4.12.2 de-bugging

[Pursuivant]

I finally got a chance to abuse the P-38, specifically the P-38H.

Thanks to TD's recent rework of this plane, it DM actually isn't that bad, although it is still far too vulnerable to control and machine gun hits.

There are no gaps between the armor glass and the armor plate - if a .303 bullet hits the armor glass, it gets stopped. If it hits armor plate, it gets stopped.

Fuel tank hits by single rifle caliber bullets actually seal after 30-60 seconds. Hits by multiple rifle caliber bullets in the same place will quickly cause a fire, but that's not unreasonable if one of the bullets is explosive or incendiary.

The engine damage model looks good - a bullet that hits the oil cooler or radiator will cause a leak, bullets that come close but miss don't. Once a leak starts, you get several minutes of flying time before overheat, and about 5 minutes of flying time before the engine starts to really suffer. You lose some power due to engine hits, but not much. It seems like it takes a LOT of damage to make an engine stop cold.

If anything, engine durability is a bit too generous; more along the lines of the P-39 than the P-40. But, at least you can lose power and actually damage the engine due to overheating, unlike the P-39.


But, here are the problems.

Control hits - like many planes in the game they are too easy to achieve, and can occur even when bullets don't hit anyplace close to control runs.

The only way I can explain it is that the game doesn't take the presence of full fuel tanks into account when determining bullet trajectories. That is, it counts fuel tanks as "empty space" rather than being filled with liquid which will slow or stop a bullet.

Likewise, the game doesn't take into account that there are layers of aluminum between a hit to the leading edge of the wing and the trailing edge, which might slow a small caliber bullet or a bit of shrapnel, or make it shatter.

Machine Gun Hits - At face value, it seems like good damage modeling to have any hit to the P-38's guns cause a jam. There's a lot of guns and ammo packed into the P-38's nose, and no armor plate in front of them or to the side, so any bullet hitting that area is likely to cause some sort of problem.

But, most of that space is taken up with ammunition. That means that there's a fair chance you'll get just one or two rounds of ammunition that are damaged by the passing bullet. This might cause an immediate jam the bullet deforms the feed chutes or ammo containers sufficiently that the bullets can't advance, but it's more likely to just create a dud bullet which will cause a jam at some time in the future.

So, for just about any hit to ammo runs by shrapnel and small caliber bullets, there should be the possibility of a) immediate stoppage/dead gun, b) unfixable stoppage after X amount of the gun's remaining ammo is fired, c) fixable stoppage after X amount of gun's remaining ammo is fired.

Hits by big bullets, explosions, and large pieces of shrapnel should create a very high chance of immediate unfixable stoppage, but with a tiny chance of the other two possibilities.

Edit: There's also two more options:

d) No effective damage. That is, the bullet hits some part of the bullet feed mechanisms or magazines which aren't currently occupied by ammunition, or otherwise manages to pass through the vital area without doing any serious damage. For example, if you shoot 50% of your ammo, and then you get a machine gun hit, the game currently treats it identically to a hit on the ammunition before you fired a shot, even though there's a 50% chance that the bullet actually just hit empty space.

e) Bang! Some sort of ammo explosion. This is extremely unlikely, even for cannon shells, but it could happen. To complicate fuel fires for tanks close to ammunition magazines, there's also the possibility of a fire causing bullets or shells to "cook off" - exploding due to heat.

A final problem which appears to be unique to the P-38 is that you can get machine gun hits for guns which have already been knocked out. That is, if both your starboard MG get hit, and you get a third bullet into a starboard machine gun, the DM counts it as a hit to a port-side machine gun!

Supercharger Hits - IL2 doesn't model supercharger damage. The P-38's turbocharger is mounted on top of the engine, with a very vulnerable flywheel protected by just an aluminum deflector and a tiny bit of armor plate on the inner side of the engines - mostly designed to keep flywheel blades from hitting the pilot. I got a few hits which realistically would have really wrecked the supercharger.

Engine Control Hits: The P-38, like other multi-engined planes had cable runs to the engine, which control things like radiator louvers, prop pitch and throttle. They're right behind the reserve fuel tanks in the inboard wing and are ripe targets for damage to the wing. But, IL2 doesn't model such things.

Here are pictures of the carnage:

http://forum.1cpublishing.eu/attachm...1&d=1404087241

The obligatory picture of long-range sniper shots by the Ace Wellington III gunners. Two hits, two machine guns down. Typically, flying the P-38 as an interceptor results in your losing half your guns in the first few attack runs!

http://forum.1cpublishing.eu/attachm...1&d=1404087241

Same flight a few seconds later. Notice a bit of engine damage to the port engine, but also another couple of bullets in the nose that took out another machine gun. The problem is, none of those bullets could have taken out the remaining two starboard machine guns, or their ammo reservoirs!

So, still 300 m out on my first attack run and I've lost 3/4 of my machine guns!

http://forum.1cpublishing.eu/attachm...1&d=1404087241

Same attack run, closer in. Crummy screenshot due to a pilot kill. Unless you go straight in at an enemy bomber the P-38 is a flying coffin because the gunners not unreasonably aim directly at the plane's center of mass; which is the cockpit.

Mind you, I'm not complaining about the PK, which I think was fair. I chose this one because I think it illustrates how easy it is to get control hits. None of the bullets came anyplace close to hitting the control runs. (Damage to the nose would have had to penetrate 1/4 to 3/8 inch - 6-10 mm - of armor plate to hit the rudder pedals or rudder pedals in the cockpit.) Furthermore, the P-38 has twin rudders, and virtually all of the damage is to plane's port side.

Realistically, assuming that just one cable run was severed, you'd get just the port side rudder fluttering or jammed, but the starboard rudder might still work. It would have to be remarkable luck or skill for less than a dozen .303 bullets to completely disable the plane's rudder control!

[Tempest123]

"Control hits - like many planes in the game they are too easy to achieve, and can occur even when bullets don't hit anyplace close to control runs."

Totally agree about this on more than a few planes, single bullets taking out ailerons and rudders etc. Hitting exactly the cable or rod would be possible but rare, probably more common to have Flak damage the sheet metal and jam the control surface.

[Pursuivant]

Originally Posted by Tempest123

Totally agree about this on more than a few planes, single bullets taking out ailerons and rudders etc. Hitting exactly the cable or rod would be possible but rare, probably more common to have Flak damage the sheet metal and jam the control surface.

Furthermore, there are exactly ZERO places on your average WW2 airplane where a single bullet - even a big one - can simultaneously wipe out rudder, ailerons and elevators, yet it happens occasionally in IL2.

You also bring up another control damage option I forgot to mention - limited range of motion for damaged control surfaces.

Currently, you have loss of control surfaces, complete loss of control of control surfaces, and partial loss of authority from control surfaces. But, you don't have limited range of motion from damaged controls, such as a rudder that you can only turn right, or which only travels through 50% of its full arc when turned to the right.

[western0221]

4.12 made a new great function, multi-turret tanks.
I'm glad in this implement.

Today's bug report is T-35 tank's weapon parameters.
T-35 has 5x turrets and in 4.12 they all work.
But thier parameters are not correct.
T-35's armaments were
- 1x 75mm cannon
- 2x 45mm cannon
- 2x 7.62mm machine-gun
in historical.
In 4.12.2 , all 5x turrets are set as 75mm cannon.
Please remap armaments.

[Pursuivant]

Troubleshooting the A-20G Damage Model:

This one has a lot of little problems, but the big problem is FUEL TANKS

A) The damage model assumes that the fuel tanks are contiguous with the plane's skin, which was not the case. That means that many glancing hits to the wing and fuselage which realistically would have missed the tanks cause punctures.

B) In a few cases, bullets which completely miss the tanks still cause fuel leaks. Like here:

http://forum.1cpublishing.eu/attachm...1&d=1404378042

Note that damage which is just to the starboard engine nachelle has somehow managed to start a leak in the outboard fuel tank!

C) As is typical for just about all the planes in the game, single rifle caliber bullets cause much worse leaks than they should and ignite fires far too often.

Remember, by about 1940, at least for most planes, self-sealing tanks weren't just self-sealing, but usually blanketed by engine exhaust.

That means that you have to tear open the fuel tank to the point that the CO2 blanket no longer is effective, you also need to have way of vaporizing the gasoline (admittedly, not hard when there's a 200 mph wind blowing through a hole in the plane's skin), and you need a spark. Sparks are a bit harder to come by since aluminum doesn't spark and only something like every 10th bullet in the belt was Incendiary, tracer or explosive.

So, two .303 bullets in rapid succession against a full fuel tank are very unlikely to cause a fire because the gas hasn't had time to spill or vaporize and the integrity of the tank is still good. A burst of .303 bullets might do the trick if it tears apart the fuel tank and the fuel then contacts a hot engines. But, that's not an option for a fuel tank mounted at a distance from the engine.

What might cause a fire is several bursts of .303 fire which tear up the tank, followed some seconds later by another big burst of .303 gunfire, which can be assumed to contain an Incendiary, Explosive or Tracer bullet in the mix, and which hits gasoline vapor.

Example of bad damage modeling here, where two .303 bullets from the same burst started a fire in a previously undamaged fuel tank.

http://forum.1cpublishing.eu/attachm...1&d=1404378818

The other bullets which hit near the burning fuel tank actually hit behind the fuel tank, so the damage looks worse than it is.

E) Fuel fires don't go out when fuel is exhausted.

Example here:

http://forum.1cpublishing.eu/attachm...1&d=1404378916

This screenshot was taken about a minute after I lost fuel due to the fire. Since the fire presumably drew fire from other tanks, it's unlikely that there was any remaining fuel in the burning tank for it to use!


F) Fuel fires also deplete the fuel from ALL tanks, not just the tank where the fire is. I know that IL2 doesn't and can't really model fuel transfer or fuel shutoffs, but it wouldn't be hard to specify that a fuel fire just takes some percentage of fuel and then goes out.

[Pursuivant]

More A-20G DM problems.

1) Pilot injury.

A) Forward Armor Plate Not Properly Modeled: In a few cases, bullets will penetrate the forward armor plate to kill the pilot - just as if the plate wasn't there. This despite the fact that most marks of A-20 had armor plate ahead of the pilot:

http://www.zenoswarbirdvideos.com/Im...otArmorDia.pdf

Unfortunately, no pictures of this one. I got three PK in quick succession and several pilot wounded results from frontal attacks which penetrated the forward armor, but didn't think to take screenshots of them. After that, I couldn't replicate the problem.

B) The armor glass in front of the pilot is mostly well modeled, with a few exceptions:

http://forum.1cpublishing.eu/attachm...1&d=1404379563

Note the hit to the edge of the armor glass, which, rather than being stopped or deflected penetrated to hit the pilot's arm. This isn't just an artifact of the cockpit model not quite matching up to the exterior model, the hit looked just as strange from the outside.

Typically, however, the armor glass does its job except at close ranges were it can be penetrated.

C) Pilot damage modeling is a bit weird. In one instance I got a bleeding wound from a bullet which just missed the pilot. At best, it would have been a grazing wound.

http://forum.1cpublishing.eu/attachm...1&d=1404380010


2) Gun damage. Unlike planes like the P-38 where just about any hit to the nose is a gun hit, it's quite difficult to hit the guns on the A-20G. While they can happen (unlike, say, the P-39 or P-400), I have to wonder if the ammunition runs for the guns could be a bit better modeled.

http://forum.1cpublishing.eu/attachm...1&d=1404380214

Notice loads of hits which might damage the guns or ammo runs, including one hit that goes straight down the barrel of one of the guns, but no damage.

3) Missing Crewman. This isn't bad DM, it's just a bug. The rear gun pivots, but no gunner is visible. He's either not modeled, or he is hiding someplace inside the plane rather than manning his gun. He is visible when it's time to bail out, though!

Finally, this technical report provides lots of good cutaway drawings of various A-20 engine, fuel and control systems:

http://legendsintheirowntime.com/A20/A20_draft_2.pdf

[Treetop64]

Just a side note on this issue regarding the SB-2:

The SB-2 had four fuel tanks, one on both sides of each engine. They were not protected with either armor or sealant. So, as it should be, taking a penetrating hit to a fuel tank will cause a leak until the tank runs dry.

The problem is that there is no way of isolating the leak to the affected tank. For example, if the no.2 tank got penetrated, the resultant leak will continue to run until all four tanks are empty. The aircraft's entire fuel supply from the remaining untouched tanks will drain out even if only one tank was damaged.

Though lovely and an enjoyable aircraft to fly, the SB-2 is already delicate enough without this crippling bug.

[TexasJG]

Version 4.12m, the “Scan MAX Range” function of FOW seems to be inoperative.

[Aviar]

Originally Posted by TexasJG

Version 4.12m, the “Scan MAX Range” function of FOW seems to be inoperative.

Just for your information, DT normally needs more proof/information regarding possible bugs.

However, if you post an example mission, I would be happy to test/confirm this issue for you.



EDIT: I made a quick test mission and it seems to be working fine in both Coop and Dogfight modes. Have you read the MDS Guide which came with the 4.10 patch?

Aviar

[Tempest123]

Originally Posted by Pursuivant

Furthermore, there are exactly ZERO places on your average WW2 airplane where a single bullet - even a big one - can simultaneously wipe out rudder, ailerons and elevators, yet it happens occasionally in IL2.

You also bring up another control damage option I forgot to mention - limited range of motion for damaged control surfaces.

Currently, you have loss of control surfaces, complete loss of control of control surfaces, and partial loss of authority from control surfaces. But, you don't have limited range of motion from damaged controls, such as a rudder that you can only turn right, or which only travels through 50% of its full arc when turned to the right.

Yup, I would like to see the instant disabling effect (which is a little arcade IMHO unless the whole surface is blown off) replaced with degrees of limited movement, i.e 20% to 70% (or whatever) of total movement through the full range of motion to simulate damaged surfaces.