Going back to my question for Jim K. in the other thread, there are a number of locks with U, V & plain-flat springs that have the mainspring arced when in the full-cock position. I was curious to understand the reasoning as to why there would be any concern over the amount of arc in the lower leg as Jim alluded to referencing Bob R's V-spring lock. As Mark E pointed out, the very nature of the U & V style mainsprings becomes convoluted with both the complexity of motion and physical characteristics of the design. Typically a flat spring is considered a short-throw linear motion but these V/U designs in lock applications also involve an amount of torsion and the varying cross sectional area also plays into the overall scheme. Looking at it purely from the resulting function, the spring does nothing more than provide power to the tumbler, the tumbler doesn't care where that power comes from or how it gets there. The resultant force of the cock movement is attributed to the geometry of the tumbler and not to the spring because as the spring is deflected further into its loaded position, the amount of stored energy increases. The same increasing load condition will be attained no matter if the power to the tumbler is being applied by a flat, compression, extension or torsion spring.
I can't do anything more than speculate on Bob's V spring without knowing if it has a second retainer on the upper leg or not as does say the U-spring in a Siler lock. In the Siler, the upper leg is anchored at the bolster and at the pin just above the U-bend rendering the upper leg into a fixed position and as such the entire load is taken by the linear deflection of the lower leg and torsion deflection of the U-bend. If the upper leg in Bob's lock does not have the second anchor point on the upper leg, such would allow compound linear deflection of both the upper and lower legs proportionally thus there would be very little force applied to the Vee as it would essentially become nothing more than the floating transition point between the upper & lower leafs. The same but to a lesser extent would apply if Bob's lock has a second anchor point on the upper leg that is further away from the Vee.
The life of any spring is primarily attributed to the amount of fatigue it sees in-use be that from cycling, deflection distance, resonance, shock-loading, ect. In the case of a flintlock, the first shock-load the mainspring sees is when the flint strikes the frizzen, the second shock-load is seen when the frizzen resistance is lost and finally when the cock makes the dead-stop at the bottom of its stroke. No matter the source of the shock-loading, there's going to be three primary nodes that take the brunt of the shock-load, one near the tumbler/stirrup, another near the bend on the lower leg and another near the first anchor point on the upper leg. Similarly the dynamic loading also produces areas of higher stress and fatigue than others, the key is to have a working balance between the design and the acceptable dynamic and shock load characteristics. Look at Dave's spring, like the Siler and other similar locks, the cross sectional area of the spring is considerably greater at the bend so the loads are distributed over a larger area which reduces the per-unit loading. On Bob's spring, there isn't a major increase in cross sectional area at the bend but rather it appears to have a fairly uniform taper to either side of the bend. In both cases the per-unit loading is reduced. In the grand scheme of things, it's nothing more than two halves of the same circle, one goes CW, the other CCW but they both accomplish the same goal of meeting at the same point. Harkening back to reference about the amount of deflection in the lower leg of Bob's lock, the amount of deflection does not pose any concern as long as it's within the working parameters of the spring material, design and per-unit loading.
As for the issue of the design being historically correct in appearance, that's a whole different thing, I'm merely commenting on the different manners in which one can obtain the same function via different design parameters. Jim, as for your comment on the strength of the spring, such can be the case but that's an issue with the specific spring because no matter if the spring is thinner but longer or shorter but fatter, the load characteristics be the exact same but the loading characteristics will be completely different. From the function standpoint only, one could easily replace both of these styles of springs with a very short eccentric S-style torsion spring low-slung on the bottom of the lock plate and it would be a lot less susceptible to the shock-load fatigue and the pre-load relaxation that is realized in flat springs although not be anywhere near historically correct.
