need evidence AIM-120 Easily driven into terrain

[Default774]

The AIM-120 is currently very easy to drive into the terrain. When flying near terrain or trees it is possible to make the AIM-120 fly into the terrain with relative ease by flying downwards for a short period of time then pulling up again. Presumably, performing such a manoeuvre puts the predicted impact point somewhere deep underground which the missile does not account for, it then flies into the terrain.

Obviously, this problem gets worse as soon as you are not flying on perfectly flat featureless terrain. As soon as trees, buildings and other map objects come into the equation driving the missile into terrain becomes much easier.

I am not asking for the missile to have perfect terrain following/avoidance capability, but some adjustments like not allowing the missile to fly below the target under some circumstances, or making the missile recognise ridiculous predicted impact points would go a long way in preventing the missile guidance from being exploited in this way.

 

120_grn_4.acmi 120_grn_3.acmi 120_gnd_2.acmi 120_gnd_1.acmi 120_grn_4.trk 120_grn_3.trk 120_grn_2.trk 120_grn_1.trk

Edited February 22, 2023 by Default774

[okopanja]

Are there any documents to demonstrate that real AIM-120 do have terrain avoidance feature?

[GGTharos]

This is not an AIM-120 problem, this is a generic RF missile problem in-game and the solution to this exists since the 50's (not necessarily implemented everywhere since the 50's, but the issue was well understood).

Altitude guidance should be pure until a certain distance or maintain the target at a certain small elevation below the horizon until the target drifts low enough, then switch to PN.   These techniques cover different use cases.   DCS isn't there yet with respect to guidance capabilities.

[okopanja]

29 minutes ago, GGTharos said:

This is not an AIM-120 problem, this is a generic RF missile problem in-game and the solution to this exists since the 50's (not necessarily implemented everywhere since the 50's, but the issue was well understood).

If you are refering to SA-2 and contemporary SAMs, you are correct, but in their case locations are preselected with known terrain configuration and them being on bottom level helped a lot.

[GGTharos]

2 minutes ago, okopanja said:

If you are refering to SA-2 and contemporary SAMs, you are correct, but in their case locations are preselected with known terrain configuration and them being on bottom level helped a lot.

I'm not, but they're included.  Terrain surveys are also not 'contemporary' IMHO.  None of what I mentioned specifically requires this.

Edited February 23, 2023 by GGTharos

[okopanja]

17 minutes ago, GGTharos said:

I'm not, but they're included.  Terrain surveys are also not 'contemporary' IMHO.  None of what I mentioned specifically requires this.

 

If you carefully look the video you will notice this occured at low distance, which means exactly what you proposed.

Can you suggest how to make missiles in DCS ground aware?

[GGTharos]

They don't need to be aware of the ground at all, that's the idea.   They just need to be conservative about taking a dive and this is all done in relation to the target.  This is accomplished by running a different guidance method on the altitude axis.  What 'short' distance is, is very subjective depending on what your purpose is.

Edited February 23, 2023 by GGTharos

[Coxy_99]

ED please create 120 laser missile that never misses, Im sorry but what world does this community live in Could you imagine a missile fired through a forest of trees the missile is not going to say to itself oh look tree we must fly over them Can we please make proper bug reports and not youtube stories thanks.

[Default774]

33 minutes ago, Coxy_99 said:

ED please create 120 laser missile that never misses, Im sorry but what world does this community live in Could you imagine a missile fired through a forest of trees the missile is not going to say to itself oh look tree we must fly over them Can we please make proper bug reports and not youtube stories thanks.

I address this exact point in the last paragraph of the OP. I am not asking for the missile to have magic terrain avoidance. I am asking for adjustments in guidance laws to reduce the chance of the missile hitting the ground when targeting diving targets. Did I not make this clear enough in my original post? If so I will add clarification.

[=BoB= David]

Problem is that AIM-120s are simulated for maybe 30% in terms of kinetics, functions and guidance in summary,  ED have thousands of these tracks, but to be honest we can only hope it will change in future...

[GGTharos]

What do you mean by kinetics?

[BIGNEWY]

If you have unclassified evidence please PM me. 

If not, you are unlikely to see a change here. 

thank you

[Pavlin_33]

3 hours ago, BIGNEWY said:

If you have unclassified evidence please PM me. 

If not, you are unlikely to see a change here. 

thank you

Was about to mention this. We just assumed that 120 has this capability.

[skywalker22]

AMRAAM has no clue about where the ground is, or any other obstacles on the way, everything on it's way can be a target. Its up to the pilot to shot it in the "correct" parameters, to make sure the missile will hit it's target.

Check here which parts consists, specially read carefully of WGU : https://www.globalsecurity.org/military/systems/munitions/aim-120-design.htm

[GGTharos]

5 hours ago, BIGNEWY said:

If you have unclassified evidence please PM me. 

If not, you are unlikely to see a change here. 

thank you

Hi @BIGNEWYthere's no such information for almost any missile whatsoever - ie. can you find any unclassified evidence that AMRAAM uses PN or APN etc for guidance?   But we know that this is basically what AAMs do and that's why ED has programmed it this way.

Likewise, issues with intercepting low-flying targets and dealing with a diving target have been known about and dealt with to some degree since the 50's.  There are specific methods for specific SAMs that we know of and are documented (eg. SA-2) and then there's the basic idea that this makes sense so use it, like most homing AAMs using PN unless otherwise specified.   

This is one of those things:  Use a different guidance algorithm in the altitude axis in order to avoid getting driven into the ground - don't need to know anything about the ground for this, it's strictly about what the target is doing and thus I'm not implying any kind of terrain following or terrain knowledge or detection - the missile will still smack a ridge the target puts it between itself and the missile, but if you dive it will not attempt to beat its target to the ground when it makes its own dive, rather it will limit itself to diving to the target altitude (or keeping the target slightly below the horizon) until it's really close (say 1nm or closer).

Edited February 25, 2023 by GGTharos

[S]

Maybe I'm a little dense here, but the further the missile is from the target, the more it will adjust its flight path, no? If it's 10nm out and sees the plane enter a 10-degree descent, it's going to calculate an intercept based on the increasing speed of the target and will lead the target more and also be moving to intercept at a lower altitude. The missile doesn't 'see' the ground, so it's really just continuously computing the impact point and steering towards that. 

IMO, if you adjusted so the missile calculates to target altitude instead of expected altitude, you're going to have people ducking under missiles instead of putting them into the ground, which is probably more awkward.

[Pavlin_33]

25 minutes ago, Бойовий Сокіл said:

What are you talking about? Of course it knows this (when it matters), as does the fighter radar. There is a big difference between mid-course guidence laws and terminal guidence laws. I have several sources on my computer and will provide the respective literature references later on. There is a reason the running anecdote is "the missile knows where it is".

Thats's a bit funky. The missile has no idea where the ground is and neither the launching aircraft (DCS ones). What could be possible is that at a certain altitude (altitude at launch minus current alt) it stops making huge intercepts. But this would only work for certain terrain.

More important point is that we have no prof that 120 has this capability. We are just guessing here. Also, do we know any A2A missile that does?

[GGTharos]

3 hours ago, S said:

Maybe I'm a little dense here, but the further the missile is from the target, the more it will adjust its flight path, no? If it's 10nm out and sees the plane enter a 10-degree descent, it's going to calculate an intercept based on the increasing speed of the target and will lead the target more and also be moving to intercept at a lower altitude. The missile doesn't 'see' the ground, so it's really just continuously computing the impact point and steering towards that. 

IMO, if you adjusted so the missile calculates to target altitude instead of expected altitude, you're going to have people ducking under missiles instead of putting them into the ground, which is probably more awkward.

The missile isn't calculating anything; it's using PN.  That means holding the target steady in the FoV and outmatching any apparent target acceleration away from the steady position.  You can have the missile aim to hold the target on the horizon if it tried to dive hard, until it's close enough to go full PN.

Edited February 26, 2023 by GGTharos

[TheFreshPrince]

If you look at the GIF in the OP, to me it looks like the missile is leading way too much. It goes Mach 2.5 and even if there was no ground, it looks like it would have overshot the target by a lot. At that distance the missile shouldn't really need to lead anyway, it could just fly "straight" (with a slight lead of course) into the target.

[Coxy_99]

I found the issue there's no manual 10 degree loft. This is why it hits the ground so easy.

[Mareno]

Am 26.2.2023 um 19:37 schrieb Coxy_99:

I found the issue there's no manual 10 degree loft. This is why it hits the ground so easy.

This has nothing to do with the issue.
 

[BIGNEWY]

folks please keep it civil. 

We have to work with data here, track replays and unclassified evidence, but I will mention it to the team. 

thanks

[skywalker22]

I found an "old" thread, with some useful information about guidance systems, worth some reading:

 

[Default774]

9 hours ago, skywalker22 said:

I found an "old" thread, with some useful information about guidance systems, worth some reading:

 

Unfortunately unless you manage to find a document that explicitly states the 120 will not in fact fly itself to intercept a predicted impact point that's 5000ft underground this is all a bit pointless sadly.

I would love to see Maestros thoughts on this (I cant figure out a way to tag his username, dont have a cyrillic keyboard)

Edited June 23, 2023 by Default774

[Default774]

@Маэстро