morph vehicle shapes when doing studies. Working with CD means we don't have to worry about units so there's no confusion over Newtons to pounds, etc. It also helps when it comes to doing quick mental math because there is only one scale to remember (ie .2 CD is very good, always, I don't have to worry about what is a good range in dimensional units, of which there can be many).

 

The only issue that might come up is comparing data across completely separate projects, but I've had to do this yet.

 

OK, this explains a bit your point of view. If you talk about cars giving only Cx (Cd) makes sense for comparision of "aerodynamic slickness" or Fd at given speed because at speeds that cars are going Cx is pretty much constant (same as A and rho).

 

Applying such conduct in aeronautics aerodynamics is simply impossible because:

1. Cd is not a constant throughout flight speeds envelope

2. rho is not constant (different altitudes, different temperatures)

3. even A can't be considered as a constant because of alfa and beta (angles of attack and slip) - simplistic approach

and last but not least velocity range is much much greater

 

 

EDIT:

 

@ IASGATG

 

Didin't you mix a bit AIM-9 versions Cd graphs? Aren't AIM-9L and M versions the same aerodynamically speaking?

To me it seems that AIM 9 L graph may be in fact for AIM-9P

Edited May 24, 2014 by spyro23

[*Rage*]

They are talking about kinematics' date=' not guidance :)[/quote']

 

Not relevant to this thread I think is the point he is making.

[IASGATG]

We just need a Cd curve for sea level that can then be used as the game knows what the A is, what the V will be and what the rho will be based off of altitude. What's the conversation we're having here? I don't understand the point of it..

[GGTharos]

The two curves from the flyout study are:

 

AIM-9L

AIM-9L VARIANT with AIM-120A/B fins and nosecone scaled to AIM-9 body.

 

These are real life shots.

 

IASGATG's graphs represent those with the AIM-9M and AIM-9L VARIANT. There's no AIM-9P.

 

Didin't you mix a bit AIM-9 versions Cd graphs? Aren't AIM-9L and M versions the same aerodynamically speaking?

To me it seems that AIM 9 L graph may be in fact for AIM-9P

[spyro23]

We just need a Cd curve for sea level that can then be used as the game knows what the A is, what the V will be and what the rho will be based off of altitude. What's the conversation we're having here? I don't understand the point of it..

 

I might have drifted a bit but again:

 

 

...... The biggest problem with this curve is Cd at sub sonic being more draggy than at supersonic. Common sense says this shouldn't be the case. (I wont go further.) ..........

 

 

I can't see any problem here. It is nothing unusual to have Cd at supersonic speed lover than at subsonic speed. When talking about science common sense approach is not always a good approach.

 

This Do Not mean that Fd at supersonic will be lower then at subsonic speed because of V squared.

 

You can make another graph Fd/V based on your Cd/V graph (you can omit A and rho) and this should show that Fd (actual drag force) at supersonic is bigger than at subsonic.

 

If you can proove that Fd gets smaller with speed then there is something wrong with Cd.

Edited May 24, 2014 by spyro23

[IASGATG]

That's actually a typo. It should be the other way around, we can forget the whole point.

 

Transonic > Super Sonic > Sub Sonic. Will correct now.

 

I misunderstood your grammar before. I thought you were questioning why it went back down again after the transonic.

 

You happy for an admin to delete the posts regarding this so we can tidy it up?

Edited May 24, 2014 by IASGATG

[spyro23]

Excuse me dragging this subject further but you lost me you last post.

 

 

............Transonic > Super Sonic > Sub Sonic. Will correct now............

 

 

Are we still talking about Cd or about Fd?

 

You happy for an admin to delete the posts regarding this so we can tidy it up?

 

Why?

[IASGATG]

Only Cd. I've never mentioned Fd.

 

And because we have two pages of posts over 1 typo that I made that's making a huge horrible mess.

[spyro23]

Clear now but:

Super Sonic Cd being < Sub Sonic Cd does not automatically mean that it is wrong ergo

....Transonic > Super Sonic > Sub Sonic. Will correct now.

is not valid.

 

It seems that we are about to circle around discussion about Cd so I'll now shut up.

Edited May 24, 2014 by spyro23

[IASGATG]

Cd at super sonic can be lower that sub sonic, however, not with the shapes and speeds we're talking about. Moreover, since surface area wont change, since density wont change, Fd will change as Cd changes. Yes V is squared but.. ... You get what I'm saying right? The conversation is mute..

 

I feel we're talking in circles for no reason. All the points to do with the graphs are still just as valid. Can we seriously delete these last two pages because they serve no purpose at all.

Edited May 25, 2014 by IASGATG

[FLANKERATOR]

Absence of proximity fuse was most staggering to me.

 

Look-down tracking and especially at that distance is not what an ER was designed for. Sure it can hit but that is going to be very difficult especially against a target that low and pulling 6G consistently.

 

What would happen if instead of AI, we had humans switching to ILV/MED or diving aggressively to force a look up? Also, at that distance, they'd have closed in for an R-73/cannon shot.

 

There are for sure things to be improved missile-wise, but maybe the priority is to fix how AI fights first.

Edited May 25, 2014 by FLANKERATOR

[TAW_Blaze]

I agree with flankerator, it's invalid to judge missile performance based on how AI use it, since AI is terrible itself. Do the test with 4-5 human surrounding you, see what's the result.

[Bimbac]

Missile performance myths

 

Hello everyone!

 

I think that discussion is all about a different subject. I understand that many of you are very frustrated about the Air-to-Air missiles performance in the latest DCS version. Frankly, me too. But the main problem lies somewhere else.

 

First, you have to assume that all missile data in this sim come from public domain, therefore they are inherently inaccurate.

 

Second, without an accurate physics simulation engine, you are unable to simulate radar performance, kinematics and aerodynamics properly, therefore even if you somehow obtained real missile data, their performance within the simulation would be different from real world.

 

Third, most of the missile performance data in DCS, concerning seeker sensitivity and tracking ability are, in my opinion, based on assumptions. Most people just ASSUME that since R-27 is older than AMRAAM, it HAS TO BE worse missile in terms of seeker performance. All this just makes me laugh.

 

Most of the people here have simply no idea or knowledge how actually these systems work at all. Most just assume. And, there is also some bias, of course. The Americans assume their missiles are superior to Russian ones, the Russians tend to think opposite. Let me tell you that R-27ER is very deadly, even today. You have to understand that R-27ER built in early 1980s is (internally) different from R-27ER made in late 1990s or early 2000s. Its performance in terms of top speed should be much better than those of the AMRAAM C, in fact, in every second of its flight, the Alamo C is twice as fast as AMRAAM. The mid-course guidance and data-link features of these missiles are currently not modeled at all, or the capabilities are severely degraded.

 

You also have to realize that DCS is still a GAME. Very accurate and well done in some aspects, indeed, but still a game. Therefore, there are some features implemented that make it easier than real life. For example, MiG-29 or Su-27 radars need an STT lock to fire and guide the R-27R, but there is no switch to a different mode or CW after it is fired, so technically, your RWR system shouldn't give you a launch warning at all. Therefore, every Eagle driver, when locked in STT by MiG or Su, would immediately have to consider the possibility of being attacked. And this is just one example.

 

So please, be patient and give ED some time to correct these deficiencies in the future. I've heard a rumor that in times of the very first Su-27 simulator in 1995, for being very precise and accurate, some people from the programing team were arrested and questioned by police and military.

 

So, let's accept the fact that because of obvious reasons, the DCS will never be perfect, but I beg the ED to make it as close to the reality as possible. The simulator deserves it, and so do we.

 

Have a good day!

[JNASova]

Bimbac,you are right about everything.Since 1.12,there is no much progres.Even much of what was good in 1.12, in the FC 2 and FC 3 is broken.Hard to understand.

Edited May 25, 2014 by =JNA=Sova

[Axion]

I agree with flankerator' date=' it's invalid to judge missile performance based on how AI use it, since AI is terrible itself. Do the test with 4-5 human surrounding you, see what's the result.[/quote']

 

I'd hazard a guess that the majority of FC players do so offline? Therefore, the use of AI in this instance becomes very valid indeed.

[TAW_Blaze]

I'd hazard a guess that the majority of FC players do so offline? Therefore, the use of AI in this instance becomes very valid indeed.

 

Inproper use of radar is a different problem than missile guidance. Or you want to fix broken AI logic by changing missile behaviour? That doesn't work like that.

[Falcon_S]

Does it even matter who launched a missile? The same is used by AI and player, radar too. This is about the behavior of missile. Why missiles so close to the target give up (and often earlier)? In the next post i will show you the problem in pictures.

Edited May 25, 2014 by Falcon_S

[IASGATG]

First, you have to assume that all missile data in this sim come from public domain, therefore they are inherently inaccurate.

 

Sorry, sarcasm is coming back out because of the dumb. If you have a classified document from 1970 that is now deemed unclassified and viewable by the public doesn't suddenly make all the information in it made up. What it makes it is valid and correct information that has since become no longer a threat of national security because of more advanced replacements. Hence why you can get all the spec on AIM-9's all the way up to the L but now further. Why you can get specs on all the AIM-7's all the way up to the F but not further.

 

Moreover, physics doesn't give a **** if you're classified or not. Laws of nature don't change because sometimes they're used under secrecy and sometimes they aren't.

 

Second, without an accurate physics simulation engine, you are unable to simulate radar performance, kinematics and aerodynamics properly, therefore even if you somehow obtained real missile data, their performance within the simulation would be different from real world.

 

If you obtain correct missile data and it says a missile behaves in a specific way under specific circumstances and you then build a curve that makes the missile mirror that, you have generated an accurate simulation of real world performance. Just because the numbers you're using in the back engine might not be right. The Cd curves, the A curves, the rho curves might all be off. The way the engine interpolates them might be off. Ultimately, if you have a set of curves that say how the missile flies in the real world, and your sim missiles fly the same, what's the problem?

 

Third, most of the missile performance data in DCS, concerning seeker sensitivity and tracking ability are, in my opinion, based on assumptions. Most people just ASSUME that since R-27 is older than AMRAAM, it HAS TO BE worse missile in terms of seeker performance. All this just makes me laugh.

 

I'd imagine they're assumed as well, except for maybe the R27 family. And yes, it is worse. Else the Russians wouldn't replace it. In the computing world, 20 years is a long ****ing time. I'm just gonna assume you're either kidding or incredibly ignorant.

 

Most of the people here have simply no idea or knowledge how actually these systems work at all. Most just assume. And, there is also some bias, of course. The Americans assume their missiles are superior to Russian ones, the Russians tend to think opposite. Let me tell you that R-27ER is very deadly, even today. You have to understand that R-27ER built in early 1980s is (internally) different from R-27ER made in late 1990s or early 2000s. Its performance in terms of top speed should be much better than those of the AMRAAM C, in fact, in every second of its flight, the Alamo C is twice as fast as AMRAAM. The mid-course guidance and data-link features of these missiles are currently not modeled at all, or the capabilities are severely degraded.

 

This is where you're again, hilariously wrong. You just dropped the assumption that most of us don't know what we're talking about. Granted, most of the people (You included) have no idea what you're talking about. You are pulling numbers out your arse without providing any sources what so ever. Since we know what the thrust output of the R-27ER is from Russian Air Force sources, and we know (Close enough) what the thrust output of the AIM-120C is from US sources, we can see who is able to go faster under what circumstances. The delta V on the ER is not more than twice that of the C. Therefore even under frictionless vacuum circumstances, the ER will not go twice as fast as the Charlie. You're just incredibly wrong or you have information I haven't seen. If you have got new information regarding the specific thrust output/fuel grain/motor specs, I'd love to see them.

 

You also have to realize that DCS is still a GAME. Very accurate and well done in some aspects, indeed, but still a game. Therefore, there are some features implemented that make it easier than real life. For example, MiG-29 or Su-27 radars need an STT lock to fire and guide the R-27R, but there is no switch to a different mode or CW after it is fired, so technically, your RWR system shouldn't give you a launch warning at all. Therefore, every Eagle driver, when locked in STT by MiG or Su, would immediately have to consider the possibility of being attacked. And this is just one example.

 

So please, be patient and give ED some time to correct these deficiencies in the future. I've heard a rumor that in times of the very first Su-27 simulator in 1995, for being very precise and accurate, some people from the programing team were arrested and questioned by police and military.

 

So, let's accept the fact that because of obvious reasons, the DCS will never be perfect, but I beg the ED to make it as close to the reality as possible. The simulator deserves it, and so do we.

 

Have a good day!

 

And then the conversation about missiles ends and so does my interest

 

Edit: A word.

Edited May 25, 2014 by IASGATG

[TAW_Blaze]

Does it even matter who launched a missile? The same is used by AI and player, radar too. This is about the behavior of missile. Why missiles so close to the target give up (and often earlier)? In the next post i will show you the problem in pictures.

 

It does. We're talking about a SARH. If you lose lock it won't guide anymore. AI behaviour can cause to lose lock easily, or, make it hard to maintain lock.

[GGTharos]

There are many possibilities, and within those, many possibilities for a bug.

 

For example:

 

- Missile is too slow (in terms of indicated air speed, just to cover many altitudes), does not have enough speed to generate required g at given altitude

 

- Target has entered the notch

 

- Target has out-gimballed the missile

 

- Target has out-rated the seeker gimbal.

 

- Target has r-maxed or r-minned the missile

This is before going into counter-measures.

 

Does it even matter who launched a missile? The same is used by AI and player, radar too. This is about the behavior of missile. Why missiles so close to the target give up (and often earlier)? In the next post i will show you the problem in pictures.

[Falcon_S]

Blaze Bro, i know these things. I count on it every time when i engange target.

 

Ok.. guided missile, seems ok:

 

 

Calculations, performed overtaking...ok:

 

 

Calculations, performed overtaking...ok:

 

 

Calculations, performed overtaking...ok:

 

 

?????????????????????????????????????????????

 

 

This happens to us too, not only AI pilot.

FC3 R-27ER test3.rar

Edited May 25, 2014 by Falcon_S
Posted ACMI

[*Rage*]

Absence of proximity fuse was most staggering to me.

 

Look-down tracking and especially at that distance is not what an ER was designed for. Sure it can hit but that is going to be very difficult especially against a target that low and pulling 6G consistently.

 

What would happen if instead of AI, we had humans switching to ILV/MED or diving aggressively to force a look up? Also, at that distance, they'd have closed in for an R-73/cannon shot.

 

There are for sure things to be improved missile-wise, but maybe the priority is to fix how AI fights first.

 

I would say the priority would be to fix missiles first than AI. Might be easier as well

 

I agree with flankerator' date=' it's invalid to judge missile performance based on how AI use it, since AI is terrible itself. Do the test with 4-5 human surrounding you, see what's the result.[/quote']

 

What?! From 10-15km as long as the lock is maintained it doesnt matter from that range. It is almost WVR. We are not talking about BVR tactics and resultant pKs. We are talking about why ERs fired in the NEZ with a maintained lock and NO COUNTERMEASURES still act stupidly.

Edited May 25, 2014 by ///Rage

[IASGATG]

Order of priority is kinematics, guidance, counter measure susceptibility and finally AI.

 

There is no point altering the AI until you have the game behaving how it should.