AP Fix

June 12, 2021

I've found a potential fix for the issue with lackluster Armor Piercing ammunition performance in the game.

First and foremost this isn't a comment on the actual Damage Model. Frankly the damage model itself is excellent. The fact you have managed to create a model that spans WW1, Tanks and WW2 is a fantastic achievement.

I think I've finally understood the issue we have with AP ammo having flown in FC for a bit. Jason's comments below we're a really good insight.

We recently overhauled our entire damage model for the engine and making further changes to it for WWI will also affect WWII. This requires some thinking and more study, but at this moment I cannot make further changes to just WWI. I see this as a long-term project to somehow change only WWI damage modeling. Even so, there are mixed opinions on this issue. Just because there are some vocal critics out there, does not make the loudest voice correct. As usual, we would need to study the issue further before any more changes are made.

In a WW1 aircraft, you have some wooden struts with a bit of fabric over them and a lot of space in between. An AP round (of pretty much any caliber within reason) is going to sail right through it and leave a nice round shaped hole. You're going to need a hell of a lot of them to do anything meaningful to the lift of the plane OR you're going to have to hit something important.

Unfortunately this carries over to their younger, but much feistier, sisters we have in the WW2 theaters. The problem is they are much more complicated and are very much not filled with space. If you take a look at a FW190s wing, you'd be hard pressed to find gaps where rounds can neatly go through:

You've got plenty of systems and additional supports to hit and create additional damage. There is absolutely no possible way to model all of this, it's frankly a waste of time and the game would run like garbage. But what we need to do is arrive at some happy medium, whereby we are simulating a half decent hit to the aircraft structure.

The problem at the moment is that every AP hit is it's worst case scenario outside of hitting one of the few modeled systems. This is why we see some ridiculous numbers for AP only ammunition. If you don't hit one of the systems, the aircraft has no ill effects irrelevant of how many rounds you put into them.

We can see this from data gathered from Finnish Virtual Pilots MP server through the month of February.

50 Cals (Wing mounted only) | Mean: 42.84, Standard deviation: 44.63, Median: 29, Mode: 2, 10th Percentile: 5.00, 90th Percentile: 94.00

This means 1 in 10 aircraft take 94 rounds or more to kill.

Now realistically these cases are the outlier, AP ammunition works fine most of the time. Going up with a weapon system that works most of the time in terms of a kill is absolutely fine as well, the issue lies with that fact that if you don't get the kill (fire, pilot kill, engine), the ammunition does almost nothing to the aircraft. If your careful with your placement it's actually possible to put 150+ rounds into a Bf109 and have it lose a few kph off it's top speed.

That's the issue that most people have. It's not about the kill, it's about the fact that nothing can happen before you finally hit the right bit.

So how much damage?

Now this is the question and is where the age old discussion gets really really grey. I don't think the current implementation of AP is an unrealistic scenario, it's just monumentally unrealistic for it to happen every single time.

We actually have a model that NATO created in 1981 to kind of use as evidence . What it shows it that on thin aircraft skin a fast moving HMG AP round is going to leave a nice round shaped hole as it's too fast and the metal is incredibly thin. What it also shows is that both a decrease in velocity and an increase in the thickness of the metal produces more damage.

Source: Design Manual for Impact Damage Tolerant Aircraft Structure

https://www.sto.nato.int › AGARD › AGARD-AG-238

Now within the realms of aircraft material, the increase isn't that significant when just considering a direct hit just to skin. But as you can see above, there is plenty of support and structures to hit and that's where you're going to see chunks flying off.

Similar to what we see in this example:

and what we can read in 90% of these:

http://www.wwiiaircraftperformance.org/p-47/p-47-encounter-reports.html

http://www.wwiiaircraftperformance.org/mustang/combat-reports.html

And now we reach another factor: tumbling

Rounds that hitting anything of significance are almost certainly going to tumble and a tumbling round is likely going to orient itself to it's most likely resting position (sideways) before it penetrates something. This means that you are looking at a lower velocity round that is 37mm wide. Sure it's not a 37mm round, but it's going to cause significant damage according to the NATO model sourced above.

It's not that we want this to happen every time, we just want it to be possible. At the moment, even with the 109 tail fix, it take an enormous amount of rounds (30+ unless delivered very quickly) to actually take any of the surfaces off. An aircraft moving in any way is going to mean it's incredibly difficult to get that many hits on a specific area.

So, the solution

The options on the table are:

devote enormous amounts of the teams time into studying the DM and finding a way to edit certain aspects without affecting the other era's and vehicles using AP ammunition
do the same as above but work in API (the actual damage will be the same we'll just see fires)
leave the DM as it is (it's fine) and slightly modify the values you put into the model to create the desired results

I think the latter is the best route.

Anyway, the proposed fix.

A few members of the community have been trying to fiddle with the ballistics and weapon files to create MODs that might potentially be used. Unfortunately most of our work has meant drastically changing the files.

This is a simple fix that changes 1 attribute: round and gun bore diameter

I’ve upscaled all of the AP only guns (.303, ShKas, MG17 & M2 .50 Cal) within the game to 150% (great thing about this is, if it's too much/too little we can easily edit that number) of their original round diameter. This has no effect on the rest of the ballistic properties. It has very limited effect on the skin damage created by the round. What it does have an effect on is the amount of rounds it takes to remove parts from the aircraft.

A very quick test to show:

I’ve not recorded extensive testing as I really can’t bear putting all that time in if this isn’t going to go anywhere, but from the extensive not recorded testing I've done you are looking at somewhere between a 30-50% drop in the number of rounds it takes to remove parts. What you also see is a slight increase in the fire chance when hitting certain parts of the aircraft.

Now I completely understand that this isn’t my product, I don’t put my name on it and you have certain standards that you want to abide by. This is very much a halfway house fix.

I also understand you will have a development methodology that this will have to go through and a pretty strict QA process.

Please feel free to use me and other members of the community to help. We are more than happy to do testing in our own time to save yours. The 109 tail fix meant a huge amount of work. You could devote a fraction of that time to implementing a simple solution and shut us all up.

But we also have to factor in that this has been in place for over a year now and with the recent language on posts from the team, it feels like it could be another year before anything is done about it.

The Mod:

I’ve attached a zip file with the mod. Unzip this to your Program Files (x86)\IL-2 Sturmovik Great Battles\data\LuaScripts folder and then Enable Mods.

It would be great if as many people as possible can test this and see what their opinion is. I’d particularly implore people who only fly on the receiving ends of .50 Cal to try this out as well (you can do a QMB and let the AI shoot you if you prefer) as you are going to be the most affected if they make a change and it’s too far in the wrong direction.

I do not want OP .50 Cals. I cannot make that clear enough. I love flying the 190 and 109 and it would frankly suck. I’ve deliberately made this a conservative change so it has more chance of being implemented.

The reason we are pushing so hard for this is we can't use specific mods only in MP, it's an all or nothing show. So we can't have this the public servers where the IL2 experience is at it's best.

Caveats:

AP rounds are not as good at creating skin damage as HE - obviously
Cannons are better than AP HMG rounds, they should be worse - obviously as well, we just shouldn’t have the scenario that 100 HMG rounds can’t do as much as 1 cannon round.

Caveat 2: Sorry Jason. I know with each of these posts that goes up your Whisky intake increases. I'm more than happy to send a case of your favorite to Vegas if this goes anywhere ?

Thanks,

Cass

worldobjects.zip

Edited June 12, 2021 by ACG_Cass

June 12, 2021

I'll chip for the Whiskey.

June 12, 2021

Thanks for doing this! I think someone actually owes you a whiskey for taking this on when you really shouldn't have to ?

June 12, 2021

This seems like a plausible fix. Good work! It sucks that I havent been able to fly the American planes really at all yet(and I'm American). I bought Bodenplatte for the p-51, but now I'm a spitfire/tempest guy ?‍♂️.

June 12, 2021

Interesting. I really appreciate the more thorough thinking.

I wonder if Ball would do more damage than AP? One would think that Ball ammunition would be more deformable and thus tumble and transfer energy sooner than AP would.

This still doesn't fix the issues with damage to ground targets (e.g. RP-3 AP is less than half as effective against ships as RP-3 HE; small ships and steam locomotives are more vulnerable to AP than to HE)... it would seem like any long term solution would need to model flooding from below waterline hits and/or give an AP a damage boost against locomotive boilers etc.

P.S. There is also the issue of the overly simplistic convergence modelling - which I suspect contributes a lot to people's dissatisfaction with the M2 (even if they don't know it).

Edited June 12, 2021 by Avimimus

June 13, 2021

Not my video, but it demonstrates why some sort of fix is desperately needed:

June 13, 2021

The lack of complexity of the DM aka "not enough systems to damage" has been discussed a lot in the AP/.50 effectiveness threads and most people agree that it's one of the problems maybe even the biggest one. I'm all for allowing for mods to fix some of these weapon problems (including OP HE effects), but the BIG problem is that most of the people that are really interested in this play online multiplayer servers. Unless 1C starts to allow the server admins to use weapon mods nothing is going to change unless 1C does something themselves. I can't see any reason why they can't allow this. Even if they don't agree with the results at least it would quell the .50 complaints and it wouldn't be a lot of work for them. People would go into a server knowing that the weapons may not be the same as the single player game. If they don't like it they can always go to a different server. I really don't see any problem with that.

Edited June 13, 2021 by BCI-Nazgul

June 14, 2021

Please don't tribalize the online "community" with mods. It killed the online aspect of the original IL2, and it will do the same thing here, and do it faster here because we have far fewer players than we did in the old days.

I'm all for fixing the AP issue, but mods are not the answer.

June 14, 2021

So, every time there is a major problem the online community should just sit around and wait two or more years for a fix (if ever)? I don't think that's much of a solution either. Normally I'd agree with you, but I'm seeing a serious disconnect between 1C's priorities and what the online customers want right now. If there was ever problem that was as thoroughly and completely researched and deserves a solution it's this one. I would have agreed wholeheartedly with you a year ago. I've quit playing this game because I'm sick of the consequences of trying to fly the planes I love with one hand tied behind my back. Now I just monitor the forum hoping for word of a fix.

Edited June 17, 2021 by BCI-Nazgul

June 14, 2021

10 hours ago, BlitzPig_EL said:

Please don't tribalize the online "community" with mods. It killed the online aspect of the original IL2, and it will do the same thing here, and do it faster here because we have far fewer players than we did in the old days.

I'm all for fixing the AP issue, but mods are not the answer.

I think the Point OP is trying to make is that the Devs can use his work as a basis for an easier interim fix.

June 14, 2021

The problem is, as I see it, that if it is done in this manner, someone that gets shot down by a modded "big bore" machine gun will then be all offended and will soon be showing up with "big bore" machine guns on his machine and a MK 108 with a 50mm bore because it can be done, and then it's off the cliff for this title. Even if this solution is done by the devs for "mods off" mode, the genie will be out of the bottle, and this leads to one conclusion, and one conclusion only, the end of the sim.

We need everyone on the same page or the sim will fall apart.

June 14, 2021

19 minutes ago, BlitzPig_EL said:

We need everyone on the same page or the sim will fall apart.

Well, judging from the amount of push back on whether or not the .50s and AP are working properly.

I'd say we are well past that point.

?

June 14, 2021

With some planes being invincible from dead six for a year and HE modeling that makes no sense, I'd say that genie is already out of the bottle.

As for everyone being on the same page, I agree. So either people need to start making better arguments for why an 80 to 1 effectiveness makes sense or they need to start backing a fix for this disparity between AP and HE rounds.

June 15, 2021

4 hours ago, BlitzPig_EL said:

The problem is, as I see it, that if it is done in this manner, someone that gets shot down by a modded "big bore" machine gun will then be all offended and will soon be showing up with "big bore" machine guns on his machine and a MK 108 with a 50mm bore because it can be done, and then it's off the cliff for this title. Even if this solution is done by the devs for "mods off" mode, the genie will be out of the bottle, and this leads to one conclusion, and one conclusion only, the end of the sim.

We need everyone on the same page or the sim will fall apart.

You're missing an important fact. The weapon mods would be SERVER based not client based. Right now there are many other things that are controlled by the server that the individual players have no control over (maps, ground target toughness, etc...) That means the "big bore" MGs would only be allowed on the planes the server admins wanted them on. So, you couldn't just show up with your own weapon mods. The only choice the player has is if they want to play on the server with the weapon settings admins have set.

June 15, 2021

Unfortunately that isn't the way the game is built. If a server is to allow one mod necessarily needs to allow all mods. This is baked into the engine.

June 15, 2021

I guess 1C would have to do a bit of programming then. That change wouldn't be too hard (I think.)

June 16, 2021

I want to take a minute to walk through the damage model presented in AGARD 238 and talk about what it does and doesn't do. Then present another damage model which is far simpler and easier to implement. It also, coincidentally, partially agrees with your Mod .

The AGARD 238 model predicts an upper and lower limit of the damage size. The damage is defined as "the diameter of an imaginary circle that just encloses the limits of cracking, perforation or spallation in the plane parallel to original surface of the sheet". Simply put, the damage size is the bullet hole + any cracks in the material. This size is referred to as LD, lateral damage and damage through the report.

The model predicts lateral damage based on

1. impact conditions, (velocity and impact angle),

2. projectile parameters, (size type and shape)

3. structural parameters, (thickness and materials).

Based on these inputs the model is capable of computing an upper damage limit and a lower damage limit.

The upper lateral damage limit , LDul, is determined by one of three equations / variables.

We compare V* to V1 and V3 to determine which equation to use. V* is the impact velocity times cosine of the impact angle to the .07 power.

And V1 is the velocity of incipient penetration for a given bullet. It's based on the target material it's thickness and the bullet and it's diameter. V3 is onset of the high velocity damage. Again, based on the proprieties of the target, it's thickness, the bullet and it's diameter.

For example the V1 for a 50 cal AP round being fired against a .25 inch plate of 2024 T-3 Aluminum is.

t = .25 (plate thickness)

d = .5 (diameter of round)

V1 = C11 (650) * ln (t/d {.5}) +1080

V1 = 650*-0.69 + 1080 = 631.5

V1 = 631.5 fps

V3 is;

V3 = (500 * (ln .5)) + 2180

V3 = (500 *-.69 )+ 2180 = 1835

V3 = 1835 fps

In the case of our 50 cal AP round Vs a .25 thick plate of 2024-T3, if V* is less than V1 (631.5), the upper limit lateral damage is 0. The round does 0 damage.

If V* is greater than or equal to V1, but less than or equal to V3, The upper damage limit is = Kθ* NIMLD, Where NIMLD is the Normal Limit Lateral Damage Model and Kθ is the the correction factor for oblique impact . Impacts in this range will do the largest amounts of damage

If V*, is greater than V3, in our example 1835, the upper damage limit is computed from HVLD (high velocity lateral damage).

If our AP round impacts at 2000 fps at angle of ~ 28.7 degrees or or less, the upper damage limit is = HVLD

As

V* > V3

2000*(27 Cos)^.7 > 1835

The high velocity damage is less than maximum amount of damage. The distribution of damage typically looks like this. The largest area of damage occurs when V* is:

V1<V*<V3 with largest amount of damage occurring at V2.

For 50 cal round impacting a .25 plate of 2024 t3, V2 would be.

V2 = 725 *(ln .5) + 1800

V2 = 725*.-69+1800

V2 = 1,299.75 fps

The lower lateral damage limit model uses a similar method to select the appropriate algorithm. Which will determine the lower limit damage size prediction.

The Problem with the AGARD model is that it's very complicated and computationally expensive. It gives us two outputs, a lower limit and a upper limit. The model doesn't tell whether the upper or lower limit is more probable. Nor does it relate the damage size to the structural integrity of the material. There is however another model which would be much simple to implement, gives a distinct output and is based on the residual strength of the panel.

Jensen in "The Ballistic Damage Characteristics and Damage Tolerance of Wing Structural Elements" developed a residual strength model for 7075-T6 aluminum panels damaged by .50 AP M2 rounds. The team fired M2 AP rounds at various angles and velocities,

Then developed a residual strength model based on fracture theory. Where Le is the effective damage (damage length)

W

Jense discovered, That the normalized effective lateral damage multiplied by the cosine of the angle of obliquity gave the lateral damage for an oblique hit.

Therefore, for a damage model, you can take the cosine of the impact angle and multiply it by the amount of normal damage to get the damage for the oblique impact.

In Il-2 the structural damage for AP rounds is = the shell diameter. To scale the damage based on the impact angle all that you would need to do is multiply the shell diameter times 1+cos of the impact angle.

Diameter = 12.7

impact angle =

damage scale = 1+(cos impact angle)

effective diameter = damage scale * Diameter.

So why does your mod match the gun camera footage from behind. Well your mod took the 50 cal ap and changed the diameter to 19.05 mm. According to Jense's methodology your mod does the equivalent damage of 12.7mm rounds impacting at an angle of 60 degrees. As 19.05/12.7 = 1.5 and the inverse cosine of .5 is 60 degrees.

Depending on the circumstance a 60 degree impact angle is not unlikely from the rear.

In short, the easiest and one of the most realistic ways to better represent the damage potential of the AP rounds is to increase the effective diameter of the round based on the impact angle.

Edited June 16, 2021 by Yak_Panther
mistakes

June 16, 2021

Generally, most complicated calculations like this can pre-calculated (either outside the program completely or one time at start up) for a limited set of data and loaded into lookup tables for the weapons, metal thickness, and firing angles you want to use. It won't be as precise, but generally it will be close enough for a decent simulation. The computing requirements at run time are much lower using this technique.

June 17, 2021

I'm taking this for a test drive tonight finally.

I'll probably be upset with myself going back to the normal .50s.

But I'm doing it.

June 17, 2021

On 6/14/2021 at 7:29 AM, BlitzPig_EL said:

Please don't tribalize the online "community" with mods. It killed the online aspect of the original IL2, and it will do the same thing here, and do it faster here because we have far fewer players than we did in the old days.

I'm all for fixing the AP issue, but mods are not the answer.

I think you are not understanding his proposal. He only created the mod as a proof of concept. His suggestion is that the devs could use a simple solution like this to create the desired outcome (ie. program the AP rounds on the backend to be handled as if they were 150% of the size they are currently registered as in the software)

Us as the end user would still see .50 listed as .50 and 7.92 listed as 7.92 etc....

The suggestion is not to require a mod and it is not about having the players or servers select a different size round in the sim. It is about how the devs could apply a 150% size modifier to the back-end logic on AP rounds to achieve a desired effect.

June 17, 2021

I've done a bit of testing with your mod, and while I'd like to do a bit more testing with a bunch of different planes for a better comparison, so far I really like this. I really wasn't a fan of the psuedo HE effect given to AP rounds in Rail's weapon mod but this seems to strike a really good balance. it reminds me a bit of a lite version of the pre-update damage model, when tearing off wings was much more common, but a bit more refined.

Funnily enough, when I was doing my testing against 109s in the P-51, I thought they were too powerful, but when I switched over to the 190 and let an AI Mustang shoot me up, I almost felt that they were underpowered. I definitely need to do more testing but if that trend continues I'd say the damage is right where it should be! lol.

I'd like to test P-47s and 51s (both with the 6 and 4 .50s mods) against both 190s and 109s, maybe some tail gunner tests and also some tests with the Tiffie and 51 as targets to admire it in DVD. I'll make another post when I have a bit more experience under my belt.

Hopefully this goes somewhere!!

June 17, 2021

Wil be soon locked

June 18, 2021

Tried it a few times now, still need to do a MP run but it felt really good. Didn't reward poor shooting at all, snap shots still rely on the golden boolet.

But if you sixed up and put a good burst on someone. It did the job without you having those random 100+ hit kills

Edited June 18, 2021 by Denum

June 18, 2021

I doubt the Devs will ever accept this solution.

June 18, 2021

Not necessarily the case. I think they've been open to reasoned approaches in the past.

One thing I wanted to make clear is that we're happy to test anything they need. I understand resources are tight at the moment and for anything to make it into their product, it would need to go through their internal QA process. But if we can get a solution to them in the best possible condition beforehand, then that will make it a much smoother process.

Only issue is we need dialogue from them in the first place and the most recent reactions to posts on the issue have become increasingly closed off. I can kind of understand with how the discussions can get sometimes.

June 19, 2021

On 6/16/2021 at 4:15 PM, Yak_Panther said:

I want to take a minute to walk through the damage model presented in AGARD 238 and talk about what it does and doesn't do. Then present another damage model which is far simpler and easier to implement. It also, coincidentally, partially agrees with your Mod .

The AGARD 238 model predicts an upper and lower limit of the damage size. The damage is defined as "the diameter of an imaginary circle that just encloses the limits of cracking, perforation or spallation in the plane parallel to original surface of the sheet". Simply put, the damage size is the bullet hole + any cracks in the material. This size is referred to as LD, lateral damage and damage through the report.

The model predicts lateral damage based on

1. impact conditions, (velocity and impact angle),

2. projectile parameters, (size type and shape)

3. structural parameters, (thickness and materials).

Based on these inputs the model is capable of computing an upper damage limit and a lower damage limit.

The upper lateral damage limit , LDul, is determined by one of three equations / variables.

We compare V* to V1 and V3 to determine which equation to use. V* is the impact velocity times cosine of the impact angle to the .07 power.

And V1 is the velocity of incipient penetration for a given bullet. It's based on the target material it's thickness and the bullet and it's diameter. V3 is onset of the high velocity damage. Again, based on the proprieties of the target, it's thickness, the bullet and it's diameter.

For example the V1 for a 50 cal AP round being fired against a .25 inch plate of 2024 T-3 Aluminum is.

t = .25 (plate thickness)

d = .5 (diameter of round)

V1 = C11 (650) * ln (t/d {.5}) +1080

V1 = 650*-0.69 + 1080 = 631.5

V1 = 631.5 fps

V3 is;

V3 = (500 * (ln .5)) + 2180

V3 = (500 *-.69 )+ 2180 = 1835

V3 = 1835 fps

In the case of our 50 cal AP round Vs a .25 thick plate of 2024-T3, if V* is less than V1 (631.5), the upper limit lateral damage is 0. The round does 0 damage.

If V* is greater than or equal to V1, but less than or equal to V3, The upper damage limit is = Kθ* NIMLD, Where NIMLD is the Normal Limit Lateral Damage Model and Kθ is the the correction factor for oblique impact . Impacts in this range will do the largest amounts of damage

If V*, is greater than V3, in our example 1835, the upper damage limit is computed from HVLD (high velocity lateral damage).

If our AP round impacts at 2000 fps at angle of ~ 28.7 degrees or or less, the upper damage limit is = HVLD

As

V* > V3

2000*(27 Cos)^.7 > 1835

The high velocity damage is less than maximum amount of damage. The distribution of damage typically looks like this. The largest area of damage occurs when V* is:

V1<V*<V3 with largest amount of damage occurring at V2.

For 50 cal round impacting a .25 plate of 2024 t3, V2 would be.

V2 = 725 *(ln .5) + 1800

V2 = 725*.-69+1800

V2 = 1,299.75 fps

The lower lateral damage limit model uses a similar method to select the appropriate algorithm. Which will determine the lower limit damage size prediction.

The Problem with the AGARD model is that it's very complicated and computationally expensive. It gives us two outputs, a lower limit and a upper limit. The model doesn't tell whether the upper or lower limit is more probable. Nor does it relate the damage size to the structural integrity of the material. There is however another model which would be much simple to implement, gives a distinct output and is based on the residual strength of the panel.

Jensen in "The Ballistic Damage Characteristics and Damage Tolerance of Wing Structural Elements" developed a residual strength model for 7075-T6 aluminum panels damaged by .50 AP M2 rounds. The team fired M2 AP rounds at various angles and velocities,

Then developed a residual strength model based on fracture theory. Where Le is the effective damage (damage length)

W

Jense discovered, That the normalized effective lateral damage multiplied by the cosine of the angle of obliquity gave the lateral damage for an oblique hit.

Therefore, for a damage model, you can take the cosine of the impact angle and multiply it by the amount of normal damage to get the damage for the oblique impact.

In Il-2 the structural damage for AP rounds is = the shell diameter. To scale the damage based on the impact angle all that you would need to do is multiply the shell diameter times 1+cos of the impact angle.

Diameter = 12.7

impact angle =

damage scale = 1+(cos impact angle)

effective diameter = damage scale * Diameter.

So why does your mod match the gun camera footage from behind. Well your mod took the 50 cal ap and changed the diameter to 19.05 mm. According to Jense's methodology your mod does the equivalent damage of 12.7mm rounds impacting at an angle of 60 degrees. As 19.05/12.7 = 1.5 and the inverse cosine of .5 is 60 degrees.

Depending on the circumstance a 60 degree impact angle is not unlikely from the rear.

In short, the easiest and one of the most realistic ways to better represent the damage potential of the AP rounds is to increase the effective diameter of the round based on the impact angle.

I understood none of this but I love it as the mod works per the conclusion.

This should be standard for the game at mininum

June 19, 2021

I think the .50 underperforms for a variety of reasons, some of which might be related to flawed mechanics.

That said, I think an equal or greater contributing factor is how unhistorical combat is in this sim compared to real life. In multiplayer, most of the targets of .50s are evading. Not just evading, but doing so beyond the limits of what probably the vast majority of human pilots would be capable of for a variety of reasons. This means the innate weaknesses in any AP round are compounded. This means getting in a 'good burst' is statistically much less likely.

Given that it's an established fact that most kills are achieved by surprise; that the majority of Luftwaffe pilots circa '44 were poorly trained and incapable of extreme evasive maneuvers even if they did spot their attacker; it's highly probable that .50s were more effective in reality than they are in the sim for the above reasons.

Just as it's easy to show clips of .50s underperforming, so can we show instances where they perform just fine. Against a sluggish target that's not evading wildly at every moment.

Observe these tests I made against a Typhoon within the last month:

Spoiler

Both examples look reasonably realistic to me. The first clip is near convergence or a little beyond, while the second one requires me to line up my wings to make accurate strikes. The only reason I fire so much in the second clip is that I'm shooting very poorly, under the target, for most of my shots. If you look at the actual shots that hit, it's pretty few. Add up all the hits towards the end of the clip, and it's what I'd call "one good burst" in a combat report.

Edited June 19, 2021 by oc2209

June 19, 2021

On 6/16/2021 at 8:15 AM, Yak_Panther said:

In Il-2 the structural damage for AP rounds is = the shell diameter. To scale the damage based on the impact angle all that you would need to do is multiply the shell diameter times 1+cos of the impact angle.

Diameter = 12.7

impact angle =

damage scale = 1+(cos impact angle)

effective diameter = damage scale * Diameter.

So why does your mod match the gun camera footage from behind. Well your mod took the 50 cal ap and changed the diameter to 19.05 mm. According to Jense's methodology your mod does the equivalent damage of 12.7mm rounds impacting at an angle of 60 degrees. As 19.05/12.7 = 1.5 and the inverse cosine of .5 is 60 degrees.

Depending on the circumstance a 60 degree impact angle is not unlikely from the rear.

In short, the easiest and one of the most realistic ways to better represent the damage potential of the AP rounds is to increase the effective diameter of the round based on the impact angle.

That does seem be a very efficiant and realizable solution to take impact angle and tumbling into account.

Thank you very much for your work and factual input.

Cheers

Edited June 19, 2021 by the_emperor

June 19, 2021

I completely understand the point you're trying to make @oc2209. I don't think anyone with any sense is arguing that the .50s don't work at all. We are simply arguing that if you aren't able to light a fire, kill the pilot or damage one of the few components modeled there is no backup in the form of actual aircraft damage. With DVD, we can show how many rounds you need in a specific area to actual have a material impact on the planes flight ability:

4 hours ago, oc2209 said:

Given that it's an established fact that most kills are achieved by surprise; that the majority of Luftwaffe pilots circa '44 were poorly trained and incapable of extreme evasive maneuvers even if they did spot their attacker; it's highly probable that .50s were more effective in reality than they are in the sim for the above reasons.

This is where .50s weakness is really shown though, at dead 6. In real life pilot gunnery was much worse than we have, they had nowhere near the control we have in IL2 and were dealing with huge wash from the plane in front + the pressure of being in a life or death situation. A couple of examples showing you putting what is definitely a lot of rounds into a Typhoons fuel tanks doesn't really prove much either. Allied planes are actually much more susceptible to .50s for some reason.

You can see plenty of examples here from the 4x early Mustangs. Clearly they aren't landing 44 rounds into the planes, which is just the mean that is required according to statistics from the most popular MP server. The fact that 1/10 planes take 94+ rounds shows the absurdity of the situation.

We just don't want situations like this where we are firing into a plane none stop, but haven't hit any of the magic buttons so nothing happens.

Imagine if HE weapons kept all of the skin damage they are capable of but weren't able to damage any components, kill the pilot or light fires without an absurd number of rounds. Most of the time it would be fine, but there would be situations where you would have to unload half your ammo at a plane just to convince it to head into the ground.

Edited June 19, 2021 by ACG_Cass

June 19, 2021

2 hours ago, ACG_Cass said:

You can see plenty of examples here from the 4x early Mustangs. Clearly they aren't landing 44 rounds into the planes

It is worth bearing in mind that guncam footage in cherry-picked, often several times.

Footage that does not show effective damage to aircraft was not usually kept at the time as it was not useful for training purposes.

Even if it was kept, it was not usually preserved for the future unless it was dramatic enough to be selected for preservation or used in newsreels.

Even if it was preserved, it might not be publicly available unless it is dramatic enough to be used in public media.

Even if it publicly available, it might not be posted in a forum thread about weapon damage unless it illustrates a point.

These multiple hurdles mean that the footage we see, is unlikely to be anywhere near a representative sample.

Edited June 19, 2021 by [DBS]Browning

June 19, 2021

3 hours ago, [DBS]Browning said:

It is worth bearing in mind that guncam footage in cherry-picked, often several times.

But if you watch the videos, there are lots of examples of the .50s not doing blockbuster damage? There's plenty out there showing that, you don't have to look hard.

You're completely right in that we should take guncam footage with a grain of salt. But there is enough US footage to show what .50 cals are capable of, and what they are not.

In an ideal world, we'd have an incredibly complex system where lots of different types of damage happen. We'd have HE shells need to hit certain sections to cause drastic skin damage and issues with their fusing. We'd have AP ammo leaving nice small neat holes on one occasion and smashing something to bits on the next.

We obviously can't have that but modelling a HMG AP round in a way that means it is completely incapable of creating meaningful damage to an aircraft outside of hitting a few modelled systems is the issue we have.

Edited June 19, 2021 by ACG_Cass

June 19, 2021

1 hour ago, ACG_Cass said:

But if you watch the videos, there are lots of examples of the .50s not doing blockbuster damage? There's plenty out there showing that, you don't have to look hard.

Well, if that's the case for these cherry-picked examples, then that certainly says something of its own.

June 19, 2021

Unimpressive gun camera film isn’t the result of hundreds of strikes producing no result (like we regularly see in game). It’s the result of an out of focus or obstructed camera, or over exposed film, or very few strikes on the target. Anytime we see a good concentration of strikes, we see a good result. Those best case scenarios should be a good model for what we want to see in game when the player lands a good concentration of hits.

Instead we regularly see stuff like this: https://combatbox.net/en/sortie/1313128/?tour=35

240 strikes on a Ju87 (at 30% accuracy) and the result was a couple fuel leaks and the gunner was killed. Those kind of results are absurd, and underline that something is terribly wrong with the current DM implementation.

June 19, 2021

7 hours ago, ACG_Cass said:

I completely understand the point you're trying to make @oc2209. I don't think anyone with any sense is arguing that the .50s don't work at all. We are simply arguing that if you aren't able to light a fire, kill the pilot or damage one of the few components modeled there is no backup in the form of actual aircraft damage. With DVD, we can show how many rounds you need in a specific area to actual have a material impact on the planes flight ability:

This is where .50s weakness is really shown though, at dead 6. In real life pilot gunnery was much worse than we have, they had nowhere near the control we have in IL2 and were dealing with huge wash from the plane in front + the pressure of being in a life or death situation. A couple of examples showing you putting what is definitely a lot of rounds into a Typhoons fuel tanks doesn't really prove much either. Allied planes are actually much more susceptible to .50s for some reason.

You can see plenty of examples here from the 4x early Mustangs. Clearly they aren't landing 44 rounds into the planes, which is just the mean that is required according to statistics from the most popular MP server. The fact that 1/10 planes take 94+ rounds shows the absurdity of the situation.

We just don't want situations like this where we are firing into a plane none stop, but haven't hit any of the magic buttons so nothing happens.

Imagine if HE weapons kept all of the skin damage they are capable of but weren't able to damage any components, kill the pilot or light fires without an absurd number of rounds. Most of the time it would be fine, but there would be situations where you would have to unload half your ammo at a plane just to convince it to head into the ground.

I highlighted the points I wanted to respond to specifically.

I never said anyone was arguing they didn't work at all. That would be quite absurd. This issue has always been a matter of degree.

I think we disagree on what constitutes "a lot of rounds." In most combat reports I've read, the pilots describe bursts of several seconds as routine. They're burning probably anywhere from 1/4 to 1/2 of their ammo per kill. The burst I used in my first clip probably amounts to less than 150 rounds, of which probably less than 30 actually hit. Similarly with the second clip, most of my shots missed. A lot of them were glancing blows before I got my wing properly lined up.

The 94 round statistic doesn't bother me, because it doesn't show (I assume) where the strikes are. It is entirely possible to ineffectually pepper any plane without concentrated strikes. It's down to bad luck that a handful of those poorly-aimed shots don't do serious damage.

When you see your guns doing serious damage, at convergence, the result is very frequently fatal. This is most often achieved at ranges less than 250m. I think a fair number of people probably have their wing convergence set too high. It's my opinion that anything over 200m is a waste under most combat conditions. This is not just my opinion, but also one shared by at least a few prominent (real life) P-47 aces.

As for the P-51 wing you showed: yes, that is a problem. However, I don't think it's a problem with the bullets' performance as much as the damage model not being able to break the P-51's wing at that location (generally at more locations across the wing than the few break points there currently are). The wing should simply fall off or break off with that many holes. If it doesn't break, then at the very least the aerodynamics should be affected. Which is a problem that making the AP ammo more damaging wouldn't solve. The problem is that the current damage model doesn't (evidently) model air flow disturbance from AP holes.

June 19, 2021

5 hours ago, [DBS]Browning said:

It is worth bearing in mind that guncam footage in cherry-picked, often several times.

Footage that does not show effective damage to aircraft was not usually kept at the time as it was not useful for training purposes.

Even if it was kept, it was not usually preserved for the future unless it was dramatic enough to be selected for preservation or used in newsreels.

Even if it was preserved, it might not be publicly available unless it is dramatic enough to be used in public media.

Even if it publicly available, it might not be posted in a forum thread about weapon damage unless it illustrates a point.

These multiple hurdles mean that the footage we see, is unlikely to be anywhere near a representative sample.

Huh

Guess that means we can discard that video of the 30mm blowing a spitfire in half that everyone uses as a trump card too then.

PERFECT! Very good!

June 19, 2021

I'm not saying anything should be discounted. It's just important to not to view it as being necessarily representative of a typical engagement.

It's not a random sample. It's selected for certain aspects, often more than once.

Edited June 19, 2021 by [DBS]Browning

June 19, 2021

@[DBS]Browning

I understand your point and its good to bring a balanced view to these things, but you have to understand the optics of yours statements.

8 hours ago, [DBS]Browning said:

Footage that does not show effective damage to aircraft was not usually kept at the time as it was not useful for training purposes.

Sure, makes complete sense.

8 hours ago, [DBS]Browning said:

Even if it was kept, it was not usually preserved for the future unless it was dramatic enough to be selected for preservation or used in newsreels.

OK, so guncam is usually the hefty stuff only, understandable.

8 hours ago, [DBS]Browning said:

Even if it was preserved, it might not be publicly available unless it is dramatic enough to be used in public media.

Wow ok, so this really only the absolute creme de la creme, Michael Bay footage.

3 hours ago, [DBS]Browning said:

5 hours ago, ACG_Cass said:

But if you watch the videos, there are lots of examples of the .50s not doing blockbuster damage? There's plenty out there showing that, you don't have to look hard.

Well, if that's the case for these cherry-picked examples, then that certainly says something of its own

What does this mean then? So you're saying these are super cherry picked for their dramatism, despite not some not being dramatic? So the USAAF just ran our of dramatic guncam footage and just stuck in a few duds? Or that 50s don't cause drama very often and are modeled well in the game?

Or, maybe, everything isn't "cherry picked".

I know your trying to provide a balanced response and are saying that guncam footage isn't reliable on it's own, but that's not whats happening. We've got an actual mathematical model, lord knows how AARs and guncam footage.

Anyway, I don't want to derail this thread into the usual poop throwing fest.

@oc2209

Understand and apologies for the sass in my response. 50s work most of the time, especially at convergence as you hit the right areas due to round saturation. The number of rounds is largely irrelevant, the problem is that all of those rounds aren't capable of delivering damage outside of landing within those special areas. So yes, 94 rounds don't matter unless you know where they are. The problem is that for 93 of those rounds the penalty for being hit is minimal.

Edited June 19, 2021 by ACG_Cass

June 19, 2021

1 hour ago, ACG_Cass said:

@oc2209

Understand and apologies for the sass in my response. 50s work most of the time, especially at convergence as you hit the right areas due to round saturation. The number of rounds is largely irrelevant, the problem is that all of those rounds aren't capable of delivering damage outside of landing within those special areas. So yes, 94 rounds don't matter unless you know where they are. The problem is that for 93 of those rounds the penalty for being hit is minimal.

No problem, I can handle a lot more sass than that.

I think maybe a good solution would be to give aerodynamic damage to AP holes at a certain ratio to HE holes.

As many people have noticed, even one 20mm HE hole can seriously affect your handling. 13mm HE are also pretty destructive, with only a handful needed to ruin your stability.

So, with that in mind, maybe some kind of system could be worked out where, for the sake of argument, say 5 AP holes would equal 1 HE hole for the purposes of affecting flight characteristics. So 20 AP strikes cumulatively would have the effect of 4 HE strikes. This would, in turn, make a plane with ~100 AP (12.7mm or above) strikes nearly unflyable.

June 20, 2021

So, we're back to AP should cause aero damage and the DM is complex to utilize AP damage effectively, which is exactly what people have been saying for 1.5 years now. Nothing new. I've read in one US report that the Germans figured it took an average of 50 13mm hits to bring down a B-17 and a long time ago I read that the average number of .50 hits to bring down a German fighter was around 14 hits. Obviously, in game results aren't tracking with this. So, it's really up to 1C to either fix the problem with detail changes to the DM and aero effects OR they can simply change the AP parameters to try and bring the kill numbers into line with the historical numbers (a statistical fix, so to speak.) Basically, there is nothing new here other than the suggestion of one statisical approach to the problem. I'm sorry if I sound negative here, but unless 1C decides to act these AP threads and .50 threads are a waste of time, IMO. My friends and I wasted (that's what I consider it now) a ton of time on demo videos and posts about this months ago and nothing has changed. I my opinion is that 1C cares little about this or doesn't even think there is a problem or it would have been fixed a long time ago. I'm not investing anymore time in this nor I am playing anymore.

Edited June 20, 2021 by BCI-Nazgul

June 21, 2021

I don't think they don't care, I think they don't have time. Based on other module delivery times there's definitely the possibility they are extremely behind.

However, we have a very reasonable short term fix here that should keep both sides fairly happy, people are willing to test it and do all the leg work also.

If giving the 109s the iron tail was an appropriate fix back then. There's little to no sound argument that this couldn't be implemented until they can add API (if I should say)

Especially because this is a suggestion that takes very little developer time.

People are getting burnt out in multiplayer because the late war American aircraft are just frustrating to use.

With the P51B/C coming that's armed with 4 .50s, there's going to a be a bit of community noise because the guns are truely that bad.

AP Fix

Recommended Posts

ACG_Cass

Denum

DBFlyguy

354thFG_Drewm3i-VR

Avimimus

357th_KW

BCI-Nazgul

BlitzPig_EL

BCI-Nazgul

gimpy117

BlitzPig_EL

Denum

-SF-Disarray

BCI-Nazgul

-SF-Disarray

BCI-Nazgul

Yak_Panther

BCI-Nazgul

Denum

SCG_Wulfe

Mtnbiker1998

[-=BP=-]Slegawsky_VR

Denum

BCI-Nazgul

ACG_Cass

=RS=EnvyC

oc2209

the_emperor

ACG_Cass

[DBS]Browning

ACG_Cass

[DBS]Browning

357th_KW

oc2209

Denum

[DBS]Browning

ACG_Cass

oc2209

BCI-Nazgul

Denum

Browse All

Activity