Russian guns balistics...

[JtD]

I'm fairly certain that they used different sources for different guns, and not even necessarily test results. I've attached the two sources I consider most likely to have been used for the ShVAK and the 151/20 AP rounds. They agree very well with in game velocities, matching time to 500m to the second digit in both cases with my test. It shows that the devs have again made their homework, in that they do base the game performance on historical documents.

 

 

But then you're right with what you say - the data might refer to different conditions. I haven't figured out what conditions exactly. However, if the data is correct, the MG151/20 AP round produces 20% more drag than the ShVAK round used for the table. Looking at the shapes of the rounds, it is very hard for me to believe that this was indeed the case. If someone could plug in the data of the two projectiles into some drag calculation software or came up with actual firing trials for the ShVAK, it would be most helpful.

Edited September 19, 2017 by JtD

[BraveSirRobin]

Question of the day

When was the last time you were "sniped" a German?

Why do people have so much trouble understanding that 109 and 190 drivers don't have to "snipe". They can just wait until they're closer before they shoot.

[Gunsmith86]

"SOVIET CANNON - A Comprehensive Study of Soviet Guns and Ammunition in Calibres 12.7mm to 57mm" is the result of more than ten years of research by Christian Koll.

 

He writes below the table that german test made under the same conditions lead to OF projectile 790 m/s ,...

 

http://russianammo.org/Russian_Ammunition_Page_25mm.html

 

This helps to clear the diffrent speeds:

815 m/s is for long barrel engine cannon and 790-800 m/s for short barrel wing and mounted cannon.

Edited September 19, 2017 by Gunsmith86

[unreasonable]

An amazing amount of data there! 

 

Just one comment - in the extract page you show it states that the propellant is "18g of VT powder".  If that is true for all types of warhead, then the MVs cannot possibly all be 800m/s in RL.

 

Which just prompts me to ask a question, which I am sure has been answered before but I have forgotten (do not get old )  - in game are there different trajectories for different ammo types, including the effect of tracers? Or do we get a single outcome per gun type?

[Gunsmith86]

An amazing amount of data there! 

 

1)Just one comment - in the extract page you show it states that the propellant is "18g of VT powder".  If that is true for all types of warhead, then the MVs cannot possibly all be 800m/s in RL.

 

2)Which just prompts me to ask a question, which I am sure has been answered before but I have forgotten (do not get old )  - in game are there different trajectories for different ammo types, including the effect of tracers? Or do we get a single outcome per gun type?

1) Yes they are all listed with 800 m/s which cant be true because of diffrent shell weight. most likely they didn´t care if they had exact values for the diffrent type of ammunition in every case. The text right belowe the table in the first picture supports this.

 

2) Don´t know its realy hard to tell if we see just spread or diffrent trajectories for different ammo types whitout a software that would show as which type of round we have fired in game.

[JtD]

Thank you very much for the information, Gunny! I find the German figures a lot more plausible than the manual figures I found in other places.

 

WRT the charge, I've seen different figures in different manuals, all were pointing to the same amount of powder no matter the shell type. The figures went up to 19.2 in a ShVAK manual.

 

The game takes different round types into account. But as far as the ShVAK goes, I'm pretty sure we only get the modern K-6 fuse, as far as I understand the earlier fuses were sorted out before or early in the war. They tended to cause explosions when still in the barrel. If so, all rounds weight around 96g, with only little differences in ballistics.

And for ~96g rounds, the given ~760m/s would seem to be a reasonable muzzle velocity to me.

Edited September 19, 2017 by JtD

[Fern]

If we set +500m as the threshold for "sniping" I would say  the day before yesterday. Trying to run away from a 109 in an La-5F at low altitude (had expended my ammo) I was gunned down from around 600m with relative ease (meaning, he didn't fire more than a couple of bursts before I had an oil leak). It was on a "normal" server, so I'm fairly certain about the distance. I've been hit by a German pursuer from much further away than that, especially when flying the MiG, but this is the most recent example.

 

Nothing wrong with that really. I presented an easy target to hit, flying straight and level, so naturally he took the shot.

 

As I've said now repeatedly: The reason you don't see German pilots taking these hail-Mary shots more often is because they don't have to.

 

I guess I should of used a different term. Dead six, a russian would of killed you. 

 

When was the last time you were "INSTAKILLED" by a German?

[Venturi]

Venturi, I will just address one point from your extraordinary rant, since it sums up your whole problem of lack of reading comprehension.

 

 

Well we can just agree we don't like each other.

As far as I'm concerned, you can't even take a formula from Wiki and translate it honestly.

Edited September 20, 2017 by Venturi

[-=PHX=-SuperEtendard]

"SOVIET CANNON - A Comprehensive Study of Soviet Guns and Ammunition in Calibres 12.7mm to 57mm" is the result of more than ten years of research by Christian Koll.

 

He writes below the table that german test made under the same conditions lead to OF projectile 790 m/s ,...

20mm shvak and B-20.jpg

 

http://russianammo.org/Russian_Ammunition_Page_25mm.html

 

This helps to clear the diffrent speeds:

815 m/s is for long barrel engine cannon and 790-800 m/s for short barrel wing and mounted cannon.

20mm shvak cannon.jpg

 

So by the chart in the webpage link, and the info in that image we would have say for the I-16 Type 24 the earlier OF/OFZ with 91g and 790 m/s, then the BZ at 96g and 750 m/s? Finally for the ShVAKs for the 1942/1943 period (Yaks, LaGGs, MiGs, etc) we would have the OZ-OZT at ~770-760 and the already mentioned BZ at 750? The text doesn't mention the BZT tracer version, but it's in the chart and we have in game an AP tracer round when firing the ShVAK in slow motion.

 

Nothing easier than that - 100% hit probability (if you wish to compare it for 50%, 75% etc. hit probability, you need to convert it first, usually with Gaussian distribution) dispersions for MG FF, MG 151/20 etc. 

 

Take note that the report notes that "The dispersion of the MG 151/20 weapons is.... extremely small" , i.e. at 1,9 mils for 100% dispersion (dunno what to say about the MG FF's shocking 1.0 for 100%, but hey, its Swiss after all).

 

MG151-20_realitycheck.png

So that report says that the maximum amount of disperison for the MG 151/20, (which I interpret as the area which contains 100% of the bullets fired) is 1,9 mils? And is 1,9 mils as the whole cone or one side from the center?

[unreasonable]

@Venturi - As per not liking you, in general,  I have no opinion since I do not know you. For all I know you are a fine member of your local chapter, and kind to old ladies.  What I do not like is the way you try to make discussions of technical subjects into personal issues.

 

On the subject of the formula:

 

 

 

If you assume the coefficient of form (i) is 1 initially, you can calculate BC for each shell type.  Hispano is higher. (Higher BC is good, other things being equal).  You can then see what i must be to make the shells come out with the same BC: the Hispano's shells would have to have a much higher i than the ShVAK.  (High i is bad......).  

 

Given the shape of the shells, my initial thought would be that actually the ShVAK's have a higher i, but the main point is that it seems highly unlikely to me that the Hispano's shells have a value for i 1.3 or more times that of the ShVAK. 

 

But as I have said, these are my expectations based on my current beliefs. If you have data that shows differently, please share.

 

For everyone else, a well written, clear article about the significance of BC and i can be found here.  

 

http://www.bergerbullets.com/form-factors-a-useful-analysis-tool/

Edited September 20, 2017 by unreasonable

[Finkeren]

I guess I should of used a different term. Dead six, a russian would of killed you.

 

When was the last time you were "INSTAKILLED" by a German?

 

I thought we were talking about Soviet guns being "lasers" with perfect accuracy? Now we are back to the idea, that they somehow also do magical amounts of damage?

 

I have been "instakilled" by Germans with just one or two hits plenty of times (that's the beauty of the DM - if just one bullet hits the right spot, you're going down), but I don't know if it has been from +500m recently.

 

You know what I haven't experienced either? Killing a German opponent with a single lucky shot from long range.

I have made a few kills this way (when no other option was available) but it has always wasted a collosal amount of ammo with repeated bursts, scoring 1-2 hits, observe the effect, shoot some more.

[unreasonable]

There are so many things going on in these anecdotes that you have to try to isolate the effects one at a time, as JtD has done in his shooting range:

 

1)Trajectory 

 

2) Spread/grouping

 

3) Damage potential per hit - based on shell type

 

4) Damage effect - based on target type and location

 

So I think it is better to leave damage out of this particular discussion - if it turns out that velocities are wrong that can affect damage, but you cannot work back from the damage to say that the velocities are wrong.

 

And that is completely leaving out the problem that MP has no turbulence or wake effects, IIRC, and people behave in odd ways. Even if 1-4 were modeled in a way that exactly matched the real world, I expect there would still be more long range shooting success than seems plausible.  

[Sturmovnik_PL]

I guess I should of used a different term. Dead six, a russian would of killed you. 

 

When was the last time you were "INSTAKILLED" by a German?

I always shoot down in that same way, insta pilot kill, or cut of my wing or tail.

[Venturi]

It says 'sufficiently small', to be exact. Thanks for posting this data. What's also interesting to note is that 'dispersion increases with barrel length, because weapons with larger muzzle velocities, due to their longer barrels, develop a tendency for barrel vibration'.

 

That's exactly the opposite of what we occasionally could hear on the forums, i.e. the Soviet guns would be more precise because of their longer barrels. It would be interesting to know how they dealt with that and how that is considered in dispersion figures. One would at least have to look at the exact mounting of the gun for any dispersion figure.

 

It is much more complex than that. Dispersion from the barrel flexion is from a harmonic vibration / waveform along the barrel. increased barrel length only adds to the harmonics if the stiffness of the barrel is not increased, and it changes the harmonic frequency. If one manages to find a harmonic "node" along the length of the barrel where the vibration is cancelled, and end the barrel right there, there will be no dispersion induced by barrel length. An alternative or supplementary way of decreasing this effect is to stiffen the barrel by adding wall thickness to it. This increases the weapon's weight rapidly however.

 

Automatic weapons usually, but not always, have greater dispersion than manually-operated weapons. For instance, the gas-cycling on a M-16 combined with a rotary bolt removes all inaccuracy involved with automatic weapons with tipping bolt-design, like the Kalashnikov. In many automatic weapons dispersion is not an unfavorable trait. For instance, a M-16 (or M249 SAW which also fires the NATO 5.56x45 cartridge with similar bolt design) will have a 2-5 minute of accuracy ANGLE at 100m, whereas the typical Ak-47 will be in the range of 5-8 MOA. In any case such dispersion is relatively minor compared to the dispersion induced by other factors, below. 

 

Much more important to dispersion than the inherent accuracy of the weapon, is the weapon's mounting. Any flexion of the weapon's mounting during firing will induce a far larger dispersion than the inherent dispersion of the weapon. Any shifting of the aiming device vis-a-vis the mounting of the weapon will induce dispersion. Any vibration through the airframe which differentially shifts the aiming device vis-a-vis the weapon mounting will induce dispersion. Additionally with weapon groups, these effects operate on each weapon independently from the others, inducing pattern dispersion.

 

The overall sum effect of all this is much greater than any inherent dispersion in the weapon itself.

 

Many of the weapons as modeled currently are far too invariable in their dispersion.

Edited September 21, 2017 by Venturi

[Guest deleted@83466]

So, in laymans terms?

[JtD]

Venturi says that inaccuracies of gunfire due to flexibility of aircraft structure and gun mountings could be more important than the inaccuracy of the gun as such. The only figures I have seen on this subject with the WW2 guns are for the Browning 0.50M2 in various installations on the B-17 and B-24, and it's absolutely true there.

 

For instance, if you fly a high-g turn and your wings are under a lot of stress (and therefore bending/twisting), there should be a noticeably increased error with wing mounted guns. Same way if your engine is running rough and the plane's shaking due to that.

Or if the mounting of the gun is poorly done, and the gun keeps vibrating inside the mount (due to the plane vibrating or due to the gun firing). Which would be the case for instance in different gun stations of the above mentioned bombers.

Edited September 22, 2017 by JtD

[1stCL/Werner]

Data which I have saved for long time.

 

 

Here are the results of a lenghty research into dispersion data, which is quite hard to come by...
The data is based on 100% diameter dispersion with 1 mil = 1/1000th of rad, the kind of mount is precised next to the weapon. 75% dispersion diameter is supposed to be half the 100% diameter which seems quite true for most weapons, this value is provided when quoted in the source (M2 data for instance).

We can clearly see the impact of the wing mounting compared to engine mounting, the later seems to have absorbed recoil and vibration much better... indeed dispersion is at least 2 times greater with wing mounted weapons.
Engine mount are the most efficient but nose mounting or cowling mounting does not provide the same amount of precision the mount being much more prone to vibration it seems.

Note that US data on the M2 is confusing since the reference data comes from a P-38 nose mounted M2, but the US manuals use the same dispersion data for wing mounted weapons. Either the P-38 mounts are really up to no good or the manuals make a wrong assumption when it comes to wing dispersion. I tend to believe the later, i think the wing mounted M2 would have had a dispersion of at least 12mils and probably more.

If you quote this data on other sites/bbs please precise the source being AAW. TIA

H means Height (or max dispersion diameter) as i previously used vertical and lateral dispersion values.

D means distance.

Units are metric.

German Weapons 
-----------------------
MG-17 Cowling mounted (Bf 109F-2 / Bf 109F-1 actual tests)
H = 0.60 / 0.8 m
D = 100 m
R/D = 60/10000 80/10000
= 6 mils / 8 mils

MG-131 Cowling mounted (Fw 190A - theorical max)
H = 1m
D = 100m
H/D = 100/10000
= 10 mils

MG-151/15 Engine mounted (Bf 109F-2 actual test)
H = 0,35 m
D = 100 m
H/D = 35/10000
= 3.5 mils

MG-FF Engine mounted (Bf 109F-1 actual test)
H = 0,2 m
D = 100m 
H/D = 20/10000
= 2 mils (very tight patern)

MG-FF Wing mounted (Bf 109E-3 actual test)
H = 0,35 m
D = 100m 
H/D = 35/10000
= 3.5 mils

MG 151/20 Engine mounted (Bf 109G-6 - theorical max)
H = 0.3m
D = 100m
H/D = 30/10000
= 3 mils

MG 151/20 Wing mounted - inner (Fw 190A - theorical max)
H = 0.7m
D = 100m
H/D = 70/10000
= 7 mils

MG 151/20 Wing mounted - outer (Fw 190A - theorical max)
H = 0.8m
D = 100m
H/D = 80/10000
= 8 mils

MK 108 Engine mounted (Ta 152 - therorical max)
H = 0.35
D = 100m
H/D = 35/10000
= 3.5 mils

Allied Weapons
------------------
M2 Nose mounted P-38 (USAAF 1944 Gunnery manual)
H = 1.88 m
D = 229 m
H/D = 188/22900
= 8.2 mils (75% = 4.1 mils)

Hispano 20mm Nose mounted P-38 (USAAF 1944 Gunnery manual)
3 mils 75%
6 mils 100% assumed

Edited September 22, 2017 by 1stCL/Werner

[Holtzauge]

Data which I have saved for long time.

 

Good collection, thanks for posting. Since no source is listed, I can add one which leads credence to the German values anyway: The Ta 152 flight manual (Flugzeughandbuch Ta 152 H-0/H-1 Schusswaffenanlage teil 8A) page 37, states the following allowable dispersion for the engine mounted Mk108 in ground testing: "100% of the rounds for an 11 round burst shall hit within H=35 cm and B=30 cm at 100 m range." (My translation).

 

That's pretty darn good considering it's 100% of the rounds. It would be reasonable to assume a normal distribution meaning most rounds land pretty much spot on at 100 m. OTOH I think this indicates that we have too good long range accuracy in-game for "some" cannon. At least that's what it seems like to me.

 

Which brings us back to the OP's question about Russian ballistics: We now have some data on the dispersion of German engine mounted cannon but AFAIK none on the Russian. However, I'm going to be radical here and suggest that there is no reason to assume that the Russian dispersion was any better. So, if this leads to a similar adjustment of German cannon as was done in roll for the Me-109 while the Russian planes were left with the Fw-190 like roll performance due to lack of "historical data" I will be disappointed.

 

So IMHO, if no solid Russian dispersion data comes up, I think it would be more reasonable to adjust the accuracy of both the German AND Russian engine mounted cannon based on the data we have, i.e. stop the tack driving at 800m. Grabs hat and starts running......

[Venturi]

3mils or 30cm at 100m is about best case scenario for that data. 

 

More representative based on that data, would be:

 

Engine mounted weapons - 5mils (50cm at 100m)

Cowling mounted weapons - 8mils (80cm at 100m)

Wing mounted weapons - 6mils (60cm at 100m)

 

Which makes sense based on the rigidity of the mounting. Wing structure will flex but will induce less dispersion than cowling mounts, which are not as firmly anchored to the axial skeleton of the aircraft. Engine mounts will be best and probably represent the inherent dispersion of the weapon. **edit- read as "best possible mounted dispersion, or baseline dispersion for aircraft based weapons of this type.** Note also that weapons which are dispersed widely along the wing have differing dispersion rates (outer having more dispersion). I would expect weapons grouped tightly together in the wing to have much less dispersion vis-a-vis each other - in other words, less group dispersion.

 

Keep in mind distance will at least proportionally increase the dispersion, and it may be greater than proportional. So for a weapon with 5mils dispersion, at 100m, 50cm dispersion. At 200m, 100cm dispersion. at 300m, 150cm dispersion. The reality is probably worse than that for several complex reasons.

 

This illustrates why having more guns is a virtue in and of itself.

Edited September 22, 2017 by Venturi

[bivalov]

Han posted this page during one of old discussions on russian forum - https://forum.il2sturmovik.ru/uploads/monthly_02_2015/post-12-0-70383500-1422957355.jpg

and said, that they have similar docs for all other soviet planes, and dispersion increases in the following order -

1) engine
2) cowling
3) wing
4) underwing
5) turret

[Holtzauge]

Han posted this page during one of old discussions on russian forum - https://forum.il2sturmovik.ru/uploads/monthly_02_2015/post-12-0-70383500-1422957355.jpg

 

and said, that they have similar docs for all other soviet planes, and dispersion increases in the following order -

 

1) engine

2) cowling

3) wing

4) underwing

5) turret

 

Thanks for the info. The list order looks just like you would expect it I think.

 

About the dispersion figure, my Russian is rather sketchy, but does it mention the test method? Is the dispersion pattern collected from single shots or from bursts? I did not see any reference to a burst duration but I may have missed that.

[Brano]

That picture is for MP-20 gun = ShVAK long barrel for engine installation. Airplane LaGG-3.

D=100m

Firing single shots,20 shots in total.Engine not running.

[Holtzauge]

OK, thanks for the translation! Then it looks like the ShVAK and the MG151/20 have similar inherent accuracy: Earlier on in this thread the MG151/20 was listed at 1.9 mils and from bivalovs picture above the ShVAK has 1.8 mils. Granted, AFAIK the condition for the MG151/20 figure are unknown but given that the MG151/20 engine mounted is listed at 3 mils above I'm assuming the 1.9 mils is the inherent accuracy.

 

However, AFAIK the list 1StCl/Werner posted above was accuarcy in bursts so that still leaves the question about the ShVAK's accuracy in bursts: Now asuming the ShVAK and MG151/20 have about the same inherent accuracy in single shots, what about in bursts? Since both were bolted to the engine I would expect them to be close but there could be differences in how the mounting was done I suppose. Engine mounts are usually dampened so when firing bursts the cannon/engine assembly would move around a bit and I suppose this could give a difference in accuracy depending on how this was done in German versus Russian planes.

Edited September 23, 2017 by Holtzauge

[Guest deleted@83466]

A very non-scientific "feeling" post:

 

I did what I said I was going to do, and flew a 109 online last night for the first time in months, and also for the very first time since the big patch.  Only a few sorties, so not much to conclude yet, but I will say that from the perspective of a mainly VVS pilot, I wish I had the firepower of the 109 in my Yak!   I didn't get the impression at all, that there was anything wimpy about the German guns compared to the VVS ones by any means, or that the VVS planes were too durable.  Like I said, just a very unscientific impression, based on just a couple of sorties in a plane that I don't consider myself up to speed on yet.  Ultimately the hard data that is being presented in threads like these will be the ultimate test to determine whether the developers got things right, but I'm just saying that for those who claim it is "obvious" that the VVS guns are overpowered, and the German guns are underpowered, I found nothing yet that endorses that impression on my own end.  I'll try to fly the 109 a lot more than I have been, and obviously I'll have to see over time how these feelings evolve or change.

Edited September 23, 2017 by Iceworm

[I./ZG1_Radick]

did you play IL2 1946 or IL2 BoS?

Edited September 23, 2017 by I./JG21-Radick

[unreasonable]

I suspect whether or not the different guns are right or not, one of the reasons that we see so many reports of German Mplayers complaining about being sniped, and so few Soviet Mplayers making the same complaint, is that the German pilots are much more likely to be flying away from their foes, thinking (often falsely, out of an exaggerated assessment of their aircraft's superiority) that they can extend or climb away.  I doubt that Soviet players do that very much, except perhaps the occasional La-5, since they know that it will not work.   So naturally they are rarely sniped - but shot down at close range.

 

It is impossible to conclude anything from personal impressions based on flying for a while - there are just too many variables.

[BraveSirRobin]

I suspect whether or not the different guns are right or not, one of the reasons that we see so many reports of German Mplayers complaining about being sniped, and so few Soviet Mplayers making the same complaint, is that the German pilots are much more likely to be flying away from their foes, thinking (often falsely, out of an exaggerated assessment of their aircraft's superiority) that they can extend or climb away.  I doubt that Soviet players do that very much, except perhaps the occasional La-5, since they know that it will not work.   So naturally they are rarely sniped - but shot down at close range.

Oddly, the German only people don't seem to want to acknowledge this explanation.

[Lusekofte]

Here you go, German sniper plane. Ace in one mission none the less with a HS 129 . 

 

 

https://forum.il2sturmovik.com/topic/31090-hs129-ace-flight/

 

 

 

is that the German pilots are much more likely to be flying away from their foes,

 

 

This is correct in my opinion and it is mentioned several places in this topic. But every time a 109 get shot down , it ends up in a FM discussion here about the opponent plane

[Holtzauge]

Some data about the sights: The German REVI AFAIK has this sight picture: The ring is 100 mils in diameter and the tick marks are at 1 deg in angle, i.e. about 17.5 mils. I measured the tick marks and I would say they are in the order of 6.3 mils in length.

 

So, since the accuracy of the MG151/20 in horizontal plane should be 3 mils, this means that if you fire a burst, all shells should as they fall along the vertical bar be no further away than a quarter of the bar width on either side. I did a few quick missions (Me-109G2, 1000 m alt at autopilot) and to me the horizontal spread in-game is larger than that but it would be good if some more people tried to see if they get the same results.

 

Next Russian LaGG: Here I believe the tick marks are at 10 mils and the ring at 140 mils: So it looks like the smaller tick is around 3.75 mils wide.

 

Using the same test procedure as for the Me-109, the LaGG shoots really tight patterns in the horizontal plane and it looks like to me the spread is more in line with the single shot no engine running data bivalov presented i.e. 1.8 mils since that is around half of the small bar width.

 

So to conclude: Looks like the in-game MG151/20 spreads more than the historical 3 mil for a burst and the LaGG has burst precision that mirrors the single shot without engine running test, i.e. very optimistic indeed.

[-=PHX=-SuperEtendard]

I did some shooting with the Bf 109 F-4 in auto level flight and recorded it in replay. Then played it in slow motion and took several pictures of the tracers (each picture when the last tracer of the previous picture disappeared). And since all of the pictures are with the same camera position and zoom level, I measured the 100 mil ring with pixels, then draw a vertical line 1,5 mils at each side of the center of the gunsight and counted how many rounds were inside and outside the 3 mils dispersion zone.

 

I measured 50 rounds in 7 pictures, because I think with 50 it would be enough to have a clear pattern, if needed I have 24 pictures in total.

 

Here are some of them, you can see there are cases with more and some with less dispersion.

 

 

 

So in the end I got from 50 rounds total, 13 were outside the 3 mil wide area, so 23% of the rounds have too much lateral dispersion.

 

Still I don't think the lateral dispersion being much of a player in long range shooting, because as I showed earlier, even at 700 meters the dispersion is less than the plane's wingspan. If my calculations are correct a 10 meter wingspan plane at 700 meters would be around 14 mils wide. I think the vertical ballistic drop is more important when shooting with center mounted weapons.

 

[VesseL]

 

...

Still I don't think the lateral dispersion being much of a player in long range shooting, because as I showed earlier, even at 700 meters the dispersion is less than the plane's wingspan. If my calculations are correct a 10 meter wingspan plane at 700 meters would be around 14 mils wide. I think the vertical ballistic drop is more important when shooting with center mounted weapons.

 

Thanks for the test. I think so too,the vertical drop is much more important. Is the vertical drop where it should be?

Edited September 24, 2017 by VesseL

[Venturi]

I did some shooting with the Bf 109 F-4 in auto level flight and recorded it in replay. Then played it in slow motion and took several pictures of the tracers (each picture when the last tracer of the previous picture disappeared). And since all of the pictures are with the same camera position and zoom level, I measured the 100 mil ring with pixels, then draw a vertical line 1,5 mils at each side of the center of the gunsight and counted how many rounds were inside and outside the 3 mils dispersion zone.

 

I measured 50 rounds in 7 pictures, because I think with 50 it would be enough to have a clear pattern, if needed I have 24 pictures in total.

 

Here are some of them, you can see there are cases with more and some with less dispersion.

 

 

 

So in the end I got from 50 rounds total, 13 were outside the 3 mil wide area, so 23% of the rounds have too much lateral dispersion.

 

Still I don't think the lateral dispersion being much of a player in long range shooting, because as I showed earlier, even at 700 meters the dispersion is less than the plane's wingspan. If my calculations are correct a 10 meter wingspan plane at 700 meters would be around 14 mils wide. I think the vertical ballistic drop is more important when shooting with center mounted weapons.

 

Yes, and what is the 95% confidence interval for the German burst dispersion data? In other words, what is the standard deviation for shot dispersion.

I have the same question for the Russian data. 

 

For all tests, to see where your 95% confidence interval is for the burst data, you should be counting how long you are holding the trigger for, because a hot and dirty barrel will result in increased dispersion. You should control this factor for all tests.

 

Until these questions are answered, all this testing is for basically nothing.

[-=PHX=-SuperEtendard]

Yes, and what is the 95% confidence interval for the German burst dispersion data? In other words, what is the standard deviation for shot dispersion.

 

 

Setting a mean, a standard deviation and a confidence interval can be done, but imho it's not needed considering we have the 100% dispersion number.

 

Besides, I don't think it's needed to consider the time between bursts because in game you don't get the overheat warning while flying at cruise speeds, even if you fire the entire ammo load of the plane in one burst, and you don't notice differences in dispersion. When firing with the plane on the ground then yes you have to watch out for gun overheating because it will pop in the technochat, and you will clearly notice increased dispersion and slower muzzle velocity as well. But I don't think when flying, the small differences in barrel temperature and dirt accumulation are modelled in the sim.

Edited September 25, 2017 by -=PHX=-SuperEtendard

[unreasonable]

Setting a mean, a standard deviation and a confidence interval can be done, but imho it's not needed considering we have the 100% dispersion number.

 

Besides, I don't think it's needed to consider the time between bursts because in game you don't get the overheat warning while flying at cruise speeds, even if you fire the entire ammo load of the plane in one burst, and you don't notice differences in dispersion. When firing with the plane on the ground then yes you have to watch out for gun overheating because it will pop in the technochat, and you will clearly notice increased dispersion and slower muzzle velocity as well. But I don't think when flying, the small differences in barrel temperature and dirt accumulation are modelled in the sim.

 

I agree that std dev etc is not needed here.

 

You may be right about the heat and fouling issues, but there is another issue with your test. If you look at the instruments when the F-4 is flying in level auto mode, you can see the bank indicator frequently moving slightly. If you go to external view you can see the rudder moving slightly as well - only about a pixel on my 4HD screen, but I am fairly sure I am not imagining it.  I assume that since the 109s have no aileron or rudder trim the AI has to make tiny adjustments to keep straight and level.  So the question is how many mils or parts thereof is the line of the aircraft moved during these adjustments? Perhaps not much - but if you want to say that dispersion is wrong because some of the shots fall a mil outside your range, you need to be sure it is the dispersion and not changes in the aircraft heading doing it.  

 

You might be able to make a mission with another a/c flying straight and level at a set range in front of you so that you can see the movement - if any - of the pipper relative to it's outline, (though given how tricky it is to make the ME produce anything exactly this might not be easy).

 

Anyway I agree with your substantive conclusion - the vertical drop over time is much larger than the increase in the group size, and much harder for the shooter to judge since it requires range estimation and knowledge of your gun's trajectory.  

[Brano]

There is a picture of MP-20 mounting in one of ShVAK operating manuals I have. Should be fixed in 3 points,one of them being described as "amortizator".

Maybe another point to mention. The barrel has been made from different type of material then usual. Not an expert on steel and steel alloys,but it was supposed to be some kind of softer/milder steel that could deal with heat and other stuff during firing better. The price to pay was shorter lifetime. There is also a complete set of material charts used for different parts of the gun in operating manual. I think JtD should have it. Im on business trip to Poland without access to the datas. It would be also nice to get hands on complete NII VVS test report,not only one page out of it. Danil has got it. We don't

[Holtzauge]

<snip>Still I don't think the lateral dispersion being much of a player in long range shooting, because as I showed earlier, even at 700 meters the dispersion is less than the plane's wingspan. If my calculations are correct a 10 meter wingspan plane at 700 meters would be around 14 mils wide. I think the vertical ballistic drop is more important when shooting with center mounted weapons.

 

First of all thanks for doing the tests. Good to see someone else conclude the MG151/20 engine cannon spread is more than the historical 3 mils.

 

However, I don't agree with yours and unreasonable's view that the too large lateral dispersion does not matter:

 

First of all even if the dispersion is still covering the span, this means that the weight of fire is diluted over a much larger area (square of the dispersion increase) meaning the chance of a hit is less. In addition, the value of hitting the wings is usually less than hitting the central fuselage in terms of damage to the target.

 

Secondly, I suspect that the in-game vertical dispersion is just as IRL coupled to the horizontal dispersion, meaning if the horizontal dispersion is more than IRL so is most likely the vertical which yet again makes hitting with a sufficient weight of fire harder.

 

Finally, I think this whole line of argumenation about if the spread is good or bad and if its useful or not since LW crates seldom shoot at 800 m is moot: If it's wrong it's wrong and should be corrected. Or else if it does not matter then why not set both the Yak and Me-109 engine cannon dispersion to 5 mils? Or should the Yak stay at around 1.8 mils? Wouldn't that then be an unfair "disadvantage" for the Yak if shooting tight groups was so bad?

 

Could use the same weird argument about the Spitfire top speed: If it's 20% lower in-game than IRL is no big deal: It can't outrun the Me-109 or Fw-190 anyway so that tactic is not used in-game ergo it's not a problem if it's 20% slow........

 

 

<snip> Im on business trip to Poland without access to the datas.<snip>

 

Well, if you could access the data when you are back that would be great.

[Brano]

Now I have access only to very good "wodka wyborova"...and waitress at hotel bar :D

[BlitzPig_EL]

Now I have access only to very good "wodka wyborova"...and waitress at hotel bar :D

 

Post of the day.

 

Be sure.

 

Enjoy.

[Brano]

Life is a question of priorities. She doesn't give a shait about ShVAK. Me neither :D

[unreasonable]

First of all thanks for doing the tests. Good to see someone else conclude the MG151/20 engine cannon spread is more than the historical 3 mils.

 

However, I don't agree with yours and unreasonable's view that the too large lateral dispersion does not matter:

 

First of all even if the dispersion is still covering the span, this means that the weight of fire is diluted over a much larger area (square of the dispersion increase) meaning the chance of a hit is less. In addition, the value of hitting the wings is usually less than hitting the central fuselage in terms of damage to the target.

 

Secondly, I suspect that the in-game vertical dispersion is just as IRL coupled to the horizontal dispersion, meaning if the horizontal dispersion is more than IRL so is most likely the vertical which yet again makes hitting with a sufficient weight of fire harder.

 

Finally, I think this whole line of argumenation about if the spread is good or bad and if its useful or not since LW crates seldom shoot at 800 m is moot: If it's wrong it's wrong and should be corrected. Or else if it does not matter then why not set both the Yak and Me-109 engine cannon dispersion to 5 mils? Or should the Yak stay at around 1.8 mils? Wouldn't that then be an unfair "disadvantage" for the Yak if shooting tight groups was so bad?

 

snip

 

I am not sure that SuperEtendard has concluded the first point correctly yet - it first has to be established that the axis of the aircraft is not moving in the flight tests.

 

Re vertical vs horizontal dispersion -  I am not saying horizontal dispersion does not matter.  Firstly, I was differentiating between dispersion - the group size at a given range, which includes the horizontal and vertical variance since I assume the group variance is circular - versus the trajectory which is the vertical position of the entire group over range, in which I assume there is no variance for a given shell type.  The second matters much more because you have to know two facts - range and trajectory - to compute the required elevation.  In contrast, irrespective of the group size, placing the pipper on the target's vertical line is always correct - in the case of the zero deflection shot the test is discussing. 

 

You say that the chance of a hit is less with a larger group. That is obviously true only in the case when the pipper is at exactly the right elevation and vertical line. People actually shooting will have an aiming error around the ideal position, and some shots that could not possibly hit with the optimum pipper position will score hits with a larger spread. We are back to the question of what should be the optimum spread for a shotgun cartridge.  In terms of what spread will give the maximum number of occasions on which at least one hit is scored, or the maximum number of hits over a sample of bursts, I am not convinced that the tightest possible spread is optimum at all.  But that can only be established if in addition to the gun data we have measured the actual aiming errors people make.  

 

The optimum group also depends on whether you are prioritizing frequent single hits, or rarer multiple hits. Given that the 109s allegedly fall apart after a single hit, it is not obvious that a very tight group for the ShVAK is an advantage at all in this sort of shot: although a too flat trajectory certainly would be, no argument there.   

 

I agree that if things can be identified as wrong they should be corrected - the problem is being sure what is wrong when in game testing is so tricky and the results are compared with historical documents that were not all compiled on the same basis.