Mitsubishi Stuka and Thunderbolt

[JtD]

Actually, the only single shot kill to the R-2800 is hitting the ignition, on top of the hub in the P-47. No armour or backup system there in real life, a clear one shot kill.

[horseback]

Originally Posted by JtD

Actually, the only single shot kill to the R-2800 is hitting the ignition, on top of the hub in the P-47. No armour or backup system there in real life, a clear one shot kill.

How big is the 'ignition' on the R-2800? What percentage of area inside the cowl (viewed from head-on) does it represent? How often does the average two or three shot burst of 7mm into the cowl from a rear gunner kill the engine?

If the answer to the last question is is a greater number than the answers to the the first questions, something is very wrong. Hitting one of the ignition coils(?) on top of the crankcase from over 200m is a classic Golden BB, and should be vanishingly rare.

It is not vanishingly rare.

cheers

horseback

[The_WOZ]

As most planes in the game, both the P-47 and the Fw-190 have complete internal collision boxes modelled. This is, individual models for each internal system with a rough shape and size.

This means that if the ignition system is shot out, is because a bullet indeed hit that small part of the engine.

Now, there's another group of planes, which includes the Stuka and the P-39/P-63, that do not have any internal collision boxes at all.
In these planes the damage to internal systems is determined procedurally every time a bullet hits the airframe.
For the Stuka there's around a 60% chance for a incendiary bullet shot to the wing root to set the wing tanks on fire.

I have once hit a P-47 with 80 20mm rounds (from a J2M) and it flew away. Other times a few hits from a 109 on the wing root will bring it down.
Even on planes with complete damage models there's some randomness thrown in to make things more interesting and realistic.
For example, back to the ignition system in the P-47, a bullet may not have enough force to knock out a magneto, but it may still sever some wiring and have the same effect.

In the case of the F6F, as a Ju88 and Betty pilot I can attest to the engine toughness to both MG and cannon fire.

[ElAurens]

I've had more "one shot" instant stops in the P 47 than any other plane in the sim. Second place goes to the glass jawed P 40.

You could probably bring the IL2 P 47 down easily with a side arm, if they were available in the sim.

[horseback]

As most planes in the game, both the P-47 and the Fw-190 have complete internal collision boxes modelled. This is, individual models for each internal system with a rough shape and size.

This means that if the ignition system is shot out, is because a bullet indeed hit that small part of the engine.

Which tells me that the shot is impossibly accurate, or that the 'collision boxes' are on the especially generous side for the P-47's (and the Hellcat's and the Corsair's) engine compartment.

As for random numbers, only God can generate a truly random number; there is always a prejudice built into any system built by men, and it is pretty obvious here.

When I run down and across the rear of a Betty at a 45 degree angle after a high 7 o'clock diving gunnery pass, and the rear gunner takes out my engine 3 out of 5 times in a QMB (my speed was in excess of 370 kts every time), that is not random.

When I approach from a level 4 o'clock, and get my engine knocked out from 450 meters as often as once in five tries, that is not random.

It is wrong.

cheers

horseback

[The_WOZ]

Let me put it this way, without these pseudo random numbers every single hit would knock out the engine, break controls, set fires, tear wings, etc, every single time

And no, the game doesn't call a special "letsporkthep47.rnd()" function, it uses de very same random number generator for the whole game. So by your logic the entire game is wrong

Also look at the size of the R-2800 distributors in this picture, each is as big as an human head.



Now, I'm not claiming that the damage model is perfect as it is, and I don't fly the P-47 so I can only tell what I have seen when flying against it online.
I would expect the engine to be somewhat more fragile, given the complex instalation with the turbocharger on the belly and all the plumbing it needs. But still it shouldn't have more probability to seize than the F6F or F4U.

Personally I wouldn't guide by getting the engine killed on a single hit by AI, it always had that supernatural hability to do tremendous damage with few rounds.

I have the same problem with the Hs-129 and AAA, it would aim directly to the engines and knock them down on a single hit, rarely it hits other thing that the very precise spot that kills the engines.

[Pursuivant]

Originally Posted by The_WOZ

Personally I wouldn't guide by getting the engine killed on a single hit by AI, it always had that supernatural hability to do tremendous damage with few rounds.

This has been changed in 4.12. If anything, AI gunnery isn't good enough, at least for Ace or Veteran fighters.

Originally Posted by The_WOZ

I have the same problem with the Hs-129 and AAA, it would aim directly to the engines and knock them down on a single hit, rarely it hits other thing that the very precise spot that kills the engines.

That might speak not so much to the accuracy of the AAA, so much as flaws with the DM for the Hs-129. TD gave us a lovely rework of this plane, but they might still be retaining old, bad DM.

[horseback]

Originally Posted by The_WOZ

Let me put it this way, without these pseudo random numbers every single hit would knock out the engine, break controls, set fires, tear wings, etc, every single time

And no, the game doesn't call a special "letsporkthep47.rnd()" function, it uses de very same random number generator for the whole game. So by your logic the entire game is wrong

Also look at the size of the R-2800 distributors in this picture, each is as big as an human head.



Now, I'm not claiming that the damage model is perfect as it is, and I don't fly the P-47 so I can only tell what I have seen when flying against it online.
I would expect the engine to be somewhat more fragile, given the complex instalation with the turbocharger on the belly and all the plumbing it needs. But still it shouldn't have more probability to seize than the F6F or F4U.

Personally I wouldn't guide by getting the engine killed on a single hit by AI, it always had that supernatural hability to do tremendous damage with few rounds.

I have the same problem with the Hs-129 and AAA, it would aim directly to the engines and knock them down on a single hit, rarely it hits other thing that the very precise spot that kills the engines.

That was why they had two distributors, for redundancy. My argument is that the Corsair and Hellcat are both at least as fragile as the P-47 (and the Hellcat is probably the worst of the trio). You might also take another look at your picture: there's a lot more space between the cylinder heads and the metal of the cowling than we would see in a comparable picture of a Focke-Wulf or Lavotchkin. The damage model for the R-2800 is plainly wrong, and the DMs for the rest of the aircraft that use it are also clearly exaggerated in light of their combat records.

As I pointed out, just making a hit inside the cowling of a fighter approaching from any angle at any distance was extremely difficult; it happens far too often, and for certain aircraft, I suspect that it happens extra often, just as certain ai aircraft always seem to be crewed by clones of Little Stevie Wonder at the gunner's position (regardless of assigned skill level) and others always have the virtual offspring of Annie Oakley and Davy Crockett at their guns.

Do a little research and see how the fuel tanks and turbosupercharger in the Thunderbolt was installed and try to remember that American fighters in general were notoriously over-engineered and built to greater stress standards than the European norm (and most definitely the Japanese norm), and that the P-47 was considered even more so. There was a heavy belly 'keel' added early in the P-47C series in order to support a large drop tank and plumbing for it; it added a lot of protection for fuel tanks that were already buried pretty deeply inside the fuselage and had oodles of the leak-proofing that was standard on less protected tanks in other aircraft of the era. See the attached picture:

That's just the turbosupercharger; it doesn't show the supporting frames or the basic structure that covered it. The majority of it was ducting and piping that was hard to hit, and wouldn't be easily or seriously punctured unless hit from the right angle (i.e., a low-probability high deflection shot). The critical components were comparatively small. It could take a licking and keep in ticking.

Carrier aircraft are designed to take the stresses of repeated carrier landings, which adds to the strength and density of the airframe, which makes them even tougher to destroy or damage.

All three of these aircraft were used extensively in close ground support and were universally acknowledged as the safest aircraft of the war for that task. Corsairs and Hellcats largely replaced the SB2C Helldiver in the divebombing role by the end of the war because the difference in accuracy was minimal and the aircraft were much more capable (and survivable) after they dropped their bombs at low level over some of the densest light flak in the world.

The FW-190's BMW and the La-5/-7 series are treated much better, as I have pointed out, and neither had a record remotely comparable, particularly when specifically assigned to ground support. The schlacht variants of the 190 were heavily rebuilt with extra armor and weapon installations; the P-47, F6F and F4U assigned to ground attack were no different from the models assigned to air combat. Hour for dedicated ground attack combat hour, all three of these American fighters were statistically safer to fly over heavily defended enemy positions.

But not in this game. The 'random numbers' that generate hits and determine damage are clearly wrong for these aircraft.

And of course, it is a given that the AI are accurate (and swivel their guns ridiculously fast) at any distance or skill setting, to the point of parody (seriously, not a lot of trained soldiers can hit a stationary target the size of a P-47's cowl opening from 400m away, much less pick off the distributors with a standard rifle, much less than that hit it with a machine gun with open sights from a moving platform). For the off-line player, that is by far the biggest hole in the sim, but there appears to be no effective way to limit their accuracy to realistic levels (that DT will admit to, but I'm still getting disabling strikes from Rookie Zeros at over 400m, and as for the Betties...).

In general, the aircraft crew ai are far too accurate at ridiculous ranges and angles, and are able to bring their guns to bear much too quickly. They consistently hit 'spots' more often than the aircraft in general. The AA on the ground also seems to be modeled as being equivalent to a battery than as an individual emplacement, and again, even the heavy guns readjust and aim at 'spots' with inhuman speed.

I've flown many Soviet and German campaigns in Il-2 Sturmovik and all its successors over the last eleven years; no German or Soviet aircraft is as susceptible to being hit or being seriously damaged by those hits as the US radial powered fighters, when you take into account their relative size and speeds. It is not a matter of tactics; you can use the QMB to assign yourself a FW 190 or La-5F/FN to attack the same formation of Betties or He-111s and take the same angles of approach and you will take fewer hits and less damage than you will with a P-47, F4U or F6F.

You will find that these aircraft are harder to keep trimmed, that their instrument displays are generally inaccurate, and even once you finally figure out how to fly them with comparable precision, you will still take more hits than with the German or Soviet radials and that those hits will do disproportionate levels of damage.

cheers

horseback

[Woke Up Dead]

Originally Posted by The_WOZ

Now, there's another group of planes, which includes the Stuka and the P-39/P-63, that do not have any internal collision boxes at all.
In these planes the damage to internal systems is determined procedurally every time a bullet hits the airframe.

Hi WOZ, do you happen to know the full list of these planes with the simplified damage model?

Someone here once posted an image of these damage boxes in a Zero, and the lack thereof in the P-39 (maybe it was you), but I couldn't track down the list of planes or the tool used to illustrate the damage boxes.

Thanks,

WokeUpDead

[ElAurens]

Long ago at the old UBI forum, Oleg did indeed say that single flak guns are modeled as a battery, to help with FPS issues in the sim.