How Long does it Take for Bearings to Flute

I agree.

GA.

VFD's can also cause problems in the motor itself,due the chopped square waves.

A motor not designed for VFD's will also suffer a shorter lifetime due to the chopped

square waves generated by the drive.

VFD designed motors have a variety of design features to prevent EDM of bearings.

Some use grounding brushes,some use grounding bearing seals,others use ceramic

coatings on bearings or ceramic bearings.

A VFD drive also requires a VFD rated cable for best results.

VFD's can also create problems in other equipment not attached to the VFD.

It generally takes several months for bearings to fail in this situation....the bigger and more important question is why was there no protection installed with the motor, it should have been ordered as a vfd driven motor, which should have included protection...

..."These discharges are so frequent (potentially millions per hour) that before long the entire bearing race surface becomes damaged with countless pits known as frosting. A phenomenon known as fluting may occur as well, producing washboard-like ridges across the frosted bearing race. Fluting causes audible noise and vibration and is an indication of a catastrophic failure mode. Regardless of the type of rolling element or raceway damage that occurs, the resulting motor failure often costs thousands of dollars in downtime and equipment failure related repair or replacement costs.

Failure rates vary widely depending on many factors, but evidence suggests that a significant portion of failures occur in only 3 to 12 months after system startup. All AC and DC motors operated by electronic drives or inverters have the potential of developing this failure in their bearings regardless of motor frame size or horsepower."...

https://www.mep.ca/Public/pdf/AEGISHandbook_English.pdf

This does seem a more likely scenario, that some damage was existing before installation, that could have been caused by any number of things....

The issue of bearing damage during storage is called “false brinelling” and is different from fluting due to EDM damage.

The rough sound may be similar though, so identifying it will depend on a visual inspection of the bearings and races.

You speak of failure on #2 but not on #1. Assuming the motors were otherwise identical and the driven equipment identical I would look for either of two possible causes:

1. A pre-installation fault that targeted #2 only. These have been mentioned by others.
2. A sufficiently different installation in terms of distance from the drive, or differences in programming parameters.

VFD's attempt to produce an output waveform that has a voltage and frequency varying simultaneously, typically in a constant or near constant linear fashion. To do this the DC is chopped and put onto the output as pulses of uniform voltage but varying width and spacing--normally with an oscillator at a high-frequency (called the carrier frequency) controlling this. The output is filtered to smooth it into a waveform that is approximately sinusoidal. However the output has a substantial burden of harmonics, that cause problems when you get to the motor. An output reactor will minimize much of this but if the cable length from the VFD to the motor is long these harmonics can cause a lot of problems. An output filter will eliminate almost all of these harmonics, but they can be large and expensive.

Your vendor did the sensible first thing--putting a grounding brush on the motor shaft. Have a good drive engineer look your system over and see if there are other steps for you, such as changing the carrier frequency, or adding filtering.

--JMM

Thanks everyone, I really appreciate the input.

I am familiar with the larger motors needing rotated during storage and keeping heaters on them, but these were only 150 HP and we have over 200 in service with this being the first occurrence at startup.

THe two at this installation were identical and used the same installation and programming as all the others. The motors were only energized with the drive for less than 5 minutes and surprisingly when we stopped the drives in coast mode the noisy (#2) motor actually spun longer than the other. So it was making more noise and seemed to be dragging on something but it really spun more freely.

I do not know the life story of these motors but I do know once they got to the fabrication shop they only sat around for about 3 months. Looking at the SN they appeared to have been manufactured within the past year

It is possible that there was damage during shipment or dropped harshly on a solid surface but that would have been before the fabrication steps. Once at the fab shop they were together almost the entire time and shipped together

You might want to put a good scope on the outputs from each drive and see if there is an undiagnosed difference. This could point to a problem within drive #2.

--JMM