GBU-28 malfunctions in mission "Bombing Range Demo"

[HoneyFox]

In this single-player mission, I try to use GBU-28 at the center pylon to engage the bunker.

With TPOD autolasing the target and using either DIRECT or AUTO mode, the GBU seems to ignore the laser spot completely and flies in a straight line (instead of going ballistic) towards some open space nearby.

 

I then go to the mission editor and check if its laser code is not matching, it is 1688, matching the TPOD's.

Finally I try to change the GBU-28 to GBU-24, and GBU-24 works. So probably there is something wrong with GBU-28? (btw, GBU-28 doesn't seem to be an available weapon for F-15E, I cannot find it in the payload editor so if I change it to GBU-24 I can no longer switch it back)

Edited June 23, 2023 by HoneyFox

[Zooker]

Same behavior experienced.

GBU-28 Issue.trk

[nthere]

I remeber that ed haven't implement gbu28 weapon function,although it's available for AI aircraft,but not for human

[Rainmaker]

-28 should not be a available. Its a mission bug 

[CH4Pz]

I had the same thing in the Artificial Reef mission, even when the laser was firing the bomb didnt track, just went full retard... you NEVER go full retard !!

[schurem]

Huh, so that's what's wrong. And here I was going thinking I was borking things up.

[HoneyFox]

One thing to add: the GBU-24 actually has some guidance issue as well, it uses insufficient gravity compensation in the mid-phase of the trajectory and this leads to trajectory being too shallow, causing the guidance law to do pull-up when getting to the final ~80m distance to the target. However. this pull-up is over-done and the bomb would fly right over the target and impact the terrain 5~10 meters farther.

GBU-10/12 however, despite using more primitive bang-bang control law, has good terminal accuracy hitting target.

 

Edit: By checking the autopilot parameters of GBU-24, I found that the parameter "Ki" (likely to be the coefficient of integral part in some PID controller) is set to 10, which seems ridiculously large (especially compared to Knav=4 & Kd=2), and is suspicious to be the culprit of the "over-done" pull-up. I might try tuning down this parameter a bit (and increase Knav for compensation as well, don't know what Kpitch does though...) and see if that could mitigate the problem.

Edit2: Seems like GBU-24 is really heavy which means it heavily relies on Ki to compensate the gravity. Perhaps a better way is to keep Ki and tune up Knav & Kd? Keep testing...

Edited July 3, 2023 by HoneyFox

[Sepdawg_]

Good to know, I thought I was screwing something up. Was making me freaking crazy.

[splash]

I have detected the same problem in Artificial Reef mission. GBU-28 ignores the laser and goes ballistic.

[QuiGon]

The GBU-28 isn't implemented yet. The one available in DCS currently is and old model for use with AI aircraft only and doesn't work properly if employed by players.

Edited July 1, 2023 by QuiGon

[HoneyFox]

Small update (probably a bit off-topic though):

For GBU-24 control law issue mentioned above, I made some tests and I think I've found a better set of autopilot parameters as follows:

		Knav				= 6, -- was 4,
		Kbias				= 3.0, -- was 0,
		Kpitch				= 2,
		Tf				= 0.2,
		Kd				= 3.5, -- was 2,
		Ki				= 15, -- was 10,

The bomb will have less oscillation at terminal distance (<100m), higher probability of hitting target directly, and is more likely to have a steeper impact angle (55+ deg or so, likely due to the increased Kbias value).

While modifying these would definitely make it not passing the server validation, I think it's still worthwhile sharing them. (probably ED could also do some more tests based on these values and further improve the bomb performance?)

Edited August 17, 2023 by HoneyFox

[skypickle]

Would these parameters be effective on an mp server that does not enforce ic if they only exist on client side?

[HoneyFox]

15 hours ago, skypickle said:

Would these parameters be effective on an mp server that does not enforce ic if they only exist on client side?

 No idea, I either play single-player mission locally with modifications, or restore the game to be vanilla when playing online. Never tried playing with modification in server that does not enforce ic before. I don't know how DCS deals with the movement/impact+damage sync for these weapons.

[draconus]

On 7/4/2023 at 7:08 PM, skypickle said:

Would these parameters be effective on an mp server that does not enforce ic if they only exist on client side?

Afaik the launched weapon is simulated on the client side.

@HoneyFoxIt's not about making the weapon fly perfect path. It has to behave as close as possible to a real counterpart - it has to reflect the real guidance and steering in both capabilities and limitations. If you have some data suggesting that it should work like you presented go on with new thread here: https://forum.dcs.world/forum/540-weapon-bugs/ or just read some that are already there.

 

[HoneyFox]

On 7/6/2023 at 4:31 PM, draconus said:

Afaik the launched weapon is simulated on the client side.

@HoneyFoxIt's not about making the weapon fly perfect path. It has to behave as close as possible to a real counterpart - it has to reflect the real guidance and steering in both capabilities and limitations. If you have some data suggesting that it should work like you presented go on with new thread here: https://forum.dcs.world/forum/540-weapon-bugs/ or just read some that are already there.

 

I wouldn't believe that the current parameters of GBU-24 in DCS is reflecting the real guidance, unless the engineers in Raytheon that designed the guidance law for it were really that dumb + the military that would receive the weapon were extremely tolerant of such high probability of missing targets.

Quote

it has to reflect the real guidance and steering in both capabilities and limitations

Reflecting real guidance & steering in both capabilities & limits is the ultimate target of realistic simulation, that doesn't mean you can/have to reach that target in one step only, even ED couldn't achieve or haven't achieved that yet in many cases (JDAM, JSOW, lots of AAM and other weapons, etc).

The current situation is, you would prefer using GBU-10 instead of GBU-24 to achieve high probability of kill, which is simply embarrassing. Since we don't have the source of the real guidance parameters (I'd suspect if ED has real guidance parameters/control diagrams for any other guided weapons as well), the next best thing to do is, to update the weapons with another set of parameters that also has no real source but at least shows closer combat performance to the real counterpart.

 

Anyway it's kinda off the topic so let it just be some info sharing post. 

Edited July 9, 2023 by HoneyFox

[RaisedByWolves]

On 6/24/2023 at 7:23 AM, CH4Pz said:

I had the same thing in the Artificial Reef mission, even when the laser was firing the bomb didnt track, just went full retard... you NEVER go full retard !!

What do you mean….. you people?

[MAXsenna]

37 minutes ago, RaisedByWolves said:

What do you mean….. you people?

You people? Where did you read that?  

[RaisedByWolves]

Thought we were quoting tropic thunder.

[Jayhawk1971]

vor 4 Stunden schrieb RaisedByWolves:

What do you mean….. you people?

What do you mean...you people?!