Further to my release of my documents into the public domain here is a paper suggesting how PM motors might work using what I call "constrained magnetostriction". The principle is that a magnetostrictive substance, when placed in a magnetic field, will try to change its shape. If that change is prevented by some siutable constraint, then internal stress will occur and that in turn will alter the magnetic properties of the material by the inverse magnetostrictive effect (Villari effect). This is offered as a suggestion for breaking the torque waveform symmetry that usually applies.
Enjoy!
Here is another suggestion for how some PM motors might work using conduction electrons in the PM to achieve some degree of dynamic magnetization movement.
Smudge