How is "armor manufacturing flaws" modelled?

[Vanir Ausf B]

apparently 1 tank per run ain't very efficient.

:eek:

Not unless you're into self-flagellation.

[Skwabie]

:eek:

Not unless you're into self-flagellation.

yea well watching the shooting itself was enjoyable for a while at least:rolleyes:

reminds me I did read most of your other tests and there was one about weapon mount vs turret front hits. From what I can see 76mm hits on the circled area are registered as "weapon mount" and they all bounced. Visually at least that part's armor is much thicker. So mayhaps it could be just a hit text naming thing..

[Vanir Ausf B]

The "weapon mount" is what the game calls the mantlet. The circled area is called the breech ring, which is bolted to the mantlet. The mantlet itself is a single plate that covers virtually the entire front turret area. The mantlet varies in thickness, but is only thin enough to be penetrated by US 76mm APCBC around the edges. But those edges are backed by a second plate 100mm thick that is the "front turret" armor proper. For a hit to penetrate the mantlet into the interior of the tank would require a direct hit on either the machine gun port or the gunner's sight. To hit the turret armor without going through the mantlet edge first (assuming equal elevation and 0° offset) would mean hitting the extreme outer edge of the front turret armor plate which would tend to deflect the projectile outward due to edge effect, or to hit the turret ring which would be hard to do without a ricochet given than it is deeply recessed.

IMO the proportion of hits on the front turret area of the Tiger that hit the "front turret" armor rather than the mantlet is much too high, particularly when the Tiger is hull down (and why being hull down so dramatically affects this proportion is a mystery in itself).

BTW, where the game graphically shows the round impacting on the tank is not reliably accurate as shown here.

[Skwabie]

...

Yep did read all that, also found the mantlet description/drawing in rexford's book.

I think probably what they did is give the "breech ring" 100+ thickness and rest of the area 100 so a 76mm shot can pen anywhere beside breech ring. Just my total guess anyway. Heck long as anyone takes pride in taking out tigers reckon there ain't much difference. Wasn't that thread you quoted dug up a bug where the upper front hull is only 15mm thick and got penned everytime, altho it was later fixed.

[Vanir Ausf B]

Heck long as anyone takes pride in taking out tigers reckon there ain't much difference.

Then why are you worrying about armor manufacturing flaws? :confused:

[Skwabie]

...

a sarcastic smiley at the end of that one would've made it clearer

[Amizaur]

Testing done. Here is the results from my shatter gap testing done on Tiger I mid

Sherman76 vs Tiger mid at 500m and 10° horizontal offset

Hits on driver plate: 105

No damage: 6 (6%)

Spalling: 47 (45%)

Partial Penetration: 50 (48%)

Penetration: 2 (2%)

And here are the results of the test I just did which was identical except for the Tigers were late models:

Sherman76 vs Tiger late at 500m and 10° horizontal offset

Hits on driver plate: 146

No damage: 11 (8%)

Spalling: 45 (31%)

Partial Penetration: 83 (57%)

Full Penetration: 7 (5%)

Not a dramatic difference, but a significant one, which is about what I would expect.

I'm not sure what the conclusion is. Is the shatter gap modelled, or not ?

In a similar test done at LONGER range, would be the proportion of penetrating hits HIGHER that from 500m test ? Because this is how the shatter gap should work.

My second though is - I wonder what it an average expected variation of results between two identical tests done with 100 samples, if chances of various events are for example 7%, 35%, 52% and 3% ?. I feel it may be well several percents. So, unfortunately, it may be just coincidence, and if the tests were done with for example 1000 samples, the results would be almost the same.

I'll try to remind/dig out some statistical formulas and calculate how much samples are needed to get results accurate to xx% with xx% certainity.

[Amizaur]

edit: I can't find correct formulas to calculate what number of samples would be needed to get results with specified accuracy and certainity.

But I wrote a simple basic program that simulated your 100 sample test.

The program just randomised a number between 0 and 100, and decided if that number means penetration, partial, spalling or no damage.

The threeshold values were set in a way:

Penetration 97 ... 100

PartialPen 45 ... 97

Spalling 7 ... 45

NoDamage 0 ... 7

... to get chances for various events shown below:

Chance4NoDamage=7

Chance4Spalling=38

Chance4ParialPen=52

Chance4Penetration=3

The above was repeated 100 times and the summarical results were printed.

Here are results of just 3 runs of that program:

PercentOfNoDamage = 4

PercentOfSpalling = 31

PercentOfPartialPen = 62

PercentOfPenetracji = 3

PercentOfNoDamage = 3

PercentOfSpalling = 35

PercentOfPartialPen = 58

PercentOfPenetracji = 4

PercentOfNoDamage = 6

PercentOfSpalling = 43

PercentOfPartialPen = 49

PercentOfPenetracji = 2

As you see the random error of measurement is quite big, the percent of Partialenetrations you get with 100 sample test may be as well 49 as 62 and this is pure statistical fluctuation in a random proces.

Unfortunately 100 samples is way too little data to get reliable results and draw any conclusions from it.

100 samples is enough to notice big difference, like between let's say 20% and 60%, but with 100 samples one cannot see a small difference between for example 55% and 65%. The measurement error is too big. It would take at least 1000 samples of data to see such small differences in probabilities.

[Vanir Ausf B]

Then someone else can do it if they wish. I don't have that much time to spend on it.

[Amizaur]

I know that doing 1000 sample test is very time consuming. I did such testing, too.

Your test is great and respect for you that you are testing things at all, trying to check/verify various things and recreate game mechanics.

Only I want to say, that 100 samples is - unfortunately - not enough to draw conclusions from them.

The differences in results up to about 10% may be as well something we look for, as a pure coincidence.

On the other hand, if we were looking for something that gives a 30% difference, then a test of even 100 samples can show it.

Fortunately, if 10 people using the same scenario done a 100-sample test, the results could be added up and treated as single 1000 sample test. Making 100 sample test is not that much time consuming, it's quite fun. Making 1000 sample test is really boring - so maybe few people would like to participate, to make a group of testers, which would do 100-200 sample tests that combined together would be 1000-2000 sample tests ? You would only make a good scenario and describe the methodology of writing down the results.

I could participate sometimes, too.

Let's see if 1000 sample test is enough - or rather, what kind of accuracy can we get with it?

I modify my program to do a serie of 1000 sample tests, let's see...

Results:

---------------------------------------------

PercentOfNoDamage = 5.7

PercentOfSpalling = 40.1

PercentOfPartialPen = 50.4

PercentOfPenetracji = 3.8

---------------------------------------------

Results:

---------------------------------------------

PercentOfNoDamage = 7.2

PercentOfSpalling = 36.4

PercentOfPartialPen = 53.3

PercentOfPenetracji = 3.1

---------------------------------------------

Results:

---------------------------------------------

PercentOfNoDamage = 6.7

PercentOfSpalling = 39.8

PercentOfPartialPen = 49.7

PercentOfPenetracji = 3.8

---------------------------------------------

The "true" (expected) values were:

NoDamage chance = 7

Spalling chance = 38

PartialPen chance = 52

Penetration chance = 3

As we can see, now the results are usually within ~+/-2% from the true value. The smaller numbers (NoDamage and Penetration) lie even within +/-1%.

I did 20 rolls, the results spreaded a little more, within 2...3%.

That means that with 1000 samples we have 0.95 certainity that the error in our test is less than about 3% (2% for small numbers).

With 100 samples we have 0.95 certainity that the error is less than about 10% (5% for the smaller numbers).

I hope this estimation could help in deciding what number of samples is needed in a test to get significant results.

[Vanir Ausf B]

You have to test both the mid and late models separately so you're looking at 2000 hits. I've done that many and more before. If you added up all the testing I did for my Tiger mantlet thread it would be in the 4-5 thousand range. But it takes entire days to do that and I just don't have the time now. I am also not feeling tremendously motivated given that I have not seen much evidence that suggests there is anything wrong with how flawed armor is modeled in the game.

[Amizaur]

I have to admit that after seeing your test results I, ehem, forgot about armor flaws and thought about shot shatter gap - wondered how to see it. That's where my question about shattering came from...

Anyway, watching the results I saw very little difference between both tests and wondered if it's enough to draw any conclusions... It showed, that the error is about 10%.

Now I tried your test setup but moved the Tigers to 1000m. I did only about 30 samples, but it was obvious that the results changed dramatically, most of the hits being ineffective (~50%), ~30% spallings and ~20% of partial penetrations. Difference in penetration between 500m and 1000m for 76mm gun is about 10% and it was EASILY visible even with small number of samples...

What difference should we expect from armor flaws modeling ? Minus 10% or mayby minus 5% of armor resistance ? It would have similar effect like changing the range by 250-500m.

I believe that such difference in armor quality would be easily noticeable in test results (the proportion between spalling/partials/full penetrations would shift by more than the possible 10% error).

I have written down 33 front superstructure hits from your initial test setup (savegame - Tiger Late, 500m):

1 ineffective hit,

7 spalling,

25 partial penetrations and

0 full penetrations

And a lot of ineffective hits on "front upper hull" (whatever it is)

[Michael Emrys]

..."front upper hull" (whatever it is)

I've always taken that to mean the vertical plate before the driver's face.

Michael

[Amizaur]

Then what all the "front superstructure" hits means ?

The "front superstructure" hits were showing as "no damage", "spalling", "partial penetrations" ect.

The "front upper hull" hits were all "no damage" with no exeptions.

I assume "front upper hull" here is the nearly horizontal plate covering the front transmission.

The strange thing is, there were a lot of "front upper hull" hits and the 70-deg plate has very small cross-section when looking from the front.

[Vanir Ausf B]

I have to admit that after seeing your test results I, ehem, forgot about armor flaws and thought about shot shatter gap - wondered how to see it. That's where my question about shattering came from...

I did more tests on shatter gap a few months ago. There was a lengthy thread on it in the CMFI forum. You can see my test results here, but the sample sizes are about the same as what I did earlier in this thread so take the results for what their worth, which may not be much but I'm done with it.

Anyway, watching the results I saw very little difference between both tests and wondered if it's enough to draw any conclusions... It showed, that the error is about 10%.

What difference should we expect from armor flaws modeling ? Minus 10% or mayby minus 5% of armor resistance ? It would have similar effect like changing the range by 250-500m.

I believe that such difference in armor quality would be easily noticeable in test results (the proportion between spalling/partials/full penetrations would shift by more than the possible 10% error).

I am assuming the reduction in armor resistance is about 10% because that is how much it was reduced on late-model Panther tanks in the CMx1 games where it was explicitly stated in the unit stats (I don't think late model Tiger Is were given flawed armor in CMx1).

[Vanir Ausf B]

The "superstructure front hull" is the driver plate. "Upper front hull" is the nearly horizontal plate below it. The latter does seem to get hit a lot considering the size of its cross section but I have always chalked that up to being an artifact of the center-of-mass aiming model.

[Michael Emrys]

I assume "front upper hull" here is the nearly horizontal plate covering the front transmission.

The strange thing is, there were a lot of "front upper hull" hits and the 70-deg plate has very small cross-section when looking from the front.

Both your points can be seen as reasonable for this particular type of tank. It would have been helpful if they had labeled that plate as 'glacis'. And you are quite right that viewed from straight on it does not have a large cross section. Could the number of hits you observed be a statistical outlier?

Michael

[Michael Emrys]

"Upper front hull" is the nearly horizontal plate below it. The latter does seem to get hit a lot considering the size of its cross section but I have always chalked that up to being an artifact of the center-of-mass aiming model.

Could be. I would expect a lot of ricochets from strikes on it. Do you observe that?

Michael

[Vanir Ausf B]

I have never seen the hit text say "ricochet" on a hit there, although shells bouncing off high into the air are common. You are right that ricochets should not be rare, at least not with solid shot AP ammo.

[Skwabie]

yep hits there, "upper frontal hull", every one of 'em are bounces. I suppose nothing goes thru that particular plate due to the extreme angle..

a shame rather that the front of the tank is shaped like that. If they could have connected the top edge of the hull to the front nose, like the JS-2 -> 2M modification, not only the frontal protection is increased, but also internal space of the tank.

Gotta marvel at works of human minds and how simple technology enhancements change them...

Test goes on, 228 Tigers destroyed so far, 4.88 shots per for tiger late, 4.67 shots per for tiger mid. All are real time.

[Amizaur]

"a shame rather that the front of the tank is shaped like that. If they could have connected the top edge of the hull to the front nose, like the JS-2 -> 2M modification, not only the frontal protection is increased, but also internal space of the tank."

They could not, from two reasons. First is quite funny - when the Tiger I design was made, they didn't have on hand an MG mount that could be installed in a sloped front plate .

It was developed later for Panther tanks.

Second, IIRC the nearly horizontal plate had to be relatively thin because it could be raised and it was some kind of opening to the front transmission compartment (correct me if I'm wrong) needed for field repairs and maintenance.

And at last - when Tiger I was designed, no one thought about any enemy gun that could penetrate the vertical 100mm plate. There was no need to inrease protection, 100mm of armor was considered enough. If they used design, they would use thinner plates . For example it could be 60mm if it was sloped to keep the effective protection level around design target of 100mm RHA .

Anyway, if they retained the 100mm thickness and made it sloped - the Tiger would be heavier than 56 tonns.

Vertical armor layout was easy to design and simple in manufacture, they could use many elements similar like in PzIV (vision slots ect) only make it thicker. The armor design wasn't any brilliant piece of engineering, they made it just thick and solid and good quality, believed it's a good decision. Later as it turned out, it wasn't. There are more optimal designs. 80mm sloped 50-60deg - like in Panther - would be a better decision. Sloped 80mm plate offered level of protection that more than 1.5x better than vertical 100mm.

IMO if the designers of the Panther tank knew how good protection will Panther's front hull 80mm/55deg armor offer in practice, the Panther would have 60mm/55deg armour instead . I'm sure they didn't plan level of protection of 150-180mm of RHA for hull, they thought about 100-120mm of RHA so similar to the level of protection offered by front turret armor.

Or maybe designers knew that, but some officials like Hitler himself, wouldn't agree for 60mm front hull armor for Panther.

I heard that Hitler was disappointed by fact, that "his new tank" Panther has "only" 80mm of front armor, he said it has to be 100mm! Only when they said to him that sloped 80mm is as good if not better than vertical 100mm, he accepted that

Back to the Tiger I basic armor design - any serious changes in existing project were unpractical (too much redesign needed if front plates were touched, it would delay production lines) and forbiden.

Instead of that, they designed an improved version, increased the front protection using sloped plates (using a newly developed MG mount) - but this project was a too-heavy tank called Tiger II, so a completly different story.

[c3k]

If you look at the Tiger I, it's just a PzIV on steroids. (Yeah, I know, but the BIG picture...) Tiger II is a Panther on steroids. Sloped, clean lines, vs. vertical, (with viewports and access panels, and bigger/thicker is better).

Metal working is labor intense. A rough piece that fits is easier to manufacture than the tight tolerance stuff. Look at how the Germans finally learned to make the overlapping sides a bit big. Far easier than grinding hardened armor plate to size. The amount of rework they did on the early pz families was amazing. The design of the Panther and Tiger II incorporated those manufacturing lessons.

But that's a side discussion. Sorry 'bout that.

Based on my AAR performance, I should stay away from any and all armor threads.

[Michael Emrys]

Anyway, if they retained the 100mm thickness and made it sloped - the Tiger would be heavier than 56 tonns.

But not by much, since the single sloped plate would be a bit smaller than the combined area of the two existing plates.

The rest of your discussion strikes me as right on, especially the part about the ease of design and construction. A lot of armor layout before the war and during the early part was pretty unsophisticated, basically just armored boxes without much thought given to ballistic properties. "You need more protection? Just make it thicker." Which actually is what they did with the Tiger.

Michael

[MikeyD]

I sometimes get the impression that armor grogs imagine WWII gunners had detailed penetration tables at their elbow in the fighting compartment and pamphlets on German armor quality handed out by the Signal Corps. But fighting on the front lines seems to have been done largely 'by the seat of the pants', to use the phrase. Direct experience, scuttlebutt, rumors, horror stories, and observation of the wreckage you pass on the road. And the occasional pamphlet put out by the Signal Corps, often involving cartoon characters. I doubt the results of British firing experiments of Tiger 1 hulls filtered immediately down to the troops in the field.

[Skwabie]

very new to all land combat/CM series myself, but don't think anybody really need the numbers when fighting the tanks where all it matters is tactics, even if you do i suppose it's some kinda mental insurance.

what's more curious is the damage modeling... 8 tigers and 8 fireflies in a 1000m standoff. the tigers usually win but why, as they both can freely penetrate each other. we see enough discussion on armor thickness and round penetrations but how is a tank considered knocked-out, more importantly how is it decided in-game..