Sunday 3 November 2013

45 mm APCR, Part 2

In 1942, Soviet engineers began experimenting with subcaliber ammunition for the 45 mm anti-tank gun. Such ammunition required tungsten, and tungsten is rare and expensive. To refresh your memory, the engineers attempted to use highly hardened steel to compensate, without much luck. Here, they came up with another solution.

"REPORT on the testing of 45 mm subcaliber armour piercing shells with experimental bimetallic cores, blueprint #3-07955

The trials are done above quota, according to paragraph 1 of the work plan of the Special Group of the AK GAU KA for 1942, presented for approval to the Chief of the AK GAU KA as #001132 on May 19th, 1942.

Over the period between May 28th and June 6th, 1942, the Special Group of the AK GAU KA performed experiments on potential partial replacement of tungsten with steel in cores made from the Reniks-6 metallo-ceramic alloy (with a tungsten carbide core) in subcaliber armour piercing shells on blueprint #3-07955.

In the first trial, the idea of a core with two heterogeneous metals was tested: the tip made of the Reniks-6 tungsten alloy, used in mass produced 45 mm APCR, and the rest of the cylindrical part from steel. The joint was done by welding with a copper base. The first trials attempted to remove 50-60% of the tungsten from the core.

1. Production of experimental bimetallic cores

Figure 1 shows the bimetallic core, which was selected for trials.

Fig 1. 

As can be seen from the figure, the front 35 mm of the core is composed of the stock Reniks-6 tungsten alloy, the rest from HV5 grade steel. The HV5 grade was selected out of practical considerations, as it was the only one present.

The mass of the components is approximately as follows:
  • Tungsten tip: ~100 grams
  • Cylindrical metal part: ~90 grams
which reduces the tungsten alloy use to 40% compared to a standard 0.270 kg core. 60% of the tungsten is saved.

Due to the prediction that the production of these shells at a factory would take along time, it was decided to produce 3-4 samples in an improvised fashion, to test out the idea. Four stock tungsten cores were chosen, 70 mm and 250 grams on average. It was decided that incisions would be made in the middle of each one, i.e. 35 mm from the ends. An attempt to make the incisions with cutters made from tungsten carbide failed, the cutters dulled and cracked, while no traces of a cut were seen on the core. A diamond tip from a Rockwell tester from the 2nd department of the Gorohovets ANIOP was used, which was never used for its intended purpose (it was missing a fragment of the diamond). The diamond let us make a groove 1.5 mm deep, but was too short to continue. The cores were then broken into two pieces by striking them with a mallet. The separation was mostly along the line cut by the diamond, but was uneven. Two of the cores had missing fragments from the section we were going to test. Using a sanding wheel, we were able to get rid of the unevenness and sharp edges of the fracture.

After that, using 45 mm HV5 steel cores meant for 76 mm subcaliber shells, cylinders 30 mm in diameter and 35 mm in length were produced. The two parts of the core were welded together at the laboratory of the 2nd department of the Gorohovets ANIOP in the following fashion: the steel part of the core was heated up in the electrical oven to a temperature of 880 degrees, then dusted with borax. As soon as the borax started to melt, the surface of the cylinder was cleared of debris using a metal brush, then was again dusted with borax, and covered with the tungsten part of the core. Small pieced of red copper were placed on the excess surface of the cylinder, and it was placed in the oven, heated up to 1100 degrees, and removed after the copper melted. In order to achieve better contact between the components of the core, the tungsten part was pressed to the steel part by hand, and then the cooled product was finished on a lathe in a manner similar to the stock tungsten core.

The average Rockwell hardness of the steel part of the core before mechanical finishing was R_C150 ~=32, which is equivalent to Brinnel H_B ~= 300, or a print size of about 3.5 mm.

The welding quality of all 4 cores was unsatisfactory, due to the fact that the parts of the cores in contact with each other had significant spaces, which were not filled with copper. Out of the worst two, one (#2) was shot, and the second (#4) was left as a sample. Photo #1 and #2 show the core in 2.25x magnification in two views, on which 'a' labels a 5 mm long space, that is 6 mm deep and 1 mm wide. 'б' marks a missing fragment, leading to a space 17 mm long, 5 mm deep, and 6 mm wide at its longest point. 

Photos #1 and #2

Masses of the produced cores:
  1. 190.6 g
  2. 182.7 g
  3. 186.2 g
  4. 180.7 g
The stock weight of the core is 270 grams. However, the average weight of the cores fired at the Gorohovets ANIOP, as well as the average weight of the cored examined, was 250 g. The loss in mass is 60-69 grams from average weight and 80-89 grams from blueprint weight.

Shell: 45 mm blueprint #3-07955, with prototype core, shell weight is 0.770 kg
Gun: 45 mm ATG, model 1932/37, #7646
Armour: 60 mm homogeneous, K=2350, #18078-8-9-43
Angle: 30 degrees
Gunpowder and charge weight: 6/7 sv 3/38 R w=350
Impact velocity: 946.8 m/s
Result: Clean penetration, one core caliber in diameter. The steel component and the tungsten component with a crack through it were found behind the armour. The dent from the casing is 7 mm.

Shell: same, 0.761 kg
Charge weight: w=0.330
Impact velocity: 904.1 m/s
Result: Clean penetration, one core caliber in diameter. The steel part of the core was found behind the plate, slightly compressed axially.

Shell: same, 0.764 kg
Charge weight: w=0.320
Impact velocity: 875.6
Result: dent. Both parts of the core were stuck in the plate. The steel part of the core was not destroyed. The rear of the plate had a bump, 5-8 mm tall. The dent from the casing is 7 mm.


Despite the crude production method and unsatisfactory quality of the weld, the bimetallic components penetrate a 60 mm plate with a coefficient K=2350 with a velocity of 904 m/s at 30 degrees. The distance that this velocity is achieved at, when shooting with a full charge (Vo = 1035-1040 m/s) is 250 m.

Therefore, the 45 mm subcaliber armour piercing shell made using blueprint #3-7955 with a bimetallic core, composed of 100 grams of tungsten and 81-91 grams of steel, mostly solves the objectives set by Artkom GAU KA (#570896, from January 25th 1942) "penetration of 60 mm of armour with resistance 2200-2400 from a distance of 250-500 meters, at an angle of 30 degrees". The distance of penetration is short by only 50 meters.

The savings of tungsten carbide alloy are approximately 60%.

The steel part of the core made from HV5 grade steel penetrates the plate in the event of a through penetration, either wholly or in fragments."

And so, using some broken lab equipment, a mallet, and some steel they had lying around, the scientists managed to put together workable subcaliber shells. Recall that all steel subcaliber shells could not penetrate a 50 mm plate at ~100 meters reliably, and could not penetrate a 60 mm plate at all. Full tungsten carbide shells could penetrate 80 mm of armour at 250-300 meters (depending on the shell type). The penetration suffers, but the resulting shell uses less than half of the tungsten necessary to produce a full one.


  1. What does resistance of 2200-2400 mean?

    1. It is a measure of the quality of armour. You will see some more values in the tables in this article, on the fifth column

    2. What should that be?
      Measurements should have units, like Brinell or Rockwell.
      otherwise its just a constant. I dont really understand that. Do you have any info on it?

    3. I'm pretty sure it's unitless. It's just a coefficient.