Discussion Forum

Dear all Comsol experts and users,

I am modeling a synchronous permanent magnet machine (surface mounted). With my simple model (Time dependent which involves the rotation of the rotor at open circuit) I can fairly, in my opinion, compute the back emf relate to one phase as well as reasonable values of electromagnetic field in order to see how much the machine is loaded magnetically speaking at open circuit (no armature reaction injected in the stator winding).
Nevertheless I have a problem with torque computation. The cogging torque should oscillate over time, but showing an average value which is zero. I used the force computation tool available in the rotating machinery subpackage paying attention to surround it with air in order to exploit the Coulomb's method for computing the torque. However I got a results which is not correct: the cogging torque is oscillating but the average value is different from zero. Moreover, in order to check the reliability of my results I computed the force also at stator side. According to the principle of action-reaction the torque should result the same: this is not true in my case: torque oscillating with a different frequency of the one acting on rotor, plus an average value, which is not zero yet.

I attach some pics down here, as the size of the file does not allow me to upload it on the website. In particular regarding the Cogging.png the green line represents the torque "felt" by the stator whereas the blue the torque computed on the rotor. Please if someone can help me out, or having suggestions I would be more than grateful.

Sincerely,
Nicola

PS Have a merry Christmas!