F.B.D.I.S.S.M - Flux.Boosted.Dual.Induction.Split.Spiral.Motor.

[Low-Q]

Femm can't simulate this properly, look at the reszlts of the simulation, all the flux is concentrated around the ends of the spiral. Because the program is 2d it can't simulate the flux lines flowing under and over the spiral which is what happens in real life.
The flux lines have to complete a closed circuit because the program is 2d the only way the magnets can complete the cycle is by looping around each end of the spiral.
I allready tried some practical experiments, and from what I can see, the spiral behaves quite differenty than that shown by Femm, It has some torque around the whole spiral as Honk predicted and not just at the extreme ends as shown by the Femm simulations.

Magnetic fields interact only in the same plane. Two magnetic fields that are crossed by 90 degrees does nor repel or attract, so 3D is not necessary. However I understand your concern regarding FEMM, but as far as I know, the magnet arrays, even if they are bent like a half "circle", it will behave as a long straight magnet to a smaller magnet. A smaller magnet will allways try to find the middle of the long magnet - an optimum resting point. As you can see in FEMM, there are few or no magnetic lines in the middle of each statormagnet array which in fact (as far as I can see) is simulating exactly that. Maybe I'm totally wrong - I hope so. Maybe Honk will succeed with his project

Vidar

[Honk]

A smaller magnet will allways try to find the middle of the long magnet - an optimum resting point.

If so, why isn't Paul Sprain magnets then focused further away from the spiral ending, than the video show us?
Well, just because the rotor magnet cannot reach a higher flux area due to the gradient spiral. It has to travel to the end to find the most flux.
If there was no shaft or the spiral was not gradient, then of course the rotor would seek to find the middle. But in this case it can't.

[acp]

Magnetic fields interact only in the same plane. Two magnetic fields that are crossed by 90 degrees does nor repel or attract, so 3D is not necessary. However I understand your concern regarding FEMM, but as far as I know, the magnet arrays, even if they are bent like a half "circle", it will behave as a long straight magnet to a smaller magnet. A smaller magnet will allways try to find the middle of the long magnet - an optimum resting point. As you can see in FEMM, there are few or no magnetic lines in the middle of each statormagnet array which in fact (as far as I can see) is simulating exactly that. Maybe I'm totally wrong - I hope so. Maybe Honk will succeed with his project

Vidar

But magnetic fields in real life are 3d not 2d. The flux lines have to complete a circuit, they have to leave and enter the same magnet. In Femm this is only possible for the middle magnets of the spiral by running around the ends of the spiral and back again, This distance is too far, consequently the fields are negated. In real life however, the magnets in the middle of the spiral can complete their cirduit by looping under and over the spiral which is a much shorter distance than looping around each end of the spiral as in the 2d case. Long lengths of magnets arrays without gaps between aer not acurately modelled in Femm.

[Honk]

We have a winner in the magnet supplier contest!
And first place goes to www.buymagnet.com.
They will deliver the magnets for $1328 US (shipping and Vat included)
http://www.overunity.com/index.php/topic,3456.msg58948.html#msg58948

[ecc]

Great Saving

By the way, buymagnet.com and ndfeb-magnet.com  seem to be one and the same manufacturing company,  Shenzen Shan, in China. They also produce small motors.

Will they supply the rotor magnets shaped according to the drawing? Are there any news on the laser cutting of the steel parts?