Canister in CMBB: Realistic or Hollywood?

[John D Salt]

Originally posted by Amedeo:

John, I don't follow your calculations here.

Right, following Amadeo's kind provision of some numbers to work with, and his polite drawing of my attention to the fact that my calculations were (as is perhaps appropriate to a discussion of canister) a lot of balls.

Here is the corrected version of my previous post.

Everything stays the same except the values for the numbers.

324 balls spread over a cone of length 200m and base diameter 40m.

Assume that the balls are independently and uniformly spread at random over the circle made by the cone at any range.

The radius and therefore area of the circle are:

Range, x___Radius, r__Area, a

50 m_________5 m_______78.5 m^2

100m________10 m______314.2 m^2

150m________15 m______706.9 m^2

200m________20 m_____1256.6 m^2

Man-sized targets in the cone of effect are assumed to be half a metre wide and 1.5 m tall, giving a target area, t, of 0.75 m^2.

The expected number of hits per man, e, can be obtained by dividing the area by 0.75 and then dividing the number of balls by the number obtained:

e = 324/(a/t)

The probability of an individual man being hit can be obtained by, first, calculating the hit probability of a single ball, p, which is 1 divided by the area divided by the target size. Obviously, this is the same as the expected number of hits for a single ball.

p = 1/(a/t)

Next, given our assumption that the balls are independently distributed, the probability, h, of an individual man being hit is given by one minus the probability of an individual ball missing, raised to the power of the number of balls.

h = 1 - ((1-p) ^ 324)

Results for each of the ranges considered are as follows:

Range, x___Expected hits per man, e____P(hit) on one man, h

50m________________3.10______________________0.96

100m_______________0.77______________________0.54

150m_______________0.34______________________0.29

200m_______________0.19______________________0.18

The difference between the hit probability and the number of expected hits per man is quite noticeable. This "over-hitting" problem would presumably become worse if the pattern was not the uniformly-distributed one assumed here.

These calculations seem to me to represent an upper bound to the casualty-causing effectiveness of canister, taking no account of cover in reducing the target are of exposed men. Similar calculations assuming smaller target areas would be easy to do along the lines given here.

As has already been mentioned, the suppressive effect of canister rounds would presumably be very high. Using the assumption that a ball coming within (say) 3 metres of a man will suppress him, it would be easy to perform further calculations on the probability of balls entering a man's "suppression space".

If sections keep proper tactical spacing with intervals of 5m between men, the killing effect of canister at close ranges will be quite low because of the small number of men in the target area. To illustrate this, consider the maximum number of men at 5m spacings that could be fitted into the cone at each range, and multiply by the hit chance against an individial man to find the expected number of casualties:

[Ooops, munged it again.

Tables snipped -- see Hat Trick's posting further down for corrected versions.]

All the best,

John.

[ May 30, 2003, 06:00 PM: Message edited by: John D Salt ]

[John D Salt]

Originally posted by Andreas:

[snips]

Since you are ignoring cover, I have trouble bringing the suggested height of 1.5m in relation with the well-known tendency of German soldiers to be right strapping lads of Aryan superhumanity with blonde hair and chiseled chins, and an average height of say 1.75m including boots and helmets.

You are not assuming that the targets belong to that famous unit of vertically challenged acolytes of Adolf, the 289th SS-Division 'Rumpelstilzchen'?

I assume that the upper 10cm of the target are protected by a canister-bullet-proof Stalhelm, that the bottom 10cm are protected by the favourable ballistic shape of steel toe-caps, and that the remaining 5cm is accounted for by the tendency of even the most Aryan of supermen to walk with a slight crouch when under canister fire.

Targets from smaller-statured units such as the one you mention, or those habitually in a crouching posture, such as Her Majesty's Royal Bengal Crouchers (as mentioned in "Beyond the Fringe"), or those that are actually members of another species such as SS Hamstertruppen, will present correspondingly smaller target areas.

To be serious for a moment, I believe there is a fairly strong tradition of taking the area of a human target as 0.5 m^2.

This may, however, be making allowance for terrain effects, and assuming that irregularities in the terrain mean that standing men are usually "shins down" (or, as a Londoner like yourself would put it, "knees-up", Mother Brown).

All the best,

John.

[Sergei]

Originally posted by Andreas:

You are not assuming that the targets belong to that famous unit of vertically challenged acolytes of Adolf, the 289th SS-Division 'Rumpelstilzchen'?

You should be well aware that during most of the year in Russia, everyone's knee-deep in mud-like substance, which Mr. Salt took properly into account.

[Marlow]

Interesting calculations for the 324 balls in a 57mm round. How about for the aprox 900 balls in the more relevant 76mm round?

[Andreas]

Marlow, probably not 100% correct, but here you go. Additionally one could expect the cone to be differently shaped due to slightly lower MV and shorter but wider barrel on the T34?

I would expect that part of overstating the effectiveness comes from CM assuming the squad is in a 20x20m area, not nicely spaced out with 5m intervals (although if you assume 2 guys per foxhole, a reduced or late-war German squad with 8 guys will just cover 20m frontage if evenly spaced at 5m. Since I never participated in proper section training (I was in the air force), I don't know what the story is for advance and assault situations.

Range, x___Expected hits per man, e____P(hit) on one man, h

50m________________8.2______________________0.999

100m_______________2.14______________________0.88

150m_______________1.69______________________0.61

200m_______________0.95______________________0.41

5m intervals we get:

Range, x____Number of targets_____Expected casualties

50m_______________2____________________1.92

100m______________3____________________2.64

150m______________4____________________1.83

200m______________5____________________2.05

2.5m intervals we get:

Range, x____Number of targets_____Expected casualties

50m_______________3____________________2.97

100m______________5____________________4.4

150m______________7____________________4.3

200m______________9____________________3.6

[ May 30, 2003, 10:15 AM: Message edited by: Andreas ]

[MikeyD]

As to the likelyhood of high canister loads in tanks, it's best not to rely too much on logic.

I read a report about a New Zealand 75mm gun Sherman going up against a Tiger. The report said he emptied his whole ammo ready rack into the beast, thirteen rounds -- SIX of them smoke! If a Sherman can have half of his ready rack filled with smoke shells there's no telling what wacky round mix a Russian tanker facing mostly infantry may have had.

[Amedeo]

The fact that the above calculations suggests that almost everything gets caught in the canister "cone" has a high probability of being hit is also illustrated by the example I provided a few posts above, of case-shot fire used to remove barbed wire obstacles!

Thus we may roughly assume that at short ranges a 57mm canister round will took 2-3 men (on the average) from attacking (i.e. moving in the open) infantry, but did onyone bother to test the effectiveness of 45mm and 57mm rounds in CMBB? The whopping ten-men-in-a-row casualties are typically inflicted by the much more common (in game terms) 76mm rounds, maybe the smaller rounds give results in accordance with the recent calculations.

Comments?

Amedeo

[JasonC]

The calculated coverage is way too high. The balls do not distribute evenly across the area of the cone. They are weighted towards its center, bell curve fashion, concentrated right along the path of aim. There is more area away from the sight line and fewer balls per unit area as well, making hits nearly certain right along the path of aim, but much less common that the uniform estimate near the edges.

The number of shells per tank seems to be following a randomized percentage per shell, giving high outliers as a result of the binominal distribution that results. Binominals have a long tail in only one direction. Right now it looks like the chance of a C round is as high as 1 in 8. That is why tanks with 100 ammo rountinely have up to 20 of them.

The 4-8 anecdotal account of the amount carried should not be read as an average of 6, turned into a percentage, and binominal-ed back up to 10-20. Instead, pick a percentage that will only rarely binominal out to 8. 1 out of 25, or 4%, would do. It would be quite rare for a tank to have none canister with that chance per shell (1/60). But it would likewise by rare to see more than 8 out of 100 as canister.

The size of exposed area for a squad of infantry in cover is very small. Just prone, it could be as low as 2m square for all 10 men combined (roughly 2 square feet per man). In a heavy building it could easily be 1/4 that (only some looking out windows or doors at any given instant). Infantry in combat is never all standing upright facing the enemy, continually.

As for the fellow talking about charging batteries, no he is not even talking about the effectiveness of canister against men charging the battery. Because by his own account he never had to use it. He is speculating, not reporting. He is counting balls and barrels and comparing them to aimed bullets.

There is plenty of actual historical evidence about the effect of canister vs. even exposed and massed infantry targets. And it does not support the "every round kills 7 men" stuff reported from CMBB tests - even without cover. In the Napoleonic era, canister was fired in huge quantities at ranges up to 500 yards against massed targets. Canister made up about a quarter of the ammo supply.

Yet Napoleonic era artillery averaged much less than 1 man hit per round fired, as can easily be seen from the fact that ammo expended by one side in major battles was much higher than total men hit on both sides, from all sources combined, including the largest cause, musketry.

E.g. at Borodino the Russians lost 44,000 men to all causes, while the French artillery fired 90,000 rounds. At Leipzig, the Allies lost 54,000 men to all causes, while the French artillery fired 200,000 rounds. The portion of the losses due to artillery was probably about 1/4 and certainly under 1/2. It therefore took on the order of 8 shots to get 1 man.

Similar craziness can result if you imagine ordinary bullets used with imaginery, calculated efficiency, instead of looking at real world totals for expenditure and effect. How lousy do you have to assume marksmenship is to W.A.G. that 10% of bullets fired should hit, when riflemen on firing ranges rountinely hit 7-9 times out of 10 at combat ranges, and MGs are expected to do better than a rifle with a 10-20 round burst?

In reality, the Germans fired 400 million rounds of 7.92mm in the Polish campaign alone. Small arms might have inflicted as high as 33-40% of Polish losses (which were 100-125K), and that still comes to 10K rounds fired per man hit. The W.A.G. would be wrong by 3 orders of magnitude.

Reduce the C rounds to 4% availability, and reduce their firepower to around 250 at 40m. They will fire like well armed full squads. Cover will reduce an FP that low to managable levels, when it is a matter of stone buildings. Infantry in the open shot at by full, well armed squads at 40m range does not generally like it very much, so the effect in the open ground case will be good enough for that tactical effect.

The present level of availability and effectiveness is not historical and falsifies tactics. You can't find Russian AARs of block clearing with canister fire, because it did not happen. In CMBB today it does.

[Hat Trick]

Originally posted by John D Salt:

324 balls spread over a cone of length 200m and base diameter 40m.

Assume that the balls are independently and uniformly spread at random over the circle made by the cone at any range.

The radius and therefore area of the circle are:

Range, x___Radius, r__Area, a

50 m_________5 m_______78.5 m^2

100m________10 m______314.2 m^2

150m________15 m______706.9 m^2

200m________20 m_____1256.6 m^2

Man-sized targets in the cone of effect are assumed to be half a metre wide and 1.5 m tall, giving a target area, t, of 0.75 m^2.

The expected number of hits per man, e, can be obtained by dividing the area by 0.75 and then dividing the number of balls by the number obtained:

e = 324/(a/t)

The probability of an individual man being hit can be obtained by, first, calculating the hit probability of a single ball, p, which is 1 divided by the area divided by the target size. Obviously, this is the same as the expected number of hits for a single ball.

p = 1/(a/t)

Next, given our assumption that the balls are independently distributed, the probability, h, of an individual man being hit is given by one minus the probability of an individual ball missing, raised to the power of the number of balls.

h = 1 - ((1-p) ^ 324)

Results for each of the ranges considered are as follows:

Range, x___Expected hits per man, e____P(hit) on one man, h

50m________________3.10______________________0.96

100m_______________0.77______________________0.54

150m_______________0.34______________________0.29

200m_______________0.19______________________0.18

The difference between the hit probability and the number of expected hits per man is quite noticeable. This "over-hitting" problem would presumably become worse if the pattern was not the uniformly-distributed one assumed here.

These calculations seem to me to represent an upper bound to the casualty-causing effectiveness of canister, taking no account of cover in reducing the target are of exposed men. Similar calculations assuming smaller target areas would be easy to do along the lines given here.

As has already been mentioned, the suppressive effect of canister rounds would presumably be very high. Using the assumption that a ball coming within (say) 3 metres of a man will suppress him, it would be easy to perform further calculations on the probability of balls entering a man's "suppression space".

If sections keep proper tactical spacing with intervals of 5m between men, the killing effect of canister at close ranges will be quite low because of the small number of men in the target area. To illustrate this, consider the maximum number of men at 5m spacings that could be fitted into the cone at each range, and multiply by the hit chance against an individial man to find the expected number of casualties:

Range, x____Number of targets_____Expected casualties

50m_______________2____________________1.92

100m______________3____________________1.62

150m______________4____________________1.16

200m______________5____________________0.90

If the target section is a bunch of slackers who keep only 2.5m intervals, then we get:

Range, x____Number of targets_____Expected casualties

50m_______________3____________________2.88

100m______________5____________________2.70

150m______________7____________________2.03

200m______________9____________________1.62

These calculations ignore enfilading effects. Even so, the levels of loss seen in CM:BB seem to me to imply very tightly-packed section formations.

All the best,

John. [/QB]

I'm not sure that the 'number of targets' figure is correct. At 50m distance, the radius is 5m, but he diameter is 10m. If the men are spaced five meters apart, three men can fit across the diameter: one on the left edge, one on the right edge, and one in the middle, 5 meters from the other two. At 200m distance, the diameter is 40m, and nine men, not five, fit within the circle. Thus, a revised table would read:

range____diameter____#of targets____casualties

50m_____10m________3_____________2.88

100m____20m________5_____________2.70

150m____30m________7_____________2.03

200m____40m________9_____________1.62

These are the same figures as previously given for 2.5 meter spacing. The new table for 2.5 meter spacing would be

range____diameter____#of targets____casualties

50m_____10m________5_____________4.80

100m____20m________9_____________4.86

150m____30m________13____________3.77

200m____40m________17____________3.06

There are (at least!) two factors that modify these figures, in addition to terrain/cover. First, this analysis assumes that the 'cone of fire' is perfectly centered, e.g., the maximum number of men are fit into the widest part of the cone at 5m or 2.5m intervals. If not so centered, the number of targets would be one less in each case -- from three to two at 50m distance with 5 meter spacing (reducing casualties from 2.88 to 1.92) and from 17 to 16 at 200m with 2.5 spacing (reducing casualties from 3.06 to 2.88).

More importantly, however, the above analysis assumes that the squad is spread out at 5m or 2.5m intervals horizontally across the front of the 'cone of fire.' This minimizes the number of units hit. A more accurate picture might be gained by assuming that the men are spread out in depth as well -- e.g., 5 meters across and 5 meters back. Not being an expert on squad formations and tactics I'm just guessing here, but if we look at the above figures we see that if the men in a squad are advancing in anything like 2 or more rows (i.e., with only one layer of depth), they would all be in the diameter of fire in every case but 5 meter spacing at 50m distance. We can then use the straight hit probabilities for each man -- 96% of the squad would be wiped out at 50m, 54% at 100m, 29% at 150m and 18% at 200m.

Of course, the above figures do not take into account terrain or cover, though it is interesting to note that in a building one would expect the benefits of cover to be at least partially offset by the fact the men are even less likely to be spread out horizontally at 5 meter intervals.

Finally, the above figures are for the 57mm gun. The 76mm has many more balls, but frankly my head hurts too much to calculate the effects right now.

P.S. If anyone can tell me how to format a table in here I would appreciate it.

[White Phosphorus]

Originally posted by JasonC:

Yet Napoleonic era artillery averaged much less than 1 man hit per round fired, as can easily be seen from the fact that ammo expended by one side in major battles was much higher than total men hit on both sides, from all sources combined, including the largest cause, musketry.

E.g. at Borodino the Russians lost 44,000 men to all causes, while the French artillery fired 90,000 rounds. At Leipzig, the Allies lost 54,000 men to all causes, while the French artillery fired 200,000 rounds. The portion of the losses due to artillery was probably about 1/4 and certainly under 1/2. It therefore took on the order of 8 shots to get 1 man.

Weren't they using grapeshot instead of canister? How did canister perform in the American Civil War?

[ May 30, 2003, 05:03 PM: Message edited by: White Phosphorus ]

[tar]

I was going to bring up the infantry in depth and breadth argument, but I was beat to the punch.

Instead, I will try an improved translation of rexford's German post:

I have a German report on 7.62cm ammo, and there is a big section entitled "Anleitung zum Schiessen mit russischen Schrapnells".

Instructions for shooting with Russian shrapnels.

"Der Streukegel hat einen Winkel von durchschnittlich 15°, das heisst, die Kugeln treffen in 80 m Enterfernung eine Flache von 20 m breite: die Lange ist je nach dem Einfallwinkel verschieden"

The cone of dispersion has an average angle of 15°, which means that the balls hitting at a distance of 80m cover an area 20m wide. The length varies depending on the angle of impact.

TThe drawing that goes with the above statement shows a 20m wide oblong shape on the ground at 80m range from the gun, with balls landing all over the shape.

"Die Zunderstellung in der Schusstafel ist so angegeben, dass der mittlere Sprengpunkt fur Ziele in der mundungswaagerechten 80 m vor dem ziel liegt. Er wind[sic] dem beobachter vom Geschutz knapp uber dem ziel (2 bis 5) erscheinen. Jnfolge der streuung wird eine kleine anzahl von geschossen unvermeidlich am boden aufschlagen. (Deshalb auch AZ). Diese sind nahezu wirkungslos, mussen aber in kauf genommen werden, da andererseits auch zu hohe sprengpunkte keine Wirkung haben".

The fuse setting given in the firing table sets the mean bursting point 80m horizontally along the line of fire before the target. It will appear to the observer of the gun to be slightly above the target (2 to 5)[m?] Because of the dispersion, a small number of the balls will inevitablely strike the ground. (Therefore also AZ [huh?]). These balls are ineffective, but have to be tolerated, since, on the other hand, a detonation that is too high above the ground will not have any effect.

[John D Salt]

Originally posted by Hat Trick:

[snips]

I'm not sure that the 'number of targets' figure is correct. At 50m distance, the radius is 5m, but he diameter is 10m.

Quite right, I have made a horlicks of it yet again.

Still, it's encouraging that people are reading this stuff...

Originally posted by Hat Trick:

There are (at least!) two factors that modify these figures, in addition to terrain/cover. First, this analysis assumes that the 'cone of fire' is perfectly centered, e.g., the maximum number of men are fit into the widest part of the cone at 5m or 2.5m intervals. If not so centered, the number of targets would be one less in each case -- from three to two at 50m distance with 5 meter spacing (reducing casualties from 2.88 to 1.92) and from 17 to 16 at 200m with 2.5 spacing (reducing casualties from 3.06 to 2.88).

This is true, but recall that the precisely-defined cone full of uniformly-distributed balls is modelling a fan with much more loosely-defined edges full of more clumpily-distributed balls (I agree with JasonC that a normal distribution would be what to expect, but I have no evidence of the patterns canister shot actually describes. If the shotgun analogy holds good, that would ISTM tend to indicate a normal distribution). Counting the men on the edges in or out is therefore a completely arbitrary matter of convention, and the numbers derived should be treated as comparative rather than absolute, as is usual in exercises like this.

Originally posted by Hat Trick:

More importantly, however, the above analysis assumes that the squad is spread out at 5m or 2.5m intervals horizontally across the front of the 'cone of fire.' This minimizes the number of units hit. A more accurate picture might be gained by assuming that the men are spread out in depth as well -- e.g., 5 meters across and 5 meters back. Not being an expert on squad formations and tactics I'm just guessing here, but if we look at the above figures we see that if the men in a squad are advancing in anything like 2 or more rows (i.e., with only one layer of depth), they would all be in the diameter of fire in every case but 5 meter spacing at 50m distance.

Good point, well made, and I claim a further argument in favour of going down to single-man granularity in the next-generation CM engine.

I've never heard of a "staggered line" formation such as you describe being used by infantry sections. The usual section formations I would expect to meet would be line, file (in fact two files, staggered), single file and arrowhead. I have read that the Germans only used line and single file, but the Russians certainly used the complete set. In the untidiness of the battlefield, I expect "column of blob" would be quite popular, and soldiers are famous for not keeping the spacings as wide as they are told to.

A final point, that has not been mentioned yet, is the effect that trajectory shape might have on the expected number of hits. It seems to me that there should be no effect, but I haven't yet convinced myself of this, and it is a curious fact of effectiveness modelling that refinements to a model almost always tends to reduce the estimate of effectiveness.

Originally posted by Hat Trick:

P.S. If anyone can tell me how to format a table in here I would appreciate it.

I've never figured out how to do it -- if anyone knows a better way than separating columns with underscores, I'd be interested to know.

All the best,

John.

[Andreas]

AZ = Aufschlagzünder = impact fuse

[Ales Dvorak]

Realistic or Hollywood?

Sometimes are realistic. What's holly in wood?

[John Kettler]

The apples appear to have become confused with the

oranges in this discussion. This is because shrapnel and canister are NOT the same thing.

As noted in my earlier post, canister operates like a giant shotgun shell, hence works by muzzle action, requiring no fuze. The canister shot begin to fan out as soon as they pass the muzzle. By contrast, shrapnel may be thought of as a gun launched shotgun, which then fires at a predetermined range, which can be thousands of yards. This is precisely why it requires a fuze. Ideally, shrapnel is fired in such a way that it arrives pointing downwards onto the target, reducing cover effects and engaging the target in depth along the axis of fire. The art of shrapnel fire lies in rapidly adjusting the fuze settings so as to optimize height, deflection and distance of burst before the target can disperse or take effective cover. The theory and practice of shrapnel gunnery is thoroughly discussed in NOTES ON TRAINING FIELD ARTILLERY DETAILS by Captains (Field Artillery) Robert M. Danford and Onorio Moretti, Yale University Press, 1917. See particularly page 68 et. seq. under "Principles of Fire."

Now that we're back on the same ordnance page, let us continue this fascinating and delightfully esoteric military discussion.

Regards,

John Kettler

[Redwolf]

I also wonder whether anybody really doubts that a squad concentrated in a 15x15 m open area gets nearly wiped out by a canister shot.

Am I the only one who thinks its is a much more interesting question whether cover should count the same against MG fire and canister shot? A head-down squad wiped out from a foxhole with a canister shot sounds much more debatable to me.

[ May 30, 2003, 10:45 PM: Message edited by: redwolf ]

[Lady Roxanne]

JasonC says, "The present level of availability and effectiveness is not historical and falsifies tactics. You can't find Russian AARs of block clearing with canister fire, because it did not happen. In CMBB today it does."

This is the real argument for toning down CMBB "cannister" round effectiveness and/or availability IMO.

Also, as it is now, a single squad in good cover has exactly one "shot" with a grenade bundle, faust, or rifle grenade before being obliterated by a single cannister round from a T-34. Once the AT weapon is used the entire squad has less than 30 seconds to live if the tank survives. Is this how it was?

[Michael Dorosh]

How does one truly measure "suppressive effect" of canister - or any weapon, for that matter?

Doesn't that depend entirely on the target and not the weapon? Specifically, the target's ability to be spooked/not spooked by muzzle flashes, smoke, rounds impacting nearby, and shock effects of near hits?

How on earth would anyone quantify that, except subjectively, and in relation to other nasty sounding/looking noises/sights?

[rexford]

The following translation by Tar seems to suggest that Russian Schr. Patr. ammo was schrapnel, since the translation refers to a fuze.

The drawings of Russian 7.62cm Schr. Patr. show a projectile with a nose piece (Dopp Z 436® russ T-6), a compartment filled with balls, and then an area under the middle compartment that appears to contain a powder substance.

I can scan the drawings and explanation of effective area and share with whoever would like to see it.

The one point that sticks out in my mind is that cannister is normally fired from guns without muzzle brakes, such as U.S. 37mm and 75mm guns and the German 75mm L24, while the 76.2mm field gun vet on the Russian Battlefield talks about case shot but his gun would seem to have had a muzzle brake.

If the 76.2mm field gun fired schrapnel, it would seem more likely to be a fuzed round than a cannister that started to spread as soon as it left the barrel (but before it left the muzzle brake).

Based on Tar's improved interpretation, which is appreciated very much, Russian schr. patr. appears to be schrapnel that explodes at a distance from the gun and then spreads out.

The Russian 45mm and 57mm anti-tank guns did not have a muzzle brake.

Originally posted by tar:

I was going to bring up the infantry in depth and breadth argument, but I was beat to the punch.

Instead, I will try an improved translation of rexford's German post:

</font><blockquote>quote:</font><hr />

I have a German report on 7.62cm ammo, and there is a big section entitled "Anleitung zum Schiessen mit russischen Schrapnells".

Instructions for shooting with Russian shrapnels.

"Der Streukegel hat einen Winkel von durchschnittlich 15°, das heisst, die Kugeln treffen in 80 m Enterfernung eine Flache von 20 m breite: die Lange ist je nach dem Einfallwinkel verschieden"

The cone of dispersion has an average angle of 15°, which means that the balls hitting at a distance of 80m cover an area 20m wide. The length varies depending on the angle of impact.

TThe drawing that goes with the above statement shows a 20m wide oblong shape on the ground at 80m range from the gun, with balls landing all over the shape.

"Die Zunderstellung in der Schusstafel ist so angegeben, dass der mittlere Sprengpunkt fur Ziele in der mundungswaagerechten 80 m vor dem ziel liegt. Er wind[sic] dem beobachter vom Geschutz knapp uber dem ziel (2 bis 5) erscheinen. Jnfolge der streuung wird eine kleine anzahl von geschossen unvermeidlich am boden aufschlagen. (Deshalb auch AZ). Diese sind nahezu wirkungslos, mussen aber in kauf genommen werden, da andererseits auch zu hohe sprengpunkte keine Wirkung haben".

The fuse setting given in the firing table sets the mean bursting point 80m horizontally along the line of fire before the target. It will appear to the observer of the gun to be slightly above the target (2 to 5)[m?] Because of the dispersion, a small number of the balls will inevitablely strike the ground. (Therefore also AZ [huh?]). These balls are ineffective, but have to be tolerated, since, on the other hand, a detonation that is too high above the ground will not have any effect. </font>

[rexford]

Originally posted by Michael Dorosh:

How does one truly measure "suppressive effect" of canister - or any weapon, for that matter?

Doesn't that depend entirely on the target and not the weapon? Specifically, the target's ability to be spooked/not spooked by muzzle flashes, smoke, rounds impacting nearby, and shock effects of near hits?

How on earth would anyone quantify that, except subjectively, and in relation to other nasty sounding/looking noises/sights?

The John Salt site has some studies on suppressive effect where projectiles above a certain velocity, which are within a certain distance from a man, have a good chance of suppressing the fellow.

Plenty of assumptions would go into who gets suppressed. Even John Wayne was occasionally pinned down by inaccurate, but close miss, fire.

And if the Duke kept his head down, lesser mortals would.

[ May 31, 2003, 03:48 AM: Message edited by: rexford ]

[rexford]

I scanned the German pages on Russian SchrPatr ammo and will e-mail them to whoever wants copies.

There are five scanned pages which cover ten pages from the manual.

E-mail me at rexford179@cs.com

One of the drawings shows the SchrPatr being fired against troops on a hillside, where the round follows a smooth trajectory to just before the target, where the SchrPatr ammo explodes ("sprengpunkt" or bursting point).

If the round must be set for a given fuze detonation range, this would introduce range estimation error into the casualty equation.

[ May 31, 2003, 04:05 AM: Message edited by: rexford ]

[rexford]

One of the German pages gives the following data for three different Russian Schr Patr projectiles:

Muzzle Velocity, Mundungswucht, Gasdruck

626 m/s, 128.6 tm, 2320 at

476 m/s, 074.4 tm, 1880 at

==========================

618 m/s, 129.6 tm, 2320 at

========================== 624 m/s, 129.0 tm, 2320 at

474 m/s, 074.4 tm, 1880 at

Higher velocity is listed as normale Ladung, lower figure as verminderte Ladung or decreased charge.

[John D Salt]

Originally posted by John Kettler:

The apples appear to have become confused with the

oranges in this discussion. This is because shrapnel and canister are NOT the same thing.

[snips]

Indeed -- although Napoleonic people who still talk about "Major Shrapnel's spherical case" might claim that they are both kinds of case-shot.

I suspect that the confusion is somewhat assisted by the similarity between the Soviet designation letters Ш (Sh) for shrapnel (Шрапнель, Shrapnel') and Щ (Shch) for canister (Картечь, Kartech') -- quite why they use the latter letter I don't know, as it doesn't appear in the word Картечь.

Things are not helped by Valeriy Potapov's curious decision to transliterate Щ as "She" on his excellent Russian Battlefield site. Then again, as far as I can see, his artillery pages ( http://www.battlefield.ru/guns/project_1.html ) do not seem to list any canister rounds.

I anyone has a reference to a source giving the characteristics of the 76mm canister round, I'd be interested to know of it, Likewise I'd be interested in further information on the German canister for the 7.5cm L/24, a round I have not seen referred to outside the world of "Squad Leader" other than a fleeting reference in von Senger und Etterlin. Oh, and, while we're collecting them, can anyone think of any other canister rounds that are believed to have existed during WW2 (apart from the US 37mm ATk and 57mm RCL)?

All the best,

John.

[John D Salt]

Originally posted by Michael Dorosh:

How does one truly measure "suppressive effect" of canister - or any weapon, for that matter?

I'm happy to leave "truly measuring" things to people who imagine that such a thing is possible. I'm interested in modelling, not measurement.

Originally posted by Michael Dorosh:

Doesn't that depend entirely on the target and not the weapon? Specifically, the target's ability to be spooked/not spooked by muzzle flashes, smoke, rounds impacting nearby, and shock effects of near hits?

How on earth would anyone quantify that, except subjectively, and in relation to other nasty sounding/looking noises/sights?

Methods that I know have been used to model suppression (as ditinct from quantifying it) include calculating whether a projectile comes within a stated distance of the target, linking suppression to the occurrence of casualties, applying fudge-factors after the fashion of Dupuy's CEV, and assessing the relative suppressive effect of different weapons through questionnaires. None of these seem particularly good. I have heard it suggested that noise levels be used too, but never heard of a model that actually uses them.

A chap called Zinn has tried to produce a "laundry-list" of factors affecting suppression, but I can't find my copy of the (1999) paper at the moment. Anyone who remembers the old WRG reaction tests will find nothing new in it, anyway.

IMHO the "official" modelling of suppression effects is lagging that of recreational wargaming by thirty years or more, largely thanks to the insistence of scrutineers that numbers used in official models have properly-certified provenance. Many combat models used to support official decision-making still do not model suppression at all.

It seems to me that you are entirely correct to emphasise the part of the target, rather than the weapon; but I wouldn't go so far as to say that the effect depends entirely on the target. Suppression is a relational action, and those always seem to cause people problems.

For my money a convincing model of suppression depends on establishing a comprehensive model of tactical situational awareness, using affective agents. The UK defence research budget currently seems to include no funding for such a project.

All the best,

John.

[John D Salt]

Originally posted by CrankyKris:

JasonC says, "The present level of availability and effectiveness is not historical and falsifies tactics. You can't find Russian AARs of block clearing with canister fire, because it did not happen. In CMBB today it does."

I agree. I have seen more pictures of the 37mm spade-mortar than I have of 76mm canister rounds.

Originally posted by CrankyKris:

[snips]

Also, as it is now, a single squad in good cover has exactly one "shot" with a grenade bundle, faust, or rifle grenade before being obliterated by a single cannister round from a T-34. Once the AT weapon is used the entire squad has less than 30 seconds to live if the tank survives. Is this how it was?

I dount it. For one thing, I would expect there to be only one or two blokes from the section actually anti-tank weapons in their hands and stalking the tank, with everyone else taking cover. Even if that were not so, at the close ranges hand-thrown anti-tank weapons are used, it should be clear that the spread of a canister round will be so narrow that it is very unlikely to hit more than one man unless they line themselves up neatly along the line of fire.

Finally, quite apart from the question of canister effectiveness, it seems to me that tanks in CM get off too lightly in close combat with infantry, as they do not seem to have any "blind" zone at extremely close range. The tank hunter's rule that the closer you are to the tank the safer you are is not reflected at all, and tank crewmen peering through their vision slits seem to have a preternatural ability to spot and shoot infantrymen dodging around within a few metres.

All the best,

John.