How 'Bout Those M3 Turrets!

[Kozure]

Originally posted by Treeburst155:

I think I'll spend some time today, throwing various AT rounds at a hull down Grant at various ranges, just for fun. I'd like to get a turret hit:upper hull hit ratio; but that might be difficult to do with a large number of tanks on a firing range.

Seperate the firing range into lanes with a single 20m wide high elevation seperator between each firing lane, e.g.

0012000000

G01200000F

0012000000

9999999999

0012000000

G01200000F

0012000000

9999999999

(not to scale)

Where the numbers are for the elevation and the G stands for Grant and the F for Flak.

It seperates the firing lanes, preventing Flaks from firing on the wrong tank; you can set a variety of ranges at the same time; and you can clearly see the effects of one Flak on one Grant for any given lane. Very easy to note hit effects.

Oh, one more note about doing this testing - turn Fog of War to "None" so that you can see all the actual results of hits.

[ January 13, 2004, 03:32 PM: Message edited by: Kozure ]

[Treeburst155]

Thanks for trying to be helpful here, Kozure. Let me say, though, I'm an old hand at running firing tests for CM.

What I wanted to do today, but didn't get to, was to test hundreds of times using a range separated into 20 firing lanes. When this is done, the detailed armor hit messages come very fast. This makes it difficult to actually record how a particular vehicle died. Was it hit in the upper hull or the turret front, etc?

The only way to see and record all the armor hit messages is to use only a small number of firing lanes. This makes running a large number of test instances very laborious.

The statisticians around here will tell you that a test needs to be run several hundred times to get within +/- 5% of the truth.

[Michael Emrys]

Originally posted by Treeburst155:

What I wanted to do today, but didn't get to, was to test hundreds of times using a range separated into 20 firing lanes. When this is done, the detailed armor hit messages come very fast. This makes it difficult to actually record how a particular vehicle died. Was it hit in the upper hull or the turret front, etc?

The only way to see and record all the armor hit messages is to use only a small number of firing lanes. This makes running a large number of test instances very laborious.

Well, you could stick with your 20 lanes, but only watch a couple at a time. This would mean rerunning the movie several times. Still laborious.

Michael

[Kozure]

Originally posted by Treeburst155:

Thanks for trying to be helpful here, Kozure. Let me say, though, I'm an old hand at running firing tests for CM.

What I wanted to do today, but didn't get to, was to test hundreds of times using a range separated into 20 firing lanes. When this is done, the detailed armor hit messages come very fast. This makes it difficult to actually record how a particular vehicle died. Was it hit in the upper hull or the turret front, etc?

The only way to see and record all the armor hit messages is to use only a small number of firing lanes. This makes running a large number of test instances very laborious.

The statisticians around here will tell you that a test needs to be run several hundred times to get within +/- 5% of the truth.

Hmmmm... well, all I did was have many lanes, and only run the vehicles up into hull down position when I had seen and recorded results for a few lanes at a time, usually about three or four tanks exposing themselves at a time. If I somehow missed one, I just rewound the film; seemed pretty easy to me.

You, being an old hand, have probably used that method already.

As for approaching the truth statistically, I guess all I was trying to demonstrate was that...

... a hull down M3 at about 700 meters is virtually unkillable from the front.

...is not the case - which I believe was the initial theme of the thread.

As you point out... just trying to be helpful. Sorry if I offended by suggesting you didn't know the ropes.

[CRSutton]

Originally posted by Kanonier Reichmann:

[QB]

Seems a tad unusual to me and if it's the case then I don't see why the U.S. ever bothered to upgrade to the Sherman when they had a far more resilient design from day one. :confused:

A seven man crew is a terrible waste of resources.

[Treeburst155]

Kozure,

I didn't take offense or mean to offend you. In fact, the reason I posted my previous post was to acknowledge your taking the time to help me get a good test going. I appreciated the effort.

The original assumption I made that the M3 would be virtually unkillable from the front when in a hull down position was proven to be wrong due to the upper hull exposure, which cannot be avoided.

The turret front however, is still a very formidable piece of armor, as can be seen by the thickness and slope statistics. 88s will usually only manage a partial penetration at 700 meters.

I've decided that this fact is merely interesting, and not worth testing because of the upper hull weakness. In other words, I don't see myself adopting different tactics due to the strong Grant turret. I only test things when I think the results will provide useful information for gameplay. The Grant will die frontally ALMOST as fast as anything else; but it will almost certainly be a hull hit that does it.

[Kanonier Reichmann]

The way I see it the Grant should, in theory, survive on the battlefield much better than any other tank provided one keeps it hull down to the enemy. Bearing in mind that only approximately 25% of all hits impact the upper hull when it's hull down then the exceptionally strong turret would take the brunt of the hits and probably deflect most if from anything 50mm or less in size. I'm wondering however if this was historically noted as being a major problem when the Germans tackled Grants? Did they regularly complain that their AFV's (and even occasionally 88's!) could barely penetrate the turrets of Grants when engaged at greater than point blank range?

To me it seems the Grant in 1942 and early '43 is the Allied equivalent of the Soviet KVI in 1941 provided they're used sensibly. Does this seem reasonable?

Regardfs

Jim R.

[Treeburst155]

I'm not convinced that 25% of all hits from AT assets to the front, on a HULL DOWN tank, are on the upper hull. Are we sure it's not 25% of all hits on a tank which is fully exposed?

If the 25% is true with hull down tanks, Grant survivability from the front while hull down should be quite a bit better than a Sherman. Perhaps I SHOULD run some tests.

[Treeburst155]

Preliminary testing indicates that, at 760 meters, a front turret penetration of a hull down Grant by an 88 will occur just as often as a partial penetration. This means that all upper hull hits and half the turret hits by an 88 will kill the Grant quickly at 760 meters. The Grant won't survive too much longer than a Sherman against an 88 at that range. Now a different gun at some different range may show the Grant significantly more resistant to death than a Sherman.

So, the in-game experience I had of seeing many turret front partial penetrations of 88 shells was just a little bad luck. Half those shells should have penetrated fully.

[ January 15, 2004, 05:47 PM: Message edited by: Treeburst155 ]

[gunnergoz]

Given that the 37mm turret was a casting and never up-armored as was the M4 turret, I'm sort of surprised that it would be given some sort of magical ability to thwart 88 rounds. I suspect it was thinner than the M4 turret, though I don't have the specs in front of me.

[Treeburst155]

I wouldn't characterize it as a "magical ability". The 88 just won't fully penetrate a Grant turret front all the time at 700+ meters. Partial penetrations are common too at that range. Looking at thickness and slope of the Grant turret front, this is understandable.

It was just a bit surprising to see 88s not fully penetrating a fairly early Allied tank at moderate range.