glad to know I am asking difficult questions. That indicates that they are probably worthwhile. I'm pondering the side discussion between yourself and Don about the shape of the impact area have some thoughts driven by modern weapon characteristics. It's may be an unexpected conclusion, but many things that turn out to be true are unexpected until fully analysized, at which point they suddenly seem "obvious".
In writing my questions I had also been thinking about what I would think the most effective mechanism would be, just from first principles. Factors I started thinking about included spread of the shot at various ranges, dispersion in the vertical plane, erosion to the barrel and muzzle, failure modes and documented use. At the time I was anticipating a cone like dispersal in both horizontal and verical planes, subject to the ballistic characteristics of the primary munition with individual shot operating within couple of standard deviations of the primary ballstic curve. I also figured that the performance individual shots would be skewed to poor performance since a certain amount of energy would be lost as those balls bounced against each other, plus smaller shot tended to have worse balistic performance over longer distance.
So preliminary thoughts I had included that it was probably beneficial for the individual balls to have their individual trajectories set before leaving the muzzle. That is if the casing burst after leaving the muzzle, then a certain number of individual balls could go in highly unexpected directions. However, I also considered that that might increase erosion. I did consider that the sabots, whether iron or would must have had very similar performance otherwise we would see that described in primary material. I also tried to recall if I had ever read that gunners were encouraged to fire cannister only when necessary in order to prlong barrel life, but could not. However, I also knew that in many cases dunners fired round shot preferentially even when within what might be considered cannister range (nod to Don, in simplicity I shall stick to one term).
One of the worst designs I could see would involve the cloud of shot losing forward speed at the muzzle (which could happen if the cannister burst on leaving the muzzle causing indivual balls to convert velocity into multiple vectors) and then have the sabot launch through the cloud of shot. This meant that the sabot would have to be designed such that it would either failure to survive passage through the bore or it would be so inefficient aerodynamically that it would lose speed faster that the balls might.
I was also wondering about what we now call first strike, grazing and last strike in machine gun theory when applying the cone of fire to the ground. In MG theory, fire by an individual gun does form something like a cone when viewed from the muzzle. However, impact on the ground is shaped more like an elongated ellipse. I think this may tie in nicely with Don's point. The explanation for the MG impact are is as follows. Imagine a cone of fire leaving the MG and following a ballistic trajectory. After a while rounds will start to impact the ground. The first rounds to strike will be those at the bottom of the cone projected forward from the barrel. The last rounds to strike will be the ones at the top of the cone projected from the barrel. These two strikes define the beginning and end of what is called the beaten zone and they are generally directly in line with the muzzle of the barrel. Other rounds which vary from the straight line possess equal energy, but slightly different vectors. The furtherest lateral movement of the outlying rounds is a function of weapon and munition design, with most MGs designed to allow a certain amount of play in the mount. Some have been historically highly accurate (considered too accurate by some) such as the Bren. Deposition of the cone of fire on the ground can be approximated as a form of diamond, or elongated ellipse. The more I think about the dynamics of the individual balls in a cannister round, the more I think a similar model could be derived to project the cone of fire onto the ground.
If this is correct, then gunners of the day should have had a good idea of a number of performance characteristics including:
- distance from first to last strike (based on calibre, slope of ground, distance to first strike)
- shape of what we would now call the beaten zone for various ranges
- maximum width of the beaten zone.
From this, I would infer that there was probably an optimum range for each calibre to fire cannister. At closer range the impact from an individual round of cannister would affect fewer and fewr files as the target closed. This brings me to another point from MG theory. MGs are not suitable for point defence for exactly this reason. Therefore we work to place MGs to enfilade other friendly units. That is MGs are sited to fire across the front of other troops so that attackers have to push through increasing or constant beaten zones instead of diminishing beaten zones.
I would be fascinated to know if such metrics existed in something like Adye or any other Aide Memoire of the period.
It is interesting to note that guns of the period typically fired in their own defence, not in defence of neighbouring guns. The exception to this was fortifications, which were designed to keep guns in defilade and to maximize enfilade fire.
This is not the first time I have considered exact parallels between period artillery and MGs.