50 cal Ballistics/Aircraft Damage Model Questions/Concerns

[NineLine]

If this is just going to turn into an ego match I will lock it...

[USARStarkey]

Here are the tracks as requested sith. Human piloted 190s.

190 damage model 1-E.trk

190 damage model 2-E.trk

190 damage model 3-E.trk

190 damage model4-E.trk

190 damage model5-E.trk

190 damage model6-E.trk

190 damage model7-E.trk

190 damage model8-E.trk

190 damage model9-E.trk

190 damage model10-E.trk

190 damage model11-E.trk

190 damage model12-E.trk

190 damage model13-E.trk

190 damage model14-E.trk

190 damage model15-E.trk

190 damage model16-E.trk

190 damage model17-E.trk

190 damage model18-E.trk

190 damage model19-E.trk

190 damage model20-E.trk

[USARStarkey]

for lols.

f15 damage model 2-E.trk

[otto]

@USARStarkey

 

i'll take a look at those traks tomorow and tell you my honest opinion.

 

 

The D9 cannon damage model seems fine to me :

 

 

I got some quick kills with the p51 vs human players(B4 dora was available so p51 vs p51) while booming them.My gun fire converged on the engine and it exploded (Might have been luck).With the d9 i's easyer because convergence is irrelevant.

Edited August 23, 2014 by otto

[9.JG27 DavidRed]

That's simply not true. The AI often takes a 10-15 second burst to down each other. This is also true of the AI FW190 firing at the AI P-51. As to human players, I know *I* have taken a LOT of hits before without going down (or even any significant trouble!). I think both the .50 and the MG151/20 are about half the strength they should be... but the .50 suffers a lot more for the loss of power. The biggest problem is that hits appear to be tracked against hit zones, and do not track (or run a probability table to adjudicate) hits against specific components. Basically, it might be possible to put 100 bullets in a wing without hitting the wing spar, but one hit in the wing spar is likely enough to snap it and kill the plane- and similar holds true for engine blocks, pilots, and (in the 190's case) cannon ammo storage. They need an actual interior ballistics model, or, failing that, a probabilistic damage model, instead of a component-fails-at-X-hitpoint-threshold model. If the model already IS probabilistic, it needs some serious tweaking.

 

what exactly is not true?.that good pilots are able to take out the enemy with a split second burst?

then better tell VIKS or the russian Firmament Fighters, because thats what they are doing for the last two years. :lol:

[otto]

After looking at USARStarkey tracks:

 

The damage model looks ok to me but:

 

In the old il2 the pilot could sustain all kinds of injury form gun fire. It was debated that the pilot was too vulnerable.So i'm not sure if this should be implemented or not in DCS.

 

The f15 looks to go down easier but does it have self sealing fuel tanks ? And of course the jet engines are exposed at the back and easy to destroy.

[USARStarkey]

After looking at USARStarkey tracks:

 

The damage model looks ok to me but:

 

In the old il2 the pilot could sustain all kinds of injury form gun fire. It was debated that the pilot was too vulnerable.So i'm not sure if this should be implemented or not in DCS.

 

The f15 looks to go down easier but does it have self sealing fuel tanks ? And of course the jet engines are exposed at the back and easy to destroy.

 

Right. Because it should take 130 hits to down a airplane statistically....against a non-maneuvering target. As for Il2, the 50s are quite a bit more lethal than in DCS in the latest il2 patches. Do you really not think the cold war era F-15 would not have self-sealing fuel tanks? Are you serious? The engines being at the rear has nothing to do with it. many of those Eagles went down due to wings being taken right off without hardly pulling the trigger.

[Robo.]

As for Il2, the 50s are quite a bit more lethal than in DCS in the latest il2 patches.

 

I respectfully disagree with the above statement. Il-2's .50 really suffer from the lack of complexity of damage model.

[Ca11 Me Karma]

In my experience mig15's have the same issue as well by having to much "armor" whenever I go against them in an F86. Never have a problem with lining up a shot, its always the countless bursts I have to put into them that becomes tedious.

[ZaltysZ]

The f15 looks to go down easier but does it have self sealing fuel tanks ?

 

http://books.google.lt/books?id=flXlAwAAQBAJ&lpg=SA1-PA13&ots=c0PuVxaLii&dq=f15%20self%20sealing%20fuel%20tanks&pg=SA1-PA13#v=onepage&q=f15%20self%20sealing%20fuel%20tanks&f=false

 

Look under "survivability".

[Pilum]

Here are the tracks as requested sith. Human piloted 190s.

 

Good that you took the time to do this! :thumbup:

 

For me it was an eyeopener to see the difference between AI and player damage model. Based on the AI statistics where in many cases 100+ hits are needed to bring down the AI and it almost always either fails catastrophically or continues to fly forever as a sieve, the player model seems impressive.

 

I have two observations after looking at the tracks: The one being that it still seems to take 100-130 rounds to make the Dora fail catastrophically in most tracks. The other being that in the few tracks where you manouver away and it took longer, the kill hit count went down to 40 and 50 in a couple of cases.

 

Now my thinking is that the Dora may already be dead after a few tens of hits in many of the tracks but because the fire keeps pouring in in most tracks we will never know.

 

So a better test of what it takes to kill a Dora with 50-cal would IMHO be to give it a few short bursts, not only from 6 O'clock but also in deflection angles. To me it looked like the engine was registering hits and the engine did quit after some time which would be expected in most cases as well I think since it's probably that in many cases it's the engine auxillaries that die and this takes some time (like in track 20).

 

If this would bring us down to the previously estimated 20 rounds for a kill I don't know and that may be a stretch but the way the rounds keep coming in in many tracks kind of clouds the issue IMHO because the engine may be dead due to leaking oil or coolant and before that shows will take time and hitting with 100 + rounds in a short time may hide this.

 

One final observation: One thing that did not come up (except the obvious empennage failures in the externals) was text messages in the debrief stating control failure due to control wire or linkage shot away which I would at least have expected in some tracks. Is this there or is it missing?

Edited August 25, 2014 by Pilum

[MiloMorai]

Why not set up some tests where one pilot flies straight (no maneuvering) and another pilot pumps rounds into the other a/c from different angles of attack.

 

The target pilot can comment on what is happening to his a/c.

 

Then switch roles.

[Pilum]

Why not set up some tests where one pilot flies straight (no maneuvering) and another pilot pumps rounds into the other a/c from different angles of attack.

 

The target pilot can comment on what is happening to his a/c.

 

Then switch roles.

 

Yep, and be sure to only shoot a few rounds and wait to see the effects before adding more.

 

One thing that would be really interesting would be to to shoot a few rounds in the P-51 radiator on full open from behind and wait and see. Will the engine seize after a period of time or won't it?

 

Maybe a job for the cowling Mg131 mounted close to the sight line in a close formation flight :smilewink:

Edited August 25, 2014 by Pilum

[USARStarkey]

Id love to do more testing, but I need someone to fly with me! My previous tester got bored. PM me if you are down.

 

 

Also note to pilum: Notice how in some tracks it just ends before the plane dies? That is because I quit whenever i noticed the plane was done. If anything in the airplane critical to flying died, I quite. In no case is the plane dead and just soaking up more hits.

[Pilum]

Also note to pilum: Notice how in some tracks it just ends before the plane dies? That is because I quit whenever i noticed the plane was done. If anything in the airplane critical to flying died, I quite. In no case is the plane dead and just soaking up more hits.

 

OK, that's good to know but again, maybe the plane was clinically dead before that? API through engine casing causing major oil leak, radiator and/or coolant circuit out of action? All these thing would kill the plane but it would take time. If you pour in 130 hits in a very short time it's difficult to know exactly how this is modeled and when you killed the plane IMHO.

 

Another thing that is strange is how few times the pilot got killed in the tracks you posted: I would expect it to be pretty unhealthy to sit like that and take 130 rounds API from dead astern since the API core from a 50-cal should have no trouble punching through back- and headrest armour.

 

This is of course unless the DCS model is so advanced that it models the rounds starting to tumble due to penetrating fuselage and hitting radio and other equipment before reaching the pilot's armour :P

[USARStarkey]

I dont know why the pilot didnt die, not something I cant control.

 

As to the engine deaths, this is hard to determine. You can tell the engine is going under certain conditions and not in others. Also, it isnt necessarily a death. Depending on how dead, you might limp back to base. When discussing the destruction of the plane, I mean practical combat damage, not effects that happen 20min later. You might "kill" a plane by making him over-stress and engine without firing a shot, and he dies 40min later on his way back to base. In the tracks I'll put up soon, a average number of hits seemed to be 80ish. Still way too high. Weirdly, every kill seems to be engine related. I daresay none of this matters however. I could post 100 tracks and ED wont change a thing.

Edited August 28, 2014 by USARStarkey

[Pilum]

I dont know why the pilot didnt die, not something I cant control.

 

As to the engine deaths, this is hard to determine. You can tell the engine is going under certain conditions and not in others. Also, it isnt necessarily a death. Depending on how dead, you might limp back to base. When discussing the destruction of the plane, I mean practical combat damage, not effects that happen 20min later. You might "kill" a plane by making him over-stress and engine without firing a shot, and he dies 40min later on his way back to base. In the tracks I'll put up soon, a average number of hits seemed to be 80ish. Still way too high. Weirdly, every kill seems to be engine related. I daresay none of this matters however. I could post 100 tracks and ED wont change a thing.

 

When you say 20 to 40 min thats sounds higher than what I had in mind and will it really take that long in DCS? I'm not an engine expert but I would expect an engine to die in a minute or two if you lose oil pressure. Maybe a few minutes more if it was the coolant that got shot. A nice (and realistic) DM would IMHO be pilots bailing out of stricken aircraft a lot earlier than the sieves the AI coast around in today.

 

Speaking of changing things, it would be nice if it became optional which DM model was used offline: either the one we have today or the plaver DM also for AI.

[USARStarkey]

I didnt mean engine failure for that specific cause. For oil pressure sure, but I meant for other reasons. Like been hit in fights and been going back to base and then 15min later the engine just died. Ive also had runaway props that went for extremely long times. I flew the length of the dog of war server with a runaway prop just last week.

[Sgt_Fresh]

Can't Shoot the Dora Down

 

I scored 88 .50 cal gun hits during a 1 v 1 against a ai D9 it leaked fuel and dark smoke, I assumed it to be an oil leak. The ai flew a constant vertical fight and I could not keep altitude with the Dora. I figured the fuel and oil leak would reduce performance but the plane got away and eventually landed.

 

Tried it again with similar results. Is the game Dora invulnerable or is there a game setting I need to adjust?

[NineLine]

I scored 88 .50 cal gun hits during a 1 v 1 against a ai D9 it leaked fuel and dark smoke, I assumed it to be an oil leak. The ai flew a constant vertical fight and I could not keep altitude with the Dora. I figured the fuel and oil leak would reduce performance but the plane got away and eventually landed.

 

Tried it again with similar results. Is the game Dora invulnerable or is there a game setting I need to adjust?

 

Visual damage isnt linked to the actual damage unfortunately... visual DM needs a revamp of sorts.

[Narushima]

I scored 88 .50 cal gun hits during a 1 v 1 against a ai D9 it leaked fuel and dark smoke, I assumed it to be an oil leak. The ai flew a constant vertical fight and I could not keep altitude with the Dora. I figured the fuel and oil leak would reduce performance but the plane got away and eventually landed.

 

Tried it again with similar results. Is the game Dora invulnerable or is there a game setting I need to adjust?

 

You can make a trigger that makes the AI plane explode after a certain damage threshold.

 

Makes it much more realistic. Check it out in the mission editor.

[NineLine]

Yes cuz most planes exploded after a certain amount of damage every time :P Much more realistic :)

[Narushima]

Yes, it is much more realistic. Now that tells you a thing or two about the AI damage model.

 

And btw, most of the time the AI goes down in a big fireball anyway, even without the trigger.

[NineLine]

Yes, it is much more realistic. Now that tells you a thing or two about the AI damage model.

 

And btw, most of the time the AI goes down in a big fireball anyway, even without the trigger.

 

Hmmm, I do fine without any trigger handicaps... dunno...

[Narushima]

Define fine. Does fine mean "historically correct" or does fine mean "good enough for me"?

 

Also, I have a strong suspicion that the damage model is somehow linked to the skill of the AI pilot. I've noticed that when I put it on the easiest setting they tend to go down faster. Could just be placebo though, and I have no way of actually testing this, so without a dev confirmation it's just a theory.

 

EDIT: Could also be that damage is applied differently depending on hardware, maybe something like per frame calculations. Would explain some things.