"Canister fire revelation" Topic

Now, with these 93 men you can begin to see in 25yd increments how many are in the direct firing lines of the cones of the 6 guns based on 7% coverage every 25yds.

25yds (7% coverage) = 6.5men potential casualties
50yds (14% coverage) = 13men potential casualties
75yds (21% coverage) = 19.5men potential casualties
100yds (28% coverage) = 26men potential casualties
125yds (35% coverage) = 32.5men potential casualties
150yds (42% coverage) = 39men potential casualties
175yds (49% coverage) = 45.5men potential casualties
200yds (56% coverage) = 52men potential casualties
225yds (63% coverage) = 58.5men potential casualties
250yds (70% coverage) = 65men potential casualties
275yds (77% coverage) = 71.5men potential casualties
300yds (84% coverage) = 78men potential casualties
325yds (91% coverage) = 84.5men potential casualties
350yds (98% coverage) = 91men potential casualties
375yds (100% coverage) = 93men potential casualties
400yds (100% coverage) = 93men potential casualties

So it can be seen clearly that it is the cones and their width when the canister makes contact with the enemy line which determines just how effective they will be. The longer out the range the more coverage and the higher the optimal coverage will be for potential hits. If you add in a small factor for stray balls or balls on erratic flight paths and even ricochet you can increase this number marginally by maybe 5-10%.

Please bare in mind, these casualties are purely based on mathermatics and not battlefield conditions. They are Optimal or ""Potential"" not actual. These numbers are what one might strive for or hope for, actual battlefield conditions will decrease these numbers, sometimes dramatically downwards. As I clearly stated, they are potential NOT actual.

Shane

"I just cannot see that canister fire was more effective at 400 yards than it was at say 150,"

The point is that, in terms of 'ball spread', there is little difference. Obviously the nearer to the target the more chance of hitting said target- and the greater the likelihood of inflicting more tissue damage- the balls being more subject to the vagueries of flight and velocity at the longer ranges.
Extreme effective range would be a good description of a six pounder firing canister at 300 metres. Adye concluded that the extreme outward perimeter of the cone was reached at around 75 yds, at which point, as I mentioned earlier, the outer section of balls would lose their kinetic energy very quickly and the trajectory would be more in a straight line- I wish I could draw on here, it'd be much easier.
So in fact, ideal range would indeed be a lot less- for the six pounder around 75 yds. At this distance the pattern of ball would be at its widest, the ball would be hitting with most force, and, due to the proximity of the target, there would be more chance of achieving a hit- or hits.
It's when attackers get closer than this distance that the effects of canister start to diminish. It's this lack of intuitiveness that is really the point here. Wargames rules that give a higher hit ratio at 30 or even 50 yds are actually getting it wrong.

I'm just reading about Grant's bouncing roundshot stick at the moment. It's 5ft long!
I should think after a long game and the accompanying 'lubrication' it could do more damage than the real thing.

Of course the above only stacks up if we are talking about one gun. If the model represents more than one actual gun then of course, due to the nature of the cone fire, there will be overlapping fire which would enable full coverage of a target at less than the figure of 75yds.

Having reread Shane's original post it would appear that our conclusions are actually somewhat different.
What we need to bear in mind is that although we are throwing around the term canister as if it were some sort of standard, there werre in fact many different types- even within the same army.
Certainly with more powerful cannon and a larger canister round, you would obviously expect this 'cone' to be somewhat larger.

Mullers numbers are in keeping with zhe period rule of thumb that the spread is 10% of the range.
I don´t know on which basis Jeffreys ssumptions are based, Jeffreys numbers are just about 2,5 % of the range.
I have seen sheets of target effects of a 24 pound howitzer and a gatling gun both shooting canister in trials around 1870, there is no Muller outer cone and a Jeffrey high density inner cone.
I have more sympathy for Muller´s numbers therefore.

But even if we take Jeffreys cone, it is not only 32 feet wide at 400 yds but its 32 feet high too, but the target is just 6 feet tall a the most, so even with Jeffreys high density cone (which indicates a 4 times better accuracy then Mullers numbers by the way) 80% of all rounds will miss the target at that distance by being overs or short misses.
And even the overs will have more or less no effect if the target is just a 3yds/meters deep line, by a rough calculation a canister ball might drop 15 cm while travelling 3m at 400 m distance, thats lousy from a modern weapons standpoint, but the flight is still fairly vertical.

Tests I have read indicated that the drop rate was actually 3-5ft in 125yds according to Jeffrey. I am very sceptical about this. This was again written by Jeffrey and open to speculation but even so there is a drop rate.

He indicated that the velocity of a British 9pndr had shown a hit rate on a target of 55% at 150yds while at 250yds the hit rate actually increased to 65%. This shows me that in the intervening 100yds between 150 to 250yds the higher balls were dropping once again and finding a mark.

I am beginning to think Jeffrey is correct in his summation that the arc or trajectory after 150yds on the higher balls causes them to lose velocity much quicker coupled with wind resistance and thus the downward arc increases dramatically.

So at 150yds you have 55% hits with a 9pndr and 65% hits at 250yds against a target regardless of whether it is a line or column. If the target is actually closer to 150yds then the higher balls simply land behind the target or into another unit following closely behind. If the target unit was a column the rear ranks might suffer some of those hits.

For my system I pulled forward the Point Blank range thus allowing a larger Effective range beginning because I do not accept Jeffrey's summation of things totally. I do believe that field conditions can make a significant difference and also believe the cone width was larger than Jeffrey suggests but maybe not as large as Muller suggests so I took a more medium view between the two.

So with my system a 6pndr Effective range begins at 50-60yds out to about 200yds while a 12pndr begins at more like 70-80yds out to 300yds.

With ranges of over 200yds with 6pndrs or 300yds with 12pndrs I also view this as Long Range and thus with much larger cones you might get full saturation and overlaps you still have a large number of the balls missing due to the large cone groupings. So at these longer ranges my factor tables for hits are severely reduced. For example factors are roughly as follows in my own tables :

PB Ranges – 9 (average coverage of between 7% to 35% small cones = large gaps)
EF Ranges – 12 (average coverage of between 36% to 70% good solid groupings)
LR Ranges – 6 (average coverage of between 71-100% but huge gaps)

So even though Long Ranges indicate complete coverage and saturation of an enemy target line you still have the situation of a huge grouping of balls with several feet possibly between each ball thus large frontages of the enemy line would be missed. I deem it unlucky for the poor targets at long range to get hit compared to grouping coverage's at Effective Ranges.

In conclusion I do believe Jeffrey has really got it correct in all he says as against the data he researched through Hughes and Muller's tests and is able to account for many of the anomalies they came up with. However, you cannot discount for stray balls and thus do have to factor them in and this I have done with the added 5-10% I added in my system.

Regards,
Shane

>>>>But even if we take Jeffrey's cone, it is not only 32 feet wide at 400 yds but its 32 feet high too,<<<<

Yes you are correct but taking a line directly through the centre of the cone to the middle of the target you have a cone up to 16ft high and below the cone centre point you have another 16ft below. If you discount the balls that land short or in the negative 13ft (allowing for 3ft for lower torso and legs) you still have the grouping for the balls which will land on target minus the 13ft above the heads.

So the square area of potential hits is around 803ft per gun cone but the area of space covered by an enemy Line within this area is a mere 192 square ft so the percentage of area covered by the enemy line is less than 24% (1/4 of the area), that of cavalry is 288 square feet or 36% (1/3 of the area) of the cone.

So if the round is firing 60 balls you would in reality get the following to occur:
Infantry 24% hits = 14.4 casualties per gun / 6 guns = 88.4 potential hits.
Cavalry 36% hits = 21.6 divided by 2 (size of horse and rider) = 10.8 casualties per gun / 6 guns = 64.8 potential hits

Multiply this by 1.5 if 90balls fired or double it if 120 balls fired.

My own conclusions really aren't far off what I posted earlier in regards to this and are what convinced me I was pretty much on the mark.

Regards,
Shane

1) Thanks for a fascinating discussion gentlemen.

2) Re. Mercer's battery at Waterloo – IIRC, a) his guns were behind a bank; b) he double-shotted with a roundshot over the canister; c) his opponents were cavalry and he halted them by building an impassable wall of downed horses (as much a result of the penetrative effects of the shot).

3) Didn't most artillery have more than one type of canister, usually with different weights/numbers of balls? Presumably there were rules or guidelines as to when each type was used (possibly heavier types would have been used against cavalry?). On that basis, might not the heavier types have had greater penetrative effect (essentially becoming small roundshot)?

The figures in the various analyses of test shootings are fascinating and undoubtedly give food for thought. However, I have one or two minor reservations.
1) It seems that all these tests were carried out with battery and target frontages parallel. If we accept that a canister ball lacked the power to penetrate more than one body, then individuals in the second and subsequent ranks will be safe from the 1,2 3 or even 8 balls that struck their front rank man. But what happens if there is any degree of enfilade? (British practice of stationing sections of guns between batallions lends itself to this sort of fire.) Given that canister shot spread (and I don't care if it is a cone or a fan or a pyramid) what happens to the individuals to the left and right rear of the font rank man? Are they not at risk of becoming casualties? If so, increase the effect of your canister fire! ;_)

2) Targets are continuous screens. A body of troops in whatever formation will have empty spaces (even given 'the touch of cloth')through which canister shot (and musket balls)will pass without hitting anyone – so reduce the effect of your canister fire.

Seriously though, I agree that for the most part rules tend to give too much emphasis to the effect of canister. My own rules may be guilty of this, but I have tried not to make canister into a 'death ray'. The emphasis is playability which yields a result that AS FAR AS WE CAN reflects what happened on the Napoleonic battlefield.

However, the bulk of ammunition available to field guns was roundshot (see Hughes "Firepower" – which also contains facsimiles of the 1813 Bengal Horse Artillery records). So, unless we want to go the book-keeping route, tracking down the number of rounds of every type of ammunition carried in axle boxes, coffrets, limbers and caissons we have to make judgement calls, that fit our idea of what happens when the 4005th Fixed Bayonets Heavy Foot are enfiladed by canister at 200 paces by 18 batteries of 64 pdr Horse Artillery Guns.

Alright, that example is deliberately ridiculous – but do you see what I'm driving at?

My own approach has been to classify artillery by the weight of its shot, irrespective of whether or not it is 'Horse' or 'Field'.

Viz:

Light Artillery: 3, 4 and 6 pdrs.

Field Artillery: 8 and 9 pdrs

Heavy Field Artillery: 12 pdrs

Unless I am playing in an historical re-fight where there is evidence to the contrary all artillery units are considered to be well-trained – that is there is no subtraction or addition to dice rolls for recruit or veteran units.

A second factor is the number of figures I have per gun detachment:

Light Artillery: 3 figures.
Light Field Artillery: 4 figures.
Heavy Field Artillery: 5 figures.

To calculate fire effect I count the number of figures plus the gun, cross reference the total with the fire factor (the same as the equivalent number of muskets, plus tactical factors – canister gets +1)and… hey presto… a number of casualties.

This is a simplified explanation as I find their is nothing more tedious than ploughing through a section of somebody else's rules , when you don't have the whole of the rules to put the section in context.

Oh yes, I limit the use of artillery to a maximum of 900 yards in the field, as it seems to me almost impossible to observe the fall of (round) shot beyond that distance (No explosion, no puff of smoke, just a spatter of earth at first graze)

BTW does anyone legislate for the penetrating power of round shot? I have seen more than one game where first line units have taken casualties from shot, but the unit 50 paces to their rear have remained unscathed.

Quarrie rules specify penetration (i.e. bounce) distances for roundshot but I never really understood how to use them.

The discussion is interesting but for me not hugely germane to the question of how to reflect artillery on the table top. If one performs exhaustive research into artillery effectiveness and comes up with an accurate answer, fine, but your game will still play wrong if you don't do the same for musketry and melee. Including one accurised element among a bunch of conjectural ones just biases outcomes in a different way.

I haven't re-read the entire thread but did the calculations of effectiveness reflect the fact that the cone of fire isn't a cone? i.e. only in a six-foot high band across the middle of the cone does anyone get hit?

Surely the effect of canister is that, at any given range, some %age of that rectangle is occupied by roughly evenly-distributed bullets. As the range increases, the dispersal gets wider and starts to permit gaps between bullets that are wide enough you could stand in one and not get hit.

Were guns fired singly to optimise the effects on those in the "cone", or were they fired in full-battery salvoes? If the latter, that would suggest the chances of surviving one gun's canister discharge at you were quite good.

Your analysis is very interesting. Since you appreciate details, I write with some information I came across once upon a time and put aside until I had more time.

Your discussion of penetration reminded me of the tables I once found. These were in the "Archiv för die Officiere der Königlich Preussischen Artillerie- und Ingenieur-Korps" from 1841 (third issue of the 12th volume; 1841 was the 6th year), starting on page 62 entitled "Versuch über die Perkussionskraft der Geschosse" (Tests of the striking power of projectiles).

I only made a copies of a few example pages to remind me as I did not have the time for full reseach and study. What I did copy might help, though.

For cannister, the tables include penetration information at different distances in Schritt (Prussia "Schritt" of about 0.8 meter). They start at 400 Schr. go up by 100's (the upper end ends when penetration stops) for the 7-lb howitzer, 10-lb howitzer and the 6-lb and 12-lb cannon (not sure which models for those into the "M"). Each cannon made separate trials with the cannister rounds with light and heavier balls (I gather the howitzers only had the heavier rounds). Let me call the two "cannister" and "grape" without starting any debate about the actual terms.

These show the number of balls going through a series of walls (96´ wide by 6´ high by 1" thick – Prussian measure where my notes say a Prussian foot was 31.385 cm and their Roth was 2.61541 cm so the walls were 30 meters wide by 1.9 meters high and 2.6 cms – about an English inch – thick). They called them walls.

Simply put, there is penetration of more than one "wall" for all howitzers and cannon with both grape and cannister. There is a statement that the equivalent of a ball to going through a man (or the equivalent: a calf) would be 4 or 5 of the 1" walls. At the same time, simply hitting a wall would not put a man out of action (although two such "hits" could). Thus, the tests required a ball to lodge in the walls and not simply hit them.

For the 6-lber the 123-ball cannister round at 400 Schr. had 26 in or through the 1st wall and 4 in or through the 2nd wall, then at 500 Schr. it had 14 in or through the 1st wall and 1 in or through the 2nd wall.

The 41-ball grape round at 400 Shr. had 13 in or though the 1st wall, 11 in or through the 2nd wall, 7 in or through the 3rd wall and 3 in or through the 4th wall; at 500 Shr. had 10 in or though the 1st wall, 5 in or through the 2nd wall, 3 in or through the 3rd wall and none in or through the 4th wall; then at 600 Shr. had 10 in or though the 1st wall, 3 in or through the 2nd wall, 1 in or through the 3rd wall and, again, none in or through the 4th wall. Finally, at 800 Schr. (nothing is give for 700 Schr) one 2 were in or through the 1st wall (and none thereafter).

They then converted these hit into a measure of "ausser Gefecht" or causing an "out of action" hit (they do not claim these equal casualties my notes show) and they do this for different types of terrain with talbes for a round and for three minutes of fire.

The table for single rounds shows 6-lber pieces with cannister (light rounds) created more the causualties of grape (heavy rounds) at 400 Schr. (26 versus 13) and 500 Schr. (14 versus 10), but then the cannister causes no effect, while the grape does (10 at 600 Shr., 6 at 700 Schr and 2 at 800 Schr). This is in completely good terrain.

For barely good terrain, only the grape (heavy balled cannister) has effect: 11 at 400 Schr., 5 at 500 Schr. 3 at 600 Schr. and 1.5 at 700 Schr.

I got these copies from the local university library's reserve library. I know I have seen that some of the series is on Google, but in a quick search I did not find this volume (as yet).

Oh, yes, there is also a table that includes tests at different distances (again, the Prussia "Schritt" of about 0.8 meter) from 5 to 400 for infantry weapons (pistol, carbine, musket, jäger and wall-mounted weapons and their penetrations. (If found I would appreciate a link.)

For your analysis, it is a same I did not copy all the text pages; still, I trust this helps. – R

Ralpher, thank you very much, your data does interest me a great deal.

If you do find the info please do not hesitate to email me at sdev2749@bigpond.net.au

Regards,
Shane

I finally found it on Google and e-mailed the address to you. Enjoy. – R

Enjoy?

Shane, Oliver Schmidt has posted some Scharnhorst trial numbers in the other canister thread on the nap history board two days ago.
In my earlier post on this thread here, I took Jeffrey´s claimed inner high density cone of 32ft at 400yds as an example.
Scharnhorst states that there is indeed an inner cone where 80% of the balls would hit, but it is much bigger than Jeffrey calculates.
he has no exact numbers for 400yds, but for 500 and 600 paces.
At 500 paces ( 375 meters/roughly 375 yards) its 100 feet, not 32 feet like in Jeffreys calculation.
At 600 paces ( 450 m/yds) the cone is 150 feet.

If we take 120 feet for 400 yds then the cone is vertically 20 times larger than a man sized target.
Horizontally the balls will theoretically be able to hit over 50 men, but only 5 to 10% of the balls will fall into the tarrget zone.
If we assume that only the inner 80% cone counts and that at this distance only heavy 40 ball canister is effective(Ralphers post), then we have 32 balls in the target zone, but only from 1 to 3 balls will hit, in the worst case not a single one.

Thank you Cptn,

The information of the Muller trials as given by Hughes does point towards this same conclusion. George says the cones were much closer together thus the percentage of cone cover along an enemy line is smaller than thought with gaps along the line due to the regulation spacing of the guns. I see (my own conclusions) that an Optimal Canister range is the distance where the cones overlap or at least touch. not the outer cones (20%) but where the inner 80% touch or overlap.

Using George's conclusions this distance is about 250-350yds but if you agree with Hughes's assessment this distance is closer to 100-150yds max. This optimal or saturation point is the peak of coverage where the maximum damage will be inflicted.

So according to the information you have the direction aims to a conclusion closer to Hughes and if so then the cones being much larger as you say means also a larger spread of fire with larger (on average) gaps between each ball increasing with each range increment. This is a two edged sword as although the coverage reaches Optimal/Saturation point quicker or at a closer range the spread of the balls actually increases thus dropping the chances of hits as you say.

My aim was to find the optimal/Saturation point (range) where the maximum casualties should be caused. The conflicting data suggests great disparities between the two thoughts and coming to an accurate conclusion as a rules writer places my in a quandary to say the least. Luckily War Games rules work on Range grouping such as for example 0-50yds – 51-100yds – 101-150yds and so on. so using this kind of thinking I can confidently place the groupings as I think are accurate and assign the appropriate factors according to what I think is pretty close to the mark.

Funnily enough I went for the middle ground between Hughes and George so I feel pretty good about it in the end. I knew if I went by George Jeffrey's data I might be too far one way and if I used Hughes data I might be too far the other so a compromise was on the cards in the end.

I would dearly love to see this data you talk of however, I thrive on data (is a major part of my daily work) and I enjoy working with figures so any new information would be gladly taken on board. Luckily my rules are all written up on Excel spreadsheets and easily corrected or changed so that new information can be factored in whenever it needs to be.

Regards,
Shane