Regarding heat created due to magnetic effect 1

[WSU2004]

I am designing magnetic coupling, I am creating torque on copper disc by putting magnets infront of it(Magnet rotor). As one of them spining other rotates because eddy current. If I put stronger magnets like N48 or higher, I am creating more torque on copper disc but at the same time i am generating more heat. Did anybody know, if I provide silicon steel ring on the face of copper disc, is it going to help me reducing the heat problem, how much it will effect on my torque level?

Regards,

 

[MagMike]

The answer depends on what you are trying to do. There are three types of magnetic couplings that I am aware of: Synchronous couplings, eddy-current couplings, hysteresis couplings.

Synchronous couplings involve two magnet assemblies arranged so that they spin together, without contact, at the same rate, up to a maximum level of torque.

Eddy-current couplings are used when the speed of one member needs to be held constant while the other member varies. Usually these involve a magnet assembly spinning relative to an electrical, non-ferromagnetic disc or ring.

Hysteresis couplings are used when one needs a predefined level of torque, irregardless of input speed. Usually these involve a magnet assembly spinning within a circumferentially oriented magnet ring.

If you are designing an eddy-current coupling, then adding a silicon steel ring will not help. It will work against the function of the device because it will be strongly attracted to the magnet.

Heat is an byproduct of eddy-current devices, there is no way to have eddy-currents without heat. The only option I can think of is to introduce some forced air cooling.

 

[MagMike]

In addition:
There is a paper published in IEEE Transactions on Magnetics this month by some researchers in the UK:
"Eddy-Current Coupling with a slotted conductor disk" by Razavi & Lamperth, page 405 (March 2006). If you are working on eddy current devices, you may find this interesting.

 

[UKpete]

Another way of looking at an eddy current coupling is to compare it with an axial field induction motor. An induction motor has a stator containing a three phase winding connected to a three phase supply, producing a rotating field which induces currents in the copper bars of the rotor, which in turn produces reaction fields and hence torque. What is different in the eddy current coupling is that the rotating field is produced by spinning magnets rather than a stationary three phase winding. Also your coupling is axial field, whereas the vast majority of induction motors are radial field (although axial field versions are available and they operate on the same principle).

The reason I'm explaining this equivalence is that you can use the same techniques as for induction motors to optimize the design. Placing a laminated steel disk BEHIND the copper disk will help increase torque because it will strengthen the field (induction motor rotors have a laminated steel core for the same reason), but of course it will not reduce the copper loss per unit torque. As Mike says in his post above, this will give an axial force which will require some kind of thrust bearing. I was going to suggest that you might balance these axial forces out by having a disk either side of the spinning magnets, but it may not be feasible.

 

[WSU2004]

Thanks Mag Mike,

I think putting slotted copper disc might help us. I am designing the eddy current coupling. As per UKPete I am going to throw some run in Faraday (FEA tool)for standard silicon steel ring like M19 & see what happens on the torque level. But getting heat out is still a big issue. Water cool conductors might help, but what is permeability of water does that effect my magnetic flux or not?

 

[UKpete]

The permeability of water is 1.0 i.e. the same as air. I'm not sure how to water cool a spinning disk, it will be indirect cooling. In induction motors etc the rotor is cooled by air convection through the airgap and possibly by conduction away along the shaft.

 

[UKpete]

WSU, I've just happened across the following quote in Gieras ("Axial Flux Perm.Mag.Brushless Machines"):

"It is difficult to manufacture a laminated rotor cage winding for a disc-type induction machine. If the cage winding is replaced with a non-magnetic high conductivity (Cu or Al) homogenous disc or steel disc coated with copper layer, the performance on the machine drastically deteriorates".

This appears to suggest that your solid disc is not very efficient, you will get better performance with a radial copper bars (the ends of which will need to be connected by a copper ring on the o/d and i/d; like a bicycle wheel).

I should add that I am not familiar with eddy current magnetic couplings, so I don't know what the "state of the art" with these is (I'm coming at it from a motor design background). And of course Gieras could be over-stating it!