Bismarck vs. Iowa

[RobertsonN]

Thanks. However, the documenation with Gkdos 100 (1940) shows the line of German naval shell development and the final step has a nose of 1.3 crh. I know Gercke stated in 1945 that they had changed to 0.93 crh because it had a good balance of oblique penetration and strong attachment of the cap. What I am unclear about is whether Bismarck and Tirpitz were equipped with the later 0.93 crh shells.

[delcyros]

If I am not misunderstanding him, his reference to 0.93crh tip mentions the armour piercing cap which was welded with a solder to the projectile´s nose. He doesn´t mention that the projectile nose itselfe is a 0.93 crh design.

But I suspect that these numbers are a bit in error:

A 38cm L/4.4 fired from the 38cm/L52 gun, against KC plates also declined by 20* obtained the following results:

20km-364mm
21.8km-350mm
25km-308mm

They may still represent actual results from shootings carried out at Meppen but the definition isn´t clear to me.

20 deg angle can be

+20 deg from the vertical
-20 deg from the vertical
+/-20 deg across the target angle

finally it´s also possible that the impact obliquity was fixed at 20 deg and striking velocity was changed according to the simulated range, ignoring the specific ballistics involved.

[alecsandros]

But I suspect that these numbers are a bit in error:...

.

Where exactly do you see the error ? The phrase was "declined by 20* ", purposefully so that you wouldn't confuse it with INCLINED, which may have been interpreted as favoring perforation at long range, given that the compounded striking obliquity would have been closer to 0*. (a shell falling at 20* against a plate inclined at 20* has a 0* vertical impact obliquity...)

[Thorsten Wahl]

according to Nathan Okuns calculations german KC 8nA) seem to have a equivalent thickness compared to US Class A of at least 106,5 %.
This relationship appears to be applicable to angles of incidence from 0 - 30 deg. advantage for KC is higher at lower angles of incidence.

So 310 mm US Class A seem to be equivalent to 294 mm KC (nA). This raw thickness seem to be penetrable at around 20000 m ~ 22 kyard (Obliqity 14°+19° ~34 degrees, impact speed ~510 m/s). maybe somwhat less if you add around 10m/s for hull plating and other non ballistic plate

[Karl Heidenreich]

Thorsten:

So 310 mm US Class A seem to be equivalent to 294 mm KC (nA). This raw thickness seem to be penetrable at around 20000 m ~ 22 kyard (Obliqity 14°+19° ~34 degrees, impact speed ~510 m/s). additional you should add around 10m/s for hull plating and other non ballistic plate

Awesome!

[alecsandros]

So 310 mm US Class A seem to be equivalent to 294 mm KC (nA). This raw thickness seem to be penetrable at around 20000 m ~ 22 kyard (Obliqity 14°+19° ~34 degrees, impact speed ~510 m/s). maybe somwhat less if you add around 10m/s for hull plating and other non ballistic plate

Krupp testing showed perforation of 308mm KC n/A declined at 20* for scaled 25km shootings. This implies a 23.7* angle of fall, and a 43.7* final obliquity angle.

I see no reason why Nathan's computations should take precedence over Krupp's ballistic trials.

[Byron Angel]

Just did a (not so quick) quick evaluation running Brian McCue's ballistic data through NaaB giving the following performance for a Bismarck 15in projectile versus USN 1935-1943 Class A armor, assuming plate inclination @ 19deg toward firer and allowing for various values of horizontal (line of fire) obliquity >

Notes -
( 1 ) "net obliquity includes angle of fall, plate inclination and horizontal obliquity components).
( 2 )"projectile intact" = passed through armor in fit condition to burst within target ship vitals.
( 3 ) "projectile ineffective" = other condition than case ( 3 ) above.

20,750 yards = 11 deg angle of fall @ 1756 f/s (535 m/s) striking velocity
Horizontal obliquity (net obliquity)
00 deg ( 30.0 deg ) = 14.8 in penetration. projectile intact.
10 deg ( 31.5 deg ) = 14.4 in penetration. projectile intact.
15 deg ( 33.2 deg ) = 14.0 in penetration. projectile intact.
20 deg ( 35.5 deg ) = 13.3 in penetration. projectile ineffective.
30 deg ( 41.4 deg ) = 11.6 in penetration. projectile ineffective.

24,605 yards = 19 deg angle of fall @ 1659 f/s (506 m/s) striking velocity
Horizontal obliquity (net obliquity)
00 deg ( 38.0 deg ) = 11.8 in penetration. projectile ineffective.
10 deg ( 39.1 deg ) = 11.6 in penetration. projectile ineffective.
15 deg ( 40.4 deg ) = 11.2 in penetration. projectile ineffective.
20 deg ( 42.2 deg ) = 10.7 in penetration. projectile ineffective.
30 deg ( 47.0 deg ) = 9.6 in penetration. projectile ineffective.

..... which leads to a general assessment that the "real world" immune zone inner limit against the Bismarck's 15 inch gun depended not only striking velocity, but also the net obliquity factor. Nathan Okun's evaluation suggests that Bismarck's projectile will break at a net obliquity > 35 deg. This suggests that, even at 20,750 yards, the horizontal (line of fire) obliquity component must be < 20 degrees to promise delivery of an intact projectile through the armor and fit to burst within the target ship's vitals.

Interesting.

[alecsandros]

Hello Byron,

With all due respect for Nathan's work, It falls short of the empirical evidence amounted in the last years. NaAB still considers KC n/A at 93-96% resistance of penetration than a British post-1930 CA plate, allthough descrpition of British tests have came to surface, showing that this was not the case at all.
Whotan is treated as a 20% EL RHA, allthough it still retained 25%. The modeling of the projectile-plate interaction does not take into consideration important factors (such as gluing of the AP cap to projectile's body, wchich was proved by US tests to deliver better results than without gluing). ANd there are many other problems!

My question is: exactly how and WHY would you want to use a computer program to defectively re-create empirical tests which wre done in years previous to the war and showed 308mm perforation at 25km against 20* declined plate?

[RobertsonN]

Thanks, gentlemen, for all the effort you've put in. The answer of perhaps a little under 20000 m taking into account structure seems plausible but I lacked the means to make a determination as precise as this. Thsis is higher than the 17600 yds against the 16/45 2240 lb shell but probably the lower angle of fall and higher velocity outdo greater mass.
At 25000 m the figure I have from Gkdos 100 (from another web site) is 335 mm KC n/A. The figure of 305 mm at 20 deg, I came to the conclusion that it probably refers to target angle. I have a formula of Nathan Okun for calculating resolved obliquity (it appears in the Hood sinking analysis) but as I spent yestday evening on another problem I did'nt have time to work that out.

[delcyros]

But I suspect that these numbers are a bit in error:...

.

Where exactly do you see the error ? The phrase was "declined by 20* ", purposefully so that you wouldn't confuse it with INCLINED, which may have been interpreted as favoring perforation at long range, given that the compounded striking obliquity would have been closer to 0*. (a shell falling at 20* against a plate inclined at 20* has a 0* vertical impact obliquity...)

Dear Alecsandros,

I see a misinterpretation with these numbers. They tend to be used without discussion of their meaning. As I posted above, as long as the definition is unclear, we don´t know what they mean.
20 deg obliquity is a rather vague explenation.

I will propose what I mean here:

[+] the only thing we know for sure is the striking velocity at this 25km range, which acc. to GKdos-100 lies at exactly at 470 m/s
[+] at 25km the angle of fall is also known to be 24 deg. It´s not known whether or not this figure is attested for in the Meppen trials or ignored.

This is the general ballistic condition. For our three possible cases we then conclude

[+] Scenario A: an impact on a 20 deg inclined plate, simulating a range of 25km: netto impact obliquity is 44 deg, striking velocity is 470/ms
[+] Scenario B: an impact on a 20 deg declined plate simulating a range of 25km; netto impact obliquity is 4 deg, striking velocity is 470m/s
[+] Scenario C: an impact assuming 20 deg target angle on a vertical plate simulating a range of 25km; netto impact obliquity is 30.86 deg, striking velocity is 470 m/s
[+] Scenario D: an impact assuming a striking velocity of 470 m/s and 20 deg obliquity on a given KC plate.

you see how four different conditions develope from this vague obliquity statement and using my GKdos-100 charts I get correspondingly four different results:
[+] for A = 240mm KC/n.A. Grenz. & 223mm KC in intact condition fit to burst
[+] for B= 395mm KC grenz. & 390mm KC in intact condition fit to burst
[+] for C= 340mm KC grenz. & 320mm KC intact condition fit to burst
[+] for D= 370mm KC grenz. & 355mm KC intact condition fit to burst

You can´t tell which one is to be meant and that´s the problem.

[alecsandros]

Dear Delycros,

I had the same puzzle when I first read David Saxton's research about Bismarck's immunity zone. He followed up and explained that the plate was declined so as to present a more difficult angle of perforation.
Following your scenarios, the tests were done following scenario A, with the final impact angle at around 44*. The shell perforated 308mm, showing a remarkable 10% decrease in penetration from the "usual" 335mm at the same range. This was because of several metallurgical processes which helped perforation during oblique attack.

The computations you are probably doing in NaAb will never replicate de exact results of empirical tests done by Krupp. SOme of the reasons of discrepancies between simulation and real life I have explained above. There are others though...

[Karl Heidenreich]

Alex:

[delcyros]

I don´t do NAab calculations for a number of reasons. I was working on the official RMA penetration curves for 38cmL4.4 APC defeating KC n.A. published in the GKdos-100. My numbers come directly from these curves, not from a computing programm. The numbers for netto impact obliquity are done by trigonometric calculation using the value for inclination of the plate, target angle and angle of fall as specified in the conditions above.
I know that both, Rheinmetall and Krupp had their own set of curves established by carrying out firing trials at Unterlüss & Meppen, respectively. But the W-office (arms) concluded that their own curves were the definitive ones based on comparative firing trials conducted at the firing range of Hillersleben.

I also tend to be a bit more sceptical with penetrations posted to exceed the official curves by almost 1/3 in this condition. This leaves us with three possibilities:

[A] the armour plate used in these trials was significantly subpar in performance to normal KC/n.A. plates (Krupp tested against many types of projectiles and plates, including pre-ww1 vintage KC samples, while official trials were conducted only against reference KC and reference Wh plates known as "special trial plates")
the projectile was much better than service ammunition to be tested at Hillersleben in aspects of obliquity penetration
[C] the penetration graphs at GKdos-100 are conservative (cross checking HKE samples and GKdos-100 curves may give a clue here)

Definitionwise, there was no "holing" definition in german terminology, only
[+]Grenz (projectile either embedded in the plate (intact) or burst/broke in holing with at least 50% of the fragments behind the plate, all impact energy depleted), &
this definition is right in between US partial penetration and US NBL and comes close to the british perforation, where a projectile needs to pass the plate by 80% but doesn´t require it to be in intact condition fit to burst
[+]Heil (projectile completely penetrates and stayed in a condition fit to burst with no significant damage to base or cavity (nose damage is accepted)
this definition matches the US effective limit. There is no congruent definition in british practice to this
[+]Sicher (small bulge or crack in the dented in plate but no plugs or fragments of either armour or projectile ejected
There is no congruent definition in either british or US definition to this. The closest is the british penetration: making a sub calibre hole in the plate with a crack formed right through it (light can be seen on the other side). The US holing limit asked for an at least calibre sized hole to be formed, requiring much more energy.



My preliminary conclusion is that for the moment the closest match to the reported penetration figureof 308mm at 25km range and 20 deg obliquity is scenario [C], using a 20deg target angle to an otherwise vertical plate, simulating a range of 25km. This results in a low deviation from official results: 308mm versus 320mm for intact penetration in both cases. Since you don´t know the remaining excess velocity from the trial, it´s a reasonably close match. Even if there would have been no excess velocity, -if You accept 2% deviation for quality in both projectile (-2%) and trial plate (+2%) than the reported result still falls in within the range of explainable deviations.
However, I haven´t seen the relevant sources and thus I am careful not to jumping from a possibility to a conclusion. Perhaps Dave can shed more light on this.

[Byron Angel]

Hello Byron,

With all due respect for Nathan's work, It falls short of the empirical evidence amounted in the last years. NaAB still considers KC n/A at 93-96% resistance of penetration than a British post-1930 CA plate, allthough descrpition of British tests have came to surface, showing that this was not the case at all.
Whotan is treated as a 20% EL RHA, allthough it still retained 25%. The modeling of the projectile-plate interaction does not take into consideration important factors (such as gluing of the AP cap to projectile's body, wchich was proved by US tests to deliver better results than without gluing). ANd there are many other problems!

My question is: exactly how and WHY would you want to use a computer program to defectively re-create empirical tests which wre done in years previous to the war and showed 308mm perforation at 25km against 20* declined plate?

..... Hi Alexandros,

To answer you questions in reverse order, I have a great respect for Nathan's intellect, background, education and his dedication to his chosen passion over many decades of study. He has long been in possession of GKDos100 and likely every other sort of related historical primary source as well (helps to work for the US government) and has digested and processed their data over many years of careful analysis. I do not take his conclusions and evaluations lightly. What others perceive as "problems" with his work may in fact simply be misunderstandings thereof. I'm not saying that I fully grasp everything he says, so I will leave it at that - I trust his workmanship.

As to the above-discussed test of the German 15in at 25,000 m, I ran that event through the McCue/NaaB process and note the following -

At 25,000 m (27,340 yds) McCue gives the German 15in a striking velocity of 1609 f/s (490 m/s) @ 23 deg angle of fall. Now comes the interpretation of the term "declined".

[ Case A ] If "declined" means that the plate was rotated 20 deg about its horizontal axis toward the firer, as if to model attack against an inclined belt, NaaB indicates that the projectile is rejected without penetration, even at zero deg horizontal obliquity.

[ Case B ] If "declined" means that the plate was rotated 20 deg about its VERTICAL axis, to represent a horizontal (line of fire) obliquity component with respect to attack upon a conventional vertical belt, the projectile easily penetrates in fully intact condition.

So ..... before everyone jumps on the "Nathan is wrong" bandwagon, which seems to have a lot of passengers, I think it important to pin down exactly what "declined" means. If Case A, then Nathan is in fact dramatically wrong, and we are confronted by a projectile capable of penetrating 4/5 caliber KC at 43 deg net obliquity i a fully intact condition - a completely mind-boggling performance. Speaking for myself, I require more unambiguous evidence before accepting this to be true. If Case B, then Nathan is not wrong.

B

[alecsandros]

The test results were confirmed in 1945 by US tests. I can't post David's bibliography here, and neither his entire paper.