4.12.2 de-bugging

[Janosch]

For all A6M5b Zero and later types, when you pick IJN or IJA as the airforce, having markings ON draws some extra clutter on the left side of fuselage. Right side is normal, they don't appear there. I have highlighted the "smudges":

They seem to be some number graphics that have gotten "lost", as they're the same colour as the fuselage number. I'm currently using vanilla 4.12.2m without any jsgme shenanigans, but I'm not 100% sure if it's my install. But is not fatal bug.

Also, the Luftwaffe Do-335 rudder number (when markings are on, of course) on the left side is drawn a bit too close to the leading edge, so it's drawn only partially.

[Baddington_VA]

B25 become fighters, and many multi engines bombers seem to have the AI characteristics of twin engined fighters.

Makes the mission a bit strange having B25 attacking enemy fighters low level after they have level bombed a target at high altitude.

I have seen that done by Ju88 A4s back in the original IL2.
I have recently found myself being chased by AI B25s and even AI Ju88A17s in IL21946.
Maybe it's not such a bad thing.

[Notorious M.i.G.]

I didn't see it noted in the 4.13 changelist (although obviously not final), but I thought I'd just drop a quick reminder that the AI refuse to engage at all when armed with X-4s. They won't even maneuver or resort to guns.

Also, the Fulmar can fire its machine guns with the wings folded, but until it gets a pit nobody will probably notice that

[Pursuivant]

Originally Posted by Baddington_VA

I have recently found myself being chased by AI B25s and even AI Ju88A17s in IL21946.
Maybe it's not such a bad thing.

I think that the new AI allows for Average or better attack planes to act like fighters against larger or slower planes, especially when the attacker's aren't carrying ordinance.

For example, I've recently had fun flying a TB-7 against a flight of Ace TBD Devastators! In that case, they dogfight fairly effectively, rather than being the usual hapless targets.

In any case, it's realistic for attack bombers to behave in this fashion as long as the opposition is inferior or evenly matched. There are numerous documented cases where patrol bombers dueled it out - like Ju-88 vs. Sunderland flying boats over the Bay of Biscay.

The only thing that's unrealistic, is that often the attackers will behave as if they were fighters with full aerobatic capability, so you see things like A-20s or B-25s doing loops and pulling high-G turns that would realistically damage the plane and its occupants.

[Woke Up Dead]

Minor bug: the "bomb bay doors open" light in the B25 does not light up when the bomb bay doors are open.

[Pursuivant]

More troubleshooting DM against 0.30 caliber guns from bombers, using the Arcade Mode to spot problems.

This time, it's the Buffalo Mk I vs. a flight of 4 Ace D3A1 "Vals" since the Buffalo is too slow to keep up with the Wellingtons.

As for the B239, the engine is incredibly fragile, with just 2-4 bullets consistently being sufficient to make it instantly lose most of its compression, and 3-5 bullets being sufficient to render the engine inoperable.

http://forum.1cpublishing.eu/attachm...1&d=1410990532

This screenshot shows damage sufficient to reduce engine power by about 25%, even though only two bullets actually hit the engine (the others passed between the cylinders or just hit the cowl).

No screenshot, but the armor glass in front of the pilot doesn't seem to have been modeled. Shots to the cockpit front go right through to kill or wound the pilot. The Buffalo I gave up a lot of performance for the sake of its armor plate, so it would be nice to actually have it.

Damage to the guns appears to be fairly well modeled, as are control cable hits since those hits don't seem to appear very often. Fuel tank leaks are fairly rare and reasonably well modeled - no leaks that result in fast loss of fuel or fires after just a few hits.

On the other hand, the damage model for the D3A1 overstates just how fragile the plane was. I was consistently getting kills (mostly central fuel tank fires) with just a few bullets - albeit possibly 0.50 caliber shots.

More to the point, the amount of damage required to trigger the "light damage" texture for the fuselage doesn't square with the amount of damage inflicted.

http://forum.1cpublishing.eu/attachm...1&d=1410991223

Notice that just 3-4 bullets in the rear fuselage (possibly .50 cal) is enough to trigger the light damage result.

Keep in mind that the D3A was a fully aerobatic dive bomber, so there was nothing inherently weak about the airframe. Lack of armor and fuel tank protection isn't the same thing as ability to take structural damage.

[Woke Up Dead]

Hi Pursuivant, are you aware that those arrows from arcade mode in your screenshots represent only the direction the bullets were traveling when they hit the first solid object, and not their actual path? So if a bullet is traveling towards the pilot's head but is stopped by the armored glass, the arrow will still go through the pilot's head.

So for example you can have a screenshot of the pilot looking like Hellraiser with all those arrows in his face, but in game all those bullets were actually stopped by the armored glass.

Also, 2-5 bullets to reduce power or completely kill an engine does not sound unrealistic. Neither do kills from "a few" bullets to unprotected fuel tanks.

Finally, the only effect of light damage on the fuselage is a small aerodynamic penalty. Again, it does not seem like an unrealistic result of taking eight or nine .50 caliber bullets like the Val in your screenshot.

[Pursuivant]

Originally Posted by Woke Up Dead

Hi Pursuivant, are you aware that those arrows from arcade mode in your screenshots represent only the direction the bullets were traveling when they hit the first solid object, and not their actual path? So if a bullet is traveling towards the pilot's head but is stopped by the armored glass, the arrow will still go through the pilot's head.

Yes. Using the arcade mode bullets is just the first step. Any screenshot that looks odd, or which results in odd damage gets compared against a 3d view of the appropriate aircraft.

So, there are plenty of cases where I've seen "arcade arrows" skewer my pilot with no ill effects because armor glass and armor are doing their job at stopping the actual bullet.

Likewise, there are plenty of screenshots I've taken where the damage subsequently makes absolute sense, such as when a bullet goes right down the barrel of a gun and IL2 records it as a gun hit.

Where it's appropriate, I call out good DM work as well as bad.

Originally Posted by Woke Up Dead

Also, 2-5 bullets to reduce power or completely kill an engine does not sound unrealistic. Neither do kills from "a few" bullets to unprotected fuel tanks.

Radial engines were noted for being able to take damage and still keep running, so it seems odd that the Buffalo series has engines that are as fragile as those of an inline engine plane such as a Spitfire or Bf-109.

Here's the problem with IL2 and engine hits.

First, there doesn't seem to be any consistency as to how damage to a particular engine is modeled. For example, damage to a Wright-Cyclone R-1820 engine mounted on a C-47 might make it behave differently than the same damage to the same engine mounted on a P-36, SBD, FM-2 or B-239.

My guess is that the creators of the B-239 DM assumed that the R-1820 wasn't a durable engine, since the Finns reported trouble with oil leaks and other problems with the used engines supplied with their planes. Since the engine didn't change for the F2A or Buffalo Mk I, whoever did the DM for those models just used the engine DM for the B-239.

The problem is that there's plenty of evidence that, apart from overheating problems, the R-1820 was a very good engine.

Second, many DM seem to just model the engine as a homogenous block, and fail to distinguish between empty space between cylinders in a radial engine, hits to the crankcase (potentially quite devastating), hits to cylinder cooling vanes (trivial damage) and hits to cylinders (damaging, but not immediately lethal).

While realistic engine damage models are probably beyond IL2's limits, the model could be tweaked a bit so that bullets that don't actually hit the engine don't damage it, and so that bullets which hit near the same place on the engine don't do any extra damage. After all, you can only destroy the same cylinder once!

Third, the bullets that are consistently killing the Wright Cyclone R-1820 engine on the Buffalo series are 0.30 caliber bullets being fired from anywhere from 50-300 meters distance, and they cause near instant engine-stoppage or serious power loss, regardless of where they hit. We're talking about bullets that make small holes and which might not have much power on them when they hit.


The problem with fuel tank hits in IL2 is that I don't think that the DM takes into account fire suppression measures, all bullets are treated as being incendiary, and self-sealing tanks aren't always well modeled. (There are a few planes where the self-sealing tanks actually work, though.)

Realistically, perhaps 1 in 5 or 10 bullets is going to be tracer, incendiary, explosive, API, or similar. The bulk of the bullets are going going to be plain ball ammo. That means that you basically only have a 1 in 5 or 1 in 10 chance of getting a hit with a bullet that has a chance to start a fire.

Next, the first bullet to go through a container of gasoline isn't likely to start a fire, since it's going through liquid (or possibly through a blanket of cold exhaust gasses protecting the empty space in the tank). That bullet is likely to create a spray of gasoline vapor, which might ignite if the bullet is explosive or incendiary, but a ball bullet on its own is just going to set things up for a subsequent bullet (or a spark, or heat from an engine) to start a fire.

Multiple ball bullet hits are most likely to further shred the fuel tank and splash the fuel around, rather than starting fires.

Third, fires start instantly and automatically appear at full size. In most cases, this is just cosmetic since IL2 does a really bad job of modeling fire damage to airframes, but fire size makes a difference when determining damage to the pilot and risk of explosion.

Realistically, what might happen is that a bullet hit to a fuel tank splashes fuel around and creates pools of uncontained gasoline that get vaporized by contact with the wind blowing through the bullet holes in the airframe.

A second bullet with explosive or incendiary qualities hits and ignites the vapor. The fire spreads more or less quickly to involve all the vapor (possibly creating an explosion if there's a lot of oxygenated vapor in a small place), then starts volatilizing and burning the remaining liquid. Fuel in the tank won't have that much oxygen to burn it, but spilled fuel is likely to burn quickly.

So, you'll get small fire to start with (sometimes well modeled by black smoke in the game) that burns the spilled fuel, followed by a big "fully involved" fire that starts to volatilize fuel stored in the tank.

Originally Posted by Woke Up Dead

Finally, the only effect of light damage on the fuselage is a small aerodynamic penalty. Again, it does not seem like an unrealistic result of taking eight or nine .50 caliber bullets like the Val in your screenshot.

True. Because of the way that IL2 "paints" light and heavy damage textures you have to ignore the appearance of the damage textures vs. their actual effect.

But, the hole made by a 0.50 caliber bullet in aluminum is going to be the thickness of a man's thumb. Exit holes are perhaps going to be a bit bigger, as might holes made if the bullet enters at an oblique angle or tumbles after impact (although bullet tumbling or fragmentation is unusual for the 0.50 BMG round).

Like you said, that's going to cause drag. My point, however, is that 8-9 bullets shouldn't be enough to cause much more than drag, like making the airframe fail under stress.

Since damage modeling is an art, it seems to me that IL2's developers have made planes that were notably vulnerable in combat for any reason excessively vulnerable to any sort of damage.

That means that planes like the TBD and D3A1, which were mostly vulnerable because they were slow and didn't have good armor and fuel protection, are potentially too vulnerable to airframe damage.