AIM-54A Terminal Guidance Flaw

[PE_Tigar]

Short description:

 

1) AIM-54A terminal guidance (intercept point calculation) does not take into account the missile's radar cone limits (thus losing contact within terminal guidance). Intercept point calculation should take radar signal cone ("basket") limitations into account.

 

2) AIM-54A turns to reacquire target after losing contact due to issue #1, but in the wrong direction (away from the target). Should turn towards the last known position of the target.

 

See the following video for illustration:

 

 

The video shows shots where this effect was the most pronounced, where basically no other explanation is possible.

 

Longer description of the test:

 

We've run four setups with all human pilots 4xF-16C vs. 2xF-14B.

 

Weapons expended - 36x AIM-54A, 33 AIM-120C, 5xAIM-9X. Kills (per setup) 1) 1xF-14B (AIM-120C) 2) 2xF-16C (AIM-54A) 3) 2xF-16C (AIM-54A - no evasive maneuvers) 1xF-14B (2xAIM-9X) 4) 1xF-14B (2xAIM-9X). Most weapons were fired within their kinetic range (i.e. the majority of missiles reached targets, or had enough energy to reach target, go active, and maneuver). Some AIM-120C timed out (battery spent) within their active range. Some AIM-120C went off target because of loft issues. Some AIM-120C were lost due to mid-course guidance issues/lack of TWS guidance.

 

Hit rate per missile: AIM-54A 11.1% (5.55% if we discount non-maneuvering targets); AIM-120C 3.03% (one missile hit, head on, ~10nm range); AIM-9X 80% (100% if we discount one inadvertent Fox 2 shot). Most Fox 3 kills were in the first two setups - none in the fourth. That leads to conclusion that the learning curve for evading missiles by using the flaws described is not steep at all - it takes an average pilot about one hour to master this "technique".

 

This post is related to AIM-54A only, there will be a separate one (unfortunately longer) with AIM-120 issues and questions.

 

Tracks and .acmi files available per request (we'll send them to devs or official beta testers, if they're interested).

[Zergburger]

have you tried this with missiles that arent so slow they are basically falling out of the sky? a phoenix doesnt handle well at 300 knots, nor do A models deal welll with chaff and notching

[PE_Tigar]

None of the shots in the video has AIM-54A traveling at 300kt - in all shots the missile has between 1.3 and 2.2 M less than 5 nm from the target. Energy bleed while performing bugged "re-acquisition maneuver" is a part of the issue. None of the missiles goes for chaff at all.

 

If you really want to contribute something you should take a closer look at the video, not just skim.

[QuiGon]

None of the shots in the video has AIM-54A traveling at 300kt - in all shots the missile has between 1.3 and 2.2 M less than 5 nm from the target. Energy bleed while performing bugged "re-acquisition maneuver" is a part of the issue. None of the missiles goes for chaff at all.

 

If you really want to contribute something you should take a closer look at the video, not just skim.

I guess you have never heard of IAS or CAS, because then you wouldn't have said something like that, otherwise, as IAS/CAS is the speed relevant for maneuvering.

Zergburger is absolutly right, that the missile is pretty slow at 350-400kts IAS/CAS, which makes it difficult for the missile to maneuver. Check the speed in TacView (left of the HUD in HUD-View).

Edited April 22, 2020 by QuiGon

[PE_Tigar]

@QuiGon: Is that so? Oh well, I guess they never told me anything about IAS and CAS at ATPL course.

 

So, again, for the attention impaired:

 

1st shot: CAS 712 - 580 (lost tracking) - 390 end of recording

 

2nd shot: CAS 745 - 560 (lost tracking) - 580 end of recording

 

3rd shot: CAS 1080 - 858

 

4th shot: CAS 760 - 585

 

5th shot: CAS 700 - 400

 

All speeds CAS, in knots. Which is, you know, nautical miles per hour. I used Mach numbers just for convenience - didn't expect people to latch onto that, and certainly didn't expect a lecture about IAS and CAS - especially not from people who didn't take time to watch the 4'40" video.

Edited April 22, 2020 by PE_Tigar

[NoJoe]

Looking at shot #3, it looks like the missile is going so slow that it has no choice but to maneuver and move the target out of its seeker field of view. I'm not sure what the seeker FOV is supposed to be, but it looks like that's what's happening. If it maintained more of a nose-on attitude, it would fall behind the F-16.

 

And that looks especially true with shot #4. The missile is just too slow compared to the Viper, especially once the Viper starts its turn away.

 

Looks to me like the Viper pilot is simply maneuvering at the end to be out of reach of the missile.

 

Not sure what happened with shots #1 and #2 though. That does seem odd that the missile maneuvered its nose away from the target and lost sight just as the Viper was jinking. Maybe the same thing, maybe not.

[PE_Tigar]

The speed of the missile in shots 3 and 4 is quite a bit higher than the speed of the target, #3 is actually the fastest missile. Suggest you look at Gs that the missile is pulling - you can actually see the turn start with Gs increasing when the target leaves the "basket". Goes for shots 3 and 4 as well. My point is that the missile should not maneuver so the target leaves the basket due to missile's maneuver itself, and when it loses the target, it should turn back to target to reacquire - not away from it.

[Noctrach]

Missiles in DCS have a scripted behaviour where they will pull 13G lead for a couple seconds from the moment they start getting notched.

This means that if the notch happens at reasonably close range the missile will be forced by this behaviour to overshoot, even moreso if the target changes direction.

 

This can be replicated by putting a missile 90 degree aspect (notching) and keeping it there at any altitude, speed or range. The 13G lead pull will expend so much energy and put it in such a poor position that a follow-up intercept is essentially impossible.

Note that this has nothing to do with seeker limitations, as missiles in DCS can reacquire during this "forced overshoot" period, resulting in an attempted U-turn once past the target.

 

 

Clear example with AIM-120C at 35,000 feet, no chaff used: https://gfycat.com/SlightPoshEmperorpenguin

 

This behaviour is identical between all missiles.

 

Besides that, a test like this... with targets jinking, chaffing and notching, introduces an incredible amount of variables that can alter the outcome.

In the future I'd recommend setting up cleaner tests, i.e. launch without notch, notch without chaff, etc. to isolate behaviours and get cleaner results.

 

 

So @quigon @nojoe while I agree the shots were very long, missile speed has nothing to do with it in this case. Simply scripted notch behaviour that can be very easily reproduced regardless of closure.

Edited April 24, 2020 by Noctrach