I recently found out that the Leopard 2A4 has an exaggerated armor values than it was supposed to
(All version except the Finnish Leopard 2A4 and Leopard 2PL).
Image of the Gun Mantlet, using DM33 from Leopard 2A5:
Compared to the cheek:
Bear in mind, this is what’s behind of those areas:
Compared to the Leo 2 PL, which obviously has better armor on the mantlet because of the armor package it has for a Leopard 2A4:
Has less armor value than the regular 2A4
I hope this gets fixed quick as this significantly improve the Leopard 2A4’s effectiveness and will cause problem in balancing statistics down the line if left unchecked for too long.
The armor is fine; it’s currently 490mm KE, which is fine taking into account the mantlet composite module PLUS trunnion (240mm thick by itself). In real life, Leopard 2s’ mantlet is not a weakspot; and I am glad this is finally reflected ingame.
The 800mm KE value the armor viewer gives is bugged and misleading; try firing at it with any shell with 500+mm pen and you will see it will punch through easily.
Another fun fact is: only 2A4 and 121’s mantlet has been updated! 2PL’s still underperforming (which is even funnier since it’s the uparmored one) and the Finnish 2A4 hasn’t been updated either.
I noticed this too, test shots seem off as well, it looks like its calculating the thickness of the strike face based off the combined volume of the mantlet if it were to a single convex hull, then adding the composite filler on top of that. if you pull the armor overlay it is reading a much higher higher, at ~470mm @0deg of HHS rather than the base 16mm @0deg.
This is further confirmed when shooting test rounds they are failing on the 16mm strike face.
No, tested it against several 500mm+ m900 and 3BM46 both failed tests reliably, KE-W was about 30% pen rate with little remaining spalling, was not until getting to M829A1 and A2 that they started to perforate the mantlet more that 50% of the time, but was unreliable and barely any damage. KE-W, M829A1, and M829A2 should be slicing right though that. this was done at 0m by they way so add any amount of distance and the fail rate shoots up.