Testing the affects of cannister using a Napoleon 12 pounder.

[Didz]

A group of ACW re-enactors have been testing the spread of heavy canister using a Napoleon 12pdr cannon, which is about the closest thing available to a French 12 pdr Field Gun.

We fired 17 rounds of Canister from an original Civil War Napoleon Field Cannon for a study on battlefield archeology. Each Canister shell contained 27 steel balls 1.5" in dia that were painted a certain color and then fired out onto a field from various cannon positions. The cannon's coordinates and the balls that were found were all documented using GPS Technology. The cannon was fired at various barrel elevations and at different levels of terrain. The data when analyzed will hopefully help archeologists determine where the cannons may have been placed during a historic battle.

I asked for a link to the result they recorded, but as predicted it looks as though the balls carry much farther than one would expect based on normal wargame standards, so it is reasonale to assume that it's use as a short range weapon was more to do with ammo conservation than effect.

Whatever, the outcome this was a far more useful test than the one conducted by the British Army for the Waterloo documentary which actually produced no usable data.

[wangrin]

I've writen about somewhere in TWC, but you can have a look to the book "Traité d'artillerie théorique et pratique: précis de la partie élémentaire et pratique", by Guillaume Piobert, 1836.

This book report several artillery and firearms tests.
For example, here are the cannister test results :

• balls spreading :
  
  
  
  column 1 : target height (meters)
  column 2 : field gun caliber
  column 3 : distance to the target (meters)
  column 4 : number( ?) of balls that hit the target compared to the line of sight.
  
  For example, at 600m, 53% of balls hit the target in an area of 1m around the line of sight.
• balls effects :
  
  
  
  column 1 : dimension of the target (height x lenght)
  column 2 : caliber of gun
  column 3 : weight of balls (gr)
  column 4 : number of balls (per layer and per box)
  column 5 : percentage of balls per shot that hit the target toatdistance of .... meters
  column 6 : percentage of balls that pass through a pine plank (27cm thick) at the distance of ... meters
• balls effects
  
  
  
  Test procedure :
  Canister have been fired from a 8pdr gun.
  Pine planks (Lenght : 4 m, height : 4 m, thickness : 3,1 cm (31 mm)) have been posted at 40m, 50m, 60m, 70m, 80m, 90m and 100m from the gun.

[Didz]

Thanks Wangrin, hopefully I've got these test results somewhere in a translated format. My main interest was to see an actual test conducted in real life.

[wangrin]

Maybe is it possible to propose a similar protocol to the re-enactor group.
The problem when only firing canister shots and looking where balls have fallen doesn't allow to define the efficiency of such weapon.
What is needed is data about spreading and penetration effect depending of distance, something that was not the goal of the re-enactors.

Did they have published their results ?

[Didz]

In my opinion the ideal test would be firing through a series of company size screens (6' x 20'). These would be ranged at intervals of say 50m, 100m, 150m etc. up to 600m and the test would record the number of balls that penetrated each screen and the widest spread of penetrations on each. The only thing better than that would be to literally arrange dummies in ranks instead of the screens and then record how many were hit.

The problem of course with counting holes is that it doesn't actually measure lethality. At closer ranges the likelihood would be that those directly in the line of the shot would be hit multiple times, whilst the spread would be quite small limiting overall casualties. At longer ranges the wider spread, plus the ricochet of shot that struck the ground short, and the fallback of shot that flew over the heads of the closer targets might actually increase lethality.

To be honest penetration is probably of little interest as getting hit by a ball is going to take a man out of the fight. Though it might have more value for naval cannon that need to penetrate the hull before being able to kill anything.

The problem with the GPS plotting of resting shot is that whilst it gives a footprint for the spread of shot it doesn't give any indication of the lethality of each shot as it traveled downrange. So, for example a shot which left the barrel in a flat trajectory and traveled downrange 600m might take out a dozen men en-route, whilst another shot resting right next to it that flew high might have passed over everyones head and simply dropped harmlessly next to it's lethal friend.

[wangrin]

Test could be composed of two sessions :
1. testing spreading : as you said, using screens ranged at intervals (each 50m)
2. testing penetration : using plank or, if possible, "thin" sand bags (to try to represent the resistance of human body), and testing each range one after an other (each 50m for example).
Advantage of "sand bag" is that it would be easier to repair

As the penetration power depended the kinetic energy, it could be interesting to measure speed. This would need high velocity cameras and a "clean" environment. I doubt it it possible.

[Didz]

2. testing penetration : using plank or, if possible, "thin" sand bags (to try to represent the resistance of human body), and testing each range one after an other (each 50m for example).

I don't think you can measure lethality using wood planks or other inert substances like sand. I've seen tests using ballistic gel, but even that is inaccurate simply because of it's uniformity. So, for example a shot hitting the target in the skull might be lethal, whilst if the same shot had hit an arm it might not take the soldier out of the fight. Bones are also a vital factor in pentration, and so unless the ballistic gel is covering a skeleton made of more solid material one will not get an accurate result of likely penetration. Bones may not stop a shot but they did use to deflect it, and at the same time bones fragments from one victim can and did injure others.

On top of that complexity, we are still assuming that all these soldiers were wandering around naked whilst getting shot at. In practice uniforms and equipment had varied potential to absorb or deflect shot, and also had varying potential to become lethal when hit by shot.

French soldiers regularly went into battle wearing their overcoats. Not because they were cold (they even wore them in Spain) but because there was enough evidence from the stories of veterans that the more layers of material you could wear the less likely you were to be killed. A great coat, over a jacket, over shirt, over a vest might not stop a cannon shot, or even a musket ball at short range. But in a battle there was enough spent shot flying around and other stuff like bone fragments, stones, and bits of kit to make that extra thickness worth the sweat of wearing it.

Prussian soldiers also recognized this and tended to wear their greatcoats rolled across their chest and back providing even more thickness over the heart, spine and one side of the neck.

On top of that you have other equipment that provided more localized resistance to the passage of shot. Such as leather crossbelts, brass chest plates, the musket itself if held across the body, the soldiers backpack crammed with spare clothing and cooking utensils. All of these things provided both some added protection and at the same time an added risk. A spent ball hitting musket stock, or belt buckle might well save the wearer from serious injury. But, at the same time a cannon shot passing through a soldier wearing all that kit might well spray the contents of that victims backpack in a wide arc of lethal cooking equipment and looted coins.

So in fact unless one was to provide the testers with hundreds of anatomically correct manikins dressed in accurate period costumes you really aren't going to get anything like accurate results for either penetration or lethality, and you may as well not bother trying because trying to equate say a plank of wood, or a sand bag with a human body is nonsense and will just skew the other results by blocking the passage of shot which shouldn't have been blocked.

For example I saw one re-enactor testing the effect of solid shot from a civil war cannon of oil drums full red dyed water, and then trying to argue that the effect was similar to that of a hit on a human being. But spectacular though it was, it would not equate in any way to the effect of firing a solid shot into a dense mass of living flesh.

I've also seen weapons tested on the carcasses of dead pigs, which are similar in body density to humans, and so produce more accurate results. But unless they were alive when hit, and dressed in full uniforms and equipment they would not produce an completely accurate impression of a solid shot passing through a body of troops. We know for example that eyewitnesses to such events describe a fine pink mist that rises above such formations caused by the arterial blood flow of dismembered bodies being dissipated by the shock wave of the ball passing through them.

We also know from real events like the King Cross bombings that most injuries in any such event are caused by objects other than the actual weapon itself. So, bone fragments, bits of metal, and wood being hit and flung at force in all directions.