Hello to everybody I would like to know more about this topic. I hope to share knowledges with people who has experience in this kind of electrical machines.

Regards

Raul-

I know that originally this was done by EPRI. They monitored mainly the 2xLine Frequency peak and trained people to take readings on all 4 sides of the transformers. The object was to be able to tell when there was a break down as the windings would then vibration more and the 2xLF peak would increase.

It is also good to look at the circulation pumps but know that you are generally not anywhere near the bearings.

Attached is an example of a couple of what I am talking about.

XFormer.pdf (63 Kb, 42 downloads) Xformer2xLF

Raul,

I have looked at this a lot. The EPRI paper talks about the blocking becoming loose if you see higher multiples of 2x line freq. What I have found is that it is very dependent on where you take the data. You are many getting different signatures based on the stiffness in the shell at different points.

It may work if you take the exact same spot over a long period of time. In truth there is just not a lot of data. There are not many XFMRs that are monitored. A very small portion of these have failed. What is needed to prove or disprove this is a monitored XFMR; a problem identified, repaired and then redo the post vibration test.

The only case I have read where this happened was in the EPRI paper. They identified loose blocking from the high amplitude multiples of 2x line freq. They had the blocking rewedged and tightened. Then they redid the analysis. The Higher multiples got higher. Did not do what they thought. To me it makes sense. They stiffened the structure so the natural freq increased. This would tend to excite higher multiples.

Hey Ron- I'd really like to visit with you on this sometime. Check out my website or drop me an email. I agree with you on the whole no data is out there. I know of one company that is doing this and I have no idea how beneficial it has been for them. They were also doing vibration on their cooling pumps for the transformers and as far as I know haven't made good call yet. Mainly cuz they won't put any money into getting their guys properly trained. It isn't the box that makes the call it is the man behind the box.

Linked below are some case histories on the same subject for a family of motors at our plant. We did indeed see harmonics of 2*LF on bearing housings of motors with loose stator coils and these frequencies went away when those motors were rewound or rewedged.

http://maintenanceforums.com/eve/forums/a/tpc/f/3751089...051039102#9051039102