AIM-120 guidance - Are CFDs Accurate?

[nighthawk2174]

So something I've been noticing for a while especially since 2.5.5 and now quite consistently since the last update is this:

Missile isn't even in the notch yet it will pull a lot of g away from the target. Quite odd and frustrating when it happens. It does seem to happen quite consistently against the AI in my tests. I haven't tested in MP yet though.

 

Trk files

https://www.dropbox.com/s/cfj6dfhg6qgnlv7/AMRAAM%20Bug3.trk?dl=0

https://www.dropbox.com/s/wb9hkkw27yr8aj6/AMRAAM%20Bug2.trk?dl=0

 

Edit:

No chaff

Edited March 20, 2020 by nighthawk2174

[Shadow KT]

Yes, I have noticed this and it definitely has been around for more than a year ....

 

It is kinda hard to replicate, but sometimes if the missile looses sights (gets notched) just as its target starts a sharp turn, the missile will yank its self

[nighthawk2174]

In this case the apparent closure rate as reported in tacview was at 145-155kts when it pulled. But yeah this has been around for a while it just seems its far easier to replicate now, I had this happen 4 times in a row.

[CypherS]

Given that this happens with the AI a lot, and that the AI knows exactly where to go to notch a missile. I'd say that the notch parameters are out of wack.

[dorianR666]

Given that this happens with the AI a lot, and that the AI knows exactly where to go to notch a missile. I'd say that the notch parameters are out of wack.

shouldnt AMRAAM guide towards the last predicted position of the target? its weird that it completely gives up after losing the target for half a second.

[nighthawk2174]

Yes even if the bandit enters the notch for a fraction of a second it should not do this. Not only should it have a track memory As its pulling at 20+ g's (where it had been previously 3g's) and the wrong direction. At the ranges seen above for a missile it should be essentially impossible to A) notch for anything but absolute fractions of a second B) Even notch the radar due to the minimal range meaning the target return would completely overwhelm any background returns.

[CypherS]

Yes even if the bandit enters the notch for a fraction of a second it should not do this. Not only should it have a track memory As its pulling at 20+ g's (where it had been previously 3g's) and the wrong direction. At the ranges seen above for a missile it should be essentially impossible to A) notch for anything but absolute fractions of a second B) Even notch the radar due to the minimal range meaning the target return would completely overwhelm any background returns.

 

It could also be that the missile is going for the chaff hence the hard pull. It doesn't make sense though; if the missile went for the chaff it shouldn't then reverse its turn back to the bandit

[nighthawk2174]

No chaff was used by the AI in the video in my first post.

[TAW_Blaze]

I've seen it happen in MP too although it's not as common since players cannot time their notches that perfect (or maybe it works differently in SP). Against the AI most of the time it goes completely stupid around pitbull. It actually dumps about 20-30% of airspeed if not more very often and puts itself in an impossible intercept scenario and/or alltogether stops tracking.

[Exorcet]

I've also seen this. I can't really pin down why it happens. Originally I thought it was due to a low closure rate but eventually I saw it occur at high speeds. The AMRAAM still needs some guidance tuning. The loft parameters also don't seem to work beyond 50 or so miles. Launch at something further than than and the missile essentially goes straight up and runs out of energy.

[TAW_Blaze]

I've also seen this. I can't really pin down why it happens. Originally I thought it was due to a low closure rate but eventually I saw it occur at high speeds. The AMRAAM still needs some guidance tuning. The loft parameters also don't seem to work beyond 50 or so miles. Launch at something further than than and the missile essentially goes straight up and runs out of energy.

 

To me it still looks like the "last known intercept point" is not existing at all. This is why you see missiles turning away from target in the notch etc. I think if this was finally implemented it would solve a lot of these problems.

 

 

Just from theoretical standpoint I cannot believe late 2000s missiles do not have any mechanisms to deal with intermittent loss of intercept data and would rather result in this completely absurd behavior. Even if you lose the track for a short time you can extrapolate based on last known trajectory, an aircraft flying at 500 kts will not instantly stop mid air or start flying backwards as indicated by some ingame behavior.

[Shadow KT]

I do believe they actually posses that ... maybe we are speaking about different thing, but from what I've seen, when a missile does this it takes the last data it had for its targets and if that was the beginning of a very sharp turn for example, it just extrapolates it and keep going as if the target has continued with that sharp turn or angular velocity or something like that, but obviously the target did not, so what we see happens

Edited March 24, 2020 by Shadow KT

[AeriaGloria]

Yeah I heard someone say that 2.5.6 introduced mid course guidance

[TAW_Blaze]

Yeah I heard someone say that 2.5.6 introduced mid course guidance

 

MCG and flying to last known intercept point are 2 totally different things..

[nighthawk2174]

I think what you may have heard is that the magic INS bug, the missile always knowing where the target is during midcourse, was probably fixed for a few missiles.

[AeriaGloria]

Thanks was a little confused, makes sense

[AeriaGloria]

I guess I was just confused becuase that’s a really big deal. To me the magic INS was how the missiles always has been, I never thought of it as a bug but as simplified missile modeling. So actually having to support missiles now, I’m surprised it wasn’t announced with everything else, unless they want it to be like the magic INS missiles never happened

[Harker]

Bump.

I get this quite frequently with the AIM-120C and honestly, it's completely illogical behavior. The missile pulls a hard turn away from the target, a little before or just before impact, for seemingly no reason.

 

Even if chaff was used or the missile was notched, it still makes no sense. Considering the target's flight path and its proximity, it should be quite easy to both reject chaff and at least fly towards the predicted intercept point for long enough to reacquire the target.

[PE_Tigar]

It does seem to happen quite consistently against the AI in my tests. I haven't tested in MP yet though.

 

Happens in MP as well, that's how the missile reacts when losing the target in notch momentarily.

[Sandman1330]

Phoenix does the same thing, it will go from 1600kts down to 600kts after a turn like that, and not have enough energy to intercept afterwards.

[nighthawk2174]

Yeah the phoneix, currently, is essentially an amraam.

[GGTharos]

To me it still looks like the "last known intercept point" is not existing at all. This is why you see missiles turning away from target in the notch etc. I think if this was finally implemented it would solve a lot of these problems.

 

Correct, it doesn't exist. The missile is flying PN. It doesn't need to exist either in this case, PN edge cases could simply be handled better.

[TAW_Blaze]

Correct, it doesn't exist. The missile is flying PN. It doesn't need to exist either in this case, PN edge cases could simply be handled better.

 

It's painful to see missiles tracking a target turning at 600 kts only to give up, pull back behind him and intercept a chaff doing 0 kts instead :D

 

If only some kind of ECCM existed to reject completely implausible target behavior..

[Druid_]

Its not just AIM120s. Their ordinance FM is a little off …… (FF 45 secs)

 

[TAW_Blaze]

Its not just AIM120s. Their ordinance FM is a little off …… (FF 45 secs)

 

 

 

It looks like it's guiding on a totally different target illuminated by something else.