Friday 31 May 2013

Accuracy Revisited

A while ago, I wrote an article comparing the accuracy of a few guns: the D-25, 88L/56, 88L/71, and ML-20. That article lacked a few details, which I will explain here.

The article talked about how the Germans calculated gun accuracy: using a percentage figure for how often you can expect to hit a 2 meter by 2.5 meter target at a certain distance. Such a measurement is useful for when you are fighting a tank that is 2 meters tall and 2.5 meters wide, and not really in any other case.

The Soviet measurement of accuracy is deviation. You will see many deviations in test documents, but the most important figure, the one that actually makes it into artillery tables and textbooks, is the expected deviation. The average deviation comes in 3 components: vertical, horizontal, and distance. For direct fire, vertical and horizontal are important, for indirect fire, distance and horizontal.

Expected deviation is similar to the 50% radius, except in only one direction. 50% of the shells will land on the closer side of the expected deviation figure, 50% of the shells will land on the further side. Using the two expected deviation figures, you can create an oval inside of which 50% of your shells will go.

Let's revisit the guns in the previous accuracy article. The D-25 has a 50% radius of 24 cm. However, you can see that the distribution of shots in that radius is not circular, it is an oval. I don't have the exact data for the D-25's average deviations, but you can clearly see that the result would be an oval that is a little shorter than it is wide. It is pretty trivial to deduce that the value of the deviations should be approximately 24 cm.

A non-circular result is very common. The ML-20 in that same article has a 50% radius of 32 cm at 1000 meters. However, the average deviations are different from that: 22 cm for vertical and 33 cm for horizontal.

Now, on to some new guns! Artillery tables contain expected deviations for lots of distances and propellant charges, but I will obviously not list all of them. The deviations will be given for 1000 meters, at a full charge, with as many shells as I have tables for. Deviations are taken from the appropriate Soviet artillery table. I have the full ones, so if you need them, ask. For each gun, I will also give the distance at which it can hit a German target 50% of the time with the regular AP shell, to compare it with the German accuracy metric.

Let's start with a classic, the 76 mm F-34 gun, mounted on the T-34-76 models 1941 and 1942:
  • HE shell OF-350, O-350A, and D-350 smoke grenade:
    • Vertical: 30 cm
    • Horizontal: 30 cm
    • Trend: horizontal deviation grows much slower than vertical deviation.
  • AP shell Br-350A and Br-350B
    • Vertical: 30 cm
    • Horizontal: 30 cm
    • Trend: horizontal deviation grows a little slower than vertical deviation.
  • HEAT shell UBP-353A
    • Vertical: 50 cm
    • Horizontal: 50 cm
    • Trend: deviations grow at the same rate.
Looking at a very extreme range, the F-34 has an average deviation of only 1 meter at 3 km. That means that the German 2x2.5 meter target would still be hit more than half the time. Jentz writes that the Tiger could hit the target 53% of the time at that range. Pretty comparable.

The 122 mm model 1938 howitzer (M-30) has the following deviations:
  • HE shells OF-462 and O-462A, smoke shell D-462
    • Vertical: 50 cm
    • Horizontal: 60 cm
    • Trend: vertical deviation grows a little faster than horizontal deviation.
The range at which the gun can hit the above German target 50% of the time is 2000 meters. 

The 45 mm model 1942 gun isn't quite that accurate (model 1932 and 1937 results are identical):
  • AP shell Br-240 and shrapnel O-240 and O-240A
    • Vertical: 50 cm
    • Horizontal: 60 cm
    • Trend: horizontal deviation grows faster than vertical.
  • APCR shell Br-240P
    • Vertical: 50 cm
    • Horizontal: 50 cm
    • Trend: horizontal and vertical deviations grow at the same rate.
The deviation gets to around 1 meter in each direction at 2000 meters, not that there is much a 45 mm gun can penetrate at that range. 

Here is a gun we haven't seen before: the T-60's ShVAK 20 mm autocannon:
  • AP incendiary and fragmentation-incendiary shells:
    • Vertical: 70 cm
    • Horizontal: 70 cm
    • Trend: vertical deviation grows faster than horizontal deviation.
The distance at which a German target has a 50% hit ratio is 1500 meters.

The 12.7 mm DShK was in a T-40 tank, that's tank gun enough for me!
  • B-32 AP bullet
    • Vertical: 76 cm
    • Horizontal: 59 cm
    • Trend: vertical grows faster than horizontal.
  • BZT-44 AP-incendiary bullet
    • Vertical: 98 cm
    • Horizontal: 67 cm
    • Trend: vertical grows faster than horizontal.
The 50% distance is also 1500 meters.

That's all I have for now on the Soviets, so now let's take a look at some foreign guns. The first up is the British 40 mm 2-pounder.
  • AP shell and AP shot
    • Vertical: 40 cm
    • Horizontal: 20 cm
    • Trend: horizontal deviation grows much faster than vertical deviation, despite starting out smaller.
This gun hits the German target 50% of the time at a distance of about 2250 meters.

Now, some American guns. First, the 75 mm M2, from the M3 Lee.
  • AP shells M61 and M72
    • Vertical: 30 cm
    • Horizontal: 30 cm
    • Trend: vertical deviation grows much faster than horizontal deviation.
  • HE shells M48 and MK1
    • Vertical: 40 cm
    • Horizontal: 30 cm
    • Trend: vertical deviation grows much faster than horizontal deviation.
The range at which this gun hits a German target 50% of the time is about 2250 meters.

The other gun on the M3 Lee, the 37 mm M5:
  • AP shell M51
    • Vertical: 40 cm
    • Horizontal: 40 cm
    • Trend: vertical and horizontal deviations grow at the same rate.
The range at which this gun hits a German target 50% of the time is a bit less than 2000 meters.

Next, the German guns. The 3.7 cm PaK gives the following results:
  • AP Pzgr
    • Vertical: 50 cm
    • Horizontal: 50 cm
    • Trend: vertical and horizontal deviations grow at the same rate.
  • HE Spgr
    • Vertical: 50 cm
    • Horizontal: 30 cm
    • Trend: vertical grows much faster than horizontal.
The gun loses enough penetration to not be of any use before it can fail to hit a 2*2.5 meter target 50% of the time, (the table only goes up to 1500 meters). 

The German heavy 105 mm model 1918 gun has the following deviations:
  • HE 19 KPS and FES (deviation given in meters)
    • Vertical: less than a meter
    • Horizontal: less than a meter
    • Trend: vertical deviation grows faster than horizontal deviation.
  • AP Pzgr. rot.
    • Vertical: 40 cm
    • Horizontal: 20 cm
    • Trend: vertical deviation grows much faster than horizontal deviation.
Due to the rapidly growing vertical deviation, it is not possible to accurately gauge the 50% distance. It would be somewhere just over 2000 meters.

In conclusion: the average deviation for most guns is more or less in the same range, 30-50 cm at 1000 meters. Even with the largest deviation (98 cm on the the worst machine gun bullet), a gun is going to hit any remotely tank-sized object more than half the time. Penetration is the issue here, as a lot of these guns can shoot much further than they can usefully penetrate anything.

The myth of superior German accuracy is, once again, proven false. The average deviation of the D-25 is comparable to the 88L/71, and the long-range accuracy of the F-34 is the same as the 88L/56.

Update: I found the complete artillery table for the 88L/71 KwK 43 gun. It is posted here

10 comments:

  1. And accuracy in the field would be determined by such factors as ability to judge range, trunnion tilt, ambient conditions (including ammunition temp) and sights/optics, leaving little room for the actual gun parameters I'd say. Scoring a long range hit would require experience, luck and quite a few rounds. Or perhaps acces to good range finder/reference. Any myth about German accuracy would hardly be founded on shooting range CEP.
    Interesting numbers though.

    ReplyDelete
  2. This article pretty much confirms the superior accuracy of large caliber guns. Actual ability to hit is still up to finding target, range, pointing the gun to it, etc. though, as Bjorn stated.

    ReplyDelete
  3. I believe this guy could use a introductory course in statistics.

    ReplyDelete
  4. The T-34/76’s one great weakness was its fire control efficiency. It suffered from the same two-man turret syndrome as other Soviet tanks in this period, namely that the tank’s commander, gun aimer, gun firer and platoon commander (if a platoon leader), were all the same person. Exacerbating this was the fact that the T-34/76 had relatively poor main gun optics quality, no turret basket, a very cramped and low turret (the gun could not depress more than three degrees severely restricting use on a reverse slope or at close range), poor turret drive reliability, no radios, and generally poor target observation and indicator devices (including no turret cupola and only one vision periscope for the tank’s commander).
    In summary, the T-34/76’s inherent fire control efficiency was so bad that even well trained and experienced tank crews were put at a severe disadvantage. For inexperienced tank crews, with no radios and probably no organised combined arms support, it was a disaster.
    So what was the result of the T-34/76’s two man turret, weak optics and poor vision devices (that is a poor overall FCE factor)? German tankers noted “T34s operated in a disorganised fashion with little coordination, or else tended to clump together like a hen with its chicks. Individual tank commanders lacked situational awareness due to the poor provision of vision devices and preoccupation with gunnery duties. A tank platoon would seldom be capable of engaging three separate targets, but would tend to focus on a single target selected by the platoon leader. As a result T-34 platoons lost the greater firepower of three independently operating tanks”.(9) The Germans noted the T-34 was very slow to find and engage targets while the Panzers could typically get off three rounds for every one fired by the T-34.
    Given this facts you can see that the gun intself is not the only factor that can determinate accuracy.If we consider only gun accuracy then yes, soviet guns are perhaps even better than german and allies guns.I belive you are confusing gun muzzle velocity with gun accuracy

    ReplyDelete
  5. Hi, would you be so kind to share the artillery tables for the F-34 and the M-30? Thanks! =)

    ReplyDelete
    Replies
    1. F-34:
      http://f-lite.ru/lfp/s018.radikal.ru/i527/1305/08/2de103cd3bc0.jpg/htm
      http://f-lite.ru/lfp/s020.radikal.ru/i721/1305/4f/1a4f1cf70482.jpg/htm
      http://f-lite.ru/lfp/s018.radikal.ru/i503/1305/37/113b757a9127.jpg/htm

      M-30: http://narod.yandex.ru/disk/59154539001.160e1f5d636ffd72a96c30676b25081f/%D0%A2%D0%A1_122-%D0%BC%D0%BC_%D0%93%D0%B0%D1%83%D0%B1_%D0%9C-30_1943.djvu.html

      Delete
  6. It seems to me that the author made a huge blunder by equalling the soviet deviation in 50% RADIUS with the german 50% ZONE.
    Zone is not radius.
    Zone is a rectangel which is slightly larger in area than TWICE the radius! The radius covers ONLY the distance between the centre of the MPI and one edge of the 50% circle. The ZONE covers the WHOLE DISTANCE between both edges.

    ReplyDelete
  7. Official soviet 122mm D25T dispersion radius figures from firing tables using BR-471 projectiles:

    1000m: 30cm x 30cm
    2000m: 60cm x 60cm
    3000m: 100cm x 90cm

    Notice these are 50% radius! Not zones as in german documents.
    For comparison 88mm L71 using Pgr39 zones, corrected for radius (radius = 1/2 diameter, 1 diameter defines max. extent of the rectangle from the centre at same height or width, respectively):

    1000m: 15cm x 25cm
    2000m: 35cm x 45cm
    3000m: 50cm x 70cm

    Bottomline, the higer accuracy of german guns is not a myth but a fact. These mistakes happen when interpretors jump to conclusions without understanding what these figures mean...

    ReplyDelete
  8. Correction, data for 122mm dispersions above for BR-471B (APBC) source is the firing table for 122mm D25T issued by the soviet military to the NVA (in russian, copy at the MGFA Potsdam).

    ReplyDelete
  9. Even if a gun shot with perfect accuracies ( 0% deviation) all shells accelerate downward at about 10m /sec/sec. at a range of 1000m range errors (remember no lasers stereoscopic sights etc WWII ) using stadia methods (marks on the reticule) will cause errors a lot bigger than a meter.

    ReplyDelete