Gau-8 AP testing (for real this time)

[Wrench]

So I've done some testing with the GAU-8 vs tanks, and I've Come to some basic conclusions:

The toughest Russian tank, as one would expect, is the T-90.

At a 45° dive angle;

 

It takes ~20 rounds of AP to kill.

The T-72 takes ~15

The T-55 takes ~10.

 

At 30° degrees, it adds about 3-5 rounds.

 

The aspect angle makes no difference whatever (tested from rear quarter and front quarter, got the same values)

 

That means that if you do a standard gun run as has been explained by various RL pilots,

45° degrees, 200 burst (Die MFer Die)

you'll fire 20 rounds of AP, and if those all hit, you'll kill a T-90.

 

For more detail, this is how I did the test:

 

I wrote a script that used the Hit event to count hits and only count AP hits.

When the tank exploded, it then outputs a text message with that count.

In this way, I could tell how many rounds it takes to kill a tank within various parameters.

For those who want to try this out for themselves, just run this is a DO SCRIPT box at mission start.

 

if not Wrench then Wrench = {} end
local hits = 0
--S_EVENT_HIT = {  id = 2,  time = Time,  initiator = Unit,  weapon = Weapon  target = Object}
Wrench.A10GunEvent = {}
function Wrench.A10GunEvent:onEvent(event)
if event.id == 2 then
	if event["weapon_name"] == "weapons.shells.GAU8_30_AP" then
		local target = event.target
		hits = hits + 1
	end
end
end
world.addEventHandler(Wrench.A10GunEvent )
--S_EVENT_DEAD = {  id = 8,  time = Time,  initiator = Unit,}
Wrench.DeadEvent = {}
function Wrench.DeadEvent:onEvent(event)
if event.id == 8 then
	trigger.action.outText(hits, 10 , true)	
end
end
world.addEventHandler(Wrench.DeadEvent )

 

I'll be doing much more exhaustive testing on this soon, but I thought I'd post my findings thus far.

[Dagger71]

interesting.I am guessing you are limiting the hits to just top armour?

[Wrench]

No, naturally a lot of the rounds are going to hit the turret, but that's pretty much what would happen operationally anyway. (meaning operationally in the sim)

At a 45° dive angle, 0° or 180° aspect, very few to 0 rounds will hit the side armor, and any rounds hitting the front armor would have hit that operationally anyway.

[Sierra99]

Wrench

 

I'd be interested in how many hits it Normally takes to reduce the target to 50% capacity or less. In real life testing many times the target was not "Destroyed" but immobilized or rendered non functional. (Round through the barrel verses round through the engine block scenarios)

 

I'd be curious how damage is modeled for Ground vehicles.

[Sierra99]

You might find this interesting...

 

https://apps.dtic.mil/dtic/tr/fulltext/u2/a522397.pdf

[Wrench]

True that an M kill is a kill, but that isn't really modeled in DCS, as a vehicle will limp along and fight long after the surviving crew should have bailed out.

You'll also notice that the vast majority of tank kills with the 30MM are C kills, given the squishy humans inside are the most vulnerable part of a tank.

So what I'm looking for is how effective is the gun in the DCS universe.

I've heard that in actual combat the average number of hits required to "kill" a tank, be that destruction, mobility, armament or crew, was 6 AP hits.

 

Given that number, and assuming most kills were M or C kills, 15 rounds for a K kill is a pretty reasonable guess.

 

I might simulate M kills with a script that would turn the AI of the tank off once it is hit with 6-10 rounds of AP, but that script wouldn't work if there are groups with multiple tanks.

 

I mostly thought I'd post this as I've seen a lot of talk about the damage model of the tank vs GAU-8 scenario, but never anyone collecting hard data.

 

I'll have to come up with a good testing methodology to collect some hard numbers in the coming days, but I thought this would be worth mentioning given the numbers I'm seeing so far.

 

One idea I've had is to 'hack' track files to change out tanks and have perfectly repeatable gun runs, but I'm not sure if tracks just replay the gun runs and re simulate the damage, or if it records the damage and simply plays it back.

I'm inclined to believe that it is simulated every time, which is why replay errors have beleaguered track files for so long.

 

Either way, interesting thought.

[Harker]

You can use UNIT AI OFF or EXPLODE UNIT in the ME to single out units in a group. It's extremely tedious if the group is large, but it can simulate an M or C kill or an ordnance explosion. You can even add a RAND chance to select one of the two. Granted though, this is more geared towards mission making.

 

Thanks for taking the time to test, btw!

[Wrench]

Harker, I wasn't aware that you could turn the AI off on a unit, this may bear fruit after all.

We'll have to see how the rest of the group responds.

I've also thought about spawning infantry when the tank gets "M Kill"ed, as though it were the crew bailing out, but we'll see.

 

This whole thing actually started with a script that would trigger a small explosion whenever a tank was hit by a 30MM AP round, but there were two problems:

1, The explosion effect, even of an explosion set to a power of 1, is big and looks weird from just a 30MM shell.

2, it really isn't that hard to kill tanks in DCS, given we have to K Kill them.

[Harker]

The spawning infantry is a nice touch, but isn't it kind of irrelevant at this point? As far as the pilot is concerned, the target is out, he's not going to do another pass for 3 or 4 crewmen. Plus, because DCS is DCS, spawning them in might cause a micro-stutter sometimes, which will let you know that you indeed killed the tank. For missions with a limited number of vehicles though, such as single tank section, it might add to the mission.

 

As for the UNIT AI OFF, the unit simply stops doing anything. It'll stop moving, engaging and reacting. Keep inn mind that if it's in a column, vehicles behind it might get stuck, because it'll block their way and they're not smart enough to go around it. Better use with other formations.

[QuiGon]

I'd be curious how damage is modeled for Ground vehicles.

It's a simple hit point system, where the entire vehicle is checked as a whole (no damaging of specific subsystems). The vehicle explodes at 0% health, but recieved damage already reduces its capabilities (e.g. 50% damage = 50% max speed, 50% accuracy, 50% reaction time).