In our experience, loose form wound motor coils do show up in the vibration as very low magnitude peaks at harmonics of 2*LF. When the condition is corrected through rewind or rewedge, that symptom disappears.

Attached is some data to support this based on theory and experience.

I would be interested to hear any comments.

twicelineharmonics.ppt (1,425 Kb, 99 downloads)

Pete,

One of my former coworkers studied a problem similar to this....with loose coils on hydro turbine generator stators. As I recall, those actually had wedges to drive in to keep the coils tightly in position. I'll see if I can track down any information.

Jon
Spintelligent Labs

Hi Jon. Yes top wedges are very common and are included in this motor. On the bigger machines like your hydro there may be more exotic mechanisms such as ripple springs as well but not here.

The original coil and wedge design was not sufficient to keep the coil tight in the slot. 5 motors were rewound. New coils after rewind have much better design, reduced side clearance, consolidated for dimensional stability over time, good quality wedge and below-wedge packing, and also vpi treatment which helps as well. Those motors no longer have those 2LF harmonics. One motor was rewedged - had top rewedges removed and replaced. Also remvoed the 2LF harmonic symptom.

All- I know I sent out a bunch of case studies. I was particularly interested to hear comments on this one. To me it seems not a normal or traditional method but there seems to be good support. We are making some decisions regarding that rewedged motor and I'm trying to consider how effective will this monitoring strategy be.

This message has been edited. Last edited by: electricpete,

Pete,

I wonder if multiplicity of 2xLF harmonics could be a good indicator of loose stator coils, the way any looseness will normally manifest itself? In this case the excitation frequency is at 2xLF. I believe other sources of 2xLF won't result in multiple harmonics. Unfortunately, it looks like you don't have data with higher Fmax. Taking TWF data with higher Fmax could also show impacting taking place in cases of looseness, in this case at 2xLF.

You can also look for coil pass frequency which is the number of stator slots times RPM (similar to RBPF) and harmonics. They will be of a relatively low level.

Howard

David - I agree on all your comments. I view this as looseness excted by 2LF force resulting in harmonics of 2LF... analogous to looseness excited by 1X resulting in harmonics of 1x. With machines all repaired there is no chance to go back and take additional data at higher Fmax. For the rewedged machine we want to watch/trend it closely we will develop a collection spec perhaps similar to a/f bearings. I am open to suggestions.

Doc - I don't understand why stator slot pass frequency would reveal coils loose in slots. What would be the mechanism?

Here is the same presentation in pdf format.
(use view/rotate to make it landscape).

twicelineharmonics.PDF (2,049 Kb, 53 downloads)