Comparing HMG Damage, and issues with .50 cals

[VBF-12_Mister26]

3 hours ago, VBF-12_KW said:

There’s also that minor issue that the .50s are coded with the incorrect type of ammo.

I think they fire all at once also instead of in a sequence like they should, what we like to call in the Army "Talking the guns"....

[Ribbon]

4 hours ago, Jayhawk said:

26, it comes the .50 passing through aircraft aluminum before energy can be transferred. I too have seen the devastation that .50's will do on solid thick targets. But when you compare a .50 passing though paper or the green man at the range to a 5.56, the only different will be a bigger hole from the .50.

It's what is reflected in the survivability study that I posted. You can't compare a .50 hitting a 3-6 inch thick mud brick house to a 1mm thick piece of aluminum. And I know from personal real life experience of my combat experiences small arms fire going directly though aircraft.

Aircraft structure isn't just 1mm thick aluminum empty shell....you may research it a bit more (aircraft structure and systems) than think about angles projectile is penetrating!

[BCI-Nazgul]

All AP rounds (including .30)  would be more effective if the aero damage was fixed and the DM model actually included all the things you can break inside a real plane modeled.   We're missing probably 75% of the things that an AP round could break in most planes, important things.   If you don't believe me just look at a cut away view of any WW II fighter and compare the things you've actually seen damage for in IL2 against all the things that could be damaged in a real plane.  If you add API to the mix the damage would be even more serious.

Edited January 11, 2021 by BCI-Nazgul

[QB.Shallot]

On 1/8/2021 at 7:44 PM, VBF-12_Mister26 said:

I think they fire all at once also instead of in a sequence like they should, what we like to call in the Army "Talking the guns"....

They do fire in perfect sync. It is correct to have the initial salvo fire at the same time since the American birds used electric triggers, but the sim does not have variable fire rate modeled, so you'll fire 'pulses' throughout a burst instead of a stream of fire like it should be. 

[von_Tom]

 

 

Oh goody, another .50 cal/109/190/LW bias thread.

 

The arguments have all been rehearsed already so let's see how long this one goes on for.

 

von Tom

[Giovanni_Giorgio]

Was there an official response from the devs about the issue?

[Creep]

57 minutes ago, von_Tom said:

 

 

Oh goody, another .50 cal/109/190/LW bias thread.

 

The arguments have all been rehearsed already so let's see how long this one goes on for.

 

von Tom

There's no need to argue, the issue is plain to see. I'm not going to stop beating on this drum until something is done to remedy the issue.

[354thFG_Panda_]

How hard it is to ONLY kill pilot with AP. This was a bug(fixed) but the difference between when this bug happened and and how it is now on the game was quite alarming. AP Bullets in this bug still go through and cause thier nano scale aerodamage. Sometimes in the testing when enough AP bullets hit a part they just fully went through like it was invisible 0_0

No mechanical or structural failures only pilot kill for AP

I will just leave this here...?

(Edit: btw the p38 in the end terminated itself in fustration)))

 

Edited January 12, 2021 by KRASNAYAPANDA

[Creep]

15 hours ago, KRASNAYAPANDA said:

How hard it is to ONLY kill pilot with AP. This was a bug(fixed) but the difference between when this bug happened and and how it is now on the game was quite alarming. AP Bullets in this bug still go through and cause thier nano scale aerodamage. Sometimes in the testing when enough AP bullets hit a part they just fully went through like it was invisible 0_0

No mechanical or structural failures only pilot kill for AP

I will just leave this here...?

(Edit: btw the p38 in the end terminated itself in fustration)))

 

clearly this is just a netcode issue

 

in case it was not clear, that was a joke

Edited January 12, 2021 by QB.Creep

[Giovanni_Giorgio]

A G-6 Late hit by by 5 (!) different opponents before being finally shot by AAA (!!!):

 

http://taw.stg2.de/pilot_sortie.php?id=11725&name=150GCT_Veltro

[Denum]

It's kinda ridiculous that it's been messed up this long.

 

I swear the single player don't make any noise because you just have to breath on NPCs with .50s and they just crash.

 

Unless I was miraculously getting PKs left right and center.

[Blackhawk_FR]

Nothing to do with 50 cal but

 

On 1/8/2021 at 9:48 PM, QB.Rails said:

How bout that 190 should spin out control when missing a vertical stabilizer for starters!

 

At that speed, I'm not sure it should automaticaly go into a spin (in other words, completely loose it's yaw stability). 

[BCI-Nazgul]

On 1/11/2021 at 7:18 AM, mincer said:

Was there an official response from the devs about the issue?

I haven't seen a comprehensive response about this in months.   The closest we've gotten is basically either deleted posts, locked threads, or something similar to, "We're not working on it, so shut up."  I fully expect this post to be deleted shortly.  The sad thing is that this whole topic could be laid to rest for while with a thoughtful response and some idea about when or if they plan to make changes.  Keep in mind that the problem isn't just .50 AP it's all AP.   My opinion is that this could easily be addressed in the short term by adding/increasing  aero damage for AP and increasing the chance for fire or explosions from AP hits while waiting for the DM to actually be capable of modeling the damage properly.  There is no reason that a statistical approach can't be taken vs. a round by round analysis in the code.   When we numerous accounts and video of planes taking 50+ hits of .50 with no immediate ill effects something is clearly wrong. 

Edited January 19, 2021 by BCI-Nazgul

[VA_chikinpickle]

On 1/8/2021 at 1:18 PM, Jayhawk said:

So to keep the discussion going with a larger audience. I don't believe that there is an issue with .50's.

 

So I've posted this before in some discord discussions a few times, but I will post it again here. I disagree with your statement. There is something wrong with the .50 caliber M2 machine guns in il2, but it isn't the AP ammo model that's the obvious problem. Not to say its perfect, but what do I know about it? (The DM). Very little. My impression of the dev team is they are very meticulous and detailed in their efforts to bring us ww2 air combat in the comfort of our homes. The pilot physiology model for example, is fantastic and reading their blog posts it is plain to see the level of thought behind it. So Im going to make the assumption that their AP ammo damage model was approached with the same meticulous effort.

 

That being said, the undeniable problem with the .50 M2 machine gun: It is not a historically accurate model of the weapon system.

 

.50 M2 machine guns in US aircraft were NOT loaded with plain armor piercing ammunition.

 

They were loaded with incendiary ammunition. Either in the form of API (Armor piercing, Incendiary) or API-T (Armor Piercing, Incendiary Tracer).

 

This is just a gut feeling, but I think that our results with pure AP ammo are closer to the mark than not. Why? Because NOBODY used pure AP ammunition in ANY fighter aircraft that Im aware of. I think there is a reason for that, and that reason is it sucked at shooting down airplanes compared to the alternatives.

 

So what is it that we should really do here? Should we argue about the details of AP ammo and make bold statements about the quality of 1C game studios hard work? While some here accuse "allied pilots" (For the record you will find that I fly all the aircraft in the game some time or other) of being bad at gunnery or telling them they need to get better? Making bold claims that "Allied pilots expect .50s to be like cannons, cannons are better!" (Of course they are, thats why cannon armed planes had 2-4 cannons and USAF fighter had 6-8 .50 cal MGs)

 

Personally, I would skip all of that. It only creates toxicity.

 

What I am asking for is only the same historical accuracy that we find in other areas of the game. USAF aircraft should have Armor Piercing, Incendiary ammunition, and API-T. This is what they had in real life. This is what they should have in il2 great battles. It will perhaps make it easier for some and more difficult for others. Those who choose to fly only aircraft from one side, but it will be a better overall experience.

 

Cheers.

Edited January 20, 2021 by VA_chikinpickle

[QB.Shallot]

Introduction:

 

The damage model changes which were made in version 4.005 (released in April 2020) changed how AP and HE ammunition interact with an aircraft’s aerodynamic and structural performance. Since those changes went live, there has been a distinct disparity between HMG’s that contain a HE component (Russian UBS 12.7mm and German MG131) and those that do not contain a HE component (US M2 .50).

 

In real life, the MG131 and UBS both contain approximately 1.5g of HE in an explosive round with a 50/50 mix of AP/HE in the belt. It is worth mentioning that this is 8.3% of the amount of HE found in the Minengeschoß shell. The US M2 .50 that is commonly found on most American aircraft in IL2 fires the incorrect ammunition for the BoBp timeframe, being just a pure AP slug, instead of the ubiquitous API ammunition that was used in the latter half of WW2. The following testing is focused on the disparity between the two classes of weapons purely within the context of airframe damage. Anything other than aerodynamic damage is out of scope for these tests. The two testing factors presented below cover damage to the wing of another aircraft either structurally (the spar is damaged enough that the wing will detach under load) or aerodynamically (enough damage was done to significantly hamper the flight characteristics). 

 

Methodology:

 

A series of tests were conducted in a controlled environment: the Combat Box training server. All tests were recorded using the recording feature and matched with the sortie logs (extracted from the server using IL2Stats). The raw server logs have been preserved as well and will be supplied upon request. Testing in this way eliminates any possibility of netcode or lag issues since the number of rounds hitting the aircraft comes from the server and not the client recording. Each test was done at the same altitude with the same atmospheric conditions. All firing was done from zero degrees angle off tail and zero closure from a range of 100-200m. After each shot, we carefully observed the effect before deciding whether or not more shots were required to reach the desired outcome for the test - this can be verified by the time between shots in the sortie logs (i.e. we were not firing any more rounds than necessary to reach the desired effect).  All tests were conducted using one of three ammo types: the German MG131 (which was recently changed to have a 50/50 mix of AP/HE ammo), the Russian UBS 7.62mm (which also has a 50/50 AP HE MIX), and the M2 .50 (which has no HE component). All tests were done shooting at the center of the left wing of each target airframe. All test data is available for download here.

 

Battery 1, Structural Damage:

 

Our first battery of tests were focused on wing breakage (how many rounds does it take to cause structural failure of the wing). All of these tests were conducted using the 109G14 as the target. The average, minimum, and maximum number of rounds required to cause structural failure of the wing are as follows for each weapon type (ordered ascending by effectiveness):

 

Weapon type	Average rounds	Minimum rounds	Maximum rounds
Russian UBS 12.7mm	24	16	35
German MG131	24.6	20	32
US M2 .50	30	23	36

 

From these tests, we can conclude that the effectiveness of each weapon type with respect to structural failure is relatively close in performance. Let it be noted that on average the MG131 and Berezin UB HMG’s are 25% more effective at causing structural damage than the Browning M2. 

 

Battery 2, Aerodynamic Damage:

 

Our second battery of tests focused on creating an aerodynamic penalty (i.e. firing the minimum number of rounds required to create an appreciable drag penalty). Firing at only one wing and looking at the visual damage, we found a direct correlation between the visual damage observed on the wing and a speed reduction of the aircraft hit. The first change in the visual modelling of a wing surface (which we will refer to as “level one damage”) creates no appreciable drag penalty (2-5 km/h). With additional shots on the wing, the next visual indication creates a significant drag penalty (which we refer to as “level two damage”). We did observe additional drag penalties when more than one section of a single wing have visual damage (e.g. the tip of the wing, the center of the wing, and on some planes the inner portion of the wing) and when components fall off (e.g. aileron, flaps, wheel covers, etc). When this occurred during a test, we removed them from our test results to keep things consistent - all of our tests for aerodynamic damage are the minimum number of shots required to achieve “level two damage” to the middle section of the wing and nothing more. Below are the effects observed for the aerodynamic damage described:

 

Airframe	Max continuous	Level one aero damage to one wing	Level two aero damage to one wing	Percentage speed loss for level two aero damage
109	454kmh	451kmh	405kmh	10.79%
P51	309mph (497kmh)	308mph (495kmh)	285mph (458kmh) / 267mph (429kmh) †	7.8% / 13.68%
190	495kmh	493kmh	465kmh	6.06%

† two levels of damage observed beyond level one for the same wing segment

 

What we discovered was that it literally is not possible to achieve level two aerodynamic damage with the M2 .50 before structural failure occurs on the wing of a Bf-109G14. For these tests, we used the FW-190D9 instead as it can sustain a lot more punishment before structural failure occurs. The average, minimum, and maximum number of rounds required to cause level two aerodynamic damage to a wing are as follows for each weapon type (ordered ascending by effectiveness): 

 

Weapon type	Average rounds	Minimum rounds	Maximum rounds
Russian UBS 12.7mm	3.4	2	5
German MG131	3.4	2	7
US M2 .50	83.4	77	87

 

Conclusion:

 

From these tests, we can conclude that the effectiveness of each weapon type with respect to structural damage is relatively close, however the aerodynamic damage is drastically different. Weapon types with an HE component immediately cause a significant aerodynamic penalty making the target aircraft slower and more difficult to maneuver. In order to achieve the same effect with AP ammo, you must shoot 2,352% more rounds than you do with HE ammo.

 

A pilot using a HMG with an HE component can cause crippling damage with a single hit as long as it contains the HE filler. This can be confirmed by using the single MG131 or single UB machine gun present on German and Russian medium bombers. Effects include a 40+ km/h speed penalty at cruise settings as well as making the aircraft far less maneuverable.

 

A pilot using the M2 .50 is hard pressed to cause any sort of crippling damage to an opponent's airframe before the structural integrity is compromised. There is little difference between an undamaged aircraft and one with level one aero damage applied to its wings, and even less if the hits happen to be on the fuselage. When fighting a Bf-109, an allied player will sooner tear the wing off of his opponent than cause an appreciable drag penalty.

 

An exact solution will not be proposed here, as we have intentionally excluded some factors for testing purposes. However, this data undeniably demonstrates the inherent weakness in the current state on the Browning M2, as well as the overmodeling of low yield HE weapons. It is our hope that the dev team will take this information and continue to improve upon this simulation.

 

While this post is tight in scope, that does not mean that the issues that encompass the Browning M2 are. Gun synchronization, convergence settings, and lack of API will continue to be unhistorical shortcomings for the weapons system until they are resolved by the developers.

 

Special thanks to starting this endeavor @QB.Creep

Thanks to the rest of QB., @VBF-12_KW and @Barnacles for their valued input.

Edited January 20, 2021 by QB.Shallot

[354thFG_Rails]

9 hours ago, JG300_Faucon said:

Nothing to do with 50 cal but

 

 

At that speed, I'm not sure it should automaticaly go into a spin (in other words, completely loose it's yaw stability). 

You're telling me at 530 kph he's going so fast that the prop is having no yaw effect on his plane? I don't buy it.

Edited January 20, 2021 by QB.Rails

[Denum]

Superb post.

 

Very scientific in your approach! 

 

Thank you for your efforts!

[RossMarBow]

Gud poast

[354thFG_Panda_]

I like the clear breakdown in this one and the highlighted differences. The one thing that could improve is more testing like maybe 10 or more times per section but I strongly feel the results would be the same based on my experiences playing both sides.

[CountZero]

Your waisting time testing they can test things faster and fix whats broken if they wonted to, 0.50 were just fine untill 4.005 changed something and it was imidiatly clear that they become mutch weeker since then, you aint gona change mineds of people after 7-8 monhs, their mineds are set in 0.50 is broken or its just fine as its now. When you have to make ammo work in tanks ww1 and ww2 airplanes one part will be brokn.

They can easy test all guns in game if they wont to fined where the problem is:

As I told you recently, I have a special dev-tool that allows me to perform thousands and thousands tests very quickly, instead of firing at an airplane in the game. Using this tool in each test I can measure the number of hits required to break a particular airframe part of the particular airplane, taking into account the direction of shooting (in 3D space) and the particular type of ammo. In the past there was an issue with this tool: the airframe was tested unloaded (meaning the zero-gravity conditions). Thus, this tool measured the number of hits taking into account only the self-strength of the airframe.
 

But here is a good news: last week I improved this tool. Now we can perform this test for the airplane in flight, at given altitude, airspeed, and G-load (in the level-turn other than 1G), while the airplane remains balanced by control surfaces. If the airplane can not be balanced at given airspeed and G-load because of the power deficiency, than we simulate a hard-break turn, meaning that the airplane reaches this G-load at this airspeed while it decelerates during the level turn.

[Creep]

2 hours ago, =[PANDA]=TheRedPanda said:

I like the clear breakdown in this one and the highlighted differences. The one thing that could improve is more testing like maybe 10 or more times per section but I strongly feel the results would be the same based on my experiences playing both sides.

I understand what you're saying, however there was very little deviation from one test to the next in terms of the number of hits required to achieve the desired effect. We did more testing than we have reported here on other aspects of the sim that I will not mention and came to the conclusion that, for those, we did need much larger sample sizes.

 

22 minutes ago, CountZero said:

Your waisting time testing they can test things faster and fix whats broken if they wonted to, 0.50 were just fine untill 4.005 changed something and it was imidiatly clear that they become mutch weeker since then, you aint gona change mineds of people after 7-8 monhs, their mineds are set in 0.50 is broken or its just fine as its now.

I disagree that it is a waste of time. What then should we do, throw our hands up into the air and walk away? Let's not be quitters.

[LemonQuat]

Get your dirty facts off my meta you pigs ?

[RedKestrel]

36 minutes ago, QB.Creep said:

I understand what you're saying, however there was very little deviation from one test to the next in terms of the number of hits required to achieve the desired effect. We did more testing than we have reported here on other aspects of the sim that I will not mention and came to the conclusion that, for those, we did need much larger sample sizes.

 

I disagree that it is a waste of time. What then should we do, throw our hands up into the air and walk away? Let's not be quitters.

This also lines up with other testing I've seen from Unreasonable, and my personal experience. 

 

Hopefully it will help that it's been presented so professionally.

[unreasonable]

This confirms most of what people have been saying, (including me ). One caveat : your "hits to damage" reflect your particular test set up, and the resulting hit distribution. This may well model an idealised in-flight attack, but it is not the absolute minimum numbers required to get the required levels of damage, which are much lower.

 

For instance: "What we discovered was that it literally is not possible to achieve level two aerodynamic damage with the M2 .50 before structural failure occurs on the wing of a Bf-109G14."

 

If you do static testing with the A20 top .50 cal, aiming the shots into one (middle) wing section you get level 2 before the wing breaks off about 1/3 of the time, after about 25 hits.  Additionally, when I test the Fw190 D wing using the same method, I get level 2 damage after about 30 hits: ie less than half your result.

 

So there is certainly some other variable in your tests: assuming that your hit recognition numbers are correct, this is possibly hits being spread over multiple hit boxes, including flaps and ailerons. 

  

 

 

[Creep]

On 1/20/2021 at 8:18 AM, unreasonable said:

This confirms most of what people have been saying, (including me ). One caveat : your "hits to damage" reflect your particular test set up, and the resulting hit distribution. This may well model an idealised in-flight attack, but it is not the absolute minimum numbers required to get the required levels of damage, which are much lower.

 

For instance: "What we discovered was that it literally is not possible to achieve level two aerodynamic damage with the M2 .50 before structural failure occurs on the wing of a Bf-109G14."

 

If you do static testing with the A20 top .50 cal, aiming the shots into one (middle) wing section you get level 2 before the wing breaks off about 1/3 of the time, after about 25 hits.  Additionally, when I test the Fw190 D wing using the same method, I get level 2 damage after about 30 hits: ie less than half your result.

 

Thanks for pointing that out, that is interesting. I wonder what about the damage model calculations change with altitude, planes in flight, etc? Were the planes that you were shooting player-controlled aircraft or static objects?

 

In any event, using bombers on the ground to focus fire on a component changes several variables in the process (altitude, airspeed, forces on the aircraft, etc). It's an interesting point, but it doesn't make me question our conclusions since we saw incredibly consistent results each time we performed the test with the same parameters.

 

One last thing... the screenshot you shared doesn't coincide with what we considered "level two" damage to be. It was only once we got to this level of visual damage that we saw an appreciable drag penalty. See below:

Since @unreasonable believes that we have misrepresented what we consider "level two" aerodynamic damage to be, I want to clarify something: our testing was done against aircraft in flight while shooting at the wing from zero degrees angle off tail approximately 100-200m away with zero closure. All of his testing was done on the ground with bombers shooting at stationary targets hitting wings at oblique angles. There are at least three critical differences between the two tests. I also want to point out that we observed the HE rounds causing "splash" damage to multiple segments of a wing with a single burst. In the above screenshot, I consider this to be "level two" aerodynamic damage on two segments of the wing (inner and middle). Here are a few more screenshots taken from the track files that we uploaded showing our testing:

In the above screenshot, a single burst of MG131 caused "level two" aerodynamic damage to the middle and outer segments of the wing. Here is another:
Notice that in the above screenshot, the damage doesn't look quite the same as the screenshot above for the same wing segment. The difference is irrelevant because our testing was about appreciable drag. It's very obvious when it happens because the plane begins to yaw towards the wing with damage. It is when that occurred that we concluded each aerodynamic test.

Edited January 31, 2021 by QB.Creep

[ACG_Cass]

8 minutes ago, unreasonable said:

If you do static testing with the A20 top .50 cal, aiming the shots into one (middle) wing section you get level 2 before the wing breaks off about 1/3 of the time, after about 25 hits.  Additionally, when I test the Fw190 D wing using the same method, I get level 2 damage after about 30 hits: ie less than half your result.

A static unloaded wing at point blank vs a flying loaded wing with a round covering distance at speed isn't going to be a comparable test.

 

There are definitely some discrepancies though so it is worth pointing out. The fact you can sheer off a wing before it having an appreciable effect on the aerodynamics of a plane does seem a bit off to me.

 

I did some similar tests with the A20 and sheering off the wing doesn't seem to take that many more rounds than sheering off surfaces from the tail section. 

 

When you run these tests on the ground it seems to bring up even more issues.

[QB.Shallot]

@unreasonable

You make a fair point, but seeing as we don't have access to dev tools, we wanted to keep the test as simple as possible to remove any potential outliers. 

Perhaps we should have said ""What we discovered was that it literally is not possible to achieve level two aerodynamic damage with the M2 .50 before structural failure occurs on the wing of a Bf-109G14, when attacking from the 6 o'clock level position."

[RedKestrel]

4 minutes ago, QB.Shallot said:

@unreasonable

You make a fair point, but seeing as we don't have access to dev tools, we wanted to keep the test as simple as possible to remove any potential outliers. 

Perhaps we should have said ""What we discovered was that it literally is not possible to achieve level two aerodynamic damage with the M2 .50 before structural failure occurs on the wing of a Bf-109G14, when attacking from the 6 o'clock level position."

If you are firing from the dead-six position it seems reasonable (pardon the pun) that the bullets would pass through the aileron and/or flaps as well as the wing, and (would do so pretty consistently.)

 

But if the case is that the control surfaces are absorbing some of the energy of the rounds, then it shows the HE rounds are still doing excessive damage, since they too would be hitting flaps or ailerons before the wing. 

Anyway, what this test best shows is the relative difference in effectiveness between the various HMGs. I wonder what would happen if you were able to test the MG131/UBS as AP only. Is it simply a matter of overperforming HE, or something with the M2 in particualar?

[Creep]

26 minutes ago, Cass said:

A static unloaded wing at point blank vs a flying loaded wing with a round covering distance at speed isn't going to be a comparable test.

 

There are definitely some discrepancies though so it is worth pointing out. The fact you can sheer off a wing before it having an appreciable effect on the aerodynamics of a plane does seem a bit off to me.

 

I did some similar tests with the A20 and sheering off the wing doesn't seem to take that many more rounds than sheering off surfaces from the tail section. 

 

When you run these tests on the ground it seems to bring up even more issues.

We didn't put it in the report, but since you bring it up... I think it is worth mentioning that for the structural tests, we didn't exert forces on the wing in-between bursts of fire. We know that sometimes only a few rounds in a wing will cause it to pop off under extreme G-loads. The goal of this test was simply to compare the ammo types for HMGs, so we thought it best to keep things as uniform as possible. And since it is difficult to be precise with the amount of G you are exerting, we thought it best to eliminate that as a variable from our testing.  

[BigGinger]

Thank you to Shallot and company for conducting this test and writing it up so well.  I'm definitely part of the group who has found the M2 .50s very frustrating in recent months, and your data perfectly aligns with my experiences with them.

[unreasonable]

Just now, QB.Creep said:

 

Thanks for pointing that out, that is interesting. I wonder what about the damage model calculations change with altitude, planes in flight, etc? Were the planes that you were shooting player-controlled aircraft or static objects?

 

In any event, using bombers on the ground to focus fire on a component changes several variables in the process (altitude, airspeed, forces on the aircraft, etc). It's an interesting point, but it doesn't make me question our conclusions since we saw incredibly consistent results each time we performed the test with the same parameters.

 

One last thing... the screenshot you shared doesn't coincide with what we considered "level two" damage to be. It was only once we got to this level of visual damage that we saw an appreciable drag penalty. See below:

 

I fully agree with your overall conclusions that the ratio of surface damage per hit between HE and AP rounds is very large, to a degree that is implausible for the small HE rounds. 

 

Static testing firing at an AI plane with engine on, on the runway.  At close range the energy is not much different from at 100-200m, and you can place 1-2 shots at a time exactly - with a bit of practice - only damaging one hit box (with AP).  While in flight a wing section will usually fall off earlier, if it is subjected to more than 1G, whether drag forces also affect the DM I know not: although I doubt it. Similarly I doubt that the hits required for surface damage are affected by aircraft speed or Gs: it seems like an unnecessary complication. 

 

Your screen shot is showing damage in two wing hit boxes, middle and inner. (As well as flaps and aileron). To me that looks like level 3 on each, but after a few tries I cannot get level 3 with .50 cal AP without the wings dropping off.  I imagine this was one of the HE test runs?  Level one is the initial "bullet holes" graphic. 2 Is a few larger holes as per my screenshot, which you can sometimes see through. 3 is widespread large holes as in your screenshot. 

 

So if that is what you were testing for, you might consider amending the description of the results. It would also help explain why your numbers are so much higher than mine! 

 

On the issue of "What should be done?" I am agnostic on the real aerodynamic effects of .50 cals in the absence of any real evidence:  I am more convinced that the damage from small HE rounds is way OTT. I certainly do not think that aerodynamic damage should be exaggerated because some MP pilots fight on in situations where a real human (or for that matter the game AI) would bail out. So I would rather the HE aero damage was scaled back considerably rather than the AP enhanced.

 

(Although where appropriate the game clearly needs an incendiary component. That is not just a .50 cal issue BTW, the 303s had them too, and so did most of the guns in the FC planes - although most of the FC grognards will hate it. They like long fights.   )   

Edited January 20, 2021 by unreasonable

[MeoW.Scharfi]

Great Post, thank you! ?

[QB.Shallot]

Quick heads up, there were some complaints that the line: "It is worth mentioning that this does not exceed the explosive mass found in a firecracker." was misleading. In retrospect, I agree, and have removed the line from the report, as it gives a false sense of the explosive power when the material is under containment. 

I've replaced it with "It is worth mentioning that this is 8.3% of the amount of HE found in the Minengeschoß shell" to provide a relative comparison. 

[Alonzo]

I think this kind of empirical testing is 100% worthwhile. The devs have a tool that simulates the damage, which is great, but that's not the same as the game actually behaving in the same way once it's running on a server, game clients are connected, and planes are in the air.

 

There seems to be a significant discrepancy in HE vs non-HE when it comes to creating aerodynamic damage. These results would at least cause me, as a developer, to check if everything's wired up the way I thought it was. This is an odd result and might imply some sort of bug or edge case in the modeling.

[CountZero]

12 hours ago, QB.Creep said:

I understand what you're saying, however there was very little deviation from one test to the next in terms of the number of hits required to achieve the desired effect. We did more testing than we have reported here on other aspects of the sim that I will not mention and came to the conclusion that, for those, we did need much larger sample sizes.

 

I disagree that it is a waste of time. What then should we do, throw our hands up into the air and walk away? Let's not be quitters.

You and others can spend weeks months on testing something that they can do in a day, and if things are ok they would have no problem show with that tool that all is all ok with AP ammo in game and problem is in players heads or in netcode.

 

11 hours ago, RedKestrel said:

If you are firing from the dead-six position it seems reasonable (pardon the pun) that the bullets would pass through the aileron and/or flaps as well as the wing, and (would do so pretty consistently.)

 

But if the case is that the control surfaces are absorbing some of the energy of the rounds, then it shows the HE rounds are still doing excessive damage, since they too would be hitting flaps or ailerons before the wing. 

Anyway, what this test best shows is the relative difference in effectiveness between the various HMGs. I wonder what would happen if you were able to test the MG131/UBS as AP only. Is it simply a matter of overperforming HE, or something with the M2 in particualar?

It can be done in mods, just replace MG131 HE ammo with its AP ammo so it will fire only 13mm AP, same to russian 12.7 with its HE replaced with its AP, and same thing would happend as with american 0.50 AP only, problem is as it was always they buffed HE and make AP more week, like they said in DDs and updates from that 4.005 somehow before 4.005 there was perception that AP ammo does more damage and HE is to week, and this is how we ended up with this week AP we have since 4.005. Its not like its mistery on what happend all is pointed out many times.

 

Edited January 21, 2021 by CountZero

[unreasonable]

Worth adding, in case anyone is confused about what the damage levels really look like.

 

Here are the three stages of surface damage on an He111 outer wing section, chosen because this is easy to get with the onboard MG81 without anything falling off. 

 

 Level 1

 Level 2

 Level 3

 

 

Here is 109 G14: middle wing hit box. First two done with MG81, last with MG131. The line of damage through the inner side of the cross means the section is close to breaking off.

 

 

 Level 1

 Level 2

 Level 3

Edited January 21, 2021 by unreasonable

[Black-Witch]

So where should the Browning .50 sit compared to the MG131 and the UB? How many rounds should be needed to cause the damage described above?

 

In the middle, between them both, with the UB being the best of the three. This according to A.G. Williams & Dr E. Gustin in "Flying guns WWII".

 

Here is the data they work from:

 

UB* 12.7mm x 108mm: 840 m/sec velocity, 13 rounds per second synchronised, 17 rounds per second un-synchronised, 48gm round with 2gm of HE.

 

M2 .50": 12.7mm x 99mm, 890 m/sec, 13 rps unsynchronised, 43gm bullet 0.9gm incendiary in the M8 round. so slightly lighter round and less Chemical energy.

 

MG131 13mm x 64mm:

AP 710 m/sec, 38.5gm 15 rounds per sec'. Lower velocity, less weight.

HE 750 m/sec, 34gm 1.2gm of HE. Lower velocity, less weight and less chemical energy than the USB.

 

The authors use a formula to get a "score" for comparing different weapons: 

 

"momentum" = projectile weight multiplied by the muzzle velocity, to calculate the kinetic energy.

 

Chemical energy: they treat HE and Incendiary as the same because fire was a big killer of aircraft, and although they "work" differently one is not more effective than the other because of the random way rounds hit different parts of an aircraft, so the score can be considered as an average when comparing weapons using HE or Incendiary or even HEI.

 

The authors say the comparison between the effect of kinetic and chemical energy is a difficult and complex subject, and quote the example of a delayed fuze HEI round, it will first cause kinetic energy as it strikes the structure, then chemical blast energy as the HE detonates, thirdly the shell fragments will inflict further kinetic energy and finally the incendiary, spread by the explosion, could cause further chemical (fire) damage.

 

Once each cartridge's score is calculated they then take into account the different rounds in a belt or magazine, they also allow for tracer rounds, which "can" reduce the amount of HE in certain shells, and finally the rate of fire is calculated, however, as in this case we are talking about a few rounds doing measurable damage I shall include only their single cartridge scores and the "average" score of rounds within a belt.

 

Here are the authors scores for these 3 HMG's.

 

MG131 AP: single cartridge score 27. comparison score 3.

MG131 HE: single cartridge score 34. comparison score 3.

M2 .50 API: single cartridge score 46. comparison score 4.5.

UB*: single cartridge score 57. comparison score 6.

 

According to the authors, the Browning .50 M8 round sits almost exactly between the MG131 and the UB* for damage potential of a single round and the average damage of several rounds hitting.

 

The authors state, "The Browning .50 M2 is an undistinguished performer, particularly when compared to the 12.7mm Berezin. The relatively small incendiary content in the .50 (0.9g compared to 2g) gives the Soviet round a flying start which it adds to with its higher rate of fire, and the .50 round is lighter as well.

 

The UB* (universal gun) came in 3 forms:

UBT, fitted in turrets.

UBK, fixed wing installation. 1050 rpm.

UBS, synchronised. 800 rpm.

 

It should be remembered that the US M8 round only became available in late '43 early '44, roughly when the P-51B's arrived, before that there were 2 rounds for the M2 .50, the M2 AP round and the M1 Incendiary round, these 2 would have been used in the P-39, P-40 and P-47 before the M8 became available.

 

Witch

 

 

 

 

 

 

[[-=BP=-]Slegawsky_VR]

Its beyond ridiculous now, 13mm against .50cal, 109s tail and the list goes on and on.

We pay good money and being denied practical use of american aircraft from the package!!!!!

Dear wonderful devs, please sell german aircraft as separate product or fix this mess.

Kind regards

[unreasonable]

Just now, Black-Witch said:

So where should the Browning .50 sit compared to the MG131 and the UB? How many rounds should be needed to cause the damage described above?

 

In the middle, between them both, with the UB being the best of the three. This according to A.G. Williams & Dr E. Gustin in "Flying guns WWII".

 

<snip>

 

 

I suspect everyone here has read the Gustin and Williams stuff: it really does not help at all in this case.

 

AP vs HE create types of damage in different ratios, which is what the game has to model to get a plausibly realistic outcome.  The OPs are concerned with the surface damage, but not the internal structural damage, so averaging out the total numbers using an index score conceals rather than adds information. You have to have a view on how much of the total energy from each round - which is reasonably easy to calculate - is transferred to the skin, passes through into the structure, or is wasted in space. Then you have to have a view on the vulnerability to damage of various aircraft components that absorb the energy, and the wider effects on the aircraft's performance.

 

G&W's index cannot do any of that.

 

(BTW: "momentum" = projectile weight multiplied by the muzzle velocity, to calculate the kinetic energy." - actually, no. The kinetic energy is = 1/2 * m * V^2 , which Gustin et all quite explicitly do not use, giving bizarre reasons. They use momentum to calculate kinetic damage. Every other source I have found that does this kind of index uses kinetic energy instead.)

 

 

Edited January 21, 2021 by unreasonable

[Creep]

On 1/19/2021 at 3:59 PM, JG300_Faucon said:

At that speed, I'm not sure it should automaticaly go into a spin (in other words, completely loose it's yaw stability). 

I thought you were a stunt pilot?