Last night I went to a new rebuilt waste burning plant to tuck in 4pcs 5kW motors for the night. Fist they was installed with one brand of motor but customer wanted "high efficency" so they were swapped to another brand being such. They have a 6Kg fan overhung and run 2980RPM. After 6 months they sound like disaster. Data indicate severe looseness 10mm/s vertical and 3 horiz at motor free end would also indicate resonance possibility. Knocktest proved nothing like that. No real indications of massive bearing failure in the normal way. Worst motor swapped, new motor runs like a dream, less than 3mm/s any direction. Old motor shaft kan be moved 1/10 mm or more vertically, not horisontally. Some info that could be gathered, bearings was angle contact bearings, not normal roller or deep grove I would be expecting. So the cast aluminium endbell of these tiny´s was beaten out of shape, primarily vertically so the bearing jumps around in the seating, I think, I have not seen it yet. Is this type of bearing common in this application? Why? Is this type of event common nowadays? These are permanently greased beasts. Out of 4 units, 2 was about in the worst state, 1 was on the way going there and 1 was reasonably ok. Not so good statistics for this application. Olov

Oli,
Not a fan of aluminum end brackets. This fan should have had longer life that it did, even with aluminum. The speed and weight of the fan suggests a potential rotor critical. Has anyone produced a rotor dynamics model for this fan? That would be my first step.
As far as 'high efficiency', there is no reason to buy a cheaper (PC) motor. There are high efficiency motors with steel end brackets.