Is T-72B3 Damaging model at close Ranges accurate compare to SB Wiki?

[Assassin 7]

So the SB Wiki saids this: Armor protection:
Frontal Turret Armor: ~520mm vs KE, ~600mm vs HEAT, +K5 ERA
Frontal Hull Armor: 190mm-400mm vs KE, 190mm-480mm vs HEAT, +K5 ERA
(SB documentation)

 

And the ammunition being fired in these cases are:

L44 DM-53-

120mm RM: DM53 APFSDS-T (L44)

4000

750

1650

1999

L44 M829A3-

120mm M256: M829A3 APFSDS-T

4000

840

1555

2003

 

Several Users have noticed that the T-72B3 seems to be taking close Range hits with no Damaging. Is this Damage Model still a WIP? or is this Normal? I am going off what the Wiki states in theory the M829A3 and DM-53 should have no issues causing damaging to the B3 when shot under 2000 meters. Here are the screenshots, some of them are under 1000 meters.

Edited August 15, 2022 by Assassin 7
added pics

[Ssnake]

The damage model is no longer "work in progress". Let's call it "consolidated" at this point (armor models in Steel Beasts never are truly "finished"; of course, if new information comes to light, we will adapt the data). Right now, this represents our best estimate.

 

As always, the locality of a hit is shaping the outcome a lot. Add to that the variability of vulnerability depending on how many rounds the vehicle may have in store, something I cannot comment on from the screenshots above.

"Luck" plays a role as well (but as always, the Lady prefers the prepared). So if you hit spots that are relatively thickly armored (front, peripheral blows) and if you then happen to NOT hit stowed ammunition in the crew compartment, the results may not appear very impressive (or overly impressive, depending on your perspective).

 

Here's an example where things wents very badly for the T-72B3 in a very short amount of time:

 

 

Your screenshots don't say anything about engagement distances but I suspect that they were >3,000...3,500m-ish?

 

So there may have been a number of contributing factors - good tactical placement of the T-72B3s, impact locations favoring the target, long engagement ranges reducing the kinetic energy of the attacking projectiles somewhat, and a fair bit of luck as well.

[Assassin 7]

The last 6 screenshots, shows the shot and then the distance which is less than 1000 meters.

Edited August 15, 2022 by Assassin 7

[Ssnake]

Yeah, I saw that later. Nevertheless, some of the hits were where the armor is strongest, at shallow anfles that multiply the LOS strength,  and at angles that don't exactly promise great damages if they go through. Not all. Sometimes, the other side simply is lucky.

[Jartsev]

12 minutes ago, Assassin 7 said:

The last 6 screenshots, shows the shot and then the distance which is around 1500 or less meters.

Damage model, ammunitions specs and their handling were not changed since 4.268.  You would see exactly same results by firing very same ammo to very same spots at very same angles against T-72B in 4.2 because LOS thickness matters. Also please note that any projectile, rocket or missile flagged as 'Tandem'(e.g. this applies to  DM53/63, M829A3) would ignore any ERA in 100% of cases in game, while in real life there are some nuances and probability might be significantly lower.

[Assassin 7]

Just now, Ssnake said:

Yeah, I saw that later. Nevertheless, some of the hits were where the armor is strongest, at shallow anfles that multiply the LOS strength,  and at angles that don't exactly promise great damages if they go through. Not all. Sometimes, the other side simply is lucky.

Ok thanks 👍

[Assassin 7]

8 minutes ago, Jartsev said:

Damage model, ammunitions specs and their handling were not changed since 4.268.  You would see exactly same results by firing very same ammo to very same spots at very angles against T-72B in 4.2 because LOS thickness matters.

I understand what your saying but from my understanding around the driver hatch and Turret Ring are some of the weakest spots for the T-Tanks as Western Tank Gunner’s are trained to shoot center mass. Seeing a shot center mass at those Ranges and not causing any damage whatsoever is hard to believe. At least a crew injured or damaged to components. But I understand your point and thanks for the responses

Edited August 15, 2022 by Assassin 7
Added at those ranges

[Jartsev]

11 minutes ago, Assassin 7 said:

I understand what your saying but from my understanding around the driver hatch and Turret Ring are some of the weakest spots for the T-Tanks as Western Tank Gunner’s are trained to shoot center mass. Seeing a shot center mass at those Ranges and not causing any damage whatsoever is hard to believe. At least a crew injured or damaged to components. But I understand your point and thanks for the responses

Simplified explanation of rather complex mechanics:
To more-or-less reliably inflict damage you need to hit internals; if projectile  slices through without touching them,  then there is a still certain probability of inflicting some damages including catastrophic ones, but that is a roll of the dice essentially, e.g. a matter of your or opponent's luck. This applies to any vehicle in game

[Assassin 7]

9 minutes ago, Jartsev said:

Simplified explanation of rather complex mechanics:
To more-or-less reliably inflict damage you need to hit internals; if projectile  slices through without touching them,  then there is a still certain probability of inflicting some damages including catastrophic ones, but that is a roll of the dice essentially, e.g. a matter of your or opponent's luck. This applies to any vehicle in game

This Video has been shown before but of course from a different T-72. As shown taken a hit from a Sabot cause structure damage and in this case if the Ammuntion compartment wasnt hit I would image that the Driver would have been killed and the Turret would not have been able to Traverse anymore. 

 

Edited August 15, 2022 by Assassin 7

[Jartsev]

36 minutes ago, Assassin 7 said:

This Video has been shown before but of course from a different T-72. As shown taken a hit from a Sabot cause structure damage and in this case if the Ammuntion compartment wasnt hit I would image that the Driver would have been killed and the Turret would not have been able to Traverse anymore. 

First of all, since 4.268 grazing impacts  (finest example is driver's  compartment roof of T-72) are essentially diverted in to vehicle's interior(LOS thickness is  reduced), if impact angle was close to zero degrees. That change was  made on your request. Now please look what I wrote above about  damage probabilities if no internal components were  hit directly. Sure  it is possible go  further with simulation of penetrator's normalization, but it is not a T-72 would suffer of such changes most.

 

Edit:
typos and grammar

Edited August 15, 2022 by Jartsev
typos and grammar

[Assassin 7]

23 minutes ago, Jartsev said:

First of all, since 4.268 grazing impacts  (finest example isdriver's  compartment roof of T-72) are essentially diverted in to vehicle's interior(LOS thickness is  reduced), if impact angle was close zero degrees. That change was  made on your request. Now please look what I wrote above about  damage probabilities if no internal components were  hit directly. Sure  it is possible go  further with simulation of penetrator's normalization, but it is not a T-72 would suffer of such changes most.

Ok something has changed with the damaging Models since this upgrade as we are seeing different results, Several of us have noticed it too. I find it hard to believe that it is the same since 4.268. As far as the T-72B3 and other taking center mass hits, ok I got it and it based on a Dice roll of luck and probability. Thanks for the answers and your Team time.  

Edited August 15, 2022 by Assassin 7
Grammar

[Ssnake]

The armor model hasn't changed (other than introducing the T-72B3 which is obviously a change). The interaction between target and projectile has changed slightly. Damage probabilities may have changed slightly, but certainly not in a radical way. To be honest, while second-guessing the damage model is legitimate, I think this is one of the cases where people read too much into single incidents to start with, and maybe also have exaggerated expectations about what a real-time simulation can deliver, or what the quality of available, unclassified base data is.

Once that a projectile perforates the outer shell, what happens inside is no longer an exact science. It can't be, with hundreds of vehicles simulated in Steel Beasts. We place components inside where we know them to be. We do not follow each and every fragment of projectile and armor material that may be bouncing around inside, possibly also tracking its temperature, shape, initial trajectory, and elastic and unelastic collisions with internal components and their shapes. We set a certain damage percentage that "feels right" and scales somewhat with the residual energy of the projectile as it enters certain compartments or components.

 

Would an APFSDS entering the turret ring area kill the driver with 90% likelihood? Yeah, maybe.

Why not 91.8%? 88.35%? 94.8882% on a Wednesday afternoon with wind coming at 5m/s from the rear left?

Who knows!

 

There are no data. Where you get declassified experimental results (such as from the Conqueror trials) they are 50 years old, or more. Above all, these experimental results cannot be directly transferred, they aren't standardized even where they are available (and mostly, they aren't).

The results from Steel Beasts can only be trusted in aggregate. Shoot a T-72 a hundred times, and in 70% of the time it'll blow up. 100% of the time, if you manage to hit the right spot. But outside of calibrated experiments, in virtual combat, lining up shots with such precision simply doesn't happen. The lesson to take home is, sometimes strange shit happens. If it happens often, there's reason for concern. But then you need to run your shootings at least two, better four dozen times.

[Assassin 7]

16 minutes ago, Ssnake said:

The armor model hasn't changed (other than introducing the T-72B3 which is obviously a change). The interaction between target and projectile has changed slightly. Damage probabilities may have changed slightly, but certainly not in a radical way. To be honest, while second-guessing the damage model is legitimate, I think this is one of the cases where people read too much into single incidents to start with, and maybe also have exaggerated expectations about what a real-time simulation can deliver, or what the quality of available, unclassified base data is.

Once that a projectile perforates the outer shell, what happens inside is no longer an exact science. It can't be, with hundreds of vehicles simulated in Steel Beasts. We place components inside where we know them to be. We do not follow each and every fragment of projectile and armor material that may be bouncing around inside, possibly also tracking its temperature, shape, initial trajectory, and elastic and unelastic collisions with internal components and their shapes. We set a certain damage percentage that "feels right" and scales somewhat with the residual energy of the projectile as it enters certain compartments or components.

 

Would an APFSDS entering the turret ring area kill the driver with 90% likelihood? Yeah, maybe.

Why not 91.8%? 88.35%? 94.8882% on a Wednesday afternoon with wind coming at 5m/s from the rear left?

Who knows!

 

There are no data. Where you get declassified experimental results (such as from the Conqueror trials) they are 50 years old, or more. Above all, these experimental results cannot be directly transferred, they aren't standardized even where they are available (and mostly, they aren't).

The results from Steel Beasts can only be trusted in aggregate. Shoot a T-72 a hundred times, and in 70% of the time it'll blow up. 100% of the time, if you manage to hit the right spot. But outside of calibrated experiments, in virtual combat, lining up shots with such precision simply doesn't happen. The lesson to take home is, sometimes strange shit happens. If it happens often, there's reason for concern. But then you need to run your shootings at least two, better four dozen times.

Understood and thanks for the response, FYI I have seen these results while playing several different missions.At first I was skeptical about the results considering I was just having a ton of bad luck but after seeing the same results at semi-close Ranges such as around 1000 meters and under, I decided to report this. I’m not honestly trying to make a big deal but it just didn’t seem right and from being a previous Armor Crewmen and Mechanic and etc. Thanks again

[Ssnake]

Well, if you do encounter it more often, I suppose we're going to look into it. That's a piece of info you didn't mention yet.

[Volcano]

I suppose it is only natural for something new to be scrutinized with a microscope. 😵  

 

First of all, that SBwiki page literally has so far just been a few minutes of hasty copy pasting and retyping from the T-72B1's page.  The stuff about the protection was something that I just typed over from what was already there, so it certainly isn't reliable enough to hold as some kind of ultimate truth. Although it does say "+K5 ERA".

 

However, the T-72B3 armor model is finished and fully functional; it is NOT some kind of work in progress. It is quite literally a T-72B1 with K5 ERA instead of K1 ERA, and some other small obvious differences. The math is exactly the same, the values are exactly the same, the damages are exactly the same. The main difference is that the K5 ERA does provide significant protection over K1 ERA and is exponentially better, especially because its wedge shaped, and provides significant KE protection over K1 ERA. Now of course, as explained, the ERA is a very tricky thing to represent, and you can't really get it 100% correct in a computer simulation, in relation to KE, HE, HEAT, but its probably about as good as it can be to get the desired effects.

 

So, there is nothing to it really, no bells, no whistles, no extra ordinary things, it is what it is - essentially not much different than the other T-72 armor models in general, but specifically the T-72B1, but with better ERA. 

 

I did a few test firings with M829A3, firing into the T-72B3 turret front thickest part of the tank (through K5 ERA), and it does penetrate with crew casualties. Those results are expected.

If you are getting no effect type impacts on the turret, then likely you are hitting it low on the turret front, just above the turret ring. In this area there is a is a horizontal "lip" on the vehicle, where the turret ring extends from and the inner turret wall extends to (connecting the two, kind of like an inner armor floor). Here you would be impacting across a sort of flat surface that extends from the turret ring itself, towards the inside of the vehicle. This would be directly similar to hitting the roof of the driver's compartment at an extreme angle. This surface is not very thick (the thickness of the turret roof), and we do cap the impact angle multiplier now since 4.2 or 4.3 (a huge improvement), but right behind that lower "lip" it would then pass into the the back wall of the front turret armor, so you would also be passing through the entirety of that too - three difference surfaces. This is the absolute worst place to hit the turret.  This is true for all T-72s, and isn't an error in modelling, although its probably at the upper limit of what a simulation can represent (a round is passing through multiple thick surfaces, at very different and extreme angles). Still, either way, real life or not, this would be the absolute thickest part of the turret.

 

 

[Assassin 7]

6 minutes ago, Volcano said:

I suppose it is only natural for something new to be scrutinized with a microscope. 😵  

 

First of all, that SBwiki page literally has so far just been a few minutes of hasty copy pasting and retyping from the T-72B1's page.  The stuff about the protection was something that I just typed over from what was already there, so it certainly isn't reliable enough to hold as some kind of ultimate truth. Although it does say "+K5 ERA".

 

However, the T-72B3 armor model is finished and fully functional; it is NOT some kind of work in progress. It is quite literally a T-72B1 with K5 ERA instead of K1 ERA, and some other small obvious differences. The math is exactly the same, the values are exactly the same, the damages are exactly the same. The main difference is that the K5 ERA does provide significant protection over K1 ERA and is exponentially better, especially because its wedge shaped, and provides significant KE protection over K1 ERA. Now of course, as explained, the ERA is a very tricky thing to represent, and you can't really get it 100% correct in a computer simulation, in relation to KE, HE, HEAT, but its probably about as good as it can be to get the desired effects.

 

So, there is nothing to it really, no bells, no whistles, no extra ordinary things, it is what it is - essentially not much different than the other T-72 armor models, but with better ERA. 

 

I did a few test firings with M829A3, firing into the T-72B3 turret front thickest part of the tank (through K5 ERA), and it does penetrate with crew casualties. Those results are expected.

If you are getting no effect type impacts on the turret, then likely you are hitting it low on the turret front, just above the turret ring. In this area there is a is a horizontal "lip" on the vehicle, where the turret ring extends from and the inner turret wall extends to (connecting the two, kind of like an inner armor floor). Here you would impacting across a sort of flat surface that extends from the turret ring itself, towards the inside of the vehicle. This would be directly similar to hitting the roof of the driver's compartment at an extreme angle. This surface is not very thick (the thickness of the turret roof), and we do cap the impact angle multiplier now since 4.2 or 4.3 (a huge improvement), but right behind that lower "lip" it would then pass into the the back wall of the front turret armor, so you would also be passing through the entirety of that too - three difference surfaces. This is the absolute worst place to hit the turret.  This is true for all T-72s, and isn't an error in modelling, although its probably at the upper limit of what a simulation can represent (a round is passing through multiple thick surfaces, at very different and extreme angles). Still, either way, real life or not, this would be the absolute thickest part of the turret.

 

 

Awesome response and thank you for the explanation 👍

[Major duck]

What we saw this Sunday game was shots that hit it in the side or in places where there are no ERA and some of those B3 just tock hit after hit after hit and kept on going but we also saw Bradleys get hit by AT-11B and sabot in the cooling without being mobility hit and 5 or 6 wampir hits.
Most of us had the feeling that B3s where easier to kill at long ranges 3+ Km then at 1km or closer.
And B3s armour havent been upgraded for 30 years so a Dm53 should slice right through

 3 clicks

3 clicks

 

 1,4 km
 

 3 clicks but on ERA and a kill

 

 3 Clicks Kill

 

 AT11B No damage 🤔
 

2 clicks No damage

 

But take a look a this AAR especially C23 it have to be an Irish crew and they are covered in four leaf clover

https://drive.google.com/file/d/1peXoAED_d1Lp8kctFp4Ek6OGyslYHG1g/view?usp=sharing

 

But like Assasin says thanks for the explanation

 

 

[Volcano]

Yes, those look fine.

 

As we can see at a glance...

 

The 1st image impact at an extreme angle in the front front fuel tank area, which it would have to pass through (at the extreme angle) before even getting anywhere into the crew compartment.

 

2nd image show a round that impacts along a nearly horizontally flat hull roof, at 3km range. So, yeah, 3km you are losing a lot of penetration power.

 

3rd image is ~1.5km but is hitting along the horizontally flat hull roof, then entering the what is near the turret ring, but at that angle it wouldn't enter the interior of the turret (it would pass through the armor layer). So, its anyone's guess whether that caused a kill in your game, but I'd say, mostly likely not.

 

4th image is of a kill you said, and it should be, because its through a flat ERA and with nothing behind it but crew area (it missed the fuel tank and horizontally flat hull roof), and actually we can see the tracer of the round behind the target (it went through it).

 

5th image looks like an impact in the turret ring area, hard to say, but as you said its a kill.

 

The last two images (which appear to be the same one) appear to impact at an angle that causes the round to impact the horizontally flat surface of the roof, after the front ERA, at 2km range. Even at 2km we are talking about a loss of at least 100mm of penetration to range, IIRC.

 

The AT-11B impact on the Leopard 2 to me is the most concerning of all of them IMO. However, judging by the diameter of the hit ray (its tiny), it implies that the ATGM did not detonate properly, or at all. Possibly it was fired too close, or it could be for some other reason - perhaps it detonated on something in front of the tank and a weakened HEAT penetration impacted the tank. Either way, even if it did explode, it is hitting at an angle into the vicinity of the skirt, and seems to also be passing into the side fuel tank all before it would reach the interior. Or it could be that the ATGM didn't detonate properly (as a bug) and exploded INSIDE the fender, as opposed to on it. Who knows, something to keep an eye on though, yes.

 

There is also some error in the AAR too, its not scientifically precise. Its a snap shot of roughly the time around the impact, with a big fat hit ray in roughly the area of impact, and of course there are all sorts of considerations going on there, with network accuracy, and so on. It's really hard to hold anything as some kind of microscope level of analysis, and the only thing that can be considered an actual obvious problem - like where a T-72 would be hit in the direct side-on flank multiple times with no damages at all and a hit ray that doesn't penetrate (things like that). 

[Assassin 7]

There is also some error in the AAR too, its not scientifically precise. Its a snap shot of roughly the time around the impact, with a big fat hit ray in roughly the area of impact, and of course there are all sorts of considerations going on there, with network accuracy, and so on. It's really hard to hold anything as some kind of microscope level of analysis, and the only thing that can be considered an actual obvious problem - like where a T-72 would be hit in the direct side-on flank multiple times with no damages at all and a hit ray that doesn't penetrate (things like that). 
 

 

would that explain these shots too? As they could actually be offset from what they show? They are less than 1000m shots

[Assassin 7]

To add: we are also seen these kind of results on other Tanks as well. Such as Tank taking direct hits to the Tracks or suspension with no damage and some times it does. Seems to be randomly happening but more often than less

[Volcano]

As I explained previously, the first shot is into that turret ring area that I already described in the thread's "solution" post.  The second shot is along the horizontally flat hull roof. This extreme angle now has a cap on how much it magnifies its thickness (since about 4.2 or so), but that doesn't mean its guaranteed to have enough energy left over after penetration to do anything, it would be up to a die roll.

 

Nothing is changed here. 

 

Edit: Now it could possibly be that this maximum-angle-cap isn't and possibly has never worked completely (its difficult to test), but IIRC, all it was intended to do is allow *some* residual energy to be left over after impacting such an extreme angle, that it might be allowed to penetrate and damage something. But that is not always going to be true (its a die roll, regardless). It's certainly the worse place to hit any vehicle (along a horizontally flat armor surface), and someone posted real world images of those situations already).  

[Assassin 7]

1 minute ago, Volcano said:

As I explained previously, the first shot is into that turret ring area that I already described.  The second shot is along the horizontally flat hull roof. This has been extreme angle now has a cap on how much it magnifies its thickness (since about 4.2 or so), but that doesn't mean its guaranteed to have enough energy left over after penetration to do anything, it would be up to a die roll.

 

Nothing is changed here. 

Ok

[Volcano]

So looking at this again now (and I can't do much more than this, sorry, work has to be done, bugs have to be fixed, etc)...

 

The maximum angle cap is 10x since version 4.2. This would make the top hull roof (40mm) into 400mm.  This is very reasonable (and is probably on the low/safe side actually), and was based off the real world photos provided before, so that the round in the photo would be allowed to barely penetrate afterwards. So keep in mind that the M829A3 would be getting -400mm reduced along such a rare impact, and then the round would be continuing on to what is likely the turret front or turret ring area, which is another ~200mm at its thinnest, but could be around 600mm if it hits more towards the turret's inner armor "lip" I already described. 

 

So let's look at the example image...

 

Let's also say it hits it along the BEST path (hull roof, then turret ring in an ideal place). So at ~1000m range, you are probably losing around 40-50mm of penetration already to range, then -400mm from the horizontally flat hull roof, then ~-200mm from the turret ring, now we are talking about having about ~150mm of penetration power remaining, which is a little weaker than a single 40mm autocannon KE round (170mm KE RHA).

 

Like I said, I can't spend much more time on this, but nothing looks out of ordinary from how it all worked before in regards to extreme angles. But at least now the extreme flat angles won't absorb 999999mm of KE RHA. 

[Volcano]

I should also mention here also that internally we are investigating whether this horizontal "lip" around the T-72 turret ring (which exists on all T-72s) could be improved my making it thinner, and this would help make it less absorbent to impacts there. But first this has to be investigated to see if it is justified or not, based on diagrams and so on.

 

And besides, its nothing new to the T-72B3 here, as its identical on all T-72 armor models. But, in the interest of perfection, we will look into it as a general improvement. Either way, I will make a comment here we decided (after the investigation) whether a reduction in thickness is justified or not. Let's see. 

 

(Either way, it doesn't apply at all to the example I explained above, though, this turret ring "lip" only applies to the original discussion earlier.)

[Assassin 7]

Thanks again and we do appreciate the investigation on this. 👍