After saving some money to build, I will be testing a device that is based on the fanner principle presented in earlier threads. This principle is mixed with the cogless alternator idea, which was also presented in earlier threads. Both of these ideas were presented by Butch LaFonte at an earlier time (his newer ideas are awesome (check them out!), but I love the fanner's simplicity).
The first part of the device works by a force acting on plates orthogonally to the magnetic field that permeates them. Since the field is in the same direction, the plates separate in the same fashion as iron particles sprinkled on paper above a magnet. This works great, but by introducing the cogless alternator idea, it makes this work so much better. By having a diametrically magnetized ring magnet rotate inside a toroid, the magnet has no cogging, as any angle looks the same to the magnet, so there is no change. By diverting the flux in the toroid, there is no torque on the ring magnet.
If the ring magnet is in one position, the flux wants to stay inside the toroid, and in another, the flux wants to go through the plates. All of this happens with no torquing on the ring magnet, which makes the fanner concept easier to play with.
Here is the FEMM simulation, which is a 'flattened' version of the design due to 3D designs not being handled well. It's the same thing as the other animation.
The design is basically two diametrically magnetized ring magnets on one shaft, with one 180 degrees rotated from the other. There is a tiny airgap between the magnets and the soft ferrite toroids (to eliminate eddy current damping), which have been cut to divert flux correctly. The plates can be iron.
With this design, the 'power stroke' happens twice per revolution, so I feel a one way bearing with a linear toothed rod and a toothed gear attached to the bearing would do well to rotate the magnetic assembly (or a 2:1 gear?). One unit alone would ONLY rotate the shaft maybe part of the way, more of them on the shaft (maybe 6 units) would be required to have a constant torque on the shaft to keep it rotating.
The toroids are cut along the top and bottom to dissuade the flux from going back to the other side of the toroid and persuade it to go through the plates instead during that position.
For the simulation, I got a net torque (via weighted stress tensor) of 0.000531 N/m, and the force (via weighted stress tensor) was up to 11 N (22 N for both sides of plates) in the Y direction for the plates.
Any ideas/thoughts/suggestions (I would especially like the input of Low-Q, as he has a very objective/logical view on these devices)? Does anyone see any problems with the design? I have ordered some large ring magnets to start testing. The only issues I might see are with the permeability of the ferrite messing things up, or the force of the plates being too small still to do any real work. Sorry for the animated MS Paint art, it took forever to draw! If anything at all, I'll provide details if it fails.
The purpose of this device is to utilize the 90 degree force of the plates, and to make them 'invisible' to the ring magnets as they happily rotate around, not seeing anything other than the toroid.
Here's a link to the animated gif, it wouldn't work below.
http://oi57.tinypic.com/105y52t.jpgTHE RING MAGNETS BELOW ARE 180 DEGREES APART, SO THE 'N' OF ONE IS AT THE SAME POSITION AS THE 'S' OF THE OTHER.