BLDC motor rotor heating

[SPIGUY1]

I ran a test to determine if any mechanical losses are significant contributors to the increased rotor temperature on a BLDC motor used for research projects. Another motor was connected to an unpowered BLDC motor with plastic material coupling. The motor was spun at about 2000rpm for 5 minutes and the temperature of the driven motor rotor was measured to be about 50C. This is a significant temperature increase that suggests there are mechanical losses that may need to be investigated further.
I am using a 6-step PWM commutation @ 20kHz. Hall sensors are used for position feedback. The phase current appears to be aligned well with the BEMF. Since the driven motor was open circuit, what losses would cause this heating?

 

[edison123]

Windage.

*Even inside a hard-boiled egg, there is a golden heart!!!*

 

[Clyde38]

What temperature was the stator at this time?

 

[SPIGUY1]

Stator winding temperature was not measured during the particular test, but it was measured in other tests and was considerably cooler than the rotor (only copper losses as expected).

 

[Clyde38]

Not knowing the features of your motor, here goes....
If this is a "typical", whatever that means PM BLDC with the rotor inside the stator (airgap ~.02"), and the stator is a slotted laminated stack with the windings in the slots then you will also have eddy current and hysterisis losses to add to your copper losses. The higher the number of magnet pole pairs, the higher the frequency and the higher the losses. If the magnets are of the conductive nature, such as NdFeB then eddy currents can also be developed in the magnets. Take these as comments to consider, not answers to your question.

 

[sreid]

Most BLDC motors have a solid steel rotor that the magnets are mounted on. The stator field threads the rotor as back iron. In a six step drive, the stator magnetic vector advances in 60 degree steps and this might cause some significant eddy currents in the rotor steel.