Monday 27 April 2015

Common Questions: KwK 43 vs D-25T Penetration

One of my earliest articles was an explanation of the difference in Soviet and German penetration standards. Now, just as then, accusations of propaganda are quickly encountered when this table pops up. After all, it's easy to Google "D-25T penetration" and "KwK 43 penetration" and compare the results. You will, of course, find a number of theoretical values derived through various methods, but tanks fight wars on the battlefield, not in Excel, so let's take a look at some practical results.

The first test that I have where the two guns compete is against the armour of the Tiger II (CAMD RF 38-11377-129). However, due to its brittle nature, it's hard to establish the clear winner here. Nevertheless, I will include this test for completeness' sake.

As you may recall from the relevant article, one 88 mm shell fired into the front of the turret at 400 meters (it took three shots to get a "fair" penetration, the others hit the edge of an existing opening) penetrated the front of the turret (180 mm) flew all the way through, and penetrated the rear (80 mm) through a weld seam. This must have been confusing for some people, as I've seen claims that the gun penetrated 260 mm and that Soviet penetration tables are a lie based on this fact, but as you can see, it didn't quite happen that way.

Now, the D-25T. Using the older pointed type AP shell at 2500 meters, it penetrates the front of the turret, making a colossal hole in the armour (700 mm by 220 mm) and, for good measure, the roof (460 mm by 300 mm). The shell does not make it through the back, but at six times the KwK 43's range, it's not the most scientific of comparisons. Let's try a target with sturdier steel, like an IS-3.

"Photo #11. Overall view of the 90 mm thick overtrack hull of the modernized IS tank after a hit from an 88 mm armour piercing shell at 400 meters. Three tracks have insignificant damage."

At the same 400 meters, the side of an IS-3 is tougher than the front of a Tiger II, eh? Let's see how close the KwK 43 has to get before penetration.

CAMD RF 38-11355-2872

At 300 meters, the KwK 43 can finally penetrate the side of the tank, but it's a "limit of rear surface" penetration, a standard where the shell only has to cause damage the rear of the plate, and not actually make it all the way through. The D-25T can achieve the same from 1100 meters, or make it clean through the plate from 600 meters. You can read more about the tests here, but the D-25T consistently penetrates the prototype hulls from a longer range than the KwK 43.

Data about how the Object 701 #1 hull stood up to these guns is also available, albeit not as detailed.

D-25 penetration limit
KwK 43 penetration limit
Lower hull side
4400 meters
3790 meters
Turret side
3400 meters
2600 meters
Data obtained from M. Kolomiets, Heavy Tank IS-4 (Moscow: Tactical Press, 2014), pp. 39, 76

Why does this happen? Aren't the numbers in the Soviet penetration table very close? They are, but ballistics is a more complicated science than this. Through a phenomenon called overmatch, a shell that is significantly thicker than the armour it strikes gets a bonus to penetration, while a shell with a small caliber compared to the armour it tries to penetrate will face a penalty. The D-25's 122 mm shell is almost 40% thicker than the KwK 43's 88 mm shell, which makes it rather more effective. This lesson was well learned by Soviet tank designers: according to NII-48 standards a post-war medium tank only needed to resist 88 mm shells, while a post-war heavy tank had to resist both 122 mm and 88 mm shells.


  1. The big problem with Soviet penetration figures is:
    -Not a very clear known methodology used(contradictory sources);
    -The plate used as standard test(cemented,very hard armor, other?Hardness used ?);
    -Use of DeMarre equation to establish penetration(OK for Soviet ammo, but cause weird values for western capped ammunition);
    The germans used a 0.44 factor to convert from a 30 degrees penetration to a 60 degrees so 8,8 cm L/71 penetrate 89mm at 60 degrees mm at 100 meters, using german standard(shell must pass close to intact) .
    If you consider that BR-471 weight 25kg and Pzg.39/43 10.4kg (less than half) is obvious that the later lose much more velocity (and penetration) over distance than BR-471.

    1. DeMarre is used after sample data has been collected for that gun to interpolate other ranges, it's not used for the whole table. That's how everyone did it.

    2. No Peter, that guy is absolutly right. I tested it in excel. I reproduced your very first penetration article table this way using the demarre formula. The result was the following: The same penetration constants were used for all guns, not only the soviet ones.

    3. Give an example of some better test protocols than this on this page.

    4. I wonder why Peter still keeps this fairytale alive. Those german penetration values which are given are calculations based on DeMarre equation.

      Even the covering page of the Document clearly states: CACULATED : (using coefficient K-2400)

      Seriously why you can't confess that those numbers of german guns are just an atempt to caluclate their penetration probability and not thrown out of actual testing?

    5. I dont get it either. Do you have more sites of the linked document?

    6. Most of the table is calculated, there are reference points from actual testing. If it was all calculated, why are there so many holes in the table? Why wouldn't it be easy to fill out with just calculations?

      If you used the 100 meter reference from that table, then there you go, you would get all the same results. Nevertheless, you can clearly tell from the practical testing that the KwK 43 is inferior in penetration to the D-25, which the table does not adequately reflect. This is why I discourage the use of this or any other table when practical testing results are available.

    7. In real tests.
      In an Aberdeen test of the upper limits of the 88mm Projectile.
      Unfortunately I don't have the year or the BHN of the armor. Though armor of 262 BHN was used in some tests of US projectiles in other tests armor BHN ranging from 224 to 444 was used.

      8" (203mm) @ 0:Two complete (NBL) penetrations with projectile passing through plate. 3124mv and 3257 mv.(550y/90y) Two ABL penetrations. Projectiles Intact. 3001 mv 3038 mv. (990y/860y)

      3 7/16" (87mm) @ 55: Two complete penetrations. both projectiles fractured. Projectile fragments passing through plate.
      One partial penetration projectile fractured.
      3334 mv on two penetrations. 3310 mv on 2.5" deep partial.

    8. And? The unknown composition/quality of the metal is precisely why I'm bringing up tests that fire at the exact same plate.

    9. Peter your tests doesnt show anything new.

      Larger shells perform better against sloped armor. This is nothing but physics. The reason therefore is the moment of inertia of the shell. This moment hampers the shell from ricocheting. The moment is significantly larger on bigger shells because it grows with both mass (m) and also with the length (l). And the length is raised to the second power, which makes this effect even more significant. The capped shell with the flat nose BR-471B makes this effect even more significant.

      Your other point is penetration on really long range. Again physics. Shells are slowed down by wind resistance. What keeps the shell flying on its way is the force of inertia F=ma. A shell with a bigger diameter has much more mass in relation to its diameter, because mass grows ³ and the diameter with ². The wind resistance is proportional to the diameter. Thus bigger shells have an easier time keeping their speed up at larger ranges. However you have to add that this only works with the capped BR-471B shell. The uncapped shell is probably worse on long range even compared to the 88 mm gun.
      However we have to keep in mind that the 8,8 cm L/71 was still better against flat surfaces on usual combat distances.

      Best Regards

    10. Peter, it may not be the same test if we don't know what happened to the 30mm outer side overtrack plate. If the 88mm had to penetrate it and the 90mm side plate and it was knocked off. Then only the 90mm inner side plate was left so the 122mm only had to penetrate it.
      That is why these ad hoc tests don't count for much.


    11. And what, if it was first torn out by kwk? In the second round when it was already torn out, kwk was still inferior to D25.

    12. These tests are 70 years old, of course they don't show anything new. As you can see, over long distances and against barely sloped plates (see the IS-4 data) the D-25 is still better.

      Also you can clearly see that there was no 30 mm plate there, unless you are claiming that the KwK 43 shell incinerated it. Also the same results persist when there is clearly no armour screens blocking the path, such as for the UFP. It's quite obvious that the penetration of the D-25 is superior.

    13. But I have to give big respect to IS 3. For which other tank on the world pay that if you dont penetrate it from the front you have not much bigger chance to penetrate it from the side?! That tank had also much thicker turret then its predecessor while it had at least the same weight!

    14. That Russian table reveals the inconsistencies of the Soviet tests. Because according to the DDR tables the velocity of the older sharp AP is 743 m/s at 300m not 600m. And 697 m/s is close to the velocity at 650m. Thus all their purported ranges are now suspect.

    15. Post-war tables with pre-war ammunition? This I gotta see.

    16. Plugging the numbers into my ballistic program shows that the MV of the 122mm gun used in the test was 796m/s. If the calculated values used DeMarre and the impact velocity numbers were also mis-calculated like these they are all wrong.

    17. What numbers? What ballistic program? What are you trying to show here? The impact velocity can be found from the D-25 ballistic table. I don't know what numbers you're trying to claim are wrong.

  2. "Using the older pointed type AP shell at 2500 meters, it penetrates the front of the turret, making a colossal hole in the armour (700 mm by 220 mm) and, for good measure, the roof (460 mm by 300 mm)."

    Do you have a picture of this magic hit? I can't find one.

    1. This article:
      There's a link to a document:
      The photo you want is on page 45.

    2. "Shot #34. Target: turret front. Shell: 122 mm AP pointed type. Propellant: reduced. Distance: 2500 m."
      This is artificially simulated the 2500 m range and hit along the turret front edge with the roof. It doesn't even look like a clean penetration, the dent or hole, if there is one is smaller than 88mm hits. The armor may have cracked off inside. See KwK 43 hits #23 and #25 for what clean penetrations look like.

    3. Have a closer look at the quoted test results. There was another 122 mm shot (number 33) against the turret front which happened before(!) the one quoted above. Maybe this one clears it up better. For some reason that part of the document was not translated yet.

      Could a native speaker translate the results of shot #33?

    4. "There was another 122 mm shot (number 33) against the turret front which happened before(!) the one quoted above. Maybe this one clears it up better."
      Thor, #33 is a shinny dent below penetration #24. According to a post here on AA #23, #24, #25 are all 88mm penetrations.

    5. Still there could be some reasonable info, for example like a weakening of the roof armor for example etc.

  3. "Through a phenomenon called overmatch, a shell that is significantly thicker than the armour it strikes gets a bonus to penetration, while a shell with a small caliber compared to the armour it tries to penetrate will face a penalty"

    I understand the overmatch penetration bonus when the shell has a larger diameter than the thickness (unadjusted for any slope modifier) of the armor, but I have a hard time understanding the penalty to shells with a smaller diameter, at least in regards to test results.

    If the 88/L71 could penetrate (using the Soviet method) 168 mm of armor at close range, as that is already considerably greater than the diameter of the shell (88 mm). Is then the overmatch penalty then built into said penetration testing result by default?

    1. As a physical phenomenon, it doesn't go anywhere during testing. If you look at the Soviet penetration table, the 30 degree figure isn't simply the 90 degree figure times cos(30), so there are some multipliers that affect penetration in play here.

  4. If we compare comparative tests with kwk 43 and soviet guns against geman sloped armour plates we can see that german guns lose their magic penetration, when they have to hit plates sloped at 55 degree angle and more.
    1,Panther 82mm at 55 degree=145 mm of EA
    kwk 43= 650 m, D10= 1500 m, D25= 2000 m
    2,IS 3 90mm at 60 degree= 180 mm of EA
    kwk 43= 300 m, D25= 1100
    back armour 60 mm at 49 degree
    kwk 43= 4000 m
    3,M 48 100 at 60 degree angle
    kwk 43= 200m, D10= 700m

    German AP caped shells have advantage against soviet non caped shells when they have to penetrate armour up to 50 degree angle but at bigger angles they lose their magic penetration rapidly. Then soviets AP shells show their adwantages in kinetic energy because 15 kg at 900 m/s and 25 kg at 800 m/s is better than 10 kg at 1000m/s and who is not able to see this big difference then he dont use common sense and there is reccomended lead rinsing from my side.

    1. You probably mean 3, M47 102mm 210 BHN at 60 degrees, KWK 43= 250m, D10 BR-412B APBC= 750m. BR-412B is post WW2.

  5. Yes it is postwar like BR 471 B and it is also APBC so it have the same characteristics and there is small difference between performance of BR 412 and BR 412B like in case of BR 471 and BR 471 B.

  6. "Photo #11. Overall view of the 90 mm thick overtrack hull of the modernized IS tank after a hit from an 88 mm armour piercing shell at 400 meters. Three tracks have insignificant damage."
    In that photo the JSIII is missing the 30mm side outboard overtrack plate. Was that torn off? With the inner 90mm at 60-degree side hull the LOS armor thickness would be 214mm.

  7. Again, here is a real table for the D-25 with a MV of 781m/s.

    At 600m the impact velocity is 705 m/s not 741 m/s. At 1100m the velocity is 643 m/s not 697 m/s.
    Even if the MV was increased by 15 m/s the impact velocities were also increased by 15 m/s they still don't match the velocities in the test.
    DeMarre needs the velocity to be correct in order to calculate penetration. If not is worthless.

    (If you can't trust the DDR who can you trust?)


    1. The goal of these penetration test was not comparing penetration abilities of the guns but testing IS 3 armour capabilities whether it meet the given requirements. So If they wanted to prove how good new tank is the increasing of nuzzle velocity is absurd and worthless.

      And here is ballistic test from NDR of D25 that agree with the numbers you cant accept.

    2. That table is for something called NBR-471. I don't know what round that is, but it's definitely not the one the Soviet table refers to. I already posted a ballistics table for the D-25 on this blog, neither the BR-471 or the BR-471B data matches this "NBR-471".

    3. The armor penetration of the NBR-471 seems close enough (+/- 2-5mm) to the BR-471 data here.
      There may be a testing methodology difference between the DDR and Soviets.
      And the table penetration here for 3000m may be in error. It more like is 85mm not 75.

      From this table the BR-471 and the NBR-471 in the DDR table lose exactly 237m/s in 2000m.
      And like I said compare the velocities of the ranges at 600m and 1100m in this table to the those in the test of CAMD RF 38-11355-2872 and you will see something is not right.

    4. Here is another reference that compares exactly:
      122mm D-25T mod. 1944
      Penetration at ranges of 500/1000/2000m in mm
      BR-471 143/117 126/102 97/79

      "Otechestvennye bronirovannye mashiny. XX vek" : nauchnoe izdanie v 4-kh tomakh/ Solyankin A.G.,

      Pavlov M.V., Pavlov I.V., Zheltov I.G./ Tom 2. "Otechestvennye bronirovannye mashiny. 1941-1945" ,

      Moscow, Exprint, 2005

    5. These tests are almost certainly BR-471B. I already posted the table for it, I don't know why you're continuing this farce. The possible answers are a) nobody in GABTU or the Red Army knew what they were doing and everyone was using the wrong data or b) only you are using the wrong data. Which do you think is more likely?

    6. OK it could be the BR-471B. But in your article you reference the turret penetrations of the King Tiger. Both hits on the turret were the BR-471 sharp nose shells. That is why I brought it up.

  8. You got it! That is the same DDR table I link to in my post. Proving my point. Q.E.D.

  9. This comment has been removed by the author.

  10. I hope that everyone is aware that 1944 test of 122mm against KT was done using A-19 not D-25. And according to some ballistic tables A-19 had greater muzzle velocity (over 800m/s) giving it about 300m advantage over D-25 (781m/s) gun. So substract 160..300m from every result of 122mm ammo in this test, to get D-25 performance (160m difference for BR-471, 300m for BR-471B).

    1. Over 800 m/s!? Wow, what ballistic tables are you looking at? Plus I already posted a D-25 ballistics table, you can familiarize yourself with it here:

      The muzzle velocity of the BR-471B in the A-19 is 800 m/s, the muzzle velocity of the BR-471B in the D-25 is 795 m/s, wow, such a huge difference. That's a grand total of a whole 2 mm according to DeMarre estimates.

  11. In two tables I've seen the muzzle velocity for D-25T was always 781m/s and for A-19 it was either 800, 801 or 803m/s. That's 20m/s of difference worth 150m of range for BR-471 (so ammo that was actually used during WW2).

    Your table says 795m/s for BR-471B, that would be only 5m/s of difference.

    Do not know which table is right, but AFAIK D-25T barrel is 245mm (2 calibres) shorter than A-19, but with muzzle brake added. That should reduce MV somewhat.

  12. Peter, something interesting: in my table muzzle velocity of BR-471, BR-471B and OF-471 is the same - 781m/s. The BR-471B has the same ballistics as OF-471 (single table for both) - guess it was designed this way, so single table and single sight scale can be used for both. Now in your table muzzle velocity of AP shell is 795m/s and at 4000m velocity drops to 518m/s. In my table BR-471B and OF-471 starts at 781m/s and at 4000m they drop to 534m/s ! Now in your table OF-471 drops from 795m/s to 553m/s at 4000m. I thought tgar your first first table is for BR-471 not BR-471B, but it's rather not - the BR-471 has much higher velocity drop (to 376m/s at 4000m in my table). Anyway, ballistics for BR-471B and OF-471 are different in your table, and identical in my table. Different ballistics should mean different aiming scales for BR-471B and OF-471 on the gunsight.

    Differences in ammo? I wonder what's the date of your table ? Which one is closer to WW2 ammo. Mine is from book dated 1947. My mentions only the Tsh-17 sight. Your mentions also TSh-21, TSh2-21, TSh2-27. For what tanks are those gunsights?


    my table: muzzle v, 4000m v, difference, time of flight

    BR-471 781 376 -405 7,5s
    BR-471B 781 534 -247 6,0s
    OF-471 781 534 -247 6,0s

    your table:

    BR-471B 795 518 -277 6,2s
    OF-471 795 553 -242 6,0s

    Probably most important data for any ballistic table is the aiming angle. Other data like time of flight, remaining velocity ect may be calculated and approximate, but aiming angle should rather be correct and verified becasue if it was wrong the table would be useless.

    Aiming angle (mils) for 4000m:

    my table:

    BR-471 53
    BR-471B 41
    OF-471 41

    your table

    BR-471B 41
    OF-471 37
    OF-471H 38

    Now that's strange. Your BR-471 and my BR-471 have identical ballistics at 4000m despite different (?) muzzle velocities. This is strong suggestion that muzzle velocities are in fact the same, but one table lists them wrong (maybe they were measured not precisely enough...). On the other hand your OF-471 has different ballistics (it's more aerodynamic) than my OF-471.

    Interesting. But so far it's hard to deduce anything.


  13. In my table the TSh-17 has a single scale called PGP that covers the ballistics of OF-471 and BR-471B shells. When firing with BR-471 shell the gunner had to use the table - for example when firing BR-471 to 2000m he should select 2200m range on PGP scale of his sight. I wonder if war-time TSh-17 on IS-2s had dedicated scale for BR-471 and only after the war, when the BR-471 was no longer used, this scale was removed. Or maybe trough the whole war gunners of IS-2s had only PGP scale with OF-471 ballistics so they had to memorise list of corrections to this scale for using BR-471 shell...

  14. Up to 1100m gunsight scale settings for BR-471 and BR-471B were the same, so the sight's PGP scale could be used directly for BR-471. From 1200m up the settings begin to difer, so when using BR-471 corrections are needed.