I keep forgetting to post this but it was brought Jo again, so here it is. This document compares early AP and APC at 0 degrees and 30 degrees. AP across the board outperformed APC, until the better ballistics of the APC took over.
Note that it says “extrapolation of the results of trials”. Its not accurate, because we know fairly well that, for example, a 5in/30° target is immune (Army limit) from 90mm AP shells at velocities under about 2800fps, unlike this chart implies.
Anyway, I came here to address some possible nuances with oversimplified models like DeMarre(but also Thompson’s and Krupp formulas): they are, in essence, approximations, valid only for a given range of values of thickness and striking velocities. They inevitably break down and start giving unrealistic results once we go too far.
Here is what I mean: here is a comparison of accurate Thickness vs. Ballistic Limit(Navy) curve build from experimental data of 76mm M79 AP and a rudimentary model, like DeMarre’s equation (but not the actual DeMarre equation, mind you) fit to match the data. As you can see below about 0,5 calibers thickness it starts to deviate more and more, despite giving excellent fit for higher thicknesses.
Which is not to say that what you’re trying to do (to design a relatively simple and accurate model for terminal ballistics) is a lost cause, just that there are some kinks to be aware of.
I’m aware of the limitations. I’m not trying to create a historically accurate simulator. I’m trying to create something reasonable that is simple to explain and simple to implement for Gaijin.
If they fixed AP British tanks would become half decent. They cant have that.
2 Likes
It would make a lot of tanks better. I know this isn’t as fun as new vehicles or new maps but it would have a significant impact on game play.
On the topic of how much penetration AP cap steals from the round:
The ratio between BLs is almost constant and equal to +5% for the capped shell, relative to uncapped shell of the same mass as penetrator w/o cap. (M62 w/o cap and windshield, but filled and fuzed should weight 12.64lb.)
Edit: It appears that this effects is greater when obliquity is increased from 0° to 20°, but the sharp drop at 30° against thick plate is difficult to explain.
From canadian data, it appears that the capped shell has a fixed +8,5% increase in critical velocity both at 0° and 30° angle:
3 Likes
The 5% checks out, according to my calculator suggestion.
I’m not going to pretend to understand all the mechanics of AP vs APC vs APHE. I’m just trying to make something that gives reasonable results and requires limited changes to the game mechanics. Gaijin chose the Demarre system for their AP penetration. I think its overly simplistic but that’s their decision. My suggestions allow them to use their simple system but generate reasonable results.
There are other formulas and calculators but they are well beyond my understanding and my ability to replicate.
Might be interesting to some:
1 Like
It’s absurd this keeps getting passed over for suggestions with fraction of the votes or replies.
3 Likes
To clarify the APCR adjustments, here is 76mm M93 using the tungsten carbide calculator.
This chart shows about 9.7” or 246mm of penetration for the M93 but it is likely using 220 BHN armor plate.
That spreadsheet is a converted form of this chart.
It’s a little complicated, but essentially the chart gives you the penetration against RHA ranging from 220 BHN to 330 BHN.
My understanding is the game uses 240 BHN as the standard. 247mm BNH at 220 would be about 241mm at 240 BHN.
1 Like
The issue with the APCR chart is the values aren’t exactly linear, there is a slight curve to the chart. It works out to a few percent so I’m not sure if it’s worth trying to figure out how to account for the curve in the DeMarre version.
I’ve requested documents to confirm but the HVAP T29E15 (M321) and APBC-T T166 were tested with the 76mm gun. Here is how performance would look. This may be a reasonable upgrade for the late US 76mm armed tanks.
Compared to M93 and M79.
1 Like
While not entirely related to how WW2-related shells function, Gaijin’s shell penetration calculator for APFSDS shells needs to be scrapped, it is inherently flawed in that a lot of shells are unnecessarily nerfed without explanation other than that Gaijin swapped to whatever the L-O equation is. Instead, Gaijin should swap to William Oddermat’s penetration calculator available at https://www.longrods.ch/perfcalc.php
The LO equation is the Lanz Odermatt formula.
I’d also like to draw attention to naval ammunitions, which suffer even more so than tank shells from the current system. While Gaijin seems to attempt to adjust their formula to match RL data of some tank shells, they completely spoil the naval shells as they work in a very different context. TLDR, the current formula is optimised for small-calibre (compared to naval shells, all tank shells are small calibre ofc) and high-velocity (600+ m/s) impact against homogenous plates, whilst naval shells IRL were optimised for relatively low impact velocities (400~600 m/s) and against face-hardened plates. Naval shells also had to passed stricter acceptance standards than tank shells IRL. The penalty from bursting charge made up by Gaijin was entirely based on tank shell performances and it severely deteriorates naval shells’ performance at low striking velocity. According to data of various navies I have, the Gaijin formula globally underestimate post WWI naval high-calibre shell’s performance by over 20%, and in some severe cases such as the British 15" APC, the formula underestimate the penetration by nearly 60% at 1400 ft/s (396 m/s) striking velocity.
14 Likes
The DeMarre formula was created to represent penetration against nickel steel ship armor in the 1890s. I don’t think the formula is a bad option, I just think Gaijin’s implementation is not great.
6 Likes
I’m not saying De Marre formula itself is a bad option either. The thing is, this formula IRL had been modified extensively over time to fit the new technologies and ofc different context, and gaijin made up their own version too but it’s entirely based on tank context. While navy shell is a very different category they insists to treat everything “equally”. One may argue that the mismatch of tank shell performance in game is somewhat within an acceptable margin but the error in calculating naval shells is way beyond the threshold to be considered as “acceptable” imo
6 Likes
That’s fair. I would work on changes for the naval version but I don’t have much data on naval guns. I have no problem with anyone using my information for a naval suggestion, or helping if there are any questions.
I just work with the information I have. This suggestion is not even all my work. There are multiple people I have worked with to gather this information and helped to get a workable solution.
2 Likes
How hard would it be to have a separate calculator for naval rounds?