I have now come up with a model that fits with the reason I asked the question, and that integrates with every single technical opinion given in this thread.
If example 1 is correct, because the surface on the inside diameter has a reverse polarity to the surface on the outside diameter, and an electric coil placed inside a non-magentised sphere induces a magnetic field with the same polarity on the outside (that model was the reason I asked the question), then the two objects will attract at the poles, the interior surface of the pole of the shell attracts to the pole of the coil, locking the objects to the rotational axis. The objects will at the same time repel at the equator, if I am correct (the field lines will run in opposite directions perpendicular to the equator. ) It would be good to have that last part validated or falsified, the rest has already been discussed I think in this thread, as this particular model conforms to example 1 which most opinions converged on.
So, the model actually seems to fit even tough the initial assumption was, it appears based on this thread, wrong.