![Image](http://www.besslerwheel.com/forum/files/ejrn_sketch_719.jpg)
If you have two FCPs following closely on one another they are wider apart near the nadir where the pendulums are travelling quickly and closer together near the zenith where the pendulums are travelling slowly.
If the bobs are connected by bellows then the gas inside the bellows will be compressed as the bobs approach the zenith and expanded as the bobs aproach the nadir.
A set of pendulums following one another at a set time interval will therefor provide an expansion/compression cycle.
Connecting bellows under compression with bellows under expansion will provide a flow of gas which can be used to drive a fan at the axle and step down the 3rd derivative precession kinetic energy (PKE) to 2nd derivative rotational kinetic energy (RKE).
Using gas pressure differences would seem to be rather impractical but one can achieve the same effect using mechanical links driving a cranked axle.
The leading pendulum will be attached to the axle and the following pendulum will be free. We can think of the leader as a stator since it is static in relation to the axle and the follower as an oscillating rotor.
One only needs to generate enough energy to get the pendulum over top dead centre and then the cycle will repeat.
The relatively small back and forth movement between the stator and rotor puts me in mind of the relatively small movement of the displacer in my model Stirling Engine. This in turn suggests the the action could be harnessed mechanically by means of a suitable linkage involving a long con-rod and cranked axle.