4.13 development update discussion and feedback

[KG26_Alpha]

Originally Posted by Tolwyn

Geforce 6600.
OpenGL 1.5

Originally Posted by Iszak

And what card is that? 3dFX VooDoo?

Originally Posted by Iszak

But Geforce 6600 IS OpenGL 2.0/2.1 compatible, so what's the problem? O_o

I'm gonna call him Trollwyn on DD TS next time Im on there lol

Originally Posted by IceFire

Actually it's all OpenGL... the DirectX mode is a wrapper only. It'd be interesting to see development of a more modern DirectX mode. It'd open up the door to all kinds of possibilities but its no small task and it would be years in the making.

Another game had a similar kind of experience. FreeSpace 2 was a space combat game that came out in the late 90s. The source was opened by the developers and the community has been busy working on it ever since. The visual effects have come a long way but we're talking a decade worth of open source development done by a pretty hardcore team. I suspect that in many ways that engine is a lot simpler than what IL-2 hooks into.

Yea I know fella I'm just going through the motions for a certain member here












.

[DD_crash]

I am sure that he would appreciate that Alpha Tolwyn has threatened to upgrade his puter when he can get someone to do it.

[Pursuivant]

I fly at lot of Quick Combat bomber intercept missions. One single-engined fighter vs. big formations of massed bombers.

That means I inevitably take hits to the front of the plane, usually clustered around the engine and cockpit, even when I use proper tactics.

To see the effects of my gunnery I usually fly in arcade mode. That means I see exactly where the bullets hit - both mine and theirs.

While I've complained about specific fighters being quite vulnerable to attacks from dead ahead before, I now think that it's a generic problem, from Bf-109 to Yak.

1) Frontal armor doesn't seem to be modeled at all for single-engined fighters. That is, armored propeller mounts, armor in front of oil or fuel tanks mounted ahead of the pilot in the nose, armored firewalls between the engine compartment and the pilot, and armor glass to the pilot's front don't seem to slow down bullets at all, even against rifle caliber bullets at several hundred meters of range.

Additionally, the "armoring" effects of bullet penetration through liquid aren't modeled at all.

For example, I regularly get pilot wounded results from bullets which pass through the fuel or oil tank bulkhead/armor, through the oil/fuel tank itself (about 20-30 cm of liquid for an oil tank) and then through the forward firewall armor!

While this is realistic for 20 mm or 0.50 caliber bullets, it seems impossible that a 0.30 caliber bullet could do the same.

PK results usually occur when a bullet goes directly through the armor glass. Again, realistic for a .50 caliber/12.7 mm bullet or a cannon shell, but not so much for a 7.62/.30/.303 bullet, especially at much more than 50 meters range.

2) There seem to be gaps in frontal armor. That is, all the plane models have a slight gap between where the armor plate for the forward firewall ends and the armor glass begins. Often, I get PK results from bullets which pass through the bottom of the cockpit frame between the armor plate and armor glass.

If that's realistic, congratulations to the modelers. I suspect, however, that there would not have been gaps, since engineers and mechanics would have closed them or engineered overlaps between armor plate and armor glass.

3) Engines, especially inline engines, seem to be extremely vulnerable to just about any damage. It doesn't matter what plane you fly, if it's got an inline engine, any hit from dead ahead through the prop boss will usually smoke it.

While this sort of damage is realistic for hits to oil and coolant radiators, and for hits from 0.50 caliber or larger bullets, it seems a bit unrealistic for shrapnel hits at anything other than point-blank range, and for rifle-caliber bullet hits to engines at ranges beyond about 100 meters.

First, the game doesn't appear to model all the parts between the propeller spinner and the engine block (perhaps another 25-50 mm of mixed aluminum, mild steel and tool steel).

Second, it doesn't appear that the game models angle of penetration, chance that the bullet will ricochet or fragment, or the inherent toughness of the engine block itself (perhaps 25-50 mm of cast iron, with about 1/10 the penetration resistance of homogenous rolled armor plate).

Since I'm not sure that the game can model angle of armor penetration, and I know it can't model the exact thickness of each engine block, it seems more reasonable to apply some sort simple penetration reduction modifier based on "angle off" from the gun to the target, which isn't applied at 0 or 90 degrees, and is maximized at 45 degrees. Additionally, there should be some sort of randomized penetration reduction (perhaps 1-20%) for any engine hit to represent chance of fragmentation, ricochet, hits to non-vital parts and the inherent toughness of the engine block.

4) It seems far too easy to blow big pieces off of bombers. For example, my target of choice these days is the Wellington, and despite its notably strong geodesic construction, the damage modeling allows me to blow the entire nose, wing or tail off using just a few 20 mm cannon shells!

The same problem applies to other notably tough aircraft such as the Ju-88, B-29 or B-17.

While I know that parts breaking off is supposed to represent the sort of catastrophic damage that the game can't properly model, it still seems unrealistic for bomber to be torn apart by anything other than a collision with another bomber, severe fire or massive explosion. I find it unrealistic that I can blow the wings off a B-17 or a Wellington with just a few 20 cannon hits.

Would it be possible to model lethal damage to bombers without modeling breaking parts? For example, would it be possible to set the threshold at which the crew bails out of AI planes and the plane becomes unflyable short of the level at which the wings come off?

[ElAurens]

Every inline engine I know of has an aluminum engine block, and I rather think that they are not 1 to 2 inches (25 to 50mm) thick.

However, I do agree that the inline engines are too fragile.

Fly the P40 for a while. I'm sure that the tail gunners could simply use
7.65 Browning Walther PPKs and bring them down with one shot.

[Janosch]

^Would you guys, like, have some raw data, like, you know, logfiles, tracks or screenshots to pass around, so others could verify your claims on the (perceived) shortcomings of damage modeling?

[Pursuivant]

Originally Posted by ElAurens

Every inline engine I know of has an aluminum engine block, and I rather think that they are not 1 to 2 inches (25 to 50mm) thick.

I believe that all WW2-era high performance aircraft engines had cast iron blocks and that cast aluminum blocks are mostly post-war. Remember, aluminum was a somewhat scarce resource for many nations during WW2, whereas mild steel was relatively common.

I admit to eyeballing the exact thickness of the engine block using cutaways of Merlin and Daimler-Benze engines. But, remember, inline aircraft engines are big (~3/4 metric tonne, nearly 1 English ton, vs. 200 kilos/300 lb. for auto engines) and are designed to deal with much greater forces than auto engines.

I'm not saying that aircraft engine blocks are bullet-proof by any means, since cast iron is relatively soft and brittle as steel goes. But they're going to be a bit tougher to penetrate than auto engines, which is mostly what you see being shot up by various guns on YouTube.

There are also four different damage states the game engine needs to model for damage effects to inline engines - no functional damage (i.e., pitting of the engine's exterior, but no penetration), penetration of the block around the gearbox (= oil leak and eventual failure or seizure of the engine due to gearbox overheating), penetration of the block around the cylinders (= coolant leak and eventual seizure of the engine due to overheating of cylinders) and the penetration of both the exterior block and one of the cylinders which results in loss of engine compression and a fuel leak in addition to the effects of a coolant leak).

As a variation on penetrating the cylinder, there's also the possibility of damaging one of the pistons, cylinder head or camshafts, or one of the spark plugs or part of the wiring harnes, which would reduce engine compression and possibly cause overheating or engine seizure without the fuel or coolant leak.

Since engines in IL2 are modeled as solid blocks of metal, it seems that the simplest way to model the different types of hits would would just be to assign random percentages of no functional damage, coolant leak, oil leak, fuel leak and compression loss/engine seizure based on bullet energy, with .50 caliber or better bullets having a chance of multiple different hits, but with .30 caliber bullets just getting one type of hit (and with fuel leak/loss of compression/engine seizure hits being very rare).

Given the fluid pressures and temperatures involved, I think that IL2 is realistic, if not a bit generous, in allowing damaged inline engine to survive as long as they do. But, since I'm ignorant about these things I'll defer to others with more experience.

[Pursuivant]

Originally Posted by Janosch

^Would you guys, like, have some raw data, like, you know, logfiles, tracks or screenshots to pass around, so others could verify your claims on the (perceived) shortcomings of damage modeling?

True, if you don't have pictures it doesn't exist!

I will try to upload pictures.

For others who wish to add pictorial evidence, set up a 16 bomber vs. 1 fighter QMB mission and turn arcade mode on.

To prove my point, be sure to use planes with weak rear defensive armament, like the Ju-88A or He-111H, or with massed rifle caliber MG in turrets, like the Wellington.

Use Ace level AI and stupid tactics like hanging out at 100-300 m right behind a bomber formation.

You'll get the results I described soon enough; loads of PK, pilot wounded and badly smoked/seized engines, usually following just a few bullet hits. As a bonus, you'll occasionally get a control cable hit (esp. for the older planes like the Bf-109 or the Soviet fighters) from hits that penetrate the engine and forward firewall.

[Pursuivant]

Pictures

#1: Yak 1B - Pilot arm wound through gap between forward armor plate/engine firewall and cockpit combing at medium-long range.

http://forum.1cpublishing.eu/attachm...1&d=1401765054

#2: P-40B just after getting "shot-shotted" at extreme range through the propeller boss by Ace He-111H-2 gunners. Note the remarkable accuracy since the nearest enemy plane is over 700 meters away! (So much for fixing "sniper" AI gunners. .) I got the "engine overheat" message within seconds after the hit, and the engine was whining badly indicating that engine seizure was going to happen just a few seconds later. This problem seems to be unique to the P-40 series.

http://forum.1cpublishing.eu/attachm...1&d=1401765076

#2 P-40B just after getting shot at medium long range by Ace He-111H-2 gunners at a slight angle off. Engine smoking badly due to oil leak with only a few minutes of life left. The bomber that inflicted the hit was over 300 meters away, just outside of the frame in the upper right hand corner. Note the remarkable grouping of two bullets within 1 foot of each other against a maneuvering target!

http://forum.1cpublishing.eu/attachm...1&d=1401765088

Happy yet?

And, mind you, I didn't have to work for these results at all. They represent 3 consecutive missions!

[ElAurens]

Pursivant, you had better do some more research,

The Merlin had an aluminum engine block with steel cylinder liners.

From Wiki:

Basic component overview (Merlin 61)

From Jane's:[25]
CylindersTwelve cylinders consisting of high-carbon steel liners set in two, two-piece cylinder blocks of cast "R.R.50" aluminium alloy having separate heads and skirts. Wet liners, ie. coolant in direct contact with external face of liners. Cylinder heads fitted with cast-iron inlet valve guides, phosphor bronze exhaust valve guides, and renewable "Silchrome" steel-alloy valve seats. Two diametrically opposed spark plugs protrude into each combustion chamber.PistonsMachined from "R.R.59" alloy forgings. Fully floating hollow gudgeon pins of hardened nickel-chrome steel. Three compression and one oil-control ring above the gudgeon pin, and one oil-control ring below.Connecting rodsH-section machined nickel-steel forgings, each pair consisting of a plain and a forked rod. The forked rod carries a nickel-steel bearing block which accommodates steel-backed lead-bronze-alloy bearing shells. The "small-end" of each rod houses a floating phosphor bronze bush.CrankshaftOne-piece, machined from a nitrogen-hardened nickel-chrome molybdenum steel forging. Statically and dynamically balanced. Seven main bearings and six throws.CrankcaseTwo aluminium-alloy castings joined together on the horizontal centreline. The upper portion bears the wheelcase, supercharger and accessories; and carries the cylinder blocks, crankshaft main bearings (split mild-steel shells lined with lead bronze alloy), and part of the housing for the airscrew reduction gear. The lower half forms an oil sump and carries the oil pumps and filters.WheelcaseAluminium casting fitted to rear of crankcase. Houses drives to the camshafts, magnetos, coolant and oil pumps, supercharger, hand and electric starters, and the electric generator.Valve gearTwo inlet and two exhaust poppet valves of "K.E.965" steel per cylinder. Both the inlet and exhaust valves have hardened "stellited" ends; while the exhaust valves also have sodium-cooled stems, and heads protected with a "Brightray" (nickel-chromium) coating. Each valve is kept closed by a pair of concentric coil-springs. A single, seven-bearing camshaft, located on the top of each cylinder head operates 24 individual steel rockers; 12 pivoting from a rocker shaft on the inner, intake side of the head to actuate the exhaust valves, the others pivoting from a shaft on the exhaust side of the head to actuate the inlet valves.

The Allison V 1710 also had an aluminum block, as did the Daimler Benz DB series, as did the Hispano and it's Russian copies, etc... etc... etc...

If these large (26+ Litre) engine had been made of cast iron they would have been far too heavy for aircraft use.

[Pursuivant]

Originally Posted by ElAurens

The Merlin had an aluminum engine block with steel cylinder liners.

I stand corrected then, and I feel stupid because I looked at that very page and somehow missed the aluminum reference. It also appears that cast aluminum block engines date to WW1 (Hispano-Suiza produced the first one).

Anyhow, that would explain why some inline engines are so vulnerable. Bullets will generally go right through aluminum, although some grades of aluminum make decent armor (the M113 APC had 3/4" of aluminum armor which prevented penetration by most small arms fire).

So, it's not so simple as figuring joules of energy vs. mm of homogenous rolled armor and dividing by some factor to get penetration of cast iron!