Armament and equipment

March 5, 2021

@Supercharger, there is a work plan, so not all at once.

March 7, 2021

@-DED-Rapidus, thanks for that information. I am aware of future improvements.

April 10, 2021

Brief description:

MG 131 shoots only HE bullets from BF 109 G-6 Late.

Detailed description, conditions:

The BF 109 G-6 (not late), G-14, K-4, and FW190 A-8, D-9 all shoot mixed AP/HE ammo from their MG 131s. The BF 109 G-6 Late seems to shoot only HE rounds from the MG 131.

I've noticed this issue while going through some multiplayer DServer logs from the month of Feburary. Specifically, I've been parsing log lines corresponding to a hit which look like:

AType:1 AMMO:(?P<ammo>[-\w]+) AID:(?P<attacker_id>\d+) TID:(?P<target_id>\d+)$

and noticed that while BULLET_GER_13x64_HE shows up as an AMMO for BF 109 G-6 Late attacker ids, BULLET_GER_13x64_AP never does.

This is based on a month of data with a few hundred engagements involving the BF 109 G-6 Late.

Additional assets (videos, screenshots, logs):

If you want DServer logs which show the issue please direct message me. I will upload and give you the DServer logs in a direct message back. The data is Finnish server logs that they've given me to test with, and the admins do not wish for the logs to be made public.

The cause of the issue might be inside luascripts/worldobjects/planes/bf109g6late.txt, note the following lines:

[GunAmmunition=0]	// 100 бронебойных и 200 осколочно-фугасных пуль 13х64 (к пулемету MG 131)
	ExpendableMass = 25.0
	ResidualMass = 0.0
	MaxMisFiresInMagazine=0
	MaxRoundsInMagazine=300
	RoundsInMagazine=300
	ReservedMagazines=0
	AmountRoundsWithOneTracer=4	// каждая 4-я пуля с трассером
	TracerIdx=0
	object0="LuaScripts/WorldObjects/Ballistics/Projectiles/BULLET_GER_13x64_HE.txt"
	object0="LuaScripts/WorldObjects/Ballistics/Projectiles/BULLET_GER_13x64_AP.txt"
	object0="LuaScripts/WorldObjects/Ballistics/Projectiles/BULLET_GER_13x64_HE.txt"
	target0="LuaScripts/WorldObjects/Ballistics/Projectiles/BULLET_GER_13x64_HE.bin"
	target0="LuaScripts/WorldObjects/Ballistics/Projectiles/BULLET_GER_13x64_AP.bin"
	target0="LuaScripts/WorldObjects/Ballistics/Projectiles/BULLET_GER_13x64_HE.bin"
	BushConfig = "LuaScripts/WorldObjects/Trash/Batch_case12-20mm.txt"
[end]

"object0" is repeated over and over, it probably should be object0,object1, etc.

Your PC config data (OS, drivers, specific software):

Windows 10, latest NVIDIA drivers.

I'm developing an update for my IL-2 Stats website mods which shows which specific bullets hit an enemy, to answer questions like "how many MG 151 rounds did my Pe-2 survive in that sortie?" and "Avg ShVAK (AP) + (HE) rounds needed to down an enemy". This is the software which made me find this bug :<

Edited April 10, 2021 by =FEW=Revolves

April 21, 2021

On 3/1/2021 at 2:35 PM, -DED-Rapidus said:

@JG4_Widukind, please do not delete the video from YouTube, we will fix bug soon/

realy??

April 22, 2021

12 hours ago, JG4_Widukind said:

realy??

Yes, on a first-come, first-served basis)

April 22, 2021

what do you mean?

April 23, 2021

I believe debug has been left switched on by accident in bullet_usa_12-7x99_ap.txt

April 25, 2021

Brief description:

All narrow-gear airframes - currently therefore applies to Spitfires only - experience frequent interruption during REAM in RRR by movement of airframe due to the slighest wind-condition on the ground

Detailed description, conditions:

When Spitfires use RRR on an airfield in the service zone, the REARM process will be frequently interrupted due to the airframe being tilted into its suspension in the slightest wind-condition on the ground. This ONLY affects the REARM process, the REFUEL process will not be interupted by the same airframe tilt in the very same conditions.
The surface condition is not relevant as the suspension rumble-tilt happens on all available surfaces.
The application of brake power (permanent press of keybind) is not relevant as it does not affect / does not hinder / does not stop the aiframe suspension fumble-tilt.
The intensity of the tilt is not relevant as the slightest presence of a wind condition and the resulting slightest of tilts aborts/interupts the REARM process (orange techchat message).

Local client hardware, ping, FPS, packet fragmentation aso are also not relevant as the interrupt is a local condition and not performance related.

The bug would affect all narrow-gear airframes but in realiter currently only affects Spitfires (Mk V, Mk IX tested, Mk XIV not tested).

As only the REARM process is affected, and the very same aiframe tossing/suspension tilt/externally caused airframe movement does neither affect the REFUEL process (both in player-agency as keybind-triggered and keybind-aborted/controled), nor the automated REPAIR process (no player agency, starts with engine off) a higher movement tolerance trigger / timeout on movement for the REARM abort seems advizable if no other solution is feasible.

WORKAROUND (temporary):

repeat REARM process / reinitiate until "REARM COMPLETE" message appears (in itself questionable, as the process is supposed to mimic the unloading and reloading of ammunition and cocking, cleaning, servicing of guns, thus the second by second step-progress being possible at all for guns as a byproduct of the as-of-yet nucleonic payload player agency feels off in itself)

Additional assets (videos, screenshots, logs):
-

Your PC config data (OS, drivers, specific software):

not bug-relevant or interdependent but dxdiag attached

DxDiag.rar

April 25, 2021

Brief description:

All narrow-gear airframes - currently therefore applies to Spitfires only - experience frequent interruption during REAM in RRR by movement of airframe due to the slighest wind-condition on the ground

Detailed description, conditions:

When Spitfires use RRR on an airfield in the service zone, the REARM process will be frequently interrupted due to the airframe being tilted into its suspension in the slightest wind-condition on the ground. This ONLY affects the REARM process, the REFUEL process will not be interupted by the same airframe tilt in the very same conditions.
The surface condition is not relevant as the suspension rumble-tilt happens on all available surfaces.
The application of brake power (permanent press of keybind) is not relevant as it does not affect / does not hinder / does not stop the aiframe suspension fumble-tilt.
The intensity of the tilt is not relevant as the slightest presence of a wind condition and the resulting slightest of tilts aborts/interupts the REARM process (orange techchat message).

Local client hardware, ping, FPS, packet fragmentation aso are also not relevant as the interrupt is a local condition and not performance related.

The bug would affect all narrow-gear airframes but in realiter currently only affects Spitfires (Mk V, Mk IX tested, Mk XIV not tested).

As only the REARM process is affected, and the very same aiframe tossing/suspension tilt/externally caused airframe movement does neither affect the REFUEL process (both in player-agency as keybind-triggered and keybind-aborted/controled), nor the automated REPAIR process (no player agency, starts with engine off) a higher movement tolerance trigger / timeout on movement for the REARM abort seems advizable if no other solution is feasible.

Bug actively observed throughout entire first quarter of 2021, first observation in 2020 after introduction of RRR iterations.

WORKAROUND (temporary):

repeat REARM process / reinitiate until "REARM COMPLETE" message appears (in itself questionable, as the process is supposed to mimic the unloading and reloading of ammunition and cocking, cleaning, servicing of guns, thus the second by second step-progress being possible at all for guns as a byproduct of the as-of-yet nucleonic payload player agency feels off in itself)

Additional assets (videos, screenshots, logs):
-

Your PC config data (OS, drivers, specific software):

not bug-relevant or interdependent but dxdiag attached

DxDiag.rar

April 26, 2021

On 4/22/2021 at 8:04 PM, JG4_Widukind said:

what do you mean?

I don't know when engineers will start fixing it, so I wrote a request not to delete the video. That's what I meant.

May 10, 2021

On 4/25/2021 at 5:46 AM, [_FLAPS_]RogoRogo said:

Brief description:

All narrow-gear airframes - currently therefore applies to Spitfires only - experience frequent interruption during REAM in RRR by movement of airframe due to the slighest wind-condition on the ground

Detailed description, conditions:

When Spitfires use RRR on an airfield in the service zone, the REARM process will be frequently interrupted due to the airframe being tilted into its suspension in the slightest wind-condition on the ground. This ONLY affects the REARM process, the REFUEL process will not be interupted by the same airframe tilt in the very same conditions.
The surface condition is not relevant as the suspension rumble-tilt happens on all available surfaces.
The application of brake power (permanent press of keybind) is not relevant as it does not affect / does not hinder / does not stop the aiframe suspension fumble-tilt.
The intensity of the tilt is not relevant as the slightest presence of a wind condition and the resulting slightest of tilts aborts/interupts the REARM process (orange techchat message).

Local client hardware, ping, FPS, packet fragmentation aso are also not relevant as the interrupt is a local condition and not performance related.

The bug would affect all narrow-gear airframes but in realiter currently only affects Spitfires (Mk V, Mk IX tested, Mk XIV not tested).

As only the REARM process is affected, and the very same aiframe tossing/suspension tilt/externally caused airframe movement does neither affect the REFUEL process (both in player-agency as keybind-triggered and keybind-aborted/controled), nor the automated REPAIR process (no player agency, starts with engine off) a higher movement tolerance trigger / timeout on movement for the REARM abort seems advizable if no other solution is feasible.

Bug actively observed throughout entire first quarter of 2021, first observation in 2020 after introduction of RRR iterations.

WORKAROUND (temporary):

repeat REARM process / reinitiate until "REARM COMPLETE" message appears (in itself questionable, as the process is supposed to mimic the unloading and reloading of ammunition and cocking, cleaning, servicing of guns, thus the second by second step-progress being possible at all for guns as a byproduct of the as-of-yet nucleonic payload player agency feels off in itself)

Additional assets (videos, screenshots, logs):
-

Your PC config data (OS, drivers, specific software):

not bug-relevant or interdependent but dxdiag attached

DxDiag.rar 15.42 kB · 0 downloads

addendum:
Trackfiles showing the bug behavior and potentially the RRR trigger itself bouncing the airframes.
Third trackfile also showing the canopy repair bug (different bug, already reported elsewhere but still present) on top if canopy is jettisoned by mistake in flight or on the ground - it cannot be replaced, only respawn solves the issue:
https://drive.google.com/drive/folders/1_5Yqdv79PuzXkSY-hjvLCPKIm9DYYOMc?usp=sharing

@-DED-Rapidus if you download these files, please short forum DM so I can delete them, thank you in advance

Edited May 10, 2021 by [_FLAPS_]RogoRogo

May 10, 2021

@[_FLAPS_]RogoRogo,thanks!

July 20, 2021

The mass and drag values on the 50 cal AP appear to be incorrect.

I have created a more historically accurate representation of the 50 cal m2 ap round.

Historic 50 Cal AP.zip

I’m attaching the files as zipped folder, that’s compatible with mod managers.

These changes do not attempt to balance the round in terms of game play. I’ve simply updated the 50 cal ap shell file, to reflect the data in historical documents. The changes are as follows.

The range penetration table has been updated. The damage values have been updated in accordance with the game standards.

The muzzle velocity has been updated for the 36 inch aircraft MG.

The drag has been reduced,

The mach where peak drag occurs has been reduced.

The mass of the round has been increased.

All the changes are based on primary source data for the round. The primary source materials for my mod are.

“Terminal ballistics data, volume III, bombs, artillery, mortar fire & rockets.” Office of the chief of Ordnance,

https://cgsc.contentdm.oclc.org/digital/collection/p4013coll8/id/2374

“Report 710 / 466, An Analysis of Firing of Cal. 50 AP Ammunition Against Homogeneous Armor Plate” Zener, Watertown Arsenal 1942

https://apps.dtic.mil/sti/pdfs/ADA953652.pdf

“Report 620, Aerodynamic Data for Spinning Projectiles” H.P. Hitchcock Ballistic Research Laboratories, Aberdeen Maryland 1947.

https://apps.dtic.mil/sti/pdfs/AD0800469.pdf

“Report 478, Ballistic Coefficients of Small Arms Bullets Of Current Production”, Karpov, Ballistic Research Laboratories, Aberdeen Maryland 1944

https://ntrl.ntis.gov/NTRL/dashboard/searchResults/titleDetail/AD491936.xhtml

“The Erosion of Guns Part Two: The Characteristics of Gun Erosion.” Burlew, National Defense Research Committee. 1942

https://apps.dtic.mil/sti/pdfs/ADB280242.pdf

“AAF Manual 200-1, Fighter Gun Harmonization”, Headquarters Army Air Forces, Washington DC 1945.

https://www.avialogs.com/trainingmanuals/army-air-forces/item/56042-aaf-200-1-manual-for-fighter-gun-harmonization

“War Department Technical Manual TM 9-225. Browning Machine Gun Caliber .50 AN-M2, Aircraft Basic” War Department, Washington D.C. 1947.

http://www.nj7p.org/Manuals/PDFs/Military/TM%209-225%2028-Jan-47%20Google.pdf

“TM 9 1907 Ballistic Data Performance Of Ammunition” War Department, Washington D.C. 1948

https://archive.org/details/TM91907BalisticDataPreformaceOfAmmunition

“Discrepancies Associated With The Drag Characteristics of Primary Fragments”, Tatom, 2010

https://apps.dtic.mil/sti/pdfs/ADA532300.pdf

“The Aerodynamic Characteristics of .50 Ball, m33, API, M8 and APIT, M20 Ammuntion”, McCoy, Ballistic Research Laboratory, Aberdeen, Maryland 1990

https://apps.dtic.mil/dtic/tr/fulltext/u2/a219106.pdf

l’ll explain my changes below.

The mass of the round has been increased.

The game currently has the mass of the shell at .0419 kg. This is too low. We have 3 separate figures for the mass of the round. They vary from 710 grains to 708 grains. BRL report 478, notes that the core of the round was changed in 1943 resulting in a reduction of the weight of the round.The Air Force uses a weight of .1015 lbs or 710.5 grains. The Ordnance Department goes with a weight of 708, Hitchcock in BRL 620 notes the sampled round is 709 grains. We’ll error conservatively and settle on a weight of 708 grains .0458kg

The muzzle velocity has been updated for the 36 inch aircraft MG.

The Game has the muzzle velocity set to 840 mps (2755.91 fps). The Air Force manual 200-100, gives the muzzle velocity of the AP round as 822.96 mps (2700 feet per second). Across multiple sources, the Department of Ordencance reports the muzzle velocity of the 36 inch barrel AN/M2 is 865.63mps (2840 fps). This is from the 50 cal AN/M2 aircraft machine gun manual issued by The Ordinance Department.

http://www.nj7p.org/Manuals/PDFs/Military/TM%209-225%2028-Jan-47%20Google.pdf

The muzzle velocity is also listed as 2840 FPS,in “Terminal ballistics data, volume III, bombs, artillery, mortar fire & rockets.” and “TM 9 1907 Ballistic Data Performance Of Ammunition”. “The Erosion of Guns, Part Two: The Characteristics of Gun Erosion” quotes the velocity as 2800. So why did the Army Airforce list the velocity as 2700 in their manuals?

“The Erosion of Guns Part Two: The Characteristics of Gun Erosion” gives us the answer.

Thus, half way through its service life, the gun would be expected to shoot at 2700 fps, due to erosion of the barrel and other parts. The Air Force is likely calibrating their sights, harmonization patterns and ballistic tables based on the performance of a well worn 50 cal, half way through it’s service life.

Therefore I’ve set the muzzle velocity to 864 mps. My justification is that this resembles a new gun and therefore follows the game’s precedent of simulating new / factory fresh specifications. I believe this to be in good faith, as none of the weapons seem to have their muzzle velocity lowered to their expected half life levels.

Changes to the Ballistic Table, Penetration and Damage.

The Watertown Arsenal Penetration tests, Report 710, are within .02 inches of The Ordnance Department's calculated penetration results presented in “Terminal ballistics data, volume III, bombs, artillery, mortar fire & rockets.” and “TM 9 1907 Ballistic Data Performance Of Ammunition”. Therefore I have based the penetration table on The Ordnance Department’s charts.

The damage stats of the guns in game are calculated by comparing the kinetic energy of a round to that of the German 20mm AP round at 100m. The German 20mm does 1000 damage to an unarmored target at this range. The formula to compute the damage of the 50 cal at a range of 100 meters works out to:

Damage = (KE 50 cal / KE German 20mm) * 1000

Where KE = .5* Mass*(V^2)

The German 20mm has a mass of .0117 kg and at 100m has a velocity of 642mps. The German 20mm has a KE of 24.11 Kilo-joules at 100 meters. The 50 cal AP, traveling at 821 mps with a mass of .0458 has a KE of 15.43 kj.

The damage of the 50 cal at 100 meters is (15.43/24.11)*1000 = 639.9, I’ve rounded up to 640.

The damage decreases due to range and armor follows the same scale as the original 50 cal rounds.

Changes to Velocity and Drag.

The Drag coefficient was changed from .45 to .4.

The Velocity where maxim drag occurs was lowered.

The Velocity of the round at 1, 100, 500 and 2000 meters has been changed.

The current drag coefficient of the round .45 more closely resembles the G5 ballistic coefficient. A G5 ballistic coefficient is not the same as the coefficient of drag, Cd / Cx. Where Cd = 2*Force of Drag / Density*Velocity^2*area. The G5 ballistic coefficient is a way of relating how much drag a given bullet has compared to a reference model. The two drag notations cannot be used interchangeably to compute a ballistic profile for a bullet.

However, the G5 drag profile can be used to compute a ballistic table based on the Siacci Method. This was how ballistic profiles were calculated to about the 1950’s when computational methods became available. The range velocity table implemented in the mod is based on the Sicacci Method. The table is based on a muzzle velocity of 863.9mps, a projectile weighing .0458kg with a G5 ballistic coefficient of .458. Resulting in the following range to velocity table

Range	Velocity
1	839.9
100	821
500	660
2000	295

The G5 profile of .458 is based on data from “Terminal ballistics data, volume III, bombs, artillery, mortar fire & rockets.”

And, “Report 478, Ballistic Coefficients of Small Arms Bullets Of Current Production”. This document, provides a computed G5 ballistic coefficient based on firing a round between two sensors. The report states that C5 and G5 functions are numerically indistinguishable.

Therefore using a G5 ballistic coefficient of .458 is a valid approach to creating a new ballistic table utilizing the Sicacci Method.

Changes to Cx, coefficient of drag.

Since the Ballistic Coefficient is not the coefficient of drag, Cx was changed to .4

Hitchcock does provide us with a drag chart in “Report 620, Aerodynamic Data for Spinning Projectiles”. However, the notion at the time was KD which is different from Cd / Cx the game uses.

Hitchcock’s KD = (4/Pi) Force *Drag / *air density* bullet diameter^2 Velocity^2.

So we have to convert it to Cd /Cx. McCoy gives the conversion factor as Cd/ Cx= (8/pi) * Kd

In “The Aerodynamic Characteristics of .50 Ball…”

The methodology is also backed by Tatom in “Discrepancies Associated With The Drag Characteristics of Primary Fragments”,

Using the data in Hitchcocks chart and converting it to Cd / Cx

Or Cx =.162* (8/pi)

Cx = .41

Hitchcock’s drag table tells us that the Peak drag occurs at ~ 1.2 mach. Therefore

Mmax=1.2

These results validate with McCoy’s 1990 testing of ballistically similar rounds. Therefore I’m comfortable with these parameters.

I believe this implementation of the 50 cal AP round to be historically accurate and inline with methodology used by the developers. I’ve tried to keep from having to interpret or guess by a heavy reliance on primary sources.

The changes to bullet_usa_12-7x99_ap.txt

//////	PhysicsBody properties
VisualImage=0,"graphics/ammo/ammoRmg.mgm",false
visualradius=4
SoundScript=""
ImageAttrs="IA_NOMINPIXELS","IA_NOROTINTERPOLATION","IA_NOCLIP"

//CollisionBody="graphics/ammo/ammo.col"

NoDirecion=true
NoCollision=true

// дальность, скорость на этой дальности (на 0м - указать на 5км/ч меньше расчетной), пары броня-урон за броней. Броню больше, чем в первой паре, не пробьет
// за базовый взят урон на дальности 100м при нулевой броне, на макс. броне урон равне 0.25 от базового, на 0.8 макс. брони урон равен 0.5 от базового
// урон при нулевой броне уменьшается с дальностью по соотношению квадрата скоростей
// скорости по дальностям взяты из лога утилиты расчета .bin файлов пули (расчет таблиц наведения), на 10км/ч меньше ресчетной
// т.к. если скорость окажется больше расчетной на данной дальности, то дамага не будет
// базовый урон расчитан по соотношению кинетических энергий на дальности 100м между данной пулей и референсной пулей SHELL_GER_20x82_AP (=1000)

Armor=1,863.9,	28.9,249,	18,479,	0,708
Armor=100,821,	24.8,234,	16,437,	0,640
Armor=500,660,	17.7,169,	11,291,	0,414 
Armor=2000,295,   6,142,	3,165,	0,188
// аэродинамические коэфициенты подобраны исходя из совпадения графика скорости по дальности для эталонной пули (посчитано на балл. калькуляторе)

Gage=12.7				//калибр
DefaultStartSpeed=864	//дефолтная начальная скорость при стрельбе из простых пушек (наземка)
MaxDistance=2500		//предельная дальность, дальше которой объект удалится
Mass=0.0458				//масса пули
Mkr=0.97				//значение числа Маха, после которого начинается повышение Cx до момента M=1, зависит от формы тела но не от размера
Mmax=1.2				//значение числа Маха, до которого завершается понижение Cx, начавшееся при M=1, зависит от формы тела но не от размера
Cx_0=0.16				//значение к-та сопротивления при (M <= Mkr), зависит от формы тела но не от размера
Cx_max=0.4				//значение к-та сопротивления при (M = 1), зависит от формы тела но не от размера
Cx_1=0.32				//значение к-та сопротивления при (M >= Mmax), зависит от формы тела но не от размера
R=0.006475				//радиус пули
LiveTime=7.0			//время жизни объекта (большее из двух: время полета на MaxDistance или время горения трассера + 3.5с)
TimeToDestroy=0			//время жизни объекта после попадания в скрытом состоянии, что бы трейл трассера не пропадал пока не расствориться
TracertShowtime = 3.63	//время горения трассера
MaxRedirections = 2		//максимальное кол-во рикошетов
RedirectedLifeTime = 1.5	//время жизни после первого рикошета
DestroyOnExplosion=false	//используется только для бронебойных пуль (default =true) с целью обеспечить применение Hit..Effect вместо Hit.., при сохранении пробоя препятствия на вылет
							//что в свою очередь требуется для "прицепления" эффектов попадания по движущейся технике, а так же гарантированной передаче попаданий в сетевой игре
BeamStopDistance = 0.10	//порог расстояния, проходимого пулей после первого попадания в конвекс с "^", свыше которого будет устанавливаться декаль "пробития"
NetworkDecals=true		//true - декали попаданий будут передаваться через сеть и отображаться на фантомах

debug=false

DefaultBulletSpeed = 864.0		//дефолтная начальная скорость при стрельбе из комплексной (авиационной) пушки неопределенной модели
BulletSpeed = "M2.50", 864.0	//начальная скорость при стрельбе из комплексного (авиационного) пулмета УБ
BulletSpeed = "M2.50 45in", 895.0	//начальная скорость при стрельбе из комплексного (авиационного) пулмета УБ

Edited September 6, 2021 by Yak_Panther
links update

July 21, 2021

The dispersion of the US 50 cal is also to low. This is not about the harmonization patterns. The discrepancy I'm pointing out is about the base accuracy of the weapon. This could be caused by either the value being set to low or a bug with implementation of the barrel heating effects.

I have implemented a mod that models the accuracy of the 50 cal AN-M2 as stated in the Air Force manuals. It is attached below as zip in a mod manger friendly way. It comes bundled with the my velocity and drag changes.

Historic dispersion + pen 50 AP.zip

This mod adds .2 degrees to the dispersion due to gun heating. Resulting in 2.75 mils of dispersion after a 50 round burst. The implementation is:

Changed mg_usa_m2-50.txt

BulletDispertionAngleData = 0.0, 0.6, 700.0, 3.5

To

BulletDispertionAngleData = 0.2, 0.6, 700.0, 3.5

In Detail Below.

In game all the aircraft with the 50 cal have the dispersion set to .038 degrees, or .6632 mils in their files.

CarriageAdditionalBulletDispertionAngle = 0.019100955, 0.03820191

The weapon file mg_usa_m2-50.txt, can add up to an additional .6 degrees of dispersion to the round.

BulletDispertionAngleData = 0.0, 0.6, 700.0, 3.5 // base VO 0.1 thousand range cold, 1 thousand hot, corrected in vepon modes 


// The law of increasing the angle of dispersion of bullets (projectiles) relative to the axis of the barrel of small arms, depending on the absolute temperature of the barrel:
// the first number is the minimum (theoretical) angle of dispersion of bullets (shells) relative to the axis of the barrel of small arms at the absolute temperature of the barrel: Т = 0 ° K = -273.15 ° С, [°], (> = 0)
// second number - the angle of dispersion of bullets (projectiles) relative to the axis of the barrel of small arms at the absolute temperature of the barrel specified by the third number, [°], (> = first number)
// the third number is the absolute temperature of the small arms barrel, at which the second number specifies the scattering angle of bullets (shells) relative to the axis of the small arms barrel, [° С], (> -273.15
// fourth number - the degree of dependence of the angle of dispersion of bullets (shells) relative to the axis of the small arms barrel from absolute temperature of the barrel of small arms, [dimensionless], (> = 1)

However this only occurs when the barrel is 700 degrees. This is where there maybe a bug. The amount of dispersion per degree of barrel heat is scaled from -273 to 700. The dispersion only increases .00061 degrees for every 1 degree that the barrel heats up. The current implementation does result in a 4 mil dispersion circle, but only after the gun fires 165 continuous rounds. This dispersion is probably to low.

According the weapon manual for the 50 Cal AN-M2, after a 165 burst the accuracy would be seriously degraded, EG the dispersion would more than 4 mil 75%.

It looks like there is either a scaling bug due to the unit of measurement or the "fix" noted in the file is a typing error. Perhaps CarriageAdditional in the aircraft file is supposed to be .38 not .038

CarriageAdditionalBulletDispertionAngle = 0.019100955, 0.3820191

After a 165 burst the change noted above would result in 10 mil circle. What the Air Force might call "serious key-holing". A 50 round burst would result in a 8.3 mil circle. The 100% dispersion circle of the AN-M2 quoted in the manuals. The 50 round burst is about the median recommended burst length due to cook off concerns. Depending on how the dispersion is implemented, .38 may also be to high. This would depend on whether the dispersion value is the mean dispersion, the 1 sigma of Gaussian distribution, or simply the max of a random value.

Sources for the Dispersion of The 50 Cal AN-M2.

In "Air Force Manual No. 64 Fighter Gunnery Rocket Firing Dive Bombing" The dispersion is listed as 4 mil 75% /

8 mil 100%.

https://drive.google.com/file/d/1wyygNo7LGMlHQ74k16CL4XhTuBvROMvH/view?usp=sharing

The same measurement of accuracy for the 50 cal is listed in AAF manual 335 25

https://books.google.com/books?id=81krAQAAMAAJ&pg

Additional Thoughts:

The dispersion rating used for US weapons, is based on a probability density function. It is given in a report about the accuracy of the 25mm M242C gun. Using it allows us to calculate the standard deviation of the rounds in mils.

https://apps.dtic.mil/dtic/tr/fulltext/u2/a193618.pdf

The Diameter of the Dispersion in mils is equal = 2 * the standard deviation of the round from the impact point in mils * the square root of Chi square value of the %

The 50 cal AN-M2 rated at 4 mil 75% would look like

4 mils = 2 * (1.2 the Standard Deviation of the round in Mils)*√2.773

The rating 4 mils 75% is saying that the standard deviation of the round is 1.2 mils. or ~ .067 degrees.

Based on the notes in files, I built a model of how I think the dispersion due to gun heating works. This is along with the primary source data is how I arrived at my figures. The model is available here.

https://docs.google.com/spreadsheets/d/1R9M0ESAPI_amWheTN25fItMNCNLnPVSUFLGhXkjVz4k/edit?usp=sharing

Edited July 22, 2021 by Yak_Panther

July 24, 2021

Hi! great information! i like first part regarding ballistics of a gun. you increased bullet weight and improved ballistics according to real data. however in that part of script there is a possibility to change round type

shellType = 8
//        SHELL_TYPE_CV = 1,// Бронебойный - CV - Armour Piercing with Cavity Bursting Charge
//        SHELL_TYPE_AP = 2,// Бронебойный сплошной - AP - Armour Piercing Solid
//        SHELL_TYPE_HV = 3,// Подкалиберный - HV - Armour Piercing Composite Rigid (High Velocity)
//        SHELL_TYPE_HT = 4,// Кумулятивный - HT - Hight Explosive Anti Tank (HEAT)
//        SHELL_TYPE_SH = 5,// Шрапнельный - SH - Shrapnel
//        SHELL_TYPE_HE = 6,// Осколочный - HE - High Explosive and Fragmental
//        SHELL_TYPE_SpHV = 7,// Подкалиберный катушечный - SpHV - Armour Piercing Spool-Shaped (Spool High Velocity)
//        SHELL_TYPE_MG = 8,// Пулемёт

from 8 i would change to 2 AP. if this makes difference in action?

but i don't think second part is correct. you missed crucial aspect that cone dispersion were measured for a single gun with a 100 rounds burst. in short 4 mils is enough accuracy on average shooting short bursts. also i think dependency accuracy vs heat is not linear.

July 26, 2021

@ronick, who is your question addressed to?

July 26, 2021

24.07.2021 в 20:57, ronick сказал:

from 8 i would change to 2 AP. if this makes difference in action?

3 часа назад, -DED-Rapidus сказал:

@ronick, who is your question addressed to?

to someone, who knows what is under hood. or to someone who tested both types. from my shooting experience with new ballistics and shelltype =2 AP it kills crew much more often.

July 26, 2021

i let myself to share a few thoughts on dispersion and accuracy. current model in game is too accurate and proposed by Yak_Panther here alternative is too inaccurate, i think.

1. from different sources it is clear that initial accuracy for gun mounted on aircraft is approximately 1mil for short burst and cold barrel(depending on gun and mount type)

but shooting on the ground!

2.we do not have additional flying vibrant aircraft dispersion coefficient (lets say +1mil more dispersion)

3. we don't need current theoretical temperature of 0 k = -273C where accuracy supposed to be perfect with zero dispersion. we already know practical 1mil value for initial burst of a gun+mount, as a reference accuracy.

4.barrel temperature should be equal to environment temperature to calculate heat effect correctly.

5. we also know from references that 50rounds is good accuracy safe recommended continuous burst with necessity to cooldown for 40seconds

6. we also know that continuous burst of 100rounds gives 8mil total discrepancy. so somewhere after 50rounds there is a rapid drop in accuracy due to overheat.

just thoughts aloud for better shooting model.

July 26, 2021

1 hour ago, ronick said:

i let myself to share a few thoughts on dispersion and accuracy. current model in game is too accurate and proposed by Yak_Panther here alternative is too inaccurate, i think.

1. from different sources it is clear that initial accuracy for gun mounted on aircraft is approximately 1mil for short burst and cold barrel(depending on gun and mount type)

but shooting on the ground!

2.we do not have additional flying vibrant aircraft dispersion coefficient (lets say +1mil more dispersion)

3. we don't need current theoretical temperature of 0 k = -273C where accuracy supposed to be perfect with zero dispersion. we already know practical 1mil value for initial burst of a gun+mount, as a reference accuracy.

4.barrel temperature should be equal to environment temperature to calculate heat effect correctly.

5. we also know from references that 50rounds is good accuracy safe recommended continuous burst with necessity to cooldown for 40seconds

6. we also know that continuous burst of 100rounds gives 8mil total discrepancy. so somewhere after 50rounds there is a rapid drop in accuracy due to overheat.

just thoughts aloud for better shooting

Could you post the sources for some of the claims please?

July 26, 2021

some refs to soviet aircraft lagg and yak are already in luascripts/worldobjekts/planes

accuracy does not exceed 1,2mil at 100meters for short burst.

some info for 50cal

https://fas.org/man/dod-101/sys/ac/equip/m296.htm

July 27, 2021

Are you aware, that you posted stuff that is used nowadays? On different platforms, different ammo and mounting?
I stopped reading in the moment i saw XM and OH58.... prototype system on heli, comparing to mounting guns into 2ww planes.
Or comparing russian MGs with russian mounting in russian plane to US ones...
We are still waiting for solid sauce.

Edited July 27, 2021 by =DMD=Honza

July 28, 2021

On 7/24/2021 at 10:57 AM, ronick said:

Hi! great information! i like first part regarding ballistics of a gun. you increased bullet weight and improved ballistics according to real data. however in that part of script there is a possibility to change round type

shellType = 8
// SHELL_TYPE_MG = 8,// Пулемёт

from 8 i would change to 2 AP. if this makes difference in action?

but i don't think second part is correct. you missed crucial aspect that cone dispersion were measured for a single gun with a 100 rounds burst. in short 4 mils is enough accuracy on average shooting short bursts. also i think dependency accuracy vs heat is not linear.

On 7/26/2021 at 5:49 AM, ronick said:

i let myself to share a few thoughts on dispersion and accuracy. current model in game is too accurate and proposed by Yak_Panther here alternative is too inaccurate, i think.

1. from different sources it is clear that initial accuracy for gun mounted on aircraft is approximately 1mil for short burst and cold barrel(depending on gun and mount type)

but shooting on the ground!

2.we do not have additional flying vibrant aircraft dispersion coefficient (lets say +1mil more dispersion)

3. we don't need current theoretical temperature of 0 k = -273C where accuracy supposed to be perfect with zero dispersion. we already know practical 1mil value for initial burst of a gun+mount, as a reference accuracy.

4.barrel temperature should be equal to environment temperature to calculate heat effect correctly.

5. we also know from references that 50rounds is good accuracy safe recommended continuous burst with necessity to cooldown for 40seconds

6. we also know that continuous burst of 100rounds gives 8mil total discrepancy. so somewhere after 50rounds there is a rapid drop in accuracy due to overheat.

just thoughts aloud for better shooting model.

You are incorrect about the accuracy of gun.

The specification for the weapon states, a 10 round bust shall have a 100% dispersion of 6.66 mils (8 inches at 100 feet) when fired from a rigid mount.

If you have a primary source to back your claims, please provide it.

The particular weapon you linked to,

https://fas.org/man/dod-101/sys/ac/equip/m296.htm

has an even greater dispersion. 8 mil 80%

http://everyspec.com/MIL-SPECS/MIL-SPECS-MIL-G/MIL-G-71047_44025/

You are misunderstanding the difference between the mean / average and the standard deviation.

Here is a good article describing Standard Deviation.

https://www.mathsisfun.com/data/standard-deviation.html

The standard deviation of the round in one axis is 1.2 mils.

This is not the same thing as the mean aka the average dispersion.

The standard deviation, also know as sigma σ. Is a measure of the variation within the data set. Technically, it is the square root of the variance. In this case, The standard deviation, describes the dispersion relative to the mean. It is the spread of the spread.

Lets say we're testing the accuracy of 2 guns, A and B, using only two shots.

Gun A's first round hits dead center. 0 error. The second shot misses the target by 10 mils.

The average / mean dispersion is 0+10 / 2 = 5 mils

The Standard Deviation is 5 mils

Gun B's first shot misses the target by 5 mils. The second shot also, misses by 5 mils.

The average / mean dispersion is 5 + 5 / 2 = 5 mils of dispersion.

The Standard Deviation is = 0.

The standard deviation is just telling you how "random" the gun is. Not how accurate it is.

The value, 1 sigma = 1.2 mils, is standard deviation in only one axis, (X or Y). It does not even represent the total standard deviation. Since the gun is assumed to disperse equally in both the X and Y axis. The complete standard deviation of the gun is 2.4 mils.

as 1.2 * 2 = 2.4

2 sigma is how it's represented in the previously cited report.

Basing a weapon's accuracy on the 1.2 mil sigma value is literally halving it's standard deviation. Inaccurately modeling the standard deviation results in large errors if the rounds are distributed normally, Gaussian distribution.

If the dispersion is modeled with a normal / Gaussian distribution we would expect the 68 % of the bullets to impact with +- 1 standard deviation. If we incorrectly set the standard deviation to 1.2 mils, only 31% of the bullets would disperse more than 1.2 mils. If the Rounds are distributed normally with a standard deviation of 1.2 mils there is a .0004 chance that the rounds disperse more than 4 mils. The 6.6 mils 100% 10 round burst military specification and the Air Force specification of 4 mils 75% both show how absurdly low a 1.2 mil dispersion is. Dispersion greater than +-4.8 mils is mathematically impossible if the standard deviation is 1.2 mils and the impacts are distributed normally. Since the gun is rated 6 mil 100% for a 10 round burst or 4 mil 75%, we know the round deviates more than 4.8 mils. Hence a dispersion of 1.2 mils is incorrect.

The 2.4 mil standard deviation agrees with the 10 round burst 6.6 mil spec, because with a standard deviation of 2.4 mils 99.994% of the rounds land within 6.66 mils, +-2.775 sigma.

The real weapon has 2.4 mil standard deviation. If the impacts are distributed normally. With the real gun, we would expect 84% of the rounds to land within +-3.4 mils and 15 % of the bullets to impact outside the 3.4 mil diameter.

My mod adds + .2 degrees of dispersion. This sets the baseline dispersion at 3.4 mil. About where 84% of where the total rounds disperse to. The mod more closely models the standard deviation of the real weapon with an accuracy rating of 4 mil 75% or a 10 burst with a 100% dispersion of 6.6 mils. This is closer to reality than the base game. Or, an implementation which models a standard deviation of 1.2 mil.

Edited July 28, 2021 by Yak_Panther
math, spelling

July 28, 2021

Hi

1. you are right on standard deviation versus total cone dispersion.

most likely they use standard deviation in code (i checked description) but questionable!

2. you are not right on the rest?

3. forget about m296 and its rigid mount. we don't know how rigid it was . for use on helicopters most likely. we have p38 shooting test that gives more value.

4. let's calculate resulting cone dispersion with proposed by you model and compare it to p38 table

carriage dispersion angle(coded in each plane) = 0.0382 (for browning 50) + your bullet historic dispersion coefficient 0.25 = 0.2882 degrees = 5.1mil = 10 mil cone for initial burst!? and goes up to 22.6mil cone over gradual heat (according to existing code coefficients)

do you take in to account that accuracy degrades a lot over heat? manual does not recommend to shoot more than 50 rounds in a burst. and from p38 shooting table we can see that only after 100rounds cone is 8mil dispersion!

5. very short. i would set bullet dispersion coefficient(consider it is sigma) from .05 to .07 for cold barrel depending on gun and mount place(more vibration for wing mount probably?)

6. it is important to match heat/dispersion angle ratio (it is 3.5currently for any gun) but i don't know formula used.

7. and cool down coefficient. now gun recovers too fast most likely.

that's it i'm out. happy with .05625 initial bullet dispersion coefficient) thanks all for attention and collaboration!

July 28, 2021

2 hours ago, ronick said:

4. let's calculate resulting cone dispersion with proposed by you model and compare it to p38 table

carriage dispersion angle(coded in each plane) = 0.0382 (for browning 50) + your bullet historic dispersion coefficient 0.25 = 0.2882 degrees = 5.1mil = 10 mil cone for initial burst!? and goes up to 22.6mil cone over gradual heat (according to existing code coefficients)

That's not how the heat effects are applied. I tested it extensively and developed a model which I shared previously.

https://docs.google.com/spreadsheets/d/1R9M0ESAPI_amWheTN25fItMNCNLnPVSUFLGhXkjVz4k/edit?usp=sharing

With my mod, after 50 rounds the total dispersion is 2.7 mils.

2 hours ago, ronick said:

manual does not recommend to shoot more than 50 rounds in a burst. and from p38 shooting table we can see that only after 100rounds cone is 8mil dispersion!

You don't know what your talking about. That burst length recommendation is meant to prevent a cook off of the ammunition due to heating. It does not have to do with the accuracy degradation. Accuracy is reduced after a continuous burst of 165 rounds.

The accuracy, velocity, cook off, and jam / stoppage all have different burst lengths limits.

The burst length that degrades the accuracy is different from the burst length that results in a cook off or a jam. The weapon's manual notes that accuracy is only degraded after a 165 round burst.

The maximum burst length that can be fired before there is a cook off is 150 rounds.

The Manual also recommends other burst lengths and cooling times in order to prevent a cook off. Those are presented in this table

A burst length greater than 150 rounds will result in a jam.

2 hours ago, ronick said:

from p38 shooting table we can see that only after 100rounds cone is 8mil dispersion!

That chart isn't illustrating a increase in dispersion due burst length. 4 mil 75% and 8 mil 100% are the same rating!

The two ratings are mathematically equal.

4 mil 75%

4 mils = 2 * (1.2 the Standard Deviation of the round in Mils)*√2.773

8 mil 100 %

8 mils = 2 * (1.2 the Standard Deviation of the round in Mils) * √11.11

The 8 mil is only a 99.61% though.

2 hours ago, ronick said:

5. very short. i would set bullet dispersion coefficient(consider it is sigma) from .05 to .07 for cold barrel depending on gun and mount place(more vibration for wing mount probably?)

You still don't understand what sigma is.

A piston engine is much greater source of vibration, than any aero eleastic deformation or flutter that a wing would experience.

Just as an FYI. According to my model and some testing. A 50 burst with my mod results in a dispersion of 2.75 Mils.

A summary of the testing and development of that model is posted below.

from

a few separate files define the accuracy parameters of the gun.

The aircraft file located in: \luascripts\worldobjects\planes\

The weapon file located in: \luascripts\worldobjects\weapons\mg_usa_m2-50.txt

Dispersion Data for the M2 .50 cal is also located in

\luascripts\worldobjects\weapons\mg_usa_m2-50.txt

Where we find:

BulletDispertionAngleData = 0.0, 0.6, 700.0, 3.5 // base VO 0.1 thousand range cold, 1 thousand hot, corrected in vepon modes

More detailed instructions are found in the same directory \luascripts\worldobjects\weapons\guninstruction.txt

BulletDispertionAngleData = 0.2, 0.5, 700.0, 6.0 

// The law of increasing the angle of dispersion of bullets (projectiles) relative to the axis of the barrel of small arms, depending on the absolute temperature of the barrel:
// the first number is the minimum (theoretical) angle of dispersion of bullets (shells) relative to the axis of the barrel of small arms at the absolute temperature of the barrel: Т = 0 ° K = -273.15 ° С, [°], (> = 0)
// second number - the angle of dispersion of bullets (projectiles) relative to the axis of the barrel of small arms at the absolute temperature of the barrel specified by the third number, [°], (> = first number)
// the third number is the absolute temperature of the small arms barrel, at which the second number specifies the scattering angle of bullets (shells) relative to the axis of the small arms barrel, [° С], (> -273.15
// fourth number - the degree of dependence of the angle of dispersion of bullets (shells) relative to the axis of the small arms barrel from absolute temperature of the barrel of small arms, [dimensionless], (> = 1)

The M2 50 cal is

BulletDispertionAngleData = 0.0, 0.6, 700.0, 3.5

Based on the instructions we can say;

When the barrel temp is -273.15 ° С, there is zero degrees of dispersion, When the barrel gets to 700c degrees the dispersion is .6 degrees. The scaling coefficient is 3.5.

My first question was how does this scaling coefficient effect the dispersion pattern. To test I put a P-39 with 50 cal only at 3000 m in Kuban Fall map. I maxed out the coefficient high and low. First we increase the value a lot.

BulletDispertionAngleData = 0.0, 0.6, 700.0, 10000

This reduce the dispersion effects to almost nothing, until the guns reach over 700c. Once the gun temp is over 700c, the dispersion went to max dispersion and above as long as the trigger was held down. The bullets would return low dispersion when the gun was allowed to cool.

Next we change the value to it's lowest setting.

BulletDispertionAngleData = 0.0, 0.6, 700.0, 1

According the instructions, 1 is the lowest value the coefficient can have. Thus we set it to 1. The result: dispersion begins to increase before the barrel reaches 700c.

This value seems to effect how much dispersion is applied per unit of barrel heat. Thus it looks like we scale the dispersion from 0.0 to .6 degrees based on the barrel temp from -273 to 700c. With result divided by the scaling factor. It may end up something like this

Dispersion in degrees = ( Barrel Temp * 0.00061 / Scaling Factor)

I then wanted to see if the guns start out -237c or are they effected by the ambient temperature. To test we'll max the scaling coeffiect so that rounds are not effected by the barrel temp until the barrel temp is above the critical temp. Are test setting looks like this.

BulletDispertionAngleData = 0.0, 90, 15, 10000

The theory of operation here is, only when the guns get above 15 c the will they start dispersing wildly 90 degrees.

To test we put a P-39 on the ground in Stalingrad in the winter. Here we get ~3 shots before the guns start to disperse. Which makes sense because the guns heat up 4.659 degrees per shot. From our weapon file.

BarrelTemperatureIncreasePerShot = 4.659

We then test on the summer map. Here the 50 cals shoot wildly from the start, As they are ambient temp which is above 15 c.

I think we can say that guns start at ambient temp, Do I want to test the alt to temp gradient in the sim and see if that has an effect on these parameters. No but I think it's sufficiently modeled.

As noted each aircraft has a value that can increase the dispersion. It's defined as, and is loacted in that aircraft's lua file.

eg: \luascripts\worldobjects\planes\_

CarriageAdditionalBulletDispertionAngle = 0.019100955, 0.03820191"

A guide to this files contents is found in:

\luascripts\worldobjects\planes\_weaponmodesinstruction.txt

CarriageAdditionalBulletDispertionAngle = 0.0, 0.0 
// Additional angles of dispersion of bullets (shells) relative to the axis of the small arms barrel, due to the design features of the small arms carriage:
 // the first number is the additional scattering angle at the first shot in the queue, [°], (> = 0) 
// second number - additional scattering angle for the second and subsequent shots in the queue, [°], (> = first number)

This looks like it adds an additional few degrees of dispersion to our dispersion. Since I have no way to test this variable, I can't tell if it's added before or after the scaling factor:

EG: is it, this for the first shot:

BulletDispertionAngleData = 0.0191, 0.6, 700.0, 3.5

or this

Dispersion in degrees + .0191 = (Delta Barrel Temp * 0.00061 / Scaling Factor)

Your setting of:

BulletDispertionAngleData = 0.1, 0.6, 700.0, 3.5

On a standard day, at an alt 2400 meters, where the temp ~0c, Your mod may change the base dispersion to:

(0.00061 * 273 + .1) / 3.5 = 0.0761 degrees = 1.32 Mils

Accounting for the added dispersion from the the carriage effects the total dispersion may equal:

(Base dispersion.0761) + (Carriage effect .0382) = 0.114 degrees =1.99 mils

or if it calculated differently:

(0.00061 * 273 + .0382 +.1 ) / 3.5= 0.087 degrees = 1.51 mils

....

We'll take your dispersion setting and calculate the dispersion step by step.

BulletDispertionAngleData = 0.1, 0.6, 700.0, 3.5

From the previous post, we know what each of these variables is supposed to do. Now let walk through how they are applied to determine the dispersion angle.

Let's look at the previous example for these dispersion settings where the barrel temperature is 0c. We determined the base dispersion before the aircraft effects were 1.32 mil or .0761 degrees. It's important to note the variables the game uses are degrees for angular measurement and degrees of Celsius for temperature. The previous equation was.

(0.00061 * 273 + .1) / 3.5 = 0.0761 degrees = 1.32 Mils

The first number in our equation (0.00061) is the base amount of dispersion per degree of barrel temp. This is computed by subtracting the max dispersion from the min and dividing by the the total number of degrees that the dispersion happens over.

Dispersion Max - Dispersion Min / Absolute Value of Temprature Degrees Min and Max

From our variables above it looks like something like this:

.6-.1 / 973.15 = 0.000513

We know the min (.1) and max dispersion(.6). The total amount of dispersion that occurs is .6 - .1 = .5

The .5 dispersion happens over the temperature range of -273.15 to 700 degrees, or 973.15 degrees.

Thus the dispersion per degree of barrel heat is; .00051. As .5/973.15 =.00051

Meaning for every 1 degree rise in barrel temperature the dispersion increases by .00051 degrees. This is where I made my error in my previous post. I used a dispersion per degree of barrel temp of .00061, because I forgot to account for your .1 in this part of the calculation. We'll see how this effects the outcome going forward.

Onto the next part of the Equation: 273. This is the difference between the barrel temp and absolute zero. The difference between 0 degrees and -273 deg is 273. Which is why we use it to calculate the dispersion at 0 degrees c. If we we're checking at 20c we would use. 273 + 20 = 293 degrees to calculate the base dispersion.

We multiply the dispersion per degree (.00051) * Delta in Barrel temp from -273 (273) to give us the temperature modified base dispersion.

.00051 * 273 = .139 degrees of temp modified base dispersion.

Back to our equation, lets now correct it and apply the variables for your mod. The equation for your mod now looks like this.
(0.00051 * 273 ) +1 = .239

The .1, is the min dispersion value from the gun setting file. We have to add this to the temperature modified base dispersion( .139) because the gun starts with .1 degree of dispersion. So the net amount of temperature modified dispersion is .1+.139 = .239 degrees of net temp effected dispersion.

We then scale net temp dispersion by the scaling coefficient defined by gun parameters. In this case 3.5. I believe that we divide the net temp dispersion (.239) by the scaling factor (3.5). This gives us the total amount of gun dispersion

.239/3.5 = .0682 degrees of calculated gun dispersion.

{(0.00051 * 273 + .1)} / 3.5 = .0683

But we still need to account for the added dispersion from the aircraft carriage variable defined in the Aircraft.txt. It's .0382 so we just added it to total computed dispersion.

{(0.00051 * 273 + .1)} / 3.5 = .0683 + .0382 = total amount of dispersion. =.1065 degrees of dispersion. To convert degrees to mills, multiply by 17.777 which give us 1.89 mils

Thus

{(0.00051 * 273 + .1)} / 3.5 = .0683

.0683 + .0382 = .1065 degrees

.1065 * 17.777 = 1.89 mils.

Therefore, your mod puts out around 1.89 mils radius of dispersion when the guns are zero degrees.

Edited July 28, 2021 by Yak_Panther

August 1, 2021

Brief desription: With the Fokker D.VIII the colimator sights are reversed. Colimator day is actually colimator night and vice versa in the mission editor.

Brief description: The track recording system sometimes lables ground units wrong. Friendly trucks for example have a red instead of a blue lable.

Brief description: All Flying Circus planes repairs processes currently get interrupted before the repairs are completed. To work around it you have to move your throttle or restart and stop the engine again. It likes to get stuck on repairing cockpit and landing gear.

Edited August 1, 2021 by J99_Sizzlorr

August 2, 2021

13 hours ago, J99_Sizzlorr said:

Brief description: All Flying Circus planes repairs processes currently get interrupted before the repairs are completed. To work around it you have to move your throttle or restart and stop the engine again. It likes to get stuck on repairing cockpit and landing gear.

Hi, is there a mission or a ready-made example?(track, video recording)

August 2, 2021

10 hours ago, -DED-Rapidus said:

Hi, is there a mission or a ready-made example?(track, video recording)

Here you go: https://www.file-upload.net/download-14651259/Tracks.rar.html

In this case the hull did not get repaired completly the upper wing was still damaged, then it moved on to repairing cockpit and got stuck until i restarted the engine, then it repaired the engine and the hull fully but got stuck on repairing cockpit again. I can not make the techno chat appear in the track recording but maybe you can.

Edited August 2, 2021 by J99_Sizzlorr

August 12, 2021

Brief Description: (Only tested with Spitfire's) When in a shallow dive, bullets and cannon shells bounce off of the surface of the ground when they shouldn't... ie. when impacting grass or earth. It makes sense that this should happen sometimes when firing at concrete or tarmac, but for me it happens most of the time regardless of surface.

Detailed Description, Conditions: Rhineland map and Kuban map, Spitfire MkXIV with machine guns deleted (just 2 x 20mm cannon) and Spitfire MkIXe with standard armament (2 x 20mm cannon, 2 x .50 cal machine guns)

PC Config: Windows 10, IL-2 4.602 and latest update 4.603 (what I mean is that the issue is the same before and after most recent update)

August 13, 2021

@Algy-Lacey,

August 13, 2021

@-DED-Rapidus Thanks for posting those videos,

The pilots in the videos are aiming at a metal target and therefore when I saw those ricochets I assumed that they were bouncing off the target, not the background terrain.

But in discussion on another thread some folks shared their experiences of live firing a machine gun at targets on varying terrain including grass and sand and they saw ricochets like we see in game, I am now convinced that what we have in game is closer to reality than I had assumed.

It still does seem that, compared to the videos, too many of the bullets ricochet in game, but that is a subjective perspective.

Happy Landings!

Algy-Lacey

September 5, 2021

On 7/22/2021 at 12:28 AM, Yak_Panther said:

The dispersion of the US 50 cal is also to low.

Sorry, but you reading the config data wrong.

It's CEP assigned in configs, not full dispercion circle.

So while source data is 8 feet full dispercion circle diameter on 2000 feet range - CEP is 8/2000/3/2=0.667e-3 parts of range (which is ~0.667 mils).

This what is configured: tan(0.03820191)=0.667e-3

On 7/22/2021 at 12:28 AM, Yak_Panther said:
Changed mg_usa_m2-50.txt
BulletDispertionAngleData = 0.0, 0.6, 700.0, 3.5 
To
BulletDispertionAngleData = 0.2, 0.6, 700.0, 3.5 

So you have added tan(0.2)=3.5e-3 to CEP. So you have incresed CEP of .M2 in 6 times over the reference ?

Guys, I'd recomend you to study what is done better and may be ask us why it's done so before doing somethimg in your amazing mods - I'm sure this would be wize.

September 5, 2021

9 hours ago, Han said:

Sorry, but you reading the config data wrong.

It's CEP assigned in configs, not full dispercion circle.

So while source data is 8 feet full dispercion circle diameter on 2000 feet range - CEP is 8/2000/3/2=0.667e-3 parts of range (which is ~0.667 mils).

This what is configured: tan(0.03820191)=0.667e-3

So you have added tan(0.2)=3.5e-3 to CEP. So you have incresed CEP of .M2 in 6 times over the reference ?

Guys, I'd recomend you to study what is done better and may be ask us why it's done so before doing somethimg in your amazing mods - I'm sure this would be wize.

I have a few questions:

1. Are the values input into the Variable BulletDispertionAngleData in degrees? The notes seem to indicate that they are.

BulletDispertionAngleData = 0.0, 0.6, 700.0, 3.5 
// base VO 0.1 thousand range cold, 1 thousand hot, corrected in vepon modes 
// The law of increasing the angle of dispersion of bullets (projectiles) relative to the axis of the barrel of small arms, depending on the absolute temperature of the barrel:
// the first number is the minimum (theoretical) angle of dispersion of bullets (shells) relative to the axis of the barrel of small arms at the absolute temperature of the barrel: Т = 0 ° K = -273.15 ° С, [°], (> = 0)
// second number - the angle of dispersion of bullets (projectiles) relative to the axis of the barrel of small arms at the absolute temperature of the barrel specified by the third number, [°], (> = first number)
// the third number is the absolute temperature of the small arms barrel, at which the second number specifies the scattering angle of bullets (shells) relative to the axis of the small arms barrel, [° С], (> -273.15
// fourth number - the degree of dependence of the angle of dispersion of bullets (shells) relative to the axis of the small arms barrel from absolute temperature of the barrel of small arms, [dimensionless], (> = 1)

2. Is the output degrees

3. How does BulletDispertionAngleData interact with the variable. Is it added to it?

CarriageAdditionalBulletDispertionAngle = 0.019100955, 0.03820191

4. If these input are degrees, are they then transformed into CEP, then output as a degree of deviation from the gun barrel?

5. When you say "CEP" are you referring to Circular Error Probable?

https://en.wikipedia.org/wiki/Circular_error_probable

If so, using the Root Mean Square (RMS) method The 50% CEP is:

CEP 50% ~ 1.774 * Sigma

For the 50 cal sigma would be 1.2 milliradians in each axis. there fore Sigma =2.4

CEP 50% = 2.4*1.774

CEP 50% = 4.08 miliradians, or .23 degrees.

The 99.7 % CEP would be= 3.408 * 2.4

99.7 % CEP = 8.1792 mils = 0.46948162

Edited September 5, 2021 by Yak_Panther

September 5, 2021

Hello,

I'm understand that digging the configs is a great deed and fun, but let's discuss not deep internal things of how game code is working but the troubles you see in game.

Anyway, CEP circle - is a circle where 50% of hits are located. In normal dispersion it's ecual to 33% of radius of full dispersion circle.

So if full dispersion is 8 feet in diameter (at 2000ft), than CEP radius is 8/3/2=1.33333 feets at 2000 ft range

atan(1.33333/2000)=0.038197° - you see it in config (a bit differ while exact full dispersion source was a bit differ).

So may be trouble that there are differ dispersion data for differ planes - possible. But we need it not for all guns, but for one specific. Is your data 74 inch at 750ft for ONE gun?

Your source on P-38H says 71.3in (5.916667ft) full dispersion at 750ft range.

It's means CEP angular radius =atan(5.916667/3/2/750)=0.075333°

It's twice larger than we have for one .50 in game. But it's twice larger, not 6 times.

So here is our old source we used for .50 in BoM (while creating P-40E).

So you see 4mils dispersion cone in text and 8ft at 2000ft drawn full dispersion circles (as they can be seen).

May be these circles are not full dispersion, but 75% hits circle, than CEP should be greater yes. But not 6 times )))

September 5, 2021

Amazing information here

Han thanks for taking the time to explain some of it.

It's beyond my understanding.

But it's very interesting!

September 5, 2021

1 hour ago, Han said:

Hello,

I'm understand that digging the configs is a great deed and fun, but let's discuss not deep internal things of how game code is working but the troubles you see in game.

Anyway, CEP circle - is a circle where 50% of hits are located. In normal dispersion it's ecual to 33% of radius of full dispersion circle.

So if full dispersion is 8 feet in diameter (at 2000ft), than CEP radius is 8/3/2=1.33333 feets at 2000 ft range

atan(1.33333/2000)=0.038197° - you see it in config (a bit differ while exact full dispersion source was a bit differ).

So may be trouble that there are differ dispersion data for differ planes - possible. But we need it not for all guns, but for one specific. Is your data 74 inch at 750ft for ONE gun?

The test was conducted with one gun installed in the nose of the P-38.

However it was the standard off accuracy used by the Air Force for sighting and harmonization on every fighter aircraft past the 1950's

On 7/21/2021 at 2:28 PM, Yak_Panther said:

The same measurement of accuracy for the 50 cal is listed in AAF manual 335 25

https://books.google.com/books?id=81krAQAAMAAJ&pg

The Circles indicate where 75% of the rounds would impact in those charts. They are the 4 mil circles.

Wouldn't CEP 50% be half the total dispersion?

Shouldn't the probability for the bullets to deviate be equal in the X and Y axis? In which case it CEP 50% would be = ( 8/3/2=1.33333 ) *2

They way you're describing it, half of a half of the dispersion modeled.

Just in case it's not clear, I'm suggesting the 50 cals are to accurate.

Edited September 5, 2021 by Yak_Panther

September 5, 2021

edit: sorry, double posted. I'll leave it so the link to the document is there for you.

Hi @Han

Hope you're good and thanks for taking the time to look into this topic.

I'll leave the maths side of things to you and @Yak_Pantherbut from a reference perspective, this the document we were working off of.

https://drive.google.com/file/d/1wyygNo7LGMlHQ74k16CL4XhTuBvROMvH/view

There are references to the 75%, 4mil pattern and on page 67 (in document) there is an example dispersion pattern of a single gun being fired from a P38.

Edited September 5, 2021 by ACG_Cass

September 5, 2021

Hi!
btw - is there something related to .30 or .303, 20 and 37mm of same kind exists about dispersion?

September 5, 2021

1 hour ago, Yak_Panther said:

Just in case it's not clear, I'm suggesting the 50 cals are to accurate.

2 hours ago, Han said:

May be these circles are not full dispersion, but 75% hits circle, than CEP should be greater yes. But not 6 times )))

September 5, 2021

Let me see what I can dig up @=FB=VikS

Dispersion data is quite difficult to find. There are some secondary sources on older forum posts I can find, but unfortunately the original source links don't work (404 error). Example: https://theairtacticalassaultgroup.com/forum/showthread.php?t=9642

I'll see if I can find the reference data that was used.

September 6, 2021

Did some digging around, Trying to find anything in an official capacity was quite hard. There's substantially more data on the American settings then the RAF

These are the little bits and pieces I've found so far. They may not be 100% accurate unfortunately and I can't verify anything yet.

The RAF tried box convergence (They referred to it as Dowding Spread)

But went back to point convergence, the mantra with the machine guns was to shoot very close, some pilots had their convergence set to 110 yards.

Data for .303 API
https://sites.google.com/site/britmilammo/-303-inch/-303-inch-incendiary?authuser=0

Data for .303 AP, they list some penetration values also.
https://sites.google.com/site/britmilammo/-303-inch/-303-inch-armour-piercing?authuser=0

Data for .303 ball
https://sites.google.com/site/britmilammo/-303-inch/-303-inch-ball---other?authuser=0

20mm cannons

Interesting things I had found,
Early Spitfires minimum cannon convergence range was 300m
Early Hurricanes (and possibly later ones also) had a minimum convergence of 400m for the outer guns

Some data on Hispano 20mm penetration, muzzle velocity and a few types of AP they had
http://www.wwiiequipment.com/index.php?option=com_content&view=article&id=96:hispano-20mm-armour-piercing-ammunition&catid=44:gunsrockets&Itemid=60

The Soviet pilots seem to have a fair bit of freedom for preferred gun harmonization. I didn't have much luck locating anything. I suspect for the iL-2 point convergence at 300-350 yards would have been the norm. I find closer then that to be troublesome as your window for strafing is very limited.

Edited September 6, 2021 by Denum

Armament and equipment

Recommended Posts

-DED-Rapidus

Supercharger

=FEW=Revolves

JG4_Widukind

-DED-Rapidus

JG4_Widukind

Stonehouse

[_FLAPS_]RogoRogo

[_FLAPS_]RogoRogo

-DED-Rapidus

[_FLAPS_]RogoRogo

-DED-Rapidus

Yak_Panther

Yak_Panther

ronick

-DED-Rapidus

ronick

ronick

354thFG_Panda_

ronick

Honza

Yak_Panther

ronick

Yak_Panther

J99_Sizzlorr

-DED-Rapidus

J99_Sizzlorr

Algy-Lacey

-DED-Rapidus

Algy-Lacey

Han

Yak_Panther

Han

Denum

Yak_Panther

ACG_Cass

=FB=VikS

Han

ACG_Cass

Denum

Create an account or sign in to comment

Create an account

Sign in

Browse All

Activity