There is a connection here between your observation about momentum and the Keenie. Also with the Milkovic proposal I made in my recent post.Kirk wrote:...
I believe the method is simple and Newtonian. There is no non linearity in position so the answer is not there. What does fit his clue is the balls working in pairs. A large ball dropped a short ways has sufficient momentum to drive a smaller ball to near the top of the wheel. Momentum is highly non-linear A 1 foot drop will give you a quarter of the momentum of a 16 foot drop. The inference is obvious.
In all three cases we are dealing with interactions between large and small masses, large and small inertias.
In the case of the Keenie it is the interaction between a single weight and the remaining weights acting as a whole.
In the case of the Milkovic proposal the differential inertia is more subtle. Superficially it appears that we merely have two masses of identical inertia. However account has to be taken of the one-way clutch. This governs the interaction between the inertia of the active freely rotating mass (the red pendulum bob in my diagram) and the passive mass (the "earthed" green and yellow bob).
The earthed bob has the inertia of the earth when subjected to a counter-clockwise force and its own inertia (equal to that of the active red bob) when subjected to a clockwise force. In effect it acts like a valve which only transmits the clockwise couple part of the active red pendulum's rotation.
I agree that the interaction is highly suggestive though I doubt many members would find it "obvious".