All:
We have a UPS that has a rotor inside rotor inductive coupling. The maint is contracted out to the manufacturer right now, but I recently got a call from an electrician about a "Vibration"... I went down there with my Equipment and found a bearing freq @ 1" sec.

The vendors vibration folks basically said that the inner rotor bearing is very difficult to catch in advance because there is no direct path to put the accelerometer.. essentially all vibration is transferred through the outer rotor bearings.

I've been asked to set-up an internal Vib program on this and was looking for any suggestions about getting accurate and timely data on rotor-in-rotor systems.

I would just do it as normal. There is a similar problem in paper machines with an internally supported suction box resting in a bearing. There is no problem to see a bearing development from the outside, also via an outer bearing. The fault pattern is so unique that you can see it well. Good Luck/Arne

On a 2500 HP EM drive I've been doing for quite some time - as Arne suggested, do as normal.

By design the bearing tolerances are not the book 'normal spec'. Generally this is proprietary information but is easily calculated and plotted. Having said that; you're looking at at higher ABEC number and with both rings rotating probably an ABEC 9 and taking the clearance to zero or very near zero as in 0.0001-3". On my machine the outer brgs are sleeve so brg frequencies generated are from the inner bearings - no brainer.

I have been able to detect brg flaws and avert failure. Tear down and inspection revealed the shot was called corrrectly.

If your bearings are both AF and frequencies are separaated far enough - then you may want to consider band filters and allow the desired frequencies to pass.