Since we have been having doubts from time to time about the “cogging” torque produced by a permanent magnet motor when it is rotated by external forces, I’ve decided to illustrate the issue and make sure this will not be a problem in our design.
The cogging torque is an undesirable torque that is produced by magnetic interlocking of stator and rotor teeth when the rotor spins, even if there are no electric currents applied. That’s something we really don’t want in an (H)EV motor, because it means waste of energy.
We initially chose the LRK motor design because it was available on the web and it “looked good”. 😉 But the more I study it, the more I admire it.
I went to the FEMM examples page, and got both the LRK model file and a simple lua script to determine the cogging torque, both created by one Edie Currants in 2004. Hmm… that name sure sounds bogus to me… Eddy Currents?…
Here is the model, an LRK motor in all it’s glory (click to enlarge):
I ran the script and made this graph (click to enlarge):
As you can see, the torque is always nearly zero when rotating the motor without current!!! 😀 This is only possible because the number of teeth in the stator and the rotor are different, so that the stator and rotor are always equally well-aligned (or misaligned), no mater which angle or position the rotor takes. There is no preferential stable position for the rotor, like there is for example in a “step motor”.
There are 2 small discontinuities in the graph; those where the points where FEMM crashed and I restarted the batch run, “more or less” from the same step… 😦 Fortunately, I now know exactly which LUA function makes the program crash, so I can avoid it if I want. 🙂 Unfortunately, it is the bitmap saving function, so there won’t be more movies until the bug is corrected! 😦
Anyway, the point is well proven. The motors won’t be “dragging” along when the car is driven by the combustion engine or coasting on the highway.
Thank you, Edie Currants! 😉