• ^Hardness is measured in BHN unless "R" for Rockwell "C" or "V" for Vickers. If none given ("?"), I assume the 225 BHN/115,000 lb/sq.in. average in the computations here, so the BHN formula has no known effect (?).
• ^

  These NBL test values are from the index card database of Dr. Allen V. Hershey, head of the US Naval Proving Ground department that performed ballistics computation and analysis during and after WWII (1943-1955, to be exact). The NBL values were given in terms of the "F" coefficients used in the generic Thompson "F" Formula developed by Dr. L.T.E. Thompson of the NPG shortly after WWI. This formula was:

  T/D = (1728.04)(W/D³)[(V/F)Cos(Ob)]²

  or, in its usual form:

  V = [F/Cos(Ob)]{(T/D)/[(1728.04)(W/D³)]}^0.5

  = (0.024055983)(D)(T/W)^0.5(F)/Cos(Ob) feet/second

  Note that the nominal plate thickness given near the top of the card in many cases is different from the T/D value given lower down, which is the exact thickness (rounded to two decimal places) where the impact center actually occurred on the plate surface – I use this latter T/D value for calculations.

• ^The "Error" column is the Possible Plus/Minus spread in the NBL about the nominal value in the column to its left. The smaller this value, the tighter the tests used were in getting an accurate NBL. The actual NBL can actually be anywhere in the range given. If there is a "--" there, then the value to its left was a single projectile impact at the velocity given, with ">" meaning that the test did not result in a complete penetration (under the NBL) and "<" meaning that it completely penetrated (over the NBL).
• ^I am using my HCWCALC Revision 1 Computer Program for 225 BHN STS/Class "B" armor. For US WWII plates, the Percent Elongation is 25 (best) and the Plate Quality is 1.0 (default). Unless otherwise noted, the AP cap adjustments used will be those of the US Navy WWII 6" Mk 35 MOD 5 AP projectile cap (AP Cap Selection #1), which is similar in tip contour and hardness pattern to other WWII US Navy AP shell caps. These are hard caps and they do not have a deep notch in the upper corner and they - except for the caps on the 6" Mk 35 MODS 9 & 10 or 8" Mk 21 MOD 5 AP shells which are all 680 BHN for most of their volumes (answer "YES" to the 600 BHN hardness question for these) - are under 600 BHN and will have the "Cap Edge Effect" default of -12% to the NBL when the program uses it (the AP cap can dig its corner into the dent it is forming in any plate that allows this effect, causing a fold that can tear open sideways easier, making a starter hole in the armor more easily), so those questions in my HCWCALC program are "NO". Unless I find otherwise (not yet), they all will shatter just like the 6" Mk 35 MOD 5 cap at 0.66 caliber plate at right-angles (those 680 BHN caps may shatter at a lower plate thickness, but I have no data). The nose shape of most later US Navy WWII AP shells is a rounded ellipse or close to it, so they have, unless specifically stated otherwise, a nose shape blunt enough to cause the "Intact Cap Effect" to occur, where an unshattered cap that is knocked off - at or above 0.0805-caliber plate thickness for any steel plate at any obliquity for all US Navy AP shells - slides sideways on the nose at high obliquity and causes a wider hole against thin armor, increasing the NBL, so this question is answered "YES". At the moment I am not attempting to adjust for projectile nose shape under the AP cap. The blunt US AP shells have a small loss in penetration ability at low obliquity (under 30 degrees) compared to the standard US Army WWII 3" M79 AP Shot baseline projectile, but get noticeably better than that - that is, a lower NBL - as obliquity goes up above 30 degrees, being best at 50-60 degrees and then gradually going back up toward the standard NBL for the M79 nose shape, reaching it at 80 degrees, because the shells are all hitting first on their lower noses at such high obliquity and by the time the long slot being torn in the plate has opened up so that the nose tip can be pushed into the plate, the details of tip shape are not as important.
• ^

  The "Percent BHN Change" column is the percent adjustment up or down of the NBL when using the British BHN Effects Formula developed during WWII using roughly 0.5-3" AP projectiles at low obliquity against British "Machinable Quality" nickel-chromium armor (their equivalent of US Army "Rolled Homogeneous Armor" used in US WWII tanks) plates of 0.25-3" thickness with BHN values of 250-450, many of the projectiles being scale models of the British uncapped 2-Pounder AP Shot projectile, the original one being 1.5648" in diameter (0.01" less than the gun barrel’s 1.5748" = 40mm bore). STS and Class "B" armor are similar, though usually softer. This formula is:

  (W/D³)(NBL)² = {747 + (43.4)(T/D)(B)^0.5 - [54000/(B₀(D) - B)]}²

  where I am adjusting the results from the HCWCALC’s NBL and 225 BHN default and where B and B₀ are BHN values (the former for the plate),

  B₀(D) = 500 - [(160)Log₁₀(D/1.5648)]

  obviously based on the 2-pdr. as the baseline standard projectile (Bo = 347.773 for 14" AP shells). This formula is found in the document of the WWII US Office of Scientific Research and Development (OSRD), National Defense Research Committer (NDRC), Division 2, entitled Technical Report Volume 1, "Effects of Impact and Explosion" (Washington GPO, 1946) (originally CONFIDENTIAL but declassified in 1960), Part 3, Terminal Ballistics. I am extrapolating the formula down to the minimum hardness of this kind of armor when fully annealed, that is, about 190-200 BHN (about Tensile Strength = 97-102 thousand lb/sq.in.). Using these BHN modifier results for projectiles much larger than 3" or plates softer than 190 BHN may give significant errors, which is one of the things that I am trying to find out here.

• ^"~" means that this value is estimated from the known other hardness/tensile strength value, based on a linear adjustment with 225 BHN = 115,000 lb/sq.in.
• ^"Lamin" means "Lamination" where the plate has one or more weak layers aligned parallel to the plate face, which divide the plate as though it was made of several different plates clamped tightly together. This weakens the plate’s total resistance – how much depends on how many layers and how weak the boundaries are. Rolling deck and turret roof plates during manufacture spreads out such weak spots, though laminations in thick plates, such as tapered lower belt and very thick turret face plates, is not due to this, since they are forged under pressure or hammers, not rollers.
• ^8.18.2"Eff" means that the damage did not damage the steel surface surrounding the cavity or adversely affect the base plug’s seal or its fuze hole. The projectile’s explosive capability is still fully "effective" after the impact. Similarly, "Inef" means "Ineffective" where the cavity is cracked or broken open somewhere ("Broke in 2" usually implies this) or the base plug is damaged considerably, no longer screwed in tightly, or its fuze hole is crushed ("Base Plug Out" automatically means Ineffective.). These will only be noted when the damage is such that it will usually cause the other damage. "Base Cracked" damage is usually where the driving band ring is cut into the projectile ("band score"), since it is weakest there.
• ^Red NBL values are those where the computed and test values conflict, when using the plate and projectile characteristics being used in my HCWCALC Program. Further analysis is needed here.
• ^This was one shot, but must have been just slightly over the NBL with no significant travel after exiting back of plate. Rare.
• ^11.111.211.311.411.511.6This is one plate (Carnegie KK405). These 6 hits barely fit plate with enough separation for good results!
• ^12.112.212.312.4This is one plate (Bethlehem 4A443A1).
• ^13.113.213.313.413.513.613.713.813.913.10Special Treatment Steel (STS) is the BuC&R/BuShips equivalent of BuOrd Class "B" homogeneous, ductile Krupp-type chromium-nickel steel armor. Normally, Carnegie/Carnegie-Illinois (part of US Steel Corporation) had the exclusive contract to make this armor for all portions of a warship requiring such anti-projectile/anti-fragment protection (decks, thin splinter screens, the belt armor of cruisers that did not use BuOrd Class "A" face-hardened armor, and so forth) that did not fall under BuOrd’s management, such as gun mounts, gun turrets, barbettes, conning towers, and anywhere Class "A" armor was used, including belt armor. Usually, STS armor was not more than 6" thick at most, usually less. These very thick plates of "Heavy" STS, tapered and single-thickness, are exceptions, being made by all three armor manufacturers, including Bethlehem Steel and the Midvale Company, who make almost all BuOrd armor of any kind. This seems to imply that BuC&R/BuShips did not have specs for armor of that thickness and created special contracts using BuOrd specs and, thus, allowing all BuOrd contractors to bid for the contracts. Homogeneous armor of the thicknesses used here are not regular battleship amidships belts, since those are all Class "A" armor at the waterline and the tapering anti-diving-projectile lower belts used in the new WWII-era US battleships are 12" thick or less at the top, getting thinner as you go downward. Thus, these STS plates must be intended for the side protection of the rudder engine rooms at the stern of the ship, though why STS and not Class "A" armor was not used here, as with the main belt, I do not know. Perhaps it was because these side plates are mostly below the waterline and can usually only be hit by diving shells with at least a circa-15-degrees or more angle of fall (flatter trajectories rarely can go any real depth before skipping off the ocean, except for the Japanese Type 88/91/1 diving AP shells of WWII, which can go deeper below the surface for longer horizontal distances at angles of fall down to near 7 degrees), so they were very rarely exposed to direct, full-velocity, unyawed hits by capped AP shells at lower angles of fall, as the waterline belt was.
• ^14.114.214.3These two shell tests are against the same plate, with/without the Cap Edge Effect being applied in the British shell case (it always is in my computations with US AP shells), where the cap at obliquities above 50 degrees or so digs a transverse notch into the plate on initial impact and makes splitting open the slot being formed in the plate easier, in the US case by -12% to the NBL (from extensive tests with the US Navy 6" Mk 35 MOD 5 AP shell, using an AP cap with a hardness and face very similar to the larger US Navy AP shells). The Hadfield AP cap is contoured to the shape of the nose much more than the US cap, so the corner where the face of the cap meets the side skirt upper edge is not as sharp as with the US Navy standard AP cap design - US cap face is a 100-degree-wide cone with a rounded tip or a flat-faced plateau covering the tip (called a "meplat" after the French term). In these tests the British projectiles ricocheted with only a 4"-deep gouge (no through opening) and broke up, indicating that they were not making any kind of noticeable notch in the plate at the indicated striking velocity, so this effect is either not happening at all or is much weaker. Note, however, that these British projectiles are much softer in their middle and lower bodies than the US AP shells, so they tend to bend when put under high stress, especially sideways stress, as here. Thus, the AP cap shape may not be the cause of this difference between US and British AP shells.
• ^Experimental Carnegie MONTANA Class "B" Barbette Plate GG609. "Upset" = Projectile body shortened and widened.

This article is copyrighted by Nathan Okun and is reproduced on NavWeaps.com with permission.