German tanks to cheap or allied tanks to expensive?

[Panzer Leo]

Well, that goes into philosophie..., i don't know either...
I made further tests. The 88L71 is in almost every reference listed with a max penetration of around 225 at 0
The 88L56 is listed with about 155 at 0
At an angle of 30, the 88L71 is listed with 200 and the 88L56 with 120.
When I test this in SPWAW, the 88L56 behaves exact as it's supposed to do...will just penetrate 120 at 30
But the 88L71 has only a penetration of about 170 at an angle of 30, not 200, as listed.
As I understand the calculation in SPWAW, everything is calculated down to an angle of 0. This means the gun penetration rate stays (at a specific range, a bit randomized) and the armor plate is calculated as if it would have 0 degrees.
This would mean, an armor plate of 200 has a resistance of about 225 at 0. But here it is a thickness of only 170 that gets this value.
When you increase the angle to 40, it goes up to
275. This means, an angle of 30 gets a bonus of 25mm and just 10 degrees more get 75mm, three times more by an increase of one third of the angle ? I know, the angle plays an important role, but this seems a bit too important...?
Is the formula supposed to work loke this ?

[john g]

Originally posted by Panzer Leo:

Is the formula supposed to work loke this ?

---

If you have trouble visualizing this, try drawing two lines on a piece of paper approx an inch (25mm) apart. Now rotate the paper 40 degrees and measure aross the lines at the 40 degree angle, you will be amazed how much more distance you have between the lines when the angle goes up. Realize that as you approach a 90 degree rotation the thickness approaches infinity. In reality it doesn't work this way 100% because not all the force is expended in a straight line as the shell impacts the armor.

The more you rotate the angle the greater the addition.

This is one reason why hills were so important during WWII, reducing the enemy armor angle by shooting down or adding to friendly armor angle by having the enemy shoot up will often make the differance between living and dying.
thanks, John.

[Kharan]

I also am a bit leery about the new OOBs, considering the source and it's prior "achievements". But if there was a list of changes, then at least we could see for ourselves.

[Panzer Leo]

You're right, John, but to me it looks like there is an unintended "jump" in the formula. Don't you think an increase of 11% for the first 30 degrees and an increase of some 60% more for the next 10 degrees to a total of 40 degrees sounds reasonable ?
Also, I would think, that the critical angle is somewhere around 45/50 degrees, as the deflected shot would not have to change direction by more then 90 degrees, when deflected, meaning a lot of energy is kept in the shot and not passed to the armor...
The second point is, that the 88L71 did penetrate up to 200mm at 30 degrees. In the game it does not. It penetrates just about 170mm at 30. This is the same for all bigger guns, so it's a formula thing...

[john g]

Originally posted by Panzer Leo:
You're right, John, but to me it looks like there is an unintended "jump" in the formula. Don't you think an increase of 11% for the first 30 degrees and an increase of some 60% more for the next 10 degrees to a total of 40 degrees sounds reasonable ?
Also, I would think, that the critical angle is somewhere around 45/50 degrees, as the deflected shot would not have to change direction by more then 90 degrees, when deflected, meaning a lot of energy is kept in the shot and not passed to the armor...
The second point is, that the 88L71 did penetrate up to 200mm at 30 degrees. In the game it does not. It penetrates just about 170mm at 30. This is the same for all bigger guns, so it's a formula thing...

---

You are making some assumptions here, 200mm at 30 of what sort of metal? What sort of armor gets a 200mm rating?

As Paul has said the armor gets transformed to equal one standard hardness armor, so 170mm of game armor may in fact be 200mm of less resistance armor. There is also a 255 limit, so as the armor approaches that number I would guess the rating is nonlinear, so that the rating follows the curve not a simple 1 to 1 rating. Do you have the complete list of criteria that they use for converting book armor thickness to game armor ratings?

Also reaize that armor testing is done with the vehicle directly in front of the gun. Very seldom is a shot at a vehicle directly head on, you will be shooting a vehicles at an oblique angle most of the time adding an additional angle to the equation.

If I really wanted to argue out picky details like this, I would be playing with the local armor miniatures group. Instead I would rather play a game that can handle the game details for me while I concentrate on commanding the force not figuring the effective armor thickness for a shot against a tank hull at some 3 angle oblique shot. Why not complain that the game assumes flat trajectory for all shots, the angle of the at round never changes vs the armor, which is yet another factor in penetrating, but not enough of a factor to worry about except at extreme ranges.
thanks, John.

[Panzer Leo]

Look at the "guns vs armor" page and you'll find your points and you'll also see, that they are included in my assumptions.
I'm not picky on some irrelevant details here, but my concern is, that the behaviour of WWII armor in combat is not really reflected as it was. I think, it should be possible to put out the heavy tanks with an adequate gun. This is how it was and that's why it should be in the game. There were no "immune" tanks in WWII.
Good you mentioned the angle. In the current formula it seems, that sloped armor of 50 degrees and higher gets really hard to penetrate and every 5 degrees make a lot difference at this level. Now try to calculate the penetration (I'm mostly relying on Lorrin's posts for slope effect) of the 88L56 vs a 70mm plate at 55 degrees. You'll see, that it does not matter wheter the 88L56 fires at 0m or 1500m, because the angle at 1500m changes, helping the penetration performence...
At the moment sloped armor seems to give some tanks a bigger bonus then they deserve...

[Paul Vebber]

Kharan: I am the ultimate author of the OOBs, I reviewed the comments EVERYONE contributed in deciding what got in and what didn't. I take you consider my ability to do that deficient? What method would you use to estimate the hundreds of data points necessary that there is no "ground truth" on? Blood was nearly spilled just trying to come to consensus on MG firepower <img src="rolleyes.gif" border="0">

Please be specific about the things that trouble you - "prior achievements"?!?!. Do you even have them (dated 24 Sept)? If not just ask...too many things have been tweaked to list- you just have to look through the files. I even lowered the vehicle FT values just for you <img src="biggrin.gif" border="0">

PzLeo:

The primary source for gun pentration is Lorrin Bird and Robert Livingstons "WWII Ballistics Armor and Gunnery". Robert is one of the primary sources of data for the guns vs armor wargamer.org site. Lorrin has been gracious enough to help me considerably with priliminary work on Combat Leader and to use include some of his formulas in that game. He has posted excellent data on the TO/E forum and is more than happy to discuss these sort of details. See my post there also.

The thing that you are not accounting for in the 170 vs 200 thing is T/D ratio. Book penetration numbers are nearly useless to use for determining if a giving piece of armor can be penetrated, because as I explain in the T/D thread up at TO&E, the penetration varies as a function of thickness to diameter ratio. The game accounts fo the variation in penetration with T/D, but it does so in "steps" because I did not have Lorrin's formulation back then and did my own curve fit to data I had to get a table of T/D adjustment factors that the game uses. The difference is no more than a couple of % between my "step values" and Lorrin's smooth functions. I only have APCBC and APCR - AP rounds are assumed APCBC and APDS assumed the same as APCR. COmbat Leader has more shell types.

Penetration in the game for 0 degrees is based on Lorrin's work where he adjusts to US 240BHN test plate and 50% probability of penetration. The game assumes that projectiles tend to fail to meet expectaiton abit more often than exceed it, and becasue there are some projectile failure mechanisms that SP:WaW does not consider that Lorrin does, the random number draw is skewed abit downward. (making it slightly less likely then 50% that "equal" pen and armor values will result in a penetration.)

As a rule of thumb, (not considering T/D) to convert a base armor and slope to "0 degree equivalent" is to take base/((cos(angle))^1.4)

The T/D is used is to adjust the "power factor" from 1.1 at very low T/D known as "overmatch" the armor tend to "plug" and fail catastrophically. At high T/d the power factor can exceed 2 as the armor tends to "flow" and compress away frm the projctile, increasing the effective resistance substantially until the ound penetrates to sufficient depth to cause the mechanis to shift to plug failure.

So depending on the T/D ratio both 200m and 170mm canbe correct penetration values. The game also seperately compute the probability of ricochet. The game also assumes that the target is not "frozen" in the same position round after round so a +/-10 Degree horizontal random factor on the horizontal angle.

The bottom line is if you agree with Lorrin and Roberts methodology, it is basically what I use (though not all the things like shatter are explicitly modeled - that will have to wait for Combat Leader). I ran several comparisons between the pridicted probability of penetration using Lorrins book, and tests using SP:WaW and the results were no worse than 15% off and most (albeit a small sample - like 7 or 8) were within 5%

If you haven't ordered WWII Ballistics: Armor and Gunner, it is required reading to understand the issues.

[ September 28, 2001: Message edited by: Paul Vebber ]</p>

[Paul Vebber]

Sorry PzLeo - I did not see you page 2 post before posting my reply.

You'll see, that it does not matter wheter the 88L56 fires at 0m or 1500m, because the angle at 1500m changes, helping the penetration performence...

---

The change in angle of attack for the incoming shell is small (a few degrees) and it is assumed that the unevenness of the ground is at least as big a factor - the vertical angle used is randomized between -6 and +3 in a "sum of 3 dice rolls" distribution.

The whole pont of T/D is exactly that the angular effect of thicker armor is more pronounced as the thickness of the armor increases and projectile diameter stays constant. The formulas in Lorrin and Roberts bookd show that.

[lnp4668]

This kinds of slightly off tangent, but will Combat Leader allows the player to select the ammo type (similar to Tiger on the Prowl)when they shoot at a target?

[Panzer Leo]

The problem I have is, that e.g. the 88L71 is after all generalization lorrin put in (and as I understand, the T/D is included in some way) it was capable of penetrating 200mm at 30 degrees.
So I tried to simulate this in the game with a custom made tank that has exactly this armor and
I was not able to recreate this penetration capability. The randomizing of the angle you mention looks like it steps in too late in the formula, as the variation in effective armor value is not nearly as much as I think it would be.
For the T/D it sounds not reasonable, that a specific ammunition is only theoretical capable of penetrating 200mm at 30, but it never makes it in reality and you still list it with 200mm not 170mm or so. As I understand Lorrin, this is to some extend included in the 200mm performence)

Example (this is rough data from Lorrin's post):

effective armor at given angle
1.15*armor listed in OOB...30
1.25.......................35
1.35.......................40
1.5........................45
1.7........................50
2.1........................55
2.7........................60

this is for APCBC and as I said just rough...

an armor of 100mm at 50 degrees would range after the angle variation of only +/- 5 from 150 to 210,
with a variation of +/-10 from 135 to 270.
The variation in the game is not nearly as much. It usually has a span of about 30mm and only decreases dramatically with a "vulnerable location hit" to about 50%.
A 88mm round fired by the 88L56 will come in on an angle 7 degrees at 2000m just calculating without additional ballistic effects, so it should be more in reality.
A decrease of 7 degrees at an high angle plate does a lot to the penetration performence as listed above. It can make up for 80mm effective armor in the example.
My point now is, that I have the feeling, that only the bonuses for sloped armor are in the formula, not it's flaws...
...I might be wrong, but something about the current way the formula behaves, makes me feel not so good about it and sloped armor is way to strong, at least for the thicker plates...

[ September 28, 2001: Message edited by: Panzer Leo ]</p>

[Kharan]

Sorry, Paul, I was ignorant that you were leading the v7 OOB development. Knowing it's you pretty much eliminates my concerns. And the "prior achievements" were not yours, although you did list them in a past thread (don't want to dig up that old dirt, probably should've just kept quiet).

I even lowered the vehicle FT values just for you <img src="biggrin.gif" border="0">

---

LOL, you just had to do that <img src="biggrin.gif" border="0"> . *grumble grumble* thanks.

[Charles2222]

I will add that I seen, but can't find it now, a diagram which I believe fprado's site had. This diagram was from a Tiger manual to all appearances. Some of the print of the diagram was very tiny, particularly the ranges it was stating, but the target showed some of the fire coming in what appeared to be between 30 and 45 degrees from that longer range (probably 3000yds).

While many of the fire demonstrated was at near flat trajectory, the longer ranges were considerably arched. It seems a lot of people would not believe what that Tiger diagram showed about the arch of their further shots. I'm of the thinking that this is because a lot of nations may had refused to train their tankers to fire at longer ranges, due to not having powerful velocity in the first place, and also because it was easier to train people to fire on flat trajectory only, plus of course the German superiority in optics had to aid in the idea of firing from great distances.

What this means, is that if the shell were to hit in a 30 degree trajectory, that suddenly the tank with the 30 degree slope is in the worst slope possibility. Some monthes ago I mentioned that in such an instance a zero degree slope on a tank would be better, that position was countered by the statement that if the 30 degree shell had not penetrated, that it would bounce for the most part down instead of up (as with flat trajectory shots on sloped armor), but the more I think about it it shouldn't matter a whole lot (I could be wrong of course).

Let's say the upper hull is hit froma 30 degree shell on a zero degree hull. On such a hull, it will either bounce into the ground, the shell would've been largely damaged on original impact, or it would hit a lower portion of the hull with less impact, where hopefully the target had just as much hull armor.

OTOH, if a flat trajectory shot were to hit a hull slope of 30, we are told it will bounce up. While there is less things to hit up, than down, there is still the chance that the turret will receive the bounce or the gun itself.

What I imagine, with a shell hitting at 30 degrees, is that there are some instances where a tanker in a Tiger, for example, was actually better off firing on well sloped tanks, such as the T34, from a further distance, where the slope was all but nullified, assuming the loss in some accuracy was worth the gain in penetration through a flat hit.

[Paul Vebber]

Thanks Kharan! I had a feeling that was the case.


Pz Leo

I will check Lorrins book (which has proceedures for calculating the angle of trajectory) but even at a 2000 yards I don't think its a great as 7 degrees. Certainly not 30 - which I'm sure was done for emphasis in the diagram.

OK I found some of Lorrin's equations in an old email:

velocity at range expressed as muzzle velocity x 2.71828^(0.7 x "k" x range in meters)
=606m/s (using -.000174=k) at 2000m

Time of Flight to 2000m(is
flight time = range (meters)/(0.5 x (muzzle velocity + velocity at range))

2000/(.5x(773+606)) = 2.9 sec

tangent gun elevation = 0.5 x 9.81 x (flight time to aimed range)squared/aimed range

=.5x9.81x2.9^2/2000 = .0206 = 1.18 degrees

There are some tweaks to thiss, but unless I made a major error it is certainly not much more than 1.25 degrees barrel elevation to get to 2000m and the angle at the other end will be slighty more, but certainly not 7 degrees.

I will double check this tonight.

Note that the random +/- 10 degrees is not a uniform random variable, but "triangular" distribution from adding several "die rolls" together so most results will be clustered toward the mid region.

Note that the table you post is for a single T/D ratio, the multipliers vary based on T/D, (though what you post seems to be close to the "rule of thumb" 1/cos(angle)^1.4 for unknown T/D.

Using this rule the 88L71 with 230mm @0 is estimated to pen 190. Is teh 170 number an "experimental result" if so it is probably due in part to the skewed distribution I mentioned in assigning an actual penetration number for a given shot (ranging in effect from +10% to -15% again in a distribution that has small tails - this represents the shot - to - shot variability in muzzle velocity, physical characteristics of the projectile, etc.

Note that a "vulnerable location hit simply reduces the effective armor after all calculations by 50%.

As to ammo type, you won't be able to select it specifically, but will be able to assign "rules" to use certain types in certain situations (based on target type, range etc) You are Regimental Commander (or Battalion)in Combat Leader, not the tank commander <img src="wink.gif" border="0">

I will shift this thread up to TO&E and Lorrin may chime in.